From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Warren Subject: Re: [PATCH 3/9] ARM: tegra: Add MIPI calibration DT entries for Tegra114 Date: Tue, 08 Oct 2013 10:17:27 -0600 Message-ID: <52543017.9030501@wwwdotorg.org> References: <1381146528-18069-1-git-send-email-treding@nvidia.com> <1381146528-18069-4-git-send-email-treding@nvidia.com> <5253249B.6010008@wwwdotorg.org> <20131008100529.GE19633@ulmo.nvidia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20131008100529.GE19633-AwZRO8vwLAwmlAP/+Wk3EA@public.gmane.org> Sender: linux-tegra-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Thierry Reding Cc: linux-tegra-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: linux-tegra@vger.kernel.org On 10/08/2013 04:05 AM, Thierry Reding wrote: > On Mon, Oct 07, 2013 at 03:16:11PM -0600, Stephen Warren wrote: >> On 10/07/2013 05:48 AM, Thierry Reding wrote: >>> Add a device node for the MIPI calibration block on Tegra114. >>> There is no need to disable it by default because it only >>> enables the clock while performing calibration and therefore >>> shouldn't be consuming any power when unused. >> >>> include/dt-bindings/gpu/host1x/tegra114-mipi.h | 11 >>> +++++++++++ >> >> Shouldn't that file be added in the same patch that defines the >> binding that header contains constants for? What are those >> constants anyway? > > I'm not sure. The values themselves are not used within the driver. > They are merely a convenience for DTS authors. > > In a nutshell, the values defined within that file are a bitmask of > the pads that need calibration. There are 9 pads (5 for CSI, 4 for > DSI) but the association to each input/output isn't immediately > obvious. One DSI output uses DSIA and DSIB pads, while the other > uses DSIC and DSID. Ah. For values that are useful when creating *.dts, but not actually part of a binding definition, I'd tend to #define them a header file or *.dtsi in the dts/ directory, rather than the dt-bindings/ directory. The whole point of dt-bindings/ was for constants that both *.dts and the driver need to share. > Perhaps in this case it isn't useful to introduce the header at all > and just using the numerical values would be fine? It's probably useful to name them, although a comment in the final *.dts would likely be fine too.