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

Ubuntu 22.04 LTS and ALSA (Crackling audio on youtube playback)
by Rohit Bhagat 10 Dec '23

10 Dec '23

Crackling audio youtube on firefox - started about 6 weeks ago (w/ Ubuntu 20.04 LTS) just upgraded to Ubuntu LTS 22.04 but problem persists. Maybe someone on ALSA team can help? Thanks,Rohit Please find attached my machine's ALSA dump at:https://alsa-project.org/db/?f=1e3ebb3358eb9200864560d1978fb4305d2b9dc0

1 0

[PATCH v1 1/3] ASoC: tas2781: Add tas2563 into header file
by Shenghao Ding 09 Dec '23

09 Dec '23

Support tas2563 Signed-off-by: Shenghao Ding <shenghao-ding(a)ti.com> --- include/sound/tas2781.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/include/sound/tas2781.h b/include/sound/tas2781.h index a6c808b22318..45be543110ac 100644 --- a/include/sound/tas2781.h +++ b/include/sound/tas2781.h @@ -1,13 +1,13 @@ /* SPDX-License-Identifier: GPL-2.0 */ // -// ALSA SoC Texas Instruments TAS2781 Audio Smart Amplifier +// ALSA SoC Texas Instruments TAS2781/TAS2563 Audio Smart Amplifier // // Copyright (C) 2022 - 2023 Texas Instruments Incorporated // https://www.ti.com // -// The TAS2781 driver implements a flexible and configurable +// The TAS2781/TAS2563 driver implements a flexible and configurable // algo coefficient setting for one, two, or even multiple -// TAS2781 chips. +// TAS2781/TAS2563 chips. // // Author: Shenghao Ding <shenghao-ding(a)ti.com> // Author: Kevin Lu <kevin-lu(a)ti.com> @@ -60,6 +60,7 @@ enum audio_device { TAS2781 = 0, + TAS2563 }; enum device_catlog_id { -- 2.34.1

4 6

[PATCH v2] ALSA: hda/tas2781: leave hda_component in usable state
by Gergo Koteles 09 Dec '23

09 Dec '23

Unloading then loading the module causes a NULL ponter dereference. The hda_unbind zeroes the hda_component, later the hda_bind tries to dereference the codec field. The hda_component is only initialized once by tas2781_generic_fixup. Set only previously modified fields to NULL. BUG: kernel NULL pointer dereference, address: 0000000000000322 Call Trace: <TASK> ? __die+0x23/0x70 ? page_fault_oops+0x171/0x4e0 ? exc_page_fault+0x7f/0x180 ? asm_exc_page_fault+0x26/0x30 ? tas2781_hda_bind+0x59/0x140 [snd_hda_scodec_tas2781_i2c] component_bind_all+0xf3/0x240 try_to_bring_up_aggregate_device+0x1c3/0x270 __component_add+0xbc/0x1a0 tas2781_hda_i2c_probe+0x289/0x3a0 [snd_hda_scodec_tas2781_i2c] i2c_device_probe+0x136/0x2e0 Fixes: 5be27f1e3ec9 ("ALSA: hda/tas2781: Add tas2781 HDA driver") CC: stable(a)vger.kernel.org Signed-off-by: Gergo Koteles <soyer(a)irl.hu> --- sound/pci/hda/tas2781_hda_i2c.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/tas2781_hda_i2c.c b/sound/pci/hda/tas2781_hda_i2c.c index fb802802939e..7dfa8aeb5fcc 100644 --- a/sound/pci/hda/tas2781_hda_i2c.c +++ b/sound/pci/hda/tas2781_hda_i2c.c @@ -612,9 +612,13 @@ static void tas2781_hda_unbind(struct device *dev, { struct tasdevice_priv *tas_priv = dev_get_drvdata(dev); struct hda_component *comps = master_data; + comps = &comps[tas_priv->index]; - if (comps[tas_priv->index].dev == dev) - memset(&comps[tas_priv->index], 0, sizeof(*comps)); + if (comps->dev == dev) { + comps->dev = NULL; + strscpy(comps->name, "", sizeof(comps->name)); + comps->playback_hook = NULL; + } tasdevice_config_info_remove(tas_priv); tasdevice_dsp_remove(tas_priv); base-commit: ffc253263a1375a65fa6c9f62a893e9767fbebfa -- 2.43.0

2 1

possible timing issue with alsa-restore.service
by GitHub issues - opened 08 Dec '23

08 Dec '23

alsa-project/alsa-utils issue #249 was opened from fajabird: not sure how to title this issue but the following happens in my case: System is a Dell XPS one (all in one PC) and I'm running latest manjaro gnome. I noticed that I had no sound although GNOME control center would show a audio device and it was not muted and volume was all the way up. So I tried alsamixer and found the device "Headphones" muted and volume at 0. Changing this did the trick and I had sound. I issued an "alsactl store" and checked that "alsa-restore.service" was not deactivated to make this persistant. Still after reboot I had no sound. But if I would issue "alsactl restore" manually it would fix the sound issue again. So I suspected that there is an timing issue. Via "systemctl edit --full alsa-restore.service" I added a 30 seconds delay before it would call "alsactl restore" and that fixed the issue permanently. But unfortunately I have no idea how to troubleshoot this properly to find the real root cause of this timing issue. Issue URL : https://github.com/alsa-project/alsa-utils/issues/249 Repository URL: https://github.com/alsa-project/alsa-utils

1 0

[PATCH] ALSA: hda/tas2781: leave hda_component in usable state
by Gergo Koteles 08 Dec '23

08 Dec '23

Unloading then loading the module causes a NULL ponter dereference. The hda_unbind zeroes the hda_component, later the hda_bind tries to dereference the codec field. The hda_component is only initialized once by tas2781_generic_fixup. Set only previously modified fields to NULL. BUG: kernel NULL pointer dereference, address: 0000000000000322 Call Trace: <TASK> ? __die+0x23/0x70 ? page_fault_oops+0x171/0x4e0 ? exc_page_fault+0x7f/0x180 ? asm_exc_page_fault+0x26/0x30 ? tas2781_hda_bind+0x59/0x140 [snd_hda_scodec_tas2781_i2c] component_bind_all+0xf3/0x240 try_to_bring_up_aggregate_device+0x1c3/0x270 __component_add+0xbc/0x1a0 tas2781_hda_i2c_probe+0x289/0x3a0 [snd_hda_scodec_tas2781_i2c] i2c_device_probe+0x136/0x2e0 Fixes: 5be27f1e3ec9 ("ALSA: hda/tas2781: Add tas2781 HDA driver") CC: stable(a)vger.kernel.org Signed-off-by: Gergo Koteles <soyer(a)irl.hu> --- sound/pci/hda/tas2781_hda_i2c.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/tas2781_hda_i2c.c b/sound/pci/hda/tas2781_hda_i2c.c index fb802802939e..ba4fdae8ec9b 100644 --- a/sound/pci/hda/tas2781_hda_i2c.c +++ b/sound/pci/hda/tas2781_hda_i2c.c @@ -612,9 +612,13 @@ static void tas2781_hda_unbind(struct device *dev, { struct tasdevice_priv *tas_priv = dev_get_drvdata(dev); struct hda_component *comps = master_data; + comps = &comps[tas_priv->index]; - if (comps[tas_priv->index].dev == dev) - memset(&comps[tas_priv->index], 0, sizeof(*comps)); + if (comps[tas_priv->index].dev == dev) { + comps->dev = NULL; + strscpy(comps->name, "", sizeof(comps->name)); + comps->playback_hook = NULL; + } tasdevice_config_info_remove(tas_priv); tasdevice_dsp_remove(tas_priv); base-commit: ffc253263a1375a65fa6c9f62a893e9767fbebfa -- 2.43.0

1 0

Bass speaker fixup for Lenovo Yoga Pro 7 14APH8
by Yevhen Hrubiian 07 Dec '23

07 Dec '23

Hello ! I've seen a few patches fixing bass speaker problem for new Lenovo (written by Philipp Jungkamp and later by Pascal Gross), for example this line in linux/sound/pci/hda/patch_realtek.c: SND_PCI_QUIRK(0x17aa, 0x3801, "Lenovo Yoga9 14IAP7", ALC287_FIXUP_YOGA9_14IAP7_BASS_SPK_PIN) Could you please add this line for a similar model with the same Problem? SND_PCI_QUIRK(0x17aa, 0x3882, "Lenovo Yoga Pro 7 14APH8", ALC287_FIXUP_YOGA9_14IAP7_BASS_SPK_PIN), I do not have experience in writing kernel patches so I hope I will not irritate you with this letter. Thanks! Yevhen Hrubiian

2 1

nhlt: Revert SSP_ANALOG device_type field
by GitHub pull_request - opened 07 Dec '23

07 Dec '23

alsa-project/alsa-utils pull request #248 was opened from andyross: This partially reverts commit 3a47ef2487ed ("topology: nhlt: intel: support more device types and directions"), which changed the default device_type in the endpoint descriptor from zero to SSP_ANALOG. This change breaks the Linux kernel NHLT parser (which AFAICT doesn't recognize SSP_ANALOG at all), producing errors like: [56458.583812] sof-audio-pci-intel-mtl 0000:00:1f.3: no matching blob for sample rate: 48000 sample width: 32 channels: 2 [56458.583833] sof-audio-pci-intel-mtl 0000:00:1f.3: failed to prepare widget dai-copier.SSP.SSP0-Codec.playback [56458.583840] sof-audio-pci-intel-mtl 0000:00:1f.3: Failed to prepare connected widgets [56458.583847] sof-audio-pci-intel-mtl 0000:00:1f.3: error: failed widget list set up for pcm 1 dir 0 [56458.583853] sof-audio-pci-intel-mtl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_hw_params on 0000:00:1f.3: -22 Revert for compatibility. Request URL : https://github.com/alsa-project/alsa-utils/pull/248 Patch URL : https://github.com/alsa-project/alsa-utils/pull/248.patch Repository URL: https://github.com/alsa-project/alsa-utils

1 0

multichannel audio -- four out of 16 channels turned off by default
by GitHub issues - opened 07 Dec '23

07 Dec '23

alsa-project/alsa-tools issue #22 was opened from jyrkialakuijala: I'm using UMC1820 and Behringer ultragain digital to have 16 channels of output. I need to use them in 48000/24 bit to use all channels, 96000 can only use 12 channels. It used to work well in early 2023. There was (likely) a software update around mid 2023 that turned off four channels automatically. I was able to fix it by issuing commands: $ amixer --card 2 cset numid=3,iface=MIXER,name='UMC1820 Output Playback Volume' 127,127,127,127,127,127,127,127,127,127,127,127,127,127,127,127 $ amixer --card 2 cset numid=1,iface=MIXER,name='UMC1820 Output Playback Switch' on,on,on,on,on,on,on,on,on,on,on,on,on,on,on,on When I turned the audio device (UMC1820) off and back on, then I need to reissue those commands to have all 16 channels in use. It was not easy to understand what was going on, it could have been a failure in the amplifier, line transformers, the ultragain box -- but turned out to be software config. When I looked at software configs I was first convinced that it was configured to 96000 rate because that also creates the same effect of only four channels being available on the extension box. It would be nice if the devices were detected properly and all channels would be on by default, not just 12 first channels. Issue URL : https://github.com/alsa-project/alsa-tools/issues/22 Repository URL: https://github.com/alsa-project/alsa-tools

1 0

[PATCH] ASoC: amd: acp: modify config flag read logic
by Vijendar Mukunda 07 Dec '23

07 Dec '23

Modify acp config flag read logic from ACP v7.0 onwards. Instead of reading from DMI table match entry, read the config flag value from BIOS ACPI table. This will remove updating DMI table when new platform support is added. Use FLAG_AMD_LEGACY_ONLY_DMIC flag as default one. Signed-off-by: Vijendar Mukunda <Vijendar.Mukunda(a)amd.com> --- sound/soc/amd/acp-config.c | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) diff --git a/sound/soc/amd/acp-config.c b/sound/soc/amd/acp-config.c index 20cee7104c2b..a9aac8f05cb9 100644 --- a/sound/soc/amd/acp-config.c +++ b/sound/soc/amd/acp-config.c @@ -19,6 +19,8 @@ #include "../sof/amd/acp.h" #include "mach-config.h" +#define ACP_7_0_REV 0x70 + static int acp_quirk_data; static const struct config_entry config_table[] = { @@ -133,15 +135,33 @@ static const struct config_entry config_table[] = { }, }; +static int snd_amd_acp_acpi_find_config(struct pci_dev *pci) +{ + const union acpi_object *obj; + int acp_flag = FLAG_AMD_LEGACY_ONLY_DMIC; + + if (!acpi_dev_get_property(ACPI_COMPANION(&pci->dev), "acp-audio-config-flag", + ACPI_TYPE_INTEGER, &obj)) + acp_flag = obj->integer.value; + + return acp_flag; +} + int snd_amd_acp_find_config(struct pci_dev *pci) { const struct config_entry *table = config_table; u16 device = pci->device; int i; - /* Do not enable FLAGS on older platforms with Rev id zero */ + /* Do not enable FLAGS on older platforms with Rev Id zero + * For platforms which has ACP 7.0 or higher, read the acp + * config flag from BIOS ACPI table and for older platforms + * read it from DMI tables. + */ if (!pci->revision) return 0; + else if (pci->revision >= ACP_7_0_REV) + return snd_amd_acp_acpi_find_config(pci); for (i = 0; i < ARRAY_SIZE(config_table); i++, table++) { if (table->device != device) -- 2.34.1

2 1

[PATCH 1/3] ASoC: cs35l45: Use modern pm_ops
by Ricardo Rivera-Matos 07 Dec '23

07 Dec '23

Make use of the recently introduced EXPORT_GPL_DEV_PM_OPS() macro, to conditionally export the runtime/system PM functions. Replace the old SET_{RUNTIME,SYSTEM_SLEEP,NOIRQ_SYSTEM_SLEEP}_PM_OPS() helpers with their modern alternatives and get rid of the now unnecessary '__maybe_unused' annotations on all PM functions. Additionally, use the pm_ptr() macro to fix the following errors when building with CONFIG_PM disabled: Signed-off-by: Ricardo Rivera-Matos <rriveram(a)opensource.cirrus.com> Acked-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> --- sound/soc/codecs/cs35l45-i2c.c | 2 +- sound/soc/codecs/cs35l45-spi.c | 2 +- sound/soc/codecs/cs35l45.c | 9 ++++----- 3 files changed, 6 insertions(+), 7 deletions(-) diff --git a/sound/soc/codecs/cs35l45-i2c.c b/sound/soc/codecs/cs35l45-i2c.c index 77e0f8750f37..bc2af1ed0fe9 100644 --- a/sound/soc/codecs/cs35l45-i2c.c +++ b/sound/soc/codecs/cs35l45-i2c.c @@ -62,7 +62,7 @@ static struct i2c_driver cs35l45_i2c_driver = { .driver = { .name = "cs35l45", .of_match_table = cs35l45_of_match, - .pm = &cs35l45_pm_ops, + .pm = pm_ptr(&cs35l45_pm_ops), }, .id_table = cs35l45_id_i2c, .probe = cs35l45_i2c_probe, diff --git a/sound/soc/codecs/cs35l45-spi.c b/sound/soc/codecs/cs35l45-spi.c index 5efb77530cc3..39e203a5f060 100644 --- a/sound/soc/codecs/cs35l45-spi.c +++ b/sound/soc/codecs/cs35l45-spi.c @@ -64,7 +64,7 @@ static struct spi_driver cs35l45_spi_driver = { .driver = { .name = "cs35l45", .of_match_table = cs35l45_of_match, - .pm = &cs35l45_pm_ops, + .pm = pm_ptr(&cs35l45_pm_ops), }, .id_table = cs35l45_id_spi, .probe = cs35l45_spi_probe, diff --git a/sound/soc/codecs/cs35l45.c b/sound/soc/codecs/cs35l45.c index b68853e42fd1..4f4df166f5f0 100644 --- a/sound/soc/codecs/cs35l45.c +++ b/sound/soc/codecs/cs35l45.c @@ -982,7 +982,7 @@ static int cs35l45_exit_hibernate(struct cs35l45_private *cs35l45) return -ETIMEDOUT; } -static int __maybe_unused cs35l45_runtime_suspend(struct device *dev) +static int cs35l45_runtime_suspend(struct device *dev) { struct cs35l45_private *cs35l45 = dev_get_drvdata(dev); @@ -999,7 +999,7 @@ static int __maybe_unused cs35l45_runtime_suspend(struct device *dev) return 0; } -static int __maybe_unused cs35l45_runtime_resume(struct device *dev) +static int cs35l45_runtime_resume(struct device *dev) { struct cs35l45_private *cs35l45 = dev_get_drvdata(dev); int ret; @@ -1466,10 +1466,9 @@ void cs35l45_remove(struct cs35l45_private *cs35l45) } EXPORT_SYMBOL_NS_GPL(cs35l45_remove, SND_SOC_CS35L45); -const struct dev_pm_ops cs35l45_pm_ops = { - SET_RUNTIME_PM_OPS(cs35l45_runtime_suspend, cs35l45_runtime_resume, NULL) +EXPORT_GPL_DEV_PM_OPS(cs35l45_pm_ops) = { + RUNTIME_PM_OPS(cs35l45_runtime_suspend, cs35l45_runtime_resume, NULL) }; -EXPORT_SYMBOL_NS_GPL(cs35l45_pm_ops, SND_SOC_CS35L45); MODULE_DESCRIPTION("ASoC CS35L45 driver"); MODULE_AUTHOR("James Schulman, Cirrus Logic Inc, <james.schulman(a)cirrus.com>"); -- 2.34.1

2 1