This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Should the patch for PR inline-asm/6162 go into 3.4?
- From: "Giovanni Bajo" <giovannibajo at libero dot it>
- To: "Ian Lance Taylor" <ian at wasabisystems dot com>,<mark at codesourcery dot com>
- Cc: <gcc at gcc dot gnu dot org>
- Date: Thu, 12 Feb 2004 01:51:07 +0100
- Subject: Re: Should the patch for PR inline-asm/6162 go into 3.4?
- References: <20040211214219.9819.qmail@gossamer.airs.com>
Ian Lance Taylor wrote:
> (Or put the PR into whatever state is appropriate
> for ``fixed on mainline, will not be fixed in earlier releases'').
Since the PR is a regression, our policy says it should stay open until the
affected branches are discontinued. The rationale is that, even if your patch
is too intrusive for the branch, it doesn't mean that someone else can't come
up with a trivial workaround somewhere in the future to fix the regression.
Of course, if the RM thinks it's better to close the bug because he reckons
that it's totally impossible to fix it in a non-intrusive way, the bug should
go to FIXED and the target milestone moved to 3.5.0. A comment can then explain
to the users that the bug won't be fixed in the branches for technical reasons.
The rule of thumb is that, when in doubt, it's always better to keep the PRs
open, they can't hurt :)
Giovanni Bajo