On Mon, Feb 17, 2025 at 07:10:54PM +0100, Jerome Brunet wrote:
> On Sat 15 Feb 2025 at 07:53, Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
>
> [...]
>
> >>
> >> >
> >> >> + int id)
> >> >> +{
> >> >> + struct auxiliary_device *auxdev;
> >> >> + int ret;
> >> >> +
> >> >> + auxdev = kzalloc(sizeof(*auxdev), GFP_KERNEL);
> >> >> + if (!auxdev)
> >> >> + return ERR_PTR(-ENOMEM);
> >> >
> >> > Ick, who cares what the error value really is? Why not just do NULL or
> >> > a valid pointer? That makes the caller much simpler to handle, right?
> >> >
> >>
> >> Sure why not
>
> I have tried the 'NULL or valid' approach. In the consumers,
> which mostly return an integer from their various init function, I got
> this weird to come up with one from NULL. EINVAL, ENOMEM, etc ... can't
> really pick one.
>
> It is actually easier to pass something along.
Ok, fair enough, thanks for trying. But I would have returned just
-ENODEV in all cases, as that's what the end result was :)
thanks,
greg k-h
