Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Message-ID: <f8130c2a-f51d-21fd-d999-886ce9559e9c@ti.com>
Date:   Wed, 6 Apr 2022 14:18:38 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.7.0
Subject: Re: [PATCH RFC V1 net-next 3/4] net: Let the active time stamping
 layer be selectable.
Content-Language: en-US
To:     Richard Cochran <richardcochran@gmail.com>,
        Andrew Lunn <andrew@lunn.ch>
CC:     Kurt Kanzenbach <kurt@linutronix.de>,
        Michael Walle <michael@walle.cc>, <davem@davemloft.net>,
        <kuba@kernel.org>, <linux-kernel@vger.kernel.org>,
        <linux@armlinux.org.uk>, <mlichvar@redhat.com>,
        <netdev@vger.kernel.org>, <qiangqing.zhang@nxp.com>,
        <vladimir.oltean@nxp.com>
References: <20220104014215.GA20062@hoboy.vegasvil.org>
 <20220404150508.3945833-1-michael@walle.cc> <YksMvHgXZxA+YZci@lunn.ch>
 <e5a6f6193b86388ed7a081939b8745be@walle.cc> <877d83rjjc.fsf@kurt>
 <ad4a8d3efbeaacf241a19bfbca5976f9@walle.cc> <87wng3pyjl.fsf@kurt>
 <defe77d9-1a41-7112-0ef6-a12aa2b725ab@ti.com> <YkxEIZfA0H8yvrzn@lunn.ch>
 <20220405154821.GB6509@hoboy.vegasvil.org>
From:   Grygorii Strashko <grygorii.strashko@ti.com>
In-Reply-To: <20220405154821.GB6509@hoboy.vegasvil.org>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 05/04/2022 18:48, Richard Cochran wrote:
> On Tue, Apr 05, 2022 at 03:29:05PM +0200, Andrew Lunn wrote:
> 
>> Maybe. Device tree is supposed to describe the hardware, not how you
>> configure the hardware. Which PTP you using is a configuration choice,
>> so i expect some people will argue it should not be in DT.
> 
> +1
> 
> Pure DT means no configuration choices.
> 
> (but you find many examples that break the rules!)
> 

My point was related to one of issues described by Michael Walle in this thread:
- supporting TS by the PHY may require also additional board support;
- phy_has_hwtstamp() defined statically by PHY drivers without taking into account board design;
- Kconfig option Doesn't really work with generic distro support and not allowed per-port cfg.

So adding smth like "hwtstamp-en" will clear identify that this particular PHY on this particular board
supports time stamping.
(or hwtstamp-full/hwtstamp-rx/hwtstamp-tx).

Of course, it will not help with default or dynamic selection of time stamping layer :(,
but it will be one problem less.

-- 
Best regards,
Grygorii, Ukraine