Re: [PATCH] arm64: dts: ls1028a: make the eMMC and SD card controllers use fixed indices

[Michael Walle <michael@walle.cc>]

Am 2020-11-24 09:47, schrieb Y.b. Lu:
> Hi Michael,
> 
>> -----Original Message-----
>> From: Michael Walle <michael@walle.cc>
>> Sent: Tuesday, November 24, 2020 4:03 PM
>> To: Y.b. Lu <yangbo.lu@nxp.com>
>> Cc: Vladimir Oltean <vladimir.oltean@nxp.com>; Shawn Guo
>> <shawnguo@kernel.org>; Leo Li <leoyang.li@nxp.com>; Rob Herring
>> <robh+dt@kernel.org>; linux-arm-kernel@lists.infradead.org;
>> devicetree@vger.kernel.org; Adrian Hunter <adrian.hunter@intel.com>; 
>> Ulf
>> Hansson <ulf.hansson@linaro.org>; linux-mmc@vger.kernel.org;
>> linux-kernel@vger.kernel.org; Ashish Kumar <ashish.kumar@nxp.com>
>> Subject: Re: [PATCH] arm64: dts: ls1028a: make the eMMC and SD card
>> controllers use fixed indices
>> 
>> Am 2020-11-24 08:41, schrieb Y.b. Lu:
>> > Hi Vladimir,
>> >
>> >> -----Original Message-----
>> >> From: Vladimir Oltean <vladimir.oltean@nxp.com>
>> >> Sent: Friday, November 20, 2020 5:30 PM
>> >> To: Y.b. Lu <yangbo.lu@nxp.com>
>> >> Cc: Shawn Guo <shawnguo@kernel.org>; Leo Li <leoyang.li@nxp.com>; Rob
>> >> Herring <robh+dt@kernel.org>; linux-arm-kernel@lists.infradead.org;
>> >> devicetree@vger.kernel.org; Adrian Hunter <adrian.hunter@intel.com>;
>> >> Ulf
>> >> Hansson <ulf.hansson@linaro.org>; linux-mmc@vger.kernel.org;
>> >> linux-kernel@vger.kernel.org; Ashish Kumar <ashish.kumar@nxp.com>;
>> >> Michael Walle <michael@walle.cc>
>> >> Subject: Re: [PATCH] arm64: dts: ls1028a: make the eMMC and SD card
>> >> controllers use fixed indices
>> >>
>> >> On Fri, Nov 20, 2020 at 02:04:02AM +0000, Y.b. Lu wrote:
>> >> > Hi Vladimir,
>> >> >
>> >> > I have already upstreamed a patch for all affected layerscape boards.
>> >> >
>> >>
>> https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.kern
>> el.org%2Fpub%2Fscm%2Flinux%2Fkernel%2Fgit%2Fshawnguo%2Flinux.git%2
>> Fcommit%2F&amp;data=04%7C01%7Cyangbo.lu%40nxp.com%7C498622ade
>> e704fc0042008d8904f6184%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0
>> %7C0%7C637418017917635725%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiM
>> C4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000
>> &amp;sdata=OciS3q%2BmP%2Bz4x1ewPHDigmUkgIZmBgUlRRTm4yaxB7s%3D
>> &amp;reserved=0?
>> >> h=imx/dt64&id=342ab37ecaf8c1b10dd3ca9a1271db29a6af0705
>> >> >
>> >> > Please check whether it works for you.
>> >>
>> >> Thanks, one can tell that I haven't done my due diligence of checking
>> >> Shawn's tree first. I'll cherry-pick that patch and carry on with my
>> >> work.
>> >>
>> >> However, the fact still remains that Michael has expressed his opinion
>> >> regarding mmcblk0 vs mmcblk1. Do you think that we could make the
>> >> aliases a per-board option instead of per-SoC? Consider that there
>> >> might
>> >> even be boards that only use SD card. It would be strange for the
>> >> block
>> >> device in that case to be called /dev/mmcblk1.
>> >
>> > I don't think it's a problem in board dts to define board specific
>> > thing, like re-defining alias, and disabling any IP it not using.
>> 
>> First, why would you put it in the architecture include anyway? That
>> is really board-specific. That is like you would say, we enable all
>> devices and a board could potentially disable it. TBH it seems that
>> this will fit your reference boards and you don't care about the
>> other ones which uses that include.
> 
> In soc dtsi, this is giving default alias for two esdhc controllers.
> This is not board specific.
> That's natural esdhc0 is mmc0 and esdhc1 is mmc1.

How could this be not board specific if there are at least three
different use cases the board can choose from - and needs three
different configurations:

(1) eMMC at /dev/mmcblk0, SD card at /dev/mmcblk1
(2) SD card at /dev/mmcblk0, eMMC at /dev/mmcblk1
(3) no eMMC at all, SD card at /dev/mmcblk0

your include only support (1). If a board needs (2) or (3) it has to
override the configuration in the _common_ include.

>> And as Vladimir pointed out, what do you do if you just have the eMMC
>> on the LS1028A. It will be mmcblk1 unless you do something like the
>> following in the board dts:
>> 
>> mmc0 = &esdhc;
>> /delete-property/ mmc1;
>> 
>> That is really cumbersome, isnt it?
> 
> The soc dtsi gives default alias to make esdhc0 as mmc0, and esdhc1 as
> mmc1, the use case just needs to consider which esdhc controller is
> used. That's fixed index for it.
> No matter how the board is designed, there are two esdhc controllers
> in soc. It's probed as mmc0 and mmc1.
> It's use case that should choose the right mmc device. It is not the
> dts that should be changed to suit use case.
> If the board owner insists to change alias to make esdhc1 as mmc0, I
> think no problem. Just do it in board dts to override the default one.

Still, why would this be enforced in the common include? What is the
advnatage here? I only see disadvantages.

-michael