Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Memory counters give segfault when attempting to use thread wild cards or numbers only total works #825

Closed
pagrubel opened this issue Aug 29, 2013 · 4 comments

Comments

@pagrubel
Copy link
Member

pagrubel@deneb:/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter /runtime{locality#0/worker-thread#*}/memory/resident --hpx:threads=2
Segmentation fault
pagrubel@deneb:
/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter /runtime{locality#0/worker-thread#0}/memory/resident --hpx:threads=2
Segmentation fault
pagrubel@deneb:~/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter /runtime{locality#0/total}/memory/resident --hpx:threads=2
/runtime{locality#0/total}/memory/resident,1,0.909590,[s],1.07708e+08,[bytes]
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.851429, 1.70286e-06
/runtime{lo

@pagrubel
Copy link
Member Author

some other counters also don't work see objects and stack recycles

/threads{locality#0/worker-thread#0}/idle-rate,1,0.917032,[s],749,[0.1%]
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.859199, 1.7184e-06
/threads{locality#0/worker-thread#0}/idle-rate,2,0.917926,[s],749,[0.1%]
pagrubel@deneb:/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/worker-thread#0}/count/cumulative -t2
/threads{locality#0/worker-thread#0}/count/cumulative,1,0.918767,[s],289504
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.858203, 1.71641e-06
/threads{locality#0/worker-thread#0}/count/cumulative,2,0.919647,[s],289510
pagrubel@deneb:
/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/worker-thread#0}/count/stolen -t2
/threads{locality#0/worker-thread#0}/count/stolen,1,0.928986,[s],6296
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.869776, 1.73955e-06
/threads{locality#0/worker-thread#0}/count/stolen,2,0.929828,[s],6304
pagrubel@deneb:/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/worker-thread#0}/count/objects -t2
Segmentation fault
pagrubel@deneb:
/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/total}/count/objects -t2
/threads{locality#0/total}/count/objects,1,0.922980,[s],398
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.862529, 1.72506e-06
/threads{locality#0/total}/count/objects,2,0.923923,[s],398
pagrubel@deneb:/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/worker-thread#0}/count/stack-recycles -t2
Segmentation fault
pagrubel@deneb:
/build/hpx_build0822$ ./bin/hpx_homogeneous_timed_task_spawn --hpx:print-counter=/threads{locality#0/total}/count/stack-recycles -t2
/threads{locality#0/total}/count/stack-recycles,1,0.939390,[s],499605
OS-threads,Tasks,Delay (micro-seconds),Total Walltime (seconds),Walltime per Task (seconds)
2, 500000, 0, 0.872973, 1.74595e-06
/threads{locality#0/total}/count/stack-recycles,2,0.940155,[s],499610

@ghost ghost assigned hkaiser Aug 29, 2013
@hkaiser
Copy link
Member

hkaiser commented Aug 29, 2013

With reagrds to the first error you're reporting, I'm seeing this error message:

{what}: couldn't create performance counter: /runtime{locality#0/worker-thread#0}/memory/resident \
    (HPX(bad_parameter)): HPX(bad_parameter)

which is describing the reason for the stop: the counter does not exist.

@hkaiser
Copy link
Member

hkaiser commented Aug 31, 2013

These segfaults are a result of a stack overflow.

@hkaiser
Copy link
Member

hkaiser commented Sep 1, 2013

Stack back traces are now collected on a larger stack which fixes the issue (see f1bffea)

@hkaiser hkaiser closed this as completed Sep 1, 2013
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants