Re: [PATCH 5/6] rtc: abx80x: create abx80x_i2c_probe()

On Wed, Jul 22, 2026 at 08:10:57PM +0000, [email protected] wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> 
> Pre-existing issues:
> - [High] RV1805 hardware workarounds are skipped during auto-detection.
> --
> 
> --- Patch [5]: [PATCH 5/6] rtc: abx80x: create abx80x_i2c_probe() ---
> commit 0661079a8b8384fb14bc67f5100b45ebe0a01c3e
> Author: Antoni Pokusinski <[email protected]>
> 
> rtc: abx80x: create abx80x_i2c_probe()
> 
> Move the I2C-specific code from abx80x_probe() to the newly created
> function. This is a preparation for introducing the support for
> ABX81X RTCs which use SPI interface.
> 
> > diff --git a/drivers/rtc/rtc-abx80x.c b/drivers/rtc/rtc-abx80x.c
> > index 7761cb353994e..6e0c14695bdf1 100644
> > --- a/drivers/rtc/rtc-abx80x.c
> > +++ b/drivers/rtc/rtc-abx80x.c
> [ ... ]
> > @@ -855,7 +820,7 @@ static int abx80x_probe(struct i2c_client *client)
> >  					ABX8XX_OUT_CTRL_EXDS,
> >  					ABX8XX_OUT_CTRL_EXDS);
> >  		if (err < 0) {
> > -			dev_err(&client->dev,
> > +			dev_err(dev,
> >  				"Unable to write output control register\n");
> >  			return -EIO;
> >  		}
> > @@ -867,26 +832,24 @@ static int abx80x_probe(struct i2c_client *client)
> >  			if (partnumber == abx80x_caps[i].pn)
> >  				break;
> >  		if (abx80x_caps[i].pn == 0) {
> > -			dev_err(&client->dev, "Unknown part: %04x\n",
> > -				partnumber);
> > +			dev_err(dev, "Unknown part: %04x\n", partnumber);
> >  			return -EINVAL;
> >  		}
> >  		part = i;
> >  	}
> 
> [Severity: High]
> This is a pre-existing issue, but are the RV1805 hardware workarounds being
> skipped during auto-detection?
> 
> Looking at abx80x_probe() in drivers/rtc/rtc-abx80x.c, the RV1805 specifics
> are configured before the part auto-detection logic resolves the hardware type:
> 
>     /* Configure RV1805 specifics */
>     if (part == RV1805) {
>         ...
>     }
> 
>     /* part autodetection */
>     if (part == ABX80X) {
>         for (i = 0; abx80x_caps[i].pn; i++)
>             if (partnumber == abx80x_caps[i].pn)
>                 break;
>         ...
>         part = i;
>     }
> 
> When the driver is bound using the generic "abracon,abx80x" compatible string
> or I2C ID, the part argument is initialized to ABX80X. Because the workarounds
> check if (part == RV1805) before part is updated to RV1805, they appear to be
> silently bypassed.
> 
> Could this cause the RV1805 RTC to accidentally enter test mode and suffer
> from extra power leakage?
Yes, this makes sense. The fix would be to simply move the autodetection before
the "Configure RV1805 specifics" section. Will add this as a separate
patch in v2.

> 
> -- 
> Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=5