Problem with input device

Joe Shaw joeshaw at
Thu Jun 3 15:24:59 PDT 2004

On Fri, 2004-06-04 at 00:15 +0200, Kristian Høgsberg wrote:
> I'll give it a shot.  I think all of the /proc info is also available 
> from ioctls on the device files, I'll use that instead of parsing.

Ah, ok, that's probably better.

> I was thinking about just changing the input class accept function from 
> class_device_accept() to a custom accept function that accepts even if 
> there is no sysfs device.  How far would that get me?

You'll need to write a compute_udi function as well, but I think that's
all you need.  If the pre_process function filled out identifying info,
then it'd work out rather nicely.


hal mailing list
hal at

More information about the Hal mailing list