Re: Strange memory corruption when using vector (works with deque)

[Reinhold Kainhofer]

Am Samstag, 21. November 2009 02:02:11 schrieben Sie:
> On Fri, 2009-11-20 at 18:35 +0100, Reinhold Kainhofer wrote:
> > While fixing bug 676, I ran into a very strange memory corruption
> > situation: Like in the tie-engraver.cc, I used vectors in
> > tie-performer.cc to store the moment together with the
> > Audio_element_info.
> >
> > Now, this always led to memory corruption as soon as I called push_back.
> > Before the push_back call, the size() of the vector was 0, after the
> > push_back, the size() would be reported as 3303820998 (instead of 1)?!?!?
> >
> > Using a deque instead of a vector works just fine.
> 
> Have you tried valgrind? 

Yes, of course (that cause my Springs.cc commit about uninitialized 
variables), but that hasn't been helpful at all...


> The memory corruption could be in a completely
> different part of the code. I have a suppression file lying around that
> eliminates most of the spurious guile warnings for valgrind if you want
> it. 

yes, that would be really great. The abundance of guile warnings doesn't make 
debugging easy...

Cheers,
Reinhold

-- 
------------------------------------------------------------------
Reinhold Kainhofer, address@hidden, http://reinhold.kainhofer.com/
 * Financial & Actuarial Math., Vienna Univ. of Technology, Austria
 * http://www.fam.tuwien.ac.at/, DVR: 0005886
 * LilyPond, Music typesetting, http://www.lilypond.org