Add initial support for platform_profile where the support is
based on availability of ASUS_THROTTLE_THERMAL_POLICY.
Because throttle_thermal_policy is used by platform_profile and is
writeable separately to platform_profile any userspace changes to
throttle_thermal_policy need to notify platform_profile.
In future throttle_thermal_policy sysfs should be removed so that
only one method controls the laptop power profile.
Signed-off-by: Luke D. Jones <luke@ljones.dev>
Link: https://lore.kernel.org/r/20210818190731.19170-2-luke@ljones.dev
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Dynamic BIOS SAR driver exposing dynamic SAR information from BIOS
The Dynamic SAR (Specific Absorption Rate) driver uses ACPI DSM
(Device Specific Method) to communicate with BIOS and retrieve
dynamic SAR information and change notifications. The driver uses
sysfs to expose this data to userspace via read and notify.
Sysfs interface is documented in detail under:
Documentation/ABI/testing/sysfs-driver-intc_sar
Signed-off-by: Shravan S <s.shravan@intel.com>
Link: https://lore.kernel.org/r/20210723211452.27995-2-s.shravan@intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
This patch fixes the bug 212671.
Althrough the Fn lock (Fn + Esc) works on Legion 5 (R7000P), its LED
light does not change with the state. This modification sets the Fn lock
state to its current value on receiving the wmi event
8FC0DE0C-B4E4-43FD-B0F3-8871711C1294 to update the LED state.
Signed-off-by: Meng Dong <whenov@gmail.com>
Link: https://lore.kernel.org/r/20210817171203.12855-1-whenov@gmail.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Substate priority levels are encoded in 4 bits in the LPM_PRI register.
This value was used as an index to an array whose element size was less
than 16, leading to the possibility of overflow should we read a larger
than expected priority. In addition to the overflow, bad values could lead
to incorrect state reporting. So rework the priority code to prevent the
overflow and perform some validation of the register. Use the priority
register values if they give an ordering of unique numbers between 0 and
the maximum number of states. Otherwise, use a default ordering instead.
Reported-by: Evgeny Novikov <novikov@ispras.ru>
Signed-off-by: David E. Box <david.e.box@linux.intel.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Link: https://lore.kernel.org/r/20210814014728.520856-1-david.e.box@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The Acer Predator Helios series (usually denoted by PHxxx-yy) features
a particular key above the keyboard named "TURBO".
The turbo key does 3 things:
1. Set all fan's speeds to TURBO mode
2. Overclocks the CPU and GPU in the safe range
3. Turn on an LED just below the turbo button
All the above actions are operating using WMI function calls,
and there is no custom OC level for turbo. It acts as a flag
for enabling turbo mode instead of telling processors to use
a specific multiply of power (e.g. 1.3x of power).
I've run some benchmark tests and it worked fine:
GpuTest 0.7.0
http://www.geeks3d.com
Module: FurMark
Normal mode Score: 7289 points (FPS: 121)
Turbo mode Score: 7675 points (FPS: 127)
Settings:
- 1920x1080 fullscreen
- antialiasing: Off
- duration: 60000 ms
Renderer:
- GeForce RTX 2060/PCIe/SSE2
- OpenGL: 4.6.0 NVIDIA 460.32.03
This feature is presented by Acer officially and should not harm
hardware in any case.
A challenging part of implementing this feature is that calling
overclock function requires knowing the exact count of fans
for CPU and GPU of each model, which to the best of my
knowledge is not available in the kernel.
So after checking the official PredatorSense application methods, it
turned out they have provided the software the list of fans in each model.
I have access to the mentioned list, and all similar PH-iii-jj can be
added easily by matching "DMI_PRODUCT_NAME".
Creating a specific file for the Acer gaming features is not possible
because the current in use WMI event GUID is required for the turbo button
and it's not possible to register multiple listeners on a single WMI event.
Signed-off-by: JafarAkhondali <jafar.akhoondali@gmail.com>
Link: https://lore.kernel.org/r/20210812125307.1749207-1-jafar.akhoondali@gmail.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
In Windows the ASUS Armory Crate program can enable or disable the
dGPU via a WMI call. This functions much the same as various Linux
methods in software where the dGPU is removed from the device tree.
However the WMI call saves the state of dGPU (enabled or not) and
this then changes the dGPU visibility in Linux with no way for
Linux users to re-enable it. We expose the WMI method so users can
see and change the dGPU ACPI state.
Signed-off-by: Luke D. Jones <luke@ljones.dev>
Link: https://lore.kernel.org/r/20210807023656.25020-3-luke@ljones.dev
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
dell-smbios is depended on by dell-laptop and that has this same table +
some extra entries for chassis-type 30, 31 and 32.
Since dell-laptop will already auto-load based on the DMI table in there
(which also is more complete) and since dell-laptop will then bring in
the dell-smbios module, the only scenario I can think of where this DMI
table inside dell-smbios-smm.c is useful is if users have the dell-laptop
module disabled and they want to use the sysfs interface offered by
dell-smbios-smm.c. But that is such a corner case, even requiring a custom
kernel build, that it does not weigh up against having this duplicate
table, which as the current state already shows can only grow stale.
Users who do hit this corner-case can always explicitly modprobe /
insmod the module.
Cc: Mario Limonciello <mario.limonciello@outlook.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20210802120734.36732-1-hdegoede@redhat.com
When numa is used to map CPU to PCI device, the optimized path to read
from cached data is not working and still calls _isst_if_get_pci_dev().
The reason is that when caching the mapping, numa information is not
available as it is read later. So move the assignment of
isst_cpu_info[cpu].numa_node before calling _isst_if_get_pci_dev().
Fixes: aa2ddd2425 ("platform/x86: ISST: Use numa node id for cpu pci dev mapping")
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20210727165052.427238-1-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The gpiod_lookup_table.table passed to gpiod_add_lookup_table() must
be terminated with an empty entry, add this.
Note we have likely been getting away with this not being present because
the GPIO lookup code first matches on the dev_id, causing most lookups to
skip checking the table and the lookups which do check the table will
find a matching entry before reaching the end. With that said, terminating
these tables properly still is obviously the correct thing to do.
Fixes: f8eb0235f6 ("x86: pcengines apuv2 gpio/leds/keys platform driver")
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20210806115515.12184-1-hdegoede@redhat.com
