This patch fixes the lockdep warning below

DEBUG_LOCKS_WARN_ON(irqs_disabled_flags(flags))
------------[ cut here ]------------
WARNING: CPU: 1 PID: 1 at linux-next/kernel/locking/lockdep.c:2876 lockdep_trace_alloc+0xe0/0xf0
 Modules linked in:

 CPU: 1 PID: 1 Comm: swapper/0 Not tainted 4.8.0-11756-g86c5152 #46
...
 Call trace:
 Exception stack(0xffff8007da837890 to 0xffff8007da8379c0)
 7880:                                   ffff8007da834000 0001000000000000
 78a0: ffff8007da837a70 ffff0000081111a0 00000000600000c5 000000000000003d
 78c0: 9374bc6a7f3c7832 0000000000381878 ffff000009db7ab8 000000000000002f
 78e0: ffff00000811aabc ffff000008be2548 ffff8007da837990 ffff00000811adf8
 7900: ffff8007da834000 00000000024080c0 00000000000000c0 ffff000009021000
 7920: 0000000000000000 0000000000000000 ffff000008c8f7c8 ffff8007da579810
 7940: 000000000000002f ffff8007da858000 0000000000000000 0000000000000001
 7960: 0000000000000001 0000000000000000 ffff00000811a468 0000000000000002
 7980: 656c62617369645f 0000000000038187 00000000000000ee ffff8007da837850
 79a0: ffff000009db50c0 ffff000009db569d 0000000000000006 ffff000089db568f
 [<ffff0000081111a0>] lockdep_trace_alloc+0xe0/0xf0
 [<ffff0000081f4950>] __kmalloc_track_caller+0x50/0x250
 [<ffff00000857c088>] devres_alloc_node+0x28/0x60
 [<ffff0000081220e0>] devm_request_threaded_irq+0x50/0xe0
 [<ffff0000087e6220>] pcc_mbox_request_channel+0x110/0x170
 [<ffff0000084b2660>] acpi_cppc_processor_probe+0x264/0x414
 [<ffff0000084ae9f4>] __acpi_processor_start+0x28/0xa0
 [<ffff0000084aeab0>] acpi_processor_start+0x44/0x54
 [<ffff00000857897c>] driver_probe_device+0x1fc/0x2b0
 [<ffff000008578ae4>] __driver_attach+0xb4/0xc0
 [<ffff00000857683c>] bus_for_each_dev+0x5c/0xa0
 [<ffff000008578110>] driver_attach+0x20/0x30
 [<ffff000008577c20>] bus_add_driver+0x110/0x230
 [<ffff000008579320>] driver_register+0x60/0x100
 [<ffff000008d478b8>] acpi_processor_driver_init+0x2c/0xb0
 [<ffff000008083168>] do_one_initcall+0x38/0x130
 [<ffff000008d20d6c>] kernel_init_freeable+0x210/0x2b4
 [<ffff000008945d90>] kernel_init+0x10/0x110
 [<ffff000008082e80>] ret_from_fork+0x10/0x50

It's because the spinlock inside pcc_mbox_request_channel() is
kept too long. This patch releases spinlock before request_irq()
and free_irq() to fix this issue  as spinlock is only needed to
protect the channel data.

Signed-off-by: Hoan Tran <hotran@apm.com>
Reviewed-by: Prashanth Prakash <pprakash@codeaurora.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>