On 10/21/2012 01:33 AM, f...@star.sr.bham.ac.uk wrote:
>> On 10/21/2012 1:29 AM, f...@star.sr.bham.ac.uk wrote:
>>> Hello,
>>>
>>> After compiling GNU Octave 3.6.3 one of the tests fails with a
>>> segmentation fault and exits. After typing "make check" the last lines
>>> of
>>> the output are:
>>>
>>>
>>>     liboctave/Sparse.cc ....................................panic:
>>> Segmentation fault -- stopping myself...
>>> make[1]: *** [check] Segmentation fault (core dumped)
>>> make[1]: Leaving directory `/home/fabian/Downloads/octave-3.6.3/test'
>>> make: *** [check] Error 2
>>>
>>>
>>> I am using the following system:
>>>
>>> No LSB modules are available.
>>> Distributor ID: Ubuntu
>>> Description:    Ubuntu 12.04 LTS
>>> Release:        12.04
>>> Codename:       precise
>>>
>>> The file config.log is attached.
>>>
>>> Any help would be appreciated. Thank you very much in advance.
>>>
>>> Kind regards,
>>>
>>> Fabián
>>>
>>
>> which version of SuiteSparse library are you using ?
>>
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Everyone hates slow websites. So do we.
>> Make your web apps faster with AppDynamics
>> Download AppDynamics Lite for free today:
>> http://p.sf.net/sfu/appdyn_sfd2d_oct
>> _______________________________________________
>> Octave-dev mailing list
>> Octave-dev@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/octave-dev
>>
>
> Hello,
>
> Thank you for your reply.
>
> It seems version 2.2.3. The command "sudo apt-cache search SuiteSparse"
> yields the following packages:
>
> libsuitesparse-metis-3.1.0 - collection of libraries for computations for
> sparse matrices
> libsuitesparse-metis-dbg - collection of libraries for computations for
> sparse matrices
> libsuitesparse-metis-dev - collection of libraries for computations for
> sparse matrices
> libamd2.2.0 - approximate minimum degree ordering library for sparse matrices
> libbtf1.1.0 - permutation to block triangular form library for sparse
> matrices
> libcamd2.2.0 - symmetric approximate minimum degree library for sparse
> matrices
> libccolamd2.7.1 - constrained column approximate library for sparse matrices
> libcholmod1.7.1 - sparse Cholesky factorization library for sparse matrices
> libcolamd2.7.1 - column approximate minimum degree ordering library for
> sparse matrices
> libcsparse2.2.3 - concise sparse matrix library
> libcxsparse2.2.3 - concise sparse matrix library (complex, int and long
> int support)
> libklu1.1.0 - circuit simulation sparse LU factorization library
> libldl2.0.1 - simple LDL' factorization library for sparse matrices
> libsuitesparse-dbg - libraries for sparse matrices computations (debugging
> symbols)
> libsuitesparse-dev - libraries for sparse matrices computations
> (development files)
> libsuitesparse-doc - libraries for sparse matrices computations
> (documentation)
> libumfpack5.4.0 - sparse LU factorization library
> r-recommended - GNU R collection of recommended packages [metapackage]
>
> Should I update, which one?


Fabián

Before digging too deep into the libraries themselves, you may want to 
get the latest version of Octave from the repository.  3.6.3 is from a 
couple months ago and there may have been some code changes related to 
sparse matrices since then.

Here is the graph at the time of 3.6.3:

http://hg.savannah.gnu.org/hgweb/octave/graph/a95432e7309c

 From there you can move forward in the graph using the +60 button on 
the right side.  Then use the browser search for "sparse".  I'm seeing 
related changesets below.  Whether any of those contain a fix, I'm not 
sure.  If the problem persists, then write back and folks can help 
isolate the problem in the library or Octave.

Dan

# Fix Octave builds with SuiteSparse >=4.0 (bug #37031)6 weeks ago

# Ensure sparse constructors have valid ridx and data indices even if 
they are zero matrices (bug #36104)

# Special-case removing rows or columns from empty sparse matricesstable

# build: Don't directly compile #included template code Sparse.cc, 
MSparse.cc5 weeks ago

# configure.ac: Fix building of Octave with ancient versions of 
SuiteSparse (3.2.X)

# when writing sparse matrices to MAT files, use nzmax of at least 1 
(bug #36603)

# error out when attempting sparse minimum solution and CXSparse is 
unavailable

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_sfd2d_oct
_______________________________________________
Octave-dev mailing list
Octave-dev@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/octave-dev

Reply via email to