From mboxrd@z Thu Jan 1 00:00:00 1970 From: Robert Jarzmik Date: Thu, 10 Dec 2015 17:34:56 +0000 Subject: Re: [PATCH v5 2/2] video: fbdev: pxafb: initial devicetree conversion Message-Id: <874mfqjjgf.fsf@belgarion.home> List-Id: References: <1447792337-13876-1-git-send-email-robert.jarzmik@free.fr> <1447792337-13876-2-git-send-email-robert.jarzmik@free.fr> <5665B3C1.1080005@ti.com> <87bna2kmp7.fsf@belgarion.home> <56699A57.9030209@ti.com> In-Reply-To: <56699A57.9030209@ti.com> (Tomi Valkeinen's message of "Thu, 10 Dec 2015 17:29:27 +0200") MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Tomi Valkeinen Cc: Rob Herring , Jean-Christophe Plagniol-Villard , linux-fbdev@vger.kernel.org, linux-kernel@vger.kernel.org Tomi Valkeinen writes: > On 07/12/15 22:50, Robert Jarzmik wrote: > >>> It seems that at least "depth" is missing from the binding document. >> You're right. >> >> Actually depth is not a "hardware" caracteristic. Moreover it's just used as an >> overlay for pxafb_set_pixfmt() to superseed var->bits_per_pixel. I'm wondering >> if the right path for of_get_pxafb_mode_info() would be to remove completely >> depth, and leave it initialized at 0 for the DT case. >> >> What do you think of this approach ? The other one would be to modify the >> binding, and yet I feel this depth doesn't belong to the binding, it's my patch >> which requires another spin IMHO. > > Yes, we should avoid non-hardware relate properties in the .dts files if > at all possible. If the driver works fine without the property, I think > it's fine to remove it. If it is required it needs to be added to the > binding document. Ok, agreed Tomi. That commits me for a v6, as the driver works fine on my tests. I'll send it in the next couple of days. Thanks for the review. -- Robert