[gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

  Hello!
  After a recent software update on my amd64-based machine, I get the
following message trying to launch Sage:

------------------------------------------------------------
Unhandled SIGSEGV: A segmentation fault occurred in Sage.
This probably occurred because a *compiled* component
of Sage has a bug in it (typically accessing invalid memory)
or is not properly wrapped with _sig_on, _sig_off.
You might want to run Sage under gdb with 'sage -gdb' to debug this.
Sage will now terminate (sorry).
------------------------------------------------------------

And then it just exits. I have no idea where to search for the reasons
of such behaviour, so decided to ask here. :)

Thanks.
Vladimir

----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[François Bissey]

>   Hello!
>   After a recent software update on my amd64-based machine, I get the
> following message trying to launch Sage:
> 
> ------------------------------------------------------------
> Unhandled SIGSEGV: A segmentation fault occurred in Sage.
> This probably occurred because a *compiled* component
> of Sage has a bug in it (typically accessing invalid memory)
> or is not properly wrapped with _sig_on, _sig_off.
> You might want to run Sage under gdb with 'sage -gdb' to debug this.
> Sage will now terminate (sorry).
> ------------------------------------------------------------
> 
> And then it just exits. I have no idea where to search for the reasons
> of such behaviour, so decided to ask here. :)
> 
Hi Vladimir,

did this upgrade include anything from the sage-on-gentoo overlay?

Francois

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

  Hello!

On Tue, 26 Oct 2010 22:38:02 +1300
François Bissey <f.r.bissey@massey.ac.nz> wrote:
> 
> did this upgrade include anything from the sage-on-gentoo overlay?
> 
  Yes. It included:
- sci-libs/cddlib-094f-r2
- sci-mathematics/singular-3.1.1.4-r1
- dev-python/cvxopt-1.1.3
- sci-mathematics/sage-4.5.3
- dev-lang/python-2.6.6-r1

On Tue, 26 Oct 2010 11:45:31 +0200
Christopher Schwan <cschwan@students.uni-mainz.de> wrote:
> 
> I need a bit more of information - do you use the latest stable (4.5.3) or the 
> release candiate version ? 
    I use Sage-4.5.3 (the stable one, I guess).

> Is there any additional text printed which gives a hint ?
    No, there is no other messages. Only the one I posted at the
beginning.

  Regards,
  Vladimir.
  
----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[François Bissey]

>   Hello!
> 
> On Tue, 26 Oct 2010 22:38:02 +1300
> 
> François Bissey <f.r.bissey@massey.ac.nz> wrote:
> > did this upgrade include anything from the sage-on-gentoo overlay?
> 
>   Yes. It included:
> - sci-libs/cddlib-094f-r2
> - sci-mathematics/singular-3.1.1.4-r1
> - dev-python/cvxopt-1.1.3
> - sci-mathematics/sage-4.5.3
> - dev-lang/python-2.6.6-r1
> 
> On Tue, 26 Oct 2010 11:45:31 +0200
> 
> Christopher Schwan <cschwan@students.uni-mainz.de> wrote:
> > I need a bit more of information - do you use the latest stable (4.5.3)
> > or the release candiate version ?
> 
>     I use Sage-4.5.3 (the stable one, I guess).
> 
> > Is there any additional text printed which gives a hint ?
> 
>     No, there is no other messages. Only the one I posted at the
> beginning.
> 
Hi Vladimir,

I believe you may be the first person to try out python 2.6.6 with sage.
You may need to rebuild cython-0.12.1 (don't use 0.13) possibly numpy, scipy, 
pexpect, ipython, matplotlib, mpmath, sympy, pynac, networx, sphinx and then 
rebuild sage-clib and sage itself again. There may be more but any one of 
these could be at fault.

Francois

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

On Wed, 27 Oct 2010 08:45:00 +1300
François Bissey <f.r.bissey@massey.ac.nz> wrote:
> 
> I believe you may be the first person to try out python 2.6.6 with sage.
> You may need to rebuild cython-0.12.1 (don't use 0.13) possibly numpy, scipy, 
> pexpect, ipython, matplotlib, mpmath, sympy, pynac, networx, sphinx and then 
> rebuild sage-clib and sage itself again. There may be more but any one of 
> these could be at fault.
> 
  I have rebuilt Cython-0.12.1 today just before starting this thread.
So, this is not because of Cython, I guess. I'll try to rebuild the
packages you mentioned one by one and find the one causing the
segfault. And then report back here.

Regards,
Vladimir

----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Christopher Schwan]

Hi,

On Tuesday 26 October 2010 22:14:12 v_2e@ukr.net wrote:
> On Wed, 27 Oct 2010 08:45:00 +1300
> 
> François Bissey <f.r.bissey@massey.ac.nz> wrote:
> > I believe you may be the first person to try out python 2.6.6 with sage.
> > You may need to rebuild cython-0.12.1 (don't use 0.13) possibly numpy,
> > scipy, pexpect, ipython, matplotlib, mpmath, sympy, pynac, networx,
> > sphinx and then rebuild sage-clib and sage itself again. There may be
> > more but any one of these could be at fault.
> 
>   I have rebuilt Cython-0.12.1 today just before starting this thread.
> So, this is not because of Cython, I guess. I'll try to rebuild the
> packages you mentioned one by one and find the one causing the
> segfault. And then report back here.

If you do not find anything, please try the following:

    USE=debug emerge -1 sage-baselayout

Also, make sure that the GNU Debugger is installed (emerge gdb). You may now 
start Sage with

   sage -gdb

which spawns Sage in a gdb session. Sage will crash just as before, but you 
may retrieve a backtrace via typing

    bt

That output would be interesting and may directly lead us to the root of the 
problem.

> 
> Regards,
> Vladimir
> 
> -----
>  <v_2e@ukr.net>

Cheers,

Christopher

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

  Hello!

On Wed, 27 Oct 2010 10:07:25 +0200
Christopher Schwan <cschwan@students.uni-mainz.de> wrote:
> 
> If you do not find anything, please try the following:
> 
>     USE=debug emerge -1 sage-baselayout
> 
> Also, make sure that the GNU Debugger is installed (emerge gdb). You may now 
> start Sage with
> 
>    sage -gdb
> 
> which spawns Sage in a gdb session. Sage will crash just as before, but you 
> may retrieve a backtrace via typing
> 
>     bt
> 
> That output would be interesting and may directly lead us to the root of the 
> problem.
> 
  I have re-emerged 
- Cython-0.12.1
- numpy
- scipy
- pexpect
- ipython
- matplotlib
- mpmath
- sympy
- pynac
- networx
-sphinx 
and rebuild sage-clib and sage after that. But it didn't help.

Then I run 'sage -gdb' and here is the output:
########## Output starts ##########
----------------------------------------------------------------------
| Sage Version 4.5.3, Release Date: 2010-09-04                       |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------
/usr/bin/sage-ipython
GNU gdb (Gentoo 7.2 p1) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html> This is free software: you are free
to change and redistribute it. There is NO WARRANTY, to the extent
permitted by law.  Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-pc-linux-gnu". For bug reporting
instructions, please see: <http://bugs.gentoo.org/>...
Reading symbols from /usr/bin/python...(no debugging symbols
found)...done. process 21451 is executing new
program: /usr/bin/python2.6 [Thread debugging using libthread_db
enabled] Python 2.6.6 (r266:84292, Oct 23 2010, 17:23:31) 
[GCC 4.4.4] on linux2
Type "help", "copyright", "credits" or "license" for more information.

Program received signal SIGSEGV, Segmentation fault.
0x00007fffe9b7c274 in __cxa_allocate_exception ()
from /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.5/libstdc++.so.6 (gdb) bt
#0  0x00007fffe9b7c274 in __cxa_allocate_exception ()
   from /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.5/libstdc++.so.6
#1  0x00007fffcabb3037 in GiNaC::function::find_function(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, unsigned int) ()
   from /usr/lib/libpynac-0.2.so.0
#2  0x00007fffca3cdb16 in ?? ()
   from /usr/lib64/python2.6/site-packages/sage/symbolic/function.so
#3  0x00007fffca3cc780 in ?? ()
   from /usr/lib64/python2.6/site-packages/sage/symbolic/function.so
#4  0x00007ffff7ad384c in ?? () from /usr/lib/libpython2.6.so.1.0
#5  0x00007ffff7a85f92 in PyObject_Call () from /usr/lib/libpython2.6.so.1.0
#6  0x00007ffff7b10b42 in PyEval_CallObjectWithKeywords ()
   from /usr/lib/libpython2.6.so.1.0
#7  0x00007ffff7a9d499 in ?? () from /usr/lib/libpython2.6.so.1.0
#8  0x00007ffff7a85f92 in PyObject_Call () from /usr/lib/libpython2.6.so.1.0
#9  0x00007fffca3c327f in ?? ()
   from /usr/lib64/python2.6/site-packages/sage/symbolic/function.so
#10 0x00007ffff7ad384c in ?? () from /usr/lib/libpython2.6.so.1.0
#11 0x00007ffff7a85f92 in PyObject_Call () from /usr/lib/libpython2.6.so.1.0
#12 0x00007ffff7b10b42 in PyEval_CallObjectWithKeywords ()
   from /usr/lib/libpython2.6.so.1.0
#13 0x00007ffff7a9d499 in ?? () from /usr/lib/libpython2.6.so.1.0
#14 0x00007ffff7a85f92 in PyObject_Call () from /usr/lib/libpython2.6.so.1.0
#15 0x00007ffff7b15fb4 in PyEval_EvalFrameEx ()
---Type <return> to continue, or q <return> to quit---q
 from /usr/lib/libpython2.6.so.1.Quit
(gdb) quit
A debugging session is active.

	Inferior 1 [process 21451] will be killed.

Quit anyway? (y or n) y
########### Output ends ###########

----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[François Bissey]

thanks for the debugging trace Vladimir.
It looks like the problem is in pynac, it may be a compatibility
problem with python 2.6.6. You have gone in uncharted territories
with this version of python. Christopher knows more about pynac
than me, he may be able to do something about it.

Thanks for the report.

Francois
 

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

  Hello!

On Thu, 28 Oct 2010 11:08:28 +1300
François Bissey <f.r.bissey@massey.ac.nz> wrote:

> thanks for the debugging trace Vladimir.
> It looks like the problem is in pynac, it may be a compatibility
> problem with python 2.6.6. You have gone in uncharted territories
> with this version of python. Christopher knows more about pynac
> than me, he may be able to do something about it.
> 
> Thanks for the report.
> 
> Francois
>  
  I just wanted to report that Sage-4.6-r1 shows exactly the same
behaviour with python-2.6.6-r1.

----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Christopher Schwan]

Hi Vladimir,

On Thursday 11 November 2010 00:00:20 v_2e@ukr.net wrote:
>   Hello!
> 
> On Thu, 28 Oct 2010 11:08:28 +1300
> 
> François Bissey <f.r.bissey@massey.ac.nz> wrote:
> > thanks for the debugging trace Vladimir.
> > It looks like the problem is in pynac, it may be a compatibility
> > problem with python 2.6.6. You have gone in uncharted territories
> > with this version of python. Christopher knows more about pynac
> > than me, he may be able to do something about it.
> > 
> > Thanks for the report.
> > 
> > Francois
> 
>   I just wanted to report that Sage-4.6-r1 shows exactly the same
> behaviour with python-2.6.6-r1.

thanks for notifying us - I hope I am able to find the flaw. Please send me your 
output of "emerge --info" !

> 
> -----
>  <v_2e@ukr.net>

Cheers,

Christopher

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Jari_42]

Hi,

Just to report that I can reproduce this exact error on my amd64 gentoo
install with Sage-4.6-r1 and python-2.6.6-r1.

Thanks,
Jari

On Thursday 11 November 2010 00:00:20 v_2e@ukr.net wrote:
>   Hello!
> 
> On Thu, 28 Oct 2010 11:08:28 +1300
> 
> François Bissey <f.r.bissey@massey.ac.nz> wrote:
> > thanks for the debugging trace Vladimir.
> > It looks like the problem is in pynac, it may be a compatibility
> > problem with python 2.6.6. You have gone in uncharted territories
> > with this version of python. Christopher knows more about pynac
> > than me, he may be able to do something about it.
> > 
> > Thanks for the report.
> > 
> > Francois
> 
>   I just wanted to report that Sage-4.6-r1 shows exactly the same
> behaviour with python-2.6.6-r1.
> 
> -----
>  <v_2e@ukr.net>

-- 
View this message in context: http://old.nabble.com/-sage-on-gentoo--Sage-4.5.3-segmentation-fault-tp30055784p30308916.html
Sent from the gentoo-science mailing list archive at Nabble.com.

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Christopher Schwan]

Hi,

On Friday 26 November 2010 12:32:35 Jari_42 wrote:
> Hi,
> 
> Just to report that I can reproduce this exact error on my amd64 gentoo
> install with Sage-4.6-r1 and python-2.6.6-r1.

Thanks for reporting! So it looks like this is not problem specific to just one 
computer - if there is anybody else who has the same problem, please write a 
mail to me (you do not need to mail to the list) and attach a file with the 
emerge --info output. I would also like to have emerge --info from Jari - 
hopefully I can find a difference to my system.

Unfortunately, I can not reproduce this error on amd64. This weekend I 
upgraded Sage in a prefixed environment to 4.6 but that went fine. After 
upgrading gcc to 4.5.1 I recompiled my entire prefix (emerge -e @world), but I 
still can not reproduce it.

For the time being, I can only recommend these general instructions:

1. make sure everything is up-to-date
2. make sure revdep-rebuild reports no package

If you have spare cpu time you may recompile world (this sounds really crazy, 
but I am running out of options ;) ) or at least Sage and it dependencies:

    emerge -C sage
    emerge --depclean -a
    emerge sage

This would make sure you do not have unintential old packages installed (this 
may have happened if I had done major changes to ebuilds without upgrading 
their revision number - this "should" not be the case, of course).

Christopher

> 
> Thanks,
> Jari
> 
> On Thursday 11 November 2010 00:00:20 v_2e@ukr.net wrote:
> >   Hello!
> > 
> > On Thu, 28 Oct 2010 11:08:28 +1300
> > 
> > François Bissey <f.r.bissey@massey.ac.nz> wrote:
> > > thanks for the debugging trace Vladimir.
> > > It looks like the problem is in pynac, it may be a compatibility
> > > problem with python 2.6.6. You have gone in uncharted territories
> > > with this version of python. Christopher knows more about pynac
> > > than me, he may be able to do something about it.
> > > 
> > > Thanks for the report.
> > > 
> > > Francois
> > > 
> >   I just wanted to report that Sage-4.6-r1 shows exactly the same
> > 
> > behaviour with python-2.6.6-r1.
> > 
> > -----
> > 
> >  <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[v_2e]

  Hello!
  I have migrated to Python-2.7.1 yesterday (by the way, I had to
rename my 
'/var/lib/layman/sage-on-gentoo/dev-lang/python/python-2.7.ebuild'
file to 
'/var/lib/layman/sage-on-gentoo/dev-lang/python/python-2.7.1.ebuild'
since there is Python-2.7.1 in the main Portage tree already.)
  And Sage shows the same behaviour as before. The same segmentation
fault.


----- 
 <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Christopher Schwan]

Hi,

On Wednesday 01 December 2010 11:14:29 v_2e@ukr.net wrote:
>   Hello!
>   I have migrated to Python-2.7.1 yesterday (by the way, I had to
> rename my
> '/var/lib/layman/sage-on-gentoo/dev-lang/python/python-2.7.ebuild'
> file to
> '/var/lib/layman/sage-on-gentoo/dev-lang/python/python-2.7.1.ebuild'
> since there is Python-2.7.1 in the main Portage tree already.)

Thanks for that info - I removed the 2.7 from our overlay because it isnt 
supported (not tested).

>   And Sage shows the same behaviour as before. The same segmentation
> fault.

Thats is as expected since Sage selects python to be 2.6.

> 
> 
> -----
>  <v_2e@ukr.net>

Re: [gentoo-science] [sage-on-gentoo] Sage-4.5.3 segmentation fault

[Christopher Schwan]

Hi!

On Tuesday 26 October 2010 11:32:23 v_2e@ukr.net wrote:
>   Hello!
>   After a recent software update on my amd64-based machine, I get the
> following message trying to launch Sage:
> 
> ------------------------------------------------------------
> Unhandled SIGSEGV: A segmentation fault occurred in Sage.
> This probably occurred because a *compiled* component
> of Sage has a bug in it (typically accessing invalid memory)
> or is not properly wrapped with _sig_on, _sig_off.
> You might want to run Sage under gdb with 'sage -gdb' to debug this.
> Sage will now terminate (sorry).
> ------------------------------------------------------------
> 
> And then it just exits. I have no idea where to search for the reasons
> of such behaviour, so decided to ask here. :)

I need a bit more of information - do you use the latest stable (4.5.3) or the 
release candiate version ? Is there any additional text printed which gives a 
hint ?

> 
> Thanks.
> Vladimir
> 
> -----
>  <v_2e@ukr.net>

Cheers,

Christopher