We just tried repeating this (deleting peaks both using the "delete"
key on the keyboard and also via the "delete" button in the selected
crosspeak list dialog) and it didn't crash, so hard to know what the
problem is. When the program seg faults it is almost guaranteed to be a
problem in the C world rather than Python world (the latter does
occasionally wipe out but rarely). Peak information is stored in both the
C world and Python world (so that the entire information does not have to
be passed down every draw) and when a peak is deleted in the Python world
the C world is told to delete it from the relevant peak list. There could
be a bug there but one would have thought we would see that here pretty
quickly. If a peak is deleted then a redraw is also forced, and if that
redraw happened before the peak was deleted from the list it's possible
something wierd could happen (not quite sure what). However the way the
code is set up the redraw should happen after the peak is deleted from the
list. Of course "should" is a dangerous word.
Did you compile the C code with -g? Otherwise I would guess the core dump
would not say anything (although perhaps gnu turns it on by default?).
Certainly worth looking at just in case.
Wayne
On Thu, 26 Feb 2004, Brian Smith wrote:
> Hi,
>
> OK, so now I know how to pick crosspeaks - now I seem to be able
> to get the program to crash & core dump quite a lot! e.g. deleting a
> crosspeak from the selected crosspeak list caused a
>
> /usr/local/ccpnmr/bin/analysis: line 1: 4576 Segmentation fault (core dumped) python -i -O $CCPNMR_TOP_DIR/ccpnmr1.0/python/ccpnmr/analysis/AnalysisGui.py $1
>
> Any way I can look at my core dump to give you more useful feedback on
> this?
>
> Brian
>
> --
> Dr. Brian O. Smith ---------------------- B.Smith at bio.gla.ac.uk
> Division of Biochemistry & Molecular Biology,
> Institute Biomedical & Life Sciences,
> Joseph Black Building, University of Glasgow, Glasgow G12 8QQ, UK.
> Tel: 0141 330 5167/6459 Fax: 0141 330 8640
>
|