August 2019 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] Possible bug report
by Harry Bloomberg 06 Aug '19

06 Aug '19

While running an amateur radio application on a Raspberry Pi 3B+ with the stretch version of Raspberrian, I see the following in my syslog: Aug 1 17:31:06 raspberrypi pulseaudio[853]: [alsa-sink-USB Audio] alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write. Aug 1 17:31:06 raspberrypi pulseaudio[853]: [alsa-sink-USB Audio] alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_usb_audio'. Please report this issue to the ALSA developers. Aug 1 17:31:06 raspberrypi pulseaudio[853]: [alsa-sink-USB Audio] alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail. So...I am reporting this to the developers as instructed by the error message. What additional information do you need? I'm not exactly sure what I was doing to trigger this error. I may have been exiting the application. Thanks, Harry Bloomberg

2 1

[alsa-devel] [PATCH 0/2] ALSA snd-usb-audio: Add Pioneer DDJ-SX3 PCM support
by Ard van Breemen 06 Aug '19

06 Aug '19

The Pioneer lacks the UAC descriptors (added to quircks). It also needs a fix in synchronisation handling for those cases that bSynchAddress == 0 Ard van Breemen (2): Skip bSynchAddress endpoint check if it is invalid. Add Pioneer DDJ-SX3 PCM quirck. sound/usb/pcm.c | 1 + sound/usb/quirks-table.h | 57 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 58 insertions(+) -- 2.17.1

2 3

[alsa-devel] [PATCH] topology: add support to parse private data for pcm
by Keyon Jie 06 Aug '19

06 Aug '19

We have private data section in struct snd_soc_tplg_pcm, but alsatplg doesn't support handling it yet, here add handling in tplg_parse_pcm() to enable it. Signed-off-by: Keyon Jie <yang.jie(a)linux.intel.com> --- src/topology/data.c | 3 +++ src/topology/pcm.c | 8 ++++++++ 2 files changed, 11 insertions(+) diff --git a/src/topology/data.c b/src/topology/data.c index b3f4421f..729ce1f4 100644 --- a/src/topology/data.c +++ b/src/topology/data.c @@ -53,6 +53,9 @@ struct snd_soc_tplg_private *get_priv_data(struct tplg_elem *elem) case SND_TPLG_TYPE_BE: priv = &elem->link->priv; break; + case SND_TPLG_TYPE_PCM: + priv = &elem->pcm->priv; + break; default: SNDERR("error: '%s': no support for private data for type %d\n", elem->id, elem->type); diff --git a/src/topology/pcm.c b/src/topology/pcm.c index 5f586dc1..c533ee68 100644 --- a/src/topology/pcm.c +++ b/src/topology/pcm.c @@ -732,6 +732,14 @@ int tplg_parse_pcm(snd_tplg_t *tplg, return err; continue; } + + /* private data */ + if (strcmp(id, "data") == 0) { + err = tplg_parse_data_refs(n, elem); + if (err < 0) + return err; + continue; + } } return 0; -- 2.20.1

2 1

[alsa-devel] [PATCH] ALSA: usb-audio: fix a memory leak bug
by Wenwen Wang 06 Aug '19

06 Aug '19

In snd_usb_get_audioformat_uac3(), a structure for channel maps 'chmap' is allocated through kzalloc() before the execution goto 'found_clock'. However, this structure is not deallocated if the memory allocation for 'pd' fails, leading to a memory leak bug. To fix the above issue, free 'fp->chmap' before returning NULL. Signed-off-by: Wenwen Wang <wenwen(a)cs.uga.edu> --- sound/usb/stream.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/usb/stream.c b/sound/usb/stream.c index 7ee9d17..e852c7f 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -1043,6 +1043,7 @@ snd_usb_get_audioformat_uac3(struct snd_usb_audio *chip, pd = kzalloc(sizeof(*pd), GFP_KERNEL); if (!pd) { + kfree(fp->chmap); kfree(fp->rate_table); kfree(fp); return NULL; -- 2.7.4

2 1

[alsa-devel] [PATCH] slimbus: qcom-ngd-ctrl: Add of_node_put() before return
by Nishka Dasgupta 06 Aug '19

06 Aug '19

Each iteration of for_each_available_child_of_node puts the previous node, but in the case of a return from the middle of the loop, there is no put, thus causing a memory leak. Hence add an of_node_put before the return in two places. Issue found with Coccinelle. Signed-off-by: Nishka Dasgupta <nishkadg.linux(a)gmail.com> --- drivers/slimbus/qcom-ngd-ctrl.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c index f3585777324c..29fbab55c3b3 100644 --- a/drivers/slimbus/qcom-ngd-ctrl.c +++ b/drivers/slimbus/qcom-ngd-ctrl.c @@ -1338,12 +1338,15 @@ static int of_qcom_slim_ngd_register(struct device *parent, continue; ngd = kzalloc(sizeof(*ngd), GFP_KERNEL); - if (!ngd) + if (!ngd) { + of_node_put(node); return -ENOMEM; + } ngd->pdev = platform_device_alloc(QCOM_SLIM_NGD_DRV_NAME, id); if (!ngd->pdev) { kfree(ngd); + of_node_put(node); return -ENOMEM; } ngd->id = id; -- 2.19.1

3 2

[alsa-devel] [PATCH v1 0/3] Allow reconfiguration of clock rate
by Jiada Wang 06 Aug '19

06 Aug '19

There is use case in Gstreamer ALSA sink, in case of changed caps Gsreatmer reconfigs and it calls snd_pcm_hw_free() before snd_pcm_prepre(). See gstreamer1.0-plugins-base/ gst-libs/gst/audio/gstaudiobasesink.c: gst_audio_base_sink_setcaps(): - gst_audio_ring_buffer_release() - gst_audio_sink_ring_buffer_release() - gst_alsasink_unprepare() - snd_pcm_hw_free() is called before - gst_audio_ring_buffer_acquire() - gst_audio_sink_ring_buffer_acquire() - gst_alsasink_prepare() - set_hwparams() - snd_pcm_hw_params() - snd_pcm_prepare() But with current implementation after clock rate is started in .prepare reconfiguration of clock rate is not allowed, unless the stream is stopped. This patch set by move stop of clock to .hw_free callback, to allow reconfiguration of clock rate. Jiada Wang (1): ASoC: rsnd: call .hw_{params,free} in pair for same stream Timo Wischer (2): ASoC: rsnd: Support hw_free() callback at DAI level ASoC: rsnd: Allow reconfiguration of clock rate sound/soc/sh/rcar/core.c | 22 +++++++++++++-- sound/soc/sh/rcar/rsnd.h | 36 ++++++++++++++++++++---- sound/soc/sh/rcar/ssi.c | 61 +++++++++++++++++++++++++++++----------- sound/soc/sh/rcar/ssiu.c | 3 +- 4 files changed, 96 insertions(+), 26 deletions(-) -- 2.19.2

3 8

[alsa-devel] [PATCH] ALSA: usb-audio: fix a memory leak bug
by Wenwen Wang 06 Aug '19

06 Aug '19

In add_new_ctl(), a mixer element structure is allocated through kzalloc() and the pointer is saved to 'elem'. Later on, a new alsa control element is created and added to this structure. In case the add process fails, i.e., the return value of snd_usb_mixer_add_control() is less than 0, the allocated structure is not freed, leading to a memory leak. To fix the above issue, free 'elem' before returning the error. Signed-off-by: Wenwen Wang <wenwen(a)cs.uga.edu> --- sound/usb/mixer_scarlett.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/usb/mixer_scarlett.c b/sound/usb/mixer_scarlett.c index 83715fd..a6c028a 100644 --- a/sound/usb/mixer_scarlett.c +++ b/sound/usb/mixer_scarlett.c @@ -562,8 +562,10 @@ static int add_new_ctl(struct usb_mixer_interface *mixer, strlcpy(kctl->id.name, name, sizeof(kctl->id.name)); err = snd_usb_mixer_add_control(&elem->head, kctl); - if (err < 0) + if (err < 0) { + kfree(elem); return err; + } if (elem_ret) *elem_ret = elem; -- 2.7.4

2 1

[alsa-devel] [PATCH v2 00/29] ASoC: add soc-component.c
by Kuninori Morimoto 06 Aug '19

06 Aug '19

Hi Mark These are v2 of soc-component.c patch-set. I want to add multi CPU DAI support to ALSA SoC. But I noticed that we want to cleanup ALSA SoC before that. These patches will do nothing from "technical" point of view, but, will makes code readable. These are focusing to "component". Actually, soc-io.c is also for component. I'm thinking that we can merge soc-io.c into soc-component.c, but do nothing by these patch-set. 1) - 4) : new patch. it removes un-used component functions 5) - 19) : small bug fixed from v1 20) - 22) : more cleanuped 23) - 29) : cares for_each_rtdcom() loop Kuninori Morimoto (29): 1) ASoC: Intel: skl-pcm: disable skl_get_time_info 2) ASoC: soc-pcm: remove soc_rtdcom_ack() 3) ASoC: soc-pcm: remove soc_rtdcom_copy_kernel() 4) ASoC: soc-pcm: remove soc_fill_silence() 5) ASoC: add soc-component.c 6) ASoC: soc-component: add snd_soc_component_get/put() 7) ASoC: soc-component: add snd_soc_component_open() 8) ASoC: soc-component: add snd_soc_component_close() 9) ASoC: soc-component: add snd_soc_component_prepare() 10) ASoC: soc-component: add snd_soc_component_hw_params() 11) ASoC: soc-component: add snd_soc_component_hw_free() 12) ASoC: soc-component: add snd_soc_component_trigger() 13) ASoC: soc-component: add snd_soc_component_suspend() 14) ASoC: soc-component: add snd_soc_component_resume() 15) ASoC: soc-component: add snd_soc_component_is_suspended() 16) ASoC: soc-component: add snd_soc_component_probe() 17) ASoC: soc-component: add snd_soc_component_remove() 18) ASoC: soc-component: add snd_soc_component_of_xlate_dai_id() 19) ASoC: soc-component: add snd_soc_component_of_xlate_dai_name() 20) ASoC: soc-component: move snd_soc_component_seq_notifier() 21) ASoC: soc-component: move snd_soc_component_stream_event() 22) ASoC: soc-component: move snd_soc_component_set_bias_level() 23) ASoC: soc-component: add snd_soc_pcm_component_pointer() 24) ASoC: soc-component: add snd_soc_pcm_component_ioctrl() 25) ASoC: soc-component: add snd_soc_pcm_component_copy_user() 26) ASoC: soc-component: add snd_soc_pcm_component_page() 27) ASoC: soc-component: add snd_soc_pcm_component_mmap() 28) ASoC: soc-component: add snd_soc_pcm_component_pcm_new() 29) ASoC: soc-component: add snd_soc_pcm_component_pcm_free() include/sound/soc-component.h | 387 ++++++++++++++++++++++++++ include/sound/soc-dapm.h | 6 - include/sound/soc.h | 306 +-------------------- sound/soc/Makefile | 2 +- sound/soc/intel/skylake/skl-pcm.c | 3 +- sound/soc/soc-component.c | 561 ++++++++++++++++++++++++++++++++++++++ sound/soc/soc-core.c | 123 ++------- sound/soc/soc-dapm.c | 28 +- sound/soc/soc-jack.c | 18 -- sound/soc/soc-pcm.c | 260 ++---------------- sound/soc/soc-utils.c | 199 -------------- 11 files changed, 1012 insertions(+), 881 deletions(-) create mode 100644 include/sound/soc-component.h create mode 100644 sound/soc/soc-component.c -- 2.7.4

3 60

[alsa-devel] [GIT PULL] ASoC updates for v5.3
by Mark Brown 05 Aug '19

05 Aug '19

The following changes since commit 87a6fe80d56be85bda093c4dd8a3e6535bda7e4a: ASoC: SOF: Intel: implement runtime idle for CNL/APL (2019-07-06 12:23:47 +0100) are available in the Git repository at: https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git tags/asoc-fix-v5.3-rc3 for you to fetch changes up to 30c21734d853dae99d05a5295a59b7e26ccd5135: ASoC: amd: acp3x: use dma address for acp3x dma driver (2019-08-02 19:04:43 +0100) ---------------------------------------------------------------- ASoC: Fixes for v5.3 A relatively large batch of mostly unremarkable fixes here, a couple of small core fixes for fairly obscure issues, more comment/email updates with no code impact than usual and a bunch of small driver fixes. The support for new sample rates in the max98373 driver is a fix for the fact that the driver declared support for those rates but would in fact return an error if these rates were selected. ---------------------------------------------------------------- Cezary Rojewski (1): MAINTAINERS: Update Intel ASoC drivers maintainers Charles Keepax (1): ASoC: dapm: Fix handling of custom_stop_condition on DAPM graph walks Cheng-Yi Chiang (1): ASoC: rockchip: Fix mono capture Christophe JAILLET (1): ASoC: Intel: Fix some acpi vs apci typo in somme comments Enric Balletbo i Serra (1): SoC: rockchip: rockchip_max98090: Enable MICBIAS for headset keypress detection Hans de Goede (1): ASoC: Intel: bytcht_es8316: Add quirk for Irbis NB41 netbook Joe Perches (1): ASoC: rt1308: Remove executable attribute from source files Kai Vehmanen (1): ASoC: SOF: Intel: hda: remove misleading error trace from IRQ thread Kirill Marinushkin (1): ASoC: Relocate my e-mail to .com domain zone Kuninori Morimoto (3): ASoC: simple_card_utils.h: care NULL dai at asoc_simple_debug_dai() ASoC: simple-card-utils: care no Platform for DPCM ASoC: audio-graph-card: add missing const at graph_get_dai_id() Marcus Cooper (1): ASoC: sun4i-i2s: Incorrect SR and WSS computation Masahiro Yamada (1): ASoC: SOF: use __u32 instead of uint32_t in uapi headers Peter Ujfalusi (2): ASoC: ti: davinci-mcasp: Fix clk PDIR handling for i2s master mode ASoC: ti: davinci-mcasp: Correct slot_width posed constraint Ricard Wanderlof (1): ASoC: Fail card instantiation if DAI format setup fails Shengjiu Wang (1): ASoC: cs42xx8: Fix MFREQ selection issue for async mode Shuming Fan (1): ASoC: rt1011: fix DC calibration offset not applying Stephan Gerhold (1): ASoC: qcom: apq8016_sbc: Fix oops with multiple DAI links Tzung-Bi Shih (1): ASoC: max98357a: use mdelay for sdmode-delay Vijendar Mukunda (2): ASoC: amd: acp3x: use dma_ops of parent device for acp3x dma driver ASoC: amd: acp3x: use dma address for acp3x dma driver Wen Yang (6): ASoC: simple-card: fix an use-after-free in simple_dai_link_of_dpcm() ASoC: simple-card: fix an use-after-free in simple_for_each_link() ASoC: audio-graph-card: fix use-after-free in graph_dai_link_of_dpcm() ASoC: audio-graph-card: fix an use-after-free in graph_get_dai_id() ASoC: samsung: odroid: fix an use-after-free issue for codec ASoC: samsung: odroid: fix a double-free issue for cpu_dai Wenwen Wang (1): ASoC: dapm: fix a memory leak bug fengchunguo (1): ASoC: max98373: add 88200 and 96000 sampling rate support MAINTAINERS | 3 +- include/sound/simple_card_utils.h | 4 + include/uapi/sound/sof/fw.h | 16 +-- include/uapi/sound/sof/header.h | 14 +-- sound/soc/amd/raven/acp3x-pcm-dma.c | 20 ++-- sound/soc/codecs/cs42xx8.c | 116 +++++++++++++++++---- sound/soc/codecs/max98357a.c | 25 +---- sound/soc/codecs/max98373.c | 6 ++ sound/soc/codecs/max98373.h | 2 + sound/soc/codecs/pcm3060-i2c.c | 4 +- sound/soc/codecs/pcm3060-spi.c | 4 +- sound/soc/codecs/pcm3060.c | 4 +- sound/soc/codecs/pcm3060.h | 2 +- sound/soc/codecs/rt1011.c | 4 +- sound/soc/codecs/rt1308.c | 0 sound/soc/codecs/rt1308.h | 0 sound/soc/generic/audio-graph-card.c | 30 +++--- sound/soc/generic/simple-card-utils.c | 7 ++ sound/soc/generic/simple-card.c | 26 ++--- sound/soc/intel/boards/bytcht_es8316.c | 8 ++ sound/soc/intel/common/soc-acpi-intel-bxt-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-byt-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-cht-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-cnl-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-glk-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-hda-match.c | 2 +- .../intel/common/soc-acpi-intel-hsw-bdw-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-icl-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-kbl-match.c | 2 +- sound/soc/intel/common/soc-acpi-intel-skl-match.c | 2 +- sound/soc/qcom/apq8016_sbc.c | 16 +-- sound/soc/rockchip/rockchip_i2s.c | 5 +- sound/soc/rockchip/rockchip_max98090.c | 32 ++++++ sound/soc/samsung/odroid.c | 8 +- sound/soc/soc-core.c | 7 +- sound/soc/soc-dapm.c | 10 +- sound/soc/sof/intel/cnl.c | 4 +- sound/soc/sof/intel/hda-ipc.c | 4 +- sound/soc/sunxi/sun4i-i2s.c | 4 +- sound/soc/ti/davinci-mcasp.c | 46 ++++++-- 40 files changed, 304 insertions(+), 147 deletions(-) mode change 100644 => 100755 sound/soc/codecs/max98373.c mode change 100644 => 100755 sound/soc/codecs/max98373.h mode change 100755 => 100644 sound/soc/codecs/rt1308.c mode change 100755 => 100644 sound/soc/codecs/rt1308.h

2 3

[alsa-devel] [PATCH 00/19] ALSA: firewire: introduce AMDTP domain
by Takashi Sakamoto 05 Aug '19

05 Aug '19

Hi, This patchset is to implement AMDTP domain I addressed in my previous patchset: https://mailman.alsa-project.org/pipermail/alsa-devel/2019-July/152430.html My work for libhinoko[1][2] allows me to sniff actual packet transmission between devices and drivers in Windows/Mac OS for long period (e.g. 1 hour). As a result, some supported devices don't follow packet sequences transferred by the drivers. They expect drivers to parse isoc packets in tx stream and recover frequency of data block then transfer isoc packets as rx stream. (I note that no specification describes this mechanism as long as I know. In the specification, clock recovery is one-way from transmitter to receivers, thus recovered clock is not necessarily used for transmission from the receiver to the transmitter.) For the clock recovery in driver side, several isoc contexts should be handled in one time to parse packets in incoming stream and to build packets in outgoing stream. The AMDTP domain is designed for this purpose. In this time, the AMDTP domain implements the functionalities to start/stop a couple of isoc contexts for some AMDTP streams. Each context still runs in own handler for scheduled hardware IRQs. I'll post further work after merged. [1] https://mailman.alsa-project.org/pipermail/alsa-devel/2019-April/147862.html [2] https://github.com/takaswie/libhinoko Takashi Sakamoto (19): ALSA: firewire-lib: add AMDTP domain structure to handle several isoc context in one interrupt callback ALSA: firewire-lib: add a kernel API to stop a couple of AMDTP streams in AMDTP domain ALSA: firewire-lib: add a kernel API to add AMDTP stream into AMDTP domain ALSA: firewire-lib: add a kernel API to start AMDTP streams in AMDTP domain ALSA: fireworks: code refactoring for initialization/destruction of AMDTP streams ALSA: fireworks: code refactoring for bus reset handler ALSA: firewire-digi00x: code refactoring for initialization/destruction of AMDTP stream ALSA: firewire-tascam: code refactoring for initialization/destruction of AMDTP stream ALSA: firewire-motu: code refactoring for initialization/destruction of AMDTP stream ALSA: fireface: code refactoring for initialization/destruction of AMDTP stream ALSA: bebob: support AMDTP domain ALSA: fireworks: support AMDTP domain ALSA: oxfw: support AMDTP domain ALSA: dice: support AMDTP domain ALSA: firewire-digi00x: support AMDTP domain ALSA: firewire-tascam: support AMDTP domain ALSA: firewire-motu: support AMDTP domain ALSA: fireface: support AMDTP domain ALSA: firewire-lib: localize kernel APIs to start/stop each AMDTP stream sound/firewire/amdtp-stream.c | 94 ++++++++++- sound/firewire/amdtp-stream.h | 20 ++- sound/firewire/bebob/bebob.h | 2 + sound/firewire/bebob/bebob_stream.c | 62 +++---- sound/firewire/dice/dice-stream.c | 32 +++- sound/firewire/dice/dice.h | 2 + sound/firewire/digi00x/digi00x-stream.c | 106 +++++++----- sound/firewire/digi00x/digi00x.h | 2 + sound/firewire/fireface/ff-stream.c | 96 ++++++----- sound/firewire/fireface/ff.h | 2 + sound/firewire/fireworks/fireworks.h | 2 + sound/firewire/fireworks/fireworks_stream.c | 176 ++++++++++---------- sound/firewire/motu/motu-stream.c | 131 +++++++-------- sound/firewire/motu/motu.h | 2 + sound/firewire/oxfw/oxfw-stream.c | 72 ++++---- sound/firewire/oxfw/oxfw.h | 2 + sound/firewire/tascam/tascam-stream.c | 123 ++++++++------ sound/firewire/tascam/tascam.h | 2 + 18 files changed, 558 insertions(+), 370 deletions(-) -- 2.20.1

2 20