#131 memory leak when -d is not present

future
closed-works-for-me
nobody
5
2014-08-15
2011-07-15
No

Hi, we're running sipp in server mode in the background for testing purposes, and noticed that it gobbles up insane amounts of memory in the long run. Running under Valgrind produced the following leak report:

450,912 bytes in 1,281 blocks are possibly lost in loss record 35 of 37
at 0x402377E: operator new(unsigned) (vg_replace_malloc.c:224)
by 0x8073122: process_message(sipp_socket*, char*, int, sockaddr_storage*) (in /usr/bin/sipp)
by 0x8079243: pollset_process(int) (in /usr/bin/sipp)
by 0x8079819: traffic_thread() (in /usr/bin/sipp)
by 0x807CF9F: main (in /usr/bin/sipp)

which hinted that incoming call objects are leaking. Interestingly, addig -d 1 to the sipp server command line removed the leak.
We're using a basic UAS responder scenario under sipp SVN revision 599.

Discussion

  • Rob Day

    Rob Day - 2012-12-15

    I ran valgrind ./sipp -sn uas, put 1,500 calls through it, and got this output, so it didn't reproduce your issue:

    ==21790==
    ==21790== HEAP SUMMARY:
    ==21790== in use at exit: 146,521 bytes in 145 blocks
    ==21790== total heap usage: 328,801 allocs, 328,656 frees, 1,344,187,437 bytes allocated
    ==21790==
    ==21790== LEAK SUMMARY:
    ==21790== definitely lost: 0 bytes in 0 blocks
    ==21790== indirectly lost: 0 bytes in 0 blocks
    ==21790== possibly lost: 0 bytes in 0 blocks
    ==21790== still reachable: 146,521 bytes in 145 blocks
    ==21790== suppressed: 0 bytes in 0 blocks
    ==21790== Rerun with --leak-check=full to see details of leaked memory
    ==21790==
    ==21790== For counts of detected and suppressed errors, rerun with: -v
    ==21790== ERROR SUMMARY: 28 errors from 6 contexts (suppressed: 2 from 2)

    If you're using a custom scenario, it would be useful to see it; otherwise I'm going to have to close this bug report as being unreproducible.

     
  • Rob Day

    Rob Day - 2012-12-15
    • labels: Execution --> Execution, unreproducible
    • milestone: --> future
     
  • Rob Day

    Rob Day - 2013-10-08
    • status: open --> closed-works-for-me
     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks