dolfin team mailing list archive

[Johan Hake <hake.dev@xxxxxxxxx>]

I pushed a small "fix" for my problem. Out commented the failing assert.
I do not know enough about the code to know whether that is the correct
fix.

There are also several more code formating things that could be fixed.
If you are going to use underscore for a variable, like _A it should be
used for a private variables and not local variables.

Johan

On 04/04/2013 12:37 PM, Corrado Maurini wrote:
> I cannot reproduce the problem neither on a fresh installed version of dolfin on quantal using gcc4.7 and boost 1.51 or 1.49.
> 
> Not being able to test is difficult to fix. I tried to generally cleanup the code. May you try with the attached patch? 
> 
> Best,
> 
> Corrado
> 
> 
> 
> 
> 
> 
> Le 2 avr. 2013 à 13:26, Johan Hake a écrit :
> 
>> On 04/02/2013 12:06 PM, Corrado Maurini wrote:
>>> I confirm that trunk is working for me on gcc 4.2 (macosx lion) and 4.4 (rhel). 
>>> Unitest and example work fine for me, without any error message, both with TAO installed or not. 
>>>
>>> I need to install a quantal distribution to test on gcc 4.7. I need some more time for this. Not sure to be able to do it before thursday, but I will try.
>>
>> That is probably fine. Not many have TAO installed and I do not think we
>> have it on the buildbot yet.
>>
>>> Thanks for fixinfg the tab, I should change my test editor ... 
>>
>> :)
>>
>>> For the assertion in the unittest do you refer to this? 
>>> self.assertAlmostEqual(assemble(F), Ly, 5)
>>> or 
>>> try:
>>>  parameters["linear_algebra_backend"] = "PETSc"
>>> except RuntimeError:
>>>  import sys; sys.exit(0)
>>>
>>> The first is taken form KrylovSolver unittest, the second from PETScSnesSolver.
>>
>> The assertion that fail is in shared_ptr. It looks like the call to
>> b->vec() at line 173 in TAOLinearBoundSolver.cpp triggers it.
>>
>> Here is the backtrace:
>>
>> #0  0x00007ffff6f1a425 in __GI_raise (sig=<optimized out>) at
>> ../nptl/sysdeps/unix/sysv/linux/raise.c:64
>> #1  0x00007ffff6f1db8b in __GI_abort () at abort.c:91
>> #2  0x00007ffff6f130ee in __assert_fail_base (fmt=<optimized out>,
>> assertion=0x7fffeb9da24f "px != 0", file=
>>    0x7fffebaa14a0
>> "/home/hake/local/include/boost/smart_ptr/shared_ptr.hpp",
>> line=<optimized out>, function=<optimized out>) at assert.c:94
>> #3  0x00007ffff6f13192 in __GI___assert_fail (assertion=0x7fffeb9da24f
>> "px != 0", file=0x7fffebaa14a0
>> "/home/hake/local/include/boost/smart_ptr/shared_ptr.hpp", line=424,
>>    function=
>>    0x7fffebaa1a60
>> <_ZZNK5boost10shared_ptrIKN6dolfin11PETScMatrixEEptEvE19__PRETTY_FUNCTION__>
>> "T* boost::shared_ptr<T>::operator->() const [with T = const
>> dolfin::PETScMatrix]") at assert.c:103
>> #4  0x00007fffeb425aae in boost::shared_ptr<dolfin::PETScMatrix
>> const>::operator-> (this=<optimized out>) at
>> /home/hake/local/include/boost/smart_ptr/shared_ptr.hpp:424
>> #5  0x00007fffeb424995 in _M_dispose (__a=..., this=<optimized out>) at
>> /usr/include/c++/4.7/bits/basic_string.h:246
>> #6  ~basic_string (this=<optimized out>, __in_chrg=<optimized out>) at
>> /usr/include/c++/4.7/bits/basic_string.h:536
>> #7  dolfin::TAOLinearBoundSolver::solve (this=this@entry=0x1e13890,
>> A1=..., x=..., b1=..., xl=..., xu=...)
>>    at
>> /home/hake/bzr/fenics/dolfin/work-trunk/dolfin/nls/TAOLinearBoundSolver.cpp:173
>> #8  0x00007fffeb425055 in dolfin::TAOLinearBoundSolver::solve
>> (this=0x1e13890, A1=..., x=..., b1=..., xl=..., xu=...)
>>    at
>> /home/hake/bzr/fenics/dolfin/work-trunk/dolfin/nls/TAOLinearBoundSolver.cpp:157
>>
>> J
>>
>>
>>> Corrado
>>>
>>> Le 2 avr. 2013 à 11:03, Johan Hake a écrit :
>>>
>>>> I gcc 4.7. I think recent gcc is more strict on things like missing
>>>> return statements.
>>>>
>>>> Johan
>>>>
>>>> On 04/02/2013 10:47 AM, Corrado Maurini wrote:
>>>>> I am having a look. It was working for me ... on macosx, and rhel5
>>>>>
>>>>> Le 2 avr. 2013 à 09:06, Johan Hake a écrit :
>>>>>
>>>>>> On 04/02/2013 09:03 AM, noreply@xxxxxxxxxxxxx
>>>>>> <mailto:noreply@xxxxxxxxxxxxx> wrote:
>>>>>>> ------------------------------------------------------------
>>>>>>> revno: 7592
>>>>>>> committer: Johan Hake <hake.dev@xxxxxxxxx <mailto:hake.dev@xxxxxxxxx>>
>>>>>>> branch nick: work-trunk
>>>>>>> timestamp: Tue 2013-04-02 08:57:33 +0200
>>>>>>> message:
>>>>>>> Misc format fixes and compile fixes for TAO interface.
>>>>>>>  -- The TAO unittest throws an assertion need to be fixed.
>>>>>>>  -- I have used TAO version 1.2
>>>>>>
>>>>>> That should be TAO 2.1.
>>>>>>
>>>>>> Johan
>>>>>>
>>>>>>> modified:
>>>>>>> dolfin/nls/TAOLinearBoundSolver.cpp
>>>>>>> dolfin/nls/TAOLinearBoundSolver.h
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> lp:dolfin
>>>>>>> https://code.launchpad.net/~dolfin-core/dolfin/trunk
>>>>>>>
>>>>>>> Your team DOLFIN Core Team is subscribed to branch lp:dolfin.
>>>>>>> To unsubscribe from this branch go to
>>>>>>> https://code.launchpad.net/~dolfin-core/dolfin/trunk/+edit-subscription
>>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Mailing list: https://launchpad.net/~dolfin
>>>>>> Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
>>>>>> <mailto:dolfin@xxxxxxxxxxxxxxxxxxx>
>>>>>> Unsubscribe : https://launchpad.net/~dolfin
>>>>>> More help   : https://help.launchpad.net/ListHelp
>>>>>
>>>>
>>>
>>
>