Linux Samsung SOC development
 help / color / mirror / Atom feed
From: Karol Lewandowski <k.lewandowsk@samsung.com>
To: arun.kk@samsung.com
Cc: Thomas Abraham <thomas.abraham@linaro.org>,
	"linux-samsung-soc@vger.kernel.org"
	<linux-samsung-soc@vger.kernel.org>,
	"devicetree-discuss@lists.ozlabs.org"
	<devicetree-discuss@lists.ozlabs.org>,
	Kukjin Kim <kgene.kim@samsung.com>,
	Kamil Debski <k.debski@samsung.com>,
	Jeongtae Park <jtp.park@samsung.com>,
	NAVEEN KRISHNA CHATRADHI <ch.naveen@samsung.com>,
	SUNIL JOSHI <joshi@samsung.com>
Subject: Re: [PATCH] ARM: EXYNOS: Add MFC device tree support
Date: Wed, 05 Sep 2012 11:42:44 +0900	[thread overview]
Message-ID: <5046BC24.8010704@samsung.com> (raw)
In-Reply-To: <31269552.67881346148504997.JavaMail.weblogic@epml04>

On 08/28/2012 07:08 PM, Arun Kumar K wrote:

> Hi Karol,
> Thanks for your comments. 
> Please find my response inline.

Hi... and sorry for so much delayed response.

>>> +
>>> +static void __init exynos5_reserve(void)
>>> +{
>>> +     s5p_mfc_reserve_mem(0x43000000, 8 << 20, 0x51000000, 8 << 20);
>>
>>
>> I think it would make sense to make this memory reservation dependent
>> on "mfc*" node being present in DTS.  It's to early to use of_* functions
>> (because tree is not populated at this stage) but fdt_* family of functions
>> work just fine.
>>
> 
> As I can see the fdt_* functions are not used in any of the ARM based SoC
> init codes. Though I can see some references in powerpc.
> The implementation and includes are present in arch/arm/boot/compressed/
> which I think cannot be used directly in mach-exynos unless we make some
> comon makefile changes. 


It looks like I was writing from memory, and I actually mixed things 
up.  To be clear this time - we can't use regular device tree handling
functions in reserve() as it's too early.  Namely, flattened device tree
is not yet converted to kernel's-natural representation.  However, 
I think we can scan fdt just fine.  To do so one just needs to use
functions declared here

  #include <linux/of_fdt.h>

Actual architecture-independent code is in drivers/of/fdt.c.  This
provides of_fdt_ family of functions. Please see below for example.


> Please clarify whether its ok to use fdt_* functions to parse the dts in 
> exynos machine init or please point me to some sample implementations
> which I can refer to.


It should be ok to use anything that works on flattened device tree
rather than its uncompressed version.  I've experimented a bit and
something like this worked for me just fine (it was around 3.3-kernel
timeframe, but I don't think that fdt api has changed):

[mach-exynos4-dt.c]

#include <linux/of_fdt.h>

int fdt_find_compat(unsigned long node, const char *uname, int depth, void *data)
{
        if (of_flat_dt_is_compatible(node, (char *)data))
                return 1;

        return 0;
}

static void __init exynos4210_dt_reserve(void)
{
        /* Reserve memory for MFC only if it's available */
        if (of_scan_flat_dt(fdt_find_compat, "samsung,s5pv210-mfc")) {
                printk(KERN_NOTICE "exynos4-dt: mfc device node found - setting up memory area for dma\n");
                s5p_mfc_reserve_mem(0x43000000, 8 << 20, 0x51000000, 8 << 20);
        }
}

[.dts]

 codec@some-addr {
      compatible = "samsung,s5pv210-mfc";
 };


So, in above code fragment I just check if mfc was defined in dts.
This could probably stay as it is.

Then I allocate _predefined_ region - and this part should be fixed.

If you have nodes like "mfc-r-size"/offset, then you could just get
this information directly from (f)dt rather than hardcoding it in the code.
Precisely, after we find compatible node we could do something like
following (untested):

unsigned long lsize, loff, rsize, roff len;
__be32 *prop;

prop = of_get_flat_dt_prop(node, "samsung,mfc-l-size", &len);
if (!prop)
   return;
lsize = of_read_ulong(prop, len/4);
...

Regards,
-- 
Karol Lewandowski | Samsung Poland R&D Center | Linux/Platform

  reply	other threads:[~2012-09-05  2:42 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-28 10:08 [PATCH] ARM: EXYNOS: Add MFC device tree support Arun Kumar K
2012-09-05  2:42 ` Karol Lewandowski [this message]
  -- strict thread matches above, loose matches on Subject: below --
2012-09-05  9:15 Arun Kumar K
2012-08-27 11:37 Arun Kumar K
2012-08-17  4:50 Arun Kumar K
2012-08-23  8:16 ` Kukjin Kim
2012-08-16 12:31 [PATCH] " Arun Kumar K
2012-08-16 12:31 ` [PATCH] ARM: EXYNOS: " Arun Kumar K
     [not found]   ` <1345120273-22913-2-git-send-email-arun.kk-Sze3O3UU22JBDgjK7y7TUQ@public.gmane.org>
2012-08-16 11:42     ` Thomas Abraham
2012-08-20  6:17   ` Karol Lewandowski

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=5046BC24.8010704@samsung.com \
    --to=k.lewandowsk@samsung.com \
    --cc=arun.kk@samsung.com \
    --cc=ch.naveen@samsung.com \
    --cc=devicetree-discuss@lists.ozlabs.org \
    --cc=joshi@samsung.com \
    --cc=jtp.park@samsung.com \
    --cc=k.debski@samsung.com \
    --cc=kgene.kim@samsung.com \
    --cc=linux-samsung-soc@vger.kernel.org \
    --cc=thomas.abraham@linaro.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