From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Drake Date: Mon, 24 Jun 2013 16:17:41 +0000 Subject: Re: [V2 5/7] video: mmp: add pitch info in mmp_win structure Message-Id: List-Id: References: <1370879574-11397-1-git-send-email-jtzhou@marvell.com> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-arm-kernel@lists.infradead.org On Mon, Jun 24, 2013 at 4:34 AM, jett zhou wrote: > pitch is used to represent line length in byte, the usage depends > on pix_fmt. > If the fmt is YUV , the pitch[0] will be Y length, pitch[1] will > be U length, pitch[2] will be V lenth. > If the fmt is RGB, the picth[0] will be line lenth, and > pitch[1]/pitch[2] will be 0 and not be used. This description is clear, thanks - hopefully you can write it with such clarity in the comment :) > For the patch rolling, do you mean combine the patch5 and patch6 > by one patch? I view patch 6 as a cleanup (consolidating and removing duplication of code), so I would leave that one separate. Patch 6 should not interact with any pitch[] stuff. Then you can write a followup patch which adds the pitch[] header, *and* modifies mmpfb_set_par() to write to pitch[], *and* acts upon pitch[] in dmafetch_set_fmt (patch 7). This way, the pitch variable is defined, documented, written to, and acted upon all in the same patch, the meaning will then be very clear. Thanks Daniel