mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Yegor Yefremov <yegorslists@googlemail.com>
To: Jan Remmet <J.Remmet@phytec.de>
Cc: barebox <barebox@lists.infradead.org>
Subject: Re: am335x: memory configuration
Date: Thu, 6 Oct 2016 15:50:24 +0200	[thread overview]
Message-ID: <CAGm1_ktk+pAF1+XLTdDa40TJesuxZk+518NCKeMhG72w1P6R7g@mail.gmail.com> (raw)
In-Reply-To: <57f65594.d4491c0a.37103.324dSMTPIN_ADDED_BROKEN@mx.google.com>

Hi Jan,

On Thu, Oct 6, 2016 at 3:45 PM, Jan Remmet <J.Remmet@phytec.de> wrote:
> On Thu, Oct 06, 2016 at 11:08:07AM +0200, Yegor Yefremov wrote:
>> I've submitted a patch [1] to support our Baltos systems back in the
>> May. There was an issue with dynamically detecting RAM size as the
>> systems can have either 256MB or 512MB modules.
>>
>> So far u-boot was able to detect the RAM size dynamically. In order to
>> do so it uses following code in arch/arm/cpu/armv7/am33xx/emif4.c:
>>
>> int dram_init(void)
>> {
>>
>> #ifndef CONFIG_SKIP_LOWLEVEL_INIT
>>         sdram_init();
>> #endif
>>
>>         /* dram_init must store complete ramsize in gd->ram_size */
>>         gd->ram_size = get_ram_size(
>>                         (void *)CONFIG_SYS_SDRAM_BASE,
>>                         CONFIG_MAX_RAM_BANK_SIZE);
>>         return 0;
>> }
>>
>> void dram_init_banksize(void)
>> {
>>         gd->bd->bi_dram[0].start = CONFIG_SYS_SDRAM_BASE;
>>         gd->bd->bi_dram[0].size = gd->ram_size;
>> }
>>
>> This peace of code will be excuted in u-boot.img (not in MLO).
>>
>> In Barebox we have get_ram_size(), that provides real RAM size and
>> am335x_sdram_size(), that returns RAM controller settings. So far DDR
>> controler will be configured for 512MB regardless of the module. Hence
>> the return value from am335x_sdram_size() cannot be used for crating
>> ram0 node.
> On our am335x boards we use MLOs with individual RAM timings and EMIF config
> values. So am335x_sdram_size works fine in barebox.bin.
> Maybe get_ram_size works if the timing settings are equal?

I've solved the problem the way Sascha proposed earlier (see the third
version of my patch sent today).

In MLO I setup memory controller to 512MB, then I check the real
amount via  get_ram_size() and if I have 256MB I just reinitialize the
memory controller with values for 256Mb i.e. rows 14 instead of 15.
With this done one can use am335x_sdram_size() in other parts of
berebox.

Thanks for the feedback.

Yegor

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

      parent reply	other threads:[~2016-10-06 13:51 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-06  9:08 Yegor Yefremov
2016-10-06 13:45 ` Jan Remmet
     [not found] ` <57f65594.d4491c0a.37103.324dSMTPIN_ADDED_BROKEN@mx.google.com>
2016-10-06 13:50   ` Yegor Yefremov [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=CAGm1_ktk+pAF1+XLTdDa40TJesuxZk+518NCKeMhG72w1P6R7g@mail.gmail.com \
    --to=yegorslists@googlemail.com \
    --cc=J.Remmet@phytec.de \
    --cc=barebox@lists.infradead.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox