2 weeks away from potential GDB 7.4.1 release...

2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

Hello,

It's been almost 3 months since the 7.4 release, and it's time to think
about whether we want to make a corrective release or not. The tentative
date is Apr 23rd. We have a few changes already in the branch, are the
problems serious enough to warrant a release? Any other changes that
should go in the 7.4 branch (please, only refer to very very safe fixes,
or fixes to critical bugs)?

Thank you,
-- 
Joel

Re: 2 weeks away from potential GDB 7.4.1 release...

[Paul Smith]

On Thu, 2012-04-12 at 08:33 -0700, Joel Brobecker wrote:
> It's been almost 3 months since the 7.4 release, and it's time to think
> about whether we want to make a corrective release or not. The tentative
> date is Apr 23rd. We have a few changes already in the branch, are the
> problems serious enough to warrant a release? Any other changes that
> should go in the 7.4 branch (please, only refer to very very safe fixes,
> or fixes to critical bugs)?

I would love to see a fix for the issue of the build failing if you
don't have makeinfo installed... I think it's already fixed but I don't
know if it was applied to the branch.  I can look up the bug # if
needed.

I have GDB embedded in a complete cross-compiler build suite and this
failure means I have to configure the build servers specially to install
makeinfo (normally it's not installed).  So it's a small thing, in
general, but it has a disproportionately large impact on my personal
setup.

Cheers!

-- 
-------------------------------------------------------------------------------
 Paul D. Smith <psmith@gnu.org>          Find some GNU make tips at:
 http://www.gnu.org                      http://make.mad-scientist.net
 "Please remain calm...I may be mad, but I am a professional." --Mad Scientist

Re: 2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

> I would love to see a fix for the issue of the build failing if you
> don't have makeinfo installed... I think it's already fixed but I don't
> know if it was applied to the branch.  I can look up the bug # if
> needed.

I am not opposed to this patch in principle, but a taking a look at
the patch would allow us to evaluate its impact. Can you repost or
send a link to that patch?

In the meantime, it seems much easier to just apply the patch locally
to your sources rather than wait for 7.4.1 or 7.5...

-- 
Joel

Re: 2 weeks away from potential GDB 7.4.1 release...

[Mike Frysinger]

[-- Attachment #1: Type: Text/Plain, Size: 777 bytes --]

On Thursday 12 April 2012 11:33:01 Joel Brobecker wrote:
> It's been almost 3 months since the 7.4 release, and it's time to think
> about whether we want to make a corrective release or not. The tentative
> date is Apr 23rd. We have a few changes already in the branch, are the
> problems serious enough to warrant a release? Any other changes that
> should go in the 7.4 branch (please, only refer to very very safe fixes,
> or fixes to critical bugs)?

there's a sim patch i think would be useful on the branch (fixes a build issue 
reported by someone shortly after the 7.4 release).  i haven't committed to a 
branch on sourceware.org before, so is there anything i should be aware of ?  
a quick search on the gdb site didn't seem to turn up anything.
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: 2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

> there's a sim patch i think would be useful on the branch (fixes a
> build issue reported by someone shortly after the 7.4 release).  i
> haven't committed to a branch on sourceware.org before, so is there
> anything i should be aware of ?  a quick search on the gdb site didn't
> seem to turn up anything.

It's the same as on HEAD. Just checkout the branch, merge the patch,
and send an email to gdb-patches to confirm the commit.

-- 
Joel

Re: 2 weeks away from potential GDB 7.4.1 release...

[Paul Smith]

On Thu, 2012-04-12 at 10:17 -0700, Joel Brobecker wrote:
> I am not opposed to this patch in principle, but a taking a look at
> the patch would allow us to evaluate its impact. Can you repost or
> send a link to that patch?

Looks like it's already fixed in the trunk:

http://sourceware.org/bugzilla/show_bug.cgi?id=13861

> In the meantime, it seems much easier to just apply the patch locally
> to your sources rather than wait for 7.4.1 or 7.5...

Well, I thought of that but the patch is to configure.ac so after I
applied it I would have to re-run autoconf to get a new configure
script... the cure is worse than the disease!! :-)

Re: 2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

> Looks like it's already fixed in the trunk:
> 
> http://sourceware.org/bugzilla/show_bug.cgi?id=13861

And following the thread, it looks it was also fixed on the 7.4
branch. Can you give one of the 7.4.x snapshots a shot to confirm it?

for instance @ ftp://sourceware.org/pub/gdb/snapshots/branch,
you have gdb-7.4.0.20120412.tar.bz2

-- 
Joel

RE: 2 weeks away from potential GDB 7.4.1 release...

[Pierre Muller]

We I committed a patch about a pascal regression

http://sourceware.org/ml/gdb-patches/2012-03/msg00599.html

http://sourceware.org/ml/gdb-cvs/2012-03/msg00205.html

I asked if I should merge it to fixes...

So I suppose that now is the time...
Is this OK?


ChangeLog entry:

2012-04-13  Pierre Muller  <muller@ics.u-strasbg.fr>

        * p-typeprint.c (pascal_type_print_method_args):
        Fix display of parameter of methods.


Index: p-typeprint.c
===================================================================
RCS file: /cvs/src/src/gdb/p-typeprint.c,v
retrieving revision 1.42.2.1
diff -c -r1.42.2.1 p-typeprint.c
*** p-typeprint.c       6 Jan 2012 04:43:21 -0000       1.42.2.1
--- p-typeprint.c       13 Apr 2012 08:04:10 -0000
***************
*** 183,190 ****
          physname += len;

          for (j = 0; j < i; ++j)
!           fputc_filtered (physname[i], stream);
!         fputs_filtered (physname, stream);

          physname += i;
          if (physname[0] != 0)
--- 183,189 ----
          physname += len;

          for (j = 0; j < i; ++j)
!           fputc_filtered (physname[j], stream);

          physname += i;
          if (physname[0] != 0)

Re: 2 weeks away from potential GDB 7.4.1 release...

[Jan Kratochvil]

On Thu, 12 Apr 2012 21:13:53 +0200, Joel Brobecker wrote:
> > Looks like it's already fixed in the trunk:
> > 
> > http://sourceware.org/bugzilla/show_bug.cgi?id=13861
> 
> And following the thread, it looks it was also fixed on the 7.4
> branch.

yes:

commit 983188c2aa5ac0f10fe3ffb8b3d326f16deb47d3
Author: Jan Kratochvil <jan.kratochvil@redhat.com>
Date:   Tue Feb 21 22:07:39 2012 +0000

    gdb/
        PR build/13638
        * configure.ac (MAKEINFO): Replace AC_CHECK_PROG by AC_ARG_VAR.
        (MAKEINFOFLAGS): Replace static string by AC_ARG_VAR.
        * configure: Regenerate.

git name-rev 983188c2aa5ac0f10fe3ffb8b3d326f16deb47d3
983188c2aa5ac0f10fe3ffb8b3d326f16deb47d3 gdb_7_4-branch~1

Sorry for the former regression.


Jan

Re: 2 weeks away from potential GDB 7.4.1 release...

[Jan Kratochvil]

Hello Joel,

On Thu, 12 Apr 2012 17:33:01 +0200, Joel Brobecker wrote:
> Any other changes that should go in the 7.4 branch (please, only refer to
> very very safe fixes, or fixes to critical bugs)?

there is still unfinished but it should make it in time.
	[patch+7.4] Fix gdbserver qXfer:libraries-svr4 regression in special cases
	http://sourceware.org/ml/gdb-patches/2012-03/msg00120.html
	http://sourceware.org/ml/gdb-patches/2012-04/msg00192.html

Still maybe there should be an evaluation of each commit and their backporting
like what happens with GCC stable branches.  I admit I do not do it even in
downstream/Fedora releases.


Thanks,
Jan

Re: 2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

> there is still unfinished but it should make it in time.
> 	[patch+7.4] Fix gdbserver qXfer:libraries-svr4 regression in special cases
> 	http://sourceware.org/ml/gdb-patches/2012-03/msg00120.html
> 	http://sourceware.org/ml/gdb-patches/2012-04/msg00192.html

No objection. If you guys think it is a safe patch, it is good enough
for me, since it fixes a regression.

> Still maybe there should be an evaluation of each commit and their
> backporting like what happens with GCC stable branches.  I admit I do
> not do it even in downstream/Fedora releases.

That would require a lot of time to complete this effort, IMO.
Anyone wanting to do it is welcome to do so, of course; but I prefer
the distributed approach where we try to remember about the branch
when posting/reviewing patches, and decide there and then whether
to backport it or not. And then ask about the few ones that we might
have missed using the 2-weeks-to-release heads up email.

-- 
Joel

Re: 2 weeks away from potential GDB 7.4.1 release...

[Joel Brobecker]

Hi Jan,

> there is still unfinished but it should make it in time.
> 	[patch+7.4] Fix gdbserver qXfer:libraries-svr4 regression in special cases
> 	http://sourceware.org/ml/gdb-patches/2012-03/msg00120.html
> 	http://sourceware.org/ml/gdb-patches/2012-04/msg00192.html

How are we doing on these? Should we delay the 7.4.1 release a bit?

-- 
Joel

Re: 2 weeks away from potential GDB 7.4.1 release...

[Jan Kratochvil]

On Thu, 19 Apr 2012 17:28:19 +0200, Joel Brobecker wrote:
> > there is still unfinished but it should make it in time.
> > 	[patch+7.4] Fix gdbserver qXfer:libraries-svr4 regression in special cases
> > 	http://sourceware.org/ml/gdb-patches/2012-03/msg00120.html
> > 	http://sourceware.org/ml/gdb-patches/2012-04/msg00192.html
> 
> How are we doing on these? Should we delay the 7.4.1 release a bit?

I am busy now with non-trunk work, I still may fix the patch above in time but
I do not consider it a blocker for 7.4.1.


Thanks,
Jan