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/46519: Missing vzeroupper


On Wed, Nov 24, 2010 at 10:53 AM, Uros Bizjak <ubizjak@gmail.com> wrote:
> On Wed, Nov 24, 2010 at 7:41 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
>
>> Here is a follow up ?patch. ?Fortran intrinsic may set TREE_THIS_VOLATILE
>> even if it does return. This patch removes the TREE_THIS_VOLATILE
>> optimization. ?OK for trunk?
>
> Looking at the testcase, I guess it is OK. Perhaps a fortran person
> can comment from the fortran POV?
>
> Uros.
>

Fortran is OK. The problem is

if (TARGET_VZEROUPPER && !TREE_THIS_VOLATILE (cfun->decl))

Even if caller, which is MAIN__, never returns, we should issue vzeroupper
when making a library call.  I am checking this one.

-- 
H.J.
---
gcc/

2010-11-24  H.J. Lu  <hongjiu.lu@intel.com>

	PR target/46519
	* config/i386/i386.c (ix86_expand_call): Don't check
	TREE_THIS_VOLATILE.

gcc/testsuite/

2010-11-24  H.J. Lu  <hongjiu.lu@intel.com>

	PR target/46519
	* gfortran.dg/pr46519-1.f: New.

Attachment: gcc-pr46519-6.patch
Description: Text document


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