Upgrade LinBox to version 1.1.7

[malb]

LinBox 1.1.7 was release in 2010, yet we are still running 1.1.6. It fixes a few issues which are relevant to Sage

• Install http://sage.math.washington.edu/home/malb/spkgs/linbox-1.1.7.spkg
• Apply attachment: trac_11718_linbox_1_1_7.patch

NOTE This SPKG re-enables the Commentator and hence might break on OSX 10.4! Needs testing.

CC: @kiwifb @cschwan @vbraun @ClementPernet @kcrisman

Component: packages: standard

Reviewer: Volker Braun, Martin Albrecht

Issue created by migration from https://trac.sagemath.org/ticket/11718

[malb]

Changed dependencies from 9511 to #9511

[malb]

Author: Martin Albrecht

[malb]

comment:5

Status update

LinBox's Sage interface changed and this causes major problems. The attached patch is supposed to address this new interface but I cannot successfully link against liblinboxsage.so any more.

I get

    846                 import matrix_complex_double_dense
    847                 return matrix_complex_double_dense.Matrix_complex_double_dense
--> 848             elif sage.rings.finite_rings.integer_mod_ring.is_IntegerModRing(R) and R.order() < matrix_modn_dense.MAX_MODULUS:
    849                 if R.order() == 2:
    850                     return matrix_mod2_dense.Matrix_mod2_dense

AttributeError: 'module' object has no attribute 'MAX_MODULUS'

However, this is not a Python problem. If I remove everything from linbox.pyx except for

cdef class Foo:
    pass

and remove all references to LinBox except for one cimport of it in matrix_modn_dense and I empty out linbox-sage.C in the LinBox !SPKG except for

#include <iostream>
#include <gmp.h>

#include <cstdlib>
#include <vector>
#include <list>

#include "linbox-sage.h"

#include <linbox/util/commentator.h>

#include <linbox/blackbox/sparse.h>

#include "linbox/element/givaro-polynomial.h"

#include <linbox/matrix/blas-matrix.h>
#include <linbox/matrix/sparse.h>
#include <linbox/vector/sparse.h>

#include <linbox/algorithms/blas-domain.h>
#include <linbox/algorithms/echelon-form.h>
#include "linbox/algorithms/gauss.h"
#include "linbox/algorithms/smith-form-adaptive.h"
#include "linbox/ffpack/ffpack.h"
#include <linbox/solutions/rank.h>          
#include <linbox/solutions/det.h>
#include <linbox/solutions/solve.h>

#include "linbox/solutions/methods.h"
#include <linbox/solutions/minpoly.h>
#include "linbox/algorithms/double-det.h"
#include <linbox/integer.h>

#include <linbox/field/gmp-rational.h>
#include <linbox/ring/givaro-polynomial.h>

I still get this problem. Hence, I figure it should have something to do with the linking. That is, I assume LinBox overwrites some symbol or corrupts some memory somehow which makes it impossible to link against it from Sage.

However, I am out of ideas and steam. Hence, I'll give up on fixing this for now.
But I will send an e-mail to the LinBox list.

PS: We also get the same UINT32_MAX error as #9511 on sage.math

[malb]

comment:7

The updated patch + SPKG seems to fix the issue, testing on sage.math now

[malb]

comment:8

The updated patch + SPKG seems to fix the issue, testing on sage.math now

[malb]

comment:9

Attachment: trac_11718_linbox_1_1_7.patch.gz

Well, we further but we don't reach the finish line:

sage -t -gdb "/home/malb/Sage/linbox/devel/sage/sage/matrix/matrix_modn_dense.pyx"
********************************************************************************
Type r at the (gdb) prompt to run the doctests.
Type bt if there is a crash to see a traceback.
********************************************************************************
GNU gdb (GDB) 7.2-debian
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-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /opt/sage-4.7.1-linbox/local/bin/python...done.
(gdb) r
Starting program: /opt/sage-4.7.1-linbox/local/bin/python /home/malb/.sage//tmp/.doctest_matrix_modn_dense.py
[Thread debugging using libthread_db enabled]

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff48bd545 in std::basic_ostream<char, std::char_traits<char> >::sentry::sentry(std::basic_ostream<char, std::char_traits<char> >&) () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#1  0x00007ffff48bdc59 in std::basic_ostream<char, std::char_traits<char> >& std::__ostream_insert<char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, char const*, long) ()
   from /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#2  0x00007ffff48be05f in std::basic_ostream<char, std::char_traits<char> >& std::operator<< <std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, char const*) ()
   from /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#3  0x00007fffdc91ef76 in ModularRandIter (F=..., frobeniusForm=..., N=3, A=0x4b56840, lda=3, c=30) at ../../linbox/randiter/modular.h:92
92                              commentator.report (10, INTERNAL_DESCRIPTION)

[kcrisman]

comment:11

Interestingly, a related change about "commentator" apparently was helpful in trying to get Sage to build on FreeBSD.

[kcrisman]

Changed keywords from none to freebsd

[nexttime]

comment:12

Just for the record:

There's a LinBox 1.1.6.p8 spkg with a couple of fixes at #12762 (currently still needing review).

[malb]

comment:13

Please close this ticket as a duplicate.

[vbraun]

Changed dependencies from #9511 to none

[vbraun]

Reviewer: Volker Braun

[vbraun]

Changed keywords from freebsd to none

[vbraun]

comment:15

Agreed

[jdemeyer]

Changed author from Martin Albrecht to none

[jdemeyer]

comment:16

Duplicate of what?

[jdemeyer]

Changed reviewer from Volker Braun to Volker Braun, Martin Albrecht

[malb]

comment:17

It's a duplicate of #12883