[llvm-commits] [PATCH]Fix a typo in X86ISelLowering.cpp
Cameron Zwarich
zwarich at apple.com
Thu Aug 4 13:14:22 PDT 2011
I didn't mean that it shouldn't be committed, just that it would be nice to have a test.
Cameron
On Aug 4, 2011, at 11:38 AM, Evan Cheng wrote:
> I agree with Duncan. The patch is obvious and I've committed it. A test case would be nice, obviously. :-)
>
> Thanks,
>
> Evan
>
> On Aug 4, 2011, at 3:04 AM, Ivan Krasin wrote:
>
>> Hi Cameron,
>>
>> On Thu, Aug 4, 2011 at 2:58 AM, Cameron Zwarich <zwarich at apple.com> wrote:
>>> Can you write a test case?
>> I would add this task into my queue, but I could not promise a fast turnaround.
>> Probably, it's fine, since this mistake already lives there for some
>> time and nobody complains about that.
>>
>> Ivan Krasin
>>
>>>
>>> Cameron
>>>
>>> ב-4 באוג 2011, בשעה 02:44, Ivan Krasin <krasin at chromium.org> כתב/ה:
>>>
>>>> Hi llvm team,
>>>>
>>>> this patch fixes a typo in lib/Target/X86/X86ISelLowering.cpp. The
>>>> credits go to http://habrahabr.ru/blogs/compilers/125626/ (Russian
>>>> hacker news)
>>>>
>>>> Ivan
>>>>
>>>> Index: lib/Target/X86/X86ISelLowering.cpp
>>>> ===================================================================
>>>> --- lib/Target/X86/X86ISelLowering.cpp (revision 136864)
>>>> +++ lib/Target/X86/X86ISelLowering.cpp (working copy)
>>>> @@ -11686,7 +11686,7 @@
>>>> // Converting this to a max would handle comparisons between positive
>>>> // and negative zero incorrectly.
>>>> if (!UnsafeFPMath &&
>>>> - !DAG.isKnownNeverZero(LHS) && !DAG.isKnownNeverZero(LHS))
>>>> + !DAG.isKnownNeverZero(LHS) && !DAG.isKnownNeverZero(RHS))
>>>> break;
>>>> Opcode = X86ISD::FMAX;
>>>> break;
>>>> <lhs-rhs-typo.patch>
>>>> _______________________________________________
>>>> llvm-commits mailing list
>>>> llvm-commits at cs.uiuc.edu
>>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>>
>>
>> _______________________________________________
>> 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