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

[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

[PATCH v2] ASoC: Intel: boards: eve: Fix DMIC records zero
by Harsha Priya 13 Aug '20

13 Aug '20

This patch adds a dapm route to provide early mclk/sclk for for DMIC on SSP0(rt5514) and Headset on SSP1(rt5663). The sclk rate for both codecs is different which is taken care of in the platform_clock_control().The platform has one mclk and one sclk. Two variables sclk0 and sclk1 are used for the two codecs on differnet ssp that use different clock rate. This change ensures the DMIC PCM port will return valid data Signed-off-by: Brent Lu <brent.lu(a)intel.com> Signed-off-by: Vamshi Krishna Gopal <vamshi.krishna.gopal(a)intel.com> Signed-off-by: Harsha Priya <harshapriya.n(a)intel.com> --- v1 -> v2: - Only one mclk with same rate is used, so changed to using one variable - dropping ssp_ prefix from sclk variable names to make them sound right. - removing a return statment that was not required - fixed commit message accordingly .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c | 146 ++++++++++++++------- 1 file changed, 97 insertions(+), 49 deletions(-) diff --git a/sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c b/sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c index b34cf6c..155f2b4 100644 --- a/sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +++ b/sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c @@ -54,7 +54,8 @@ struct kbl_codec_private { struct list_head hdmi_pcm_list; struct snd_soc_jack kabylake_hdmi[2]; struct clk *mclk; - struct clk *sclk; + struct clk *sclk0; + struct clk *sclk1; }; enum { @@ -77,13 +78,29 @@ static const struct snd_kcontrol_new kabylake_controls[] = { }; static int platform_clock_control(struct snd_soc_dapm_widget *w, - struct snd_kcontrol *k, int event) + struct snd_kcontrol *k, int event, int ssp_num) { struct snd_soc_dapm_context *dapm = w->dapm; struct snd_soc_card *card = dapm->card; struct kbl_codec_private *priv = snd_soc_card_get_drvdata(card); + struct clk *sclk; + unsigned long sclk_rate; int ret = 0; + switch (ssp_num) { + case 0: + sclk = priv->sclk0; + sclk_rate = 6144000; + break; + case 1: + sclk = priv->sclk1; + sclk_rate = 3072000; + break; + default: + dev_err(card->dev, "Invalid ssp_num %d\n", ssp_num); + return -EINVAL; + } + /* * MCLK/SCLK need to be ON early for a successful synchronization of * codec internal clock. And the clocks are turned off during @@ -91,38 +108,48 @@ static int platform_clock_control(struct snd_soc_dapm_widget *w, */ switch (event) { case SND_SOC_DAPM_PRE_PMU: - /* Enable MCLK */ ret = clk_set_rate(priv->mclk, 24000000); if (ret < 0) { - dev_err(card->dev, "Can't set rate for mclk, err: %d\n", - ret); - return ret; + dev_err(card->dev, "Can't set rate for mclk for ssp%d, err: %d\n", + ssp_num, ret); + return ret; } - ret = clk_prepare_enable(priv->mclk); - if (ret < 0) { - dev_err(card->dev, "Can't enable mclk, err: %d\n", ret); - return ret; + if (!__clk_is_enabled(priv->mclk)) { + /* Enable MCLK */ + ret = clk_prepare_enable(priv->mclk); + if (ret < 0) { + dev_err(card->dev, "Can't enable mclk for ssp%d, err: %d\n", + ssp_num, ret); + return ret; + } } - /* Enable SCLK */ - ret = clk_set_rate(priv->sclk, 3072000); + ret = clk_set_rate(sclk, sclk_rate); if (ret < 0) { - dev_err(card->dev, "Can't set rate for sclk, err: %d\n", - ret); + dev_err(card->dev, "Can't set rate for sclk for ssp%d, err: %d\n", + ssp_num, ret); clk_disable_unprepare(priv->mclk); return ret; } - ret = clk_prepare_enable(priv->sclk); - if (ret < 0) { - dev_err(card->dev, "Can't enable sclk, err: %d\n", ret); - clk_disable_unprepare(priv->mclk); + if (!__clk_is_enabled(sclk)) { + /* Enable SCLK */ + ret = clk_prepare_enable(sclk); + if (ret < 0) { + dev_err(card->dev, "Can't enable sclk for ssp%d, err: %d\n", + ssp_num, ret); + clk_disable_unprepare(priv->mclk); + return ret; + } } break; case SND_SOC_DAPM_POST_PMD: - clk_disable_unprepare(priv->mclk); - clk_disable_unprepare(priv->sclk); + if (__clk_is_enabled(priv->mclk)) + clk_disable_unprepare(priv->mclk); + + if (__clk_is_enabled(sclk)) + clk_disable_unprepare(sclk); break; default: return 0; @@ -131,6 +158,18 @@ static int platform_clock_control(struct snd_soc_dapm_widget *w, return 0; } +static int platform_clock_control_ssp0(struct snd_soc_dapm_widget *w, + struct snd_kcontrol *k, int event) +{ + return platform_clock_control(w, k, event, 0); +} + +static int platform_clock_control_ssp1(struct snd_soc_dapm_widget *w, + struct snd_kcontrol *k, int event) +{ + return platform_clock_control(w, k, event, 1); +} + static const struct snd_soc_dapm_widget kabylake_widgets[] = { SND_SOC_DAPM_HP("Headphone Jack", NULL), SND_SOC_DAPM_MIC("Headset Mic", NULL), @@ -139,15 +178,17 @@ static const struct snd_soc_dapm_widget kabylake_widgets[] = { SND_SOC_DAPM_MIC("DMIC", NULL), SND_SOC_DAPM_SPK("HDMI1", NULL), SND_SOC_DAPM_SPK("HDMI2", NULL), - SND_SOC_DAPM_SUPPLY("Platform Clock", SND_SOC_NOPM, 0, 0, - platform_clock_control, SND_SOC_DAPM_PRE_PMU | + SND_SOC_DAPM_SUPPLY("Platform Clock SSP0", SND_SOC_NOPM, 0, 0, + platform_clock_control_ssp0, SND_SOC_DAPM_PRE_PMU | + SND_SOC_DAPM_POST_PMD), + SND_SOC_DAPM_SUPPLY("Platform Clock SSP1", SND_SOC_NOPM, 0, 0, + platform_clock_control_ssp1, SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMD), - }; static const struct snd_soc_dapm_route kabylake_map[] = { /* Headphones */ - { "Headphone Jack", NULL, "Platform Clock" }, + { "Headphone Jack", NULL, "Platform Clock SSP1" }, { "Headphone Jack", NULL, "HPOL" }, { "Headphone Jack", NULL, "HPOR" }, @@ -156,7 +197,7 @@ static const struct snd_soc_dapm_route kabylake_map[] = { { "Right Spk", NULL, "Right BE_OUT" }, /* other jacks */ - { "Headset Mic", NULL, "Platform Clock" }, + { "Headset Mic", NULL, "Platform Clock SSP1" }, { "IN1P", NULL, "Headset Mic" }, { "IN1N", NULL, "Headset Mic" }, @@ -180,6 +221,7 @@ static const struct snd_soc_dapm_route kabylake_map[] = { { "ssp0 Rx", NULL, "Right HiFi Capture" }, /* DMIC */ + { "DMIC", NULL, "Platform Clock SSP0" }, { "DMIC1L", NULL, "DMIC" }, { "DMIC1R", NULL, "DMIC" }, { "DMIC2L", NULL, "DMIC" }, @@ -666,7 +708,7 @@ static int kabylake_set_bias_level(struct snd_soc_card *card, if (!component || strcmp(component->name, RT5514_DEV_NAME)) return 0; - if (IS_ERR(priv->mclk)) + if (IS_ERR(priv->mclk0)) return 0; /* @@ -757,6 +799,28 @@ static struct snd_soc_card kabylake_audio_card = { .late_probe = kabylake_card_late_probe, }; +static int kabylake_audio_clk_get(struct device *dev, const char *id, + struct clk **clk) +{ + int ret = 0; + + if (!clk) + return -EINVAL; + + *clk = devm_clk_get(dev, id); + if (IS_ERR(*clk)) { + ret = PTR_ERR(*clk); + if (ret == -ENOENT) { + dev_info(dev, "Failed to get %s, defer probe\n", id); + return -EPROBE_DEFER; + } + + dev_err(dev, "Failed to get %s with err:%d\n", id, ret); + } + + return ret; +} + static int kabylake_audio_probe(struct platform_device *pdev) { struct kbl_codec_private *ctx; @@ -777,33 +841,17 @@ static int kabylake_audio_probe(struct platform_device *pdev) dmic_constraints = mach->mach_params.dmic_num == 2 ? &constraints_dmic_2ch : &constraints_dmic_channels; - ctx->mclk = devm_clk_get(&pdev->dev, "ssp1_mclk"); - if (IS_ERR(ctx->mclk)) { - ret = PTR_ERR(ctx->mclk); - if (ret == -ENOENT) { - dev_info(&pdev->dev, - "Failed to get ssp1_mclk, defer probe\n"); - return -EPROBE_DEFER; - } - - dev_err(&pdev->dev, "Failed to get ssp1_mclk with err:%d\n", - ret); + ret = kabylake_audio_clk_get(&pdev->dev, "ssp0_sclk", &ctx->sclk0); + if (ret != 0) return ret; - } - ctx->sclk = devm_clk_get(&pdev->dev, "ssp1_sclk"); - if (IS_ERR(ctx->sclk)) { - ret = PTR_ERR(ctx->sclk); - if (ret == -ENOENT) { - dev_info(&pdev->dev, - "Failed to get ssp1_sclk, defer probe\n"); - return -EPROBE_DEFER; - } + ret = kabylake_audio_clk_get(&pdev->dev, "ssp1_mclk", &ctx->mclk); + if (ret != 0) + return ret; - dev_err(&pdev->dev, "Failed to get ssp1_sclk with err:%d\n", - ret); + ret = kabylake_audio_clk_get(&pdev->dev, "ssp1_sclk", &ctx->sclk1); + if (ret != 0) return ret; - } return devm_snd_soc_register_card(&pdev->dev, &kabylake_audio_card); } -- 2.7.4

5 9

[PATCH 0/3] Modernize tasklet callback API
by Kees Cook 12 Aug '20

12 Aug '20

Hi, This is the infrastructure changes to prepare the tasklet API for conversion to passing the tasklet struct as the callback argument instead of an arbitrary unsigned long. The first patch details why this is useful (it's the same rationale as the timer_struct changes from a bit ago: less abuse during memory corruption attacks, more in line with existing ways of doing things in the kernel, save a little space in struct, etc). Notably, the existing tasklet API use is much less messy, so there is less to clean up. It's not clear to me which tree this should go through... Greg since it starts with a USB clean-up, -tip for timer or interrupt, or if I should just carry it. I'm open to suggestions, but if I don't hear otherwise, I'll just carry it. My goal is to have this merged for v5.9-rc1 so that during the v5.10 development cycle the new API will be available. The entire tree of changes is here[1] currently, but to split it up by maintainer the infrastructure changes need to be landed first. Review and Acks appreciated! :) Thanks, -Kees [1] https://github.com/allenpais/tasklets/commits/tasklets_V2 Kees Cook (2): usb: gadget: udc: Avoid tasklet passing a global treewide: Replace DECLARE_TASKLET() with DECLARE_TASKLET_OLD() Romain Perier (1): tasklet: Introduce new initialization API drivers/input/keyboard/omap-keypad.c | 2 +- drivers/input/serio/hil_mlc.c | 2 +- drivers/net/wan/farsync.c | 4 +-- drivers/s390/crypto/ap_bus.c | 2 +- drivers/staging/most/dim2/dim2.c | 2 +- drivers/staging/octeon/ethernet-tx.c | 2 +- drivers/tty/vt/keyboard.c | 2 +- drivers/usb/gadget/udc/snps_udc_core.c | 6 ++--- drivers/usb/host/fhci-sched.c | 2 +- include/linux/interrupt.h | 37 ++++++++++++++++++++++---- kernel/backtracetest.c | 2 +- kernel/debug/debug_core.c | 2 +- kernel/irq/resend.c | 2 +- kernel/softirq.c | 18 ++++++++++++- net/atm/pppoatm.c | 2 +- net/iucv/iucv.c | 2 +- sound/drivers/pcsp/pcsp_lib.c | 2 +- 17 files changed, 66 insertions(+), 25 deletions(-) -- 2.25.1

10 26

USB audio cuts 2-3 secs at start
by Stuart Naylor 11 Aug '20

11 Aug '20

http://alsa-project.org/db/?f=6b107178c9939f235a390c07cb5c1f71d6a6e0c0 Its an Anker power conf and for some reason the first couple of seconds of aplay do not play via USB. Alsa-info above anyone and tips or ideas? Stuart

2 2

[PATCH 0/4] ASoC: Intel: KMB: TDM Enablement patches
by Michael Sit Wei Hong 10 Aug '20

10 Aug '20

This patch series is to enable multiple features on the Keembay Platform Michael Sit Wei Hong (4): ASoC: Intel: KMB: Add 8kHz audio support ASoC: Intel: KMB: Rework disable channel function ASoC: Intel: KMB: Enable TDM audio capture dt-bindings: sound: intel,keembay-i2s: Add channel-max property .../bindings/sound/intel,keembay-i2s.yaml | 8 + sound/soc/intel/keembay/kmb_platform.c | 137 +++++++++++++----- sound/soc/intel/keembay/kmb_platform.h | 1 + 3 files changed, 112 insertions(+), 34 deletions(-) -- 2.17.1

3 14

pcm_meter.c issue at s16_update
by Pavel Hofman 09 Aug '20

09 Aug '20

Hi, I am debugging the following problem with the PCM METER API: An application with meter/scope configured in .asoundrc hits a high CPU load sometimes when a new playback stream is opened. E.g. in MPD when opening a new radio stream. The 100% CPU core load takes tens of seconds. Debugging the issue revealed this: TL;DR: at decrease of [status.appl_ptr - status.delay] between consequent runs the size of buffer to convert by the built-in s16 scope is set from usual 100s to huge value of pcm->boundary (1.5G), causing a very long processing at 100% core load. Details: Every scope setup uses a default s16 scope defined in alsa-lib, which converts samples to s16 so that the user-provided scopes can access data via snd_pcm_scope_s16_get_channel_buffer buffers with converted samples. The s16 scope converts samples in a loop, in frames chunks, decrementing size variable by frames until zero. The value of size is hundreds samples max at each s16_update call, calculated in https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_meter.c#L1… . A debug added after that line shows this: @@ -1098,14 +1103,18 @@ static void s16_update(snd_pcm_scope_t *scope) snd_pcm_sframes_t size; snd_pcm_uframes_t offset; size = meter->now - s16->old; + fprintf(stderr, "s16_update 1: meter->now %ld, s16->old %ld, size %ld\n", meter->now, s16->old, size); if (size < 0) size += spcm->boundary; offset = s16->old % meter->buf_size; s16_update 1: meter->now 2154673, s16->old 2153771, size 902 s16_update 1: meter->now 2155579, s16->old 2154673, size 906 s16_update 1: meter->now 2156487, s16->old 2155579, size 908 s16_update 1: meter->now 2157391, s16->old 2156487, size 904 s16_update 1: meter->now 2158298, s16->old 2157391, size 907 s16_update 1: meter->now 2159203, s16->old 2158298, size 905 s16_update 1: meter->now 2160091, s16->old 2159203, size 888 s16_update 1: meter->now 2160976, s16->old 2160091, size 885 s16_update 1: meter->now 2161924, s16->old 2160976, size 948 s16_update 1: meter->now 2162829, s16->old 2161924, size 905 s16_update 1: meter->now 2163733, s16->old 2162829, size 904 s16_update 1: meter->now 2164594, s16->old 2163733, size 861 s16_update 1: meter->now 2165542, s16->old 2164594, size 948 s16_update 1: meter->now 2166448, s16->old 2165542, size 906 s16_update 1: meter->now 2167352, s16->old 2166448, size 904 s16_update 1: meter->now 2168259, s16->old 2167352, size 907 s16_update 1: meter->now 2169165, s16->old 2168259, size 906 s16_update 1: meter->now 2170050, s16->old 2169165, size 885 s16_update 1: meter->now 2170936, s16->old 2170050, size 886 s16_update 1: meter->now 2171877, s16->old 2170936, size 941 s16_update 1: meter->now 2172783, s16->old 2171877, size 906 s16_update 1: meter->now 2173690, s16->old 2172783, size 907 s16_update 1: meter->now 2174596, s16->old 2173690, size 906 s16_update 1: meter->now 2175502, s16->old 2174596, size 906 s16_update 1: meter->now 2176407, s16->old 2175502, size 905 s16_update 1: meter->now 2177311, s16->old 2176407, size 904 s16_update 1: meter->now 2178215, s16->old 2177311, size 904 However, sometimes meter->now < s16->old, causing size < 0, for which this line applies https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_meter.c#L1… size += spcm->boundary and size becomes a huge number, because hw:0 pcm -> boundary is 1445068800. The loop at https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_meter.c#L1… takes tens of seconds, untill the huge buffer is converted to s16. Usually the negative difference is quite small: s16_update 1: meter->now 988, s16->old 21139, size -20151 s16_update 1: meter->now 870, s16->old 18865, size -17995 s16_update 1: meter->now 1065, s16->old 20786, size -19721 s16_update 1: meter->now 701, s16->old 24839, size -24138 s16_update 1: meter->now 782, s16->old 869, size -87 s16_update 1: meter->now 763, s16->old 841, size -78 s16_update 1: meter->now 963, s16->old 22525, size -21562 s16_update 1: meter->now 665, s16->old 832, size -167 s16_update 1: meter->now 745, s16->old 865, size -120 s16_update 1: meter->now 859, s16->old 1002, size -143 s16_update 1: meter->now 935, s16->old 117452, size -116517 Yet in all these cases the size variable is incremented by 1445068800, which clearly does not make sense. Now the question is how it happens that meter->now < s16->old, when meter->now is calculated by now = status.appl_ptr - status.delay https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_meter.c#L2… and s16->old is set to meter->now at the end of each s16_update call https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_meter.c#L1… . The debug around the event looks like this, suddenly meter->now drops down: s16_update 1: meter->now 2567498, s16->old 2566593, size 905 s16_update 1: meter->now 2568401, s16->old 2567498, size 903 s16_update 1: meter->now 20786, s16->old 20786, size 0 s16_update 1: meter->now 1065, s16->old 20786, size -19721 s16_update 1: meter->now 24839, s16->old 24838, size 1 s16_update 1: meter->now 701, s16->old 24839, size -24138 s16_update 1: meter->now 1253162, s16->old 701, size 1252461 s16_update 1: meter->now 1255148, s16->old 1253162, size 1986 .......... s16_update 1: meter->now 11136, s16->old 10261, size 875 s16_update 1: meter->now 22525, s16->old 22524, size 1 s16_update 1: meter->now 963, s16->old 22525, size -21562 s16_update 1: meter->now 1270914, s16->old 963, size 1269951 s16_update 1: meter->now 1272917, s16->old 1270914, size 2003 Could it perhaps be an XRUN, causing the difference (status.appl_ptr - status.delay) to actually decrease compared to the previous run? I really do not know how/when such decrease can happen. But if it is a standard situation, than IMO the size of to-be converted buffer by the s16 scope should not be set to the huge pcm->boundary. Thanks a lot for help. Best regards, Pavel.

3 11

[PATCH v4 00/12] ASoC: qcom: Add support for SC7180 lpass variant
by Rohit kumar 07 Aug '20

07 Aug '20

This patch chain add audio support for SC7180 soc by doing the required modification in existing common lpass-cpu/lpass-platform driver. This also fixes some concurrency issue. Changes since v3: - Fixed yaml documentation comments and make dt_binding_check issues. - Moved general fixes out of sc7180 specific patches as suggested by Srinivas. - Update clock-names to make it same as existing platforms. Ajit Pandey (4): ASoC: qcom: Add common array to initialize soc based core clocks ASoC: qcom: lpass-platform: Replace card->dev with component->dev include: dt-bindings: sound: Add sc7180-lpass bindings header ASoC: qcom: lpass-sc7180: Add platform driver for lpass audio Rohit kumar (8): ASoC: qcom: lpass-cpu: Move ahbix clk to platform specific function ASoC: qcom: lpass-platform: fix memory leak ASoC: qcom: lpass: Use regmap_field for i2sctl and dmactl registers ASoC: qcom: lpass-cpu: fix concurrency issue dt-bindings: sound: lpass-cpu: Add sc7180 lpass cpu node ASoC: qcom: lpass-cpu: Use platform_get_resource ASoC: qcom: lpass-platform: Use platform_get_irq dt-bindings: sound: lpass-cpu: Move to yaml format .../devicetree/bindings/sound/qcom,lpass-cpu.txt | 79 -------- .../devicetree/bindings/sound/qcom,lpass-cpu.yaml | 185 ++++++++++++++++++ include/dt-bindings/sound/sc7180-lpass.h | 10 + sound/soc/qcom/Kconfig | 5 + sound/soc/qcom/Makefile | 2 + sound/soc/qcom/lpass-apq8016.c | 86 ++++++-- sound/soc/qcom/lpass-cpu.c | 204 ++++++++++--------- sound/soc/qcom/lpass-ipq806x.c | 67 +++++++ sound/soc/qcom/lpass-lpaif-reg.h | 157 ++++++++------- sound/soc/qcom/lpass-platform.c | 155 +++++++++++---- sound/soc/qcom/lpass-sc7180.c | 216 +++++++++++++++++++++ sound/soc/qcom/lpass.h | 63 +++++- 12 files changed, 930 insertions(+), 299 deletions(-) delete mode 100644 Documentation/devicetree/bindings/sound/qcom,lpass-cpu.txt create mode 100644 Documentation/devicetree/bindings/sound/qcom,lpass-cpu.yaml create mode 100644 include/dt-bindings/sound/sc7180-lpass.h create mode 100644 sound/soc/qcom/lpass-sc7180.c -- Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc., is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.

4 27

[PATCH v5 00/11] Add ASoC AHUB components for Tegra210 and later
by Sameer Pujar 07 Aug '20

07 Aug '20

Overview ======== Audio Processing Engine (APE) comprises of Audio DMA (ADMA) and Audio Hub (AHUB) unit. AHUB is a collection of hardware accelerators for audio pre-processing and post-processing. It also includes a programmable full crossbar for routing audio data across these accelerators. This series exposes some of these below mentioned HW devices as ASoC components for Tegra platforms from Tegra210 onwards. * ADMAIF : The interface between ADMA and AHUB * XBAR : Crossbar for routing audio samples across various modules * I2S : Inter-IC Sound Controller * DMIC : Digital Microphone * DSPK : Digital Speaker Following is the summary of current series. * Add YAML DT binding documentation for above mentioned modules. * Helper function for ACIF programming is exposed for Tegra210 and later. * Add ASoC driver components for each of the above modules. * Build ACONNECT and ADMA drivers which are essential to realize audio use case. * Add DT entries for above components for Tegra210, Tegra186 and Tegra194. As per the suggestion in [0] audio graph based sound card support is pushed in a separate series. [0] https://lkml.org/lkml/2020/6/27/4 Changelog ========= v4 -> v5 -------- * Common changes - simple-card driver changes are dropped. Changes are migrated to audio graph card and are moved to a separate series as suggested. - '#sound-dai-cells' property is not needed for planned audio graph card Hence dropped from documentation and related DT binding of component drivers. - CIF and DAP DAIs are added for I/O drivers (DMIC, DSPK, I2S) to represent DAI links using audio graph card. Similary DAIs are added in AHUB driver to describe endpoints in audio crossbar. Routing is updated to reflect the same in drivers. v3 -> v4 -------- * [1/23] "ASoC: dt-bindings: tegra: Add DT bindings for Tegra210" - Removed multiple examples and retained one example per doc - Fixed as per inputs on the previous series - Tested bindings with 'make dt_binding_check/dtbs_check' * [2/23] "ASoC: tegra: Add support for CIF programming" - No change * Common changes (for patch [3/10] to [7/10]) - Mixer control overrides, for PCM parameters (rate, channel, bits), in each driver are dropped. - Updated routing as per DPCM usage - Minor changes related to formatting * New changes (patch [8/23] to [18/23] and patch [23/23]) - Based on discussions in following threads DPCM is used for Tegra Audio. https://lkml.org/lkml/2020/2/20/91 https://lkml.org/lkml/2020/4/30/519 - The simple-card driver is used for Tegra Audio and accordingly some enhancements are made in simple-card and core drivers. - Patch [8/23] to [18/23] are related to simple-card and core changes. - Patch [23/23] adds sound card support to realize complete audio path. This is based on simple-card driver with proposed enhancements. - Re-ordered patches depending on above v2 -> v3 -------- * [1/10] "dt-bindings: sound: tegra: add DT binding for AHUB - Updated licence - Removed redundancy w.r.t items/const/enum - Added constraints wherever needed with "pattern" property * [2/10] "ASoC: tegra: add support for CIF programming" - Removed tegra_cif.c - Instead added inline helper function in tegra_cif.h * common changes (for patch [3/10] to [7/10]) - Replace LATE system calls with Normal sleep - Remove explicit RPM suspend in driver remove() call - Use devm_kzalloc() instead of devm_kcalloc() for single element - Replace 'ret' with 'err' for better reading - Consistent error printing style across drivers - Minor formating fixes * [8/10] "arm64: tegra: add AHUB components for few Tegra chips" - no change * [9/10] "arm64: tegra: enable AHUB modules for few Tegra chips" - no change * [10/10] "arm64: defconfig: enable AHUB components for Tegra210 and later" (New patch) - Enables ACONNECT and AHUB components. With this AHUB and components are registered with ASoC core. v1 -> v2 -------- * [1/9] "dt-bindings: sound: tegra: add DT binding for AHUB" - no changes * [2/9] "ASoC: tegra: add support for CIF programming" - removed CIF programming changes for legacy chips. - this patch now exposes helper function for CIF programming, which can be used on Tegra210 later. - later tegra_cif.c can be extended for legacy chips as well. - updated commit message accordingly * [3/9] "ASoC: tegra: add Tegra210 based DMIC driver" - removed unnecessary initialization of 'ret' in probe() * [4/9] "ASoC: tegra: add Tegra210 based I2S driver" - removed unnecessary initialization of 'ret' in probe() - fixed indentation - added consistent bracing for if-else clauses - updated 'rx_fifo_th' type to 'unsigned int' - used BIT() macro for defines like '1 << {x}' in tegra210_i2s.h * [5/9] "ASoC: tegra: add Tegra210 based AHUB driver" - used of_device_get_match_data() to get 'soc_data' and removed explicit of_match_device() - used devm_platform_ioremap_resource() and removed explicit platform_get_resource() - fixed indentation for devm_snd_soc_register_component() - updated commit message - updated commit message to reflect compatible binding for Tegra186 and Tegra194. * [6/9] "ASoC: tegra: add Tegra186 based DSPK driver" - removed unnecessary initialization of 'ret' in probe() - updated 'max_th' to 'unsigned int' - shortened lengthy macro names to avoid wrapping in tegra186_dspk_wr_reg() and to be consistent * [7/9] "ASoC: tegra: add Tegra210 based ADMAIF driver" - used of_device_get_match_data() and removed explicit of_match_device() - used BIT() macro for defines like '1 << {x}' in tegra210_admaif.h - updated commit message to reflect compatible binding for Tegra186 and Tegra194. * [8/9] "arm64: tegra: add AHUB components for few Tegra chips" - no change * [9/9] "arm64: tegra: enable AHUB modules for few Tegra chips" - no change * common changes for patch [3/9] to [7/9] - sorted headers in alphabetical order - moved MODULE_DEVICE_TABLE() right below *_of_match table - removed macro DRV_NAME - removed explicit 'owner' field from platform_driver structure - added 'const' to snd_soc_dai_ops structure Sameer Pujar (11): ASoC: dt-bindings: tegra: Add DT bindings for Tegra210 ASoC: tegra: Add support for CIF programming ASoC: tegra: Add Tegra210 based DMIC driver ASoC: tegra: Add Tegra210 based I2S driver ASoC: tegra: Add Tegra210 based AHUB driver ASoC: tegra: Add Tegra186 based DSPK driver ASoC: tegra: Add Tegra210 based ADMAIF driver arm64: defconfig: Build AHUB component drivers arm64: defconfig: Build ADMA and ACONNECT driver arm64: tegra: Enable ACONNECT, ADMA and AGIC on Jetson Nano arm64: tegra: Add DT binding for AHUB components .../bindings/sound/nvidia,tegra186-dspk.yaml | 83 +++ .../bindings/sound/nvidia,tegra210-admaif.yaml | 111 +++ .../bindings/sound/nvidia,tegra210-ahub.yaml | 136 ++++ .../bindings/sound/nvidia,tegra210-dmic.yaml | 83 +++ .../bindings/sound/nvidia,tegra210-i2s.yaml | 101 +++ arch/arm64/boot/dts/nvidia/tegra186.dtsi | 217 +++++- arch/arm64/boot/dts/nvidia/tegra194.dtsi | 225 +++++- arch/arm64/boot/dts/nvidia/tegra210-p3450-0000.dts | 12 + arch/arm64/boot/dts/nvidia/tegra210.dtsi | 140 ++++ arch/arm64/configs/defconfig | 8 + sound/soc/tegra/Kconfig | 56 ++ sound/soc/tegra/Makefile | 10 + sound/soc/tegra/tegra186_dspk.c | 442 +++++++++++ sound/soc/tegra/tegra186_dspk.h | 70 ++ sound/soc/tegra/tegra210_admaif.c | 800 ++++++++++++++++++++ sound/soc/tegra/tegra210_admaif.h | 162 ++++ sound/soc/tegra/tegra210_ahub.c | 676 +++++++++++++++++ sound/soc/tegra/tegra210_ahub.h | 127 ++++ sound/soc/tegra/tegra210_dmic.c | 455 ++++++++++++ sound/soc/tegra/tegra210_dmic.h | 82 +++ sound/soc/tegra/tegra210_i2s.c | 812 +++++++++++++++++++++ sound/soc/tegra/tegra210_i2s.h | 126 ++++ sound/soc/tegra/tegra_cif.h | 65 ++ sound/soc/tegra/tegra_pcm.c | 235 +++++- sound/soc/tegra/tegra_pcm.h | 21 +- 25 files changed, 5251 insertions(+), 4 deletions(-) create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra186-dspk.yaml create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra210-admaif.yaml create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra210-ahub.yaml create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra210-dmic.yaml create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra210-i2s.yaml create mode 100644 sound/soc/tegra/tegra186_dspk.c create mode 100644 sound/soc/tegra/tegra186_dspk.h create mode 100644 sound/soc/tegra/tegra210_admaif.c create mode 100644 sound/soc/tegra/tegra210_admaif.h create mode 100644 sound/soc/tegra/tegra210_ahub.c create mode 100644 sound/soc/tegra/tegra210_ahub.h create mode 100644 sound/soc/tegra/tegra210_dmic.c create mode 100644 sound/soc/tegra/tegra210_dmic.h create mode 100644 sound/soc/tegra/tegra210_i2s.c create mode 100644 sound/soc/tegra/tegra210_i2s.h create mode 100644 sound/soc/tegra/tegra_cif.h -- 2.7.4

4 21