[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAOesGMhxs0A-YTXpS9Lqk_sn2=Q5jaCM2+mjEuvtwSX9Y49eMw@mail.gmail.com>
Date: Fri, 8 Nov 2019 14:29:31 -0800
From: Olof Johansson <olof@...om.net>
To: Lorenzo Pieralisi <lorenzo.pieralisi@....com>
Cc: Bjorn Helgaas <helgaas@...nel.org>,
Stephen Rothwell <sfr@...b.auug.org.au>,
Arnd Bergmann <arnd@...db.de>,
ARM <linux-arm-kernel@...ts.infradead.org>,
Linux Next Mailing List <linux-next@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Xiaowei Bao <xiaowei.bao@....com>,
Hou Zhiqiang <Zhiqiang.Hou@....com>,
Fabio Estevam <festevam@...il.com>,
Shawn Guo <shawnguo@...nel.org>
Subject: Re: linux-next: manual merge of the pci tree with the arm-soc tree
On Fri, Nov 8, 2019 at 3:07 AM Lorenzo Pieralisi
<lorenzo.pieralisi@....com> wrote:
>
> On Thu, Nov 07, 2019 at 03:18:01PM -0600, Bjorn Helgaas wrote:
> > On Thu, Nov 07, 2019 at 10:27:20AM -0800, Olof Johansson wrote:
> > > On Wed, Nov 6, 2019 at 2:46 PM Stephen Rothwell <sfr@...b.auug.org.au> wrote:
> > > >
> > > > Hi all,
> > > >
> > > > Today's linux-next merge of the pci tree got a conflict in:
> > > >
> > > > arch/arm64/boot/dts/freescale/fsl-ls1028a.dtsi
> > > >
> > > > between commit:
> > > >
> > > > 68e36a429ef5 ("arm64: dts: ls1028a: Move thermal-zone out of SoC")
> > > >
> > > > from the arm-soc tree and commit:
> > > >
> > > > 8d49ebe713ab ("arm64: dts: ls1028a: Add PCIe controller DT nodes")
> > >
> > > Bjorn, we ask that driver subsystem maintainers don't pick up DT
> > > changes since it causes conflicts like these.
> > >
> > > Is it easy for you to drop this patch, or are we stuck with it?
> > > Ideally it should never have been sent to you in the first place. :(
> >
> > Lorenzo, is it feasible for you to drop it from your pci/layerscape
> > branch and repush it? If so, I can redo the merge into my "next"
> > branch.
>
> Done. Should we ignore all dts updates from now onwards ?
Thanks!
Indeed, dts updates should only go in through the platform maintainers
(i.e. through soc tree), unless there are strong reasons to bring them
in through driver trees.
If there's a need for a dt-include to be shared between driver and
dts, getting them on a stable branch that's merged through both trees
is usually the best way. Reach out when that happens and we can
coordinate.
Regards,
-Olof
Powered by blists - more mailing lists