On 9/13/21 12:04 PM, Evgenii Shatokhin wrote:
> On 12.09.2021 18:03, Vasily Averin wrote:
>> @@ -183,7 +187,11 @@ static __always_inline struct cgroup_subsys_state
>> *__ub_get_css(struct user_bean
>> if (css)
>> css_put(css);
>> - css_get(root_css);
>> + if ((ub == &ub0) || (root_css != ub0.ub_bound_css[idx]))
>
> Is it possible that root_css != ub0.ub_bound_css[idx] at this point? Some
> kind of race changing ub0.ub_bound_css[idx] under our feet? At the first
> glance, it seems unlikely.
You are right, this looks incorrect.
Thank you,
Vasily Averin
_______________________________________________
Devel mailing list
[email protected]
https://lists.openvz.org/mailman/listinfo/devel
