Control: forwarded -1 https://savannah.gnu.org/bugs/?func=detailitem&item_id=25089
Hi Vincent, Vincent Lefevre wrote: > On 2021-08-01 22:46:57 +0200, Axel Beckert wrote: > > I've seen this zsh zombies at least Buster, too. See > > http://savannah.gnu.org/bugs/?func=detailitem&item_id=25089 > > > > Could your issue be related or even identical to the one above in the > > upstream bug tracker? (I currently assume so, hence the "confirmed" > > tag.) > > There aren't many details, but this seems to be the same bug: Thanks for confirming my gut feeling. > > Unfortunately so far I wasn't able to reproduce this on demand. > > Currently I can reproduce the bug all the time on this VM Thanks for the detail that this also happens on a VM. > when starting new screen sessions. Note: I haven't rebooted the VM > yet (in case this is related to some kernel status). Indeed, I just did one try with the .screenrc I built for testing this bug (which more or less opens around 90 zsh screen windows. About 20 to 30 or so became zombies until I closed the last one. > > And for some reason, so far this issue — if seen at all — has mostly > > been seen when Screen runs on virtual machines. The reason for this is > > though unclear, at least to me. > > I would say that this could have an influence on the behavior in the > case of a race condition. [...] > I'm wondering whether the issue is due to what happens with the clone. > I suspect that there is a wait4 that catches this clone, but not the > terminated shell process, hence the corresponding zombie. Thanks for your analysis. One more analysis pointing towards a race condition and/or a logical error > > Vincent Lefevre wrote: > > > It seems more or less reproducbile: > > > > That would be good news! […] > > Would you mind providing that .screenrc-mutt? > > This doens't matter: the problem is reproducible with a normal > "screen", even if I remove the .screenrc file. > > > And now easy or how long does it take until zombies appear? > > I don't understand what you mean: Typo: s/now/how/ > as soon as I quit a shell running from a screen window, I get a > zombie (same PID). Yep, already got that from the remainder of this reply of yours. But nevertheless it made me trying this again and indeed, I was immediately able to reproduce it (already rebooted into 4.19.194-3). So something indeed seems to have worsened the situation. So I will start experimenting with the patches suggested by upstream in the upstream bug report again. Regards, Axel -- ,''`. | Axel Beckert <a...@debian.org>, https://people.debian.org/~abe/ : :' : | Debian Developer, ftp.ch.debian.org Admin `. `' | 4096R: 2517 B724 C5F6 CA99 5329 6E61 2FF9 CD59 6126 16B5 `- | 1024D: F067 EA27 26B9 C3FC 1486 202E C09E 1D89 9593 0EDE
