[llvm] TTI: Fix special casing vectorization costs of saturating add/sub (PR #97463)

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Tue Jul 2 12:39:01 PDT 2024


https://github.com/arsenm created https://github.com/llvm/llvm-project/pull/97463

None

>From fd197af9717589741108795b729627658211b52a Mon Sep 17 00:00:00 2001
From: Matt Arsenault <Matthew.Arsenault at amd.com>
Date: Mon, 13 Jul 2020 21:07:35 -0400
Subject: [PATCH] TTI: Fix special casing vectorization costs of saturating
 add/sub

---
 llvm/include/llvm/CodeGen/BasicTTIImpl.h      |  48 +-
 .../Analysis/CostModel/AArch64/arith-ssat.ll  |  32 +-
 .../Analysis/CostModel/AArch64/arith-usat.ll  |  32 +-
 .../CostModel/AArch64/sve-intrinsics.ll       |   8 +-
 .../CostModel/ARM/active_lane_mask.ll         |  24 +-
 .../test/Analysis/CostModel/ARM/arith-ssat.ll | 428 +++++------
 .../test/Analysis/CostModel/ARM/arith-usat.ll | 404 +++++-----
 .../Analysis/CostModel/RISCV/int-sat-math.ll  |  32 +-
 .../CostModel/X86/arith-ssat-codesize.ll      | 700 ++++++++---------
 .../CostModel/X86/arith-ssat-latency.ll       | 704 +++++++++---------
 .../CostModel/X86/arith-ssat-sizelatency.ll   | 702 ++++++++---------
 .../test/Analysis/CostModel/X86/arith-ssat.ll | 440 +++++------
 .../CostModel/X86/arith-usat-codesize.ll      | 704 +++++++++---------
 .../CostModel/X86/arith-usat-latency.ll       | 704 +++++++++---------
 .../CostModel/X86/arith-usat-sizelatency.ll   | 704 +++++++++---------
 .../test/Analysis/CostModel/X86/arith-usat.ll | 296 ++++----
 .../LoopVectorize/AArch64/intrinsiccost.ll    |   2 +-
 .../LoopVectorize/ARM/mve-saddsatcost.ll      |   4 +-
 .../SLPVectorizer/X86/arith-add-ssat.ll       | 418 ++++++++---
 .../SLPVectorizer/X86/arith-add-usat.ll       |  54 --
 .../SLPVectorizer/X86/arith-sub-ssat.ll       | 418 ++++++++---
 .../SLPVectorizer/X86/arith-sub-usat.ll       |  48 +-
 22 files changed, 3604 insertions(+), 3302 deletions(-)

diff --git a/llvm/include/llvm/CodeGen/BasicTTIImpl.h b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
index 4f1dc9f991c06..2e0f5ee6ea6bd 100644
--- a/llvm/include/llvm/CodeGen/BasicTTIImpl.h
+++ b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
@@ -2129,45 +2129,17 @@ class BasicTTIImplBase : public TargetTransformInfoImplCRTPBase<T> {
       return Cost;
     }
     case Intrinsic::sadd_sat:
-    case Intrinsic::ssub_sat: {
-      Type *CondTy = RetTy->getWithNewBitWidth(1);
-
-      Type *OpTy = StructType::create({RetTy, CondTy});
-      Intrinsic::ID OverflowOp = IID == Intrinsic::sadd_sat
-                                     ? Intrinsic::sadd_with_overflow
-                                     : Intrinsic::ssub_with_overflow;
-      CmpInst::Predicate Pred = CmpInst::ICMP_SGT;
-
-      // SatMax -> Overflow && SumDiff < 0
-      // SatMin -> Overflow && SumDiff >= 0
-      InstructionCost Cost = 0;
-      IntrinsicCostAttributes Attrs(OverflowOp, OpTy, {RetTy, RetTy}, FMF,
-                                    nullptr, ScalarizationCostPassed);
-      Cost += thisT()->getIntrinsicInstrCost(Attrs, CostKind);
-      Cost += thisT()->getCmpSelInstrCost(BinaryOperator::ICmp, RetTy, CondTy,
-                                          Pred, CostKind);
-      Cost += 2 * thisT()->getCmpSelInstrCost(BinaryOperator::Select, RetTy,
-                                              CondTy, Pred, CostKind);
-      return Cost;
-    }
+      ISD = ISD::SADDSAT;
+      break;
+    case Intrinsic::ssub_sat:
+      ISD = ISD::SSUBSAT;
+      break;
     case Intrinsic::uadd_sat:
-    case Intrinsic::usub_sat: {
-      Type *CondTy = RetTy->getWithNewBitWidth(1);
-
-      Type *OpTy = StructType::create({RetTy, CondTy});
-      Intrinsic::ID OverflowOp = IID == Intrinsic::uadd_sat
-                                     ? Intrinsic::uadd_with_overflow
-                                     : Intrinsic::usub_with_overflow;
-
-      InstructionCost Cost = 0;
-      IntrinsicCostAttributes Attrs(OverflowOp, OpTy, {RetTy, RetTy}, FMF,
-                                    nullptr, ScalarizationCostPassed);
-      Cost += thisT()->getIntrinsicInstrCost(Attrs, CostKind);
-      Cost +=
-          thisT()->getCmpSelInstrCost(BinaryOperator::Select, RetTy, CondTy,
-                                      CmpInst::BAD_ICMP_PREDICATE, CostKind);
-      return Cost;
-    }
+      ISD = ISD::UADDSAT;
+      break;
+    case Intrinsic::usub_sat:
+      ISD = ISD::USUBSAT;
+      break;
     case Intrinsic::smul_fix:
     case Intrinsic::umul_fix: {
       unsigned ExtSize = RetTy->getScalarSizeInBits() * 2;
diff --git a/llvm/test/Analysis/CostModel/AArch64/arith-ssat.ll b/llvm/test/Analysis/CostModel/AArch64/arith-ssat.ll
index 2267d9b88c970..d79eee53ecb48 100644
--- a/llvm/test/Analysis/CostModel/AArch64/arith-ssat.ll
+++ b/llvm/test/Analysis/CostModel/AArch64/arith-ssat.ll
@@ -33,23 +33,23 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; RECIP-LABEL: 'add'
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V3I32 = call <3 x i32> @llvm.sadd.sat.v3i32(<3 x i32> undef, <3 x i32> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -59,23 +59,23 @@ define i32 @add(i32 %arg) {
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SIZE-LABEL: 'add'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V3I32 = call <3 x i32> @llvm.sadd.sat.v3i32(<3 x i32> undef, <3 x i32> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -142,22 +142,22 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; RECIP-LABEL: 'sub'
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -167,22 +167,22 @@ define i32 @sub(i32 %arg) {
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SIZE-LABEL: 'sub'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/AArch64/arith-usat.ll b/llvm/test/Analysis/CostModel/AArch64/arith-usat.ll
index 5a131f23847b1..c022a79ec2f38 100644
--- a/llvm/test/Analysis/CostModel/AArch64/arith-usat.ll
+++ b/llvm/test/Analysis/CostModel/AArch64/arith-usat.ll
@@ -32,22 +32,22 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; RECIP-LABEL: 'add'
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -57,22 +57,22 @@ define i32 @add(i32 %arg) {
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SIZE-LABEL: 'add'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -138,22 +138,22 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; RECIP-LABEL: 'sub'
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -163,22 +163,22 @@ define i32 @sub(i32 %arg) {
 ; RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SIZE-LABEL: 'sub'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll b/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll
index 1ff280d75b4e9..07e63e4cf25a0 100644
--- a/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll
+++ b/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll
@@ -691,8 +691,8 @@ define void @get_lane_mask() #0 {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef)
@@ -701,8 +701,8 @@ define void @get_lane_mask() #0 {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
 ; TYPE_BASED_ONLY-LABEL: 'get_lane_mask'
diff --git a/llvm/test/Analysis/CostModel/ARM/active_lane_mask.ll b/llvm/test/Analysis/CostModel/ARM/active_lane_mask.ll
index cdd61c147fdad..204f05a767e93 100644
--- a/llvm/test/Analysis/CostModel/ARM/active_lane_mask.ll
+++ b/llvm/test/Analysis/CostModel/ARM/active_lane_mask.ll
@@ -3,18 +3,18 @@
 
 define void @get_lane_mask() {
 ; CHECK-LABEL: 'get_lane_mask'
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %mask_v8i1_i16 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i16(i16 undef, i16 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %mask_v4i1_i16 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i16(i16 undef, i16 undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %mask_v2i1_i16 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 240 for instruction: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 208 for instruction: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 208 for instruction: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %mask_v8i1_i16 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %mask_v4i1_i16 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %mask_v2i1_i16 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
   %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
index ba1d163882bcb..804dc89b7808e 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
@@ -37,72 +37,72 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 define i32 @add(i32 %arg) {
 ; V8M-RECIP-LABEL: 'add'
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 704 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-RECIP-LABEL: 'add'
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 117 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; MVE-RECIP-LABEL: 'add'
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 298 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 594 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -111,73 +111,73 @@ define i32 @add(i32 %arg) {
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; V8M-SIZE-LABEL: 'add'
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 133 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-SIZE-LABEL: 'add'
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 105 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; MVE-SIZE-LABEL: 'add'
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 61 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 150 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -243,72 +243,72 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 define i32 @sub(i32 %arg) {
 ; V8M-RECIP-LABEL: 'sub'
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 704 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-RECIP-LABEL: 'sub'
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 117 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; MVE-RECIP-LABEL: 'sub'
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 298 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 594 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -317,73 +317,73 @@ define i32 @sub(i32 %arg) {
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; V8M-SIZE-LABEL: 'sub'
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 133 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-SIZE-LABEL: 'sub'
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 105 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; MVE-SIZE-LABEL: 'sub'
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 61 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 150 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 43 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
index 3b755f9a55f4f..3575b74d27a1a 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
@@ -36,73 +36,73 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; V8M-RECIP-LABEL: 'add'
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 704 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-RECIP-LABEL: 'add'
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; MVE-RECIP-LABEL: 'add'
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 296 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -111,73 +111,73 @@ define i32 @add(i32 %arg) {
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; V8M-SIZE-LABEL: 'add'
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-SIZE-LABEL: 'add'
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; MVE-SIZE-LABEL: 'add'
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 146 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -242,73 +242,73 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; V8M-RECIP-LABEL: 'sub'
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 352 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 704 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-RECIP-LABEL: 'sub'
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; MVE-RECIP-LABEL: 'sub'
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 296 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
@@ -317,73 +317,73 @@ define i32 @sub(i32 %arg) {
 ; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; V8M-SIZE-LABEL: 'sub'
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; NEON-SIZE-LABEL: 'sub'
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; MVE-SIZE-LABEL: 'sub'
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 146 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
 ; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/RISCV/int-sat-math.ll b/llvm/test/Analysis/CostModel/RISCV/int-sat-math.ll
index 541534d2778aa..cdf33a11e2f14 100644
--- a/llvm/test/Analysis/CostModel/RISCV/int-sat-math.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/int-sat-math.ll
@@ -3,7 +3,7 @@
 
 define void @sadd.sat() {
 ; CHECK-LABEL: 'sadd.sat'
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %1 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %1 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -12,7 +12,7 @@ define void @sadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x i8> @llvm.sadd.sat.nxv4i8(<vscale x 4 x i8> undef, <vscale x 4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x i8> @llvm.sadd.sat.nxv8i8(<vscale x 8 x i8> undef, <vscale x 8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x i8> @llvm.sadd.sat.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %10 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %10 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %11 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %12 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %13 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
@@ -21,7 +21,7 @@ define void @sadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 4 x i16> @llvm.sadd.sat.nxv4i16(<vscale x 4 x i16> undef, <vscale x 4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 8 x i16> @llvm.sadd.sat.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 16 x i16> @llvm.sadd.sat.nxv16i16(<vscale x 16 x i16> undef, <vscale x 16 x i16> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %19 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %19 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %20 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %21 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %22 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
@@ -30,7 +30,7 @@ define void @sadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %25 = call <vscale x 4 x i32> @llvm.sadd.sat.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 8 x i32> @llvm.sadd.sat.nxv8i32(<vscale x 8 x i32> undef, <vscale x 8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 16 x i32> @llvm.sadd.sat.nxv16i32(<vscale x 16 x i32> undef, <vscale x 16 x i32> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %28 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %28 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %29 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %30 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %31 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
@@ -80,7 +80,7 @@ define void @sadd.sat() {
 
 define void @uadd.sat() {
 ; CHECK-LABEL: 'uadd.sat'
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %1 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %1 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -89,7 +89,7 @@ define void @uadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x i8> @llvm.uadd.sat.nxv4i8(<vscale x 4 x i8> undef, <vscale x 4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x i8> @llvm.uadd.sat.nxv8i8(<vscale x 8 x i8> undef, <vscale x 8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x i8> @llvm.uadd.sat.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %10 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %10 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %11 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %12 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %13 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
@@ -98,7 +98,7 @@ define void @uadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 4 x i16> @llvm.uadd.sat.nxv4i16(<vscale x 4 x i16> undef, <vscale x 4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 8 x i16> @llvm.uadd.sat.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 16 x i16> @llvm.uadd.sat.nxv16i16(<vscale x 16 x i16> undef, <vscale x 16 x i16> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %19 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %19 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %20 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %21 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %22 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
@@ -107,7 +107,7 @@ define void @uadd.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %25 = call <vscale x 4 x i32> @llvm.uadd.sat.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 8 x i32> @llvm.uadd.sat.nxv8i32(<vscale x 8 x i32> undef, <vscale x 8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 16 x i32> @llvm.uadd.sat.nxv16i32(<vscale x 16 x i32> undef, <vscale x 16 x i32> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %28 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %28 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %29 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %30 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %31 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
@@ -157,7 +157,7 @@ define void @uadd.sat() {
 
 define void @usub.sat() {
 ; CHECK-LABEL: 'usub.sat'
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %1 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %1 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -166,7 +166,7 @@ define void @usub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x i8> @llvm.usub.sat.nxv4i8(<vscale x 4 x i8> undef, <vscale x 4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x i8> @llvm.usub.sat.nxv8i8(<vscale x 8 x i8> undef, <vscale x 8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x i8> @llvm.usub.sat.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %10 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %10 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %11 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %12 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %13 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
@@ -175,7 +175,7 @@ define void @usub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 4 x i16> @llvm.usub.sat.nxv4i16(<vscale x 4 x i16> undef, <vscale x 4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 8 x i16> @llvm.usub.sat.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 16 x i16> @llvm.usub.sat.nxv16i16(<vscale x 16 x i16> undef, <vscale x 16 x i16> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %19 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %19 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %20 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %21 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %22 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
@@ -184,7 +184,7 @@ define void @usub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %25 = call <vscale x 4 x i32> @llvm.usub.sat.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 8 x i32> @llvm.usub.sat.nxv8i32(<vscale x 8 x i32> undef, <vscale x 8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 16 x i32> @llvm.usub.sat.nxv16i32(<vscale x 16 x i32> undef, <vscale x 16 x i32> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %28 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %28 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %29 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %30 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %31 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
@@ -234,7 +234,7 @@ define void @usub.sat() {
 
 define void @ssub.sat() {
 ; CHECK-LABEL: 'ssub.sat'
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %1 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %1 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
@@ -243,7 +243,7 @@ define void @ssub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x i8> @llvm.ssub.sat.nxv4i8(<vscale x 4 x i8> undef, <vscale x 4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x i8> @llvm.ssub.sat.nxv8i8(<vscale x 8 x i8> undef, <vscale x 8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x i8> @llvm.ssub.sat.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %10 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %10 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %11 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %12 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %13 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
@@ -252,7 +252,7 @@ define void @ssub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 4 x i16> @llvm.ssub.sat.nxv4i16(<vscale x 4 x i16> undef, <vscale x 4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 8 x i16> @llvm.ssub.sat.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 16 x i16> @llvm.ssub.sat.nxv16i16(<vscale x 16 x i16> undef, <vscale x 16 x i16> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %19 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %19 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %20 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %21 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %22 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
@@ -261,7 +261,7 @@ define void @ssub.sat() {
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %25 = call <vscale x 4 x i32> @llvm.ssub.sat.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 8 x i32> @llvm.ssub.sat.nxv8i32(<vscale x 8 x i32> undef, <vscale x 8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 16 x i32> @llvm.ssub.sat.nxv16i32(<vscale x 16 x i32> undef, <vscale x 16 x i32> undef)
-; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %28 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %28 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %29 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %30 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %31 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-ssat-codesize.ll b/llvm/test/Analysis/CostModel/X86/arith-ssat-codesize.ll
index 21c99555c5df3..669456bcda343 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-ssat-codesize.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-ssat-codesize.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-ssat-latency.ll b/llvm/test/Analysis/CostModel/X86/arith-ssat-latency.ll
index 1d398d9555fbf..659e45841c0fe 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-ssat-latency.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-ssat-latency.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 101 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 101 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-ssat-sizelatency.ll b/llvm/test/Analysis/CostModel/X86/arith-ssat-sizelatency.ll
index 4d1fa7c5cf55c..bdc98a3b9635a 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-ssat-sizelatency.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-ssat-sizelatency.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 93 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 93 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 93 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 93 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 49 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-ssat.ll b/llvm/test/Analysis/CostModel/X86/arith-ssat.ll
index 4bbd28edb434b..52f751477badf 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-ssat.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-ssat.ll
@@ -34,209 +34,209 @@ declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 89 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 178 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
@@ -287,209 +287,209 @@ declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 37 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 73 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-usat-codesize.ll b/llvm/test/Analysis/CostModel/X86/arith-usat-codesize.ll
index e92b8102a5a3f..3a7314532f7f4 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-usat-codesize.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-usat-codesize.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-usat-latency.ll b/llvm/test/Analysis/CostModel/X86/arith-usat-latency.ll
index 9d9e5843c2285..bd031517529f8 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-usat-latency.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-usat-latency.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-usat-sizelatency.ll b/llvm/test/Analysis/CostModel/X86/arith-usat-sizelatency.ll
index 1912822bdcb65..bce002c5ff5e3 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-usat-sizelatency.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-usat-sizelatency.ll
@@ -34,212 +34,212 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
@@ -287,212 +287,212 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 91 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 179 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/X86/arith-usat.ll b/llvm/test/Analysis/CostModel/X86/arith-usat.ll
index 75950e40c2a00..6534176d63953 100644
--- a/llvm/test/Analysis/CostModel/X86/arith-usat.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-usat.ll
@@ -34,209 +34,209 @@ declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
 ; SSE2-LABEL: 'add'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'add'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'add'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'add'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'add'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'add'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'add'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'add'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'add'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'add'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'add'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
@@ -287,209 +287,209 @@ declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
 ; SSE2-LABEL: 'sub'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'sub'
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'sub'
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'sub'
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'sub'
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512F-LABEL: 'sub'
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512F-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512F-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512BW-LABEL: 'sub'
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512BW-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512DQ-LABEL: 'sub'
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; AVX512DQ-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SLM-LABEL: 'sub'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'sub'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'sub'
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll b/llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll
index 0f26092f510ca..a60f246c9e860 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll
@@ -7,7 +7,7 @@ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
 target triple = "aarch64--linux-gnu"
 
 ; CHECK-COST-LABEL: sadd
-; CHECK-COST: Found an estimated cost of 6 for VF 1 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
+; CHECK-COST: Found an estimated cost of 10 for VF 1 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 ; CHECK-COST: Found an estimated cost of 4 for VF 2 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 ; CHECK-COST: Found an estimated cost of 1 for VF 4 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 ; CHECK-COST: Found an estimated cost of 1 for VF 8 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
diff --git a/llvm/test/Transforms/LoopVectorize/ARM/mve-saddsatcost.ll b/llvm/test/Transforms/LoopVectorize/ARM/mve-saddsatcost.ll
index 0515ccea94b5b..44d17d09307b3 100644
--- a/llvm/test/Transforms/LoopVectorize/ARM/mve-saddsatcost.ll
+++ b/llvm/test/Transforms/LoopVectorize/ARM/mve-saddsatcost.ll
@@ -7,8 +7,8 @@ target datalayout = "e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64"
 target triple = "thumbv8.1m.main-arm-none-eabi"
 
 ; CHECK-COST-LABEL: arm_offset_q15
-; CHECK-COST: Found an estimated cost of 7 for VF 1 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
-; CHECK-COST: Found an estimated cost of 36 for VF 2 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
+; CHECK-COST: Found an estimated cost of 1 for VF 1 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
+; CHECK-COST: Found an estimated cost of 26 for VF 2 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 ; CHECK-COST: Found an estimated cost of 8 for VF 4 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 ; CHECK-COST: Found an estimated cost of 2 for VF 8 For instruction:   %1 = tail call i16 @llvm.sadd.sat.i16(i16 %0, i16 %offset)
 
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/arith-add-ssat.ll b/llvm/test/Transforms/SLPVectorizer/X86/arith-add-ssat.ll
index 24c5fcb068086..1e8d45c89182b 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/arith-add-ssat.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/arith-add-ssat.ll
@@ -62,75 +62,60 @@ define void @add_v8i64() {
 ; SSE-NEXT:    ret void
 ;
 ; SLM-LABEL: @add_v8i64(
-; SLM-NEXT:    [[A0:%.*]] = load i64, ptr @a64, align 8
-; SLM-NEXT:    [[A1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
-; SLM-NEXT:    [[A2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; SLM-NEXT:    [[A3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
-; SLM-NEXT:    [[A4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; SLM-NEXT:    [[A5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
-; SLM-NEXT:    [[A6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; SLM-NEXT:    [[A7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
-; SLM-NEXT:    [[B0:%.*]] = load i64, ptr @b64, align 8
-; SLM-NEXT:    [[B1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
-; SLM-NEXT:    [[B2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; SLM-NEXT:    [[B3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
-; SLM-NEXT:    [[B4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; SLM-NEXT:    [[B5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
-; SLM-NEXT:    [[B6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; SLM-NEXT:    [[B7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
-; SLM-NEXT:    [[R0:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A0]], i64 [[B0]])
-; SLM-NEXT:    [[R1:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A1]], i64 [[B1]])
-; SLM-NEXT:    [[R2:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A2]], i64 [[B2]])
-; SLM-NEXT:    [[R3:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A3]], i64 [[B3]])
-; SLM-NEXT:    [[R4:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A4]], i64 [[B4]])
-; SLM-NEXT:    [[R5:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A5]], i64 [[B5]])
-; SLM-NEXT:    [[R6:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A6]], i64 [[B6]])
-; SLM-NEXT:    [[R7:%.*]] = call i64 @llvm.sadd.sat.i64(i64 [[A7]], i64 [[B7]])
-; SLM-NEXT:    store i64 [[R0]], ptr @c64, align 8
-; SLM-NEXT:    store i64 [[R1]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
-; SLM-NEXT:    store i64 [[R2]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; SLM-NEXT:    store i64 [[R3]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
-; SLM-NEXT:    store i64 [[R4]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; SLM-NEXT:    store i64 [[R5]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
-; SLM-NEXT:    store i64 [[R6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; SLM-NEXT:    store i64 [[R7]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
+; SLM-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; SLM-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; SLM-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; SLM-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; SLM-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; SLM-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; SLM-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; SLM-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; SLM-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; SLM-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
 ; SLM-NEXT:    ret void
 ;
-; AVX1-LABEL: @add_v8i64(
-; AVX1-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
-; AVX1-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
-; AVX1-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
-; AVX1-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
-; AVX1-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
-; AVX1-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
-; AVX1-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; AVX1-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; AVX1-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
-; AVX1-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; AVX1-NEXT:    ret void
-;
-; AVX2-LABEL: @add_v8i64(
-; AVX2-NEXT:    [[TMP1:%.*]] = load <4 x i64>, ptr @a64, align 8
-; AVX2-NEXT:    [[TMP2:%.*]] = load <4 x i64>, ptr @b64, align 8
-; AVX2-NEXT:    [[TMP3:%.*]] = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> [[TMP1]], <4 x i64> [[TMP2]])
-; AVX2-NEXT:    store <4 x i64> [[TMP3]], ptr @c64, align 8
-; AVX2-NEXT:    [[TMP4:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; AVX2-NEXT:    [[TMP5:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; AVX2-NEXT:    [[TMP6:%.*]] = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> [[TMP4]], <4 x i64> [[TMP5]])
-; AVX2-NEXT:    store <4 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; AVX2-NEXT:    ret void
+; AVX-LABEL: @add_v8i64(
+; AVX-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; AVX-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; AVX-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; AVX-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; AVX-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; AVX-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; AVX-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; AVX-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; AVX-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; AVX-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
+; AVX-NEXT:    ret void
 ;
 ; AVX512-LABEL: @add_v8i64(
-; AVX512-NEXT:    [[TMP1:%.*]] = load <8 x i64>, ptr @a64, align 8
-; AVX512-NEXT:    [[TMP2:%.*]] = load <8 x i64>, ptr @b64, align 8
-; AVX512-NEXT:    [[TMP3:%.*]] = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> [[TMP1]], <8 x i64> [[TMP2]])
-; AVX512-NEXT:    store <8 x i64> [[TMP3]], ptr @c64, align 8
+; AVX512-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; AVX512-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; AVX512-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; AVX512-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; AVX512-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; AVX512-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; AVX512-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; AVX512-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; AVX512-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; AVX512-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
 ; AVX512-NEXT:    ret void
 ;
 ; AVX256BW-LABEL: @add_v8i64(
@@ -180,59 +165,271 @@ define void @add_v8i64() {
 
 define void @add_v16i32() {
 ; SSE-LABEL: @add_v16i32(
-; SSE-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
-; SSE-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
-; SSE-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
-; SSE-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
-; SSE-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
-; SSE-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
-; SSE-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
-; SSE-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; SSE-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; SSE-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; SSE-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; SSE-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; SSE-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; SSE-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; SSE-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; SSE-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; SSE-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; SSE-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; SSE-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; SSE-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; SSE-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; SSE-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; SSE-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; SSE-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; SSE-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; SSE-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; SSE-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; SSE-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; SSE-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; SSE-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; SSE-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; SSE-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; SSE-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; SSE-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; SSE-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; SSE-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; SSE-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; SSE-NEXT:    [[R0:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A0]], i32 [[B0]])
+; SSE-NEXT:    [[R1:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A1]], i32 [[B1]])
+; SSE-NEXT:    [[R2:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A2]], i32 [[B2]])
+; SSE-NEXT:    [[R3:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A3]], i32 [[B3]])
+; SSE-NEXT:    [[R4:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A4]], i32 [[B4]])
+; SSE-NEXT:    [[R5:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A5]], i32 [[B5]])
+; SSE-NEXT:    [[R6:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A6]], i32 [[B6]])
+; SSE-NEXT:    [[R7:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A7]], i32 [[B7]])
+; SSE-NEXT:    [[R8:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A8]], i32 [[B8]])
+; SSE-NEXT:    [[R9:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A9]], i32 [[B9]])
+; SSE-NEXT:    [[R10:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A10]], i32 [[B10]])
+; SSE-NEXT:    [[R11:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A11]], i32 [[B11]])
+; SSE-NEXT:    [[R12:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A12]], i32 [[B12]])
+; SSE-NEXT:    [[R13:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A13]], i32 [[B13]])
+; SSE-NEXT:    [[R14:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A14]], i32 [[B14]])
+; SSE-NEXT:    [[R15:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A15]], i32 [[B15]])
+; SSE-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; SSE-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; SSE-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; SSE-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; SSE-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; SSE-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; SSE-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; SSE-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; SSE-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; SSE-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; SSE-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; SSE-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; SSE-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SSE-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; SSE-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; SSE-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; SSE-NEXT:    ret void
 ;
 ; SLM-LABEL: @add_v16i32(
-; SLM-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
-; SLM-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
-; SLM-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
-; SLM-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
-; SLM-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
-; SLM-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
-; SLM-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
-; SLM-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
-; SLM-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
-; SLM-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; SLM-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; SLM-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; SLM-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; SLM-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; SLM-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; SLM-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; SLM-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; SLM-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; SLM-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; SLM-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; SLM-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; SLM-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; SLM-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; SLM-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; SLM-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; SLM-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; SLM-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; SLM-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; SLM-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; SLM-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; SLM-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; SLM-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; SLM-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; SLM-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; SLM-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; SLM-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; SLM-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; SLM-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; SLM-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; SLM-NEXT:    [[R0:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A0]], i32 [[B0]])
+; SLM-NEXT:    [[R1:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A1]], i32 [[B1]])
+; SLM-NEXT:    [[R2:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A2]], i32 [[B2]])
+; SLM-NEXT:    [[R3:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A3]], i32 [[B3]])
+; SLM-NEXT:    [[R4:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A4]], i32 [[B4]])
+; SLM-NEXT:    [[R5:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A5]], i32 [[B5]])
+; SLM-NEXT:    [[R6:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A6]], i32 [[B6]])
+; SLM-NEXT:    [[R7:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A7]], i32 [[B7]])
+; SLM-NEXT:    [[R8:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A8]], i32 [[B8]])
+; SLM-NEXT:    [[R9:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A9]], i32 [[B9]])
+; SLM-NEXT:    [[R10:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A10]], i32 [[B10]])
+; SLM-NEXT:    [[R11:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A11]], i32 [[B11]])
+; SLM-NEXT:    [[R12:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A12]], i32 [[B12]])
+; SLM-NEXT:    [[R13:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A13]], i32 [[B13]])
+; SLM-NEXT:    [[R14:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A14]], i32 [[B14]])
+; SLM-NEXT:    [[R15:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A15]], i32 [[B15]])
+; SLM-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; SLM-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; SLM-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; SLM-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; SLM-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; SLM-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; SLM-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; SLM-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; SLM-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; SLM-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; SLM-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; SLM-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; SLM-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SLM-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; SLM-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; SLM-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; SLM-NEXT:    ret void
 ;
 ; AVX-LABEL: @add_v16i32(
-; AVX-NEXT:    [[TMP1:%.*]] = load <8 x i32>, ptr @a32, align 4
-; AVX-NEXT:    [[TMP2:%.*]] = load <8 x i32>, ptr @b32, align 4
-; AVX-NEXT:    [[TMP3:%.*]] = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> [[TMP1]], <8 x i32> [[TMP2]])
-; AVX-NEXT:    store <8 x i32> [[TMP3]], ptr @c32, align 4
-; AVX-NEXT:    [[TMP4:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; AVX-NEXT:    [[TMP5:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; AVX-NEXT:    [[TMP6:%.*]] = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
-; AVX-NEXT:    store <8 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; AVX-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; AVX-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; AVX-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; AVX-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; AVX-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; AVX-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; AVX-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; AVX-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; AVX-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; AVX-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; AVX-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; AVX-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; AVX-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; AVX-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; AVX-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; AVX-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; AVX-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; AVX-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; AVX-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; AVX-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; AVX-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; AVX-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; AVX-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; AVX-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; AVX-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; AVX-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; AVX-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; AVX-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; AVX-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; AVX-NEXT:    [[R0:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A0]], i32 [[B0]])
+; AVX-NEXT:    [[R1:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A1]], i32 [[B1]])
+; AVX-NEXT:    [[R2:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A2]], i32 [[B2]])
+; AVX-NEXT:    [[R3:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A3]], i32 [[B3]])
+; AVX-NEXT:    [[R4:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A4]], i32 [[B4]])
+; AVX-NEXT:    [[R5:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A5]], i32 [[B5]])
+; AVX-NEXT:    [[R6:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A6]], i32 [[B6]])
+; AVX-NEXT:    [[R7:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A7]], i32 [[B7]])
+; AVX-NEXT:    [[R8:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A8]], i32 [[B8]])
+; AVX-NEXT:    [[R9:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A9]], i32 [[B9]])
+; AVX-NEXT:    [[R10:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A10]], i32 [[B10]])
+; AVX-NEXT:    [[R11:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A11]], i32 [[B11]])
+; AVX-NEXT:    [[R12:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A12]], i32 [[B12]])
+; AVX-NEXT:    [[R13:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A13]], i32 [[B13]])
+; AVX-NEXT:    [[R14:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A14]], i32 [[B14]])
+; AVX-NEXT:    [[R15:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A15]], i32 [[B15]])
+; AVX-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; AVX-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; AVX-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; AVX-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; AVX-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; AVX-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; AVX-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; AVX-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; AVX-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; AVX-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; AVX-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; AVX-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; AVX-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; AVX-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; AVX-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; AVX-NEXT:    ret void
 ;
 ; AVX512-LABEL: @add_v16i32(
-; AVX512-NEXT:    [[TMP1:%.*]] = load <16 x i32>, ptr @a32, align 4
-; AVX512-NEXT:    [[TMP2:%.*]] = load <16 x i32>, ptr @b32, align 4
-; AVX512-NEXT:    [[TMP3:%.*]] = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> [[TMP1]], <16 x i32> [[TMP2]])
-; AVX512-NEXT:    store <16 x i32> [[TMP3]], ptr @c32, align 4
+; AVX512-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; AVX512-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; AVX512-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; AVX512-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; AVX512-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; AVX512-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; AVX512-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; AVX512-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; AVX512-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; AVX512-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; AVX512-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; AVX512-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; AVX512-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; AVX512-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; AVX512-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; AVX512-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; AVX512-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; AVX512-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; AVX512-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; AVX512-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; AVX512-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; AVX512-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; AVX512-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; AVX512-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; AVX512-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; AVX512-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; AVX512-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; AVX512-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; AVX512-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; AVX512-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; AVX512-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; AVX512-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; AVX512-NEXT:    [[R0:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A0]], i32 [[B0]])
+; AVX512-NEXT:    [[R1:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A1]], i32 [[B1]])
+; AVX512-NEXT:    [[R2:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A2]], i32 [[B2]])
+; AVX512-NEXT:    [[R3:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A3]], i32 [[B3]])
+; AVX512-NEXT:    [[R4:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A4]], i32 [[B4]])
+; AVX512-NEXT:    [[R5:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A5]], i32 [[B5]])
+; AVX512-NEXT:    [[R6:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A6]], i32 [[B6]])
+; AVX512-NEXT:    [[R7:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A7]], i32 [[B7]])
+; AVX512-NEXT:    [[R8:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A8]], i32 [[B8]])
+; AVX512-NEXT:    [[R9:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A9]], i32 [[B9]])
+; AVX512-NEXT:    [[R10:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A10]], i32 [[B10]])
+; AVX512-NEXT:    [[R11:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A11]], i32 [[B11]])
+; AVX512-NEXT:    [[R12:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A12]], i32 [[B12]])
+; AVX512-NEXT:    [[R13:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A13]], i32 [[B13]])
+; AVX512-NEXT:    [[R14:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A14]], i32 [[B14]])
+; AVX512-NEXT:    [[R15:%.*]] = call i32 @llvm.sadd.sat.i32(i32 [[A15]], i32 [[B15]])
+; AVX512-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; AVX512-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; AVX512-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; AVX512-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; AVX512-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; AVX512-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; AVX512-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; AVX512-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; AVX512-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX512-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; AVX512-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; AVX512-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; AVX512-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; AVX512-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; AVX512-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; AVX512-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; AVX512-NEXT:    ret void
 ;
   %a0  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 0 ), align 4
@@ -805,3 +1002,6 @@ define void @add_v64i8() {
   store i8 %r63, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 63), align 1
   ret void
 }
+;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
+; AVX1: {{.*}}
+; AVX2: {{.*}}
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/arith-add-usat.ll b/llvm/test/Transforms/SLPVectorizer/X86/arith-add-usat.ll
index fab022d691c07..913a3aa1de676 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/arith-add-usat.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/arith-add-usat.ll
@@ -26,41 +26,6 @@ declare i16 @llvm.uadd.sat.i16(i16, i16)
 declare i8  @llvm.uadd.sat.i8 (i8 , i8 )
 
 define void @add_v8i64() {
-; SSE-LABEL: @add_v8i64(
-; SSE-NEXT:    [[A0:%.*]] = load i64, ptr @a64, align 8
-; SSE-NEXT:    [[A1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
-; SSE-NEXT:    [[A2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; SSE-NEXT:    [[A3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
-; SSE-NEXT:    [[A4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; SSE-NEXT:    [[A5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
-; SSE-NEXT:    [[A6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; SSE-NEXT:    [[A7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
-; SSE-NEXT:    [[B0:%.*]] = load i64, ptr @b64, align 8
-; SSE-NEXT:    [[B1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
-; SSE-NEXT:    [[B2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; SSE-NEXT:    [[B3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
-; SSE-NEXT:    [[B4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; SSE-NEXT:    [[B5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
-; SSE-NEXT:    [[B6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; SSE-NEXT:    [[B7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
-; SSE-NEXT:    [[R0:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A0]], i64 [[B0]])
-; SSE-NEXT:    [[R1:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A1]], i64 [[B1]])
-; SSE-NEXT:    [[R2:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A2]], i64 [[B2]])
-; SSE-NEXT:    [[R3:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A3]], i64 [[B3]])
-; SSE-NEXT:    [[R4:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A4]], i64 [[B4]])
-; SSE-NEXT:    [[R5:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A5]], i64 [[B5]])
-; SSE-NEXT:    [[R6:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A6]], i64 [[B6]])
-; SSE-NEXT:    [[R7:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[A7]], i64 [[B7]])
-; SSE-NEXT:    store i64 [[R0]], ptr @c64, align 8
-; SSE-NEXT:    store i64 [[R1]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
-; SSE-NEXT:    store i64 [[R2]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; SSE-NEXT:    store i64 [[R3]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
-; SSE-NEXT:    store i64 [[R4]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; SSE-NEXT:    store i64 [[R5]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
-; SSE-NEXT:    store i64 [[R6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; SSE-NEXT:    store i64 [[R7]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
-; SSE-NEXT:    ret void
-;
 ; AVX-LABEL: @add_v8i64(
 ; AVX-NEXT:    [[TMP1:%.*]] = load <4 x i64>, ptr @a64, align 8
 ; AVX-NEXT:    [[TMP2:%.*]] = load <4 x i64>, ptr @b64, align 8
@@ -115,25 +80,6 @@ define void @add_v8i64() {
 }
 
 define void @add_v16i32() {
-; SSE-LABEL: @add_v16i32(
-; SSE-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
-; SSE-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
-; SSE-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
-; SSE-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
-; SSE-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
-; SSE-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
-; SSE-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
-; SSE-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
-; SSE-NEXT:    ret void
-;
 ; AVX-LABEL: @add_v16i32(
 ; AVX-NEXT:    [[TMP1:%.*]] = load <8 x i32>, ptr @a32, align 4
 ; AVX-NEXT:    [[TMP2:%.*]] = load <8 x i32>, ptr @b32, align 4
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-ssat.ll b/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-ssat.ll
index afaab8b8ca642..4aa02c6b65ccc 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-ssat.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-ssat.ll
@@ -62,75 +62,60 @@ define void @sub_v8i64() {
 ; SSE-NEXT:    ret void
 ;
 ; SLM-LABEL: @sub_v8i64(
-; SLM-NEXT:    [[A0:%.*]] = load i64, ptr @a64, align 8
-; SLM-NEXT:    [[A1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
-; SLM-NEXT:    [[A2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; SLM-NEXT:    [[A3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
-; SLM-NEXT:    [[A4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; SLM-NEXT:    [[A5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
-; SLM-NEXT:    [[A6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; SLM-NEXT:    [[A7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
-; SLM-NEXT:    [[B0:%.*]] = load i64, ptr @b64, align 8
-; SLM-NEXT:    [[B1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
-; SLM-NEXT:    [[B2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; SLM-NEXT:    [[B3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
-; SLM-NEXT:    [[B4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; SLM-NEXT:    [[B5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
-; SLM-NEXT:    [[B6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; SLM-NEXT:    [[B7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
-; SLM-NEXT:    [[R0:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A0]], i64 [[B0]])
-; SLM-NEXT:    [[R1:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A1]], i64 [[B1]])
-; SLM-NEXT:    [[R2:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A2]], i64 [[B2]])
-; SLM-NEXT:    [[R3:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A3]], i64 [[B3]])
-; SLM-NEXT:    [[R4:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A4]], i64 [[B4]])
-; SLM-NEXT:    [[R5:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A5]], i64 [[B5]])
-; SLM-NEXT:    [[R6:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A6]], i64 [[B6]])
-; SLM-NEXT:    [[R7:%.*]] = call i64 @llvm.ssub.sat.i64(i64 [[A7]], i64 [[B7]])
-; SLM-NEXT:    store i64 [[R0]], ptr @c64, align 8
-; SLM-NEXT:    store i64 [[R1]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
-; SLM-NEXT:    store i64 [[R2]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; SLM-NEXT:    store i64 [[R3]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
-; SLM-NEXT:    store i64 [[R4]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; SLM-NEXT:    store i64 [[R5]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
-; SLM-NEXT:    store i64 [[R6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; SLM-NEXT:    store i64 [[R7]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
+; SLM-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; SLM-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; SLM-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; SLM-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; SLM-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; SLM-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; SLM-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; SLM-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; SLM-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; SLM-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; SLM-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; SLM-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
 ; SLM-NEXT:    ret void
 ;
-; AVX1-LABEL: @sub_v8i64(
-; AVX1-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
-; AVX1-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
-; AVX1-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
-; AVX1-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
-; AVX1-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
-; AVX1-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; AVX1-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
-; AVX1-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; AVX1-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; AVX1-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; AVX1-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
-; AVX1-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; AVX1-NEXT:    ret void
-;
-; AVX2-LABEL: @sub_v8i64(
-; AVX2-NEXT:    [[TMP1:%.*]] = load <4 x i64>, ptr @a64, align 8
-; AVX2-NEXT:    [[TMP2:%.*]] = load <4 x i64>, ptr @b64, align 8
-; AVX2-NEXT:    [[TMP3:%.*]] = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> [[TMP1]], <4 x i64> [[TMP2]])
-; AVX2-NEXT:    store <4 x i64> [[TMP3]], ptr @c64, align 8
-; AVX2-NEXT:    [[TMP4:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; AVX2-NEXT:    [[TMP5:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; AVX2-NEXT:    [[TMP6:%.*]] = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> [[TMP4]], <4 x i64> [[TMP5]])
-; AVX2-NEXT:    store <4 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; AVX2-NEXT:    ret void
+; AVX-LABEL: @sub_v8i64(
+; AVX-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; AVX-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; AVX-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; AVX-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; AVX-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; AVX-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; AVX-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; AVX-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; AVX-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; AVX-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; AVX-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; AVX-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
+; AVX-NEXT:    ret void
 ;
 ; AVX512-LABEL: @sub_v8i64(
-; AVX512-NEXT:    [[TMP1:%.*]] = load <8 x i64>, ptr @a64, align 8
-; AVX512-NEXT:    [[TMP2:%.*]] = load <8 x i64>, ptr @b64, align 8
-; AVX512-NEXT:    [[TMP3:%.*]] = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> [[TMP1]], <8 x i64> [[TMP2]])
-; AVX512-NEXT:    store <8 x i64> [[TMP3]], ptr @c64, align 8
+; AVX512-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; AVX512-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; AVX512-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; AVX512-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; AVX512-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; AVX512-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; AVX512-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; AVX512-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; AVX512-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; AVX512-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; AVX512-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; AVX512-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
 ; AVX512-NEXT:    ret void
 ;
 ; AVX256BW-LABEL: @sub_v8i64(
@@ -180,59 +165,271 @@ define void @sub_v8i64() {
 
 define void @sub_v16i32() {
 ; SSE-LABEL: @sub_v16i32(
-; SSE-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
-; SSE-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
-; SSE-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
-; SSE-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
-; SSE-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
-; SSE-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
-; SSE-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
-; SSE-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; SSE-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; SSE-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; SSE-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; SSE-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; SSE-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; SSE-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; SSE-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; SSE-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; SSE-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; SSE-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; SSE-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; SSE-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; SSE-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; SSE-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; SSE-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; SSE-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; SSE-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; SSE-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; SSE-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; SSE-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; SSE-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; SSE-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; SSE-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; SSE-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; SSE-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; SSE-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; SSE-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; SSE-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; SSE-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; SSE-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; SSE-NEXT:    [[R0:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A0]], i32 [[B0]])
+; SSE-NEXT:    [[R1:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A1]], i32 [[B1]])
+; SSE-NEXT:    [[R2:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A2]], i32 [[B2]])
+; SSE-NEXT:    [[R3:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A3]], i32 [[B3]])
+; SSE-NEXT:    [[R4:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A4]], i32 [[B4]])
+; SSE-NEXT:    [[R5:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A5]], i32 [[B5]])
+; SSE-NEXT:    [[R6:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A6]], i32 [[B6]])
+; SSE-NEXT:    [[R7:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A7]], i32 [[B7]])
+; SSE-NEXT:    [[R8:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A8]], i32 [[B8]])
+; SSE-NEXT:    [[R9:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A9]], i32 [[B9]])
+; SSE-NEXT:    [[R10:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A10]], i32 [[B10]])
+; SSE-NEXT:    [[R11:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A11]], i32 [[B11]])
+; SSE-NEXT:    [[R12:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A12]], i32 [[B12]])
+; SSE-NEXT:    [[R13:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A13]], i32 [[B13]])
+; SSE-NEXT:    [[R14:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A14]], i32 [[B14]])
+; SSE-NEXT:    [[R15:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A15]], i32 [[B15]])
+; SSE-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; SSE-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; SSE-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; SSE-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; SSE-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; SSE-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; SSE-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; SSE-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; SSE-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; SSE-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; SSE-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; SSE-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; SSE-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SSE-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; SSE-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; SSE-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; SSE-NEXT:    ret void
 ;
 ; SLM-LABEL: @sub_v16i32(
-; SLM-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
-; SLM-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
-; SLM-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
-; SLM-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
-; SLM-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
-; SLM-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
-; SLM-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
-; SLM-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
-; SLM-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
-; SLM-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
-; SLM-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
-; SLM-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; SLM-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; SLM-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; SLM-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; SLM-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; SLM-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; SLM-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; SLM-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; SLM-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; SLM-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; SLM-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; SLM-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; SLM-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; SLM-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; SLM-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; SLM-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; SLM-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; SLM-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; SLM-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; SLM-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; SLM-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; SLM-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; SLM-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; SLM-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; SLM-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; SLM-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; SLM-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; SLM-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; SLM-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; SLM-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; SLM-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; SLM-NEXT:    [[R0:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A0]], i32 [[B0]])
+; SLM-NEXT:    [[R1:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A1]], i32 [[B1]])
+; SLM-NEXT:    [[R2:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A2]], i32 [[B2]])
+; SLM-NEXT:    [[R3:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A3]], i32 [[B3]])
+; SLM-NEXT:    [[R4:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A4]], i32 [[B4]])
+; SLM-NEXT:    [[R5:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A5]], i32 [[B5]])
+; SLM-NEXT:    [[R6:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A6]], i32 [[B6]])
+; SLM-NEXT:    [[R7:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A7]], i32 [[B7]])
+; SLM-NEXT:    [[R8:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A8]], i32 [[B8]])
+; SLM-NEXT:    [[R9:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A9]], i32 [[B9]])
+; SLM-NEXT:    [[R10:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A10]], i32 [[B10]])
+; SLM-NEXT:    [[R11:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A11]], i32 [[B11]])
+; SLM-NEXT:    [[R12:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A12]], i32 [[B12]])
+; SLM-NEXT:    [[R13:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A13]], i32 [[B13]])
+; SLM-NEXT:    [[R14:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A14]], i32 [[B14]])
+; SLM-NEXT:    [[R15:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A15]], i32 [[B15]])
+; SLM-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; SLM-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; SLM-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; SLM-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; SLM-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; SLM-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; SLM-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; SLM-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; SLM-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; SLM-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; SLM-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; SLM-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; SLM-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; SLM-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; SLM-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; SLM-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; SLM-NEXT:    ret void
 ;
 ; AVX-LABEL: @sub_v16i32(
-; AVX-NEXT:    [[TMP1:%.*]] = load <8 x i32>, ptr @a32, align 4
-; AVX-NEXT:    [[TMP2:%.*]] = load <8 x i32>, ptr @b32, align 4
-; AVX-NEXT:    [[TMP3:%.*]] = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> [[TMP1]], <8 x i32> [[TMP2]])
-; AVX-NEXT:    store <8 x i32> [[TMP3]], ptr @c32, align 4
-; AVX-NEXT:    [[TMP4:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
-; AVX-NEXT:    [[TMP5:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
-; AVX-NEXT:    [[TMP6:%.*]] = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
-; AVX-NEXT:    store <8 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; AVX-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; AVX-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; AVX-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; AVX-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; AVX-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; AVX-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; AVX-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; AVX-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; AVX-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; AVX-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; AVX-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; AVX-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; AVX-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; AVX-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; AVX-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; AVX-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; AVX-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; AVX-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; AVX-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; AVX-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; AVX-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; AVX-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; AVX-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; AVX-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; AVX-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; AVX-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; AVX-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; AVX-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; AVX-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; AVX-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; AVX-NEXT:    [[R0:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A0]], i32 [[B0]])
+; AVX-NEXT:    [[R1:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A1]], i32 [[B1]])
+; AVX-NEXT:    [[R2:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A2]], i32 [[B2]])
+; AVX-NEXT:    [[R3:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A3]], i32 [[B3]])
+; AVX-NEXT:    [[R4:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A4]], i32 [[B4]])
+; AVX-NEXT:    [[R5:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A5]], i32 [[B5]])
+; AVX-NEXT:    [[R6:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A6]], i32 [[B6]])
+; AVX-NEXT:    [[R7:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A7]], i32 [[B7]])
+; AVX-NEXT:    [[R8:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A8]], i32 [[B8]])
+; AVX-NEXT:    [[R9:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A9]], i32 [[B9]])
+; AVX-NEXT:    [[R10:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A10]], i32 [[B10]])
+; AVX-NEXT:    [[R11:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A11]], i32 [[B11]])
+; AVX-NEXT:    [[R12:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A12]], i32 [[B12]])
+; AVX-NEXT:    [[R13:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A13]], i32 [[B13]])
+; AVX-NEXT:    [[R14:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A14]], i32 [[B14]])
+; AVX-NEXT:    [[R15:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A15]], i32 [[B15]])
+; AVX-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; AVX-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; AVX-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; AVX-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; AVX-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; AVX-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; AVX-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; AVX-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; AVX-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; AVX-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; AVX-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; AVX-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; AVX-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; AVX-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; AVX-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; AVX-NEXT:    ret void
 ;
 ; AVX512-LABEL: @sub_v16i32(
-; AVX512-NEXT:    [[TMP1:%.*]] = load <16 x i32>, ptr @a32, align 4
-; AVX512-NEXT:    [[TMP2:%.*]] = load <16 x i32>, ptr @b32, align 4
-; AVX512-NEXT:    [[TMP3:%.*]] = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> [[TMP1]], <16 x i32> [[TMP2]])
-; AVX512-NEXT:    store <16 x i32> [[TMP3]], ptr @c32, align 4
+; AVX512-NEXT:    [[A0:%.*]] = load i32, ptr @a32, align 4
+; AVX512-NEXT:    [[A1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1), align 4
+; AVX512-NEXT:    [[A2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2), align 4
+; AVX512-NEXT:    [[A3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3), align 4
+; AVX512-NEXT:    [[A4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
+; AVX512-NEXT:    [[A5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5), align 4
+; AVX512-NEXT:    [[A6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6), align 4
+; AVX512-NEXT:    [[A7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7), align 4
+; AVX512-NEXT:    [[A8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
+; AVX512-NEXT:    [[A9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9), align 4
+; AVX512-NEXT:    [[A10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
+; AVX512-NEXT:    [[A11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
+; AVX512-NEXT:    [[A12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
+; AVX512-NEXT:    [[A13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
+; AVX512-NEXT:    [[A14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
+; AVX512-NEXT:    [[A15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
+; AVX512-NEXT:    [[B0:%.*]] = load i32, ptr @b32, align 4
+; AVX512-NEXT:    [[B1:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1), align 4
+; AVX512-NEXT:    [[B2:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2), align 4
+; AVX512-NEXT:    [[B3:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3), align 4
+; AVX512-NEXT:    [[B4:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
+; AVX512-NEXT:    [[B5:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5), align 4
+; AVX512-NEXT:    [[B6:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6), align 4
+; AVX512-NEXT:    [[B7:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7), align 4
+; AVX512-NEXT:    [[B8:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
+; AVX512-NEXT:    [[B9:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9), align 4
+; AVX512-NEXT:    [[B10:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
+; AVX512-NEXT:    [[B11:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
+; AVX512-NEXT:    [[B12:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
+; AVX512-NEXT:    [[B13:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
+; AVX512-NEXT:    [[B14:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
+; AVX512-NEXT:    [[B15:%.*]] = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
+; AVX512-NEXT:    [[R0:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A0]], i32 [[B0]])
+; AVX512-NEXT:    [[R1:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A1]], i32 [[B1]])
+; AVX512-NEXT:    [[R2:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A2]], i32 [[B2]])
+; AVX512-NEXT:    [[R3:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A3]], i32 [[B3]])
+; AVX512-NEXT:    [[R4:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A4]], i32 [[B4]])
+; AVX512-NEXT:    [[R5:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A5]], i32 [[B5]])
+; AVX512-NEXT:    [[R6:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A6]], i32 [[B6]])
+; AVX512-NEXT:    [[R7:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A7]], i32 [[B7]])
+; AVX512-NEXT:    [[R8:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A8]], i32 [[B8]])
+; AVX512-NEXT:    [[R9:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A9]], i32 [[B9]])
+; AVX512-NEXT:    [[R10:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A10]], i32 [[B10]])
+; AVX512-NEXT:    [[R11:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A11]], i32 [[B11]])
+; AVX512-NEXT:    [[R12:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A12]], i32 [[B12]])
+; AVX512-NEXT:    [[R13:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A13]], i32 [[B13]])
+; AVX512-NEXT:    [[R14:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A14]], i32 [[B14]])
+; AVX512-NEXT:    [[R15:%.*]] = call i32 @llvm.ssub.sat.i32(i32 [[A15]], i32 [[B15]])
+; AVX512-NEXT:    store i32 [[R0]], ptr @c32, align 4
+; AVX512-NEXT:    store i32 [[R1]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1), align 4
+; AVX512-NEXT:    store i32 [[R2]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2), align 4
+; AVX512-NEXT:    store i32 [[R3]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3), align 4
+; AVX512-NEXT:    store i32 [[R4]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
+; AVX512-NEXT:    store i32 [[R5]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5), align 4
+; AVX512-NEXT:    store i32 [[R6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6), align 4
+; AVX512-NEXT:    store i32 [[R7]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7), align 4
+; AVX512-NEXT:    store i32 [[R8]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
+; AVX512-NEXT:    store i32 [[R9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9), align 4
+; AVX512-NEXT:    store i32 [[R10]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
+; AVX512-NEXT:    store i32 [[R11]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
+; AVX512-NEXT:    store i32 [[R12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
+; AVX512-NEXT:    store i32 [[R13]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
+; AVX512-NEXT:    store i32 [[R14]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
+; AVX512-NEXT:    store i32 [[R15]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
 ; AVX512-NEXT:    ret void
 ;
   %a0  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 0 ), align 4
@@ -805,3 +1002,6 @@ define void @sub_v64i8() {
   store i8 %r63, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 63), align 1
   ret void
 }
+;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
+; AVX1: {{.*}}
+; AVX2: {{.*}}
diff --git a/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-usat.ll b/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-usat.ll
index 3510863c88930..2638b682a6809 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-usat.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/arith-sub-usat.ll
@@ -27,38 +27,22 @@ declare i8  @llvm.usub.sat.i8 (i8 , i8 )
 
 define void @sub_v8i64() {
 ; SSE-LABEL: @sub_v8i64(
-; SSE-NEXT:    [[A0:%.*]] = load i64, ptr @a64, align 8
-; SSE-NEXT:    [[A1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
-; SSE-NEXT:    [[A2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
-; SSE-NEXT:    [[A3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
-; SSE-NEXT:    [[A4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
-; SSE-NEXT:    [[A5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
-; SSE-NEXT:    [[A6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
-; SSE-NEXT:    [[A7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
-; SSE-NEXT:    [[B0:%.*]] = load i64, ptr @b64, align 8
-; SSE-NEXT:    [[B1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
-; SSE-NEXT:    [[B2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
-; SSE-NEXT:    [[B3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
-; SSE-NEXT:    [[B4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
-; SSE-NEXT:    [[B5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
-; SSE-NEXT:    [[B6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
-; SSE-NEXT:    [[B7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
-; SSE-NEXT:    [[R0:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A0]], i64 [[B0]])
-; SSE-NEXT:    [[R1:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A1]], i64 [[B1]])
-; SSE-NEXT:    [[R2:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A2]], i64 [[B2]])
-; SSE-NEXT:    [[R3:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A3]], i64 [[B3]])
-; SSE-NEXT:    [[R4:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A4]], i64 [[B4]])
-; SSE-NEXT:    [[R5:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A5]], i64 [[B5]])
-; SSE-NEXT:    [[R6:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A6]], i64 [[B6]])
-; SSE-NEXT:    [[R7:%.*]] = call i64 @llvm.usub.sat.i64(i64 [[A7]], i64 [[B7]])
-; SSE-NEXT:    store i64 [[R0]], ptr @c64, align 8
-; SSE-NEXT:    store i64 [[R1]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
-; SSE-NEXT:    store i64 [[R2]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
-; SSE-NEXT:    store i64 [[R3]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
-; SSE-NEXT:    store i64 [[R4]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
-; SSE-NEXT:    store i64 [[R5]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
-; SSE-NEXT:    store i64 [[R6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
-; SSE-NEXT:    store i64 [[R7]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
+; SSE-NEXT:    [[TMP1:%.*]] = load <2 x i64>, ptr @a64, align 8
+; SSE-NEXT:    [[TMP2:%.*]] = load <2 x i64>, ptr @b64, align 8
+; SSE-NEXT:    [[TMP3:%.*]] = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> [[TMP1]], <2 x i64> [[TMP2]])
+; SSE-NEXT:    store <2 x i64> [[TMP3]], ptr @c64, align 8
+; SSE-NEXT:    [[TMP4:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
+; SSE-NEXT:    [[TMP5:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
+; SSE-NEXT:    [[TMP6:%.*]] = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> [[TMP4]], <2 x i64> [[TMP5]])
+; SSE-NEXT:    store <2 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
+; SSE-NEXT:    [[TMP7:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
+; SSE-NEXT:    [[TMP8:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
+; SSE-NEXT:    [[TMP9:%.*]] = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]])
+; SSE-NEXT:    store <2 x i64> [[TMP9]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
+; SSE-NEXT:    [[TMP10:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
+; SSE-NEXT:    [[TMP11:%.*]] = load <2 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
+; SSE-NEXT:    [[TMP12:%.*]] = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> [[TMP10]], <2 x i64> [[TMP11]])
+; SSE-NEXT:    store <2 x i64> [[TMP12]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
 ; SSE-NEXT:    ret void
 ;
 ; AVX-LABEL: @sub_v8i64(



More information about the llvm-commits mailing list