Issues for 24.3

Issues for 24.3

[Glenn Morris]

Here are some issues that I think (may) need addressing for 24.3.
If you have any more, or think any of these should be left for a
subsequent releases, please comment.

If you can fix any of these, please followup to the relevant bug
address.


Tabulated list mode header mouse-1 doesn't sort any more:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13560

Crash related to new profiler?
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545

Changing binding of mouse-1 leaves mark-active:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13523

If save-buffer is interrupted during the interactive coding prompt,
the original file is missing. Seems hard to fix, and present since 23.1,
so probably too late for 24.3.
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522


Build failures:

Doesn't build on ia64. Fixed? Need more info. Probably can be worked
around by lowering optimization, or setting GC_MARK_STACK.
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11971

Doesn't build on Mac OS X with gcc 4.6+. Maybe too late for 24.3.
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9927

Doesn't build on IRIX. I doubt this can be fixed except by someone with
IRIX, and only ~ 1 person seems to be an Emacs IRIX user, so probably
not that important to fix. Happens with older releases too.
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9684


Misc issues that seem to be unreproducible:

Problem sending mails:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13546

Mac OS X opens with tiny window, then crashes (?):
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13428

MS Windows crash:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145

Background colour issues:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12448

Redisplay issues:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906

MS Windows frame raising:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094

Mac OS X modeline not full width:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13014

Re: Issues for 24.3

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Date: Sat, 26 Jan 2013 22:20:09 -0500
> 
> Crash related to new profiler?
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545

A recipe was requested, but no response.

> If save-buffer is interrupted during the interactive coding prompt,
> the original file is missing. Seems hard to fix, and present since 23.1,
> so probably too late for 24.3.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522

Agreed.

> MS Windows crash:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145

This one is a duplicate of 13262, so it should be closed.

> Redisplay issues:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906

I won't hold my breath for this one, since I cannot reproduce it on
any machine where I can debug.  Michael Albinus produced some trace
data, and I'm studying it, but it isn't easy (and I'm not even sure
there's only one bug here).

> MS Windows frame raising:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094

Without a reproducible recipe, no chance of solving this one.  The OP
keeps referencing old reports "with changes" as the bug seemingly keep
evolving; they should instead post a clear recipe starting with "emacs -Q".

Yet another profiler crash [Was: Re: Issues for 24.3]

[Dmitry Antipov]

On 01/27/2013 10:36 AM, Eli Zaretskii wrote:

>> Crash related to new profiler?
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545
>
> A recipe was requested, but no response.

This crash is likely caused by C stack overflow and may be really hard
to reproduce. I was unable to do it, but found another profiler crash
which is much easier to reproduce (Bug#13573).

Dmitry

Re: Yet another profiler crash [Was: Re: Issues for 24.3]

[YAMAMOTO Mitsuharu]

>>>>> On Mon, 28 Jan 2013 09:37:14 +0400, Dmitry Antipov <dmantipov@yandex.ru> said:

>>> Crash related to new profiler?
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545
>> 
>> A recipe was requested, but no response.

> This crash is likely caused by C stack overflow and may be really
> hard to reproduce.

Can we make sure that keys (backtraces?) in the hash tables used in
the profiler are always acyclic?  Otherwise, using the `equal' test
for such keys are wrong and may lead to stack overflow, because an
error signal triggers another hash table operation in the profiler.

I noticed that the emacs-24 branch used the `equal' test, but the
trunk used a specialized function (cmpfn_profiler) for comparing keys.
Probably the issue is fixed in the trunk.

				     YAMAMOTO Mitsuharu
				mituharu@math.s.chiba-u.ac.jp

Re: Issues for 24.3

[Stefan Monnier]

> Crash related to new profiler?
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545

I think we can release 24.3 without fixing this bug: it's not
a regression since the profiler didn't work in Emacs-24.1 either.


        Stefan

Re: Issues for 24.3

[Glenn Morris]

Status update:

I don't feel that many more pretests are needed.

IMO the only real outstanding issue is 13523.
Please have a think about this issue if you can.

Mac OS X maybe: 13428 (need info).

MS Windows maybe: 13527 (need info).

query-replace issue: 13579 (I haven't read this properly)


Glenn Morris wrote:

> Tabulated list mode header mouse-1 doesn't sort any more:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13560

Hopefully fixed.

> Crash related to new profiler?
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545

Probably deferred to after 24.3.

> Changing binding of mouse-1 leaves mark-active:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13523

This is the one real remaining issue that bothers me.

> If save-buffer is interrupted during the interactive coding prompt,
> the original file is missing. Seems hard to fix, and present since 23.1,
> so probably too late for 24.3.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522

Hopefully fixed in trunk.

> Build failures:
>
> Doesn't build on ia64. Fixed? Need more info. Probably can be worked
> around by lowering optimization, or setting GC_MARK_STACK.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11971
>
> Doesn't build on Mac OS X with gcc 4.6+. Maybe too late for 24.3.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9927
>
> Doesn't build on IRIX. I doubt this can be fixed except by someone with
> IRIX, and only ~ 1 person seems to be an Emacs IRIX user, so probably
> not that important to fix. Happens with older releases too.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9684

Not likely to be any progress on any of these.

> Misc issues that seem to be unreproducible:
>
> Problem sending mails:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13546

Maybe fixed.

> Mac OS X opens with tiny window, then crashes (?):
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13428

Need info.

> MS Windows crash:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145

Was previously fixed.

> Background colour issues:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12448
>
> Redisplay issues:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906

Not likely to be any progress on any of these.

> MS Windows frame raising:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094

Not an issue for 24.3.

> Mac OS X modeline not full width:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13014

Resolved.


New additions:

MS Windows process issue:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13527

query-replace thing
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13579

RE: Issues for 24.3

[Drew Adams]

> I don't feel that many more pretests are needed.
> IMO the only real outstanding issue is 13523.

These regressions are still broken in the pretest, AFAIK.

 8429: `flush-lines' does not flush all it should
13061: redefining a function no longer cancels `debug-on-entry'
11314: incorrect Lisp indentation for `if' with `@@@@@@'
11325: wrong order in output of `substitute-command-keys'
11364: mouse selection then cursor movement no longer extends 
       selection in `delete-selection-mode'

Re: Issues for 24.3

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Date: Thu, 31 Jan 2013 23:06:55 -0500
> 
> MS Windows maybe: 13527 (need info).

Given that only one user reported this, I'd say this isn't a critical
issue.  (I would, of course, be happy to dig into it given more
information.)  I've already re-read the relevant sources, trying to
find some loophole, but didn't find anything evident.  As I don't use
Org too much, it could be that the problem is revealed by something
that doesn't happen in my day-to-day usage.

> New additions:
> 
> MS Windows process issue:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13527

See above.

Re: Issues for 24.3

[Alan Mackenzie]

Hello, Glenn.

On Sat, Jan 26, 2013 at 10:20:09PM -0500, Glenn Morris wrote:

> Here are some issues that I think (may) need addressing for 24.3.
> If you have any more, or think any of these should be left for a
> subsequent releases, please comment.

I currently have two bugs, almost fixed, 1.5 of which are regressions.  I
think the fixes should go into 24.3.  They are:

11152: c-doc-comment-style stops working as a style variable.  This is a
regression in define-globalized-minor-mode.

11749: 24.1; C-mode indentation gives wrong-type-argument error.  This is
an error in CC Mode's state cache.

11152 is currently under negotiation between Stefan and myself.  The fix
for 11749 is currently out with an OP (Michael Welsh Duggan) whom I've
asked for feedback.

-- 
Alan Mackenzie (Nuremberg, Germany).

RE: Issues for 24.3

[Drew Adams]

> > I don't feel that many more pretests are needed.
> > IMO the only real outstanding issue is 13523.
> 
> These regressions are still broken in the pretest, AFAIK.
> 
>  8429: `flush-lines' does not flush all it should
> 13061: redefining a function no longer cancels `debug-on-entry'
> 11314: incorrect Lisp indentation for `if' with `@@@@@@'
> 11325: wrong order in output of `substitute-command-keys'
> 11364: mouse selection then cursor movement no longer extends 
>        selection in `delete-selection-mode'

Oh, and this one too:

13149: Emacs thinks file was changed outside Emacs, but it was not

RE: Issues for 24.3

[Drew Adams]

> Oh, and this one too:
> 13149: Emacs thinks file was changed outside Emacs, but it was not

Oops, sorry about that one.  I'm not sure that is a problem in the pretest.  It
is a problem in the trunk, at least.

Re: Issues for 24.3

[Ivan Kanis]

> Glenn wrote:
>
> IMO the only real outstanding issue is 13523.
> Please have a think about this issue if you can.
>
> Mac OS X maybe: 13428 (need info).

I seem to be the only one running OS X 10.4. It doesn't bother me that
you make a release wit this bug. It can be fixed afterward.
-- 
Ivan Kanis
http://ivan.kanis.fr

Rule your mind or it will rule you.
    -- Horace

Re: Issues for 24.3

[Dmitry Gutov]

Glenn Morris <rgm@gnu.org> writes:

> Here are some issues that I think (may) need addressing for 24.3.
> If you have any more, or think any of these should be left for a
> subsequent releases, please comment.

Performance regression with uncompiled ruby-mode:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13605

Could someone macroexpansion-savvy take a look?

Re: Issues for 24.3

[Glenn Morris]

Eli Zaretskii wrote:

> Given that only one user reported this, I'd say this isn't a critical
> issue.  (I would, of course, be happy to dig into it given more
> information.)  I've already re-read the relevant sources, trying to
> find some loophole, but didn't find anything evident.  As I don't use
> Org too much, it could be that the problem is revealed by something
> that doesn't happen in my day-to-day usage.

OK; thanks for looking at it.

Re: Issues for 24.3

[Glenn Morris]

Dmitry Gutov wrote:

> Performance regression with uncompiled ruby-mode:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13605

I don't see how that can be a regression when eager macro expansion does
not exist prior to 24.3. Anyway, the speed of uncompiled code is not
that important.

Re: Issues for 24.3

[Glenn Morris]

Alan Mackenzie wrote:

> 11152: c-doc-comment-style stops working as a style variable.  This is a
> regression in define-globalized-minor-mode.

It feels too late to change define-globalized-minor-mode in emacs-24 at
this stage. Either we can find a cc-mode-specific solution, or live with
the problem in 24.3.

> 11749: 24.1; C-mode indentation gives wrong-type-argument error.  This is
> an error in CC Mode's state cache.

Seems like the patch gives good results, so probably you should install
it to whichever branch you think appropriate. I don't think anyone other
than you can evaluate how safe a cc-mode fix is.


Then I think it's probably time for another pretest.

Re: Issues for 24.3

[J Irving]

FWIW, I still see horrible full screen toolbar rendering in the .92 pretest, very much like what I reported here:

   http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13444

OS X is not a high priority platform for emacs as I understand it, but if there's a hope of improvement before 24.3 is released, I will attempt to reproduce that bug precisely with the pretest.

This is not a regression per se, since I don't believe Emacs has previously officially supported any kind of full screen mode on OS X. (I've previously used an old-style fullscreen hack via homebrew with the NS port, and more recently have been using Mitsuharu-san's awesome Mac port which has a good - also old-style, non-Lion - fullscreen mode.) Nonetheless, the lack of a viable full screen mode is critical to me :)

cheers, J

On 2013-01-26, at 10:20 PM, Glenn Morris <rgm@gnu.org> wrote:

> 
> Here are some issues that I think (may) need addressing for 24.3.
> If you have any more, or think any of these should be left for a
> subsequent releases, please comment.
> 
> If you can fix any of these, please followup to the relevant bug
> address.
> 
> 
> Tabulated list mode header mouse-1 doesn't sort any more:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13560
> 
> Crash related to new profiler?
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545
> 
> Changing binding of mouse-1 leaves mark-active:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13523
> 
> If save-buffer is interrupted during the interactive coding prompt,
> the original file is missing. Seems hard to fix, and present since 23.1,
> so probably too late for 24.3.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522
> 
> 
> Build failures:
> 
> Doesn't build on ia64. Fixed? Need more info. Probably can be worked
> around by lowering optimization, or setting GC_MARK_STACK.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11971
> 
> Doesn't build on Mac OS X with gcc 4.6+. Maybe too late for 24.3.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9927
> 
> Doesn't build on IRIX. I doubt this can be fixed except by someone with
> IRIX, and only ~ 1 person seems to be an Emacs IRIX user, so probably
> not that important to fix. Happens with older releases too.
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9684
> 
> 
> Misc issues that seem to be unreproducible:
> 
> Problem sending mails:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13546
> 
> Mac OS X opens with tiny window, then crashes (?):
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13428
> 
> MS Windows crash:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145
> 
> Background colour issues:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12448
> 
> Redisplay issues:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906
> 
> MS Windows frame raising:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094
> 
> Mac OS X modeline not full width:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13014
> 

Re: Issues for 24.3

[Dmitry Gutov]

On 02.02.2013 7:15, Glenn Morris wrote:
> Dmitry Gutov wrote:
>
>> Performance regression with uncompiled ruby-mode:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13605
>
> I don't see how that can be a regression when eager macro expansion does
> not exist prior to 24.3. Anyway, the speed of uncompiled code is not
> that important.

Indeed, sorry, not that kind of regression. Still, though, when 
uncompiled, the function is 2-3 times slower than the 24.2 version.

If you think it's fine, that's okay, just wanted to bring it to attention.

Re: Issues for 24.3

[Alan Mackenzie]

Hi, Glenn.

On Fri, Feb 01, 2013 at 10:19:08PM -0500, Glenn Morris wrote:
> Alan Mackenzie wrote:

> > 11152: c-doc-comment-style stops working as a style variable.  This is a
> > regression in define-globalized-minor-mode.

> It feels too late to change define-globalized-minor-mode in emacs-24 at
> this stage. Either we can find a cc-mode-specific solution, or live with
> the problem in 24.3.

OK.  Stefan agrees with you.  I'll see what I can manage.

> > 11749: 24.1; C-mode indentation gives wrong-type-argument error.  This is
> > an error in CC Mode's state cache.

> Seems like the patch gives good results, so probably you should install
> it to whichever branch you think appropriate. I don't think anyone other
> than you can evaluate how safe a cc-mode fix is.

I've committed it to emacs-24.  Maybe, just maybe, that particular bit of
code will now have achieved some stability.

> Then I think it's probably time for another pretest.

:-)

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Issues for 24.3

[Jan Djärv]

Hello.

This is fixed in the trunk. As it is not a regression, the release branch is not appropriate.

You can select between old/new fullscreen by customizing ns-use-native-fullscreen.

     Jan D.

2 feb 2013 kl. 04:48 skrev J Irving <j@lollyshouse.ca>:

> FWIW, I still see horrible full screen toolbar rendering in the .92 pretest, very much like what I reported here:
> 
>   http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13444
> 
> OS X is not a high priority platform for emacs as I understand it, but if there's a hope of improvement before 24.3 is released, I will attempt to reproduce that bug precisely with the pretest.
> 
> This is not a regression per se, since I don't believe Emacs has previously officially supported any kind of full screen mode on OS X. (I've previously used an old-style fullscreen hack via homebrew with the NS port, and more recently have been using Mitsuharu-san's awesome Mac port which has a good - also old-style, non-Lion - fullscreen mode.) Nonetheless, the lack of a viable full screen mode is critical to me :)
> 
> cheers, J
> 
> On 2013-01-26, at 10:20 PM, Glenn Morris <rgm@gnu.org> wrote:
> 
>> 
>> Here are some issues that I think (may) need addressing for 24.3.
>> If you have any more, or think any of these should be left for a
>> subsequent releases, please comment.
>> 
>> If you can fix any of these, please followup to the relevant bug
>> address.
>> 
>> 
>> Tabulated list mode header mouse-1 doesn't sort any more:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13560
>> 
>> Crash related to new profiler?
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545
>> 
>> Changing binding of mouse-1 leaves mark-active:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13523
>> 
>> If save-buffer is interrupted during the interactive coding prompt,
>> the original file is missing. Seems hard to fix, and present since 23.1,
>> so probably too late for 24.3.
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522
>> 
>> 
>> Build failures:
>> 
>> Doesn't build on ia64. Fixed? Need more info. Probably can be worked
>> around by lowering optimization, or setting GC_MARK_STACK.
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11971
>> 
>> Doesn't build on Mac OS X with gcc 4.6+. Maybe too late for 24.3.
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9927
>> 
>> Doesn't build on IRIX. I doubt this can be fixed except by someone with
>> IRIX, and only ~ 1 person seems to be an Emacs IRIX user, so probably
>> not that important to fix. Happens with older releases too.
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9684
>> 
>> 
>> Misc issues that seem to be unreproducible:
>> 
>> Problem sending mails:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13546
>> 
>> Mac OS X opens with tiny window, then crashes (?):
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13428
>> 
>> MS Windows crash:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145
>> 
>> Background colour issues:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12448
>> 
>> Redisplay issues:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906
>> 
>> MS Windows frame raising:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094
>> 
>> Mac OS X modeline not full width:
>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13014
> 

Re: Issues for 24.3

[J Irving]

On 2013-02-05, at 7:22 AM, Jan Djärv <jan.h.d@swipnet.se> wrote:

> Hello.
> 
> This is fixed in the trunk. As it is not a regression, the release branch is not appropriate.
> 
> You can select between old/new fullscreen by customizing ns-use-native-fullscreen.

Ah, that I did not know.

Thanks for the update.

cheers, J


> 2 feb 2013 kl. 04:48 skrev J Irving <j@lollyshouse.ca>:
> 
>> FWIW, I still see horrible full screen toolbar rendering in the .92 pretest, very much like what I reported here:
>> 
>>  http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13444
>> 
>> OS X is not a high priority platform for emacs as I understand it, but if there's a hope of improvement before 24.3 is released, I will attempt to reproduce that bug precisely with the pretest.
>> 
>> This is not a regression per se, since I don't believe Emacs has previously officially supported any kind of full screen mode on OS X. (I've previously used an old-style fullscreen hack via homebrew with the NS port, and more recently have been using Mitsuharu-san's awesome Mac port which has a good - also old-style, non-Lion - fullscreen mode.) Nonetheless, the lack of a viable full screen mode is critical to me :)
>> 
>> cheers, J
>> 
>> On 2013-01-26, at 10:20 PM, Glenn Morris <rgm@gnu.org> wrote:
>> 
>>> 
>>> Here are some issues that I think (may) need addressing for 24.3.
>>> If you have any more, or think any of these should be left for a
>>> subsequent releases, please comment.
>>> 
>>> If you can fix any of these, please followup to the relevant bug
>>> address.
>>> 
>>> 
>>> Tabulated list mode header mouse-1 doesn't sort any more:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13560
>>> 
>>> Crash related to new profiler?
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13545
>>> 
>>> Changing binding of mouse-1 leaves mark-active:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13523
>>> 
>>> If save-buffer is interrupted during the interactive coding prompt,
>>> the original file is missing. Seems hard to fix, and present since 23.1,
>>> so probably too late for 24.3.
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13522
>>> 
>>> 
>>> Build failures:
>>> 
>>> Doesn't build on ia64. Fixed? Need more info. Probably can be worked
>>> around by lowering optimization, or setting GC_MARK_STACK.
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11971
>>> 
>>> Doesn't build on Mac OS X with gcc 4.6+. Maybe too late for 24.3.
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9927
>>> 
>>> Doesn't build on IRIX. I doubt this can be fixed except by someone with
>>> IRIX, and only ~ 1 person seems to be an Emacs IRIX user, so probably
>>> not that important to fix. Happens with older releases too.
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9684
>>> 
>>> 
>>> Misc issues that seem to be unreproducible:
>>> 
>>> Problem sending mails:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13546
>>> 
>>> Mac OS X opens with tiny window, then crashes (?):
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13428
>>> 
>>> MS Windows crash:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8145
>>> 
>>> Background colour issues:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12448
>>> 
>>> Redisplay issues:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12906
>>> 
>>> MS Windows frame raising:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13094
>>> 
>>> Mac OS X modeline not full width:
>>> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13014
>> 
>