Re: [ANNOUNCE] udev 171

[Paul Bender <pebender@san.rr.com>]

On 5/26/2011 11:47 PM, Gabor Z. Papp wrote:
> * Kay Sievers<kay.sievers@vrfy.org>:
>
> | udev 171
> | ====
> | Bugfixes.
>
> [...]
>
> | udev 169
> | ====
> | Bugfixes.
>
> | We require at least Linux kernel 2.6.32 now.
>
>    CC     extras/input_id/input_id.o
> extras/input_id/input_id.c: In function 'test_key':
> extras/input_id/input_id.c:167: error: 'BTN_TRIGGER_HAPPY' undeclared (first use in this function)
> extras/input_id/input_id.c:167: error: (Each undeclared identifier is reported only once
> extras/input_id/input_id.c:167: error: for each function it appears in.)
> make[2]: *** [extras/input_id/input_id.o] Error 1
> make[1]: *** [all-recursive] Error 1
> make: *** [all] Error 2
>
> $ uname -a
> Linux gzp 2.6.32.41 #1 SMP PREEMPT Tue May 24 09:17:23 CEST 2011 i686 GNU/Linux

I believe that BTN_TRIGGER_HAPPY was added to linux/input.h in the 
2.6.34 kernel.

Given that the udev function in question is attempting to differentiate 
between KEY_* and BTN_* events and given that KEY_* and BTN_* events are 
intermixed numerically, I cannot think of a reliable solution (even the 
current implementation is not reliable).

When I wrote eventlircd, my solution to this problem (I wanted to 
separate keyboard events from mice and joystick events) was to write an 
awk script that created a look up array for KEY_* versus BTN_* by 
parsing linux/input.h. I used the autoconf archives macro 
AX_ABSOLUTE_HEADER to locate linux/input.h at build time in the 
configure phase. While not ideal, it was the best hack I could come up with.