[commit] Fix prototype related breakage in go32-nat.c

[commit] Fix prototype related breakage in go32-nat.c

[Eli Zaretskii]

Didn't we agree at the time that the prototypes should be corrected as
well?

Committed.


2009-04-11  Eli Zaretskii  <eliz@gnu.org>

	* go32-nat.c (go32_attach, go32_detach:) Fix prototypes to be
	consistent with the change from 2009-03-17.

Index: gdb/go32-nat.c
===================================================================
RCS file: /cvs/src/src/gdb/go32-nat.c,v
retrieving revision 1.67
diff -u -r1.67 go32-nat.c
--- gdb/go32-nat.c	17 Mar 2009 19:28:08 -0000	1.67
+++ gdb/go32-nat.c	11 Apr 2009 17:55:06 -0000
@@ -167,8 +167,8 @@
 static int prog_has_started = 0;
 static void go32_open (char *name, int from_tty);
 static void go32_close (int quitting);
-static void go32_attach (char *args, int from_tty);
-static void go32_detach (char *args, int from_tty);
+static void go32_attach (struct target_ops *ops, char *args, int from_tty);
+static void go32_detach (struct target_ops *ops, char *args, int from_tty);
 static void go32_resume (struct target_ops *ops,
 			 ptid_t ptid, int step,
 			 enum target_signal siggnal);

Re: [commit] Fix prototype related breakage in go32-nat.c

[Pedro Alves]

On Saturday 11 April 2009 18:59:20, Eli Zaretskii write:
> Didn't we agree at the time that the prototypes should be corrected as
> well?

You're probably confused...  those functions were changed
much prior to that agreement:

 http://sourceware.org/ml/gdb-patches/2008-08/msg00403.html
 http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/go32-nat.c.diff?r1=1.60&r2=1.61&cvsroot=src&f=h

Obviously, it was a simple lapse, one that wouldn't happen if
these functions weren't needlessly forward declared.

(and we agreed to *keep* the prototypes.  Obviously, if they
are to keep, they need to be adjusted, there's no need to agree
on that)

> Committed.
> 
> 
> 2009-04-11  Eli Zaretskii  <eliz@gnu.org>
> 
> 	* go32-nat.c (go32_attach, go32_detach:) Fix prototypes to be

                                              ^ Typo.

> 	consistent with the change from 2009-03-17.

                                        ^^^^^^^^^^
Not that it matter a thing, but, ... wrong date, the change of
the 17th only concerned target_kill.

> 
> Index: gdb/go32-nat.c
> ===================================================================
> RCS file: /cvs/src/src/gdb/go32-nat.c,v
> retrieving revision 1.67
> diff -u -r1.67 go32-nat.c
> --- gdb/go32-nat.c	17 Mar 2009 19:28:08 -0000	1.67
> +++ gdb/go32-nat.c	11 Apr 2009 17:55:06 -0000
> @@ -167,8 +167,8 @@
>  static int prog_has_started = 0;
>  static void go32_open (char *name, int from_tty);
>  static void go32_close (int quitting);
> -static void go32_attach (char *args, int from_tty);
> -static void go32_detach (char *args, int from_tty);
> +static void go32_attach (struct target_ops *ops, char *args, int from_tty);
> +static void go32_detach (struct target_ops *ops, char *args, int from_tty);
>  static void go32_resume (struct target_ops *ops,
>  			 ptid_t ptid, int step,
>  			 enum target_signal siggnal);
> 



-- 
Pedro Alves

Re: [commit] Fix prototype related breakage in go32-nat.c

[Eli Zaretskii]

> From: Pedro Alves <pedro@codesourcery.com>
> Date: Sat, 11 Apr 2009 19:16:09 +0100
> 
> On Saturday 11 April 2009 18:59:20, Eli Zaretskii write:
> > Didn't we agree at the time that the prototypes should be corrected as
> > well?
> 
> You're probably confused...  those functions were changed
> much prior to that agreement:

Sorry.  It's easy to become confused when modified functions are not
mentioned in the logs.

> it was a simple lapse, one that wouldn't happen if these functions
> weren't needlessly forward declared.

We agreed to keep the prototypes, so I'm really surprised that this
issue is brought up again.  What is the purpose of reviving this dead
horse?  As you point out correctly, the problem happened before the
dispute, so why did you need this ``told you so'' reminder?

Btw, it's obvious that no one tried to compile go32-nat.c since then.
How can we be sure we don't break targets if they are not regularly
built?  Didn't Andrew set up such a build at the time?

> (and we agreed to *keep* the prototypes.  Obviously, if they
> are to keep, they need to be adjusted, there's no need to agree
> on that)

We don't need to agree on that, but we do need to keep them adjusted,
and I think it's the job of whoever makes the change to take care of
that.

> > 2009-04-11  Eli Zaretskii  <eliz@gnu.org>
> > 
> > 	* go32-nat.c (go32_attach, go32_detach:) Fix prototypes to be
> 
>                                               ^ Typo.
> 
> > 	consistent with the change from 2009-03-17.
> 
>                                         ^^^^^^^^^^
> Not that it matter a thing, but, ... wrong date, the change of
> the 17th only concerned target_kill.

Thanks, I fixed these.

Re: [commit] Fix prototype related breakage in go32-nat.c

[Pedro Alves]

On Saturday 11 April 2009 20:27:23, Eli Zaretskii wrote:
> > From: Pedro Alves <pedro@codesourcery.com>
> > Date: Sat, 11 Apr 2009 19:16:09 +0100
> > 
> > On Saturday 11 April 2009 18:59:20, Eli Zaretskii write:
> > > Didn't we agree at the time that the prototypes should be corrected as
> > > well?
> > 
> > You're probably confused...  those functions were changed
> > much prior to that agreement:
> 
> Sorry.  It's easy to become confused when modified functions are not
> mentioned in the logs.
> 
> > it was a simple lapse, one that wouldn't happen if these functions
> > weren't needlessly forward declared.
> 
> We agreed to keep the prototypes, so I'm really surprised that this
> issue is brought up again.  What is the purpose of reviving this dead
> horse?  As you point out correctly, the problem happened before the
> dispute, so why did you need this ``told you so'' reminder?

Huh, what's with *that* attitude?  You know *now* that this happened
before the dispute, but you didn't when you posted the patch and
the question "Didn't we agree at the time that the prototypes
should be corrected as well?" --- why did you need to ask it?

> Btw, it's obvious that no one tried to compile go32-nat.c since then.
> How can we be sure we don't break targets if they are not regularly
> built?

Right, last time I built it was on August 2008, which resulted in
the wiki page at <http://sourceware.org/gdb/wiki/BuildingOnDJGPP>.   I
can only assume that people interested in keeping the port alive will be
the people that build it most frequently.  And I assumed that
the port maintainer would build it regularly.

> Didn't Andrew set up such a build at the time?

I have no idea what talking about here.  If you're talking
about Cagney, all his work predates me.

> > (and we agreed to *keep* the prototypes.  Obviously, if they
> > are to keep, they need to be adjusted, there's no need to agree
> > on that)
> 
> We don't need to agree on that, but we do need to keep them adjusted,
> and I think it's the job of whoever makes the change to take care of
> that.

Obviously, we're saying the same thing.  However, typos, lapses and
mistakes do happen.

> Thanks, I fixed these.

Thanks.

-- 
Pedro Alves

Re: [commit] Fix prototype related breakage in go32-nat.c

[Eli Zaretskii]

> From: Pedro Alves <pedro@codesourcery.com>
> Date: Sat, 11 Apr 2009 21:02:57 +0100
> 
> You know *now* that this happened before the dispute, but you didn't
> when you posted the patch and the question "Didn't we agree at the
> time that the prototypes should be corrected as well?" --- why did
> you need to ask it?

Because it seemed to me that the change which failed to update the
prototypes was the subject of that dispute.  But that was due to a
honest mistake, whereas the comment about ``needless forward
declarations'' was not.

> I can only assume that people interested in keeping the port alive
> will be the people that build it most frequently.  And I assumed
> that the port maintainer would build it regularly.

I'm doing my best.  I'm alone.  There's no one to come and help.  It's
not easy, what with my non-existent free time, of which I invest a
large portion into timely reviews of the GDB docs patches (which of
course has priority over taking care of the DJGPP build).

Back in August I thought that the DJGPP port had no users, because I
didn't hear from the DJGPP news group for a long time.  But it turned
out this was because of some snafu between my account on gnu.org and
DJ's server, which DJ fixed a couple of months ago, and since then
I've seen that the forum is alive and kicking, and people are
evidently still using DJGPP ports.  So I'm trying to revive GDB as
well, after I did the same with Emacs.

It's not easy, but I'm doing it.  All I'm asking for is a bit of
sympathy, and a small effort here and there.  Is that too much to ask
from a community to which I contribute every single week?

Re: [commit] Fix prototype related breakage in go32-nat.c

[Eli Zaretskii]

> From: Pedro Alves <pedro@codesourcery.com>
> Date: Sat, 11 Apr 2009 21:02:57 +0100
> 
> <http://sourceware.org/gdb/wiki/BuildingOnDJGPP>

Thanks for posting that page, I forgot about it.

As it happens, I already found the reasons for all the problems you
describe there (and for a few more), and will be posting patches for
them some time next week.

Once the problems are solved in CVS, is there something we should do
with the above page?

Re: [commit] Fix prototype related breakage in go32-nat.c

[Tom Tromey]

>>>>> "Eli" == Eli Zaretskii <eliz@gnu.org> writes:

Eli> It's not easy, but I'm doing it.  All I'm asking for is a bit of
Eli> sympathy, and a small effort here and there.  Is that too much to ask
Eli> from a community to which I contribute every single week?

No, it is not too much to ask.  However, respect is a two-way street.
This patch could easily have gone in without finger-pointing.  I'm
sure whoever neglected to update the prototypes did not do it on
purpose.

Tom

Re: [commit] Fix prototype related breakage in go32-nat.c

[Eli Zaretskii]

> Cc: Pedro Alves <pedro@codesourcery.com>, gdb-patches@sourceware.org
> From: Tom Tromey <tromey@redhat.com>
> Date: Mon, 13 Apr 2009 10:48:39 -0600
> 
> This patch could easily have gone in without finger-pointing.  I'm
> sure whoever neglected to update the prototypes did not do it on
> purpose.

I'm sorry.  I saw the compilation error, recalled the painful argument
about the prototypes, and jumped to the wrong conclusions.

In any case, I didn't mean any finger-pointing, just to express a mild
annoyance that something that was so painfully agreed upon didn't
happen.  As you might imagine, it didn't take more than a few seconds
to fix the problem, so the annoyance couldn't last too long.