From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Subject: Re: [PATCH v2 2/2] IIO: st_accel_i2c.c: Use probe_new() instead of probe() To: Andy Shevchenko , Nikolaus Voss Cc: Jonathan Cameron , Hartmut Knaack , Lars-Peter Clausen , Peter Meerwald-Stadler , Lorenzo Bianconi , Linus Walleij , Xiongfeng Wang , linux-iio@vger.kernel.org, Linux Kernel Mailing List , nv@vosn.de References: <82c6f53cfa03f9bc7c0adfc423ae65fc986a1d25.1530599660.git.nikolaus.voss@loewensteinmedical.de> From: Javier Martinez Canillas Message-ID: <10258a21-db42-2c4e-91d6-e9227e11f53b@redhat.com> Date: Wed, 4 Jul 2018 12:49:29 +0200 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 List-ID: Hi Andy, On 07/04/2018 12:19 PM, Andy Shevchenko wrote: > Summon Javier to the discussion. > For my opinion he is an expert in this topic. I wouldn't call myself an expert in anything to be honest :) [snip] > >> The table is not used by the driver, but is necessary to >> >> a) bind an i2c device declared via i2c_board_info with type field set >> to one of the names of the i2c_device_id table to this driver >> b) bind an i2c device declared via DT or ACPI but with no match in of_id/ >> acpi_id table but an i2c_device_id table match to this driver (fallback >> matching) >> c) create the right modaliases at compile time for this driver to make >> module auto-loading work in case of a) and b) > I think Nikolaus is correct, assuming that the driver can be used on legacy platforms that register the I2C devices using board files / platform data. In that case you still need a I2C device ID table for (a) and (c) as he said. I don't buy on (b) though, that's a bug in my opinion. If you register an I2C device via DT then you must have a OF device ID entry that matches the device and the same for ACPI. You can't rely on the I2C device table to do the match. I would also remove the struct i2c_device_id .driver_data fields from the I2C device ID table, since are not used and just makes reading the code confusing (only the struct i2c_device_id .name is used as far as I can see). > Javier, just a summary of the above. Nikolaus switched one driver to > use ->probe_new() hook and left i2c ID table at the same time. > My understanding that this table is not anymore in use. > > But I have to admit I didn't see entire picture of this. Can you shed a light? > So to shed some light, in the past even {OF,ACPI}-only drivers needed an I2C ID table because: 1) the .probe callback had a struct i2c_device_id * parameter and 2) the I2C core always reported a modalias of the form i2c: even for devices registered via OF. The .probe_new callbacks solves (1) and the I2C core now reports of:N*T*Cfoo* solving (2). So the I2C device table isn't required anymore for {OF,ACPI}-only drivers, but it's still required for drivers that support legacy board files that calls i2c_register_board_info() directly. Same for the old .probe callback, it's needed if struct i2c_device_id .driver_data is used in the probe function. Best regards, -- Javier Martinez Canillas Software Engineer - Desktop Hardware Enablement Red Hat