search

RRZE-HPC / likwid

Performance monitoring and benchmarking suite
https://hpc.fau.de/research/tools/likwid/
GNU General Public License v3.0
1.63k stars 226 forks source link

[BUG] likwid's child processes survive when likwid parent is terminated #611

Open jdomke opened 6 months ago

jdomke commented 6 months ago 
$ killall cholesky_EXTRALARGE; timeout --foreground --kill-after=5s 20s likwid-perfctr --marker -C 20 -g MEM polybench/linear-algebra/solvers/cholesky/cholesky_EXTRALARGE ; ps aux | grep cholesky_EXTRALARGE
--------------------------------------------------------------------------------
CPU name:
CPU type:       Fujitsu A64FX
CPU clock:      0.00 GHz
--------------------------------------------------------------------------------
                 356:   domke      86551 97.6  0.8 274880 267968 pts/0   R    17:41   0:23 polybench/linear-algebra/solvers/cholesky/cholesky_EXTRALARGE

Expected behavior: no cholesky_EXTRALARGE found

(i found some signal handler code in likwid which "should" terminate its children upon receiving SigINT, but even with "timeout -s SIGINT..." i'm still seeing cholesky_EXTRALARGE running afterwards)

TomTheBear commented 5 months ago

Thanks for reporting. Yes this might happen in some situations where the waiting for the child process fails. I will take a look.