April 2024 - Alsa-devel - mailman.alsa-project.org

[PATCH v2] slimbus: qcom-ngd-ctrl: Add timeout for wait operation
by Viken Dadhaniya 11 Apr '24

11 Apr '24

In current driver qcom_slim_ngd_up_worker() indefinitely waiting for ctrl->qmi_up completion object. This is resulting in workqueue lockup on Kthread. Added wait_for_completion_interruptible_timeout to allow the thread to wait for specific timeout period and bail out instead waiting infinitely. Fixes: a899d324863a ("slimbus: qcom-ngd-ctrl: add Sub System Restart support") Cc: stable(a)vger.kernel.org Reviewed-by: Konrad Dybcio <konrad.dybcio(a)linaro.org> Signed-off-by: Viken Dadhaniya <quic_vdadhani(a)quicinc.com> --- v1 -> v2: - Remove macro and add value inline. - add fix, cc and review tag. --- --- drivers/slimbus/qcom-ngd-ctrl.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c index efeba8275a66..a09a26bf4988 100644 --- a/drivers/slimbus/qcom-ngd-ctrl.c +++ b/drivers/slimbus/qcom-ngd-ctrl.c @@ -1451,7 +1451,11 @@ static void qcom_slim_ngd_up_worker(struct work_struct *work) ctrl = container_of(work, struct qcom_slim_ngd_ctrl, ngd_up_work); /* Make sure qmi service is up before continuing */ - wait_for_completion_interruptible(&ctrl->qmi_up); + if (!wait_for_completion_interruptible_timeout(&ctrl->qmi_up, + msecs_to_jiffies(MSEC_PER_SEC))) { + dev_err(ctrl->dev, "QMI wait timeout\n"); + return; + } mutex_lock(&ctrl->ssr_lock); qcom_slim_ngd_enable(ctrl, true); -- QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation

2 1

[PATCH] slimbus: qcom-ctrl: fix module autoloading
by Krzysztof Kozlowski 11 Apr '24

11 Apr '24

Add MODULE_DEVICE_TABLE(), so the module could be properly autoloaded based on the alias from of_device_id table. Pin controllers are considered core components, so usually they are built-in, however these Signed-off-by: Krzysztof Kozlowski <krzk(a)kernel.org> --- drivers/slimbus/qcom-ctrl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/slimbus/qcom-ctrl.c b/drivers/slimbus/qcom-ctrl.c index 400b7b385a44..f7d04b50c973 100644 --- a/drivers/slimbus/qcom-ctrl.c +++ b/drivers/slimbus/qcom-ctrl.c @@ -718,6 +718,7 @@ static const struct of_device_id qcom_slim_dt_match[] = { { .compatible = "qcom,slim", }, {} }; +MODULE_DEVICE_TABLE(of, qcom_slim_dt_match); static struct platform_driver qcom_slim_driver = { .probe = qcom_slim_probe, -- 2.34.1

3 2

[PATCH] slimbus: Convert to platform remove callback returning void
by Uwe Kleine-König 11 Apr '24

11 Apr '24

The .remove() callback for a platform driver returns an int which makes many driver authors wrongly assume it's possible to do error handling by returning an error code. However the value returned is ignored (apart from emitting a warning) and this typically results in resource leaks. To improve here there is a quest to make the remove callback return void. In the first step of this quest all drivers are converted to .remove_new(), which already returns void. Eventually after all drivers are converted, .remove_new() will be renamed to .remove(). Trivially convert the slimbus drivers from always returning zero in the remove callback to the void returning variant. Signed-off-by: Uwe Kleine-König <u.kleine-koenig(a)pengutronix.de> --- drivers/slimbus/qcom-ctrl.c | 5 ++--- drivers/slimbus/qcom-ngd-ctrl.c | 11 ++++------- 2 files changed, 6 insertions(+), 10 deletions(-) diff --git a/drivers/slimbus/qcom-ctrl.c b/drivers/slimbus/qcom-ctrl.c index 400b7b385a44..7d632fad1300 100644 --- a/drivers/slimbus/qcom-ctrl.c +++ b/drivers/slimbus/qcom-ctrl.c @@ -626,7 +626,7 @@ static int qcom_slim_probe(struct platform_device *pdev) return ret; } -static int qcom_slim_remove(struct platform_device *pdev) +static void qcom_slim_remove(struct platform_device *pdev) { struct qcom_slim_ctrl *ctrl = platform_get_drvdata(pdev); @@ -635,7 +635,6 @@ static int qcom_slim_remove(struct platform_device *pdev) clk_disable_unprepare(ctrl->rclk); clk_disable_unprepare(ctrl->hclk); destroy_workqueue(ctrl->rxwq); - return 0; } /* @@ -721,7 +720,7 @@ static const struct of_device_id qcom_slim_dt_match[] = { static struct platform_driver qcom_slim_driver = { .probe = qcom_slim_probe, - .remove = qcom_slim_remove, + .remove_new = qcom_slim_remove, .driver = { .name = "qcom_slim_ctrl", .of_match_table = qcom_slim_dt_match, diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c index efeba8275a66..a5aa364f562e 100644 --- a/drivers/slimbus/qcom-ngd-ctrl.c +++ b/drivers/slimbus/qcom-ngd-ctrl.c @@ -1675,14 +1675,12 @@ static int qcom_slim_ngd_ctrl_probe(struct platform_device *pdev) return ret; } -static int qcom_slim_ngd_ctrl_remove(struct platform_device *pdev) +static void qcom_slim_ngd_ctrl_remove(struct platform_device *pdev) { platform_driver_unregister(&qcom_slim_ngd_driver); - - return 0; } -static int qcom_slim_ngd_remove(struct platform_device *pdev) +static void qcom_slim_ngd_remove(struct platform_device *pdev) { struct qcom_slim_ngd_ctrl *ctrl = platform_get_drvdata(pdev); @@ -1697,7 +1695,6 @@ static int qcom_slim_ngd_remove(struct platform_device *pdev) kfree(ctrl->ngd); ctrl->ngd = NULL; - return 0; } static int __maybe_unused qcom_slim_ngd_runtime_idle(struct device *dev) @@ -1740,7 +1737,7 @@ static const struct dev_pm_ops qcom_slim_ngd_dev_pm_ops = { static struct platform_driver qcom_slim_ngd_ctrl_driver = { .probe = qcom_slim_ngd_ctrl_probe, - .remove = qcom_slim_ngd_ctrl_remove, + .remove_new = qcom_slim_ngd_ctrl_remove, .driver = { .name = "qcom,slim-ngd-ctrl", .of_match_table = qcom_slim_ngd_dt_match, @@ -1749,7 +1746,7 @@ static struct platform_driver qcom_slim_ngd_ctrl_driver = { static struct platform_driver qcom_slim_ngd_driver = { .probe = qcom_slim_ngd_probe, - .remove = qcom_slim_ngd_remove, + .remove_new = qcom_slim_ngd_remove, .driver = { .name = QCOM_SLIM_NGD_DRV_NAME, .pm = &qcom_slim_ngd_dev_pm_ops, base-commit: 33e1d31873f87d119e5120b88cd350efa68ef276 -- 2.43.0

2 2

[PATCH v2] slimbus: qcom-ngd-ctrl: Reduce auto suspend delay
by Viken Dadhaniya 11 Apr '24

11 Apr '24

Currently we have auto suspend delay of 1s which is very high and it takes long time to driver for runtime suspend after use case is done. Hence to optimize runtime PM ops, reduce auto suspend delay to 100ms. Signed-off-by: Viken Dadhaniya <quic_vdadhani(a)quicinc.com> --- v1 -> v2: - Remove macro and add value inline. - Link to v1: https://lore.kernel.org/linux-arm-msm/d784e46d-974d-4bf3-a2d3-491e7ad19701@… --- --- drivers/slimbus/qcom-ngd-ctrl.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c index efeba8275a66..5e132273c6c4 100644 --- a/drivers/slimbus/qcom-ngd-ctrl.c +++ b/drivers/slimbus/qcom-ngd-ctrl.c @@ -81,7 +81,6 @@ #define SLIM_USR_MC_DISCONNECT_PORT 0x2E #define SLIM_USR_MC_REPEAT_CHANGE_VALUE 0x0 -#define QCOM_SLIM_NGD_AUTOSUSPEND MSEC_PER_SEC #define SLIM_RX_MSGQ_TIMEOUT_VAL 0x10000 #define SLIM_LA_MGR 0xFF @@ -1571,7 +1570,7 @@ static int qcom_slim_ngd_probe(struct platform_device *pdev) platform_set_drvdata(pdev, ctrl); pm_runtime_use_autosuspend(dev); - pm_runtime_set_autosuspend_delay(dev, QCOM_SLIM_NGD_AUTOSUSPEND); + pm_runtime_set_autosuspend_delay(dev, 100); pm_runtime_set_suspended(dev); pm_runtime_enable(dev); pm_runtime_get_noresume(dev); -- QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation

3 2

[PATCH v1 0/7] Add features, fixes and laptops for CS35L41 HDA
by Stefan Binding 11 Apr '24

11 Apr '24

This chain adds a number of improvements for CS35L41, including a new firmware loading feature, as well as some additional cleanup. Patch 1 adds support for a new firmware tuning file which is used to set special tuning parameters used by the firmware. The HP Omen laptop added in patches 2 and 3 use such a feature. Patch 4 changes the DSP1RX5/6 Source used by the firmware to depend on the boost type. A similar change in the ASoC driver will be needed later. Patches 5 and 6 are cleanup patches, with the first of those re-using cs-amp-lib for calibration which as added for CS35L56 drivers, ensuring that all laptops are using the correct calibration values. Patch 7 fixes a small error in the CS35L41 Property table for the Lenovo Legion slim 7 16ARHA7. Richard Fitzgerald (1): ALSA: hda: cs35l41: Remove redundant argument to cs35l41_request_firmware_file() Stefan Binding (6): ALSA: hda: cs35l41: Set the max PCM Gain using tuning setting ALSA: hda: cs35l41: Support HP Omen models without _DSD ALSA: hda/realtek: Add quirks for HP Omen models using CS35L41 ALSA: hda: cs35l41: Update DSP1RX5/6 Sources for DSP config ALSA: hda: cs35l41: Use shared cs-amp-lib to apply calibration ALSA: hda: cs35l41: Remove Speaker ID for Lenovo Legion slim 7 16ARHA7 include/sound/cs35l41.h | 5 + sound/pci/hda/Kconfig | 2 + sound/pci/hda/cs35l41_hda.c | 347 +++++++++++++++++++-------- sound/pci/hda/cs35l41_hda.h | 6 + sound/pci/hda/cs35l41_hda_property.c | 8 +- sound/pci/hda/patch_realtek.c | 4 + 6 files changed, 265 insertions(+), 107 deletions(-) -- 2.34.1

2 9

[PATCH] ASoC: wm_adsp: Include array_size.h
by Richard Fitzgerald 10 Apr '24

10 Apr '24

Explicitly #include array_size.h for the ARRAY_SIZE() macro. Signed-off-by: Richard Fitzgerald <rf(a)opensource.cirrus.com> --- sound/soc/codecs/wm_adsp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index 7d5c096e06cd..9a89682ebc1f 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -7,6 +7,7 @@ * Author: Mark Brown <broonie(a)opensource.wolfsonmicro.com> */ +#include <linux/array_size.h> #include <linux/ctype.h> #include <linux/module.h> #include <linux/moduleparam.h> -- 2.39.2

2 1

[PATCH] ASoC: cs35l56: Include array_size.h
by Richard Fitzgerald 10 Apr '24

10 Apr '24

Explicitly #include array_size.h for the source files that use ARRAY_SIZE(). Signed-off-by: Richard Fitzgerald <rf(a)opensource.cirrus.com> --- sound/soc/codecs/cs35l56-shared.c | 1 + sound/soc/codecs/cs35l56.c | 1 + 2 files changed, 2 insertions(+) diff --git a/sound/soc/codecs/cs35l56-shared.c b/sound/soc/codecs/cs35l56-shared.c index ec1d95e57061..3aa39f5923eb 100644 --- a/sound/soc/codecs/cs35l56-shared.c +++ b/sound/soc/codecs/cs35l56-shared.c @@ -5,6 +5,7 @@ // Copyright (C) 2023 Cirrus Logic, Inc. and // Cirrus Logic International Semiconductor Ltd. +#include <linux/array_size.h> #include <linux/firmware/cirrus/wmfw.h> #include <linux/gpio/consumer.h> #include <linux/regmap.h> diff --git a/sound/soc/codecs/cs35l56.c b/sound/soc/codecs/cs35l56.c index 6331b8c6136e..dfd703d9e12f 100644 --- a/sound/soc/codecs/cs35l56.c +++ b/sound/soc/codecs/cs35l56.c @@ -6,6 +6,7 @@ // Cirrus Logic International Semiconductor Ltd. #include <linux/acpi.h> +#include <linux/array_size.h> #include <linux/completion.h> #include <linux/debugfs.h> #include <linux/delay.h> -- 2.39.2

2 1

[PATCH v4 00/18] SoC: Cleanup MediaTek soundcard machine drivers
by AngeloGioacchino Del Regno 10 Apr '24

10 Apr '24

Changes in v4: - Rebased over next-20240409 - Dropped the first 4 patches from v3 as were already picked by Mark Changes in v3: - Added audio-routing names in enum in all yaml files - Added mention of disallowing old and new properties together in commit message of bindings patches - Fixed validation errors with sound-card-common.yaml inclusion due to missing model property in examples - Removed `else` enforcing headset-codec/speaker-codecs requirement if xxx-dai-link not present to avoid future commit noise as the deprecated statement will disallow deprecated properties as required Changes in v2: - Bindings: Changed link-name/codec/clk-provider to remove `items` and leave just the enum - Moved .*-dai-link pattern additionalProperties after `type: object` - Added ref to sound-card-common.yaml - Fixed dai-link-xxx -> xxx-dai-link typo in example comment This series performs a cleanup of most of the MediaTek AFE drivers and soundcard machine drivers, reducing code duplication and setting a base to be consistent with their devicetree bindings, as many of those are using different properties and nodes for no good reason. Summarizing: - Commonizes functions and ops where possible - Adds a common probe mechanism, increasing maintainability of soundcard drivers for older MediaTek SoCs - Migrates all drivers to support the new bindings - Obviously retains compatibility with old device trees - Reduces machine-specific parameters hardcoding in drivers - Can now set machine-specific params in device tree - Uses the `audio-routing` and `dai-link` nodes like some other non-MediaTek SoC sound drivers - Imposes consistency between MediaTek ASoC machine soundcard drivers bindings - Reduces code size and greatly reduces the amount of code that will be required for newer drivers (retaining compatibility with the old bindings was costly in terms of code size, otherwise this series would've removed ~1000 more lines, or something along that line). This series was (manually) tested on MT8173, MT8192, MT8195 and MT8186 Chromebooks. AngeloGioacchino Del Regno (18): ASoC: mediatek: Add common machine soundcard driver probe mechanism ASoC: mediatek: common: Constify struct mtk_sof_priv ASoC: mediatek: mt8188: Migrate to mtk_soundcard_common_probe ASoC: mediatek: mt8195: Migrate to mtk_soundcard_common_probe ASoC: mediatek: mt8192: Migrate to mtk_soundcard_common_probe ASoC: mediatek: mt8186: Migrate to mtk_soundcard_common_probe ASoC: mediatek: Add common snd_soc_ops .startup() callback ASoC: mediatek: mt8195: Migrate to the common mtk_soundcard_startup ASoC: mediatek: mt8192: Migrate to the common mtk_soundcard_startup ASoC: mediatek: mt8186-rt1019: Migrate to the common mtk_soundcard_startup ASoC: mediatek: Add common mtk_afe_component_probe callback ASoC: mediatek: Use common mtk_afe_pcm_platform with common probe cb ASoC: mediatek: mt8186: Unify mt8186-mt6366 machine drivers ASoC: dt-bindings: mt8195: Document audio-routing and dai-link subnode ASoC: dt-bindings: mt8192: Document audio-routing and dai-link subnode ASoC: dt-bindings: mt8186: Document audio-routing and dai-link subnode arm64: dts: mediatek: mt8195-cherry: Specify sound DAI links and routing arm64: dts: mediatek: mt8186-corsola: Specify sound DAI links and routing .../sound/mt8186-mt6366-da7219-max98357.yaml | 131 +- .../sound/mt8186-mt6366-rt1019-rt5682s.yaml | 120 +- .../sound/mt8192-mt6359-rt1015-rt5682.yaml | 139 +- .../bindings/sound/mt8195-mt6359.yaml | 134 ++ .../boot/dts/mediatek/mt8186-corsola.dtsi | 42 +- .../boot/dts/mediatek/mt8195-cherry.dtsi | 45 + sound/soc/mediatek/Kconfig | 24 +- .../mediatek/common/mtk-afe-platform-driver.c | 18 + .../soc/mediatek/common/mtk-dsp-sof-common.c | 15 +- .../soc/mediatek/common/mtk-dsp-sof-common.h | 1 - sound/soc/mediatek/common/mtk-soc-card.h | 7 +- .../mediatek/common/mtk-soundcard-driver.c | 199 +++ .../mediatek/common/mtk-soundcard-driver.h | 42 + sound/soc/mediatek/mt6797/mt6797-afe-pcm.c | 14 +- sound/soc/mediatek/mt7986/mt7986-afe-pcm.c | 14 +- sound/soc/mediatek/mt8183/mt8183-afe-pcm.c | 14 +- sound/soc/mediatek/mt8186/Makefile | 3 +- sound/soc/mediatek/mt8186/mt8186-afe-pcm.c | 19 +- sound/soc/mediatek/mt8186/mt8186-dai-adda.c | 2 +- .../mt8186/mt8186-mt6366-da7219-max98357.c | 1189 ----------------- ...t6366-rt1019-rt5682s.c => mt8186-mt6366.c} | 578 ++++---- sound/soc/mediatek/mt8188/mt8188-afe-pcm.c | 21 +- sound/soc/mediatek/mt8188/mt8188-mt6359.c | 203 +-- sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 25 +- .../mt8192/mt8192-mt6359-rt1015-rt5682.c | 301 ++--- sound/soc/mediatek/mt8195/mt8195-afe-pcm.c | 21 +- sound/soc/mediatek/mt8195/mt8195-mt6359.c | 487 +++---- 27 files changed, 1613 insertions(+), 2195 deletions(-) delete mode 100644 sound/soc/mediatek/mt8186/mt8186-mt6366-da7219-max98357.c rename sound/soc/mediatek/mt8186/{mt8186-mt6366-rt1019-rt5682s.c => mt8186-mt6366.c} (72%) -- 2.44.0

2 32

Thinkpad with ALC1318 has a chance of damaging the IC
by Kailang 10 Apr '24

10 Apr '24

Hi Takashi, Thinkpad with ALC1318 has a chance of damaging the IC at S4 resume. BR, Kailang

1 0

[PATCH v2 00/13] ASoC: ti: davinci-i2s: Add features to McBSP driver
by Bastien Curutchet 10 Apr '24

10 Apr '24

This series aims to add some features to McBSP driver. Convert bindings from .txt to .yaml. Add possibility to use an external clock as sample rate generator's input. Add handling of new formats (TDM, S24_LE, BP_FC). Enable the detection of unexpected frame pulses. Set the clock free-running mode according to SND_SOC_DAIFMT_[GATED/CONT] configuration in DAI format. Add ti,T1-framing[tx/rx] properties in DT. They allow to set the data delay to two bit-clock periods. This has been tested on a platform designed off of the DAVINCI/OMAP-L138 connected to 3 daisy-chained AD7767. An external clock drives the sample rate generator through the CLKS pin. The hardware I have only allowed me to test acquisition side of McBSP. It is connected to a 6 channels TDM and acts as Bit clock provider and Frame clock consumer. Change log v1 -> v2: PATCH 1 (bindings): * Drop power-domains property's description * Drop the unused label 'mcbsp0' in example * Add <> around each entry of the 'dmas' property * Add 'Reviewed-by: Rob Herring <robh(a)kernel.org>' PATCH 2 (bindings): * Drop the 'ti,enable-sync-err' flag * Drop the 'ti,disable-free-run' flag * Add 'Reviewed-by: Rob Herring <robh(a)kernel.org>' PATCH 4: * In probe() use dev_err for fixed error PATCH 7 (TDM): * set playback.max_channels to 128 * Add a check on tx_mask as the one done for rx_mask * Allow TDM with BP_FP format PATCH 9: * Detection of unexpected frame pulses is enabled by default PATCH 10: * Free-running mode is selected by the DAI format through SND_SOC_DAIFMT_[CONT/GATED] PATCH 12: * drop the 'ti,drive-dx' property * add 'ti,T1-framing-[rx/tx]' properties PATCH 13: * Drop the drive_dx part * Add support for 'T1 framing' with data delay set to 2 bit-clock periods Bastien Curutchet (13): ASoC: dt-bindings: davinci-mcbsp: convert McBSP bindings to yaml schema ASoC: dt-bindings: davinci-mcbsp: Add optional clock ASoC: ti: davinci-i2s: Remove the unused clk_input_pin attribute ASoC: ti: davinci-i2s: Replace dev_err with dev_err_probe ASoC: ti: davinci-i2s: Use external clock to drive sample rate generator ASoC: ti: davinci-i2s: Delete unnecessary assignment ASoC: ti: davinci-i2s: Add TDM support ASoC: ti: davinci-i2s: Add handling of BP_FC format ASoC: ti: davinci-i2s: Enable unexpected frame pulses detection ASoC: ti: davinci-i2s: Link free-run mode to SND_SOC_DAIFMT_[GATED/CONT] ASoC: ti: davinci-i2s: Add S24_LE to supported formats ASoC: dt-bindings: davinci-mcbsp: Add the 'ti,T1-framing-{rx/tx}' flags ASoC: ti: davinci-i2s: Add T1 framing support .../bindings/sound/davinci-mcbsp.txt | 50 ---- .../bindings/sound/davinci-mcbsp.yaml | 113 +++++++ include/linux/platform_data/davinci_asp.h | 15 - sound/soc/ti/davinci-i2s.c | 278 ++++++++++++++---- 4 files changed, 333 insertions(+), 123 deletions(-) delete mode 100644 Documentation/devicetree/bindings/sound/davinci-mcbsp.txt create mode 100644 Documentation/devicetree/bindings/sound/davinci-mcbsp.yaml -- 2.44.0

4 16