Describe the inactivity detection additionally using the free-fall
register. Due to the controversial discussions on the mailing list, this
section of the documentation will be committed separately to allow for a
more focused and detailed elaboration of the topic.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-8-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Inactivity and free-fall events are essentially the same type of sensor
events. Therefore, inactivity detection (normally set for periods between 1
and 255 seconds) can be extended for shorter durations to support free-fall
detection.
For periods shorter than 1 second, the driver automatically configures the
threshold and duration using the free-fall register. For periods longer
than 1 second, it uses the inactivity threshold and duration using the
inactivity registers.
When using the free-fall register, the link bit is not set, which means
auto-sleep cannot be enabled if activity detection is also active.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-6-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Enable AC/DC coupling configuration for activity and inactivity detection
by setting the AC/DC bit. Extend existing magnitude-based detection with
adaptive AC-coupled mode.
Use DC-coupled mode to compare acceleration samples directly against
configured thresholds. Use AC-coupled mode to compare samples against a
reference taken at the start of activity detection. Implement DC-coupled
events using MAG, and AC-coupled events using MAG_ADAPTIVE.
Expose configuration of thresholds and periods via separate sysfs handles.
Note that both coupling modes share the same sensor registers, so activity
or inactivity detection cannot be configured for both AC and DC
simultaneously. Apply the most recently configured mode.
Simplify event handling and support adaptive AC-coupling.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-5-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Add support for the sensor’s inactivity feature in the driver. When both
activity and inactivity detection are enabled, the sensor sets a link bit
that ties the two functions together. This also enables auto-sleep mode,
allowing the sensor to automatically enter sleep state upon detecting
inactivity.
Inactivity detection relies on a configurable threshold and a specified
time period. If sensor measurements remain below the threshold for the
defined duration, the sensor transitions to the inactivity state.
When an Output Data Rate (ODR) is set, the inactivity time period is
automatically adjusted to a sensible default. Higher ODRs result in shorter
inactivity timeouts, while lower ODRs allow longer durations-within
reasonable upper and lower bounds. This is important because features like
auto-sleep operate effectively only between 12.5 Hz and 400 Hz. These
defaults are applied when the sample rate is modified, but users can
override them by explicitly setting a custom inactivity timeout.
Similarly, configuring the g-range provides default threshold values for
both activity and inactivity detection. These are implicit defaults meant
to simplify configuration, but they can also be manually overridden as
needed.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-4-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Enable the sensor to detect activity and trigger interrupts accordingly.
Activity events are determined based on a threshold, which is initialized
to a sensible default during probe. This default value is adopted from the
legacy ADXL345 input driver to maintain consistent behavior.
The combination of activity detection, ODR configuration, and range
settings lays the groundwork for the activity/inactivity hysteresis
mechanism, which will be implemented in a subsequent patch. As such,
portions of this patch prepare switch-case structures to support those
upcoming changes.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-3-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Simplify setting the tap suppress bit by applying regmap_assign_bit(). As a
result, the defines can be reorganized for better clarity.
Signed-off-by: Lothar Rubusch <l.rubusch@gmail.com>
Link: https://patch.msgid.link/20250727210014.27766-2-l.rubusch@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
There is no txt file for it, add yaml for invensense,icp10100
which is already used in the driver. Also, document other compatibles
for ICP-101xx family and add invensense,icp10100 as a fallback.
Acked-by: Jean-Baptiste Maneyrol <jean-baptiste.maneyrol@tdk.com>
Signed-off-by: Rodrigo Gobbi <rodrigo.gobbi.7@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://patch.msgid.link/20250801221424.132305-1-rodrigo.gobbi.7@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Add support for debugfs_reg_access through the driver's iio_info structure
to enable low-level register read/write access for debugging.
Signed-off-by: Akshay Jindal <akshayaj.lkd@gmail.com>
Link: https://patch.msgid.link/20250728161445.13261-1-akshayaj.lkd@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Put the filter types in alphabetical order by name. This makes it easier
to find a specific filter type when looking through the documentation.
Signed-off-by: David Lechner <dlechner@baylibre.com>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20250710-iio-adc-ad7137-add-filter-support-v1-4-acffe401c4d2@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Drop the call to INIT_LIST_HEAD(&cb_buff->buffer.demux_list). That's
already done in iio_buffer_init(&cb_buff->buffer).
Signed-off-by: Nuno Sá <nuno.sa@analog.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Link: https://patch.msgid.link/20250725-iio-minor-cleanup-v1-1-4e561372142e@analog.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Driver is getting clock and almost immediately enabling it, with the
devm_request_irq() as the only relevant code executed between, thus the
probe path and cleanups can be simplified with devm_clk_get_enabled().
Move devm_request_irq() earlier, so the interrupt handler will be
registered before clock is enabled. This might be important in case
regulator supplies are enabled by other device driver and this device
raises interrupt immediately after clock sarts ticking.
The change does not reverse cleanup paths - first regulator will be
disabled, then clock and finally interrupt handler freed.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://patch.msgid.link/20250713-iio-clk-get-enabled-v1-1-70abc1f9ce6c@linaro.org
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
dynaimage,al3010 and dynaimage,al3320a are I2C devices, so their
interrupts are not routed to GIC and rarely first interrupt cell is <0>.
This looks like copy-paste from some SoC block, so drop it and also for
readability use known defines instead of hard-coding interrupt flag.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: David Heidelberg <david@ixit.cz>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250724111345.47889-8-krzysztof.kozlowski@linaro.org
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
The interrupt line to GIC is IRQ_TYPE_LEVEL_HIGH, so use that instead of
"none". Also replace the hard-coded GIC_SPI flag.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com>
Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250724111345.47889-7-krzysztof.kozlowski@linaro.org
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
For readability use known defines instead of hard-coding GPIO or
interrupt flag.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Acked-by: Matti Vaittinen <mazziesaccount@gmail.com>
Link: https://patch.msgid.link/20250724111345.47889-6-krzysztof.kozlowski@linaro.org
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Drop includes of headers which example code does not use. No functional
impact.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: David Heidelberg <david@ixit.cz>
Acked-by: Matti Vaittinen <mazziesaccount@gmail.com>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250724111345.47889-5-krzysztof.kozlowski@linaro.org
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Change the ti-adc081c driver to use individual model structures instead
of an array. This reduces the verbosity of the code. Also, the data is
now const as it should have been in the first place. The ADCxx1C_MODEL()
macro is dropped to be consistent with similar model definitions in
other drivers.
Signed-off-by: David Lechner <dlechner@baylibre.com>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20250721-iio-const-data-11-v2-1-c3fec12511ee@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Move the scan struct to the stack instead of being in the driver state
struct. The buffer is only used in a single function and does not need
to be DMA-safe so it does not need to exist outside of that function's
scope.
Signed-off-by: David Lechner <dlechner@baylibre.com>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20250721-iio-use-more-iio_declare_buffer_with_ts-v2-1-f8fb11b8add8@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Use a stack allocated scan struct in srf08_trigger_handler(). Since the
scan buffer isn't used outside of this function and doesn't need to be
DMA-safe, it doesn't need to be in struct srf08_data. We can also
eliminate an extra local variable for the return value of
srf08_read_ranging() by using scan.chan directly.
Reviewed-by: Andreas Klinger <ak@it-klinger.de>
Signed-off-by: David Lechner <dlechner@baylibre.com>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20250721-iio-use-more-iio_declare_buffer_with_ts-6-v2-1-8b66e5b4e75a@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Use a stack-allocated struct in sx9500_trigger_handler() to hold the
IIO buffer scan data. Since the scan buffer isn't used outside of this
function, it doesn't need to be in struct sx9500_data.
By always allocating enough space for the maximum number of channels,
we can avoid having to reallocate the buffer each time buffered reads
are enabled.
Signed-off-by: David Lechner <dlechner@baylibre.com>
Link: https://patch.msgid.link/20250722-iio-use-more-iio_declare_buffer_with_ts-4-v2-1-9e566f3a4c6a@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Use a stack allocated struct for the scan data instead of using the
driver state to store the struct. The scan data is not used outside of
the interrupt handler function so the struct does not need to exist
outside of that scope.
Signed-off-by: David Lechner <dlechner@baylibre.com>
Link: https://patch.msgid.link/20250722-iio-proximity-mb1232-use-stack-allocated-scan-struct-v1-1-b4ef77e9ddea@baylibre.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
The SARADC controller in most Rockchip SoCs are part of power domains
that are always powered on, i.e. PD_BUS or PD_PERI. These always powered
on power domains have typically not been described in the device tree.
Because these power domains have been left out of the device tree there
has not been any real need to properly describe the power domain of the
SARADC controller.
On RK3528 the SARADC controller is part of the PD_VPU power domain.
Add support to describe an optional power-domains for the SARADC
controller in Rockchip SoCs.
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250723085654.2273324-4-jonas@kwiboo.se
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
avg sample info is a bit field coded inside the following
bits: 5,6,7 and 8 of a device status register.
Channel num info the same, but over bits: 1, 2 and 3.
Mask both values in order to avoid touching other register bits,
since the first info (avg sample), came from DT.
Signed-off-by: Rodrigo Gobbi <rodrigo.gobbi.7@gmail.com>
Reviewed-by: David Lechner <dlechner@baylibre.com>
Link: https://patch.msgid.link/20250717221559.158872-1-rodrigo.gobbi.7@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Use kvfree() to fix the following Coccinelle/coccicheck warning reported
by kfree_mismatch.cocci:
WARNING kvmalloc is used to allocate this memory at line 10398
Fixes: f728c17fc9 ("accel/habanalabs/gaudi2: move HMMU page tables to device memory")
Reported-by: Qianfeng Rong <rongqianfeng@vivo.com>
Closes: https://patch.msgid.link/20250808085530.233737-1-rongqianfeng@vivo.com
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
[lukas: acknowledge Qianfeng, adjust Thorsten's domain, add Fixes tag]
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Cc: stable@vger.kernel.org # v6.9+
Link: https://patch.msgid.link/20240820231028.136126-1-thorsten.blum@toblux.com
Remove the redundant __GFP_ZERO flag from kzalloc() since kzalloc()
inherently zeroes memory.
Signed-off-by: Qianfeng Rong <rongqianfeng@vivo.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The NEED_* macros do an implicit goto in case the safety bounds checks
fail. Add the unlikely hints as this is the error case and not a hot
path. The final assembly is slightly shorter and some jumps got
reordered according to the hints.
text data bss dec hex filename
2294 16 0 2310 906 pre/lzo1x_decompress_safe.o
2277 16 0 2293 8f5 post/lzo1x_decompress_safe.o
text data bss dec hex filename
3444 48 0 3492 da4 pre/lzo1x_compress_safe.o
3372 48 0 3420 d5c post/lzo1x_compress_safe.o
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
It is better to replace ns_to_ktime() with us_to_ktime(),
which can make the code clearer.
Signed-off-by: Xichao Zhao <zhao.xichao@vivo.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Compile-testing this driver is only possible when the AMBA bus driver is
available in the kernel:
x86_64-linux-ld: drivers/char/hw_random/nomadik-rng.o: in function `nmk_rng_remove':
nomadik-rng.c:(.text+0x67): undefined reference to `amba_release_regions'
x86_64-linux-ld: drivers/char/hw_random/nomadik-rng.o: in function `nmk_rng_probe':
nomadik-rng.c:(.text+0xee): undefined reference to `amba_request_regions'
x86_64-linux-ld: nomadik-rng.c:(.text+0x18d): undefined reference to `amba_release_regions'
The was previously implied by the 'depends on ARCH_NOMADIK', but needs to be
specified for the COMPILE_TEST case.
Fixes: d5e93b3374 ("hwrng: Kconfig - Add helper dependency on COMPILE_TEST")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The SNP_VLEK_LOAD IOCTL currently fails due to sev_cmd_buffer_len()
returning the default expected buffer length of 0 instead of the correct
value, which would be sizeof(struct sev_user_data_snp_vlek_load). Add
specific handling for SNP_VLEK_LOAD so the correct expected size is
returned.
Reported-by: Diego GonzalezVillalobos <Diego.GonzalezVillalobos@amd.com>
Cc: Diego GonzalezVillalobos <Diego.GonzalezVillalobos@amd.com>
Fixes: 332d2c1d71 ("crypto: ccp: Add the SNP_VLEK_LOAD command")
Signed-off-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
sg_nents_for_len() returns an int which is negative in case of error.
Fixes: 472b04444c ("crypto: keembay - Add Keem Bay OCS HCU driver")
Signed-off-by: Thomas Fourier <fourier.thomas@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Add slice-specific command queue counters for QAT GEN6 devices to monitor
utilization metrics, including wait time, execution duration, and release
events.
Update the documentation to reflect the new command queue counter
functionality.
Co-developed-by: George Abraham P <george.abraham.p@intel.com>
Signed-off-by: George Abraham P <george.abraham.p@intel.com>
Signed-off-by: Vijay Sundar Selvamani <vijay.sundar.selvamani@intel.com>
Signed-off-by: Suman Kumar Chakraborty <suman.kumar.chakraborty@intel.com>
Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Add a new performance counter that measures the average ring buffer idle
duration.
This metric is now included in the telemetry counters exposed via
debugfs for QAT GEN6 devices.
Update the documentation to reflect the new idle duration counter
Co-developed-by: George Abraham P <george.abraham.p@intel.com>
Signed-off-by: George Abraham P <george.abraham.p@intel.com>
Signed-off-by: Vijay Sundar Selvamani <vijay.sundar.selvamani@intel.com>
Signed-off-by: Suman Kumar Chakraborty <suman.kumar.chakraborty@intel.com>
Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
There is a spelling mistake in the module description text. Fix it.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Acked-by: Maxime Méré <maxime.mere@foss.st.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
To enable ciphertext hiding, it must be specified in the SNP_INIT_EX
command as part of SNP initialization.
Modify the sev_platform_init_args structure, which is used as input to
sev_platform_init(), to include a field that, when non-zero,
indicates that ciphertext hiding should be enabled and specifies the
maximum ASID that can be used for an SEV-SNP guest.
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ashish Kalra <ashish.kalra@amd.com>
Reviewed-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Implement an API that checks the overall feature support for SEV-SNP
ciphertext hiding.
This API verifies both the support of the SEV firmware for the feature
and its enablement in the platform's BIOS.
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ashish Kalra <ashish.kalra@amd.com>
Reviewed-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The FEATURE_INFO command provides hypervisors with a programmatic means
to learn about the supported features of the currently loaded firmware.
This command mimics the CPUID instruction relative to sub-leaf input and
the four unsigned integer output values. To obtain information
regarding the features present in the currently loaded SEV firmware,
use the SNP_FEATURE_INFO command.
Cache the SNP platform status and feature information from CPUID
0x8000_0024 in the sev_device structure. If SNP is enabled, utilize
this cached SNP platform status for the API major, minor and build
version.
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ashish Kalra <ashish.kalra@amd.com>
Reviewed-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Cache the SEV platform status into sev_device structure and use this
cached SEV platform status for api_major/minor/build.
The platform state is unique between SEV and SNP and hence needs to be
tracked independently.
Remove the state field from sev_device structure and instead track SEV
state from the cached SEV platform status.
Suggested-by: Tom Lendacky <thomas.lendacky@amd.com>
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ashish Kalra <ashish.kalra@amd.com>
Reviewed-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Define new bit-field definitions returned by SNP_PLATFORM_STATUS command
such as new capabilities like SNP_FEATURE_INFO command availability,
ciphertext hiding enabled and capability.
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ashish Kalra <ashish.kalra@amd.com>
Reviewed-by: Kim Phillips <kim.phillips@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
I will be leaving Intel soon, Koby Elbaz & Konstantin Sinyuk will take
the role of habanalabs driver maintainers.
Signed-off-by: Yaron Avizrat <yaron.avizrat@intel.com>
Reviewed-by: Koby Elbaz <koby.elbaz@intel.com>
Reviewed-by: Konstantin Sinyuk <konstantin.sinyuk@intel.com>
Link: https://patch.msgid.link/DM4PR11MB55491ACAA33DF29CEF3C67DAE935A@DM4PR11MB5549.namprd11.prod.outlook.com
[lukas: wrap to 72 chars]
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Drop 'inline' from all the *_mod_init_arch() functions so that the
compiler will warn about any bugs where they are unused due to not being
wired up properly. (There are no such bugs currently, so this just
establishes a more robust convention for the future. Of course, these
functions also tend to get inlined anyway, regardless of the keyword.)
Link: https://lore.kernel.org/r/20250816020240.431545-1-ebiggers@kernel.org
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Tony Nguyen says:
====================
ice: implement SRIOV VF Active-Active LAG
Dave Ertman says:
Implement support for SRIOV VFs over an Active-Active link aggregate.
The same restrictions apply as are in place for the support of
Active-Backup bonds.
- the two interfaces must be on the same NIC
- the FW LLDP engine needs to be disabled
- the DDP package that supports VF LAG must be loaded on device
- the two interfaces must have the same QoS config
- only the first interface added to the bond will have VF support
- the interface with VFs must be in switchdev mode
With the additional requirement of
- the version of the FW on the NIC needs to have VF Active/Active support
The balancing of traffic between the two interfaces is done on a queue
basis. Taking the queues allocated to all of the VFs as a whole, one
half of them will be distributed to each interface. When a link goes
down, then the queues allocated to the down interface will migrate to
the active port. When the down port comes back up, then the same
queues as were originally assigned there will be moved back.
Patch 1 cleans up void pointer casts
Patch 2 utilizes bool over u8 when appropriate
Patch 3 adds a driver prefix to a LAG define
Patch 4 pre-move a function to reduce delta in implementation patch
Patch 5 cleanup variable initialization in declaration block
Patch 6 cleanup capability parsing for LAG feature
Patch 7 is the implementation of the new functionality
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue:
ice: Implement support for SRIOV VFs across Active/Active bonds
ice: cleanup capabilities evaluation
ice: Cleanup variable initialization in LAG code
ice: move LAG function in code to prepare for Active-Active
ice: Add driver specific prefix to LAG defines
ice: replace u8 elements with bool where appropriate
ice: Remove casts on void pointers in LAG code
====================
Link: https://patch.msgid.link/20250814230855.128068-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The CI has hit a couple of cases of:
RUN global.data_steal ...
tls.c:2762:data_steal:Expected recv(cfd, buf2, sizeof(buf2), MSG_DONTWAIT) (20000) == -1 (-1)
data_steal: Test terminated by timeout
FAIL global.data_steal
Looks like the 2msec sleep is not long enough. Make the sleep longer,
and then instead of second sleep wait for the thieving process to exit.
That way we can be sure it called recv() before us.
While at it also avoid trying to steal more than a record, this seems
to be causing issues in manual testing as well.
Fixes: d7e82594a4 ("selftests: tls: test TCP stealing data from under the TLS socket")
Link: https://patch.msgid.link/20250814194323.2014650-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
During raid resync, if a disk becomes faulty, the operation is
briefly interrupted. The MD_RECOVERY_RECOVER flag triggered by
the disk failure causes sync_action to incorrectly show "recover"
instead of "resync". The same issue affects reshape operations.
Reproduction steps:
mdadm -Cv /dev/md1 -l1 -n4 -e1.2 /dev/sd{a..d} // -> resync happened
mdadm -f /dev/md1 /dev/sda // -> resync interrupted
cat sync_action
-> recover
Add progress checks in md_sync_action() for resync/recover/reshape
to ensure the interface correctly reports the actual operation type.
Fixes: 4b10a3bc67 ("md: ensure resync is prioritized over recovery")
Signed-off-by: Zheng Qixing <zhengqixing@huawei.com>
Link: https://lore.kernel.org/linux-raid/20250816002534.1754356-3-zhengqixing@huaweicloud.com
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
