psi: remove 500ms min window size limitation for triggers
authorSuren Baghdasaryan <surenb@google.com>
Fri, 3 Mar 2023 01:13:46 +0000 (17:13 -0800)
committerPeter Zijlstra <peterz@infradead.org>
Mon, 8 May 2023 08:58:38 +0000 (10:58 +0200)
commit519fabc7aaba3f0847cf37d5f9a5740c370eb777
tree137180a2e095b896faefbcd381c0f2594e4c9923
parent044f0e27dec6e30bb8875a4a12c5f2594964e93f
psi: remove 500ms min window size limitation for triggers

Current 500ms min window size for psi triggers limits polling interval
to 50ms to prevent polling threads from using too much cpu bandwidth by
polling too frequently. However the number of cgroups with triggers is
unlimited, so this protection can be defeated by creating multiple
cgroups with psi triggers (triggers in each cgroup are served by a single
"psimon" kernel thread).
Instead of limiting min polling period, which also limits the latency of
psi events, it's better to limit psi trigger creation to authorized users
only, like we do for system-wide psi triggers (/proc/pressure/* files can
be written only by processes with CAP_SYS_RESOURCE capability). This also
makes access rules for cgroup psi files consistent with system-wide ones.
Add a CAP_SYS_RESOURCE capability check for cgroup psi file writers and
remove the psi window min size limitation.

Suggested-by: Sudarshan Rajagopalan <quic_sudaraja@quicinc.com>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Link: https://lore.kernel.org/all/cover.1676067791.git.quic_sudaraja@quicinc.com/
kernel/cgroup/cgroup.c
kernel/sched/psi.c