Re: [PATCH -next] md:ensure mddev->reconfig_mutex is hold when try to get mddev->sync_thread

[Li Lingfeng <lilingfeng@huaweicloud.com>]

Thanks for your advice, I will send a new patch soon.

在 2023/7/28 15:21, Paul Menzel 写道:
> Dear Li,
>
>
> Thank you for your patch.
>
> I notice two minor things in the summary:
>
> 1.  Please add a space after the colon
> 2.  “is hold” should be “is held”.
>
> Maybe even shorter:
>
> md: Hold mddev->reconfig_mutex when trying to get mddev->sync_thread
>
> Am 27.07.23 um 09:20 schrieb Li Lingfeng:
>> Commit ba9d9f1a707f ("Revert "md: unlock mddev before reap 
>> sync_thread in
>> action_store"") removed the scenario of calling md_unregister_thread()
>> without holding mddev->reconfig_mutex, so add a lock holding check 
>> before
>> acquiring mddev->sync_thread.
>
> Maybe also mention, that this is done by passing `mdev` to 
> `md_unregister_thread()`.
>
> Plesae add a Fixes: tag.
>
This is not a bugfix patch , so I don't think it's necessary to add a 
fix tag.
>> Signed-off-by: Li Lingfeng <lilingfeng@huaweicloud.com>
>> ---
>>   drivers/md/md-cluster.c  | 8 ++++----
>>   drivers/md/md.c          | 9 +++++----
>>   drivers/md/md.h          | 2 +-
>>   drivers/md/raid1.c       | 4 ++--
>>   drivers/md/raid10.c      | 2 +-
>>   drivers/md/raid5-cache.c | 2 +-
>>   drivers/md/raid5.c       | 2 +-
>>   7 files changed, 15 insertions(+), 14 deletions(-)
>>
>> diff --git a/drivers/md/md-cluster.c b/drivers/md/md-cluster.c
>> index 3d9fd74233df..1e26eb223349 100644
>> --- a/drivers/md/md-cluster.c
>> +++ b/drivers/md/md-cluster.c
>> @@ -952,8 +952,8 @@ static int join(struct mddev *mddev, int nodes)
>>       return 0;
>>   err:
>>       set_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state);
>> -    md_unregister_thread(&cinfo->recovery_thread);
>> -    md_unregister_thread(&cinfo->recv_thread);
>> +    md_unregister_thread(mddev, &cinfo->recovery_thread);
>> +    md_unregister_thread(mddev, &cinfo->recv_thread);
>>       lockres_free(cinfo->message_lockres);
>>       lockres_free(cinfo->token_lockres);
>>       lockres_free(cinfo->ack_lockres);
>> @@ -1015,8 +1015,8 @@ static int leave(struct mddev *mddev)
>>           resync_bitmap(mddev);
>>         set_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state);
>> -    md_unregister_thread(&cinfo->recovery_thread);
>> -    md_unregister_thread(&cinfo->recv_thread);
>> +    md_unregister_thread(mddev, &cinfo->recovery_thread);
>> +    md_unregister_thread(mddev, &cinfo->recv_thread);
>>       lockres_free(cinfo->message_lockres);
>>       lockres_free(cinfo->token_lockres);
>>       lockres_free(cinfo->ack_lockres);
>> diff --git a/drivers/md/md.c b/drivers/md/md.c
>> index a3d98273b295..5c3c19b8d509 100644
>> --- a/drivers/md/md.c
>> +++ b/drivers/md/md.c
>> @@ -6258,7 +6258,7 @@ static void mddev_detach(struct mddev *mddev)
>>           mddev->pers->quiesce(mddev, 1);
>>           mddev->pers->quiesce(mddev, 0);
>>       }
>> -    md_unregister_thread(&mddev->thread);
>> +    md_unregister_thread(mddev, &mddev->thread);
>>       if (mddev->queue)
>>           blk_sync_queue(mddev->queue); /* the unplug fn references 
>> 'conf'*/
>>   }
>> @@ -7990,9 +7990,10 @@ struct md_thread *md_register_thread(void 
>> (*run) (struct md_thread *),
>>   }
>>   EXPORT_SYMBOL(md_register_thread);
>>   -void md_unregister_thread(struct md_thread __rcu **threadp)
>> +void md_unregister_thread(struct mddev *mddev, struct md_thread 
>> __rcu **threadp)
>>   {
>> -    struct md_thread *thread = rcu_dereference_protected(*threadp, 
>> true);
>> +    struct md_thread *thread = rcu_dereference_protected(*threadp,
>> + lockdep_is_held(&mddev->reconfig_mutex));
>>         if (!thread)
>>           return;
>> @@ -9484,7 +9485,7 @@ void md_reap_sync_thread(struct mddev *mddev)
>>       bool is_reshaped = false;
>>         /* resync has finished, collect result */
>> -    md_unregister_thread(&mddev->sync_thread);
>> +    md_unregister_thread(mddev, &mddev->sync_thread);
>>       atomic_inc(&mddev->sync_seq);
>>         if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery) &&
>> diff --git a/drivers/md/md.h b/drivers/md/md.h
>> index 8ae957480976..9bcb77bca963 100644
>> --- a/drivers/md/md.h
>> +++ b/drivers/md/md.h
>> @@ -761,7 +761,7 @@ extern struct md_thread *md_register_thread(
>>       void (*run)(struct md_thread *thread),
>>       struct mddev *mddev,
>>       const char *name);
>> -extern void md_unregister_thread(struct md_thread __rcu **threadp);
>> +extern void md_unregister_thread(struct mddev *mddev, struct 
>> md_thread __rcu **threadp);
>>   extern void md_wakeup_thread(struct md_thread __rcu *thread);
>>   extern void md_check_recovery(struct mddev *mddev);
>>   extern void md_reap_sync_thread(struct mddev *mddev);
>> diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
>> index 23d211969565..581dfbdfca89 100644
>> --- a/drivers/md/raid1.c
>> +++ b/drivers/md/raid1.c
>> @@ -3152,7 +3152,7 @@ static int raid1_run(struct mddev *mddev)
>>        * RAID1 needs at least one disk in active
>>        */
>>       if (conf->raid_disks - mddev->degraded < 1) {
>> -        md_unregister_thread(&conf->thread);
>> +        md_unregister_thread(mddev, &conf->thread);
>>           ret = -EINVAL;
>>           goto abort;
>>       }
>> @@ -3179,7 +3179,7 @@ static int raid1_run(struct mddev *mddev)
>>         ret = md_integrity_register(mddev);
>>       if (ret) {
>> -        md_unregister_thread(&mddev->thread);
>> +        md_unregister_thread(mddev, &mddev->thread);
>>           goto abort;
>>       }
>>       return 0;
>> diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
>> index 16aa9d735880..6188b71186f4 100644
>> --- a/drivers/md/raid10.c
>> +++ b/drivers/md/raid10.c
>> @@ -4320,7 +4320,7 @@ static int raid10_run(struct mddev *mddev)
>>       return 0;
>>     out_free_conf:
>> -    md_unregister_thread(&mddev->thread);
>> +    md_unregister_thread(mddev, &mddev->thread);
>>       raid10_free_conf(conf);
>>       mddev->private = NULL;
>>   out:
>> diff --git a/drivers/md/raid5-cache.c b/drivers/md/raid5-cache.c
>> index 47ba7d9e81e1..ce9b42fd54b9 100644
>> --- a/drivers/md/raid5-cache.c
>> +++ b/drivers/md/raid5-cache.c
>> @@ -3171,7 +3171,7 @@ void r5l_exit_log(struct r5conf *conf)
>>       /* Ensure disable_writeback_work wakes up and exits */
>>       wake_up(&conf->mddev->sb_wait);
>>       flush_work(&log->disable_writeback_work);
>> -    md_unregister_thread(&log->reclaim_thread);
>> +    md_unregister_thread(conf->mddev, &log->reclaim_thread);
>>         conf->log = NULL;
>>   diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
>> index 4cdb35e54251..f41f9b712d3d 100644
>> --- a/drivers/md/raid5.c
>> +++ b/drivers/md/raid5.c
>> @@ -8107,7 +8107,7 @@ static int raid5_run(struct mddev *mddev)
>>         return 0;
>>   abort:
>> -    md_unregister_thread(&mddev->thread);
>> +    md_unregister_thread(mddev, &mddev->thread);
>>       print_raid5_conf(conf);
>>       free_conf(conf);
>>       mddev->private = NULL;