U-Boot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Michael Walle <michael@walle.cc>
To: u-boot@lists.denx.de
Subject: [PATCH v4 2/3] env: Access Environment in SPI flashes before relocation
Date: Sat, 31 Oct 2020 01:25:16 +0100	[thread overview]
Message-ID: <f3601b07cc1286e74e81fcbaab31fbe1@walle.cc> (raw)
In-Reply-To: <20201031000703.GZ5340@bill-the-cat>

Hi,

Am 2020-10-31 01:07, schrieb Tom Rini:
[..]
>> > +static int env_sf_init(void)
>> > +{
>> > +#if defined(INITENV) && (CONFIG_ENV_ADDR != 0x0)
>> > +	return env_sf_init_addr();
>> > +#elif defined(CONFIG_ENV_SPI_EARLY)
>> > +	return env_sf_init_early();
>> > +#endif
>> > +	/*
>> > +	 * we must return with 0 if there is nothing done,
>> > +	 * else env_set_inited() get not called in env_init()
>> > +	 */
>> > +	return 0;
>> > +}
>> > +
>> >  U_BOOT_ENV_LOCATION(sf) = {
>> >  	.location	= ENVL_SPI_FLASH,
>> >  	ENV_NAME("SPIFlash")
>> >  	.load		= env_sf_load,
>> >  	.save		= CONFIG_IS_ENABLED(SAVEENV) ? ENV_SAVE_PTR(env_sf_save) :
>> > NULL,
>> > -#if defined(INITENV) && (CONFIG_ENV_ADDR != 0x0)
>> >  	.init		= env_sf_init,
>> > -#endif
>> 
>> This will break my board (the new kontron sl28). Before the
>> change, the environment is loaded from SPI, after this patch
>> the environment won't be loaded anymore. If I delete the
>> .init callback, the behavior is the same as before.
>> 
>> The problem here is that returning 0 in env_sf_init() is not
>> enough because if the .init callback is not set, env_init() will
>> have ret defaulting to -ENOENT and in this case will load the
>> default environment and setting env_valid to ENV_VALID. I didn't
>> follow the whole call chain then. But this will then eventually
>> lead to loading the actual environment in a later stage.
> 
> Ugh.  The games we play in the env area really just need to be
> rewritten.  So at this point I've merged the series, should I revert or
> do you see an easy fix for your platform?  Thanks!

Mh, my board cannot be the only one with a late environment
from SPI flash, can it?

Returning 0 will call env_set_inited() and returning -ENOENT
will call the second part. I can't tell why it was done that
way. So I don't have a simple fix.

But I guess the following ugly ifdeffery could do it:

#if (defined(INITENV) && (CONFIG_ENV_ADDR != 0x0)) || 
defined(CONFIG_ENV_SPI_EARLY)
	.init		= env_sf_init,
#endif

I can try that tomorrow. Also the comment about the return
value should be removed (?).

-michael

  reply	other threads:[~2020-10-31  0:25 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-10-10  8:28 [PATCH v4 0/3] env: Access Environment in SPI flashes before relocation Heiko Schocher
2020-10-10  8:28 ` [PATCH v4 1/3] env: split env_import_redund() into 2 functions Heiko Schocher
2020-10-12  3:35   ` Simon Glass
2020-10-30 18:46   ` Tom Rini
2020-10-10  8:28 ` [PATCH v4 2/3] env: Access Environment in SPI flashes before relocation Heiko Schocher
2020-10-12  3:35   ` Simon Glass
2020-10-30 18:46   ` Tom Rini
2020-10-30 23:51   ` Michael Walle
2020-10-31  0:07     ` Tom Rini
2020-10-31  0:25       ` Michael Walle [this message]
2020-11-02  5:20       ` Heiko Schocher
2020-10-10  8:28 ` [PATCH v4 3/3] imx6: enable early spi environment access on aristainetos boards Heiko Schocher
2020-10-30 18:46   ` Tom Rini

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=f3601b07cc1286e74e81fcbaab31fbe1@walle.cc \
    --to=michael@walle.cc \
    --cc=u-boot@lists.denx.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