Re: New timeofday subsystem: Lockups
On Fri, 2005-07-22 at 02:08 -0600, Frank Sorenson wrote: > I'm not sure whether this is an issue with John's TOD patches, John's > NTP rework, or Nish's softtimer patches, but something in this > combination seems to be locking up my system frequently. Often, it will > completely hang during boot, and it periodically hangs while starting X > or even just under normal (unstressed) use. > > During several of the boot-ups, the NMI Watchdog has caught lockups. > Here is the output of one of those lockups (hand-copied, so hopefully no > mistakes): Thanks for the detailed bug report! Hmmm. It looks like the TSC interpolator is deadlocking. I'll try to dig in and hunt that one down (although forgive me if I don't get back to you until sometime next week). Thanks again for the great testing and reporting! -john - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
New timeofday subsystem: Lockups
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 John, Nish, others: I'm not sure whether this is an issue with John's TOD patches, John's NTP rework, or Nish's softtimer patches, but something in this combination seems to be locking up my system frequently. Often, it will completely hang during boot, and it periodically hangs while starting X or even just under normal (unstressed) use. During several of the boot-ups, the NMI Watchdog has caught lockups. Here is the output of one of those lockups (hand-copied, so hopefully no mistakes): NMI Watchdog detected LOCKUP on CPU0, eip c01248bb, registers: Modules linked in: ipw2200 ieee80211 ieee80211_crypt CPU:0 EIP:0060:[] Not tainted VLI EFLAGS: 0046 (2.6.13-rc3-skas3-v9-pre7-fs6) EIP is at get_jiffies_64+0x2b/0x40 eax: ebx: ffbf503ecx: 00011c84 edx: esi: edi: ebp: c05cba00 esp: c065ff00 ds: 007b es: 007b ss: 0068 Process swapper (pid: 0, threadinfo=c065f000 task=c0552b80) Stack: 94ad 88e3 c03b01d7 c06aca50 c0330d90 67347457 000c 88e1 c0558080 c065f000 f7c42de4 0082 c05cba00 c013e908 c06ac9c0 0080 0aa2 0aa2 fffb7e00 f7c42de4 0082 Call Trace: [] read_tsc_interp+0x17/0x120 [] __ide_do_rw_disk+0x200/0540 [] do_monotonic_clock+0x28/0x130 [] add_timer+0x16/0x60 [] ide_set_handler+0x28/0x60 [] task_in_intr+0x4f/0x100 [] ide_intr+0x95/0x1d0 [] task_in_intr+0x0/0x100 [] handle_IRQ_event+0x33/0x70 [] __do_IRQ+0x53/0xa0 = [] read_tsc_interp+0x17/0x120 [] common_interrupt+0x1a/0x20 [] alloc_pidmap+0x8b/0x1f0 [] tsc_interp_sync+0x4f/0xa0 [] tsc_interp_sync+0x0/0xa0 [] run_timer_softirq+0xb4/0x1d0 [] __do_softirq+0x42/0xa0 [] do_softirq+0x4e/0x60 === [] irq_exit+0x35/0x40 [] do_IRQ+0x5a/0xa0 [] common_interrupt+0x1a/0x20 [] acpi_processor_idle+0x10e/0x28e [] default_idle+0x0/0x30 [] cpu_idle+0x34/0x50 [] start_kernel+0x186/0x1f0 Code: 8b 0d 64 da 68 c0 56 53 90 8d b4 26 00 00 00 00 89 c8 8b 1d 80 5e 55 c0 8b 35 84 5e 55 c0 89 ca 8b 0d 64 da 68 c0 83 e2 01 31 c8 <09> c2 75 e1 89 d8 89 f2 5b 5e c3 90 90 90 90 90 90 90 90 90 90 console shuts up ... Hopefully someone can make something out of this! Any ideas? Thanks, Frank - -- Frank Sorenson - KD7TZK Systems Manager, Computer Science Department Brigham Young University [EMAIL PROTECTED] -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFC4KloaI0dwg4A47wRAiosAKD7Hn+nzJEizqKvXDaXIfXw0T+0RACgrQnF NikzhdmXzPjL0Bi2D2aBOk4= =EN1O -END PGP SIGNATURE- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
New timeofday subsystem: Lockups
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 John, Nish, others: I'm not sure whether this is an issue with John's TOD patches, John's NTP rework, or Nish's softtimer patches, but something in this combination seems to be locking up my system frequently. Often, it will completely hang during boot, and it periodically hangs while starting X or even just under normal (unstressed) use. During several of the boot-ups, the NMI Watchdog has caught lockups. Here is the output of one of those lockups (hand-copied, so hopefully no mistakes): NMI Watchdog detected LOCKUP on CPU0, eip c01248bb, registers: Modules linked in: ipw2200 ieee80211 ieee80211_crypt CPU:0 EIP:0060:[c01248bb] Not tainted VLI EFLAGS: 0046 (2.6.13-rc3-skas3-v9-pre7-fs6) EIP is at get_jiffies_64+0x2b/0x40 eax: ebx: ffbf503ecx: 00011c84 edx: esi: edi: ebp: c05cba00 esp: c065ff00 ds: 007b es: 007b ss: 0068 Process swapper (pid: 0, threadinfo=c065f000 task=c0552b80) Stack: 94ad 88e3 c03b01d7 c06aca50 c0330d90 67347457 000c 88e1 c0558080 c065f000 f7c42de4 0082 c05cba00 c013e908 c06ac9c0 0080 0aa2 0aa2 fffb7e00 f7c42de4 0082 Call Trace: [c03b01d7] read_tsc_interp+0x17/0x120 [c0330d90] __ide_do_rw_disk+0x200/0540 [c013e908] do_monotonic_clock+0x28/0x130 [c01283e6] add_timer+0x16/0x60 [c0328738] ide_set_handler+0x28/0x60 [c032c7df] task_in_intr+0x4f/0x100 [c0327405] ide_intr+0x95/0x1d0 [c032c790] task_in_intr+0x0/0x100 [c014cae3] handle_IRQ_event+0x33/0x70 [c014cbe6] __do_IRQ+0x53/0xa0 = [c03b01d7] read_tsc_interp+0x17/0x120 [c0104b86] common_interrupt+0x1a/0x20 [c013007b] alloc_pidmap+0x8b/0x1f0 [c03b016f] tsc_interp_sync+0x4f/0xa0 [c03b0120] tsc_interp_sync+0x0/0xa0 [c0128884] run_timer_softirq+0xb4/0x1d0 [c0124912] __do_softirq+0x42/0xa0 [c010688e] do_softirq+0x4e/0x60 === [c0124a35] irq_exit+0x35/0x40 [c010674a] do_IRQ+0x5a/0xa0 [c0104b86] common_interrupt+0x1a/0x20 [c02b16fa] acpi_processor_idle+0x10e/0x28e [c0102030] default_idle+0x0/0x30 [c01020d4] cpu_idle+0x34/0x50 [c0621856] start_kernel+0x186/0x1f0 Code: 8b 0d 64 da 68 c0 56 53 90 8d b4 26 00 00 00 00 89 c8 8b 1d 80 5e 55 c0 8b 35 84 5e 55 c0 89 ca 8b 0d 64 da 68 c0 83 e2 01 31 c8 09 c2 75 e1 89 d8 89 f2 5b 5e c3 90 90 90 90 90 90 90 90 90 90 console shuts up ... Hopefully someone can make something out of this! Any ideas? Thanks, Frank - -- Frank Sorenson - KD7TZK Systems Manager, Computer Science Department Brigham Young University [EMAIL PROTECTED] -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFC4KloaI0dwg4A47wRAiosAKD7Hn+nzJEizqKvXDaXIfXw0T+0RACgrQnF NikzhdmXzPjL0Bi2D2aBOk4= =EN1O -END PGP SIGNATURE- - To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: New timeofday subsystem: Lockups
On Fri, 2005-07-22 at 02:08 -0600, Frank Sorenson wrote: I'm not sure whether this is an issue with John's TOD patches, John's NTP rework, or Nish's softtimer patches, but something in this combination seems to be locking up my system frequently. Often, it will completely hang during boot, and it periodically hangs while starting X or even just under normal (unstressed) use. During several of the boot-ups, the NMI Watchdog has caught lockups. Here is the output of one of those lockups (hand-copied, so hopefully no mistakes): Thanks for the detailed bug report! Hmmm. It looks like the TSC interpolator is deadlocking. I'll try to dig in and hunt that one down (although forgive me if I don't get back to you until sometime next week). Thanks again for the great testing and reporting! -john - To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/