This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Update expanders to use optimize_insn_for_size predicate.
On Sun, Aug 3, 2008 at 4:01 PM, Jan Hubicka <jh@suse.cz> wrote:
>> On Sat, Aug 2, 2008 at 9:15 PM, Jan Hubicka <jh@suse.cz> wrote:
>> > Hi,
>> > this patch replace some of optimize_size tests by
>> > optimize_insn_for_speed_p predicate so we can make decisions on per-BB
>> > granuality.
>>
>> Ok.
>>
>> Btw - can we revert the inlining changes caused by the function-specific
>> changes? I think Meissner is waiting for your feedback.
>
> We need to benchmark if they really cause the SPECfp regression and if
> they do I would also like to figure out what cold functions are
> appearing in the benchmarks and why.
Are you planning to do this somewhen soon? These regressions really shadow
the effects of the tuples merge if there are any. Note that the SPEC
2k6 regressions
are even more prominent.
I'd rather just revert these changes as they are not really part of
the function-specific
optimization infrastructure and should have been merged as a separate patch.
Richard.