This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH]: PR c/34921 Misalign stack variable referenced by nested function
Richard Sandiford wrote:
> "H.J. Lu" <hjl.tools@gmail.com> writes:
>
>>> >
>>> > 2008-01-22 Joey Ye <joey.ye@intel.com>
>>> > H.J. Lu <hongjiu.lu@intel.com>
>>> >
>>> > * gcc.c-torture/execute/nest-align-1.c: New test case.
>>> >
>>> >
>>> This is FAILing on mipsel-linux:
>>>
>>> http://gcc.gnu.org/ml/gcc-testresults/2008-02/msg01622.html
>>>
>>>
>> It may fail on targets whose preferred stack boundary is less than 16byte.
>>
>
> OK, let's get the ball rolling then...
>
> Tested on x86_64-linux-gnu, where the test still runs, and
> mipsisa64-elfoabi, where it doesn't. Applied.
>
> (I don't think XFAILing is appropriate. It's the kind of thing
> that could work accidentally (and unpredictably) in some run-time
> environments. Better just to skip.)
>
Thanks Richard.
Why doesn't __attribute__ (( aligned(16) )) issue a warning if the
specified alignment cannot be guaranteed?
I think the stack alignment branch is working on this issue for x86, but
I am not sure about the availability of a general solution.
David Daney