Re: [RFC][PATCH] f2fs: Avoid print false deadlock messages.

[majianpeng <majianpeng@gmail.com>]

On 05/15/2013 04:28 PM, Peter Zijlstra wrote:
> On Wed, May 15, 2013 at 02:58:53PM +0800, majianpeng wrote:
>> When mounted the f2fs, kernel will print the following messages:
>>
>> [  105.533038] =============================================
>> [  105.533065] [ INFO: possible recursive locking detected ]
>> [  105.533088] 3.10.0-rc1+ #101 Not tainted
>> [  105.533105] ---------------------------------------------
>> [  105.533127] mount/5833 is trying to acquire lock:
>> [  105.533147]  (&sbi->fs_lock[i]){+.+...}, at: [<ffffffffa02017a6>] write_checkpoint+0xb6/0xaf0 [f2fs]
>> [  105.533204]
>> [  105.533204] but task is already holding lock:
>> [  105.533228]  (&sbi->fs_lock[i]){+.+...}, at: [<ffffffffa02017a6>] write_checkpoint+0xb6/0xaf0 [f2fs]
>> [  105.533278]
>> [  105.533278] other info that might help us debug this:
>> [  105.533305]  Possible unsafe locking scenario:
>> [  105.533305]
>> [  105.533329]        CPU0
>> [  105.533341]        ----
>> [  105.533353]   lock(&sbi->fs_lock[i]);
>> [  105.533373]   lock(&sbi->fs_lock[i]);
>> [  105.533394]
>> [  105.533394]  *** DEADLOCK ***
>> [  105.533394]
>> By adding some messages, i found this problem because the gcc
>> optimizing. For those codes:
>>>        for (i = 0; i < NR_GLOBAL_LOCKS; i++)
>>>                mutex_init(&sbi->fs_lock[i]);
>> The defination of mutex_init is:
>>> #define mutex_init(mutex)
>>> do {
>>>
>>>        static struct lock_class_key __key;
>>>
>>>                                                                      
>>>        __mutex_init((mutex), #mutex, &__key);
>>>
>>> } while (0)
>> Because the optimizing of gcc, there are only one __key rather than
>> NR_GLOBAL_LOCKS times.
> Its not a gcc specific optimization, any C compiler would. Its also very
> much on purpose.
>
>> Add there is other problems about lockname.Using 'for()' the lockname is
>> the same which is '&sbi->fs_lock[i]'.If it met problem about
>> mutex-operation, it can't find which one.
>>
>> Although my patch can work,i think it's not best.Because if
>> NR_GLOBAL_LOCKS changed, we may leak to change this.
>>
>> BTY, if who know how to avoid optimize, please tell me. Thanks!
Thanks your answer! Your patch looks good.
> There isn't. What you typically want to do is annotate the lock site.
> In particular it looks like mutex_lock_all() is the offensive piece of
> code (horrible function name though; the only redeeming thing being that
> f2fs.h isn't likely to be included elsewhere).
>
> One thing you can do here is modify it to look like:
>
> static inline void mutex_lock_all(struct f2fs_sb_info *sbi)
> {
> 	int i;
>
> 	for (i = 0; i < NR_GLOBAL_LOCKS; i++) {
> 		/*
> 		 * This is the only time we take multiple fs_lock[]
> 		 * instances; the order is immaterial since we
> 		 * always hold cp_mutex, which serializes multiple
> 		 * such operations.
> 		 */
> 		mutex_lock_nest_lock(&sbi->fs_lock[i], &sbi->cp_mutex);
> 	}
> }
>
> That tells the lock validator that it is ok to lock multiple instances
> of the fs_lock[i] class because the lock order is guarded by cp_mutex.
> While your patch also works, it has multiple down-sides; its easy to get
> out of sync when you modify NR_GLOBAL_LOCKS; also it consumes more
> static lockdep resources (lockdep has to allocate all its resources
> from static arrays since allocating memory also uses locks -- recursive
> problem).
>
Yes, but there is a problem if fs_block[] met deadlock. How to find which one?
Because the lock->name is the same.

Thanks!
Jianpeng Ma