| Age | Commit message (Collapse) | Author |
|---|
|
Add Intel uncore frequency driver support for Pantherlake, Wildcatlake
and Novalake processors.
Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Link: https://patch.msgid.link/20251022211733.3565526-1-sathyanarayanan.kuppuswamy@linux.intel.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
|
|
The newer HP Omen laptops, such as Omen 16-wf1xxx, use the same
WMI-based thermal profile interface as Victus 16-r1000 and 16-s1000
models.
Add the DMI board name "8C78" to the victus_s_thermal_profile_boards
list to enable proper fan and thermal mode control.
Tested on: HP Omen 16-wf1xxx (board 8C78)
Result:
* Fan RPMs are readable
* echo 0 | sudo tee /sys/devices/platform/hp-wmi/hwmon/*/pwm1_enable
allows the fans to run on max RPM.
Signed-off-by: Krishna Chomal <krishna.chomal108@gmail.com>
Link: https://patch.msgid.link/20251018111001.56625-1-krishna.chomal108@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
|
|
HONOR MagicBook X16/X14 models produced in 2025 cannot use the Print
Screen and YOYO keys properly, with the system reporting them as
unknown key presses (codes: 0x028b and 0x028e).
To resolve this, a key_entry is added for both the HONOR Print Screen
key and the HONOR YOYO key, ensuring they function correctly on these
models.
Signed-off-by: Ston Jia <ston.jia@outlook.com>
Link: https://patch.msgid.link/20251029051804.220111-1-ston.jia@outlook.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
|
|
thermal profile support
This patch adds Victus 16-r0 (8bbe) and Victus 16-s0(8bd4, 8bd5) laptop
DMI board name into existing list
Signed-off-by: Edip Hazuri <edip@medip.dev>
Link: https://patch.msgid.link/20251015181042.23961-3-edip@medip.dev
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
|
|
Currently, the NUM_CONTROLLERS macro is used to limit the amount of memory
controllers (UMCs) available per node. The number of UMCs available per node,
however, is already cached by the max_mcs variable of struct amd64_pvt.
Allocate the relevant data structures dynamically using the variable instead
of static allocation through the macro.
The max_mcs variable is used for legacy systems too. These systems have a max
of 2 controllers. Since the default value of max_mcs, set in per_family_init(),
is 2, these legacy systems are also covered.
Signed-off-by: Avadhut Naik <avadhut.naik@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20251106015727.1987246-1-avadhut.naik@amd.com
|
|
The camera subsystem for the MSM8939 is the same as MSM8916 except with
3 CSID instead of 2, and some higher clock rates.
As a quirk, this SoC needs writing values to 2 VFE VBIF registers
(see downstream msm8939-camera.dtsi vbif-{regs,settings} properties).
This fixes black stripes across sensor and garbage in CSID TPG outputs.
Add support for the MSM8939 camera subsystem.
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Vincent Knecht <vincent.knecht@mailoo.org>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Some devices need writing values to VFE VBIF registers.
Add helper functions to do this.
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Vincent Knecht <vincent.knecht@mailoo.org>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Add a configuration for all CSI lanes into D-PHY bus mode on Qualcomm
SM8650 CAMSS CSIPHY IPs.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Add the basic support of CAMSS IP on Qualcomm SM8650 SoC powered boards.
SM8650 CAMSS provides:
- 6 x CSIPHY,
- 3 x CSID, 2 x CSID Lite,
- 3 x VFE, 2 x VFE Lite.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
On hardware architectures where a single CAMNOC module is split into
two, one for each of the real time (RT) and non real time (NRT) modules
within camera sub system, processing VFE output over the AXI bus
requires enabling and setting the appropriate clock rate for the RT
CAMNOC. This change lays the groundwork for supporting such
configurations.
Signed-off-by: Hangxiang Ma <hangxiang.ma@oss.qualcomm.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Vijay Kumar Tumati <vijay.tumati@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Replace the hardcoded buffer count value with a macro to enable
operating on these buffers elsewhere in the CAMSS driver based on this
count. Some of the hardware architectures require deferring the AUP and
REG update until after the CSID configuration and this macro is expected
to be useful in such scenarios.
Signed-off-by: Hangxiang Ma <hangxiang.ma@oss.qualcomm.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The driver has never supported anything but OF probing so drop the
unused platform module alias.
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The static TPDM function as a dummy source, however, it is essential
to enable the port connected to the TPDA and configure the element size.
Without this, the TPDA cannot correctly receive trace data from the
static TPDM. Since the static TPDM does not require MMIO mapping to
access its registers, a clock controller is not mandatory for its
operation.
Signed-off-by: Jie Gan <jie.gan@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20251028-add_static_tpdm_support-v4-2-84e21b98e727@oss.qualcomm.com
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
|
|
Currently, the ctl_name string is statically assigned based on the family and
model of the SOC when the amd64_edac module is loaded.
The same, however, is not exactly needed as the string can be generated and
assigned at runtime through scnprintf().
Remove all static assignments and generate the string at runtime. Also,
cleanup the switch cases which became defunct and consolidate identical cases.
Signed-off-by: Avadhut Naik <avadhut.naik@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20251106015727.1987246-1-avadhut.naik@amd.com
|
|
Merge series from AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>:
This series adds support for three new MediaTek PMICs: MT6316, MT6363
and MT6373 and their variants - used in board designs featuring the
MediaTek MT8196 Chromebook SoC, or the MT6991 Dimensity 9400 Smartphone
SoC.
In particular, MT6316 is a regulator, but the MT6363 and MT6373 PMICs
are multi-function devices, as they have and expose multiple sub-devices;
moreover, some of those also contain an interrupt controller, managing
internal IPs interrupts: for those, a chained interrupt handler is
registered, which parent is the SPMI controller itself.
This series adds support for all of the MT6316 regulator variants and
for MT6363, MT6373 SPMI PMICs and their interrupt controller.
|
|
Merge series from Srinivas Kandagatla <srinivas.kandagatla@oss.qualcomm.com>:
This patchset has 4 fixes and some enhancements to the Elite DSP driver
support.
Fixes includes
- setting correct flags for expected behaviour of appl_ptr
- fix closing of copp instances
- fix buffer alignment.
- fix state checks before closing asm stream
Enhancements include:
- adding q6asm_get_hw_pointer and ack callback support
- simplify code via __free(kfree) mechanism.
- use spinlock guards
- few cleanups discovered during doing above 2.
There is another set of updates comming soon, which will add support
for early memory mapping and few more modules support in audioreach.
|
|
Stephen Rothwell reports htmldocs warnings when merging drm-misc tree:
Documentation/gpu/drm-mm:40: include/drm/ttm/ttm_device.h:225: ERROR: Unknown target name: "ttm_allocation". [docutils]
Documentation/gpu/drm-mm:43: drivers/gpu/drm/ttm/ttm_device.c:202: ERROR: Unknown target name: "ttm_allocation". [docutils]
Documentation/gpu/drm-mm:73: include/drm/ttm/ttm_pool.h:68: ERROR: Unknown target name: "ttm_allocation_pool". [docutils]
Documentation/gpu/drm-mm:76: drivers/gpu/drm/ttm/ttm_pool.c:1070: ERROR: Unknown target name: "ttm_allocation_pool". [docutils]
Fix these by adding missing wildcard on TTM_ALLOCATION_* and
TTM_ALLOCATION_POOL_* in @alloc_flags description.
Fixes: 0af5b6a8f8dd ("drm/ttm: Replace multiple booleans with flags in pool init")
Fixes: 77e19f8d3297 ("drm/ttm: Replace multiple booleans with flags in device init")
Fixes: 402b3a865090 ("drm/ttm: Add an allocation flag to propagate -ENOSPC on OOM")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Closes: https://lore.kernel.org/linux-next/20251105161838.55b962a3@canb.auug.org.au/
Signed-off-by: Bagas Sanjaya <bagasdotme@gmail.com>
Acked-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tvrtko Ursulin <tursulin@ursulin.net>
Link: https://lore.kernel.org/r/20251106005217.14026-1-bagasdotme@gmail.com
|
|
Add sanity check in iris_vb2_stop_streaming. If inst->state is
already IRIS_INST_ERROR, we should skip the stream_off operation
because it would still send packets to the firmware.
In iris_kill_session, inst->state is set to IRIS_INST_ERROR and
session_close is executed, which will kfree(inst_hfi_gen2->packet).
If stop_streaming is called afterward, it will cause a crash.
Fixes: 11712ce70f8e5 ("media: iris: implement vb2 streaming ops")
Cc: stable@vger.kernel.org
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Signed-off-by: Wangao Wang <wangao.wang@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
[bod: remove qcom from patch title]
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Improve the condition used to determine when input internal buffers need
to be reconfigured during streamon on the capture port. Previously, the
check relied on the INPUT_PAUSE sub-state, which was also being set
during seek operations. This led to input buffers being queued multiple
times to the firmware, causing session errors due to duplicate buffer
submissions.
This change introduces a more accurate check using the FIRST_IPSC and
DRC sub-states to ensure that input buffer reconfiguration is triggered
only during resolution change scenarios, such as streamoff/on on the
capture port. This avoids duplicate buffer queuing during seek
operations.
Fixes: c1f8b2cc72ec ("media: iris: handle streamoff/on from client in dynamic resolution change")
Cc: stable@vger.kernel.org
Reported-by: Val Packett <val@packett.cool>
Closes: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/4700
Signed-off-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The iris_v4l2_file_ops structure is never modified. Make it const.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Abel Vesa <abel.vesa@linaro.org>
[bod: drop qcom from patch title for consistency with other patches]
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
As a part of migrating code from the old Venus driver to the new Iris
one, add support for the SC7280 platform. It is very similar to SM8250,
but it (currently) uses no reset controls (there is an optional
GCC-generated reset, it will be added later) and no AON registers
region. Extend the VPU ops to support optional clocks and skip the AON
shutdown for this platform.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
In preparation to adding more Gen1 platforms, which will share a
significant amount of data, rename the SM8250 platform file to
iris_platform_gen1.c.
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Now as we have removed PIPE value from inst_fw_caps_dec there should be
no difference between inst_fw_caps of QCS8300 and SM8550+. Drop the
QCS8300-specific tables and use generic one instead.
The differences between QCS8300 and SM8550 data comes from a
non-conflict merge of commit d22037f3fd33 ("media: iris: Set platform
capabilities to firmware for encoder video device") (which added .set
callbacks), and commit 6bdfa3f947a7 ("media: iris: Add platform-specific
capabilities for encoder video device") (which added QCS8300 data, but
not the callbacks).
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The value of the PIPE property depends on the number of pipes available
on the platform and is frequently the only difference between several
fw_caps. In order to reduce duplication, use num_vpp_pipe from the
iris_platform_data rather than hardcoding the value into the fw_cap.
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
[bod: duplciation -> duplication]
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Make all struct iris_platform_data instances constant, they are not
modified at runtime.
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Make all struct platform_inst_fw_cap instances constant, they are not
modified at runtime.
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Introduce handling for the QC08C format in the encoder. QC08C
format is NV12 with UBWC compression. Update format checks and
configuration to enable encoding to QC08C streams.
Signed-off-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Introduce handling for the QC08C format in the decoder. QC08C format is
NV12 with UBWC compression. Update format checks and configuration to
enable decoding of QC08C streams.
Signed-off-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Add handling for the HFI_PROP_OPB_ENABLE property, which allows enabling
or disabling split mode in the firmware. When HFI_PROP_OPB_ENABLE is set
to true, the firmware activates split mode for output picture buffers
(OPB). The OPB format is determined by the HFI_PROP_COLOR_FORMAT
property, supporting NV12 or QC08C formats.
Signed-off-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The driver has never supported anything but OF probing so drop the
unused platform module aliases.
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The encoder and decoder platform devices are registered by the venus
driver as children of the venus device, but even if someone were to mess
this up no amount of probe deferring is going to conjure up a parent.
Relatedly, the venus driver sets its driver data before registering the
child devices and if this ever breaks we want to learn about it by
failing probe.
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
Reviewed-by: Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Use the newly added of_reserved_mem_region_to_resource() function to
handle "memory-region" properties.
Reviewed-by: Dikshita Agarwal <quic_dikshita@quicinc.com>
Reviewed-by: Ming Qian <ming.qian@oss.nxp.com>
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bod@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistentcy cannot be addressed without refactoring the API.
alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.
This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.
This change adds a new WQ_PERCPU flag to explicitly request
alloc_workqueue() to be per-cpu when WQ_UNBOUND has not been specified.
With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.
Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.
CC: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251101163121.78400-6-marco.crivellari@suse.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistentcy cannot be addressed without refactoring the API.
alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.
This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.
This change adds a new WQ_PERCPU flag to explicitly request
alloc_workqueue() to be per-cpu when WQ_UNBOUND has not been specified.
With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.
Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.
CC: Yishai Hadas <yishaih@nvidia.com>
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251101163121.78400-5-marco.crivellari@suse.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistentcy cannot be addressed without refactoring the API.
alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.
This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.
This change adds a new WQ_PERCPU flag to explicitly request
alloc_workqueue() to be per-cpu when WQ_UNBOUND has not been specified.
With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.
Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.
CC: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251101163121.78400-4-marco.crivellari@suse.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistentcy cannot be addressed without refactoring the API.
alloc_workqueue() treats all queues as per-CPU by default, while unbound
workqueues must opt-in via WQ_UNBOUND.
This default is suboptimal: most workloads benefit from unbound queues,
allowing the scheduler to place worker threads where they’re needed and
reducing noise when CPUs are isolated.
This change adds a new WQ_PERCPU flag to explicitly request
alloc_workqueue() to be per-cpu when WQ_UNBOUND has not been specified.
With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU.
Once migration is complete, WQ_UNBOUND can be removed and unbound will
become the implicit default.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251101163121.78400-3-marco.crivellari@suse.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistency cannot be addressed without refactoring the API.
system_unbound_wq should be the default workqueue so as not to enforce
locality constraints for random work whenever it's not required.
Adding system_dfl_wq to encourage its use when unbound work should be used.
The old system_unbound_wq will be kept for a few release cycles.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251101163121.78400-2-marco.crivellari@suse.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Need to take an SRQ lock in poll_cq before moving SRQ tail.
Signed-off-by: Jay Bhat <jay.bhat@intel.com>
Reviewed-by: Jacob Moroni <jmoroni@google.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Link: https://patch.msgid.link/20251031021726.1003-7-tatyana.e.nikolova@intel.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
TX wait skbs need to be completed when they are done. PCIe part does this
inside rtw89_pci_tx_status() during RPP processing. Other HCIs use a
mechanism based on C2H firmware messages.
Store TX wait skbs inside TX report queue so that it'll be possible to
identify completed items inside the C2H handler. Try to do this as
similar to PCIe path as possible. When the corresponding TX wait skb is
found inside TX report queue, unlink it from there and call
rtw89_core_tx_wait_complete() to mark the completion.
If the callee waiting for the completion has already timed out, the TX
wait skb is placed into TX wait list (like PCIe part does).
It's important that during HCI reset all pending TX wait frames should be
completed inside hci.ops->reset method before calling
rtw89_tx_wait_list_clear().
Found by Linux Verification Center (linuxtesting.org).
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-11-pchelkin@ispras.ru
|
|
In order to provide TX reports for the management queue rtw89 should
configure the firmware. Do this with SET_CMC_TBL_MGQ_RPT_EN() for the
WiFi6 chips and with CCTLINFO_G7_W0_MGQ_RPT_EN flag for the WiFi7 ones.
Suggested-by: Bitterblue Smith <rtl8821cerfe2@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-10-pchelkin@ispras.ru
|
|
Frames flagged with IEEE80211_TX_CTL_REQ_TX_STATUS mean the driver has to
report to mac80211 stack whether AP sent ACK for the null frame/probe
request or not. It's not implemented in USB part of the driver yet.
PCIe HCI has its own way of getting TX status incorporated into RPP
feature, and it's always enabled there. Other HCIs need a different
scheme based on processing C2H messages.
Thus define a .tx_rpt_enabled flag indicating which HCIs need to enable a
TX report feature. Currently it is USB only.
Toggle a bit in the TX descriptor and place flagged skbs in a fix-sized
queue to wait for a message from the firmware. Firmware maintains a 4-bit
sequence number for required frames hence the queue can contain just 16
elements simultaneously. That's enough for normal driver / firmware
communication. If the firmware crashes for any reason and doesn't provide
TX reports in time, driver will handle this and report the obsolete frames
as dropped.
rtw89 also has a new feature providing a TX report for each transmission
attempt. Ignore a failed TX status reported by the firmware until retry
limit is reached or successful status appears. When there is no success
and the retry limit is reached, report the frame up to the wireless stack
as failed eventually.
HCI reset should stop all pending TX activity so forcefully flush the
queue there.
Found by Linux Verification Center (linuxtesting.org).
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-9-pchelkin@ispras.ru
|
|
During HCI reset all pending TX URBs should be canceled. Use anchor to
keep track of them and have an ability to cancel them synchronously.
Note however that canceling RX URBs can't be done here in
rtw89_usb_ops_reset() as it breaks driver initialization.
Found by Linux Verification Center (linuxtesting.org).
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-8-pchelkin@ispras.ru
|
|
TX type FWCMD is used to download firmware and send H2C commands, and
it only fill few fields of TX description, such as desc_info->pkt_size.
Therefore, early return the TX type FWCMD.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Link: https://patch.msgid.link/20251104135720.321110-7-pchelkin@ispras.ru
|
|
strcpy() is deprecated; use the safer strscpy() instead.
The destination buffer is only zero-initialized for the first iteration
and since strscpy() guarantees its NUL termination anyway, remove
zero-initializing 'eng_type'.
Link: https://github.com/KSPP/linux/issues/88
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
PCI device 0x115A is similar to pspv5, except it doesn't have platform
access mailbox support.
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
The previous version check made it difficult to support newer major
versions (e.g., v6.0) without adding extra checks/macros. Update the
logic to only reject v5.0 and allow future versions without additional
changes.
Signed-off-by: Gaurav Kashyap <gaurav.kashyap@oss.qualcomm.com>
Signed-off-by: Jingyi Wang <jingyi.wang@oss.qualcomm.com>
Reviewed-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Fix cocci warnings to use %pe to print PTR_ERR().
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Julia Lawall <julia.lawall@inria.fr>
Closes: https://lore.kernel.org/r/202510231229.Z6TduqZy-lkp@intel.com/
Signed-off-by: Harsh Jain <h.jain@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
rtw89 has several ways of handling TX status report events. The first one
is based on RPP feature which is used by PCIe HCI. The other one depends
on firmware sending a corresponding C2H message, quite similar to what
rtw88 has.
Toggle a bit in the TX descriptor to indicate to the firmware that TX
report for the frame is expected. This will allow handling TX wait skbs
and the ones flagged with IEEE80211_TX_CTL_REQ_TX_STATUS correctly.
Do the bulk of the patch according to the vendor driver for RTL8851BU.
However, there are slight differences in C2H message format between
different types of chips. RTL885xB ones follow format V0. RTL8852C has
format V1, and RTL8922AU has format V2.
Found by Linux Verification Center (linuxtesting.org).
Suggested-by: Bitterblue Smith <rtl8821cerfe2@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-6-pchelkin@ispras.ru
|
|
The qm_get_qos_value() function calls bus_find_device_by_name() which
increases the device reference count, but fails to call put_device()
to balance the reference count and lead to a device reference leak.
Add put_device() calls in both the error path and success path to
properly balance the reference count.
Found via static analysis.
Fixes: 22d7a6c39cab ("crypto: hisilicon/qm - add pci bdf number check")
Cc: stable@vger.kernel.org
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Reviewed-by: Longfang Liu <liulongfang@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Pass TX status value directly into rtw89_core_tx_wait_complete(). This
will make it a bit in sync with further patches and will give flexibility
in future work. Also use scope based RCU locking which simplifies the
code of the function.
Found by Linux Verification Center (linuxtesting.org).
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20251104135720.321110-5-pchelkin@ispras.ru
|
