linux-kernel - Re: [PATCH] ARM: dts: sun7i: a20: fix MMC index enumeration

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <40e45abf-7640-49a6-814b-9924510395c1@benettiengineering.com>
Date: Wed, 14 Jan 2026 18:50:00 +0100
From: Giulio Benetti <giulio.benetti@...ettiengineering.com>
To: Andre Przywara <andre.przywara@....com>
Cc: Rob Herring <robh@...nel.org>, Krzysztof Kozlowski <krzk+dt@...nel.org>,
 Conor Dooley <conor+dt@...nel.org>, Chen-Yu Tsai <wens@...nel.org>,
 Jernej Skrabec <jernej.skrabec@...il.com>,
 Samuel Holland <samuel@...lland.org>, devicetree@...r.kernel.org,
 linux-arm-kernel@...ts.infradead.org, linux-sunxi@...ts.linux.dev,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH] ARM: dts: sun7i: a20: fix MMC index enumeration

On 1/14/26 15:06, Andre Przywara wrote:
> On Wed, 14 Jan 2026 11:50:04 +0100
> Giulio Benetti <giulio.benetti@...ettiengineering.com> wrote:
> 
> Hi Giulio,
> 
>> Currently, the OLinuXino-A20-Micro experiences non-deterministic MMC
>> enumeration during boot. When both an SDIO WiFi module (MMC1) and
>> an SD card (MMC0) are present, the kernel may swap their indices.
>>
>> Explicitly define mmc* aliases to ensure consistent device naming
>> and prevent mapping conflicts.
> 
> So this comes up every now and then, but for sunxi it was decided to not
> support this. The enumeration of devices in the kernel is never guaranteed
> to follow some order, and this is widely accepted for SCSI (/dev/sda) and
> NVMe. Distros are coping with this forever since, using more sustainable
> designations like UUIDs or labels.

I see,

> If you don't want to change the way you communicate the root device to the
> kernel, you can actually use U-Boot's DTs, which adds the indicies for its
> own purposes, and can pass this on to the kernel. Just don't load a DT
> from storage (which you shouldn't be doing anyways), and just use
> $fdtcontroladdr as the DT address when booting the kernel.

Thanks for pointing, let's see if Buildroot Maintainers accept this
patch as is, otherwise I'll probably go with UUID.

Thank you for taking care and explaining.

Best regards
Giulio

> Cheers,
> Andre
> 
>> Signed-off-by: Giulio Benetti <giulio.benetti@...ettiengineering.com>
>> ---
>>   arch/arm/boot/dts/allwinner/sun7i-a20.dtsi | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/arch/arm/boot/dts/allwinner/sun7i-a20.dtsi b/arch/arm/boot/dts/allwinner/sun7i-a20.dtsi
>> index 5f44f09c5545..a958400bcd7c 100644
>> --- a/arch/arm/boot/dts/allwinner/sun7i-a20.dtsi
>> +++ b/arch/arm/boot/dts/allwinner/sun7i-a20.dtsi
>> @@ -55,6 +55,10 @@ / {
>>   	#size-cells = <1>;
>>   
>>   	aliases {
>> +		mmc0 = &mmc0;
>> +		mmc1 = &mmc1;
>> +		mmc2 = &mmc2;
>> +		mmc3 = &mmc3;
>>   		ethernet0 = &gmac;
>>   	};
>>   
>