On Sun, Apr 21, 2013 at 12:12:14AM +0800, Lai Jiangshan wrote:
> > I want an assertion here.
>
> freezing codes is very simple for verifying.
I still want it (and your patch is removing it). The usual cases are
fine but things can get messy on edge cases like freeze failure, bugs
in freezer code
Please forget all my other patches.
But these 1/7 and 2/7 __WQ_FREEZING patches can be in 3.10
On Thu, Apr 4, 2013 at 10:12 PM, Tejun Heo wrote:
> Hello, Lai.
>
> On Thu, Apr 04, 2013 at 10:05:32AM +0800, Lai Jiangshan wrote:
>> @@ -4757,25 +4747,16 @@ void thaw_workqueues(void)
>> {
>> s
Hello, Lai.
On Thu, Apr 04, 2013 at 10:05:32AM +0800, Lai Jiangshan wrote:
> @@ -4757,25 +4747,16 @@ void thaw_workqueues(void)
> {
> struct workqueue_struct *wq;
> struct pool_workqueue *pwq;
> - struct worker_pool *pool;
> - int pi;
>
> mutex_lock(&wq_pool_mutex);
>
freezing is nothing related to pools, but POOL_FREEZING adds a connection,
and causes freeze_workqueues_begin() and thaw_workqueues() complicated.
Since freezing is workqueue instance attribute, so we introduce __WQ_FREEZING
to wq->flags instead and remove POOL_FREEZING.
Signed-off-by: Lai Jiangs
4 matches
Mail list logo
