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: [RFC][PATCH] backport faster dominance info patch from tree-ssa


On Sun, 28 Dec 2003, Steven Bosscher wrote:
> Unfortunately it does not fix any PRs, but it does speed up the
> compiler somewhat (~49s down from ~51.5s (~5%) for Geralds PR8361
> testcase [...]

5% on that testcase is nothing to sneer at! :-)

> which is in a way also a bugfix.

It definitely is bug fix, for PR8361 is a high priority regression
for GCC 3.4, so given your rationale...

> 1. it is a speedup, and we need lots of those ;-)
> 2. the bulk of the changes are in dominance and et-forest,
>    so this patch is mostly just an interface change
> 3. it has already had lots of testing on the tree-ssa branch.

...I really think this should be considered even at stage 1.

(Would you mind adding a reference to your patch to the PR?)

Thanks,
Gerald


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