[X86] Add Issue #61104 test case

Shows the failure of combineBitcastvxi1 to sign-extend a vXi1 allones vselect operand

There's a number of other problems in Issue #61104 still to address, but this one has an easy quick fix

diff --git a/llvm/test/CodeGen/X86/bitcast-and-setcc-256.ll b/llvm/test/CodeGen/X86/bitcast-and-setcc-256.ll
index 225ebac..3b04da9 100644 (file)
--- a/llvm/test/CodeGen/X86/bitcast-and-setcc-256.ll
+++ b/llvm/test/CodeGen/X86/bitcast-and-setcc-256.ll
@@ -445,6 +445,81 @@ define i8 @v8i32_or_and(<8 x i32> %a, <8 x i32> %b, <8 x i32> %c, <8 x i32> %d,
   ret i8 %res
 }
 
+; PR61104 - peek through vselect allones operand
+; combineBitcastvxi1 may be called before the fold vselect(c,-1,x) --> or(c,x) vXi1 fold
+
+define i8 @v8i32_or_select(<8 x i32> %a0, <8 x i32> %a1, <8 x i32> %a2) {
+; SSE2-SSSE3-LABEL: v8i32_or_select:
+; SSE2-SSSE3:       # %bb.0:
+; SSE2-SSSE3-NEXT:    pcmpeqd %xmm3, %xmm1
+; SSE2-SSSE3-NEXT:    pcmpeqd %xmm2, %xmm0
+; SSE2-SSSE3-NEXT:    packssdw %xmm1, %xmm0
+; SSE2-SSSE3-NEXT:    packssdw %xmm5, %xmm4
+; SSE2-SSSE3-NEXT:    por %xmm0, %xmm4
+; SSE2-SSSE3-NEXT:    packsswb %xmm4, %xmm4
+; SSE2-SSSE3-NEXT:    pmovmskb %xmm4, %eax
+; SSE2-SSSE3-NEXT:    # kill: def $al killed $al killed $eax
+; SSE2-SSSE3-NEXT:    retq
+;
+; AVX1-LABEL: v8i32_or_select:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vpcmpeqd %xmm1, %xmm0, %xmm3
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpcmpeqd %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vextractf128 $1, %ymm2, %xmm1
+; AVX1-NEXT:    vpor %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vpor %xmm2, %xmm3, %xmm1
+; AVX1-NEXT:    vpackssdw %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    vpacksswb %xmm0, %xmm0, %xmm0
+; AVX1-NEXT:    vpmovmskb %xmm0, %eax
+; AVX1-NEXT:    # kill: def $al killed $al killed $eax
+; AVX1-NEXT:    vzeroupper
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: v8i32_or_select:
+; AVX2:       # %bb.0:
+; AVX2-NEXT:    vpcmpeqd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpcmpgtd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpor %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpackssdw %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpacksswb %xmm0, %xmm0, %xmm0
+; AVX2-NEXT:    vpmovmskb %xmm0, %eax
+; AVX2-NEXT:    # kill: def $al killed $al killed $eax
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
+;
+; AVX512F-LABEL: v8i32_or_select:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    vpcmpeqd %ymm1, %ymm0, %k0
+; AVX512F-NEXT:    vpxor %xmm0, %xmm0, %xmm0
+; AVX512F-NEXT:    vpcmpgtd %ymm2, %ymm0, %k1
+; AVX512F-NEXT:    korw %k1, %k0, %k0
+; AVX512F-NEXT:    kmovw %k0, %eax
+; AVX512F-NEXT:    # kill: def $al killed $al killed $eax
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512BW-LABEL: v8i32_or_select:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    vpcmpeqd %ymm1, %ymm0, %k0
+; AVX512BW-NEXT:    vpxor %xmm0, %xmm0, %xmm0
+; AVX512BW-NEXT:    vpcmpgtd %ymm2, %ymm0, %k1
+; AVX512BW-NEXT:    korw %k1, %k0, %k0
+; AVX512BW-NEXT:    kmovd %k0, %eax
+; AVX512BW-NEXT:    # kill: def $al killed $al killed $eax
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+  %cmp = icmp eq <8 x i32> %a0, %a1
+  %slt = icmp slt <8 x i32> %a2, zeroinitializer
+  %sel = select <8 x i1> %cmp, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, <8 x i1> %slt
+  %res = bitcast <8 x i1> %sel to i8
+  ret i8 %res
+}
+
+
 define i8 @v8f32_and(<8 x float> %a, <8 x float> %b, <8 x float> %c, <8 x float> %d) {
 ; SSE2-SSSE3-LABEL: v8f32_and:
 ; SSE2-SSSE3:       # %bb.0: