This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] Fix PR43464: update loop closed SSA form once copy prop is done
On Wed, Mar 31, 2010 at 03:26, Richard Guenther <rguenther@suse.de> wrote:
> On Tue, 30 Mar 2010, Sebastian Pop wrote:
>
>> On Tue, Mar 30, 2010 at 10:59, Sebastian Pop <sebpop@gmail.com> wrote:
>> > On Tue, Mar 30, 2010 at 08:52, Zdenek Dvorak <rakdver@kam.mff.cuni.cz> wrote:
>> >> Hi,
>> >>
>> >>> > sure -- just check whether at least one predecessor edge of the block in that the phi node is located
>> >>> > is a loop exit (this is unnecessarily conservative, of course, but so is the current code),
>> >>>
>> >>> Something like the following? ?(I don't want to rely on recorded
>> >>> loop exits)
>> >>
>> >> almost -- this would also return true for loop entry edges. ?I'd suggest
>> >> loop_exit_edge_p (e->src->loop_father, e)
>> >>
>> >> Zdenek
>> >>
>> >>> Sebastian - does this fix your testcase?
>> >
>> > Yes, the attached patch with the suggestion from Zdenek fixes the
>> > problem on the graphite branch. ?I will commit this to the branch for
>> > extra testing on the SPEC 2006.
>>
>> The patch passed bootstrap and test (all languages except Java)
>> on amd64-linux on the Graphite branch.
>
> The patch is ok for trunk.
The SPEC 2006 passed on the graphite branch without regressions.
I will commit the patch later today.
Sebastian