linux.git/drivers/mailbox/pcc.c, branch v4.13

mailbox: pcc: Fix crash when request PCC channel 0

2017-07-26T00:11:47+00:00

When PCCT is not available, kernel crashes as below when requests PCC
channel 0. This patch fixes this issue.

[    0.920454] PCCT header not found.
...
[    8.031309] Unable to handle kernel NULL pointer dereference at virtual address 00000010
[    8.031310] [0000000000000010] user address but active_mm is swapper
[    8.031312] Internal error: Oops: 96000004 [#1] PREEMPT SMP
[    8.031313] Modules linked in:
[    8.031316] CPU: 31 PID: 1 Comm: swapper/0 Tainted: G        W       4.13.0-rc1 #18
[    8.031317] Hardware name: AppliedMicro(R) 07/20/2017
[    8.031318] task: ffff809ef3b08000 task.stack: ffff809ef3b10000
[    8.031322] PC is at pcc_mbox_request_channel+0x8c/0x160
[    8.031325] LR is at xgene_slimpro_i2c_probe+0x1c0/0x378
[    8.031326] pc : [] lr : [] pstate: 00000045
[    8.031327] sp : ffff809ef3b13bd0
[    8.031327] x29: ffff809ef3b13bd0 x28: ffff000008ed90a0
[    8.031329] x27: ffff000009091000 x26: ffff000008e50470
[    8.031330] x25: ffff000008ed9100 x24: ffff809eefd9ac30
[    8.031332] x23: 0000000000000000 x22: ffff0000090e3e10
[    8.031333] x21: ffff0000090e3000 x20: 0000000000000000
[    8.031335] x19: 0000000000000000 x18: 0000000000087ffc
[    8.031336] x17: 2fe48d76a78303f0 x16: 0000000000087ffc
[    8.031337] x15: ffff000000000000 x14: 0000000000000000
[    8.031339] x13: 0000000000000000 x12: 0000000000000018
[    8.031340] x11: 0000000000000018 x10: 0101010101010101
[    8.031342] x9 : 0000000000000000 x8 : 7f7f7f7f7f7f7f7f
[    8.031343] x7 : fefefefeff6b646d x6 : 0000008080808080
[    8.031345] x5 : 0000000000000000 x4 : 0000000000000001
[    8.031346] x3 : 0000000000000000 x2 : ffff000008819b64
[    8.031348] x1 : 0000000000000000 x0 : 0000000000000000
...
[    8.031393] Call trace:
[    8.031394] Exception stack(0xffff809ef3b13a00 to 0xffff809ef3b13b30)
[    8.031395] 3a00: 0000000000000000 0001000000000000 ffff809ef3b13bd0 ffff000008899450
[    8.031397] 3a20: ffff809f7e1f9a10 ffff000008f60be0 0000000000000001 ffff809ef3b13b7c
[    8.031398] 3a40: ffff809f7e1f9a10 0000000000000000 ffff000009091000 0000000000000003
[    8.031399] 3a60: ffff000009091000 0000000000000003 ffff809ef3b13a80 ffff0000084e0794
[    8.031400] 3a80: ffff809ef3b13a90 ffff00000850bb64 ffff809ef3b13ad0 ffff00000850bf34
[    8.031402] 3aa0: 0000000000000000 0000000000000000 ffff000008819b64 0000000000000000
[    8.031403] 3ac0: 0000000000000001 0000000000000000 0000008080808080 fefefefeff6b646d
[    8.031404] 3ae0: 7f7f7f7f7f7f7f7f 0000000000000000 0101010101010101 0000000000000018
[    8.031405] 3b00: 0000000000000018 0000000000000000 0000000000000000 ffff000000000000
[    8.031406] 3b20: 0000000000087ffc 2fe48d76a78303f0
[    8.031409] [] pcc_mbox_request_channel+0x8c/0x160
[    8.031410] [] xgene_slimpro_i2c_probe+0x1c0/0x378
[    8.031413] [] platform_drv_probe+0x50/0xbc
[    8.031414] [] driver_probe_device+0x21c/0x2d0
[    8.031416] [] __driver_attach+0xac/0xb0
[    8.031417] [] bus_for_each_dev+0x58/0x98
[    8.031418] [] driver_attach+0x20/0x28
[    8.031419] [] bus_add_driver+0x1c8/0x22c
[    8.031421] [] driver_register+0x60/0xf4
[    8.031422] [] __platform_driver_register+0x4c/0x54
[    8.031425] [] xgene_slimpro_i2c_driver_init+0x18/0x20
[    8.031426] [] do_one_initcall+0x38/0x124
[    8.031429] [] kernel_init_freeable+0x190/0x22c
[    8.031431] [] kernel_init+0x10/0xfc
[    8.031432] [] ret_from_fork+0x10/0x50
[    8.031434] Code: cb030e63 8b030013 b140067f 54fffda8 (f9400a61)
[    8.031448] ---[ end trace 14eb48a4e1e1f9fb ]---

Signed-off-by: Hoan Tran 
Acked-by: Prashanth Prakash 
Signed-off-by: Rafael J. Wysocki

ACPICA: Add support for new PCCT subtables

2017-06-12T12:58:39+00:00

ACPICA commit e7b817e3c405a4fb9ae9ee7ae4992b8c1f20d284

Extended PCC Subspaces (types 3 and 4)

Link: https://github.com/acpica/acpica/commit/e7b817e3
Signed-off-by: David E. Box 
Signed-off-by: Bob Moore 
Signed-off-by: Lv Zheng 
Signed-off-by: Rafael J. Wysocki

Merge branches 'acpica' and 'acpi-scan'

2016-12-22T13:34:24+00:00

* acpica:
  ACPI / osl: Remove deprecated acpi_get_table_with_size()/early_acpi_os_unmap_memory()
  ACPI / osl: Remove acpi_get_table_with_size()/early_acpi_os_unmap_memory() users
  ACPICA: Tables: Allow FADT to be customized with virtual address
  ACPICA: Tables: Back port acpi_get_table_with_size() and early_acpi_os_unmap_memory() from Linux kernel

* acpi-scan:
  ACPI: do not warn if _BQC does not exist

ACPI / osl: Remove acpi_get_table_with_size()/early_acpi_os_unmap_memory() users

2016-12-21T01:36:38+00:00

This patch removes the users of the deprectated APIs:
 acpi_get_table_with_size()
 early_acpi_os_unmap_memory()
The following APIs should be used instead of:
 acpi_get_table()
 acpi_put_table()

The deprecated APIs are invented to be a replacement of acpi_get_table()
during the early stage so that the early mapped pointer will not be stored
in ACPICA core and thus the late stage acpi_get_table() won't return a
wrong pointer. The mapping size is returned just because it is required by
early_acpi_os_unmap_memory() to unmap the pointer during early stage.

But as the mapping size equals to the acpi_table_header.length
(see acpi_tb_init_table_descriptor() and acpi_tb_validate_table()), when
such a convenient result is returned, driver code will start to use it
instead of accessing acpi_table_header to obtain the length.

Thus this patch cleans up the drivers by replacing returned table size with
acpi_table_header.length, and should be a no-op.

Reported-by: Dan Williams 
Signed-off-by: Lv Zheng 
Signed-off-by: Rafael J. Wysocki

mailbox: PCC: Fix lockdep warning when request PCC channel

2016-11-14T21:07:38+00:00

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
 [] lockdep_trace_alloc+0xe0/0xf0
 [] __kmalloc_track_caller+0x50/0x250
 [] devres_alloc_node+0x28/0x60
 [] devm_request_threaded_irq+0x50/0xe0
 [] pcc_mbox_request_channel+0x110/0x170
 [] acpi_cppc_processor_probe+0x264/0x414
 [] __acpi_processor_start+0x28/0xa0
 [] acpi_processor_start+0x44/0x54
 [] driver_probe_device+0x1fc/0x2b0
 [] __driver_attach+0xb4/0xc0
 [] bus_for_each_dev+0x5c/0xa0
 [] driver_attach+0x20/0x30
 [] bus_add_driver+0x110/0x230
 [] driver_register+0x60/0x100
 [] acpi_processor_driver_init+0x2c/0xb0
 [] do_one_initcall+0x38/0x130
 [] kernel_init_freeable+0x210/0x2b4
 [] kernel_init+0x10/0x110
 [] 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 
Reviewed-by: Prashanth Prakash 
Signed-off-by: Rafael J. Wysocki

mailbox: pcc: Support HW-Reduced Communication Subspace type 2

2016-08-30T22:37:20+00:00

ACPI 6.1 has a PCC HW-Reduced Communication Subspace type 2 intended for
use on HW-Reduce ACPI Platform, which requires read-modify-write sequence
to acknowledge doorbell interrupt. This patch provides the implementation
for the Communication Subspace Type 2.

Signed-off-by: Hoan Tran 
Reviewed-by: Prashanth Prakash 
Signed-off-by: Rafael J. Wysocki

Merge branches 'pm-cpufreq', 'pm-cpuidle' and 'acpi-cppc'

2016-04-08T19:46:05+00:00

* pm-cpufreq:
  cpufreq: dt: Drop stale comment
  cpufreq: intel_pstate: Documenation for structures
  cpufreq: intel_pstate: fix inconsistency in setting policy limits
  intel_pstate: Avoid extra invocation of intel_pstate_sample()
  intel_pstate: Do not set utilization update hook too early

* pm-cpuidle:
  intel_idle: Add KBL support
  intel_idle: Add SKX support
  intel_idle: Clean up all registered devices on exit.
  intel_idle: Propagate hot plug errors.
  intel_idle: Don't overreact to a cpuidle registration failure.
  intel_idle: Setup the timer broadcast only on successful driver load.
  intel_idle: Avoid a double free of the per-CPU data.
  intel_idle: Fix dangling registration on error path.
  intel_idle: Fix deallocation order on the driver exit path.
  intel_idle: Remove redundant initialization calls.
  intel_idle: Fix a helper function's return value.
  intel_idle: remove useless return from void function.

* acpi-cppc:
  mailbox: pcc: Don't access an unmapped memory address space

mailbox: pcc: Don't access an unmapped memory address space

2016-04-06T23:25:28+00:00

The acpi_pcc_probe() may end up accessing memory outside of the PCCT
table space causing the kernel panic(). Increment the pcct_entry
pointer after parsing 'HW-reduced Communications Subspace' to fix
the problem. This change also enables the parsing of subtable at
index 0.

Signed-off-by: Shanker Donthineni 
Acked-by: Ashwin Chaugule 
Signed-off-by: Rafael J. Wysocki

Merge branches 'acpi-processor' and 'acpi-cppc'

2016-03-14T13:20:33+00:00

* acpi-processor:
  ACPI / sleep: move acpi_processor_sleep to sleep.c
  ACPI / processor : add support for ACPI0010 processor container
  ACPI / processor_idle: replace PREFIX with pr_fmt

* acpi-cppc:
  ACPI / CPPC: use MRTT/MPAR to decide if/when a req can be sent
  ACPI / CPPC: replace writeX/readX to PCC with relaxed version
  mailbox: pcc: optimized pcc_send_data
  ACPI / CPPC: optimized cpc_read and cpc_write
  ACPI / CPPC: Optimize PCC Read Write operations

mailbox: pcc: optimized pcc_send_data

2016-03-09T22:35:29+00:00

pcc_send_data() can be invoked during the execution of performance
critical code as in cppc_cpufreq driver. With acpi_* APIs, the
doorbell register accessed in pcc_send_data() if present in system
memory will be searched (in cached virt to phys addr mapping),
mapped, read/written and then unmapped. These operations take
significant amount of time.

This patch maps the performance critical doorbell register
during init and then reads/writes to it directly using the
mapped virtual address. This patch + similar changes to CPPC
acpi driver reduce the time per freq. transition from around
200us to about 20us for the CPPC cpufreq driver

Signed-off-by: Prashanth Prakash 
Acked-by: Ashwin Chaugule 
Signed-off-by: Rafael J. Wysocki