[U-Boot] [PATCH 5/5] dra7xx: Add dra72_evm_defconfig using CONFIG_DM

[Simon Glass]

Hi Tom.

On 2 August 2015 at 17:29, Tom Rini <trini at konsulko.com> wrote:
> On Sun, Aug 02, 2015 at 05:22:30PM -0600, Simon Glass wrote:
>> Hi Tom,
>>
>> On 2 August 2015 at 17:07, Tom Rini <trini at konsulko.com> wrote:
>> > On Sun, Aug 02, 2015 at 03:29:12PM -0600, Simon Glass wrote:
>> >> Hi Tom,
>> >>
>> >> On 31 July 2015 at 17:55, Tom Rini <trini at konsulko.com> wrote:
>> >> > - Import various DT files for DRA7 / DR72x / dra72-evm from Linux Kernel
>> >> >   v4.1
>> >> > - Add config file for this board, enable DM and DM_GPIO
>> >> >
>> >> > Cc: Simon Glass <sjg at chromium.org>
>> >> > Signed-off-by: Tom Rini <trini at konsulko.com>
>> >> >
>> >> > ---
>> >> > For some reason I can't yet figure out, enabling CONFIG_DM_SERIAL seems
>> >> > to be triggering the no console panic failure path, but I can't seem to
>> >> > figure out why.
>> >>
>> >> There is a bug in it but I suspect your problem is elsewhere:
>> >>
>> >> http://patchwork.ozlabs.org/patch/498648/
>> >
>> > I had that applied locally based on your hint, no help :(
>> >
>> >> You could enable the debug UART and print characters to see where it fails...
>> >
>> > I need to, yes, but one of my tests was commenting out the
>> > panic_str(...) and that went from a loop (since I didn't have
>> > CONFIG_PANIC_HANG set) to a hard crash I'm guessing a bit later on.
>>
>> If I could repeat this on BBB I could try to track it down. But I
>> suspect it is something else.
>
> Indeed, I was hoping it was a few things but I couldn't make
> am335x-gp-evm fail.
>
>> Is this in U-Boot proper? It really should pick up the stdout-path
>> right at the top of serial_find_console_or_panic() and return.
>
> Full U-Boot.  SPL_DM is off.  And yes, since we set stdout-path and have
> everything else set right in the DT it's very puzzling :(

Well if you can get a debug UART:

- check the return value of uclass_get_device_by_of_offset(). It
should be 0 assuming that you have u-boot,dm-pre-reloc set for your
console UART node
- if not, then lists_bind_fdt() should return 0, and then
device_probe() should return 0. For the former, something is horribly
wrong. For the latter check if your driver's probe function is
returning something odd...

Regards,
Simon