evdev problems

[Magnus Vigerlöf]

On Wednesday 21 March 2007 22:58, Daniel Stone wrote:
[...]
> > What's best, to send the patches to the list or register a bug? Other
> > options?
>
> There are no real plans to change this, so if you send patches to the
> list, I'll merge them.  (I know I have some others pending -- sorry,
> been busy.)  That'd be great, thanks.

Thanks, the list it is. I hope the patches I'll make won't add too much to 
your backlog..

> > I'm also concerned about which memory actually gets freed when the
> > remove-request arrives for a hot-plugged device. But I haven't got that
> > far yet as the wacom driver is no saint in that part either.
>
> Yeah, it all really depends on the DDX and on the device.  As far as I
> know, the DIX part doesn't have any leaks there.

Good to know. I'll see what pops up as I try to get the wacom driver working.. 
I hope there won't be that much..

> > The 'old limits' part is the 20-device limit (inputInfo.numDevices) which
> > is increased for every call to xf86ActivateDevice [5,6] but never
> > decreased. When it get past 20 FatalError is called even if there's maybe
> > only 4 devices in the list... Would it be desireable to have an ever
> > increasing counter or is it possible to reuse the ids as soon as they are
> > freed?
>
> Oh god, I didn't even realise that.  We can bump the limit up by having
> it dynamically allocated.

:) Yes, but would it be a problem if we reused the newly freed device id as 
soon as possible (i.e. is the id a significant part of the connection to the 
X-clients or something)?

[...]
> > There's a proposal on the wiki [7] which outlines the functions that
> > could be used for what purpose for input devices. Last edit is Sept '05,
> > so one wonder is this is a good reference to use when updating the
> > current device drivers?
>
> No, it was never implemented at all, and doesn't necessarily bear much
> relation to what's been done.  To be honest, the input side of things
> isn't really documented at all.

Hmm... I'll put some documentation work on my todo-list once I understand 
enough of the calls needed for initialization and deinitialization. Is the 
status the same for the option functions? (Not knowing could explain why it's 
been used in differents ways and has some smaller leaks, fixes for frees of 
non-heap strings?)

> Thanks for your work!

No prob.. Without your hotplug work I wouldn't be here, so thank you for 
setting ball in motion..

Cheers
  Magnus