The AudioDSP firmware's GAIN module has same initialization payload as
PEAKVOL and user volume setting can be applied up-front. Update existing
code to account for PEAKVOL and GAIN both.
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://patch.msgid.link/20250217102115.3539427-5-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver exposes volume kcontrols if PEAKVOL/GAIN module is present
in the streaming path. Currently there is no control over their default
values including the effect that may accompany the volume change event.
Add template for PEAKVOL/GAIN module which holds all the information
needed to address the limitation.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://patch.msgid.link/20250217102115.3539427-4-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
To provide minimal support for multi-channel kcontrols i.e.: more than
stereo configuration, store the number of channels specified within the
SectionControlMixer. The field is part of the topology standard,
currently skipped by the ASoC core.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://patch.msgid.link/20250217102115.3539427-3-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Fields ->ops.info and ->type of struct snd_soc_tplg_ctl_hdr denote
info-operation type and control type respectively. These are two
different pieces of information. The info type is represented by
SND_SOC_TPLG_CTL_xxx and SND_SOC_TPLG_DAPM_CTL_xxx on UAPI side whereas
for control type it is SND_SOC_TPLG_TYPE_xxx (mixer, bytes or enum).
The type of the kcontrol to be created is currently guessed based on the
value of the ->ops.info. Use the ->type instead to correct and simplify
the code. With this change ops.info() can be customized by sound drivers
utilizing the ASoC-topology just like ops.get() and ops.put() can be.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://patch.msgid.link/20250217102115.3539427-2-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
While processing TPC stats received from firmware, there are chances that
the tpc_stats might not be filled and the data is not available. This can
happen under two scenarios. First, when firmware sends a non-zero event
count before event count 0. When this happens, tpc_stats will be checked
for data before memory allocation and the tpc_stats will be unavailable.
Second, when memory allocation failed when event count received is 0 and
the firmware still sends a non-zero event. When this happens, memory will
not be allocated for tpc_stats though event count is 0, so when non-zero
event count is received, tpc_stats will be empty. There are checks to
validate if tpc_stats variable is filled that are used in two subsequent
places, but these are placed after tpc_stats is dereference without
checking if it is NULL or has valid data.
Fix this by removing the mentioned checks and adding a NULL check after
assigning tpc_stats to check if it is valid.
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00173-QCAHKSWPL_SILICONZ-1
Closes: https://scan7.scan.coverity.com/#/project-view/52668/11354?selectedIssue=1637145
Fixes: f0c3bb78e4 ("wifi: ath12k: Add Support to Parse TPC Event from Firmware")
Signed-off-by: Roopni Devanathan <quic_rdevanat@quicinc.com>
Reviewed-by: Vasanthakumar Thiagarajan <vasanthakumar.thiagarajan@oss.qualcomm.com>
Link: https://patch.msgid.link/20250221041250.769491-1-quic_rdevanat@quicinc.com
Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Commit 3952657848 ("wifi: ath12k: Use mac80211 vif's link_conf instead of
bss_conf") aims at, where applicable, replacing all usage of vif's bss_conf
with link specific bss_conff, but missed one instance in
ath12k_mac_vif_cache_flush(). This results in wrong configurations passed
to ath12k_mac_bss_info_changed() when the link in question is not the default
link.
Change to use the link specific bss_conf to fix this issue.
Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0-03427-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1.15378.4
Fixes: 3952657848 ("wifi: ath12k: Use mac80211 vif's link_conf instead of bss_conf")
Signed-off-by: Baochen Qiang <quic_bqiang@quicinc.com>
Link: https://patch.msgid.link/20241209024146.3282-1-quic_bqiang@quicinc.com
Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
With the DPCM case, the backend only support capture or playback, then
the linked frontend can only support capture or playback, but frontend
can't automatically enable only capture or playback, it needs the input
from dt-binding.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Reviewed-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Link: https://patch.msgid.link/20250217021715.284951-3-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Refer to audio graph card, add playback-only and capture-only property
for imx-audio-card.yaml for the case that only playback or capture is
supported.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Reviewed-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250217021715.284951-2-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The Rockchip power-domain controller also plans to make use of
per-domain regulators similar to the MediaTek power-domain controller.
Since existing DTs are missing the regulator information, the kernel
should fallback to the automatically created dummy regulator if
necessary. Thus the version without the _optional suffix is needed.
The Rockchip driver plans to use the managed version, but to be
consistent with existing code the unmanaged version is added at the
same time.
Tested-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Link: https://patch.msgid.link/20250220-rk3588-gpu-pwr-domain-regulator-v6-1-a4f9c24e5b81@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
The `objs` suffix is reserved for user-space tools. Use the `y` suffix
instead, which is usually used for kernel drivers.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Reviewed-by: Jithu Joseph <jithu.joseph@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250218194113.26589-5-kuurtb@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The `objs` suffix is reserved for user-space tools. Use the `y` suffix
instead, which is usually used for kernel drivers.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250218194113.26589-4-kuurtb@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The `objs` suffix is reserved for user-space tools. Use the `y` suffix
instead, which is usually used for kernel drivers.
While at it, fix alignment in AMD PMC and PMF Makefiles.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250218194113.26589-3-kuurtb@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The `objs` suffix is reserved for user-space tools. Use the `y` suffix
instead, which is usually used for kernel drivers.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250218194113.26589-2-kuurtb@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The last use of one_qsfp_write() was removed in 2016's
commit 145dd2b399 ("IB/hfi1: Always turn on CDRs for low power QSFP
modules")
Remove it.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Link: https://patch.msgid.link/20250223215543.153312-1-linux@treblig.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Remove hard-coded strings by using the str_enable_disable() helper
function.
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Acked-by: Tzung-Bi Shih <tzungbi@kernel.org>
Link: https://patch.msgid.link/20250220120100.1530-2-thorsten.blum@linux.dev
Signed-off-by: Mark Brown <broonie@kernel.org>
Remove hard-coded strings by using the str_enabled_disabled() helper
function.
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Reviewed-by: Andrei Simion <andrei.simion@microchip.com>
Link: https://patch.msgid.link/20250222225925.539840-2-thorsten.blum@linux.dev
Signed-off-by: Mark Brown <broonie@kernel.org>
CONFIG_SND_SOC_SOF_AMD_COMMON is a hidden option that is only selected by
other options. It can't have a direct depends on AMD_NODE because select
can't pick another option automatically.
This was attempted to be fixed in commit b47834ee44 ("ASoC: SOF: amd:
Add depends on CPU_SUP_AMD") but this just masked the issue as it was found
in another config.
Instead move the `depends on AMD_NODE` out of SND_SOC_SOF_AMD_COMMON to
all the consumers and drop `depends on CPU_SUP_AMD`.
Fixes: b47834ee44 ("ASoC: SOF: amd: Add depends on CPU_SUP_AMD")
Fixes: f120cf33d2 ("ASoC: SOF: amd: Use AMD_NODE")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202502220104.a8P6ApQN-lkp@intel.com/
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/20250221181840.2639793-1-superm1@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
The ES8328 codec driver, which is also used for the ES8388 chip that
appears to have an identical register map, claims that the output can
either take the route from DAC->Mixer->Output or through DAC->Output
directly. To the best of what I could find, this is not true, and
creates problems.
Without DACCONTROL17 bit index 7 set for the left channel, as well as
DACCONTROL20 bit index 7 set for the right channel, I cannot get any
analog audio out on Left Out 2 and Right Out 2 respectively, despite the
DAPM routes claiming that this should be possible. Furthermore, the same
is the case for Left Out 1 and Right Out 1, showing that those two don't
have a direct route from DAC to output bypassing the mixer either.
Those control bits toggle whether the DACs are fed (stale bread?) into
their respective mixers. If one "unmutes" the mixer controls in
alsamixer, then sure, the audio output works, but if it doesn't work
without the mixer being fed the DAC input then evidently it's not a
direct output from the DAC.
ES8328/ES8388 are seemingly not alone in this. ES8323, which uses a
separate driver for what appears to be a very similar register map,
simply flips those two bits on in its probe function, and then pretends
there is no power management whatsoever for the individual controls.
Fair enough.
My theory as to why nobody has noticed this up to this point is that
everyone just assumes it's their fault when they had to unmute an
additional control in ALSA.
Fix this in the es8328 driver by removing the erroneous direct route,
then get rid of the playback switch controls and have those bits tied to
the mixer's widget instead, which until now had no register to play
with.
Fixes: 567e4f9892 ("ASoC: add es8328 codec driver")
Signed-off-by: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
Link: https://patch.msgid.link/20250222-es8328-route-bludgeoning-v1-1-99bfb7fb22d9@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Even when COMPAT is compiled out, we still have to pass
ctx->compat to __import_iovec(). Replace the read with an indirection
with a constant when the kernel doesn't support compat.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
Link: https://lore.kernel.org/r/2819df9c8533c36b46d7baccbb317a0ec89da6cd.1740400452.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Update gpio-virtuser to use the new dev-sync-probe helper functions for
synchronized platform device creation, reducing code duplication.
No functional change.
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250221133501.2203897-4-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Update gpio-sim to use the new dev-sync-probe helper functions for
synchronized platform device creation, reducing code duplication.
No functional change.
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250221133501.2203897-3-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Both gpio-sim and gpio-virtuser share a mechanism to instantiate a
platform device, wait for probe completion, and retrieve the probe
success or error status synchronously. With gpio-aggregator planned to
adopt this approach for its configfs interface, it's time to factor
out the common code.
Add dev-sync-probe.[ch] to house helper functions used by all such
implementations.
No functional change.
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250221133501.2203897-2-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Variable allocated by charlcd_alloc() should be released
by charlcd_free(). The following patch changed kfree() to
charlcd_free() to fix an API misuse.
Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org>
Fixes: 718e05ed92 ("auxdisplay: Introduce hd44780_common.[ch]")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Variable allocated by charlcd_alloc() should be released
by charlcd_free(). The following patch changed kfree() to
charlcd_free() to fix an API misuse.
Fixes: 718e05ed92 ("auxdisplay: Introduce hd44780_common.[ch]")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Merge the immutable branch rockchip into next, to allow it to be tested
together with the changes that are targeted for v6.15.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
To resolve the build error with undefined reference to
`arm_smccc_1_1_get_conduit', let's add a build dependency to
HAVE_ARM_SMCCC_DISCOVERY.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Fixes: 61eeb96787 ("pmdomain: rockchip: Check if SMC could be handled by TA")
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Let the driver core manage the lifetime of the HWMON device, by
registering it inside tpacpi_hwmon_pdriver's probe and using
devm_hwmon_device_register_with_groups().
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Reviewed-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Tested-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Link: https://lore.kernel.org/r/20250215000302.19753-3-kuurtb@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
It was reported that if subdrivers assigned devres resources inside
ibm_init_struct's .init callbacks, driver binding would fail with the
following error message:
platform thinkpad_acpi: Resources present before probing
Let the driver core manage the lifetimes of the subdrivers and children
devices, by initializing them inside tpacpi_driver's .probe callback.
This is appropriate because these subdrivers usually expose sysfs groups
and the driver core manages this automatically to avoid races.
One immediate benefit of this, is that we are now able to use devres
inside .init subdriver callbacks.
platform_create_bundle is specifically used because it makes the
driver's probe type synchronous and returns an ERR_PTR if attachment
failed.
Additionally, to make error handling simpler, allocate the input device
using devm_input_allocate_device().
Reported-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Closes: https://lore.kernel.org/platform-driver-x86/20250208091438.5972-1-mpearson-lenovo@squebb.ca/#t
Reviewed-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Tested-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250215000302.19753-2-kuurtb@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Remove unneeded extern keyword in header file. Functions are
extern by default so extern keyword is not unnecessary for
function declaration.
Signed-off-by: Xi Pardee <xi.pardee@linux.intel.com>
Link: https://lore.kernel.org/r/20250214214416.10150-4-xi.pardee@linux.intel.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Remove unnecessary declarations in header file. Variable that are
used by only one .c file are removed from header file and changed
to be static in their corresponding .c file.
Signed-off-by: Xi Pardee <xi.pardee@linux.intel.com>
Link: https://lore.kernel.org/r/20250214214416.10150-3-xi.pardee@linux.intel.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Add Panther Lake support to intel_pmc_core driver
Signed-off-by: Rajvi Jingar <rajvi.jingar@linux.intel.com>
Signed-off-by: Xi Pardee <xi.pardee@linux.intel.com>
Link: https://lore.kernel.org/r/20250214214416.10150-2-xi.pardee@linux.intel.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Warn WMI driver developers to not use GUID-based and non-GUID-based
functions for querying WMI data blocks and handling WMI events
simultaneously on the same device, as this will corrupt the WMI device
state and might thus lead to erratic behaviour.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-9-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
After performing some tests with a custom SSDT table available at
https://github.com/Wer-Wolf/acpi-wmi-ssdt i found out that Windows
also enables data block collection even when the data block is
being set.
Emulate this behaviour to avoid confusing the ACPI firmware.
The bus-based API already implements this behaviour, so only the
legacy GUID-based API needs to be changed.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-8-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The handling of the WExx ACPI methods used for enabling and disabling
WMI events has multiple flaws:
- the ACPI methods are called even when the WMI device has not been
marked as expensive.
- WExx ACPI methods might be called for inappropriate WMI devices.
- the error code AE_NOT_FOUND is treated as success.
The handling of the WCxx ACPI methods used for enabling and disabling
WMI data blocks is also flawed:
- WMI data blocks are enabled and disabled for every single "query"
operation. This is racy and inefficient.
Unify the handling of both ACPI methods by introducing a common
helper function for enabling and disabling WMI devices.
Also enable/disable WMI data blocks during probe/remove and shutdown
to match the handling of WMI events.
Legacy GUID-based functions still have to enable/disable the WMI
device manually and thus still suffer from a potential race condition.
Since those functions are deprecated and suffer from various other
flaws this issue is purposefully not fixed.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-7-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Use devm_add_action_or_reset() to disable the WMI device instead of
manually calling wmi_method_enable() to prepare for future changes
inside the WMI data block handlign code.
The reason for this is that we have to make sure that all
devres-managed resources are released first because some might still
want to access the underlying WMI device.
Because devres-managed resources are not released during shutdown
we still have to manually disable the WMI device in this case.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-6-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The WMI core already knows the instance count of a WMI guid.
Use this information instead of querying all possible instances
which is slow and might be unreliable.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-5-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Since the driver already binds to LENOVO_BIOS_SETTING_GUID, using
wmidev_block_query() inside tlmi_setting() allows for faster
access to BIOS settings.
Reviewed-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-4-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Move the ACPI buffer handling out of tlmi_extract_output_string()
and instead pass the unpacked ACPI object to prepare for future
changes.
Reviewed-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-3-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Since the driver already binds to HP_WMI_NUMERIC_SENSOR_GUID, using
wmidev_block_query() allows for faster sensor access.
Acked-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: James Seo <james@equiv.tech>
Reviewed-by: James Seo <james@equiv.tech>
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250216193251.866125-2-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
When a bio with REQ_PREFLUSH is submitted to dm, __send_empty_flush()
generates a flush_bio with REQ_OP_WRITE | REQ_PREFLUSH | REQ_SYNC,
which causes the flush_bio to be throttled by wbt_wait().
An example from v5.4, similar problem also exists in upstream:
crash> bt 2091206
PID: 2091206 TASK: ffff2050df92a300 CPU: 109 COMMAND: "kworker/u260:0"
#0 [ffff800084a2f7f0] __switch_to at ffff80004008aeb8
#1 [ffff800084a2f820] __schedule at ffff800040bfa0c4
#2 [ffff800084a2f880] schedule at ffff800040bfa4b4
#3 [ffff800084a2f8a0] io_schedule at ffff800040bfa9c4
#4 [ffff800084a2f8c0] rq_qos_wait at ffff8000405925bc
#5 [ffff800084a2f940] wbt_wait at ffff8000405bb3a0
#6 [ffff800084a2f9a0] __rq_qos_throttle at ffff800040592254
#7 [ffff800084a2f9c0] blk_mq_make_request at ffff80004057cf38
#8 [ffff800084a2fa60] generic_make_request at ffff800040570138
#9 [ffff800084a2fae0] submit_bio at ffff8000405703b4
#10 [ffff800084a2fb50] xlog_write_iclog at ffff800001280834 [xfs]
#11 [ffff800084a2fbb0] xlog_sync at ffff800001280c3c [xfs]
#12 [ffff800084a2fbf0] xlog_state_release_iclog at ffff800001280df4 [xfs]
#13 [ffff800084a2fc10] xlog_write at ffff80000128203c [xfs]
#14 [ffff800084a2fcd0] xlog_cil_push at ffff8000012846dc [xfs]
#15 [ffff800084a2fda0] xlog_cil_push_work at ffff800001284a2c [xfs]
#16 [ffff800084a2fdb0] process_one_work at ffff800040111d08
#17 [ffff800084a2fe00] worker_thread at ffff8000401121cc
#18 [ffff800084a2fe70] kthread at ffff800040118de4
After commit 2def2845cc ("xfs: don't allow log IO to be throttled"),
the metadata submitted by xlog_write_iclog() should not be throttled.
But due to the existence of the dm layer, throttling flush_bio indirectly
causes the metadata bio to be throttled.
Fix this by conditionally adding REQ_IDLE to flush_bio.bi_opf, which makes
wbt_should_throttle() return false to avoid wbt_wait().
Signed-off-by: Jinliang Zheng <alexjlzheng@tencent.com>
Reviewed-by: Tianxiang Peng <txpeng@tencent.com>
Reviewed-by: Hao Peng <flyingpeng@tencent.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Clear provisional refcount values and count free/allocated blocks in
one integrated loop. Process 8 aligned bytes at a time instead of
every byte individually.
On an Intel i7-11850H this reduces the CPU time needed to process a
loaded refcount block by a factor of about 5-6. On a large system the
refcount loading may be the largest factor in device startup time.
Signed-off-by: Ken Raeburn <raeburn@redhat.com>
Signed-off-by: Matthew Sakai <msakai@redhat.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Lists are the new rings, so update all remaining references to rings to
talk about lists.
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Matthew Sakai <msakai@redhat.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>