MAX31760 is a precision fan speed controller with nonvolatile lookup table.
Device has one internal and one external temperature sensor support.
Controls two fans and measures their speeds. Generates hardware alerts when
programmable max and critical temperatures are exceeded.
Signed-off-by: Ibrahim Tilki <Ibrahim.Tilki@analog.com>
Reviewed-by: Nurettin Bolucu <Nurettin.Bolucu@analog.com>
Link: https://lore.kernel.org/r/20220910171945.48088-2-Ibrahim.Tilki@analog.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Fix: "ERROR: space required before the open parenthesis '('"
All of the errors were introduced before this series of patches.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-22-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
This change adds debugfs to read and write temperature sensor coefficients
- g, h, j and cal5.
The coefficients can vary between product and product, so it can be very
useful to be able to modify them on the fly during the calibration
process.
e.g.:
cat /sys/kernel/debug/940f23d0000.pvt/ts_coeff_cal5
4096
echo 83000 > sys/kernel/debug/940f23d0000.pvt/ts_coeff_g
Signed-off-by: Eliav Farber <farbere@amazon.com>
Link: https://lore.kernel.org/r/20220908152449.35457-21-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Use thermal coefficients from the device tree if they exist.
Otherwise, use default values according to the series (5 or 6).
All coefficients can be used or only part of them.
The coefficients shall be used for fine tuning the default values since
coefficients can vary between product and product.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-20-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add optional temperature coefficient properties:
*) moortec,ts-coeff-g
*) moortec,ts-coeff-h
*) moortec,ts-coeff-cal5
*) moortec,ts-coeff-j
If defined they shall be used instead of defaults.
The coefficients were added to device tree on top of the series property
(which can be used to select between series 5 and series 6), because
coefficients can vary between product and product, and code defaults might
not be accurate enough.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-19-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The current equation used in code is aligned to series 5:
T = G + H * (n / cal5 - 0.5) + J * F
Where:
G = 60, H = 200, cal5 = 4094, J = -0.1, F = frequency clock in MHz
Series 6 has a slightly different equation:
T = G + H * (n / cal5 - 0.5)
and a different set of coefficients:
G = 57.4, H = 249.4, cal5 = 4096
This change supports equation and coefficients for both series.
(for series 6, J is set to 0).
The series is determined according to “moortec,ts-series” property in
the device tree.
If absent, series 5 is assumed to be the default.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-18-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add optional "moortec,ts-series" property to define the temperature
equation and coefficients that shall be used to convert the digital
output to value in milli-Celsius.
Supported series: 5 (default) and 6.
Series 5:
T = G + H * (n / cal5 - 0.5) + J * F
Where: G = 60, H = 200, cal5 = 4094, J = -0.1, F = frequency clock in MHz
Series 6:
T = G + H * (n / cal5 - 0.5)
Where: G = 57.4, H = 249.4, cal5 = 4096
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-17-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The SENSORS_ASPEED is part of the Aspeed silicon so it makes
sense to depend on ARCH_ASPEED and for compile testing.
Signed-off-by: Peter Robinson <pbrobinson@gmail.com>
Link: https://lore.kernel.org/r/20220916120936.372591-1-pbrobinson@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Modify the equation and coefficients used to convert the digital output
to temperature according to series 5 of the Moortec Embedded Temperature
Sensor (METS) datasheet:
T = G + H * (n / cal5 - 0.5) + J * F
Where:
*) G = 60, H = 200, cal5 = 4094, J = -0.1.
*) F = frequency clock in MHz.
*) n is the digital output.
In code, the G, H and J coefficients are multiplied by a factor of 1000
to get the temperature in milli-Celsius.
Final result is clamped in case it exceeds min/max thresholds.
Change is done since it is unclear where the current equation and
coefficients came from.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-16-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for mr76006 pre-scaler which provides divide-by-2 scaling
of the input voltage, so that it can be presented to the VM for
measurement within its range (the VM input range is limited from -0.1V
to 1V).
The driver reads from the device-tree all the channels that use the
mr76006 pre-scaler and multiplies the voltage result by a factor of 2,
to represent to the user with the actual voltage input source.
Channels that are not in the device-tree are multiplied by a factor
of 1.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-15-farbere@amazon.com
[groeck: Addressed conflicts against commit d59eacaac9]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for mr76006 pre-scaler which provides divide-by-2 scaling of
the input voltage, so that it can be presented to the VM for measurement
within its range (the VM input range is limited to -0.1V to 1V).
The new "moortec,vm-pre-scaler-x2" property lists the channels that use
the mr76006 pre-scaler.
The driver will use this list to multiply the voltage result by 2, to
present to the user with the actual voltage input source.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-14-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add active channel support per voltage monitor.
The number of active channels is read from the device-tree.
When absent in device-tree, all channels are assumed to be used.
This shall be useful to expose sysfs only for inputs that are connected
to a voltage source.
Setting number of active channels to 0, means that entire VM sensor is
not used.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-13-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add optional "moortec,vm-active-channels" property to define the number
of active channels per VM.
This shall be useful to avoid exposing sysfs for reading inputs that are
not connected to any voltage source.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-12-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Static analyzers report:
drivers/hwmon/emc2305.c:194 emc2305_set_cur_state()
warn: impossible condition '(val > 255) => (0-255 > 255)'
'val' is u8 and thus can never be larger than 255. In theory
the operation calculating 'val' could result in a value larger
than 255, but this won't happen because its parameter has already
been range checked and it is guaranteed that the result never exceeds
255. Remove the unnecessary value check.
Cc: Michael Shych <michaelsh@nvidia.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Don't fail the probe function and don't deassert the reset controller if
a "reset" property doesn't exist in the device tree.
Change is done for SOCs that don't support a reset controller.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220908152449.35457-10-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Change "resets" property to be optional instead of required, for SOCs
that don't support a reset controller.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Acked-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-9-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
This changes adds a detailed description for the mr75203 controller and
for some of the analog IPs controlled by it.
Signed-off-by: Eliav Farber <farbere@amazon.com>
Acked-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220908152449.35457-8-farbere@amazon.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Extend aquacomputer_d5next driver to expose various hardware
sensors of the Aquacomputer High Flow Next flow sensor, which
communicates through a proprietary USB HID protocol.
The High Flow Next exposes +5V voltages, water quality, conductivity
and flow readings. A temperature sensor can be connected to it, in
which case it provides its reading and an estimation of the
dissipated/absorbed power in the liquid cooling loop.
Additionally, serial number and firmware version are exposed through
debugfs.
Registry offsets were discovered and tested by users on Github [1] [2].
[1] https://github.com/aleksamagicka/aquacomputer_d5next-hwmon/issues/8
[2] https://github.com/aleksamagicka/aquacomputer_d5next-hwmon/pull/34
Signed-off-by: Aleksa Savic <savicaleksa83@gmail.com>
Link: https://lore.kernel.org/r/20220907100739.806571-1-savicaleksa83@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
On systems with more than one tps23861, creating the debugfs directory
for additional devices fails with
debugfs: Directory 'tps23861' with parent '/' already present!
To resolve this, include the hwmon device name in the directory name.
Since the name is unique, this guarantees that the debugfs directory
is unique.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20220907015405.16547-2-mr.nuke.me@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
When reading the 'port_status' debugfs entry, some I2C registers were
read more than once. This looks inefficient in an I2C trace.
To reduce I2C traffic, update tps23861_port_status_show() to only read
each register once. Indexing the port number from 0 instead of 1 also
allows simplifying things a bit.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Link: https://lore.kernel.org/r/20220907015405.16547-1-mr.nuke.me@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add driver for Microchip EMC2301/2/3/5 RPM-based PWM Fan Speed Controller.
Modify Makefile and Kconfig to support Microchip EMC2305 RPM-based
PWM Fan Speed Controller.
Signed-off-by: Michael Shych <michaelsh@nvidia.com>
Reviewed-by: Vadim Pasternak <vadimp@nvidia.com>
Link: https://lore.kernel.org/r/20220810171552.56417-3-michaelsh@nvidia.com
[groeck: Drop unnecessary () around DIV_ROUND_CLOSEST()]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Introduce platform data structure for EM2305 driver to allow configuration
device PWMs and thermal zones by passing required platform data
to the driver. If no platform data is provided, the driver is supposed
to work with default settings.
Signed-off-by: Michael Shych <michaelsh@nvidia.com>
Reviewed-by: Vadim Pasternak <vadimp@nvidia.com>
Link: https://lore.kernel.org/r/20220810171552.56417-2-michaelsh@nvidia.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The MAX31790 has a tach input enable bit in each fan's configuration
register. This is only enabled by the driver if RPM mode is selected,
but the driver doesn't provide a way to independently enable tachometer
input regardless of the regulator mode.
By adding the fanN_enable sysfs files, we can decouple the tach input
from the regulator mode. Also update the documentation.
Signed-off-by: Justin Ledford <justinledford@google.com>
Link: https://lore.kernel.org/r/20220829195930.2521755-1-justinledford@google.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
There are no users outside of PWM core of the of_pwm_get().
Make it static.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20220826172642.16404-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The devm_of_pwm_get() has recently lost its single user, drop
the dead API as well.
Note, the new code should use either plain pwm_get() or managed
devm_pwm_get() or devm_fwnode_pwm_get() APIs.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20220826172642.16404-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
There is no need to call OF specific devm_of_pwm_get() since
the device node parameter duplicates in the device parameter.
Hence we may safely replace it by plain devm_pwm_get() call.
This allows to drop devm_of_pwm_get() as no more users will be.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20220826172642.16404-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The devm_clk_get_enabled() helper:
- calls devm_clk_get()
- calls clk_prepare_enable() and registers what is needed in order to
call clk_disable_unprepare() when needed, as a managed resource.
This simplifies the code, the error handling paths and avoid the need of
a dedicated function used with devm_add_action_or_reset().
Based on my test with allyesconfig, this reduces the .o size from:
text data bss dec hex filename
2419 1472 0 3891 f33 drivers/hwmon/sparx5-temp.o
down to:
2155 1472 0 3627 e2b drivers/hwmon/sparx5-temp.o
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/cfe4c965074b5ecbe03830b05e038b4594c7b970.1661336689.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
When dell-smm-hwmon is loaded on a machine with a buggy BIOS
with the option "force" being enabled, it wrongly prints
that the buggy features where disabled. This may cause
users to wrongly assume that the driver still protects them
from these BIOS bugs even with "force" being enabled.
Replace the messages with two messages each which are depending
on the value of the "force" parameter. The messages which are
being printed when "force" is not set use dev_notice() instead
of dev_warn() since they only serve as a notice.
Tested on a Dell Inspiron 3505.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20220822174053.8750-3-W_Armin@gmx.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Convert the module to be property provider agnostic and allow
it to be used on non-OF platforms.
Include mod_devicetable.h explicitly to replace the dropped of.h
which included mod_devicetable.h indirectly.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20220826173700.17395-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add reporting if the PSU is running in single or multi rail mode via
ocpmode debugfs entry. Also update the documentation and driver comments
accordingly.
Signed-off-by: Wilken Gottwalt <wilken.gottwalt@posteo.net>
Link: https://lore.kernel.org/r/YvS9PZKr0xqFqJny@monster.localdomain
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for reading virtual temperature sensors for the D5 Next, Octo,
Quadro and Farbwerk 360.
Virtual temperature sensors are written to the device by the user, pulling
from an arbitrary value source. Writing to them is not yet reverse
engineered, so the only way to set them for now is to use the official
software.
Signed-off-by: Aleksa Savic <savicaleksa83@gmail.com>
Link: https://lore.kernel.org/r/20220817121441.112198-1-savicaleksa83@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
This driver utilises a WMI interface found in AMD 500 series ASUS boards,
to read EC registers. But it turned out that ASUS abandoned the
interface, as it disappeared from Intel 600 series boards. Additionally,
the WMI interface was incredibly slow. Therefore this driver was deprecated
in favor of the asus_ec_sensors driver, which supports more boards, more
sensors, and is faster.
Signed-off-by: Eugene Shalygin <eugene.shalygin@gmail.com>
Link: https://lore.kernel.org/r/20220720072016.102086-2-eugene.shalygin@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
In gsc_hwmon_get_devtree_pdata(), we should call of_node_get() before
the of_find_compatible_node() which will automatically call
of_node_put() for the 'from' argument.
Fixes: 3bce5377ef ("hwmon: Add Gateworks System Controller support")
Signed-off-by: Liang He <windhl@126.com>
Co-developed-by: Mengda Chen <chenmengda2009@163.com>
Signed-off-by: Mengda Chen <chenmengda2009@163.com>
Link: https://lore.kernel.org/r/20220916154708.3084515-1-chenmengda2009@163.com
Cc: stable@vger.kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The offsets for setting speeds of fans connected to Quadro are off by one.
Set them to their correct values.
The offsets as shown point to registers for setting the fan control mode,
which will be explored in future patches, but slipped in here. When
setting fan speeds, the resulting values were overlapping, which made the
fans still run in my initial testing.
Fixes: cdbe34da01 ("hwmon: (aquacomputer_d5next) Add support for Aquacomputer Quadro fan controller")
Signed-off-by: Aleksa Savic <savicaleksa83@gmail.com>
Link: https://lore.kernel.org/r/20220914114327.6941-1-savicaleksa83@gmail.com
Cc: stable@vger.kenrel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
- Remove unused scripts/gcc-ld script
- Add zstd support to scripts/extract-ikconfig
- Check 'make headers' for UML
- Fix scripts/mksysmap to ignore local symbols
-----BEGIN PGP SIGNATURE-----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=FJba
-----END PGP SIGNATURE-----
Merge tag 'kbuild-fixes-v6.0-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Remove unused scripts/gcc-ld script
- Add zstd support to scripts/extract-ikconfig
- Check 'make headers' for UML
- Fix scripts/mksysmap to ignore local symbols
* tag 'kbuild-fixes-v6.0-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
mksysmap: Fix the mismatch of 'L0' symbols in System.map
kbuild: disable header exports for UML in a straightforward way
scripts/extract-ikconfig: add zstd compression support
scripts: remove obsolete gcc-ld script
- Disable in-kernel BTI when compiling with GCC, as it makes invalid
assumptions about the distance between functions which has led to
crashes when calling modules on a CPU with BTI support.
- Remove bogus TIF_SME flag management if memory allocation fails in the
ptrace code.
- Fix the resume path when configured for 52-bit virtual addressing.
-----BEGIN PGP SIGNATURE-----
iQFEBAABCgAuFiEEPxTL6PPUbjXGY88ct6xw3ITBYzQFAmMcmCUQHHdpbGxAa2Vy
bmVsLm9yZwAKCRC3rHDchMFjNJEPB/9hhMNrX0yAc64otI99DRBuuZjlGWm5tLVj
A5dByt0bvca6XIE/5nZpduYcSGBORL4Ss29mbYteWlaU89TlLEHaedOpDTX0xj5g
61H8UiK0YDX9w/7hbPZw7Uxdvw30LJSxszWZ4Ucst9CYTq+mMbbhgf8h7/BDmBD+
OBpkOlouW4fSw15/UlbgBu6UTAVkbWNcWxzwzOVAdqFYzQ75eMnmOEkEZ8b/YBAA
wZ+0kFmpjGkWymZKlgTxLfQ3eVdKwlrFzBliDtbFoFNuWyyDQ2e1qDQKFZvST4ul
NMVlF19K9NKGVlxl2++rafbg5BQ+Y+0RJ2Hujf72W6sPaThJUDM2
=7sa1
-----END PGP SIGNATURE-----
Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Will Deacon:
"Three small arm64 fixes, all related to optional architecture
extensions: BTI, SME and 52-bit virtual addressing:
- Disable in-kernel BTI when compiling with GCC, as it makes invalid
assumptions about the distance between functions which has led to
crashes when calling modules on a CPU with BTI support
- Remove bogus TIF_SME flag management if memory allocation fails in
the ptrace code
- Fix the resume path when configured for 52-bit virtual addressing"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: mm: fix resume for 52-bit enabled builds
arm64/ptrace: Don't clear calling process' TIF_SME on OOM
arm64/bti: Disable in kernel BTI when cross section thunks are broken
Including:
- Intel VT-d fixes from Lu Baolu
- Boot kdump kernels with VT-d scalable mode on
- Calculate the right page table levels
- Fix two recursive locking issues
- Fix a lockdep splat issue
- AMD IOMMU fixes:
- Fix for completion-wait command to use full 64 bits of data
- Fix PASID related issue where GPU sound devices failed to
initialize
- Fix for Virtio-IOMMU to report correct caching behavior, needed for
use with VFIO
-----BEGIN PGP SIGNATURE-----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=+h6p
-----END PGP SIGNATURE-----
Merge tag 'iommu-fixes-v6.0-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu
Pull iommu fixes from Joerg Roedel:
- Intel VT-d fixes from Lu Baolu:
- Boot kdump kernels with VT-d scalable mode on
- Calculate the right page table levels
- Fix two recursive locking issues
- Fix a lockdep splat issue
- AMD IOMMU fixes:
- Fix for completion-wait command to use full 64 bits of data
- Fix PASID related issue where GPU sound devices failed to
initialize
- Fix for Virtio-IOMMU to report correct caching behavior, needed for
use with VFIO
* tag 'iommu-fixes-v6.0-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
iommu: Fix false ownership failure on AMD systems with PASID activated
iommu/vt-d: Fix possible recursive locking in intel_iommu_init()
iommu/virtio: Fix interaction with VFIO
iommu/vt-d: Fix lockdep splat due to klist iteration in atomic context
iommu/vt-d: Fix recursive lock issue in iommu_flush_dev_iotlb()
iommu/vt-d: Correctly calculate sagaw value of IOMMU
iommu/vt-d: Fix kdump kernels boot failure with scalable mode
iommu/amd: use full 64-bit value in build_completion_wait()
- fix for octeon irq setup problem
- fix compiler warning for new CONFIG option
- switch to SPARSEMEM_EXTREME for all platforms selecting SPARSEMEM
-----BEGIN PGP SIGNATURE-----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=djqb
-----END PGP SIGNATURE-----
Merge tag 'mips-fixes_6.0_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux
Pull MIPS fixes from Thomas Bogendoerfer:
- fix for loongson32 starup hang
- fix for octeon irq setup problem
- fix compiler warning for new CONFIG option
- switch to SPARSEMEM_EXTREME for all platforms selecting SPARSEMEM
* tag 'mips-fixes_6.0_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
mips: Select SPARSEMEM_EXTREME
MIPS: OCTEON: irq: Fix octeon_irq_force_ciu_mapping()
MIPS: octeon: Get rid of preprocessor directives around RESERVE32
MIPS: loongson32: ls1c: Fix hang during startup
The AMD IOMMU driver cannot activate PASID mode on a RID without the RID's
translation being set to IDENTITY. Further it requires changing the RID's
page table layout from the normal v1 IOMMU_DOMAIN_IDENTITY layout to a
different v2 layout.
It does this by creating a new iommu_domain, configuring that domain for
v2 identity operation and then attaching it to the group, from within the
driver. This logic assumes the group is already set to the IDENTITY domain
and is being used by the DMA API.
However, since the ownership logic is based on the group's domain pointer
equaling the default domain to detect DMA API ownership, this causes it to
look like the group is not attached to the DMA API any more. This blocks
attaching drivers to any other devices in the group.
In a real system this manifests itself as the HD-audio devices on some AMD
platforms losing their device drivers.
Work around this unique behavior of the AMD driver by checking for
equality of IDENTITY domains based on their type, not their pointer
value. This allows the AMD driver to have two IDENTITY domains for
internal purposes without breaking the check.
Have the AMD driver properly declare that the special domain it created is
actually an IDENTITY domain.
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: stable@vger.kernel.org
Fixes: 512881eacf ("bus: platform,amba,fsl-mc,PCI: Add device DMA ownership management")
Reported-by: Takashi Iwai <tiwai@suse.de>
Tested-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/0-v1-ea566e16b06b+811-amd_owner_jgg@nvidia.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The global rwsem dmar_global_lock was introduced by commit 3a5670e8ac
("iommu/vt-d: Introduce a rwsem to protect global data structures"). It
is used to protect DMAR related global data from DMAR hotplug operations.
The dmar_global_lock used in the intel_iommu_init() might cause recursive
locking issue, for example, intel_iommu_get_resv_regions() is taking the
dmar_global_lock from within a section where intel_iommu_init() already
holds it via probe_acpi_namespace_devices().
Using dmar_global_lock in intel_iommu_init() could be relaxed since it is
unlikely that any IO board must be hot added before the IOMMU subsystem is
initialized. This eliminates the possible recursive locking issue by moving
down DMAR hotplug support after the IOMMU is initialized and removing the
uses of dmar_global_lock in intel_iommu_init().
Fixes: d5692d4af0 ("iommu/vt-d: Fix suspicious RCU usage in probe_acpi_namespace_devices()")
Reported-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/894db0ccae854b35c73814485569b634237b5538.1657034828.git.robin.murphy@arm.com
Link: https://lore.kernel.org/r/20220718235325.3952426-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
