| Summary: | interrupting bitbake doesn't kill all child processes | ||
|---|---|---|---|
| Product: | [Build System, Metadata & Runtime] BitBake | Reporter: | Stefan Stanacar <sstncr> |
| Component: | bitbake | Assignee: | valentin <valentin.popa> |
| Status: | VERIFIED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | Medium+ | CC: | jessica.zhang, poky.bs.watcher, poky.watcher, ross.burton, valentin.popa |
| Version: | unspecified | ||
| Target Milestone: | 1.5 | ||
| Hardware: | x86 | ||
| OS: | Multiple | ||
| Whiteboard: | |||
| OS type for building Yocto: | --- | Type of Regression: | --- |
| Verified: | Documentation change: | No (bug/feature does not impact docs) | |
I'm sorry for the ps truncated output, (narrow terminal and didn't used -w option on ps -ef) in step 2: Currently 1 running tasks (164 of 164): 0: core-image-sato-1.0-r0 do_testimage (pid 10211) stefans 9992 9989 0 12:05 pts/0 00:00:00 ping 127.0.0.1 stefans 10120 4676 7 12:07 pts/0 00:00:03 python /home/stefans/z/poky/bitbake/bin/bitbake core-image-sato -c testimage stefans 10169 10120 37 12:07 pts/0 00:00:15 python /home/stefans/z/poky/bitbake/bin/bitbake core-image-sato -c testimage stefans 10191 10169 2 12:07 pts/0 00:00:00 python /home/stefans/z/poky/bitbake/bin/bitbake-worker decafbad stefans 10211 10191 99 12:07 pts/0 00:00:27 python /home/stefans/z/poky/bitbake/bin/bitbake-worker decafbad in step 3: stefans 10211 1 99 12:07 pts/0 00:01:31 python /home/stefans/z/poky/bitbake/bin/bitbake-worker decafbad stefans 10214 10211 0 12:07 pts/0 00:00:00 ping 127.0.0.1 poky commit master:590010a6525b0e1bc1de73e794764e23404591df As another data point, we discovered a bug when we dropped pre python 2.7.3 support in that our multiprocessing pool workaround was always being used. There is a slim chance that now we require 2.7.3, the issue might be resolved... (In reply to comment #3) > As another data point, we discovered a bug when we dropped pre python 2.7.3 > support in that our multiprocessing pool workaround was always being used. > There is a slim chance that now we require 2.7.3, the issue might be > resolved... [Sorry, wrong bug for this comment :(] Hi Cristiana and Ross In the bug triage meeting, we decide to reassign this bug to Cristiana but very likely need Ross's help. So please work together. Patch sent on bitbake-devel. Merged: http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=888d1f35ee19665741bf3a78627cf2a8d3d6824b Verified in master:3c5b6af99110bd7f64f5240a3e9d156f0593c497 |
Double Ctrl+C doesn't seem to kill all bitbake's sub-processes: steps to reproduce: 1. Create a simple class with a simple task, e.g: python do_testimage() { import subprocess, time subprocess.Popen("ping 127.0.0.1", shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT) # do something for a while endtime = time.time() + 1000 while time.time() < endtime: pass } addtask testimage 2. bitbake something -c testimage: NOTE: Executing RunQueue Tasks Currently 1 running tasks (164 of 164): 0: core-image-sato-1.0-r0 do_testimage (pid 9762) Meanwhile check ps output stefans 9671 4676 5 12:02 pts/0 00:00:03 python /home/stefans/z/poky/bitb stefans 9720 9671 29 12:02 pts/0 00:00:15 python /home/stefans/z/poky/bitb stefans 9742 9720 1 12:02 pts/0 00:00:00 python /home/stefans/z/poky/bitb stefans 9762 9742 98 12:02 pts/0 00:00:39 python /home/stefans/z/poky/bitb stefans 9765 9762 0 12:02 pts/0 00:00:00 ping 127.0.0.1 3. Double Ctrl+C bitbake - build stops However running bitbake <something> again ERROR: Only one copy of bitbake should be run against a build directory Check ps output again. I still have: stefans 9762 1 99 12:02 pts/0 00:01:11 python /home/stefans/z/poky/bitb stefans 9765 9762 0 12:02 pts/0 00:00:00 ping 127.0.0.1 I have to kill the ping process then the python process ends.