mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-10-25 14:02:52 +02:00
f9f7ca4f85
We have a number of useful lowering strategies for VBROADCAST instructions (both from memory and register element 0) which the 128-bit form of the MOVDDUP instruction can make use of. This patch tweaks lowerVectorShuffleAsBroadcast to enable it to broadcast 2f64 args using MOVDDUP as well. It does require a slight tweak to the lowerVectorShuffleAsBroadcast mechanism as the existing MOVDDUP lowering uses isShuffleEquivalent which can match binary shuffles that can lower to (unary) broadcasts. Differential Revision: http://reviews.llvm.org/D17680 llvm-svn: 262478
60 lines
1.6 KiB
LLVM
60 lines
1.6 KiB
LLVM
; RUN: llc < %s -mcpu=corei7 -march=x86 -mattr=+sse4.1 | FileCheck %s
|
|
|
|
define void @t1(float* %R, <4 x float>* %P1) nounwind {
|
|
; CHECK-LABEL: t1:
|
|
; CHECK: # BB#0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %ecx
|
|
; CHECK-NEXT: movss 12(%ecx), %xmm0
|
|
; CHECK-NEXT: movss %xmm0, (%eax)
|
|
; CHECK-NEXT: retl
|
|
|
|
%X = load <4 x float>, <4 x float>* %P1
|
|
%tmp = extractelement <4 x float> %X, i32 3
|
|
store float %tmp, float* %R
|
|
ret void
|
|
}
|
|
|
|
define float @t2(<4 x float>* %P1) nounwind {
|
|
; CHECK-LABEL: t2:
|
|
; CHECK: # BB#0:
|
|
; CHECK-NEXT: pushl %eax
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: movddup {{.*#+}} xmm0 = mem[0,0]
|
|
; CHECK-NEXT: movss %xmm0, (%esp)
|
|
; CHECK-NEXT: flds (%esp)
|
|
; CHECK-NEXT: popl %eax
|
|
; CHECK-NEXT: retl
|
|
|
|
%X = load <4 x float>, <4 x float>* %P1
|
|
%tmp = extractelement <4 x float> %X, i32 2
|
|
ret float %tmp
|
|
}
|
|
|
|
define void @t3(i32* %R, <4 x i32>* %P1) nounwind {
|
|
; CHECK-LABEL: t3:
|
|
; CHECK: # BB#0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %ecx
|
|
; CHECK-NEXT: movl 12(%ecx), %ecx
|
|
; CHECK-NEXT: movl %ecx, (%eax)
|
|
; CHECK-NEXT: retl
|
|
|
|
%X = load <4 x i32>, <4 x i32>* %P1
|
|
%tmp = extractelement <4 x i32> %X, i32 3
|
|
store i32 %tmp, i32* %R
|
|
ret void
|
|
}
|
|
|
|
define i32 @t4(<4 x i32>* %P1) nounwind {
|
|
; CHECK-LABEL: t4:
|
|
; CHECK: # BB#0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: movl 12(%eax), %eax
|
|
; CHECK-NEXT: retl
|
|
|
|
%X = load <4 x i32>, <4 x i32>* %P1
|
|
%tmp = extractelement <4 x i32> %X, i32 3
|
|
ret i32 %tmp
|
|
}
|