[petsc-dev] Memory Logging Problem

[Matthew Knepley]

On Tue, Mar 15, 2016 at 6:38 AM, Jed Brown <jed at jedbrown.org> wrote:

> Lisandro Dalcin <dalcinl at gmail.com> writes:
>
> > On 13 March 2016 at 04:21, Matthew Knepley <knepley at gmail.com> wrote:
> >> You can fix the SEGV by moving the library unloading to after the malloc
> >> dump. However,
> >> it then records that the memory from loading the library is still
> unfreed.
> >
> > Or let's do it the Python way and never ever dlclose() dynamic libraries
> :-)
>
> If I understand correctly, this would be no different from calling
> dlclose after dealing with the malloc_dump.  And I think it's the right
> solution because otherwise we have to copy __func__ instead of just
> referencing it.  What is the "memory from loading the library"?
>

The memory is the struct telling us what the libraries are to dclose. So
either we
would copy that to memory we do not log, or we ignore that memory in the
logging,
which is what I did.

   Matt

-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20160315/eec6454c/attachment.html>