This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] argv needs to be writable
- To: Zack Weinberg <zack at wolery dot cumb dot org>
- Subject: Re: [PATCH] argv needs to be writable
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Wed, 14 Jun 2000 12:57:28 -0600
- cc: "David O'Brien" <obrien at NUXI dot com>, gcc-patches at gcc dot gnu dot org
- Reply-To: law at cygnus dot com
In message <20000614104113.G26867@wolery.cumb.org>you write:
> On Wed, Jun 14, 2000 at 10:47:35AM -0600, Jeffrey A Law wrote:
> >
> > In message <20000614095105.D26867@wolery.cumb.org>you write:
> > > On Wed, Jun 14, 2000 at 09:19:37AM -0700, David O'Brien wrote:
> > > > On Wed, Jun 14, 2000 at 09:53:29AM -0600, Jeffrey A Law wrote:
> > > > > Actually, I would prefer you to find out if any of the lang_speci
> fic_dr
> > > iver
> > > > > routines actually modify argv (if you already did, then just tell
> me wh
> > > ich
> > > > > one).
> > > >
> > > > gccspec.c(lang_specific_driver) does not, but
> > > > cppspec.c(lang_specific_driver) does. Of course the signature of t
> hese
> > > > two functions must match.
> > >
> > > cppspec.c could probably be fixed to not modify argv
> > Someone can certainly do that, then re-constify argv at that point. Howe
> ver
> > we need to settle on the signature of this stuff since it bleeds into oth
> er
> > front-ends that aren't necessarily in our source tree (like gpc for examp
> le).
>
> Hm, thinking about it a bit more, is there any advantage to having
> argv be const char ** (or const char *const *)? Or to having it not
> be?
I don't believe so at the current time other than the warning police.
jeff