On Fri, Jun 03, 2022 at 11:51:34AM +0200, Uladzislau Rezki wrote:
> > On Thu, Jun 02, 2022 at 10:06:43AM +0200, Uladzislau Rezki (Sony) wrote:
> > > From: "Joel Fernandes (Google)" <joel@xxxxxxxxxxxxxxxxx>
> > >
> > > monitor_todo is not needed as the work struct already tracks
> > > if work is pending. Just use that to know if work is pending
> > > using schedule_delayed_work() helper.
> > >
> > > Signed-off-by: Uladzislau Rezki (Sony) <urezki@xxxxxxxxx>
> > > Signed-off-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> > > ---
> > > kernel/rcu/tree.c | 33 ++++++++++++++++-----------------
> > > 1 file changed, 16 insertions(+), 17 deletions(-)
> > >
> > > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> > > index 222d59299a2a..fd16c0b46d9e 100644
> > > --- a/kernel/rcu/tree.c
> > > +++ b/kernel/rcu/tree.c
> > > @@ -3295,7 +3295,6 @@ struct kfree_rcu_cpu_work {
> > > * @krw_arr: Array of batches of kfree_rcu() objects waiting for a grace period
> > > * @lock: Synchronize access to this structure
> > > * @monitor_work: Promote @head to @head_free after KFREE_DRAIN_JIFFIES
> > > - * @monitor_todo: Tracks whether a @monitor_work delayed work is pending
> > > * @initialized: The @rcu_work fields have been initialized
> > > * @count: Number of objects for which GP not started
> > > * @bkvcache:
> > > @@ -3320,7 +3319,6 @@ struct kfree_rcu_cpu {
> > > struct kfree_rcu_cpu_work krw_arr[KFREE_N_BATCHES];
> > > raw_spinlock_t lock;
> > > struct delayed_work monitor_work;
> > > - bool monitor_todo;
> > > bool initialized;
> > > int count;
> > >
> > > @@ -3500,6 +3498,18 @@ static void kfree_rcu_work(struct work_struct *work)
> > > }
> > > }
> > >
> > > +static bool
> > > +need_offload_krc(struct kfree_rcu_cpu *krcp)
> > > +{
> > > + int i;
> > > +
> > > + for (i = 0; i < FREE_N_CHANNELS; i++)
> > > + if (krcp->bkvhead[i])
> > > + return true;
> > > +
> > > + return !!krcp->head;
> > > +}
> >
> > Thanks for modifying my original patch to do this, and thanks for giving me
> > the attribution for the patch. This function is a nice addition.
> >
> It was you who did it :) Actually the second patch depends on it therefore
> i decided to upload it on behalf of you with slight modification hoping that
> you would not mind.
Yes I don't mind at all :) Thank you again for seeing it through!
> > For the patch in its entirety:
> > Reviewed-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> >
> Thanks for the review!
Sure, any time. By the way I am out in the Carribean next week. I will catch
you all the week after.
Quick update on my side on the lazy CB stuff, I made some progress on using
the bypass lists for lazy CBs, its looking good and builds now. I fixed bug
in my code where idle loop was flushing lazy CBs on its way to idle.. I think
its probably O(workingdays) away from v2 posting assuming all goes well.
thanks,
- Joel
