Re: [Gc-linux-devel] Some coordination

[Free T. C. <Fre...@fr...>]

Michael Steil wrote:
> On 23.01.2004, at 09:49, Free The Cube wrote:
> 
>> You're right :) !!!! But the "temporary" driver is written :), it need 
>> more tests (it's a port of the GCLIB driver so it should work 
>> correclty (i hope) :)
>> I will work on the SI driver...
> 
> 
> Great. Jut send a kernel patch (preferably 2.6.1) to the mailing list 
> and we'll put it in the CVS for everyone to see it.
> ...and you need an SF CVS account yourself then.
> 

I can't send you a patch against 2.6.1, I'm at my work office ;) (no 
kernel sources, and can't spend time to download, test, ...) but I can 
send you the source file tested with linux 2.4.XX :) !

In my version, there are 2 parts :
  - the GCPAD driver
  - some functions (not finished) to test it with my 86 box...
I removed the last one in the attached file.

We still must adjust some values such as 'gcpad[i].dev.absmin' 
'gcpad[i].dev.absmax' !!! If i remember correctly, the trigger values go 
from -32767 to 32768, it is not really what we want (0 to 32768).

This driver (if it works :) ) is not really good :
  - the active loop wastes CPU time, even if i didn't feel this into my 
x86 box.
  - it doesn't use the SI layer
  - and the worst : it is not tested :) !

let us not be so pessimistic : it uses the standart input layer :D...

Free The Cube.