Hi Frieder,
On 19.07.21 13:57, Frieder Schrempf wrote:
Hi Stefano,
On 17.07.21 14:08, Stefano Babic wrote:
Hi Frieder,
On 13.07.21 11:14, Frieder Schrempf wrote:
From: Frieder Schrempf frieder.schrempf@kontron.de
This adds support for i.MX6UL/ULL-based evaluation kits with SoMs by Kontron Electronics GmbH.
Currently there are the following SoM flavors (SoM-Line): * N6310: SOM with i.MX6UL-2, 256MB RAM, 256MB SPI NAND * N6311: SOM with i.MX6UL-2, 512MB RAM, 512MB SPI NAND * N6411: SOM with i.MX6ULL, 512MB RAM, 512MB SPI NAND
And the according evaluation boards (Board-Line): * N6310-S: Baseboard with SOM N6310, eMMC, display (optional), ... * N6311-S: Baseboard with SOM N6311, eMMC, display (optional), ... * N6411-S: Baseboard with SOM N6411, eMMC, display (optional), ...
Currently U-Boot describes i.MX6UL and i.MX6ULL through separate config options at compile-time. Though the differences are so minor, that for the scope of these SoMs we just use a single defconfig that is compatible with both SoCs.
Signed-off-by: Frieder Schrempf frieder.schrempf@kontron.de Reviewed-by: Stefano Babic sbabic@denx.de
Fixes in v2: * Add MAINTAINERS file * Rename board directory * Use binman to generate FIT
Something weird happens here. kontron_mx6ul is built well, however this breaks the rest of i.MX6UL boards. The point is when another board is built, but your DT is built as well due to:
+ imx6ul-pico-pi.dtb \ + imx6ul-kontron-n631x-s.dtb \ + imx6ull-kontron-n641x-s.dtb
In fact, I get :
Building current source for 5 boards (32 threads, 7 jobs per thread) arm: + mx6ul_14x14_evk +(mx6ul_14x14_evk) Error: arch/arm/dts/imx6ul-kontron-n6x1x-s-u-boot.dtsi:57.27-28 syntax error +(mx6ul_14x14_evk) FATAL ERROR: Unable to parse input tree +(mx6ul_14x14_evk) make[3]: *** [scripts/Makefile.lib:332: arch/arm/dts/imx6ul-kontron-n631x-s.dtb] Error 1 +(mx6ul_14x14_evk) make[3]: *** [scripts/Makefile.lib:332: arch/arm/dts/imx6ull-kontron-n641x-s.dtb] Error 1 +(mx6ul_14x14_evk) make[2]: *** [dts/Makefile:44: arch-dtbs] Error 2 +(mx6ul_14x14_evk) make[1]: *** [Makefile:1141: dts/dt.dtb] Error 2 +(mx6ul_14x14_evk) make: *** [Makefile:171: sub-make] Error 2 arm: + mx6ul_9x9_evk +(mx6ul_9x9_evk) Error: arch/arm/dts/imx6ul-kontron-n6x1x-s-u-boot.dtsi:57.27-28 syntax error +(mx6ul_9x9_evk) FATAL ERROR: Unable to parse input tree +(mx6ul_9x9_evk) make[3]: *** [scripts/Makefile.lib:332: arch/arm/dts/imx6ul-kontron-n631x-s.dtb] Error 1 +(mx6ul_9x9_evk) make[3]: *** [scripts/Makefile.lib:332: arch/arm/dts/imx6ull-kontron-n641x-s.dtb] Error 1 +(mx6ul_9x9_evk) make[2]: *** [dts/Makefile:44: arch-dtbs] Error 2 +(mx6ul_9x9_evk) make[1]: *** [Makefile:1141: dts/dt.dtb] Error 2 +(mx6ul_9x9_evk) make: *** [Makefile:171: sub-make] Error 2 3 0 2 /5 0:00:05 : mx6ul_9x9_evk
And same errors with other MX6UL boards. Just to check this, I hacked the Makefile as follows:
diff --git a/arch/arm/dts/Makefile b/arch/arm/dts/Makefile index 6030ff4aeb..89317f3979 100644 --- a/arch/arm/dts/Makefile +++ b/arch/arm/dts/Makefile @@ -809,7 +809,9 @@ dtb-$(CONFIG_MX6UL) += \ imx6ul-liteboard.dtb \ imx6ul-phytec-segin-ff-rdk-nand.dtb \ imx6ul-pico-hobbit.dtb \ - imx6ul-pico-pi.dtb \ + imx6ul-pico-pi.dtb
+dtb-$(CONFIG_TARGET_KONTRON_MX6UL) += \ imx6ul-kontron-n631x-s.dtb \ imx6ull-kontron-n641x-s.dtb
Note: if this could be an accepted solution, I have no problems to do it - on at91 is accepted, too. But I am wondering what is happening, it looks like a CONFIG_ is missing, but in the other boards and not yours.
Any idea ?
It looks like when boards don't have CONFIG_FIT set then CONFIG_FIT_EXTERNAL_OFFSET is not defined, but my devicetree uses this and is compiled for all of the boards as it is under CONFIG_MX6UL.
Instead of separating the dts, we could also add a "#if defined(CONFIG_FIT)" around the binman nodes in the dts. I saw that other boards also do this (e. g. socfpga_soc64_fit-u-boot.dtsi).
Yes, you're right. Anyway, it looks like just to make build happy, and I do not see any added value if Kontron's DT are compiled without the section. But as this is done for other boards, too, it is fine.
Shall I send an update with this fix?
Yes, please. Just the last two patches, I have already applied the first ones.
Regards, Stefano
Best regards Frieder