Re: [PATCH 7/7] media: i2c: dw9719: Fix power on/off sequence

[Sakari Ailus <sakari.ailus@linux.intel.com>]

On Thu, Sep 18, 2025 at 11:31:18PM +0200, André Apitzsch wrote:
> Hi Sakari,
> 
> Am Mittwoch, dem 17.09.2025 um 13:10 +0300 schrieb Sakari Ailus:
> > Hi André, Val,
> > 
> > On Tue, Sep 16, 2025 at 05:08:44PM -0300, Val Packett wrote:
> > > 
> > > On 9/15/25 5:48 PM, André Apitzsch wrote:
> > > > Hi Sakari,
> > > > 
> > > > @Val, please see below.
> > > > 
> > > > Am Montag, dem 18.08.2025 um 07:44 +0000 schrieb Sakari Ailus:
> > > > > Hi André,
> > > > > 
> > > > > On Sun, Aug 17, 2025 at 07:09:26PM +0200, André Apitzsch via B4
> > > > > Relay wrote:
> > > > > >   	u64 val;
> > > > > >   	int ret;
> > > > > >   	int err;
> > > > > > @@ -109,13 +116,15 @@ static int dw9719_power_up(struct
> > > > > > dw9719_device *dw9719, bool detect)
> > > > > >   	if (ret)
> > > > > >   		return ret;
> > > > > > -	/* Jiggle SCL pin to wake up device */
> > > > > > -	reg_pwr = (dw9719->model == DW9718S) ? DW9718S_PD :
> > > > > > DW9719_CONTROL;
> > > > > > -	cci_write(dw9719->regmap, reg_pwr, DW9719_SHUTDOWN,
> > > > > > &ret);
> > > > > > -	fsleep(100);
> > > > > > +	/*
> > > > > > +	 * Need 100us to transition from SHUTDOWN to
> > > > > > STANDBY.
> > > > > > +	 * Jiggle the SCL pin to wake up the device (even
> > > > > > when the
> > > > > > regulator
> > > > > > +	 * is shared) and wait double the time to be sure,
> > > > > > then
> > > > > > retry the write.
> > > > > Why double? Isn't the datasheet correct when it comes to the
> > > > > power-on sequence?
> > > > > 
> > > > I haven't noticed any problems during power-up of DW9761.
> > > > However, according to the commit message, there seems be an issue
> > > > with DW9718S. But I don't own the device and cannot test it.
> > > > 
> > > > Maybe Val can provided some additional information.
> > > 
> > > I haven't had access to the datasheet for the DW9718S, so this was
> > > all deduced experimentally. By "to be sure" I meant that I
> > > literally raised the timeout "just in case", not based on actual
> > > issues.
> > > 
> > > The actually important change was expecting the failure on the
> > > write and not erroring out.
> > 
> > Ack.
> > 
> > André: could you add this information to the comment as well?
> 
> Could you clarify which information.

What you now have in the commit message. It's relevant for the driver.

-- 
Sakari Ailus