Skip to contentSkip to conversation

New: BUG: soft lockup detected on CPU#0!

Bugme-new Andrew Morton 3 participants 3 messages Nov 16, 2007 – Nov 19, 2007 1 message not in the archive ACPI

lore
Share & tools
lore

Andrew Morton

author
#1 lore raw
In reply to a message not in the archive

(switched to email. Please reply via emailed reply-to-all, not via the bugzilla web interface).

Quoted 17 lines · On Fri, 16 Nov 2007 14:56:13 -0800 (PST) [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=9398

           Summary: BUG: soft lockup detected on CPU#0!
           Product: Other
           Version: 2.5
     KernelVersion: 2.6.23.8
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: Other
        AssignedTo: [email protected]
        ReportedBy: [email protected]
Quoted 1 line

Most recent kernel where this bug did not occur: 2.6.23.0

Ow. Are you sure? This is a regression which was added into the 2.6.23 stable tree?

Quoted 30 lines

Distribution: Debian stable
Hardware Environment: Pentium M laptop
Software Environment: Debian stable
Problem Description:

Steps to reproduce:

i2c-adapter i2c-5: SMBus Quick command not supported, can't probe for chips i2c-adapter i2c-6: SMBus Quick command not supported, can't probe for chips
pcmcia: Detected deprecated PCMCIA ioctl usage from process: discover.
pcmcia: This interface will soon be removed from the kernel; please expect breakage unless you upgrade to new tools.
pcmcia: see http://www.kernel.org/pub/linux/utils/kernel/pcmcia/pcmcia.html for details.
eth0: link down
ieee80211_crypt: registered algorithm 'TKIP'

BUG: soft lockup detected on CPU#0!
 [<c0140bf4>] softlockup_tick+0x90/0xaf
 [<c0120a17>] update_process_times+0x32/0x54
 [<c012edfd>] tick_periodic+0x6e/0x78
 [<c012ee16>] tick_handle_periodic+0xf/0x5d
 [<c0126390>] insert_work+0x59/0x5c
 [<c012ef3c>] tick_do_broadcast+0x1f/0x3f
 [<c012f05a>] tick_do_periodic_broadcast+0x1a/0x31
 [<c012f08c>] tick_handle_periodic_broadcast+0x1b/0x5b
 [<c01272fb>] __rcu_process_callbacks+0x112/0x170
 [<c0106e83>] timer_interrupt+0x34/0x3d
 [<c0140e66>] handle_IRQ_event+0x1a/0x3f
 [<c01420df>] handle_level_irq+0x77/0xd0
 [<c01061c5>] do_IRQ+0x75/0x8c
Quoted 8 lines
 [<c0220d76>] acpi_hw_register_write+0x11b/0x14b
 [<c0104713>] common_interrupt+0x23/0x28
 [<c0110000>] mc_sysdev_remove+0x2b/0x4b
 [<c0232ddf>] acpi_processor_idle+0x22f/0x398
 [<c0102344>] cpu_idle+0x43/0x71
 [<c03d29de>] start_kernel+0x250/0x255
 [<c03d2317>] unknown_bootoption+0x0/0x195
 =======================

Looks like acpi_hw_register_write() has locked up. Or someone is continuously calling acpi_hw_register_write().

I assume that the mc_sysdev_remove() in there is just stack garbage. To confirm this could you please set CONFIG_MICROCODE=n and retest?

Also, it would be interesting to test whether we have introduced this bug into 2.6.24-rc2 (or -rc3, if it's out).

Thanks.

#2 lore raw
Quoted 7 lines · Andrew Morton
Quoted 4 lines
     KernelVersion: 2.6.23.8

Most recent kernel where this bug did not occur: 2.6.23.0

Ow. Are you sure? This is a regression which was added into the 2.6.23 stable tree?

Can you see if this shows up in 2.6.23.2 or 2.6.23.3? Those are the releases that did core kernel changes.

#3 lore raw
In reply to Andrew Morton
Quoted 67 lines · Andrew Morton

(switched to email. Please reply via emailed reply-to-all, not via the bugzilla web interface).

Quoted 17 lines · On Fri, 16 Nov 2007 14:56:13 -0800 (PST) [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=9398

           Summary: BUG: soft lockup detected on CPU#0!
           Product: Other
           Version: 2.5
     KernelVersion: 2.6.23.8
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: Other
        AssignedTo: [email protected]
        ReportedBy: [email protected]
Quoted 1 line

Most recent kernel where this bug did not occur: 2.6.23.0

Ow. Are you sure? This is a regression which was added into the 2.6.23 stable tree?

Quoted 30 lines

Distribution: Debian stable
Hardware Environment: Pentium M laptop
Software Environment: Debian stable
Problem Description:

Steps to reproduce:

i2c-adapter i2c-5: SMBus Quick command not supported, can't probe for chips i2c-adapter i2c-6: SMBus Quick command not supported, can't probe for chips
pcmcia: Detected deprecated PCMCIA ioctl usage from process: discover.
pcmcia: This interface will soon be removed from the kernel; please expect breakage unless you upgrade to new tools.
pcmcia: see http://www.kernel.org/pub/linux/utils/kernel/pcmcia/pcmcia.html for details.
eth0: link down
ieee80211_crypt: registered algorithm 'TKIP'

BUG: soft lockup detected on CPU#0!
 [<c0140bf4>] softlockup_tick+0x90/0xaf
 [<c0120a17>] update_process_times+0x32/0x54
 [<c012edfd>] tick_periodic+0x6e/0x78
 [<c012ee16>] tick_handle_periodic+0xf/0x5d
 [<c0126390>] insert_work+0x59/0x5c
 [<c012ef3c>] tick_do_broadcast+0x1f/0x3f
 [<c012f05a>] tick_do_periodic_broadcast+0x1a/0x31
 [<c012f08c>] tick_handle_periodic_broadcast+0x1b/0x5b
 [<c01272fb>] __rcu_process_callbacks+0x112/0x170
 [<c0106e83>] timer_interrupt+0x34/0x3d
 [<c0140e66>] handle_IRQ_event+0x1a/0x3f
 [<c01420df>] handle_level_irq+0x77/0xd0
 [<c01061c5>] do_IRQ+0x75/0x8c
Quoted 8 lines
 [<c0220d76>] acpi_hw_register_write+0x11b/0x14b
 [<c0104713>] common_interrupt+0x23/0x28
 [<c0110000>] mc_sysdev_remove+0x2b/0x4b
 [<c0232ddf>] acpi_processor_idle+0x22f/0x398
 [<c0102344>] cpu_idle+0x43/0x71
 [<c03d29de>] start_kernel+0x250/0x255
 [<c03d2317>] unknown_bootoption+0x0/0x195
 =======================
Quoted 9 lines

Looks like acpi_hw_register_write() has locked up. Or someone is continuously calling acpi_hw_register_write().

I assume that the mc_sysdev_remove() in there is just stack garbage. To confirm this could you please set CONFIG_MICROCODE=n and retest?

Also, it would be interesting to test whether we have introduced this bug into 2.6.24-rc2 (or -rc3, if it's out).

The softlockup detector was patched in 2.6.23.2; before that it was broken for a few releases.

patch a115d5caca1a2905ba7a32b408a6042b20179aaa in mainline