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

[alsa-devel] Bug report for changes in ucm2 with chtrt5645 card on Lenovo Miix 320
by Tim Schumacher 15 Jan '20

15 Jan '20

Hi alsa devs, I'm on Arch Linux on a rather uncommon laptop (Lenovo Miix 320) and my sound stopped working on the upgrade from alsa-lib 1.1.9 to 1.2.1. wabbits on IRC helped me track down the problem to a change in the /usr/share/alsa/ucm (now /usr/share/alsa/ucm2) files. It seems like /usr/share/alsa/ucm2/chtrt5645/HiFi-dmic2.conf and a symlink at /usr/share/alsa/ucm2/chtrt5645/LENOVO-80XF-LenovoMIIX320_10ICR-LNVNB161216.conf were specifically added for this laptop but in the contrary they break the sound for me. After the upgrade to alsa-lib 1.2.1 the sound is not working. The commands play no sound: pasuspender -- speaker-test --nloops=1 --channels=2 --test=wav --device=hw:0,0 pasuspender -- speaker-test --nloops=1 --channels=2 --test=wav --device=hw:0,1 And this is the alsa-info.sh output http://alsa-project.org/db/?f=f883910a5c5101b4b1ea4202d1fe84ccd139f796 After deleting the /usr/share/alsa/ucm2/chtrt5645/LENOVO-80XF-LenovoMIIX320_10ICR-LNVNB161216.conf symlink both commands from above play sound and this is the alsa-info.sh output http://alsa-project.org/db/?f=e759eb9118a191b6c3b8c021fed58abc9cf95076 Thanks. - Tim

2 6

[alsa-devel] [PATCH] ALSA: hda/analog - Minor optimization for SPDIF mux connections
by Takashi Iwai 15 Jan '20

15 Jan '20

AD HD-audio codec driver has a few code lines invoking snd_get_num_conns() and using its return value as the array index without checking. This is basically safe in all those places; at the second and later calls snd_get_num_conns() returns the value cached from the first invocation, hence the value is always consistent. However, it looks a bit confusing as if a lack of the proper check. This patch introduces a new field num_smux_conns in ad198x_spec for simplifying the code. Now we store and refer to the value more locally without invoking the extra function at each time. Reported-by: Colin King <colin.king(a)canonical.com> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/pci/hda/patch_analog.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/sound/pci/hda/patch_analog.c b/sound/pci/hda/patch_analog.c index 88c46b051d14..2132b2acec4d 100644 --- a/sound/pci/hda/patch_analog.c +++ b/sound/pci/hda/patch_analog.c @@ -28,6 +28,7 @@ struct ad198x_spec { hda_nid_t eapd_nid; unsigned int beep_amp; /* beep amp value, set via set_beep_amp() */ + int num_smux_conns; }; @@ -453,8 +454,7 @@ static int ad1983_auto_smux_enum_info(struct snd_kcontrol *kcontrol, struct ad198x_spec *spec = codec->spec; static const char * const texts2[] = { "PCM", "ADC" }; static const char * const texts3[] = { "PCM", "ADC1", "ADC2" }; - hda_nid_t dig_out = spec->gen.multiout.dig_out_nid; - int num_conns = snd_hda_get_num_conns(codec, dig_out); + int num_conns = spec->num_smux_conns; if (num_conns == 2) return snd_hda_enum_helper_info(kcontrol, uinfo, 2, texts2); @@ -481,7 +481,7 @@ static int ad1983_auto_smux_enum_put(struct snd_kcontrol *kcontrol, struct ad198x_spec *spec = codec->spec; unsigned int val = ucontrol->value.enumerated.item[0]; hda_nid_t dig_out = spec->gen.multiout.dig_out_nid; - int num_conns = snd_hda_get_num_conns(codec, dig_out); + int num_conns = spec->num_smux_conns; if (val >= num_conns) return -EINVAL; @@ -512,6 +512,7 @@ static int ad1983_add_spdif_mux_ctl(struct hda_codec *codec) num_conns = snd_hda_get_num_conns(codec, dig_out); if (num_conns != 2 && num_conns != 3) return 0; + spec->num_smux_conns = num_conns; if (!snd_hda_gen_add_kctl(&spec->gen, NULL, &ad1983_auto_smux_mixer)) return -ENOMEM; return 0; @@ -730,10 +731,12 @@ static int ad1988_auto_smux_enum_info(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo) { struct hda_codec *codec = snd_kcontrol_chip(kcontrol); + struct ad198x_spec *spec = codec->spec; static const char * const texts[] = { "PCM", "ADC1", "ADC2", "ADC3", }; - int num_conns = snd_hda_get_num_conns(codec, 0x0b) + 1; + int num_conns = spec->num_smux_conns; + if (num_conns > 4) num_conns = 4; return snd_hda_enum_helper_info(kcontrol, uinfo, num_conns, texts); @@ -756,7 +759,7 @@ static int ad1988_auto_smux_enum_put(struct snd_kcontrol *kcontrol, struct ad198x_spec *spec = codec->spec; unsigned int val = ucontrol->value.enumerated.item[0]; struct nid_path *path; - int num_conns = snd_hda_get_num_conns(codec, 0x0b) + 1; + int num_conns = spec->num_smux_conns; if (val >= num_conns) return -EINVAL; @@ -847,6 +850,7 @@ static int ad1988_add_spdif_mux_ctl(struct hda_codec *codec) num_conns = snd_hda_get_num_conns(codec, 0x0b) + 1; if (num_conns != 3 && num_conns != 4) return 0; + spec->num_smux_conns = num_conns; for (i = 0; i < num_conns; i++) { struct nid_path *path = snd_array_new(&spec->gen.paths); -- 2.16.4

1 0

[alsa-devel] ABI breakage in alsa-lib 1.2
by Jordi Mallach 15 Jan '20

15 Jan '20

[Please Cc me on replies, I'm not currently subscribed to alsa-devel] Hello, I'm one of the ALSA Debian maintainers, and some weeks ago I looked into updating our ALSA packages to 1.2.1. I quickly noticed there's been some fundamental change to libasound2, one that appears to break ABI and should have warranted an soname bump at least. With the split of the snd_tplg API to the new libatopology2, libasound2 has been left missing that part of the API that was part of the asound.h header, etc. This could have been a somewhat lesser issue if libasound2 linked to libatopology, but unfortunately this goes in the other direction, so the ABI and API breakage is not avoided. I cannot update alsa-lib in this state. I could hack around and patch it to reinstate the missing api in libasound, but this sounds like pain and suffering for everyone involved. How should we move forward with this? I am sorry I only come up with this 2 months after the release when I had identified the problem many weeks ago, I've been extremely busy. Thanks, Jordi -- Jordi Mallach <jordi(a)debian.org> Debian Project

2 1

[alsa-devel] [PATCH RFC 0/3] ASoC: Add Multi CPU DAI support
by Bard liao 15 Jan '20

15 Jan '20

From: Bard Liao <yung-chuan.liao(a)linux.intel.com> As discussed in [1], ASoC core supports multi codec DAIs on a DAI link. However it does not do so for CPU DAIs. So, add support for multi CPU DAIs on a DAI Link by adding multi CPU DAI in Card instantiation, suspend and resume functions, PCM ops, stream handling functions and DAPM. [1]: https://www.spinics.net/lists/alsa-devel/msg71369.html Shreyas NC (3): ASoC: Add initial support for multiple CPU DAIs ASoC: Add multiple CPU DAI support for PCM ops ASoC: Add multiple CPU DAI support in DAPM include/sound/soc.h | 15 ++ sound/soc/soc-core.c | 183 ++++++++++-------- sound/soc/soc-dapm.c | 131 +++++++------ sound/soc/soc-pcm.c | 432 ++++++++++++++++++++++++++++--------------- 4 files changed, 483 insertions(+), 278 deletions(-) -- 2.17.1

4 7

[alsa-devel] [PATCH] ALSA: hda - fix out of bounds read on spec->smux_paths
by Colin King 14 Jan '20

14 Jan '20

From: Colin Ian King <colin.king(a)canonical.com> It is possible for the call to snd_hda_get_num_conns to fail and return a negative error code that gets assigned to num_conns. In that specific case, the check of very large values of val against num_conns will not fail the -EINVAL check and later on an out of bounds array read on spec->smux_paths will occur. Fix this by sanity checking for an error return from the call to snd_hda_get_num_conns. Addresses-Coverity: ("Out-of-bounds read") Fixes: 272f3ea31776 ("ALSA: hda - Add SPDIF mux control to AD codec auto-parser") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> --- sound/pci/hda/patch_analog.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/patch_analog.c b/sound/pci/hda/patch_analog.c index 88c46b051d14..399561369495 100644 --- a/sound/pci/hda/patch_analog.c +++ b/sound/pci/hda/patch_analog.c @@ -756,9 +756,11 @@ static int ad1988_auto_smux_enum_put(struct snd_kcontrol *kcontrol, struct ad198x_spec *spec = codec->spec; unsigned int val = ucontrol->value.enumerated.item[0]; struct nid_path *path; - int num_conns = snd_hda_get_num_conns(codec, 0x0b) + 1; + int num_conns = snd_hda_get_num_conns(codec, 0x0b); - if (val >= num_conns) + if (num_conns < 0) + return num_conns; + if (val >= num_conns + 1) return -EINVAL; if (spec->cur_smux == val) return 0; -- 2.24.0

3 2

[alsa-devel] [PATCH 0/5] soundwire: intel: add DAI callbacks
by Pierre-Louis Bossart 14 Jan '20

14 Jan '20

The existing mainline code is missing most of the DAI callbacks needed for a functional implementation, and the existing ones need to be modified to provide the relevant information to SOF drivers. As suggested by Vinod, these patches are shared first - with the risk that they are separated from the actual DAI enablement, so reviewers might wonder why they are needed in the first place. For reference, the complete set of 100+ patches required for SoundWire on Intel platforms is available here: https://github.com/thesofproject/linux/pull/1692 Pierre-Louis Bossart (2): soundwire: intel: rename res field as link_res soundwire: intel: free all resources on hw_free() Rander Wang (3): soundwire: intel: add prepare support in sdw dai driver soundwire: intel: add trigger support in sdw dai driver soundwire: intel: add sdw_stream_setup helper for .startup callback drivers/soundwire/intel.c | 198 ++++++++++++++++++++++++++++++++++---- 1 file changed, 178 insertions(+), 20 deletions(-) base-commit: b637124800a157c4df3699d1137d8533394f7678 -- 2.20.1

2 7

[alsa-devel] how to use DPCM to alter audio path between headphones and speaker
by Tim Harvey 14 Jan '20

14 Jan '20

Greetings, I'm working with a board that has a TI TLV320AIC3105 codec connected to an IMX6 SSI with a headphone jack and a TS3A227E headphone detect chip (HPROUT/HPLOUT), mono speaker amp output (RIGHT_LOP/RIGHT_LOM), and mic in (MIC1L). It seems that Dynamic PCM allows for the automatic routing of audio to change between headphone out and speaker out depending on jack insertion/removal event but I'm not clear how to configure that. Can this be done via simple-audio-card bindings or does it require a modified machine driver? Any pointers or references are greatly appreciated! Best Regards, Tim

1 0

[alsa-devel] [PATCH 0/4] ASoC: msm8916-wcd-analog: MIC BIAS fixes/additions
by Stephan Gerhold 14 Jan '20

14 Jan '20

This patch series fixes some bugs in the MIC BIAS implementation in msm8916-wcd-analog. Finally, it adds support for MIC BIAS Internal3, which is needed for msm8916-longcheer-l8150. Stephan Gerhold (4): ASoC: msm8916-wcd-analog: Fix selected events for MIC BIAS External1 ASoC: msm8916-wcd-analog: Fix MIC BIAS Internal1 ASoC: msm8916-wcd-analog: Simplify MIC BIAS Internal ASoC: msm8916-wcd-analog: Add MIC BIAS Internal3 sound/soc/codecs/msm8916-wcd-analog.c | 59 ++++++++++----------------- 1 file changed, 21 insertions(+), 38 deletions(-) -- 2.24.1

3 12

[alsa-devel] [PATCH] ASoC: sgtl5000: add multi endpoint support
by Marek Vasut 14 Jan '20

14 Jan '20

Support multiple endpoints on SGTL5000 codec port when used in of_graph context. This patch allows to share the codec port between two CPU DAIs. Example: Custom STM32MP157C board uses SGTL5000 audio codec. This codec is connected to two serial audio interfaces, which are configured either as rx or tx. >From AsoC point of view the topolgy is the following: // 2 CPU DAIs (SAI2A/B), 1 Codec (SGTL5000) Playback: CPU-A-DAI(slave) -> (master)CODEC-DAI/port0 Record: CPU-B-DAI(slave) <- (master)CODEC-DAI/port0 In the DT two endpoints have to be associated to the codec port: sgtl5000_port: port { sgtl5000_tx_endpoint: endpoint@0 { remote-endpoint = <&sai2a_endpoint>; }; sgtl5000_rx_endpoint: endpoint@1 { remote-endpoint = <&sai2b_endpoint>; }; }; However, when the audio graph card parses the codec nodes, it expects to find DAI interface indexes matching the endpoints indexes. The current patch forces the use of DAI id 0 for both endpoints, which allows to share the codec DAI between the two CPU DAIs for playback and capture streams respectively. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Mark Brown <broonie(a)kernel.org> --- NOTE: This is the same as commit ad6bb3067c4d ("ASoC: cs42l51: add multi endpoint support") --- sound/soc/codecs/sgtl5000.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/sound/soc/codecs/sgtl5000.c b/sound/soc/codecs/sgtl5000.c index aa1f9637d8950..93da6b58c97df 100644 --- a/sound/soc/codecs/sgtl5000.c +++ b/sound/soc/codecs/sgtl5000.c @@ -1513,6 +1513,13 @@ static int sgtl5000_probe(struct snd_soc_component *component) return ret; } +static int sgtl5000_of_xlate_dai_id(struct snd_soc_component *component, + struct device_node *endpoint) +{ + /* return dai id 0, whatever the endpoint index */ + return 0; +} + static const struct snd_soc_component_driver sgtl5000_driver = { .probe = sgtl5000_probe, .set_bias_level = sgtl5000_set_bias_level, @@ -1522,6 +1529,7 @@ static const struct snd_soc_component_driver sgtl5000_driver = { .num_dapm_widgets = ARRAY_SIZE(sgtl5000_dapm_widgets), .dapm_routes = sgtl5000_dapm_routes, .num_dapm_routes = ARRAY_SIZE(sgtl5000_dapm_routes), + .of_xlate_dai_id = sgtl5000_of_xlate_dai_id, .suspend_bias_off = 1, .idle_bias_on = 1, .use_pmdown_time = 1, -- 2.24.1

2 1

[alsa-devel] [PATCH] soundwire: bus: fix device number leak on errors
by Pierre-Louis Bossart 14 Jan '20

14 Jan '20

If the programming of the dev_number fails due to an IO error, a new device_number will be assigned, resulting in a leak. Make sure we only assign a device_number once per Slave device. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> --- drivers/soundwire/bus.c | 37 ++++++++++++++++++++++------------- include/linux/soundwire/sdw.h | 4 +++- 2 files changed, 26 insertions(+), 15 deletions(-) diff --git a/drivers/soundwire/bus.c b/drivers/soundwire/bus.c index be5d437058ed..8f973e70e6f7 100644 --- a/drivers/soundwire/bus.c +++ b/drivers/soundwire/bus.c @@ -456,26 +456,35 @@ static int sdw_get_device_num(struct sdw_slave *slave) static int sdw_assign_device_num(struct sdw_slave *slave) { int ret, dev_num; + bool new_device = false; /* check first if device number is assigned, if so reuse that */ if (!slave->dev_num) { - mutex_lock(&slave->bus->bus_lock); - dev_num = sdw_get_device_num(slave); - mutex_unlock(&slave->bus->bus_lock); - if (dev_num < 0) { - dev_err(slave->bus->dev, "Get dev_num failed: %d\n", - dev_num); - return dev_num; + if (!slave->dev_num_sticky) { + mutex_lock(&slave->bus->bus_lock); + dev_num = sdw_get_device_num(slave); + mutex_unlock(&slave->bus->bus_lock); + if (dev_num < 0) { + dev_err(slave->bus->dev, "Get dev_num failed: %d\n", + dev_num); + return dev_num; + } + slave->dev_num = dev_num; + slave->dev_num_sticky = dev_num; + new_device = true; + } else { + slave->dev_num = slave->dev_num_sticky; } - } else { + } + + if (!new_device) dev_info(slave->bus->dev, - "Slave already registered dev_num:%d\n", + "Slave already registered, reusing dev_num:%d\n", slave->dev_num); - /* Clear the slave->dev_num to transfer message on device 0 */ - dev_num = slave->dev_num; - slave->dev_num = 0; - } + /* Clear the slave->dev_num to transfer message on device 0 */ + dev_num = slave->dev_num; + slave->dev_num = 0; ret = sdw_write(slave, SDW_SCP_DEVNUMBER, dev_num); if (ret < 0) { @@ -485,7 +494,7 @@ static int sdw_assign_device_num(struct sdw_slave *slave) } /* After xfer of msg, restore dev_num */ - slave->dev_num = dev_num; + slave->dev_num = slave->dev_num_sticky; return 0; } diff --git a/include/linux/soundwire/sdw.h b/include/linux/soundwire/sdw.h index b7c9eca4332a..b451bb622335 100644 --- a/include/linux/soundwire/sdw.h +++ b/include/linux/soundwire/sdw.h @@ -546,7 +546,8 @@ struct sdw_slave_ops { * @debugfs: Slave debugfs * @node: node for bus list * @port_ready: Port ready completion flag for each Slave port - * @dev_num: Device Number assigned by Bus + * @dev_num: Current Device Number, values can be 0 or dev_num_sticky + * @dev_num_sticky: one-time static Device Number assigned by Bus * @probed: boolean tracking driver state * @probe_complete: completion utility to control potential races * on startup between driver probe/initialization and SoundWire @@ -575,6 +576,7 @@ struct sdw_slave { struct list_head node; struct completion *port_ready; u16 dev_num; + u16 dev_num_sticky; bool probed; struct completion probe_complete; struct completion enumeration_complete; -- 2.20.1

2 2