Re: [FEniCS] Strange problem with failing unit test

Fri, 15 Nov 2013 09:18:30 -0800 

On 2013-11-15 15:41, Chris Richardson wrote:

This has just appeared today at
https://bugs.launchpad.net/ubuntu/+source/gcc-4.8/+bug/1246802
The new package fixes the nth_element problem. With the updated package, all tests pass for me. 

Hopefully it will move from 'proposed' to 'updates' soon.

Garth





Hello Garth, or anyone else affected,

Accepted gcc-4.8 into saucy-proposed. The package will build now and
be available at
http://launchpad.net/ubuntu/+source/gcc-4.8/4.8.1-10ubuntu9 in a few
hours, and then in the -proposed repository.

Please help us by testing this new package. See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation how
to enable and use -proposed. Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this
bug, mentioning the version of the package you tested, and change the
tag from verification-needed to verification-done. If it does not fix
the bug for you, please add a comment stating that, and change the tag
to verification-failed. In either case, details of your testing will
help us make a better decision.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you
in advance!



On 14/11/2013 12:26, Anders Logg wrote:

This might somehow be related to nth_element, but the real problem was
a bug in computing distances to 3D triangles which has been fixed now.

--
Anders


On Thu, Nov 14, 2013 at 08:13:07AM +0100, Johan Hake wrote:
On Wed, Nov 13, 2013 at 11:16 PM, Garth N. Wells <[email protected]> wrote:
Everyone could add some weight to the Ubuntu std::nth_element bug report:
https://bugs.launchpad.net/ubuntu/+source/gcc-4.8/+bug/1246802 

    Not sure if it will help.
Unfortunately I do not think it will help. We probably need to get the "heat" 
up. I am surprised it does not affect more people/projects.

Johan





    Garth

    On 13 Nov 2013, at 20:00, Anders Logg <[email protected]> wrote:

    > The plot thickens...
    >
> I see now that nth_element is used in more than one place. I'll take a 
    > look tomorrow and replace it by boost::nth_element throughout.
    >
    >
    >
    > On Wed, Nov 13, 2013 at 07:51:46PM +0100, Johan Hake wrote:
    >> After returning:
    >>
    >>  /usr/include/c++/4.8/bits/stl_algo.h
    >>
>> to its original state the unit test passed locally. At least now Anders 
    has
    >> something to look into ;)
    >>
    >> J
    >>
    >>
>> On Wed, Nov 13, 2013 at 7:35 PM, Johan Hake <[email protected]> wrote: 
    >>
>> I removed rounding-math, std=c++98, and used Release instead of 
    Developer
    >>    mode but I still get the same error.
    >>
    >>    I am on saucy
    >>
    >>    J
    >>
    >>
    >>    On Wed, Nov 13, 2013 at 6:28 PM, Jan Blechta <
    [email protected]>
    >>    wrote:
    >>
    >>        On Wed, 13 Nov 2013 18:07:58 +0100
    >>        Johan Hake <[email protected]> wrote:
    >>
>>> I have a patched stl library as suggested by Anders. Is that the 
    >>
    >>        I can't see such a suggestion in this thread.
    >>
    >>> common denominator here?
    >>
    >>        I don't know but my DOLFIN is compiled with
    >>
    >>          -O2 -g
    >>
    >>        while yours is possibly with
    >>
    >>          -O2 -g -std=c++98 -Wall -Werror -pedantic -pipe
    >>
>> because of Developer build type. (I abandoned Developer build 
    because
>> of some CGAL related errors.) Is it a possible cause? Should I 
    try with
    >>
    >>          -std=c++98
    >>
    >>        ?
    >>
    >>        Jan
    >>
    >>>
    >>> Johan
    >>>
    >>>
    >>> On Wed, Nov 13, 2013 at 5:50 PM, Jan Blechta
    >>> <[email protected]>wrote:
    >>>
    >>>> I can't reproduce it.
    >>>>
    >>>> Jan
    >>>>
    >>>>
    >>>> On Wed, 13 Nov 2013 14:56:37 +0100
    >>>> Anders Logg <[email protected]> wrote:
    >>>>
>>>>> The unit test geometry/BoundingBoxTree.py fails on my machines 
    >>>>> (with master) but seem to pass on the buildbots.
    >>>>>
    >>>>> The test in question fails with:
    >>>>>
    >>>>> Traceback (most recent call last):
    >>>>>  File "BoundingBoxTree.py", line 372, in
    >>>>>  test_compute_closest_entity_3d
    >>>>>    self.assertEqual(entity, reference[0])
    >>>>> AssertionError: 5 != 4
    >>>>>
>>>>> Can someone else check if this test fails on your machines as 
    >>>>> well?
    >>>>>
    >>>>
    >>>> _______________________________________________
    >>>> fenics mailing list
    >>>> [email protected]
    >>>> http://fenicsproject.org/mailman/listinfo/fenics
    >>>>
    >>
    >>
    >>
    >>
    >>
    >
    >> _______________________________________________
    >> fenics mailing list
    >> [email protected]
    >> http://fenicsproject.org/mailman/listinfo/fenics
    >
    > _______________________________________________
    > fenics mailing list
    > [email protected]
    > http://fenicsproject.org/mailman/listinfo/fenics




_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics

_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics

_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics

Reply via email to