Bugzilla – Bug 2281
brite-generic-example generates valgrind errors; won't work on Mac OS X
Last modified: 2016-02-04 14:23:26 EST
./waf --valgrind --run brite-generic-example shows leaks and exits with error I did not check the MPI example but suspect the same for that one as well.
I'm not seeing it exit with an error however the BRITE library itself leaks memory like a sieve. I'm pretty sure this came up before and it was decided not to fix/change BRITE. However if that is desired, I'm happy to work on it. Is BRITE actually useful to people? Is there other desired functionality? Is it worth reworking BRITE and ingesting it into the ns-3 codebase? Also by default BRITE doesn't work on Macs. This is probably just an issue with the Makefile though not producing a .dylib.
(In reply to Brian Swenson from comment #1) > I'm not seeing it exit with an error however the BRITE library itself leaks > memory like a sieve. I'm pretty sure this came up before and it was decided > not to fix/change BRITE. However if that is desired, I'm happy to work on > it. > > Is BRITE actually useful to people? > Is there other desired functionality? > Is it worth reworking BRITE and ingesting it into the ns-3 codebase? > > Also by default BRITE doesn't work on Macs. This is probably just an issue > with the Makefile though not producing a .dylib. From my perspective, I would not bother to fix BRITE valgrind issues, if the issue is there and not in the ns-3 code. It would be useful to make it work on Macs if easy to do so. I would tend to leave as is (it may be useful as is, but we are not yet sure it is useful enough to rework it more thoroughly into the ns-3 codebase).
Marking patch wanted in case someone wants to provide a patch to make it work on MAC OS X. I suspect we will mark the valgrind issue as WONTFIX.