GDB 7.10 release 2015-07-17 status update?

GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

Hello,

It's been almost 2 weeks since we branched, and there has been a number
of issues fixed since then.

On the wiki (https://sourceware.org/gdb/wiki/GDB_7.10_Release),
we have 1 TODO item left:

  - Eval with KeithS regressions due to GCC 5.1 (regressions caused by
    GCC rather than GDB)

    For this one, the latest (private) update I got from Keith was
    that GDB was likely in the clear - mostly GCC issues? Keith?

We still have 2 "Maybe"'s:

  - [GCC review/Jan] PR compile/18485 (set debug compile: Display GCC driver
    filename)
  - [GCC review/Jan] PR compile/18486 (Add 'set compile-gcc')

  If the reason for it being a "Maybe" is time, we can probably give you
  a little extra time (see below).

I've moved the following item to the Exclude list, as per Pedro's
recommendation:

  - PR gdb/18600 (After forking and threads spawning, gdb leaves newly
    created threads stopped) (too delicate)

David Edelson reported that the buildBot was reporting a lot of
failures around the time we cut the branch. Has anyone looked at those
since then? When I did, I was looking at the waterfall view, and
several builds looked like they were failing right at the beginning,
so I didn't know what to make of it. Also, that's a lot of info to
grok, so I didn't have time to look further.

What I can say is that I've tested the branch with a number of
platforms, not as exhaustively of course, but so far, it looks
pretty good.

What I think we should do is give it another couple of weeks, and
see if other reports come up. It would be good, also, if we had
some info about the buildBOT failures, and whether the release
might be affected by any regression it found.

So, I propose a re-sync on Aug 3rd.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Luis Machado]

Hi Joel,

On 07/17/2015 02:53 PM, Joel Brobecker wrote:
> Hello,
>
> It's been almost 2 weeks since we branched, and there has been a number
> of issues fixed since then.
>
> On the wiki (https://sourceware.org/gdb/wiki/GDB_7.10_Release),
> we have 1 TODO item left:
>
>    - Eval with KeithS regressions due to GCC 5.1 (regressions caused by
>      GCC rather than GDB)
>
>      For this one, the latest (private) update I got from Keith was
>      that GDB was likely in the clear - mostly GCC issues? Keith?
>
> We still have 2 "Maybe"'s:
>
>    - [GCC review/Jan] PR compile/18485 (set debug compile: Display GCC driver
>      filename)
>    - [GCC review/Jan] PR compile/18486 (Add 'set compile-gcc')
>
>    If the reason for it being a "Maybe" is time, we can probably give you
>    a little extra time (see below).
>
> I've moved the following item to the Exclude list, as per Pedro's
> recommendation:
>
>    - PR gdb/18600 (After forking and threads spawning, gdb leaves newly
>      created threads stopped) (too delicate)
>
> David Edelson reported that the buildBot was reporting a lot of
> failures around the time we cut the branch. Has anyone looked at those
> since then? When I did, I was looking at the waterfall view, and
> several builds looked like they were failing right at the beginning,
> so I didn't know what to make of it. Also, that's a lot of info to
> grok, so I didn't have time to look further.
>
> What I can say is that I've tested the branch with a number of
> platforms, not as exhaustively of course, but so far, it looks
> pretty good.
>
> What I think we should do is give it another couple of weeks, and
> see if other reports come up. It would be good, also, if we had
> some info about the buildBOT failures, and whether the release
> might be affected by any regression it found.
>
> So, I propose a re-sync on Aug 3rd.
>

Do we consider the fact that mingw32 (and maybe other non-GNU) builds 
have been broken since this libiberty sync a blocker/critical issue?

https://sourceware.org/ml/gdb-patches/2015-06/msg00405.html

Last we heard of it was here...

https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html

But we received no word from Yunlian Jiang yet.

Maybe it should be added to the wiki too?

Re: GDB 7.10 release 2015-07-17 status update?

[Keith Seitz]

On 07/17/2015 10:53 AM, Joel Brobecker wrote:
>   - Eval with KeithS regressions due to GCC 5.1 (regressions caused by
>     GCC rather than GDB)
> 
>     For this one, the latest (private) update I got from Keith was
>     that GDB was likely in the clear - mostly GCC issues? Keith?

That is correct. If there are gdb issues, it is merely some regexps in
the test suite that need updating.

Keith

Re: GDB 7.10 release 2015-07-17 status update?

[Jan Kratochvil]

On Fri, 17 Jul 2015 19:53:28 +0200, Joel Brobecker wrote:
> We still have 2 "Maybe"'s:
> 
>   - [GCC review/Jan] PR compile/18485 (set debug compile: Display GCC driver
>     filename)
>   - [GCC review/Jan] PR compile/18486 (Add 'set compile-gcc')
> 
>   If the reason for it being a "Maybe" is time, we can probably give you
>   a little extra time (see below).

I have pinged those now:
	ping: [PATCH v3 1/4] libcc1: Introduce GCC_FE_VERSION_1
	https://gcc.gnu.org/ml/gcc-patches/2015-07/msg01563.html

But I do not think it matters much, there are other more important parts of
the 'compile' project.

The release really should not be delayed in any way for it.
It could be also already rejected that 7.11 will be enough for it.


Jan

Re: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> Do we consider the fact that mingw32 (and maybe other non-GNU) builds have
> been broken since this libiberty sync a blocker/critical issue?
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00405.html
> 
> Last we heard of it was here...
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
> 
> But we received no word from Yunlian Jiang yet.
> 
> Maybe it should be added to the wiki too?

It's strange, it builds fine for me on Windows/MinGW - but perhaps
the problem doesn't happen if you use MinGW64, which is what I think
my compiler provides.

But, back to your question, very definitely. If we can't even build
on Windows, we should at least decide what to do next. I haven't seen
the error; do you happen to have a link handy? Do you happen to have
a link?

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> >   - [GCC review/Jan] PR compile/18485 (set debug compile: Display GCC driver
> >     filename)
> >   - [GCC review/Jan] PR compile/18486 (Add 'set compile-gcc')
> > 
> >   If the reason for it being a "Maybe" is time, we can probably give you
> >   a little extra time (see below).
> 
> I have pinged those now:
> 	ping: [PATCH v3 1/4] libcc1: Introduce GCC_FE_VERSION_1
> 	https://gcc.gnu.org/ml/gcc-patches/2015-07/msg01563.html
> 
> But I do not think it matters much, there are other more important parts of
> the 'compile' project.
> 
> The release really should not be delayed in any way for it.
> It could be also already rejected that 7.11 will be enough for it.

OK, in that case, let's just delete those entries from the wiki,
and if you get the patches in master in time, you can always ask
for the patches to also be ported to 7.10.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Luis Machado]

On 07/17/2015 03:32 PM, Joel Brobecker wrote:
>> Do we consider the fact that mingw32 (and maybe other non-GNU) builds have
>> been broken since this libiberty sync a blocker/critical issue?
>>
>> https://sourceware.org/ml/gdb-patches/2015-06/msg00405.html
>>
>> Last we heard of it was here...
>>
>> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
>>
>> But we received no word from Yunlian Jiang yet.
>>
>> Maybe it should be added to the wiki too?
>
> It's strange, it builds fine for me on Windows/MinGW - but perhaps
> the problem doesn't happen if you use MinGW64, which is what I think
> my compiler provides.
>
> But, back to your question, very definitely. If we can't even build
> on Windows, we should at least decide what to do next. I haven't seen
> the error; do you happen to have a link handy? Do you happen to have
> a link?
>

Jan-Benedict posted one example here:

https://sourceware.org/ml/gdb-patches/2015-06/msg00650.html

I see something similar related to the implicit declaration of asprintf.

Re: GDB 7.10 release 2015-07-17 status update?

[Eli Zaretskii]

> Date: Fri, 17 Jul 2015 15:14:47 -0300
> From: Luis Machado <lgustavo@codesourcery.com>
> CC: <keiths@redhat.com>, Jan Kratochvil <jan.kratochvil@redhat.com>,	<yunlian@google.com>
> 
> Do we consider the fact that mingw32 (and maybe other non-GNU) builds 
> have been broken since this libiberty sync a blocker/critical issue?

It's not broken: I've built the pretest with that problem in it, and
it does build, albeit with a (harmless) warning.

> Last we heard of it was here...
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html

The release tarball builds without -Werror, so this warning is just a
warning, since asprintf does return an int.

Re: GDB 7.10 release 2015-07-17 status update?

[Eli Zaretskii]

> Date: Fri, 17 Jul 2015 11:32:46 -0700
> From: Joel Brobecker <brobecker@adacore.com>
> Cc: gdb-patches@sourceware.org, keiths@redhat.com,	Jan Kratochvil <jan.kratochvil@redhat.com>, yunlian@google.com
> 
> > Do we consider the fact that mingw32 (and maybe other non-GNU) builds have
> > been broken since this libiberty sync a blocker/critical issue?
> > 
> > https://sourceware.org/ml/gdb-patches/2015-06/msg00405.html
> > 
> > Last we heard of it was here...
> > 
> > https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
> > 
> > But we received no word from Yunlian Jiang yet.
> > 
> > Maybe it should be added to the wiki too?
> 
> It's strange, it builds fine for me on Windows/MinGW - but perhaps
> the problem doesn't happen if you use MinGW64, which is what I think
> my compiler provides.

Are you sure you didn't see a warning about asprintf while building
bfd?  It's easy to miss that.

Re: GDB 7.10 release 2015-07-17 status update?

[Eli Zaretskii]

> Date: Fri, 17 Jul 2015 15:37:59 -0300
> From: Luis Machado <lgustavo@codesourcery.com>
> CC: <gdb-patches@sourceware.org>, <keiths@redhat.com>, Jan Kratochvil	<jan.kratochvil@redhat.com>, <yunlian@google.com>
> 
> > But, back to your question, very definitely. If we can't even build
> > on Windows, we should at least decide what to do next. I haven't seen
> > the error; do you happen to have a link handy? Do you happen to have
> > a link?
> >
> 
> Jan-Benedict posted one example here:
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00650.html
> 
> I see something similar related to the implicit declaration of asprintf.

I see it also, but it's just a warning that doesn't affect the
correctness of the code.

It would be nice to fix that before the relase, of course.

Re: GDB 7.10 release 2015-07-17 status update?

[Simon Marchi]

On 15-07-17 01:53 PM, Joel Brobecker wrote:
> I've moved the following item to the Exclude list, as per Pedro's
> recommendation:
> 
>   - PR gdb/18600 (After forking and threads spawning, gdb leaves newly
>     created threads stopped) (too delicate)

Have I missed something? Pedro said that about changing the behavior of
when the prompt returns when an inferior exits (see [1]).  I don't recall
him saying that about the fix for 18600 though.  I was actually planning
early next week to formalize and submit the patches he sent in that thread,
hoping they could find their way into 7.10.

[1] https://sourceware.org/ml/gdb-patches/2015-07/msg00384.html

Re: GDB 7.10 release 2015-07-17 status update?

[Sergio Durigan Junior]

On Friday, July 17 2015, Joel Brobecker wrote:

> David Edelson reported that the buildBot was reporting a lot of
> failures around the time we cut the branch. Has anyone looked at those
> since then? When I did, I was looking at the waterfall view, and
> several builds looked like they were failing right at the beginning,
> so I didn't know what to make of it. Also, that's a lot of info to
> grok, so I didn't have time to look further.
>
> What I can say is that I've tested the branch with a number of
> platforms, not as exhaustively of course, but so far, it looks
> pretty good.

Heya,

As has been explained before, our testsuite contains some racy tests,
and therefore our BuildBot is just spilling lots of FAIL's from them.  I
have been trying to work on a solution to help address this problem;
meanwhile, other are looking at the logs (mostly Yao, I think) and
reporting whenever they see something there.

> What I think we should do is give it another couple of weeks, and
> see if other reports come up. It would be good, also, if we had
> some info about the buildBOT failures, and whether the release
> might be affected by any regression it found.

Yeah, I agree.  I'll investigate the logs and see if they contain
something suspicious.  Unfortunately, due to the recent problem
(reported at
<https://sourceware.org/ml/gdb-patches/2015-07/msg00506.html>) the build
logs are not available anymore for a number of builds...  But the
BuildBot logs (i.e., the e-mails it sends to gdb-testers -- those have
not been affected) should be enough to determine if there's something
strange.

Thanks,

-- 
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: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> Jan-Benedict posted one example here:
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00650.html
> 
> I see something similar related to the implicit declaration of asprintf.

I see. It would be nice to fix those, but those are not fatal
for the release, because the default on the release branch is
to turn -Werror off; so users will get the warning, but it will
not be fatal unless they configure with --with-werror.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Luis Machado]

On 07/17/2015 05:48 PM, Joel Brobecker wrote:
>> Jan-Benedict posted one example here:
>>
>> https://sourceware.org/ml/gdb-patches/2015-06/msg00650.html
>>
>> I see something similar related to the implicit declaration of asprintf.
>
> I see. It would be nice to fix those, but those are not fatal
> for the release, because the default on the release branch is
> to turn -Werror off; so users will get the warning, but it will
> not be fatal unless they configure with --with-werror.
>

Fair enough. I'm ok with that.

Re: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> > I've moved the following item to the Exclude list, as per Pedro's
> > recommendation:
> > 
> >   - PR gdb/18600 (After forking and threads spawning, gdb leaves newly
> >     created threads stopped) (too delicate)
> 
> Have I missed something? Pedro said that about changing the behavior of
> when the prompt returns when an inferior exits (see [1]).  I don't recall
> him saying that about the fix for 18600 though.  I was actually planning
> early next week to formalize and submit the patches he sent in that thread,
> hoping they could find their way into 7.10.
> 
> [1] https://sourceware.org/ml/gdb-patches/2015-07/msg00384.html

OK, I must have gotten confused and did not realize there were two
issues being discussed. I've put the item back on the todo.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Simon Marchi]

On 15-07-17 04:57 PM, Joel Brobecker wrote:
> OK, I must have gotten confused and did not realize there were two
> issues being discussed. I've put the item back on the todo.

Ok thanks!  I'll try to send it early next week, so that it's not too
last minute.

Simon

Re: GDB 7.10 release 2015-07-17 status update?

[Yao Qi]

Joel Brobecker <brobecker@adacore.com> writes:

> David Edelson reported that the buildBot was reporting a lot of
> failures around the time we cut the branch. Has anyone looked at those
> since then? When I did, I was looking at the waterfall view, and
> several builds looked like they were failing right at the beginning,
> so I didn't know what to make of it. Also, that's a lot of info to
> grok, so I didn't have time to look further.

Yes, I looked at them last week.  I can't reproduce most of them on my
machine.  For fails I can reproduce, I identified the patches related to
these fails, and ping the authors.

  - fails in gdb.arch/i386-biarch-core.exp are fixed by Jan
  https://sourceware.org/ml/gdb-patches/2015-07/msg00476.html

  - GDBserver crashes when killing a multi-thread process.  Remind Pedro
  to commit his patch.
  https://sourceware.org/ml/gdb-patches/2015-03/msg00182.html

  - fails in attach-pie-noexec.exp fail on native-extended-gdbserver are
    fixed by me.
  https://sourceware.org/ml/gdb-patches/2015-07/msg00367.html

  - fails in gdb.base/gnu_vector.exp are being discussed.
  https://sourceware.org/ml/gdb-patches/2015-07/msg00315.html
  These fails are not fixed, but they are test case problem rather than
  GDB's problem.  I still don't have a chance to think about how to make
  different architectures/os happy for this test case.  IMO, these fails
  shouldn't block 7.10 release.

  - fails in gdb.server/wrapper.exp
  I am working on a patch series to fix these fails, and probably I'll
  send them out today or tomorrow.
  I'd like to fix it before 7.10 release.

-- 
Yao (齐尧)

Re: GDB 7.10 release 2015-07-17 status update?

[Iain Buclaw]

On 17 July 2015 at 20:37, Luis Machado <lgustavo@codesourcery.com> wrote:
> On 07/17/2015 03:32 PM, Joel Brobecker wrote:
>>>
>>> Do we consider the fact that mingw32 (and maybe other non-GNU) builds
>>> have
>>> been broken since this libiberty sync a blocker/critical issue?
>>>
>>> https://sourceware.org/ml/gdb-patches/2015-06/msg00405.html
>>>
>>> Last we heard of it was here...
>>>
>>> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
>>>
>>> But we received no word from Yunlian Jiang yet.
>>>
>>> Maybe it should be added to the wiki too?
>>
>>
>> It's strange, it builds fine for me on Windows/MinGW - but perhaps
>> the problem doesn't happen if you use MinGW64, which is what I think
>> my compiler provides.
>>
>> But, back to your question, very definitely. If we can't even build
>> on Windows, we should at least decide what to do next. I haven't seen
>> the error; do you happen to have a link handy? Do you happen to have
>> a link?
>>
>
> Jan-Benedict posted one example here:
>
> https://sourceware.org/ml/gdb-patches/2015-06/msg00650.html
>
> I see something similar related to the implicit declaration of asprintf.

I know that Pedro suggested fixing up bfd's configure.ac, has this
been tried out?

https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html

I can send in a patch, but admittedly I'll be making the change blind.

Re: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> Yes, I looked at them last week.  I can't reproduce most of them on my
> machine.  For fails I can reproduce, I identified the patches related to
> these fails, and ping the authors.
> 
>   - fails in gdb.arch/i386-biarch-core.exp are fixed by Jan
>   https://sourceware.org/ml/gdb-patches/2015-07/msg00476.html
> 
>   - GDBserver crashes when killing a multi-thread process.  Remind Pedro
>   to commit his patch.
>   https://sourceware.org/ml/gdb-patches/2015-03/msg00182.html
> 
>   - fails in attach-pie-noexec.exp fail on native-extended-gdbserver are
>     fixed by me.
>   https://sourceware.org/ml/gdb-patches/2015-07/msg00367.html
> 
>   - fails in gdb.base/gnu_vector.exp are being discussed.
>   https://sourceware.org/ml/gdb-patches/2015-07/msg00315.html
>   These fails are not fixed, but they are test case problem rather than
>   GDB's problem.  I still don't have a chance to think about how to make
>   different architectures/os happy for this test case.  IMO, these fails
>   shouldn't block 7.10 release.
> 
>   - fails in gdb.server/wrapper.exp
>   I am working on a patch series to fix these fails, and probably I'll
>   send them out today or tomorrow.
>   I'd like to fix it before 7.10 release.

Thanks for looking into those, Yao.

The tentative date for relase is early Aug, so just let me know if
you need a little more time.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Joel Brobecker]

> I know that Pedro suggested fixing up bfd's configure.ac, has this
> been tried out?
> 
> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
> 
> I can send in a patch, but admittedly I'll be making the change blind.

I've been trying to reproduce, to test your patch, but I've been hitting
all sort of weird issues that have nothing to do with this and now I am
running out of time :-(.

It would be good to get this fixed on master, but as said previously,
it should not be critical for the gdb-7.10-branch.

-- 
Joel

Re: GDB 7.10 release 2015-07-17 status update?

[Pedro Alves]

On 07/17/2015 08:03 PM, Eli Zaretskii wrote:
>> Date: Fri, 17 Jul 2015 15:14:47 -0300
>> From: Luis Machado <lgustavo@codesourcery.com>
>> CC: <keiths@redhat.com>, Jan Kratochvil <jan.kratochvil@redhat.com>,	<yunlian@google.com>
>>
>> Do we consider the fact that mingw32 (and maybe other non-GNU) builds 
>> have been broken since this libiberty sync a blocker/critical issue?
> 
> It's not broken: I've built the pretest with that problem in it, and
> it does build, albeit with a (harmless) warning.
> 
>> Last we heard of it was here...
>>
>> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
> 
> The release tarball builds without -Werror, so this warning is just a
> warning, since asprintf does return an int.

It's also a varargs function though; varargs functions
must have a visible prototype.  I wouldn't be surprised if this
resulted in a crash/corruption on some targets/abis.

ISTR seeing a patch for this, but I can't find it now.

Did I just imagine it?

Thanks,
Pedro Alves

Re: GDB 7.10 release 2015-07-17 status update?

[Iain Buclaw]

On 24 July 2015 at 11:26, Pedro Alves <palves@redhat.com> wrote:
> On 07/17/2015 08:03 PM, Eli Zaretskii wrote:
>>> Date: Fri, 17 Jul 2015 15:14:47 -0300
>>> From: Luis Machado <lgustavo@codesourcery.com>
>>> CC: <keiths@redhat.com>, Jan Kratochvil <jan.kratochvil@redhat.com>, <yunlian@google.com>
>>>
>>> Do we consider the fact that mingw32 (and maybe other non-GNU) builds
>>> have been broken since this libiberty sync a blocker/critical issue?
>>
>> It's not broken: I've built the pretest with that problem in it, and
>> it does build, albeit with a (harmless) warning.
>>
>>> Last we heard of it was here...
>>>
>>> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
>>
>> The release tarball builds without -Werror, so this warning is just a
>> warning, since asprintf does return an int.
>
> It's also a varargs function though; varargs functions
> must have a visible prototype.  I wouldn't be surprised if this
> resulted in a crash/corruption on some targets/abis.
>
> ISTR seeing a patch for this, but I can't find it now.
>
> Did I just imagine it?
>

Do you mean: https://sourceware.org/ml/gdb-patches/2015-07/msg00570.html

Iain

Re: GDB 7.10 release 2015-07-17 status update?

[Pedro Alves]

On 07/24/2015 03:30 PM, Iain Buclaw wrote:
> On 24 July 2015 at 11:26, Pedro Alves <palves@redhat.com> wrote:
>> On 07/17/2015 08:03 PM, Eli Zaretskii wrote:
>>>> Date: Fri, 17 Jul 2015 15:14:47 -0300
>>>> From: Luis Machado <lgustavo@codesourcery.com>
>>>> CC: <keiths@redhat.com>, Jan Kratochvil <jan.kratochvil@redhat.com>, <yunlian@google.com>
>>>>
>>>> Do we consider the fact that mingw32 (and maybe other non-GNU) builds
>>>> have been broken since this libiberty sync a blocker/critical issue?
>>>
>>> It's not broken: I've built the pretest with that problem in it, and
>>> it does build, albeit with a (harmless) warning.
>>>
>>>> Last we heard of it was here...
>>>>
>>>> https://sourceware.org/ml/gdb-patches/2015-06/msg00654.html
>>>
>>> The release tarball builds without -Werror, so this warning is just a
>>> warning, since asprintf does return an int.
>>
>> It's also a varargs function though; varargs functions
>> must have a visible prototype.  I wouldn't be surprised if this
>> resulted in a crash/corruption on some targets/abis.
>>
>> ISTR seeing a patch for this, but I can't find it now.
>>
>> Did I just imagine it?
>>
> 
> Do you mean: https://sourceware.org/ml/gdb-patches/2015-07/msg00570.html
> 

Ah, yes.  Looks good to me, though bfd is owned by binutils; the
patch needs to be sent to the binutils@ list.

Thanks,
Pedro Alves