Re: [PATCH 1/1 FROM FIXED] Revert "HID: fix unit exponent parsing"

[Nikolai Kondrashov <spbnick@gmail.com>]

On 10/09/2013 09:42 PM, Nikolai Kondrashov wrote:
>> I would say that the current approach (without the revert) is exactly this:
>> - if the data is stored on only 1 byte ( if (!(raw_value&
>> 0xfffffff0))), do the two's complement -> any value less than 7 will
>> be the same, above are considered as negative.
>> - if not, then use the raw value.
>
> It is not exactly what I suggested. It also considers anything above 15 to be
> a normal integer. However, it might be a cleaner way.
>
> All-in-all, I'd say that the relevant hid-core.c code should have its comment
> fixed, and the hid-input.c (hidinput_calc_abs_res) change needs to be reverted
> as it (incorrectly) takes the component unit power into account for resolution
> calculation and makes the "unit" item value handling harder to comprehend.

On a second glance at the test data, hid-core.c needs to be fixed as well, as
for the non-nibble case it loses the sign by reading the item value as
unsigned. So a perfectly valid 1 byte 0xFD value becomes 253, instead of -3.

I'll include that into the patch.

Sincerely,
Nick