Bug 27424 - No input devices work and then mapping doesn't work
Summary: No input devices work and then mapping doesn't work
Status: RESOLVED INVALID
Alias: None
Product: xorg
Classification: Unclassified
Component: Input/evdev (show other bugs)
Version: unspecified
Hardware: Other Linux (All)
: medium major
Assignee: Peter Hutterer
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-04-02 08:00 UTC by Dave Gilbert
Modified: 2010-04-21 21:53 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Xorg.0.log with broken keyboard mapping (36.67 KB, text/plain)
2010-04-02 08:00 UTC, Dave Gilbert
no flags Details

Description Dave Gilbert 2010-04-02 08:00:22 UTC
Created attachment 34615 [details]
Xorg.0.log with broken keyboard mapping

Hi,
  I'm running Ubuntu Lucid Linux beta with xorg-edgers on x86-64 with a USB mouse and a PS/2 keyboard (Model M, GB layout) and have just pulled in an update that has broken input in a couple of ways that may be related.

The current version shows as: 2:1.7.99.2~git20091220.0cb638dc-0ubuntu0tormod

X started fine but I got no input - mouse or keyboard; the display was still updated and I could start an xterm onto it; both keyboard/mouse were shown in the output of hal-device (although Lucid is udev not hal)

There were no 'Adding device' lines in the /var/log/Xorg.0.log

I added a xorg.conf with:


Section "ServerFlags"
	Option "AutoAddDevices" "false"
EndSection

and that got me a mouse but no keyboard, so I added:


Section "InputDevice"
        Identifier      "Generic Keyboard"
        Driver          "EvDev"
        Option          "XkbRules"      "evdev"
        Option          "XkbModel"      "pc102"
        Option          "XkbLayout"     "gb"
#        Option          "XkbOptions"    "lv3:ralt_switch"
	Option 		"Device"	"/dev/input/by-path/platform-i8042-serio-0-event-kbd"
EndSection

now I have a keyboard, but the keys in the arrow/insert/printscreen block are completely mismapped (up arrow seems to be print screen and other hilarity).

That latter symptom feels like bug 24557 but shouldn't the XkbRules option work around that?

Dave
Comment 1 Ed Tomlinson 2010-04-02 08:09:54 UTC
I've got similar problems here.

Section "InputDevice"
        Identifier  "Keyboard0"
        Driver      "evdev
       option      "Device" "/dev/input/event0"
#        option      "CoreKeyboard"
        option      "XkbModel" "pc105"
        option      "XkbLayout" "ca"
        option      "XkbVariant" "fr"
        Option      "XkbOptions" "terminate:ctrl_alt_bksp"
EndSection

My up and down keys do not work.  Interestingly xev shows the correct keys pressed...

I am using gentoo with xorg 1.8
Comment 2 Dave Gilbert 2010-04-02 08:11:30 UTC
Also, synergy makes these keys work on the remote synergy client but not on the local machine running the server.
Comment 3 Ed Tomlinson 2010-04-02 08:26:27 UTC
Using the old kbd driver (1.4) bypasses the problem here - would be nice to be able to use evdev for keyboards though...
Comment 4 Dave Gilbert 2010-04-02 10:23:31 UTC
Thanks to Sarvatt on #ubuntu-x for spotting I'd got an ancient package in my xorg-edgers installation; so for me that's fixed - but that leaves Ed's problem which had the same symptom as mine from the mapping.
Ed: Perhaps you've got a mismatched package somewhere as well?

Dave
Comment 5 Ed Tomlinson 2010-04-02 14:58:52 UTC
What specific package caused the problem?   And what files did it contain?  This install has been around a while its possible there is some crud that needs cleaning here too.
Comment 6 Ed Tomlinson 2010-04-03 05:27:53 UTC
Rebuild libX11 helped here.  Now only the 'UP' key is failing.  Interestingly xev shows no KeyPress event when up is pressed:

FocusOut event, serial 34, synthetic NO, window 0x1a00001,
    mode NotifyGrab, detail NotifyAncestor

FocusIn event, serial 34, synthetic NO, window 0x1a00001,
    mode NotifyUngrab, detail NotifyAncestor

KeymapNotify event, serial 34, synthetic NO, window 0x0,
    keys:  8   0   0   0   0   0   0   0   0   0   0   0   0   4294967168 0   0   
           0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   

KeyRelease event, serial 34, synthetic NO, window 0x1a00001,
    root 0x108, subw 0x0, time 72962426, (281,42), root:(285,65),
    state 0x0, keycode 111 (keysym 0xff52, Up), same_screen YES,
    XLookupString gives 0 bytes: 
    XFilterEvent returns: False

Ideas?
Comment 7 Ed Tomlinson 2010-04-04 06:36:40 UTC
Looks like libXi needed to be updated to the latest (git) version too.

All is working now here.
Comment 8 Peter Hutterer 2010-04-15 21:40:29 UTC
can we close this bug then?

there are some issues with keyboard handling that clients (namely metacity) don't handle well, so if you disable hotplugging but still use evdev, you're going to run into issues with different keymaps on the VCP and the physical device. We've covered the standard use-cases but some special ones cannot be handled easily.

Dave - I guess the initial lack of devices may have to do with the switch to udev, just guessing though.
Comment 9 Ed Tomlinson 2010-04-16 04:38:14 UTC
In both cases, the problem turned out to be a miss match between packages (we talked on irc).  Here libXi needed to be from git along with a few others.

As far as I am concerned this can be closed.
Comment 10 Peter Hutterer 2010-04-21 21:53:48 UTC
thanks, closing as invalid then, caused by some package mismatch.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.