hpet driver backtraces when hyperthreading enabled

Bug #473702 reported by Mike Pastore
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
New
Undecided
Unassigned
Nominated for Karmic by Lukas Koranda
Nominated for Lucid by Lukas Koranda

Bug Description

Binary package hint: linux-image-2.6.31-14-generic

Ubuntu 9.10 Desktop AMD64

The following appears in dmesg when Hyperthreading is enabled in the BIOS:

[ 1.859689] ------------[ cut here ]------------
[ 1.859695] WARNING: at /build/buildd/linux-2.6.31/arch/x86/kernel/hpet.c:390 hpet_next_event+0x75/0x90()
[ 1.859697] Hardware name:
[ 1.859698] Modules linked in:
[ 1.859700] Pid: 0, comm: swapper Not tainted 2.6.31-14-generic #48-Ubuntu
[ 1.859701] Call Trace:
[ 1.859706] [<ffffffff8105e788>] warn_slowpath_common+0x78/0xb0
[ 1.859708] [<ffffffff8105e7cf>] warn_slowpath_null+0xf/0x20
[ 1.859710] [<ffffffff81034335>] hpet_next_event+0x75/0x90
[ 1.859712] [<ffffffff8103437b>] hpet_legacy_next_event+0xb/0x10
[ 1.859715] [<ffffffff810856df>] clockevents_program_event+0x4f/0x90
[ 1.859718] [<ffffffff81086bd0>] tick_dev_program_event+0x40/0xd0
[ 1.859720] [<ffffffff81086566>] tick_broadcast_oneshot_control+0x116/0x120
[ 1.859723] [<ffffffff81085e50>] tick_notify+0x130/0x1a0
[ 1.859725] [<ffffffff8152c197>] notifier_call_chain+0x47/0x90
[ 1.859728] [<ffffffff8107d841>] raw_notifier_call_chain+0x11/0x20
[ 1.859730] [<ffffffff810854e4>] clockevents_notify+0x34/0xa0
[ 1.859733] [<ffffffff812d7876>] lapic_timer_state_broadcast+0x41/0x43
[ 1.859735] [<ffffffff812d7dd3>] acpi_idle_enter_bm+0x185/0x2bf
[ 1.859737] [<ffffffff812d7c40>] ? acpi_idle_enter_c1+0xae/0xbc
[ 1.859739] [<ffffffff813fe40b>] cpuidle_idle_call+0x9b/0xf0
[ 1.859742] [<ffffffff81010e12>] cpu_idle+0xb2/0x100
[ 1.859745] [<ffffffff815239f6>] start_secondary+0xa9/0xab
[ 1.859748] ---[ end trace 24e97c3b18305fae ]---

It does not appear when Hyperthreading is disabled in the BIOS.

I have a strong suspicion that this bug will only appear on Core i7-8xx chips. This is the Nehalem/Lynnfield chip, Socket LGA-1156, with the new P55 Express chipset. There are currently two models: the 860 (2.8 GHz) and the 870 (2.93 GHz). There is a third chip for this socket, the Core i5-750 (2.66 GHz), but it does not support Hyperthreading, so it would not be affected.

Revision history for this message
Mike Pastore (mwpastore) wrote :
Revision history for this message
Mike Pastore (mwpastore) wrote :
Revision history for this message
Mike Pastore (mwpastore) wrote :
Revision history for this message
Mike Pastore (mwpastore) wrote :
Revision history for this message
Mike Pastore (mwpastore) wrote :

lspci is the same with ht disabled

Revision history for this message
Mike Pastore (mwpastore) wrote :
Revision history for this message
Yucong Sun (sunyucong) wrote :

happened to me as well, both 2.6.28 & 2.6.31

cpu is an single quad core with HT X3450

[ 2.158796] ------------[ cut here ]------------
[ 2.158808] WARNING: at /build/buildd/linux-2.6.31/arch/x86/kernel/hpet.c:390 hpet_next_event+0x75/0x90()
[ 2.158817] Hardware name: PowerEdge T310
[ 2.158821] Modules linked in:
[ 2.158829] Pid: 0, comm: swapper Not tainted 2.6.31-14-server #48-Ubuntu
[ 2.158837] BIOS EDD facility v0.16 2004-Jun-25, 0 devices found
[ 2.158842] EDD information not available.
[ 2.158848] Call Trace:
[ 2.158858] [<ffffffff8105e618>] warn_slowpath_common+0x78/0xb0
[ 2.158865] [<ffffffff8105e65f>] warn_slowpath_null+0xf/0x20
[ 2.158871] [<ffffffff81034335>] hpet_next_event+0x75/0x90
[ 2.158877] [<ffffffff8103437b>] hpet_legacy_next_event+0xb/0x10
[ 2.158885] [<ffffffff810851df>] clockevents_program_event+0x4f/0x90
[ 2.158893] [<ffffffff810866d0>] tick_dev_program_event+0x40/0xd0
[ 2.158900] [<ffffffff81086066>] tick_broadcast_oneshot_control+0x116/0x120
[ 2.158907] [<ffffffff81085950>] tick_notify+0x130/0x1a0
[ 2.158916] [<ffffffff81529887>] notifier_call_chain+0x47/0x90
[ 2.158923] [<ffffffff8107d331>] raw_notifier_call_chain+0x11/0x20
[ 2.158929] [<ffffffff81084fe4>] clockevents_notify+0x34/0xa0
[ 2.158937] [<ffffffff812d5b6a>] lapic_timer_state_broadcast+0x41/0x43
[ 2.158943] [<ffffffff812d60c7>] acpi_idle_enter_bm+0x185/0x2bf
[ 2.158949] [<ffffffff812d5f34>] ? acpi_idle_enter_c1+0xae/0xbc
[ 2.158958] [<ffffffff813fc13b>] cpuidle_idle_call+0x9b/0xf0
[ 2.158966] [<ffffffff81010e02>] cpu_idle+0xb2/0x100
[ 2.158972] [<ffffffff81521166>] start_secondary+0xa9/0xab
[ 2.158977] ---[ end trace decc02a3eb18b513 ]---

Revision history for this message
Vadim Zeitlin (vz-ubuntu) wrote :

Also happens with i860 on Gigabyte GA-P55-UD5 motherboard.

This problem is discussed in this LKML thread: http://thread.gmane.org/gmane.linux.kernel/913374/

Revision history for this message
Lukas Koranda (lkoranda) wrote :

Intel(R) Core(TM) i7 CPU Q 720 @ 1.60GHz on P55 chipset

Lucid Beta2 2.6.32-19-generic x86_64

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.