[llvm] e5c64b4 - [CostModel][AMDGPU] Fix intrinsics costs estimations.

Daniil Fukalov via llvm-commits llvm-commits at lists.llvm.org
Mon Dec 13 06:17:56 PST 2021


Author: Daniil Fukalov
Date: 2021-12-13T17:17:34+03:00
New Revision: e5c64b45be396a3f6e1fa7c6a3ffb7b5614edaca

URL: https://github.com/llvm/llvm-project/commit/e5c64b45be396a3f6e1fa7c6a3ffb7b5614edaca
DIFF: https://github.com/llvm/llvm-project/commit/e5c64b45be396a3f6e1fa7c6a3ffb7b5614edaca.diff

LOG: [CostModel][AMDGPU] Fix intrinsics costs estimations.

1. Fixed costs inconsistency for llvm.fma.vXf16 instinsiscs.
2. Added tests for llvm.sadd.sat, llvm.ssub.sat, llvm.uadd.sat, llvm.usub.sat
   intrisics since they have special processing in cost model.
3. Minor intrisics' costs tests updat and refinement.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D115385

Added: 
    llvm/test/Analysis/CostModel/AMDGPU/arith-ssat.ll
    llvm/test/Analysis/CostModel/AMDGPU/arith-usat.ll

Modified: 
    llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
    llvm/test/Analysis/CostModel/AMDGPU/fma.ll
    llvm/test/Analysis/CostModel/AMDGPU/reduce-and.ll
    llvm/test/Analysis/CostModel/AMDGPU/reduce-or.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp b/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
index 4a31a598fc789..09c5eb192e1f6 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
@@ -691,40 +691,6 @@ GCNTTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
     return BaseT::getIntrinsicInstrCost(ICA, CostKind);
 
   Type *RetTy = ICA.getReturnType();
-  EVT OrigTy = TLI->getValueType(DL, RetTy);
-  if (!OrigTy.isSimple()) {
-    if (CostKind != TTI::TCK_CodeSize)
-      return BaseT::getIntrinsicInstrCost(ICA, CostKind);
-
-    // TODO: Combine these two logic paths.
-    if (ICA.isTypeBasedOnly())
-      return getTypeBasedIntrinsicInstrCost(ICA, CostKind);
-
-    unsigned RetVF =
-        (RetTy->isVectorTy() ? cast<FixedVectorType>(RetTy)->getNumElements()
-                             : 1);
-    const IntrinsicInst *I = ICA.getInst();
-    const SmallVectorImpl<const Value *> &Args = ICA.getArgs();
-    FastMathFlags FMF = ICA.getFlags();
-    // Assume that we need to scalarize this intrinsic.
-
-    // Compute the scalarization overhead based on Args for a vector
-    // intrinsic. A vectorizer will pass a scalar RetTy and VF > 1, while
-    // CostModel will pass a vector RetTy and VF is 1.
-    InstructionCost ScalarizationCost = InstructionCost::getInvalid();
-    if (RetVF > 1) {
-      ScalarizationCost = 0;
-      if (!RetTy->isVoidTy())
-        ScalarizationCost +=
-            getScalarizationOverhead(cast<VectorType>(RetTy), true, false);
-      ScalarizationCost +=
-          getOperandsScalarizationOverhead(Args, ICA.getArgTypes());
-    }
-
-    IntrinsicCostAttributes Attrs(ICA.getID(), RetTy, ICA.getArgTypes(), FMF, I,
-                                  ScalarizationCost);
-    return getIntrinsicInstrCost(Attrs, CostKind);
-  }
 
   // Legalize the type.
   std::pair<InstructionCost, MVT> LT = TLI->getTypeLegalizationCost(DL, RetTy);

diff  --git a/llvm/test/Analysis/CostModel/AMDGPU/arith-ssat.ll b/llvm/test/Analysis/CostModel/AMDGPU/arith-ssat.ll
new file mode 100644
index 0000000000000..54636e70502ba
--- /dev/null
+++ b/llvm/test/Analysis/CostModel/AMDGPU/arith-ssat.ll
@@ -0,0 +1,370 @@
+; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW-SIZE %s
+; END.
+
+declare i64        @llvm.sadd.sat.i64(i64, i64)
+declare <2 x i64>  @llvm.sadd.sat.v2i64(<2 x i64>, <2 x i64>)
+declare <4 x i64>  @llvm.sadd.sat.v4i64(<4 x i64>, <4 x i64>)
+declare <5 x i64>  @llvm.sadd.sat.v5i64(<5 x i64>, <5 x i64>)
+declare <8 x i64>  @llvm.sadd.sat.v8i64(<8 x i64>, <8 x i64>)
+
+declare i32        @llvm.sadd.sat.i32(i32, i32)
+declare <2 x i32>  @llvm.sadd.sat.v2i32(<2 x i32>, <2 x i32>)
+declare <4 x i32>  @llvm.sadd.sat.v4i32(<4 x i32>, <4 x i32>)
+declare <8 x i32>  @llvm.sadd.sat.v8i32(<8 x i32>, <8 x i32>)
+declare <9 x i32>  @llvm.sadd.sat.v9i32(<9 x i32>, <9 x i32>)
+declare <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32>, <16 x i32>)
+
+declare i16        @llvm.sadd.sat.i16(i16, i16)
+declare <2 x i16>  @llvm.sadd.sat.v2i16(<2 x i16>, <2 x i16>)
+declare <4 x i16>  @llvm.sadd.sat.v4i16(<4 x i16>, <4 x i16>)
+declare <8 x i16>  @llvm.sadd.sat.v8i16(<8 x i16>, <8 x i16>)
+declare <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16>, <16 x i16>)
+declare <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16>, <17 x i16>)
+declare <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16>, <32 x i16>)
+
+declare i8         @llvm.sadd.sat.i8(i8,  i8)
+declare <2 x i8>   @llvm.sadd.sat.v2i8(<2 x i8>, <2 x i8>)
+declare <4 x i8>   @llvm.sadd.sat.v4i8(<4 x i8>, <4 x i8>)
+declare <8 x i8>   @llvm.sadd.sat.v8i8(<8 x i8>, <8 x i8>)
+declare <16 x i8>  @llvm.sadd.sat.v16i8(<16 x i8>, <16 x i8>)
+declare <32 x i8>  @llvm.sadd.sat.v32i8(<32 x i8>, <32 x i8>)
+declare <33 x i8>  @llvm.sadd.sat.v33i8(<33 x i8>, <33 x i8>)
+declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
+
+define i32 @add(i32 %arg) {
+; FAST-LABEL: 'add'
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.sadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.sadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V17I16 = call <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.sadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; SLOW-LABEL: 'add'
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.sadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.sadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V17I16 = call <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.sadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; FAST-SIZE-LABEL: 'add'
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.sadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.sadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V17I16 = call <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.sadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; SLOW-SIZE-LABEL: 'add'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.sadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.sadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V17I16 = call <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.sadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-SIZE-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)
+  %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+  %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+  %V5I64 = call <5 x i64> @llvm.sadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+  %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+
+  %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+  %V2I32  = call <2 x i32>  @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+  %V4I32  = call <4 x i32>  @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+  %V8I32  = call <8 x i32>  @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+  %V9I32  = call <9 x i32>  @llvm.sadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+  %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+
+  %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+  %V2I16  = call <2 x i16>  @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+  %V4I16  = call <4 x i16>  @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+  %V8I16  = call <8 x i16>  @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+  %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+  %V17I16 = call <17 x i16> @llvm.sadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+  %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+
+  %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+  %V2I8  = call <2 x i8>  @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+  %V4I8  = call <4 x i8>  @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+  %V8I8  = call <8 x i8>  @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+  %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+  %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+  %V33I8 = call <33 x i8> @llvm.sadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+  %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+
+  ret i32 undef
+}
+
+declare i64        @llvm.ssub.sat.i64(i64, i64)
+declare <2 x i64>  @llvm.ssub.sat.v2i64(<2 x i64>, <2 x i64>)
+declare <4 x i64>  @llvm.ssub.sat.v4i64(<4 x i64>, <4 x i64>)
+declare <5 x i64>  @llvm.ssub.sat.v5i64(<5 x i64>, <5 x i64>)
+declare <8 x i64>  @llvm.ssub.sat.v8i64(<8 x i64>, <8 x i64>)
+
+declare i32        @llvm.ssub.sat.i32(i32, i32)
+declare <2 x i32>  @llvm.ssub.sat.v2i32(<2 x i32>, <2 x i32>)
+declare <4 x i32>  @llvm.ssub.sat.v4i32(<4 x i32>, <4 x i32>)
+declare <8 x i32>  @llvm.ssub.sat.v8i32(<8 x i32>, <8 x i32>)
+declare <9 x i32>  @llvm.ssub.sat.v9i32(<9 x i32>, <9 x i32>)
+declare <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32>, <16 x i32>)
+
+declare i16        @llvm.ssub.sat.i16(i16, i16)
+declare <2 x i16>  @llvm.ssub.sat.v2i16(<2 x i16>, <2 x i16>)
+declare <4 x i16>  @llvm.ssub.sat.v4i16(<4 x i16>, <4 x i16>)
+declare <8 x i16>  @llvm.ssub.sat.v8i16(<8 x i16>, <8 x i16>)
+declare <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16>, <16 x i16>)
+declare <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16>, <17 x i16>)
+declare <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16>, <32 x i16>)
+
+declare i8         @llvm.ssub.sat.i8(i8,  i8)
+declare <2 x i8>   @llvm.ssub.sat.v2i8(<2 x i8>, <2 x i8>)
+declare <4 x i8>   @llvm.ssub.sat.v4i8(<4 x i8>, <4 x i8>)
+declare <8 x i8>   @llvm.ssub.sat.v8i8(<8 x i8>, <8 x i8>)
+declare <16 x i8>  @llvm.ssub.sat.v16i8(<16 x i8>, <16 x i8>)
+declare <32 x i8>  @llvm.ssub.sat.v32i8(<32 x i8>, <32 x i8>)
+declare <33 x i8>  @llvm.ssub.sat.v33i8(<33 x i8>, <33 x i8>)
+declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
+
+define i32 @sub(i32 %arg) {
+; FAST-LABEL: 'sub'
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.ssub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.ssub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V17I16 = call <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.ssub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; SLOW-LABEL: 'sub'
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.ssub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.ssub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V17I16 = call <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.ssub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; FAST-SIZE-LABEL: 'sub'
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.ssub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.ssub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V17I16 = call <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.ssub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; SLOW-SIZE-LABEL: 'sub'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.ssub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.ssub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V17I16 = call <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.ssub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-SIZE-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)
+  %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+  %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+  %V5I64 = call <5 x i64> @llvm.ssub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+  %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+
+  %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+  %V2I32  = call <2 x i32>  @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+  %V4I32  = call <4 x i32>  @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+  %V8I32  = call <8 x i32>  @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+  %V9I32  = call <9 x i32>  @llvm.ssub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+  %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+
+  %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+  %V2I16  = call <2 x i16>  @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+  %V4I16  = call <4 x i16>  @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+  %V8I16  = call <8 x i16>  @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+  %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+  %V17I16 = call <17 x i16> @llvm.ssub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+  %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+
+  %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+  %V2I8  = call <2 x i8>  @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+  %V4I8  = call <4 x i8>  @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+  %V8I8  = call <8 x i8>  @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+  %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+  %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+  %V33I8 = call <33 x i8> @llvm.ssub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+  %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+
+  ret i32 undef
+}

diff  --git a/llvm/test/Analysis/CostModel/AMDGPU/arith-usat.ll b/llvm/test/Analysis/CostModel/AMDGPU/arith-usat.ll
new file mode 100644
index 0000000000000..3fe7583041e9a
--- /dev/null
+++ b/llvm/test/Analysis/CostModel/AMDGPU/arith-usat.ll
@@ -0,0 +1,370 @@
+; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefixes=FAST %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefixes=FAST-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW-SIZE %s
+; END.
+
+declare i64        @llvm.uadd.sat.i64(i64, i64)
+declare <2 x i64>  @llvm.uadd.sat.v2i64(<2 x i64>, <2 x i64>)
+declare <4 x i64>  @llvm.uadd.sat.v4i64(<4 x i64>, <4 x i64>)
+declare <5 x i64>  @llvm.uadd.sat.v5i64(<5 x i64>, <5 x i64>)
+declare <8 x i64>  @llvm.uadd.sat.v8i64(<8 x i64>, <8 x i64>)
+
+declare i32        @llvm.uadd.sat.i32(i32, i32)
+declare <2 x i32>  @llvm.uadd.sat.v2i32(<2 x i32>, <2 x i32>)
+declare <4 x i32>  @llvm.uadd.sat.v4i32(<4 x i32>, <4 x i32>)
+declare <8 x i32>  @llvm.uadd.sat.v8i32(<8 x i32>, <8 x i32>)
+declare <9 x i32>  @llvm.uadd.sat.v9i32(<9 x i32>, <9 x i32>)
+declare <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32>, <16 x i32>)
+
+declare i16        @llvm.uadd.sat.i16(i16, i16)
+declare <2 x i16>  @llvm.uadd.sat.v2i16(<2 x i16>, <2 x i16>)
+declare <4 x i16>  @llvm.uadd.sat.v4i16(<4 x i16>, <4 x i16>)
+declare <8 x i16>  @llvm.uadd.sat.v8i16(<8 x i16>, <8 x i16>)
+declare <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16>, <16 x i16>)
+declare <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16>, <17 x i16>)
+declare <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16>, <32 x i16>)
+
+declare i8         @llvm.uadd.sat.i8(i8,  i8)
+declare <2 x i8>   @llvm.uadd.sat.v2i8(<2 x i8>, <2 x i8>)
+declare <4 x i8>   @llvm.uadd.sat.v4i8(<4 x i8>, <4 x i8>)
+declare <8 x i8>   @llvm.uadd.sat.v8i8(<8 x i8>, <8 x i8>)
+declare <16 x i8>  @llvm.uadd.sat.v16i8(<16 x i8>, <16 x i8>)
+declare <32 x i8>  @llvm.uadd.sat.v32i8(<32 x i8>, <32 x i8>)
+declare <33 x i8>  @llvm.uadd.sat.v33i8(<33 x i8>, <33 x i8>)
+declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
+
+define i32 @add(i32 %arg) {
+; FAST-LABEL: 'add'
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.uadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.uadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V17I16 = call <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.uadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; SLOW-LABEL: 'add'
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.uadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.uadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V17I16 = call <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.uadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; FAST-SIZE-LABEL: 'add'
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.uadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.uadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V17I16 = call <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.uadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; SLOW-SIZE-LABEL: 'add'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.uadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.uadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V17I16 = call <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.uadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-SIZE-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)
+  %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+  %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+  %V5I64 = call <5 x i64> @llvm.uadd.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+  %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+
+  %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+  %V2I32  = call <2 x i32>  @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+  %V4I32  = call <4 x i32>  @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+  %V8I32  = call <8 x i32>  @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+  %V9I32  = call <9 x i32>  @llvm.uadd.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+  %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+
+  %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+  %V2I16  = call <2 x i16>  @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+  %V4I16  = call <4 x i16>  @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+  %V8I16  = call <8 x i16>  @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+  %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+  %V17I16 = call <17 x i16> @llvm.uadd.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+  %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+
+  %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+  %V2I8  = call <2 x i8>  @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+  %V4I8  = call <4 x i8>  @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+  %V8I8  = call <8 x i8>  @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+  %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+  %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+  %V33I8 = call <33 x i8> @llvm.uadd.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+  %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+
+  ret i32 undef
+}
+
+declare i64        @llvm.usub.sat.i64(i64, i64)
+declare <2 x i64>  @llvm.usub.sat.v2i64(<2 x i64>, <2 x i64>)
+declare <4 x i64>  @llvm.usub.sat.v4i64(<4 x i64>, <4 x i64>)
+declare <5 x i64>  @llvm.usub.sat.v5i64(<5 x i64>, <5 x i64>)
+declare <8 x i64>  @llvm.usub.sat.v8i64(<8 x i64>, <8 x i64>)
+
+declare i32        @llvm.usub.sat.i32(i32, i32)
+declare <2 x i32>  @llvm.usub.sat.v2i32(<2 x i32>, <2 x i32>)
+declare <4 x i32>  @llvm.usub.sat.v4i32(<4 x i32>, <4 x i32>)
+declare <8 x i32>  @llvm.usub.sat.v8i32(<8 x i32>, <8 x i32>)
+declare <9 x i32>  @llvm.usub.sat.v9i32(<9 x i32>, <9 x i32>)
+declare <16 x i32> @llvm.usub.sat.v16i32(<16 x i32>, <16 x i32>)
+
+declare i16        @llvm.usub.sat.i16(i16, i16)
+declare <2 x i16>  @llvm.usub.sat.v2i16(<2 x i16>, <2 x i16>)
+declare <4 x i16>  @llvm.usub.sat.v4i16(<4 x i16>, <4 x i16>)
+declare <8 x i16>  @llvm.usub.sat.v8i16(<8 x i16>, <8 x i16>)
+declare <16 x i16> @llvm.usub.sat.v16i16(<16 x i16>, <16 x i16>)
+declare <17 x i16> @llvm.usub.sat.v17i16(<17 x i16>, <17 x i16>)
+declare <32 x i16> @llvm.usub.sat.v32i16(<32 x i16>, <32 x i16>)
+
+declare i8         @llvm.usub.sat.i8(i8,  i8)
+declare <2 x i8>   @llvm.usub.sat.v2i8(<2 x i8>, <2 x i8>)
+declare <4 x i8>   @llvm.usub.sat.v4i8(<4 x i8>, <4 x i8>)
+declare <8 x i8>   @llvm.usub.sat.v8i8(<8 x i8>, <8 x i8>)
+declare <16 x i8>  @llvm.usub.sat.v16i8(<16 x i8>, <16 x i8>)
+declare <32 x i8>  @llvm.usub.sat.v32i8(<32 x i8>, <32 x i8>)
+declare <33 x i8>  @llvm.usub.sat.v33i8(<33 x i8>, <33 x i8>)
+declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
+
+define i32 @sub(i32 %arg) {
+; FAST-LABEL: 'sub'
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.usub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.usub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V17I16 = call <17 x i16> @llvm.usub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.usub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; SLOW-LABEL: 'sub'
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V5I64 = call <5 x i64> @llvm.usub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-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)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V9I32 = call <9 x i32> @llvm.usub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V17I16 = call <17 x i16> @llvm.usub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V33I8 = call <33 x i8> @llvm.usub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 264 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
+;
+; FAST-SIZE-LABEL: 'sub'
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.usub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.usub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; FAST-SIZE-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V17I16 = call <17 x i16> @llvm.usub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-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)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.usub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; SLOW-SIZE-LABEL: 'sub'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V5I64 = call <5 x i64> @llvm.usub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+; SLOW-SIZE-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V9I32 = call <9 x i32> @llvm.usub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V17I16 = call <17 x i16> @llvm.usub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-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)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V33I8 = call <33 x i8> @llvm.usub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLOW-SIZE-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)
+  %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+  %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+  %V5I64 = call <5 x i64> @llvm.usub.sat.v5i64(<5 x i64> undef, <5 x i64> undef)
+  %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+
+  %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+  %V2I32  = call <2 x i32>  @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
+  %V4I32  = call <4 x i32>  @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+  %V8I32  = call <8 x i32>  @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+  %V9I32  = call <9 x i32>  @llvm.usub.sat.v9i32(<9 x i32> undef, <9 x i32> undef)
+  %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+
+  %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+  %V2I16  = call <2 x i16>  @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
+  %V4I16  = call <4 x i16>  @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+  %V8I16  = call <8 x i16>  @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+  %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+  %V17I16 = call <17 x i16> @llvm.usub.sat.v17i16(<17 x i16> undef, <17 x i16> undef)
+  %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+
+  %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+  %V2I8  = call <2 x i8>  @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
+  %V4I8  = call <4 x i8>  @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+  %V8I8  = call <8 x i8>  @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+  %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+  %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+  %V33I8 = call <33 x i8> @llvm.usub.sat.v33i8(<33 x i8> undef, <33 x i8> undef)
+  %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+
+  ret i32 undef
+}

diff  --git a/llvm/test/Analysis/CostModel/AMDGPU/fma.ll b/llvm/test/Analysis/CostModel/AMDGPU/fma.ll
index 55ea14537c9cc..cec33c25c6157 100644
--- a/llvm/test/Analysis/CostModel/AMDGPU/fma.ll
+++ b/llvm/test/Analysis/CostModel/AMDGPU/fma.ll
@@ -1,28 +1,34 @@
 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
-; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=GFX9,GFX90A %s
-; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900  -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=GFX9,GFX900 %s
-; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mattr=-half-rate-64-ops < %s | FileCheck -check-prefixes=SLOW %s
-; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=GFX9-SIZE,GFX90A-SIZE %s
-; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=GFX9-SIZE,SIZE %s
-; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mattr=-half-rate-64-ops < %s | FileCheck -check-prefixes=SIZE,SLOW-SIZE %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST,SLOWF64,NOPACKEDF32 %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST,FASTF64,PACKEDF32 %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST,SLOWF64,NOPACKEDF32 %s
+; RUN: opt -cost-model -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx1010 -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST-SIZE,SLOWF64-SIZE,NOPACKEDF32-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx90a -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST-SIZE,FASTF64-SIZE,PACKEDF32-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 -mattr=+half-rate-64-ops < %s | FileCheck -check-prefixes=FAST-SIZE,SLOWF64-SIZE,NOPACKEDF32-SIZE %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa < %s | FileCheck -check-prefixes=SLOW-SIZE %s
 ; END.
 
 define amdgpu_kernel void @fma_f32() #0 {
-; GFX90A-LABEL: 'fma_f32'
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
-;
-; GFX900-LABEL: 'fma_f32'
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
+; NOPACKEDF32-LABEL: 'fma_f32'
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
+; NOPACKEDF32-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
+;
+; PACKEDF32-LABEL: 'fma_f32'
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
+; PACKEDF32-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
 ; SLOW-LABEL: 'fma_f32'
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
@@ -30,72 +36,98 @@ define amdgpu_kernel void @fma_f32() #0 {
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
-; GFX90A-SIZE-LABEL: 'fma_f32'
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
-;
-; SIZE-LABEL: 'fma_f32'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+; NOPACKEDF32-SIZE-LABEL: 'fma_f32'
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
+; NOPACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
+; PACKEDF32-SIZE-LABEL: 'fma_f32'
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
+; PACKEDF32-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
+; SLOW-SIZE-LABEL: 'fma_f32'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
   %f32 = call float @llvm.fma.f32(float undef, float undef, float undef) #1
   %v2f32 = call <2 x float> @llvm.fma.v2f32(<2 x float> undef, <2 x float> undef, <2 x float> undef) #1
   %v3f32 = call <3 x float> @llvm.fma.v3f32(<3 x float> undef, <3 x float> undef, <3 x float> undef) #1
   %v4f32 = call <4 x float> @llvm.fma.v4f32(<4 x float> undef, <4 x float> undef, <4 x float> undef) #1
   %v5f32 = call <5 x float> @llvm.fma.v5f32(<5 x float> undef, <5 x float> undef, <5 x float> undef) #1
+  %v8f32 = call <8 x float> @llvm.fma.v8f32(<8 x float> undef, <8 x float> undef, <8 x float> undef) #1
+  %v9f32 = call <9 x float> @llvm.fma.v9f32(<9 x float> undef, <9 x float> undef, <9 x float> undef) #1
   ret void
 }
 
 define amdgpu_kernel void @fma_f64() #0 {
-; GFX90A-LABEL: 'fma_f64'
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
-; GFX90A-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
-;
-; GFX900-LABEL: 'fma_f64'
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
-; GFX900-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
+; SLOWF64-LABEL: 'fma_f64'
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; SLOWF64-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
+;
+; FASTF64-LABEL: 'fma_f64'
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; FASTF64-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
 ; SLOW-LABEL: 'fma_f64'
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
-; SLOW-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
-; GFX90A-SIZE-LABEL: 'fma_f64'
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
-; GFX90A-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
-;
-; SIZE-LABEL: 'fma_f64'
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
-; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+; SLOWF64-SIZE-LABEL: 'fma_f64'
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; SLOWF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
+; FASTF64-SIZE-LABEL: 'fma_f64'
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; FASTF64-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
+; SLOW-SIZE-LABEL: 'fma_f64'
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %v3f64 = call <3 x double> @llvm.fma.v3f64(<3 x double> undef, <3 x double> undef, <3 x double> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f64 = call <4 x double> @llvm.fma.v4f64(<4 x double> undef, <4 x double> undef, <4 x double> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %v5f64 = call <5 x double> @llvm.fma.v5f64(<5 x double> undef, <5 x double> undef, <5 x double> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
   %f64 = call double @llvm.fma.f64(double undef, double undef, double undef) #1
   %v2f64 = call <2 x double> @llvm.fma.v2f64(<2 x double> undef, <2 x double> undef, <2 x double> undef) #1
@@ -106,29 +138,35 @@ define amdgpu_kernel void @fma_f64() #0 {
 }
 
 define amdgpu_kernel void @fma_f16() #0 {
-; GFX9-LABEL: 'fma_f16'
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f16 = call <2 x half> @llvm.fma.v2f16(<2 x half> undef, <2 x half> undef, <2 x half> undef) #2
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
-; GFX9-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
+; FAST-LABEL: 'fma_f16'
+; FAST-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f16 = call <2 x half> @llvm.fma.v2f16(<2 x half> undef, <2 x half> undef, <2 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v16f16 = call <16 x half> @llvm.fma.v16f16(<16 x half> undef, <16 x half> undef, <16 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %v17f16 = call <17 x half> @llvm.fma.v17f16(<17 x half> undef, <17 x half> undef, <17 x half> undef) #2
+; FAST-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
 ; SLOW-LABEL: 'fma_f16'
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v2f16 = call <2 x half> @llvm.fma.v2f16(<2 x half> undef, <2 x half> undef, <2 x half> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
-; SLOW-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %v16f16 = call <16 x half> @llvm.fma.v16f16(<16 x half> undef, <16 x half> undef, <16 x half> undef) #2
+; SLOW-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %v17f16 = call <17 x half> @llvm.fma.v17f16(<17 x half> undef, <17 x half> undef, <17 x half> undef) #2
 ; SLOW-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret void
 ;
-; GFX9-SIZE-LABEL: 'fma_f16'
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f16 = call <2 x half> @llvm.fma.v2f16(<2 x half> undef, <2 x half> undef, <2 x half> undef) #2
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
-; GFX9-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+; FAST-SIZE-LABEL: 'fma_f16'
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v2f16 = call <2 x half> @llvm.fma.v2f16(<2 x half> undef, <2 x half> undef, <2 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v16f16 = call <16 x half> @llvm.fma.v16f16(<16 x half> undef, <16 x half> undef, <16 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %v17f16 = call <17 x half> @llvm.fma.v17f16(<17 x half> undef, <17 x half> undef, <17 x half> undef) #2
+; FAST-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
 ; SLOW-SIZE-LABEL: 'fma_f16'
 ; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #2
@@ -136,6 +174,8 @@ define amdgpu_kernel void @fma_f16() #0 {
 ; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #2
 ; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #2
 ; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %v16f16 = call <16 x half> @llvm.fma.v16f16(<16 x half> undef, <16 x half> undef, <16 x half> undef) #2
+; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %v17f16 = call <17 x half> @llvm.fma.v17f16(<17 x half> undef, <17 x half> undef, <17 x half> undef) #2
 ; SLOW-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
   %f16 = call half @llvm.fma.f16(half undef, half undef, half undef) #1
@@ -143,6 +183,8 @@ define amdgpu_kernel void @fma_f16() #0 {
   %v3f16 = call <3 x half> @llvm.fma.v3f16(<3 x half> undef, <3 x half> undef, <3 x half> undef) #1
   %v4f16 = call <4 x half> @llvm.fma.v4f16(<4 x half> undef, <4 x half> undef, <4 x half> undef) #1
   %v5f16 = call <5 x half> @llvm.fma.v5f16(<5 x half> undef, <5 x half> undef, <5 x half> undef) #1
+  %v16f16 = call <16 x half> @llvm.fma.v16f16(<16 x half> undef, <16 x half> undef, <16 x half> undef) #1
+  %v17f16 = call <17 x half> @llvm.fma.v17f16(<17 x half> undef, <17 x half> undef, <17 x half> undef) #1
   ret void
 }
 
@@ -151,6 +193,8 @@ declare <2 x float> @llvm.fma.v2f32(<2 x float>, <2 x float>, <2 x float>) #1
 declare <3 x float> @llvm.fma.v3f32(<3 x float>, <3 x float>, <3 x float>) #1
 declare <4 x float> @llvm.fma.v4f32(<4 x float>, <4 x float>, <4 x float>) #1
 declare <5 x float> @llvm.fma.v5f32(<5 x float>, <5 x float>, <5 x float>) #1
+declare <8 x float> @llvm.fma.v8f32(<8 x float>, <8 x float>, <8 x float>) #1
+declare <9 x float> @llvm.fma.v9f32(<9 x float>, <9 x float>, <9 x float>) #1
 
 declare double @llvm.fma.f64(double, double, double) #1
 declare <2 x double> @llvm.fma.v2f64(<2 x double>, <2 x double>, <2 x double>) #1
@@ -163,6 +207,8 @@ declare <2 x half> @llvm.fma.v2f16(<2 x half>, <2 x half>, <2 x half>) #1
 declare <3 x half> @llvm.fma.v3f16(<3 x half>, <3 x half>, <3 x half>) #1
 declare <4 x half> @llvm.fma.v4f16(<4 x half>, <4 x half>, <4 x half>) #1
 declare <5 x half> @llvm.fma.v5f16(<5 x half>, <5 x half>, <5 x half>) #1
+declare <16 x half> @llvm.fma.v16f16(<16 x half>, <16 x half>, <16 x half>) #1
+declare <17 x half> @llvm.fma.v17f16(<17 x half>, <17 x half>, <17 x half>) #1
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readnone }

diff  --git a/llvm/test/Analysis/CostModel/AMDGPU/reduce-and.ll b/llvm/test/Analysis/CostModel/AMDGPU/reduce-and.ll
index 8c71d508c0608..1b9b9ffd73fe6 100644
--- a/llvm/test/Analysis/CostModel/AMDGPU/reduce-and.ll
+++ b/llvm/test/Analysis/CostModel/AMDGPU/reduce-and.ll
@@ -13,6 +13,7 @@ define i32 @reduce_i1(i32 %arg) {
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32 = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 65 for instruction: %V64 = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 130 for instruction: %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)
+; ALL-NEXT:  Cost Model: Found an estimated cost of 260 for instruction: %V256 = call i1 @llvm.vector.reduce.and.v256i1(<256 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
 ;
 ; ALL-SIZE-LABEL: 'reduce_i1'
@@ -24,6 +25,7 @@ define i32 @reduce_i1(i32 %arg) {
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32 = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 65 for instruction: %V64 = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 129 for instruction: %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)
+; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 257 for instruction: %V256 = call i1 @llvm.vector.reduce.and.v256i1(<256 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %V1   = call i1 @llvm.vector.reduce.and.v1i1(<1 x i1> undef)
@@ -34,6 +36,7 @@ define i32 @reduce_i1(i32 %arg) {
   %V32  = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)
+  %V256 = call i1 @llvm.vector.reduce.and.v256i1(<256 x i1> undef)
   ret i32 undef
 }
 
@@ -45,3 +48,4 @@ declare i1 @llvm.vector.reduce.and.v16i1(<16 x i1>)
 declare i1 @llvm.vector.reduce.and.v32i1(<32 x i1>)
 declare i1 @llvm.vector.reduce.and.v64i1(<64 x i1>)
 declare i1 @llvm.vector.reduce.and.v128i1(<128 x i1>)
+declare i1 @llvm.vector.reduce.and.v256i1(<256 x i1>)

diff  --git a/llvm/test/Analysis/CostModel/AMDGPU/reduce-or.ll b/llvm/test/Analysis/CostModel/AMDGPU/reduce-or.ll
index e530d0481e32b..1868e7eb7e803 100644
--- a/llvm/test/Analysis/CostModel/AMDGPU/reduce-or.ll
+++ b/llvm/test/Analysis/CostModel/AMDGPU/reduce-or.ll
@@ -13,6 +13,7 @@ define i32 @reduce_i1(i32 %arg) {
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32 = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 65 for instruction: %V64 = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 130 for instruction: %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)
+; ALL-NEXT:  Cost Model: Found an estimated cost of 260 for instruction: %V256 = call i1 @llvm.vector.reduce.or.v256i1(<256 x i1> undef)
 ; ALL-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: ret i32 undef
 ;
 ; ALL-SIZE-LABEL: 'reduce_i1'
@@ -24,6 +25,7 @@ define i32 @reduce_i1(i32 %arg) {
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32 = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 65 for instruction: %V64 = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 129 for instruction: %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)
+; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 257 for instruction: %V256 = call i1 @llvm.vector.reduce.or.v256i1(<256 x i1> undef)
 ; ALL-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %V1   = call i1 @llvm.vector.reduce.or.v1i1(<1 x i1> undef)
@@ -34,6 +36,7 @@ define i32 @reduce_i1(i32 %arg) {
   %V32  = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)
+  %V256 = call i1 @llvm.vector.reduce.or.v256i1(<256 x i1> undef)
   ret i32 undef
 }
 
@@ -45,3 +48,4 @@ declare i1 @llvm.vector.reduce.or.v16i1(<16 x i1>)
 declare i1 @llvm.vector.reduce.or.v32i1(<32 x i1>)
 declare i1 @llvm.vector.reduce.or.v64i1(<64 x i1>)
 declare i1 @llvm.vector.reduce.or.v128i1(<128 x i1>)
+declare i1 @llvm.vector.reduce.or.v256i1(<256 x i1>)


        


More information about the llvm-commits mailing list