This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Stage 2
- From: Mark Mitchell <mark at codesourcery dot com>
- To: Andrew Pinski <pinskia at physics dot uc dot edu>
- Cc: Ziemowit Laski <zlaski at apple dot com>, gcc at gcc dot gnu dot org
- Date: Tue, 06 Jul 2004 11:46:25 -0700
- Subject: Re: Stage 2
- Organization: CodeSourcery, LLC
- References: <200407061827.i66IRoX08985@tin.geop.uc.edu>
Andrew Pinski wrote:
Ziemowit Laski wrote:
Mark,
Will it still be possible to merge in objc-improvements-branch during
Stage 2?
Yes, probably, since we already reviewed this in some detail.
Make sure that you don't cause a compile-time performance regression for
plain C++; I'm currently seeing approximately 15% regression *since June
16th*. Lest anyone start invoking Moore's Law, let me point out that
this would result in doubling compile times every four months or so.
Can you pinpoint when the regression happened. I will note that Kenner's
patch came in during this time frame which could have caused this regression.
Not too easily.
This is on OS X, measuing Qt compile times. During much of this period,
the compiler would not build at all on this platform, or would not build
Qt. However, I think that it's likely that the regression happened
before the 24th, which would definitely include Kenner's patch, and I am
indeed very suspicious of that patch.
Would you be at all interested in helping to track this down, as a
fellow OS X user?
--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com