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

[PATCH v3 0/5] Tegra234 APE support
by Sameer Pujar 17 Feb '22

17 Feb '22

This adds APE support on Tegra234 and enables sound card on Jetson AGX Orin platform. Changelog ========= v2->v3 ------ * Fixed DT build error related to power-domain and iommu reference. Reported-by: kernel test robot <lkp(a)intel.com> v1->v2 ------ * Rebased patches on latest linux-next Mohan Kumar (1): ASoC: tegra: Update AHUB driver for Tegra234 Sameer Pujar (4): dt-bindings: Add Tegra234 APE support dt-bindings: Document Tegra234 APE support arm64: tegra: Add audio devices on Tegra234 arm64: tegra: APE sound card for Jetson AGX Orin .../bindings/bus/nvidia,tegra210-aconnect.yaml | 1 + .../bindings/dma/nvidia,tegra210-adma.yaml | 4 +- .../bindings/interrupt-controller/arm,gic.yaml | 1 + .../bindings/sound/nvidia,tegra186-dspk.yaml | 4 +- .../bindings/sound/nvidia,tegra210-admaif.yaml | 4 +- .../bindings/sound/nvidia,tegra210-adx.yaml | 1 + .../bindings/sound/nvidia,tegra210-ahub.yaml | 1 + .../bindings/sound/nvidia,tegra210-amx.yaml | 3 + .../bindings/sound/nvidia,tegra210-dmic.yaml | 1 + .../bindings/sound/nvidia,tegra210-i2s.yaml | 1 + .../bindings/sound/nvidia,tegra210-mixer.yaml | 1 + .../bindings/sound/nvidia,tegra210-mvc.yaml | 1 + .../bindings/sound/nvidia,tegra210-sfc.yaml | 1 + .../dts/nvidia/tegra234-p3737-0000+p3701-0000.dts | 1783 ++++++++++++++++++++ arch/arm64/boot/dts/nvidia/tegra234.dtsi | 432 +++++ include/dt-bindings/clock/tegra234-clock.h | 74 +- include/dt-bindings/memory/tegra234-mc.h | 7 + include/dt-bindings/power/tegra234-powergate.h | 9 + sound/soc/tegra/tegra210_ahub.c | 146 +- sound/soc/tegra/tegra210_ahub.h | 4 +- 20 files changed, 2473 insertions(+), 6 deletions(-) create mode 100644 include/dt-bindings/power/tegra234-powergate.h -- 2.7.4

3 8

[PATCH v17 0/2] mtk: add ADSP mailbox controller for MT8195
by allen-kh.cheng 15 Feb '22

15 Feb '22

From: Allen-KH Cheng <Allen-KH.Cheng(a)mediatek.com> Mediatek ADSP IPC is used to send notification or short message between processors with dsp. It will place the message to the share buffer and will access the ADSP mailbox registers to kick dsp. Two mailboxes used to send notification or short message between processors with dsp. This patchset was tested and confirmed on MT8195 cherry board. Based on linux-next changes since v16: - fix WARNING: modpost: missing MODULE_LICENSE() in drivers/mailbox /mtk-adsp-mailbox.o. Add MODULE_LICENSE in the last line. - Due to WARNING: Missing or malformed SPDX-License-Identifier tag in line 1 in checkpatch, we don't remove SPDX-License in line 1. changes since v15: - add paragraphs to describe the config symbols fully changes since v14: - add inline for get_mtk_adsp_mbox_priv changes since v13: - rebase on v5.16-rc8 - add back ptr check from of_device_get_match_data changes since v12: - remove of_device_get_match_data ptr check: of_device_get_match_data(dev) will never going to return NULL. driver probe with compatible mediatek,mt8195-adsp-mbox. changes since v11: - remove useless MODULE_LICENSE changes since v10: - split up v9 into two separate submissions changes since v9: - rename adsp_mbox_chan_ops to mtk_adsp_mbox_chan_ops changes since v8: - remove struct adsp_mbox_ch_info - add get_mtk_adsp_mbox_priv - use mtk_adsp_mbox_priv va_mboxreg address in adsp mbox driver - add struct mtk_adsp_mbox_cfg for DSP mbox register offset - remove adsp mbox register offset hard code define - remove mtk-adsp-ipc.h reference in adsp mbox driver changes since v7: - add mtk prefix for adsp ipc functions - rename adsp_mbox_ch_info to mtk_adsp_mbox_ch_info - remove incorrect reviewers in commit message changes since v6: - dt-bindings: change example dtsi node. - rename config MTK_ADSP_IPC_MBOX to MTK_ADSP_MBOX - remove unused variable - add reviewers changes since v5: - remove some redundant code changes since v4: - use switch ... case in adsp_ipc_recv - add error handling path for chan_name pointer - refine some code to be concise changes since v3: - reorder MTK_ADSP_IPC_MBOX config - remove some redundant code - remove lock in mtk-adsp-mailbox changes since v2: - separate adsp_mailbox into two instances changes since v1: - fix dt_binding_check error Allen-KH Cheng (2): dt-bindings: mailbox: add mtk adsp-mbox document mailbox: mediatek: add support for adsp mailbox controller .../bindings/mailbox/mtk,adsp-mbox.yaml | 52 ++++++ drivers/mailbox/Kconfig | 9 + drivers/mailbox/Makefile | 2 + drivers/mailbox/mtk-adsp-mailbox.c | 176 ++++++++++++++++++ 4 files changed, 239 insertions(+) create mode 100644 Documentation/devicetree/bindings/mailbox/mtk,adsp-mbox.yaml create mode 100644 drivers/mailbox/mtk-adsp-mailbox.c -- 2.18.0

2 4

[PATCH] ALSA: pcm: accept the OPEN state for snd_pcm_stop()
by Jaroslav Kysela 14 Feb '22

14 Feb '22

It may be useful to completely invalidate streaming under some circumstances like the USB gadget detach. This case is a bit different than the complete disconnection. The applications should be notified that the PCM streaming is no longer available, but the recovery may be expected. This patch adds support for SNDRV_PCM_STATE_OPEN passed to snd_pcm_stop() function. Only the hw_free ioctl is allowed to free the buffers in this state for a full recovery operation or the PCM file handle must be closed. In the OPEN state, ioctls return EBADFD error (with the added hw_free exception above). The applications which are not aware about this new state transition (and recovery) will fail with an error. This operation is expected. Link: https://lore.kernel.org/alsa-devel/4e17c99b-1d8a-8ebe-972c-9b1299952ff8@ivi… Cc: Pavel Hofman <pavel.hofman(a)ivitera.com> Signed-off-by: Jaroslav Kysela <perex(a)perex.cz> --- include/sound/pcm.h | 1 + sound/core/pcm.c | 1 + sound/core/pcm_native.c | 12 +++++++++++- 3 files changed, 13 insertions(+), 1 deletion(-) diff --git a/include/sound/pcm.h b/include/sound/pcm.h index 33451f8ff755..4de1c2c91109 100644 --- a/include/sound/pcm.h +++ b/include/sound/pcm.h @@ -467,6 +467,7 @@ struct snd_pcm_substream { /* -- assigned files -- */ int ref_count; atomic_t mmap_count; + atomic_t queued_hw_free; unsigned int f_flags; void (*pcm_release)(struct snd_pcm_substream *); struct pid *pid; diff --git a/sound/core/pcm.c b/sound/core/pcm.c index 6fd3677685d7..8dc7e99b2113 100644 --- a/sound/core/pcm.c +++ b/sound/core/pcm.c @@ -694,6 +694,7 @@ int snd_pcm_new_stream(struct snd_pcm *pcm, int stream, int substream_count) snd_pcm_group_init(&substream->self_group); list_add_tail(&substream->link_list, &substream->self_group.substreams); atomic_set(&substream->mmap_count, 0); + atomic_set(&substream->queued_hw_free, 0); prev = substream; } return 0; diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c index 621883e71194..118ad3f26f4a 100644 --- a/sound/core/pcm_native.c +++ b/sound/core/pcm_native.c @@ -686,10 +686,13 @@ static int snd_pcm_hw_params(struct snd_pcm_substream *substream, snd_pcm_stream_lock_irq(substream); switch (runtime->status->state) { case SNDRV_PCM_STATE_OPEN: + if (atomic_read(&substream->queued_hw_free)) + goto __badfd; case SNDRV_PCM_STATE_SETUP: case SNDRV_PCM_STATE_PREPARED: break; default: +__badfd: snd_pcm_stream_unlock_irq(substream); return -EBADFD; } @@ -829,6 +832,7 @@ static int do_hw_free(struct snd_pcm_substream *substream) result = substream->ops->hw_free(substream); if (substream->managed_buffer_alloc) snd_pcm_lib_free_pages(substream); + atomic_set(&substream->queued_hw_free, 0); return result; } @@ -1454,6 +1458,8 @@ static void snd_pcm_post_stop(struct snd_pcm_substream *substream, } wake_up(&runtime->sleep); wake_up(&runtime->tsleep); + if (state == SNDRV_PCM_STATE_OPEN) + atomic_set(&substream->queued_hw_free, 1); } static const struct action_ops snd_pcm_action_stop = { @@ -1469,6 +1475,9 @@ static const struct action_ops snd_pcm_action_stop = { * * The state of each stream is then changed to the given state unconditionally. * + * If the requested state is OPEN, the stream is invalidated and + * the application must call hw_free to recover the operation. + * * Return: Zero if successful, or a negative error code. */ int snd_pcm_stop(struct snd_pcm_substream *substream, snd_pcm_state_t state) @@ -2637,7 +2646,8 @@ void snd_pcm_release_substream(struct snd_pcm_substream *substream) snd_pcm_drop(substream); if (substream->hw_opened) { - if (substream->runtime->status->state != SNDRV_PCM_STATE_OPEN) + if (substream->runtime->status->state != SNDRV_PCM_STATE_OPEN || + atomic_read(&substream->queued_hw_free)) do_hw_free(substream); substream->ops->close(substream); substream->hw_opened = 0; -- 2.33.1

5 7

[PATCH 00/19] soundwire: stream: cleanup of 'stream' support
by Bard Liao 11 Feb '22

11 Feb '22

This series revisits the SoundWire 'sdw_stream' support to split allocation and configuration steps. This is necessary if for example the routines are called multiple times from the hw_params stage. This also helps with better error handling. Pierre-Louis Bossart (19): soundwire: stream: remove unused parameter in sdw_stream_add_slave soundwire: stream: add slave runtime to list earlier soundwire: stream: simplify check on port range soundwire: stream: add alloc/config/free helpers for ports soundwire: stream: split port allocation and configuration loops soundwire: stream: split alloc and config in two functions soundwire: stream: add 'slave' prefix for port range checks soundwire: stream: group sdw_port and sdw_master/slave_port functions soundwire: stream: simplify sdw_alloc_master_rt() soundwire: stream: split sdw_alloc_master_rt() in alloc and config soundwire: stream: move sdw_alloc_slave_rt() before 'master' helpers soundwire: stream: split sdw_alloc_slave_rt() in alloc and config soundwire: stream: group sdw_stream_ functions soundwire: stream: rename and move master/slave_rt_free routines soundwire: stream: move list addition to sdw_slave_alloc_rt() soundwire: stream: separate alloc and config within sdw_stream_add_xxx() soundwire: stream: introduce sdw_slave_rt_find() helper soundwire: stream: sdw_stream_add_ functions can be called multiple times soundwire: stream: make enable/disable/deprepare idempotent drivers/soundwire/stream.c | 960 +++++++++++++++++++++---------------- 1 file changed, 547 insertions(+), 413 deletions(-) -- 2.17.1

2 20

[PATCH v2 0/6] ASoC: dt-bindings: samsung: convert to dtschema
by Krzysztof Kozlowski 11 Feb '22

11 Feb '22

Hi, The patchset is based on Rob's sound-dai changes: https://lore.kernel.org/all/20220126231427.1638089-1-robh@kernel.org/ Changes since v1: 1. Correct samsung,snow cpu/sound-dai. Best regards, Krzysztof Krzysztof Kozlowski (6): ASoC: dt-bindings: samsung,aries-wm8994: require sound-dai property ASoC: dt-bindings: samsung,arndale: convert to dtschema ASoC: dt-bindings: samsung,arndale: document ALC5631 ASoC: dt-bindings: samsung,smdk5250: convert to dtschema ASoC: dt-bindings: samsung,snow: convert to dtschema ASoC: dt-bindings: samsung,tm2: convert to dtschema .../devicetree/bindings/sound/arndale.txt | 25 ------ .../bindings/sound/samsung,aries-wm8994.yaml | 4 + .../bindings/sound/samsung,arndale.yaml | 45 +++++++++++ .../bindings/sound/samsung,smdk-wm8994.txt | 14 ---- .../bindings/sound/samsung,smdk5250.yaml | 38 +++++++++ .../bindings/sound/samsung,snow.yaml | 74 +++++++++++++++++ .../bindings/sound/samsung,tm2-audio.txt | 42 ---------- .../bindings/sound/samsung,tm2.yaml | 80 +++++++++++++++++++ .../devicetree/bindings/sound/snow.txt | 31 ------- 9 files changed, 241 insertions(+), 112 deletions(-) delete mode 100644 Documentation/devicetree/bindings/sound/arndale.txt create mode 100644 Documentation/devicetree/bindings/sound/samsung,arndale.yaml delete mode 100644 Documentation/devicetree/bindings/sound/samsung,smdk-wm8994.txt create mode 100644 Documentation/devicetree/bindings/sound/samsung,smdk5250.yaml create mode 100644 Documentation/devicetree/bindings/sound/samsung,snow.yaml delete mode 100644 Documentation/devicetree/bindings/sound/samsung,tm2-audio.txt create mode 100644 Documentation/devicetree/bindings/sound/samsung,tm2.yaml delete mode 100644 Documentation/devicetree/bindings/sound/snow.txt -- 2.32.0

3 13

[PATCH] soundwire: Replace acpi_bus_get_device()
by Rafael J. Wysocki 10 Feb '22

10 Feb '22

From: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Replace acpi_bus_get_device() that is going to be dropped with acpi_fetch_acpi_dev(). No intentional functional impact. Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> --- drivers/soundwire/intel_init.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) Index: linux-pm/drivers/soundwire/intel_init.c =================================================================== --- linux-pm.orig/drivers/soundwire/intel_init.c +++ linux-pm/drivers/soundwire/intel_init.c @@ -180,7 +180,8 @@ static struct sdw_intel_ctx if (!res) return NULL; - if (acpi_bus_get_device(res->handle, &adev)) + adev = acpi_fetch_acpi_dev(res->handle); + if (!adev) return NULL; if (!res->count) @@ -294,13 +295,13 @@ err: static int sdw_intel_startup_controller(struct sdw_intel_ctx *ctx) { - struct acpi_device *adev; + struct acpi_device *adev = acpi_fetch_acpi_dev(ctx->handle); struct sdw_intel_link_dev *ldev; u32 caps; u32 link_mask; int i; - if (acpi_bus_get_device(ctx->handle, &adev)) + if (!adev) return -EINVAL; /* Check SNDWLCAP.LCOUNT */

2 1

[PATCH] soundwire: bus: add dev_warn() messages to track UNATTACHED devices
by Bard Liao 10 Feb '22

10 Feb '22

From: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> In rare cases, some devices seem to lose sync and never re-attach on the bus. This seems to happen only when there are more than one device per link, which suggests either an electrical issue, a race condition or a state machine issue. Add two dev_warn() messages to identify the sequence by which the devices become UNATTACHED. BugLink: https://github.com/thesofproject/linux/issues/3063 BugLink: https://github.com/thesofproject/linux/issues/3325 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Rander Wang <rander.wang(a)intel.com> Signed-off-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> --- drivers/soundwire/bus.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/soundwire/bus.c b/drivers/soundwire/bus.c index 67369e941d0d..354d3f89366f 100644 --- a/drivers/soundwire/bus.c +++ b/drivers/soundwire/bus.c @@ -1749,8 +1749,11 @@ int sdw_handle_slave_status(struct sdw_bus *bus, continue; if (status[i] == SDW_SLAVE_UNATTACHED && - slave->status != SDW_SLAVE_UNATTACHED) + slave->status != SDW_SLAVE_UNATTACHED) { + dev_warn(&slave->dev, "Slave %d state check1: UNATTACHED, status was %d\n", + i, slave->status); sdw_modify_slave_status(slave, SDW_SLAVE_UNATTACHED); + } } if (status[0] == SDW_SLAVE_ATTACHED) { @@ -1785,6 +1788,9 @@ int sdw_handle_slave_status(struct sdw_bus *bus, if (slave->status == SDW_SLAVE_UNATTACHED) break; + dev_warn(&slave->dev, "Slave %d state check2: UNATTACHED, status was %d\n", + i, slave->status); + sdw_modify_slave_status(slave, SDW_SLAVE_UNATTACHED); break; -- 2.17.1

2 1

[PATCH] soundwire: intel: fix wrong register name in intel_shim_wake
by Bard Liao 10 Feb '22

10 Feb '22

From: Libin Yang <libin.yang(a)intel.com> When clearing the sdw wakests status, we should use SDW_SHIM_WAKESTS. Fixes: 4a17c441c7cb ("soundwire: intel: revisit SHIM programming sequences.") Signed-off-by: Libin Yang <libin.yang(a)intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Rander Wang <rander.wang(a)intel.com> Signed-off-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> --- drivers/soundwire/intel.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/soundwire/intel.c b/drivers/soundwire/intel.c index 122f7a29d8ca..63101f1ba271 100644 --- a/drivers/soundwire/intel.c +++ b/drivers/soundwire/intel.c @@ -448,8 +448,8 @@ static void intel_shim_wake(struct sdw_intel *sdw, bool wake_enable) /* Clear wake status */ wake_sts = intel_readw(shim, SDW_SHIM_WAKESTS); - wake_sts |= (SDW_SHIM_WAKEEN_ENABLE << link_id); - intel_writew(shim, SDW_SHIM_WAKESTS_STATUS, wake_sts); + wake_sts |= (SDW_SHIM_WAKESTS_STATUS << link_id); + intel_writew(shim, SDW_SHIM_WAKESTS, wake_sts); } mutex_unlock(sdw->link_res->shim_lock); } -- 2.17.1

2 1

[PATCH 1/4] ALSA: hda/realtek: fix mute/micmute LEDs for HP 855 G8
by Jeremy Szu 09 Feb '22

09 Feb '22

The HP EliteBook 855 G8 Notebook PC is using ALC285 codec which needs ALC285_FIXUP_HP_MUTE_LED fixup to make it works. After applying the fixup, the mute/micmute LEDs work good. Signed-off-by: Jeremy Szu <jeremy.szu(a)canonical.com> --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 552e2cb73291..9d68f591c6bf 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -8291,6 +8291,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x103c, 0x87f7, "HP Spectre x360 14", ALC245_FIXUP_HP_X360_AMP), SND_PCI_QUIRK(0x103c, 0x8846, "HP EliteBook 850 G8 Notebook PC", ALC285_FIXUP_HP_GPIO_LED), SND_PCI_QUIRK(0x103c, 0x884c, "HP EliteBook 840 G8 Notebook PC", ALC285_FIXUP_HP_GPIO_LED), + SND_PCI_QUIRK(0x103c, 0x8896, "HP EliteBook 855 G8 Notebook PC", ALC285_FIXUP_HP_MUTE_LED), SND_PCI_QUIRK(0x103c, 0x8898, "HP EliteBook 845 G8 Notebook PC", ALC285_FIXUP_HP_LIMIT_INT_MIC_BOOST), SND_PCI_QUIRK(0x1043, 0x103e, "ASUS X540SA", ALC256_FIXUP_ASUS_MIC), SND_PCI_QUIRK(0x1043, 0x103f, "ASUS TX300", ALC282_FIXUP_ASUS_TX300), -- 2.31.1

3 32

[PATCH] ASoC: fsl_sai: Enable combine mode soft
by Sascha Hauer 09 Feb '22

09 Feb '22

The fsl_sai driver calculates the number of pins used and enables multiple channels if necessary. This means the SAI expects data in one FIFO per pin. The SDMA engine only services a single FIFO, so multi pin support doesn't work at all. This patch enables the software combine mode in chips that support it. With this the SAI presents only a single FIFO to the outside and distributes the data into the different FIFOs internally. Signed-off-by: Sascha Hauer <s.hauer(a)pengutronix.de> --- sound/soc/fsl/fsl_sai.c | 9 +++++++++ sound/soc/fsl/fsl_sai.h | 1 + 2 files changed, 10 insertions(+) diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c index 10544fa27dc0..cab015d96889 100644 --- a/sound/soc/fsl/fsl_sai.c +++ b/sound/soc/fsl/fsl_sai.c @@ -517,6 +517,10 @@ static int fsl_sai_hw_params(struct snd_pcm_substream *substream, FSL_SAI_CR5_FBT_MASK, val_cr5); } + if (sai->soc_data->pins > 1) + regmap_update_bits(sai->regmap, FSL_SAI_xCR4(tx, ofs), + FSL_SAI_CR4_FCOMB_MASK, FSL_SAI_CR4_FCOMB_SOFT); + regmap_update_bits(sai->regmap, FSL_SAI_xCR3(tx, ofs), FSL_SAI_CR3_TRCE_MASK, FSL_SAI_CR3_TRCE((1 << pins) - 1)); @@ -1195,6 +1199,7 @@ static const struct fsl_sai_soc_data fsl_sai_vf610_data = { .use_imx_pcm = false, .use_edma = false, .fifo_depth = 32, + .pins = 1, .reg_offset = 0, .mclk0_is_mclk1 = false, .flags = 0, @@ -1204,6 +1209,7 @@ static const struct fsl_sai_soc_data fsl_sai_imx6sx_data = { .use_imx_pcm = true, .use_edma = false, .fifo_depth = 32, + .pins = 1, .reg_offset = 0, .mclk0_is_mclk1 = true, .flags = 0, @@ -1213,6 +1219,7 @@ static const struct fsl_sai_soc_data fsl_sai_imx7ulp_data = { .use_imx_pcm = true, .use_edma = false, .fifo_depth = 16, + .pins = 2, .reg_offset = 8, .mclk0_is_mclk1 = false, .flags = PMQOS_CPU_LATENCY, @@ -1222,6 +1229,7 @@ static const struct fsl_sai_soc_data fsl_sai_imx8mq_data = { .use_imx_pcm = true, .use_edma = false, .fifo_depth = 128, + .pins = 8, .reg_offset = 8, .mclk0_is_mclk1 = false, .flags = 0, @@ -1231,6 +1239,7 @@ static const struct fsl_sai_soc_data fsl_sai_imx8qm_data = { .use_imx_pcm = true, .use_edma = true, .fifo_depth = 64, + .pins = 1, .reg_offset = 0, .mclk0_is_mclk1 = false, .flags = 0, diff --git a/sound/soc/fsl/fsl_sai.h b/sound/soc/fsl/fsl_sai.h index 9aaf231bc024..410f6e6a9137 100644 --- a/sound/soc/fsl/fsl_sai.h +++ b/sound/soc/fsl/fsl_sai.h @@ -223,6 +223,7 @@ struct fsl_sai_soc_data { bool use_edma; bool mclk0_is_mclk1; unsigned int fifo_depth; + unsigned int pins; unsigned int reg_offset; unsigned int flags; }; -- 2.30.2

2 2