Subject: Re: [PATCH] driver/i2c/mux: Add register based mux i2c-mux-reg
To: ext York Sun <yorksun@freescale.com>, wsa@the-dreams.de
References: <1434475692-4611-1-git-send-email-yorksun@freescale.com>
 <55818C3C.8000704@nokia.com> <5581A3CA.4050805@freescale.com>
Cc: linux-kernel@vger.kernel.org, linux-i2c@vger.kernel.org,
        Peter Korsgaard <peter.korsgaard@barco.com>
From: Alexander Sverdlin <alexander.sverdlin@nokia.com>
Message-ID: <55827A1E.4020608@nokia.com>
Date: Thu, 18 Jun 2015 09:58:22 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101
 Thunderbird/38.0.1
MIME-Version: 1.0
In-Reply-To: <5581A3CA.4050805@freescale.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1164
Lines: 28

Hello!

On 17/06/15 18:43, ext York Sun wrote:
>> > Yeah, this is really bad idea. You maybe want something like
>> > __iomem "cookie" here instead of this bare pointer.
> Let me try.
> 

Could you think about different access widths, please?
Not all buses are 32-bits wide and even on 64-bit CPUs one might have 16-bit bus
and 32 bits accesses are not allowed or perform two accesses, etc...

So to cover the use-cases which I see one needs to have a possibility to select between
__raw_writeb()/__raw_writew()/__raw_writel()/__raw_writeq() (now that I'm thinking about
it, I think these native-Endianness functions are preferred and if one has a bus with
different Endianness he should think about the conversion in the reg property of subnodes).

Very important is readback with corresponding __raw_read*(), but maybe you want to do
this optional via additional DT property...

-- 
Best regards,
Alexander Sverdlin.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/