keyboards events generated wrongly create issues

xev displays those

KeyPress event, serial 31, synthetic NO, window 0x3c00001,
    root 0x69, subw 0x0, time 4199027, (781,-45), root:(788,6),
    state 0x100, keycode 101 (keysym 0x0, NoSymbol), same_screen YES,
    XLookupString gives 0 bytes:
    XmbLookupString gives 0 bytes:
    XFilterEvent returns: False

KeyRelease event, serial 31, synthetic NO, window 0x3c00001,
    root 0x69, subw 0x0, time 4199064, (782,-45), root:(789,6),
    state 0x100, keycode 101 (keysym 0x0, NoSymbol), same_screen YES,
    XLookupString gives 0 bytes:
    XFilterEvent returns: False

One of the commenters on Debian bug #374026 reported a nearly identical problem, that started for him after this X upgrade:

2008-01-04 13:24:54 upgrade xserver-xorg-input-kbd 1:1.2.0-1+1.2.1 1:1.2.2-3

However, it seems that reporter was 'me too-ing' his issue onto another keyboard bug that doesn't seem to actually be the same issue. However, some of the troubleshooting steps suggested in that thread (and other Debian bugs) would be worthwhile to do here too:

1. Check if the timing of the -kbd update for Ubuntu matches about when you remember starting to see the issue. If it does, then try downgrading your -kbd to the prior release and see if the issue goes away. If so, I'll review the changes and see if we can identify a patch from that.

2. Are you running KVM or another VM? If so, see if there is a relation between working in that, and then losing the modifier keys. If this is the case, perhaps Soren would have more advice for further troubleshooting. Also try `setxkbmap -rules xorg -layout "fr"` to see if that forces things to kick back on.

3. In searching for 'keyboard modifier' in Debian BTS, I note a number of issues particular to the "fr" keyboard map, although none are a match to your issue. However, this could be tested by changing to a different keyboard map for a day or two and see whether the modifiers get lost there too.

4. Another user reported a year ago an issue when changing the system clock. Try setting the time to 10 minutes prior, then hold a modifier key down while running ntpdate to re-sync the time. (It appears the issue was solved and a fix pushed in xserver 1.2, but given the number of changes in the input code for xserver 1.4, a regression is certainly a possibility.)

Let me know how this testing goes, and we'll determine the correct direction from there.

Fujitsu and I are getting this as well, on an EN_US keyboard (i believe).

This has been happening for months, so it could well have been from that update.

I've used the "setxkbmap -rules xorg -layout us" before, and it didn't fix the problem. I'm not running KVM or another VM.

Upstream blames the kernel :)

I haven't seen anything like this with the evdev driver (input-hotplug ftw).

This has been striking me particularly badly over the past 48 hours. In most instances, I just get an endless series of tabs thrown at the focussed window, which makes debugging a bit difficult. I note that I have been holding down tab most times that this starts. There is no noticeable difference in frequency of occurrence between the latest 2.6.24 and 2.6.22 kernels.

I managed to run xev (by typing into Gajim, which fortunately ignores tabs, and pasting into a terminal), and it showed something much like seb128's, except that it was tabs.

Intel-based Dell Inspiron 630m here, if that's useful.

Hi All,

Just curious if you can attach your dmesg output after you start noticing this issue on a 2.6.24 kernel? I'm just curious if anything interesting gets logged that's kernel related. Also, Timo you mentioned that upstream blames the kernel. Do you have a reference you can point us to regarding this claim?

Leann: only a reply I got on #xorg-devel from Daniel Stone, so no. I wonder if showkey would show the same on console.

Moving to the xserver for now. Bug 190615 seems similar.

after typing 'dmesg', i get

[ 2313.780663] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.786696] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.790326] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.790874] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.795818] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.798396] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.801211] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.952068] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2313.956056] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.201861] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.270893] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.272512] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.273047] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.443729] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.446654] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.447789] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.449296] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.452262] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)
[ 2315.452526] eth1: TKIP decrypt failed for RX frame from 00:0f:b5:e6:6d:dc (res=-3)

endlessly repeated, with the numbers incrementing.

Is there any way of forcing the kernel to ignore keycode 101 events, or something of that nature? This is getting really annoying.

Hey, this is probably a dupe of bug 194214 (or the other way around), could you check the latest xorg-server upload if you still have this bug?