July 2023 - Alsa-devel - mailman.alsa-project.org

[PATCH 0/8] ASoC/SOF/Intel/AMD: cleanups for GCC11 -fanalyzer checks
by Pierre-Louis Bossart 01 Aug '23

01 Aug '23

GCC11 provides an '-fanalyzer' static analysis option which does not provide too many false-positives. This patch cleans-up known problematic code paths to help enable this capability in CI. We've used this for about a month already. Pierre-Louis Bossart (8): ASoC: SOF: sof-client-probes-ipc4: add checks to prevent static analysis warnings ASoC: SOF: ipc3: add checks to prevent static analysis warnings ASoC: SOF: topology: simplify code to prevent static analysis warnings ASoC: SOF: imx: remove error checks on NULL ipc ASoC: SOF: mediatek: remove error checks on NULL ipc ASoC: Intel: bdw_rt286: add checks to avoid static analysis warnings ASoC: Intel: atom: remove static analysis false positive ASoC: amd: acp5x-mach:add checks to avoid static analysis warnings include/linux/firmware/imx/dsp.h | 6 ------ include/linux/firmware/mediatek/mtk-adsp-ipc.h | 6 ------ sound/soc/amd/vangogh/acp5x-mach.c | 3 +++ sound/soc/intel/atom/sst/sst_stream.c | 13 +++++++------ sound/soc/intel/boards/bdw_rt286.c | 6 ++++++ sound/soc/sof/ipc3.c | 2 +- sound/soc/sof/sof-client-probes-ipc4.c | 9 +++++++++ sound/soc/sof/topology.c | 5 +++-- 8 files changed, 29 insertions(+), 21 deletions(-) -- 2.39.2

2 9

Audio output heavily distorted on Xen PV dom0 if not recording (Intel ADL)
by Marek Marczykowski-Górecki 01 Aug '23

01 Aug '23

Hi, I have a rather interesting issue on HP EliteBook 1040, with Intel ADL HDA. Sometimes (often) audio output is heavily distorted, but then opening pavucontrol or just launching `parecord /dev/null` fixes the issue instantly. But, when the mixer or recording is closed, then few seconds later issue comes back. You can hear this at https://cloud.marmarek.net/s/46prqZnwrqDkBdD Initially recording was running, then about 10s stopped and you can hear the issue few seconds later. Then I starter recording again, and the issue is gone. This is how it should sound: https://youtube.com/watch?v=4Tr0otuiQuU I do not see any kernel messages when issue happens. The issue happens on Linux 6.1.26, 6.2.10, and 6.3.2. I haven't tested other versions. It happens only when running under Xen (in PV dom0). The very same Linux started directly works fine. Full dmesg of both native and Xen boots is at https://gist.github.com/marmarek/548ee79282446344a172109bcd943930 When running Linux 6.3.2 there, I noticed that just after boot first playback is okay, but if I start and stop recording it breaks. And then after some time of idle it fixes itself again. I have not observed this behavior on 6.2.10 (it was broken initially too), but I could be also doing some things differently. I have few other ADL-based systems (but not HP), and they do not show this issue, only this HP laptop is affected. Example dmesg from another ADL-based system: https://gist.github.com/marmarek/38c5ba3ec58f9300fbc1842ddc9fa4f5 I'm not sure if that's the relevant difference, but comparing dmesg from working and not working systems, I see those on boot for the problematic case: snd_hda_intel 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040100 snd_hda_intel 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040100 sof-audio-pci-intel-tgl 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver sof-audio-pci-intel-tgl 0000:00:1f.3: enabling device (0000 -> 0002) xen: registering gsi 16 triggering 0 polarity 1 Already setup the GSI :16 sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if 0x040100 sof-audio-pci-intel-tgl 0000:00:1f.3: bound 0000:00:02.0 (ops i915_audio_component_bind_ops [i915]) sof-audio-pci-intel-tgl 0000:00:1f.3: use msi interrupt mode sof-audio-pci-intel-tgl 0000:00:1f.3: hda codecs found, mask 5 sof-audio-pci-intel-tgl 0000:00:1f.3: using HDA machine driver skl_hda_dsp_generic now sof-audio-pci-intel-tgl 0000:00:1f.3: DMICs detected in NHLT tables: 2 sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware info: version 2:2:0-57864 sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware: ABI 3:22:1 Kernel ABI 3:23:0 sof-audio-pci-intel-tgl 0000:00:1f.3: unknown sof_ext_man header type 3 size 0x30 sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware info: version 2:2:0-57864 sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware: ABI 3:22:1 Kernel ABI 3:23:0 sof-audio-pci-intel-tgl 0000:00:1f.3: Topology: ABI 3:22:1 Kernel ABI 3:23:0 skl_hda_dsp_generic skl_hda_dsp_generic: ASoC: Parent card not yet available, widget card binding deferred cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: Firmware version: 3 cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: cirrus/cs35l41-dsp1-spk-prot-103c8974.wmfw: Fri 27 Aug 2021 14:58:19 W. Europe Daylight Time cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: Firmware: 400a4 vendor: 0x2 v0.43.1, 2 algorithms cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: 0: ID cd v29.63.1 XM@94 YM@e cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: 1: ID f20b v0.1.0 XM@176 YM@0 cs35l41-hda spi0-CSC3551:00-cs35l41-hda.0: DSP1: spk-prot: E:\Amp Tuning\HP\840\0527\103C8972_220527.bin snd_hda_codec_realtek ehdaudio0D0: bound spi0-CSC3551:00-cs35l41-hda.0 (ops cs35l41_hda_comp_ops [snd_hda_scodec_cs35l41]) cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: Firmware version: 3 cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: cirrus/cs35l41-dsp1-spk-prot-103c8974.wmfw: Fri 27 Aug 2021 14:58:19 W. Europe Daylight Time cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: Firmware: 400a4 vendor: 0x2 v0.43.1, 2 algorithms cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: 0: ID cd v29.63.1 XM@94 YM@e cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: 1: ID f20b v0.1.0 XM@176 YM@0 cs35l41-hda spi0-CSC3551:00-cs35l41-hda.1: DSP1: spk-prot: E:\Amp Tuning\HP\840\0527\103C8972_220527.bin snd_hda_codec_realtek ehdaudio0D0: bound spi0-CSC3551:00-cs35l41-hda.1 (ops cs35l41_hda_comp_ops [snd_hda_scodec_cs35l41]) Any ideas? -- Best Regards, Marek Marczykowski-Górecki Invisible Things Lab

3 6

[PATCH 0/9] ALSA: hda/cs35l56: Various bugfixes
by Richard Fitzgerald 01 Aug '23

01 Aug '23

A collection of various bugfixes to the cs35l56 driver. Richard Fitzgerald (9): ALSA: hda/cs35l56: Complete firmware reboot before calling cs_dsp_run() ALSA: hda/cs35l56: Do not mark cache dirty after REINIT ALSA: hda/cs35l56: Call cs_dsp_power_down() before reloading firmware ALSA: hda/cs35l56: Always power-up and start cs_dsp ALSA: hda/cs35l56: Call cs_dsp_power_down() before calling cs_dsp_remove() ALSA: hda/cs35l56: cs_dsp_power_down() on cs35l56_hda_fw_load() error path ALSA: hda/cs35l56: Do not download firmware over existing RAM firmware ALSA: hda/cs35l56: Fail if .bin not found and firmware not patched ALSA: hda/cs35l56: Reject I2C alias addresses sound/pci/hda/cs35l56_hda.c | 91 ++++++++++++++++++++++++++----------- 1 file changed, 65 insertions(+), 26 deletions(-) -- 2.30.2

2 10

[PATCH 0/7] usb: Add USB MIDI 2.0 Gadget Function Driver
by Takashi Iwai 01 Aug '23

01 Aug '23

Hi, This is a patch set for adding a USB gadget function driver for USB MIDI 2.0 interface. The MIDI 2.0 support in the host side was already enabled since 6.5 kernel in snd-usb-audio driver. Now, this provides a similar gadget functionality using the existing ALSA UMP helper functions. The gadget driver provides both MIDI 2.0 and MIDI 1.0 fallback interfaces with one or more loopback UMP rawmidi devices. The configfs implementation is flexible enough to allow users to set up multiple UMP Endpoints and Function Blocks. The first patch adds the bare f_midi2 driver without configfs support, followed by the configfs support, MIDI 1.0 interface support and a few other extensions and documentation. Takashi === Takashi Iwai (7): usb: gadget: Add support for USB MIDI 2.0 function driver usb: gadget: midi2: Add configfs support usb: gadget: midi2: Dynamically create MIDI 1.0 altset descriptors usb: gadget: midi2: MIDI 1.0 interface (altset 0) support usb: gadget: midi2: Add testing documentation usb: gadget: midi2: Add "Operation Mode" control usb: gadget: midi2: More flexible MIDI 1.0 configuration .../ABI/testing/configfs-usb-gadget-midi2 | 54 + Documentation/usb/gadget-testing.rst | 154 + drivers/usb/gadget/Kconfig | 18 + drivers/usb/gadget/function/Makefile | 2 + drivers/usb/gadget/function/f_midi2.c | 2883 +++++++++++++++++ drivers/usb/gadget/function/u_midi2.h | 81 + 6 files changed, 3192 insertions(+) create mode 100644 Documentation/ABI/testing/configfs-usb-gadget-midi2 create mode 100644 drivers/usb/gadget/function/f_midi2.c create mode 100644 drivers/usb/gadget/function/u_midi2.h -- 2.35.3

3 10

[PATCH] ASoC: soc-core.c: use devm_snd_soc_register_dai()
by Kuninori Morimoto 01 Aug '23

01 Aug '23

It is still using snd_soc_{un}register_dai() manually. Let's use devm_snd_soc_register_dai(). Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- sound/soc/soc-core.c | 42 ++++++++++-------------------------------- 1 file changed, 10 insertions(+), 32 deletions(-) diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index a5b96c17633a..7284363b9088 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -2552,48 +2552,28 @@ struct snd_soc_dai *snd_soc_register_dai(struct snd_soc_component *component, EXPORT_SYMBOL_GPL(snd_soc_register_dai); /** - * snd_soc_unregister_dais - Unregister DAIs from the ASoC core - * - * @component: The component for which the DAIs should be unregistered - */ -static void snd_soc_unregister_dais(struct snd_soc_component *component) -{ - struct snd_soc_dai *dai, *_dai; - - for_each_component_dais_safe(component, dai, _dai) - snd_soc_unregister_dai(dai); -} - -/** - * snd_soc_register_dais - Register a DAI with the ASoC core + * devm_snd_soc_register_dais - Register a DAI with the ASoC core * * @component: The component the DAIs are registered for * @dai_drv: DAI driver to use for the DAIs * @count: Number of DAIs */ -static int snd_soc_register_dais(struct snd_soc_component *component, - struct snd_soc_dai_driver *dai_drv, - size_t count) +static int devm_snd_soc_register_dais(struct snd_soc_component *component, + struct snd_soc_dai_driver *dai_drv, + size_t count) { struct snd_soc_dai *dai; unsigned int i; - int ret; for (i = 0; i < count; i++) { - dai = snd_soc_register_dai(component, dai_drv + i, count == 1 && - component->driver->legacy_dai_naming); - if (dai == NULL) { - ret = -ENOMEM; - goto err; - } + dai = devm_snd_soc_register_dai(component->dev, + component, dai_drv + i, count == 1 && + component->driver->legacy_dai_naming); + if (!dai) + return -ENOMEM; } return 0; - -err: - snd_soc_unregister_dais(component); - - return ret; } #define ENDIANNESS_MAP(name) \ @@ -2644,8 +2624,6 @@ static void snd_soc_del_component_unlocked(struct snd_soc_component *component) { struct snd_soc_card *card = component->card; - snd_soc_unregister_dais(component); - if (card) snd_soc_unbind_card(card, false); @@ -2696,7 +2674,7 @@ int snd_soc_add_component(struct snd_soc_component *component, } } - ret = snd_soc_register_dais(component, dai_drv, num_dai); + ret = devm_snd_soc_register_dais(component, dai_drv, num_dai); if (ret < 0) { dev_err(component->dev, "ASoC: Failed to register DAIs: %d\n", ret); -- 2.25.1

1 0

[PATCH] ASoC: codecs/hda.c: use devm_snd_soc_register_dai()
by Kuninori Morimoto 01 Aug '23

01 Aug '23

It is still using snd_soc_{un}register_dai() manually. Let's use devm_snd_soc_register_dai(). Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- sound/soc/codecs/hda.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/sound/soc/codecs/hda.c b/sound/soc/codecs/hda.c index d57b043d6bfe..551bd0cbfc44 100644 --- a/sound/soc/codecs/hda.c +++ b/sound/soc/codecs/hda.c @@ -99,7 +99,7 @@ static int hda_codec_register_dais(struct hda_codec *codec, struct snd_soc_compo list_for_each_entry(pcm, &codec->pcm_list_head, list) { struct snd_soc_dai *dai; - dai = snd_soc_register_dai(component, drvs, false); + dai = devm_snd_soc_register_dai(component->dev, component, drvs, false); if (!dai) { dev_err(component->dev, "register dai for %s failed\n", pcm->name); return -EINVAL; @@ -108,7 +108,6 @@ static int hda_codec_register_dais(struct hda_codec *codec, struct snd_soc_compo ret = snd_soc_dapm_new_dai_widgets(dapm, dai); if (ret < 0) { dev_err(component->dev, "create widgets failed: %d\n", ret); - snd_soc_unregister_dai(dai); return ret; } @@ -135,7 +134,6 @@ static void hda_codec_unregister_dais(struct hda_codec *codec, for_each_pcm_streams(stream) snd_soc_dapm_free_widget(snd_soc_dai_get_widget(dai, stream)); - snd_soc_unregister_dai(dai); break; } } -- 2.25.1

1 0

[PATCH 0/2] ASoC: wm8960: Clocking tweaks
by Mark Brown 01 Aug '23

01 Aug '23

A couple of minor clocking improvements that make the wm8960 driver work a bit better with generic cards. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- Mark Brown (2): ASoC: wm8960: Read initial MCLK rate from clock API ASoC: wm8960: Make automatic the default clocking mode sound/soc/codecs/wm8960.c | 8 ++++++++ sound/soc/codecs/wm8960.h | 4 ++-- 2 files changed, 10 insertions(+), 2 deletions(-) --- base-commit: 5d0c230f1de8c7515b6567d9afba1f196fb4e2f4 change-id: 20230730-asoc-wm8960-clk-6c7415ee22b4 Best regards, -- Mark Brown <broonie(a)kernel.org>

2 5

params_free functions don't work
by GitHub issues - opened 31 Jul '23

31 Jul '23

alsa-project/alsa-lib issue #341 was opened from imyxh: The following code segfaults on the call to `snd_pcm_hw_params_free` on my machine: ```c #include <alsa/asoundlib.h> void main() { snd_pcm_hw_params_t *params; snd_pcm_hw_params_alloca(&params); snd_pcm_hw_params_free(params); } ``` The same goes for `s/hw/sw/g`. Am I doing something wrong? In more complex examples it seems to fail with munmap_chunk complaining about an invalid pointer, or free complaining about an invalid size. How are the `*_params_free` functions meant to be used? Issue URL : https://github.com/alsa-project/alsa-lib/issues/341 Repository URL: https://github.com/alsa-project/alsa-lib

1 0

[PATCH] ASoC: codecs: rx-macro: Enable RX path clock to match POST_PMD
by Krzysztof Kozlowski 31 Jul '23

31 Jul '23

The driver disables RX path clock in SND_SOC_DAPM_POST_PMD of rx_macro_enable_interp_clk() and rx_macro_enable_mix_path(). Make the code symmetrical by enabling the clock in SND_SOC_DAPM_PRE_PMU. This also matches downstream Qualcomm driver for Qualcomm SM8550 SoC. Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski(a)linaro.org> --- sound/soc/codecs/lpass-rx-macro.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/sound/soc/codecs/lpass-rx-macro.c b/sound/soc/codecs/lpass-rx-macro.c index 685ca95ef4a9..a758ef4f575a 100644 --- a/sound/soc/codecs/lpass-rx-macro.c +++ b/sound/soc/codecs/lpass-rx-macro.c @@ -2529,6 +2529,9 @@ static int rx_macro_enable_interp_clk(struct snd_soc_component *component, CDC_RX_PATH_PGA_MUTE_MASK, 0x1); snd_soc_component_write_field(component, dsm_reg, CDC_RX_RXn_DSM_CLK_EN_MASK, 0x1); + snd_soc_component_update_bits(component, main_reg, + CDC_RX_PATH_CLK_EN_MASK, + CDC_RX_PATH_CLK_ENABLE); snd_soc_component_update_bits(component, rx_cfg2_reg, CDC_RX_RXn_HPF_CUT_FREQ_MASK, 0x03); rx_macro_load_compander_coeff(component, rx, interp_idx, event); @@ -2595,6 +2598,9 @@ static int rx_macro_enable_mix_path(struct snd_soc_dapm_widget *w, switch (event) { case SND_SOC_DAPM_PRE_PMU: rx_macro_enable_interp_clk(component, event, w->shift); + snd_soc_component_update_bits(component, mix_reg, + CDC_RX_RXn_MIX_CLK_EN_MASK, + CDC_RX_RXn_MIX_CLK_EN_MASK); break; case SND_SOC_DAPM_POST_PMU: snd_soc_component_write(component, gain_reg, -- 2.34.1

2 1

[PATCH 0/5] ASoC: rsnd: add multi Component support
by Kuninori Morimoto 31 Jul '23

31 Jul '23

Hi Mark These are part of "ASoC: add multi Component support" patch-set. The patch-set supports below case as "2 Cards". +-- basic board --------+ |+--------+ | || CPU ch0| <--> CodecA | || ch1| <-+ | |+--------+ | | +-------------|---------+ +-- expansion board ----+ | | | | +-> CodecB| +-----------------------+ Renesas sound driver and its Doc part were held for observation. Rob mentioned that "definitions" vs "$defs". But I got error on "$defs", no error on "definitions". I believe this change is not mandatory. He also mentioned that "reg" is missing, but I also believe that "reg" is automatically handled somehow/somewhere (I'm not sure detail, but other reviewer indicated it before). He also mentioned that "ports" and "port", but added new "ports" needs special handling. Using "ports" vs "port" are different, not compatible on this driver. This means we need both on Doc. Thus, there is no update for these. Kuninori Morimoto (5): ASoC: rsnd: use DAI driver ID instead of DAI ID ASoC: rsnd: cleanup rsnd_dai_of_node() ASoC: rsnd: enable multi Component support for Audio Graph Card/Card2 ASoC: dt-bindings: renesas,rsnd.yaml: add common port-def ASoC: dt-bindings: renesas,rsnd.yaml: enable multi ports for multi Component support .../bindings/sound/renesas,rsnd.yaml | 60 +++---- sound/soc/sh/rcar/core.c | 146 ++++++++++++------ sound/soc/sh/rcar/rsnd.h | 4 + 3 files changed, 134 insertions(+), 76 deletions(-) -- 2.25.1

2 6