Hi,
On 29/08/13 14:31, Miles Rout wrote:
Can you reproduce the bug yourself on a local copy you've compiled
with debugging symbols and no optimisation?
yes - good suggestion. This is the normal approach. Until you get to
where I am right now,
where the debug version on my development system does not crash, but the
release version on the
customer's site does crash.
It is a multithreaded program - 100 concurrent threads is normal. This
runs a bit differently on valgrind
(in comparison to the release version without valgrind)
"Some" hours have gone into running it with valgrind, and checking for
all corruptions..
I think the next step is to go over the recent code additions and test
them again...
Rather, develop more extreme tests and try to elicit a failure..
Cheers,
Derek.
--
========================================
Q How did you make it work?
A The usual story, got everything right.
_______________________________________________
Linux-users mailing list
[email protected]
http://lists.canterbury.ac.nz/mailman/listinfo/linux-users
