On Thu, Feb 9, 2017 at 6:06 AM, Mark Brown <broo...@kernel.org> wrote: > On Thu, Feb 09, 2017 at 09:48:09AM +0800, Shawn Guo wrote: >> On Wed, Feb 08, 2017 at 06:52:07PM -0600, Rob Herring wrote: > >> > Personally, I'd prefer "dt-bindings: sound: blah...", but not enough to >> > argue with Mark about it. If that is not the prefix, then it should at >> > least have "binding" in the subject. > >> +1 > >> The prefix of sound bindings is quite unique from other subsystems. >> Looking at the prefix of Documentation/devicetree/bindings/sound/ >> commits, I'm always confused whether it's a pure binding commit >> or just submitted as part of the driver patch. I feel that we >> kinda lose the point of having a prefix. > > That's really not what's happening reliably - other subsystems also seem > to have a bunch of things prefixed for the subsystem and the DT specific > prefixes are all over the shop, people seem to be making them up at > random.
I'm getting more picky about the subject and splitting bindings to a separate patch, but generally only when I have other comments. And I've had to get some maintainers to stop combining commits as they apply them. Maybe get_maintainers.pl could spit out the desired prefix and checkpatch check it. Evidently, running "git log --oneline" is too hard. > If DT binding review were something that reliably and > consistently happened and didn't affect the subsystem I'd perhaps buy it > but for run of the mill stuff it seems like getting things reviewed in > the subsystem is more important. I review everything that gets sent to the DT list unless maintainers apply it first. I'll still comment afterwards if there's anything significant (or I missed that it was applied :)). Rob