On Tue, Jun 14, 2022 at 7:54 PM Jakub Kicinski <kuba@xxxxxxxxxx> wrote: > > On Tue, 14 Jun 2022 15:48:43 -0700 Krzysztof Kozlowski wrote: > > > I have seen a mixed set of the convention for dts patches. They are following > > > both routes i.e device tree or subsystem repos provided acked from device > > > tree maintainer. If there is preference for device tree repo then I can drop > > > net-next from subject prefix and resend it for the dt repo. > > > > If you got Ack from Devicetree bindings maintainer (Rob Herring or me), > > then feel free to take it via net-next. I think, it is actually > > preferred, unless this is some fix which needs to go via DT (Rob's) tree. > > > > If you don't have these acks, then better don't take it :) unless it's > > really waiting too long on the lists. I hope it's not that case. > > GTK, thanks. I'm also often confused by the correct tree for DT patches. It is documented in Documentation/devicetree/bindings/submitting-patches.rst, section II. The default is bindings should go thru subsystem trees. Given netdev's tendency to apply bindings before DT review and even semi-automated checks, but skipping standalone patches, I haven't minded picking them up. Rob
