[cfe-commits] r150238 - in /cfe/trunk: lib/CodeGen/CGCall.cpp test/CodeGen/arm-arguments.c

[Evgeniy Stepanov]

Thanks, disabling assertions helped reproduce the failure.

The bot mostly skips CMake steps, and it makes it hard to find out the
exact configuration.

On Fri, Feb 10, 2012 at 2:12 PM, Chandler Carruth <chandlerc at google.com> wrote:
> On Fri, Feb 10, 2012 at 12:03 AM, Evgeniy Stepanov
> <eugeni.stepanov at gmail.com> wrote:
>>
>> Hm, I can't reproduce this on my machine. Btw, is "release mode" the
>> same as --enable-optimized?
>
>
> No, you need to disable assertions as well.
>
>>
>> I guess I'll have to blindly loosen the check until the bot is happy :(
>
>
> Look at the printout from the bot -- it shows the variation that fails.
>
>>
>>
>> On Fri, Feb 10, 2012 at 1:46 PM, Chandler Carruth <chandlerc at google.com>
>> wrote:
>> > On Thu, Feb 9, 2012 at 11:30 PM, Evgeniy Stepanov
>> > <eugeni.stepanov at gmail.com> wrote:
>> >>
>> >> Author: eugenis
>> >> Date: Fri Feb 10 03:30:15 2012
>> >> New Revision: 150238
>> >>
>> >> URL: http://llvm.org/viewvc/llvm-project?rev=150238&view=rev
>> >> Log:
>> >> Fix function prolog codegen whe coerce-to type is a struct.
>> >>
>> >> This changes function prolog in such a way as to avoid out-of-bounds
>> >> stack store in the case when coerce-to type has a larger storage size
>> >> than the real argument type.
>> >>
>> >> Fixes PR11905.
>> >
>> >
>> > This is breaking some of our build bots:
>> >
>> >
>> > http://bb.pgr.jp/builders/cmake-clang-i686-msvc10/builds/7223/steps/test_clang/logs/Clang%20%3A%3A%20CodeGen___arm-arguments.c
>> >
>> > It looks like these tests fail in release mode, you should write your
>> > file
>> > check to check more loosely.
>> >
>
>