[U-Boot] [PATCH] arm: imx: Switch Wandboard to use config_distro_bootcmd.h.

[XoD]

Hello,
I have now tested upstream uboot-imx with OpenELEC with SPL suppport for
wandbard.

It's work great without your patch. but with your path, OpenELEC don't boot.

The default uboot config assume than kernel file name is zImage.
But OpenELEC use a filename KERNEL.
The actual version of OpenELEC use a uEnv.txt file with this line :
zImage=/KERNEL
bootfile=/KERNEL
But this no longer work with your patch.

How I can configure uboot to open a file named KERNEL instead of zImage (at
build time or at runtime) ?

Thank you.

2015-05-20 9:36 GMT+02:00 XoD <xoddark at gmail.com>:

> Ok, thank you for your response.
>
> I waiting SPL for wandboard too.
> I like to add Wandboard support in OpenELEC (and Lakka) with SPL (to not
> have separate img for Wandboard dual and Wandboard quad.)
>
> XoD
>
> 2015-05-19 22:42 GMT+02:00 Vagrant Cascadian <vagrant at debian.org>:
>
>> On 2015-05-19, XoD wrote:
>> > Any new of this ? it can be merged ?
>>
>> I think my submitted patch was a bit too invasive by removing most of
>> the environment, and have reworked the patch to leave much of the
>> environment:
>>
>>
>> https://anonscm.debian.org/cgit/collab-maint/u-boot.git/tree/debian/patches/wandboard/config_distro_bootcmd.patch?h=experimental-2015.04
>>
>> I haven't yet reworked it for resubmission, but I'd be happy to do so.
>>
>> I was hoping to see the wandboard SPL support added before reworking, as
>> that will require a few minor changes to the config_distro_bootcmd patch
>> as well:
>>
>>   https://patchwork.ozlabs.org/patch/471092/
>>
>>
>> > I have tested and successfully boot a fedora on a Wandboard Quad with
>> this.
>> >
>> > The improvement of the console variable management can be done with an
>> > other patch ?
>>
>> It would be nice if switching to config_distro_bootcmd.h was not
>> dependent on sorting out the console variable switch... though I'd be
>> fine with switching the default console to include the baudrate as well.
>>
>>
>> > Thank you for working on this.
>>
>> Thanks for testing!
>>
>> live well,
>>   vagrant
>>
>>
>> > 2015-03-29 15:05 GMT+02:00 Tom Rini <trini at konsulko.com>:
>> >
>> >> On Sat, Mar 28, 2015 at 02:15:38PM +0100, Karsten Merker wrote:
>> >> > On Fri, Mar 27, 2015 at 06:24:43PM -0700, Vagrant Cascadian wrote:
>> >> >
>> >> > > This allows for more flexible and standardized boot across multiple
>> >> > > platforms. Remove most redundant legacy boot environment.
>> >> > >
>> >> > > Cc: Otavio Salvador <otavio at ossystems.com.br>
>> >> > > Signed-off-by: Vagrant Cascadian <vagrant at debian.org>
>> >> > > ---
>> >> > >
>> >> > >  include/configs/wandboard.h | 139
>> >> ++++++--------------------------------------
>> >> > >  1 file changed, 17 insertions(+), 122 deletions(-)
>> >> > >
>> >> > > diff --git a/include/configs/wandboard.h
>> b/include/configs/wandboard.h
>> >> > [...]
>> >> > >  #define CONFIG_EXTRA_ENV_SETTINGS \
>> >> > > -   "script=boot.scr\0" \
>> >> > > -   "image=zImage\0" \
>> >> > >     "console=ttymxc0\0" \
>> >> >
>> >> > Hello,
>> >> >
>> >> > regarding the boot environment standardization there is still the
>> >> > open topic of standardizing the console variable format for
>> >> > serial consoles - most platforms include the console baudrate in
>> >> > the console variable (e.g. "console=ttyS0,115200") while some
>> >> > others, in particular the i.MX6 platforms, do not. This means
>> >> > that distributions like Debian currently need to add special-case
>> >> > handling for i.MX6-based platforms in their boot scripts which
>> >> > goes against the idea of having one generic boot script for all
>> >> > platforms that use config_distro_bootcmd.h.
>> >> >
>> >> > It would be nice if the i.MX6 platforms could - while adopting
>> >> > config_distro_bootcmd.h and thereby changing their default
>> >> > environment to a large extend - also change their console
>> >> > variable from console=ttymxc0 to console=ttymxc0,115200.
>> >>
>> >> Yes please.  And Karsten can you do a patch that updates the README to
>> >> note that as an expectation?  Thanks!
>> >>
>> >> --
>> >> Tom
>>
>
>