From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 2/5] mtip32xx: handle arbitrary size bios Date: Mon, 25 Nov 2013 22:10:18 -0800 Message-ID: <20131126061018.GC6434@infradead.org> References: <1385418633-9753-1-git-send-email-kmo@daterainc.com> <1385418633-9753-3-git-send-email-kmo@daterainc.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, axboe@kernel.dk, Asai Thambi S P , Sam Bradshaw To: Kent Overstreet Return-path: Received: from bombadil.infradead.org ([198.137.202.9]:38381 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751749Ab3KZGKT (ORCPT ); Tue, 26 Nov 2013 01:10:19 -0500 Content-Disposition: inline In-Reply-To: <1385418633-9753-3-git-send-email-kmo@daterainc.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Mon, Nov 25, 2013 at 02:30:30PM -0800, Kent Overstreet wrote: > We get a measurable performance increase by handling this in the driver when > we're already looping over the biovec, instead of handling it separately in > generic_make_request() (or bio_add_page() originally) Given that Jens has a patch queue up to convert the driver to blk-mq adjusting the blk-mq infrastructure to handle this might be a better plan forward..