[linux-next] BUG: sleeping function called from invalid context at kernel/irq/manage.c:110

[Sergey Senozhatsky]

Hello Thomas,

next-20170228

at resume after suspend:

[   45.383425] PM: noirq suspend of devices complete after 16.311 msecs
[   45.383795] ACPI: Preparing to enter system sleep state S3
[   45.384114] PM: Saving platform NVS memory
[   45.384167] Disabling non-boot CPUs ...
[   45.411548] Broke affinity for irq 25
[   45.412608] smpboot: CPU 1 is now offline
[   45.437328] Broke affinity for irq 26
[   45.438364] smpboot: CPU 2 is now offline
[   45.467319] Broke affinity for irq 27
[   45.468355] smpboot: CPU 3 is now offline
[   45.470112] ACPI: Low-level resume complete
[   45.470497] PM: Restoring platform NVS memory
[   45.470906] Suspended for 6.778 seconds
...
[   45.470914] BUG: sleeping function called from invalid context at kernel/irq/manage.c:110
[   45.470936] in_atomic(): 0, irqs_disabled(): 1, pid: 324, name: systemd-sleep
[   45.470950] 5 locks held by systemd-sleep/324:
[   45.470959]  #0:  (sb_writers#5){.+.+.+}, at: [<ffffffff81174220>] vfs_write+0xef/0x189
[   45.470984]  #1:  (&of->mutex){+.+.+.}, at: [<ffffffff811d8c47>] kernfs_fop_write+0xd7/0x169
[   45.471007]  #2:  (s_active#68){.+.+.+}, at: [<ffffffff811d8c4f>] kernfs_fop_write+0xdf/0x169
[   45.471030]  #3:  (pm_mutex){+.+.+.}, at: [<ffffffff810937ec>] pm_suspend+0x10d/0x68a
[   45.471052]  #4:  (acpi_scan_lock){+.+.+.}, at: [<ffffffff812c60c4>] acpi_scan_lock_acquire+0x17/0x19
[   45.471077] irq event stamp: 20382
[   45.471084] hardirqs last  enabled at (20381): [<ffffffff81163355>] kfree+0x19e/0x26a
[   45.471101] hardirqs last disabled at (20382): [<ffffffff81092bb8>] arch_suspend_disable_irqs+0xf/0x11
[   45.471121] softirqs last  enabled at (19414): [<ffffffff81586a93>] __do_softirq+0x34b/0x454
[   45.471139] softirqs last disabled at (19389): [<ffffffff8104c00c>] irq_exit+0x49/0x99
[   45.471156] CPU: 0 PID: 324 Comm: systemd-sleep Not tainted 4.10.0-next-20170228-dbg-00004-gcdefdec38154-dirty #1091
[   45.471200] Call Trace:
[   45.471207]  dump_stack+0x68/0x92
[   45.471215]  ? __synchronize_hardirq+0x31/0x54
[   45.471225]  ___might_sleep+0x206/0x21d
[   45.471233]  __might_sleep+0x77/0x7e
[   45.471241]  synchronize_irq+0x38/0x8b
[   45.471248]  ? __irq_put_desc_unlock+0x1e/0x3e
[   45.471258]  ? __disable_irq_nosync+0x44/0x4a
[   45.471267]  disable_irq+0x1c/0x1f
[   45.471275]  hpet_msi_resume+0x4e/0xa9
[   45.471283]  clockevents_tick_resume+0x14/0x19
[   45.471292]  tick_resume_local+0x32/0x53
[   45.471300]  tick_resume+0x13/0x15
[   45.471308]  timekeeping_resume+0x19d/0x1b3
[   45.471318]  syscore_resume+0x134/0x2ee
[   45.471326]  ? syscore_resume+0x134/0x2ee
[   45.471335]  suspend_devices_and_enter+0x62f/0xb14
[   45.471345]  pm_suspend+0x5f2/0x68a
[   45.471352]  state_store+0xa7/0xce
[   45.471360]  kobj_attr_store+0x10/0x1a
[   45.471367]  sysfs_kf_write+0x47/0x52
[   45.471375]  kernfs_fop_write+0x11c/0x169
[   45.471383]  __vfs_write+0x26/0xf5
[   45.471392]  ? rcu_read_lock_sched_held+0x64/0x6c
[   45.471401]  ? rcu_sync_lockdep_assert+0x2c/0x52
[   45.471411]  ? __sb_start_write+0x19b/0x1c6
[   45.471420]  vfs_write+0x100/0x189
[   45.471427]  SyS_write+0x51/0x8e
[   45.471434]  entry_SYSCALL_64_fastpath+0x18/0xad
[   45.471443] RIP: 0033:0x7f1ddb930530
[   45.471450] RSP: 002b:00007ffd9a3521a8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[   45.471467] RAX: ffffffffffffffda RBX: 00007f1ddbbedb38 RCX: 00007f1ddb930530
[   45.471480] RDX: 0000000000000004 RSI: 0000564c49e34390 RDI: 0000000000000004
[   45.471494] RBP: 000000000000270f R08: 0000564c49e34240 R09: 00007f1ddc226180
[   45.471508] R10: 0000000000001010 R11: 0000000000000246 R12: 00007f1ddbbedb38
[   45.471522] R13: 0000000000001010 R14: 0000564c49e34380 R15: 00007f1ddbbedae0


kernel/irq/manage.c:110 is

	wait_event(desc->wait_for_threads, !atomic_read(&desc->threads_active));

	-ss