The U-Boot on gitorious is more of a nostalgia piece, we actually
abandoned the repo a long time ago.
I am recommending this, and hoping Linaro stick to it: Linaro, Ubuntu,
Debian or any concerned distribution has absolutely NO business
building or flashing U-Boot for the boards.
You do not need to know codenames for the boards as described in U-Boot.
You do not need to even HAVE the source.
It should not be a requirement for a board to have it's firmware built
within the distribution it runs. This is, to my eyes, an absolutely
braindead endeavor.
You do not have to flash it.
We can provide the source code to anyone who asks, but I do ask in
return that you develop support for mainline, and Genesi will decide
if we ship it on our hardware. We have to maintain a correctly
operating consumer product here, it is not "usual" for any PC to
actively be reflashing and recompiling, or even providing firmware
binaries along with the operating system, whether they run Linux or
not.
The only suggestion I see in this thread that makes any sense is
documenting the bootcmd we have, and I will tell you it does the
following very simple task (which we will freely admit, we stripped
and modified from flash-kernel in Ubuntu, for Dove and OMAP3) in
pseudocode:
for each (mmc0, mmc1, ide)
if exists boot.scr
load it
if valid-image
run it
fi
fi
end
That's it. That's all it needs to do. What confused us was that
flash-kernel actively modified the bootcmd variable for Dove/OMAP to
insert this boot procedure. We decided it was silly, and this would be
better to ship it by default, so that all a boot.script has to do is
load kernel to loadaddr
load ramdisk to ramdiskaddr
setenv bootargs foo bar
bootm loadaddr ramdiskaddr
(there are two pertinent variables in U-boot that help us flash new
versons, "model" and a build "version" which is constructed from the
exact date and time it built, but you don't need to care about those.
We ship a single kernel for both boards now, so knowing which model it
is is ONLY relevant to U-Boot itself).
--
Matt Sealey
matt@genesi-usa.com
Product Development Analyst, Genesi USA, Inc.
2011/2/8 Eric Miao
eric.miao@canonical.com:
> On Thu, Feb 3, 2011 at 6:44 AM, Loïc Minier
loic.minier@linaro.org wrote:
>> On Wed, Feb 02, 2011, Per Förlin wrote:
>>> I made a new attempt bringing up my imx today, this time I reconnected
>>> the keyboard and then I could recover u-boot without any trouble. I
>>> disconnected the keyboard when changing the DIP switches.
>>
>> Per, upstream u-boot got support in git for efikamx some hours ago; I
>> managed to boot it this very morning! make efikamx_config and make,
>> then write u-boot.imx to SD at offset 0x400; boot with 4 DIP switches
>> to off.
>
> Hi All,
>
> Sorry being late on this.
>
> I guess we need some where to document all these tricks, which will be of
> great help to other developers. At least the below information from what I
> understand:
>
> 1. A bit explanation of the codenames, efikasb for smartbook and efikamx
> for smart top? As we may need to follow a consistent naming scheme.
>
> 2. The three possible boot up methods: a) Internal SPI NOR flash, b) the
> MicroSD card behind the battery and c) the normal SD card at the left
> side, not really sure about the situation on Efika MX (smart top) as
> I don't have one in my hands
>
> 3. The DIP switch states corresponding to each boot up method above
>
> 4. The boot sequence of the Internal SPI NOR flash, as it looks to me that
> it's trying to find boot.scr from either the MicroSD or SD card on the
> first partition? Also it would be helpful to document the envionment
> variables of this specific u-boot.
>
> 5. If Linaro is going to support u-boot for Efika MX/SB, it's better to
> follow what is in upstream. However, there could be some differences
> between the u-boot in the recovery/installing image downloadable from
> powerdeveloper.org and the one in upstream. We might need to figure
> out those differences and see how to handle them.
>
> 6. As Loic mentioned, would be of great help if the process of generating
> the recovery/installer images are documented. And linaro-media-create
> might get to be aware if there are any tricks.
>
> Matt,
>
> Maybe I missed the search, do you know if the information above is already
> documented somewhere? And if not, I'm not sure if a page on wiki.linaro.org
> will be fine?
>
> Some other small problems I met with u-boot on Efika are:
>
> 1. upstream u-boot building error for efikamx:
>
> efikamx.c: In function ‘board_init’:
> efikamx.c:624:27: error: ‘MACH_TYPE_MX51_LANGE51’ undeclared (first
> use in this function)
> efikamx.c:624:27: note: each undeclared identifier is reported only
> once for each function it appears in
> make[1]: *** [efikamx.o] Error 1
> make[1]: Leaving directory `/home/ycmiao/linaro/uboot-imx/board/efikamx'
> make: *** [board/efikamx/libefikamx.o] Error 2
>
> 2. check-out of git://gitorious.org/efikamx/uboot-efikamx.git didn't seem
> to build for a 'make efikasb_config'?
>
> Finally, it would be nice if a GUI support is possible as most users (even
> developers) don't have a serial cable. But that could be postponed to a
> later implementation.
>
> - eric
>
>>
>> I think we will have this in .debs soon; mainline kernel lack some
>> important stuff still though
>>
>> Cheers
>> --
>> Loïc Minier
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev@lists.linaro.org
>>
http://lists.linaro.org/mailman/listinfo/linaro-dev
>>
>
