From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Fri, 1 Jun 2018 01:43:02 -0700 From: Christoph Hellwig To: Ming Lei Cc: Jens Axboe , Kent Overstreet , Christoph Hellwig , Alexander Viro , David Sterba , Huang Ying , linux-kernel@vger.kernel.org, linux-block@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, Theodore Ts'o , "Darrick J . Wong" , Coly Li , Filipe Manana Subject: Re: [RESEND PATCH V5 00/33] block: support multipage bvec Message-ID: <20180601084302.GB30694@infradead.org> References: <20180525034621.31147-1-ming.lei@redhat.com> <20180525045306.GB8740@kmo-pixel> <8aa4276d-c0bc-3266-aa53-bf08a2e5ab5c@kernel.dk> <20180527072332.GA18240@ming.t460p> <20180528023042.GC26790@ming.t460p> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180528023042.GC26790@ming.t460p> Sender: owner-linux-mm@kvack.org List-ID: On Mon, May 28, 2018 at 10:30:43AM +0800, Ming Lei wrote: > I hate the conversion too, but looks it can't be avoided since > bio_for_each_segment_all() has to be changed. I guess you mean what is currently bio_for_each_page_all in your tree which now takes a bvec_iter_all agument? We could just try to hide the bia variable initially under an unlike to be used name, and then slowly move to the proper bio_for_each_page_all API unhiding it with the rename. But I think your current bio_for_each_segment_all should just go away. All three users really should be using better abstractions.