On Tue, Jan 17, 2012 at 2:16 PM, John Cremona <john.crem...@gmail.com> wrote:
> On 17 January 2012 23:04, William Stein <wst...@gmail.com> wrote:
>> On Tue, Jan 17, 2012 at 1:47 PM, John Cremona <john.crem...@gmail.com> wrote:
>>>>> After building Sage on OS X 10.7, "make test" did this:
>>>>
>>>> For the record, running tests with "sage -t devel/sage/sage" yields
>>>> hundred(s) of failing files:
>>>>
>>>> http://wstein.org/home/wstein/tmp/test-sage-5.0.beta1-osx10.7.txt
>>>>
>>>> It could be that most of these boil down to some code at the core of
>>>> PARI (the bezout function) being miscompiled.
>>>>
>>>> There is a discussion about this from August 2011 here:
>>>>
>>>> http://pari.math.u-bordeaux.fr/archives/pari-dev-1108/msg00000.html
>>>>
>>>> It unfortunately appears based on the mailing list -- and searching
>>>> through the ** 5 months ** since then -- that nothing at all was done
>>>> to try to fix the problem, even though I had setup access to 10.7 for
>>>> the Pari developers. Probably the main problem is that the machine I
>>>> setup -- sqrt5.math.washington.edu -- is on the CS network, and for
>>>> some reason it keeps getting kicked off.
>>>>
>>>> This really sucks. Maybe I can rewrite their bezout to get around
>>>> the problem.
>>>
>>> William, I am in the same building as Karim and Bill for the rest of
>>> this week. Can you explain to me what the problem is (*) so I can
>>> talk to them? They have a Pari-Days all next week in Bordeaux and I
>>> may persuade them that this is worth them spending time on.
>>>
>>> (*) OK so I read through that thread. Maybe a personal reminder that
>>> this is still outstanding would help...
>>
>> Since this is a compiler bug, I'm not sure what is the best thing to
>> do at present.
>>
>> Let me run the full test suite of Sage first, having built PARI with
>> -O0 and see what happens. If that works, I will just need to figure
>> out how to build only bezout* with -O0 (or -O1). I'm guessing the
>> best thing the PARI devs could do at this point is report a bug to
>> Apple (maybe Justin Walker could file a bug report to Apple about
>> their compiler)? Also, suggestions about how to properly build only
>> *one* file with -O1 would be helpful.
>
> I would do that by adding a specific target into the Makefile for that
> one file with a custom compile command which would then be used
> instead of the default pattern to make a .o from a .c. (Several years
> ago I had one file which crashed a compiler at some optimazation
> level, so I did this to compile just that file at a lower level.)
>
> John
Thanks John -- good idea.
To follow up on my previous message, when I rebuilt pari with
SAGE_DEBUG="yes" (so -O0), then ran the test suite, only the following
tests fail:
sage -t devel/sage/sage/combinat/schubert_polynomial.py #
Killed/crashed
sage -t devel/sage/sage/combinat/tableau.py # Time out
sage -t devel/sage/sage/combinat/root_system/weyl_characters.py # Time
out
sage -t devel/sage/sage/combinat/sf/hall_littlewood.py # Time out
sage -t devel/sage/sage/combinat/sf/homogeneous.py # Time out
sage -t devel/sage/sage/combinat/sf/ns_macdonald.py # Time out
sage -t devel/sage/sage/combinat/sf/schur.py # Time out
sage -t devel/sage/sage/combinat/sf/sfa.py # Time out
sage -t devel/sage/sage/matrix/matrix_double_dense.pyx # 4 doctests
failed
sage -t devel/sage/sage/matrix/matrix2.pyx # 4 doctests failed
sage -t devel/sage/sage/libs/symmetrica/schur.pxi # Time out
sage -t devel/sage/sage/rings/polynomial/polynomial_element.pyx # 1
doctests failed
sage -t devel/sage/sage/rings/qqbar.py # Time out
sage -t devel/sage/sage/tests/interrupt.pyx # 13 doctests failed
The obvious next thing to do is try building symmetrica with less
optimization and trying the above tests.
-- William
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
