Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[willie]

18. Nov 2016 08:49 by d...@linuxfromscratch.org:


>
>
> On 11/17/2016 02:47 AM, > wil...@tuta.io>  wrote:
>> Some Chromium versions break with system libs but 54.0.284.100 works
>> fine with the above system libs
>>
> Last time I looked, .90, all of these were broken in some way. libvpx 
> required --enable-experimental (I forget what extended feature it actually 
> needed, something like ??tp in vp9), libpng won't work with the apng patch, 
> libjpeg_turbo was a no go, and ICU had multiple issues. I'm building 100 now 
> as per current instructions, but will go back and put the system versions 
> back in if these are fixed. I don't see anything that fixes any of these bugs 
> in the changelog, but if you have some insight, I'd appreciate it.
>
> https://chromium.googlesource.com/chromium/src/+log/54.0.2840.90..54.0.2840.100?pretty=fuller=1
>
> Thanks for the heads up.
>
> --DJ
>
>

Might be worth noticing I'm currently using most of the build instructions from 
Arch Linux instead of the BLFS version. That might explain why the system libs 
work now.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[DJ Lucas]

On 11/17/2016 02:47 AM, wil...@tuta.io wrote:

Some Chromium versions break with system libs but 54.0.284.100 works
fine with the above system libs

Last time I looked, .90, all of these were broken in some way. libvpx 
required --enable-experimental (I forget what extended feature it 
actually needed, something like ??tp in vp9), libpng won't work with the 
apng patch, libjpeg_turbo was a no go, and ICU had multiple issues. I'm 
building 100 now as per current instructions, but will go back and put 
the system versions back in if these are fixed. I don't see anything 
that fixes any of these bugs in the changelog, but if you have some 
insight, I'd appreciate it.


https://chromium.googlesource.com/chromium/src/+log/54.0.2840.90..54.0.2840.100?pretty=fuller=1

Thanks for the heads up.

--DJ

--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[willie]

> The API whose change prevents libreoffice to build was announced. There is a
> fix in the book.
>
> The segfaults with the mozilla apps is due to a regression. This kind of
> things happen... There is a patch which allows firefox and thunderbird to run.
> I have to build seamonkey (never done that before), to check it works.
>
> For chromium, internal ICU is used in the book. For sgdisk, ICU should not be
> needed as of version 1.0.1 (see ticket #8527).
>
> So, I propose to revert back to 58.1 (with the patch of course). Would it be
> OK? Note that apps compiled with 57.1 seem to work with the 58.1 library, at
> least FF and TB. No need to recompile AFAICT.
>
> Regards
> Pierre


Haven't followed the book for a while and I'm pleasantly surprised to find 
Chromium. However, I've been using Chromium for a good while now with the 
system libs ICU, libjpeg, libpng and libvpx.





Some Chromium versions break with system libs but 54.0.284.100 works fine with 
the above system libs

-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 16/11/2016 21:37, Ken Moffat wrote:

On Wed, Nov 16, 2016 at 09:28:46PM +0100, Pierre Labastie wrote:

So, I propose to revert back to 58.1 (with the patch of course). Would it be
OK? Note that apps compiled with 57.1 seem to work with the 58.1 library, at
least FF and TB. No need to recompile AFAICT.


Sounds good to me - I'm impressed by how quickly you found the fix
for the regression.

ĸen


Thanks, I was greatly helped by the "Known Issues" paragraph in 
http://site.icu-project.org/download/58#TOC-ICU4C-Download...


Pierre



--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 16/11/2016 21:36, Douglas R. Reno wrote:



On Wed, Nov 16, 2016 at 2:28 PM, Pierre Labastie 
> wrote:


On 16/11/2016 09:31, wil...@tuta.io  wrote:
> 14. Nov 2016 20:35 by renodr2...@gmail.com
 >:
>
>
>
> On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs

> >> wrote:
>
> Douglas R. Reno wrote:
>
> On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat
>  >> wrote:
>
> On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ
Lucas wrote:
>
> On Sat, Nov 12, 2016 at
06:56:08PM +0100,
> Pierre Labastie wrote:
>
>
> OK, I think I found the culprit: went
back to
> ICU-57.1, compiled FF
> against it, and no segfaults for about
one hour...
>
> I'll try to find where the ICU API changed.
> Pierre
>
>
> All over the place, 58 has not been pretty
for many packages.
> This might be a good start:
> https://bugzilla.mozilla.org/show_bug.cgi?id=1299615

>
>
> For the overall changes,
> http://source.icu-project.org/repos/icu/icu/tags/release-58-

> 1/APIChangeReport.html
>
> But like DJ says, it ain't pretty. The page
linking to that,
> http://site.icu-project.org/download/58#TOC-Common-Changes

> summariese these as
>
> API Changes since ICU4C 57
>
> ·   The API change report looks large and noisy
this time. Most of
>  the changes are from the following:
>  · The layout engine code was removed,
resulting in a long
> list
>of Removed API.
>  · C++ APIs that used const StringPiece& now
pass/return
>StringPiece by value. All affected
functions show as being
>Removed with their old signature and as
Added with
> their new
>signature. (#12140)
>  · Most COUNT and LIMIT enum constants have been
> deprecated. (#12420)
>
>
> I'm rebuilding Firefox, Seamonkey, and Thunderbird
with ICU-57.1 as we
> speak. Probably about 6 hours until I have any news.
>
> If this fixes it, do we need to revert to 57.1?
>
>
> I would say yes, the revert is needed.  We need to keep
an eye on it
> though.  If it bothers enough upstream upstream
developers, then thy
> will probably create some compatibility functions.
>
> Usually in this situation though. old APIs are marked
deprecated for a
> while (years) before they are removed.

The API whose change prevents libreoffice to build was announced.
There is a
fix in the book.

The segfaults with the mozilla apps is due to a regression. This
kind of
things happen... There is a patch which allows firefox and
thunderbird to run.
I have to build seamonkey (never done that before), to check it works.

For chromium, internal ICU is used in the book. For sgdisk, ICU
should not be
needed as of version 1.0.1 (see ticket #8527).

So, I propose to revert back to 58.1 (with the patch of course).
Would it be
OK? Note that apps compiled with 57.1 seem to work with the 58.1
library, at
least FF and TB. No need to recompile AFAICT.

Regards
Pierre



Can we wait a couple days? Evince, Epiphany, and WebkitGTK+ all use 
those heavily, and I would like to make sure that those all work fine 
with 58.1 as well. They'll be updated in the next 48 hours.


If you want to revert earlier though, go ahead. I have no problems 
with it, other than that I'll have to interrupt my current build to 
move a script with the patch over.


I'll wait, no prob. I've accepted a bunch of tickets not long ago... I 
can test webkitGTK with midori, if you wish.


Pierre
-- 

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Bruce Dubbs]

Pierre Labastie wrote:

On 16/11/2016 09:31, wil...@tuta.io wrote:

14. Nov 2016 20:35 by renodr2...@gmail.com :



 On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs > wrote:

 Douglas R. Reno wrote:

 On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat
 > wrote:

 On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:

 On Sat, Nov 12, 2016 at 06:56:08PM +0100,
 Pierre Labastie wrote:


 OK, I think I found the culprit: went back to
 ICU-57.1, compiled FF
 against it, and no segfaults for about one hour...

 I'll try to find where the ICU API changed.
 Pierre


 All over the place, 58 has not been pretty for many 
packages.
 This might be a good start:
 https://bugzilla.mozilla.org/show_bug.cgi?id=1299615


 For the overall changes,
 http://source.icu-project.org/repos/icu/icu/tags/release-58-
 1/APIChangeReport.html

 But like DJ says, it ain't pretty. The page linking to that,
 http://site.icu-project.org/download/58#TOC-Common-Changes
 summariese these as

 API Changes since ICU4C 57

 ·   The API change report looks large and noisy this time. 
Most of
  the changes are from the following:
  · The layout engine code was removed, resulting in a long
 list
of Removed API.
  · C++ APIs that used const StringPiece& now pass/return
StringPiece by value. All affected functions show as 
being
Removed with their old signature and as Added with
 their new
signature. (#12140)
  · Most COUNT and LIMIT enum constants have been
 deprecated. (#12420)


 I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 
as we
 speak. Probably about 6 hours until I have any news.

 If this fixes it, do we need to revert to 57.1?


 I would say yes, the revert is needed.   We need to keep an eye on it
 though.  If it bothers enough upstream upstream developers, then thy
 will probably create some compatibility functions.

 Usually in this situation though. old APIs are marked deprecated for a
 while (years) before they are removed.


The API whose change prevents libreoffice to build was announced. There is a
fix in the book.

The segfaults with the mozilla apps is due to a regression. This kind of
things happen... There is a patch which allows firefox and thunderbird to run.
I have to build seamonkey (never done that before), to check it works.

For chromium, internal ICU is used in the book. For sgdisk, ICU should not be
needed as of version 1.0.1 (see ticket #8527).

So, I propose to revert back to 58.1 (with the patch of course). Would it be
OK? Note that apps compiled with 57.1 seem to work with the 58.1 library, at
least FF and TB. No need to recompile AFAICT.


Redoing version 58 is fine.  The purpose of -dev is to get things in and 
working and occasional breakage is expected.


  -- Bruce


--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Wed, Nov 16, 2016 at 09:28:46PM +0100, Pierre Labastie wrote:
> 
> So, I propose to revert back to 58.1 (with the patch of course). Would it be
> OK? Note that apps compiled with 57.1 seem to work with the 58.1 library, at
> least FF and TB. No need to recompile AFAICT.
> 
Sounds good to me - I'm impressed by how quickly you found the fix
for the regression.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Wed, Nov 16, 2016 at 2:28 PM, Pierre Labastie 
wrote:

> On 16/11/2016 09:31, wil...@tuta.io wrote:
> > 14. Nov 2016 20:35 by renodr2...@gmail.com  >:
> >
> >
> >
> > On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs  > > wrote:
> >
> > Douglas R. Reno wrote:
> >
> > On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat
> > >
> wrote:
> >
> > On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
> >
> > On Sat, Nov 12, 2016 at 06:56:08PM +0100,
> > Pierre Labastie wrote:
> >
> >
> > OK, I think I found the culprit: went back to
> > ICU-57.1, compiled FF
> > against it, and no segfaults for about one
> hour...
> >
> > I'll try to find where the ICU API changed.
> > Pierre
> >
> >
> > All over the place, 58 has not been pretty for many
> packages.
> > This might be a good start:
> > https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
> >
> >
> > For the overall changes,
> > http://source.icu-project.org/
> repos/icu/icu/tags/release-58-
> > 1/APIChangeReport.html
> >
> > But like DJ says, it ain't pretty. The page linking to
> that,
> > http://site.icu-project.org/
> download/58#TOC-Common-Changes
> > summariese these as
> >
> > API Changes since ICU4C 57
> >
> > ·   The API change report looks large and noisy this
> time. Most of
> >  the changes are from the following:
> >  · The layout engine code was removed, resulting in
> a long
> > list
> >of Removed API.
> >  · C++ APIs that used const StringPiece& now
> pass/return
> >StringPiece by value. All affected functions show
> as being
> >Removed with their old signature and as Added with
> > their new
> >signature. (#12140)
> >  · Most COUNT and LIMIT enum constants have been
> > deprecated. (#12420)
> >
> >
> > I'm rebuilding Firefox, Seamonkey, and Thunderbird with
> ICU-57.1 as we
> > speak. Probably about 6 hours until I have any news.
> >
> > If this fixes it, do we need to revert to 57.1?
> >
> >
> > I would say yes, the revert is needed.   We need to keep an eye
> on it
> > though.  If it bothers enough upstream upstream developers, then
> thy
> > will probably create some compatibility functions.
> >
> > Usually in this situation though. old APIs are marked deprecated
> for a
> > while (years) before they are removed.
>
> The API whose change prevents libreoffice to build was announced. There is
> a
> fix in the book.
>
> The segfaults with the mozilla apps is due to a regression. This kind of
> things happen... There is a patch which allows firefox and thunderbird to
> run.
> I have to build seamonkey (never done that before), to check it works.
>
> For chromium, internal ICU is used in the book. For sgdisk, ICU should not
> be
> needed as of version 1.0.1 (see ticket #8527).
>
> So, I propose to revert back to 58.1 (with the patch of course). Would it
> be
> OK? Note that apps compiled with 57.1 seem to work with the 58.1 library,
> at
> least FF and TB. No need to recompile AFAICT.
>
> Regards
> Pierre
>
>
>

Can we wait a couple days? Evince, Epiphany, and WebkitGTK+ all use those
heavily, and I would like to make sure that those all work fine with 58.1
as well. They'll be updated in the next 48 hours.

If you want to revert earlier though, go ahead. I have no problems with it,
other than that I'll have to interrupt my current build to move a script
with the patch over.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 16/11/2016 09:31, wil...@tuta.io wrote:
> 14. Nov 2016 20:35 by renodr2...@gmail.com :
> 
> 
> 
> On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs  > wrote:
> 
> Douglas R. Reno wrote:
> 
> On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat
> > wrote:
> 
> On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
> 
> On Sat, Nov 12, 2016 at 06:56:08PM +0100,
> Pierre Labastie wrote:
> 
> 
> OK, I think I found the culprit: went back to
> ICU-57.1, compiled FF
> against it, and no segfaults for about one hour...
> 
> I'll try to find where the ICU API changed.
> Pierre
> 
> 
> All over the place, 58 has not been pretty for many 
> packages.
> This might be a good start:
> https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
> 
> 
> For the overall changes,
> http://source.icu-project.org/repos/icu/icu/tags/release-58-
> 1/APIChangeReport.html
> 
> But like DJ says, it ain't pretty. The page linking to that,
> http://site.icu-project.org/download/58#TOC-Common-Changes
> summariese these as
> 
> API Changes since ICU4C 57
> 
> ·   The API change report looks large and noisy this time. 
> Most of
>  the changes are from the following:
>  · The layout engine code was removed, resulting in a long
> list
>of Removed API.
>  · C++ APIs that used const StringPiece& now pass/return
>StringPiece by value. All affected functions show as 
> being
>Removed with their old signature and as Added with
> their new
>signature. (#12140)
>  · Most COUNT and LIMIT enum constants have been
> deprecated. (#12420)
> 
> 
> I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 
> as we
> speak. Probably about 6 hours until I have any news.
> 
> If this fixes it, do we need to revert to 57.1?
> 
> 
> I would say yes, the revert is needed.   We need to keep an eye on it
> though.  If it bothers enough upstream upstream developers, then thy
> will probably create some compatibility functions.
> 
> Usually in this situation though. old APIs are marked deprecated for a
> while (years) before they are removed.

The API whose change prevents libreoffice to build was announced. There is a
fix in the book.

The segfaults with the mozilla apps is due to a regression. This kind of
things happen... There is a patch which allows firefox and thunderbird to run.
I have to build seamonkey (never done that before), to check it works.

For chromium, internal ICU is used in the book. For sgdisk, ICU should not be
needed as of version 1.0.1 (see ticket #8527).

So, I propose to revert back to 58.1 (with the patch of course). Would it be
OK? Note that apps compiled with 57.1 seem to work with the 58.1 library, at
least FF and TB. No need to recompile AFAICT.

Regards
Pierre

-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[willie]

14. Nov 2016 20:35 by renodr2...@gmail.com:


>
>
> On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs <> bruce.du...@gmail.com> > 
> wrote:
>
>> Douglas R. Reno wrote:
>>
>>> On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat <>>> zarniwh...@ntlworld.com>>> 
>>> > wrote:
>>>
 On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:

 On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:

>>
>>
>> OK, I think I found the culprit: went back to ICU-57.1, compiled FF
>> against it, and no segfaults for about one hour...
>>
>> I'll try to find where the ICU API changed.
>> Pierre
>>
>
> All over the place, 58 has not been pretty for many packages.
> This might be a good start:
> https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
>
>>
>>
 For the overall changes,
 http://source.icu-project.org/repos/icu/icu/tags/release-58-
 1/APIChangeReport.html

 But like DJ says, it ain't pretty. The page linking to that,
 http://site.icu-project.org/download/58#TOC-Common-Changes
 summariese these as

 API Changes since ICU4C 57

 ·   The API change report looks large and noisy this time. Most of
      the changes are from the following:
      · The layout engine code was removed, resulting in a long list
        of Removed API.
      · C++ APIs that used const StringPiece& now pass/return
        StringPiece by value. All affected functions show as being
        Removed with their old signature and as Added with their new
        signature. (#12140)
      · Most COUNT and LIMIT enum constants have been deprecated. (#12420)

>>
>>
>>> I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 as we
>>> speak. Probably about 6 hours until I have any news.
>>>
>>> If this fixes it, do we need to revert to 57.1?
>>>
>>
>> I would say yes, the revert is needed.   We need to keep an eye on it 
>> though.  If it bothers enough upstream upstream developers, then thy will 
>> probably create some compatibility functions.
>>
>> Usually in this situation though. old APIs are marked deprecated for a while 
>> (years) before they are removed.
>>
>>   -- Bruce
>>
>>
>
> Confirmed that Firefox, Thunderbird, and Seamonkey all work with ICU-57.1. 
> Final permission to revert to 57.1? 




ICU 58.1 works fine for me, except with the Mozilla family of products.





For thunderbird and Firefox it's enough to disable with-system-icu, after that 
the segfaults go away. Other programs which depend on ICU work fine with the 
new version. I've tested Chromium 54, Evince and Libreoffice and all are stable 
so far.





But I don't see problems with downgrading ICU either.

-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Sun, Nov 13, 2016 at 5:44 PM, Bruce Dubbs  wrote:

> Douglas R. Reno wrote:
>
>> On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat 
>> wrote:
>>
>>> On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
>>>
 On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
>>>
>>
> OK, I think I found the culprit: went back to ICU-57.1, compiled FF
> against it, and no segfaults for about one hour...
>
> I'll try to find where the ICU API changed.
> Pierre
>

 All over the place, 58 has not been pretty for many packages.
 This might be a good start:
 https://bugzilla.mozilla.org/show_bug.cgi?id=1299615

>>>
> For the overall changes,
>>> http://source.icu-project.org/repos/icu/icu/tags/release-58-
>>> 1/APIChangeReport.html
>>>
>>> But like DJ says, it ain't pretty. The page linking to that,
>>> http://site.icu-project.org/download/58#TOC-Common-Changes
>>> summariese these as
>>>
>>> API Changes since ICU4C 57
>>>
>>> ·   The API change report looks large and noisy this time. Most of
>>>  the changes are from the following:
>>>  · The layout engine code was removed, resulting in a long list
>>>of Removed API.
>>>  · C++ APIs that used const StringPiece& now pass/return
>>>StringPiece by value. All affected functions show as being
>>>Removed with their old signature and as Added with their new
>>>signature. (#12140)
>>>  · Most COUNT and LIMIT enum constants have been deprecated. (#12420)
>>>
>>
> I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 as we
>> speak. Probably about 6 hours until I have any news.
>>
>> If this fixes it, do we need to revert to 57.1?
>>
>
> I would say yes, the revert is needed.   We need to keep an eye on it
> though.  If it bothers enough upstream upstream developers, then thy will
> probably create some compatibility functions.
>
> Usually in this situation though. old APIs are marked deprecated for a
> while (years) before they are removed.
>
>   -- Bruce
>
>
>
Confirmed that Firefox, Thunderbird, and Seamonkey all work with ICU-57.1.
Final permission to revert to 57.1?
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[akhiezer]

> Date: Mon, 14 Nov 2016 15:00:13 +
> From: Ken Moffat <zarniwh...@ntlworld.com>
> Subject: Re: [blfs-dev] Seamonkey, Firefox,
>  and Thunderbird all cause segfaults with the latest SVN build
>
> On Mon, Nov 14, 2016 at 09:23:11AM +, akhiezer wrote:
> > > Date: Sun, 13 Nov 2016 21:59:07 +
> > > From: Ken Moffat <zarniwh...@ntlworld.com>
> > > Subject: Re: [blfs-dev] Seamonkey, Firefox,
> > >  and Thunderbird all cause segfaults with the latest SVN build
> > >
> > .
> > .
> > >
> > > I'm surprised that links cuts it at all for you - I find it a pain
> > > for googling error messages.  [...]
> > 
> > 
> > Any particular reason? Hopefully you can at least copy (e.g. via gpm)
> > from shell, and shift-paste into links; and similarly shift-select
> > in links, and then paste in shell. Or is it re tabs, or flipping
> > back'n'forth to shell, or what? Just interested.
> > 
> Most error messages find matches on list archives.  Doing this from
> Xorg, i n a 100x40 terminal, a google similar to yesterday's, for
> for
>  redefinition of ???struct iphdr???
> gives me 14 lines of "menu" (All/Images/Videos/etc/etc each on one
> line) with the first three matches.  And as I scroll down it is
> comparatively difficult to parse the three or four lines of plain
> text.
>


Maybe lynx's (out-of-the-box) colouring could help in the parsing?


akh




--
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Mon, Nov 14, 2016 at 09:23:11AM +, akhiezer wrote:
> > Date: Sun, 13 Nov 2016 21:59:07 +
> > From: Ken Moffat <zarniwh...@ntlworld.com>
> > Subject: Re: [blfs-dev] Seamonkey, Firefox,
> >  and Thunderbird all cause segfaults with the latest SVN build
> >
>   .
>   .
> >
> > I'm surprised that links cuts it at all for you - I find it a pain
> > for googling error messages.  [...]
> 
> 
> Any particular reason? Hopefully you can at least copy (e.g. via gpm)
> from shell, and shift-paste into links; and similarly shift-select
> in links, and then paste in shell. Or is it re tabs, or flipping
> back'n'forth to shell, or what? Just interested.
> 
Most error messages find matches on list archives.  Doing this from
Xorg, i n a 100x40 terminal, a google similar to yesterday's, for
for
 redefinition of ‘struct iphdr’
gives me 14 lines of "menu" (All/Images/Videos/etc/etc each on one
line) with the first three matches.  And as I scroll down it is
comparatively difficult to parse the three or four lines of plain
text.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[akhiezer]

> Date: Sun, 13 Nov 2016 21:59:07 +
> From: Ken Moffat <zarniwh...@ntlworld.com>
> Subject: Re: [blfs-dev] Seamonkey, Firefox,
>  and Thunderbird all cause segfaults with the latest SVN build
>
.
.
>
> I'm surprised that links cuts it at all for you - I find it a pain
> for googling error messages.  [...]


Any particular reason? Hopefully you can at least copy (e.g. via gpm)
from shell, and shift-paste into links; and similarly shift-select
in links, and then paste in shell. Or is it re tabs, or flipping
back'n'forth to shell, or what? Just interested.



akh





--
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Bruce Dubbs]

Douglas R. Reno wrote:

On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat  wrote:

On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:

On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:



OK, I think I found the culprit: went back to ICU-57.1, compiled FF
against it, and no segfaults for about one hour...

I'll try to find where the ICU API changed.
Pierre


All over the place, 58 has not been pretty for many packages.
This might be a good start:
https://bugzilla.mozilla.org/show_bug.cgi?id=1299615



For the overall changes,
http://source.icu-project.org/repos/icu/icu/tags/release-58-
1/APIChangeReport.html

But like DJ says, it ain't pretty. The page linking to that,
http://site.icu-project.org/download/58#TOC-Common-Changes
summariese these as

API Changes since ICU4C 57

·   The API change report looks large and noisy this time. Most of
 the changes are from the following:
 · The layout engine code was removed, resulting in a long list
   of Removed API.
 · C++ APIs that used const StringPiece& now pass/return
   StringPiece by value. All affected functions show as being
   Removed with their old signature and as Added with their new
   signature. (#12140)
 · Most COUNT and LIMIT enum constants have been deprecated. (#12420)



I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 as we
speak. Probably about 6 hours until I have any news.

If this fixes it, do we need to revert to 57.1?


I would say yes, the revert is needed.   We need to keep an eye on it 
though.  If it bothers enough upstream upstream developers, then thy will 
probably create some compatibility functions.


Usually in this situation though. old APIs are marked deprecated for a 
while (years) before they are removed.


  -- Bruce

--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[DJ Lucas]

If possible, I'd suggest using the in tree copy for now.

--DJ


On November 13, 2016 5:34:50 PM CST, "Douglas R. Reno"  
wrote:
>On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat 
>wrote:
>
>> On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
>> >
>> > > > > On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie
>wrote:
>> > > > > > >
>> > > OK, I think I found the culprit: went back to ICU-57.1, compiled
>FF
>> > > against it, and no segfaults for about one hour...
>> > >
>> > > I'll try to find where the ICU API changed.
>> > > Pierre
>> >
>> > All over the place, 58 has not been pretty for many packages.
>> > This might be a good start:
>> > https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
>> >
>> > --DJ
>> >
>> For the overall changes,
>> http://source.icu-project.org/repos/icu/icu/tags/release-58-
>> 1/APIChangeReport.html
>>
>> But like DJ says, it ain't pretty. The page linking to that,
>> http://site.icu-project.org/download/58#TOC-Common-Changes
>> summariese these as
>>
>> API Changes since ICU4C 57
>>
>> ·   The API change report looks large and noisy this time. Most of
>> the changes are from the following:
>> · The layout engine code was removed, resulting in a long list
>>   of Removed API.
>> · C++ APIs that used const StringPiece& now pass/return
>>   StringPiece by value. All affected functions show as being
>>   Removed with their old signature and as Added with their new
>>   signature. (#12140)
>> · Most COUNT and LIMIT enum constants have been deprecated.
>(#12420)
>>
>> ĸen
>> --
>> `I shall take my mountains', said Lu-Tze. `The climate will be good
>> for them.' -- Small Gods
>> --
>> http://lists.linuxfromscratch.org/listinfo/blfs-dev
>> FAQ: http://www.linuxfromscratch.org/blfs/faq.html
>> Unsubscribe: See the above information page
>>
>
>I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 as we
>speak. Probably about 6 hours until I have any news.
>
>If this fixes it, do we need to revert to 57.1?
>
>
>
>
>-- 
>http://lists.linuxfromscratch.org/listinfo/blfs-dev
>FAQ: http://www.linuxfromscratch.org/blfs/faq.html
>Unsubscribe: See the above information page

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Sun, Nov 13, 2016 at 4:07 PM, Ken Moffat  wrote:

> On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
> >
> > > > > On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
> > > > > > >
> > > OK, I think I found the culprit: went back to ICU-57.1, compiled FF
> > > against it, and no segfaults for about one hour...
> > >
> > > I'll try to find where the ICU API changed.
> > > Pierre
> >
> > All over the place, 58 has not been pretty for many packages.
> > This might be a good start:
> > https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
> >
> > --DJ
> >
> For the overall changes,
> http://source.icu-project.org/repos/icu/icu/tags/release-58-
> 1/APIChangeReport.html
>
> But like DJ says, it ain't pretty. The page linking to that,
> http://site.icu-project.org/download/58#TOC-Common-Changes
> summariese these as
>
> API Changes since ICU4C 57
>
> ·   The API change report looks large and noisy this time. Most of
> the changes are from the following:
> · The layout engine code was removed, resulting in a long list
>   of Removed API.
> · C++ APIs that used const StringPiece& now pass/return
>   StringPiece by value. All affected functions show as being
>   Removed with their old signature and as Added with their new
>   signature. (#12140)
> · Most COUNT and LIMIT enum constants have been deprecated. (#12420)
>
> ĸen
> --
> `I shall take my mountains', said Lu-Tze. `The climate will be good
> for them.' -- Small Gods
> --
> http://lists.linuxfromscratch.org/listinfo/blfs-dev
> FAQ: http://www.linuxfromscratch.org/blfs/faq.html
> Unsubscribe: See the above information page
>

I'm rebuilding Firefox, Seamonkey, and Thunderbird with ICU-57.1 as we
speak. Probably about 6 hours until I have any news.

If this fixes it, do we need to revert to 57.1?
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Nov 13, 2016 4:05 PM, "Ken Moffat"  wrote:
>
> On Sun, Nov 13, 2016 at 02:51:23PM -0600, Douglas R. Reno wrote:
> >
> > Any disadvantages if I downgrade to ICU-57.1? I have SO many updates
out in
> > BLFS land and one in LFS! Links won't cut it forever.
>
> AFAICS (ignoring the "clean up" changes in 58.1 which have caused
> problems), the main other change in 58 seems to have been to move to
> unicode-9.0.  The changes there were generally for old glyphs or for
> emojis.  It seems that 57 is solid, and the big distros do not seem
> to have started to use it.
>
> I'm surprised that links cuts it at all for you - I find it a pain
> for googling error messages.  I would go with 57 for the moment.
>
> ĸen
> --
> `I shall take my mountains', said Lu-Tze. `The climate will be good
> for them.' -- Small Gods

I have a couple Windows PCs for that. Thats all that they're good for! :-)
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sun, Nov 13, 2016 at 12:55:01PM -0600, DJ Lucas wrote:
> 
> > > > On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
> > > > > > 
> > OK, I think I found the culprit: went back to ICU-57.1, compiled FF
> > against it, and no segfaults for about one hour...
> > 
> > I'll try to find where the ICU API changed.
> > Pierre
> 
> All over the place, 58 has not been pretty for many packages.
> This might be a good start:
> https://bugzilla.mozilla.org/show_bug.cgi?id=1299615
> 
> --DJ
> 
For the overall changes,
http://source.icu-project.org/repos/icu/icu/tags/release-58-1/APIChangeReport.html

But like DJ says, it ain't pretty. The page linking to that,
http://site.icu-project.org/download/58#TOC-Common-Changes
summariese these as

API Changes since ICU4C 57

·   The API change report looks large and noisy this time. Most of
the changes are from the following:
· The layout engine code was removed, resulting in a long list
  of Removed API.
· C++ APIs that used const StringPiece& now pass/return
  StringPiece by value. All affected functions show as being
  Removed with their old signature and as Added with their new
  signature. (#12140)
· Most COUNT and LIMIT enum constants have been deprecated. (#12420)

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sun, Nov 13, 2016 at 02:51:23PM -0600, Douglas R. Reno wrote:
> 
> Any disadvantages if I downgrade to ICU-57.1? I have SO many updates out in
> BLFS land and one in LFS! Links won't cut it forever.

AFAICS (ignoring the "clean up" changes in 58.1 which have caused
problems), the main other change in 58 seems to have been to move to
unicode-9.0.  The changes there were generally for old glyphs or for
emojis.  It seems that 57 is solid, and the big distros do not seem
to have started to use it.

I'm surprised that links cuts it at all for you - I find it a pain
for googling error messages.  I would go with 57 for the moment.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sun, Nov 13, 2016 at 03:29:14PM +0100, Pierre Labastie wrote:
> On 13/11/2016 09:27, Pierre Labastie wrote:
> > On 12/11/2016 23:33, Ken Moffat wrote:
> > > On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
> > > > > Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I 
> > > > > can
> > > > > shuffle versions ;). Then I'll try previous sqlite, previous NSS, and 
> > > > > previous
> > > > > harfbuzz.
> > > > > 
> OK, I think I found the culprit: went back to ICU-57.1, compiled FF against
> it, and no segfaults for about one hour...
> 
> I'll try to find where the ICU API changed.
> Pierre

If only the mailserver was not so slow!  I've just sent a reply with
details of what I had and had not updated, and guessed that icu might
be the problem.  That (and other reported problems) was also why I
did not update to icu58 - but I didn't want to question it if the
problem was clearly elsewhere.

Google is not finding many distros using icu 58.  Apparently debian
experimental has/had it, also cygwin.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Nov 13, 2016 2:47 PM, "Ken Moffat"  wrote:
>
> On Sun, Nov 13, 2016 at 09:27:48AM +0100, Pierre Labastie wrote:
> >
> > Sorry for not being clear. I meant to "answer" the sentence "Let me try
to get
> > back to GTK-3.20. Then I'll try previous sqlite, previous NSS, and
previous
> > harfbuzz".
> >
> > I should have said something like "whatever the version of those
packages, FF
> > still segfaults. And whatever the default search engine, it segfaults
when I
> > type the first letter in the search field."
> >
> > But I was wondering whether the render engines could explain the
segfault. Now
> > I have tried to get back to swrast, and the segfault is still here...
> >
> > Pierre
> Thanks for explaining.  I've got a fresh build under way.
> Unfortunately, my attempt to add '-g' to _my_ CFLAGS in LFS was in
> effective (luser error) but I've got them in BLFS.
>
> Not quite there yet, fix on the way for net-tools-CVS with 4.8
> headers.  Glad I was doing htat in chroot.
>
> ĸen
> --

Any disadvantages if I downgrade to ICU-57.1? I have SO many updates out in
BLFS land and one in LFS! Links won't cut it forever.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sat, Nov 12, 2016 at 09:22:15AM +0100, Pierre Labastie wrote:

(replying to the earlier post)
> 
> For the record, my DE/DM is LXDE/openbox. The engine for the search field is
> google (the default). I tried to move to duckduckgo, but it opened the
> duckduckgo page. I could search something, then it crashed (segfaulted) soon
> after. I did not pursue further.
> 
My build is finished (my normal packages, but only as far as firefox),
and it seems to work.

The base system (probably irrelevant, since I think you were on
LFS-7.10) is LFS-svn-20161022, with BLFS packages updated to r17862
(20161009 where you added xdg-user-dirs.

Along with that, the following ARE up to date:

curl-7.51.0, firefox-49.0.2, nspr-4.13.1, nss-3.27.1, sqlite-3.15.0.

I have used the default google, also en.wikipedia and duckduckgo,
all to search for firefox.  I have used https:// to google,
kernel.org, amazon.

Looking at what is NOT up to date in this build (from BLFS trac) I
noticed the following:

poppler-0.48.0 - I have not built (older) poppler yet
vala-0.34.1 - I still have 0.32.1
gnutls-3.5.5 - I still have 3.5.3
sudo-1.8.18p1 - I still have 1.8.17p1
icu-58.1 - I still have 57.1
libpng-1.6.26 - I still have 1.6.26
libXi-1.7.8 - I still have 1.7.7
harfbuzz-1.3.3 - I still have 1.3.2

Obviously, there might also be something else which I have not
built, but which you guys did build.

I can't test thunderbird (no spare mail account), and for the moment
I don't see any point in testing seamonkey.

Oddly, while the build was running last night I used firefox on the
7.10 host and late-on (possibly in the early part of BLFS) the
"smart bar" suggestions suddenly started spreading out to left and
right of the main firefox window, and remained there.  Killed
firefox, the spread-out windows at the edges remained but eventually
disappeared. It happened again, but the spread-out parts later
disappeared.  It (the host) was fine a couple of hours ago when I
searched re net-tools, and the new one is fine.

I have a suspicion that icu-58.1 might be the problem - it is
linked by libxul.so.  For firefox, I'm using system harfbuzz - it
seems unlikely that moving that from 1.3.2 to 1.3.3 will break
things, and in any case we do not use system harfbuzz in seamonkey
or thunderbird.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[DJ Lucas]

On 11/13/2016 08:29 AM, Pierre Labastie wrote:

On 13/11/2016 09:27, Pierre Labastie wrote:

On 12/11/2016 23:33, Ken Moffat wrote:

On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:

Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I
can
shuffle versions ;). Then I'll try previous sqlite, previous NSS,
and previous
harfbuzz.


OK, I think I found the culprit: went back to ICU-57.1, compiled FF
against it, and no segfaults for about one hour...

I'll try to find where the ICU API changed.
Pierre


All over the place, 58 has not been pretty for many packages.
This might be a good start:
https://bugzilla.mozilla.org/show_bug.cgi?id=1299615

--DJ



--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sun, Nov 13, 2016 at 09:27:48AM +0100, Pierre Labastie wrote:
> 
> Sorry for not being clear. I meant to "answer" the sentence "Let me try to get
> back to GTK-3.20. Then I'll try previous sqlite, previous NSS, and previous
> harfbuzz".
> 
> I should have said something like "whatever the version of those packages, FF
> still segfaults. And whatever the default search engine, it segfaults when I
> type the first letter in the search field."
> 
> But I was wondering whether the render engines could explain the segfault. Now
> I have tried to get back to swrast, and the segfault is still here...
> 
> Pierre
Thanks for explaining.  I've got a fresh build under way.
Unfortunately, my attempt to add '-g' to _my_ CFLAGS in LFS was in
effective (luser error) but I've got them in BLFS.

Not quite there yet, fix on the way for net-tools-CVS with 4.8
headers.  Glad I was doing htat in chroot.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 13/11/2016 09:27, Pierre Labastie wrote:

On 12/11/2016 23:33, Ken Moffat wrote:

On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:

Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
shuffle versions ;). Then I'll try previous sqlite, previous NSS, and previous
harfbuzz.

OK, I think I found the culprit: went back to ICU-57.1, compiled FF 
against it, and no segfaults for about one hour...


I'll try to find where the ICU API changed.
Pierre
--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 12/11/2016 23:33, Ken Moffat wrote:
> On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
>>
>>>
>>> Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
>>> shuffle versions ;). Then I'll try previous sqlite, previous NSS, and 
>>> previous
>>> harfbuzz.
>>>
>> Nothing worked...
>> Could it have something to do with the xorg video driver? I usually use KMS
>> with swrast. But on this machine, which has integrated intel graphics, I use
>> the intel drivers.
>>
>> Pierre
>>
> I interpret "nothing worked" to mean "no video in Xorg".  If I've
> misunderstood you, please clarify, and ignore the rest of this reply.

Sorry for not being clear. I meant to "answer" the sentence "Let me try to get
back to GTK-3.20. Then I'll try previous sqlite, previous NSS, and previous
harfbuzz".

I should have said something like "whatever the version of those packages, FF
still segfaults. And whatever the default search engine, it segfaults when I
type the first letter in the search field."

But I was wondering whether the render engines could explain the segfault. Now
I have tried to get back to swrast, and the segfault is still here...

Pierre
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

On Nov 12, 2016 3:42 PM, "Pierre Labastie"  wrote:
>
> On 12/11/2016 09:22, Pierre Labastie wrote:
> > On 12/11/2016 05:20, Ken Moffat wrote:
> >> On Fri, Nov 11, 2016 at 10:59:28PM +0100, Pierre Labastie wrote:
> >>> On 11/11/2016 20:01, Douglas R. Reno wrote:
>  Greetings,
> 
>  I am unable to get Seamonkey, Firefox, or Thunderbird to work
properly with
>  the latest SVN of BLFS.
> 
>  They all segfault with references to libpthread (expected),
libnspr4, libnss,
>  and libxul.
> 
>  I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr ->
nss), and
>  when I get home, I'll attempt to build a debug build to see if I can
debug
>  this issue further.
> 
>  Different applications give me different behavior:
> 
>  Firefox will spuriously crash (probably background rendering...
again)
>  Thunderbird will crash when trying to send mail
>  Seamonkey will crash before it finishes starting.
> 
>  I'll try to investigate this later, but this is all I've been able
to do for days.
> 
> 
> >>>
> >>> Same issue here with Firefox (49.0.2) [...]
> >>
> >> On 7.10 with recent sqlite, nspr, nss it seems fine.  But I'm still
> >> on gtk+-3.20 and the old glib.
> >>
>
> >
> > Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
> > shuffle versions ;). Then I'll try previous sqlite, previous NSS, and
previous
> > harfbuzz.
> >
> Nothing worked...
> Could it have something to do with the xorg video driver? I usually use
KMS
> with swrast. But on this machine, which has integrated intel graphics, I
use
> the intel drivers.
>
> Pierre
>
>

I'm not 100% sure. If it helps, I'm running an nvidia card with the nouveau
drivers.

--Douglas R. Reno, bashing his head over a wall due to FF, TB, and SM.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[DJ Lucas]

On 11/12/2016 11:56 AM, Pierre Labastie wrote:

On 12/11/2016 09:22, Pierre Labastie wrote:

On 12/11/2016 05:20, Ken Moffat wrote:

On Fri, Nov 11, 2016 at 10:59:28PM +0100, Pierre Labastie wrote:

On 11/11/2016 20:01, Douglas R. Reno wrote:

Greetings,

I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
the latest SVN of BLFS.

They all segfault with references to libpthread (expected), libnspr4, libnss,
and libxul.

I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss), and
when I get home, I'll attempt to build a debug build to see if I can debug
this issue further.

Different applications give me different behavior:

Firefox will spuriously crash (probably background rendering... again)
Thunderbird will crash when trying to send mail
Seamonkey will crash before it finishes starting.

I'll try to investigate this later, but this is all I've been able to do for 
days.




Same issue here with Firefox (49.0.2) [...]


On 7.10 with recent sqlite, nspr, nss it seems fine.  But I'm still
on gtk+-3.20 and the old glib.





Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
shuffle versions ;). Then I'll try previous sqlite, previous NSS, and previous
harfbuzz.


Nothing worked...
Could it have something to do with the xorg video driver? I usually use KMS
with swrast. But on this machine, which has integrated intel graphics, I use
the intel drivers.

Pierre


I'm not on latest and greatest as far as system goes, been a bit slow to 
replace this one as I've been digging into no /usr/lib64, but running 
FF-49.0.2 and TB-45.4.0 fine. My current install list is too big to send 
to list. System was originally built right at a month ago 10/10, 
linux-4.8.1. Install list doesn't include the LFS packages, but has 
build order and BLFS packages used. is at 
http://www.linuxfromscratch.org/~dj/buildlist.txt. Also running with 
Intel drivers.


Maybe it can help. I probably won't start on a new build for a couple of 
days yet. On a mission to install *everything* in the book, but 
certificates got in the way. :-)


--DJ

--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Sat, Nov 12, 2016 at 06:56:08PM +0100, Pierre Labastie wrote:
> 
> > 
> > Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
> > shuffle versions ;). Then I'll try previous sqlite, previous NSS, and 
> > previous
> > harfbuzz.
> > 
> Nothing worked...
> Could it have something to do with the xorg video driver? I usually use KMS
> with swrast. But on this machine, which has integrated intel graphics, I use
> the intel drivers.
> 
> Pierre
> 
I interpret "nothing worked" to mean "no video in Xorg".  If I've
misunderstood you, please clarify, and ignore the rest of this reply.

Anything relevant in the Xorg log, or indeed in the system log (or
journalctl if this is systemd) ?

For systems *without* a DM,
 startx 2>&1 | tee xlog
will capture any errors which might otherwise disappear.

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 12/11/2016 09:22, Pierre Labastie wrote:
> On 12/11/2016 05:20, Ken Moffat wrote:
>> On Fri, Nov 11, 2016 at 10:59:28PM +0100, Pierre Labastie wrote:
>>> On 11/11/2016 20:01, Douglas R. Reno wrote:
 Greetings,

 I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
 the latest SVN of BLFS.

 They all segfault with references to libpthread (expected), libnspr4, 
 libnss,
 and libxul.

 I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss), 
 and
 when I get home, I'll attempt to build a debug build to see if I can debug
 this issue further.

 Different applications give me different behavior:

 Firefox will spuriously crash (probably background rendering... again)
 Thunderbird will crash when trying to send mail
 Seamonkey will crash before it finishes starting.

 I'll try to investigate this later, but this is all I've been able to do 
 for days.


>>>
>>> Same issue here with Firefox (49.0.2) [...]
>>
>> On 7.10 with recent sqlite, nspr, nss it seems fine.  But I'm still
>> on gtk+-3.20 and the old glib.
>>

> 
> Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
> shuffle versions ;). Then I'll try previous sqlite, previous NSS, and previous
> harfbuzz.
> 
Nothing worked...
Could it have something to do with the xorg video driver? I usually use KMS
with swrast. But on this machine, which has integrated intel graphics, I use
the intel drivers.

Pierre


-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 12/11/2016 05:20, Ken Moffat wrote:
> On Fri, Nov 11, 2016 at 10:59:28PM +0100, Pierre Labastie wrote:
>> On 11/11/2016 20:01, Douglas R. Reno wrote:
>>> Greetings,
>>>
>>> I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
>>> the latest SVN of BLFS.
>>>
>>> They all segfault with references to libpthread (expected), libnspr4, 
>>> libnss,
>>> and libxul.
>>>
>>> I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss), and
>>> when I get home, I'll attempt to build a debug build to see if I can debug
>>> this issue further.
>>>
>>> Different applications give me different behavior:
>>>
>>> Firefox will spuriously crash (probably background rendering... again)
>>> Thunderbird will crash when trying to send mail
>>> Seamonkey will crash before it finishes starting.
>>>
>>> I'll try to investigate this later, but this is all I've been able to do 
>>> for days.
>>>
>>>
>>
>> Same issue here with Firefox (49.0.2) (*), but I have not seen anything with
>> Thunderbird, even when sending mail (I do not use seamonkey). Interestingly,
>> with exactly the same build, FF 48.0.2 seems to work perfectly. Maybe it has
>> something to do with the gtk3 patch, which was used in FF 48, and is not
>> applied with FF 49.
>>
>> (*) one sure way I can crash it is to write into the "search" field.
>>
>> Pierre
> 
> Thanks for the thoughts (I regard this as a *serious* problem) - but
> the patch (originally from fedora and also used at Arch) is no longer
> applied by them - it used to fix gtk+-3.20 changes, now a version is
> upstream.
> 
> My newest system is still 3.10 (I've got a newer LFS build, but I
> was expecting a new version of xorg-server would be coming soon (and
> I saw a comment the other month suggesting there is a new variant of
> one of the libs - I forget the details, but it looked like a new lib
> we don't use, possibly something like fonts2), so I've been deferring
> the xorg build.
> 
> On 7.10 with recent sqlite, nspr, nss it seems fine.  But I'm still
> on gtk+-3.20 and the old glib.
> 
> Which is your default engine for the Search field ?  Mine is
> wikipedia.
> 
> You say you have ff-48.0.2 working on the same build - is that a
> typo ?  I only see 48.0 in the releases. How about 49.0 ?

Thanks. I agree it is serious, but I am not sure I did not do something wrong...

It is not a typo : we had FF-48.0.2 in stable BLFS 7.10 (and actually, the
link in that page still works). But this reminds me that the base of this
build is stable LFS 7.10. Only BLFS is current SVN. I am not sure the changed
LFS packages versions are important for FF, though.

Let me try to get back to GTK-3.20 (I'm happy I use porg, so that I can
shuffle versions ;). Then I'll try previous sqlite, previous NSS, and previous
harfbuzz.

For the record, my DE/DM is LXDE/openbox. The engine for the search field is
google (the default). I tried to move to duckduckgo, but it opened the
duckduckgo page. I could search something, then it crashed (segfaulted) soon
after. I did not pursue further.

Regards,
Pierre
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Ken Moffat]

On Fri, Nov 11, 2016 at 10:59:28PM +0100, Pierre Labastie wrote:
> On 11/11/2016 20:01, Douglas R. Reno wrote:
> > Greetings,
> > 
> > I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
> > the latest SVN of BLFS.
> > 
> > They all segfault with references to libpthread (expected), libnspr4, 
> > libnss,
> > and libxul.
> > 
> > I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss), and
> > when I get home, I'll attempt to build a debug build to see if I can debug
> > this issue further.
> > 
> > Different applications give me different behavior:
> > 
> > Firefox will spuriously crash (probably background rendering... again)
> > Thunderbird will crash when trying to send mail
> > Seamonkey will crash before it finishes starting.
> > 
> > I'll try to investigate this later, but this is all I've been able to do 
> > for days.
> > 
> > 
> 
> Same issue here with Firefox (49.0.2) (*), but I have not seen anything with
> Thunderbird, even when sending mail (I do not use seamonkey). Interestingly,
> with exactly the same build, FF 48.0.2 seems to work perfectly. Maybe it has
> something to do with the gtk3 patch, which was used in FF 48, and is not
> applied with FF 49.
> 
> (*) one sure way I can crash it is to write into the "search" field.
> 
> Pierre

Thanks for the thoughts (I regard this as a *serious* problem) - but
the patch (originally from fedora and also used at Arch) is no longer
applied by them - it used to fix gtk+-3.20 changes, now a version is
upstream.

My newest system is still 3.10 (I've got a newer LFS build, but I
was expecting a new version of xorg-server would be coming soon (and
I saw a comment the other month suggesting there is a new variant of
one of the libs - I forget the details, but it looked like a new lib
we don't use, possibly something like fonts2), so I've been deferring
the xorg build.

On 7.10 with recent sqlite, nspr, nss it seems fine.  But I'm still
on gtk+-3.20 and the old glib.

Which is your default engine for the Search field ?  Mine is
wikipedia.

You say you have ff-48.0.2 working on the same build - is that a
typo ?  I only see 48.0 in the releases. How about 49.0 ?

ĸen
-- 
`I shall take my mountains', said Lu-Tze. `The climate will be good
for them.' -- Small Gods
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: [blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Pierre Labastie]

On 11/11/2016 20:01, Douglas R. Reno wrote:
> Greetings,
> 
> I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
> the latest SVN of BLFS.
> 
> They all segfault with references to libpthread (expected), libnspr4, libnss,
> and libxul.
> 
> I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss), and
> when I get home, I'll attempt to build a debug build to see if I can debug
> this issue further.
> 
> Different applications give me different behavior:
> 
> Firefox will spuriously crash (probably background rendering... again)
> Thunderbird will crash when trying to send mail
> Seamonkey will crash before it finishes starting.
> 
> I'll try to investigate this later, but this is all I've been able to do for 
> days.
> 
> 

Same issue here with Firefox (49.0.2) (*), but I have not seen anything with
Thunderbird, even when sending mail (I do not use seamonkey). Interestingly,
with exactly the same build, FF 48.0.2 seems to work perfectly. Maybe it has
something to do with the gtk3 patch, which was used in FF 48, and is not
applied with FF 49.

(*) one sure way I can crash it is to write into the "search" field.

Pierre
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

[blfs-dev] Seamonkey, Firefox, and Thunderbird all cause segfaults with the latest SVN build

[Douglas R. Reno]

Greetings,

I am unable to get Seamonkey, Firefox, or Thunderbird to work properly with
the latest SVN of BLFS.

They all segfault with references to libpthread (expected), libnspr4,
libnss, and libxul.

I've recompiled the entire NSS stack (p11-kit -> sqlite -> nspr -> nss),
and when I get home, I'll attempt to build a debug build to see if I can
debug this issue further.

Different applications give me different behavior:

Firefox will spuriously crash (probably background rendering... again)
Thunderbird will crash when trying to send mail
Seamonkey will crash before it finishes starting.

I'll try to investigate this later, but this is all I've been able to do
for days.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page