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

[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

[PATCH 0/2] Untested TAS2562 power setting fixes
by Martin Povišer 05 Sep '22

05 Sep '22

The tas2562 driver does the same thing with the setting of PWR_CTRL field as the tas2764/tas2770 drivers were doing. Link: https://lore.kernel.org/alsa-devel/20220808141246.5749-1-povik+lin@cutebit.… Link: https://lore.kernel.org/alsa-devel/20220825140241.53963-1-povik+lin@cutebit… These are blindly written patches without testing since I don't have the hardware. (I even tried TI's formal sample request program but was refused there. CCing @ti.com addresses I found on other series recently submitted.) Martin Povišer (2): ASoC: tas2562: Drop conflicting set_bias_level power setting ASoC: tas2562: Fix mute/unmute sound/soc/codecs/tas2562.c | 90 ++++++++++++-------------------------- 1 file changed, 29 insertions(+), 61 deletions(-) -- 2.33.0

2 5

[PATCH v6 0/2] lib/string_helpers: Introduce parse_int_array_user()
by Cezary Rojewski 05 Sep '22

05 Sep '22

Continuation of recent upstream discussion [1] regarding user string tokenization. First, parse_int_array_user() is introduced to allow for splitting specified user string into a sequence of integers. Makes use of get_options() internally so the parsing logic is not duplicated. With that done, redundant parts of the sound driver are removed. Originally similar functionality was added for the SOF sound driver. As more users are on the horizon, it is desirable to update existing string_helpers code and provide a unified solution. Changes in v6: - minor improvements in SOF code: dropped superfluous array-empty checks as suggested by Andy Changes in v5: - fixed kernel doc for parse_int_array_user() Changes in v4: - renamed the function to parse_int_array_user() - at the name several local variable names have been reworded to match the above Changes in v3: - relocated tokenize_user_input() implementation to string_helpers as requested by Matthew Changes in v2: - reused get_options() so no parsing logic is duplicated - simplified __user variant with help of memdup_user_nul() Both suggested by Andy, thanks for thorough review [1]: https://lore.kernel.org/alsa-devel/20220707091301.1282291-1-cezary.rojewski… Cezary Rojewski (2): lib/string_helpers: Introduce parse_int_array_user() ASoC: SOF: Remove strsplit_u32() and tokenize_input() include/linux/string_helpers.h | 2 + lib/string_helpers.c | 44 +++++++++++++ sound/soc/sof/sof-client-probes.c | 104 +++++------------------------- 3 files changed, 61 insertions(+), 89 deletions(-) -- 2.25.1

2 3

[PATCH] ASoC: SOF: Introduce function sof_of_machine_select
by Chunxu Li 05 Sep '22

05 Sep '22

>From current design in sof_machine_check the SOF can only support ACPI type machine. In sof_machine_check if there is no ACPI machine exist, the function will return -ENODEV directly, that's we don't expected if we do not base on ACPI machine. So we add a new function named sof_of_machine_select that we can pass sof_machine_check and obtain info required by snd_sof_new_platform_drv. Signed-off-by: Chunxu Li <chunxu.li(a)mediatek.com> --- sound/soc/sof/sof-audio.c | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) diff --git a/sound/soc/sof/sof-audio.c b/sound/soc/sof/sof-audio.c index c18e723435bd..ea9663d448eb 100644 --- a/sound/soc/sof/sof-audio.c +++ b/sound/soc/sof/sof-audio.c @@ -785,6 +785,28 @@ int sof_dai_get_bclk(struct snd_soc_pcm_runtime *rtd) } EXPORT_SYMBOL(sof_dai_get_bclk); +static struct snd_sof_of_mach *sof_of_machine_select(struct snd_sof_dev *sdev) +{ + struct snd_sof_pdata *sof_pdata = sdev->pdata; + const struct sof_dev_desc *desc = sof_pdata->desc; + struct snd_sof_of_mach *mach = desc->of_machines; + + if (!mach) + return NULL; + + for (; mach->compatible; mach++) { + if (of_machine_is_compatible(mach->compatible)) { + sof_pdata->tplg_filename = mach->sof_tplg_filename; + if (mach->fw_filename) + sof_pdata->fw_filename = mach->fw_filename; + + return mach; + } + } + + return NULL; +} + /* * SOF Driver enumeration. */ @@ -795,6 +817,7 @@ int sof_machine_check(struct snd_sof_dev *sdev) struct snd_soc_acpi_mach *mach; if (!IS_ENABLED(CONFIG_SND_SOC_SOF_FORCE_NOCODEC_MODE)) { + const struct snd_sof_of_mach *of_mach; /* find machine */ mach = snd_sof_machine_select(sdev); @@ -804,6 +827,12 @@ int sof_machine_check(struct snd_sof_dev *sdev) return 0; } + of_mach = sof_of_machine_select(sdev); + if (of_mach) { + sof_pdata->of_machine = of_mach; + return 0; + } + if (!IS_ENABLED(CONFIG_SND_SOC_SOF_NOCODEC)) { dev_err(sdev->dev, "error: no matching ASoC machine driver found - aborting probe\n"); return -ENODEV; -- 2.25.1

3 2

[PATCH v2] ASoC: wm_adsp: Handle optional legacy support
by Cristian Ciocaltea 05 Sep '22

05 Sep '22

The tracing capabilities for the speaker protection fw enabled via commit c55b3e46cb99 ("ASoC: wm_adsp: Add trace caps to speaker protection FW") are not be available on all platforms, such as the Valve's Steam Deck which is based on the Halo Core DSP. As a consequence, whenever the firmware is loaded, a rather misleading 'Failed to parse legacy: -19' error message is written to the kernel ring buffer: [ 288.977412] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: Firmware version: 3 [ 288.978002] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: cs35l41-dsp1-spk-prot.wmfw: Fri 02 Apr 2021 21:03:50 W. Europe Daylight Time [ 289.094065] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: Firmware: 400a4 vendor: 0x2 v0.33.0, 2 algorithms [ 289.095073] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: 0: ID cd v29.53.0 XM@94 YM@e [ 289.095665] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: 1: ID f20b v0.0.1 XM@170 YM@0 [ 289.096275] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: Protection: C:\Users\ocanavan\Desktop\cirrusTune_july2021.bin [ 291.172383] steamdeck kernel: cs35l41 spi-VLV1776:01: DSP1: Failed to parse legacy: -19 Update wm_adsp_buffer_init() to print a more descriptive info message when wm_adsp_buffer_parse_legacy() returns -ENODEV. Fixes: c55b3e46cb99 ("ASoC: wm_adsp: Add trace caps to speaker protection FW") Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea(a)collabora.com> --- Changes in v2: - Add a generic handling of the issue, not bound to a particular firmware, as suggested by Charles - Update commit message, including the subject, to reflect the changes sound/soc/codecs/wm_adsp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index cfaa45ede916..8a2e9771bb50 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1602,7 +1602,9 @@ static int wm_adsp_buffer_init(struct wm_adsp *dsp) if (list_empty(&dsp->buffer_list)) { /* Fall back to legacy support */ ret = wm_adsp_buffer_parse_legacy(dsp); - if (ret) + if (ret == -ENODEV) + adsp_info(dsp, "Legacy support not available\n"); + else if (ret) adsp_warn(dsp, "Failed to parse legacy: %d\n", ret); } base-commit: 568035b01cfb107af8d2e4bd2fb9aea22cf5b868 -- 2.37.2

3 2