On 1/24/25 9:47 AM, chenridong Wrote:
On 2025/1/24 1:47, Abel Wu wrote:
The commit b824766504e4 ("cgroup/rstat: add force idle show helper")
retrieves forceidle_time outside cgroup_rstat_lock for non-root cgroups
which can be potentially inconsistent with other stats.
Rather than reverting that commit, fix it in a way that retains the
effort of cleaning up the ifdef-messes.
Fixes: b824766504e4 ("cgroup/rstat: add force idle show helper")
Signed-off-by: Abel Wu <wuyun.abel@xxxxxxxxxxxxx>
---
kernel/cgroup/rstat.c | 29 +++++++++++++----------------
1 file changed, 13 insertions(+), 16 deletions(-)
diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index 5877974ece92..c2784c317cdd 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -613,36 +613,33 @@ static void cgroup_force_idle_show(struct seq_file *seq, struct cgroup_base_stat
void cgroup_base_stat_cputime_show(struct seq_file *seq)
{
struct cgroup *cgrp = seq_css(seq)->cgroup;
- u64 usage, utime, stime, ntime;
+ struct cgroup_base_stat bstat;
if (cgroup_parent(cgrp)) {
cgroup_rstat_flush_hold(cgrp);
- usage = cgrp->bstat.cputime.sum_exec_runtime;
+ bstat = cgrp->bstat;
Thank you for finding that.
In my version 2, I used to assign cgrp->bstat to bstat.
This is Tj's comment:
https://lore.kernel.org/linux-kernel/ZoQ2ti7nnz9EJSc3@xxxxxxxxxxxxxxx/
Thanks for pointing out. Using memcpy() is fine for me, but I see
cgroup_base_stat_flush() is using the same pattern, and if we don't
want copy like this, it would be better unify them in a separate
patch.
But IMHO in either way, I don't think reading forceidle time outside
cgroup_rstat_lock is the right thing to do.
Best Regards,
Abel
