On Friday 10 August 2007 11:03:45 Klaas Gadeyne wrote:
> For more infomation about this bug, visit
> <https://svn.fmtc.be/bugzilla/orocos/show_bug.cgi?id=444>
>
> Klaas Gadeyne <klaas gadeyne [..] > changed:
>
> What |Removed |Added
> --------------------------------------------------------------------------
> CC| |klaas gadeyne [..]
>
>
>
> --- Comment #1 from Klaas Gadeyne <klaas gadeyne [..] > 2007-08-10
> 11:03:45 --- (In reply to comment #0)
>
> > While extending the matrixwrapper test I noticed that the resizing of
> > SymmetricMatrix causes some troubles when using BOOST.
> >
> > The test code I used is:
> > // SYMMETRICMATRIX
> > SymmetricMatrix Ks(r+2); Ks = v;
> > SymmetricMatrix Ks_resize(r); Ks_resize = v;
> > CPPUNIT_ASSERT_EQUAL(Ks.rows(), r+2);
> > CPPUNIT_ASSERT_EQUAL(Ks.columns(), r+2);
> > Ks.resize(r);
>
> What is v in this test program (a double variable, or a matrix?)
a double.

> > While executing the test for BOOST the following error is returned:
> >
> > Assertion failed in file /usr/include/boost/numeric/ublas/storage.hpp at
> > line 182:
> > i < size_
> > .E
> >
> > ##Failure Location unknown## : Error
> > Test name: MatrixwrapperTest::testMatrixwrapperValue
> > uncaught exception of type boost::numeric::ublas::bad_index
> > - bad index
> >
> > Failures !!!
> > Run: 1 Failure total: 1 Failures: 0 Errors: 1
> > _
> > (furhtermore ddd is unable to backtrace the problem)
>
> You didn't have a segfault, so you won't get a backtrace. You might
> put a breakpoint in the surroundings of the assertion though.
>
> > This operation is causing no troubles when using LTI.
> >
> > The code of the resizing in matrix_BOOST.cpp is as follows (very similar
> > to resizing of a normal Matrix which is causing no problems):
> > __
> > void
> > MySymmetricMatrix::resize(unsigned int i, bool copy, bool initialize)
> > {
> > BoostSymmetricMatrix & temp = (BoostSymmetricMatrix &) (*this);
> > temp.resize(i, copy);
> > }
> > _
> >
> > Any ideas?
>
> Try stepping down in the resize function?