On 27/02/2023 15:39, Dmitry Rokosov wrote: > Hello Krzysztof! > > On Thu, Feb 23, 2023 at 10:09:25AM +0100, Krzysztof Kozlowski wrote: > > [...] > >>>> describes following compontents: CPU, GIC, IRQ, Timer, UART, >>>> PIN controller. It's capable of booting up into >>>> the serial console. >>>> >>>> This is based on arm64 version of meson-a1.dtsi. >>>> >>>> Signed-off-by: Alexey Romanov <avromanov@xxxxxxxxxxxxxx> >>>> --- >>>> arch/arm/boot/dts/meson-a1.dtsi | 151 ++++++++++++++++++++++++++++++++ >>> >>> There is such file and there is such DTS/hardware support. I don't see >>> any reason why entire DTSI should be duplicated. What's more, your >>> commit does not explain it - does not justify duplication. >> >> One more comment - I think you just added dead code. It's >> uncompilable/untestable. Otherwise, please share how to build this DTSI >> without DTS. > > You are right, Alexey doesn't provide any exact *.dts file for any > board, and *.dtsi file should be included somewhere, otherwise this is > dead code. > Unfortunately, our internal board *.dts file is useless for kernel > community, cause there is not any chance to burn locally compiled kernel > to our product due to secureboot protection. > But I think there is one possible option. We have reference Amlogic > boards somewhere in the office. So we can test 32-bit configuration on > it and prepare proper *.dts file for that. What do you think, it > reasanoble? You just need to provide valid board which works in 32-bit mode. Anyway duplicating DTSI is a no-go and we do not do it in other platforms. Best regards, Krzysztof
