[llvm-commits] [PATCH] optimization of x == -y for X86
Chad Rosier
mcrosier at apple.com
Fri Apr 27 14:00:03 PDT 2012
Manman,
On Apr 27, 2012, at 11:22 AM, Jay Foad wrote:
> On 27 April 2012 19:12, manman ren <mren at apple.com> wrote:
>>
>> This patch optimizes codegen for if(x==-y) on x86
>> from
>> negl %esi
>> cmpl %esi, %edi
>> je .LBB0_2
>> to
>> addl %esi, %edi
>> je .L4
>
>> Index: test/CodeGen/X86/neg_cmp.ll
>> ===================================================================
>> --- test/CodeGen/X86/neg_cmp.ll (revision 0)
>> +++ test/CodeGen/X86/neg_cmp.ll (revision 0)
>> @@ -0,0 +1,21 @@
>> +; RUN: llc < %s -march=x86-64 | FileCheck %s
>> +
>> +; rdar://11245199
>
> Can you mention PR 12545 please? I don't think I have access to rdar:// (do I?).
LGTM; Once you've added the PR number, I'd be happy to commit. Also, please make sure to update the status of both the radar and the PR (see: http://llvm.org/bugs/show_bug.cgi?id=12545). You'll need to setup a Bugzilla account if you haven't already done so.
Jay,
Thanks for pointing this out as radar information is not publicly viewable.
Chad
> Thanks for fixing it!
>
> Jay.
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
More information about the llvm-commits
mailing list