This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: PATCH: PR target/46195: r165965 regresses i386 darwin


On Thu, Oct 28, 2010 at 7:11 PM, Jack Howarth <howarth@bromo.med.uc.edu> wrote:
> On Thu, Oct 28, 2010 at 10:33:14AM -0700, H.J. Lu wrote:
>> Hi,
>>
>> On Darwin, although long double is aligned to 16byte, it is aligned to
>> 4byte when passed on stack.
>>
>> Jack, please try this on Darwin. OK for trunk if there are no
>> regressions on Darwin and Linux?
>
> H.J.,
> ? I can confirm that all of the regressions introduced by r165965
> are eliminated on x86_64-apple-darwin10...
>
> http://gcc.gnu.org/ml/gcc-testresults/2010-10/msg02294.html
>
> Thanks for fixing this issue.
> ? ? ? ? ? ? ?Jack
>
>>
>> Thanks.
>>
>>
>> H.J.
>> ---
>> 2010-10-28 ?H.J. Lu ?<hongjiu.lu@intel.com>
>>
>> ? ? ? PR target/46195
>> ? ? ? * config/i386/i386.c (bigger_than_4byte_aligned_p): New.
>> ? ? ? (ix86_function_arg_boundary): Align long double parameters on
>> ? ? ? stack to 4byte in 32bit.

There are regressions on Linux nor Darwin.  OK for trunk?

Thanks.

-- 
H.J.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]