KDGKBENT and unicode

[tike64]

Dear linux-kernel,

I am trying to read keyboard in raw mode (K_MEDIUMRAW, to be exact). I 
hope to avoid reading all the keymaps from the kernel to the application 
for keycode translation. Therefore I call KDGKBENT ioctl for every 
keycode I get from the tty. Basically this works except, when the 
translation result is an unicode keysym or should I say the unicode 
value happens to be greater than 0xFF, I get K_HOLE.

This observation seems to be in harmony with the drivers/char/vt_ioctl.c 
code. There I learned that, if I had the keyboard in the K_UNICODE mode, 
I could get directly the 16 bit unicode values.

So I have two options:

1) Read the keymaps in K_UNICODE mode and do the translations myself 
while in K_MEDIUMRAW mode

2) For each keycode switch to K_UNICODE mode, ask the translation from 
kernel, switch back to K_MEDIUMRAW mode and read the next keycode (the 
continuous switching might have severe side effects).

The question is am I correct so far or am I miserably confused? Aren't 
there any smarter ways to do the translation?

The next question probably must be how do I handle CAPS with kernel 
keymaps? As the CAPS sensitivity is encoded in the type field of keysym 
value, it is lost when the value is 16 bit unicode. I am seeing this 
phenomenon in ordinary VT when I manage to load some keysyms as 16 bit 
unicode.

--

Timo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/