Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 23 Feb 2018 16:07:35 +0100
From:   Simon Horman <horms@verge.net.au>
To:     Geert Uytterhoeven <geert@linux-m68k.org>
Cc:     Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>,
        jacopo mondi <jacopo@jmondi.org>,
        Jacopo Mondi <jacopo+renesas@jmondi.org>,
        Magnus Damm <magnus.damm@gmail.com>,
        Rob Herring <robh+dt@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@vger.kernel.org>,
        Linux ARM <linux-arm-kernel@lists.infradead.org>,
        Linux-Renesas <linux-renesas-soc@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2 19/19] ARM64: dts: r8a77965: Add EtherAVB device node
Message-ID: <20180223150734.7ogbdjczcm4hb3oj@verge.net.au>
References: <cf157266-5ea7-6795-1e31-0d1ccf85e7e8@cogentembedded.com>
 <20180221100716.GF7203@w540>
 <cbeb1b6b-1b02-7665-0f1f-847ab8de69e3@cogentembedded.com>
 <20180221173106.vgyshbz3kkc4p3cu@verge.net.au>
 <321c6c8c-8853-ac00-7567-c0be82b510f4@cogentembedded.com>
 <35959135-cb3b-040a-1944-26845e0c07b0@cogentembedded.com>
 <20180221182340.wvdvgtv5pw7pvd3v@verge.net.au>
 <eaf3c72b-a482-86cf-88e7-4af457f74558@cogentembedded.com>
 <20180222155411.gnn42poawtlf3rze@verge.net.au>
 <CAMuHMdWj1v6WmfLY5JttZ+LMpSn2XvjgBS95Ag9TBohRRU1qTA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAMuHMdWj1v6WmfLY5JttZ+LMpSn2XvjgBS95Ag9TBohRRU1qTA@mail.gmail.com>
Organisation: Horms Solutions BV
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Feb 23, 2018 at 08:06:11AM +0100, Geert Uytterhoeven wrote:
> Hi Simon,
> 
> On Thu, Feb 22, 2018 at 4:54 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Wed, Feb 21, 2018 at 09:53:58PM +0300, Sergei Shtylyov wrote:
> >> On 02/21/2018 09:23 PM, Simon Horman wrote:
> >>
> >> > ...
> >> >
> >> >>>>>>>> +                    clocks = <&cpg CPG_MOD 812>;
> >> >>>>>>>> +                    power-domains = <&sysc 32>;
> >> >>>>>>>> +                    resets = <&cpg 812>;
> >> >>>>>>>> +                    phy-mode = "rgmii-txid";
> >> >>>>>>>
> >> >>>>>>>    Why not just "rgmii"? TX delay is a board specific detail, no?
> >> >>>>>>>
> >> >>>>>> I admit I took this one straight from r8a7796 dtsi.
> >> >>>>>> Would you like to me resend and change this?
> >> >>>>>
> >> >>>>>    Yes, unless Simon would fix it while merging...
> >> >>>>
> >> >>>> Can I confirm the desired change is s/rgmii-txid/rgmii/ ?
> >> >>>
> >> >>>    Yes.
> >> >>
> >> >>    Apparently that means that this prop should be overridden in the board file
> >> >> (which may not be an easy task given the board is Salvator-XS again).
> >> >>
> >> >> [...]
> >> >
> >> > Can we override it in r8a77965-salvator-x.dts or r8a77965-salvator-xs.dts?
> >>
> >>    In salvator-common.dtsi most probably -- it has the PHY data for Ether AVB.
> >>
> >> > I feel that I'm missing an important point here.
> >>
> >>    Well, r8a779{5|6}.dtsi also have phy-mode = "rgmii-txid" (which was
> >>    unjustified in my current understanding). Thus such board override
> >>    wouldn't hurt them. But we lack a patch modifying salvator-common.dtsi
> >>    in htis series, so I'm now thinking a respin of this series is needed
> >>    anyway... sorry for being unclear. :-)
> >
> > While I've applied other patches in this series I have
> > not applied this one - mainly to allow this discussion to conclude.
> >
> > Is an appropriate solution to do the following?
> >
> > 1) Atomically update
> >    a) r8a779{5|6}.dtsi to use "rgmii" and
> >    b) salvator-common.dtsi to use "rgmii-txid"
> > 2) Update this patch to use "rgmii"
> 
> No need for atomics.
> It can all be done by patches touching single files only:
>   1) Add "rgmii-txid" to board .dts(i) files,
>   2) Change s/rgmii-txid/rgmii/ in SoC .dtsi files.

Sounds good.

Jacopo would you care to:

a) Implement the above rgmii-txid / rgmii update.
b) Rebase this patch on top of that update and [PATCH v2 0/7] R-Car M3-N
   DTS fixes