From mboxrd@z Thu Jan  1 00:00:00 1970
From: sebastian.hesselbarth@gmail.com (Sebastian Hesselbarth)
Date: Mon, 19 Aug 2013 01:01:54 +0200
Subject: [RFC v1 5/5] ARM: mvebu: add board init for Armada 1500
In-Reply-To: <201308172108.38824.arnd@arndb.de>
References: <1376682098-10580-1-git-send-email-sebastian.hesselbarth@gmail.com>
 <1376682098-10580-6-git-send-email-sebastian.hesselbarth@gmail.com>
 <20130816204831.GX13964@titan.lakedaemon.net>
 <201308172108.38824.arnd@arndb.de>
Message-ID: <52115262.70402@gmail.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On 08/17/2013 09:08 PM, Arnd Bergmann wrote:
> On Friday 16 August 2013, Jason Cooper wrote:
>>> +
>>> +#define ARMADA_1500_REG_BASE_VIRT	0xf6000000
>>> +#define ARMADA_1500_REG_BASE_SIZE	0x03000000
>>> +
>>> +static struct map_desc armada_1500_io_desc[] __initdata = {
>>> +	{
>>> +		.virtual	= ARMADA_1500_REG_BASE_VIRT,
>>> +		.pfn		= __phys_to_pfn(ARMADA_1500_REG_BASE_VIRT),
>>> +		.length		= ARMADA_1500_REG_BASE_SIZE,
>>> +		.type		= MT_DEVICE,
>>> +	},
>>> +};
>
> You should really try to find out what driver uses this. If you have a requirement
> that VIRT == PHYS here, the most likely explanation is that some driver accidentally
> uses readl/writel on the physical address rather than on the result of ioremap.
>
> You can try shrinking the area using bisection until you have found the offending
> driver based on the address.

While bringing up timer and irq, I had no luck without that mapping,
but I didn't try without later on. There is no driver requiring it,
as I either only use mainline drivers that never knew about A1500
before, or I have written it on my own (dw-apb-ictl).

I don't think it is required and will try to remove it, now that I
know I have set up core stuff correctly and it boots.

>>> +static void __init armada_1500_timer_and_clk_init(void)
>>> +{
>>> +	of_clk_init(NULL);
>>> +	clocksource_of_init();
>>> +}
>>> +
>>> +static void __init armada_1500_dt_init(void)
>>> +{
>>> +	l2x0_of_init(0x70c00000, 0xfeffffff);
>
> New platforms should call this as 'l2x0_of_init(0, 0);' and get the bits from DT.

Ok.

> Note that we should really change the common code to do both the of_clk_init()
> and the l2x0_of_init() automatically, but that needs to be done with some care,
> in order to not break any of the existing platforms. Would you be able to do
> one of the two? We can then get the next person that wants to add a platform
> to do the last one ;-)

Scary but that reduces armada-1500.c to MACH descriptor and DT
compatible only ;) I don't expect any weird hacks required for
it anyway.

Sebastian