This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Ping: patches for -mms-bitfields
- From: Roger Sayle <roger at eyesopen dot com>
- To: Kaz Kojima <kkojima at rr dot iij4u dot or dot jp>, Eric Christopher <echristo at apple dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Fri, 14 Jul 2006 21:20:23 -0600 (MDT)
- Subject: Re: Ping: patches for -mms-bitfields
On Fri, 14 Jul 2006, Eric Christopher wrote:
> > Fix PR middle-end/28160: Bogus error with -mms-bitfields
> > http://gcc.gnu.org/ml/gcc-patches/2006-06/msg01472.html
> >
> > Fix PR middle-end/28161: Wrong bit field layout with -mms-bitfield
> > http://gcc.gnu.org/ml/gcc-patches/2006-06/msg01473.html
>
> I've taken a look at these and they're fine by me - you'll need an
> official maintainer for the file to approve them though.
As Eric's confirmed that the logic is correct, these patches are
OK for mainline. Although the bugzilla PRs don't list these as
regressions in the summary/title, the "known to work" field lists
4.1.1, which would make these mainline regressions, and therefore
suitable during stage 3.
I can't tell from my reading of the PRs whether the failures in the
g++ testsuite occur with default testing, or only when -mms-bitfields
is explicitly used? It might make sense to duplicate the affected
testcases, and check them with -mms-bitfields to prevent this behavior
regressing, if we're not regularly testing this functionality already.
Many thanks for fixing this, and to Eric for checking that the fixes
are reasonable.
Roger
--