diff options
| author | Clement Courbet <courbet@google.com> | 2018-12-13 09:56:19 +0000 |
|---|---|---|
| committer | Clement Courbet <courbet@google.com> | 2018-12-13 09:56:19 +0000 |
| commit | 76f4ae109266fae1d17cade097a14216bc10822e (patch) | |
| tree | 25d0590be62d8c00aa7c8ed652d23869ad92a3ae /llvm/test/CodeGen/X86/memcpy-from-string.ll | |
| parent | a257639a6935a2c63377784c5c9c3b73864a2582 (diff) | |
| download | bcm5719-llvm-76f4ae109266fae1d17cade097a14216bc10822e.tar.gz bcm5719-llvm-76f4ae109266fae1d17cade097a14216bc10822e.zip | |
[CodeGen] Allow mempcy/memset to generate small overlapping stores.
Summary:
All targets either just return false here or properly model `Fast`, so I
don't think there is any reason to prevent CodeGen from doing the right
thing here.
Subscribers: nemanjai, javed.absar, eraman, jsji, llvm-commits
Differential Revision: https://reviews.llvm.org/D55365
llvm-svn: 349016
Diffstat (limited to 'llvm/test/CodeGen/X86/memcpy-from-string.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/memcpy-from-string.ll | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/llvm/test/CodeGen/X86/memcpy-from-string.ll b/llvm/test/CodeGen/X86/memcpy-from-string.ll index 454667276b4..af88ffa843d 100644 --- a/llvm/test/CodeGen/X86/memcpy-from-string.ll +++ b/llvm/test/CodeGen/X86/memcpy-from-string.ll @@ -16,8 +16,7 @@ target triple = "x86_64-unknown-linux-gnu" define void @foo(i8* %tmp2) { ; X86-LABEL: foo: ; X86: # %bb.0: -; X86-NEXT: movb $0, 6(%rdi) -; X86-NEXT: movw $15212, 4(%rdi) # imm = 0x3B6C +; X86-NEXT: movl $3894379, 3(%rdi) # imm = 0x3B6C6B ; X86-NEXT: movl $1802117222, (%rdi) # imm = 0x6B6A2066 ; X86-NEXT: retq call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp2, i8* getelementptr inbounds ([10 x i8], [10 x i8]* @0, i64 0, i64 3), i64 7, i1 false) |
