[X86, AVX] try to lowerVectorShuffleAsElementInsertion() for all 256-bit vector sub-types

I suggested this change in D7898 (http://llvm.org/viewvc/llvm-project?view=revision&revision=231354)

It improves the v4i64 case although not optimally. This AVX codegen:

  vmovq {{.*#+}} xmm0 = mem[0],zero
  vxorpd %ymm1, %ymm1, %ymm1
  vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3]

Becomes:

  vmovsd {{.*#+}} xmm0 = mem[0],zero

Unfortunately, this doesn't completely solve PR22685. There are still at least 2 problems under here:

    We're not handling v32i8 / v16i16.
    We're not getting the FP / int domains right for instruction selection.

But since this patch alone appears to do no harm, reduces code duplication, and helps v4i64, 
I'm submitting this patch ahead of fixing the above.

Differential Revision: http://reviews.llvm.org/D8341

llvm-svn: 233704

1 files changed, 7 insertions, 6 deletions

diff --git a/llvm/test/CodeGen/X86/2012-1-10-buildvector.ll b/llvm/test/CodeGen/X86/2012-1-10-buildvector.ll
index a9b8cc6c62a..ece28bfb6c5 100644
--- a/llvm/test/CodeGen/X86/2012-1-10-buildvector.ll
+++ b/llvm/test/CodeGen/X86/2012-1-10-buildvector.ll
@@ -3,25 +3,26 @@
 target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f80:128:128-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32-S32"
 target triple = "i686-pc-win32"
 
-;CHECK-LABEL: bad_cast:
+; CHECK-LABEL: bad_cast:
 define void @bad_cast() {
 entry:
   %vext.i = shufflevector <2 x i64> undef, <2 x i64> undef, <3 x i32> <i32 0, i32 1, i32 undef>
   %vecinit8.i = shufflevector <3 x i64> zeroinitializer, <3 x i64> %vext.i, <3 x i32> <i32 0, i32 3, i32 4>
   store <3 x i64> %vecinit8.i, <3 x i64>* undef, align 32
-;CHECK: ret
+; CHECK: ret
   ret void
 }
 
 
-;CHECK-LABEL: bad_insert:
+; CHECK-LABEL: bad_insert:
 define void @bad_insert(i32 %t) {
 entry:
-;CHECK: vxorps %ymm1, %ymm1, %ymm1
-;CHECK-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3,4,5,6,7]
+; CHECK:      vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; CHECK-NEXT: vmovaps %ymm0 
+; CHECK:      ret
+
   %v2 = insertelement <8 x i32> zeroinitializer, i32 %t, i32 0
   store <8 x i32> %v2, <8 x i32> addrspace(1)* undef, align 32
-;CHECK: ret
   ret void
 }