Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
Date:   Sun, 12 Jun 2022 21:23:47 +0200
From:   Stephen Kitt <steve@sk2.org>
To:     Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc:     Alessandro Zummo <a.zummo@towertech.it>,
        Dianlong Li <long17.cool@163.com>,
        Wolfram Sang <wsa@kernel.org>, linux-rtc@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] rtc: use simple i2c probe
In-Reply-To: <YqSnwf7paZa1+uuh@mail.local>
References: <20220610162346.4134094-1-steve@sk2.org>
 <YqSnwf7paZa1+uuh@mail.local>
User-Agent: Roundcube Webmail/1.4.13
Message-ID: <2bd6ab78f649ba282c5369d0e969fea2@sk2.org>
Content-Type: text/plain; charset=UTF-8;
 format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hi Alexandre,

Le 11/06/2022 16:48, Alexandre Belloni a écrit :
> On 10/06/2022 18:23:43+0200, Stephen Kitt wrote:
>> All these drivers have an i2c probe function which doesn't use the
>> "struct i2c_device_id *id" parameter, so they can trivially be
>> converted to the "probe_new" style of probe with a single argument.
>> 
> 
> I think you should explain why you want to do that as the trend is to 
> do
> the exact opposite to allow support for those RTCS on x86 systems.

Indeed, I should have given more context. The idea is to continue the 
transition started with 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b8a1a4cd5a98a2adf8dfd6902cd98e57d910ee12 
(in 2016...). I²C drivers using probe() involve a call to i2c_match_id: 
in drivers/i2c/i2c-core-base.c,

         /*
          * When there are no more users of probe(),
          * rename probe_new to probe.
          */
         if (driver->probe_new)
                 status = driver->probe_new(client);
         else if (driver->probe)
                 status = driver->probe(client,
                                        i2c_match_id(driver->id_table, 
client));
         else
                 status = -EINVAL;

Many drivers don't actually need the second parameter, so instead of 
having probe() with both parameters, the goal is to switch to a probe 
function with only "struct i2c_client *". Probe functions that *do* need 
the "struct i2c_device_id" can call i2c_match_id themselves (as is done 
currently with of_match_id).

I discussed this briefly with Wolfram beginning of June at Kernel 
Recipes, and as I understood it the plan was still to continue with this 
transition (in fact, the plan was for *me* to continue with this 
transition).

Do you have an example of a change requiring id to support RTCs on x86 
systems?

Regards,

Stephen