0005653: segmentation fault after systemctl restart sogo - SOGo | BTS

ID	Project	Category	View Status	Date Submitted	Last Update

0005653	SOGo	with SOGo	public	2022-12-01 09:15	2022-12-01 09:15

Reporter	ds	Assigned To
Priority	normal	Severity	crash	Reproducibility	always
Status	new	Resolution	open
Platform	Server	OS	Debian	OS Version	11
Product Version	5.7.1

Summary	0005653: segmentation fault after systemctl restart sogo
Description	Problem: `systemctl restart sogo` (also systemctl stop sogo) can cause a segmentation fault entry in Kernel log (journal -k). Sometimes it happens and sometimes not (see below). Kernel log: `kernel: sogod[359451]: segfault at 7ffd8f32dee3 ip 00007fd865928c2c sp 00007ffd0f32ddf8 error 4 in libc-2.31.so[7fd8657db000+15a000] kernel: Code: 28 7f 4f 01 62 e1 fe 28 7f 54 17 ff 62 e1 fe 28 7f 5c 17 fe c3 48 39 f7 0f 87 cb 00 00 00 0f 84 b3 fe ff ff 62 e1 fe 28 6f 26 <62> e1 fe 28 6f 6c 16 ff 62 e1 fe 28 6f 74 16 fe 62 e1 fe 28 6f 7c ...` Hypothesis: The init-Skript may cause the problem. Systemd uses the /etc/init.d/sogo script by means of the systemd-sysv-generator. In its start) and restart) - Section there is the line `start-stop-daemon -c $USER --stop --pidfile $PIDFILE --retry=TERM/20/KILL/5 --oknodo`. The TERM timeout value (here 20) has an impact on causing the segmentation faults because it might be to low in some cases.
Steps To Reproduce	For reproducibility, this parameter (TERM/20) can be set to a lower value, e.g. to 1, which results in segmentation fault happening at every restart/stop. The line `[WOWatchDog]> all children exited. We now terminate.` will then never show up anymore in the sogo.log.
Tags	No tags attached.

Date Modified	Username	Field	Change
2022-12-01 09:15	ds	New Issue