Re: No mouse on console

[Brad Boyer <flar@allandria.com>]

On Sat, Jun 21, 2003 at 07:53:55PM +0200, Wolfgang Pfeiffer wrote:
> Has anyone managed to get a USB mouse working on a Powerbook G4
> on the console?

It's not on a Powerbook, but I have a USB mouse working fine on
the console of my PowerMac G4 (AGP Graphics). It's pretty simple.

> Output in /var/log/syslog says (as it seems for disconnecting and then
> reconnecting the mouse to the USB connector):
> 
> ----------------------
> Jun 21 18:48:51 debby kernel: usb.c: USB disconnect on device 2
> Jun 21 18:48:58 debby kernel: hub.c: USB new device connect on bus1/1, assigned device number 3
> Jun 21 18:48:58 debby kernel: input4: Logitech USB-PS/2 Optical Mouse on usb1:3.0
> -----------------------

This is just the kernel drivers finding the device. Here's mine:

Jun 21 13:23:59 nene kernel: hub.c: USB new device connect on bus2/1/4, assigned device number 87
Jun 21 13:23:59 nene kernel: input1: Kensington Kensington USB/PS2 Trackball on usb2:87.0

I have a slightly more complicated USB setup, but it's the same basic stuff.

> gpm 1.19.6-12 is installed

I have gpm 1.19.6-12.1, so yours should be recent enough.

> gpmconfig says:
> --------------------
> Device: /dev/gpmdata
> Type: imps2
> Repeat_Type: raw
> --------------------

This is most likely your problem. Mine says this:

Device: /dev/input/mice
Type: autops2
Repeat_Type: ms3

> I *think* I had it connected to /dev/psaux before, but I'm not sure of
> that ...
> So far, none of the settings I tried got the mouse up and running

Try changing the device to /dev/input/mice, which is the output of
all mouse-like devices controlled by the generic input layer (includes
both USB and ADB in 2.4.x). The device /dev/gpmdata is almost certainly
wrong. That's usually used to feed mouse events from gpm to XFree86 in
a slightly old-fashioned way.

> Strange:
> 
> At the end of gpmconfig I'm told by this routine that the
> mouse interface server : gpm
> is first stopped, then started, but all this without any error message.

I suspect it's being stupid about not giving any errors when it
fails due to having the wrong device file.

> Or, as another possibility to solve this:
> 
> Is the Debian-Version I'm running -- 3.0 r1, stable -- so broken,
> incomplete, or whatever, that it will make sense to upgrade to testing?
> 
> After all there are more things here that do not work.
> 
> Could be very well that all this and other mess here is due to my lack of
> experience with Debian ... but if the stuff is simply buggy then please
> someone let me know ...

I have testing on my G4, but stable should be fine for most stuff. In
fact, there are problems with testing. They're just different problems.

	Brad Boyer
	flar@allandria.com