Re: [PATCH v2] i2c-hid: introduce HID over i2c specification implementation

[Jian-Jhong Ding <jj_ding-9cfG7bMpBgR9nmWX13WWKA@public.gmane.org>]

Benjamin Tissoires <benjamin.tissoires-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> writes:
> Hi JJ,
>
> On Thu, Oct 18, 2012 at 11:07 AM, Jian-Jhong Ding <jj_ding-9cfG7bMpBgR9nmWX13WWKA@public.gmane.org> wrote:
>> Hi Benjamin,
>>
>> Some suggestions to make the error messages more "human", and a little
>> question on the return value of i2c_hid_fetch_hid_descriptor.  Please see below:
>
> I fully agree with the more "human" error messages.
>
> However, for i2c_hid_fetch_hid_descriptor return values, I'm affraid I
> can't use -EINVAL.
>
> Jean Delvare (one of the i2c maintainers) told in his review of the v1:
> "
> These should all be -ENODEV in this function
> [i2c_hid_fetch_hid_descriptor]: the device isn't what you
> expected. EINVAL is for invalid argument.
> "

I must have missed that mail.  Thank you for pointing this out.

-JJ

> So ENODEV is the right return value.
>
> Anyway, thanks for the review.
>
> Cheers,
> Benjamin