Hello Uwe,
Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx> writes:
> Now that all drivers are converted to the (new) .probe() callback, the
> temporary .probe_new() can go away. \o/
>
For some definition of "temporary" :)
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx>
> ---
> Hello,
>
> I was planning to send out this patch for inclusion into next after
> v6.5-rc1 as I thought there were only three drivers left to be converted
> before (on top of today's next/master). While preparing to submit this
> patch I noticed that a few more .probe_new were introduced in next even
> after commit 03c835f498b5 ("i2c: Switch .probe() to not take an id
> parameter"). I just sent out patches for all drivers in next that are
> still using .probe_new. These are:
>
> - w1: ds2482: Switch back to use struct i2c_driver's .probe()
> https://lore.kernel.org/lkml/20230612072807.839689-1-u.kleine-koenig@xxxxxxxxxxxxxx
> Krzysztof Kozlowski already signaled this won't go into 6.5-rc via
> his tree, and he's ok if it goes in via i2c then.
>
> - drm/i2c: Switch i2c drivers back to use .probe()
> https://lore.kernel.org/dri-devel/20230611202740.826120-1-u.kleine-koenig@xxxxxxxxxxxxxx
> Currently applied to
>
> https://anongit.freedesktop.org/git/drm/drm-misc.git drm-misc-next
>
> as d13b5d2b2b45. This won't make it onto v6.5-rc1 either.
>
Right, sorry for not getting that one in time...
[...]
> Until we're there, feel free to already look at this patch. I was unsure
> if I should split of the change to Documentation/hwmon/pmbus.rst. As
> .probe() already works in v6.3-rc2 this hunk could go in already
> earlier, e.g. via the hwmon tree. While thinking about that, the hunk
> for Documentation/i2c/writing-clients.rst could go in before, too.
> So tell me your thoughts, then I can split as desired.
>
I think that would be better to split indeed.
> Best regards
> Uwe
>
[...]
> - union {
> /* Standard driver model interfaces */
> - int (*probe)(struct i2c_client *client);
> - /*
> - * Legacy callback that was part of a conversion of .probe().
> - * Today it has the same semantic as .probe(). Don't use for new
> - * code.
> - */
> - int (*probe_new)(struct i2c_client *client);
> - };
> + int (*probe)(struct i2c_client *client);
> void (*remove)(struct i2c_client *client);
>
Thanks for your persistence to get to this point!
Reviewed-by: Javier Martinez Canillas <javierm@xxxxxxxxxx>
--
Best regards,
Javier Martinez Canillas
Core Platforms
Red Hat
