Gregor Boirie [Wed, 20 Apr 2016 17:23:43 +0000 (19:23 +0200)]
iio:core: mounting matrix support
Expose a rotation matrix to indicate userspace the chip placement with
respect to the overall hardware system. This is needed to adjust
coordinates sampled from a sensor chip when its position deviates from the
main hardware system.
Final coordinates computation is delegated to userspace since:
* computation may involve floating point arithmetics ;
* it allows an application to combine adjustments with arbitrary
transformations.
This 3 dimentional space rotation matrix is expressed as 3x3 array of
strings to support floating point numbers. It may be retrieved from a
"[<dir>_][<type>_]mount_matrix" sysfs attribute file. It is declared into a
device / driver specific DTS property or platform data.
Signed-off-by: Gregor Boirie <gregor.boirie@parrot.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
iio: Add resource managed APIs devm_iio_channel_{get,release) in devres
Add following APIs in the list of managed resources of IIO:
devm_iio_channel_get()
devm_iio_channel_get_all()
devm_iio_channel_release()
devm_iio_channel_release_all()
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
iio: core: Add devm_ APIs for iio_channel_{get,release}_all
Some of kernel driver uses the IIO framework to get the sensor
value via ADC or IIO HW driver. The client driver get iio channel
by iio_channel_get_all() and release it by calling
iio_channel_release_all().
Add resource managed version (devm_*) of these APIs so that if client
calls the devm_iio_channel_get_all() then it need not to release it
explicitly, it can be done by managed device framework when driver
get un-binded.
This reduces the code in error path and also need of .remove callback in
some cases.
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
iio: core: Add devm_ APIs for iio_channel_{get,release}
Some of kernel driver uses the IIO framework to get the sensor
value via ADC or IIO HW driver. The client driver get iio channel
by iio_channel_get() and release it by calling iio_channel_release().
Add resource managed version (devm_*) of these APIs so that if client
calls the devm_iio_channel_get() then it need not to release it explicitly,
it can be done by managed device framework when driver get un-binded.
This reduces the code in error path and also need of .remove callback in
some cases.
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
If no channels are enabled when we run generic_buffer on a
device, add a command-line option to just enable all of them,
run the sampling and disable them all again afterwards.
This is extremely useful when I'm low-level testing my
sensors with interrupts and triggers, sample session:
Some types of ST Sensors can be connected to the same IRQ line
as other peripherals using open drain. Add a device tree binding
and a sensor data property to flip the right bit in the interrupt
control register to enable open drain mode on the INT line.
If the line is set to be open drain, also tag on IRQF_SHARED
to the IRQ flags when requesting the interrupt, as the whole
point of using open drain interrupt lines is to share them with
more than one peripheral (wire-or).
Cc: devicetree@vger.kernel.org Cc: Giuseppe Barba <giuseppe.barba@st.com> Cc: Denis Ciocca <denis.ciocca@st.com> Acked-by: Rob Herring <rob@kernel.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Linus Walleij [Thu, 24 Mar 2016 13:18:05 +0000 (14:18 +0100)]
iio: st_sensors: verify interrupt event to status
This makes all ST sensor drivers check that they actually have
new data available for the requested channel(s) before claiming
an IRQ, by reading the status register (which is conveniently
the same for all ST sensors) and check that the channel has new
data before proceeding to read it and fill the buffer.
This way sensors can share an interrupt line: it can be flaged
as shared and then the sensor that did not fire will return
NO_IRQ, and the sensor that fired will handle the IRQ and
return IRQ_HANDLED.
Cc: Giuseppe Barba <giuseppe.barba@st.com> Cc: Denis Ciocca <denis.ciocca@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Linus Walleij [Thu, 24 Mar 2016 13:18:04 +0000 (14:18 +0100)]
iio: st_sensors: read each channel individually
The current buffer read code tries to optimize reads from the
sensor data registers by issuing a single read operation across
all the indata registers.
This doesn't work: when the LIS331DL accelerometer sensor is
configured to open drain, active low interrupt mode, this will
just clear the XDA (X-axis data available) bit in the STATUS_REG
register (0x27), while YDA, ZDA and even ZYXDA remain set to 1,
and the internal logic of the sensor holds the DRDY (INT1) line
asserted (the value of the status register is 0xee).
If we instead issue one read operation per enabled channel
(X, Y, Z) things start working and we can use open drain and
active low interrupts.
Note that a backported patch fixing this issue will be heading
via the fixes branch but changes in this file already in staging-next
will make that patch 'look' rather different. The code in here
is the correct one when that clash hits.
Cc: Giuseppe Barba <giuseppe.barba@st.com> Cc: Denis Ciocca <denis.ciocca@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Stefan Wahren [Thu, 14 Apr 2016 15:48:27 +0000 (15:48 +0000)]
iio: mxs-lradc: move STMP reset out of ADC init
This patch moves the STMP reset out of ADC init function so as to remove
the possiblity of an error return which will be necessary for PM ops support
patches to follow.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com> Tested-by: Marek Vasut <marex@denx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Stefan Wahren [Thu, 14 Apr 2016 15:48:26 +0000 (15:48 +0000)]
iio: mxs-lradc: move TS config into suitable function
This patch moves the touchscreen type configuration into
a more suitable function. Btw this simplifies PM ops later.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com> Reviewed-by: Marek Vasut <marex@denx.de> Tested-by: Marek Vasut <marex@denx.de> Acked-by: Dmitry Torokhov <dmitry.torokhov@gmail.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Using this requires software triggers like CONFIG_IIO_HRTIMER_TRIGGER.
The device can be configured to do internal periodic sampling but does not
offer some sort of interrupt on data ready. Interrupts can only trigger when
values get out of a specific range.
Signed-off-by: Crestez Dan Leonard <leonard.crestez@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
This is a reimplementation of the old misc device driver for the
ROHM BH1780 ambient light sensor (drivers/misc/bh1780gli.c).
Differences from the old driver:
- Uses the IIO framework
- Uses runtime PM to idle the hardware after 5 seconds
- No weird custom power management from userspace
- No homebrewn values in sysfs
This uses the same (undocumented) device tree compatible-string
as the old driver ("rohm,bh1780gli").
Cc: Arnd Bergmann <arnd@arndb.de> Cc: Ulf Hansson <ulf.hansson@linaro.org> Cc: Daniel Mack <daniel@caiaq.de> Cc: Peter Meerwald-Stadler <pmeerw@pmeerw.net> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
staging:iio:adis16240: Set self_test_no_autoclear flag
The ADIS16201 does not automatically clear the self test flag bit the self
test has been, so clear it manually. Otherwise we'll see a offset caused by
the self-test bias on the output values during normal operation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
staging:iio:adis16209: Set self_test_no_autoclear flag
The ADIS16201 does not automatically clear the self test flag bit the self
test has been, so clear it manually. Otherwise we'll see a offset caused by
the self-test bias on the output values during normal operation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
staging:iio:adis16203: Set self_test_no_autoclear flag
The ADIS16201 does not automatically clear the self test flag bit the self
test has been, so clear it manually. Otherwise we'll see a offset caused by
the self-test bias on the output values during normal operation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
staging:iio:adis16201: Set self_test_no_autoclear flag
The ADIS16201 does not automatically clear the self test flag bit the self
test has been, so clear it manually. Otherwise we'll see a offset caused by
the self-test bias on the output values during normal operation.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
iio:adis: Add support for manual self-test flag clear
Some variants of the devices from the ADIS family don't auto-clear the
self-test bit after the self-test has completed. Instead we have to
manually clear. Add support for this to the ADIS library.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Daniel Baluta [Fri, 15 Apr 2016 15:06:56 +0000 (18:06 +0300)]
iio: imu: Add initial support for Bosch BMI160
BMI160 is an Inertial Measurement Unit (IMU) which provides acceleration
and angular rate measurement. It also offers a secondary I2C interface
for connecting a magnetometer sensor (usually BMM160).
Current driver offers support for accelerometer and gyroscope readings
via sysfs or via buffer interface using an external trigger (e.g.
hrtimer). Data is retrieved from IMU via I2C or SPI interface.
Datasheet is at:
http://www.mouser.com/ds/2/783/BST-BMI160-DS000-07-786474.pdf
Signed-off-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Harald Geyer [Sun, 17 Jan 2016 16:13:31 +0000 (16:13 +0000)]
iio: dht11: Improve logging
* Unify log messages
* Add more DEBUG messages
Apparently this driver is working unreliably on some platforms that I can't
test. Therefore I want an easy way for bug reporters to provide useful
information without making the driver too chatty by default.
Signed-off-by: Harald Geyer <harald@ccbib.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Paul Cercueil [Tue, 5 Apr 2016 07:46:19 +0000 (09:46 +0200)]
iio: dac: Add support for the AD5592R/AD5593R ADCs/DACs
This patch adds support for the AD5592R (spi) and AD5593R (i2c)
ADC/DAC/GPIO devices.
Signed-off-by: Paul Cercueil <paul.cercueil@analog.com> Signed-off-by: Michael Hennerich <michael.hennerich@analog.com> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
The "ch->ch_bd" is already assined to "bd" but this is only
for checking null or MAGIC number.
in the dgnc_tty_ioctl function, bd can be used for referencing
to board_ops structure.
Tetsuo Handa [Tue, 8 Mar 2016 11:01:32 +0000 (20:01 +0900)]
android,lowmemorykiller: Don't abuse TIF_MEMDIE.
Currently, lowmemorykiller (LMK) is using TIF_MEMDIE for two purposes.
One is to remember processes killed by LMK, and the other is to
accelerate termination of processes killed by LMK.
But since LMK is invoked as a memory shrinker function, there still
should be some memory available. It is very likely that memory
allocations by processes killed by LMK will succeed without using
ALLOC_NO_WATERMARKS via TIF_MEMDIE. Even if their allocations cannot
escape from memory allocation loop unless they use ALLOC_NO_WATERMARKS,
lowmem_deathpending_timeout can guarantee forward progress by choosing
next victim process.
On the other hand, mark_oom_victim() assumes that it must be called with
oom_lock held and it must not be called after oom_killer_disable() was
called. But LMK is calling it without holding oom_lock and checking
oom_killer_disabled. It is possible that LMK calls mark_oom_victim()
due to allocation requests by kernel threads after current thread
returned from oom_killer_disabled(). This will break synchronization
for PM/suspend.
This patch introduces per a task_struct flag for remembering processes
killed by LMK, and replaces TIF_MEMDIE with that flag. By applying this
patch, assumption by mark_oom_victim() becomes true.
Merge tag 'iio-for-4.7a' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-next
Jonathan writes:
First round of IIO new device support, features and cleanups for the 4.7 cycle.
New core support
* UV light modifier (for intensity)
* UV light index channel type.
New device support
* hp206c barometer and altimeter
- new driver.
* mcp4131 potentiometer
- new driver supporting lots of parts from Microchip.
* mma8452
- FXLS8471Q support
- NXP LPC18XX SOC ADC
- new driver.
- NXP LPC18XX SOC DAC
- new driver.
- rockchip_saradc
- support rk3399
* st accel
- h3lis331dl support
Staging driver removals
* adis16204
- obsolete part making it hard to get parts to test the driver in order
to clean it up.
* adis16220
- obsolete part making it hard to get the parts test the driver in order
to clean it up.
Features
* core
- convenience functions to claim / release direct access to the device.
Makes more consistent handling of this corner easier. Used in ad7192 driver.
* ak8975
- power regulator support.
* at91-sama5d2
- differential channel support.
* mma8452
- runtime pm support
- drop device specific autosleep and use the runtime pm one instead.
* ms5611
- DT bindings
- oversampling ratio support
Cleanups and minor fixes
* MAINTAINERS
- Peter got married - hence name change!
* Documentation
- Fix a typo in in_proximity_raw description.
- Add some missing docs for iio_buffer_access_funcs.
* Tools
- update iio_event_monitor names to match new stuff.
- make generic_buffer look for triggers ending in -trigger as we let these in
for a number of drivers a long time back and now it is a fairly common
option.
Drivers
* staging wide
- convert bare unsigned usage to unsigned int to comply with coding style.
* non staging wide:
- since boiler plate gpio handling of interrupts has been moved into the
ACPI core we don't need to include gpio/consumer.h in a load of drivers so
drop it.
* ad7606
- fix an endian casting sparse warning.
* ak8975
- fix a possible unitialized warning from gcc.
- drop and unused field left over from earlier cleanups
- fix a missing regulator_disable on exit.
* at91-sama5d2
- typo and indentation
- missing IOMEM dependency.
- cleanup mode register usage by avoidling erasing whole thing when changing
the sampling frequency.
* bmc150
- use the core demux and available_scan_masks to simplify buffer handling
- optimize the transfers in the trigger handler now we have a magic function
to emulate bulk reads (under circumstances met here). This matters with some
rather dumb i2c adapters in particular.
- use a single regmap_conf for all bus types as they were all the same.
* bmg160
- use the core demux and available_scan_masks to simplify the buffer handling
- optimize the transfers in the trigger handler now we have a magic funciton
to emulate bulk rads (under circumstances met here).
- drop gpio interrupt probing from the driver (ACPI) as now handled by the
ACPI core.
* ina2xx-adc
- update the CALIB register when RShunt changes.
- fix scale for VShunt - in reality this error canceled out when used.
* isl29028
- use regmap to retrieve the struct device instead of carrying a second
copy of it around.
* kxcjk-1013
- use core demux
- optimize i2c transfers in the trigger handler.
* mcp4531
- refactor to use a pointer to access model parameters instead of indexing
into the array each time.
* mma8452
- style fixes
- avoid swtiching to active whenever the config changes
- add missin i2c_device_id for mma8451
* mpu6050
- fix possible NULL dereference.
- fix the name / chip_id used when ACPI used (otherwise reports as NULL).
* ms5611
- fix a missing regulator_disable that left the regulator on during removal.
* mxc4005
- drop gpio interrupt handling for ACPI case from driver as the core now
handles this case.
* st-sensors
- note that there are only ever a maximum of 3 axis on current st-sensors
so just allocate a fixed sized buffer big enough for that.
* tpl0102
- change the i2c_check_functionality condition to bring it inline with other
IIO users as EOPNOTSUPP.
* tsl2563
- replace deprecated flush_scheduled_work
Irina Tirdea [Tue, 29 Mar 2016 12:21:21 +0000 (15:21 +0300)]
iio: accel: bmc150: use common definition for regmap conf
bmc150_i2c_regmap_conf is defined three times (in bmc150-accel-core.c,
bmc150-accel-i2c.c and and bmc150-accel-spi.c), although the
definition is the same.
Use one common definition for bmc150_i2c_regmap_conf in all
included files.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
iio: buffer: add missing descriptions in iio_buffer_access_funcs
The members buffer_group and attrs of iio_buffer_access_funcs have no
descriptions for the documentation. Adding them.
Fixes: 08e7e0adaa17 ("iio: buffer: Allocate standard attributes in the core") Signed-off-by: Luis de Bethencourt <luisbg@osg.samsung.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Linus Walleij [Thu, 24 Mar 2016 08:39:14 +0000 (09:39 +0100)]
iio: tools: make generic_buffer look for "-trigger"
All the ST Sensors use the old "<foo>-trigger" rather than the
standard "<foo>-devN" new standard suffix for triggers. Now much
to do about it since it is ABI, but make the testing tools
recognize it too.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Irina Tirdea [Thu, 24 Mar 2016 09:05:09 +0000 (11:05 +0200)]
iio: remove gpio interrupt probing from drivers that use a single interrupt
Commit 845c877009cf014b ("i2c / ACPI: Assign IRQ for devices that have
GpioInt automatically") automatically assigns the first ACPI GPIO
interrupt in client->irq, so we can remove the probing code from
drivers that use only one interrupt.
Commit 0f0796509c07c1c7 ("iio: remove gpio interrupt probing from drivers
that use a single interrupt") removes gpio interrupt probing from most
drivers. This patch cleans the remaining ones.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Irina Tirdea [Thu, 24 Mar 2016 09:08:38 +0000 (11:08 +0200)]
iio: remove unused gpio consumer.h include
GPIO handling code has been removed from the drivers (since
this is now handled by the ACPI core) in commit 0f0796509c07 ("iio:
remove gpio interrupt probing from drivers that use a single interrupt").
Remove the include for linux/gpio/consumer.h since it is no longer
used.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Adriana Reus [Thu, 24 Mar 2016 09:29:31 +0000 (11:29 +0200)]
iio: accel: kxcjk-1013: optimize i2c transfers in trigger handler
Some i2c busses (e.g.: Synopsys DesignWare I2C adapter) need to
enable/disable the bus at each i2c transfer and must wait for
the enable/disable to happen before sending the data.
When reading data in the trigger handler, the kxcjk-1013 accel driver
does one i2c transfer for each axis. This has an impact on the
frequency of the accelerometer at high sample rates due to additional
delays introduced by the i2c bus at each transfer.
Reading all axis values in one i2c transfer reduces the delays
introduced by the i2c bus. Uses i2c_smbus_read_i2c_block_data_or_emulated
that will fallback to reading each axis as a separate word in case i2c
block read is not supported.
Signed-off-by: Adriana Reus <adriana.reus@intel.com> Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Acked-by: Jonathan Cameron <jic23@kernel.org> Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Adriana Reus [Thu, 24 Mar 2016 09:29:30 +0000 (11:29 +0200)]
iio: accel: kxcjk-1013: use available_scan_masks
Use available_scan_masks to allow the iio core to select
the data to send to userspace depending on which axes are
enabled, instead of doing this in the driver's interrupt
handler.
Irina Tirdea [Thu, 24 Mar 2016 09:29:29 +0000 (11:29 +0200)]
iio: accel: bmg160: optimize transfers in trigger handler
Some i2c busses (e.g.: Synopsys DesignWare I2C adapter) need to
enable/disable the bus at each i2c transfer and must wait for
the enable/disable to happen before sending the data.
When reading data in the trigger handler, the bmg160 gyro driver does
one bus transfer for each axis. This has an impact on the frequency
of the accelerometer at high sample rates due to additional delays
introduced by the bus at each transfer.
Reading all axis values in one bus transfer reduces the delays
introduced by the bus.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Irina Tirdea [Thu, 24 Mar 2016 09:29:28 +0000 (11:29 +0200)]
iio: gyro: bmg160: use available_scan_masks
Use available_scan_masks to allow the iio core to select
the data to send to userspace depending on which axes are
enabled, instead of doing this in the driver's interrupt
handler.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Irina Tirdea [Thu, 24 Mar 2016 09:29:27 +0000 (11:29 +0200)]
iio: accel: bmc150: optimize transfers in trigger handler
Some i2c busses (e.g.: Synopsys DesignWare I2C adapter) need to
enable/disable the bus at each i2c transfer and must wait for
the enable/disable to happen before sending the data.
When reading data in the trigger handler, the bmc150 accel driver does
one bus transfer for each axis. This has an impact on the frequency
of the accelerometer at high sample rates due to additional delays
introduced by the bus at each transfer.
Reading all axis values in one bus transfer reduces the delays
introduced by the bus.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Irina Tirdea [Thu, 24 Mar 2016 09:29:26 +0000 (11:29 +0200)]
iio: accel: bmc150: use available_scan_masks
Use available_scan_masks to allow the iio core to select
the data to send to userspace depending on which axes are
enabled, instead of doing this in the driver's interrupt
handler.
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
Daeseok Youn [Thu, 31 Mar 2016 08:03:59 +0000 (17:03 +0900)]
staging: dgnc: clean up dgnc_input function
This is for fixing checkpatch.pl warning about
"Alignment should match open parenthesis" but if that is
fixed, code line is over 80 characters.
I think "ch->ch_rqueue + tail + i" could be declared once in
the begining of loop.
James Simmons [Thu, 31 Mar 2016 14:18:36 +0000 (10:18 -0400)]
staging: lustre: libcfs: move memory_pressure functions to libcfs_prim.h
Long ago libcfs_prim.h was used for userland code which is why
memory_pressure_*() handling is in both libcfs_prim.h and
linux-mem.h headers. So lets just move the memory_pressure_*()
to libcfs_prim.h.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6245
Reviewed-on: http://review.whamcloud.com/13841 Reviewed-by: frank zago <fzago@cray.com> Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com> Reviewed-by: John L. Hammond <john.hammond@intel.com> Reviewed-by: Oleg Drokin <oleg.drokin@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Vitaly Fertman [Wed, 30 Mar 2016 23:49:06 +0000 (19:49 -0400)]
staging/lustre/ldlm: Solve a race for LRU lock cancel
This patch solves a race condition that the lock may be used again
after LRU cancellation policy check. In that case, the lock may have
locked or dirty pages that makes the policy check totally useless.
The problem is solved by checking l_last_used at cancellation time
therefore it can make sure that the lock has not been used.
Vitaly Fertman [Wed, 30 Mar 2016 23:49:04 +0000 (19:49 -0400)]
staging/lustre/ldlm: restore the ELC for enqueue
after LU-4300 enqueue does not ELC anymore, however if enqueue is
agressive (ls -la of a large dir) we may exceed lru-resize limit
quickly because LRUR shrinker and recalc are called not so often.
ELC is to be restored in enqueue.
ELC also should check for the lock weight, in addition to LRUR.
ELC can also keep "skipped" locks, i.e. once checked for the weight
and left in the lru - let LRUR take care about them later.
LRUR is to be left untouched, no weight logic, otherwise LU-5727
appears and OPEN locks do not get canceled.
Oleg Drokin [Wed, 30 Mar 2016 23:49:01 +0000 (19:49 -0400)]
staging/lustre/llite: Remove unused vui_local_lock field
vvp_io_setattr_lock is the only user that sets it, but it's
never checked anywhere, so could go away.
Also get rid of enum ccc_setattr_lock_type that becomes unused.
John Hammond [Wed, 30 Mar 2016 23:49:00 +0000 (19:49 -0400)]
staging/lustre/llite: Move several declarations to llite_internal.h
Move several declarations between llite_internal.h and vvp_internal.h
with the goal of reserving the latter header for functions that
pertain to vvp_{device,object,page,...}.
Signed-off-by: John L. Hammond <john.hammond@intel.com> Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-on: http://review.whamcloud.com/13714
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5971 Reviewed-by: Bobi Jam <bobijam@hotmail.com> Reviewed-by: James Simmons <uja.ornl@gmail.com> Signed-off-by: Oleg Drokin <green@linuxhacker.ru> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
John Hammond [Wed, 30 Mar 2016 23:48:57 +0000 (19:48 -0400)]
staging/lustre/llite: rename struct ccc_thread_info to vvp_thread_info
struct ccc_thread_info is used in the VVP parts of llite so rename
it struct vvp_thread_info. Rename supporting functions accordingly.
Move init code from lcommon_cl.c to vvp_dev.c
Signed-off-by: John L. Hammond <john.hammond@intel.com> Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-on: http://review.whamcloud.com/13714
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5971 Reviewed-by: Bobi Jam <bobijam@hotmail.com> Reviewed-by: James Simmons <uja.ornl@gmail.com> Signed-off-by: Oleg Drokin <green@linuxhacker.ru> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
John L. Hammond [Wed, 30 Mar 2016 23:48:51 +0000 (19:48 -0400)]
staging/lustre/llite: merge ccc_io and vvp_io
Move the contents of struct vvp_io into struct ccc_io, delete the
former, and rename the latter to struct vvp_io. Rename various ccc_io
related functions to use vvp rather than ccc.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/13351
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5971 Reviewed-by: Lai Siyao <lai.siyao@intel.com> Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com> Signed-off-by: Oleg Drokin <green@linuxhacker.ru> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
John L. Hammond [Wed, 30 Mar 2016 23:48:50 +0000 (19:48 -0400)]
staging/lustre:llite: remove struct ll_ra_read
Ever since removal of the the unused function ll_ra_read_get(),
the struct ll_ra_read members lrr_reader and lrr_linkage and
the struct ll_readahead_state member ras_read_beads unnecessary
so remove them.
In struct vvp_io replace the struct ll_ra_read cui_bead member with
cui_ra_start and cui_ra_count.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/13347
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5971 Reviewed-by: Bobi Jam <bobijam@hotmail.com> Reviewed-by: Lai Siyao <lai.siyao@intel.com> Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com> Signed-off-by: Oleg Drokin <green@linuxhacker.ru> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / linux-beck.git / log