Re: [PATCH] net: ipa: Remove redundant dev_err()

[Alex Elder]

On 12/11/22 10:34 AM, Heiner Kallweit wrote:
> On 11.12.2022 15:47, Kang Minchul wrote:
> > Function dev_err() is redundant because platform_get_irq_byname()
> > already prints an error.
> >
> > Signed-off-by: Kang Minchul <tegongkang@xxxxxxxxx>
> > ---
> >   drivers/net/ipa/gsi.c | 5 +----
> >   1 file changed, 1 insertion(+), 4 deletions(-)
> >
> > diff --git a/drivers/net/ipa/gsi.c b/drivers/net/ipa/gsi.c
> > index 55226b264e3c..585cfd3f9ec0 100644
> > --- a/drivers/net/ipa/gsi.c
> > +++ b/drivers/net/ipa/gsi.c
> > @@ -1967,11 +1967,8 @@ int gsi_init(struct gsi *gsi, struct platform_device *pdev, bool prefetch,
> >   
> >   	/* Get the GSI IRQ and request for it to wake the system */
> >   	ret = platform_get_irq_byname(pdev, "gsi");
> > -	if (ret <= 0) {
> > -		dev_err(gsi->dev,
> > -			"DT error %d getting \"gsi\" IRQ property\n", ret);
> > +	if (ret <= 0)
>
> According to the function description it can't return 0.

That's great!  I explicitly checked for 0 because at the
time I couldn't *prove* that 0 was an invalid return, and
there was nothing obvious in the function saying so.  I
*thought* it was invalid but lacked the guidance in the
code to know for sure.

Here is the commit that fixed that:
  a85a6c86c25be driver core: platform: Clarify that IRQ 0 is invalid

And it turns out that this particular block of code got
moved but not modified after that comment commit:
  0b8d676108451 net: ipa: request GSI IRQ later

Anyway, I'm very pleased this can be simplified.

					-Alex

> You can further simplify the code.
> And you patch should be annotated net-next.
>
> >   		return ret ? : -EINVAL;
> > -	}
> >   	irq = ret;
> >   
> >   	ret = request_irq(irq, gsi_isr, 0, "gsi", gsi);