Re: [PATCH v2 1/2] blk-iocost: add refcounting for iocg

[Yu Kuai]

Hi,

在 2023/01/13 8:53, Tejun Heo 写道:
> Hello,
>
> On Thu, Jan 12, 2023 at 02:18:15PM +0800, Yu Kuai wrote:
> > remove the blkcg_deactivate_policy() from rq_qos_exit() from deleting
> > the device, and delay the policy cleanup and free to blkg_destroy_all().
> > Then the policies(other than bfq) can only call pd_free_fn() from
> > blkg_destroy(), and it's easy to guarantee the order. For bfq, it can
> > stay the same since bfq has refcounting itself.
> >
> > Then for the problem that ioc can be freed in pd_free_fn(), we can fix
> > it by freeing ioc in ioc_pd_free() for root blkg instead of
> > rq_qos_exit().
> >
> > What do you think?
>
> That would remove the ability to dynamically remove an rq_qos policy, right?
> We don't currently do it but given that having an rq_qos registered comes
> with perf overhead, it's something we might want to do in the future - e.g.

Yes, that make sense, remove ioc and other policies dynamically.

> only activate the policy when the controller is actually enabled. So, idk.
> What's wrong with synchronizing the two removal paths? blkcg policies are
> combinations of cgroups and block device configurations, so having exit
> paths from both sides is kinda natural.

I still can't figure out how to synchronizing them will a mutex. Maybe
I'm being foolish...

Thanks,
Kuai