From 48ef68c20695f301dbb65d0b667228e190502a4d Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Sat, 18 Apr 2015 18:50:55 +0000
Subject: [PATCH] [X86][SSE] Force execution domain on float/double unpack
 shuffle tests.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@235259 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/stack-folding-fp-avx1.ll  | 32 ++++++++++++++++------
 test/CodeGen/X86/stack-folding-fp-sse42.ll | 16 ++++++++---
 2 files changed, 36 insertions(+), 12 deletions(-)

diff --git a/test/CodeGen/X86/stack-folding-fp-avx1.ll b/test/CodeGen/X86/stack-folding-fp-avx1.ll
index 18cd4179de8..c7c1fc94638 100644
--- a/test/CodeGen/X86/stack-folding-fp-avx1.ll
+++ b/test/CodeGen/X86/stack-folding-fp-avx1.ll
@@ -1699,7 +1699,9 @@ define <2 x double> @stack_fold_unpckhpd(<2 x double> %a0, <2 x double> %a1) {
   ;CHECK:       vunpckhpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 1, i32 3>
-  ret <2 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
+  ret <2 x double> %3
 }
 
 define <4 x double> @stack_fold_unpckhpd_ymm(<4 x double> %a0, <4 x double> %a1) {
@@ -1707,7 +1709,9 @@ define <4 x double> @stack_fold_unpckhpd_ymm(<4 x double> %a0, <4 x double> %a1)
   ;CHECK:       vunpckhpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
-  ret <4 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
+  ret <4 x double> %3
 }
 
 define <4 x float> @stack_fold_unpckhps(<4 x float> %a0, <4 x float> %a1) {
@@ -1715,7 +1719,9 @@ define <4 x float> @stack_fold_unpckhps(<4 x float> %a0, <4 x float> %a1) {
   ;CHECK:       vunpckhps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x float> %a0, <4 x float> %a1, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
-  ret <4 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <4 x float> %3
 }
 
 define <8 x float> @stack_fold_unpckhps_ymm(<8 x float> %a0, <8 x float> %a1) {
@@ -1723,7 +1729,9 @@ define <8 x float> @stack_fold_unpckhps_ymm(<8 x float> %a0, <8 x float> %a1) {
   ;CHECK:       vunpckhps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <8 x float> %a0, <8 x float> %a1, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
-  ret <8 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <8 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <8 x float> %3
 }
 
 define <2 x double> @stack_fold_unpcklpd(<2 x double> %a0, <2 x double> %a1) {
@@ -1731,7 +1739,9 @@ define <2 x double> @stack_fold_unpcklpd(<2 x double> %a0, <2 x double> %a1) {
   ;CHECK:       vunpcklpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 0, i32 2>
-  ret <2 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
+  ret <2 x double> %3
 }
 
 define <4 x double> @stack_fold_unpcklpd_ymm(<4 x double> %a0, <4 x double> %a1) {
@@ -1739,7 +1749,9 @@ define <4 x double> @stack_fold_unpcklpd_ymm(<4 x double> %a0, <4 x double> %a1)
   ;CHECK:       vunpcklpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
-  ret <4 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
+  ret <4 x double> %3
 }
 
 define <4 x float> @stack_fold_unpcklps(<4 x float> %a0, <4 x float> %a1) {
@@ -1747,7 +1759,9 @@ define <4 x float> @stack_fold_unpcklps(<4 x float> %a0, <4 x float> %a1) {
   ;CHECK:       vunpcklps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x float> %a0, <4 x float> %a1, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
-  ret <4 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <4 x float> %3
 }
 
 define <8 x float> @stack_fold_unpcklps_ymm(<8 x float> %a0, <8 x float> %a1) {
@@ -1755,7 +1769,9 @@ define <8 x float> @stack_fold_unpcklps_ymm(<8 x float> %a0, <8 x float> %a1) {
   ;CHECK:       vunpcklps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <8 x float> %a0, <8 x float> %a1, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
-  ret <8 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <8 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <8 x float> %3
 }
 
 define <2 x double> @stack_fold_xorpd(<2 x double> %a0, <2 x double> %a1) {
diff --git a/test/CodeGen/X86/stack-folding-fp-sse42.ll b/test/CodeGen/X86/stack-folding-fp-sse42.ll
index c26cc9df989..95f0c3d3a18 100644
--- a/test/CodeGen/X86/stack-folding-fp-sse42.ll
+++ b/test/CodeGen/X86/stack-folding-fp-sse42.ll
@@ -1035,7 +1035,9 @@ define <2 x double> @stack_fold_unpckhpd(<2 x double> %a0, <2 x double> %a1) {
   ;CHECK:       unpckhpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 1, i32 3>
-  ret <2 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
+  ret <2 x double> %3
 }
 
 define <4 x float> @stack_fold_unpckhps(<4 x float> %a0, <4 x float> %a1) {
@@ -1043,7 +1045,9 @@ define <4 x float> @stack_fold_unpckhps(<4 x float> %a0, <4 x float> %a1) {
   ;CHECK:       unpckhps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x float> %a0, <4 x float> %a1, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
-  ret <4 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <4 x float> %3
 }
 
 define <2 x double> @stack_fold_unpcklpd(<2 x double> %a0, <2 x double> %a1) {
@@ -1051,7 +1055,9 @@ define <2 x double> @stack_fold_unpcklpd(<2 x double> %a0, <2 x double> %a1) {
   ;CHECK:       unpcklpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 0, i32 2>
-  ret <2 x double> %2
+  ; fadd forces execution domain
+  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
+  ret <2 x double> %3
 }
 
 define <4 x float> @stack_fold_unpcklps(<4 x float> %a0, <4 x float> %a1) {
@@ -1059,7 +1065,9 @@ define <4 x float> @stack_fold_unpcklps(<4 x float> %a0, <4 x float> %a1) {
   ;CHECK:       unpcklps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
   %2 = shufflevector <4 x float> %a0, <4 x float> %a1, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
-  ret <4 x float> %2
+  ; fadd forces execution domain
+  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
+  ret <4 x float> %3
 }
 
 define <2 x double> @stack_fold_xorpd(<2 x double> %a0, <2 x double> %a1) {
-- 
2.34.1