August 2018 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH] sound/core.c: fix wrong return code of rsnd_probe() when devm_kzalloc() fails
by Jiecheng Wu 18 Aug '18

18 Aug '18

Function rsnd_probe() defined in sound/soc/sh/rcar/core.c calls devm_kzalloc(). As devm_kzalloc() may return NULL when OOM happens, the caller should return -ENOMEM when devm_kzalloc() fails. --- sound/soc/sh/rcar/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sh/rcar/core.c b/sound/soc/sh/rcar/core.c index f8425d8..fd92906 100644 --- a/sound/soc/sh/rcar/core.c +++ b/sound/soc/sh/rcar/core.c @@ -1502,7 +1502,7 @@ static int rsnd_probe(struct platform_device *pdev) */ priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); if (!priv) - return -ENODEV; + return -ENOMEM; priv->pdev = pdev; priv->flags = (unsigned long)of_device_get_match_data(dev); -- 2.6.4

1 0

[alsa-devel] [PATCH v6 0/7] Enable HDA Codec support on Intel Platforms
by Pierre-Louis Bossart 17 Aug '18

17 Aug '18

Many Intel platforms (SKL, KBL) etc. in the market supports enhanced audio capabilities which also includes DSP processing. The default HDaudio legacy driver does not allow for the use of the DSP, this patch set makes it possible while reusing existing code for HDAudio codecs and without significant changes to the legacy driver. This v6 is based on Marks' for-next branch - the merge of topic/drm_audio_component is not included here. Tests were run successfully on multiple platforms (Dell XPS13, KBL NUC, APL NUC and LeafHill reference board). Both the HDaudio and HDMI outputs were tested. Credits: all the initial code was written by Rakesh Ughreja, the rebase to broonie/for-next, cleanups and additional tests were done by Pierre Bossart. TODO in future update: 1. fix the HDMI jack detection which only works after the mixer values are set, which isn't practical for headless devices always connected (this is a problem in the hdac_hdmi codec that was present before this series was submitted) 2. when the screen goes blank, HDMI playback stops. This isn't related to this patchset since we rely on the existing hdac_hdmi codec, but will need additional investigations. Changes v6 (comments from Mark and Takashi) - simplify Kconfig options, remove FORCE for now and make code simpler - change SPDIX style to C++ for C files - add log message if invalid configuration reached - remove static int counter, added private counter in context instead - fixed a couple of indendation issues to make checkpatch.pl --strict happy Changes v5 (comments from Vinod) - fix SPDIF style across patches - patch 1: update commit message to remove reference to "fix" - patch 3: remove platform name assignment and Xmas tree style - patch 7: simplify error handling, add missing _put() on errors, remove tests on ops, add switch statement, fix alignment - patch 8: fix indentation issues, use bool instead of tristate Changes v4: - rebase/update on Takashi's topic/drm_audio_component branch - changes in the HDaudio detection to avoid adding a fake ACPI ID - new Kconfigs to control HDaudio detection Changes v3: - port to component model - additional tests on ApolloLake and KabyLake NUC devices - cleanups (alignment, typos, etc) Changes v2: - Resolved review comments and rebased to latest kernel. - added module load support for codec drivers. Pierre-Louis Bossart (2): ASoC: Intel: common: add table for HDA-based platforms ASoC: Intel: Skylake: add option to control HDAudio + DSP usage Rakesh Ughreja (5): ASoC: Intel: Boards: Machine driver for SKL+ w/ HDAudio codecs ASoC: Intel: Skylake: use HDAudio if ACPI enumeration fails ASoC: Intel: Skylake: add HDA BE DAIs ASoC: Intel: Skylake: use hda_bus instead of hdac_bus ASoC: hdac_hda: add asoc extension for legacy HDA codec drivers include/sound/soc-acpi-intel-match.h | 6 + sound/pci/hda/hda_bind.c | 12 + sound/soc/codecs/Kconfig | 5 + sound/soc/codecs/Makefile | 2 + sound/soc/codecs/hdac_hda.c | 485 ++++++++++++++++++ sound/soc/codecs/hdac_hda.h | 24 + sound/soc/intel/Kconfig | 9 + sound/soc/intel/boards/Kconfig | 9 + sound/soc/intel/boards/Makefile | 2 + sound/soc/intel/boards/skl_hda_dsp_common.c | 127 +++++ sound/soc/intel/boards/skl_hda_dsp_common.h | 38 ++ sound/soc/intel/boards/skl_hda_dsp_generic.c | 181 +++++++ sound/soc/intel/common/Makefile | 3 +- .../intel/common/soc-acpi-intel-hda-match.c | 40 ++ sound/soc/intel/skylake/skl-pcm.c | 70 ++- sound/soc/intel/skylake/skl.c | 104 +++- sound/soc/intel/skylake/skl.h | 12 +- 17 files changed, 1102 insertions(+), 27 deletions(-) create mode 100644 sound/soc/codecs/hdac_hda.c create mode 100644 sound/soc/codecs/hdac_hda.h create mode 100644 sound/soc/intel/boards/skl_hda_dsp_common.c create mode 100644 sound/soc/intel/boards/skl_hda_dsp_common.h create mode 100644 sound/soc/intel/boards/skl_hda_dsp_generic.c create mode 100644 sound/soc/intel/common/soc-acpi-intel-hda-match.c -- 2.17.1

1 8

[alsa-devel] [PATCH v2 1/2] ASoC: adau17x1: Unused exported functions changed to internal
by Robert Rosengren 17 Aug '18

17 Aug '18

adau17x1_setup_firmware and adau17x1_has_dsp is only used internally, so making them static instead of exported. Change-Id: I4882513ab457966b0371ac2c1024550ca6edbe0c Signed-off-by: Robert Rosengren <robertr(a)axis.com> --- sound/soc/codecs/adau17x1.c | 9 +++++---- sound/soc/codecs/adau17x1.h | 4 ---- 2 files changed, 5 insertions(+), 8 deletions(-) diff --git a/sound/soc/codecs/adau17x1.c b/sound/soc/codecs/adau17x1.c index 57169b8ff14e..2f2afb4c0c88 100644 --- a/sound/soc/codecs/adau17x1.c +++ b/sound/soc/codecs/adau17x1.c @@ -60,6 +60,9 @@ static const struct snd_kcontrol_new adau17x1_controls[] = { SOC_ENUM("Mic Bias Mode", adau17x1_mic_bias_mode_enum), }; +static int adau17x1_setup_firmware(struct snd_soc_component *component, + unsigned int rate); + static int adau17x1_pll_event(struct snd_soc_dapm_widget *w, struct snd_kcontrol *kcontrol, int event) { @@ -313,7 +316,7 @@ static const struct snd_soc_dapm_route adau17x1_no_dsp_dapm_routes[] = { { "Capture", NULL, "Right Decimator" }, }; -bool adau17x1_has_dsp(struct adau *adau) +static bool adau17x1_has_dsp(struct adau *adau) { switch (adau->type) { case ADAU1761: @@ -324,7 +327,6 @@ bool adau17x1_has_dsp(struct adau *adau) return false; } } -EXPORT_SYMBOL_GPL(adau17x1_has_dsp); static int adau17x1_set_dai_pll(struct snd_soc_dai *dai, int pll_id, int source, unsigned int freq_in, unsigned int freq_out) @@ -836,7 +838,7 @@ bool adau17x1_volatile_register(struct device *dev, unsigned int reg) } EXPORT_SYMBOL_GPL(adau17x1_volatile_register); -int adau17x1_setup_firmware(struct snd_soc_component *component, +static int adau17x1_setup_firmware(struct snd_soc_component *component, unsigned int rate) { int ret; @@ -880,7 +882,6 @@ int adau17x1_setup_firmware(struct snd_soc_component *component, return ret; } -EXPORT_SYMBOL_GPL(adau17x1_setup_firmware); int adau17x1_add_widgets(struct snd_soc_component *component) { diff --git a/sound/soc/codecs/adau17x1.h b/sound/soc/codecs/adau17x1.h index e6fe87beec07..98a3b6f5bc96 100644 --- a/sound/soc/codecs/adau17x1.h +++ b/sound/soc/codecs/adau17x1.h @@ -68,10 +68,6 @@ int adau17x1_resume(struct snd_soc_component *component); extern const struct snd_soc_dai_ops adau17x1_dai_ops; -int adau17x1_setup_firmware(struct snd_soc_component *component, - unsigned int rate); -bool adau17x1_has_dsp(struct adau *adau); - #define ADAU17X1_CLOCK_CONTROL 0x4000 #define ADAU17X1_PLL_CONTROL 0x4002 #define ADAU17X1_REC_POWER_MGMT 0x4009 -- 2.11.0

4 5

[alsa-devel] [RFC PATCH] ASoC: core: Optimise suspend/resume of DAPM widgets
by Jon Hunter 16 Aug '18

16 Aug '18

For soundcards that have several DAI links and many DAPM widgets the time taken for snd_soc_suspend to execute has been observed to be several milliseconds. The time is largely spent executing dapm_power_widgets() for each for the DAI links that need to be suspended. Given that dapm_power_widgets() is called again after suspending/resuming the DAI links, one way to optimise the suspend/resume time is to avoid calling dapm_power_widgets() for each DAI link and reduces the suspend time significantly. Please note that this has been observed on the Tegra210 Jetson TX1 platform which is not currently supported in the mainline for audio but has been tested with out-of-tree patches to enable I2S audio. For Tegra210 the audio drivers (for I2S, etc) are implemented as component drivers to offer greater flexibility with dynamic audio routing within the Tegra Audio Processing Engine (APE) from userspace. The downside of this flexibility is that now there are a lof more DAI links and so the time to suspend the soundcard can take several milliseconds. In the resume path, it is not clear if there could be any issues from not sync'ing the DAPM power state until after unmuting and resuming the CPU DAI drivers, because this will happen later with this change. Signed-off-by: Jon Hunter <jonathanh(a)nvidia.com> --- It is not 100% clear to me if there are any downsides to this approach but I am looking to find away to avoid calling dapm_power_widgets() so many times in the suspend path when there are many DAI links. Unfortunately, I don't think that I can simply mark some of the DAI links with 'ignore_suspend' because it could be possible that they are active. include/sound/soc-dapm.h | 2 ++ sound/soc/soc-core.c | 18 ++++-------------- sound/soc/soc-dapm.c | 43 ++++++++++++++++++++++++++++++++++++++++--- 3 files changed, 46 insertions(+), 17 deletions(-) diff --git a/include/sound/soc-dapm.h b/include/sound/soc-dapm.h index af9ef16cc34d..5dbd601154c4 100644 --- a/include/sound/soc-dapm.h +++ b/include/sound/soc-dapm.h @@ -427,6 +427,8 @@ void snd_soc_dapm_reset_cache(struct snd_soc_dapm_context *dapm); /* dapm events */ void snd_soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream, int event); +void snd_soc_dapm_stream_suspend(struct snd_soc_pcm_runtime *rtd, int stream); +void snd_soc_dapm_stream_resume(struct snd_soc_pcm_runtime *rtd, int stream); void snd_soc_dapm_shutdown(struct snd_soc_card *card); /* external DAPM widget events */ diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index 81b27923303d..03c5fa95afb8 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -519,13 +519,8 @@ int snd_soc_suspend(struct device *dev) if (rtd->dai_link->ignore_suspend) continue; - snd_soc_dapm_stream_event(rtd, - SNDRV_PCM_STREAM_PLAYBACK, - SND_SOC_DAPM_STREAM_SUSPEND); - - snd_soc_dapm_stream_event(rtd, - SNDRV_PCM_STREAM_CAPTURE, - SND_SOC_DAPM_STREAM_SUSPEND); + snd_soc_dapm_stream_suspend(rtd, SNDRV_PCM_STREAM_PLAYBACK); + snd_soc_dapm_stream_suspend(rtd, SNDRV_PCM_STREAM_CAPTURE); } /* Recheck all endpoints too, their state is affected by suspend */ @@ -637,13 +632,8 @@ static void soc_resume_deferred(struct work_struct *work) if (rtd->dai_link->ignore_suspend) continue; - snd_soc_dapm_stream_event(rtd, - SNDRV_PCM_STREAM_PLAYBACK, - SND_SOC_DAPM_STREAM_RESUME); - - snd_soc_dapm_stream_event(rtd, - SNDRV_PCM_STREAM_CAPTURE, - SND_SOC_DAPM_STREAM_RESUME); + snd_soc_dapm_stream_resume(rtd, SNDRV_PCM_STREAM_PLAYBACK); + snd_soc_dapm_stream_resume(rtd, SNDRV_PCM_STREAM_CAPTURE); } /* unmute any active DACs */ diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c index 7e96793050c9..7d4f75d29d55 100644 --- a/sound/soc/soc-dapm.c +++ b/sound/soc/soc-dapm.c @@ -4201,7 +4201,7 @@ void snd_soc_dapm_connect_dai_link_widgets(struct snd_soc_card *card) } static void soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream, - int event) + int event, bool power_widgets) { int i; @@ -4209,7 +4209,8 @@ static void soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream, for (i = 0; i < rtd->num_codecs; i++) soc_dapm_dai_stream_event(rtd->codec_dais[i], stream, event); - dapm_power_widgets(rtd->card, event); + if (power_widgets) + dapm_power_widgets(rtd->card, event); } /** @@ -4229,7 +4230,43 @@ void snd_soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream, struct snd_soc_card *card = rtd->card; mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME); - soc_dapm_stream_event(rtd, stream, event); + soc_dapm_stream_event(rtd, stream, event, true); + mutex_unlock(&card->dapm_mutex); +} + +/** + * snd_soc_dapm_stream_suspend - send a suspend stream event to the dapm core + * @rtd: PCM runtime data + * @stream: stream name + * + * Sends a suspend stream event to the dapm core. + * + * Returns 0 for success else error. + */ +void snd_soc_dapm_stream_suspend(struct snd_soc_pcm_runtime *rtd, int stream) +{ + struct snd_soc_card *card = rtd->card; + + mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME); + soc_dapm_stream_event(rtd, stream, SND_SOC_DAPM_STREAM_SUSPEND, false); + mutex_unlock(&card->dapm_mutex); +} + +/** + * snd_soc_dapm_stream_resume - send a resume stream event to the dapm core + * @rtd: PCM runtime data + * @stream: stream name + * + * Sends a resume stream event to the dapm core. + * + * Returns 0 for success else error. + */ +void snd_soc_dapm_stream_resume(struct snd_soc_pcm_runtime *rtd, int stream) +{ + struct snd_soc_card *card = rtd->card; + + mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME); + soc_dapm_stream_event(rtd, stream, SND_SOC_DAPM_STREAM_RESUME, false); mutex_unlock(&card->dapm_mutex); } -- 2.7.4

2 6

[alsa-devel] [ANN] Linux Media Summit 2018 together with ELC-EU
by Mauro Carvalho Chehab 16 Aug '18

16 Aug '18

Hi all, We're organizing this year's media summit to happen together with the Embedded Linux Conference - Europe, on Thursday, October 25 2018: https://events.linuxfoundation.org/events/elc-openiot-europe-2018/ Media Summit is the premier forum to discuss the Linux multimedia development for cameras, audio and video streaming devices, analog/digital TV support, remote controller and HDMI Consumer Electronics Control (CEC) at the Linux Kernel and its related userspace APIs. The ones interested on participating on it should be registered at ELC or OSS Europe, and indicate the participation at the media summit via Linux Foundation's register site: https://events.linuxfoundation.org/events/elc-openiot-europe-2018/register/ We're also asking for the themes to be discussed there. Please send your proposals to the linux-media(a)vger.kernel.org mailing list. In order to make easier for us to track, please add a [RFP] at the beginning of the email's subject. Hoping to see you there, Regards, Mauro

1 0

[alsa-devel] [PATCH] ASoC: max98373: Added speaker FS gain cotnrol register to volatile.
by Ryan Lee 16 Aug '18

16 Aug '18

Signed-off-by: Ryan Lee <ryans.lee(a)maximintegrated.com> --- Changes : Added Speaker FS gain maximum control(0x203E) register to volatile. sound/soc/codecs/max98373.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/max98373.c b/sound/soc/codecs/max98373.c index 92b7125..2764fae 100644 --- a/sound/soc/codecs/max98373.c +++ b/sound/soc/codecs/max98373.c @@ -520,6 +520,7 @@ static bool max98373_volatile_reg(struct device *dev, unsigned int reg) { switch (reg) { case MAX98373_R2000_SW_RESET ... MAX98373_R2009_INT_FLAG3: + case MAX98373_R203E_AMP_PATH_GAIN: case MAX98373_R2054_MEAS_ADC_PVDD_CH_READBACK: case MAX98373_R2055_MEAS_ADC_THERM_CH_READBACK: case MAX98373_R20B6_BDE_CUR_STATE_READBACK: -- 2.7.4

2 1

[alsa-devel] [PATCH 1/2] ASoC: rt5514: Fix the issue of the delay volume applied again
by Oder Chiou 16 Aug '18

16 Aug '18

After our evaluation, we need to modify the default values to make sure the volume applied immediately. Signed-off-by: Oder Chiou <oder_chiou(a)realtek.com> --- sound/soc/codecs/rt5514.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/rt5514.c b/sound/soc/codecs/rt5514.c index dca82dd..32fe76c 100644 --- a/sound/soc/codecs/rt5514.c +++ b/sound/soc/codecs/rt5514.c @@ -64,8 +64,8 @@ static const struct reg_sequence rt5514_patch[] = { {RT5514_ANA_CTRL_LDO10, 0x00028604}, {RT5514_ANA_CTRL_ADCFED, 0x00000800}, {RT5514_ASRC_IN_CTRL1, 0x00000003}, - {RT5514_DOWNFILTER0_CTRL3, 0x10000352}, - {RT5514_DOWNFILTER1_CTRL3, 0x10000352}, + {RT5514_DOWNFILTER0_CTRL3, 0x10000342}, + {RT5514_DOWNFILTER1_CTRL3, 0x10000342}, }; static const struct reg_default rt5514_reg[] = { @@ -92,10 +92,10 @@ static const struct reg_default rt5514_reg[] = { {RT5514_ASRC_IN_CTRL1, 0x00000003}, {RT5514_DOWNFILTER0_CTRL1, 0x00020c2f}, {RT5514_DOWNFILTER0_CTRL2, 0x00020c2f}, - {RT5514_DOWNFILTER0_CTRL3, 0x10000352}, + {RT5514_DOWNFILTER0_CTRL3, 0x10000342}, {RT5514_DOWNFILTER1_CTRL1, 0x00020c2f}, {RT5514_DOWNFILTER1_CTRL2, 0x00020c2f}, - {RT5514_DOWNFILTER1_CTRL3, 0x10000352}, + {RT5514_DOWNFILTER1_CTRL3, 0x10000342}, {RT5514_ANA_CTRL_LDO10, 0x00028604}, {RT5514_ANA_CTRL_LDO18_16, 0x02000345}, {RT5514_ANA_CTRL_ADC12, 0x0000a2a8}, -- 2.7.4

2 2

[alsa-devel] [PATCH] ALSA: info: Check for integer overflow in snd_info_entry_write()
by Erick Reyes 16 Aug '18

16 Aug '18

Commit 4adb7bcbcb69 ("ALSA: core: Use seq_file for text proc file reads") heavily refactored ALSA procfs and fixed the overflow as a side-effect, so this fix only applies to kernels < 4.2 and there is no upstream equivalent snd_info_entry_write() resizes the buffer with an unsigned long size argument that gets truncated because resize_info_buffer() takes the size parameter as an unsigned int. On 64-bit kernels, this causes the following copy_to_user() to write out-of-bounds if (pos + count) can't be represented by an unsigned int. Signed-off-by: Siqi Lin <siqilin(a)google.com> Signed-off-by: Erick Reyes <erickreyes(a)google.com> --- sound/core/info.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/core/info.c b/sound/core/info.c index 9f404e965ea2..08832c973a53 100644 --- a/sound/core/info.c +++ b/sound/core/info.c @@ -253,6 +253,7 @@ static ssize_t snd_info_entry_write(struct file *file, const char __user *buffer struct snd_info_buffer *buf; ssize_t size = 0; loff_t pos; + unsigned long realloc_size; data = file->private_data; if (snd_BUG_ON(!data)) @@ -261,7 +262,8 @@ static ssize_t snd_info_entry_write(struct file *file, const char __user *buffer pos = *offset; if (pos < 0 || (long) pos != pos || (ssize_t) count < 0) return -EIO; - if ((unsigned long) pos + (unsigned long) count < (unsigned long) pos) + realloc_size = (unsigned long) pos + (unsigned long) count; + if (realloc_size < (unsigned long) pos || realloc_size > UINT_MAX) return -EIO; switch (entry->content) { case SNDRV_INFO_CONTENT_TEXT: -- 2.18.0.865.gffc8e1a3cd6-goog

3 2

[alsa-devel] Alsa test application which does both record and render
by Chakravarthi Pradeep 16 Aug '18

16 Aug '18

Dear ALSA Team. I developed ALSA driver which record the audio data from PCI. I would like to test it using unit test application which can record and render frame by frame. Is there any ALSA unit test application which can be used for testing ? Thanks and Regards Chakravarthi Pradeep.K

1 0

[alsa-devel] [PATCH] control_hw: Fix issue when applying seccomp policy
by Hsin-Yu Chao 16 Aug '18

16 Aug '18

When seccomp policy is applied to filter ioctl syscall with SNDRV_CTL_IOCTL_TLV_COMMAND, SNDRV_CTL_IOCTL_TLV_READ and SNDRV_CTL_IOCTL_TLV_WRITE in whiltelist, alsa-lib still breaks in at snd_ctl_hw_elem_tlv(). Fix the problem by passing unsigned int to ioctl. Signed-off-by: Hsin-Yu Chao <hychao(a)chromium.org> --- src/control/control_hw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/control/control_hw.c b/src/control/control_hw.c index 68eca522..b54d65f2 100644 --- a/src/control/control_hw.c +++ b/src/control/control_hw.c @@ -215,7 +215,7 @@ static int snd_ctl_hw_elem_tlv(snd_ctl_t *handle, int op_flag, unsigned int numid, unsigned int *tlv, unsigned int tlv_size) { - int inum; + unsigned int inum; snd_ctl_hw_t *hw = handle->private_data; struct snd_ctl_tlv *xtlv; -- 2.18.0.865.gffc8e1a3cd6-goog

2 3