All of lore.kernel.org
 help / color / mirror / Atom feed
From: Daiane Angolini <daiane.angolini@freescale.com>
To: Carlos Rafael Giani <dv@pseudoterminal.org>,
	"meta-freescale@yoctoproject.org"
	<meta-freescale@yoctoproject.org>
Subject: Re: About customizing the image_types_fsl class
Date: Mon, 20 Jan 2014 12:12:00 -0200	[thread overview]
Message-ID: <52DD2EB0.5030208@freescale.com> (raw)
In-Reply-To: <52DCF67C.6030503@pseudoterminal.org>

On 20-01-2014 08:12, Carlos Rafael Giani wrote:
> Hello,
>
> I am working on revised hummingboard and cubox-i patches.
>
> One issue that has come up is its u-boot support.
> There is a new forked u-boot version for these machines, with SPL
> support. The way it is supposed to be built differs from the regular
> u-boot.imx generation.
> Building produces two files, u-boot.img and SPL. The SPL has to be
> flashed first, the u-boot.img right after.
>
> The details are here:
> http://imx.solid-run.com/wiki/index.php?title=Building_the_kernel_and_u-boot_for_the_CuBox-i_and_the_HummingBoard
>
> Excerpt:
>     Flashing SPL -  sudo dd if=SPL of=/dev/sdX bs=512 seek=2
>     Flashing u-boot.img as raw to the micro SD -  sudo dd if=u-boot.img
> of=/dev/sdX bs=1K seek=42
>
>
> I have been thinking about how to adapt this for meta-fsl-arm-extra. I
> essentially have to derive my own class from image_types_fsl and provide
> a new SDcard generation function. And to do that, I would have to copy &
> paste large parts of the existing mx6 sdcard generation command. This is
> not exactly clean.

I think the question here is how "standard" will SPL be for imx. How 
many boards has already SPL support *now*?

I think it's something we need to start including, because it's the next 
standard, however, we must make both working in parallel (spl and non-spl).

And, I would say, it's better to include the additional source code for 
SPL support directly to image_types_fsl instead of derivative it only on 
meta-fsl-arm-extra

>
> One other detail, which is less important but still present, is that
> these machines do _not_ expect the uImage to be outside of the
> partitions. They just load the uImage from the first partition by default.
>
> I know mainline u-boot got hummingboard and cubox-i support, but first I
> want to use something that has been tested by the machine vendors.
> (Plus, I am not sure how stable the current git mainline of u-boot is,
> and OE still uses 2013.10).

Overall I choose u-boot mainline always. It is our default bootloader, 
at least in general lines.

The u-boot mainline hummingboard stability can be known with simple 
test. And any additional support may be included. It's only a matter of 
planing.

2014.01 is about to be released, and u-boot-fslc is about to be update 
to that version. And we may thing about backport any accepted patch to 
2014.01 if it's planned only to 2014.04.

Conclusion: I think the best is u-boot mainline, even if it need some 
rework, it's the best long-term option, in my point of view.


>
> Suggestions? Comments?

Let's wait for more suggestions.


>
> cheers,
>    Carlos
>
>
> _______________________________________________
> meta-freescale mailing list
> meta-freescale@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/meta-freescale
>

Regards,
-- 
Daiane



  reply	other threads:[~2014-01-20 14:17 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-20 10:12 About customizing the image_types_fsl class Carlos Rafael Giani
2014-01-20 14:12 ` Daiane Angolini [this message]
2014-01-20 14:41   ` Carlos Rafael Giani
2014-01-20 15:00     ` Daiane Angolini
2014-01-20 17:40       ` Carlos Rafael Giani
2014-01-20 19:48         ` Fabio Estevam
2014-01-21 13:04           ` Carlos Rafael Giani
2014-01-21 13:22             ` Fabio Estevam
2014-01-21 13:28             ` Daiane.Angolini
2014-01-20 19:54         ` John Weber
2014-01-20 20:20           ` Otavio Salvador

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=52DD2EB0.5030208@freescale.com \
    --to=daiane.angolini@freescale.com \
    --cc=dv@pseudoterminal.org \
    --cc=meta-freescale@yoctoproject.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.