This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch] ATTR_NORETURN -> ATTRIBUTE_NORETURN unification [Re: [patch] ATTR_* -> ATTRIBUTE_* unification]
- From: Pedro Alves <pedro at codesourcery dot com>
- To: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- Cc: gdb-patches at sourceware dot org, Mark Kettenis <mark dot kettenis at xs4all dot nl>
- Date: Mon, 3 May 2010 00:41:55 +0100
- Subject: Re: [patch] ATTR_NORETURN -> ATTRIBUTE_NORETURN unification [Re: [patch] ATTR_* -> ATTRIBUTE_* unification]
- References: <20100430181605.GA19190@host0.dyn.jankratochvil.net> <20100502213636.GA21869@host0.dyn.jankratochvil.net> <20100502231941.GA27802@host0.dyn.jankratochvil.net>
On Monday 03 May 2010 00:19:41, Jan Kratochvil wrote:
> On Sun, 02 May 2010 23:36:36 +0200, Jan Kratochvil wrote:
> > I will post some similar NORETURN + ATTR_NORETURN -> ATTRIBUTE_NORETURN.
> >
> > Going to drop NORETURN as it is only for gcc < 2.7, ansidecl.h does not
> > support it and a missing noreturn flag can only affect code size, performance
> > and affect compilation diagnostics.
>
> error_no_arg was present both in command.h and cli/cli-cmds.h. Moving it the
> other way would require #include changes.
>
> mips_error will start to be effectively declared as noreturn.
>
> No regressions on {x86_64,x86_64-m32,i686}-fedora12-linux-gnu.
>
This is okay. Thanks for doing this.
> Similar work is needed in gdb/gdbserver/.
>
> doc/ talks about .mh files; but when similar macros are already used from
> include/ for bfd/ etc. it probably was inappropriate to do from .mh.
Yeah.
--
Pedro Alves