[jaunty] Keyboard and Mouse Frozen on Login Screen

The update a few moments ago is promising. Mouse now moving on the login screen, unfortunately, keyboard and screen still freezing on the login screen.

I have exactly the same problem

Last update fixed things for me

just did an upgrade, still not working. rebooted a couple of times and also tried kdm and slim but problem still exists.
changing status back to confirmed

Keyboard and mouse broke for me too when I upgraded Jaunty and rebooted yesterday (I had been upgrading/rebooting daily). Both USB and PS/2 interfaces are affected for the kdm and gdm login screen. I dropped to the recovery console, started the SSH server and then resumed booting and was able to copy logs off using ssh while it was at the login screen. Xorg.0.log is attached which reports mouse and keyboard could not be found. The command hal-device reports that it couldn't connect to hald. The command
 sudo hald --daemon=no --use-syslog --verbose=yes
reports:
*** [DIE] mmap_cache.c:di_rules_init():79 : Couldn't mmap file '/var/cache/hald/fdi-cache', errno=22: Invalid argument

The command ls -la /var/cache/hald/fdi-cache reports:
-rw-r--r-- 1 root root 0 2009-02-27 21:43 /var/cache/hald/fdi-cache

The command apt-get --reinstall install hal-info hangs at the point where it tries to start the hal daemon.

2.6.28-7-generic kernel, MSI KN72 motherboard (nForce2), ATI 9200SE graphics card (default driver).

Well despite hanging, following reinstalling hal-info above I don't have the problem any more.

Well, despite NO hanging, following reinstall hal-info did not solve the problem for me, I get

Runner started - allowed paths are '/usr/lib/hal:/usr/lib/hal/scripts:/usr/bin'
Error binding udev_event socket: Address already in use

on uname -r 2.6.28-8-generic

I am running on Asus M2NBP-VM CSM . Neither ps/2 nor USB mice work - but work fine in e.g. PuppyLinux 4.1.2

Graphical progs work remotely via ssh -X, so it is definetely a hal/kernel/xorg problem.

Update: from time to time, I can move the mouse a little, but I get no cursor afterwards. In fact, if the supposed cursor is on something clickable, it might even react to it. Occurs with the login screen and within kde (gnome not yet installed)

I tried acpi=off as boot option, but that means noise and wasting energy - even now I think that also the cpufreqd is not running properly. Anyway, that did not help.

It feels like a mouse polling (timing) problem now on ps/2 and a "HID Device" issue on USB and a powernowd problem for the fan running all time and the system probably at max. freq....

Please feel free to demand any more logs - but I might just exchange the board and have a look what happens with a newer chipset than this 2year-old one nforce 5 or 6 or whatever...

By the way, not sure if it was lenny's or jaunty's mini.iso installer - something broke an unimportant lvm vgroup when trying to ext3 (or ext4 for jaunty) a seemingly unused reserve partition - sad. And the trigger to go jaunty was that something breaks writing aids in OO 3.0.1 when installing from the .ppa - ts...

My comments above were for the 2.6.28-7-generic kernel which works for me. With 2.6.28-8-generic my machine freezes either shortly before the login screen appears or up to 3 seconds after. Sometimes I can move the mouse on the login screen before it freezes, but no cursor afterwards and any existing ssh connection is lost. dbus and hal do start at the recovery console and hal-device reports the mouse. After a freeze and reboot, the Xorg.0.log is zero in size. Only /var/log/auth.log and /var/log/udev have been written to in /var/log. This has been the case for me for a few weeks now - I think since 2.6.28-8-24 arrived. 5 yr old MSI K7N2 nForce2 motherboard.

By "no cursor", I meant the text cursor stops flashing.

So Antex and me experience the same Problems on different hardware - which both is nforce and AMD am I right?
As Antex runs on an older kernel, I wil try to install the same 2.6.28-7 and see if it there's an improvement...

Cant't install via synaptic anything older than 2.6.28-8 or 2.6.36-3 for rt. Booting rt gives me bug #337356. Tried booting the server kernel also - same behaviour than linux-image-generic. I'm afraid that's it for today unless nobody demands specific logs...

Yes, my processor is AMD Athlon XP 2800+

If I remove powernowd from the boot sequence, the freeze with 2.6.28-8 doesn't happen. Running /etc/init.d/powernowd start gives instant lockup. linux_2.6.28-8.24.diff.gz does mention cpufreq-nforce2.o.

Strange, I had no powernowd installed - maybe that's why my machine seemed a little loud.

I installed sysfsutils and added

devices/system/cpu/cpu0/cpufreq/scaling_governor=ondemand
devices/system/cpu/cpu1/cpufreq/scaling_governor=ondemand

and also set this manually to avoid rebooting. I get reported that my athlon x2 4400+ is now running at 1 Ghz instead of 2.3. Do I trust this info? Will check wattage with/witout....

I also searched for cpufreq and found /lib/modules/2.6.28-8-generic/kernel/arch/x86/kernel/cpu/cpufreq:
p4-clockmod.ko speedstep-lib.ko

I tried to lsmod if any cpufreq-deamon was running - nope. I tried to load some - nope.

Antex, what lsmod / lsmod -l gives you? What governors are there? I think some amd-specific things are missing in this kernel build. Maybe that's our problem? In other words, if I sudo apt-get purge powernowd and sudo apt-get install powernowd no amd-powernowd modules appear or load? Unless generic cpufreqd is builtin into the kernel?

I am not sure if this is a new bug - same thing with 2008.04 and my other machine - a phenom - also had to get freq-scaling working manually - or related, because for me with or without scaling I have no mouse (can be moved sometime and then cursor disappears for good) - but machine still works remotely / headlessly so far o.k.

Available Mods

Maybe something related to this bug #332170

From the error message mentioning hald/fdi-cache and the report that reinstalling hal-info resolved the issue, it sounds a lot like this is a hal-info bug rather than X.

See also https://wiki.ubuntu.com/X/Troubleshooting/HalBreaksKeyboardAndMouse

Dax's bug title and first post sounded like my hal-info problem but his second post sounds like my powernowd problem *if* by "screen still freezing" he means the text cursor stops flashing in kdm. Dax?

My XP2800+ machine install was from a Jaunty Desktop Alpha 4 ISO. A default install of a Desktop Alpha 2 on an XP2400+ machine VMWare VM has powernowd but a Desktop Alpha 5 does not. So I guess only people with certain CPUs upgrading from pre-Alpha 5 install will get hit by the powernowd freeze. For example, my CPU does not support frequency scaling and I don't think Dax's Celeron does either.

Grizzly, did you install off Alpha 5? If you are saying that you can work on your Jaunty machine remotely then we have different symptoms because I can't ssh remotely into the box after the kdm login screen text cursor stops flashing. The mouse cursor doesn't disappear it just doesn't move. All I can do is a hard reset. Does the KDM login screen text cursor stop flashing for you?

lsmod output is attached. I am seeing CONFIG_CPU_FREQ_TABLE=y in the 2.6.28-8.24 kernel config diff so I think that means the modules are now built-in. They are available in 2.6.28-7 but not available in 2.6.28-8 apart from /lib/modules/2.6.28-8-generic/kernel/arch/x86/kernel/cpu/cpufreq: e_powersaver.ko.

Looking at the powernowd script, at one point it seems to decide whether the modules are built in by checking whether the architecture is ppc and exits sharply if it is. So I think that powernowd either hasn't caught up with the changes being made to the kernel or the plan is to remove it from the repo.

If I install powernowd on my Alpha 5 VM, there is no freeze. And seeing as uninstalling powernowd on my XP2800+ machine both brings it into line with Alpha 5 and fixes the problem, I don't think I should take this further.

Actually, I installed via this "mini.iso" (similar to the debian one) put onto an usb stick by unetbootin. So everything besides the initrd basically (even the installer itself) was fetched from the state it was in on March 1st. So hard to say if I had "Alpha4" or "Alpha5". Everyday I do a script I called "ubup.sh" which basically cleans the apt-cache, autoremoves everything possibly there, dist-upgrades and autocleans again, while using a squid proxy on my "kamikaze" router to speed up things...

Today came an update of the kernel again (8-13 or something), now I am obviously running on the server kernel.

To veryfy (or falsify) that it is a transitional problem, I will just install the system now again.

And no, I do not have a frozen keyboard and remote access works o.k. But my mouse pointer and text cursor also dissappears after moving the ps/2 mouse a little. USB mouse is still dead - Sometimes I even manage to selet the user field with a click bevor my cursor dies...

I think, there are two or three things messed up, I still get the complaints about nvidia nforce2 and the synaptics init failure even in the virtual machine (on my other ibex box, running the 32bit version there)

I will also simply do a clean reinstall and see what's up then - keys etc. saved in svn anyway...

Gratulations Antex if you work fine now - I don't

Reinstalled from scratch via mini iso from March 4th - keyboard works, mouse not neither on usb. I think this must be a HAL problem, because the USB pendrive I installed from seemingly is recognised:

Bus 002 Device 002: ID 058f:6387 Alcor Micro Corp. Transcend JetFlash Flash Drive
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

Now I have Gnome instead of KDE can use the ps/2 mouse and move it - it appears three or four times for a short while on screen and then disapears for good, clicking also sometimes works.

But - I found something I think I read about, have to check - dmesg | grep input

[ 13.041965] input: ETPS/2 Elantech Touchpad as /devices/platform/i8042/serio1/input/input5

This is WRONG - must be imps2 generic mouse or something - it's a simple logitech one.

I found the following related to 2.6.28 and ps/2 issues with Elantech

- bug #315882 - Elantech driver too old to function
- bug #123775 - older but interesting, because it is the outher way round, here the touchpad is rec. as wheelmouse
  in my case, the wheelmouse is recognised as a touchbad

Still no Idea why the USB mouse is not lighting up at all...

So I will have to look if I can tweak something in xorg.conf (which is empty for now - I hate that I can't even use ctrl-backspace any more and must use this unzap-flag argh...

Grizzly, have you eliminated the possibility that it is one of the startup scripts in the multi-threaded boot that is killing your PS/2 mouse? e.g. Have you tried by-passing some of them by doing /etc/init.d/dbus start, /etc/init.d/hal start, startx from the recovery console? (You said earlier you got the problem within kde but I wasn't sure if you got there quite normally via auto-logon.)

Hi,

don't know if this will help, but as I'm affected as well. After boot (linux-image 2.6.28.8.8) keyboard is working, mouse not (logitech usb connected via converter on PS/2 port, so PS/2).

But, forcing psmouse proto to imps get my mouse back:

sudo rmmod psmouse
sudo modprobe psmouse proto=imps

@Xavier this sounds interesting with reloading the modules - would it help to explicitely configure imps2 in xorg.conf as well.
I am on the road this weekend, but will check back tomorrow.

@Anthex also sounds interesting - will try the recovery console. The problem does not have anything to do with kde-wm or gnome-wm - did a complete reinstall and took ubuntu-desktop instead of kubuntu-desktop to see if it made any difference. As there is none, I really think it is startup scripts, xorg, hal-infor or the kernel... I'll keep you updated.

I made a fresh install from alternate Ubuntu Studio Jaunty X86_64 alpha 5 a week ago. Sometimes I have no keyboard and mouse at GDM. I am not sure, but I think this bug generates logs in /var/log/gdm. I have 5 logs there and I think they match situation and time the bug appeared. Rebooting usually solves the problem, only once I had to reboot twice..

@Antex: recovery console and
> sudo /etc/init.d/dbus start
> sudo /etc/init.d/hal start
> startx
does not improve anything

@Xavier
This helped as a workaround - still my USB Mouse does not light up, but the one via PS/2 adapter starts to work! Yippee...

@Timo Michel
I have the same in GDM's :0.log
> Warning: Type "ONE_LEVEL" has 1 levels, but <RALT> has 2 symbols
> Ignoring extra symbols
(but only once, and I don't have
ddxSigGiveUp: Closing log

That is, working with dist-upgraded from yesterday.

I will now dist-upgrade to today, set up my card driver and if no one protests, add Xaviers workaround to the bug description (for ps/2)

Looks like different bugs for me: probs with USB-low-speed device AND ps/2 mouse... - for me ps/2 keyboard alwas worked, will try usb now.

Any idea (dmesg) where to look for psmouse module errors? Maybe a race condition? So not hal, but kernel problem?

I dist-upgraded to 2.6.28.9 and now I can only login with failsafe gnome session.
Normal login gives

> /etc/X11/Xsession.d/60x11-localhost: 4: Syntax error: Bad fd number

from this bug #340697... and moved the ampersand to the end of the line...

I found this in my dmesg as related to imps/psmouse reload

[ 33.610209] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 34.294951] psmouse.c: Touchpad at isa0060/serio1/input0 lost synchronization, throwing 2 bytes away.
[ 34.803045] psmouse.c: resync failed, issuing reconnect request
[ 35.489893] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 35.622326] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 35.638041] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 35.678491] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 35.689965] psmouse.c: Touchpad at isa0060/serio1/input0 lost sync at byte 4
[ 35.689968] psmouse.c: issuing reconnect request
[ 45.708144] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[ 45.708150] Bluetooth: BNEP filters: protocol multicast

so - still the system thinks it sees a touchpad - and again it wants to do something with bluetooth which the board does not have :-(

I put the rmmod lsmod sequence in /etc/rc.local - now the ps/2 mouse works even at the login screen

Any comments whether it is a hal bug or a kernel or a startup script bug? (VirtualBox-Jaunty runs fine single-cored - real system doesnt (dual cora = race?)

Hi there,

Concerning the ps/2 mouse which get recognized as Elantech touchpad, it's apparently a know kernel issue, and only affect some logitech mouse.

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=9ab7b25e6a30d2292bd6d4913b71c918ee1e21b4

There're already a bug on LP for this specific issue #318722, so should be better to comment on that bug for logitech / elantech missdetection.

Meanwhile, my USB still mouse didn't work.

Dist-upgraded as every day - nothing changed so far / ps2 mouse which is a logitech - other I can't say, because the sticker is scrubed away after long years of ubuntu ;-) ist still a touchpad - but otherwise works.
USB mouse doees not light up at all - this one is a RX250 ... so I will have a look at bug #318722

@Xavier - on writing issue #ID, lp does not link, only when one writes bug #ID it makes a hyperlink. I'll put that to "help improve lp"

Hi guys, same for me

I'm running Jaunty on live CD
USB Logitech mouse
connexion to KVM switch, which is PS2, thru PS2/USB adaptor

sudo rmmod psmouse
sudo modprobe psmouse proto=imps
has been the solution to make it work, no frozen mouse anymore, hope the developers will correct the bug

Something screwed up my panel (no panel in gnome at all) so I reinstalled, running into new bugs such as grub2 has no grub.cfg where to edit boot options. Howerver, now ps/2 works out of the box, usb mouse via usb still does not.

But in xorg.log I have ps/2 logitech now

[ 5.963646] (II) ImExPS/2 Logitech Wheel Mouse: Found 5 mouse buttons
[ 5.963659] (II) ImExPS/2 Logitech Wheel Mouse: Found x and y relative axes
[ 5.963666] (II) ImExPS/2 Logitech Wheel Mouse: Configuring as mouse
[ 5.963679] (**) ImExPS/2 Logitech Wheel Mouse: YAxisMapping: buttons 4 and 5
[ 5.963687] (**) ImExPS/2 Logitech Wheel Mouse: EmulateWheelButton: 4, EmulateWheelInertia: 10, emulateWheelTimeout: 200
[ 5.963709] (II) XINPUT: Adding extended input device "ImExPS/2 Logitech Wheel Mouse" (type: MOUSE)
[ 5.963734] (**) ImExPS/2 Logitech Wheel Mouse: (accel) keeping acceleration scheme 1
[ 5.963742] (**) ImExPS/2 Logitech Wheel Mouse: (accel) filter chain progression: 2.00
[ 5.963754] (**) ImExPS/2 Logitech Wheel Mouse: (accel) filter stage 0: 20.00 ms
[ 5.963767] (**) ImExPS/2 Logitech Wheel Mouse: (accel) set acceleration profile 0

So - that's a start - elantech touchpad users seems to have gotten further as well..

I also have this problem. The keyboard and mouse work properly until X starts, then this is found in the log:

[ 424.382322] hald[3985]: segfault at 0 ip 0000000000435b05 sp 00007fffc3760990 error 4 in hald[400000+57000]

Also i get the message: Internal Error, Failed to initialize HAL! at the beginning of the startup, which would make sense, if hald has segfaulted as X is starting.

I have the same problem (keyboard/mouse die after X starts) after dist upgrading (-8 to -11 kernel, x64). Booting into single user mode and starting dbus and hald manually I get the same hald segfault as above.

Hello,

I put a comment following bug 342456 - https://bugs.launchpad.net/bugs/342156

Initial message :

"- On my desktop PC, with Jaunty 64 Alpha 6, no problem until login screen, but when I am on Ubuntu session, mouse does not work correctly.

I can moove it and cursor follows the moves on the screen, but almost all clicks do not work.

For examle, on the panel with some icons to launch programs like Firefox, Pidgin... I can click to one, it launches the software, but after that I can only re-click to the same icon, no other icons or menu works.

Moreover, even if program starts, click on the menu bars or buttons has no effect, (no bookmark on Firefox, no file selection on Nautilus...), but keyboard works fine (I can launch terminal with Ctrl - Alt - F1).

What is surprising is that the first occur of this bug was not at startup, but during a session - I was installing another OS in Virtualbox and mouse began to not react anymore.

- On my laptop with Jaunty 32 Alpha 5, I was doing the update to Alpha 6 and mouse freezing occured during installation of updates, when I restart, it does not go further than login screen - I think for this one I will have to reinstall.

I have for the 2 computers USB mouse - Wireless optical Logitech and normal Dell ones."

I noticed that last time it happened when the piles (battery) of my wireless mouse were about to be changed.

I put in joined file results from "dmesg" command as asked for normal use, USB mouse device unplugged, and Piles (battery) of wireless mouse off (so mouse inactive) but USB device on.

I hope I was clear.

Best regards.

Reinstalling from the alpha 5 cd and dist-upgrading directly to -11 fixed the issue for me. And I no longer need to boot with i8042.nopnp for the keyboard/mouse to work. Weird, perhaps a problem applying one of the intermediate updates.

I think I have the same issue, but it's only when I'm boot into the -rt kernel (that's 2.6.28-3-rt) When I boot into the generic (2.6.28-11-generic) kernel mouse and keyboard functionality is fine. Go in with the real-time kernel and sometimes it'll freeze straight after logging in, sometimes I might get a good 10 - 15 minutes activity before it freezes up. I know the system is still running because the system time still moves and it responds to me pressing my power button.

Now I'm quite a beginner at this stuff and I've never really had any input into bug-fixing so if anyone could point me in the right direction of any logs or data that might help find a fix you'd have to walk me though it. But I'm willing to get any info required.

Gareth

I have a Dell vostro 1310. I get keyboard freezes intermittently, requiring me to reboot via the power key (I push the power key and the machine reboots, which is also odd). I'm using the generic kernel. So far, the second boot has always worked.

I've just encountered the same problem with the release version of Jaunty.

It's been working fine since Thursday (yes, I had to have it right away :)

As of an hour ago, every time I try to log into Jaunty, the keyboard and mouse freeze up at the (X windows) login prompt. That is -- both the USB keyboard and mouse and the main keyboard and mouse on my notebook (MSI clone with AMD Sempron, ATI Radeon).

The Red light is lit on the optical mouse, but it doesn't work.

I've tried starting in recovery mode, and checking log files, but nothing seems out of the ordinary.

Has anyone else had this problem with post-release Jaunty?

Please ignore my previous post -- I had tried to quiet dbus daemon based on another bug report:

  <allow send_interface="org.freedesktop.DBus.Properties" send_type="method_call" send_member="Get"/>

which hosed it instead.

You can try disabling hotplug from Xorg.

Add the following to your xorg.conf
Section "ServerFlags"
# Option "DontZap" "False" ## Uncomment this only is you wish to enable Ctrl+Alt+Backspace to kill X
     Option "AutoAddDevices" "False"
EndSection

This option lets X and hal use the information about your input devices from xorg.conf.

PS. This issue has nothing to do with the nvidia kernel module (the one from nvidia site).

please let me know if this solved your problem. Greetings.

On the machine concerned I went back to debian (testing), fast booting is negated by having to reboot. I may try ubuntu 9.04 again if there is an X upgrade and a gnome upgrade to restore email indexing. Editorial comment/venting: This X bug confirms my suspicion that Ubuntu is again seduced by shiny and new software that doesn't always work very well.
------Original Message------
From: CarloSalvatore
Sender: <email address hidden>
To: Tim Richardson
ReplyTo: Bug 335297
Subject: [Bug 335297] Re: [jaunty] Keyboard and Mouse Frozen on Login Screen
Sent: Apr 28, 2009 17:49

please let me know if this solved your problem. Greetings.

--
[jaunty] Keyboard and Mouse Frozen on Login Screen
https://bugs.launchpad.net/bugs/335297
You received this bug notification because you are a direct subscriber
of the bug.

Sent via BlackBerry® from Vodafone

I'm getting the same problem as Tim. I'm on a Dell Vostro 1710. The problem seems completely erratic right now because sometimes when I boot up I will have neither the keyboard nor the mouse, and sometimes I will have both. They become non-responsive instantly at the login window. I have tried to boot into the recovery console, but the keyboard seems to randomly freeze there as well by the time you choose root terminal, root terminal with networking, etc.

I have nvidia graphics in my vostro. I still have ubuntu 9.04 on another dell laptop, with intel graphics; no problems.
------Original Message------
From: Daniel
Sender: <email address hidden>
To: Tim Richardson
ReplyTo: Bug 335297
Subject: [Bug 335297] Re: [jaunty] Keyboard and Mouse Frozen on Login Screen
Sent: Apr 28, 2009 22:51

I'm getting the same problem as Tim. I'm on a Dell Vostro 1710. The
problem seems completely erratic right now because sometimes when I boot
up I will have neither the keyboard nor the mouse, and sometimes I will
have both. They become non-responsive instantly at the login window. I
have tried to boot into the recovery console, but the keyboard seems to
randomly freeze there as well by the time you choose root terminal, root
terminal with networking, etc.

--
[jaunty] Keyboard and Mouse Frozen on Login Screen
https://bugs.launchpad.net/bugs/335297
You received this bug notification because you are a direct subscriber
of the bug.

Sent via BlackBerry® from Vodafone

I too am experiencing this problem, on a Dell Vostro 1510.

Just as an update: if I have the laptop plugged in when I turn it on, the first time it comes to the login window the keyboard and mouse will be unresponsive. I hit the power button to shut it down normally, and reboot, and it then works just fine every time. On the other hand, if I don't plug it in, I can reboot it as many times as I want and it will always have a locked keyboard and mouse.

I often experience this problem when I boot into kernel 2.6.28-11-generic, but if I select 2.6.27-11-generic instead, it works correctly every time. Can anyone else confirm this?

I have the exact same error after upgrading from 8.10 to 9.04 on my Dell Vostro 1310 laptop.
Every time I boot the machine, when the power cord has been disconnected, I am unable to login due to a non-responsive keyboard at the login screen. I can always succeed restarting the machine using an USB mouse, and the login screen always works after these warm reboots.
I have tried once booting up on kernel 2.6.27-11-generic and it worked. I will continue testing on this then, as it is quite annoying with all these reboots.
As of now I do regret upgrading to 9.04, the older version worked better for me - but I still love Ubuntu.

Hello, I had the same problem with my Dell XPS 1330 under Jaunty.
This problem did not happen immediately after the upgrade from Intrepid, but a few days later, with apparently no reasons.
Both mouse and keyboard froze at login screen, and I was unable to make them work by rebooting.
However, I resolved this problem. I am not really sure of the steps you should follow because I made many many tries. I think that the following commands made the miracle:

sudo update-rc.d -f dbus remove
sudo update-rc.d dbus defaults
sudo update-rc.d -f hal remove
sudo update-rc.d hal defaults

they reset the dbus and hal starting service defaults in rc.d

If these commands do not work, keep in mind that before executing them I:

- reinstalled both dbus and hal
- reconfigured the xorg.conf to defaults with dpkg-reconfigure xserver-xorg

In addition, I was able to "unlock" mouse and keyboard at login screen by adding the following lines to xorg.conf

Section "ServerFlags"
       Option "AllowEmptyInput" "false"
EndSection

however, when logging in, a message told me that HAL could not be started. I started it via console by typing

sudo /etc/init.d/hal start

and HAL started correctly, but mouse and keyboard behaviours were very strange, e.g. by pressing a letter, I obtained the letter repeated twice. After resetting dbus and hal runlevels I commented the three lines added in xorg.conf without any error.

I hope that my post will be useful ;)

I've also been running update Jaunty for a week or so and it's been fine, but tonight I ran updates while fiddling with amarok and my system froze. DRT. I had no choice but to power cycle and upon rebooting it froze again right after entering my password and hitting enter. I rebooted again to the same behavior. I have keyboard and mouse when GDM shows up so I can switch to a terminal window, but I don't know what to do since I have no internet at this point (wireless). The previous post isn't inspiring me because he tried many different things. If I switch back to the GDM window and enter my password and hit enter it insta-freezes again. I'm writing this from Hardy which thankfully I still have installed for situations just like this. I eagerly await a better fix for this completely debilitating bug.

chaz, you can try to execute only the following commands, without trying everything I did:

sudo update-rc.d -f dbus remove
sudo update-rc.d dbus defaults
sudo update-rc.d -f hal remove
sudo update-rc.d hal defaults

I found a posible solution for this bug at:

    https://bugs.launchpad.net/ubuntu/+bug/341094

I have a DELL VOSTRO 1510, and add "i8042.reset i8042.nomux i8042.nopnp i8042.noloop" to boot parameters in GRUB.

this work fine for my up to now. (two days).

i have a dell vostro 1710 and have been experiencing the same lack of stabilty since upgrading .

i've been with ubuntu for several years and this is the worst release yet :( :( :(

There is clearly something going wrong with :
ACPI - power management is not making sense anymore and my battery is discharging faster than it used to
MOUSE and KEYBOARD - mostly fixed with the gotos grub params , still not as stable as it used to be
AUDIO - pulseaudio , it still sucks and is still not properly configured
DBUS - clearly getting random signals from ACPI and breaking
GDM - i can login using the latest ubuntu kernel

This is a serious regression.

This worked for me

sudo update-rc.d -f dbus remove
sudo update-rc.d dbus defaults
sudo update-rc.d -f hal remove
sudo update-rc.d hal defaults

I have a P4 homemade... can't remember what's in. if you want info please tell me what logs
I did an dist upgrade from hardy (through online update manager) to

Ubuntu 9.04, kernel 2.6.28-11-generic

Mouse and keyboard were frozen. I went to recovery mode and managed to get in X with
/etc/init.d/dbus start
/etc/init.d/hal start
startx

Then I was able to google to this bugs launchpad and found the solution! (I tried adding noht to /boot/grub/menu.lst but made no changes)
So thanks a lot to you all and especially botticchio, and now I'm registered !

I'm still surprised at how long this issue has been going on and that no corrected updates were available through update manager. Anyway thanks again and long live.....

For me the situation is the same as for Jesper Steinicke Würtz (Dell Vostro 1310 with problems after upgrade.

As I thought this was a Bug in X.org I (erroneously) filed a bug report: Bug #378866

I've included the output of lspci for further information.

Besides not being able to login I don't find any error messages in the logs (no HAL crash etc.).

And, as others have already noted: this is _VERY_ annoying. I also regret having updated to 9.04.

Oliver

Debian unstable does not have this problem.I've been using sid for a few weeks now on my Vostro. It's the same X version as Jaunty (I suppose a bit more recent but basically the same version). However, the kernel is 2.6.29, newer than Ubuntu 9.04. This bug caused me to go back to Debian, and for sure Debian is fine on this laptop, both sid and testing, which is still on kernel 2.6.26 and the older X.

Maybe it's kernel related?

I think my problem is related to this post. I bought an PS/2 to usb adapter. The keyboard and mouse work only sometimes. Is the consensus that I should use?
sudo update-rc.d -f dbus remove
sudo update-rc.d dbus defaults
sudo update-rc.d -f hal remove
sudo update-rc.d hal defaults
Thanks!

Still having the problem with my PS2 keyboard and mouse (they are connected to a PS2 to USB adapter). Any further suggestions?
Thanks!

I've just installed clean 9.04. Everything were okay until I did an upgrade. Reboot caused my mouse and keyboard unusable. In recovery mode keyboard is working fine. I noticed also that during installation of server 9.04 and (after that) xorg I received dpkg error on hald. It was on the same machine, so probably there is a problem with HAL here. The interesting thing is that my keyboard and mouse is also unusable under X after clean installation of archlinux.

I spotted this:

kernel: hald[3654]: segfault at 0 ip 0000000000433725 sp 00007fffebd4f2b0 error 4 in hald...

In my case this is probably because of another bug: https://bugs.launchpad.net/hal/+bug/361689

This is happening when you install mdadm and try to use software RAID arrays. I have two of them: one mirror and one stripe. Anyway, the patch for the above bug is still not published for 9.04. So right now I am disconnected from my data on my RAID arrays, which is bad. :/

After testing it for a while, I can now confirm that the problem does not happen for me with kernel 2.6.27-11-generic.

So this seems to be a bug in 2.7.28

Oliver

Grmpf. Of course I mean 2.6.28.

Hello, I had the same problem after reinstalling jaunty because of problems between KDE and Gnome in the same machine. When I tried "rmmod psmouse" I got an error so I checked modules. The problem was I have kernel 2.6.28-13 but I had the directory /lib/modules/2.6.28-11-generic.

I made

cp -R /lib/modules/2.6.28-11-generic /lib/modules/2.6.28-13-generic

and I solved the problem.

I hope it will be useful for you.

Hello,

  I can't confirm what SnowTDM is writing. For all kernel versions I
have a separate modules directory (2.6.27-11, 2.6.28-11, 2.6.28-13). The
only working kernel version is 2.6.27-11. Of course I haven't tried to
run modinfo with the non-working kernels to see if there's some kind of
problem.

  BTW: I doubt that it is a good idea to simply copy kernel modules to
new versions.

  Oliver

I had the same problem. When X starts, no keyboard and no mouse but the pointer (in the login box) works and the sound.

My problem was the /etc/init.d/hal script. It was empty!!!! And then hal did not start.

I reinstall hal (sudo aptitude reinstall hal) in the recover mode and all works again.

Just an update:

I have now booted a few times successfully into kernel 2.6.28-13-generic #45-Ubuntu, but a few minutes ago I had the same problem (lockup at boot at the gdm prompt). So the problem is still there (I'm constantly updating my system). Right now working again with 2.6.27-11.

ojo@hermes:~$ grep "hal" /var/log/Xorg.0.log <-- current working configuration
(II) config/hal: Adding input device AT Translated Set 2 keyboard
(II) config/hal: Adding input device Video Bus
(II) config/hal: Adding input device PS/2 Mouse
(II) config/hal: Adding input device AlpsPS/2 ALPS GlidePoint
(II) config/hal: Adding input device Macintosh mouse button emulation
ojo@hermes:~$ grep "hal" /var/log/Xorg.0.log.old <-- last boot
(II) config/hal: Adding input device Video Bus
(II) config/hal: Adding input device Macintosh mouse button emulation

I am also getting the non-responsive keyboard/mouse. Dual Opteron, 64-bit Ubuntu desktop 9.04. Worked fine until the latest update and reboot. Xorg.0.log seems to indicate hald hadn't started. I ssh'd into the machine to see that indeed hald hadn't started. I have tried reinstalling hal and hal-info to no effect. The /etc/init.d/hal script looks fine and the symlink (S24hal) is in rc2.d. I can manually start it with 'sudo /etc/rc2.d/S24hal start' . Once manually started, all seems to work fine. I am at a loss as to why it doesn't start automatically.

running 'sudo hald --daemon=no --use-syslog --verbose=yes' gives a few error messages about an acl-list problem (error opening/creating /var/run/hald/acl-list).

Not sure if that is important.

Update for 2.6.28-14: not working.
See attached Xorg.log

I had intrepid ubuntu server 8.10. Upgraded to jaunty 9.04 and rebooted.

Mouse and keyboard was non-responsive at login screen of gnome. My computer was a dead duck.

I took the installation CD and mounted the HD to a shell only session so I could use the keyboard.
There was no hal process (ps -edalf|grep hal).

The commands that I think helped me was:

apt-get -f install # took quite some time
apt-get --reinstall install hal hal-info

I then rebooted, and It worked. I did a dist-upgrade and rebooted again. Only problem was that Firefox complained about google not reachable. I restarted Firefox and now my Ubuntu jaunty seems to work.

The full sequence of commands I typed in the recovery shell was:

apt-get --reinstall install dbus # FAILED!
apt-get --reinstall install hal hal-info #FAILED!
apt-get -f install # took quite some time
apt-get --reinstall install hal hal-info

The commands commented with "FAILED!" requested me to run "apt-get -f install", which seemed to do the trick.

I'm using an Intel Pentium (2 cores, 2500 I think) and ps mouse and keyboard (green + lilac connectors).

Update for 2.6.28-15: NOT WORKING.

http://www.freedesktop.org/wiki/Software/hal/