6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-07-21 02:29:22 +00:00
Code Issues Projects Releases Wiki Activity

[X86, AVX] add an exedepfix entry for vmovq == vmovlps == vmovlpd

Browse Source 
This is the AVX extension of r235014:
http://llvm.org/viewvc/llvm-project?view=revision&revision=235014

Review:
http://reviews.llvm.org/D8691



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@235210 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Sanjay Patel 2015-04-17 17:02:37 +00:00
parent 80c84fe4ff
commit c7b16819e8
5 changed files with 7 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
lib/Target/X86/X86InstrInfo.cpp
View File

@ -6015,7 +6015,7 @@ static const uint16_t ReplaceableInstrs[][3] = {
{ X86::VMOVAPSrr, X86::VMOVAPDrr, X86::VMOVDQArr },
{ X86::VMOVUPSmr, X86::VMOVUPDmr, X86::VMOVDQUmr },
{ X86::VMOVUPSrm, X86::VMOVUPDrm, X86::VMOVDQUrm },
// TODO: Add the AVX versions of MOVLPSmr
{ X86::VMOVLPSmr, X86::VMOVLPDmr, X86::VMOVPQI2QImr },
{ X86::VMOVNTPSmr, X86::VMOVNTPDmr, X86::VMOVNTDQmr },
{ X86::VANDNPSrm, X86::VANDNPDrm, X86::VPANDNrm },
{ X86::VANDNPSrr, X86::VANDNPDrr, X86::VPANDNrr },

2
test/CodeGen/X86/avx-intrinsics-x86.ll
View File

@ -856,7 +856,7 @@ define void @test_x86_sse2_storel_dq(i8* %a0, <4 x i32> %a1) {
; CHECK-LABEL: test_x86_sse2_storel_dq:
; CHECK: # BB#0:
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
; CHECK-NEXT: vmovq %xmm0, (%eax)
; CHECK-NEXT: vmovlps %xmm0, (%eax)
; CHECK-NEXT: retl
call void @llvm.x86.sse2.storel.dq(i8* %a0, <4 x i32> %a1)
ret void

7
test/CodeGen/X86/exedeps-movq.ll
View File

@ -19,12 +19,7 @@ define void @store_floats(<4 x float> %x, i64* %p) {
; AVX-LABEL: store_floats:
; AVX: # BB#0:
; AVX-NEXT: vaddps %xmm0, %xmm0, %xmm0
; !!! FIXME - the AVX version is not handled correctly.
; AVX-NEXT: vmovq %xmm0, (%rdi)
; AVX-NEXT: vmovlps %xmm0, (%rdi)
; AVX-NEXT: retq
%a = fadd <4 x float> %x, %x
%b = shufflevector <4 x float> %a, <4 x float> undef, <2 x i32> <i32 0, i32 1>

4
test/CodeGen/X86/load-slice.ll
View File

@ -28,7 +28,7 @@
; Swap Imm and Real.
; STRESS-NEXT: vinsertps $16, [[RES_Imm]], [[RES_Real]], [[RES_Vec:%xmm[0-9]+]]
; Put the results back into out[out_start].
; STRESS-NEXT: vmovq [[RES_Vec]], ([[BASE]])
; STRESS-NEXT: vmovlps [[RES_Vec]], ([[BASE]])
;
; Same for REGULAR, we eliminate register bank copy with each slices.
; REGULAR-LABEL: t1:
@ -43,7 +43,7 @@
; Swap Imm and Real.
; REGULAR-NEXT: vinsertps $16, [[RES_Imm]], [[RES_Real]], [[RES_Vec:%xmm[0-9]+]]
; Put the results back into out[out_start].
; REGULAR-NEXT: vmovq [[RES_Vec]], ([[BASE]])
; REGULAR-NEXT: vmovlps [[RES_Vec]], ([[BASE]])
define void @t1(%class.Complex* nocapture %out, i64 %out_start) {
entry:
%arrayidx = getelementptr inbounds %class.Complex, %class.Complex* %out, i64 %out_start

4
test/CodeGen/X86/widen_load-1.ll
View File

@ -9,8 +9,8 @@
; SSE: movaps %xmm0, (%rsp)
; SSE: callq killcommon
; AVX: vmovdqa compl+128(%rip), %xmm0
; AVX: vmovdqa %xmm0, (%rsp)
; AVX: vmovaps compl+128(%rip), %xmm0
; AVX: vmovaps %xmm0, (%rsp)
; AVX: callq killcommon
@compl = linkonce global [20 x i64] zeroinitializer, align 64 ; <[20 x i64]*> [#uses=1]