Dell server by any chance?<div><br></div><div>I have a similar problem with a TE220B in a Dell 1950 III server - i've seen several other people having issues with digium cards in dell servers as well.</div><div><br></div>
<div>I've actually done something similar to what you have done - isolated the TE220B onto its own IRQ and set processor affinity for all the IRQs to particular cores... so far I haven't had kernel pancs since doing this, but its still a little too early to say if it has fixed the issue 100% or not.</div>
<div><br></div><div>--</div><div>Matt<br><br><div class="gmail_quote">On Mon, Mar 29, 2010 at 8:30 PM, James Lamanna <span dir="ltr"><<a href="mailto:jlamanna@gmail.com">jlamanna@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi,<br>
I'm trying to figure out the cause of a soft lockup I experienced:<br>
<br>
Mar 29 09:38:24 pstn1 kernel: BUG: soft lockup - CPU#0 stuck for 10s!<br>
[asterisk:32029]<br>
Mar 29 09:38:24 pstn1 kernel: Pid: 32029, comm: asterisk<br>
Mar 29 09:38:24 pstn1 kernel: EIP: 0060:[<c046e7fe>] CPU: 0<br>
Mar 29 09:38:24 pstn1 kernel: EIP is at kfree+0x68/0x6c<br>
Mar 29 09:38:24 pstn1 kernel: EFLAGS: 00000286 Tainted: GF<br>
(2.6.18-128.1.10.el5 #1)<br>
Mar 29 09:38:24 pstn1 kernel: EAX: 00000029 EBX: f7ff9380 ECX:<br>
f7fff880 EDX: c11ff9a0<br>
Mar 29 09:38:24 pstn1 kernel: ESI: 00000286 EDI: cffcda00 EBP:<br>
e5e10c80 DS: 007b ES: 007b<br>
Mar 29 09:38:24 pstn1 kernel: CR0: 80050033 CR2: b7ce39e0 CR3:<br>
0f911000 CR4: 000006d0<br>
Mar 29 09:38:24 pstn1 kernel: [<c05b067c>] kfree_skbmem+0x8/0x61<br>
Mar 29 09:38:24 pstn1 kernel: [<c05e9aaf>] __udp_queue_rcv_skb+0x4a/0x51<br>
Mar 29 09:38:24 pstn1 kernel: [<c05ad993>] release_sock+0x44/0x91<br>
Mar 29 09:38:24 pstn1 kernel: [<c05ea939>] udp_sendmsg+0x44e/0x514<br>
Mar 29 09:38:24 pstn1 kernel: [<c05efdec>] inet_sendmsg+0x35/0x3f<br>
Mar 29 09:38:24 pstn1 kernel: [<c05ab30c>] sock_sendmsg+0xce/0xe8<br>
Mar 29 09:38:24 pstn1 kernel: [<c043464f>] autoremove_wake_function+0x0/0x2d<br>
Mar 29 09:38:24 pstn1 kernel: [<c04ea17b>] copy_from_user+0x17/0x5d<br>
Mar 29 09:38:24 pstn1 kernel: [<c04ea3a1>] copy_to_user+0x31/0x48<br>
Mar 29 09:38:24 pstn1 kernel: [<f89ab141>] zt_chan_read+0x1e0/0x20b [zaptel]<br>
Mar 29 09:38:24 pstn1 kernel: [<c04ea195>] copy_from_user+0x31/0x5d<br>
Mar 29 09:38:24 pstn1 kernel: [<c05ac4c4>] sys_sendto+0x116/0x140<br>
Mar 29 09:38:24 pstn1 kernel: [<c0415d4f>] flush_tlb_page+0x74/0x77<br>
Mar 29 09:38:24 pstn1 kernel: [<c0461331>] do_wp_page+0x3bf/0x40a<br>
Mar 29 09:38:24 pstn1 kernel: [<c04284f1>] current_fs_time+0x4a/0x55<br>
Mar 29 09:38:24 pstn1 kernel: [<c0488f9b>] touch_atime+0x60/0x91<br>
Mar 29 09:38:24 pstn1 kernel: [<c047d9d0>] pipe_readv+0x315/0x321<br>
Mar 29 09:38:24 pstn1 kernel: [<c05acde4>] sys_socketcall+0x106/0x19e<br>
Mar 29 09:38:24 pstn1 kernel: [<c0404f17>] syscall_call+0x7/0xb<br>
Mar 29 09:38:24 pstn1 kernel: =======================<br>
<br>
<br>
This occurred during a "high load" period (52 calls across 3 PRI spans).<br>
<br>
A couple days ago I moved the interrupts for my PRI card to CPU0 from<br>
CPU3, because CPU3 was handling everything else:<br>
CPU0 CPU1 CPU2 CPU3<br>
0: 306 0 0 3684057379 IO-APIC-edge timer<br>
1: 0 0 0 13468 IO-APIC-edge i8042<br>
8: 0 0 0 3 IO-APIC-edge rtc<br>
9: 0 0 0 0 IO-APIC-level acpi<br>
12: 0 0 0 4 IO-APIC-edge i8042<br>
169: 0 0 0 0 IO-APIC-level uhci_hcd:usb2<br>
177: 0 0 0 18392593 IO-APIC-level ata_piix<br>
185: 0 0 0 1 IO-APIC-level ehci_hcd:usb1<br>
193: 0 0 0 0 IO-APIC-level uhci_hcd:usb3<br>
201: 0 0 0 2090021759 IO-APIC-level eth0<br>
209: 149621223 0 0 3534419461 IO-APIC-level wct4xxp<br>
<br>
<br>
(The CPU3 number for wct4xxp is not increasing any more).<br>
<br>
What is the interrupt distribution of other people's systems?<br>
Before I made this change I was having a problem with D-channels<br>
dropping occasionally, so I thought it might be an interrupt/load<br>
issue.<br>
<br>
Thank you.<br>
<br>
-- James<br>
<br>
--<br>
_____________________________________________________________________<br>
-- Bandwidth and Colocation Provided by <a href="http://www.api-digital.com" target="_blank">http://www.api-digital.com</a> --<br>
New to Asterisk? Join us for a live introductory webinar every Thurs:<br>
<a href="http://www.asterisk.org/hello" target="_blank">http://www.asterisk.org/hello</a><br>
<br>
asterisk-users mailing list<br>
To UNSUBSCRIBE or update options visit:<br>
<a href="http://lists.digium.com/mailman/listinfo/asterisk-users" target="_blank">http://lists.digium.com/mailman/listinfo/asterisk-users</a><br>
</blockquote></div><br></div>