mir/criu
2
0
Fork 0
mirror of https://github.com/checkpoint-restore/criu synced 2025-08-22 01:51:51 +00:00
Code Issues Releases Wiki Activity

freeze_processes: implement kludges for cgroup v1

Browse Source 
Cgroup v1 freezer has always been problematic, failing to freeze a
cgroup.

In runc, we have implemented a few kludges to increase the chance of
succeeding, but those are used when runc freezes a cgroup for its own
purposes (for "runc pause" and to modify device properties for cgroup
v1).

When criu is used, it fails to freeze a cgroup from time to time
(see [1], [2]). Let's try adding kludges similar to ones in runc.

Alas, I have absolutely no way to test this, so please review carefully.

[1]: https://github.com/opencontainers/runc/issues/4273
[2]: https://github.com/opencontainers/runc/issues/4457

Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This commit is contained in:
Kir Kolyshkin 2024-12-12 17:34:17 -08:00 committed by Andrei Vagin
parent 9c3c095cfe
commit 7c66617d0e
1 changed files with 31 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
criu/seize.c
View File

@ -539,6 +539,34 @@ err:
return exit_code;
}
static void cgroupv1_freezer_kludges(int fd, int iter, const struct timespec *req) {
/* As per older kernel docs (freezer-subsystem.txt before
* the kernel commit ef9fe980c6fcc1821), if FREEZING is seen,
* userspace should either retry or thaw. While current
* kernel cgroup v1 docs no longer mention a need to retry,
* even recent kernels can't reliably freeze a cgroup v1.
*
* Let's keep asking the kernel to freeze from time to time.
* In addition, do occasional thaw/sleep/freeze.
*
* This is still a game of chances (the real fix belongs to the kernel)
* but these kludges might improve the probability of success.
*
* Cgroup v2 does not have this problem.
*/
switch (iter % 32) {
case 9:
case 20:
freezer_write_state(fd, FROZEN);
break;
case 31:
freezer_write_state(fd, THAWED);
nanosleep(req, NULL);
freezer_write_state(fd, FROZEN);
break;
}
}
static int freeze_processes(void)
{
int fd, exit_code = -1;
@ -597,6 +625,9 @@ static int freeze_processes(void)
if (state == FROZEN || i++ == nr_attempts || alarm_timeouted())
break;
if (!cgroup_v2)
cgroupv1_freezer_kludges(fd, i, &req);
nanosleep(&req, NULL);
}