September 2022 - Alsa-devel - mailman.alsa-project.org

[PATCH] ASoC: fsl_asrc: Add initialization finishing check in runtime resume
by Shengjiu Wang 06 Sep '22

06 Sep '22

If the initialization is not finished, then filling input data to the FIFO may fail. So it is better to add initialization finishing check in the runtime resume for suspend & resume case. And consider the case of three instances working in parallel, increase the retry times to 50 for more initialization time. Signed-off-by: Shengjiu Wang <shengjiu.wang(a)nxp.com> --- sound/soc/fsl/fsl_asrc.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c index aa5edf32d988..bae263b90ac1 100644 --- a/sound/soc/fsl/fsl_asrc.c +++ b/sound/soc/fsl/fsl_asrc.c @@ -20,6 +20,7 @@ #define IDEAL_RATIO_DECIMAL_DEPTH 26 #define DIVIDER_NUM 64 +#define INIT_TRY_NUM 50 #define pair_err(fmt, ...) \ dev_err(&asrc->pdev->dev, "Pair %c: " fmt, 'A' + index, ##__VA_ARGS__) @@ -579,7 +580,7 @@ static void fsl_asrc_start_pair(struct fsl_asrc_pair *pair) { struct fsl_asrc *asrc = pair->asrc; enum asrc_pair_index index = pair->index; - int reg, retry = 10, i; + int reg, retry = INIT_TRY_NUM, i; /* Enable the current pair */ regmap_update_bits(asrc->regmap, REG_ASRCTR, @@ -592,6 +593,10 @@ static void fsl_asrc_start_pair(struct fsl_asrc_pair *pair) reg &= ASRCFG_INIRQi_MASK(index); } while (!reg && --retry); + /* FIXME: Doesn't treat initialization timeout as error */ + if (!retry) + dev_warn(&asrc->pdev->dev, "initialization isn't finished\n"); + /* Make the input fifo to ASRC STALL level */ regmap_read(asrc->regmap, REG_ASRCNCR, &reg); for (i = 0; i < pair->channels * 4; i++) @@ -1257,6 +1262,7 @@ static int fsl_asrc_runtime_resume(struct device *dev) { struct fsl_asrc *asrc = dev_get_drvdata(dev); struct fsl_asrc_priv *asrc_priv = asrc->private; + int reg, retry = INIT_TRY_NUM; int i, ret; u32 asrctr; @@ -1295,6 +1301,17 @@ static int fsl_asrc_runtime_resume(struct device *dev) regmap_update_bits(asrc->regmap, REG_ASRCTR, ASRCTR_ASRCEi_ALL_MASK, asrctr); + /* Wait for status of initialization for every enabled pairs */ + do { + udelay(5); + regmap_read(asrc->regmap, REG_ASRCFG, &reg); + reg = (reg >> ASRCFG_INIRQi_SHIFT(0)) & 0x7; + } while ((reg != ((asrctr >> ASRCTR_ASRCEi_SHIFT(0)) & 0x7)) && --retry); + + /* FIXME: Doesn't treat initialization timeout as error */ + if (!retry) + dev_warn(dev, "initialization isn't finished\n"); + return 0; disable_asrck_clk: -- 2.34.1

3 6

snd_pcm_readi() takes to much time.
by GitHub issues - opened 06 Sep '22

06 Sep '22

alsa-project/alsa-lib issue #270 was opened from coder0201200: I am using alsa open source library for my project to read and write audio data. For that I am using snd_pcm_readi function to read audio data. I am able to read audio data. But issue is that sometimes it will take so much time to read data.(200-230 msec). Normally this function executed in 1-3 ms that is normal case, but after every 2-4 frames it is taking 200-300 msec. So this is affecting my whole application process. I have also try to set this call as nonblock with the help of snd_pcm_nonblock function. With that I have observed some improvements but still after 2-3 frames it is taking 100msec time to get a data. Issue URL : https://github.com/alsa-project/alsa-lib/issues/270 Repository URL: https://github.com/alsa-project/alsa-lib

1 0

[PATCH v1 1/2] ALSA: hda: cs35l41: Call put_device() in the scope of get_device()
by Andy Shevchenko 06 Sep '22

06 Sep '22

When put_device() is called in another function it's hard to realize that and easy to "fix" the code in a wrong way. Instead, move put_device() to be in the same scope as get_device(), so we prevent appearance of any attempts to "fix" the code. Signed-off-by: Andy Shevchenko <andriy.shevchenko(a)linux.intel.com> --- sound/pci/hda/cs35l41_hda.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/sound/pci/hda/cs35l41_hda.c b/sound/pci/hda/cs35l41_hda.c index 15e2a0009080..12e955931044 100644 --- a/sound/pci/hda/cs35l41_hda.c +++ b/sound/pci/hda/cs35l41_hda.c @@ -1154,7 +1154,6 @@ static int cs35l41_no_acpi_dsd(struct cs35l41_hda *cs35l41, struct device *physd hw_cfg->gpio2.func = CS35L41_INTERRUPT; hw_cfg->gpio2.valid = true; hw_cfg->valid = true; - put_device(physdev); if (strncmp(hid, "CLSA0100", 8) == 0) { hw_cfg->bst_type = CS35L41_EXT_BOOST_NO_VSPK_SWITCH; @@ -1204,9 +1203,10 @@ static int cs35l41_hda_read_acpi(struct cs35l41_hda *cs35l41, const char *hid, i property = "cirrus,dev-index"; ret = device_property_count_u32(physdev, property); - if (ret <= 0) - return cs35l41_no_acpi_dsd(cs35l41, physdev, id, hid); - + if (ret <= 0) { + ret = cs35l41_no_acpi_dsd(cs35l41, physdev, id, hid); + goto err_put_physdev; + } if (ret > ARRAY_SIZE(values)) { ret = -EINVAL; goto err; @@ -1295,8 +1295,9 @@ static int cs35l41_hda_read_acpi(struct cs35l41_hda *cs35l41, const char *hid, i return 0; err: - put_device(physdev); dev_err(cs35l41->dev, "Failed property %s: %d\n", property, ret); +err_put_physdev: + put_device(physdev); return ret; } -- 2.35.1

2 2

[PATCH] ALSA: hda: beep: Simplify keep-power-at-enable behavior
by Takashi Iwai 06 Sep '22

06 Sep '22

The recent fix for IDT codecs to keep the power up while the beep is enabled can be better integrated into the beep helper code. This patch cleans up the code with refactoring. Fixes: 414d38ba8710 ("ALSA: hda/sigmatel: Keep power up while beep is enabled") Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/pci/hda/hda_beep.c | 15 +++++++++++++-- sound/pci/hda/hda_beep.h | 1 + sound/pci/hda/patch_sigmatel.c | 25 ++----------------------- 3 files changed, 16 insertions(+), 25 deletions(-) diff --git a/sound/pci/hda/hda_beep.c b/sound/pci/hda/hda_beep.c index 53a2b89f8983..e63621bcb214 100644 --- a/sound/pci/hda/hda_beep.c +++ b/sound/pci/hda/hda_beep.c @@ -118,6 +118,12 @@ static int snd_hda_beep_event(struct input_dev *dev, unsigned int type, return 0; } +static void turn_on_beep(struct hda_beep *beep) +{ + if (beep->keep_power_at_enable) + snd_hda_power_up_pm(beep->codec); +} + static void turn_off_beep(struct hda_beep *beep) { cancel_work_sync(&beep->beep_work); @@ -125,6 +131,8 @@ static void turn_off_beep(struct hda_beep *beep) /* turn off beep */ generate_tone(beep, 0); } + if (beep->keep_power_at_enable) + snd_hda_power_down_pm(beep->codec); } /** @@ -140,7 +148,9 @@ int snd_hda_enable_beep_device(struct hda_codec *codec, int enable) enable = !!enable; if (beep->enabled != enable) { beep->enabled = enable; - if (!enable) + if (enable) + turn_on_beep(beep); + else turn_off_beep(beep); return 1; } @@ -167,7 +177,8 @@ static int beep_dev_disconnect(struct snd_device *device) input_unregister_device(beep->dev); else input_free_device(beep->dev); - turn_off_beep(beep); + if (beep->enabled) + turn_off_beep(beep); return 0; } diff --git a/sound/pci/hda/hda_beep.h b/sound/pci/hda/hda_beep.h index a25358a4807a..db76e3ddba65 100644 --- a/sound/pci/hda/hda_beep.h +++ b/sound/pci/hda/hda_beep.h @@ -25,6 +25,7 @@ struct hda_beep { unsigned int enabled:1; unsigned int linear_tone:1; /* linear tone for IDT/STAC codec */ unsigned int playing:1; + unsigned int keep_power_at_enable:1; /* set by driver */ struct work_struct beep_work; /* scheduled task for beep event */ struct mutex mutex; void (*power_hook)(struct hda_beep *beep, bool on); diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 7f340f18599c..a794a01a68ca 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -4311,6 +4311,8 @@ static int stac_parse_auto_config(struct hda_codec *codec) if (codec->beep) { /* IDT/STAC codecs have linear beep tone parameter */ codec->beep->linear_tone = spec->linear_tone_beep; + /* keep power up while beep is enabled */ + codec->beep->keep_power_at_enable = 1; /* if no beep switch is available, make its own one */ caps = query_amp_caps(codec, nid, HDA_OUTPUT); if (!(caps & AC_AMPCAP_MUTE)) { @@ -4444,28 +4446,6 @@ static int stac_suspend(struct hda_codec *codec) return 0; } - -static int stac_check_power_status(struct hda_codec *codec, hda_nid_t nid) -{ -#ifdef CONFIG_SND_HDA_INPUT_BEEP - struct sigmatel_spec *spec = codec->spec; -#endif - int ret = snd_hda_gen_check_power_status(codec, nid); - -#ifdef CONFIG_SND_HDA_INPUT_BEEP - if (nid == spec->gen.beep_nid && codec->beep) { - if (codec->beep->enabled != spec->beep_power_on) { - spec->beep_power_on = codec->beep->enabled; - if (spec->beep_power_on) - snd_hda_power_up_pm(codec); - else - snd_hda_power_down_pm(codec); - } - ret |= spec->beep_power_on; - } -#endif - return ret; -} #else #define stac_suspend NULL #endif /* CONFIG_PM */ @@ -4478,7 +4458,6 @@ static const struct hda_codec_ops stac_patch_ops = { .unsol_event = snd_hda_jack_unsol_event, #ifdef CONFIG_PM .suspend = stac_suspend, - .check_power_status = stac_check_power_status, #endif }; -- 2.35.3

1 0

[PATCH] ALSA: hda: Once again fix regression of page allocations with IOMMU
by Takashi Iwai 06 Sep '22

06 Sep '22

The last fix for trying to recover the regression on AMD platforms, unfortunately, leaded to yet another regression: it turned out that IOMMUs don't like the usage of raw page allocations. This is yet another attempt for addressing the log saga; at this time, we re-use the existing buffer allocation mechanism with SG-pages although we require only single pages. The SG buffer allocation itself was confirmed to work for stream buffers, so it's relatively easy to adapt for other places. The only problem is: although the HD-audio code is accessing the address directly via dmab->address field, SG-pages don't set up it. For the ease of adaption, we now set up the dmab->addr field from the address of the first page as default, so that it can run with the HD-audio driver code as-is without the excessive call of snd_sgbuf_get_addr() multiple times; that's the only change in the memalloc helper side. The rest is nothing but a flip of the dma_type field in the HD-audio side. Fixes: a8d302a0b770 ("ALSA: memalloc: Revive x86-specific WC page allocations again") Reported-by: Mikhail Gavrilov <mikhail.v.gavrilov(a)gmail.com> Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov(a)gmail.com> Cc: <stable(a)vger.kernel.org> Link: https://lore.kernel.org/r/CABXGCsO+kB2t5QyHY-rUe76npr1m0-5JOtt8g8SiHUo34ur7… Link: https://bugzilla.kernel.org/show_bug.cgi?id=216112 Link: https://bugzilla.kernel.org/show_bug.cgi?id=216363 Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/core/memalloc.c | 9 +++++++-- sound/pci/hda/hda_intel.c | 2 +- 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/sound/core/memalloc.c b/sound/core/memalloc.c index b665ac66ccbe..cfcd8eff4139 100644 --- a/sound/core/memalloc.c +++ b/sound/core/memalloc.c @@ -543,10 +543,13 @@ static void *snd_dma_noncontig_alloc(struct snd_dma_buffer *dmab, size_t size) dmab->dev.need_sync = dma_need_sync(dmab->dev.dev, sg_dma_address(sgt->sgl)); p = dma_vmap_noncontiguous(dmab->dev.dev, size, sgt); - if (p) + if (p) { dmab->private_data = sgt; - else + /* store the first page address for convenience */ + dmab->addr = snd_sgbuf_get_addr(dmab, 0); + } else { dma_free_noncontiguous(dmab->dev.dev, size, sgt, dmab->dev.dir); + } return p; } @@ -780,6 +783,8 @@ static void *snd_dma_sg_fallback_alloc(struct snd_dma_buffer *dmab, size_t size) if (!p) goto error; dmab->private_data = sgbuf; + /* store the first page address for convenience */ + dmab->addr = snd_sgbuf_get_addr(dmab, 0); return p; error: diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index a77165bd92a9..b20694fd69de 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -1817,7 +1817,7 @@ static int azx_create(struct snd_card *card, struct pci_dev *pci, /* use the non-cached pages in non-snoop mode */ if (!azx_snoop(chip)) - azx_bus(chip)->dma_type = SNDRV_DMA_TYPE_DEV_WC; + azx_bus(chip)->dma_type = SNDRV_DMA_TYPE_DEV_WC_SG; if (chip->driver_type == AZX_DRIVER_NVIDIA) { dev_dbg(chip->card->dev, "Enable delay in RIRB handling\n"); -- 2.35.3

1 0

[PATCH] ASoC: max98390: Remove unnecessary amp on/off conrtol
by Steve Lee 06 Sep '22

06 Sep '22

The Amp is already control in userspace before trigger calibrate function. Remove unnecessary control in calibrate function. Signed-off-by: Steve Lee <steve.lee.analog(a)gmail.com> --- sound/soc/codecs/max98390.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/sound/soc/codecs/max98390.c b/sound/soc/codecs/max98390.c index 91c0bf3d76fc..99e769190568 100644 --- a/sound/soc/codecs/max98390.c +++ b/sound/soc/codecs/max98390.c @@ -826,9 +826,6 @@ static int max98390_dsm_calibrate(struct snd_soc_component *component) struct max98390_priv *max98390 = snd_soc_component_get_drvdata(component); - regmap_write(max98390->regmap, MAX98390_R203A_AMP_EN, 0x81); - regmap_write(max98390->regmap, MAX98390_R23FF_GLOBAL_EN, 0x01); - regmap_read(max98390->regmap, THERMAL_RDC_RD_BACK_BYTE1, &rdc); regmap_read(max98390->regmap, @@ -847,9 +844,6 @@ static int max98390_dsm_calibrate(struct snd_soc_component *component) dev_info(component->dev, "rdc resistance about %d.%02d ohm, reg=0x%X temp reg=0x%X\n", rdc_integer, rdc_factor, rdc_cal_result, temp); - regmap_write(max98390->regmap, MAX98390_R23FF_GLOBAL_EN, 0x00); - regmap_write(max98390->regmap, MAX98390_R203A_AMP_EN, 0x80); - return 0; } -- 2.17.1

3 4

[PATCH] ALSA: usb-audio: Fix an out-of-bounds bug in __snd_usb_parse_audio_interface()
by Dongxiang Ke 06 Sep '22

06 Sep '22

There may be a bad USB audio device with a USB ID of (0x04fa, 0x4201) and the number of it's interfaces less than 4, an out-of-bounds read bug occurs when parsing the interface descriptor for this device. Fix this by checking the number of interfaces. Signed-off-by: Dongxiang Ke <kdx.glider(a)gmail.com> --- sound/usb/stream.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/usb/stream.c b/sound/usb/stream.c index ceb93d798182..40ce8a1cb318 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -1105,7 +1105,7 @@ static int __snd_usb_parse_audio_interface(struct snd_usb_audio *chip, * Dallas DS4201 workaround: It presents 5 altsettings, but the last * one misses syncpipe, and does not produce any sound. */ - if (chip->usb_id == USB_ID(0x04fa, 0x4201)) + if (chip->usb_id == USB_ID(0x04fa, 0x4201) && num >= 4) num = 4; for (i = 0; i < num; i++) { -- 2.30.2

2 1

[PATCH] ALSA: hda/tegra: Align BDL entry to 4KB boundary
by Mohan Kumar 06 Sep '22

06 Sep '22

AZA HW may send a burst read/write request crossing 4K memory boundary. The 4KB boundary is not guaranteed by Tegra HDA HW. Make SW change to include the flag AZX_DCAPS_4K_BDLE_BOUNDARY to align BDLE to 4K boundary. Signed-off-by: Mohan Kumar <mkumard(a)nvidia.com> --- sound/pci/hda/hda_tegra.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/sound/pci/hda/hda_tegra.c b/sound/pci/hda/hda_tegra.c index 7debb2c76aa6..976a112c7d00 100644 --- a/sound/pci/hda/hda_tegra.c +++ b/sound/pci/hda/hda_tegra.c @@ -474,7 +474,8 @@ MODULE_DEVICE_TABLE(of, hda_tegra_match); static int hda_tegra_probe(struct platform_device *pdev) { const unsigned int driver_flags = AZX_DCAPS_CORBRP_SELF_CLEAR | - AZX_DCAPS_PM_RUNTIME; + AZX_DCAPS_PM_RUNTIME | + AZX_DCAPS_4K_BDLE_BOUNDARY; struct snd_card *card; struct azx *chip; struct hda_tegra *hda; -- 2.17.1

2 1

[PATCH v3 0/4] ASoC platform driver for Apple MCA
by Martin Povišer 06 Sep '22

06 Sep '22

Hi all, sending what should be the final touches on Apple MCA driver. It most likely goes without saying but please do not merge the DT additions into the ASoC tree. Martin Changes since v2: - fix Mark's style nit - order compatibles in schema - redo the DT additions for t8103 - remove stray unlock in mca_be_prepare (and rename the locking patch) Link: https://lore.kernel.org/asahi/20220819125430.4920-1-povik+lin@cutebit.org/T… Changes since v1: - reflect the set_fmt provider/consumer change in 6.0-rc1 - fix the probe function requesting optional reset - unroll 'dma-names' in schema - fix binding schema/code discrepancy in register ranges - minor things Link: https://lore.kernel.org/asahi/20220808224153.3634-1-povik+lin@cutebit.org/T… Changes since 'macaudio RFC v2' [0]: - addition of locking (extra commit) - transition to set_bclk_ratio (instead of getting the bclk ratio from set_sysclk) - using shared reset control and documenting the reset in binding - formatting, comments, and a minor fix to hw driving Link: https://lore.kernel.org/asahi/20220606191910.16580-1-povik+lin@cutebit.org/… Martin Povišer (4): dt-bindings: sound: Add Apple MCA I2S transceiver arm64: dts: apple: t8103: Add MCA and its support ASoC: apple: mca: Start new platform driver ASoC: apple: mca: Add locking .../devicetree/bindings/sound/apple,mca.yaml | 131 ++ MAINTAINERS | 8 + arch/arm64/boot/dts/apple/t8103-jxxx.dtsi | 4 + arch/arm64/boot/dts/apple/t8103.dtsi | 73 ++ sound/soc/Kconfig | 1 + sound/soc/Makefile | 1 + sound/soc/apple/Kconfig | 9 + sound/soc/apple/Makefile | 3 + sound/soc/apple/mca.c | 1167 +++++++++++++++++ 9 files changed, 1397 insertions(+) create mode 100644 Documentation/devicetree/bindings/sound/apple,mca.yaml create mode 100644 sound/soc/apple/Kconfig create mode 100644 sound/soc/apple/Makefile create mode 100644 sound/soc/apple/mca.c -- 2.33.0

4 9

amidi: --send shouldn't require a seekable file
by GitHub issues - opened 06 Sep '22

06 Sep '22

alsa-project/alsa-utils issue #167 was opened from BasedUser: Just like `--receive`, `--send` should instead send events to a given port until the sent file returns EOF, or `amidi` is terminated. Pipes and FIFOs cannot currently be used to create a persistent MIDI source. Issue URL : https://github.com/alsa-project/alsa-utils/issues/167 Repository URL: https://github.com/alsa-project/alsa-utils

1 0