- 21 Apr, 2024 2 commits
-
-
Luca Ceresoli authored
To make the widget debugfs files more informative, add a line showing the widget type string. Keeping backward compatibility is nice to have being debugfs, and ease of parsing by both humans and software is also good. To maximize both with a reasonable effort add a new line without thouching the already complex format of the first line. The syntax is meant to be a key/value pair. The existing vizdapm tool continues working after this change, ignoring the new line. The new format is: Left ADC: Off in 1 out 0 - R2(0x2) mask 0x2 stream Left HiFi Capture inactive widget-type adc out "static" "Capture" "cs42l51.0-004a" in "static" "Left PGA" "cs42l51.0-004a" Signed-off-by:
Luca Ceresoli <luca.ceresoli@bootlin.com> Reviewed-by:
Alexandre Belloni <alexandre.belloni@bootlin.com> Link: https://lore.kernel.org/r/20240416-vizdapm-ng-v1-2-5d33c0b57bc5@bootlin.comSigned-off-by:
Mark Brown <broonie@kernel.org>
-
Luca Ceresoli authored
debugfs entries for DAPM widgets have the following form: Left ADC: Off in 1 out 0 - R2(0x2) mask 0x2 stream Left HiFi Capture inactive out "static" "Capture" in "static" "Left PGA" Lines with the "in" and "out" prefixes describe routes from/to other widgets presenting the path name and the connected widget name. This is ambiguous in case of cards having multiple widgets with the same name in different components. For example the STM32MP157A-DK1 board (arch/arm/boot/dts/st/stm32mp15xx-dkx.dtsi) has a "Capture" widget in both the "cs42l51.0-004a" and the "hdmi-audio-codec.1.auto" components. Avoid the ambiguity by adding the component name to the "in" and "out" lines. Add the new field at the end to minimize backward compatibility issues. The existing vizdapm tool continues working after this change. The output becomes: Left ADC: Off in 1 out 0 - R2(0x2) mask 0x2 stream Left HiFi Capture inactive out "static" "Capture" "cs42l51.0-004a" in "static" "Left PGA" "cs42l51.0-004a" Signed-off-by:
-
- 20 Apr, 2024 1 commit
-
-
Mark Brown authored
Merge series from Cezary Rojewski <cezary.rojewski@intel.com>: The existing code can be both improved and simplified. To make this change easier to manage, first add new implementation and then remove deadcode in a separate patch. Simplification achieved with: - reduce the amount of resources requested by the driver i.e.: IPC and CLDMA request_irq() merged into one - reduce the number of DSP ops from 2 to 1: irq_handler/thread() vs dsp_interrupt() - drop ambiguity around CLDMA interrupt, let skl.c handle that explicitly as it is the only user With that done, switch to the new implementation and remove unused members. While the change is non-trivial, from functional perspective status quo is achieved.
-
- 19 Apr, 2024 3 commits
-
-
Cezary Rojewski authored
Most IRQ-related code is duplicated in the driver. Switch to the new implementation and remove unused members. While the change is non-trivial, from functional perspective status quo is achieved. Signed-off-by:
Cezary Rojewski <cezary.rojewski@intel.com> Link: https://lore.kernel.org/r/20240419084857.2719593-3-cezary.rojewski@intel.comSigned-off-by:
-
Cezary Rojewski authored
The existing code can be both improved and simplified. To make this change easier to manage, first add new implementation and then remove deadcode in a separate patch. Simplification achieved with: - reduce the amount of resources requested by the driver i.e.: IPC and CLDMA request_irq() merged into one - reduce the number of DSP ops from 2 to 1: irq_handler/thread() vs dsp_interrupt() - drop ambiguity around CLDMA interrupt, let skl.c handle that explicitly as it is the only user Signed-off-by:
-
Mohammad Shehar Yaar Tausif authored
Convert NVIDIA Tegra20 DAS (Digital Audio Switch) binding to schema. Signed-off-by:
Mohammad Shehar Yaar Tausif <sheharyaar48@gmail.com> Reviewed-by:
Krzysztof Kozlowski <krzk@kernel.org> Link: https://lore.kernel.org/r/20240418163326.58365-1-sheharyaar48@gmail.comSigned-off-by:
-
- 17 Apr, 2024 4 commits
-
-
John Watts authored
pcm_formats should be a u64 as it is a SNDRV_PCM_FMTBIT_* not a SNDRV_PCM_FORMAT_*. Also fix a small grammar error while we're here. Reported-by:
kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202404170103.ySYwieqi-lkp@intel.com/Signed-off-by:
John Watts <contact@jookia.org> Link: https://lore.kernel.org/r/20240417-sunxi_s32_fix-v1-1-d82e451565c0@jookia.orgSigned-off-by:
-
Frank Li authored
Add ref: dai-common.yaml to fix below warning. arch/arm/boot/dts/nxp/imx/imx6q-sabreauto.dtb: esai@2024000: Unevaluated properties are not allowed ('#sound-dai-cells' was unexpected) Reviewed-by:Rob Herring <robh@kernel.org> Signed-off-by:
Frank Li <Frank.Li@nxp.com> Link: https://lore.kernel.org/r/20240416-esai_arm_dts_warning-v2-2-879e59c0c3b8@nxp.comSigned-off-by:
-
Frank Li authored
fsl,fifo-depth have default value 64 in driver(sound/soc/fsl/fsl_esai.c). fsl,esai-synchronous is flag(bool) type. It doesn't make sense to put flag type into 'required'. Fix warning: arch/arm/boot/dts/nxp/imx/imx6q-sabreauto.dtb: esai@2024000: 'fsl,fifo-depth' is a required property arch/arm/boot/dts/nxp/imx/imx6q-sabreauto.dtb: esai@2024000: 'fsl,esai-synchronous' is a required property Signed-off-by:
-
Mark Brown authored
Merge series from AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>: Changes in v5: - Cleaned up MT8186 soundcard migration commit which erroneously had leftovers from development - Changed the mtk_pcm_constraints_data structure to hold pointers to snd_pcm_hw_constraint_list, as to really reuse the const data - Tested again on all of the listed MTK platforms. Changes in v4: - Rebased over next-20240409 - Dropped the first 4 patches from v3 as were already picked by Mark - Fixed W=1 build issue 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 +- .../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 +++---- 25 files changed, 1597 insertions(+), 2190 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
-
- 16 Apr, 2024 17 commits
-
-
Steven Rostedt authored
The string SND_SOC_DAPM_DIR_OUT is printed in the snd_soc_dapm_path trace event instead of its value: (((REC->path_dir) == SND_SOC_DAPM_DIR_OUT) ? "->" : "<-") User space cannot parse this, as it has no idea what SND_SOC_DAPM_DIR_OUT is. Use TRACE_DEFINE_ENUM() to convert it to its value: (((REC->path_dir) == 1) ? "->" : "<-") So that user space tools, such as perf and trace-cmd, can parse it correctly. Reported-by:
Steven Rostedt (Google) <rostedt@goodmis.org> Link: https://lore.kernel.org/r/20240416000303.04670cdf@rorschach.local.homeSigned-off-by:
-
AngeloGioacchino Del Regno authored
Document the dai-link subnodes and the audio-routing property, allowing to describe machine specific audio hardware and links in device tree. While at it, also deprecate the old properties which were previously used with the driver's partially hardcoded configuration. Also, since describing machine specific audio hardware and links replaces the now deprecated old logic doing the same in a driver hardcoded fashion, it is not allowed to have both the old and new properties together. Reviewed-by:
AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Link: https://lore.kernel.org/r/20240416071410.75620-17-angelogioacchino.delregno@collabora.comSigned-off-by:
-
AngeloGioacchino Del Regno authored
Document the dai-link subnodes and the audio-routing property, allowing to describe machine specific audio hardware and links in device tree. While at it, also deprecate the old properties which were previously used with the driver's partially hardcoded configuration. Also, since describing machine specific audio hardware and links replaces the now deprecated old logic doing the same in a driver hardcoded fashion, it is not allowed to have both the old and new properties together. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Document the dai-link subnodes and the audio-routing property, allowing to describe machine specific audio hardware and links in device tree. While at it, also deprecate the old properties which were previously used with driver hardcoded configuration. Also, since describing machine specific audio hardware and links replaces the now deprecated old logic doing the same in a driver hardcoded fashion, it is not allowed to have both the old and new properties together. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Copy the few differences from mt8186-mt6166-da7219-max98357 in the mt8186-mt6366-rt1019-rt5682s driver to greatly reduce code duplication; since now the driver is meant to support MT8186 with the MT6366 PMIC codec and various combinations of I2S codecs, rename the driver to mt8186-mt6366 for consistency with MT8195 and MT8188, and rename the configuration option to SND_SOC_MT8186_MT6366. Since right now there is no machine using the da7219-max98357 yet, the snd_soc_dapm_route array was omitted as it's now possible to specify the audio routing in device trees instead. While at it, also add the missing sentinel comment to the last entry of the of_device_id array. Signed-off-by:
-
AngeloGioacchino Del Regno authored
Since the mtk-afe-platform-driver generic mtk_afe_pcm_platform now has a common .probe() callback, there is no reason to keep duplicating this function over and over in the SoC specific AFE-PCM drivers: switch over to register with the common bits instead. Note that MT8186 was left out of this because it is registering some extra sinegen controls in the AFE-PCM probe callback and needs extra cleanups to be able to use the common bits. Reviewed-by:
Alexandre Mergnat <amergnat@baylibre.com> Signed-off-by:
-
AngeloGioacchino Del Regno authored
Multiple MediaTek AFE PCM component drivers are using their own .probe() callback, but most of those are simply duplicated functions as they are doing exactly the same thing over and over. Add a common probe callback for this component to reduce duplication. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add a const mtk_pcm_constraints_data struct array with all of the (again, constant) constraints for all of the supported usecases, remove the duplicated functions and call mtk_soundcard_startup() instead in all of the .startup() callbacks. Signed-off-by:
-
AngeloGioacchino Del Regno authored
Add a const mtk_pcm_constraints_data struct array with all of the (again, constant) constraints for all of the supported usecases, remove the duplicated functions and call mtk_soundcard_startup() instead in all of the .startup() callbacks. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add a const mtk_pcm_constraints_data struct array with all of the (again, constant) constraints for all of the supported usecases, remove the duplicated functions and call mtk_soundcard_startup() instead in all of the .startup() callbacks. Reviewed-by:
-
AngeloGioacchino Del Regno authored
MediaTek platforms are typically setting PCM rate and channels constraints for playback, capture and HDMI/DisplayPort playback: commonize the startup callback by adding the PCM constraints data to the mtk_platform_card_data structure and by reusing the common mtk_soundcard_startup() function for all of them by getting back the parameters from the aforementioned struct. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add mtk_soundcard_pdata platform data for the MediaTek common sound card probe mechanism, including a driver/soc-specific probe extension (used for bits that cannot be commonized hence specific to this driver), and change the probe function to mtk_soundcard_common_probe. This is also adding the possibility of specifying the links and routing with the audio-routing property and (x)-dai-link nodes in device trees to stop hardcoding machine specific links in the card driver assupported by the common probe function, but support for legacy device trees is retained with a legacy_probe function, which is used only in case the new properties are not found. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add mtk_soundcard_pdata platform data for the MediaTek common sound card probe mechanism, including a driver/soc-specific probe extension (used for bits that cannot be commonized hence specific to this driver), and change the probe function to mtk_soundcard_common_probe. This is also adding the possibility of specifying the links and routing with the audio-routing property and (x)-dai-link nodes in device trees to stop hardcoding machine specific links in the card driver assupported by the common probe function, but support for legacy device trees is retained with a legacy_probe function, which is used only in case the new properties are not found. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add mtk_soundcard_pdata platform data for the MediaTek common sound card probe mechanism, including a driver/soc-specific probe extension (used for bits that cannot be commonized hence specific to this driver), and change the probe function to mtk_soundcard_common_probe. This is also adding the possibility of specifying the links and routing with the audio-routing property and (x)-dai-link nodes in device trees to stop hardcoding machine specific links in the card driver assupported by the common probe function, but support for legacy device trees is retained with a legacy_probe function, which is used only in case the new properties are not found. Signed-off-by:
-
AngeloGioacchino Del Regno authored
Add mtk_soundcard_pdata platform data for the MediaTek common sound card probe mechanism, including a driver/soc-specific probe extension (used for bits that cannot be commonized hence specific to this driver), and change the probe function to mtk_soundcard_common_probe. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Apart from a dai_link_list variable, the mtk_sof_priv currently holds data that never gets modified during runtime. Constify the mtk_sof_priv structure and move the SOF dai_link_list as sof_dai_link_list in struct mtk_soc_card_data, which is a structure that already holds the card's machine specific, runtime modified data. This allows to safely pass the mtk_sof_priv structure as platform data for the commonized card probe mechanism. Reviewed-by:
-
AngeloGioacchino Del Regno authored
Add a common machine soundcard driver probe function that supports both DSP and AFE-direct usecases and also provides a hook for legacy machine soundcard driver probe mechanisms. Note that the hook is there because, even for legacy probe, a lot of the actual code can still be commonized, hence still reducing duplication for the legacy devicetree retrocompatibility cases. This common probe function deprecates all of the inconsistent previous probe mechanisms and aims to settle all of the MediaTek card drivers on consistent and common devicetree properties describing wanted DAIs, device specific DAI configuration and DAI links to codecs found on each device/board. Reviewed-by:
-
- 15 Apr, 2024 8 commits
-
-
Shengjiu Wang authored
The wm8904 codec is used on an i.MX95 Toradex board. Signed-off-by:
Shengjiu Wang <shengjiu.wang@nxp.com> Link: https://lore.kernel.org/r/1713165456-3494-3-git-send-email-shengjiu.wang@nxp.comSigned-off-by:
-
Shengjiu Wang authored
wm8904 codec is used on i.MX95 Toradex board Signed-off-by:
-
John Watts authored
The I2S cores used in the H3 onwards support 32-bit sample rates. Support these by adding a per-variant PCM format list. Signed-off-by:
Jernej Skrabec <jernej.skrabec@gmail.com> Link: https://lore.kernel.org/r/20240403-sunxi_s32-v2-1-29ebf6ad590a@jookia.orgSigned-off-by:
-
Krzysztof Kozlowski authored
Neither core nor the driver modifes 'struct regulator_ops', so it can be const for code safety. Signed-off-by:
-
Krzysztof Kozlowski authored
The ID table already has respective entry and MODULE_DEVICE_TABLE and creates proper alias for platform driver. Having another MODULE_ALIAS causes the alias to be duplicated. Signed-off-by:
-
Krzysztof Kozlowski authored
The ID table already has respective entry and MODULE_DEVICE_TABLE and creates proper alias for platform driver. Having another MODULE_ALIAS causes the alias to be duplicated. Signed-off-by:
-
end.to.start authored
Add support microphone from Acer Aspire A315-24P and for some other similar devices with such vendor Signed-off-by:
"end.to.start" <end.to.start@mail.ru> Link: https://lore.kernel.org/r/20240412075533.10214-1-end.to.start@mail.ruSigned-off-by:
-
Mark Brown authored
Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>: More cleanups from Brent, notably the removal of the redundant cml_rt1011_rt5682 machine driver, fixes for SoundWire platforms and changes to sof_rt5682 to allow for 96+ sampling rates. For the rest of this kernel cycle, we are still working on SoundWire updates for MeteorLake (usual missing ACPI signature required for topology selection and jack detection information). We'll provide those patches as soon as they are reviewed/validated.
-
- 14 Apr, 2024 5 commits
-
-
Kuninori Morimoto authored
It is missing generic compatible for R-Car Gen4 Signed-off-by:
Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Link: https://lore.kernel.org/r/8734rrcewo.wl-kuninori.morimoto.gx@renesas.comSigned-off-by:
-
Kuninori Morimoto authored
Add document for R-Car V4M (R8A779H0). Signed-off-by:
-
Dan Carpenter authored
There are two issues here: 1) The get_device() needs a matching put_device() on error paths. 2) The "if (!ret)" was supposed to be "if (ret)". I re-arranged the code a bit to do the allocation before the get_device(). Fixes: ef7784e4 ("ASoC: soc-card: Add KUnit test case for snd_soc_card_get_kcontrol") Signed-off-by:
Dan Carpenter <dan.carpenter@linaro.org> Reviewed-by:
Richard Fitzgerald <rf@opensource.cirrus.com> Link: https://lore.kernel.org/r/450dd21a-b24b-48ba-9aa4-c02e4617852f@moroto.mountainSigned-off-by:
-
Fabio Estevam authored
Document fsl,imx25-pdk-sgtl5000 to fix the following dt-schema warning: imx25-pdk.dtb: sound: compatible: 'oneOf' conditional failed, one must be fixed: ['fsl,imx25-pdk-sgtl5000', 'fsl,imx-audio-sgtl5000'] is too long Fixes: 4189b542 ("ASoC: dt-bindings: fsl-asoc-card: convert to YAML") Signed-off-by:
Fabio Estevam <festevam@denx.de> Acked-by:
-
Bard Liao authored
When 96KHz sample rate is used, and MCLK is 24.576MHz, we will need pll_in = 24576000 and pll_out = 49152000 which is not supported by RT5682S_PLL2. Use RT5682S_PLL1 in this case. We don't test sample rate because RT5682S_PLL2 doesn't support 24.576MHz input and in the MCLK = 24.576MHz, sample rate = 48KHz case, i.e. pll_in == pll_out, PLL will not be used at all. Reviewed-by:
Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by:
Bard Liao <yung-chuan.liao@linux.intel.com> Signed-off-by:
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20240411220347.131267-13-pierre-louis.bossart@linux.intel.comSigned-off-by:
-
