From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PREEMPT-RT] [REPOST PATCH 1/5] scsi: bnx2i: convert to kworker Date: Mon, 7 Nov 2016 17:38:48 +0100 Message-ID: <20161107163848.GA1421@lst.de> References: <20160817101003.30055-1-bigeasy@linutronix.de> <20161107161901.GA12297@infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from verein.lst.de ([213.95.11.211]:57910 "EHLO newverein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753123AbcKGQiu (ORCPT ); Mon, 7 Nov 2016 11:38:50 -0500 Content-Disposition: inline In-Reply-To: Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Thomas Gleixner Cc: Christoph Hellwig , Sebastian Andrzej Siewior , "James E.J. Bottomley" , "Martin K. Petersen" , linux-scsi@vger.kernel.org, Chad Dupuis , rt@linutronix.de, QLogic-Storage-Upstream@qlogic.com, Johannes Thumshirn , Christoph Hellwig On Mon, Nov 07, 2016 at 05:31:03PM +0100, Thomas Gleixner wrote: > On Mon, 7 Nov 2016, Christoph Hellwig wrote: > > > It seems like the whole damn driver should just use threaded interrupts. > > Of course it's a giant beast and not just the iSCSI one. But even > > if we don't go all the way I'd much prefer workqueues. kthread work > > is simply the worst API ever and I'd prefer to not have it proliferate. > > That's what the patch is doing. It uses INIT_WORK() and > schedule_work[_on](). I can't find any reference to kthread work. You're right it does - between the incorrect subject and the fact that it still keeps the linked list of items arounds instead of fully using the workqueue infrastructure I was a bit confused before my first coffee this morning. Same applies to patch 2.