xf86-input-synaptics 1.0.99.3

[Peter Hutterer]

On Wed, Mar 04, 2009 at 09:27:34PM +0000, Colin Guthrie wrote:
> 'Twas brillig, and Colin Guthrie at 04/03/09 14:47 did gyre and gimble:
> > 'Twas brillig, and Colin Guthrie at 04/03/09 14:40 did gyre and gimble:
> >> 'Twas brillig, and Peter Hutterer at 04/03/09 06:58 did gyre and gimble:
> >>> Another snapshot before the release since a number of fixes went into this
> >>> one.
> >>>
> >>> Most notably, syndaemon updated to use device properties by default. And a fix
> >>> that caused synclient with properties to fail on 64-bit machines.
> >> Seems this is segv'ing for me :s
> >>
> >> Confirmed by another user too, (both of us are 64 bit).
> >>
> >> The device is currently configured in xorg.conf (not via hal yet as per 
> >> my previous mail), but only has Option SHMConfig on set and nothing more.
> >>
> >> (II) Synaptics touchpad driver version 1.0.99.3
> >> (--) SynapticsMouse1 auto-dev sets device to /dev/input/event2
> >> (**) Option "Device" "/dev/input/event2"
> >> (II) SynapticsMouse1: x-axis range 1472 - 5472
> >> (II) SynapticsMouse1: y-axis range 1408 - 4448
> >> (II) SynapticsMouse1: pressure range 0 - 255
> >> (II) SynapticsMouse1: finger width range 0 - 0
> >> (II) SynapticsMouse1: buttons: left right middle double triple
> >> (--) SynapticsMouse1 touchpad found
> >> (**) SynapticsMouse1: always reports core events
> >>
> >> Backtrace:
> >> 0: /etc/X11/X(xorg_backtrace+0x26) [0x4eca36]
> >> 1: /etc/X11/X(xf86SigHandler+0x3e) [0x480a3e]
> >> 2: /lib64/libc.so.6 [0x7fd0ae38dab0]
> >> 3: /lib64/libc.so.6(strlen+0x30) [0x7fd0ae3d7710]
> >> 4: /etc/X11/X(xf86ActivateDevice+0x6c) [0x4918bc]
> >> 5: /etc/X11/X [0x491a66]
> >> 6: /etc/X11/X(InitInput+0x40) [0x468790]
> >> 7: /etc/X11/X(main+0x37a) [0x42ecaa]
> >> 8: /lib64/libc.so.6(__libc_start_main+0xe6) [0x7fd0ae37a446]
> >> 9: /etc/X11/X [0x42e179]
> >>
> >> Fatal server error:
> >> Caught signal 11.  Server aborting
> > 
> > Forgot to say this is with xserver 1.6
> 
> Did some digging... turns out reverting this commit fixed it up:
> http://cgit.freedesktop.org/xorg/driver/xf86-input-synaptics/commit/?id=c719553dac875824b2d2a8f7714a89998ab4828d
> 
> As I said I (and my other victim) are both 64 bit, so I suspect the 
> "funny things" you mention are happening even with this commit :s

Thanks, fixed now with 1.0.99.4. A missing #define resulted in the struct
sizes being different in the driver and the server. Quite entertaining.

Cheers,
  Peter