[llvm] d74533a - [X86] Add sse4.1 RUNs lines to the min/max reduction cost model tests.
Craig Topper via llvm-commits
llvm-commits at lists.llvm.org
Sun Mar 29 16:11:33 PDT 2020
Author: Craig Topper
Date: 2020-03-29T16:05:35-07:00
New Revision: d74533a18b8c442f2e17a4c326393ecc36931244
URL: https://github.com/llvm/llvm-project/commit/d74533a18b8c442f2e17a4c326393ecc36931244
DIFF: https://github.com/llvm/llvm-project/commit/d74533a18b8c442f2e17a4c326393ecc36931244.diff
LOG: [X86] Add sse4.1 RUNs lines to the min/max reduction cost model tests.
Mostly this matches the sse4.2 we already had command lines for.
Except in the i64 case since sse4.1 doesn't have pcmpgtq.
Added:
Modified:
llvm/test/Analysis/CostModel/X86/reduce-fmax.ll
llvm/test/Analysis/CostModel/X86/reduce-fmin.ll
llvm/test/Analysis/CostModel/X86/reduce-smax.ll
llvm/test/Analysis/CostModel/X86/reduce-smin.ll
llvm/test/Analysis/CostModel/X86/reduce-umax.ll
llvm/test/Analysis/CostModel/X86/reduce-umin.ll
Removed:
################################################################################
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-fmax.ll b/llvm/test/Analysis/CostModel/X86/reduce-fmax.ll
index c189adad8000..2f6ae6e60b2b 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-fmax.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-fmax.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -60,14 +61,14 @@ define i32 @reduce_f32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmax.v32f32(<32 x float> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_f32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmax.v1f32(<1 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call float @llvm.experimental.vector.reduce.fmax.v2f32(<2 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call float @llvm.experimental.vector.reduce.fmax.v4f32(<4 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call float @llvm.experimental.vector.reduce.fmax.v8f32(<8 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16 = call float @llvm.experimental.vector.reduce.fmax.v16f32(<16 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmax.v32f32(<32 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_f32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmax.v1f32(<1 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call float @llvm.experimental.vector.reduce.fmax.v2f32(<2 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call float @llvm.experimental.vector.reduce.fmax.v4f32(<4 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call float @llvm.experimental.vector.reduce.fmax.v8f32(<8 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16 = call float @llvm.experimental.vector.reduce.fmax.v16f32(<16 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmax.v32f32(<32 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX-LABEL: 'reduce_f32'
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmax.v1f32(<1 x float> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-fmin.ll b/llvm/test/Analysis/CostModel/X86/reduce-fmin.ll
index 6ef95c416efb..917da71af9d9 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-fmin.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-fmin.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -60,14 +61,14 @@ define i32 @reduce_f32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmin.v32f32(<32 x float> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_f32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmin.v1f32(<1 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call float @llvm.experimental.vector.reduce.fmin.v2f32(<2 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call float @llvm.experimental.vector.reduce.fmin.v4f32(<4 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call float @llvm.experimental.vector.reduce.fmin.v8f32(<8 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16 = call float @llvm.experimental.vector.reduce.fmin.v16f32(<16 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmin.v32f32(<32 x float> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_f32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmin.v1f32(<1 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call float @llvm.experimental.vector.reduce.fmin.v2f32(<2 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call float @llvm.experimental.vector.reduce.fmin.v4f32(<4 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call float @llvm.experimental.vector.reduce.fmin.v8f32(<8 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16 = call float @llvm.experimental.vector.reduce.fmin.v16f32(<16 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V32 = call float @llvm.experimental.vector.reduce.fmin.v32f32(<32 x float> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX-LABEL: 'reduce_f32'
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.experimental.vector.reduce.fmin.v1f32(<1 x float> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-smax.ll b/llvm/test/Analysis/CostModel/X86/reduce-smax.ll
index a578ad67f723..cf2ac7ee1e64 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-smax.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-smax.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -25,6 +26,14 @@ define i32 @reduce_i64(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.smax.v16i64(<16 x i64> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
+; SSE41-LABEL: 'reduce_i64'
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.smax.v1i64(<1 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.smax.v2i64(<2 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.smax.v4i64(<4 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.smax.v8i64(<8 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.smax.v16i64(<16 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
; SSE42-LABEL: 'reduce_i64'
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.smax.v1i64(<1 x i64> undef)
; SSE42-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.smax.v2i64(<2 x i64> undef)
@@ -82,13 +91,13 @@ define i32 @reduce_i32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smax.v32i32(<32 x i32> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smax.v2i32(<2 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.smax.v4i32(<4 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.smax.v8i32(<8 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.smax.v16i32(<16 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smax.v32i32(<32 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smax.v2i32(<2 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.smax.v4i32(<4 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.smax.v8i32(<8 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.smax.v16i32(<16 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smax.v32i32(<32 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i32'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smax.v2i32(<2 x i32> undef)
@@ -141,14 +150,14 @@ define i32 @reduce_i16(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smax.v64i16(<64 x i16> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i16'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smax.v2i16(<2 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.smax.v4i16(<4 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.smax.v8i16(<8 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.smax.v16i16(<16 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.smax.v32i16(<32 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smax.v64i16(<64 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i16'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smax.v2i16(<2 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.smax.v4i16(<4 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.smax.v8i16(<8 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.smax.v16i16(<16 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.smax.v32i16(<32 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smax.v64i16(<64 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i16'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smax.v2i16(<2 x i16> undef)
@@ -225,15 +234,15 @@ define i32 @reduce_i8(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smax.v128i8(<128 x i8> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i8'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smax.v2i8(<2 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.smax.v4i8(<4 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.smax.v8i8(<8 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.smax.v16i8(<16 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.smax.v32i8(<32 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.smax.v64i8(<64 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smax.v128i8(<128 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i8'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smax.v2i8(<2 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.smax.v4i8(<4 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.smax.v8i8(<8 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.smax.v16i8(<16 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.smax.v32i8(<32 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.smax.v64i8(<64 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smax.v128i8(<128 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i8'
; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smax.v2i8(<2 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-smin.ll b/llvm/test/Analysis/CostModel/X86/reduce-smin.ll
index 01263daea9be..e91517708819 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-smin.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-smin.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -25,6 +26,14 @@ define i32 @reduce_i64(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.smin.v16i64(<16 x i64> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
+; SSE41-LABEL: 'reduce_i64'
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.smin.v1i64(<1 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.smin.v2i64(<2 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.smin.v4i64(<4 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.smin.v8i64(<8 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.smin.v16i64(<16 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
; SSE42-LABEL: 'reduce_i64'
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.smin.v1i64(<1 x i64> undef)
; SSE42-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.smin.v2i64(<2 x i64> undef)
@@ -82,13 +91,13 @@ define i32 @reduce_i32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smin.v32i32(<32 x i32> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smin.v2i32(<2 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.smin.v4i32(<4 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.smin.v8i32(<8 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.smin.v16i32(<16 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smin.v32i32(<32 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smin.v2i32(<2 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.smin.v4i32(<4 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.smin.v8i32(<8 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.smin.v16i32(<16 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.smin.v32i32(<32 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i32'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.smin.v2i32(<2 x i32> undef)
@@ -141,14 +150,14 @@ define i32 @reduce_i16(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smin.v64i16(<64 x i16> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i16'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smin.v2i16(<2 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.smin.v4i16(<4 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.smin.v8i16(<8 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.smin.v16i16(<16 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.smin.v32i16(<32 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smin.v64i16(<64 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i16'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smin.v2i16(<2 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.smin.v4i16(<4 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.smin.v8i16(<8 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.smin.v16i16(<16 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.smin.v32i16(<32 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.smin.v64i16(<64 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i16'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.smin.v2i16(<2 x i16> undef)
@@ -225,15 +234,15 @@ define i32 @reduce_i8(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smin.v128i8(<128 x i8> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i8'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smin.v2i8(<2 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.smin.v4i8(<4 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.smin.v8i8(<8 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.smin.v16i8(<16 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.smin.v32i8(<32 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.smin.v64i8(<64 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smin.v128i8(<128 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i8'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smin.v2i8(<2 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.smin.v4i8(<4 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.smin.v8i8(<8 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.smin.v16i8(<16 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.smin.v32i8(<32 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.smin.v64i8(<64 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.smin.v128i8(<128 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i8'
; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.smin.v2i8(<2 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-umax.ll b/llvm/test/Analysis/CostModel/X86/reduce-umax.ll
index f77915e1b356..da43b61c522c 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-umax.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-umax.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -25,6 +26,14 @@ define i32 @reduce_i64(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.umax.v16i64(<16 x i64> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
+; SSE41-LABEL: 'reduce_i64'
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.umax.v1i64(<1 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.umax.v2i64(<2 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 18 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.umax.v4i64(<4 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 36 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.umax.v8i64(<8 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 72 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.umax.v16i64(<16 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
; SSE42-LABEL: 'reduce_i64'
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.umax.v1i64(<1 x i64> undef)
; SSE42-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.umax.v2i64(<2 x i64> undef)
@@ -82,13 +91,13 @@ define i32 @reduce_i32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umax.v32i32(<32 x i32> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umax.v2i32(<2 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.umax.v4i32(<4 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.umax.v8i32(<8 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.umax.v16i32(<16 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umax.v32i32(<32 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umax.v2i32(<2 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.umax.v4i32(<4 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.umax.v8i32(<8 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.umax.v16i32(<16 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umax.v32i32(<32 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i32'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umax.v2i32(<2 x i32> undef)
@@ -141,14 +150,14 @@ define i32 @reduce_i16(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umax.v64i16(<64 x i16> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i16'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umax.v2i16(<2 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.umax.v4i16(<4 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.umax.v8i16(<8 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.umax.v16i16(<16 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.umax.v32i16(<32 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umax.v64i16(<64 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i16'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umax.v2i16(<2 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.umax.v4i16(<4 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.umax.v8i16(<8 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.umax.v16i16(<16 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.umax.v32i16(<32 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umax.v64i16(<64 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i16'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umax.v2i16(<2 x i16> undef)
@@ -225,15 +234,15 @@ define i32 @reduce_i8(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umax.v128i8(<128 x i8> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i8'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umax.v2i8(<2 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.umax.v4i8(<4 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.umax.v8i8(<8 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.umax.v16i8(<16 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.umax.v32i8(<32 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.umax.v64i8(<64 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umax.v128i8(<128 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i8'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umax.v2i8(<2 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.umax.v4i8(<4 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.umax.v8i8(<8 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.umax.v16i8(<16 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.umax.v32i8(<32 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.umax.v64i8(<64 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umax.v128i8(<128 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i8'
; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umax.v2i8(<2 x i8> undef)
diff --git a/llvm/test/Analysis/CostModel/X86/reduce-umin.ll b/llvm/test/Analysis/CostModel/X86/reduce-umin.ll
index ce35e8b24848..2406a3a4c648 100644
--- a/llvm/test/Analysis/CostModel/X86/reduce-umin.ll
+++ b/llvm/test/Analysis/CostModel/X86/reduce-umin.ll
@@ -1,7 +1,8 @@
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+ssse3 | FileCheck %s --check-prefixes=CHECK,SSE,SSSE3
-; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE42
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.1 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE41
+; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+sse4.2 | FileCheck %s --check-prefixes=CHECK,SSE,SSE4,SSE42
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx2 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -cost-model -mtriple=x86_64-apple-darwin -analyze -mattr=+avx512f | FileCheck %s --check-prefixes=CHECK,AVX512,AVX512F
@@ -25,6 +26,14 @@ define i32 @reduce_i64(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.umin.v16i64(<16 x i64> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
+; SSE41-LABEL: 'reduce_i64'
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.umin.v1i64(<1 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.umin.v2i64(<2 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 18 for instruction: %V4 = call i64 @llvm.experimental.vector.reduce.umin.v4i64(<4 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 36 for instruction: %V8 = call i64 @llvm.experimental.vector.reduce.umin.v8i64(<8 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 72 for instruction: %V16 = call i64 @llvm.experimental.vector.reduce.umin.v16i64(<16 x i64> undef)
+; SSE41-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+;
; SSE42-LABEL: 'reduce_i64'
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call i64 @llvm.experimental.vector.reduce.umin.v1i64(<1 x i64> undef)
; SSE42-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V2 = call i64 @llvm.experimental.vector.reduce.umin.v2i64(<2 x i64> undef)
@@ -82,13 +91,13 @@ define i32 @reduce_i32(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 48 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umin.v32i32(<32 x i32> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i32'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umin.v2i32(<2 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.umin.v4i32(<4 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.umin.v8i32(<8 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.umin.v16i32(<16 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umin.v32i32(<32 x i32> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i32'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umin.v2i32(<2 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i32 @llvm.experimental.vector.reduce.umin.v4i32(<4 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.experimental.vector.reduce.umin.v8i32(<8 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16 = call i32 @llvm.experimental.vector.reduce.umin.v16i32(<16 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32 = call i32 @llvm.experimental.vector.reduce.umin.v32i32(<32 x i32> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i32'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i32 @llvm.experimental.vector.reduce.umin.v2i32(<2 x i32> undef)
@@ -141,14 +150,14 @@ define i32 @reduce_i16(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umin.v64i16(<64 x i16> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i16'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umin.v2i16(<2 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.umin.v4i16(<4 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.umin.v8i16(<8 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.umin.v16i16(<16 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.umin.v32i16(<32 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umin.v64i16(<64 x i16> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i16'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umin.v2i16(<2 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4 = call i16 @llvm.experimental.vector.reduce.umin.v4i16(<4 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8 = call i16 @llvm.experimental.vector.reduce.umin.v8i16(<8 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.experimental.vector.reduce.umin.v16i16(<16 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32 = call i16 @llvm.experimental.vector.reduce.umin.v32i16(<32 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.experimental.vector.reduce.umin.v64i16(<64 x i16> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i16'
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2 = call i16 @llvm.experimental.vector.reduce.umin.v2i16(<2 x i16> undef)
@@ -225,15 +234,15 @@ define i32 @reduce_i8(i32 %arg) {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umin.v128i8(<128 x i8> undef)
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
-; SSE42-LABEL: 'reduce_i8'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umin.v2i8(<2 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.umin.v4i8(<4 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.umin.v8i8(<8 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.umin.v16i8(<16 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.umin.v32i8(<32 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.umin.v64i8(<64 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umin.v128i8(<128 x i8> undef)
-; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
+; SSE4-LABEL: 'reduce_i8'
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umin.v2i8(<2 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i8 @llvm.experimental.vector.reduce.umin.v4i8(<4 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i8 @llvm.experimental.vector.reduce.umin.v8i8(<8 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i8 @llvm.experimental.vector.reduce.umin.v16i8(<16 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V32 = call i8 @llvm.experimental.vector.reduce.umin.v32i8(<32 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V64 = call i8 @llvm.experimental.vector.reduce.umin.v64i8(<64 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V128 = call i8 @llvm.experimental.vector.reduce.umin.v128i8(<128 x i8> undef)
+; SSE4-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_i8'
; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.experimental.vector.reduce.umin.v2i8(<2 x i8> undef)
More information about the llvm-commits
mailing list