Re: [PATCH 1/4] Input: adp5588-keys: bail on returned error

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Wed, Oct 02, 2024 at 01:23:18PM +0200, Nuno Sá wrote:
> On Wed, 2024-10-02 at 04:09 -0700, Dmitry Torokhov wrote:
> > Hi Nuno,
> > 
> > On Wed, Oct 02, 2024 at 12:51:50PM +0200, Nuno Sa via B4 Relay wrote:
> > > @@ -455,8 +457,16 @@ static int adp5588_gpio_add(struct adp5588_kpad *kpad)
> > >  	for (i = 0; i <= ADP5588_BANK(ADP5588_MAXGPIO); i++) {
> > >  		kpad->dat_out[i] = adp5588_read(kpad->client,
> > >  						GPIO_DAT_OUT1 + i);
> > > +		if (kpad->dat_out[i] < 0)
> > > +			return kpad->dat_out[i];
> > > +
> > >  		kpad->dir[i] = adp5588_read(kpad->client, GPIO_DIR1 + i);
> > > +		if (kpad->dir[i] < 0)
> > > +			return kpad->dir[i];
> > > +
> > >  		kpad->pull_dis[i] = adp5588_read(kpad->client, GPIO_PULL1 + i);
> > > +		if (kpad->pull_dis[i] < 0)
> > > +			return kpad->pull_dis[i];
> > 
> > 
> > Unfortunately all these are u8 so they will never be negative. You need
> > to do the adp5588_read() refactor first and then (or maybe together) add
> > error checking.
> > 
> 
> Ahh crap... Completely missed that. Yeah, will see what looks better... Thanks for
> catching this.
> 
> BTW, this is also wrong in the adp5589 series.

I didn't get that far there ;)

Thanks.

-- 
Dmitry