Like other Lenovo models the IdeaPad V510-15IKB does not have an hw
rfkill switch. This results in hard-blocked radios after boot, resulting
in always blocked radios rendering them unusable.

Add the IdeaPad V510-15IKB to the no_hw_rfkill DMI list and allows using
the built-in radios.
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

This adds a debugfs consumer for the exported kernel API
intel_pmc_read_s0ix_residency. This debugfs entry reads S0ix residency
directly from the PMC hardware counters.

TEST:
- echo freeze > /sys/power/state
- Wake the system, read the S0ix residency i.e.
  cat /sys/kernel/debug/telemetry/s0ix_residency_usec
Signed-off-by: NShanth Murthy <shanth.murthy@intel.com>
Signed-off-by: NRajneesh Bhardwaj <rajneesh.bhardwaj@intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

There are bunch of "goto out;" paths where we don't set the error code.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

This fixes an oops found while testing load/unload of the
intel_telemetry_debugfs module. module_init uses register_pm_notifier
for PM callbacks, but unregister_pm_notifier was missing from
module_exit.

 [ 97.481860] BUG: unable to handle kernel paging request at ffffffffa006f010
 [ 97.489742] IP: blocking_notifier_chain_register+0x3a/0xa0
 [ 97.495898] PGD 2e0a067
 [ 97.495899] PUD 2e0b063
 [ 97.498737] PMD 179e29067
 [ 97.501573] PTE 0

 [ 97.508423] Oops: 0000 1 PREEMPT SMP
 [ 97.512724] Modules linked in: intel_telemetry_debugfs intel_rapl gpio_keys dwc3 udc_core intel_telemetry_pltdrv intel_punit_ipc intel_telemetry_core rtc_cmos efivars x86_pkg_temp_thermal iwlwifi snd_hda_codec_hdmi soc_button_array btusb cfg80211 btrtl mei_me hci_uart btbcm mei btintel i915 bluetooth intel_pmc_ipc snd_hda_intel spi_pxa2xx_platform snd_hda_codec dwc3_pci snd_hda_core tpm_tis tpm_tis_core tpm efivarfs
 [ 97.558453] CPU: 0 PID: 889 Comm: modprobe Not tainted 4.11.0-rc6-intel-dev-bkc #1
 [ 97.566950] Hardware name: Intel Corp. Joule DVT3/SDS, BIOS GTPP181A.X64.0143.B30.1701132137 01/13/2017
 [ 97.577518] task: ffff8801793a21c0 task.stack: ffff8801793f0000
 [ 97.584162] RIP: 0010:blocking_notifier_chain_register+0x3a/0xa0
 [ 97.590903] RSP: 0018:ffff8801793f3c58 EFLAGS: 00010286
 [ 97.596802] RAX: ffffffffa006f000 RBX: ffffffff81e3ea20 RCX: 0000000000000000
 [ 97.604812] RDX: ffff880179eaf210 RSI: ffffffffa0131000 RDI: ffffffff81e3ea20
 [ 97.612821] RBP: ffff8801793f3c68 R08: 0000000000000006 R09: 000000000000005c
 [ 97.620847] R10: 0000000000000000 R11: 0000000000000006 R12: ffffffffa0131000
 [ 97.628855] R13: 0000000000000000 R14: ffff880176e35f48 R15: ffff8801793f3ea8
 [ 97.636865] FS: 00007f7eeba07700(0000) GS:ffff88017fc00000(0000) knlGS:0000000000000000
 [ 97.645948] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 [ 97.652423] CR2: ffffffffa006f010 CR3: 00000001775ef000 CR4: 00000000003406f0
 [ 97.660423] Call Trace:
 [ 97.663166] ? 0xffffffffa0031000
 [ 97.666885] register_pm_notifier+0x18/0x20
 [ 97.671581] telemetry_debugfs_init+0x92/0x1000
Signed-off-by: NPriyalee Kushwaha <priyalee.kushwaha@intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

attribute_groups are not supposed to change at runtime. All functions
working with attribute_groups provided by <linux/sysfs.h> work with const
attribute_group.

The content of sony_laptop_input_index does not change and is declared
as a static global array.

Constify spic_attribute_group and sony_laptop_input_index.

File size before:
   text	   data	    bss	    dec	    hex	filename
  31273	   5176	    372	  36821	   8fd5	drivers/platform/x86/sony-laptop.o

File size After adding 'const':
   text	   data	    bss	    dec	    hex	filename
  31337	   5112	    372	  36821	   8fd5	drivers/platform/x86/sony-laptop.o
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
[dvhart: updated commit message, includes fix suggested by Arnd Bergmann]
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Using a dedicated Kconfig option for enabling debugging means the user
may be forced to recompile their kernel in order to gather debugging
information, which is inconvenient.  Replace custom debugging
infrastructure with standard logging functions, taking advantage of
dynamic debug.  Replace a pr_info() call inside an ACPI callback with an
acpi_handle_info() call.

The following mapping was used:

  - FUJLAPTOP_DBG_ERROR -> acpi_handle_err()
  - FUJLAPTOP_DBG_WARN  -> acpi_handle_info() / dev_info()
  - FUJLAPTOP_DBG_INFO  -> acpi_handle_debug()
  - FUJLAPTOP_DBG_TRACE -> acpi_handle_debug() / dev_dbg()

This means that some events which used to only be logged when the user
explicitly requested it will now be logged by default:

  - ACPI method evaluation errors,
  - unknown ACPI notification codes,
  - unknown hotkey scancodes.

The first type of events should happen rarely, if ever at all.  The rest
is interesting from driver development perspective as their presence in
the logs will mean the driver is unaware of certain events, handling of
which should be implemented.
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

acpi_ns_initialize_devices(), which is called during system-wide ACPI
initialization, already detects and calls all _INI methods belonging to
objects present in ACPI tables.  There is no need to call these methods
again every time the module is loaded because they only initialize
status flags and hotkey-related variables; status flags are effectively
constants, hotkey-related variables may be assigned non-zero values
before acpi_fujitsu_laptop_add() is called, but that does not really
matter as we drain the scancodes queued in the firmware's ring buffer
before doing anything else.

Remove sections of code which invoke and check evaluation status of the
_INI methods belonging to the ACPI devices handled by the driver.
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Calling acpi_bus_update_power() for ACPI devices FUJ02B1 and FUJ02E3 is
pointless as they are not power manageable (neither _PS0 nor _PR0 is
defined for any of them), which causes their power state to be inherited
from their parent devices.  Given the ACPI paths of these two devices
(\_SB.PCI0.LPCB.FJEX, \_SB.FEXT), their parent devices are also not
power manageable.  These parent devices will thus have their power state
initialized to ACPI_STATE_D0, which in turn causes the power state for
both FUJ02B1 and FUJ02E3 to always be ACPI_STATE_D0 ("on").

Remove relevant acpi_bus_update_power() calls along with parts of debug
messages that they were supposed to have an effect on.
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

In the case of brightness-related FUJ02B1 ACPI device, initializing the
input device associated with it identically as acpi-video initializes
its input device makes sense.  However, using the same data for the
input device associated with the FUJ02E3 ACPI device makes little sense,
because the latter has nothing to do with video and assigning an
arbitrary product ID to it is redundant.
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

No formatting is needed when setting ACPI device name and class, so
switch to using strcpy() for this purpose.
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Do not check whether the pointer passed to ACPI add callbacks is NULL as
it is earlier dereferenced anyway in the bus-level probe callback,
acpi_device_probe().

Do not check the value of acpi_disabled in fujitsu_init(), because it is
already done by acpi_bus_register_driver(), which is the first function
called by fujitsu_init().
Signed-off-by: NMichał Kępień <kernel@kempniu.pl>
Reviewed-by: NJonathan Woithe <jwoithe@just42.net>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

File size before:
   text	   data	    bss	    dec	    hex	filename
   5396	   5016	     85	  10497	   2901	drivers/platform/x86/msi-laptop.o

File size After adding 'const':
   text	   data	    bss	    dec	    hex	filename
   5524	   4888	     85	  10497	   2901	drivers/platform/x86/msi-laptop.o
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

File size before:
   text	   data	    bss	    dec	    hex	filename
   9934	   1136	      2	  11072	   2b40	drivers/platform/x86/eeepc-laptop.o

File size After adding 'const':
   text	   data	    bss	    dec	    hex	filename
   9998	   1072	      2	  11072	   2b40	drivers/platform/x86/eeepc-laptop.o
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

File size before:
   text	   data	    bss	    dec	    hex	filename
   1471	    528	      8	   2007	    7d7	drivers/platform/x86/toshiba_haps.o

File size After adding 'const':
   text	   data	    bss	    dec	    hex	filename
   1519	    464	      8	   1991	    7c7	drivers/platform/x86/toshiba_haps.o
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Pali recently noticed that WMI instances are zero indexed.

The only reason that these calls all worked properly is because the ASL
didn't verify the instance number.
Signed-off-by: NLouis Davis <louis.davis@dell.com>
Signed-off-by: NMario Limonciello <mario.limonciello@dell.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Pali recently noticed that WMI instances are zero indexed.

The only reason that these calls all worked properly is because the ASL
didn't verify the instance number.
Signed-off-by: NMario Limonciello <mario.limonciello@dell.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Declare thermal_cooling_device_ops structure as const as it is only passed
as an argument to the function thermal_cooling_device_register and this
argument is of type const. So, declare the structure as const.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Declare thermal_cooling_device_ops structure as const as it is only passed
as an argument to the function thermal_cooling_device_register and this
argument is of type const. So, declare the structure as const.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Dell Latitude 3160 does not have keyboard backlight, but there is a
sysfs interface for it, which does nothing at all.

KBD_LED_ON_TOKEN is the only token can be found. Since it doesn't have
KBD_LED_OFF_TOKEN or KBD_LED_AUTO_*_TOKEN, it should be safe to assume
at least two tokens should be present to support keyboard backlight.
Not all models have ON token - they may have multiple AUTO tokens instead.

Models which do not use SMBIOS token to control keyboard backlight, also
have this issue. Brightness level is 0 on these models. Verified on Dell
Inspiron 3565.

Reports keyboard backlight is supported only when at least two modes are
present.
Signed-off-by: NKai-Heng Feng <kai.heng.feng@canonical.com>
Reviewed-by: NPali Rohár <pali.rohar@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Pali Rohár found that there have some wmi query/evaluation
code that they used 'one' as the first WMI instance number.
But the number is indexed from zero that it's must less than
the instance_count in _WDG.

This patch changes those instance number from one to zero.

Cc: Darren Hart <dvhart@infradead.org>
Cc: Andy Shevchenko <andy@infradead.org>
Reviewed-by: NPali Rohár <pali.rohar@gmail.com>
Signed-off-by: N"Lee, Chun-Yi" <jlee@suse.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Lenovo Legion Y720-15IKBN is yet another Lenovo model that does not
have an hw rfkill switch, resulting in wifi always reported as hard
blocked.

Add the model to the list of models without rfkill switch.
Signed-off-by: NOlle Liljenzin <olle@liljenzin.se>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Lenovo Legion Y520-15IKBN is yet another Lenovo model that does not
have an hw rfkill switch, resulting in wifi always reported as hard
blocked.

Add the model to the list of models without rfkill switch.
Signed-off-by: NOlle Liljenzin <olle@liljenzin.se>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Add touchscreen info for the Point of View mobii wintab p800w tablet.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Add touchscreen info for Pipo W2S tablet.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

object_id and notify_id are in one union structure and their meaning is
defined by flags. Therefore do not print notify_id for non-event block and
do not print object_id for event block. Remove also reserved member as it
does not have any defined meaning or type yet.

As object_id and notify_id union members overlaps and have different types,
it caused that kernel print to dmesg binary data. This patch eliminates it.
Signed-off-by: NPali Rohár <pali.rohar@gmail.com>
Reviewed-by: NAndy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Add const to rfkill_ops structures that are only passed as an argument
to the functions rfkill_alloc or samsung_new_rfkill. These arguments are
of type const, so such structures can be annotated with const.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Add const to rfkill_ops structure as it is only passed as an argument
to the functions rfkill_alloc. This argument is of type const,
so annotate the structure with const.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Add const to rfkill_ops structures that are only passed as an argument
to the functions rfkill_alloc or samsung_new_rfkill. These arguments are
of type const, so such structures can be annotated with const.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Add copyright statements for Andy Lutomirski and Darren Hart (VMware)
for their contributions to the WMI bus infrastructure and the creation
of the wmi-bmof driver.
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>
Cc: Andy Lutomirski <luto@kernel.org>

If a machine reports a RF Button in the communication button device
bitmap, we need to remove it before calling Get Device Status otherwise
it will return the "Undefined device" (0xE2) error code.

Although this may be a BIOS bug, we don't really need to get or set the
RF Button status. The status indicator LED embedded in the button is
controlled by firmware logic, depending on the status of the wireless
radios present on the machine (WiFi || WWAN).

This commit fixes the wireless status indicator LED on the Acer
TravelMate P648-G2-MG, and cleans the following message from the kernel
log: "Get Current Device Status failed: 0xe2 - 0x0".
Signed-off-by: NJoão Paulo Rechi Vita <jprvita@endlessm.com>
Reviewed-by: N"Lee, Chun-Yi" <joeyli.kernel@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

The struct pcc_keyinput is not used in panasonic-laptop and in
anywhere in kernel, and it can be removed.
Signed-off-by: NAlex Hung <alex.hung@canonical.com>
Acked-by: NHarald Welte <laforge@gnumonks.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

The variable is used uninitialized which might come into unexpected
behaviour on some Samsung laptops.

Initialize it to 0xffff which seems a proper value for non-supported
feature.
Reported-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Move some initialization out of _init and into _probe.
Update signatures and logic to use the wmi bus and device structures.
Signed-off-by: NAndy Lutomirski <luto@kernel.org>
[dvhart: drop deprecated sparse_keymap_free, order declarations, add commit msg]
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

Many laptops (and maybe servers?) have embedded WMI Binary MOF metadata.
We do not yet have open-source tools for processing the data, although
one is in the works thanks to Pali:

	https://github.com/pali/bmfdec

There is currently no interface to get the data in the first place. By
exposing it, we facilitate the development of new tools.

This is based on the original work of Andy Lutomirski <luto@kernel.org>,
but contains several modifications in response to various reviews.
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: NAndy Shevchenko <andy.shevchenko@gmail.com>

The Microsoft WMI documentation requires all data blocks to implement
the Query Control Method (WQxx). If we encounter a data block not
implementing this control method, issue a warning, and ignore the data
block. Remove the "readable" attribute as all data blocks must be
readable (query-able).

Be consistent with the language in the documentation, replace the
"writable" attribute with "setable".

Simplify (flatten) the control flow of wmi_create_device a bit while
we are updating it for the above changes.
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Some subdrivers need to access sibling devices. This gives them a
clean way to do so.
Signed-off-by: NAndy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

We already have the PNP glue to instantiate platform devices for the
ACPI devices that WMI drives. WMI should therefore attach to the
platform device, not the ACPI node.
Signed-off-by: NAndy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

wmi_query_block is unnecessarily indirect. Add a straightforward
method for wmi bus drivers to use to read block data.
Signed-off-by: NAndy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

As a platform driver, acpi_driver.notify will not be available,
so use acpi_install_notify_handler as we will be converting to a
platform driver.

This gives event drivers a simple way to handle events. It
also seems closer to what the Windows docs suggest that Windows
does: it sounds like, in Windows, the mapper is responsible for
called _WED before dispatching to the subdriver.
Signed-off-by: NAndy Lutomirski <luto@kernel.org>
[dvhart: merge two development commits and update commit message]
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NDarren Hart (VMware) <dvhart@infradead.org>

At some point, we will want sub-drivers to get references to other
devices on the same WMI bus. This change is needed to avoid races.

This ends up simplifying the setup code and fixing some leaks, too.

This is based on the original work of Andy Lutomirski <luto@kernel.org>,
but includes several modifications, many in response to review from
Michał Kępień <kernel@kempniu.pl>:

  https://www.spinics.net/lists/platform-driver-x86/msg08201.htmlSigned-off-by: NDarren Hart (VMware) <dvhart@infradead.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Mario Limonciello <mario_limonciello@dell.com>
Cc: Pali Rohár <pali.rohar@gmail.com>
Cc: linux-kernel@vger.kernel.org
Cc: platform-driver-x86@vger.kernel.org
Cc: linux-acpi@vger.kernel.org
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>