lockdep

[Jay Cliburn]

[jcliburn at osprey ~]$ uname -rm
2.6.17-1.2391.fc6 x86_64

After updating FC6T1 in the past 12 hours, I see the following lockdep
message.

=======================================================
[ INFO: possible circular locking dependency detected ]
-------------------------------------------------------
cpuspeed/1527 is trying to acquire lock:
 (&policy->lock){--..}, at: [<ffffffff802681da>] mutex_lock+0x2a/0x2e

but task is already holding lock:
 (cpucontrol){--..}, at: [<ffffffff802681da>] mutex_lock+0x2a/0x2e

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #2 (cpucontrol){--..}:
       [<ffffffff802ab51c>] lock_acquire+0x7a/0xa1
       [<ffffffff80267ffc>] __mutex_lock_slowpath+0xeb/0x29f
       [<ffffffff802681d9>] mutex_lock+0x29/0x2e
       [<ffffffff802af1cc>] __lock_cpu_hotplug+0x3c/0x5f
       [<ffffffff802af209>] lock_cpu_hotplug+0xa/0xd
       [<ffffffff8041ad7c>] __cpufreq_driver_target+0x1a/0x81
       [<ffffffff8041c03d>] cpufreq_governor_userspace+0x1e9/0x22c
       [<ffffffff8041a734>] __cpufreq_governor+0x74/0x107
       [<ffffffff8041a99c>] __cpufreq_set_policy+0x1d5/0x1e7
       [<ffffffff8041a9e9>] cpufreq_set_policy+0x3b/0x97
       [<ffffffff8041b5c5>] cpufreq_add_dev+0x3ac/0x57b
       [<ffffffff803bae72>] sysdev_driver_register+0xa7/0x13a
       [<ffffffff8041a5e0>] cpufreq_register_driver+0xc1/0x1a1
       [<ffffffff8027fecc>] powernowk8_init+0x7e/0x88
       [<ffffffff8026f864>] init+0x1fc/0x3cd
       [<ffffffff80262bdd>] child_rip+0x7/0x12

-> #1 (userspace_mutex){--..}:
       [<ffffffff802ab51c>] lock_acquire+0x7a/0xa1
       [<ffffffff80267ffc>] __mutex_lock_slowpath+0xeb/0x29f
       [<ffffffff802681d9>] mutex_lock+0x29/0x2e
       [<ffffffff8041beb9>] cpufreq_governor_userspace+0x65/0x22c
       [<ffffffff8041a734>] __cpufreq_governor+0x74/0x107
       [<ffffffff8041a93b>] __cpufreq_set_policy+0x174/0x1e7
       [<ffffffff8041a9e9>] cpufreq_set_policy+0x3b/0x97
       [<ffffffff8041b5c5>] cpufreq_add_dev+0x3ac/0x57b
       [<ffffffff803bae72>] sysdev_driver_register+0xa7/0x13a
       [<ffffffff8041a5e0>] cpufreq_register_driver+0xc1/0x1a1
       [<ffffffff8027fecc>] powernowk8_init+0x7e/0x88
       [<ffffffff8026f864>] init+0x1fc/0x3cd
       [<ffffffff80262bdd>] child_rip+0x7/0x12

-> #0 (&policy->lock){--..}:
       [<ffffffff802ab51c>] lock_acquire+0x7a/0xa1
       [<ffffffff80267ffc>] __mutex_lock_slowpath+0xeb/0x29f
       [<ffffffff802681d9>] mutex_lock+0x29/0x2e
       [<ffffffff8041ab93>] store_scaling_governor+0x14e/0x19c
       [<ffffffff8027765a>] store+0x4b/0x66
       [<ffffffff803105f7>] sysfs_write_file+0xd0/0x103
       [<ffffffff80217528>] vfs_write+0xce/0x175
       [<ffffffff80217e16>] sys_write+0x46/0x70
       [<ffffffff80261c8d>] system_call+0x7d/0x83

other info that might help us debug this:

1 lock held by cpuspeed/1527:
 #0:  (cpucontrol){--..}, at: [<ffffffff802681da>] mutex_lock+0x2a/0x2e

stack backtrace:

Call Trace:
 [<ffffffff80270865>] show_trace+0xaa/0x23d
 [<ffffffff80270a0d>] dump_stack+0x15/0x17
 [<ffffffff802a9bda>] print_circular_bug_tail+0x6c/0x77
 [<ffffffff802ab299>] __lock_acquire+0x8a2/0xaab
 [<ffffffff802ab51d>] lock_acquire+0x7b/0xa1
 [<ffffffff80267ffd>] __mutex_lock_slowpath+0xec/0x29f
 [<ffffffff802681da>] mutex_lock+0x2a/0x2e
 [<ffffffff8041ab94>] store_scaling_governor+0x14f/0x19c
 [<ffffffff8027765b>] store+0x4c/0x66
 [<ffffffff803105f8>] sysfs_write_file+0xd1/0x103
 [<ffffffff80217529>] vfs_write+0xcf/0x175
 [<ffffffff80217e17>] sys_write+0x47/0x70
 [<ffffffff80261c8e>] system_call+0x7e/0x83