#9723: Sage will not start on 64-bit Solaris 10 - M4RI issue ??
------------------------+---------------------------------------------------
Reporter: drkirkby | Owner: drkirkby
Type: defect | Status: new
Priority: major | Milestone: sage-4.5.3
Component: solaris | Keywords:
Author: | Upstream: N/A
Reviewer: | Merged:
Work_issues: |
------------------------+---------------------------------------------------
Changes (by drkirkby):
* cc: mpatel, ddrake (added)
Old description:
> == Hardware/Software ==
> * [http://www.oracle.com/us/products/servers-storage/servers/sparc-
> enterprise/t-series/031584.htm Sun T5240]
> * [http://www.oracle.com/us/products/servers-
> storage/microelectronics/031459.htm T2 PLUS] processors running at 1167
> MHz. (16 cores and 128 hardware threads in total).
> * 32 GB RAM
> * No swap space
> * gcc 4.4.1 configured to use the Sun linker and assembler.
> * Sage 4.5.3.alpha0, with the following four .spkg files, which are
> necessary to allow Sage to build on 64-bit on Solaris/!OpenSolaris on
> both SPARC and x64.
> * ECL - #9643. This was not necessary for 64-bit SPARC, but it would
> have been for 64-bit Solaris or !OpenSolaris x86 builds, and was included
> in this build.
> * ATLAS #9508
> * Singular #9397 (Despite the ticket's says !OpenSolaris, there is a
> critical 64-bit fix for Solaris 10 on SPARC too. Many of the problems
> first observed on a 64-bit !OpenSolaris port also affect the 64-bit
> Solaris 10 SPARC port).
> * zn_poly #9358 (Again, despite what the ticket's title is, this is
> also another fix which is essential for Sage to build 64-bit on Solaris
> 10 SPARC)
>
> == Background ==
> Although a fully stable copy of Sage on 64-bit Solaris 10 SPARC has never
> been built, Sage has at least built and able to do computations. See for
> example [http://groups.google.co.uk/group/sage-devel/msg/63dd0f2d01bcfe4e
> factoring 323232323923321 on 64-bit Solaris 10 SPARC].
>
> As shown above, Sage 4.5.0 could be made to work, though I believe more
> recent versions have worked too. However, version earlier needed a few
> more new .spkg files, as some such as Pynac (#7861), libfpll (#7864),
> Numpy (#8086) etc were only merged in 4.5.3.alpha0.
>
> I've previously built Sage 64-bit on SPARC on both
> {{{t2.math.washington.edu}}} and on one of my own Solaris 10 SPARC
> systems - either my Sun Blade 1000 or my Sun Blade 2000 - I can not
> recall which).
>
> == The problem ==
>
> 4.5.3.alpha0, with the four .spkg files mentioned above, builds on
> Solaris 10 SPARC. The file install.log shows:
>
> {{{
> To install gap, gp, singular, etc., scripts
> in a standard bin directory, start sage and
> type something like
> sage: install_scripts('/usr/local/bin')
> at the Sage command prompt.
>
> To build the documentation, run
> make doc
>
> Sage build/upgrade complete!
> }}}
>
> However, the build fails to start at all on {{{t2.math.washington.edu}}},
> whereas a previous Sage was just about usable, though it was unstable.
> Instead 4.5.3.alpha0 fails with:
>
> {{{
> kir...@t2:64 ~/t2/64/sage-4.5.3.alpha0$ ./sage
> ----------------------------------------------------------------------
> | Sage Version 4.5.3.alpha0, Release Date: 2010-08-09 |
> | Type notebook() for the GUI, and license() for information. |
> ----------------------------------------------------------------------
> **********************************************************************
> * *
> * Warning: this is a prerelease version, and it may be unstable. *
> * *
> **********************************************************************
> ---------------------------------------------------------------------------
> ImportError Traceback (most recent call
> last)
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/IPython/ipmaker.pyc in force_import(modname)
> 64 reload(sys.modules[modname])
> 65 else:
> ---> 66 __import__(modname)
> 67
> 68
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/ipy_profile_sage.py
> in <module>()
> 5 preparser(True)
> 6
> ----> 7 import sage.all_cmdline
> 8 sage.all_cmdline._init_cmdline(globals())
> 9
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/sage/all_cmdline.py in <module>()
> 12 try:
> 13
> ---> 14 from sage.all import *
> 15 from sage.calculus.predefined import x
> 16 preparser(on=True)
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/sage/all.py in <module>()
> 71
> 72 from sage.rings.all import *
> ---> 73 from sage.matrix.all import *
> 74
> 75 # This must come before Calculus -- it initializes the Pynac
> library.
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/sage/matrix/all.py in <module>()
> ----> 1
> 2
> 3 from matrix_space import MatrixSpace, is_MatrixSpace
> 4 from constructor import matrix, Matrix, random_matrix,
> diagonal_matrix, identity_matrix, block_matrix, block_diagonal_matrix,
> jordan_block, zero_matrix
> 5 from matrix import is_Matrix
> 6 from berlekamp_massey import berlekamp_massey
> 7
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/sage/matrix/matrix_space.pyc in <module>()
> 35 import matrix_generic_sparse
> 36
> ---> 37 import matrix_modn_dense
> 38 import matrix_modn_sparse
> 39
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/matrix_integer_dense.pxd
> in init sage.matrix.matrix_modn_dense
> (sage/matrix/matrix_modn_dense.c:14829)()
>
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/matrix_mod2_dense.pxd
> in init sage.matrix.matrix_integer_dense
> (sage/matrix/matrix_integer_dense.c:39010)()
>
> ImportError: ld.so.1: python: fatal: relocation error: file
> /rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
> packages/sage/matrix/matrix_mod2_dense.so: symbol mzd_lqup: referenced
> symbol not found
> Error importing ipy_profile_sage - perhaps you should run %upgrade?
> WARNING: Loading of ipy_profile_sage failed.
> }}}
> == Likely cause ==
> mzd_lqup issues reported have been part of M4RI problems, but M4RI has
> not been updated recently to my knowledge.
New description:
== Hardware/Software ==
* [http://www.oracle.com/us/products/servers-storage/servers/sparc-
enterprise/t-series/031584.htm Sun T5240]
* [http://www.oracle.com/us/products/servers-
storage/microelectronics/031459.htm T2 PLUS] processors running at 1167
MHz. (16 cores and 128 hardware threads in total).
* 32 GB RAM
* No swap space
* gcc 4.4.1 configured to use the Sun linker and assembler.
* Sage 4.5.3.alpha0, with the following four .spkg files, which are
necessary to allow Sage to build on 64-bit on Solaris/!OpenSolaris on both
SPARC and x64.
* ECL - #9643. This was not necessary for 64-bit SPARC, but it would
have been necessary for 64-bit Solaris x86 or 64-bit !OpenSolaris, and was
included in this build.
* ATLAS #9508
* Singular #9397 (Despite the ticket's says !OpenSolaris, this is a
critical 64-bit fix for Solaris 10 on SPARC too. Many of the problems
first observed on a 64-bit !OpenSolaris port also affect the 64-bit
Solaris 10 SPARC port).
* zn_poly #9358 (Again, despite what the ticket's title is, this is
also another fix which is essential for Sage to build 64-bit on Solaris 10
SPARC)
== Background ==
Although a fully stable copy of Sage on 64-bit Solaris 10 SPARC has never
been built, Sage has at least built and able to do computations. See for
example [http://groups.google.co.uk/group/sage-devel/msg/63dd0f2d01bcfe4e
factoring 323232323923321 on 64-bit Solaris 10 SPARC].
As shown above, Sage 4.5.0 could be made to work, though I believe more
recent versions have worked too. However, version earlier needed a few
more new .spkg files, as some such as Pynac (#7861), libfpll (#7864),
Numpy (#8086) etc were only merged in 4.5.3.alpha0.
I've previously built Sage 64-bit on SPARC on both
{{{t2.math.washington.edu}}} and on one of my own Solaris 10 SPARC systems
- either my Sun Blade 1000 or my Sun Blade 2000 - I can not recall which).
== The problem ==
4.5.3.alpha0, with the four .spkg files mentioned above, builds on Solaris
10 SPARC. The file install.log shows:
{{{
To install gap, gp, singular, etc., scripts
in a standard bin directory, start sage and
type something like
sage: install_scripts('/usr/local/bin')
at the Sage command prompt.
To build the documentation, run
make doc
Sage build/upgrade complete!
}}}
However, the build fails to start at all on {{{t2.math.washington.edu}}},
whereas a previous Sage was just about usable, though it was unstable.
Instead 4.5.3.alpha0 fails with:
{{{
kir...@t2:64 ~/t2/64/sage-4.5.3.alpha0$ ./sage
----------------------------------------------------------------------
| Sage Version 4.5.3.alpha0, Release Date: 2010-08-09 |
| Type notebook() for the GUI, and license() for information. |
----------------------------------------------------------------------
**********************************************************************
* *
* Warning: this is a prerelease version, and it may be unstable. *
* *
**********************************************************************
---------------------------------------------------------------------------
ImportError Traceback (most recent call
last)
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/IPython/ipmaker.pyc in force_import(modname)
64 reload(sys.modules[modname])
65 else:
---> 66 __import__(modname)
67
68
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/ipy_profile_sage.py
in <module>()
5 preparser(True)
6
----> 7 import sage.all_cmdline
8 sage.all_cmdline._init_cmdline(globals())
9
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/sage/all_cmdline.py in <module>()
12 try:
13
---> 14 from sage.all import *
15 from sage.calculus.predefined import x
16 preparser(on=True)
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/sage/all.py in <module>()
71
72 from sage.rings.all import *
---> 73 from sage.matrix.all import *
74
75 # This must come before Calculus -- it initializes the Pynac
library.
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/sage/matrix/all.py in <module>()
----> 1
2
3 from matrix_space import MatrixSpace, is_MatrixSpace
4 from constructor import matrix, Matrix, random_matrix,
diagonal_matrix, identity_matrix, block_matrix, block_diagonal_matrix,
jordan_block, zero_matrix
5 from matrix import is_Matrix
6 from berlekamp_massey import berlekamp_massey
7
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/sage/matrix/matrix_space.pyc in <module>()
35 import matrix_generic_sparse
36
---> 37 import matrix_modn_dense
38 import matrix_modn_sparse
39
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/matrix_integer_dense.pxd
in init sage.matrix.matrix_modn_dense
(sage/matrix/matrix_modn_dense.c:14829)()
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/bin/matrix_mod2_dense.pxd
in init sage.matrix.matrix_integer_dense
(sage/matrix/matrix_integer_dense.c:39010)()
ImportError: ld.so.1: python: fatal: relocation error: file
/rootpool2/local/kirkby/t2/64/sage-4.5.3.alpha0/local/lib/python2.6/site-
packages/sage/matrix/matrix_mod2_dense.so: symbol mzd_lqup: referenced
symbol not found
Error importing ipy_profile_sage - perhaps you should run %upgrade?
WARNING: Loading of ipy_profile_sage failed.
}}}
== Likely cause ==
mzd_lqup issues reported have been part of M4RI problems, but M4RI has not
been updated recently to my knowledge.
--
Comment:
I'm cc'ing the release managers, who might know if the library patch got
merged by mistake, without the library, or what else might have caused
this.
Dave
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/9723#comment:6>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sage-trac?hl=en.
