[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20171009144702.GB3301751@devbig577.frc2.facebook.com>
Date: Mon, 9 Oct 2017 07:47:02 -0700
From: Tejun Heo <tj@...nel.org>
To: Boqun Feng <boqun.feng@...il.com>
Cc: linux-kernel@...r.kernel.org, Josef Bacik <josef@...icpanda.com>,
Peter Zijlstra <peterz@...radead.org>,
Lai Jiangshan <jiangshanlai@...il.com>
Subject: Re: [PATCH workqueue/for-4.14-fixes] workqueue: replace
pool->manager_arb mutex with a flag
Hello, Boqun.
On Mon, Oct 09, 2017 at 10:21:17PM +0800, Boqun Feng wrote:
> > +static DECLARE_WAIT_QUEUE_HEAD(wq_manager_wait); /* wait for manager to go away */
>
> I think this wait_queue_head better be a per-pool one rather than shared
> among pools?
It should be fine either way. All the involved operations are pretty
low frequency.
> > @@ -3338,7 +3332,10 @@ static void put_unbound_pool(struct worker_pool *pool)
> > if (pool->detach_completion)
> > wait_for_completion(pool->detach_completion);
> >
> > - mutex_unlock(&pool->manager_arb);
> > + spin_lock_irq(&pool->lock);
> > + pool->flags &= ~POOL_MANAGER_ACTIVE;
> > + wake_up(&wq_manager_wait);
> > + spin_unlock_irq(&pool->lock);
> >
>
> Is the above code necesarry? IIUC, we are going to free the pool
> entirely, so whether manager is active is pointless here and no one is
> waiting for the ->flags of *this* pool to be !POOL_MANAGER_ACTIVE.
>
> Am I missing something subtle here?
Ah, true. I'll drop the above chunk.
Thanks.
--
tejun
Powered by blists - more mailing lists