[Letux-kernel] Raspi3B+ support in Letux Kernel

H. Nikolaus Schaller hns at goldelico.com
Wed Jan 9 10:54:28 CET 2019


> Am 09.01.2019 um 10:21 schrieb H. Nikolaus Schaller <hns at goldelico.com>:
> 
> Hi,
> 
>> Am 08.01.2019 um 21:59 schrieb H. Nikolaus Schaller <hns at goldelico.com>:
>> 
>> Hi all,
>> 
>> With that I now get a __symbols__ section in the DTB. But there
>> is no __overrides__. Strange, how that is generated from the github
>> tree. Maybe it is explicitly in some .dtsi there?
> 
> I have now done a couple of additional tests (on Udoo neo with SD card reader):
> 
> First test:
> 
> 	dtc -s -I dtb -O dts /media/letux/boot/bcm2710-rpi-3-b-plus.dtb >bcm2710.dts
> 	dtc -s -I dts -O dtb bcm2710.dts >/media/letux/boot/bcm2710-rpi-3-b-plus.dtb
> 	dtc -s -I dtb -O dts /media/letux/boot/bcm2710-rpi-3-b-plus.dtb >bcm2710-2.dts
> 	diff bcm2710.dts bcm2710-2.dts
> 	-> no diffs
> 
> This shows that the non-kernel dtc does not loose any information.
> 
> And: this dtb boots!
> 
> Next test:
> 
> 	vi bcm2710.dts
> 		remove the __overrides__ section
> 	dtc -s -I dts -O dtb bcm2710.dts >/media/letux/boot/bcm2710-rpi-3-b-plus.dtb
> 
> This dtb still boots which means the __overrides__ section is not important.
> 
> 3rd test:
> 
> 	vi bcm2710.dts
> 		remove the __symbols__ section
> 	dtc -s -I dts -O dtb bcm2710.dts >/media/letux/boot/bcm2710-rpi-3-b-plus.dtb
> 
> Now ... this also boots!
> 
> Which means that a missing -@ is not the problem...
> 
> So it must be a detail in the upstream dts sources itself. Maybe a missing pinmux for the UART?
> Or a differently set clock or power supply? Or the swapped uart numbers/names/aliases.
> 
> I still wonder how the upstream developers of Raspi patches do boot on their devices...

Maybe studying https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=f08f58a2bf689
gives more insights...

BR,
Nikolaus



More information about the Letux-kernel mailing list