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;
Subject: Re: [PATCH] Remove blkg node after destroying blkg
To:     Tao Su <tao1.su@linux.intel.com>
Cc:     linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
        tj@kernel.org, josef@toxicpanda.com, axboe@kernel.dk,
        "yukuai (C)" <yukuai3@huawei.com>
References: <20230425075911.839539-1-tao1.su@linux.intel.com>
 <aa5de32c-c92b-d032-e9bb-83d2436ff72c@huawei.com>
 <ZEegQCCZ96ij6mw5@linux.bj.intel.com>
From:   Yu Kuai <yukuai1@huaweicloud.com>
Message-ID: <6d486f85-87ed-fa35-00cb-4c37fef17536@huaweicloud.com>
Date:   Tue, 25 Apr 2023 19:09:52 +0800
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.8.0
MIME-Version: 1.0
In-Reply-To: <ZEegQCCZ96ij6mw5@linux.bj.intel.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hi,

在 2023/04/25 17:41, Tao Su 写道:
> On Tue, Apr 25, 2023 at 04:09:34PM +0800, Yu Kuai wrote:
>> Hi,
>>
>> 在 2023/04/25 15:59, Tao Su 写道:
>>> Kernel hang when poweroff or reboot, due to infinite restart in function
>>> blkg_destroy_all. It will goto restart label when a batch of blkgs are
>>> destroyed, but not remove blkg node in blkg_list. So the blkg_list is
>>> same in every 'restart' and result in kernel hang.
>>>
>>> By adding list_del to remove blkg node after destroying, can solve this
>>> kernel hang issue and satisfy the previous will to 'restart'.
>>>
>>> Reported-by: Xiangfei Ma <xiangfeix.ma@intel.com>
>>> Tested-by: Xiangfei Ma <xiangfeix.ma@intel.com>
>>> Tested-by: Farrah Chen <farrah.chen@intel.com>
>>> Signed-off-by: Tao Su <tao1.su@linux.intel.com>
>>> ---
>>>    block/blk-cgroup.c | 1 +
>>>    1 file changed, 1 insertion(+)
>>>
>>> diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c
>>> index bd50b55bdb61..960eb538a704 100644
>>> --- a/block/blk-cgroup.c
>>> +++ b/block/blk-cgroup.c
>>> @@ -530,6 +530,7 @@ static void blkg_destroy_all(struct gendisk *disk)
>>>    		spin_lock(&blkcg->lock);
>>>    		blkg_destroy(blkg);
>>> +		list_del(&blkg->q_node);
>>
>> blkg should stay on the queue list until blkg_free_workfn(), otherwise
>> parent blkg can be freed before child, which will cause some known
>> issue.
> 
> Yes, directly removing blkg node is not appropriate, which I noticed some
> comments in blkg_destroy(), thanks for pointing out this issue.
> 
>>
>> I think this hung happens when total blkg is greater than
>> BLKG_DESTROY_BATCH_SIZE, right?
> 
> Yes, you are right.
> 
>>
>> Can you try if following patch fix your problem?
> 
> This patch can also fix my problem, and indeed is a more secure way.

Thanks for the test, for a better solution, I think 'blkcg_mutex' can
be used to protect 'blkg->q_node' list instead of 'queue_lock', so that
the 'restart' can be removed because softlockup can be avoided.

Thanks,
Kuai

> 
> Thanks,
> Tao
> 
>>
>> index 1c1ebeb51003..0ecb4cce8af2 100644
>> --- a/block/blk-cgroup.c
>> +++ b/block/blk-cgroup.c
>> @@ -527,6 +527,9 @@ static void blkg_destroy_all(struct gendisk *disk)
>>          list_for_each_entry_safe(blkg, n, &q->blkg_list, q_node) {
>>                  struct blkcg *blkcg = blkg->blkcg;
>>
>> +               if (hlist_unhashed(&blkg->blkcg_node))
>> +                       continue;
>> +
>>                  spin_lock(&blkcg->lock);
>>                  blkg_destroy(blkg);
>>                  spin_unlock(&blkcg->lock);
>>
>>>    		spin_unlock(&blkcg->lock);
>>>    		/*
>>>
> 
> .
>