[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <571E0785.1060200@arm.com>
Date: Mon, 25 Apr 2016 13:03:17 +0100
From: Vladimir Murzin <vladimir.murzin@....com>
To: Arnd Bergmann <arnd@...db.de>
Cc: Matthias Brugger <mbrugger@...e.com>, sudeep.holla@....com,
lorenzo.pieralisi@....com, linux@....linux.org.uk,
liviu.dudau@....com, linux-kernel@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH 3/5] ARM: dts: introduce MPS2 AN385/AN386
On 25/04/16 12:32, Arnd Bergmann wrote:
> On Monday 25 April 2016 11:15:35 Vladimir Murzin wrote:
>>>> +
>>>> +#include "mps2.dtsi"
>>>> +
>>>> +/ {
>>>> + model = "ARM MPS2 Application Note 385/386";
>>>> + compatible = "arm,mps2";
>>>> +
>>>> + aliases {
>>>> + serial0 = &uart0;
>>>> + };
>>>> +
>>>> + chosen {
>>>> + bootargs = "rdinit=/sbin/init earlycon";
>>>
>>> The rootfs should be independent from the board.
>>> In my opinion, this type of bootargs should not be part of the DTS.
>>
>> I have no strong opinion on this, but quick grep shows plenty places
>> where the same happens, and even [1] has rootfs specific option as an
>> example. So if there is no strong call to remove rdinit from bootargs,
>> I'd prefer to keep it as it makes testing easier for me.
>>
>> [1] Documentation/devicetree/booting-without-of.txt
>>
>
> I think "rdinit=/sbin/init" is redundant, that is what we try running
> anyway, so I don't see any downside in removing it.
Fair point.
@Sudeep could you fixup it locally or I need to respin?
Cheers
Vladimir
>
> Arnd
>
>
Powered by blists - more mailing lists