mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: "Tomaž Šolc" <tomaz.solc@klevio.com>
To: Roland Hieber <rhi@pengutronix.de>
Cc: "barebox@lists.infradead.org" <barebox@lists.infradead.org>,
	Moritz Augsburger <ml+barebox@moritz.augsburger.name>
Subject: Re: Accessing MMC/SD card storage on Raspberry Pi
Date: Tue, 12 Feb 2019 16:31:05 +0100	[thread overview]
Message-ID: <f803fb6f-763b-7027-a11e-086ad48b0054@klevio.com> (raw)
In-Reply-To: <20190212112240.sjrrrcdmysdyr3iy@pengutronix.de>

Hi

On 12. 02. 19 12:22, Roland Hieber wrote:
> If I remember correctly, there was a patch changing that behaviour... Ah
> yes, commit abc0447440d6 [0]:
> 
>      ARM: dts: bcm2835-rpi: re-enable booting from SD card
> 
>      Kernel commit a19adf8d86e822eba502486524127595372c85f6 [0] changed the
>      RPi1 device tree from &sdhci to &sdhost, which was imported in barebox
>      commit d14b844b08635c717fb52a294ed8d6872e260315. It seems that barebox
>      does not have a driver for brcm,bcm2835-sdhost (yet), which resulted in
>      barebox unable to boot from SD card. However, the driver for
>      brcm,bcm2835-sdhci worked fine in the past.
> 
>      Although the upstream change was made for good reasons, the simplest
>      change for barebox for now is to revert to using &sdhci and disable
>      &sdhost.
> 
> [0]: https://git.pengutronix.de/cgit/barebox/commit/?h=master&id=abc0447440d6
> 
> The "solution" was a small device tree fix. Could this be related to
> your problem? Unfortunately no one here has a RPi Compute Module, but
> there were already similar questions in the last month. (Cc Moritz)

Ah, thanks for pointing this out. That commit only touched device tree 
for R.Pi 1. I tried to do something similar on Compute Module 3 by 
adding the following into bcm2837-rpi-3.dts:

&sdhci {
	pinctrl-0 = <&emmc_gpio48>;
	/delete-property/ mmc-pwrseq;
	/delete-property/ non-removable;
	/delete-property/ #address-cells;
	/delete-property/ #size-cells;
	/delete-node/ wifi@1;
};

&sdhost {
	status = "disabled";
};

Unfortunately it fails to boot like that with the last things printed 
out being:

initcall-> rpi_devices_init+0x0/0x138
register_device: bcm2835_fb0
bcm2835_mci 3f300000.sdhci@7e300000.of: Wanted 400000 hz, returning 
divider 313 (626) which yields 399361 hz
bcm2835_mci 3f300000.sdhci@7e300000.of: Changing bus width to 1
bcm2835_mci 3f300000.sdhci@7e300000.of: IO settings: bus width=0, 
frequency=400000 Hz
bcm2835_mci 3f300000.sdhci@7e300000.of: Changing bus width to 1
bcm2835_mci 3f300000.sdhci@7e300000.of: IO settings: bus width=0, 
frequency=400000 Hz

One possible reason for this might be that pin functions aren't 
re-assigned. I think pinctrl isn't implemented for bcm2835. It probably 
works on R.Pi 1 because there firmware already assigns correct pin 
functions by default before jumping to Bearbox.

I'll see if I can implement pinctrl. It shouldn't be much work since 
most things already seem to be implemented in gpio-bcm2835.c.

Best regards
Tomaž

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

      reply	other threads:[~2019-02-12 15:31 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-02-11 15:09 Tomaž Šolc
2019-02-12 11:22 ` Roland Hieber
2019-02-12 15:31   ` Tomaž Šolc [this message]

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=f803fb6f-763b-7027-a11e-086ad48b0054@klevio.com \
    --to=tomaz.solc@klevio.com \
    --cc=barebox@lists.infradead.org \
    --cc=ml+barebox@moritz.augsburger.name \
    --cc=rhi@pengutronix.de \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox