[X86][AVX] Added tests to load+broadcast non-zero'th vector elements

Baseline for an upcoming patch for PR23022

llvm-svn: 254898

2 files changed, 375 insertions, 3 deletions

diff --git a/llvm/test/CodeGen/X86/avx-vbroadcast.ll b/llvm/test/CodeGen/X86/avx-vbroadcast.ll
index bfc9149b107..5c0f43da876 100644
--- a/llvm/test/CodeGen/X86/avx-vbroadcast.ll
+++ b/llvm/test/CodeGen/X86/avx-vbroadcast.ll
@@ -102,6 +102,159 @@ entry:
   ret <4 x i32> %vecinit6.i
 }
 
+; FIXME: Pointer adjusted broadcasts
+
+define <4 x i32> @load_splat_4i32_4i32_1111(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i32_4i32_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpshufd {{.*#+}} xmm0 = mem[1,1,1,1]
+; CHECK-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 1, i32 1, i32 1, i32 1>
+  ret <4 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_4i32_33333333(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8i32_4i32_33333333:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = mem[3,3,3,3]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-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 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+  ret <8 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_8i32_55555555(<8 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8i32_8i32_55555555:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %ymm0
+; CHECK-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-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 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
+  ret <8 x i32> %ret
+}
+
+define <4 x float> @load_splat_4f32_4f32_1111(<4 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f32_4f32_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = mem[1,1,1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x float>, <4 x float>* %ptr
+  %ret = shufflevector <4 x float> %ld, <4 x float> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x float> %ret
+}
+
+define <8 x float> @load_splat_8f32_4f32_33333333(<4 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8f32_4f32_33333333:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = mem[3,3,3,3]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x float>, <4 x float>* %ptr
+  %ret = shufflevector <4 x float> %ld, <4 x float> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+  ret <8 x float> %ret
+}
+
+define <8 x float> @load_splat_8f32_8f32_55555555(<8 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8f32_8f32_55555555:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %ymm0
+; CHECK-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <8 x float>, <8 x float>* %ptr
+  %ret = shufflevector <8 x float> %ld, <8 x float> undef, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
+  ret <8 x float> %ret
+}
+
+define <2 x i64> @load_splat_2i64_2i64_1111(<2 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_2i64_2i64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpshufd {{.*#+}} xmm0 = mem[2,3,2,3]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x i64>, <2 x i64>* %ptr
+  %ret = shufflevector <2 x i64> %ld, <2 x i64> undef, <2 x i32> <i32 1, i32 1>
+  ret <2 x i64> %ret
+}
+
+define <4 x i64> @load_splat_4i64_2i64_1111(<2 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i64_2i64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %xmm0
+; CHECK-NEXT:    vmovhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x i64>, <2 x i64>* %ptr
+  %ret = shufflevector <2 x i64> %ld, <2 x i64> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x i64> %ret
+}
+
+define <4 x i64> @load_splat_4i64_4i64_2222(<4 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i64_4i64_2222:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovapd (%rdi), %ymm0
+; CHECK-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; CHECK-NEXT:    vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x i64>, <4 x i64>* %ptr
+  %ret = shufflevector <4 x i64> %ld, <4 x i64> undef, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
+  ret <4 x i64> %ret
+}
+
+define <2 x double> @load_splat_2f64_2f64_1111(<2 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_2f64_2f64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %xmm0
+; CHECK-NEXT:    vmovhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x double>, <2 x double>* %ptr
+  %ret = shufflevector <2 x double> %ld, <2 x double> undef, <2 x i32> <i32 1, i32 1>
+  ret <2 x double> %ret
+}
+
+define <4 x double> @load_splat_4f64_2f64_1111(<2 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f64_2f64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %xmm0
+; CHECK-NEXT:    vmovhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x double>, <2 x double>* %ptr
+  %ret = shufflevector <2 x double> %ld, <2 x double> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x double> %ret
+}
+
+define <4 x double> @load_splat_4f64_4f64_2222(<4 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f64_4f64_2222:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovapd (%rdi), %ymm0
+; CHECK-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; CHECK-NEXT:    vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; CHECK-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x double>, <4 x double>* %ptr
+  %ret = shufflevector <4 x double> %ld, <4 x double> undef, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
+  ret <4 x double> %ret
+}
+
 ; Unsupported vbroadcasts
 
 define <2 x i64> @G(i64* %ptr) nounwind uwtable readnone ssp {
diff --git a/llvm/test/CodeGen/X86/avx2-vbroadcast.ll b/llvm/test/CodeGen/X86/avx2-vbroadcast.ll
index 418707cdc23..186f5087365 100644
--- a/llvm/test/CodeGen/X86/avx2-vbroadcast.ll
+++ b/llvm/test/CodeGen/X86/avx2-vbroadcast.ll
@@ -172,6 +172,225 @@ entry:
   ret <4 x i64> %q3
 }
 
+; FIXME: Pointer adjusted broadcasts
+
+define <16 x i8> @load_splat_16i8_16i8_1111111111111111(<16 x i8>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_16i8_16i8_1111111111111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <16 x i8>, <16 x i8>* %ptr
+  %ret = shufflevector <16 x i8> %ld, <16 x i8> undef, <16 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <16 x i8> %ret
+}
+
+define <32 x i8> @load_splat_32i8_16i8_11111111111111111111111111111111(<16 x i8>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_32i8_16i8_11111111111111111111111111111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
+; CHECK-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <16 x i8>, <16 x i8>* %ptr
+  %ret = shufflevector <16 x i8> %ld, <16 x i8> undef, <32 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <32 x i8> %ret
+}
+
+define <32 x i8> @load_splat_32i8_32i8_11111111111111111111111111111111(<32 x i8>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_32i8_32i8_11111111111111111111111111111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %ymm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
+; CHECK-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <32 x i8>, <32 x i8>* %ptr
+  %ret = shufflevector <32 x i8> %ld, <32 x i8> undef, <32 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <32 x i8> %ret
+}
+
+define <8 x i16> @load_splat_8i16_8i16_11111111(<8 x i16>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8i16_8i16_11111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; CHECK-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 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <8 x i16> %ret
+}
+
+define <16 x i16> @load_splat_16i16_8i16_1111111111111111(<8 x i16>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_16i16_8i16_1111111111111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; CHECK-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <8 x i16>, <8 x i16>* %ptr
+  %ret = shufflevector <8 x i16> %ld, <8 x i16> undef, <16 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <16 x i16> %ret
+}
+
+define <16 x i16> @load_splat_16i16_16i16_1111111111111111(<16 x i16>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_16i16_16i16_1111111111111111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %ymm0
+; CHECK-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; CHECK-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; CHECK-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 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+  ret <16 x i16> %ret
+}
+
+define <4 x i32> @load_splat_4i32_4i32_1111(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i32_4i32_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpshufd {{.*#+}} xmm0 = mem[1,1,1,1]
+; CHECK-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 1, i32 1, i32 1, i32 1>
+  ret <4 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_4i32_33333333(<4 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8i32_4i32_33333333:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpbroadcastd LCPI15_0(%rip), %ymm1
+; CHECK-NEXT:    vpermd %ymm0, %ymm1, %ymm0
+; CHECK-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 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+  ret <8 x i32> %ret
+}
+
+define <8 x i32> @load_splat_8i32_8i32_55555555(<8 x i32>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8i32_8i32_55555555:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpbroadcastd LCPI16_0(%rip), %ymm0
+; CHECK-NEXT:    vpermd (%rdi), %ymm0, %ymm0
+; CHECK-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 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
+  ret <8 x i32> %ret
+}
+
+define <4 x float> @load_splat_4f32_4f32_1111(<4 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f32_4f32_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermilps {{.*#+}} xmm0 = mem[1,1,1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x float>, <4 x float>* %ptr
+  %ret = shufflevector <4 x float> %ld, <4 x float> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x float> %ret
+}
+
+define <8 x float> @load_splat_8f32_4f32_33333333(<4 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8f32_4f32_33333333:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %xmm0

+; CHECK-NEXT:    vbroadcastss LCPI18_0(%rip), %ymm1

+; CHECK-NEXT:    vpermps %ymm0, %ymm1, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x float>, <4 x float>* %ptr
+  %ret = shufflevector <4 x float> %ld, <4 x float> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+  ret <8 x float> %ret
+}
+
+define <8 x float> @load_splat_8f32_8f32_55555555(<8 x float>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_8f32_8f32_55555555:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vbroadcastss LCPI19_0(%rip), %ymm0
+; CHECK-NEXT:    vpermps (%rdi), %ymm0, %ymm0
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <8 x float>, <8 x float>* %ptr
+  %ret = shufflevector <8 x float> %ld, <8 x float> undef, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
+  ret <8 x float> %ret
+}
+
+define <2 x i64> @load_splat_2i64_2i64_1111(<2 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_2i64_2i64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpshufd {{.*#+}} xmm0 = mem[2,3,2,3]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x i64>, <2 x i64>* %ptr
+  %ret = shufflevector <2 x i64> %ld, <2 x i64> undef, <2 x i32> <i32 1, i32 1>
+  ret <2 x i64> %ret
+}
+
+define <4 x i64> @load_splat_4i64_2i64_1111(<2 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i64_2i64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovdqa (%rdi), %xmm0
+; CHECK-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[1,1,1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x i64>, <2 x i64>* %ptr
+  %ret = shufflevector <2 x i64> %ld, <2 x i64> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x i64> %ret
+}
+
+define <4 x i64> @load_splat_4i64_4i64_2222(<4 x i64>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4i64_4i64_2222:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermq {{.*#+}} ymm0 = mem[2,2,2,2]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x i64>, <4 x i64>* %ptr
+  %ret = shufflevector <4 x i64> %ld, <4 x i64> undef, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
+  ret <4 x i64> %ret
+}
+
+define <2 x double> @load_splat_2f64_2f64_1111(<2 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_2f64_2f64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovaps (%rdi), %xmm0
+; CHECK-NEXT:    vmovhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x double>, <2 x double>* %ptr
+  %ret = shufflevector <2 x double> %ld, <2 x double> undef, <2 x i32> <i32 1, i32 1>
+  ret <2 x double> %ret
+}
+
+define <4 x double> @load_splat_4f64_2f64_1111(<2 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f64_2f64_1111:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vmovapd (%rdi), %xmm0
+; CHECK-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[1,1,1,1]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <2 x double>, <2 x double>* %ptr
+  %ret = shufflevector <2 x double> %ld, <2 x double> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
+  ret <4 x double> %ret
+}
+
+define <4 x double> @load_splat_4f64_4f64_2222(<4 x double>* %ptr) nounwind uwtable readnone ssp {
+; CHECK-LABEL: load_splat_4f64_4f64_2222:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    vpermpd {{.*#+}} ymm0 = mem[2,2,2,2]
+; CHECK-NEXT:    retq
+entry:
+  %ld = load <4 x double>, <4 x double>* %ptr
+  %ret = shufflevector <4 x double> %ld, <4 x double> undef, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
+  ret <4 x double> %ret
+}
+
 ; make sure that we still don't support broadcast double into 128-bit vector
 ; this used to crash
 define <2 x double> @I(double* %ptr) nounwind uwtable readnone ssp {
@@ -242,13 +461,13 @@ define void @crash() nounwind alwaysinline {
 ; CHECK:       ## BB#0: ## %WGLoopsEntry
 ; CHECK-NEXT:    xorl %eax, %eax
 ; CHECK-NEXT:    testb %al, %al
-; CHECK-NEXT:    je LBB13_1
+; CHECK-NEXT:    je LBB31_1
 ; CHECK-NEXT:  ## BB#2: ## %ret
 ; CHECK-NEXT:    retq
 ; CHECK-NEXT:    .align 4, 0x90
-; CHECK-NEXT:  LBB13_1: ## %footer349VF
+; CHECK-NEXT:  LBB31_1: ## %footer349VF
 ; CHECK-NEXT:    ## =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    jmp LBB13_1
+; CHECK-NEXT:    jmp LBB31_1
 WGLoopsEntry:
   br i1 undef, label %ret, label %footer329VF