July 2020 - Alsa-devel - mailman.alsa-project.org

[PATCH 0/2] Add period size constraint for Atom Chromebook
by Brent Lu 21 Aug '20

21 Aug '20

Two different constraints are implemented: one is in platform's CPU DAI to enforce period sizes which are already used in Android BSP. The other is in Atom Chromebook's machine driver to use 240 as period size. Brent Lu (1): ASoC: intel: atom: Add period size constraint Yu-Hsuan Hsu (1): ASoC: Intel: Add period size constraint on strago board sound/soc/intel/atom/sst-mfld-platform-pcm.c | 15 +++++++++++++++ sound/soc/intel/boards/cht_bsw_max98090_ti.c | 14 +++++++++++++- sound/soc/intel/boards/cht_bsw_rt5645.c | 14 +++++++++++++- 3 files changed, 41 insertions(+), 2 deletions(-) -- 2.7.4

7 53

[alsa-devel] Attempting to understand odd snd-usb-audio code and behavior
by Keith A. Milner 21 Aug '20

21 Aug '20

Over the last few weeks I've been trying to dig through the snd-usb-audio code to try to get an understanding of it. I've used ftrace and also a local version which I have heavily polluted with kprintf(KERN_INFO statements so I can track what it's doing in practice. I've been playing with this with some USB audio interfaces I have some which work, and some which don't, and trying to understand the differences and, ultimately, what I can do to make the non-functioning ones work. I've got to the point where I have a fairly good understanding of some parts, but I'm struggling to understand why the code is behaving as it does in practice. My observation when plugging the USB device in is that, as expected, it tries to build the audio device, and add components to it including attaching capture and playback streams. It seems to go through this phase without too many surprises, and we get to the standard dmesg output: usbcore: registered new interface driver snd-usb-audio After this, it gets confusing. It seems to open and close the PCMs (e.g, snd_usb_capture_open) several times in a row. Each time it sets up the hardware info, checks and sets the format and then closes the PCM. I don't see anything obviously different between most of these cycles. Here's an example my logging output (hopefully this will make sense): Aug 22 20:04:16 kerneldev kernel: [ 923.140789] Called: pcm.c - snd_usb_capture_open for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.140789] Called: pcm.c - snd_usb_pcm_open for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.140789] Setting up hw info Aug 22 20:04:16 kerneldev kernel: [ 923.140789] Checking format: Aug 22 20:04:16 kerneldev kernel: [ 923.140790] Interface number = 2 Aug 22 20:04:16 kerneldev kernel: [ 923.140790] alt setting = 1 Aug 22 20:04:16 kerneldev kernel: [ 923.140790] alt setting index = 1 Aug 22 20:04:16 kerneldev kernel: [ 923.140790] format type = 1 Aug 22 20:04:16 kerneldev kernel: [ 923.140790] frame size = 0 Aug 22 20:04:16 kerneldev kernel: [ 923.140791] Setting pt=125 ptmin=125 Aug 22 20:04:16 kerneldev kernel: [ 923.140792] Checking for unusual bit rates Aug 22 20:04:16 kerneldev kernel: [ 923.140794] Called: init.c - snd_card_file_remove Aug 22 20:04:16 kerneldev kernel: [ 923.140921] Called: pcm.c - snd_usb_hw_params for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.140945] Setting format: Aug 22 20:04:16 kerneldev kernel: [ 923.140945] Bytes per period = 8816 Aug 22 20:04:16 kerneldev kernel: [ 923.140945] Frames per period = 1102 Aug 22 20:04:16 kerneldev kernel: [ 923.140946] Periods per buffer = 4 Aug 22 20:04:16 kerneldev kernel: [ 923.140946] Channels = 2 Aug 22 20:04:16 kerneldev kernel: [ 923.140946] Rate = 44100 Aug 22 20:04:16 kerneldev kernel: [ 923.141279] Called: endpoint.c - snd_usb_add_endpoint of type 0 with direction 1 Aug 22 20:04:16 kerneldev kernel: [ 923.141292] Called: pcm.c - snd_usb_pcm_prepare for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.141293] Called: pcm.c - configure_endpoint Aug 22 20:04:16 kerneldev kernel: [ 923.141294] Called: endpoint.c - snd_usb_endpoint_set_params for endpoint 142, type 0 Aug 22 20:04:16 kerneldev kernel: [ 923.141294] Setting endpoint params for endpoint num 142 Aug 22 20:04:16 kerneldev kernel: [ 923.141294] channels=2 Aug 22 20:04:16 kerneldev kernel: [ 923.141294] period bytes=8816 Aug 22 20:04:16 kerneldev kernel: [ 923.141295] frames per period=1102 Aug 22 20:04:16 kerneldev kernel: [ 923.141295] periods per buffer=4 Aug 22 20:04:16 kerneldev kernel: [ 923.141296] Called: endpoint.c - data_ep_set_params for endpoint 142, type 0 Aug 22 20:04:16 kerneldev kernel: [ 923.141301] Called: pcm.c - snd_usb_pcm_prepare for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.142768] Called: init.c - snd_card_file_add Aug 22 20:04:16 kerneldev kernel: [ 923.142772] Called: init.c - snd_card_file_remove Aug 22 20:04:16 kerneldev kernel: [ 923.142774] Called: init.c - snd_card_file_add Aug 22 20:04:16 kerneldev kernel: [ 923.142830] Called: init.c - snd_card_file_remove Aug 22 20:04:16 kerneldev kernel: [ 923.142836] Called: pcm.c - snd_usb_hw_free for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.142837] Called: endpoint.c - snd_usb_endpoint_deactivate for endpoint 142, type 0 Aug 22 20:04:16 kerneldev kernel: [ 923.142846] Called: pcm.c - snd_usb_capture_close for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.146760] Called: init.c - snd_card_file_remove Aug 22 20:04:16 kerneldev kernel: [ 923.146834] Called: init.c - snd_card_file_add Aug 22 20:04:16 kerneldev kernel: [ 923.146838] Called: init.c - snd_card_file_remove Aug 22 20:04:16 kerneldev kernel: [ 923.146841] Called: init.c - snd_card_file_add Aug 22 20:04:16 kerneldev kernel: [ 923.146845] Called: init.c - snd_card_file_add Aug 22 20:04:16 kerneldev kernel: [ 923.146848] Called: pcm.c - snd_usb_capture_open for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.146849] Called: pcm.c - snd_usb_pcm_open for pcm substream 0 Aug 22 20:04:16 kerneldev kernel: [ 923.146849] Setting up hw info It seems to do this several times, before the behavior changing slightly, as it now starts to try to set the endpoint parameters and seems to leave the stream open. However, it then tries to open the stream and set the endpoint parameters again resulting in errors: Aug 22 20:22:00 kerneldev kernel: [ 1987.452945] Called: endpoint.c - snd_usb_endpoint_set_params for endpoint 142, type 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.452945] Setting endpoint params for endpoint num 142 Aug 22 20:22:00 kerneldev kernel: [ 1987.452945] channels=2 Aug 22 20:22:00 kerneldev kernel: [ 1987.452945] period bytes=1768 Aug 22 20:22:00 kerneldev kernel: [ 1987.452946] frames per period=221 Aug 22 20:22:00 kerneldev kernel: [ 1987.452946] periods per buffer=593 Aug 22 20:22:00 kerneldev kernel: [ 1987.452947] usb 1-1: Unable to change format on ep #8e: already in use Maybe this is driven by hotplug or something, but it seems to be performing a lot of pointless activities. Eventually it seems to try to open the PCM playback or capture stream (with some devices it opens playback, with others capture) and then checks the pointer: Aug 22 20:22:00 kerneldev kernel: [ 1987.470194] Called: pcm.c - snd_usb_substream_playback_trigger for pcm substream 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470198] Called: pcm.c - snd_usb_pcm_pointer for pcm substream 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470198] Pointer is 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470656] Called: pcm.c - snd_usb_pcm_pointer for pcm substream 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470657] Pointer is 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470660] Called: pcm.c - snd_usb_pcm_pointer for pcm substream 0 Aug 22 20:22:00 kerneldev kernel: [ 1987.470660] Pointer is 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.473567] Called: pcm.c - snd_usb_pcm_pointer for pcm substream 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.473569] Pointer is 0 The above is from a non-working device. For a working device the pointer increments. Eventually this activity ends: Aug 22 20:22:05 kerneldev kernel: [ 1992.473575] Called: pcm.c - snd_usb_substream_playback_trigger for pcm substream 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.473576] Called: endpoint.c - snd_usb_endpoint_stop for endpoint 142, type 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.473578] Called: endpoint.c - snd_usb_endpoint_stop for endpoint 13, type 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.473602] Called: pcm.c - snd_usb_hw_free for pcm substream 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.503287] Called: endpoint.c - snd_usb_endpoint_deactivate for endpoint 142, type 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.503288] Called: endpoint.c - snd_usb_endpoint_deactivate for endpoint 13, type 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.503323] Called: pcm.c - snd_usb_playback_close for pcm substream 0 Aug 22 20:22:05 kerneldev kernel: [ 1992.503656] Called: init.c - snd_card_file_remove This seems as if it might be some sort of test of the PCMs, to make sure they are working. Certainly in non-working devices the pcm_pointer stays at zero and in working devices it increments. It also seems to try harder with some devices than others, even though the device descriptors seem very similar. Can anyone explain any of this to me, particularly why it's exhibiting this behaviour. I've uploaded one of the logs from a non-working device to Pastebin if anyone wants to look at a full example: https://pastebin.com/uBZEuQ57 Also, is this the best way to approach this, or are there other tools/techniques that can help? I've experimented with ftrace and that was useful, but didn't give the level of detail I wanted. Cheers, Keith

3 10

[PATCH] ASoC: Intel: skl_hda_dsp_generic: Fix NULLptr dereference in autosuspend delay
by Mateusz Gorski 19 Aug '20

19 Aug '20

Different modules for HDMI codec are used depending on the "hda_codec_use_common_hdmi" option being enabled or not. Driver private context for both of them is different. This leads to null-pointer dereference error when driver tries to set autosuspend delay for HDMI codec while the option is off (hdac_hdmi module is used for HDMI). Change the string in conditional statement to "ehdaudio0D0" to ensure that only the HDAudio codec is handled by this function. Fixes: 5bf73b1b1dec ("ASoC: intel/skl/hda - fix oops on systems without i915 audio codec") Signed-off-by: Mateusz Gorski <mateusz.gorski(a)linux.intel.com> --- sound/soc/intel/boards/skl_hda_dsp_generic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/intel/boards/skl_hda_dsp_generic.c b/sound/soc/intel/boards/skl_hda_dsp_generic.c index ca4900036ead..bc50eda297ab 100644 --- a/sound/soc/intel/boards/skl_hda_dsp_generic.c +++ b/sound/soc/intel/boards/skl_hda_dsp_generic.c @@ -181,7 +181,7 @@ static void skl_set_hda_codec_autosuspend_delay(struct snd_soc_card *card) struct snd_soc_dai *dai; for_each_card_rtds(card, rtd) { - if (!strstr(rtd->dai_link->codecs->name, "ehdaudio")) + if (!strstr(rtd->dai_link->codecs->name, "ehdaudio0D0")) continue; dai = asoc_rtd_to_codec(rtd, 0); hda_pvt = snd_soc_component_get_drvdata(dai->component); -- 2.17.1

6 7

[PATCH v2 0/3] Fix Kconfig dependency issue with DMAENGINES selection
by Laurent Pinchart 19 Aug '20

19 Aug '20

Hello, This small series fixes a Kconfig dependency issue with the recently merged Xilixn DPSUB DRM/KMS driver. The fix is in patch 3/3, but requires a separate fixes in patches 1/3 and 2/3 to avoid circular dependencies: drivers/i2c/Kconfig:8:error: recursive dependency detected! drivers/i2c/Kconfig:8: symbol I2C is selected by FB_DDC drivers/video/fbdev/Kconfig:63: symbol FB_DDC depends on FB drivers/video/fbdev/Kconfig:12: symbol FB is selected by DRM_KMS_FB_HELPER drivers/gpu/drm/Kconfig:80: symbol DRM_KMS_FB_HELPER depends on DRM_KMS_HELPER drivers/gpu/drm/Kconfig:74: symbol DRM_KMS_HELPER is selected by DRM_ZYNQMP_DPSUB drivers/gpu/drm/xlnx/Kconfig:1: symbol DRM_ZYNQMP_DPSUB depends on DMA_ENGINE drivers/dma/Kconfig:44: symbol DMA_ENGINE depends on DMADEVICES drivers/dma/Kconfig:6: symbol DMADEVICES is selected by SND_SOC_SH4_SIU sound/soc/sh/Kconfig:30: symbol SND_SOC_SH4_SIU is selected by SND_SIU_MIGOR sound/soc/sh/Kconfig:60: symbol SND_SIU_MIGOR depends on I2C For a resolution refer to Documentation/kbuild/kconfig-language.rst subsection "Kconfig recursive dependency limitations" Due to the DPSUB driver being merged in v5.9, this is a candidate fix for v5.9 as well. 1/3 and 2/3 can be merged independently, 3/3 depends on the first two. What's the best course of action, can I merge this all in a single tree, or should the rapidio and ASoC patches be merged independently early in the -rc cycle, and the DRM patch later on top ? I don't expect conflicts (especially in 2/3 and 3/3), so merging the whole series in one go would be simpler in my opinion. Compared to v1, commit messages have been fixed to mention DMADEVICES instead of DMAENGINES. Laurent Pinchart (3): rapidio: Replace 'select' DMADEVICES 'with depends on' ASoC: sh: Replace 'select' DMADEVICES 'with depends on' drm: xlnx: dpsub: Fix DMADEVICES Kconfig dependency drivers/gpu/drm/xlnx/Kconfig | 1 + drivers/rapidio/Kconfig | 2 +- sound/soc/sh/Kconfig | 2 +- 3 files changed, 3 insertions(+), 2 deletions(-) -- Regards, Laurent Pinchart

3 10

[PATCH RFC] ASoC: wm8994: Avoid attempts to read unreadable registers
by Sylwester Nawrocki 18 Aug '20

18 Aug '20

The driver supports WM1811, WM8994, WM8958 devices but according to documentation and the regmap definitions the WM8958_DSP2_* registers are only available on WM8958. In current code these registers are being accessed as if they were available on all the three chips. When starting playback on WM1811 CODEC multiple errors like: "wm8994-codec wm8994-codec: ASoC: error at soc_component_read_no_lock on wm8994-codec: -5" can be seen, which is caused by attempts to read an unavailable WM8958_DSP2_PROGRAM register. The issue has been uncovered by recent commit "e2329ee ASoC: soc-component: add soc_component_err()". This patch adds a check in wm8958_aif_ev() callback so the DSP2 handling is only done for WM8958. Signed-off-by: Sylwester Nawrocki <s.nawrocki(a)samsung.com> --- sound/soc/codecs/wm8958-dsp2.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/sound/soc/codecs/wm8958-dsp2.c b/sound/soc/codecs/wm8958-dsp2.c index 68a3b48..3bce9a1 100644 --- a/sound/soc/codecs/wm8958-dsp2.c +++ b/sound/soc/codecs/wm8958-dsp2.c @@ -412,8 +412,12 @@ int wm8958_aif_ev(struct snd_soc_dapm_widget *w, struct snd_kcontrol *kcontrol, int event) { struct snd_soc_component *component = snd_soc_dapm_to_component(w->dapm); + struct wm8994 *control = dev_get_drvdata(component->dev->parent); int i; + if (control->type != WM8958) + return 0; + switch (event) { case SND_SOC_DAPM_POST_PMU: case SND_SOC_DAPM_PRE_PMU: -- 2.7.4

3 2

[PATCH v3 00/10] ASoC: qdsp6: add gapless compressed audio support
by Srinivas Kandagatla 18 Aug '20

18 Aug '20

This patchset adds gapless compressed audio support on q6asm. Gapless on q6asm is implemented using 2 streams in a single q6asm session. First few patches such as stream id per each command, gapless flags and silence meta data are for preparedness for adding gapless support. Last patch implements copy callback to allow finer control over buffer offsets, specially in partial drain cases. This patchset is tested on RB3 aka DB845c platform. This patchset as it is will support gapless however QDSP can also support switching decoders on a single stream. Patches to support such feature are send in different patchset which involves adding generic interfaces. Thanks, srini Changes since v2:(mostly suggested by Pierre) - removed unnessary kernel style comments, - moved TIMESTAMP flag to respective patch. - move preparatory code from gapless support patch to new one. - fix subject prefix of one patch. - add comments to clarify valid stream_ids Srinivas Kandagatla (10): ASoC: q6asm: rename misleading session id variable ASoC: q6asm: make commands specific to streams ASoC: q6asm: use flags directly from q6asm-dai ASoC: q6asm: add length to write command token ASoC: q6asm: add support to remove intial and trailing silence ASoC: q6asm: add support to gapless flag in q6asm open ASoC: q6asm-dai: add next track metadata support ASoC: q6asm-dai: prepare set params to accept profile change ASoC: q6asm-dai: add gapless support ASoC: q6asm-dai: add support to copy callback sound/soc/qcom/qdsp6/q6asm-dai.c | 414 +++++++++++++++++++++++-------- sound/soc/qcom/qdsp6/q6asm.c | 169 +++++++++---- sound/soc/qcom/qdsp6/q6asm.h | 49 ++-- 3 files changed, 469 insertions(+), 163 deletions(-) -- 2.21.0

1 10

[PATCH v2 0/8] ASoC: sun50i-codec-analog: Cleanup and power management
by Samuel Holland 18 Aug '20

18 Aug '20

This series performs some minor cleanup on the driver for the analog codec in the Allwinner A64, and hooks up the existing mute switches to DAPM widgets, in order to provide improved power management. Changes since v1: - Collected Acked-by/Reviewed-by tags - Used SOC_MIXER_NAMED_CTL_ARRAY to avoid naming a widget "Earpiece" Samuel Holland (8): ASoC: sun50i-codec-analog: Fix duplicate use of ADC enable bits ASoC: sun50i-codec-analog: Gate the amplifier clock during suspend ASoC: sun50i-codec-analog: Group and sort mixer routes ASoC: sun50i-codec-analog: Make headphone routes stereo ASoC: sun50i-codec-analog: Enable DAPM for headphone switch ASoC: sun50i-codec-analog: Make line out routes stereo ASoC: sun50i-codec-analog: Enable DAPM for line out switch ASoC: sun50i-codec-analog: Enable DAPM for earpiece switch sound/soc/sunxi/sun50i-codec-analog.c | 176 ++++++++++++++++---------- 1 file changed, 111 insertions(+), 65 deletions(-) -- 2.26.2

2 9

[PATCH 00/13] soundwire: intel: add power management support
by Bard Liao 18 Aug '20

18 Aug '20

This series adds power management support for Intel soundwire links. Bard Liao (1): soundwire: intel: reinitialize IP+DSP in .prepare(), but only when resuming Pierre-Louis Bossart (10): soundwire: intel: Add basic power management support soundwire: intel: add pm_runtime support soundwire: intel: reset pm_runtime status during system resume soundwire: intel: fix race condition on system resume soundwire: intel: call helper to reset Slave states on resume soundwire: intel: pm_runtime idle scheduling soundwire: intel: add CLK_STOP_TEARDOWN for pm_runtime suspend soundwire: intel: add CLK_STOP_NOT_ALLOWED support soundwire: intel_init: handle power rail dependencies for clock stop mode soundwire: intel: support clock_stop mode without quirks Rander Wang (2): soundwire: intel: add CLK_STOP_BUS_RESET support soundwire: intel: refine runtime pm for SDW_INTEL_CLK_STOP_BUS_RESET drivers/soundwire/cadence_master.h | 4 + drivers/soundwire/intel.c | 448 ++++++++++++++++++++++++++++- drivers/soundwire/intel.h | 2 + drivers/soundwire/intel_init.c | 19 +- 4 files changed, 465 insertions(+), 8 deletions(-) -- 2.17.1

3 21

[PATCH v3 1/3] ASoC: samsung: Document DT bindings for Midas sound subsystem
by Sylwester Nawrocki 17 Aug '20

17 Aug '20

This patch adds documentation of DT biding for the Midas sound complex. Partially based on the *txt version by Simon Shields <simon(a)lineageos.org>. Signed-off-by: Sylwester Nawrocki <s.nawrocki(a)samsung.com> --- Changes for v3: - change the sound-dai property type in cpu, codec nodes from phandle-array to phandle Changes for v2: - fix wrong *-gpios entries in the example, - mark sound-dai properties as required. --- .../bindings/sound/samsung,midas-audio.yaml | 108 +++++++++++++++++++++ 1 file changed, 108 insertions(+) create mode 100644 Documentation/devicetree/bindings/sound/samsung,midas-audio.yaml diff --git a/Documentation/devicetree/bindings/sound/samsung,midas-audio.yaml b/Documentation/devicetree/bindings/sound/samsung,midas-audio.yaml new file mode 100644 index 0000000..1c755de --- /dev/null +++ b/Documentation/devicetree/bindings/sound/samsung,midas-audio.yaml @@ -0,0 +1,108 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/sound/samsung,midas-audio.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung Midas audio complex with WM1811 codec + +maintainers: + - Sylwester Nawrocki <s.nawrocki(a)samsung.com> + +properties: + compatible: + const: samsung,midas-audio + + model: + $ref: /schemas/types.yaml#/definitions/string + description: The user-visible name of this sound complex. + + cpu: + type: object + properties: + sound-dai: + $ref: /schemas/types.yaml#/definitions/phandle + description: phandle to the I2S controller + required: + - sound-dai + + codec: + type: object + properties: + sound-dai: + $ref: /schemas/types.yaml#/definitions/phandle + description: phandle to the WM1811 CODEC + required: + - sound-dai + + samsung,audio-routing: + $ref: /schemas/types.yaml#/definitions/non-unique-string-array + description: | + List of the connections between audio components; each entry is + a pair of strings, the first being the connection's sink, the second + being the connection's source; valid names for sources and sinks are + the WM1811's pins (as documented in its binding), and the jacks + on the board: HP, SPK, Main Mic, Sub Mic, Headset Mic. + + mic-bias-supply: + description: Supply for the micbias on the Main microphone + + submic-bias-supply: + description: Supply for the micbias on the Sub microphone + + fm-sel-gpios: + description: GPIO pin for FM selection + + lineout-sel-gpios: + description: GPIO pin for line out selection + +required: + - compatible + - model + - cpu + - codec + - samsung,audio-routing + - mic-bias-supply + - submic-bias-supply + +additionalProperties: false + +examples: + - | + #include <dt-bindings/gpio/gpio.h> + + sound { + compatible = "samsung,midas-audio"; + model = "Midas"; + + fm-sel-gpios = <&gpaa0 3 GPIO_ACTIVE_HIGH>; + + mic-bias-supply = <&mic_bias_reg>; + submic-bias-supply = <&submic_bias_reg>; + + samsung,audio-routing = + "HP", "HPOUT1L", + "HP", "HPOUT1R", + + "SPK", "SPKOUTLN", + "SPK", "SPKOUTLP", + "SPK", "SPKOUTRN", + "SPK", "SPKOUTRP", + + "RCV", "HPOUT2N", + "RCV", "HPOUT2P", + + "IN1LP", "Main Mic", + "IN1LN", "Main Mic", + "IN1RP", "Sub Mic", + "IN1LP", "Sub Mic"; + + cpu { + sound-dai = <&i2s0>; + }; + + codec { + sound-dai = <&wm1811>; + }; + + }; -- 2.7.4

3 5

[PATCH v2] soundwire: master: enable pm runtime
by Bard Liao 17 Aug '20

17 Aug '20

The hierarchy of soundwire devices is platform device -> M device -> S device. A S device is physically attached on the platform device. So the platform device should be resumed when a S device is resumed. As the bridge of platform device and S device, we have to implement runtime pm on M driver. We have set runtime pm ops in M driver already, but still need to enable runtime pm. Signed-off-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> --- Changes in v2: - Edit the commit message to explain why we need the patch. --- drivers/soundwire/master.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/soundwire/master.c b/drivers/soundwire/master.c index 5f0b2189defe..3488bb824e84 100644 --- a/drivers/soundwire/master.c +++ b/drivers/soundwire/master.c @@ -154,6 +154,7 @@ int sdw_master_device_add(struct sdw_bus *bus, struct device *parent, bus->dev = &md->dev; bus->md = md; + pm_runtime_enable(&bus->md->dev); device_register_err: return ret; } @@ -166,6 +167,7 @@ int sdw_master_device_add(struct sdw_bus *bus, struct device *parent, */ int sdw_master_device_del(struct sdw_bus *bus) { + pm_runtime_disable(&bus->md->dev); device_unregister(bus->dev); return 0; -- 2.17.1

2 1