[llvm] 64cfb8a - [NFC][ARM] Add intrinsic code size runs

Sam Parker via llvm-commits llvm-commits at lists.llvm.org
Tue May 26 01:42:25 PDT 2020


Author: Sam Parker
Date: 2020-05-26T09:41:54+01:00
New Revision: 64cfb8a864cf98dcd762a26d03cba95145b9aa41

URL: https://github.com/llvm/llvm-project/commit/64cfb8a864cf98dcd762a26d03cba95145b9aa41
DIFF: https://github.com/llvm/llvm-project/commit/64cfb8a864cf98dcd762a26d03cba95145b9aa41.diff

LOG: [NFC][ARM] Add intrinsic code size runs

Add code size analysis of arithmetic intrinsics.

Added: 
    

Modified: 
    llvm/test/Analysis/CostModel/ARM/arith-overflow.ll
    llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
    llvm/test/Analysis/CostModel/ARM/arith-usat.ll
    llvm/test/Analysis/CostModel/ARM/reduce-add.ll

Removed: 
    


################################################################################
diff  --git a/llvm/test/Analysis/CostModel/ARM/arith-overflow.ll b/llvm/test/Analysis/CostModel/ARM/arith-overflow.ll
index 66a03888c0d4..b50aa97643e2 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-overflow.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-overflow.ll
@@ -1,7 +1,10 @@
 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
-; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M
-; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON
-; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE
+; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-RECIP
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-SIZE
 
 declare {i64, i1}              @llvm.sadd.with.overflow.i64(i64, i64)
 declare {<2 x i64>, <2 x i1>}  @llvm.sadd.with.overflow.v2i64(<2 x i64>, <2 x i64>)
@@ -24,62 +27,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.sadd.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.sadd.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sadd(i32 %arg) {
-; V8M-LABEL: 'sadd'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'sadd'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'sadd'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 582 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'sadd'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'sadd'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'sadd'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 582 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'sadd'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'sadd'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'sadd'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.sadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.sadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.sadd.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.sadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.sadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.sadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.sadd.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.sadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.sadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.sadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.sadd.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.sadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.sadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.sadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.sadd.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.sadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
@@ -125,62 +185,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.uadd.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.uadd.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @uadd(i32 %arg) {
-; V8M-LABEL: 'uadd'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'uadd'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'uadd'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'uadd'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'uadd'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'uadd'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'uadd'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'uadd'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'uadd'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.uadd.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.uadd.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.uadd.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.uadd.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.uadd.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.uadd.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.uadd.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.uadd.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.uadd.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.uadd.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.uadd.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.uadd.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.uadd.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.uadd.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.uadd.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
@@ -226,62 +343,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.ssub.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.ssub.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @ssub(i32 %arg) {
-; V8M-LABEL: 'ssub'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'ssub'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'ssub'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 582 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'ssub'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'ssub'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'ssub'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 166 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 582 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'ssub'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'ssub'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'ssub'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.ssub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.ssub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.ssub.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.ssub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.ssub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.ssub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.ssub.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.ssub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.ssub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.ssub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.ssub.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.ssub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.ssub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.ssub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.ssub.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.ssub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
@@ -327,62 +501,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.usub.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.usub.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @usub(i32 %arg) {
-; V8M-LABEL: 'usub'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'usub'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'usub'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'usub'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'usub'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'usub'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'usub'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'usub'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'usub'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.usub.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.usub.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.usub.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.usub.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.usub.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.usub.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.usub.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.usub.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.usub.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.usub.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.usub.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.usub.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.usub.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.usub.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.usub.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
@@ -428,62 +659,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.smul.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.smul.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @smul(i32 %arg) {
-; V8M-LABEL: 'smul'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 156 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 312 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 224 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 224 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'smul'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 154 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 300 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'smul'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 86 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 316 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1208 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 424 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 424 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'smul'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 156 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 312 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 224 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 224 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'smul'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 81 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 154 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 300 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'smul'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 768 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 86 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 316 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1208 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 424 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 424 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'smul'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'smul'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'smul'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.smul.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.smul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
@@ -529,62 +817,119 @@ declare {<32 x i8>, <32 x i1>}  @llvm.umul.with.overflow.v32i8(<32 x i8>, <32 x
 declare {<64 x i8>, <64 x i1>}  @llvm.umul.with.overflow.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @umul(i32 %arg) {
-; V8M-LABEL: 'umul'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 296 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 152 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 304 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'umul'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 75 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 146 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'umul'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 248 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 624 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 312 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1200 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 416 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 416 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'umul'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 74 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 296 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 152 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 304 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 192 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 384 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'umul'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 75 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 146 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'umul'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 248 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 624 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 312 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1200 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 416 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 416 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'umul'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'umul'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'umul'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call { <2 x i64>, <2 x i1> } @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %I64 = call {i64, i1} @llvm.umul.with.overflow.i64(i64 undef, i64 undef)
   %V2I64 = call {<2 x i64>, <2 x i1>} @llvm.umul.with.overflow.v2i64(<2 x i64> undef, <2 x i64> undef)

diff  --git a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
index 59687ce72249..bc2ac44215e0 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
@@ -1,7 +1,10 @@
 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
-; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M
-; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON
-; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE
+; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-RECIP
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-SIZE
 
 declare i64        @llvm.sadd.sat.i64(i64, i64)
 declare <2 x i64>  @llvm.sadd.sat.v2i64(<2 x i64>, <2 x i64>)
@@ -24,62 +27,119 @@ declare <32 x i8>  @llvm.sadd.sat.v32i8(<32 x i8>, <32 x i8>)
 declare <64 x i8>  @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
-; V8M-LABEL: 'add'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 216 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 1152 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'add'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 216 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1152 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'add'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 123 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-RECIP-LABEL: 'add'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 123 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'add'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 118 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 302 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1046 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-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)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-RECIP-LABEL: 'add'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 118 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 302 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1046 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'add'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'add'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'add'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-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)
@@ -125,62 +185,119 @@ declare <32 x i8>  @llvm.ssub.sat.v32i8(<32 x i8>, <32 x i8>)
 declare <64 x i8>  @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
-; V8M-LABEL: 'sub'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 216 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 1152 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'sub'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 216 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 1152 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'sub'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 123 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'sub'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 118 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 302 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1046 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'sub'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'sub'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 51 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 123 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-SIZE-LABEL: 'sub'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'sub'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 118 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 302 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1046 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-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)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-SIZE-LABEL: 'sub'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-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)

diff  --git a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
index 92c313f4ab1c..c4f09937b47e 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
@@ -1,7 +1,10 @@
 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
-; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M
-; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON
-; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE
+; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-RECIP
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-SIZE
 
 declare i64        @llvm.uadd.sat.i64(i64, i64)
 declare <2 x i64>  @llvm.uadd.sat.v2i64(<2 x i64>, <2 x i64>)
@@ -24,62 +27,119 @@ declare <32 x i8>  @llvm.uadd.sat.v32i8(<32 x i8>, <32 x i8>)
 declare <64 x i8>  @llvm.uadd.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @add(i32 %arg) {
-; V8M-LABEL: 'add'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'add'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'add'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-RECIP-LABEL: 'add'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'add'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-RECIP-LABEL: 'add'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'add'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'add'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'add'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.uadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.uadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-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)
@@ -125,62 +185,119 @@ declare <32 x i8>  @llvm.usub.sat.v32i8(<32 x i8>, <32 x i8>)
 declare <64 x i8>  @llvm.usub.sat.v64i8(<64 x i8>, <64 x i8>)
 
 define i32 @sub(i32 %arg) {
-; V8M-LABEL: 'sub'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'sub'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 160 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 320 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; NEON-RECIP-LABEL: 'sub'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; MVE-RECIP-LABEL: 'sub'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'sub'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'sub'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-SIZE-LABEL: 'sub'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'sub'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 448 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-SIZE-LABEL: 'sub'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2I64 = call <2 x i64> @llvm.usub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I64 = call <4 x i64> @llvm.usub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-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)

diff  --git a/llvm/test/Analysis/CostModel/ARM/reduce-add.ll b/llvm/test/Analysis/CostModel/ARM/reduce-add.ll
index 73614ab5ece8..089061253d82 100644
--- a/llvm/test/Analysis/CostModel/ARM/reduce-add.ll
+++ b/llvm/test/Analysis/CostModel/ARM/reduce-add.ll
@@ -1,32 +1,59 @@
 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
-; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M
-; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON
-; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE
+; RUN: opt -cost-model -analyze -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-RECIP
+; RUN: opt -cost-model -analyze -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-RECIP
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON-SIZE
+; RUN: opt -cost-model -analyze -cost-kind=code-size -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE-SIZE
 
 define i32 @reduce_i64(i32 %arg) {
-; V8M-LABEL: 'reduce_i64'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'reduce_i64'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'reduce_i64'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 55 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 107 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-RECIP-LABEL: 'reduce_i64'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 55 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 107 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'reduce_i64'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 202 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 730 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-RECIP-LABEL: 'reduce_i64'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 202 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 730 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
+; V8M-SIZE-LABEL: 'reduce_i64'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; NEON-SIZE-LABEL: 'reduce_i64'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
+; MVE-SIZE-LABEL: 'reduce_i64'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.add.v4i64(<4 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.add.v8i64(<8 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.add.v16i64(<16 x i64> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %V1  = call i64 @llvm.experimental.vector.reduce.add.v1i64(<1 x i64> undef)
   %V2  = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> undef)
@@ -37,105 +64,65 @@ define i32 @reduce_i64(i32 %arg) {
 }
 
 define i32 @reduce_i32(i32 %arg) {
-; V8M-LABEL: 'reduce_i32'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.add.v2i32(<2 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.add.v4i32(<4 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.add.v8i32(<8 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.add.v16i32(<16 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.add.v32i32(<32 x i32> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; NEON-LABEL: 'reduce_i32'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.add.v2i32(<2 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.add.v4i32(<4 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.add.v8i32(<8 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.add.v16i32(<16 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 228 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.add.v32i32(<32 x i32> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-; MVE-LABEL: 'reduce_i32'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.add.v2i32(<2 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.add.v4i32(<4 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 234 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.add.v8i32(<8 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 622 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.add.v16i32(<16 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2166 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.add.v32i32(<32 x i32> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
-;
-  %V2  = call i32 @llvm.experimental.vector.reduce.add.v2i32(<2 x i32> undef)
-  %V4  = call i32 @llvm.experimental.vector.reduce.add.v4i32(<4 x i32> undef)
-  %V8  = call i32 @llvm.experimental.vector.reduce.add.v8i32(<8 x i32> undef)
-  %V16 = call i32 @llvm.experimental.vector.reduce.add.v16i32(<16 x i32> undef)
-  %V32 = call i32 @llvm.experimental.vector.reduce.add.v32i32(<32 x i32> undef)
-  ret i32 undef
-}
-
-define i32 @reduce_i16(i32 %arg) {
-; V8M-LABEL: 'reduce_i16'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.add.v2i16(<2 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.add.v4i16(<4 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.add.v8i16(<8 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.add.v16i16(<16 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.add.v32i16(<32 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 190 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.add.v64i16(<64 x i16> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-RECIP-LABEL: 'reduce_i32'
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 190 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 382 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; V8M-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'reduce_i16'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.add.v2i16(<2 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.add.v4i16(<4 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 150 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.add.v8i16(<8 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 199 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.add.v16i16(<16 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 297 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.add.v32i16(<32 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 493 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.add.v64i16(<64 x i16> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-RECIP-LABEL: 'reduce_i32'
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 150 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 391 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 488 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 682 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 1070 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; NEON-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'reduce_i16'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.add.v2i16(<2 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.add.v4i16(<4 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 782 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.add.v8i16(<8 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 1168 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.add.v16i16(<16 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 2708 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.add.v32i16(<32 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 8860 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.add.v64i16(<64 x i16> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-RECIP-LABEL: 'reduce_i32'
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 782 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 4120 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 5658 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 11806 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 36390 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; MVE-RECIP-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
-  %V2  = call i16 @llvm.experimental.vector.reduce.add.v2i16(<2 x i16> undef)
-  %V4  = call i16 @llvm.experimental.vector.reduce.add.v4i16(<4 x i16> undef)
-  %V8  = call i16 @llvm.experimental.vector.reduce.add.v8i16(<8 x i16> undef)
-  %V16 = call i16 @llvm.experimental.vector.reduce.add.v16i16(<16 x i16> undef)
-  %V32 = call i16 @llvm.experimental.vector.reduce.add.v32i16(<32 x i16> undef)
-  %V64 = call i16 @llvm.experimental.vector.reduce.add.v64i16(<64 x i16> undef)
-  ret i32 undef
-}
-
-define i32 @reduce_i8(i32 %arg) {
-; V8M-LABEL: 'reduce_i8'
-; V8M-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 190 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 382 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
-; V8M-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; V8M-SIZE-LABEL: 'reduce_i32'
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; V8M-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; NEON-LABEL: 'reduce_i8'
-; NEON-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 150 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 391 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 488 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 682 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 1070 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
-; NEON-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; NEON-SIZE-LABEL: 'reduce_i32'
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; NEON-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
-; MVE-LABEL: 'reduce_i8'
-; MVE-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 782 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 4120 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 5658 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 11806 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 36390 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
-; MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; MVE-SIZE-LABEL: 'reduce_i32'
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.add.v8i8(<8 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.add.v16i8(<16 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.add.v32i8(<32 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.add.v64i8(<64 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.add.v128i8(<128 x i8> undef)
+; MVE-SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
   %V2   = call i8 @llvm.experimental.vector.reduce.add.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.experimental.vector.reduce.add.v4i8(<4 x i8> undef)


        


More information about the llvm-commits mailing list