3 weeks to GDB 7.9.1 release?

3 weeks to GDB 7.9.1 release?

[Joel Brobecker]

Hello everyone,

Planned date for 7.9.1 re-spin is Wed May, 13th. It appears that
it's been a fairly smooth sailing for the gdb-7.9-branch since
GDB 7.9 was released... Our release wiki page
(https://sourceware.org/gdb/wiki/GDB_7.9_Release) shows that only
one issue was fixed, and we only have 1 that has been identified
as blocking:

   [Doug?] PR18066 argument "word" seems broken in Command.complete
   (text, word)

Anything else that should be seen as blocking for 7.9.1?

This is a good opportunity to remind everyone that patches to
release branches, once the first official release has been made,
must be documented on that page. Otherwise, I will not be able
to provide the full list of improvements that this release brings.
I reviewed the commits on that branch, and identified 2 patches that
have been pushed without following the full procedure, and I will be
emailing those authors privately.  But the branch is polluted by
a number of date-update commits, so if you know you've made a change
and forgot to follow the procedure, it's not too late: You need a PR,
which you can create if you don't have one already, and then just add
a one-line description to the wiki page shown above (that description
can simply be the subject line of the PR).

Thank you,
-- 
Joel

Re: 3 weeks to GDB 7.9.1 release?

[Eli Zaretskii]

> Date: Wed, 22 Apr 2015 06:32:31 -0700
> From: Joel Brobecker <brobecker@adacore.com>
> 
> Planned date for 7.9.1 re-spin is Wed May, 13th. It appears that
> it's been a fairly smooth sailing for the gdb-7.9-branch since
> GDB 7.9 was released... Our release wiki page
> (https://sourceware.org/gdb/wiki/GDB_7.9_Release) shows that only
> one issue was fixed, and we only have 1 that has been identified
> as blocking:
> 
>    [Doug?] PR18066 argument "word" seems broken in Command.complete
>    (text, word)
> 
> Anything else that should be seen as blocking for 7.9.1?

Not really blocking, but would someone please help me get these
annoyances fixed upstream, so that I don't need to continue fixing
them in every GDB release?

  http://lists.gnu.org/archive/html/bug-binutils/2015-03/msg00115.html

TIA

Re: 3 weeks to GDB 7.9.1 release?

[Joel Brobecker]

> Not really blocking, but would someone please help me get these
> annoyances fixed upstream, so that I don't need to continue fixing
> them in every GDB release?
> 
>   http://lists.gnu.org/archive/html/bug-binutils/2015-03/msg00115.html

IIUC (quick read only), you need libiberty patches to be pushed?
libiberty is owned by GCC, and so such pushes should just be pushed
to GCC's SVN first, and then can be backported to binutils-gdb...
Do you have SVN write access to GCC?

-- 
Joel

Re: 3 weeks to GDB 7.9.1 release?

[Eli Zaretskii]

> Date: Wed, 22 Apr 2015 07:31:20 -0700
> From: Joel Brobecker <brobecker@adacore.com>
> Cc: gdb-patches@sourceware.org
> 
> > Not really blocking, but would someone please help me get these
> > annoyances fixed upstream, so that I don't need to continue fixing
> > them in every GDB release?
> > 
> >   http://lists.gnu.org/archive/html/bug-binutils/2015-03/msg00115.html
> 
> IIUC (quick read only), you need libiberty patches to be pushed?
> libiberty is owned by GCC, and so such pushes should just be pushed
> to GCC's SVN first, and then can be backported to binutils-gdb...

Yes, I know.  That message had DJ Delorie CC'ed.

> Do you have SVN write access to GCC?

No.  That's the problem, as the patches were already approved.

RE: 3 weeks to GDB 7.9.1 release?

[Pierre Muller]

  Would the patch to gdb/p-exp.y that fixes issue PR 127815
https://sourceware.org/bugzilla/show_bug.cgi?id=17815

be acceptable without the testsuite change?

  This patch was accepted recently by Doug, and committed to trunk.
https://sourceware.org/ml/gdb-patches/2015-04/msg00735.html

The patch [part 1/3] itself is at:
https://sourceware.org/ml/gdb-patches/2015-02/msg00111.html

Or would the application of the three parts of the patch be more acceptable?
The two other parts are at:
https://sourceware.org/ml/gdb-patches/2015-02/msg00112.html
and
https://sourceware.org/ml/gdb-patches/2015-02/msg00113.html


  I would really like to have it in 7.9.1
as it is completely pascal language specific.

Pierre Muller
as pascal language maintainer

> -----Message d'origine-----
> De : gdb-patches-owner@sourceware.org [mailto:gdb-patches-
> owner@sourceware.org] De la part de Joel Brobecker
> Envoyé : mercredi 22 avril 2015 15:33
> À : gdb-patches@sourceware.org
> Objet : 3 weeks to GDB 7.9.1 release?
> 
> Hello everyone,
> 
> Planned date for 7.9.1 re-spin is Wed May, 13th. It appears that
> it's been a fairly smooth sailing for the gdb-7.9-branch since
> GDB 7.9 was released... Our release wiki page
> (https://sourceware.org/gdb/wiki/GDB_7.9_Release) shows that only
> one issue was fixed, and we only have 1 that has been identified
> as blocking:
> 
>    [Doug?] PR18066 argument "word" seems broken in Command.complete
>    (text, word)
> 
> Anything else that should be seen as blocking for 7.9.1?
> 
> This is a good opportunity to remind everyone that patches to
> release branches, once the first official release has been made,
> must be documented on that page. Otherwise, I will not be able
> to provide the full list of improvements that this release brings.
> I reviewed the commits on that branch, and identified 2 patches that
> have been pushed without following the full procedure, and I will be
> emailing those authors privately.  But the branch is polluted by
> a number of date-update commits, so if you know you've made a change
> and forgot to follow the procedure, it's not too late: You need a PR,
> which you can create if you don't have one already, and then just add
> a one-line description to the wiki page shown above (that description
> can simply be the subject line of the PR).
> 
> Thank you,
> --
> Joel

Re: 3 weeks to GDB 7.9.1 release?

[Joel Brobecker]

>   Would the patch to gdb/p-exp.y that fixes issue PR 127815
> https://sourceware.org/bugzilla/show_bug.cgi?id=17815
> 
> be acceptable without the testsuite change?
> 
>   This patch was accepted recently by Doug, and committed to trunk.
> https://sourceware.org/ml/gdb-patches/2015-04/msg00735.html
> 
> The patch [part 1/3] itself is at:
> https://sourceware.org/ml/gdb-patches/2015-02/msg00111.html
> 
> Or would the application of the three parts of the patch be more acceptable?
> The two other parts are at:
> https://sourceware.org/ml/gdb-patches/2015-02/msg00112.html
> and
> https://sourceware.org/ml/gdb-patches/2015-02/msg00113.html

It's mostly your's and Doug's decision. The general idea is to
only take patches that are safe and fix an important issue. But
if it's only affecting Pascal, then we can be a little more
relaxed about it.

Regarding the testsuite patches, since they cannot harm the debugger,
they should be part of the set being considered for inclusion in
the gdb-7.9-branch.

If the decision is to include the patch, please remember to update
the wiki page as well!

-- 
Joel

Re: 3 weeks to GDB 7.9.1 release?

[Doug Evans]

On Wed, Apr 22, 2015 at 10:14 AM, Joel Brobecker <brobecker@adacore.com> wrote:
>>   Would the patch to gdb/p-exp.y that fixes issue PR 127815
>> https://sourceware.org/bugzilla/show_bug.cgi?id=17815
>>
>> be acceptable without the testsuite change?
>>
>>   This patch was accepted recently by Doug, and committed to trunk.
>> https://sourceware.org/ml/gdb-patches/2015-04/msg00735.html
>>
>> The patch [part 1/3] itself is at:
>> https://sourceware.org/ml/gdb-patches/2015-02/msg00111.html
>>
>> Or would the application of the three parts of the patch be more acceptable?
>> The two other parts are at:
>> https://sourceware.org/ml/gdb-patches/2015-02/msg00112.html
>> and
>> https://sourceware.org/ml/gdb-patches/2015-02/msg00113.html
>
> It's mostly your's and Doug's decision. The general idea is to
> only take patches that are safe and fix an important issue. But
> if it's only affecting Pascal, then we can be a little more
> relaxed about it.

Agreed.
I think this and the tests can go into the branch.

> Regarding the testsuite patches, since they cannot harm the debugger,
> they should be part of the set being considered for inclusion in
> the gdb-7.9-branch.
>
> If the decision is to include the patch, please remember to update
> the wiki page as well!
>
> --
> Joel

RE: 3 weeks to GDB 7.9.1 release?

[Pierre Muller]

 Hi Doug,

thanks for the approval,
I just committed the series to gdb-7.9-branch
https://sourceware.org/ml/gdb-cvs/2015-05/msg00001.html


By the way, my patches in trunk were applied on top of your fix from April 6. 2015.

concerning  lib/pascal.exp 
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commitdiff;h=01b622d4c7536f20622bc0e23ff78b7fe37bc1e3

  It would seem logical to also merge your fix.

  Joel,
you mentioned something about NEWS entry?

Does a fix in 7.9.1 need a NEWS entry in that branch,
even if it has no NEWS entry in trunk?

Pierre Muller



> -----Message d'origine-----
> De : gdb-patches-owner@sourceware.org [mailto:gdb-patches-
> owner@sourceware.org] De la part de Doug Evans
> Envoyé : samedi 2 mai 2015 17:34
> À : Joel Brobecker
> Cc : Pierre Muller; gdb-patches
> Objet : Re: 3 weeks to GDB 7.9.1 release?
> 
> On Wed, Apr 22, 2015 at 10:14 AM, Joel Brobecker
> <brobecker@adacore.com> wrote:
> >>   Would the patch to gdb/p-exp.y that fixes issue PR 127815
> >> https://sourceware.org/bugzilla/show_bug.cgi?id=17815
> >>
> >> be acceptable without the testsuite change?
> >>
> >>   This patch was accepted recently by Doug, and committed to trunk.
> >> https://sourceware.org/ml/gdb-patches/2015-04/msg00735.html
> >>
> >> The patch [part 1/3] itself is at:
> >> https://sourceware.org/ml/gdb-patches/2015-02/msg00111.html
> >>
> >> Or would the application of the three parts of the patch be more
> acceptable?
> >> The two other parts are at:
> >> https://sourceware.org/ml/gdb-patches/2015-02/msg00112.html
> >> and
> >> https://sourceware.org/ml/gdb-patches/2015-02/msg00113.html
> >
> > It's mostly your's and Doug's decision. The general idea is to
> > only take patches that are safe and fix an important issue. But
> > if it's only affecting Pascal, then we can be a little more
> > relaxed about it.
> 
> Agreed.
> I think this and the tests can go into the branch.
> 
> > Regarding the testsuite patches, since they cannot harm the debugger,
> > they should be part of the set being considered for inclusion in
> > the gdb-7.9-branch.
> >
> > If the decision is to include the patch, please remember to update
> > the wiki page as well!
> >
> > --
> > Joel

Re: 3 weeks to GDB 7.9.1 release?

[Joel Brobecker]

>   Joel,
> you mentioned something about NEWS entry?
> 
> Does a fix in 7.9.1 need a NEWS entry in that branch,
> even if it has no NEWS entry in trunk?

It needs to be documented in the release's wiki page:
(https://sourceware.org/gdb/wiki/GDB_7.9_Release - "Fixes in 7.91"
section). If you do not have edit privileges, please send me your
wiki user ID, and I will add it to the authorized list of editors.

Thank you,
-- 
Joel

Re: 3 weeks to GDB 7.9.1 release?

[Sergio Durigan Junior]

On Wednesday, April 22 2015, Joel Brobecker wrote:

> Hello everyone,
>
> Planned date for 7.9.1 re-spin is Wed May, 13th. It appears that
> it's been a fairly smooth sailing for the gdb-7.9-branch since
> GDB 7.9 was released... Our release wiki page
> (https://sourceware.org/gdb/wiki/GDB_7.9_Release) shows that only
> one issue was fixed, and we only have 1 that has been identified
> as blocking:
>
>    [Doug?] PR18066 argument "word" seems broken in Command.complete
>    (text, word)

This bug has been fixed by:

  commit 6d62641c832525382336c1b04731d85cb6c398e7
  Author: Sergio Durigan Junior <sergiodj@redhat.com>
  Date:   Wed Apr 8 18:27:10 2015 -0400

So I closed it and updated the wiki.

On a related note, it doesn't seem to me that this bug should have been
considered as a blocker for this release.  It seems to be a small issue
that shouldn't block a GDB release :-).  Maybe Doug had other reasons to
consider it a blocker?

Cheers,

-- 
Sergio
GPG key ID: 237A 54B1 0287 28BF 00EF  31F4 D0EB 7628 65FC 5E36
Please send encrypted e-mail if possible
http://sergiodj.net/

Re: 3 weeks to GDB 7.9.1 release?

[Doug Evans]

On Fri, May 1, 2015 at 3:23 PM, Sergio Durigan Junior
<sergiodj@redhat.com> wrote:
> On Wednesday, April 22 2015, Joel Brobecker wrote:
>
>> Hello everyone,
>>
>> Planned date for 7.9.1 re-spin is Wed May, 13th. It appears that
>> it's been a fairly smooth sailing for the gdb-7.9-branch since
>> GDB 7.9 was released... Our release wiki page
>> (https://sourceware.org/gdb/wiki/GDB_7.9_Release) shows that only
>> one issue was fixed, and we only have 1 that has been identified
>> as blocking:
>>
>>    [Doug?] PR18066 argument "word" seems broken in Command.complete
>>    (text, word)
>
> This bug has been fixed by:
>
>   commit 6d62641c832525382336c1b04731d85cb6c398e7
>   Author: Sergio Durigan Junior <sergiodj@redhat.com>
>   Date:   Wed Apr 8 18:27:10 2015 -0400
>
> So I closed it and updated the wiki.
>
> On a related note, it doesn't seem to me that this bug should have been
> considered as a blocker for this release.  It seems to be a small issue
> that shouldn't block a GDB release :-).  Maybe Doug had other reasons to
> consider it a blocker?

I think I put it there to not forget about it - I wasn't sure it was a
blocker, and I figured better safe than sorry.

Re: 3 weeks to GDB 7.9.1 release?

[Joel Brobecker]

Hello again,

1 more week to go until the tentative date we set for GDB 7.9.1
release!

There is currently only one issue identified as blocking:

    * [Doug] PR18285 ptype expr-with-xmethod -> SEGV
      (a patch has been pushed to master, so should normally not take
      too long before going in, hopefully)

If there is anything else that you think we should fix before release,
please let us know.

Thank you!
-- 
Joel