[X86][SSE] Added tests where we should be trying to widen a load+splat into a broadcast

llvm-svn: 276527

diff --git a/llvm/test/CodeGen/X86/widened-broadcast.ll b/llvm/test/CodeGen/X86/widened-broadcast.ll
new file mode 100644 (file)
index 0000000..0c71db5
--- /dev/null
+++ b/llvm/test/CodeGen/X86/widened-broadcast.ll
@@ -0,0 +1,149 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE42
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f | FileCheck %s --check-prefix=AVX --check-prefix=AVX512 --check-prefix=AVX512F
+
+; Widened shuffle broadcast loads
+
+define <4 x i32> @load_splat_4i32_4i32_0101(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; SSE-LABEL: load_splat_4i32_4i32_0101:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pshufd {{.*#+}} xmm0 = mem[0,1,0,1]
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: load_splat_4i32_4i32_0101:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = mem[0,1,0,1]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_splat_4i32_4i32_0101:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpbroadcastq (%rdi), %xmm0
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: load_splat_4i32_4i32_0101:
+; AVX512:       # BB#0: # %entry
+; AVX512-NEXT:    vpbroadcastq (%rdi), %xmm0
+; AVX512-NEXT:    retq
+entry:
+  %ld = load <4 x i32>, <4 x i32>* %ptr
+  %ret = shufflevector <4 x i32> %ld, <4 x i32> undef, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
+  ret <4 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_4i32_01010101(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; SSE-LABEL: load_splat_8i32_4i32_01010101:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pshufd {{.*#+}} xmm0 = mem[0,1,0,1]
+; SSE-NEXT:    movdqa %xmm0, %xmm1
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: load_splat_8i32_4i32_01010101:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = mem[0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_splat_8i32_4i32_01010101:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vmovaps (%rdi), %xmm0
+; AVX2-NEXT:    vbroadcastsd %xmm0, %ymm0
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: load_splat_8i32_4i32_01010101:
+; AVX512:       # BB#0: # %entry
+; AVX512-NEXT:    vmovaps (%rdi), %xmm0
+; AVX512-NEXT:    vbroadcastsd %xmm0, %ymm0
+; AVX512-NEXT:    retq
+entry:
+  %ld = load <4 x i32>, <4 x i32>* %ptr
+  %ret = shufflevector <4 x i32> %ld, <4 x i32> undef, <8 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1>
+  ret <8 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_8i32_01010101(<8 x i32>* %ptr) nounwind uwtable readnone ssp {
+; SSE-LABEL: load_splat_8i32_8i32_01010101:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pshufd {{.*#+}} xmm0 = mem[0,1,0,1]
+; SSE-NEXT:    movdqa %xmm0, %xmm1
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: load_splat_8i32_8i32_01010101:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vmovapd (%rdi), %ymm0
+; AVX1-NEXT:    vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_splat_8i32_8i32_01010101:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vbroadcastsd (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: load_splat_8i32_8i32_01010101:
+; AVX512:       # BB#0: # %entry
+; AVX512-NEXT:    vbroadcastsd (%rdi), %ymm0
+; AVX512-NEXT:    retq
+entry:
+  %ld = load <8 x i32>, <8 x i32>* %ptr
+  %ret = shufflevector <8 x i32> %ld, <8 x i32> undef, <8 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1>
+  ret <8 x i32> %ret
+}
+
+define <8 x i16> @load_splat_8i16_8i16_01010101(<8 x i16>* %ptr) nounwind uwtable readnone ssp {
+; SSE-LABEL: load_splat_8i16_8i16_01010101:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pshufd {{.*#+}} xmm0 = mem[0,0,0,0]
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: load_splat_8i16_8i16_01010101:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = mem[0,0,0,0]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_splat_8i16_8i16_01010101:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vbroadcastss (%rdi), %xmm0
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: load_splat_8i16_8i16_01010101:
+; AVX512:       # BB#0: # %entry
+; AVX512-NEXT:    vbroadcastss (%rdi), %xmm0
+; AVX512-NEXT:    retq
+entry:
+  %ld = load <8 x i16>, <8 x i16>* %ptr
+  %ret = shufflevector <8 x i16> %ld, <8 x i16> undef, <8 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1>
+  ret <8 x i16> %ret
+}
+
+define <16 x i16> @load_splat_16i16_16i16_01010101(<16 x i16>* %ptr) nounwind uwtable readnone ssp {
+; SSE-LABEL: load_splat_16i16_16i16_01010101:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pshufd {{.*#+}} xmm0 = mem[0,0,0,0]
+; SSE-NEXT:    movdqa %xmm0, %xmm1
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: load_splat_16i16_16i16_01010101:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vmovaps (%rdi), %ymm0
+; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_splat_16i16_16i16_01010101:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vbroadcastss (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: load_splat_16i16_16i16_01010101:
+; AVX512:       # BB#0: # %entry
+; AVX512-NEXT:    vbroadcastss (%rdi), %ymm0
+; AVX512-NEXT:    retq
+entry:
+  %ld = load <16 x i16>, <16 x i16>* %ptr
+  %ret = shufflevector <16 x i16> %ld, <16 x i16> undef, <16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1>
+  ret <16 x i16> %ret
+}
+