This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH 11/15] Rename functions and make nonstatic as necessary
- From: Gary Benson <gbenson at redhat dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: Siva Chandra <sivachandra at google dot com>, gdb-patches <gdb-patches at sourceware dot org>
- Date: Wed, 18 Jun 2014 15:09:52 +0100
- Subject: Re: [PATCH 11/15] Rename functions and make nonstatic as necessary
- Authentication-results: sourceware.org; auth=none
- References: <1403014378-4349-1-git-send-email-gbenson at redhat dot com> <1403014378-4349-12-git-send-email-gbenson at redhat dot com> <CAGyQ6gwYauHffA+R8yd=LoRSB2VPZdEDy7HsQ0-dt+jqxuhFHg at mail dot gmail dot com> <20140618132832 dot GA6225 at blade dot nx> <53A19A53 dot 6030107 at redhat dot com>
Pedro Alves wrote:
> On 06/18/2014 02:28 PM, Gary Benson wrote:
> > Siva Chandra wrote:
> > > On Tue, Jun 17, 2014 at 7:12 AM, Gary Benson <gbenson@redhat.com> wrote:
> > > > This commit renames the functions that are to be shared.
> > > > Functions to be shared that were static are made nonstatic.
> > >
> > > I thinking making the functions nonstatic in this part was
> > > not correct. The build currently fails when built with
> > > "-Werror=missing-prototypes" as 14/15 has not yet been
> > > committed.
> >
> > My apologies Siva.
> >
> > Pedro, should I commit the remaining piece now, and work on the
> > i386_insert_hw_breakpoint as a separate patch?
>
> As you've committed the header patch already, I think the issue
> is actually already fixed?
Ah, ok, I misunderstood. I can stop rushing!
I actually wish I hadn't committed the header patch now, the
i386_{insert,remove}_hw_breakpoint refactoring you spotted
means those functions didn't need exposing at all.
I'm working on a patch to revert part of that change: remove the
prototypes, make them static again, *and* rename the now static
functions them back to their original names. The last point isn't
strictly necessary, but I don't like that the refactoring would
have renamed some functions it didn't need to. I want to put the
old names back so that people familiar with that code don't see
more disruption than strictly necessary.
I was going to mail it alone (I thought Siva was stuck) but since
he's not I'll mail it as part of a three-patch series to complete
the refactoring.
Thanks,
Gary
--
http://gbenson.net/