From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 2/1] dm: do not allocate any mempools for blk-mq request-based DM Date: Tue, 28 Apr 2015 08:28:43 +0200 Message-ID: <20150428062843.GA19065@lst.de> References: <1429957432-20672-1-git-send-email-hch@lst.de> <20150428005950.GA31039@redhat.com> <20150428010303.GB31039@redhat.com> Reply-To: device-mapper development Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20150428010303.GB31039@redhat.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Mike Snitzer Cc: Jens Axboe , dm-devel@redhat.com, Christoph Hellwig List-Id: dm-devel.ids On Mon, Apr 27, 2015 at 09:03:04PM -0400, Mike Snitzer wrote: > Do not allocate the io_pool mempool for blk-mq request-based DM > (DM_TYPE_MQ_REQUEST_BASED) in dm_alloc_rq_mempools(). > > Also refine __bind_mempools() to have more precise awareness of which > mempools each type of DM device uses -- avoids mempool churn when > reloading DM tables (particularly for DM_TYPE_REQUEST_BASED). Btw, I looked into this code and didn't dare to touch it before I understood how we deal with the case of dm-mpath using blk-mq and low level driver not or the other way around. As far as I can see we'd need the request mempool as long as the low level driver does not use dm-mq, independent of what dm-mpath itsel uses. The code doesn't seem to handle this right, but as mentioned I'm not 100% sure and need to dive deeper into this. Is there a place enforcing dm-mpath is using the same request type as the underlying devices?