[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20191209165800.GB27229@linux.vnet.ibm.com>
Date: Mon, 9 Dec 2019 22:28:00 +0530
From: Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
To: Jens Axboe <axboe@...nel.dk>
Cc: Boaz Harrosh <boaz@...xistor.com>, Phil Auld <pauld@...hat.com>,
Ming Lei <ming.lei@...hat.com>,
Peter Zijlstra <peterz@...radead.org>,
Dave Chinner <david@...morbit.com>,
linux-block@...r.kernel.org, linux-fsdevel@...r.kernel.org,
linux-xfs@...r.kernel.org, linux-kernel@...r.kernel.org,
Jeff Moyer <jmoyer@...hat.com>,
Dave Chinner <dchinner@...hat.com>,
Eric Sandeen <sandeen@...hat.com>,
Christoph Hellwig <hch@....de>, Ingo Molnar <mingo@...hat.com>,
Tejun Heo <tj@...nel.org>,
Vincent Guittot <vincent.guittot@...aro.org>
Subject: Re: single aio thread is migrated crazily by scheduler
* Jens Axboe <axboe@...nel.dk> [2019-11-21 09:19:29]:
> >
> > I wish there was a flag to wake_up() or to the event object that says
> > to relinquish the remaning of the time-slice to the waiter on same
> > CPU, since I will be soon sleeping.
>
> Isn't that basically what wake_up_sync() is?
>
Workqueue don't seem to be using wait_queue_head which is needed when using
wake_up_sync and its related APIs. Also wake_up_sync would work when the
waking task seems to hand off the CPU and goes to sleep/block. However here,
i.e in fsperf case atleast, the waking thread continues to run after it has
woken the per cpu workqueue.
--
Thanks and Regards
Srikar Dronamraju
Powered by blists - more mailing lists