Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Sender: Tejun Heo <htejun@gmail.com>
Date:   Wed, 4 Jan 2023 11:45:42 -1000
From:   Tejun Heo <tj@kernel.org>
To:     Yu Kuai <yukuai1@huaweicloud.com>
Cc:     hch@infradead.org, josef@toxicpanda.com, axboe@kernel.dk,
        cgroups@vger.kernel.org, linux-block@vger.kernel.org,
        linux-kernel@vger.kernel.org, yukuai3@huawei.com,
        yi.zhang@huawei.com
Subject: Re: [PATCH v2 2/2] blk-iocost: add refcounting for ioc
Message-ID: <Y7XzhqyJYBDfZhnF@slm.duckdns.org>
References: <20221227125502.541931-1-yukuai1@huaweicloud.com>
 <20221227125502.541931-3-yukuai1@huaweicloud.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20221227125502.541931-3-yukuai1@huaweicloud.com>
Precedence: bulk

On Tue, Dec 27, 2022 at 08:55:02PM +0800, Yu Kuai wrote:
> Root cause is that blkg_free() can be asynchronously, and it can race
> with delete device:
> 
> T1			T2		T3
> //delete device
> del_gendisk
>  bdi_unregister
>   bdi_remove_from_list
>    synchronize_rcu_expedited
> 
> 			//rmdir cgroup
> 			blkcg_destroy_blkgs
> 			 blkg_destroy
> 			  percpu_ref_kill
> 			   blkg_release
> 			    call_rcu
>  rq_qos_exit
>   ioc_rqos_exit
>    kfree(ioc)
> 					__blkg_release
> 					 blkg_free
> 					  blkg_free_workfn
> 					   pd_free_fn
> 					    ioc_pd_free
> 					     spin_lock_irqsave
> 
> Fix the problem by add refcounting for ioc, and iocg will grab reference
> of ioc, so that ioc won't be freed until all the iocg is exited.

Ditto, why do this in iocost instead of blk-cgroup core?

Thanks.

-- 
tejun