[PATCH] selftests/cgroup: Fix intermittent test_cgfreezer_ptrace test failures

[PATCH] selftests/cgroup: Fix intermittent test_cgfreezer_ptrace test failures

[Waiman Long]

It is found that the test_cgfreezer_ptrace test of test_freezer can
intermittently fail on some architectures like arm64 and ppc64.

After further tracing of the mechanics of the test_cgfreezer_ptrace test,
it is found that the ptrace(PTRACE_DETACH) call temporaily unfreezes the
cgroup and then freezes it again afterward in the detaching process. The
reading of the frozen flag from cgroup.events is done from a different
process running maybe on a different CPU. As a result, racing is possible
and the intermediate unfrozen state can be read leading to occasional test
failures especially on architectures with a weak memory model like arm64.

Fix that by adding a short 1 ms delay before reading the frozen state
to ensure that the final frozen value will be read.

By running test_freezer 100 times in a loop, there were 28
test_cgfreezer_ptrace failures out of 100 on an arm64 test system before
the patch. After applying the patch, there was no test failure at all
in 100 runs of test_freezer.

Signed-off-by: Waiman Long <longman@redhat.com>
---
 tools/testing/selftests/cgroup/test_freezer.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/tools/testing/selftests/cgroup/test_freezer.c b/tools/testing/selftests/cgroup/test_freezer.c
index 0569e93fa6b0..8e2cca74d212 100644
--- a/tools/testing/selftests/cgroup/test_freezer.c
+++ b/tools/testing/selftests/cgroup/test_freezer.c
@@ -625,6 +625,18 @@ static int test_cgfreezer_ptrace(const char *root)
 	if (ptrace(PTRACE_DETACH, pid, NULL, NULL))
 		goto cleanup;
 
+	/*
+	 * The ptrace(PTRACE_DETACH) call will temporaily unfreeze the cgroup
+	 * and then freeze it again afterward in the detaching process. The
+	 * reading of the frozen flag from cgroup.events is done from a
+	 * different process running maybe on a different CPU. As a result,
+	 * racing is possible and the intermediate unfrozen state can be read
+	 * leading to occasional test failure especially on architectures with
+	 * a weak memory model like arm64. This intermittent test failure can
+	 * be avoided by adding a 1ms short delay before reading the frozen
+	 * state.
+	 */
+	usleep(1000);
 	if (cg_check_frozen(cgroup, true))
 		goto cleanup;
 
-- 
2.55.0

Re: [PATCH] selftests/cgroup: Fix intermittent test_cgfreezer_ptrace test failures

[Tejun Heo]

Hello, Waiman.

> +	usleep(1000);
>  	if (cg_check_frozen(cgroup, true))
>  		goto cleanup;

A fixed 1ms sleep only hides the race. On a loaded machine or a slow arch
(you mention ppc64) the refreeze can take longer than 1ms, and the test
reads the unfrozen state and fails anyway.

The freezer test already has a way to wait for this properly.
cg_prepare_for_wait() sets up an inotify watch on cgroup.events and
cg_wait_for() blocks until it changes. cg_enter_and_wait_for_frozen() shows
the pattern: loop cg_wait_for() then cg_check_frozen(). The cgroup always
ends up frozen, so looping until cg_check_frozen() reports frozen is
reliable and doesn't depend on timing.

Can you respin using that instead of usleep()?

Also, temporaily -> temporarily, in both the changelog and the comment.

Thanks.

-- 
tejun

Re: [PATCH] selftests/cgroup: Fix intermittent test_cgfreezer_ptrace test failures

[Tao Cui]

在 2026/7/15 06:09, Tejun Heo 写道:
> Hello, Waiman.
> 
>> +	usleep(1000);
>>  	if (cg_check_frozen(cgroup, true))
>>  		goto cleanup;
> 
> A fixed 1ms sleep only hides the race. On a loaded machine or a slow arch
> (you mention ppc64) the refreeze can take longer than 1ms, and the test
> reads the unfrozen state and fails anyway.
> 
> The freezer test already has a way to wait for this properly.
> cg_prepare_for_wait() sets up an inotify watch on cgroup.events and
> cg_wait_for() blocks until it changes. cg_enter_and_wait_for_frozen() shows
> the pattern: loop cg_wait_for() then cg_check_frozen(). The cgroup always
> ends up frozen, so looping until cg_check_frozen() reports frozen is
> reliable and doesn't depend on timing.
> 
> Can you respin using that instead of usleep()?
> 
> Also, temporaily -> temporarily, in both the changelog and the comment.
> 

Hi Tejun, Waiman,

I ran into a similar issue a while back and can add a data point on
the reproducibility: when running the full cgroup selftest suite, a
few cases -- test_cgfreezer_ptrace and test_cgfreezer_stopped -- fail
intermittently (around 40-70% on VMs), yet each failing case passes
reliably when run on its own. That points at state carried across
tests rather than a per-test bug, which is also why a fixed sleep/retry
is fragile.

On the "unfrozen state" above, I traced where CGRP_FROZEN actually gets
cleared. When the frozen tracee is woken by PTRACE_INTERRUPT,
JOBCTL_TRAP_STOP takes priority over JOBCTL_TRAP_FREEZE in get_signal(),
so the task enters ptrace_stop() instead of going back to
do_freezer_trap(). With debug printk in cgroup_update_frozen_flag() and
cgroup_leave_frozen(), the cg_test_ptrace trace shows:

  0 -> 1  nr_frozen=1  task_count=1   /* initial freeze            */
  1 -> 0  nr_frozen=0  task_count=1   /* dec, task still in cgroup */
  0 -> 1  nr_frozen=0  task_count=0   /* task left cgroup          */

The 1 -> 0 step (nr_frozen 1->0 while task_count is still 1) is
cgroup_dec_frozen_cnt(), called from cgroup_leave_frozen(true) at the
end of ptrace_stop() (signal.c:2479) when the tracee is woken by
PTRACE_DETACH. That clears CGRP_FROZEN until the task loops back into
do_freezer_trap() and re-enters the frozen state -- the transient
unfrozen window the test hits.

As a kernel-side attempt I changed cgroup_enter_frozen() so it no longer
returns early when current->frozen is already true: css_set_lock is
taken before the check and, if CGRP_FREEZE is still set,
cgroup_update_frozen() is called to re-verify the cgroup frozen state
when a frozen task is handed off to ptrace.

Due to other work I've had to pause this investigation for now, so I'm
sharing the above as a data point rather than a finished fix.

Thanks,
Tao

> Thanks.
> 

Re: [PATCH] selftests/cgroup: Fix intermittent test_cgfreezer_ptrace test failures

[Waiman Long]

On 7/14/26 6:09 PM, Tejun Heo wrote:
> Hello, Waiman.
>
>> +	usleep(1000);
>>   	if (cg_check_frozen(cgroup, true))
>>   		goto cleanup;
> A fixed 1ms sleep only hides the race. On a loaded machine or a slow arch
> (you mention ppc64) the refreeze can take longer than 1ms, and the test
> reads the unfrozen state and fails anyway.
>
> The freezer test already has a way to wait for this properly.
> cg_prepare_for_wait() sets up an inotify watch on cgroup.events and
> cg_wait_for() blocks until it changes. cg_enter_and_wait_for_frozen() shows
> the pattern: loop cg_wait_for() then cg_check_frozen(). The cgroup always
> ends up frozen, so looping until cg_check_frozen() reports frozen is
> reliable and doesn't depend on timing.
>
> Can you respin using that instead of usleep()?
>
> Also, temporaily -> temporarily, in both the changelog and the comment.
>
> Thanks.

Thanks for the suggestion. Will revise the patch as suggested.

Cheers,
Longman