Bug 119210 - Haldaemon dies and does not start correctly
Summary: Haldaemon dies and does not start correctly
Status: VERIFIED DUPLICATE of bug 121913
Alias: None
Product: SUSE LINUX 10.0
Classification: openSUSE
Component: Hotplug (show other bugs)
Version: Preview 1
Hardware: Other All
: P5 - None : Normal
Target Milestone: ---
Assignee: Danny Kukawka
QA Contact: E-mail List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-09-28 13:05 UTC by Jason Kasper
Modified: 2005-10-18 00:29 UTC (History)
0 users

See Also:
Found By: Other
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
output from "lshal > /tmp/lshal.out 2>&1" (59.68 KB, text/plain)
2005-10-14 10:11 UTC, Jason Kasper
Details
output from sudo su -c "hald --daemon=no --retain-privileges --verbose=yes > /tmp/hald.out 2>&1 &" (136.23 KB, text/plain)
2005-10-14 12:49 UTC, Jason Kasper
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Jason Kasper 2005-09-28 13:05:34 UTC
This is coming as a side-result of bug #106897.  It seems that at some point, 
hal is dying and if I notice that it's not running and do an /etc/init.d/
haldaemon stop and start, and then check to see if it's running with "ps waux | 
grep hal", I don't see any hal processes running.  This might be affecting bug 
#106897 in that powersave depends on hal.

I just forced powersave to stop/start and did a suspend/resume on my laptop and 
I notice that now (after suspend/resume), /etc/init.d/haldaemon start does seem 
to work:

> ps waux | grep hal
root     29540  7.4  0.3   4960  3676 ?        Ss   08:59   0:00 /usr/sbin/hald 
--daemon=yes --retain-privileges
root     29560  0.0  0.0   1792   612 ?        S    08:59   0:00 hald-addon-
storage
Comment 1 Danny Al-Gaaf 2005-10-06 13:43:19 UTC
Please update to the last version of SUSE Linux. Preview 1 is to old to 
reproduce this here, we had to many changes since this version.
Comment 2 Danny Al-Gaaf 2005-10-10 13:32:07 UTC
any news? Can I close the bug?
Comment 3 Jason Kasper 2005-10-10 16:37:00 UTC
Well, I've updated to 10.0-final, and I have hal-0.5.4-6 installed.  I don't 
know if you want to close this or not, since hald is not running currently, and 
this is after 12 hours of uptime.  

Is hald supposed to be running at all times?  If it is, then I think there's a 
bug somewhere that is causing it to crash/die/etc., and I'd like to help figure 
out why this is happening.  How can I debug this?

If it's not supposed to be running at all times, and it's acceptable that at 
some point it stops running, then I guess you can close this bug?

Comment 4 Danny Al-Gaaf 2005-10-10 17:00:29 UTC
Yes, HAL should always run. I have no idea, why it dies at your machine, we 
never had such a problem here. Could you restart hal with 'rchal restart' ? 

- Is there any message in /var/log/message related to hal (try: 'grep hal /var/
log/messages')?
- Is there any other hal process ('ps aux | grep hal')?
- Is dbus running?
- Did you updated all packages from YOU ?
Comment 5 Jason Kasper 2005-10-10 17:12:45 UTC
- There are no running hal processes, per "ps waux | grep hal"

> grep hal /var/log/messages
Oct  8 21:41:16 linux hal-subfs-mount[4521]: MOUNTPOINT:: /media/floppy
Oct  8 21:41:16 linux hal-subfs-mount[4521]: Collected mount options and 
Called(0) /bin/mount -t subfs -o fs=floppyfss,sync,procuid,nosuid,nodev,exec /
dev/fd0 "/media/floppy"
Oct  8 21:41:16 linux hal-subfs-mount[4807]: SYMLINKS:: disk/by-path/pci-0000:
00:1f.1-ide-1:0 dvd cdrom
Oct  8 21:41:16 linux hal-subfs-mount[4807]: MOUNT_POINT:: /media/dvd
Oct  8 21:41:16 linux hal-subfs-mount[4807]: MOUNTPOINT:: /media/dvd
Oct  8 21:41:16 linux hal-subfs-mount[4807]: Collected mount options and 
Called(0) /bin/mount -t subfs -o fs=cdfss,ro,procuid,nosuid,nodev,exec,
iocharset=utf8 /dev/hdc "/media/dvd"
Oct  8 21:42:26 linux hal-subfs-mount[5187]: By hald-subfs-mount created link /
media/SU1000_001 got removed.
Oct  8 23:11:56 gandalf hal-subfs-mount[13160]: Device: /dev/hdc already 
mounted, exit now!
Oct  9 00:11:17 gandalf hal-subfs-mount[16736]: Device: /dev/hdc already 
mounted, exit now!
Oct  9 00:11:58 gandalf hal-subfs-mount[16747]: Device: /dev/hdc already 
mounted, exit now!

> ps waux | grep dbus
100       3765  0.0  0.1   3520  1544 ?        Ss   Oct09   0:01 /usr/bin/dbus-
daemon --system

- after restarting hal with "sudo rchal restart", after I "ps waux | grep hal", 
it's still not running.  This is what I've seen before too--once hald dies, it 
doesn't seem to be restartable.

And yes, I'm fully updated from YOU.
Comment 6 Danny Al-Gaaf 2005-10-10 18:57:15 UTC
Is there maybe the pidfile of hal /var/run/hal (haldaemon.pid) after the 
haldaemon died? If so, remove the pid file before restart.
Comment 7 Jason Kasper 2005-10-10 19:03:45 UTC
Okay, after having done that....

`--> ps waux | grep hal
root     30564  9.7  0.3   5028  3720 ?        Ss   15:02   0:01 /usr/sbin/hald 
--daemon=yes --retain-privileges
root     30591  0.0  0.0   1796   612 ?        S    15:02   0:00 hald-addon-
storage

So, it starts up okay after that manual intervention.  Shouldn't an extra step 
be made in the startup script to see if the pid that is in that file is actually 
valid before not doing anything?
Comment 8 Danny Al-Gaaf 2005-10-10 19:19:21 UTC
The startup script do this normaly. If the haldaemon died what say 'rchal 
start'? Is there a information about the pid file?
Comment 9 Jason Kasper 2005-10-13 20:50:27 UTC
Hi Danny.  Here's what I see:

$ sudo rchal  start
Removing stale PID file /var/run/hal/haldaemon.pid.
Starting HAL daemon                                                   done

Yet after that, when I $(ps waux | grep hal), I see that there is no hal process 
running.
Comment 10 Danny Al-Gaaf 2005-10-14 09:56:27 UTC
This is really strange. Could you start hal manually in this case (hald --
daemon=yes --retain-privileges)? 

Could you attach the output of lshal if HAL is correct running? 
Comment 11 Jason Kasper 2005-10-14 10:10:29 UTC
Sure.  Here's what I see:

.-(~)---------------------------------------------------------(gideon@localhost)
-
`--> sudo hald --daemon=yes --retain-privileges
Password:
.-(~)---------------------------------------------------------(gideon@localhost)
-
`--> ps waux | grep hal
root     12149  9.1  0.3   5024  3720 ?        Ss   06:04   0:01 hald --
daemon=yes --retain-privileges
root     12174  0.0  0.0   1792   612 ?        S    06:04   0:00 hald-addon-
storage

So, it starts.  It did seem to take a good couple of seconds for hald to start.  
In /var/log/messages, I see this:

Oct 14 06:04:12 linux sudo:   gideon : TTY=pts/5 ; PWD=/home/gideon ; USER=root
 ; COMMAND=/usr/sbin/hald --daemon=yes --retain-privileges
Oct 14 06:04:15 linux kernel: floppy0: Unable to grab DMA2 for the floppy driver


I'm not sure if the floppy0 error has something to do with hal dying....

I'll attach the output of lshal next.  Thanks for your help Danny!!  =:)
Comment 12 Jason Kasper 2005-10-14 10:11:43 UTC
Created attachment 54059 [details]
output from "lshal > /tmp/lshal.out 2>&1"
Comment 13 Jason Kasper 2005-10-14 12:49:25 UTC
Created attachment 54088 [details]
output from sudo su -c "hald --daemon=no --retain-privileges --verbose=yes > /tmp/hald.out 2>&1 &"

Okay, I just rebooted and noticed that hal did not start again at boot time.  I ran the following command and am attaching the output.

sudo su -c "hald --daemon=no --retain-privileges --verbose=yes > /tmp/hald.out 2>&1 &"

I've noticed that the very last output from this has this:

08:34:12.343 [I] physdev.c:1443: phys_add: subsys=usb sysfs_path=/sys/devices/pci0000:00/0000:00:1d.0/usb1/1-1/1-1.2, parent=0x080a17b0
08:34:12.343 [E] util.c:409: Cannot open '/sys/devices/pci0000:00/0000:00:1d.0/usb1/1-1/1-1.2/serial'
08:34:12.343 [E] util.c:409: Cannot open '/sys/devices/pci0000:00/0000:00:1d.0/usb1/1-1/1-1.2/serial'
scandir: No such file or directory
08:34:12.358 [I] device_info.c:1370: *** Matched file /usr/share/hal/fdi/information/10freedesktop/10-wireless-mice.fdi
scandir: No such file or directory
scandir: No such file or directory
08:34:12.364 [I] physdev.c:1368: Add callouts completed udi=/org/freedesktop/Hal/devices/usb_device_46d_c50e_noserial
08:34:12.364 [I] hald.c:89: Added device to GDL; udi=/org/freedesktop/Hal/devices/usb_device_46d_c50e_noserial
08:34:12.368 [E] util.c:684: Couldn't spawn 'hald-addon-usb-csr' err=Failed to execute child process "hald-addon-usb-csr" (No such file or directory)!

HTH!!
Comment 14 Danny Al-Gaaf 2005-10-14 14:12:15 UTC
This is what I supposed as a possible reason for the crash. There is already a 
YOU update in the pipe for this problem. see bug #121913.

As workaround until the YOU update is available, remove /usr/share/hal/fdi/
information/10freedesktop/10-wireless-mice.fdi



*** This bug has been marked as a duplicate of 121913 ***
Comment 15 Ihno Krumreich 2005-10-18 00:29:32 UTC
CLosed.