Re: [PATCH -next] mtd: docg3: Fix passing zero to 'PTR_ERR' warning in doc_probe_device

[Boris Brezillon]

On Wed, 23 Jan 2019 17:43:58 +0800
YueHaibing <yuehaibing@xxxxxxxxxx> wrote:

> On 2019/1/23 17:12, Boris Brezillon wrote:
> > On Wed, 23 Jan 2019 14:58:27 +0800
> > YueHaibing <yuehaibing@xxxxxxxxxx> wrote:
> >   
> >> Fix a static code checker warning:
> >> drivers/mtd/devices/docg3.c:1875
> >>  doc_probe_device() warn: passing zero to 'ERR_PTR'
> >>
> >> Fixes: ae9d4934b2d7 ("mtd: docg3: add multiple floor support")
> >> Signed-off-by: YueHaibing <yuehaibing@xxxxxxxxxx>
> >> ---
> >>  drivers/mtd/devices/docg3.c | 2 +-
> >>  1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/mtd/devices/docg3.c b/drivers/mtd/devices/docg3.c
> >> index 4c94fc0..60ddc38 100644
> >> --- a/drivers/mtd/devices/docg3.c
> >> +++ b/drivers/mtd/devices/docg3.c
> >> @@ -1872,7 +1872,7 @@ doc_probe_device(struct docg3_cascade *cascade, int floor, struct device *dev)
> >>  nomem2:
> >>  	kfree(docg3);
> >>  nomem1:
> >> -	return ERR_PTR(ret);
> >> +	return ret ? ERR_PTR(ret) : NULL;
> >>  }
> >>    
> > 
> > This is not the right fix, you should instead set ret to -EINVAL:  
> 
> Thanks , will fix it in v2.
> 
> And seems doc_probe_device cannot return NULL. The NULL check of doc_probe_device
> in docg3_probe also can been removed, right?

Hm, looks like the ret = 0 was intentional (to identify when ID is
invalid). Let's just keep your initial patch.