[PATCH] D35320: [X86][SSE] Add support for extending bool vectors bitcasted from scalars.

[Elena Demikhovsky via Phabricator via llvm-commits]

delena added a comment.

What should be returned at the end according to the calling conventions - zext or sext? In some cases you add shift-right in order to return zext value. In the AVX-512 cases the last inst is vpmovm2* that means sext.



================
Comment at: test/CodeGen/X86/bitcast-int-to-vector-bool.ll:24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    movzbl %dil, %eax
+; AVX1-NEXT:    vmovq %rax, %xmm0
----------------
One "mov" is enough here. you, probably, use "zext" instead of "anyext". (I did not look at the code yet)


================
Comment at: test/CodeGen/X86/bitcast-int-to-vector-bool.ll:120
+; AVX1-NEXT:    vmovd %edi, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,0,1,1]
----------------
pshufb can be used for broadcasting i8.


================
Comment at: test/CodeGen/X86/bitcast-int-to-vector-bool.ll:133
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm1 = [1,2,4,8,16,32,64,128]
+; AVX2-NEXT:    vpand %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpcmpeqw %xmm1, %xmm0, %xmm0
----------------
why load is not folded in vpand?


Repository:
  rL LLVM

https://reviews.llvm.org/D35320