From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg Kroah-Hartman Subject: Re: SPDX conversion under scripts/dtc/ of Linux Kernel Date: Wed, 19 Jun 2019 14:59:48 +0200 Message-ID: <20190619125948.GA27090@kroah.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Masahiro Yamada Cc: David Gibson , Rob Herring , linux-spdx@vger.kernel.org, Devicetree Compiler , DTML , Linux Kernel Mailing List , Thomas Gleixner , Frank Rowand List-Id: devicetree@vger.kernel.org On Wed, Jun 19, 2019 at 07:23:19PM +0900, Masahiro Yamada wrote: > Hi. > > In this development cycle of Linux kernel, > lots of files were converted to use SPDX > instead of the license boilerplate. > > However. > > Some files were imported from a different project, > and are periodically synchronized with the upstream. > Have we discussed what to do about this case? > > > For example, scripts/dtc/ is the case. > > The files in scripts/dtc/ are synced with the upstream > device tree compiler. > > Rob Herring periodically runs scripts/dtc/update-dtc-source.sh > to import outcome from the upstream. > > > The upstream DTC has not adopted SPDX yet. > > Some files in Linux (e.g. scripts/dtc/dtc.c) > have been converted to SPDX. > > So, they are out of sync now. > > The license boilerplate will come back > when Rob runs scripts/dtc/update-dtc-source.sh > next time. > > What shall we do? > > [1] Convert upstream DTC to SPDX > > This will be a happy solution if it is acceptable in DTC. > Since we cannot push the decision of the kernel to a different > project, this is totally up to David Gibson. That's fine with me :) > [2] Change scripts/dtc/update-dtc-source.sh to > take care of the license block somehow That would also be good. > [3] Go back to license boilerplate, and keep the files > synced with the upstream > (and scripts/dtc/ should be excluded from the > SPDX conversion tool.) nothing is being excluded from the SPDX conversions, sorry. The goal is to do this for every file in the kernel tree. Otherwise it's pointless. > Or, what else? Rob remembers to keep those first lines of the files intact when doing the next sync? thanks, greg k-h