Hi all. I'm a happy owner of a Dell XPS 1530 laptop where I use an external mouse and usually keep the internal touchpad disabled. Fedora bugzilla link: https://bugzilla.redhat.com/show_bug.cgi?id=448656 The bugzilla entry has more text, but I've tried to distil the info here so one doesn't have to wade over several pages of postings in order to get to the info. Kristian Högsberg Cc'd as he's the assigned redhat resource for that particular bug. The problem: Some time ago (I'm not sure when as I've been suspending the computer rather than rebooting it) I started noticing that I got some really weird input from the keyboard without me pressing any keys. The most frequently "fake-pressed" key is the equivalent of F7 (^[[18~). It seems to always generate escape-sequences similar to those of pressing either of the F-keys for some reason (not a blessing when 14000 help-browser windows suddenly try to spawn after the touchpad generating F1 sequences under Gnome). The mouse-pointer also jumps around the screen like a wild monkey randomly activating left or right button at near-tachyon speed. So far I haven't noticed it scrolling or middle-clicking. The troubleshooting: At first I thought it was the motherboard acting up (and got it replaced as it actually broke down 4 days after the problems started). After that I upgraded from Fedora 8 to Fedora 9, and none of the fedora9-shipped kernels work. Suspecting the technician was a bit clumsy or perhaps managed to short-circuit the touchpad, I loaded up Windows Vista, which works perfectly, ruling out hardware errors. Unfortunately, I need this laptop for more than watching porn and playing games, so running vista isn't really an option. I've tried booting latest master from linux-input.git (eff88f9927b8a54b378b4d4863eef488612b6dd7, v2.6.26-rc9-80-geff88f9 as of this writing), but to no avail. Finally, I located an ancient kernel (2.6.23.1-42) that doesn't have this particular problem but exhibits a much more common and less serious problem (tapping the touchpad doesn't work - everything else is fine). The next released kernel from fedora (2.6.24.3-12) doesn't work. dmesg output gets flooded with entries like this: psmouse.c: bad data from KBC - bad parity <about 10-30 of those> followed by input: Virtual ThinkFinger Keyboard as /devices/virtual/input/input11 psmouse.c: bad data from KBC - bad parity psmouse.c: Mouse at isa0060/serio1/input0 lost synchronization, throwing 1 bytes away. psmouse.c: failed to re-enable mouse on isa0060/serio1 psmouse.c: resync failed, issuing reconnect request psmouse.c: Failed to reset mouse on isa0060/serio1 input: PS/2 Generic Mouse as /devices/platform/i8042/serio1/input/input12 psmouse.c: Failed to enable mouse on isa0060/serio1 It loops over every 1.5-3 minutes or so. Progress so far:
From what information I've been able to gather from simple printk() debugging
is that it seems the kernel doesn't get further than the initial check in alps_detect on the touchpad but falls back to some version of "generic ps/2" mode. The patches the fedora team apply to their kernels are numerous, and I haven't been able to find a listing showing which git commit any particular fedora kernel is built upon. Not very many of them touch the drivers/input section though, and those that do do it either for the macbook touchpad driver or to suppress some user-scaring message. I've set my hopes to finding the bug somewhere in the range v2.6.23..v2.6.24, and that's what I'm currently bisecting. So far it's slow going though as I'm not very familiar with kernel hacking and I don't know a quick way to boot a kernel so that all the hardware with 100% certainty gets a hard reset. Any hints, help or pointers would be much appreciated. TIA -- Andreas Ericsson andreas.ericsson@xxxxxx OP5 AB www.op5.se Tel: +46 8-230225 Fax: +46 8-230231 -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
