Sage 8.8.rc2 released
95 views
Skip to first unread message
Volker Braun
Jun 19, 2019, 5:54:06 PM6/19/19
to sage-release
As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html
8084b69595 (tag: 8.8.rc2, trac/develop) Updated SageMath version to 8.8.rc2
ab2ee449bb Trac #27961: openblas 0.3.6 vs. OS X on some processors
57e1ba71f9 Trac #27950: Replace source activate by conda activate in conda install guide
8df690fd2a (tag: 8.8.rc1) Updated SageMath version to 8.8.rc1
Paul Masson
Jun 19, 2019, 8:48:25 PM6/19/19
to sage-release
In an incremental build from beta 7 on macOX Mojave (10.14.5), I'm still encountering a crash-inducing issue with openblas. The relevant part of the crash log is
ImportError: dlopen(/Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/sage/matrix/matrix_rational_dense.so, 2): Library not loaded: /Users/Masson/Downloads/GitHub/sage/local/lib/libopenblas_haswellp-r0.3.5.dylib
Referenced from: /Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/sage/matrix/matrix_rational_dense.so
Reason: image not found
I can work around it by creating a second copy of libopenblas and changing its last digit to 5, but why would I need two copies of the library?
François Bissey
Jun 19, 2019, 8:52:19 PM6/19/19
to 'Justin C. Walker' via sage-release
You shouldn't need two copies. It is more like a dependency issue.
sage/matrix/matrix_rational_dense.so (and may be other) should be
re-built during the incremental build but isn’t/aren't.
Touching the relevant .pyx file and using `./sage -b` should fix it.
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> To post to this group, send email to sage-r...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-release.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/d04fa013-2143-4da5-a1c2-9036d1943fa4%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
sage/matrix/matrix_rational_dense.so (and may be other) should be
re-built during the incremental build but isn’t/aren't.
Touching the relevant .pyx file and using `./sage -b` should fix it.
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> To post to this group, send email to sage-r...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-release.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/d04fa013-2143-4da5-a1c2-9036d1943fa4%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
John H Palmieri
Jun 19, 2019, 9:25:30 PM6/19/19
to sage-release
This problem is being discussed at https://trac.sagemath.org/ticket/28008. As François says, you can touch matrix_rational_dense.pyx and do "sage -b", or you can do "sage -ba" to rebuild all of Sage's library files, in case other ones are also affected.
On Wednesday, June 19, 2019 at 5:52:19 PM UTC-7, François Bissey wrote:
You shouldn't need two copies. It is more like a dependency issue.
sage/matrix/matrix_rational_dense.so (and may be other) should be
re-built during the incremental build but isn’t/aren't.
Touching the relevant .pyx file and using `./sage -b` should fix it.
> On 20/06/2019, at 12:48 PM, Paul Masson <paulm...@comcast.net> wrote:
>
> In an incremental build from beta 7 on macOX Mojave (10.14.5), I'm still encountering a crash-inducing issue with openblas. The relevant part of the crash log is
>
> ImportError: dlopen(/Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/sage/matrix/matrix_rational_dense.so, 2): Library not loaded: /Users/Masson/Downloads/GitHub/sage/local/lib/libopenblas_haswellp-r0.3.5.dylib
> Referenced from: /Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/sage/matrix/matrix_rational_dense.so
> Reason: image not found
>
> I can work around it by creating a second copy of libopenblas and changing its last digit to 5, but why would I need two copies of the library?
>
> On Wednesday, June 19, 2019 at 2:54:06 PM UTC-7, Volker Braun wrote:
> As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html
>
>
> 8084b69595 (tag: 8.8.rc2, trac/develop) Updated SageMath version to 8.8.rc2
> ab2ee449bb Trac #27961: openblas 0.3.6 vs. OS X on some processors
> 57e1ba71f9 Trac #27950: Replace source activate by conda activate in conda install guide
> 8df690fd2a (tag: 8.8.rc1) Updated SageMath version to 8.8.rc1
>
>
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-r...@googlegroups.com.
Paul Masson
Jun 19, 2019, 9:41:49 PM6/19/19
to sage-release
That did indeed fix the issue. Thanks guys!
Eric Gourgoulhon
Jun 21, 2019, 5:10:01 AM6/21/19
to sage-release
On Ubuntu 18.04 running on bi-Xeon E5-2623 (8 cores) + 16 GB RAM:
- python2:
- incremental build (from 8.8.rc1, with -j16) completed, with Sphinx complaining during the doc build:
formatargspec() is now deprecated. Please use sphinx.util.inspect.Signature instead
- all tests from ptestlong passed
- python3:
- incremental build (from 8.8.rc1, with -j16) completed, with the same Sphinx deprecation warning as with python 2.
- tests from ptest-python3 passed, except for 2 doctests
in src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py:
sage -t --long --warn-long 51.8 src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py
**********************************************************************
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 1372, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H = HyperellipticCurve(t^5 + a*t)
Exception raised:
Traceback (most recent call last):
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1105, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function[7]>", line 1, in <module>
H = HyperellipticCurve(t**Integer(5) + a*t)
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/schemes/hyperelliptic_curves/constructor.py", line 220, in HyperellipticCurve
raise ValueError("In characteristic 2, argument h (= %s) must be non-zero."%h)
ValueError: In characteristic 2, argument h (= 0) must be non-zero.
**********************************************************************
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 1373, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H.zeta_function()
Expected:
(81*x^4 + 72*x^3 + 32*x^2 + 8*x + 1)/(9*x^2 - 10*x + 1)
Got:
(16*x^4 + 8*x^3 + x^2 + 2*x + 1)/(4*x^2 - 5*x + 1)
**********************************************************************
1 item had failures:
2 of 16 in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
[373 tests, 2 failures, 61.45 s]
**********************************************************************
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 1372, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H = HyperellipticCurve(t^5 + a*t)
Exception raised:
Traceback (most recent call last):
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1105, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function[7]>", line 1, in <module>
H = HyperellipticCurve(t**Integer(5) + a*t)
File "/home/eric/sage/py3/local/lib/python3.7/site-packages/sage/schemes/hyperelliptic_curves/constructor.py", line 220, in HyperellipticCurve
raise ValueError("In characteristic 2, argument h (= %s) must be non-zero."%h)
ValueError: In characteristic 2, argument h (= 0) must be non-zero.
**********************************************************************
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 1373, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H.zeta_function()
Expected:
(81*x^4 + 72*x^3 + 32*x^2 + 8*x + 1)/(9*x^2 - 10*x + 1)
Got:
(16*x^4 + 8*x^3 + x^2 + 2*x + 1)/(4*x^2 - 5*x + 1)
**********************************************************************
1 item had failures:
2 of 16 in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
[373 tests, 2 failures, 61.45 s]
Both doctests are passed when run standalone:
./sage -t --long --warn-long 51.8 src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py
Running doctests with ID 2019-06-21-11-12-55-4862dc1d.
Git branch: threejs_vector_field-27932
Using --optional=build,dochtml,memlimit,mpir,python2,sage
Doctesting 1 file.
sage -t --long --warn-long 51.8 src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py
[373 tests, 55.46 s]
----------------------------------------------------------------------
All tests passed!
----------------------------------------------------------------------
Total time for all tests: 55.6 seconds
cpu time: 55.5 seconds
cumulative wall time: 55.5 seconds
Running doctests with ID 2019-06-21-11-12-55-4862dc1d.
Git branch: threejs_vector_field-27932
Using --optional=build,dochtml,memlimit,mpir,python2,sage
Doctesting 1 file.
sage -t --long --warn-long 51.8 src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py
[373 tests, 55.46 s]
----------------------------------------------------------------------
All tests passed!
----------------------------------------------------------------------
Total time for all tests: 55.6 seconds
cpu time: 55.5 seconds
cumulative wall time: 55.5 seconds
Dima Pasechnik
Jun 21, 2019, 7:36:15 AM6/21/19
to sage-release
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py",
line 1373, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H.zeta_function()
Expected:
(81*x^4 + 72*x^3 + 32*x^2 + 8*x + 1)/(9*x^2 - 10*x + 1)
Got:
(16*x^4 + 8*x^3 + x^2 + 2*x + 1)/(4*x^2 - 5*x + 1)
looks like a Heisenbug to me. These are really different answers.
line 1373, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.zeta_function
Failed example:
H.zeta_function()
Expected:
(81*x^4 + 72*x^3 + 32*x^2 + 8*x + 1)/(9*x^2 - 10*x + 1)
Got:
(16*x^4 + 8*x^3 + x^2 + 2*x + 1)/(4*x^2 - 5*x + 1)
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To post to this group, send email to sage-r...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-release.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/96f32b96-d540-4035-b77f-c9e42e59b5ba%40googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-release.
Reply all
Reply to author
Forward
0 new messages