kernel panic making system unusable: "do_IRQ: [number] No irq handler for vector"

ninjagecko · #1 3rd February 2007, 06:06 PM

Hello.

I'm running into kernel panics involving unhandleable IRQs. This means I have to constantly hard-reboot the computer and makes it unusable. Could anyone please give insight into what the problem might be?

Symptoms:

I'm new to Fedora Core and just installed FC6 on a new quad-core computer, but I've been using it for 12-24 hours before this started happened; it has thus far only happened while I've been listening to music.

Kernel Version:
$ uname -a
2.6.19-1.2895.fc6 #1 SMP

Do people think this is a kernel problem, distro problem, ALSA problem, hardware problem, etc...?

Should I downgrade my kernel? If so, how do I do this using yum?

Thanks,
Mats

related topics:

http://www.fedoraforum.org/forum/sho...ghlight=do_IRQ (same processor & kernel!)

http://www.fedoraforum.org/forum/sho...ghlight=do_IRQ

---------------------------------
--- first time

[I was in the process of playing an mp3 on Banshee, while having firefox open and doing folding@home; the computer hard-freezes, and I don't think I can move my mouse even, though I can't remember that well]

Feb 2 21:28:25 cloudfall dhclient: bound to 18.243.6.119 -- renewal in 1514 seconds.
Feb 2 21:37:09 cloudfall kernel: do_IRQ: 3.79 No irq handler for vector
Feb 2 21:37:18 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 00000002
Feb 2 21:37:18 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 0016201a
Feb 2 21:37:26 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 0000001e
Feb 2 21:37:26 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 0016201b
Feb 2 21:37:29 cloudfall shutdown[23109]: shutting down for system halt
Feb 2 21:43:59 cloudfall syslogd 1.4.1: restart.
Feb 2 21:43:59 cloudfall kernel: klogd 1.4.1, log source = /proc/kmsg started.

[when I reboot, I get dumped into shell #1 and have to ctrl-alt-f7 back to X... nothing else is amiss]

--- second time

[was in the process of playing an mp3 on Amarok while copying files to disk with firefox open; very rarely (only twice so far) an odd bug crops up where my computer will randomly stop being able to play media files nomatter which app/decoder I use. a little while later after the bug kicks in, I get this dumped to my console while my screen freezes:]

Feb 3 11:34:28 cloudfall kernel: do_IRQ: 3.196 No irq handler for vector
Feb 3 11:34:38 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 000959b6
Feb 3 11:34:39 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 00000000
Feb 3 11:34:46 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 000959b7
Feb 3 11:34:47 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 0000001e

[this time was able to move the cursor, but do nothing else, nor go to a terminal; below I hit the power button, but have to hold it down or else the system won't reboot]

Feb 3 11:34:48 cloudfall shutdown[23051]: shutting down for system halt

[I get dumped to text mode right when I would have gotten to a login prompt, again]

--- third time:
[I was using Amarok started playing a song. The song started stuttering and repeating just like the first time this ever happened (on Banshee using GStreamer). I was actually able to very slowly kill a window and invoke alt-f2 before my computer stopped being responsive.]

Feb 3 11:37:58 cloudfall gconfd (mats-3328): starting (version 2.14.0), pid 3328 user 'mats'
Feb 3 11:37:58 cloudfall gconfd (mats-3328): Resolved address "xml:readonly:/etc/gconf/gconf.xml.mandatory" to a read-only configuration source at position 0
Feb 3 11:37:58 cloudfall gconfd (mats-3328): Resolved address "xml:readwrite:/home/mats/.gconf" to a writable configuration source at position 1
Feb 3 11:37:58 cloudfall gconfd (mats-3328): Resolved address "xml:readonly:/etc/gconf/gconf.xml.defaults" to a read-only configuration source at position 2
Feb 3 11:38:01 cloudfall scim-bridge: Another agent is running...
Feb 3 11:38:01 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 11:38:02 cloudfall scim-bridge: Another agent is running...
Feb 3 11:38:02 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 11:38:03 cloudfall scim-bridge: Another agent is running...
...
...
... [repeats for a long time]
...
...
Feb 3 12:13:30 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 12:15:17 cloudfall scim-bridge: Another agent is running...
Feb 3 12:15:17 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 12:15:28 cloudfall kernel: do_IRQ: 1.122 No irq handler for vector
Feb 3 12:15:40 cloudfall scim-bridge: Another agent is running...
Feb 3 12:15:40 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 12:19:22 cloudfall syslogd 1.4.1: restart.

[I *don't* get dumped to text mode right before my boot prompt.]

--- fourth time:
[The only app I had open was Firefox. I open up Banshee... and my console says "kernel...do_IRQ:..." -- and everything freezes. I have to hard-reboot yet again. This happened the exact second I hit "enter" to invoke Banshee.]

...
... [repeats]
...
Feb 3 12:22:16 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 12:22:16 cloudfall scim-bridge: Another agent is running...
Feb 3 12:22:16 cloudfall scim-bridge: Failed to allocate the agent. Exitting...
Feb 3 12:22:17 cloudfall kernel: do_IRQ: 1.130 No irq handler for vector
Feb 3 12:22:25 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 00005861
Feb 3 12:22:26 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 00000002
Feb 3 12:22:33 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 00005862
Feb 3 12:22:34 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 0000001e
Feb 3 12:22:41 cloudfall kernel: NVRM: Xid (0001:00): 16, Head 00000000 Count 00005863
Feb 3 12:22:41 cloudfall shutdown[4041]: shutting down for system halt
Feb 3 12:22:42 cloudfall kernel: NVRM: Xid (0001:00): 8, Channel 00000020
Feb 3 12:22:42 cloudfall smartd[3030]: smartd received signal 15: Terminated
Feb 3 12:22:42 cloudfall smartd[3030]: smartd is exiting (exit status 0)
Feb 3 12:22:43 cloudfall avahi-daemon[2906]: Got SIGTERM, quitting.
Feb 3 12:22:43 cloudfall avahi-daemon[2906]: Leaving mDNS multicast group on interface eth0.IPv6 with address fe80::21a:92ff:fe0a:60fd.
Feb 3 12:22:43 cloudfall avahi-daemon[2906]: Leaving mDNS multicast group on interface eth0.IPv4 with address 18.243.2.30.
Feb 3 12:22:44 cloudfall hcid[2580]: Got disconnected from the system message bus
Feb 3 12:22:44 cloudfall rpc.statd[2519]: Caught signal 15, un-registering and exiting.
Feb 3 12:22:45 cloudfall pcscd: pcscdaemon.c:529:signal_trap() Preparing for suicide
Feb 3 12:22:45 cloudfall pcscd: hotplug_libusb.c:361:HPEstablishUSBNotifications() Hotplug stopped
Feb 3 12:23:46 cloudfall syslogd 1.4.1: restart.

[I get dumped to text mode login before a boot prompt again.]

ninjagecko · #2 3rd February 2007, 07:32 PM

Happened again...

"xml:readonly:/etc/gconf/gconf.xml.defaults" to a read-only configuration source at position 2
Feb 3 13:21:21 cloudfall kernel: do_IRQ: 1.112 No irq handler for vector
Feb 3 13:23:23 cloudfall syslogd 1.4.1: restart.

When I rebooted, I had hard drive problems. Hard rebooted again and I've had no problems yet.

Instructions of how to downgrade the kernel using Yum would be greatly appreciated. I can't even find the kernel package in yum!

ninjagecko · #3 4th February 2007, 01:47 AM

update: It seems that downgrading my kernel to 2.6.18-2... stopped the IRQ problems!
However, now I can't find a way to install the nvidia drivers... compiling works but xgl fails to load; the livna repositories don't support 2.6.18 (or at least, I have no idea how to download a slightly-older version of a package; I think I've tried eveything right about now...)