[AVX-512] Re-generate tests that were updated for r290663 without usi…

…ng update_llc_test_checks.py so duplicate check lines weren't merged. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@290868 91177308-0d34-0410-b5e6-96231b3b80d8
simoatze · Jan 3, 2017 · ddf520d · ddf520d
1 parent 6bb8971
commit ddf520d
Show file tree

Hide file tree

Showing 4 changed files with 295 additions and 410 deletions.
diff --git a/test/CodeGen/X86/avx2-vbroadcast.ll b/test/CodeGen/X86/avx2-vbroadcast.ll
@@ -1105,55 +1105,30 @@ define <4 x double> @splat_concat4(double %d) {
 ; Those test cases exerce the latter.
 
 define void @isel_crash_16b(i8* %cV_R.addr) {
-; X32-AVX2-LABEL: isel_crash_16b:
-; X32-AVX2:       ## BB#0: ## %eintry
-; X32-AVX2-NEXT:    subl $60, %esp
-; X32-AVX2-NEXT:  Lcfi0:
-; X32-AVX2-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX2-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX2-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX2-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX2-NEXT:    vpbroadcastb (%eax), %xmm1
-; X32-AVX2-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    addl $60, %esp
-; X32-AVX2-NEXT:    retl
-;
-; X64-AVX2-LABEL: isel_crash_16b:
-; X64-AVX2:       ## BB#0: ## %eintry
-; X64-AVX2-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X64-AVX2-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    movb (%rdi), %al
-; X64-AVX2-NEXT:    vmovd %eax, %xmm1
-; X64-AVX2-NEXT:    vpbroadcastb %xmm1, %xmm1
-; X64-AVX2-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    retq
-;
-; X32-AVX512VL-LABEL: isel_crash_16b:
-; X32-AVX512VL:       ## BB#0: ## %eintry
-; X32-AVX512VL-NEXT:    subl $60, %esp
-; X32-AVX512VL-NEXT:  Lcfi0:
-; X32-AVX512VL-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX512VL-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX512VL-NEXT:    vpbroadcastb (%eax), %xmm1
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    addl $60, %esp
-; X32-AVX512VL-NEXT:    retl
+; X32-LABEL: isel_crash_16b:
+; X32:       ## BB#0: ## %eintry
+; X32-NEXT:    subl $60, %esp
+; X32-NEXT:  Lcfi0:
+; X32-NEXT:    .cfi_def_cfa_offset 64
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    vxorps %xmm0, %xmm0, %xmm0
+; X32-NEXT:    vmovaps %xmm0, (%esp)
+; X32-NEXT:    vpbroadcastb (%eax), %xmm1
+; X32-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
+; X32-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
+; X32-NEXT:    addl $60, %esp
+; X32-NEXT:    retl
 ;
-; X64-AVX512VL-LABEL: isel_crash_16b:
-; X64-AVX512VL:       ## BB#0: ## %eintry
-; X64-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X64-AVX512VL-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    movb (%rdi), %al
-; X64-AVX512VL-NEXT:    vmovd %eax, %xmm1
-; X64-AVX512VL-NEXT:    vpbroadcastb %xmm1, %xmm1
-; X64-AVX512VL-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    retq
+; X64-LABEL: isel_crash_16b:
+; X64:       ## BB#0: ## %eintry
+; X64-NEXT:    vxorps %xmm0, %xmm0, %xmm0
+; X64-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    movb (%rdi), %al
+; X64-NEXT:    vmovd %eax, %xmm1
+; X64-NEXT:    vpbroadcastb %xmm1, %xmm1
+; X64-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    retq
 eintry:
   %__a.addr.i = alloca <2 x i64>, align 16
   %__b.addr.i = alloca <2 x i64>, align 16
@@ -1277,55 +1252,30 @@ eintry:
 }
 
 define void @isel_crash_8w(i16* %cV_R.addr) {
-; X32-AVX2-LABEL: isel_crash_8w:
-; X32-AVX2:       ## BB#0: ## %entry
-; X32-AVX2-NEXT:    subl $60, %esp
-; X32-AVX2-NEXT:  Lcfi4:
-; X32-AVX2-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX2-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX2-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX2-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX2-NEXT:    vpbroadcastw (%eax), %xmm1
-; X32-AVX2-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    addl $60, %esp
-; X32-AVX2-NEXT:    retl
-;
-; X64-AVX2-LABEL: isel_crash_8w:
-; X64-AVX2:       ## BB#0: ## %entry
-; X64-AVX2-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X64-AVX2-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    movw (%rdi), %ax
-; X64-AVX2-NEXT:    vmovd %eax, %xmm1
-; X64-AVX2-NEXT:    vpbroadcastw %xmm1, %xmm1
-; X64-AVX2-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
-; X64-AVX2-NEXT:    retq
-;
-; X32-AVX512VL-LABEL: isel_crash_8w:
-; X32-AVX512VL:       ## BB#0: ## %entry
-; X32-AVX512VL-NEXT:    subl $60, %esp
-; X32-AVX512VL-NEXT:  Lcfi4:
-; X32-AVX512VL-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX512VL-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX512VL-NEXT:    vpbroadcastw (%eax), %xmm1
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    addl $60, %esp
-; X32-AVX512VL-NEXT:    retl
+; X32-LABEL: isel_crash_8w:
+; X32:       ## BB#0: ## %entry
+; X32-NEXT:    subl $60, %esp
+; X32-NEXT:  Lcfi4:
+; X32-NEXT:    .cfi_def_cfa_offset 64
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    vxorps %xmm0, %xmm0, %xmm0
+; X32-NEXT:    vmovaps %xmm0, (%esp)
+; X32-NEXT:    vpbroadcastw (%eax), %xmm1
+; X32-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
+; X32-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
+; X32-NEXT:    addl $60, %esp
+; X32-NEXT:    retl
 ;
-; X64-AVX512VL-LABEL: isel_crash_8w:
-; X64-AVX512VL:       ## BB#0: ## %entry
-; X64-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X64-AVX512VL-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    movw (%rdi), %ax
-; X64-AVX512VL-NEXT:    vmovd %eax, %xmm1
-; X64-AVX512VL-NEXT:    vpbroadcastw %xmm1, %xmm1
-; X64-AVX512VL-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
-; X64-AVX512VL-NEXT:    retq
+; X64-LABEL: isel_crash_8w:
+; X64:       ## BB#0: ## %entry
+; X64-NEXT:    vxorps %xmm0, %xmm0, %xmm0
+; X64-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    movw (%rdi), %ax
+; X64-NEXT:    vmovd %eax, %xmm1
+; X64-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X64-NEXT:    vmovaps %xmm0, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
+; X64-NEXT:    retq
 entry:
   %__a.addr.i = alloca <2 x i64>, align 16
   %__b.addr.i = alloca <2 x i64>, align 16
@@ -1605,24 +1555,24 @@ eintry:
 }
 
 define void @isel_crash_2q(i64* %cV_R.addr) {
-; X32-AVX2-LABEL: isel_crash_2q:
-; X32-AVX2:       ## BB#0: ## %entry
-; X32-AVX2-NEXT:    subl $60, %esp
-; X32-AVX2-NEXT:  Lcfi12:
-; X32-AVX2-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX2-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX2-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX2-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX2-NEXT:    movl (%eax), %ecx
-; X32-AVX2-NEXT:    movl 4(%eax), %eax
-; X32-AVX2-NEXT:    vmovd %ecx, %xmm1
-; X32-AVX2-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
-; X32-AVX2-NEXT:    vpinsrd $2, %ecx, %xmm1, %xmm1
-; X32-AVX2-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm1
-; X32-AVX2-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX2-NEXT:    addl $60, %esp
-; X32-AVX2-NEXT:    retl
+; X32-LABEL: isel_crash_2q:
+; X32:       ## BB#0: ## %entry
+; X32-NEXT:    subl $60, %esp
+; X32-NEXT:  Lcfi12:
+; X32-NEXT:    .cfi_def_cfa_offset 64
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    vxorps %xmm0, %xmm0, %xmm0
+; X32-NEXT:    vmovaps %xmm0, (%esp)
+; X32-NEXT:    movl (%eax), %ecx
+; X32-NEXT:    movl 4(%eax), %eax
+; X32-NEXT:    vmovd %ecx, %xmm1
+; X32-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $2, %ecx, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm1
+; X32-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
+; X32-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
+; X32-NEXT:    addl $60, %esp
+; X32-NEXT:    retl
 ;
 ; X64-AVX2-LABEL: isel_crash_2q:
 ; X64-AVX2:       ## BB#0: ## %entry
@@ -1635,25 +1585,6 @@ define void @isel_crash_2q(i64* %cV_R.addr) {
 ; X64-AVX2-NEXT:    vmovdqa %xmm1, -{{[0-9]+}}(%rsp)
 ; X64-AVX2-NEXT:    retq
 ;
-; X32-AVX512VL-LABEL: isel_crash_2q:
-; X32-AVX512VL:       ## BB#0: ## %entry
-; X32-AVX512VL-NEXT:    subl $60, %esp
-; X32-AVX512VL-NEXT:  Lcfi12:
-; X32-AVX512VL-NEXT:    .cfi_def_cfa_offset 64
-; X32-AVX512VL-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, (%esp)
-; X32-AVX512VL-NEXT:    movl (%eax), %ecx
-; X32-AVX512VL-NEXT:    movl 4(%eax), %eax
-; X32-AVX512VL-NEXT:    vmovd %ecx, %xmm1
-; X32-AVX512VL-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
-; X32-AVX512VL-NEXT:    vpinsrd $2, %ecx, %xmm1, %xmm1
-; X32-AVX512VL-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm1
-; X32-AVX512VL-NEXT:    vmovaps %xmm0, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    vmovdqa %xmm1, {{[0-9]+}}(%esp)
-; X32-AVX512VL-NEXT:    addl $60, %esp
-; X32-AVX512VL-NEXT:    retl
-;
 ; X64-AVX512VL-LABEL: isel_crash_2q:
 ; X64-AVX512VL:       ## BB#0: ## %entry
 ; X64-AVX512VL-NEXT:    vxorps %xmm0, %xmm0, %xmm0

diff --git a/test/CodeGen/X86/subvector-broadcast.ll b/test/CodeGen/X86/subvector-broadcast.ll
@@ -1349,6 +1349,44 @@ define <16 x i32> @test_broadcast_4i32_16i32_chain(<4 x i32>* %p0, <4 x float>*
 @gb4 = global <8 x i64> zeroinitializer, align 8
 
 define void @fallback_broadcast_v4i64_to_v8i64(<4 x i64> %a, <8 x i64> %b) {
+; X32-AVX1-LABEL: fallback_broadcast_v4i64_to_v8i64:
+; X32-AVX1:       ## BB#0: ## %entry
+; X32-AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; X32-AVX1-NEXT:    vmovaps {{.*#+}} ymm4 = [1,0,2,0,3,0,4,0]
+; X32-AVX1-NEXT:    vextractf128 $1, %ymm4, %xmm5
+; X32-AVX1-NEXT:    vpaddq %xmm5, %xmm3, %xmm3
+; X32-AVX1-NEXT:    vpaddq %xmm4, %xmm0, %xmm0
+; X32-AVX1-NEXT:    vinsertf128 $1, %xmm3, %ymm0, %ymm0
+; X32-AVX1-NEXT:    vextractf128 $1, %ymm2, %xmm3
+; X32-AVX1-NEXT:    vpaddq %xmm5, %xmm3, %xmm3
+; X32-AVX1-NEXT:    vpaddq %xmm4, %xmm2, %xmm2
+; X32-AVX1-NEXT:    vinsertf128 $1, %xmm3, %ymm2, %ymm2
+; X32-AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; X32-AVX1-NEXT:    vpaddq %xmm5, %xmm3, %xmm3
+; X32-AVX1-NEXT:    vpaddq %xmm4, %xmm1, %xmm1
+; X32-AVX1-NEXT:    vinsertf128 $1, %xmm3, %ymm1, %ymm1
+; X32-AVX1-NEXT:    vandps %ymm4, %ymm1, %ymm1
+; X32-AVX1-NEXT:    vandps %ymm4, %ymm2, %ymm2
+; X32-AVX1-NEXT:    vmovups %ymm0, _ga4
+; X32-AVX1-NEXT:    vmovups %ymm2, _gb4+32
+; X32-AVX1-NEXT:    vmovups %ymm1, _gb4
+; X32-AVX1-NEXT:    vzeroupper
+; X32-AVX1-NEXT:    retl
+;
+; X32-AVX2-LABEL: fallback_broadcast_v4i64_to_v8i64:
+; X32-AVX2:       ## BB#0: ## %entry
+; X32-AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [1,0,2,0,3,0,4,0]
+; X32-AVX2-NEXT:    vpaddq %ymm3, %ymm0, %ymm0
+; X32-AVX2-NEXT:    vpaddq %ymm3, %ymm2, %ymm2
+; X32-AVX2-NEXT:    vpaddq %ymm3, %ymm1, %ymm1
+; X32-AVX2-NEXT:    vpand %ymm3, %ymm1, %ymm1
+; X32-AVX2-NEXT:    vpand %ymm3, %ymm2, %ymm2
+; X32-AVX2-NEXT:    vmovdqu %ymm0, _ga4
+; X32-AVX2-NEXT:    vmovdqu %ymm2, _gb4+32
+; X32-AVX2-NEXT:    vmovdqu %ymm1, _gb4
+; X32-AVX2-NEXT:    vzeroupper
+; X32-AVX2-NEXT:    retl
+;
 ; X32-AVX512-LABEL: fallback_broadcast_v4i64_to_v8i64:
 ; X32-AVX512:       ## BB#0: ## %entry
 ; X32-AVX512-NEXT:    vpaddq LCPI26_0, %ymm0, %ymm0
@@ -1359,6 +1397,45 @@ define void @fallback_broadcast_v4i64_to_v8i64(<4 x i64> %a, <8 x i64> %b) {
 ; X32-AVX512-NEXT:    vmovdqu64 %zmm1, _gb4
 ; X32-AVX512-NEXT:    retl
 ;
+; X64-AVX1-LABEL: fallback_broadcast_v4i64_to_v8i64:
+; X64-AVX1:       ## BB#0: ## %entry
+; X64-AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; X64-AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [3,4]
+; X64-AVX1-NEXT:    vpaddq %xmm4, %xmm3, %xmm3
+; X64-AVX1-NEXT:    vmovdqa {{.*#+}} xmm5 = [1,2]
+; X64-AVX1-NEXT:    vpaddq %xmm5, %xmm0, %xmm0
+; X64-AVX1-NEXT:    vinsertf128 $1, %xmm3, %ymm0, %ymm0
+; X64-AVX1-NEXT:    vmovaps {{.*#+}} ymm3 = [1,2,3,4]
+; X64-AVX1-NEXT:    vextractf128 $1, %ymm2, %xmm6
+; X64-AVX1-NEXT:    vpaddq %xmm4, %xmm6, %xmm6
+; X64-AVX1-NEXT:    vpaddq %xmm5, %xmm2, %xmm2
+; X64-AVX1-NEXT:    vinsertf128 $1, %xmm6, %ymm2, %ymm2
+; X64-AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm6
+; X64-AVX1-NEXT:    vpaddq %xmm4, %xmm6, %xmm4
+; X64-AVX1-NEXT:    vpaddq %xmm5, %xmm1, %xmm1
+; X64-AVX1-NEXT:    vinsertf128 $1, %xmm4, %ymm1, %ymm1
+; X64-AVX1-NEXT:    vandps %ymm3, %ymm1, %ymm1
+; X64-AVX1-NEXT:    vandps %ymm3, %ymm2, %ymm2
+; X64-AVX1-NEXT:    vmovups %ymm0, {{.*}}(%rip)
+; X64-AVX1-NEXT:    vmovups %ymm2, _gb4+{{.*}}(%rip)
+; X64-AVX1-NEXT:    vmovups %ymm1, {{.*}}(%rip)
+; X64-AVX1-NEXT:    vzeroupper
+; X64-AVX1-NEXT:    retq
+;
+; X64-AVX2-LABEL: fallback_broadcast_v4i64_to_v8i64:
+; X64-AVX2:       ## BB#0: ## %entry
+; X64-AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [1,2,3,4]
+; X64-AVX2-NEXT:    vpaddq %ymm3, %ymm0, %ymm0
+; X64-AVX2-NEXT:    vpaddq %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vpaddq %ymm3, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpand %ymm3, %ymm1, %ymm1
+; X64-AVX2-NEXT:    vpand %ymm3, %ymm2, %ymm2
+; X64-AVX2-NEXT:    vmovdqu %ymm0, {{.*}}(%rip)
+; X64-AVX2-NEXT:    vmovdqu %ymm2, _gb4+{{.*}}(%rip)
+; X64-AVX2-NEXT:    vmovdqu %ymm1, {{.*}}(%rip)
+; X64-AVX2-NEXT:    vzeroupper
+; X64-AVX2-NEXT:    retq
+;
 ; X64-AVX512-LABEL: fallback_broadcast_v4i64_to_v8i64:
 ; X64-AVX512:       ## BB#0: ## %entry
 ; X64-AVX512-NEXT:    vmovdqa {{.*#+}} ymm2 = [1,2,3,4]
@@ -1383,6 +1460,20 @@ entry:
 @gb2 = global <8 x double> zeroinitializer, align 8
 
 define void @fallback_broadcast_v4f64_to_v8f64(<4 x double> %a, <8 x double> %b) {
+; X32-AVX-LABEL: fallback_broadcast_v4f64_to_v8f64:
+; X32-AVX:       ## BB#0: ## %entry
+; X32-AVX-NEXT:    vmovapd {{.*#+}} ymm3 = [1.000000e+00,2.000000e+00,3.000000e+00,4.000000e+00]
+; X32-AVX-NEXT:    vaddpd %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vaddpd %ymm3, %ymm2, %ymm2
+; X32-AVX-NEXT:    vaddpd %ymm3, %ymm1, %ymm1
+; X32-AVX-NEXT:    vdivpd %ymm3, %ymm1, %ymm1
+; X32-AVX-NEXT:    vdivpd %ymm3, %ymm2, %ymm2
+; X32-AVX-NEXT:    vmovupd %ymm0, _ga2
+; X32-AVX-NEXT:    vmovupd %ymm2, _gb2+32
+; X32-AVX-NEXT:    vmovupd %ymm1, _gb2
+; X32-AVX-NEXT:    vzeroupper
+; X32-AVX-NEXT:    retl
+;
 ; X32-AVX512-LABEL: fallback_broadcast_v4f64_to_v8f64:
 ; X32-AVX512:       ## BB#0: ## %entry
 ; X32-AVX512-NEXT:    vmovapd {{.*#+}} ymm2 = [1.000000e+00,2.000000e+00,3.000000e+00,4.000000e+00]
@@ -1394,6 +1485,20 @@ define void @fallback_broadcast_v4f64_to_v8f64(<4 x double> %a, <8 x double> %b)
 ; X32-AVX512-NEXT:    vmovupd %zmm1, _gb2
 ; X32-AVX512-NEXT:    retl
 ;
+; X64-AVX-LABEL: fallback_broadcast_v4f64_to_v8f64:
+; X64-AVX:       ## BB#0: ## %entry
+; X64-AVX-NEXT:    vmovapd {{.*#+}} ymm3 = [1.000000e+00,2.000000e+00,3.000000e+00,4.000000e+00]
+; X64-AVX-NEXT:    vaddpd %ymm3, %ymm0, %ymm0
+; X64-AVX-NEXT:    vaddpd %ymm3, %ymm2, %ymm2
+; X64-AVX-NEXT:    vaddpd %ymm3, %ymm1, %ymm1
+; X64-AVX-NEXT:    vdivpd %ymm3, %ymm1, %ymm1
+; X64-AVX-NEXT:    vdivpd %ymm3, %ymm2, %ymm2
+; X64-AVX-NEXT:    vmovupd %ymm0, {{.*}}(%rip)
+; X64-AVX-NEXT:    vmovupd %ymm2, _gb2+{{.*}}(%rip)
+; X64-AVX-NEXT:    vmovupd %ymm1, {{.*}}(%rip)
+; X64-AVX-NEXT:    vzeroupper
+; X64-AVX-NEXT:    retq
+;
 ; X64-AVX512-LABEL: fallback_broadcast_v4f64_to_v8f64:
 ; X64-AVX512:       ## BB#0: ## %entry
 ; X64-AVX512-NEXT:    vmovapd {{.*#+}} ymm2 = [1.000000e+00,2.000000e+00,3.000000e+00,4.000000e+00]