August 31, 200916 yr I'm getting this message in the log file about every 3 minutes: Tower kernel: Hangcheck: hangcheck value past margin! I'm not seeing any other errors (Log file attached) I just moved to a new MB so this is the first time I've seen this. A quick forum search didn't help. There were some references to it.. but it never really was explained. The other times there were other errors associated with it.. Jim Aug 30 15:37:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:40:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:43:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:46:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:49:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:52:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:55:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 15:58:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:01:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:04:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:07:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:10:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:13:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:16:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:19:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:22:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:25:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:28:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:31:30 Tower kernel: Hangcheck: hangcheck value past margin! Aug 30 16:34:30 Tower kernel: Hangcheck: hangcheck value past margin!
September 1, 200916 yr Author I am also seeing some weirdness where the network drops out for a while and then comes back. When that happens I get in the log: Aug 31 15:40:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 15:43:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 16:19:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 16:22:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 16:58:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:01:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:02:25 Tower kernel: r8169: eth0: link up Aug 31 17:04:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:07:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:46:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:49:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:52:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 17:54:07 Tower kernel: r8169: eth0: link up Aug 31 18:12:33 Tower emhttp: shcmd (94): /usr/sbin/hdparm -y /dev/hda >/dev/null Aug 31 18:47:37 Tower kernel: r8169: eth0: link up Aug 31 18:49:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 18:52:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 18:55:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 18:58:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:01:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:04:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:07:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:10:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:13:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:16:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:19:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:22:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:43:07 Tower dhcpcd[1208]: sending DHCP_REQUEST for 192.168.2.80 to 192.168.2.1 Aug 31 19:43:07 Tower dhcpcd[1208]: dhcpIPaddrLeaseTime=86400 in DHCP server response. Aug 31 19:43:07 Tower dhcpcd[1208]: DHCP_ACK received from (192.168.2.1) Aug 31 19:55:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 19:58:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:01:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:04:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:07:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:10:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:13:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:16:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:19:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:22:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:25:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:28:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:31:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:34:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:37:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:40:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:43:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:46:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:49:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:52:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:52:34 Tower emhttp: shcmd (95): /usr/sbin/hdparm -y /dev/hda >/dev/null Aug 31 20:55:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 20:56:34 Tower emhttp: shcmd (96): /usr/sbin/hdparm -y /dev/hdb >/dev/null Aug 31 20:56:34 Tower emhttp: shcmd (97): /usr/sbin/hdparm -y /dev/sde >/dev/null Aug 31 20:58:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:01:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:04:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:06:35 Tower emhttp: shcmd (98): /usr/sbin/hdparm -y /dev/sdd >/dev/null Aug 31 21:07:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:10:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:13:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:16:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:19:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:22:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:25:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:28:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:31:31 Tower kernel: Hangcheck: hangcheck value past margin! Aug 31 21:34:31 Tower kernel: Hangcheck: hangcheck value past margin! It seems like I lose the connection and then it comes back! Could they be related? IS there something wrong with this MB? Should I RMA it?
September 1, 200916 yr See here: it is a linux kernel issue... http://www.lubby.org/stable/index.php?page=Databases.result2&ArticleID=77 You might try the current release if it handles your MB better than 4.4.2. Otherwise, some of the boot codes might be needed. Your network interface seems to be involved with this series of error messages: Aug 31 06:37:49 Tower kernel: ------------[ cut here ]------------ Aug 31 06:37:49 Tower kernel: WARNING: at net/sched/sch_generic.c:219 dev_watchdog+0xf0/0x16d() Aug 31 06:37:49 Tower kernel: NETDEV WATCHDOG: eth0 (r8169): transmit timed out Aug 31 06:37:49 Tower kernel: Modules linked in: md_mod ide_disk ahci r8169 pata_jmicron libata jmicron ide_core Aug 31 06:37:49 Tower kernel: Pid: 0, comm: swapper Not tainted 2.6.27.7-unRAID #3 Aug 31 06:37:49 Tower kernel: [<c011cc08>] warn_slowpath+0x61/0x86 Aug 31 06:37:49 Tower kernel: [<c011545b>] enqueue_task+0xa/0x14 Aug 31 06:37:49 Tower kernel: [<c01154eb>] activate_task+0x16/0x1b Aug 31 06:37:49 Tower kernel: [<c011816c>] try_to_wake_up+0x11c/0x125 Aug 31 06:37:49 Tower kernel: [<c012b594>] autoremove_wake_function+0xc/0x2b Aug 31 06:37:49 Tower kernel: [<c01157c0>] __wake_up_common+0x34/0x58 Aug 31 06:37:49 Tower kernel: [<c011603d>] __wake_up+0x29/0x39 Aug 31 06:37:49 Tower kernel: [<c015185a>] dma_pool_free+0xde/0x128 Aug 31 06:37:49 Tower kernel: [<c0130a7b>] clocksource_get_next+0x39/0x3f Aug 31 06:37:49 Tower kernel: [<c012fafa>] update_wall_time+0x584/0x71d Aug 31 06:37:49 Tower kernel: [<c020e218>] strlcpy+0x14/0x41 Aug 31 06:37:49 Tower kernel: [<c02c2491>] dev_watchdog+0xf0/0x16d Aug 31 06:37:49 Tower kernel: [<c0123c56>] lock_timer_base+0x19/0x35 Aug 31 06:37:49 Tower kernel: [<c0123d5d>] __mod_timer+0x8d/0x97 Aug 31 06:37:49 Tower kernel: [<c02bd127>] neigh_periodic_timer+0x10f/0x11f Aug 31 06:37:49 Tower kernel: [<c02c23a1>] dev_watchdog+0x0/0x16d Aug 31 06:37:49 Tower kernel: [<c01239be>] run_timer_softirq+0x107/0x15a Aug 31 06:37:49 Tower kernel: [<c01204b9>] __do_softirq+0x6c/0xcf Aug 31 06:37:49 Tower kernel: [<c012054e>] do_softirq+0x32/0x36 Aug 31 06:37:49 Tower kernel: [<c0105179>] do_IRQ+0x54/0x67 Aug 31 06:37:49 Tower kernel: [<c01035a3>] common_interrupt+0x23/0x28 Aug 31 06:37:49 Tower kernel: [<c01200d8>] sys_adjtimex+0x1a/0x4f Aug 31 06:37:49 Tower kernel: [<c0242f6c>] acpi_idle_enter_simple+0x15a/0x1c2 Aug 31 06:37:49 Tower kernel: [<c029fe13>] cpuidle_idle_call+0x55/0x86 Aug 31 06:37:49 Tower kernel: [<c01019e1>] cpu_idle+0xbd/0xd5 Aug 31 06:37:49 Tower kernel: ======================= Aug 31 06:37:49 Tower kernel: ---[ end trace aeb6fd3cb4230753 ]---
September 1, 200916 yr Author I'm also running my system really slow to save power. It's a 1.8GHz CPU running at 900MHz Might that have anything to do with it? I have reset the system to normal and I'll see how it goes. I was really hoping to underclock to save power. Jim
September 1, 200916 yr Author Interesting.... After I bumped my CPU speed back to normal.. I didn't see any of the hangcheck warnings... It ran all night without getting one.. I'm wondering why slowing the speed down would introduce those errors. As I understand it... It's a descrepnecy between two different kinds of timers. Maybe when running slow, one of the timers isn't calculated correctly? I assume one is a hardware timer and the other is counted with CPU cycles? any thoughts on how I can debug this? Or should I just punt on this.. I'll see if I still get my network drops at the faster cpu/FSB/Memory I want to be able to run slow! *pout*
September 5, 200916 yr The Lubby site that Joe linked mentions that this is comparing the hangcheck time with the TSC timer, and your posted syslog indicates that you had an unstable TSC timer: Aug 29 22:07:47 Tower kernel: Marking TSC unstable due to TSC halts in idle This line shows the HangCheck setup: Aug 29 22:07:47 Tower kernel: Hangcheck: starting hangcheck timer 0.9.0 (tick is 180 seconds, margin is 60 seconds). That certainly corresponds to the precise 3 minute frequency of your syslog messages, but a margin of 60 seconds?!? I can see underclocking causing delays of milliseconds, but longer than 60 seconds??!!?? That is very hard to believe! You might check to see if you still have an unstable TSC with your normal clocking. Then you might try underclocking less radically, try 1.50GHz, then 1.20GHz ... An unstable TSC is not uncommon, and the kernel detects and works around it normally, but yours may be too far off. Aug 31 06:37:49 Tower kernel: ------------[ cut here ]------------ Aug 31 06:37:49 Tower kernel: WARNING: at net/sched/sch_generic.c:219 dev_watchdog+0xf0/0x16d() Aug 31 06:37:49 Tower kernel: NETDEV WATCHDOG: eth0 (r8169): transmit timed out There are several users with this problem, with the "NETDEV WATCHDOG" and "transmit timed out" and some network disruption, all using the r8169 driver. The best advice seems to be to install a better NIC (such as an Intel PRO/1000 card) and disable the onboard Realtek NIC. Search watchdog and r8169 for other posts about this.
Archived
This topic is now archived and is closed to further replies.