2 weeks to GDB 7.10 tentative branching date!

2 weeks to GDB 7.10 tentative branching date!

[Joel Brobecker]

Hello,

Just a quick reminder that we've set the tentative branching
date for the GDB 7.10 release to June 15th, which is only
a couple of weeks away.

As in previous releases, we have a wiki page where we track all
blocking issues before we can cut the branch:

    https://sourceware.org/gdb/wiki/GDB_7.10_Release

If you know of anything that we should hold the branch for, please
let us know. There are typically regressions, important features
that need a little extra time to be completely integrated in master,
etc.

Thank you!
-- 
Joel

Re: 2 weeks to GDB 7.10 tentative branching date!

[Jan Kratochvil]

On Mon, 01 Jun 2015 23:37:09 +0200, Joel Brobecker wrote:
> As in previous releases, we have a wiki page where we track all
> blocking issues before we can cut the branch:
> 
>     https://sourceware.org/gdb/wiki/GDB_7.10_Release

Here is a list of some recent pending patches of mine.  They do not have filed
PRs and I haven't filed PRs just for the purpose of the Wiki page above.
It find unclear what should / should not be considered a blocker.

[patchv2] compile: Fix crash on cv-qualified self-reference
https://sourceware.org/ml/gdb-patches/2015-05/msg00434.html
Message-ID: <20150516132555.GB17266@host1.jankratochvil.net>

[patchv6] compile: Use also inferior munmap
https://sourceware.org/ml/gdb-patches/2015-05/msg00432.html
Message-ID: <20150516131850.GA16793@host1.jankratochvil.net>

[PATCH v4 1/2] compile: set debug compile: Display GCC driver filename
https://sourceware.org/ml/gdb-patches/2015-05/msg00628.html
Message-ID: <20150524185752.31862.62312.stgit@host1.jankratochvil.net>
[PATCH v5 2/2] compile: Add 'set compile-gcc'
https://sourceware.org/ml/gdb-patches/2015-05/msg00630.html
Message-ID: <20150525185046.GA30823@host1.jankratochvil.net>

[patch] Fix jit-reader.h for multilib
https://sourceware.org/ml/gdb-patches/2015-01/msg00122.html
Message-ID: <20150107173634.GA13295@host2.jankratochvil.net>

more a RFC, unclear what is the right fix:
Re: [patch] Do not close BFDs, breaking deleted inferior shlibs
https://sourceware.org/ml/gdb-patches/2015-02/msg00457.html
Message-ID: <20150217192731.GA24843@host1.jankratochvil.net>


Jan

Re: 2 weeks to GDB 7.10 tentative branching date!

[Phil Muldoon]

On 02/06/15 09:12, Jan Kratochvil wrote:
> On Mon, 01 Jun 2015 23:37:09 +0200, Joel Brobecker wrote:
>> As in previous releases, we have a wiki page where we track all
>> blocking issues before we can cut the branch:
>>
>>     https://sourceware.org/gdb/wiki/GDB_7.10_Release
> Here is a list of some recent pending patches of mine.  They do not have filed
> PRs and I haven't filed PRs just for the purpose of the Wiki page above.
> It find unclear what should / should not be considered a blocker.
>
> [patchv2] compile: Fix crash on cv-qualified self-reference
> https://sourceware.org/ml/gdb-patches/2015-05/msg00434.html
> Message-ID: <20150516132555.GB17266@host1.jankratochvil.net>

I think is a definite blocker.

>
> [patchv6] compile: Use also inferior munmap
> https://sourceware.org/ml/gdb-patches/2015-05/msg00432.html
> Message-ID: <20150516131850.GA16793@host1.jankratochvil.net>
>
> [PATCH v4 1/2] compile: set debug compile: Display GCC driver filename
> https://sourceware.org/ml/gdb-patches/2015-05/msg00628.html
> Message-ID: <20150524185752.31862.62312.stgit@host1.jankratochvil.net>
> [PATCH v5 2/2] compile: Add 'set compile-gcc'
> https://sourceware.org/ml/gdb-patches/2015-05/msg00630.html
> Message-ID: <20150525185046.GA30823@host1.jankratochvil.net>
These would be nice to have, but I cant seem them blocking?

Not concerned of those below, personally.

>
> [patch] Fix jit-reader.h for multilib
> https://sourceware.org/ml/gdb-patches/2015-01/msg00122.html
> Message-ID: <20150107173634.GA13295@host2.jankratochvil.net>
>
> more a RFC, unclear what is the right fix:
> Re: [patch] Do not close BFDs, breaking deleted inferior shlibs
> https://sourceware.org/ml/gdb-patches/2015-02/msg00457.html
> Message-ID: <20150217192731.GA24843@host1.jankratochvil.net>
>
>
> Jan

Re: 2 weeks to GDB 7.10 tentative branching date!

[Yao Qi]

On 02/06/15 09:12, Jan Kratochvil wrote:
> [PATCH v4 1/2] compile: set debug compile: Display GCC driver filename
> https://sourceware.org/ml/gdb-patches/2015-05/msg00628.html
> Message-ID:<20150524185752.31862.62312.stgit@host1.jankratochvil.net>

This was reviewed and approved by Pedro here
https://sourceware.org/ml/gdb-patches/2015-05/msg00711.html

-- 
Yao (齐尧)

Re: 2 weeks to GDB 7.10 tentative branching date!

[Jan Kratochvil]

On Tue, 02 Jun 2015 18:21:12 +0200, Yao Qi wrote:
> On 02/06/15 09:12, Jan Kratochvil wrote:
> >[PATCH v4 1/2] compile: set debug compile: Display GCC driver filename
> This was reviewed and approved by Pedro here
> https://sourceware.org/ml/gdb-patches/2015-05/msg00711.html

Yes, sorry about that - it is still waiting on its GCC part which confused me.


[patchv2] compile: Fix crash on cv-qualified self-reference
https://sourceware.org/ml/gdb-patches/2015-05/msg00434.html
Message-ID: <20150516132555.GB17266@host1.jankratochvil.net>

[patchv6] compile: Use also inferior munmap
https://sourceware.org/ml/gdb-patches/2015-05/msg00432.html
Message-ID: <20150516131850.GA16793@host1.jankratochvil.net>
 - maybe it is already reviewed by Pedro in
   https://sourceware.org/ml/gdb-patches/2015-05/msg00107.html
   but the patch got simplified so it was not approved as is.

[patch] Fix jit-reader.h for multilib
https://sourceware.org/ml/gdb-patches/2015-01/msg00122.html
Message-ID: <20150107173634.GA13295@host2.jankratochvil.net>

Re: [patch] Do not close BFDs, breaking deleted inferior shlibs
https://sourceware.org/ml/gdb-patches/2015-02/msg00457.html
Message-ID: <20150217192731.GA24843@host1.jankratochvil.net>
 - more a RFC, unclear what is the right fix:


Jan

Re: 2 weeks to GDB 7.10 tentative branching date!

[Philippe Waroquiers]

On Mon, 2015-06-01 at 14:37 -0700, Joel Brobecker wrote:
> Hello,
> 
> Just a quick reminder that we've set the tentative branching
> date for the GDB 7.10 release to June 15th, which is only
> a couple of weeks away.
> 
> As in previous releases, we have a wiki page where we track all
> blocking issues before we can cut the branch:
> 
>     https://sourceware.org/gdb/wiki/GDB_7.10_Release
> 
> If you know of anything that we should hold the branch for, please
> let us know. There are typically regressions, important features
> that need a little extra time to be completely integrated in master,
> etc.
> 
> Thank you!
It would be nice to have something that fixes the wrong
executable files being picked up by GDB.
See thread
https://sourceware.org/ml/gdb-patches/2015-05/msg00090.html

For the moment, the behaviour is really strange when detaching
and attaching to another process.
That does not only happen with gdbserver, I think
the same bug also happens with 'simple' attach/detach/attach.

Philippe

Re: 2 weeks to GDB 7.10 tentative branching date!

[Joel Brobecker]

> > [patchv2] compile: Fix crash on cv-qualified self-reference
> > https://sourceware.org/ml/gdb-patches/2015-05/msg00434.html
> > Message-ID: <20150516132555.GB17266@host1.jankratochvil.net>
> 
> I think is a definite blocker.

Agreed.

> > [patchv6] compile: Use also inferior munmap
> > https://sourceware.org/ml/gdb-patches/2015-05/msg00432.html
> > Message-ID: <20150516131850.GA16793@host1.jankratochvil.net>
> >
> > [PATCH v4 1/2] compile: set debug compile: Display GCC driver filename
> > https://sourceware.org/ml/gdb-patches/2015-05/msg00628.html
> > Message-ID: <20150524185752.31862.62312.stgit@host1.jankratochvil.net>
> > [PATCH v5 2/2] compile: Add 'set compile-gcc'
> > https://sourceware.org/ml/gdb-patches/2015-05/msg00630.html
> > Message-ID: <20150525185046.GA30823@host1.jankratochvil.net>
> These would be nice to have, but I cant seem them blocking?
> 
> Not concerned of those below, personally.

Same here, but I think there is a chance that we can backport
the fixes to the branch prior to making the release, as they should
only affect the "compile" feature?

-- 
Joel

Re: 2 weeks to GDB 7.10 tentative branching date!

[Jan Kratochvil]

On Wed, 03 Jun 2015 19:06:10 +0200, Joel Brobecker wrote:
> Same here, but I think there is a chance that we can backport
> the fixes to the branch prior to making the release, as they should
> only affect the "compile" feature?

Yes.


Jan

Re: 2 weeks to GDB 7.10 tentative branching date!

[Joel Brobecker]

> On Wed, 03 Jun 2015 19:06:10 +0200, Joel Brobecker wrote:
> > Same here, but I think there is a chance that we can backport
> > the fixes to the branch prior to making the release, as they should
> > only affect the "compile" feature?
> 
> Yes.

I've "opened" the sections that deal with updates after branch
creation. Can you add the first item in the "before branch creation"
section, and the rest if the "GDB 7.10 Release" section? You can use
the GDB 7.9 release page as a guideline for the format:
https://sourceware.org/gdb/wiki/GDB_7.9_Release

Thanks!

-- 
Joel

Re: 2 weeks to GDB 7.10 tentative branching date!

[Joel Brobecker]

> It would be nice to have something that fixes the wrong
> executable files being picked up by GDB.
> See thread
> https://sourceware.org/ml/gdb-patches/2015-05/msg00090.html
> 
> For the moment, the behaviour is really strange when detaching
> and attaching to another process.
> That does not only happen with gdbserver, I think
> the same bug also happens with 'simple' attach/detach/attach.

FWIW, it would be nice, but given that this is not a regression,
It would be hard justifying the release because of it. At least
not by very much.

What I propose is putting this item on the "Maybe" list, and try
to get the patch in by mid June. If, by then, the patch or the
desired behavior is still under discussion, then re-evaluate our
situation. In particular, if the patch is really close, then wait.
Otherwise, give the patch more time to mature, and cut the branch
without it. Does it sound reasonable?

-- 
Joel

Re: 2 weeks to GDB 7.10 tentative branching date!

[Doug Evans]

On Mon, Jun 1, 2015 at 2:37 PM, Joel Brobecker <brobecker@adacore.com> wrote:
> Hello,
>
> Just a quick reminder that we've set the tentative branching
> date for the GDB 7.10 release to June 15th, which is only
> a couple of weeks away.
>
> As in previous releases, we have a wiki page where we track all
> blocking issues before we can cut the branch:
>
>     https://sourceware.org/gdb/wiki/GDB_7.10_Release
>
> If you know of anything that we should hold the branch for, please
> let us know. There are typically regressions, important features
> that need a little extra time to be completely integrated in master,
> etc.
>
> Thank you!
> --
> Joel

It would be great to resolve pr 16253 for 7.10.
We reverted a fix in the lead up to 7.8, hoping to revisit it
afterwards, but we never did. :-(
[details in the pr]

https://sourceware.org/bugzilla/show_bug.cgi?id=16253

Maybe a two step solution would be ok?

For 7.10:

1) re-add keiths patch, with the following modifications:
   a) add a STRUCT vs VAR_DOMAIN indicator to .gdb_index
   b) bump .gdb_index version to 9
   c) modify keith's patch so that it only is enabled when either
       .gdb_index is not in use or its version >= 9
       (IOW we'd still have to keep symbol_matches_domain,
       in some form, but use it conditionally)

For later:

2) when all producers we care about can produce v9 indices,
    remove hack 1c

I'm aware of the u/i consequences of this approach
(whether examples like in 16253 would work would be
dependent upon whether the index is in use and its version).

Another thought I had is that symbol lookup has changed a
bit since 7.8. It may be possible to just reapply keith's
patch and avoid the massive perf regression differently.

Comments?

GDB Bugzilla Target Milestone 7.10 [Re: 2 weeks to GDB 7.10 tentative branching date!]

[Jan Kratochvil]

On Wed, 03 Jun 2015 19:32:42 +0200, Joel Brobecker wrote:
> Can you add the first item in the "before branch creation" section,

GDB Bugzilla does not have "Target Milestone" = "7.10".


Jan

Re: 2 weeks to GDB 7.10 tentative branching date!

[Philippe Waroquiers]

On Wed, 2015-06-03 at 10:38 -0700, Joel Brobecker wrote:
> FWIW, it would be nice, but given that this is not a regression,
> It would be hard justifying the release because of it. At least
> not by very much.
Effectively, this is a 'would be nice'.

> 
> What I propose is putting this item on the "Maybe" list, and try
> to get the patch in by mid June. If, by then, the patch or the
> desired behavior is still under discussion, then re-evaluate our
> situation. In particular, if the patch is really close, then wait.
> Otherwise, give the patch more time to mature, and cut the branch
> without it. Does it sound reasonable?
So, yes, for sure, what you propose is reasonable.

Philippe

Re: 2 weeks to GDB 7.10 tentative branching date!

[Doug Evans]

On Wed, Jun 3, 2015 at 11:08 AM, Doug Evans <dje@google.com> wrote:
>...
> https://sourceware.org/bugzilla/show_bug.cgi?id=16253
>
> Maybe a two step solution would be ok?
>
> For 7.10:
>
> 1) re-add keiths patch, with the following modifications:
>    a) add a STRUCT vs VAR_DOMAIN indicator to .gdb_index
>    b) bump .gdb_index version to 9
>    c) modify keith's patch so that it only is enabled when either
>        .gdb_index is not in use or its version >= 9
>        (IOW we'd still have to keep symbol_matches_domain,
>        in some form, but use it conditionally)

For reference sake,
the index that is proposed to go into DWARF puts the
symbol's DWARF tag into the index.
That provides more info than what .gdb_index has,
and should be enough to avoid this issue
(I'd still really like to see a gdb implementation of
this proposal before it's too late).

GDB 7.10 new branching date: Jun 29th, 2015 (was: "Re: 2 weeks to GDB 7.10 tentative branching date!")

[Joel Brobecker]

Hello,

We're at the tentative branching date for GDB 7.10, but I think we
should give some patches a little extra time for getting in before
we branch. The problem is that I'm going to be away next week, so
the wait will have to be a couple of weeks.

    >>  New branching date: June 29th.  <<

The GDB 7.10 release wiki page:
https://sourceware.org/gdb/wiki/GDB_7.10_Release

For those of you who have pending patches, please ping the Global
Maintainers who have been following your patches so far. For
the patches with no apparently designated Global Maintainer, please
ping Pedro, Doug, Eli and I (the "FSF-appointed maintainers"), so
we can find someone to help.

Thank you,
-- 
Joel

Re: GDB 7.10 new branching date: Jun 29th, 2015

[Yao Qi]

Joel Brobecker <brobecker@adacore.com> writes:

>     >>  New branching date: June 29th.  <<

Hi Joel,
When will we create the branch?

-- 
Yao (齐尧)

Re: GDB 7.10 new branching date: Jun 29th, 2015

[Joel Brobecker]

> When will we create the branch?

How about today? There was one pending item, but it got moved to
the "maybe" section of the 7.10 release TODO. So we should be good
to go.

I was really hoping that the patch issuing a warning when a discrepancy
between objfile being loaded in memory and actual executable being
debug is detected would make it... I'm wondering where we are at with
this one!

-- 
Joel

Re: GDB 7.10 new branching date: Jun 29th, 2015

[Gary Benson]

Joel Brobecker wrote:
> > When will we create the branch?
> 
> How about today? There was one pending item, but it got moved to
> the "maybe" section of the 7.10 release TODO. So we should be good
> to go.
> 
> I was really hoping that the patch issuing a warning when a
> discrepancy between objfile being loaded in memory and actual
> executable being debug is detected would make it... I'm wondering
> where we are at with this one!

Is that the attach-attach series (make user-specified filenames
sticky)?  I'm sorry, I didn't realize you were waiting for it.
I don't have anything ready to go and I'm on vacation for a week
after today.  I'll see what I can do now...

Gary

-- 
http://gbenson.net/

Re: GDB 7.10 new branching date: Jun 29th, 2015

[Yao Qi]

Joel Brobecker <brobecker@adacore.com> writes:

> How about today? There was one pending item, but it got moved to
> the "maybe" section of the 7.10 release TODO. So we should be good
> to go.

Yes, it is good to create branch soon, as there is no blocking issues.

>
> I was really hoping that the patch issuing a warning when a discrepancy
> between objfile being loaded in memory and actual executable being
> debug is detected would make it... I'm wondering where we are at with
> this one!

Do you have a link about these patches?  According to your description
here, I feel it is a useful feature, but not critical that we must
include it in 7.10 release.

-- 
Yao (齐尧)

Re: GDB 7.10 new branching date: Jun 29th, 2015

[Joel Brobecker]

> > I was really hoping that the patch issuing a warning when a discrepancy
> > between objfile being loaded in memory and actual executable being
> > debug is detected would make it... I'm wondering where we are at with
> > this one!
> 
> Do you have a link about these patches?  According to your description
> here, I feel it is a useful feature, but not critical that we must
> include it in 7.10 release.

Agreed. I will probably create the branch later today.

-- 
Joel