summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2025-03-16spi: sg2044-nor: Convert to dev_err_probe()Andy Shevchenko
One of the cases in sg2044_spifmc_probe() may be converted to use dev_err_probe(). Do it. While at it, use local device pointer in all such calls and drop unneeded __func__ parameter as dev_err_probe() is assumed to be called only during probe phase. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://patch.msgid.link/20250313111423.322775-3-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16spi: sg2044-nor: Fully convert to device managed resourcesAndy Shevchenko
The driver has a wrong order of the cleaning up the resources, i.e. it first will destroy the mutex and only then free the SPI which might still use it. Fix this by switching to devm_mutex_init(). Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://patch.msgid.link/20250313111423.322775-2-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16regulator: pf9453: add PMIC PF9453 supportJoy Zou
Support new PMIC PF9453, which is totally difference with PCA9450. So create new file for it. The PF9453 is a single chip Power Management IC (PMIC) specifically designed for i.MX 91 processor. It provides power supply solutions for IoT (Internet of Things), smart appliance, and portable applications where size and efficiency are critical. The device provides four high efficiency step-down regulators, three LDOs, one 400 mA load switch and 32.768 kHz crystal oscillator driver. Signed-off-by: Joy Zou <joy.zou@nxp.com> Signed-off-by: Frank Li <Frank.Li@nxp.com> Link: https://patch.msgid.link/20250314-pf9453-v5-2-ab0cf1f871b0@nxp.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16regulator: dt-bindings: pca9450: Add nxp,pf9453 compatible stringFrank Li
Add the compatible string "nxp,pf9453" for the PF9453 regulator. The PF9453 is similar to the PCA9460 but supports only LDO1, LDO2, LDO_SVNS, and BUCK[1-4]. Restrict LDO and BUCK numbers for nxp,pf9453 and keep the same restriction for other compatible strings. Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Frank Li <Frank.Li@nxp.com> Link: https://patch.msgid.link/20250314-pf9453-v5-1-ab0cf1f871b0@nxp.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: cs35l56: Remove redundant 'flush_workqueue()' callsChen Ni
'destroy_workqueue()' already drains the queue before destroying it, so there is no need to flush it explicitly. Remove the redundant 'flush_workqueue()' calls. This was generated with coccinelle: @@ expression E; @@ -flush_workqueue(E); destroy_workqueue(E); Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://patch.msgid.link/20250312072635.1429870-1-nichen@iscas.ac.cn Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: codecs: wcd937x: fix a potential memory leak in wcd937x_soc_codec_probe()Haoxiang Li
When snd_soc_dapm_new_controls() or snd_soc_dapm_add_routes() fails, wcd937x_soc_codec_probe() returns without releasing 'wcd937x->clsh_info', which is allocated by wcd_clsh_ctrl_alloc. Add wcd_clsh_ctrl_free() to prevent potential memory leak. Fixes: 313e978df7fc ("ASoC: codecs: wcd937x: add audio routing and Kconfig") Cc: stable@vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com> Link: https://patch.msgid.link/20250226085050.3584898-1-haoxiang_li2024@163.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: soc-pcm: tidyup function name to snd_soc_dpcm_be_can_xxx()Kuninori Morimoto
We have similar but different function names snd_soc_dpcm_fe_can_xxx() snd_soc_dpcm_be_can_xxx() snd_soc_dpcm_can_be_xxx() ~~~~~~ Let's unified these to can_xx Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Link: https://patch.msgid.link/87plip7ie4.wl-kuninori.morimoto.gx@renesas.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: amd: acp: Fix leak in acp_pci_probe()Dan Carpenter
There needs to be some cleanup on this error path. We can't just return directly. Fixes: aaf7a668bb38 ("ASoC: amd: acp: Add new interrupt handle callbacks in acp_common_hw_ops") Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org> Link: https://patch.msgid.link/3dad80cb-e177-45aa-97ac-df9c98a47d94@stanley.mountain Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: tlv320adc3xxx: remove unnecessary NULL check before ↵Chen Ni
clk_disable_unprepare() clk_disable_unprepare() already checks NULL by using IS_ERR_OR_NULL. Remove unneeded NULL check for adc3xxx->mclk here. Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://patch.msgid.link/20250312034337.1235378-1-nichen@iscas.ac.cn Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: tegra: Use non-atomic timeout for ADX status registerRitu Chaudhary
ADX startup() callback uses atomic poll timeout on ADX status register. This is unnecessary because: - The startup() callback itself is non-atomic. - The subsequent timeout call in the same function already uses a non-atomic version. Using atomic version can hog CPU when it is not really needed, so replace it with non-atomic version. Fixes: a99ab6f395a9e ("ASoC: tegra: Add Tegra210 based ADX driver") Signed-off-by: Ritu Chaudhary <rituc@nvidia.com> Signed-off-by: Sheetal <sheetal@nvidia.com> Link: https://patch.msgid.link/20250311062010.33412-1-sheetal@nvidia.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: sma1307: Add NULL check in sma1307_setting_loaded()Chenyuan Yang
All varibale allocated by kzalloc and devm_kzalloc could be NULL. Multiple pointer checks and their cleanup are added. This issue is found by our static analysis tool Signed-off-by: Chenyuan Yang <chenyuan0y@gmail.com> Link: https://patch.msgid.link/20250311015714.1333857-1-chenyuan0y@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: mt8365: remove unnecessary NULL check before clk_disable_unprepare()Chen Ni
clk_disable_unprepare() already checks NULL by using IS_ERR_OR_NULL. Remove unneeded NULL check for clk here. Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://patch.msgid.link/20250312032600.1235158-1-nichen@iscas.ac.cn Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: ti: davinci-i2s: remove unnecessary NULL check before ↵Chen Ni
clk_disable_unprepare() clk_disable_unprepare() already checks NULL by using IS_ERR_OR_NULL. Remove unneeded NULL check for dev->ext_clk here. Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://patch.msgid.link/20250312033509.1235268-1-nichen@iscas.ac.cn Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Add support for GE Entity propertiesCharles Keepax
Add support for parsing the Group Entity properties from DisCo/ACPI. Group Entities allow control of several other Entities, typically Selector Units, from a single control. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-7-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Add SDCA Control Range data access helperCharles Keepax
SDCA Ranges are two dimensional arrays of data associated with controls, add a helper to provide an x,y access mechanism to the data and a helper to locate a specific value inside a range. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-6-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Add type flag for ControlsCharles Keepax
SDCA Controls come in a variety of data formats, to simplify later parsing work out this data type as the control is parsed and stash it for later use. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-5-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Allow naming of imp def controlsCharles Keepax
Implementation defined controls will not be present in the large list of known controls for SDCA. The driver should not return an error for these, because it is perfectly legal to have implementation defined controls. Update the handling to instead generate a generic name. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-4-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Use __free() to manage local buffersCharles Keepax
Use the cleanup.h helpers to manage some local buffers, this cleans up the error paths a little. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-3-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: SDCA: Tidy up initialization write parsingCharles Keepax
Slightly neaten up the initialization write code to overlay a struct rather than shifting the pointer along manually. This also removes the Sparse warning: sound/soc/sdca/sdca_functions.c:233:36: warning: cast to restricted __le32 Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20250312172205.4152686-2-ckeepax@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: pcm6240: Drop bogus code handling IRQ as GPIOLinus Walleij
The current code for the IRQ in pcm6240 makes no sense: it looks up an IRQ with of_irq_get(), treat it as a GPIO by issuing gpio_request(), gpio_direction_input() and gpio_to_irq() on it. This is just wrong, if the device tree assigns the IRQ from a GPIO number this is just incorrect: it is clearly stated that GPIO providers and IRQ providers are orthogonal. It is possible to look up an IRQ to a corresponding GPIO line but this is taking an IRQ and pretending it's a GPIO, which is just semantically wrong. Drop the offending code and treat the IRQ that we get from the device tree as any other IRQ, see for example other codec drivers. The DT bindings for this codec does not have any in-tree DTS files, which may explain why things are weird. As a bonus, this moves the driver away from the legacy <linux/gpio.h> include. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Link: https://patch.msgid.link/20250312-pcm-codecs-v1-3-41ffc4f8fc5c@linaro.org Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: pcm3008: Convert to GPIO descriptorsLinus Walleij
This converts the PCM3008 driver to look up the GPIO lines using descriptors. Apparently there are no in-tree users of the platform data struct, so users need to adopt. New users can associate the GPIO lines with the platform device "pcm3008-codec". Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Link: https://patch.msgid.link/20250312-pcm-codecs-v1-2-41ffc4f8fc5c@linaro.org Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: pcm1681: Drop unused includeLinus Walleij
This driver includes the legacy GPIO header <linux/gpio.h> but does not use any symbols from it so drop the include. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Link: https://patch.msgid.link/20250312-pcm-codecs-v1-1-41ffc4f8fc5c@linaro.org Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: mediatek: mt8188-mt6359: Add accdet headset jack detect supportNícolas F. R. A. Prado
Enable headset jack detection for MT8188 platforms that use the MT6359 ACCDET block for it, indicated by the mediatek,accdet property in DT. For those platforms, register a jack and initialize the ACCDET block to report jack events through it. Co-developed-by: Zoran Zhan <zoran.zhan@mediatek.com> Signed-off-by: Zoran Zhan <zoran.zhan@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Link: https://patch.msgid.link/20250306-mt8188-accdet-v3-4-7828e835ff4b@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: mediatek: mt6359: Add stub for mt6359_accdet_enable_jack_detectNícolas F. R. A. Prado
Add a stub for mt6359_accdet_enable_jack_detect() to prevent linker failures in the machine sound drivers calling it when CONFIG_SND_SOC_MT6359_ACCDET is not enabled. Suggested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Link: https://patch.msgid.link/20250306-mt8188-accdet-v3-3-7828e835ff4b@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: mediatek: common: Handle mediatek,accdet propertyNícolas F. R. A. Prado
Handle the optional mediatek,accdet property. When present, retrieve the sound component from its phandle, so the machine sound driver can use it to register the audio jack and initialize the MT6359 ACCDET for jack detection. Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Link: https://patch.msgid.link/20250306-mt8188-accdet-v3-2-7828e835ff4b@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: dt-bindings: mediatek,mt8188-mt6359: Add mediatek,accdetNícolas F. R. A. Prado
Add a mediatek,accdet phandle property to allow getting a reference to the MT6359 ACCDET block, which is responsible for detecting jack insertion/removal. Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Reviewed-by: Rob Herring (Arm) <robh@kernel.org> Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Link: https://patch.msgid.link/20250306-mt8188-accdet-v3-1-7828e835ff4b@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: codecs: Add aw88166 amplifier driverWeidong Wang
The driver is for amplifiers aw88166 of Awinic Technology Corporation. The AW88166 is a high efficiency digital Smart K audio amplifier Signed-off-by: Weidong Wang <wangweidong.a@awinic.com> Link: https://patch.msgid.link/20250312120100.9730-3-wangweidong.a@awinic.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16ASoC: dt-bindings: Add schema for "awinic,aw88166"Weidong Wang
Add the awinic,aw88166 property to support the aw88166 chip. Signed-off-by: Weidong Wang <wangweidong.a@awinic.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://patch.msgid.link/20250312120100.9730-2-wangweidong.a@awinic.com Signed-off-by: Mark Brown <broonie@kernel.org>
2025-03-16Linux 6.14-rc7Linus Torvalds
2025-03-16Merge tag 'media/v6.14-3' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media Pull media fix from Mauro Carvalho Chehab: "rtl2832 driver regression fix" * tag 'media/v6.14-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: media: rtl2832_sdr: assign vb2 lock before vb2_queue_init
2025-03-16Merge tag 'i2c-for-6.14-rc7' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux Pull i2c fixes from Wolfram Sang: - omap: fix irq ACKS to avoid irq storming and system hang - ali1535, ali15x3, sis630: fix error path at probe exit * tag 'i2c-for-6.14-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux: i2c: sis630: Fix an error handling path in sis630_probe() i2c: ali15x3: Fix an error handling path in ali15x3_probe() i2c: ali1535: Fix an error handling path in ali1535_probe() i2c: omap: fix IRQ storms
2025-03-16Merge tag 'trace-v6.14-rc5' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace Pull tracing fix from Steven Rostedt: "Fix ref count of trace_array in error path of histogram file open Tracing instances have a ref count to keep them around while files within their directories are open. This prevents them from being deleted while they are used. The histogram code had some files that needed to take the ref count and that was added, but the error paths did not decrement the ref counts. This caused the instances from ever being removed if a histogram file failed to open due to some error" * tag 'trace-v6.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: tracing: Correct the refcount if the hist/hist_debug file fails to open
2025-03-16perf/core: Use sysfs_emit() instead of scnprintf()XieLudan
Follow the advice in Documentation/filesystems/sysfs.rst: "- show() should only use sysfs_emit() or sysfs_emit_at() when formatting the value to be returned to user space." No change in functionality intended. [ mingo: Updated the changelog ] Signed-off-by: XieLudan <xie.ludan@zte.com.cn> Signed-off-by: Ingo Molnar <mingo@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20250315141738452lXIH39UJAXlCmcATCzcBv@zte.com.cn
2025-03-15Merge tag 'usb-6.14-rc7' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb Pull USB fixes from Greg KH: "Here are some small USB and Thunderbolt driver fixes and new usb-serial device ids. Included in here are: - new usb-serial device ids - typec driver bugfix - thunderbolt driver resume bugfix All of these have been in linux-next with no reported issues" * tag 'usb-6.14-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: usb: typec: tcpm: fix state transition for SNK_WAIT_CAPABILITIES state in run_state_machine() USB: serial: ftdi_sio: add support for Altera USB Blaster 3 thunderbolt: Prevent use-after-free in resume from hibernate USB: serial: option: fix Telit Cinterion FE990A name USB: serial: option: add Telit Cinterion FE990B compositions USB: serial: option: match on interface class for Telit FN990B
2025-03-15Merge tag 'input-for-v6.14-rc6' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input Pull input updates from Dmitry Torokhov: - several new device IDs added to xpad game controller driver - support for imagis IST3038H variant of chip added to imagis touch controller driver - a fix for GPIO allocation for ads7846 touch controller driver - a fix for iqs7222 driver to properly support status register - a fix for goodix-berlin touch controller driver to use the right name for the regulator - more i8042 quirks to better handle several old Clevo devices. * tag 'input-for-v6.14-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: MAINTAINERS: Remove myself from the goodix touchscreen maintainers Input: iqs7222 - preserve system status register Input: i8042 - swap old quirk combination with new quirk for more devices Input: i8042 - swap old quirk combination with new quirk for several devices Input: i8042 - add required quirks for missing old boardnames Input: i8042 - swap old quirk combination with new quirk for NHxxRZQ Input: xpad - rename QH controller to Legion Go S Input: xpad - add support for TECNO Pocket Go Input: xpad - add support for ZOTAC Gaming Zone Input: goodix-berlin - fix vddio regulator references Input: goodix-berlin - fix comment referencing wrong regulator Input: imagis - add support for imagis IST3038H dt-bindings: input/touchscreen: imagis: add compatible for ist3038h Input: xpad - add multiple supported devices Input: xpad - add 8BitDo SN30 Pro, Hyperkin X91 and Gamesir G7 SE controllers Input: ads7846 - fix gpiod allocation Input: wdt87xx_i2c - fix compiler warning
2025-03-15Merge tag 'rust-fixes-6.14-3' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux Pull rust fixes from Miguel Ojeda: "Toolchain and infrastructure: - Disallow BTF generation with Rust + LTO - Improve rust-analyzer support 'kernel' crate: - 'init' module: remove 'Zeroable' implementation for a couple types that should not have it - 'alloc' module: fix macOS failure in host test by satisfying POSIX alignment requirement - Add missing '\n's to 'pr_*!()' calls And a couple other minor cleanups" * tag 'rust-fixes-6.14-3' of git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux: scripts: generate_rust_analyzer: add uapi crate scripts: generate_rust_analyzer: add missing include_dirs scripts: generate_rust_analyzer: add missing macros deps rust: Disallow BTF generation with Rust + LTO rust: task: fix `SAFETY` comment in `Task::wake_up` rust: workqueue: add missing newline to pr_info! examples rust: sync: add missing newline in locked_by log example rust: init: add missing newline to pr_info! calls rust: error: add missing newline to pr_warn! calls rust: docs: add missing newline to printing macro examples rust: alloc: satisfy POSIX alignment requirement rust: init: fix `Zeroable` implementation for `Option<NonNull<T>>` and `Option<KBox<T>>` rust: remove leftover mentions of the `alloc` crate
2025-03-15Merge tag 'fsnotify_for_v6.14-rc7' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs Pull fsnotify reverts from Jan Kara: "Syzbot has found out that fsnotify HSM events generated on page fault can be generated while we already hold freeze protection for the filesystem (when you do buffered write from a buffer which is mmapped file on the same filesystem) which violates expectations for HSM events and could lead to deadlocks of HSM clients with filesystem freezing. Since it's quite late in the cycle we've decided to revert changes implementing HSM events on page fault for now and instead just generate one event for the whole range on mmap(2) so that HSM client can fetch the data at that moment" * tag 'fsnotify_for_v6.14-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs: Revert "fanotify: disable readahead if we have pre-content watches" Revert "mm: don't allow huge faults for files with pre content watches" Revert "fsnotify: generate pre-content permission event on page fault" Revert "xfs: add pre-content fsnotify hook for DAX faults" Revert "ext4: add pre-content fsnotify hook for DAX faults" fsnotify: add pre-content hooks on mmap()
2025-03-15ALSA: hda/realtek: Always honor no_shutup_pinsTakashi Iwai
The workaround for Dell machines to skip the pin-shutup for mic pins introduced alc_headset_mic_no_shutup() that is replaced from the generic snd_hda_shutup_pins() for certain codecs. The problem is that the call is done unconditionally even if spec->no_shutup_pins is set. This seems causing problems on other platforms like Lenovo. This patch corrects the behavior and the driver honors always spec->no_shutup_pins flag and skips alc_headset_mic_no_shutup() if it's set. Fixes: dad3197da7a3 ("ALSA: hda/realtek - Fixup headphone noise via runtime suspend") Reported-and-tested-by: Oleg Gorobets <oleg.goro@gmail.com> Link: https://patch.msgid.link/20250315143020.27184-1-tiwai@suse.de Signed-off-by: Takashi Iwai <tiwai@suse.de>
2025-03-15Revert "sched/core: Reduce cost of sched_move_task when config autogroup"Dietmar Eggemann
This reverts commit eff6c8ce8d4d7faef75f66614dd20bb50595d261. Hazem reported a 30% drop in UnixBench spawn test with commit eff6c8ce8d4d ("sched/core: Reduce cost of sched_move_task when config autogroup") on a m6g.xlarge AWS EC2 instance with 4 vCPUs and 16 GiB RAM (aarch64) (single level MC sched domain): https://lkml.kernel.org/r/20250205151026.13061-1-hagarhem@amazon.com There is an early bail from sched_move_task() if p->sched_task_group is equal to p's 'cpu cgroup' (sched_get_task_group()). E.g. both are pointing to taskgroup '/user.slice/user-1000.slice/session-1.scope' (Ubuntu '22.04.5 LTS'). So in: do_exit() sched_autogroup_exit_task() sched_move_task() if sched_get_task_group(p) == p->sched_task_group return /* p is enqueued */ dequeue_task() \ sched_change_group() | task_change_group_fair() | detach_task_cfs_rq() | (1) set_task_rq() | attach_task_cfs_rq() | enqueue_task() / (1) isn't called for p anymore. Turns out that the regression is related to sgs->group_util in group_is_overloaded() and group_has_capacity(). If (1) isn't called for all the 'spawn' tasks then sgs->group_util is ~900 and sgs->group_capacity = 1024 (single CPU sched domain) and this leads to group_is_overloaded() returning true (2) and group_has_capacity() false (3) much more often compared to the case when (1) is called. I.e. there are much more cases of 'group_is_overloaded' and 'group_fully_busy' in WF_FORK wakeup sched_balance_find_dst_cpu() which then returns much more often a CPU != smp_processor_id() (5). This isn't good for these extremely short running tasks (FORK + EXIT) and also involves calling sched_balance_find_dst_group_cpu() unnecessary (single CPU sched domain). Instead if (1) is called for 'p->flags & PF_EXITING' then the path (4),(6) is taken much more often. select_task_rq_fair(..., wake_flags = WF_FORK) cpu = smp_processor_id() new_cpu = sched_balance_find_dst_cpu(..., cpu, ...) group = sched_balance_find_dst_group(..., cpu) do { update_sg_wakeup_stats() sgs->group_type = group_classify() if group_is_overloaded() (2) return group_overloaded if !group_has_capacity() (3) return group_fully_busy return group_has_spare (4) } while group if local_sgs.group_type > idlest_sgs.group_type return idlest (5) case group_has_spare: if local_sgs.idle_cpus >= idlest_sgs.idle_cpus return NULL (6) Unixbench Tests './Run -c 4 spawn' on: (a) VM AWS instance (m7gd.16xlarge) with v6.13 ('maxcpus=4 nr_cpus=4') and Ubuntu 22.04.5 LTS (aarch64). Shell & test run in '/user.slice/user-1000.slice/session-1.scope'. w/o patch w/ patch 21005 27120 (b) i7-13700K with tip/sched/core ('nosmt maxcpus=8 nr_cpus=8') and Ubuntu 22.04.5 LTS (x86_64). Shell & test run in '/A'. w/o patch w/ patch 67675 88806 CONFIG_SCHED_AUTOGROUP=y & /sys/proc/kernel/sched_autogroup_enabled equal 0 or 1. Reported-by: Hazem Mohamed Abuelfotoh <abuehaze@amazon.com> Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org> Tested-by: Hagar Hemdan <hagarhem@amazon.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: https://lore.kernel.org/r/20250314151345.275739-1-dietmar.eggemann@arm.com
2025-03-15sched/uclamp: Optimize sched_uclamp_used static key enablingXuewen Yan
Repeat calls of static_branch_enable() to an already enabled static key introduce overhead, because it calls cpus_read_lock(). Users may frequently set the uclamp value of tasks, triggering the repeat enabling of the sched_uclamp_used static key. Optimize this and avoid repeat calls to static_branch_enable() by checking whether it's enabled already. [ mingo: Rewrote the changelog for legibility ] Signed-off-by: Xuewen Yan <xuewen.yan@unisoc.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Christian Loehle <christian.loehle@arm.com> Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org> Link: https://lore.kernel.org/r/20250219093747.2612-2-xuewen.yan@unisoc.com
2025-03-15sched/uclamp: Use the uclamp_is_used() helper instead of open-coding itXuewen Yan
Don't open-code static_branch_unlikely(&sched_uclamp_used), we have the uclamp_is_used() wrapper around it. [ mingo: Clean up the changelog ] Signed-off-by: Xuewen Yan <xuewen.yan@unisoc.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Hongyan Xia <hongyan.xia2@arm.com> Reviewed-by: Christian Loehle <christian.loehle@arm.com> Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org> Link: https://lore.kernel.org/r/20250219093747.2612-1-xuewen.yan@unisoc.com
2025-03-15Merge tag 'i2c-host-fixes-6.14-rc7' of ↵Wolfram Sang
git://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux into i2c/for-current i2c-host-fixes for v6.14-rc7 - omap: fixed irq ACKS to avoid irq storming and system hang. - ali1535, ali15x3, sis630: fixed error path at probe exit.
2025-03-14Merge tag 'v6.14-rc6-smb3-server-fixes' of git://git.samba.org/ksmbdLinus Torvalds
Pull smb server fixes from Steve French: - Two fixes for oplock break/lease races * tag 'v6.14-rc6-smb3-server-fixes' of git://git.samba.org/ksmbd: ksmbd: prevent connection release during oplock break notification ksmbd: fix use-after-free in ksmbd_free_work_struct
2025-03-14Merge tag 'v6.14-rc6-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6Linus Torvalds
Pull smb client fixes from Steve French: "Six smb3 client fixes, all also for stable" * tag 'v6.14-rc6-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6: smb: client: Fix match_session bug preventing session reuse cifs: Fix integer overflow while processing closetimeo mount option cifs: Fix integer overflow while processing actimeo mount option cifs: Fix integer overflow while processing acdirmax mount option cifs: Fix integer overflow while processing acregmax mount option smb: client: fix regression with guest option
2025-03-14Merge tag 'bcachefs-2025-03-14.2' of git://evilpiepirate.org/bcachefsLinus Torvalds
Pull another bcachefs hotfix from Kent Overstreet: - fix 32 bit build breakage * tag 'bcachefs-2025-03-14.2' of git://evilpiepirate.org/bcachefs: bcachefs: fix build on 32 bit in get_random_u64_below()
2025-03-14bcachefs: fix build on 32 bit in get_random_u64_below()Kent Overstreet
bare 64 bit divides not allowed, whoops arm-linux-gnueabi-ld: drivers/char/random.o: in function `__get_random_u64_below': drivers/char/random.c:602:(.text+0xc70): undefined reference to `__aeabi_uldivmod' Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-15tracing: tprobe-events: Fix leakage of module refcountMasami Hiramatsu (Google)
When enabling the tracepoint at loading module, the target module refcount is incremented by find_tracepoint_in_module(). But it is unnecessary because the module is not unloaded while processing module loading callbacks. Moreover, the refcount is not decremented in that function. To be clear the module refcount handling, move the try_module_get() callsite to trace_fprobe_create_internal(), where it is actually required. Link: https://lore.kernel.org/all/174182761071.83274.18334217580449925882.stgit@devnote2/ Fixes: 57a7e6de9e30 ("tracing/fprobe: Support raw tracepoints on future loaded modules") Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org> Cc: stable@vger.kernel.org
2025-03-15tracing: tprobe-events: Fix to clean up tprobe correctly when module unloadMasami Hiramatsu (Google)
When unloading module, the tprobe events are not correctly cleaned up. Thus it becomes `fprobe-event` and never be enabled again even if loading the same module again. For example; # cd /sys/kernel/tracing # modprobe trace_events_sample # echo 't:my_tprobe foo_bar' >> dynamic_events # cat dynamic_events t:tracepoints/my_tprobe foo_bar # rmmod trace_events_sample # cat dynamic_events f:tracepoints/my_tprobe foo_bar As you can see, the second time my_tprobe starts with 'f' instead of 't'. This unregisters the fprobe and tracepoint callback when module is unloaded but marks the fprobe-event is tprobe-event. Link: https://lore.kernel.org/all/174158724946.189309.15826571379395619524.stgit@mhiramat.tok.corp.google.com/ Fixes: 57a7e6de9e30 ("tracing/fprobe: Support raw tracepoints on future loaded modules") Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
2025-03-14Merge tag 'xfs-fixes-6.14-rc7' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linuxLinus Torvalds
Pull xfs cleanup from Carlos Maiolino: "Use abs_diff instead of XFS_ABSDIFF" * tag 'xfs-fixes-6.14-rc7' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: xfs: Use abs_diff instead of XFS_ABSDIFF
2025-03-14KVM: arm64: Create each pKVM hyp vcpu after its corresponding host vcpuFuad Tabba
Instead of creating and initializing _all_ hyp vcpus in pKVM when the first host vcpu runs for the first time, initialize _each_ hyp vcpu in conjunction with its corresponding host vcpu. Some of the host vcpu state (e.g., system registers and traps values) is not initialized until the first time the host vcpu is run. Therefore, initializing a hyp vcpu before its corresponding host vcpu has run for the first time might not view the complete host state of these vcpus. Additionally, this behavior is inline with non-protected modes. Acked-by: Will Deacon <will@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Signed-off-by: Fuad Tabba <tabba@google.com> Link: https://lore.kernel.org/r/20250314111832.4137161-5-tabba@google.com Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
generated by cgit (git 2.34.1) at 2025-09-17 02:44:43 +0000