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

Poor performace on mmap reading arm64 on audio device
by Michael Nazzareno Trimarchi 23 Nov '20

23 Nov '20

Hi all I'm trying to figure out how to increase performance on audio reading using the mmap interface. Right now what I understand it's that allocation comes from core/memalloc.c ops that allocate the memory for dma under driver/dma. The reference platform I have is an imx8mm and the allocation in arm64 is: 0xffff800011ff5000-0xffff800012005000 64K PTE RW NX SHD AF UXN MEM/NORMAL-NC This is the reason that is allocated for dma interface. Now access linear on the multichannel interface the performance is bad but worse if I try to access a channel a time on read. So it looks like it is better to copy the block using memcpy on a cached area and then operate on a single channel sample. If it's correct what I'm saying the mmap_begin and mmap_commit basically they don't do anything on cache level so the page mapping and way is used is always the same. Can the interface be modified to allow cache the area during read and restore in the commit phase? Michael

2 6

[PATCH v5 0/6] Tegra210 audio graph card
by Sameer Pujar 23 Nov '20

23 Nov '20

This series adds audio graph based sound card support for Tegra210 platforms like Jetson-TX1 an Jetson-Nano. The following preparatory audio graph enhancement series is already merged. * https://patchwork.kernel.org/project/alsa-devel/list/?series=375629&state=* Following are the summary of changes: * Add graph/audio-graph based schemas or schema updates for Tegra210 component and machine drivers. * Add Tegra audio graph machine driver. * Add required DT support for Jetson-TX1/Nano. This work is based on earlier discussion of DPCM usage for Tegra and simple card driver updates. * https://lkml.org/lkml/2020/4/30/519 * https://lkml.org/lkml/2020/6/27/4 This series has dependency over following graph and audio-graph series. * http://patchwork.ozlabs.org/project/devicetree-bindings/patch/2020110220365… * https://patchwork.kernel.org/project/alsa-devel/list/?series=382009&state=* Changelog ========= v4 -> v5 -------- * Audio graph related changes were sent in separate v5 series as mentioned above and are dropped from current series. * Graph and audio graph doc patches are dropped from this series and are sent separately as mentioned above. * Minor change with phandle label for TX1 and Nano platform DT files. * No changes in other patches. v3 -> v4 -------- * Added new patches to convert graph.txt and audio-graph-card.txt to corresponding json-schema files. Later these references are used in Tegra audio graph schema. * AHUB component binding docs are updated to reflect the usage of ports/port/endpoint * More common stuff is moved into graph_parse_of() and this is used by both generic and Tegra audio graph. * DT binding for Tegra audio graph is updated to included "ports { }" * As per the suggestion 'void *data' member is dropped from 'asoc_simple_priv' and instead container method is used to maintain required custom data internal to Tegra audio graph. v2 -> v3 -------- * Dropped new compatible addition in generic graph driver after reviewing it with Morimoto-san. Instead added Tegra audio graph driver and new compatibles are added in the same. * Added new patches to expose new members for customization in audio graph driver. * Added new patch for Tegra audio graph driver and related documentation. * Minor change in below commit where mutex version of helper is used "ASoC: audio-graph: Identify 'no_pcm' DAI links for DPCM" * DT binding is updated to use the newly exposed compatibles * No changes in other patches v1 -> v2 -------- * Re-organized ports/endpoints description for ADMAIF and XBAR. Updated DT patches accordingly. * After above change, multiple Codec endpoint support is not required and hence dropped for now. This will be considered separately if at all required in future. * Re-ordered patches in the series. Sameer Pujar (6): ASoC: dt-bindings: tegra: Add graph bindings ASoC: dt-bindings: tegra: Add json-schema for Tegra audio graph card ASoC: tegra: Add audio graph based card driver arm64: defconfig: Enable Tegra audio graph card driver arm64: tegra: Audio graph header for Tegra210 arm64: tegra: Audio graph sound card for Jetson Nano and TX1 .../sound/nvidia,tegra-audio-graph-card.yaml | 187 +++++++++++++++ .../bindings/sound/nvidia,tegra186-dspk.yaml | 6 + .../bindings/sound/nvidia,tegra210-admaif.yaml | 6 + .../bindings/sound/nvidia,tegra210-ahub.yaml | 13 +- .../bindings/sound/nvidia,tegra210-dmic.yaml | 6 + .../bindings/sound/nvidia,tegra210-i2s.yaml | 6 + .../boot/dts/nvidia/tegra210-audio-graph.dtsi | 153 ++++++++++++ arch/arm64/boot/dts/nvidia/tegra210-p2371-2180.dts | 262 +++++++++++++++++++++ arch/arm64/boot/dts/nvidia/tegra210-p3450-0000.dts | 146 ++++++++++++ arch/arm64/configs/defconfig | 1 + sound/soc/tegra/Kconfig | 9 + sound/soc/tegra/Makefile | 2 + sound/soc/tegra/tegra_audio_graph_card.c | 255 ++++++++++++++++++++ 13 files changed, 1050 insertions(+), 2 deletions(-) create mode 100644 Documentation/devicetree/bindings/sound/nvidia,tegra-audio-graph-card.yaml create mode 100644 arch/arm64/boot/dts/nvidia/tegra210-audio-graph.dtsi create mode 100644 sound/soc/tegra/tegra_audio_graph_card.c -- 2.7.4

3 11

[PATCH] ALSA: usb-audio : Add support for the Pioneer DJ DDJ-RR controller.
by Daniel 21 Nov '20

21 Nov '20

--- sound/usb/quirks-table.h | 56 +++++++++++++++++++++++++++++++++++++++- 1 file changed, 55 insertions(+), 1 deletion(-) diff --git a/sound/usb/quirks-table.h b/sound/usb/quirks-table.h index 3c1697f6b60c..a39233cb4d72 100644 --- a/sound/usb/quirks-table.h +++ b/sound/usb/quirks-table.h @@ -3530,7 +3530,61 @@ AU0828_DEVICE(0x2040, 0x7270, "Hauppauge", "HVR-950Q"), } } }, - +{ + /* + * PIONEER DJ DDJ-RR + * PCM is 6 channels out & 4 channels in @ 44.1 fixed + */ + USB_DEVICE_VENDOR_SPEC(0x2b73, 0x000d), + .driver_info = (unsigned long) &(const struct snd_usb_audio_quirk) { + .ifnum = QUIRK_ANY_INTERFACE, + .type = QUIRK_COMPOSITE, + .data = (const struct snd_usb_audio_quirk[]) { + { + .ifnum = 0, + .type = QUIRK_AUDIO_FIXED_ENDPOINT, + .data = &(const struct audioformat) { + .formats = SNDRV_PCM_FMTBIT_S24_3LE, + .channels = 6, //Master, Headphones & Booth + .iface = 0, + .altsetting = 1, + .altset_idx = 1, + .endpoint = 0x01, + .ep_attr = USB_ENDPOINT_XFER_ISOC| + USB_ENDPOINT_SYNC_ASYNC, + .rates = SNDRV_PCM_RATE_44100, + .rate_min = 44100, + .rate_max = 44100, + .nr_rates = 1, + .rate_table = (unsigned int[]) { 44100 } + } + }, + { + .ifnum = 0, + .type = QUIRK_AUDIO_FIXED_ENDPOINT, + .data = &(const struct audioformat) { + .formats = SNDRV_PCM_FMTBIT_S24_3LE, + .channels = 4, //2x RCA inputs (CH1 & CH2) + .iface = 0, + .altsetting = 1, + .altset_idx = 1, + .endpoint = 0x82, + .ep_attr = USB_ENDPOINT_XFER_ISOC| + USB_ENDPOINT_SYNC_ASYNC| + USB_ENDPOINT_USAGE_IMPLICIT_FB, + .rates = SNDRV_PCM_RATE_44100, + .rate_min = 44100, + .rate_max = 44100, + .nr_rates = 1, + .rate_table = (unsigned int[]) { 44100 } + } + }, + { + .ifnum = -1 + } + } + } +}, { /* * Pioneer DJ DJM-900NXS2 -- 2.29.2

3 2

[PATCH V2] ALSA: emu10k1: Use dma_set_mask_and_coherent to simplify code
by Christophe JAILLET 21 Nov '20

21 Nov '20

'pci_set_dma_mask()' + 'pci_set_consistent_dma_mask()' can be replaced by an equivalent 'dma_set_mask_and_coherent()' which is much less verbose. Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> --- v2: do not update the dev_err message, to keep consistency with other dev_err --- sound/pci/emu10k1/emu10k1x.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/pci/emu10k1/emu10k1x.c b/sound/pci/emu10k1/emu10k1x.c index def8161cde4c..f2b6fd434474 100644 --- a/sound/pci/emu10k1/emu10k1x.c +++ b/sound/pci/emu10k1/emu10k1x.c @@ -894,8 +894,8 @@ static int snd_emu10k1x_create(struct snd_card *card, if ((err = pci_enable_device(pci)) < 0) return err; - if (pci_set_dma_mask(pci, DMA_BIT_MASK(28)) < 0 || - pci_set_consistent_dma_mask(pci, DMA_BIT_MASK(28)) < 0) { + + if (dma_set_mask_and_coherent(&pci->dev, DMA_BIT_MASK(28)) < 0) { dev_err(card->dev, "error to set 28bit mask DMA\n"); pci_disable_device(pci); return -ENXIO; -- 2.27.0

2 1

[PATCH] ALSA: aloop: Constify ops structs
by Rikard Falkeborn 21 Nov '20

21 Nov '20

The only usage of the ops field in the loopback_cable struct is to call its members, the field it self is never changed. Make it a pointer to const. This allows us to constify two static loopback_ops structs to allow the compiler to put them in read-only memory. Signed-off-by: Rikard Falkeborn <rikard.falkeborn(a)gmail.com> --- sound/drivers/aloop.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/drivers/aloop.c b/sound/drivers/aloop.c index c91356326699..702f91b9c60f 100644 --- a/sound/drivers/aloop.c +++ b/sound/drivers/aloop.c @@ -105,7 +105,7 @@ struct loopback_cable { unsigned int running; unsigned int pause; /* timer specific */ - struct loopback_ops *ops; + const struct loopback_ops *ops; /* If sound timer is used */ struct { int stream; @@ -1021,7 +1021,7 @@ static int loopback_jiffies_timer_open(struct loopback_pcm *dpcm) return 0; } -static struct loopback_ops loopback_jiffies_timer_ops = { +static const struct loopback_ops loopback_jiffies_timer_ops = { .open = loopback_jiffies_timer_open, .start = loopback_jiffies_timer_start, .stop = loopback_jiffies_timer_stop, @@ -1172,7 +1172,7 @@ static int loopback_snd_timer_open(struct loopback_pcm *dpcm) /* stop_sync() is not required for sound timer because it does not need to be * restarted in loopback_prepare() on Xrun recovery */ -static struct loopback_ops loopback_snd_timer_ops = { +static const struct loopback_ops loopback_snd_timer_ops = { .open = loopback_snd_timer_open, .start = loopback_snd_timer_start, .stop = loopback_snd_timer_stop, -- 2.29.2

2 1

[PATCH] ALSA: emu10k1: Use dma_set_mask_and_coherent to simplify code
by Christophe JAILLET 21 Nov '20

21 Nov '20

'pci_set_dma_mask()' + 'pci_set_consistent_dma_mask()' can be replaced by an equivalent 'dma_set_mask_and_coherent()' which is much less verbose. With at it replace the 'card->dev' parameter passed to a 'dev_err()' call by an equivalent '&pci->dev' which is more consistent with the test just one line above. Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> --- sound/pci/emu10k1/emu10k1x.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/pci/emu10k1/emu10k1x.c b/sound/pci/emu10k1/emu10k1x.c index def8161cde4c..f2b6fd434474 100644 --- a/sound/pci/emu10k1/emu10k1x.c +++ b/sound/pci/emu10k1/emu10k1x.c @@ -894,9 +894,9 @@ static int snd_emu10k1x_create(struct snd_card *card, if ((err = pci_enable_device(pci)) < 0) return err; - if (pci_set_dma_mask(pci, DMA_BIT_MASK(28)) < 0 || - pci_set_consistent_dma_mask(pci, DMA_BIT_MASK(28)) < 0) { - dev_err(card->dev, "error to set 28bit mask DMA\n"); + + if (dma_set_mask_and_coherent(&pci->dev, DMA_BIT_MASK(28)) < 0) { + dev_err(&pci->dev, "error to set 28bit mask DMA\n"); pci_disable_device(pci); return -ENXIO; } -- 2.27.0

2 1

Announcement for alsa-gobject v0.2.0 release
by Takashi Sakamoto 21 Nov '20

21 Nov '20

Hi, This is an announcement for the second official release of alsa-gobject project[1]. The release is available in below URL: * https://github.com/alsa-project/alsa-gobject/releases/tag/v0.2.0 It includes some enhancements and bug fixes. * Some implementations of GLib GQuark is exported as error domain and corresponding enumerations are added. This improves error reporting in each libraries. It closes issue #47[2]. * Selection of substream in ALSARawmidi.StreamPair.open() now works well. This closes issue #53[3]. * Prototypes of some methods in ALSACtl.Card object for data of TLV(Type-Length-Value) in ALSA control interface are changed without backward compatibility. Corresponding exported symbols are maintained by new symbol version. In detail, refer to #55[4]. * Error handling is changed so that function call with invalid argument just returns with warnings instead of generating GError, according to convention for GLib application. Memory allocation failure is supposed to be unrecoverable error and aborts program, as well. * It's possible to build for older version Linux kernel, therefore build dependency is now Linux kernel v4.5 or later. For the new release, documentations are also updated. They're auto-generated by gtk-doc[5] from function annotations: * https://alsa-project.github.io/alsa-gobject-docs/ The new release of Rust crates are available as well in repository. They're auto-generated by gtk-rs/gir[6]: * https://github.com/alsa-project/alsa-gobject-rs/releases/tag/v0.2.0 I note that ALSA PCM interface is not supported yet. Please refer to my announcement for v0.1.0 for the reason[7]. [1] https://github.com/alsa-project/alsa-gobject [2] https://github.com/alsa-project/alsa-gobject/issues/47 [3] https://github.com/alsa-project/alsa-gobject/issues/53 [4] https://github.com/alsa-project/alsa-gobject/pull/55 [5] https://developer.gnome.org/gtk-doc-manual/ [6] https://github.com/gtk-rs/gir [7] https://mailman.alsa-project.org/pipermail/alsa-devel/2020-June/169488.html Regards Takashi Sakamoto

1 0

[PATCH] ASoC: SOF: IPC: fix implicit type overflow
by Kai Vehmanen 20 Nov '20

20 Nov '20

From: Curtis Malainey <cujomalainey(a)chromium.org> Implicit values may have a length of 15bits (s16) so we need to declare the proper size so we don't get undefined behaviour. This appears to be arch and compiler dependent. This commit is to keep the headers aligned between the firmware and kernel. UBSan discovered this bug in the firmware. Signed-off-by: Curtis Malainey <cujomalainey(a)chromium.org> Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski(a)linux.intel.com> Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> --- include/sound/sof/header.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/sound/sof/header.h b/include/sound/sof/header.h index 2d35997ace40..571e1dd54b89 100644 --- a/include/sound/sof/header.h +++ b/include/sound/sof/header.h @@ -31,12 +31,12 @@ /* Global Message - Generic */ #define SOF_GLB_TYPE_SHIFT 28 -#define SOF_GLB_TYPE_MASK (0xf << SOF_GLB_TYPE_SHIFT) +#define SOF_GLB_TYPE_MASK (0xfL << SOF_GLB_TYPE_SHIFT) #define SOF_GLB_TYPE(x) ((x) << SOF_GLB_TYPE_SHIFT) /* Command Message - Generic */ #define SOF_CMD_TYPE_SHIFT 16 -#define SOF_CMD_TYPE_MASK (0xfff << SOF_CMD_TYPE_SHIFT) +#define SOF_CMD_TYPE_MASK (0xfffL << SOF_CMD_TYPE_SHIFT) #define SOF_CMD_TYPE(x) ((x) << SOF_CMD_TYPE_SHIFT) /* Global Message Types */ -- 2.28.0

2 1

[PATCH 0/2] ASoC: mt6359: Fix regulator_dev_lookup() fails for id "LDO_VAUD18"
by Jiaxin Yu 20 Nov '20

20 Nov '20

This series fixed "LDO_VAUD18-supply" regulator register fail. We can see the error log "mt6359-sound supply LDO_VAUD18 not found, using dummy regulator" when register the DAPM widget "LDO_VAUD18" Otherwise, the power can not be turned on correctly when recording. Jiaxin Yu (2): ASoC: mediatek: mt6359: Fix regulator_dev_lookup() fails for id "LDO_VAUD18" dt-bindings: mediatek: mt6359: remove unused property for mt6359 Documentation/devicetree/bindings/sound/mt6359.yaml | 9 --------- sound/soc/codecs/mt6359.c | 7 ++++--- 2 files changed, 4 insertions(+), 12 deletions(-) -- 2.18.0

2 3

[PATCH] ASoC: qcom: sm8250: fix HDMI audio playback
by Dmitry Baryshkov 20 Nov '20

20 Nov '20

Current code does not setup CPU dai (causing -EIO errors on playback) and does not pass SND_SOC_DAIFMT_I2S to codec fmt (causing i2s-hifi errors). Fix both errors to enable HDMI audio playback on SM8250. Tested on RB5 platform. Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> Fixes: aa2e2785545a ("ASoC: qcom: sm8250: add sound card qrb5165-rb5 support") Cc: Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> --- sound/soc/qcom/sm8250.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/soc/qcom/sm8250.c b/sound/soc/qcom/sm8250.c index 35c9ad11edff..314e3e566d6d 100644 --- a/sound/soc/qcom/sm8250.c +++ b/sound/soc/qcom/sm8250.c @@ -36,6 +36,7 @@ static int sm8250_be_hw_params_fixup(struct snd_soc_pcm_runtime *rtd, static int sm8250_snd_startup(struct snd_pcm_substream *substream) { + unsigned int fmt = SND_SOC_DAIFMT_CBS_CFS; unsigned int codec_dai_fmt = SND_SOC_DAIFMT_CBS_CFS; struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai *cpu_dai = asoc_rtd_to_cpu(rtd, 0); @@ -43,10 +44,11 @@ static int sm8250_snd_startup(struct snd_pcm_substream *substream) switch (cpu_dai->id) { case TERTIARY_MI2S_RX: - codec_dai_fmt |= SND_SOC_DAIFMT_NB_NF; + codec_dai_fmt |= SND_SOC_DAIFMT_NB_NF | SND_SOC_DAIFMT_I2S; snd_soc_dai_set_sysclk(cpu_dai, Q6AFE_LPASS_CLK_ID_TER_MI2S_IBIT, MI2S_BCLK_RATE, SNDRV_PCM_STREAM_PLAYBACK); + snd_soc_dai_set_fmt(cpu_dai, fmt); snd_soc_dai_set_fmt(codec_dai, codec_dai_fmt); break; default: -- 2.29.2

3 2