upstream/opnsense-src
1
0
Fork 0
mirror of https://github.com/opnsense/src.git synced 2026-06-12 10:10:24 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

reap_kill_subtree_once(): handle proctree_lock unlock in reap_kill_proc()

Browse source 
Recorded reaper might loose its reaper status, so we should not assert
it, but check and avoid signalling if this happens.

Reported and tested by:	pho
Reviewed by:	markj
Sponsored by:	The FreeBSD Foundation
MFC after:	2 week
Differential revision:	https://reviews.freebsd.org/D35310
This commit is contained in:
Konstantin Belousov 2022-05-10 00:41:23 +03:00
parent addf103ce6
commit 1d4abf2cfa
1 changed files with 13 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
sys/kern/kern_procctl.c
View file

@ -370,8 +370,20 @@ reap_kill_subtree_once(struct thread *td, struct proc *p, struct proc *reaper,
TAILQ_INIT(&tracker);
reap_kill_sched(&tracker, reaper);
while ((t = TAILQ_FIRST(&tracker)) != NULL) {
MPASS((t->parent->p_treeflag & P_TREE_REAPER) != 0);
TAILQ_REMOVE(&tracker, t, link);
/*
* Since reap_kill_proc() drops proctree_lock sx, it
* is possible that the tracked reaper is no longer.
* In this case the subtree is reparented to the new
* reaper, which should handle it.
*/
if ((t->parent->p_treeflag & P_TREE_REAPER) == 0) {
free(t, M_TEMP);
res = true;
continue;
}
LIST_FOREACH(p2, &t->parent->p_reaplist, p_reapsibling) {
if (t->parent == reaper &&
(rk->rk_flags & REAPER_KILL_SUBTREE) != 0 &&