From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Bart Van Assche To: "hch@lst.de" , "axboe@kernel.dk" CC: "linux-block@vger.kernel.org" , "ming.lei@redhat.com" Subject: Re: [PATCH 4/6] mtip32xx: convert internal command issue to block IO path Date: Tue, 2 May 2017 15:42:22 +0000 Message-ID: <1493739741.2552.3.camel@sandisk.com> References: <1493398473-29872-1-git-send-email-axboe@fb.com> <1493398473-29872-5-git-send-email-axboe@fb.com> <20170502072810.GA11582@lst.de> <20170502151607.GA22963@lst.de> In-Reply-To: Content-Type: text/plain; charset="iso-8859-1" MIME-Version: 1.0 List-ID: On Tue, 2017-05-02 at 09:25 -0600, Jens Axboe wrote: > On 05/02/2017 09:16 AM, Christoph Hellwig wrote: > > Any reason for the move from ->end_io_data to ->special? I thought > > that ->special was something we'd get rid of sooner or later now > > that we can have additional per-cmd data even for !mq. >=20 > With the switch to blk_execute_rq(), we can't be using end_io_data > and end_io, as we use that internally for the wakeup. So I have to > stuff it somewhere. >=20 > The obvious option would be to move it to mtip_cmd, but we can't > safely access that prior to having a driver tag assigned, which doesn't > happen until we end up in our ->queue_rq(). So we need to stuff it > somewhere. Hello Jens, Do you think it would be a good idea to allow blk_get_request() callers to specify that a driver tag has to be allocated even if a scheduler has been configured? That would make it possible to store completion data in mtip_cmd for the mtip driver. Thanks, Bart.=