[plug] Loss of X mouse control

Greg Gamble gregg at maths.uwa.edu.au
Wed Sep 23 19:29:03 WST 1998

According to Leon Brooks:
> Greg Gamble wrote:
> >   XFree86 3.3.2 patch level 3
> Using XFree86 here
> >   gpm-Linux 1.1.3
> Not using GPM here
> > when one has a
> >   Leadtek WinFast 3D S680  (AGP - if that matters)
> > video card which is 4MB, has chipset:
> >   S3 ViRGE/GX2
> Using an S3 with Virge/GX chipset (only known bug is that it says it
> does sync-on-green but doesn't actually do it).
> > and uses the
> >   XF86_SVGA
> Using this.
> > server. The file /etc/XF86Config has Pointer section consisting
> > of:
> Using a serial (not PS/2 or Bus) mouse.

  I have now discovered why I had a problem ... when I had used
the program `setup' I had wrongly defined my mouse as a MouseSystems
type and so at boot gpm was called with the switch `-t ms' when it
should have been `-t ps2'. Thus I didn't have mouse-control outside X
... being a newbie I didn't realise that this is the effect that gpm
had. My problem when `startx' was invoked was caused by X and gpm
fighting for control of the mouse. The "solution" that I posted 
effectively disabled gpm, so that startx worked. The correct solution
was to have

  gpm -R -m /dev/mouse -t ps2 

in the bootup /etc/rc.d/rc.local script and none of the other rubbish
I posted and to edit the /usr/X11R6/bin/startx script to have 

  gpm -k

at the beginning to kill gpm ... and to have

  gpm -R -m /dev/mouse -t ps2 

at the end to restart gpm on exit from X ... that was mentioned
somewhere in some documentation ... so it was no bug, just me being an
inept newbie. However the problem with the above "correct solution" is
that now startx can only properly be run by root (since gpm can only
be killed etc. by the superuser) ... this is not too much of a problem
since I'm the only user on my machine, but it is still an annoying
restriction ... is there any way around it?

  Greg Gamble <gregg at maths.uwa.edu.au>

More information about the plug mailing list