mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2025-01-31 20:51:52 +01:00
21418c85a1
If the mask input to getV4X86ShuffleImm8 only refers to a single source element (+ undefs) then canonicalize to a full broadcast. getV4X86ShuffleImm8 defaults to inline values for undefs, which can be useful for shuffle widening/narrowing but does leave SimplifyDemanded* calls thinking the shuffle depends on unnecessary elements. I'm still investigating what we should do more generally to avoid these undemanded elements, but broadcast cases was a simpler win.
24 lines
742 B
LLVM
24 lines
742 B
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=i686-unknown -mattr=+sse2 | FileCheck %s --check-prefix=X32
|
|
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+sse2 | FileCheck %s --check-prefix=X64
|
|
|
|
; Verify that we don't fail when shift by zero is encountered.
|
|
|
|
define i64 @test1(<2 x i64> %a) {
|
|
; X32-LABEL: test1:
|
|
; X32: # %bb.0: # %entry
|
|
; X32-NEXT: movd %xmm0, %eax
|
|
; X32-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,1,1]
|
|
; X32-NEXT: movd %xmm0, %edx
|
|
; X32-NEXT: retl
|
|
;
|
|
; X64-LABEL: test1:
|
|
; X64: # %bb.0: # %entry
|
|
; X64-NEXT: movq %xmm0, %rax
|
|
; X64-NEXT: retq
|
|
entry:
|
|
%c = shl <2 x i64> %a, <i64 0, i64 2>
|
|
%d = extractelement <2 x i64> %c, i32 0
|
|
ret i64 %d
|
|
}
|