This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: bootstrap problems
- To: Gabriel Dos Reis <gdr at codesourcery dot com>
- Subject: Re: bootstrap problems
- From: Jan Hubicka <jh at suse dot cz>
- Date: Thu, 7 Jun 2001 12:45:21 +0200
- Cc: Jan Hubicka <jh at suse dot cz>, Jeffrey Oldham <oldham at codesourcery dot com>, gcc at gcc dot gnu dot org
- References: <20010605091247.A7113@codesourcery.com> <20010606132120.F5609@atrey.karlin.mff.cuni.cz> <20010606084820.A8350@codesourcery.com> <20010607120702.G27948@atrey.karlin.mff.cuni.cz> <flzobkoa44.fsf@sel.cmla.ens-cachan.fr>
> This is not the first time this type of problem is showing up.
Thats point of my email (see 2). The last time it happened, the latent problem
was kept in tree. It is now in 3.0 and we are unlikely to fix it properly,
as fix is involved. That is sad :(
It only affects ARM, but still I consider it as failure.
>
> I think, no matter how good is your patch, if it prevents people from
> bootstrapping and continuing their work then it ought to be reverted
> until a patch is found for the latent bug.
I've just commited one.
There is probably another latent bug, but I will first need to set up the
environment for full mips bootstrap, as Jeffrey didn't sent me the
command line and I can't reproduce it at the moment.
I will take care to bring mips back to bootstrapland till night.
>
> We should not take the exponential-growth road of branching. We do
> already have a developement branch and we should keep it as the
> referential development source for everybody.
Thats why I suggest the system of tags. It will keep thinks linear.
Honza
>
> -- Gaby
> CodeSourcery, LLC http://www.codesourcery.com