lkml.org 
[lkml]   [2014]   [Apr]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH] workqueue: fix double unlock bug
From
2014-04-14 15:50 GMT+09:00 Lai Jiangshan <laijs@cn.fujitsu.com>:
> On 04/14/2014 08:58 AM, Daeseok Youn wrote:
>>
>> mutex_unlock() and put_pwq_unlocked() do not need to be called
>> when alloc_unbound_pwq() is failed.
>>
>> Signed-off-by: Daeseok Youn <daeseok.youn@gmail.com>
>> ---
>> kernel/workqueue.c | 2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
>> index 0ee63af..e6e9f6a 100644
>> --- a/kernel/workqueue.c
>> +++ b/kernel/workqueue.c
>> @@ -4100,7 +4100,7 @@ static void wq_update_unbound_numa(struct workqueue_struct *wq, int cpu,
>> if (!pwq) {
>> pr_warning("workqueue: allocation failed while updating NUMA affinity of \"%s\"\n",
>> wq->name);
>> - goto out_unlock;
>> + return;
>> }
>>
>> /*
>
>
> Nice catch!!!
>
> The supposed correct behavior is documented in the head of
> this function. We forgot to do it.
>
> * If NUMA affinity can't be adjusted due to memory allocation failure, it
> * falls back to @wq->dfl_pwq which may not be optimal but is always
> * correct.
>
> Could you use the following code instead of "goto out_unlock":
> mutex_lock(&wq->mutex);
> if (pwq == wq->dfl_pwq)
> goto out_unlock;
> else
> goto use_dfl_pwq;
>
> Correct&BAD. There are two blocks of suck code in this function:
> if (pwq == wq->dfl_pwq)
> goto out_unlock;
> else
> goto use_dfl_pwq;
>
> You can replace both these two blocks code to the following code:
> goto use_dfl_pwq;
OK. I will remove that "if-else" condition and just use "goto use_dfl_pwq" and
send this patch as V2.

Thanks.
Daeseok Youn
>
> The result is the same as before except it adds some small overhead.
> I don't care the small overhead in this function.
>
> Thanks
> Lai
>


\
 
 \ /
  Last update: 2014-04-14 11:21    [W:0.431 / U:0.752 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site