- 23 Nov, 2018 8 commits
-
-
Gao Xiang authored
Let's make sure that the one registering a workgroup will also take the primary work lock at first for two reasons: 1) There's no need to introduce such a race window (and consequently overhead) between registering and locking, other tasks could break in by chance, and the race seems unnecessary (no benefit at all); 2) It's better to take the primary work when a workgroup is registered to apply the cache managed policy, for example, if some other tasks break in, it could turn into the in-place decompression rather than use as the cached decompression. Reviewed-by:Chao Yu <yuchao0@huawei.com> Signed-off-by:
Gao Xiang <gaoxiang25@huawei.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Gao Xiang authored
`z_erofs_vle_workgroup' is heavily generated in the decompression, for example, it resets 32 bytes redundantly for 64-bit platforms even through Z_EROFS_VLE_INLINE_PAGEVECS + Z_EROFS_CLUSTER_MAX_PAGES, default 4, pages are stored in `z_erofs_vle_workgroup'. As an another example, `struct mutex' takes 72 bytes for our kirin 64-bit platforms, it's unnecessary to be reseted at first and be initialized each time. Let's avoid filling all `z_erofs_vle_workgroup' with 0 at first since most fields are reinitialized to meaningful values later, and pagevec is no need to initialized at all. Reviewed-by:
-
Gao Xiang authored
Just like other generic locks, insert a full barrier in case of memory reorder. Reviewed-by:
-
Gao Xiang authored
There are two minor issues in the current freeze interface: 1) Freeze interfaces have not related with CONFIG_DEBUG_SPINLOCK, therefore fix the incorrect conditions; 2) For SMP platforms, it should also disable preemption before doing atomic_cmpxchg in case that some high priority tasks preempt between atomic_cmpxchg and disable_preempt, then spin on the locked refcount later. Reviewed-by: -
Gao Xiang authored
It's better to use atomic_cond_read_relaxed, which is implemented in hardware instructions to monitor a variable changes currently for ARM64, instead of open-coded busy waiting. Reviewed-by:
-
Gao Xiang authored
When the managed cache is enabled, the last reference count of a workgroup must be used for its workstation. Otherwise, it could lead to incorrect (un)freezes in the reclaim path, and it would be harmful. A typical race as follows: Thread 1 (In the reclaim path) Thread 2 workgroup_freeze(grp, 1) refcnt = 1 ... workgroup_unfreeze(grp, 1) refcnt = 1 workgroup_get(grp) refcnt = 2 (x) workgroup_put(grp) refcnt = 1 (x) ...unexpected behaviors * grp is detached but still used, which violates cache-managed freeze constraint. Reviewed-by: -
Gao Xiang authored
It's better not to positively BUG_ON the kernel, however developers need a way to locate issues as soon as possible. DBG_BUGON is introduced and it could only crash when EROFS_FS_DEBUG (EROFS developping feature) is on. It is helpful for developers to find and solve bugs quickly by eng builds. Previously, DBG_BUGON is defined as ((void)0) if EROFS_FS_DEBUG is off, but some unused variable warnings as follows could occur: drivers/staging/erofs/unzip_vle.c: In function `init_alway:': drivers/staging/erofs/unzip_vle.c:61:33: warning: unused variable `work' [-Wunused-variable] struct z_erofs_vle_work *const work = ^~~~ Fix it to #define DBG_BUGON(x) ((void)(x)). Reviewed-by: -
Gao Xiang authored
`trace_erofs_readpage' was designed for .readpage() interface hook in order to trace the detailed synchronized read rather than the internal implementation `z_erofs_do_read_page' which both .readpage() and .readpages() use. It seems the tracepoint was placed to a wrong place by mistake and it should be moved to the right place. Fixes: 284db12c ("staging: erofs: add trace points for reading zipped data") Reviewed-by:
Chen Gong <gongchen4@huawei.com> Reviewed-by:
-
- 22 Nov, 2018 11 commits
-
-
Michael Straube authored
Rename struct field SwLed0 to avoid CamelCase. SwLed0 -> sw_led Signed-off-by:
Michael Straube <straube.linux@gmail.com> Signed-off-by:
-
Michael Straube authored
Rename the function LedControl8188eu to avoid CamelCase. LedControl8188eu -> led_control_8188eu Signed-off-by:
-
Michael Straube authored
The comment for blink_handler() adds no extra information in addition to the function name, so remove it. Signed-off-by:
-
Michael Straube authored
Rename the function BlinkHandler to avoid CamelCase. BlinkHandler -> blink_handler Signed-off-by:
-
Michael Straube authored
Rename the function SwLedOff to avoid CamelCase. SwLedOff -> sw_led_off Signed-off-by:
-
Michael Straube authored
Rename the function SwLedOn to avoid CamelCase. SwLedOn -> sw_led_on Signed-off-by:
-
Michael Straube authored
Rename the variable LedCfg to avoid CamleCase. LedCfg -> led_cfg Signed-off-by:
-
Michael Straube authored
Add spaces around '&' to follow kernel coding style. Reported by checkpatch. Signed-off-by:
-
Michael Straube authored
Remove comments in rtl8188eu_led.c. The function names are well descriptive. Signed-off-by:
-
Michael Straube authored
Remove unnecessary parentheses reported by checkpatch. Signed-off-by:
-
Greg Kroah-Hartman authored
Merge tag 'iio-for-4.21a' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-testing Jonathan writes: First set of new device support, features and cleanups for IIO in the 4.21 cycle Along with the headline feature of 5 new drivers, we have the substantial addition of auxilliary sensor support on the lsm6sdx parts for ST. There has also been a good set of staging cleanup in this period with more underway. An ever increasing number of devices supported with just a new ID which is a good sign that at least some manufacturers are continuing to stabilise their interfaces. New device support, * ad7124 - New driver supporting Analog Devices' ad7124-4 and ad7124-8 parts with the inevitable DT binding. * ad7949 - New driver supporting Analog Devices' ad7949, AD7682 and AD7689 ADCs. * rm3100 - New driver supporting PNIs RM3100 magnometer with bindings and vendor prefix. * ti-dac7311 - New driver supporting DAC7311, DAC6311 and DAC5311 TI DACs, with DT bindings. * vcnl5035 - New driver supporting the light sensor part of the VCNL4035, with DT bindings Features, * bindings - Add a generic ADC channel binding as we keep reinventing this wheel. * adc128s052 - Add IDs for additional pin compatible parts. - Add APCI ID seen on E3940 UP squared boards. * ad_sigma_delta - Allow for custom data register overiding default. * kxcjk1013 - Add KIOX0009 ACPI ID as seen on the Acer One 10. * lsm6dsx - Rework leading to... - External sensor support using the built in I2C master. - Initial support for a slave lis2mdl magnetometer. * meson-saradc - Add temperature sensor support and bindings. * st_magn - New ID for lsm9dsl_magn with bindings - New ID for lis3de accelerometer * tpl0102 - Add supprot for IIO_AVAIL_RANGE to report the range available from this device to userspace and in kernel users. Cleanups and minor fixes * tools - Allow outside specification of CFLAGS * ad2s90 - Handle and spi_read error. - Handle spi_setup failure - Drop a pointless assignment. - Prevent a potentail race by moving device registration to after all other setup. - Add missing scale attribute. - Add a sanity check on channel type before trying to read it. * ad2s1210 - Move to modern gpio descriptors. - Drop a gpioin flag which made no sense as far as we can tell. - Add dt table (bindings doc to follow when this is ready for moving out of staging). * ad5933 - Drop camel-case naming of ext_clk_hz. - White space fixes. * ad7150 - Local variable to shorten overly long line. - Alignment and line break fixes. * ad7280a - Handle an error path that was previously ignored. - Use crc8.h to build the crc table replacing custom code. - Avoid unecessary cast. - Power down the device if an error happens in probe - Use devm routines to simplify probe and remove. * ad7606 - Alignment fixes. * ad7780 - This worked as long as by coincidence an uninitialized value was 0. Lets not rely on that. - Ensure gain update is only used with the ad778x chips that actually support it. - Tidy up pattern mask generation. - Read regulator when scale is requested (which should be infrequent) as it might have changed from initialization. * ad7816 - Move to modern gpio descriptors - Don't use a busy_pin for ad7818 as there isn't one. - Ensure RD/WR and CONVST pins are outputs (previously they were brought up as inputs which doesn't seem to make any sense) - DT id table. * adc128s052 - SPDX * adt7316 - Alignment fix. - Fix data reading. When using I2C the driver never actually used the value read. This has been broken a very long time hence no rush to fix it now + the driver is undergoing a lot of cleanup. - Sanity check that the i2c read didn't fail to actually read anything. * dpot-dac - Mark a switch full through with slightly different text so that gcc doesn't warn on it. * gyro-adc - Fix a wrong file in the MAINTAINERS entry and add binding doc to the listed files. * ina2xx - Add some early returns to clarify error paths in switch. * lsm6dsx - MAINTAINERS entry. * max11100 - SPDX * max9611 - SPDX * mcp4131 - use of_device_get_match_data in preference to spi_get_device_id approach. * rcar-adc - SPDX * sc27xx - Add ADC conversion timeout support to avoid possible fault. * ssp_sensors - Don't free managed resources manually. * st-magn - Add a comment to avoid future confusion over when to use -magn postfix (on multi chip in package parts) - Add BDU register for LIS3MDL where it seems to have been missed. * st-sensors - Minor spelling, grammar etc fixes. * tpl0102 - Use a pointer rather than an index of an array to improve conciseness. * tag 'iio-for-4.21a' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio: (80 commits) Staging: iio: adt7316: Add an extra check for 'ret' equals to 0 Staging: iio: adt7316: Fix i2c data reading, set the data field dt-bindings: iio: adc: Add docs for ad7124 iio: adc: Add ad7124 support dt-bindings: iio: adc: Add common ADCs properties to a separate file iio: ad_sigma_delta: Allow to provide custom data register address staging: iio: ad7816: Add device tree table. iio: imu: st_lsm6dsx: add entry in MAINTAINERS file iio: potentiometer: mcp4131: use of_device_get_match_data() staging: iio: adc: ad7280a: use devm_* APIs staging: iio: adc: ad7280a: power down the device on error in probe dt-bindings: iio: imu: st_lsm6dsx: add support to i2c pullup resistors iio: imu: st_lsm6dsx: add hw FIFO support to i2c controller iio: imu: st_lsm6dsx: add st_lsm6dsx_push_tagged_data routine iio: imu: st_lsm6dsx: add i2c embedded controller support iio: imu: st_lsm6dsx: introduce st_lsm6dsx_sensor_set_enable routine iio: imu: st_lsm6dsx: introduce ST_LSM6DSX_ID_EXT sensor ids iio: imu: st_lsm6dsx: remove static from st_lsm6dsx_set_watermark iio: imu: st_lsm6dsx: reload trimming parameter at bootstrap iio: imu: st_lsm6dsx: introduce locked read/write utility routines ...
-
- 20 Nov, 2018 14 commits
-
-
André Almeida authored
Using multiple license representation as shown here [1]. [1] https://spdx.org/sites/cpstandard/files/pages/files/using_spdx_license_list_short_identifiers.pdfSigned-off-by:
André Almeida <andrealmeid@riseup.net> Acked-by:
Stefan Wahren <stefan.wahren@i2se.com> Signed-off-by:
-
André Almeida authored
As one can see in a typedef at linux/types.h:103, s32 and int32_t are both a 32 bits signed integer. s32 is preferred because it is defined by the kernel maintainers and it prevents portability issues. Changed to meet checkpatch criteria. Signed-off-by:
-
Cristian Sicilia authored
Align parameters to the opened parentesis. Signed-off-by:
Cristian Sicilia <sicilia.cristian@gmail.com> Reviewed-by:
-
Cristian Sicilia authored
Comparisons should place the constant on the right side of the test. Signed-off-by:
-
Cristian Sicilia authored
Replace equal to NULL with logic unary operator, and removing not equal to NULL comparison. Signed-off-by:
-
Tim Collier authored
Remove the unused crcend variable from the crcimage function. Signed-off-by:
Tim Collier <osdevtc@gmail.com> Signed-off-by:
-
Cristian Sicilia authored
Fix function naming and parenthesis. Signed-off-by:
-
Cristian Sicilia authored
The line is split up to stay in 80 characters- Signed-off-by:
-
Josenivaldo Benito Jr authored
Correct following errors reported by checkpath.pl: ERROR: space required before the open parenthesis '(' #265: FILE: drivers/staging/rtl8723bs/hal/rtl8723bs_recv.c:265: + if(!precvframe) ')' Also similar errors on line 274 and 283. Signed-off-by:Josenivaldo Benito Jr <jrbenito@benito.qsl.br> Signed-off-by:
-
Cristian Sicilia authored
The parameter is sent to next line to stay in 80 characters Signed-off-by:
-
Cristian Sicilia authored
Added before parameters some spaces to match open parenthesis on previous line. Signed-off-by:
-
Michael Straube authored
The SPDX identifier is a legally binding shorthand, which can be used instead of the full boiler plate text. All source and header files have SPDX identifiers. Remove the GPL boiler plate text to reduce file size. Signed-off-by:
-
Greg Kroah-Hartman authored
This reverts commit 57d05368. Johan writes: This commit doesn't even do what it claims. It leaves the two spaces before tabs in place and instead removes a tab that was used to indent a continuation line, and therefore also hurts readability while introducing an inconsistency. Reported-by:
Johan Hovold <johan@kernel.org> Cc: Ioannis Valasakis <code@wizofe.uk> Signed-off-by:
-
Greg Kroah-Hartman authored
This reverts commit fcb733b5. Johan writes: This patch removes the double newlines that were used as section headers in the lengthy protocol header file, and therefore hurts readability. So I'm reverting it. Reported-by:
-
- 17 Nov, 2018 6 commits
-
-
Shreeya Patel authored
ret = 0 indicates a case of no error but no data read from the bus which is an invalid case. This case doesn't ever happen in reality. It should perhaps be handled for correctness though. Signed-off-by:
Shreeya Patel <shreeya.patel23498@gmail.com> Signed-off-by:
Jonathan Cameron <Jonathan.Cameron@huawei.com>
-
Shreeya Patel authored
adt7316_i2c_read function nowhere sets the data field. It is necessary to have an appropriate value for it. Hence, assign the value stored in 'ret' variable to data field. This is an ancient bug, and as no one seems to have noticed, probably no sense in applying it to stable. Signed-off-by:
-
Stefan Popa authored
Add support for Analog Devices AD7124 4-channels and 8-channels ADC. Signed-off-by:
Stefan Popa <stefan.popa@analog.com> Reviewed-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Stefan Popa authored
The ad7124-4 and ad7124-8 are a family of 4 and 8 channel sigma-delta ADCs with 24-bit precision and reference. Three power modes are available which in turn affect the output data rate: * Full power: 9.38 SPS to 19,200 SPS * Mid power: 2.34 SPS to 4800 SPS * Low power: 1.17 SPS to 2400 SPS The ad7124-4 can be configured to have four differential inputs, while ad7124-8 can have 8. Moreover, ad7124 also supports per channel configuration. Each configuration consists of gain, reference source, output data rate and bipolar/unipolar configuration. Datasheets: Link: http://www.analog.com/media/en/technical-documentation/data-sheets/AD7124-4.pdf Link: http://www.analog.com/media/en/technical-documentation/data-sheets/ad7124-8.pdfSigned-off-by:
-
Stefan Popa authored
There are several ADC drivers that depend on the same device tree bindings. Rather than continue to duplicate the properties, this patch adds a common adc binding document that can be referenced. For beginning, only two properties are documented. Signed-off-by:
-
Lars-Peter Clausen authored
Some newer devices from the Sigma-Delta ADC family do have their data register at a different address than the current default address. Add a parameter to the ad_sigma_delta_info struct which allows to override the default address. Signed-off-by:
Lars-Peter Clausen <lars@metafoo.de> Signed-off-by:
-
- 16 Nov, 2018 1 commit
-
-
Nishad Kamdar authored
Add device tree table for matching vendor ID. Signed-off-by:
Nishad Kamdar <nishadkamdar@gmail.com> Signed-off-by:
-
