On Thu, Aug 22, 2019 at 10:20:32AM +0200, Sebastian Andrzej Siewior wrote:
> On 2019-08-21 08:53:29 [-0700], Tejun Heo wrote:
> > On Fri, Aug 16, 2019 at 01:18:14PM +0200, Sebastian Andrzej Siewior wrote:
> > > I was looking at the lifetime of the the ->css_rstat_flush() to see if
> > > cgroup_rstat_cpu_lock should remain a raw_spinlock_t. I didn't find any
> > > users and is unused since it was introduced in commit
> > > 8f53470bab042 ("cgroup: Add cgroup_subsys->css_rstat_flush()")
> > >
> > > Remove the css_rstat_flush callback because it has no users.
> > >
> > > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
> >
> > Yeah, I'd rather keep it. The patch which used this didn't get merged
> > but the use case is valid and it will likely be used for different
> > cases.
>
> I was afraid that the inner while loop in cgroup_rstat_flush_locked()
> could get too long with the css_rstat_flush() here. Especially if it
> acquires spin locks. But since this is not currently happening...
> Any objections to the remaining series if I drop this patch?
Nack for the whole series. Please stop mixing interface and locking
changes.
Thanks.
--
tejun
