Skip to content

Commit

Permalink
cgroup/cpuset: Fix load balance state in update_partition_sd_lb()
Browse files Browse the repository at this point in the history
Commit a86ce68 ("cgroup/cpuset: Extract out CS_CPU_EXCLUSIVE
& CS_SCHED_LOAD_BALANCE handling") adds a new helper function
update_partition_sd_lb() to update the load balance state of the
cpuset. However the new load balance is determined by just looking at
whether the cpuset is a valid isolated partition root or not.  That is
not enough if the cpuset is not a valid partition root but its parent
is in the isolated state (load balance off). Update the function to
set the new state to be the same as its parent in this case like what
has been done in commit c8c9262 ("cgroup/cpuset: Inherit parent's
load balance state in v2").

Fixes: a86ce68 ("cgroup/cpuset: Extract out CS_CPU_EXCLUSIVE & CS_SCHED_LOAD_BALANCE handling")
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
  • Loading branch information
Waiman-Long authored and htejun committed Sep 18, 2023
1 parent d24f059 commit 6fcdb01
Showing 1 changed file with 11 additions and 1 deletion.
12 changes: 11 additions & 1 deletion kernel/cgroup/cpuset.c
Original file line number Diff line number Diff line change
Expand Up @@ -1304,13 +1304,23 @@ static int update_partition_exclusive(struct cpuset *cs, int new_prs)
*
* Changing load balance flag will automatically call
* rebuild_sched_domains_locked().
* This function is for cgroup v2 only.
*/
static void update_partition_sd_lb(struct cpuset *cs, int old_prs)
{
int new_prs = cs->partition_root_state;
bool new_lb = (new_prs != PRS_ISOLATED);
bool rebuild_domains = (new_prs > 0) || (old_prs > 0);
bool new_lb;

/*
* If cs is not a valid partition root, the load balance state
* will follow its parent.
*/
if (new_prs > 0) {
new_lb = (new_prs != PRS_ISOLATED);
} else {
new_lb = is_sched_load_balance(parent_cs(cs));
}
if (new_lb != !!is_sched_load_balance(cs)) {
rebuild_domains = true;
if (new_lb)
Expand Down

0 comments on commit 6fcdb01

Please sign in to comment.