Re: [PATCH] backlight: ktz8866: Convert to i2c's .probe_new()

[Jianhua Lu <lujianhua000@gmail.com>]

On Sat, Jan 28, 2023 at 05:16:13PM +0100, Uwe Kleine-König wrote:
> Hello,
> 
> On Sat, Jan 28, 2023 at 10:14:09PM +0800, Jianhua Lu wrote:
> > On Sat, Jan 28, 2023 at 02:32:39PM +0100, Uwe Kleine-König wrote:
> > > On Sat, Jan 28, 2023 at 08:36:28AM +0800, Jianhua Lu wrote:
> > > > I prefer that you pack this commit to the i2c-tree commit that drops
> > > > old .probe(). 
> > > 
> > > That's fine for me. Can I interpret this as an Ack for this patch?
> > 
> > Yes, but can't get my A-b directly, this patch should be ignored and 
> > resend it within the i2c-tree patch series or split it to two patch
> > series.
> 
> I'm not sure if I understand you correctly. Up to know I though you want
> the patch as is go in together with the patch that modifies struct
> i2c_driver such that the PR has in two separate commits:
> 
> 	i2c: Modify .probe() to not take an id parameter
> 	backlight: ktz8866: Convert to i2c's .probe_new()
This is case 1, the case 2 should be:
	Patch 1: i2c: Modify .probe() to not take an id parameter
	Patch 2: backlight: ktz8866: Convert to i2c's .probe_new()
					 'subsystem': 'i2c driver name': Convert to i2c's .probe_new()
					 ...
> 
> Did I understand that right?
> 
> In that case an Ack by you would be fine and welcome.
> 
> I don't want to squash the changes to the ktz8866 driver into the patch
> that modifies struct i2c_driver, as this needlessly clutters the commit,
> if it's that what you wanted. (There are more than 1000 i2c drivers and
> the others are not converted in a single lockstep, too.)
Do't squash this patch, I'd like you send a series patch instead of
a single patch.
> 
> Best regards
> Uwe
> 
> -- 
> Pengutronix e.K.                           | Uwe Kleine-König            |
> Industrial Linux Solutions                 | https://www.pengutronix.de/ |