[PATCH] rpi: Set FDT for RPi CM4 to the IO Board one

[Ariel D'Alessandro]

Hi Matthias,

On 9/19/22 08:37, Matthias Brugger wrote:
> Hi Ariel,
> 
> On 16/09/2022 17:13, Ariel D'Alessandro wrote:
>> For the RPi CM4 (Compute Module 4), we currently try to load the dtb
>> file bcm2711-rpi-cm4.dtb, which is not built by the upstream kernel.
>>
>> Instead, the only CM4 dtb file provided by linux upstream is the
>> bcm2711-rpi-cm4-io.dtb, so let's use that.
>>
>> Signed-off-by: Ariel D'Alessandro <ariel.dalessandro at collabora.com>
>> ---
>>   board/raspberrypi/rpi/rpi.c | 4 ++--
>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/board/raspberrypi/rpi/rpi.c b/board/raspberrypi/rpi/rpi.c
>> index 17b8108cc8..b88f80ce05 100644
>> --- a/board/raspberrypi/rpi/rpi.c
>> +++ b/board/raspberrypi/rpi/rpi.c
>> @@ -168,8 +168,8 @@ static const struct rpi_model
>> rpi_models_new_scheme[] = {
>>           true,
>>       },
>>       [0x14] = {
>> -        "Compute Module 4",
>> -        DTB_DIR "bcm2711-rpi-cm4.dtb",
>> +        "Compute Module 4 IO Board",
>> +        DTB_DIR "bcm2711-rpi-cm4-io.dtb",
> 
> We had that discussion in the past. It is supposed that most CM4s will
> be used with a custom board. I don't think it's a good idea to tie the
> ID of the module to a specific IO board.

Ah, sorry I didn't notice that previous thread. Your point makes sense.
I guess another question would be: to avoid this inconsistency, does it
make sense to have a generic bcm2711-rpi-cm4.dtb built in the kernel?

Thanks!
Ariel