This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [v3] libstdc++/24071
- From: Eric Botcazou <ebotcazou at libertysurf dot fr>
- To: Benjamin Kosnik <bkoz at redhat dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Tue, 27 Sep 2005 12:04:16 +0200
- Subject: Re: [v3] libstdc++/24071
- References: <20050926205836.200a857f.bkoz@redhat.com>
> This reverts to the previous, slightly twisted logic in mt_alloc
> initialization. It should fix the new solaris fails.
As confirmed privately, this works fine on the 4.0 branch. I think the long
term fix are propositions #3 and #4 of your message in the PR.
> I anxiously await the next, monthly installment of solaris test results
> for gcc mainline posted on gcc-testresults, which should verify this claim.
A bit of irony, right? :-)
That's not the best approach for dealing with this kind of problems. If you
think a patch may have unexpected effects on Solaris or if the patch does
address a Solaris-specific problem, then send it to me beforehand and I'd
happy to test it. Paolo did so a few times and this worked well.
FYI I usually test mainline on a weekly basis for 6 SPARC/Solaris platforms,
although I don't always post the results. I can't do more since there are 2
other open branches, I'm the only one regularly testing on these platforms
and the machines are not dedicated to GCC development. But I stop everything
when a release is about to be made off one of the branch and a spate of RCs is
issued in a matter of days; that's why the most recent results I have for
mainline are dated 09/12 (and are clean for libstdc++-v3).
--
Eric Botcazou