[PATCH] D149249: [X86] Handle multiple use freeze(undef) in LowerAVXCONCAT_VECTORS as zero vectors

[Simon Pilgrim via Phabricator via llvm-commits]

This revision was landed with ongoing or failed builds.
This revision was automatically updated to reflect the committed changes.
Closed by commit rGee52af74d8e5: [X86] Handle multiple use freeze(undef) in LowerAVXCONCAT_VECTORS as zero… (authored by RKSimon).

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D149249/new/

https://reviews.llvm.org/D149249

Files:
  llvm/lib/Target/X86/X86ISelLowering.cpp
  llvm/test/CodeGen/X86/avx512-intrinsics.ll
  llvm/test/CodeGen/X86/avx512fp16-intrinsics.ll


Index: llvm/test/CodeGen/X86/avx512fp16-intrinsics.ll
===================================================================
--- llvm/test/CodeGen/X86/avx512fp16-intrinsics.ll
+++ llvm/test/CodeGen/X86/avx512fp16-intrinsics.ll
@@ -1231,10 +1231,7 @@
 define <32 x half> @test_mm512_castph128_ph512_freeze(<8 x half> %a0) nounwind {
 ; CHECK-LABEL: test_mm512_castph128_ph512_freeze:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm1
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; CHECK-NEXT:    vinsertf64x4 $1, %ymm1, %zmm0, %zmm0
+; CHECK-NEXT:    vmovaps %xmm0, %xmm0
 ; CHECK-NEXT:    retq
   %a1 = freeze <8 x half> poison
   %res = shufflevector <8 x half> %a0, <8 x half> %a1, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
Index: llvm/test/CodeGen/X86/avx512-intrinsics.ll
===================================================================
--- llvm/test/CodeGen/X86/avx512-intrinsics.ll
+++ llvm/test/CodeGen/X86/avx512-intrinsics.ll
@@ -7495,10 +7495,7 @@
 define <8 x double> @test_mm256_castpd128_pd256_freeze(<2 x double> %a0) nounwind {
 ; CHECK-LABEL: test_mm256_castpd128_pd256_freeze:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm1
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; CHECK-NEXT:    vinsertf64x4 $1, %ymm1, %zmm0, %zmm0
+; CHECK-NEXT:    vmovaps %xmm0, %xmm0
 ; CHECK-NEXT:    ret{{[l|q]}}
   %a1 = freeze <2 x double> poison
   %res = shufflevector <2 x double> %a0, <2 x double> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3>
@@ -7520,10 +7517,7 @@
 define <16 x float> @test_mm256_castps128_ps512_freeze(<4 x float> %a0) nounwind {
 ; CHECK-LABEL: test_mm256_castps128_ps512_freeze:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm1
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; CHECK-NEXT:    vinsertf64x4 $1, %ymm1, %zmm0, %zmm0
+; CHECK-NEXT:    vmovaps %xmm0, %xmm0
 ; CHECK-NEXT:    ret{{[l|q]}}
   %a1 = freeze <4 x float> poison
   %res = shufflevector <4 x float> %a0, <4 x float> %a1, <16x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 4, i32 5, i32 6, i32 7, i32 4, i32 5, i32 6, i32 7>
@@ -7545,10 +7539,7 @@
 define <8 x i64> @test_mm512_castsi128_si512_freeze(<2 x i64> %a0) nounwind {
 ; CHECK-LABEL: test_mm512_castsi128_si512_freeze:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm1
-; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
-; CHECK-NEXT:    vinsertf64x4 $1, %ymm1, %zmm0, %zmm0
+; CHECK-NEXT:    vmovaps %xmm0, %xmm0
 ; CHECK-NEXT:    ret{{[l|q]}}
   %a1 = freeze <2 x i64> poison
   %res = shufflevector <2 x i64> %a0, <2 x i64> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 2, i32 3>
Index: llvm/lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- llvm/lib/Target/X86/X86ISelLowering.cpp
+++ llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -11663,8 +11663,13 @@
     SDValue SubVec = Op.getOperand(i);
     if (SubVec.isUndef())
       continue;
-    if (ISD::isFreezeUndef(SubVec.getNode()) && SubVec.hasOneUse())
-      ++NumFreezeUndef;
+    if (ISD::isFreezeUndef(SubVec.getNode())) {
+        // If the freeze(undef) has multiple uses then we must fold to zero.
+        if (SubVec.hasOneUse())
+          ++NumFreezeUndef;
+        else
+          ++NumZero;
+    }
     else if (ISD::isBuildVectorAllZeros(SubVec.getNode()))
       ++NumZero;
     else {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D149249.517121.patch
Type: text/x-patch
Size: 3927 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230426/10a9e6b5/attachment.bin>