This is the mail archive of the
mailing list for the GCC project.
Re: PATCH: testsuite/objc.dg/bitfield-2.m requires signed char
- From: Richard Earnshaw <rearnsha at cambridge dot arm dot com>
- To: Stan Shebs <shebs at apple dot com>
- Cc: Richard dot Earnshaw at arm dot com, gcc-patches at gcc dot gnu dot org, Ziemowit Laski <zlaski at apple dot com>
- Date: Tue, 17 Sep 2002 11:37:48 +0100
- Subject: Re: PATCH: testsuite/objc.dg/bitfield-2.m requires signed char
- Organization: ARM Ltd.
- Reply-to: Richard dot Earnshaw at arm dot com
> Richard Earnshaw wrote:
> >The objc test bitfield-2.m fails on machines that don't have char signed
> >by default. This includes ARM and probably several other machines.
> >rather than mangle the test in a way which is probably not desirable, it
> >seams best to force the compiler to used signed chars for this case.
> >2002-09-16 Richard Earnshaw <firstname.lastname@example.org>
> > * objc.dg/bitfield-2.m (dg-options): Add -fsigned-char.
> Amusing that this runs, seeing as how the NeXT runtime hasn't been ported to
> ARM, but in any case yes, this is OK.
There wouldn't have been much point in making the change if it didn't fix
a test failure for me...
As a general point, if tests like this are relying on some system library
(or whatever), then the exp file should be detecting this requirement and
not running the test on platforms that don't have it.