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 PR39806] Potential fix for spurious code generation

On Mon, Nov 30, 2009 at 4:06 PM, Jakub Jelinek <jakub@redhat.com> wrote:
> On Mon, Nov 30, 2009 at 05:00:10PM +0200, Olga Golovanevsky wrote:
>> > If you are never walking the hashtable via htab_traverse drop
>> > the patch, otherwise install it (it's still bogus then but less so).
>> >
>> > Richard.
>>
>> I do walk. Committed.
>
> Then you might want something like Richard's approach in tree-into-ssa.c from
> http://gcc.gnu.org/ml/gcc-patches/2009-11/msg01618.html, otherwise you risk
> -g/-g0 code generation differences.

Only for the walks where you generate either variables or code from
of course.  Analysis walks are fine I think.

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