- Sound-open-firmware - mailman.alsa-project.org

[Sound-open-firmware] [PATCH v4 0/4] Add MT8186 ADSP dt-binding
by Tinghan Shen 23 Jun '22

23 Jun '22

v3 -> v4: 1. Update commit message of patch 1/4 2. Add review tag to path 3/4 v2 -> v3: 1. Change mbox-names to rx/tx for both mt8186 and mt8195. 2. Update description of mbox-names 3. Use static string array instead of kasprintf 4. Align clock names in dsp driver with dt-bindings v1 -> v2: 1. Change mbox-names from mbox0/mbox1 to rep/req for both mt8186 and mt8195. 2. rename clock-names and remove unused headers Tinghan Shen (4): dt-bindings: dsp: mediatek: Use meaningful names for mbox firmware: mediatek: Use meaningful names for mbox dt-bindings: dsp: mediatek: Add mt8186 dsp document ASoC: SOF: mediatek: Align mt8186 clock names with dt-bindings .../bindings/dsp/mediatek,mt8186-dsp.yaml | 91 +++++++++++++++++++ .../bindings/dsp/mediatek,mt8195-dsp.yaml | 10 +- drivers/firmware/mtk-adsp-ipc.c | 36 +++----- sound/soc/sof/mediatek/mt8186/mt8186-clk.c | 4 +- 4 files changed, 110 insertions(+), 31 deletions(-) create mode 100644 Documentation/devicetree/bindings/dsp/mediatek,mt8186-dsp.yaml -- 2.18.0

3 6

[Sound-open-firmware] ASoC: SOF: Race condition in ipc.c
by noman pouigt 22 Jun '22

22 Jun '22

Folks, I have borrowed part of SOF architecture for my own DSP framework development as the memory on the DSP is extremely small and wouldn't be able to support SOF. Currently I am running into a race condition as below: CPU DSP PCM_TRIGGER_START sof_ipc_send_msg ----> <------Immediate ACK ipc3_wait_tx_done (wait_event_timeout) <------ POSITION update snd_pcm_period_elapsed As you can see TRIGGER_START didn't even finish and waiting for it to complete in ipc3_wait_tx_done function. However, before it could complete the position interrupt was issued which results in calling period_elapsed function. In order to fix this I assume below is called in SOF framework: schedule_work(&spcm->stream[substream->stream].period_elapsed_work); How is this design working? If the interrupt is coming too fast from the DSP than the associated function with this schedule_work will not get called as the scheduler will not get time to schedule the workqueue and elapsed function will not be called thereby not increasing the hw_ptr. How is the flow control for data transfer achieved? I am facing the above problem in my design. I am wondering if I can simply add one more IPC command(don't call wait_event_interruptible for this) after TRIGGER_START to start the streaming.This way schedule_work for updating period_elapsed function can be avoided and it can be called in an interrupt context. Thanks, Variksla

2 2

[Sound-open-firmware] [PATCH v3 0/4] Add MT8186 ADSP dt-binding
by Tinghan Shen 20 Jun '22

20 Jun '22

v2 -> v3: 1. Change mbox-names to rx/tx for both mt8186 and mt8195. 2. Update description of mbox-names 3. Use static string array instead of kasprintf 4. Align clock names in dsp driver with dt-bindings v1 -> v2: 1. Change mbox-names from mbox0/mbox1 to rep/req for both mt8186 and mt8195. 2. rename clock-names and remove unused headers Tinghan Shen (4): dt-bindings: dsp: mediatek: Use meaningful names for mbox firmware: mediatek: Use meaningful names for mbox dt-bindings: dsp: mediatek: Add mt8186 dsp document ASoC: SOF: mediatek: Align mt8186 clock names with dt-bindings .../bindings/dsp/mediatek,mt8186-dsp.yaml | 91 +++++++++++++++++++ .../bindings/dsp/mediatek,mt8195-dsp.yaml | 10 +- drivers/firmware/mtk-adsp-ipc.c | 36 +++----- sound/soc/sof/mediatek/mt8186/mt8186-clk.c | 4 +- 4 files changed, 110 insertions(+), 31 deletions(-) create mode 100644 Documentation/devicetree/bindings/dsp/mediatek,mt8186-dsp.yaml -- 2.18.0

2 9

[Sound-open-firmware] [ANNOUNCE] SOF v2.2-rc1
by Girdwood, Liam R 17 Jun '22

17 Jun '22

The first release candidate for SOF v2.2 is now available for download. This release moves SOF closer to fully using native Zephyr RTOS and also includes performance and stability improvements over v2.1. Release notes and binaries available here. https://github.com/thesofproject/sof/releases/tag/v2.2-rc1 Final v2.2 due before end of June. Thanks Liam

1 0

[Sound-open-firmware] [PATCH] ASoC: SOF: ipc4-topology: Fix error code in sof_ipc4_volume_put()
by Dan Carpenter 16 Jun '22

16 Jun '22

The sof_ipc4_volume_put() function returns type bool so returning -ENOENT means returning true. Return false instead. Fixes: 955e84fc0b6d ("ASoC: SOF: ipc4-topology: Add control IO ops") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> --- sound/soc/sof/ipc4-control.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/ipc4-control.c b/sound/soc/sof/ipc4-control.c index 95ee121dd3cf..0d5a578c3496 100644 --- a/sound/soc/sof/ipc4-control.c +++ b/sound/soc/sof/ipc4-control.c @@ -142,7 +142,7 @@ static bool sof_ipc4_volume_put(struct snd_sof_control *scontrol, if (!widget_found) { dev_err(scomp->dev, "Failed to find widget for kcontrol %s\n", scontrol->name); - return -ENOENT; + return false; } ret = sof_ipc4_set_volume_data(sdev, swidget, scontrol); -- 2.35.1

3 2

[Sound-open-firmware] [PATCH] ASoC: SOF: mediatek: Fix error code in probe
by Dan Carpenter 15 Jun '22

15 Jun '22

This should return PTR_ERR() instead of IS_ERR(). Fixes: e0100bfd383c ("ASoC: SOF: mediatek: Add mt8186 ipc support") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> --- sound/soc/sof/mediatek/mt8186/mt8186.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/mediatek/mt8186/mt8186.c b/sound/soc/sof/mediatek/mt8186/mt8186.c index 3333a0634e29..e006532caf2f 100644 --- a/sound/soc/sof/mediatek/mt8186/mt8186.c +++ b/sound/soc/sof/mediatek/mt8186/mt8186.c @@ -392,7 +392,7 @@ static int mt8186_dsp_probe(struct snd_sof_dev *sdev) PLATFORM_DEVID_NONE, pdev, sizeof(*pdev)); if (IS_ERR(priv->ipc_dev)) { - ret = IS_ERR(priv->ipc_dev); + ret = PTR_ERR(priv->ipc_dev); dev_err(sdev->dev, "failed to create mtk-adsp-ipc device\n"); goto err_adsp_off; } -- 2.35.1

5 4

[Sound-open-firmware] [PATCH 1/2] ASoC: SOF: ipc-msg-injector: Propagate write errors correctly
by Dan Carpenter 07 Jun '22

07 Jun '22

This code is supposed to propagate errors from simple_write_to_buffer() or return -EFAULT if "size != count". However "size" needs to be signed for the code to work correctly and the case where "size == 0" is not handled correctly. Fixes: 066c67624d8c ("ASoC: SOF: ipc-msg-injector: Add support for IPC4 messages") Fixes: 2f0b1b013bbc ("ASoC: SOF: debug: Add support for IPC message injection") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> --- sound/soc/sof/sof-client-ipc-msg-injector.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/sound/soc/sof/sof-client-ipc-msg-injector.c b/sound/soc/sof/sof-client-ipc-msg-injector.c index 03490a4d4ae7..030cb97d7713 100644 --- a/sound/soc/sof/sof-client-ipc-msg-injector.c +++ b/sound/soc/sof/sof-client-ipc-msg-injector.c @@ -150,7 +150,7 @@ static ssize_t sof_msg_inject_dfs_write(struct file *file, const char __user *bu { struct sof_client_dev *cdev = file->private_data; struct sof_msg_inject_priv *priv = cdev->data; - size_t size; + ssize_t size; int ret; if (*ppos) @@ -158,8 +158,10 @@ static ssize_t sof_msg_inject_dfs_write(struct file *file, const char __user *bu size = simple_write_to_buffer(priv->tx_buffer, priv->max_msg_size, ppos, buffer, count); + if (size < 0) + return size; if (size != count) - return size > 0 ? -EFAULT : size; + return -EFAULT; memset(priv->rx_buffer, 0, priv->max_msg_size); @@ -179,7 +181,7 @@ static ssize_t sof_msg_inject_ipc4_dfs_write(struct file *file, struct sof_client_dev *cdev = file->private_data; struct sof_msg_inject_priv *priv = cdev->data; struct sof_ipc4_msg *ipc4_msg = priv->tx_buffer; - size_t size; + ssize_t size; int ret; if (*ppos) @@ -192,8 +194,10 @@ static ssize_t sof_msg_inject_ipc4_dfs_write(struct file *file, size = simple_write_to_buffer(&ipc4_msg->header_u64, sizeof(ipc4_msg->header_u64), ppos, buffer, count); + if (size < 0) + return size; if (size != sizeof(ipc4_msg->header_u64)) - return size > 0 ? -EFAULT : size; + return -EFAULT; count -= size; if (!count) { @@ -201,8 +205,10 @@ static ssize_t sof_msg_inject_ipc4_dfs_write(struct file *file, size = simple_write_to_buffer(ipc4_msg->data_ptr, priv->max_msg_size, ppos, buffer, count); + if (size < 0) + return size; if (size != count) - return size > 0 ? -EFAULT : size; + return -EFAULT; } ipc4_msg->data_size = count; -- 2.35.1

3 4

[Sound-open-firmware] [PATCH] ASoC: SOF: Fix potential NULL pointer dereference
by Amadeusz Sławiński 07 Jun '22

07 Jun '22

Cleanup path for sof_prepare_widgets_in_path() should check if unprepare callback exists before calling it, instead it checks if it does not exist. Fix the check. Fixes: 370014c8197a ("ASoC: SOF: Add a prepare op to IPC topology widget ops") Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski(a)linux.intel.com> --- sound/soc/sof/sof-audio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/sof-audio.c b/sound/soc/sof/sof-audio.c index cd41d0599cef..789d0a1f1a77 100644 --- a/sound/soc/sof/sof-audio.c +++ b/sound/soc/sof/sof-audio.c @@ -323,7 +323,7 @@ sof_prepare_widgets_in_path(struct snd_sof_dev *sdev, struct snd_soc_dapm_widget p->walking = false; if (ret < 0) { /* unprepare the source widget */ - if (!widget_ops[widget->id].ipc_unprepare && swidget->prepared) { + if (widget_ops[widget->id].ipc_unprepare && swidget->prepared) { widget_ops[widget->id].ipc_unprepare(swidget); swidget->prepared = false; } -- 2.25.1

3 2

[Sound-open-firmware] [PATCH v2] ASoC: SOF: Fix potential NULL pointer dereference
by Amadeusz Sławiński 06 Jun '22

06 Jun '22

Cleanup path for sof_prepare_widgets_in_path() should check if unprepare callback exists before calling it, instead it checks if it does not exist. Fix the check. Fixes: 66344c6d9211 ("ASoC: SOF: Add a prepare op to IPC topology widget ops") Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> --- v2 : Fix Fixes hash, apparently I looked at wrong tree --- sound/soc/sof/sof-audio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/sof-audio.c b/sound/soc/sof/sof-audio.c index 8d740635a4bb..28976098a89e 100644 --- a/sound/soc/sof/sof-audio.c +++ b/sound/soc/sof/sof-audio.c @@ -318,7 +318,7 @@ sof_prepare_widgets_in_path(struct snd_sof_dev *sdev, struct snd_soc_dapm_widget p->walking = false; if (ret < 0) { /* unprepare the source widget */ - if (!widget_ops[widget->id].ipc_unprepare && swidget->prepared) { + if (widget_ops[widget->id].ipc_unprepare && swidget->prepared) { widget_ops[widget->id].ipc_unprepare(swidget); swidget->prepared = false; } -- 2.25.1

1 0

Re: [Sound-open-firmware] [PATCH] ASoC: SOF: Fix potential NULL pointer dereference
by Pierre-Louis Bossart 06 Jun '22

06 Jun '22

On 6/6/22 06:31, Mark Brown wrote: > On Thu, Jun 02, 2022 at 03:57:57PM +0200, Amadeusz Sławiński wrote: >> Cleanup path for sof_prepare_widgets_in_path() should check if unprepare >> callback exists before calling it, instead it checks if it does not >> exist. Fix the check. >> >> Fixes: 370014c8197a ("ASoC: SOF: Add a prepare op to IPC topology widget ops") > > This commit does not exist. Indeed, this SHA was for the SOF tree, the upstream version is this: 66344c6d92113 ("ASoC: SOF: Add a prepare op to IPC topology widget ops")

1 0