lockdep splat during suspend

lockdep splat during suspend

[Borislav Petkov]

Hi dudes,

this is what I'm seeing on 4.9-rc5+tip/master during suspend. (And yes,
that printk thing is nuts):

[   87.471336] hib.sh (2591): drop_caches: 3
[   87.779292] PM: Syncing filesystems ... 
[   87.788772] done.
[   87.788853] Freezing user space processes ... (elapsed 0.001 seconds) done.
[   87.791447] PM: Preallocating image memory... done (allocated 105192 pages)
[   88.074812] PM: Allocated 420768 kbytes in 0.28 seconds (1502.74 MB/s)
[   88.074852] Freezing remaining freezable tasks ... (elapsed 0.001 seconds) done.
[   88.080163] ACPI : EC: event blocked
[   88.334267] PM: freeze of devices complete after 256.808 msecs
[   88.353487] PM: late freeze of devices complete after 19.202 msecs
[   88.354728] ACPI : EC: interrupt blocked
[   88.355761] PM: noirq freeze of devices complete after 2.257 msecs
[   88.355777] Disabling non-boot CPUs ...
[   88.378463] Broke affinity for irq 16
[   88.378468] Broke affinity for irq 18
[   88.378471] Broke affinity for irq 19
[   88.379835] smpboot: CPU 1 is now offline

[   88.402872] ======================================================
[   88.402873] [ INFO: possible circular locking dependency detected ]
[   88.402874] 4.9.0-rc5+ #1 Not tainted
[   88.402875] -------------------------------------------------------
[   88.402875] hib.sh/2591 is trying to acquire lock:
[   88.402876]  (
[   88.402877] console_lock
[   88.402877] ){+.+.+.}
[   88.402883] , at: 
[   88.402883] [<ffffffff810d26de>] console_cpu_notify+0xe/0x20
[   88.402883] 
               but task is already holding lock:
[   88.402884]  (
[   88.402884] cpu_hotplug.lock
[   88.402884] ){+.+.+.}
[   88.402886] , at: 
[   88.402886] [<ffffffff8106bd52>] cpu_hotplug_begin+0x72/0xc0
[   88.402887] 
               which lock already depends on the new lock.

[   88.402887] 
               the existing dependency chain (in reverse order) is:
[   88.402888] 
               -> #3
[   88.402888]  (
[   88.402888] cpu_hotplug.lock
[   88.402889] ){+.+.+.}
[   88.402889] :
[   88.402892]        
[   88.402892] [<ffffffff810c11e2>] lock_acquire+0xb2/0x200
[   88.402895]        
[   88.402895] [<ffffffff817582c4>] mutex_lock_nested+0x54/0x600
[   88.402898]        
[   88.402898] [<ffffffff8106a063>] get_online_cpus+0x53/0x70
[   88.402901]        
[   88.402901] [<ffffffff810def14>] _rcu_barrier+0xa4/0x160
[   88.402902]        
[   88.402902] [<ffffffff810df025>] rcu_barrier+0x15/0x20
[   88.402905]        
[   88.402905] [<ffffffff8164d024>] netdev_run_todo+0x64/0x300
[   88.402907]        
[   88.402907] [<ffffffff8165a9de>] rtnl_unlock+0xe/0x10
[   88.402909]        
[   88.402909] [<ffffffffa02f6bff>] 0xffffffffa02f6bff
[   88.402910]        
[   88.402910] [<ffffffffa04e83f8>] 0xffffffffa04e83f8
[   88.402911]        
[   88.402911] [<ffffffffa04e07cb>] 0xffffffffa04e07cb
[   88.402914]        
[   88.402914] [<ffffffffa0264d29>] msr_exit+0x962/0xc39 [msr]
[   88.402914]        
[   88.402914] [<ffffffffa04f68cf>] 0xffffffffa04f68cf
[   88.402917]        
[   88.402917] [<ffffffff81103985>] SyS_delete_module+0x185/0x250
[   88.402918]        
[   88.402919] [<ffffffff8175e1ee>] entry_SYSCALL_64_fastpath+0x1c/0xb1
[   88.402919] 
               -> #2
[   88.402920]  (
[   88.402920] rcu_preempt_state.barrier_mutex
[   88.402920] ){+.+...}
[   88.402920] :
[   88.402922]        
[   88.402922] [<ffffffff810c11e2>] lock_acquire+0xb2/0x200
[   88.402924]        
[   88.402924] [<ffffffff817582c4>] mutex_lock_nested+0x54/0x600
[   88.402925]        
[   88.402925] [<ffffffff810deea6>] _rcu_barrier+0x36/0x160
[   88.402927]        
[   88.402927] [<ffffffff810df025>] rcu_barrier+0x15/0x20
[   88.402930]        
[   88.402930] [<ffffffff814b4183>] i915_gem_shrink_all+0x23/0x30
[   88.402932]        
[   88.402932] [<ffffffff814b0c4a>] i915_gem_freeze+0x2a/0x50
[   88.402935]        
[   88.402935] [<ffffffff8146ab32>] i915_pm_freeze+0x22/0x30
[   88.402938]        
[   88.402938] [<ffffffff8137c29e>] pci_pm_freeze+0x5e/0xe0
[   88.402940]        
[   88.402940] [<ffffffff81546fae>] dpm_run_callback+0x4e/0x2d0
[   88.402942]        
[   88.402942] [<ffffffff81547c56>] __device_suspend+0xf6/0x2c0
[   88.402943]        
[   88.402943] [<ffffffff81547e3f>] async_suspend+0x1f/0xa0
[   88.402945]        
[   88.402945] [<ffffffff81092237>] async_run_entry_fn+0x37/0xe0
[   88.402947]        
[   88.402947] [<ffffffff81088018>] process_one_work+0x1e8/0x730
[   88.402948]        
[   88.402948] [<ffffffff810885a8>] worker_thread+0x48/0x4e0
[   88.402950]        
[   88.402950] [<ffffffff8108ed04>] kthread+0xf4/0x110
[   88.402952]        
[   88.402952] [<ffffffff8175e477>] ret_from_fork+0x27/0x40
[   88.402952] 
               -> #1
[   88.402952]  (
[   88.402953] &dev->struct_mutex
[   88.402953] ){+.+.+.}
[   88.402953] :
[   88.402955]        
[   88.402955] [<ffffffff810c11e2>] lock_acquire+0xb2/0x200
[   88.402957]        
[   88.402957] [<ffffffff817582c4>] mutex_lock_nested+0x54/0x600
[   88.402960]        
[   88.402960] [<ffffffff8150ba85>] intel_fbdev_pan_display+0x35/0x70
[   88.402962]        
[   88.402962] [<ffffffff813a130b>] fb_pan_display+0xdb/0x170
[   88.402964]        
[   88.402964] [<ffffffff8139b870>] bit_update_start+0x20/0x50
[   88.402965]        
[   88.402965] [<ffffffff8139872f>] fbcon_switch+0x3bf/0x610
[   88.402968]        
[   88.402968] [<ffffffff814040a1>] redraw_screen+0x151/0x220
[   88.402969]        
[   88.402969] [<ffffffff81397eb6>] fbcon_do_set_font+0x246/0x3e0
[   88.402970]        
[   88.402970] [<ffffffff81398324>] fbcon_set_font+0x1f4/0x240
[   88.402972]        
[   88.402973] [<ffffffff81409209>] con_font_op+0x3e9/0x490
[   88.402974]        
[   88.402974] [<ffffffff813fb21d>] vt_ioctl+0x42d/0x13d0
[   88.402976]        
[   88.402976] [<ffffffff813eed8e>] tty_ioctl+0x3ae/0xee0
[   88.402979]        
[   88.402979] [<ffffffff812040a3>] do_vfs_ioctl+0x93/0x6a0
[   88.402980]        
[   88.402980] [<ffffffff812046f1>] SyS_ioctl+0x41/0x70
[   88.402982]        
[   88.402982] [<ffffffff8175e1ee>] entry_SYSCALL_64_fastpath+0x1c/0xb1
[   88.402982] 
               -> #0
[   88.402983]  (
[   88.402983] console_lock
[   88.402983] ){+.+.+.}
[   88.402983] :
[   88.402985]        
[   88.402985] [<ffffffff810c0b5d>] __lock_acquire+0x11cd/0x1440
[   88.402987]        
[   88.402987] [<ffffffff810c11e2>] lock_acquire+0xb2/0x200
[   88.402989]        
[   88.402989] [<ffffffff810cf77c>] console_lock+0x3c/0x60
[   88.402991]        
[   88.402991] [<ffffffff810d26de>] console_cpu_notify+0xe/0x20
[   88.402992]        
[   88.402992] [<ffffffff8106aa59>] cpuhp_invoke_callback+0xa9/0x8b0
[   88.402993]        
[   88.402993] [<ffffffff8106b472>] cpuhp_down_callbacks+0x42/0x80
[   88.402995]        
[   88.402995] [<ffffffff817532a9>] _cpu_down+0xa9/0x110
[   88.402997]        
[   88.402997] [<ffffffff8106c466>] freeze_secondary_cpus+0xb6/0x3e0
[   88.402998]        
[   88.402998] [<ffffffff810c8731>] hibernation_snapshot+0x201/0x360
[   88.403000]        
[   88.403000] [<ffffffff810c8f82>] hibernate+0x152/0x200
[   88.403001]        
[   88.403001] [<ffffffff810c5e86>] state_store+0xd6/0xe0
[   88.403003]        
[   88.403003] [<ffffffff8133f71f>] kobj_attr_store+0xf/0x20
[   88.403006]        
[   88.403006] [<ffffffff8126f645>] sysfs_kf_write+0x45/0x60
[   88.403008]        
[   88.403008] [<ffffffff8126e8f5>] kernfs_fop_write+0x135/0x1c0
[   88.403010]        
[   88.403010] [<ffffffff811ee1a8>] __vfs_write+0x28/0x120
[   88.403011]        
[   88.403011] [<ffffffff811ef2b8>] vfs_write+0xb8/0x1b0
[   88.403013]        
[   88.403013] [<ffffffff811f0609>] SyS_write+0x49/0xa0
[   88.403014]        
[   88.403014] [<ffffffff8175e1ee>] entry_SYSCALL_64_fastpath+0x1c/0xb1
[   88.403014] 
               other info that might help us debug this:

[   88.403015] Chain exists of:
                 
[   88.403015] console_lock
[   88.403016]  --> 
[   88.403016] rcu_preempt_state.barrier_mutex
[   88.403017]  --> 
[   88.403017] cpu_hotplug.lock
[   88.403017] 

[   88.403017]  Possible unsafe locking scenario:

[   88.403018]        CPU0                    CPU1
[   88.403018]        ----                    ----
[   88.403018]   lock(
[   88.403019] cpu_hotplug.lock
[   88.403019] );
[   88.403020]                                lock(
[   88.403020] rcu_preempt_state.barrier_mutex
[   88.403020] );
[   88.403021]                                lock(
[   88.403021] cpu_hotplug.lock
[   88.403021] );
[   88.403021]   lock(
[   88.403022] console_lock
[   88.403022] );
[   88.403022] 
                *** DEADLOCK ***

[   88.403023] 8 locks held by hib.sh/2591:
[   88.403023]  #0: 
[   88.403024]  (
[   88.403024] sb_writers
[   88.403025] #5
[   88.403025] ){.+.+.+}
[   88.403026] , at: 
[   88.403026] [<ffffffff811ef394>] vfs_write+0x194/0x1b0
[   88.403026]  #1: 
[   88.403027]  (
[   88.403027] &of->mutex
[   88.403027] ){+.+.+.}
[   88.403029] , at: 
[   88.403029] [<ffffffff8126e8c1>] kernfs_fop_write+0x101/0x1c0
[   88.403029]  #2: 
[   88.403030]  (
[   88.403030] s_active
[   88.403031] #197
[   88.403031] ){.+.+.+}
[   88.403032] , at: 
[   88.403032] [<ffffffff8126e8c9>] kernfs_fop_write+0x109/0x1c0
[   88.403033]  #3: 
[   88.403033]  (
[   88.403033] pm_mutex
[   88.403034] ){+.+.+.}
[   88.403035] , at: 
[   88.403035] [<ffffffff810c8e71>] hibernate+0x41/0x200
[   88.403036]  #4: 
[   88.403036]  (
[   88.403036] device_hotplug_lock
[   88.403037] ){+.+.+.}
[   88.403038] , at: 
[   88.403038] [<ffffffff81536a77>] lock_device_hotplug+0x17/0x20
[   88.403039]  #5: 
[   88.403039]  (
[   88.403040] cpu_add_remove_lock
[   88.403040] ){+.+.+.}
[   88.403041] , at: 
[   88.403041] [<ffffffff8106c3d7>] freeze_secondary_cpus+0x27/0x3e0
[   88.403042]  #6: 
[   88.403042]  (
[   88.403042] cpu_hotplug.dep_map
[   88.403043] ){++++++}
[   88.403044] , at: 
[   88.403044] [<ffffffff8106bce5>] cpu_hotplug_begin+0x5/0xc0
[   88.403044]  #7: 
[   88.403044]  (
[   88.403045] cpu_hotplug.lock
[   88.403045] ){+.+.+.}
[   88.403046] , at: 
[   88.403046] [<ffffffff8106bd52>] cpu_hotplug_begin+0x72/0xc0
[   88.403046] 
               stack backtrace:
[   88.403048] CPU: 0 PID: 2591 Comm: hib.sh Not tainted 4.9.0-rc5+ #1
[   88.403049] Hardware name: LENOVO 2320CTO/2320CTO, BIOS G2ET86WW (2.06 ) 11/13/2012
[   88.403049] Call Trace:
[   88.403053]  dump_stack+0x67/0x92
[   88.403055]  print_circular_bug+0x1e3/0x250
[   88.403056]  __lock_acquire+0x11cd/0x1440
[   88.403059]  ? debug_lockdep_rcu_enabled+0x1d/0x20
[   88.403061]  lock_acquire+0xb2/0x200
[   88.403063]  ? console_cpu_notify+0xe/0x20
[   88.403065]  ? register_console+0x380/0x380
[   88.403066]  console_lock+0x3c/0x60
[   88.403068]  ? console_cpu_notify+0xe/0x20
[   88.403069]  console_cpu_notify+0xe/0x20
[   88.403070]  cpuhp_invoke_callback+0xa9/0x8b0
[   88.403072]  ? __flow_cache_shrink+0x140/0x140
[   88.403073]  cpuhp_down_callbacks+0x42/0x80
[   88.403075]  _cpu_down+0xa9/0x110
[   88.403076]  freeze_secondary_cpus+0xb6/0x3e0
[   88.403077]  hibernation_snapshot+0x201/0x360
[   88.403078]  hibernate+0x152/0x200
[   88.403080]  state_store+0xd6/0xe0
[   88.403081]  kobj_attr_store+0xf/0x20
[   88.403082]  sysfs_kf_write+0x45/0x60
[   88.403084]  kernfs_fop_write+0x135/0x1c0
[   88.403086]  __vfs_write+0x28/0x120
[   88.403087]  ? preempt_count_sub+0xa1/0x110
[   88.403089]  ? __sb_start_write+0xea/0x200
[   88.403090]  ? vfs_write+0x194/0x1b0
[   88.403091]  vfs_write+0xb8/0x1b0
[   88.403092]  ? trace_hardirqs_on_caller+0x12d/0x1b0
[   88.403093]  SyS_write+0x49/0xa0
[   88.403095]  entry_SYSCALL_64_fastpath+0x1c/0xb1
[   88.403096] RIP: 0033:0x7f37f6dc51c0
[   88.403097] RSP: 002b:00007ffe98060948 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[   88.403098] RAX: ffffffffffffffda RBX: 0000000000000046 RCX: 00007f37f6dc51c0
[   88.403098] RDX: 0000000000000005 RSI: 00000000013f5408 RDI: 0000000000000001
[   88.403099] RBP: 0000000000000004 R08: 00007f37f7088780 R09: 00007f37f78d8700
[   88.403100] R10: 0000000000000004 R11: 0000000000000246 R12: 00000000004be368
[   88.403100] R13: 0000000000002002 R14: 00000000013fe400 R15: 0000000000000070
[   88.417290] Broke affinity for irq 16
[   88.417293] Broke affinity for irq 18
[   88.417295] Broke affinity for irq 19
[   88.418315] smpboot: CPU 2 is now offline
[   88.461454] Broke affinity for irq 1
[   88.461460] Broke affinity for irq 8
[   88.461464] Broke affinity for irq 9
[   88.461468] Broke affinity for irq 12
[   88.461473] Broke affinity for irq 16
[   88.461477] Broke affinity for irq 18
[   88.461480] Broke affinity for irq 19
[   88.461485] Broke affinity for irq 23
[   88.461489] Broke affinity for irq 26
[   88.461493] Broke affinity for irq 27
[   88.462528] smpboot: CPU 3 is now offline
[   88.473978] PM: Creating hibernation image:
[   88.733315] PM: Need to copy 104116 pages
[   88.475760] Suspended for 19.000 seconds
[   88.475851] Enabling non-boot CPUs ...
[   88.487470] x86: Booting SMP configuration:
[   88.487479] smpboot: Booting Node 0 Processor 1 APIC 0x1
[   88.492153]  cache: parent cpu1 should not be sleeping
[   88.492571] CPU1 is up
[   88.507516] smpboot: Booting Node 0 Processor 2 APIC 0x2
[   88.511244]  cache: parent cpu2 should not be sleeping
[   88.511626] CPU2 is up
[   88.527572] smpboot: Booting Node 0 Processor 3 APIC 0x3
[   88.531225]  cache: parent cpu3 should not be sleeping
[   88.531617] CPU3 is up
[   88.537517] ACPI : EC: interrupt unblocked
[   88.540582] sdhci-pci 0000:02:00.0: MMC controller base frequency changed to 50Mhz.
[   88.583150] PM: noirq restore of devices complete after 45.897 msecs
[   88.583639] PM: early restore of devices complete after 0.459 msecs
[   88.625421] usb usb1: root hub lost power or was reset
[   88.625432] usb usb2: root hub lost power or was reset
[   88.625608] usb usb3: root hub lost power or was reset
[   88.625883] usb usb4: root hub lost power or was reset
[   88.626879] ACPI : button: The lid device is not compliant to SW_LID.
[   88.626910] ACPI : EC: event unblocked
[   88.627295] rtc_cmos 00:02: System wakeup disabled by ACPI
[   88.629508] ehci-pci 0000:00:1a.0: cache line size of 64 is not supported
[   88.629783] ehci-pci 0000:00:1d.0: cache line size of 64 is not supported
[   88.636192] sd 2:0:0:0: [sdb] Starting disk
[   88.636228] sd 0:0:0:0: [sda] Starting disk
[   88.963083] usb 4-1: reset high-speed USB device number 2 using ehci-pci
[   88.963085] usb 3-1: reset high-speed USB device number 2 using ehci-pci
[   88.974052] ata5: SATA link down (SStatus 0 SControl 300)
[   88.974093] ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[   88.974136] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[   88.974583] ata3.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[   88.974596] ata3.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[   88.975119] ata1.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[   88.975132] ata1.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[   88.975549] ata3.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[   88.975562] ata3.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[   88.975887] ata3.00: configured for UDMA/100
[   88.977083] ata1.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[   88.977112] ata1.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[   88.977937] ata1.00: configured for UDMA/133
[   89.671711] PM: restore of devices complete after 1046.321 msecs
[   89.673226] Restarting tasks ... done.
[   89.673738] usb 1-1: USB disconnect, device number 2
[   89.823083] usb 1-1: new low-speed USB device number 3 using xhci_hcd
[   89.969125] usb 1-1: New USB device found, idVendor=046d, idProduct=c045
[   89.969135] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[   89.969141] usb 1-1: Product: USB-PS/2 Optical Mouse
[   89.969146] usb 1-1: Manufacturer: Logitech
[   89.974053] input: Logitech USB-PS/2 Optical Mouse as /devices/pci0000:00/0000:00:14.0/usb1/1-1/1-1:1.0/0003:046D:C045.0002/input/input20
[   90.031487] hid-generic 0003:046D:C045.0002: input,hidraw0: USB HID v1.10 Mouse [Logitech USB-PS/2 Optical Mouse] on usb-0000:00:14.0-1/input0
[   90.340587] psmouse serio1: synaptics: queried max coordinates: x [..5768], y [..5062]
[   90.371618] psmouse serio1: synaptics: queried min coordinates: x [1174..], y [790..]
[   90.421219] e1000e: eth0 NIC Link is Up 100 Mbps Full Duplex, Flow Control: Rx/Tx

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

Re: lockdep splat during suspend

[Sebastian Andrzej Siewior]

On 2016-11-17 13:43:58 [+0100], Borislav Petkov wrote:
> Hi dudes,
Hi Boris,

> this is what I'm seeing on 4.9-rc5+tip/master during suspend. (And yes,
> that printk thing is nuts):

assuming this happens always, could you please try to revert
  90b14889d2f9 ("kernel/printk: Convert to hotplug state machine")
and see if it still there? I am curious if this is new or old splat.

Sebastian

Re: lockdep splat during suspend

[Thomas Gleixner]

On Thu, 17 Nov 2016, Sebastian Andrzej Siewior wrote:

> On 2016-11-17 13:43:58 [+0100], Borislav Petkov wrote:
> > Hi dudes,
> Hi Boris,
> 
> > this is what I'm seeing on 4.9-rc5+tip/master during suspend. (And yes,
> > that printk thing is nuts):
> 
> assuming this happens always, could you please try to revert
>   90b14889d2f9 ("kernel/printk: Convert to hotplug state machine")
> and see if it still there? I am curious if this is new or old splat.

It's a new one and that patch simply missed the fact, that the actions in
the original notifier were all w/o FROZEN, but the conversion dropped
that. Which probably explains the splat because suspend/resume is where the
FROZEN stuff happens.

Boris, does the patch below fix the issue for you ?

Thanks,

	tglx

8<-----------------------

--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -2194,8 +2194,10 @@ void resume_console(void)
  */
 static int console_cpu_notify(unsigned int cpu)
 {
-	console_lock();
-	console_unlock();
+	if (!cpuhp_tasks_frozen) {
+		console_lock();
+		console_unlock();
+	}
 	return 0;
 }
 

Re: lockdep splat during suspend

[Borislav Petkov]

On Thu, Nov 17, 2016 at 05:31:55PM +0100, Thomas Gleixner wrote:
> --- a/kernel/printk/printk.c
> +++ b/kernel/printk/printk.c
> @@ -2194,8 +2194,10 @@ void resume_console(void)
>   */
>  static int console_cpu_notify(unsigned int cpu)
>  {
> -	console_lock();
> -	console_unlock();
> +	if (!cpuhp_tasks_frozen) {
> +		console_lock();
> +		console_unlock();
> +	}
>  	return 0;
>  }

Reported-and-tested-by: Borislav Petkov <bp@suse.de>

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

[tip:smp/hotplug] kernel/printk: Block cpuhotplug callback when tasks are frozen

[tip-bot for Thomas Gleixner]

Commit-ID:  f97960fbddd98e1b7bfba469b8510b49a62b4bc5
Gitweb:     http://git.kernel.org/tip/f97960fbddd98e1b7bfba469b8510b49a62b4bc5
Author:     Thomas Gleixner <tglx@linutronix.de>
AuthorDate: Thu, 17 Nov 2016 17:31:55 +0100
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Thu, 17 Nov 2016 19:44:58 +0100

kernel/printk: Block cpuhotplug callback when tasks are frozen

The recent conversion of the console hotplug notifier to the state machine
missed the fact, that the notifier only operated on the non frozen
transitions. As a consequence the console_lock/unlock() pair is also
invoked during suspend, which results in a lockdep warning.

Restore the previous state by making the lock/unlock conditional on
!tasks_frozen.

Fixes: 90b14889d2f9 ("kernel/printk: Convert to hotplug state machine")
Reported-and-tested-by: Borislav Petkov <bp@alien8.de>
Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1611171729320.3645@nanos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

---
 kernel/printk/printk.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c
index 4487ffc..37893da 100644
--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -2194,8 +2194,10 @@ void resume_console(void)
  */
 static int console_cpu_notify(unsigned int cpu)
 {
-	console_lock();
-	console_unlock();
+	if (!cpuhp_tasks_frozen) {
+		console_lock();
+		console_unlock();
+	}
 	return 0;
 }