Messages in this thread | | | Date | Wed, 7 May 2014 23:30:35 +0800 | Subject | Re: [PATCH 04/10] workqueue: destroy worker directly in the idle timeout handler | From | Lai Jiangshan <> |
| |
On Wed, May 7, 2014 at 9:41 PM, Tejun Heo <tj@kernel.org> wrote: > On Wed, May 07, 2014 at 09:38:39PM +0800, Lai Jiangshan wrote: >> On Wed, May 7, 2014 at 9:12 PM, Tejun Heo <tj@kernel.org> wrote: >> > Hello, Lai. >> > >> > On Wed, May 07, 2014 at 03:10:20PM +0800, Lai Jiangshan wrote: >> >> 1) complete() can't be called inside attach_mutex due to the worker >> >> shouldn't access to the pool after complete(). >> > >> > Sure, complete it after releasing the lock. Shutdown can't complete >> > before the completion gets completed, right? >> > >> >> 2) put_unbound_pool() may called from get_unbound_pool(), we need to add >> >> an additional check and avoid the wait_for_completion() if so. >> >> Do you accept if I remove put_unbound_pool() from get_unbound_pool() >> and use several freeing code instead? > > Hah? How much extra complexity are we talking about? It's a single > if, no?
DECLARE_COMPLETION_ONSTACK(completion); #1
...
while ((worker = first_worker(pool))) { destroy_worker(worker); pool->detach_completion = &completion; #2 }
... unlock;
if (pool->detach_completion) wait_for_completion(); #3
One thing is separated into 3 places and about 5~7lines. I hope a single wait_for_completion() or single wait_event().
get_unbound_pool(): fail: if (pool) put_unbound_pool(pool);
I think we can change it into:
fail: if (pool) { if (pool->id >= 0) idr_remove(&worker_pool_idr, pool->id); call_rcu_sched(&pool->rcu, rcu_free_pool); }
Thanks, Lai
> > -- > tejun > -- > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/
| |