Am 23.04.2017 um 12:41 schrieb Andreas Färber:
Am 22.04.2017 um 16:51 schrieb Heinrich Schuchardt:
On 04/22/2017 03:10 PM, Andreas Färber wrote:
Hi,
Am 22.04.2017 um 14:59 schrieb Heinrich Schuchardt:
On 01/15/2017 07:52 PM, Andreas Färber wrote:
Use the generic "distro" boot framework to enable automatic DHCP boot.
The fdt and kernel addresses are adopted from downstream; ramdisk and scriptaddr addresses were chosen arbitrarily.
Signed-off-by: Andreas Färber afaerber@suse.de
...
diff --git a/include/configs/odroid-c2.h b/include/configs/odroid-c2.h index 1d5b3d5..bf7e1af 100644 --- a/include/configs/odroid-c2.h +++ b/include/configs/odroid-c2.h @@ -12,6 +12,8 @@ #define CONFIG_CONS_INDEX 0 #define CONFIG_BAUDRATE 115200
+#define MESON_FDTFILE_SETTING "fdtfile=amlogic/meson-gxbb-odroidc2.dtb\0"
Hello Andreas,
do you remember why you chose to put meson-gxbb-odroidc2.dtb into subdirectory amlogic?
I chose that directory name when I added GXBaby support to the kernel:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch...
Note that we had a discussion about this a while ago when I sent patches to amend the distro framework's dtb search paths, and Tom's decision was to instead hardcode the vendor dirs inside $fdtfile.
The following mail by Tom Rini just mentions the the file name should match between Linux and U-Boot but does not ask for a vendor subdirectory hardcoded in fdtfile: https://lists.denx.de/pipermail/u-boot/2016-April/251709.html
And the dragonboard ended up with fdtfile=apq8016-sbc.dtb without vendor directory.
Could you, please, refer me to the mail with the decision that you mentioned.
Quoting "Re: [U-Boot] [PATCH 4/6] efi_loader: Improve .dtb search for arm64" from Jul 12 2016 20:05: "Please re-work the series around this idea."
... and as documented there I even brought up the point that not all distros might stick to that structure. Tom considered that unlikely. ;)
Regards, Andreas