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

[PATCH 1/3] ASoC: Intel: atom: Take the drv->lock mutex before calling sst_send_slot_map()
by Hans de Goede 03 Apr '20

03 Apr '20

sst_send_slot_map() uses sst_fill_and_send_cmd_unlocked() because in some places it is called with the drv->lock mutex already held. So it must always be called with the mutex locked. This commit adds missing locking in the sst_set_be_modules() code-path. Fixes: 24c8d14192cc ("ASoC: Intel: mrfld: add DSP core controls") Signed-off-by: Hans de Goede <hdegoede(a)redhat.com> --- sound/soc/intel/atom/sst-atom-controls.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/soc/intel/atom/sst-atom-controls.c b/sound/soc/intel/atom/sst-atom-controls.c index baef461a99f1..2c3798034b1d 100644 --- a/sound/soc/intel/atom/sst-atom-controls.c +++ b/sound/soc/intel/atom/sst-atom-controls.c @@ -966,7 +966,9 @@ static int sst_set_be_modules(struct snd_soc_dapm_widget *w, dev_dbg(c->dev, "Enter: widget=%s\n", w->name); if (SND_SOC_DAPM_EVENT_ON(event)) { + mutex_lock(&drv->lock); ret = sst_send_slot_map(drv); + mutex_unlock(&drv->lock); if (ret) return ret; ret = sst_send_pipe_module_params(w, k); -- 2.26.0

3 6

alsa-gobject: add library to operate ALSA Sequencer character device without queue feature
by GitHub pull_request - edited 03 Apr '20

03 Apr '20

alsa-project/alsa-gobject pull request #10 was edited from takaswie: This request is to add a new library, libalsaseq, to operate ALSA Sequencer character device. In design of ALSA Sequencer, one character device is added to Linux system. Usually a special file, '/dev/snd/seq', is used for the character device. Userspace applications can operate ALSA Sequencer via the character device. The library supports functionalities of ALSA Sequencer but without queue-related ones. The library includes some GObject-derived object. ALSASeq.UserClient maintains file descriptor of ALSA Sequencer character device, thus it's the center of user operation. The other objects represent data structures in UAPI of ALSA Sequencer. * ALSASeqSystemInfo - struct snd_seq_system_info * ALSASeqClientInfo - struct snd_seq_client_info * ALSASeqClientPool - struct snd_seq_client_pool * ALSASeqPortInfo - struct snd_seq_port_info * ALSASeqSubscribeData - struct snd_seq_port_subscribe The library includes some GObject boxed object just to represent data structures. ALSASeq.Addr is a thin wrapper to struct snd_seq_addr, and ALSASeq.Tstamp is a thin wrapper to struct snd_seq_timestamp. For operations without instances of ALSASeq.UserClient, the library has some global APIs. ALSA Sequencer has some macros. They're represented as GObject enumerations and flags. * ALSASeq.ClientType - one of NO_CLIENT, USER_CLIENT, KERNEL_CLIENT * ALSASeq.FilterAttrFlag - SNDRV_SEQ_FILTER_XXX * ALSASeq.PortCapFlag - SNDRV_PORT_CAP_XXX * ALSASeq.PortAttrFlag - NDRV_SEQ_PORT_TYPE_XXX * ALSASeq.EventType - SNDRV_SEQ_EVENT_XXX * ALSASeq.EventTimestampMode - SNDRV_SEQ_TIME_STAMP_XXX * ALSASeq.EventTimeMode - SNDRV_SEQ_TIME_MODE_XXX * ALSASeq.EventLengthMode - SNDRV_SEQ_EVENT_LENGTH_XXX * ALSASeq.EventPriorityMode - SNDRV_SEQ_PRIORITY_XXX * ALSASeq.PortSubscribeFlag - SNDRV_SEQ_PORT_SUBS_XXX * ALSASeq.QuerySubscribeType - SNDRV_SEQ_QUERY_SUBS_XXX Some specific values for addressing are also represented by GObject enumerations and flags. * ALSASeq.SpecificAddress - SNDRV_SEQ_ADDRESS_XXX * ALSASeq.SpecificClientId - SNDRV_SEQ_CLIENT_XXX * ALSASeq.SpecificPortId - SNDRV_SEQ_PORT_SYSTEM_XXX * ALSASeq.SpecificQueueId - SNDRV_SEQ_QUEUE_XXX ALSA Sequencer has many types of event. The library has an GObject abstract object, ALSASeq.Event. Two objects are derived from it: * ALSASeq.EventFixed * ALSASeq.EventVariable The former has some ways to access to the data of event. The data is represented by GObject boxed object if it's complicated structure: * ALSASeq.EventDataConnect - struct snd_seq_connect * ALSASeq.EventDataNote - struct snd_seq_ev_note * ALSASeq.EventDataCtl - struct snd_seq_ev_ctrl * ALSASeq.EventDataResult - struct snd_seq_result * ALSASeq.EventDataQueue - struct snd_seq_ev_queue_control The latter is for events to have variable length of data blob. ALSASeq.UserClient has 'schedule_event' API to schedule arbitrary event. It also has 'handle-event' GObject signal for applications to handle asynchronous event as an instance of ALSASeq.EventFixed or ALSASeq.EventVariable. To handle the GObject signal, applications should get GSource from the instance and process GMainContext, typically by call of GMainLoop. The library doesn't have support for below structures since they're useless practically: * struct snd_seq_event_bounce * struct snd_seq_running_info In my later work, queue functions will be added. Request URL : https://github.com/alsa-project/alsa-gobject/pull/10 Patch URL : https://github.com/alsa-project/alsa-gobject/pull/10.patch Repository URL: https://github.com/alsa-project/alsa-gobject

1 0

[PATCH 0/1] reviving ack callback for 5.7(latest)
by Gyeongtaek Lee 03 Apr '20

03 Apr '20

ACK callback is useful to notify application pointer in DMA buffer to DSP, because it is called after application pointer is updated in ALSA. DSP can use the pointer to control preprocessing, detect underrun and so on. I'd like to revive it from 5.4 if it is possible. This patch is created in for-linus branch. Gyeongtaek Lee (1): ASoC: soc-component: add snd_soc_component_ack() include/sound/soc-component.h | 3 +++ sound/soc/soc-component.c | 14 ++++++++++++++ sound/soc/soc-pcm.c | 2 ++ 3 files changed, 19 insertions(+) base-commit: 52ac54e1de31af2d40f10dd4e711e774655e091b -- 2.21.0

3 2

[PATCH] quirk for Pioneer DJ DJM-250MK2 usb sound card
by František Kučera 03 Apr '20

03 Apr '20

Hello, I have created a quirk that adds support for the Pioneer DJ DJM-250MK2 mixer (acts like a USB sound card, but Vendor Specific Class). Output (from computer to card) is working: 8 channels, 48 000 Hz, S24_3LE. Input is not working. I tried, but I was unable to make it working even under an officially supported OS with Pioneer drivers - Audacity or Mixxx see the channels, but there is no signal. It probably works only with Pioneer proprietary application. But I hope that even mere output support will be useful for someone. (it is usable for playback, just not for DVS) Franta Signed-off-by: František Kučera <franta-linux(a)frantovo.cz> --- sound/usb/quirks-table.h | 42 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) diff --git a/sound/usb/quirks-table.h b/sound/usb/quirks-table.h index d187aa6d50db..dcaf9eed9a41 100644 --- a/sound/usb/quirks-table.h +++ b/sound/usb/quirks-table.h @@ -3592,5 +3592,47 @@ AU0828_DEVICE(0x2040, 0x7270, "Hauppauge", "HVR-950Q"), } } }, +{ + /* + * Pioneer DJ DJM-250MK2 + * PCM is 8 channels out @ 48 fixed (endpoints 0x01). + * The output from computer to the mixer is usable. + * + * The input (phono or line to computer) is not working. + * It should be at endpoint 0x82 and probably also 8 channels, + * but it seems that it works only with Pioneer proprietary software. + * Even on officially supported OS, the Audacity was unable to record + * and Mixxx to recognize the control vinyls. + */ + USB_DEVICE_VENDOR_SPEC(0x2b73, 0x0017), + .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 = 8, // outputs + .iface = 0, + .altsetting = 1, + .altset_idx = 1, + .endpoint = 0x01, + .ep_attr = USB_ENDPOINT_XFER_ISOC| + USB_ENDPOINT_SYNC_ASYNC, + .rates = SNDRV_PCM_RATE_48000, + .rate_min = 48000, + .rate_max = 48000, + .nr_rates = 1, + .rate_table = (unsigned int[]) { 48000 } + } + }, + { + .ifnum = -1 + } + } + } +}, #undef USB_DEVICE_VENDOR_SPEC

4 6

[PATCH] ALSA: pcm: oss: Fix regression by buffer overflow fix (again)
by Takashi Iwai 03 Apr '20

03 Apr '20

[ This is essentially the same fix as commit ae769d355664, but it's adapted to the latest code for 5.7; hence it contains no Fixes or other tags for avoid backport confusion -- tiwai ] The recent fix for the OOB access in PCM OSS plugins (commit f2ecf903ef06: "ALSA: pcm: oss: Avoid plugin buffer overflow") caused a regression on OSS applications. The patch introduced the size check in client and slave size calculations to limit to each plugin's buffer size, but I overlooked that some code paths call those without allocating the buffer but just for estimation. This patch fixes the bug by skipping the size check for those code paths while keeping checking in the actual transfer calls. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/core/oss/pcm_plugin.c | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/sound/core/oss/pcm_plugin.c b/sound/core/oss/pcm_plugin.c index fbda4ebf38b3..59d62f05658f 100644 --- a/sound/core/oss/pcm_plugin.c +++ b/sound/core/oss/pcm_plugin.c @@ -197,7 +197,8 @@ int snd_pcm_plugin_free(struct snd_pcm_plugin *plugin) } static snd_pcm_sframes_t calc_dst_frames(struct snd_pcm_substream *plug, - snd_pcm_sframes_t frames) + snd_pcm_sframes_t frames, + bool check_size) { struct snd_pcm_plugin *plugin, *plugin_next; @@ -209,7 +210,7 @@ static snd_pcm_sframes_t calc_dst_frames(struct snd_pcm_substream *plug, if (frames < 0) return frames; } - if (frames > plugin->buf_frames) + if (check_size && frames > plugin->buf_frames) frames = plugin->buf_frames; plugin = plugin_next; } @@ -217,13 +218,14 @@ static snd_pcm_sframes_t calc_dst_frames(struct snd_pcm_substream *plug, } static snd_pcm_sframes_t calc_src_frames(struct snd_pcm_substream *plug, - snd_pcm_sframes_t frames) + snd_pcm_sframes_t frames, + bool check_size) { struct snd_pcm_plugin *plugin, *plugin_prev; plugin = snd_pcm_plug_last(plug); while (plugin && frames > 0) { - if (frames > plugin->buf_frames) + if (check_size && frames > plugin->buf_frames) frames = plugin->buf_frames; plugin_prev = plugin->prev; if (plugin->src_frames) { @@ -242,9 +244,9 @@ snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, snd_p return -ENXIO; switch (snd_pcm_plug_stream(plug)) { case SNDRV_PCM_STREAM_PLAYBACK: - return calc_src_frames(plug, drv_frames); + return calc_src_frames(plug, drv_frames, false); case SNDRV_PCM_STREAM_CAPTURE: - return calc_dst_frames(plug, drv_frames); + return calc_dst_frames(plug, drv_frames, false); default: snd_BUG(); return -EINVAL; @@ -257,9 +259,9 @@ snd_pcm_sframes_t snd_pcm_plug_slave_size(struct snd_pcm_substream *plug, snd_pc return -ENXIO; switch (snd_pcm_plug_stream(plug)) { case SNDRV_PCM_STREAM_PLAYBACK: - return calc_dst_frames(plug, clt_frames); + return calc_dst_frames(plug, clt_frames, false); case SNDRV_PCM_STREAM_CAPTURE: - return calc_src_frames(plug, clt_frames); + return calc_src_frames(plug, clt_frames, false); default: snd_BUG(); return -EINVAL; @@ -622,7 +624,7 @@ snd_pcm_sframes_t snd_pcm_plug_write_transfer(struct snd_pcm_substream *plug, st src_channels = dst_channels; plugin = next; } - return snd_pcm_plug_client_size(plug, frames); + return calc_src_frames(plug, frames, true); } snd_pcm_sframes_t snd_pcm_plug_read_transfer(struct snd_pcm_substream *plug, struct snd_pcm_plugin_channel *dst_channels_final, snd_pcm_uframes_t size) @@ -632,7 +634,7 @@ snd_pcm_sframes_t snd_pcm_plug_read_transfer(struct snd_pcm_substream *plug, str snd_pcm_sframes_t frames = size; int err; - frames = snd_pcm_plug_slave_size(plug, frames); + frames = calc_src_frames(plug, frames, true); if (frames < 0) return frames; -- 2.16.4

1 0

[PATCH] ALSA: pcm: oss: Fix regression by buffer overflow fix
by Takashi Iwai 03 Apr '20

03 Apr '20

The recent fix for the OOB access in PCM OSS plugins (commit f2ecf903ef06: "ALSA: pcm: oss: Avoid plugin buffer overflow") caused a regression on OSS applications. The patch introduced the size check in client and slave size calculations to limit to each plugin's buffer size, but I overlooked that some code paths call those without allocating the buffer but just for estimation. This patch fixes the bug by skipping the size check for those code paths while keeping checking in the actual transfer calls. Fixes: f2ecf903ef06 ("ALSA: pcm: oss: Avoid plugin buffer overflow") Tested-and-reported-by: Jari Ruusu <jari.ruusu(a)gmail.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/core/oss/pcm_plugin.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) diff --git a/sound/core/oss/pcm_plugin.c b/sound/core/oss/pcm_plugin.c index 752d078908e9..50c35ecc8953 100644 --- a/sound/core/oss/pcm_plugin.c +++ b/sound/core/oss/pcm_plugin.c @@ -196,7 +196,9 @@ int snd_pcm_plugin_free(struct snd_pcm_plugin *plugin) return 0; } -snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, snd_pcm_uframes_t drv_frames) +static snd_pcm_sframes_t plug_client_size(struct snd_pcm_substream *plug, + snd_pcm_uframes_t drv_frames, + bool check_size) { struct snd_pcm_plugin *plugin, *plugin_prev, *plugin_next; int stream; @@ -209,7 +211,7 @@ snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, snd_p if (stream == SNDRV_PCM_STREAM_PLAYBACK) { plugin = snd_pcm_plug_last(plug); while (plugin && drv_frames > 0) { - if (drv_frames > plugin->buf_frames) + if (check_size && drv_frames > plugin->buf_frames) drv_frames = plugin->buf_frames; plugin_prev = plugin->prev; if (plugin->src_frames) @@ -222,7 +224,7 @@ snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, snd_p plugin_next = plugin->next; if (plugin->dst_frames) drv_frames = plugin->dst_frames(plugin, drv_frames); - if (drv_frames > plugin->buf_frames) + if (check_size && drv_frames > plugin->buf_frames) drv_frames = plugin->buf_frames; plugin = plugin_next; } @@ -231,7 +233,9 @@ snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, snd_p return drv_frames; } -snd_pcm_sframes_t snd_pcm_plug_slave_size(struct snd_pcm_substream *plug, snd_pcm_uframes_t clt_frames) +static snd_pcm_sframes_t plug_slave_size(struct snd_pcm_substream *plug, + snd_pcm_uframes_t clt_frames, + bool check_size) { struct snd_pcm_plugin *plugin, *plugin_prev, *plugin_next; snd_pcm_sframes_t frames; @@ -252,14 +256,14 @@ snd_pcm_sframes_t snd_pcm_plug_slave_size(struct snd_pcm_substream *plug, snd_pc if (frames < 0) return frames; } - if (frames > plugin->buf_frames) + if (check_size && frames > plugin->buf_frames) frames = plugin->buf_frames; plugin = plugin_next; } } else if (stream == SNDRV_PCM_STREAM_CAPTURE) { plugin = snd_pcm_plug_last(plug); while (plugin) { - if (frames > plugin->buf_frames) + if (check_size && frames > plugin->buf_frames) frames = plugin->buf_frames; plugin_prev = plugin->prev; if (plugin->src_frames) { @@ -274,6 +278,18 @@ snd_pcm_sframes_t snd_pcm_plug_slave_size(struct snd_pcm_substream *plug, snd_pc return frames; } +snd_pcm_sframes_t snd_pcm_plug_client_size(struct snd_pcm_substream *plug, + snd_pcm_uframes_t drv_frames) +{ + return plug_client_size(plug, drv_frames, false); +} + +snd_pcm_sframes_t snd_pcm_plug_slave_size(struct snd_pcm_substream *plug, + snd_pcm_uframes_t clt_frames) +{ + return plug_slave_size(plug, clt_frames, false); +} + static int snd_pcm_plug_formats(const struct snd_mask *mask, snd_pcm_format_t format) { @@ -630,7 +646,7 @@ snd_pcm_sframes_t snd_pcm_plug_write_transfer(struct snd_pcm_substream *plug, st src_channels = dst_channels; plugin = next; } - return snd_pcm_plug_client_size(plug, frames); + return plug_client_size(plug, frames, true); } snd_pcm_sframes_t snd_pcm_plug_read_transfer(struct snd_pcm_substream *plug, struct snd_pcm_plugin_channel *dst_channels_final, snd_pcm_uframes_t size) @@ -640,7 +656,7 @@ snd_pcm_sframes_t snd_pcm_plug_read_transfer(struct snd_pcm_substream *plug, str snd_pcm_sframes_t frames = size; int err; - frames = snd_pcm_plug_slave_size(plug, frames); + frames = plug_slave_size(plug, frames, true); if (frames < 0) return frames; -- 2.16.4

1 0

[PATCH v2 1/1] ASoC: soc-component: add snd_soc_component_ack()
by Gyeongtaek Lee 03 Apr '20

03 Apr '20

Current ALSA SoC is directly using component->driver->ops->xxx, thus, it is deep nested, and makes code difficult to read, and is not good for encapsulation. This patch adds new snd_soc_component_ack() and use it. Ack callback is revived because some ASoC driver is using it. Signed-off-by: Gyeongtaek Lee <gt82.lee(a)samsung.com> --- include/sound/soc-component.h | 1 + sound/soc/soc-component.c | 18 ++++++++++++++++++ sound/soc/soc-pcm.c | 2 ++ 3 files changed, 21 insertions(+) diff --git a/include/sound/soc-component.h b/include/sound/soc-component.h index 5d80b2eef525..50b079272b47 100644 --- a/include/sound/soc-component.h +++ b/include/sound/soc-component.h @@ -381,6 +381,7 @@ struct page *snd_soc_pcm_component_page(struct snd_pcm_substream *substream, unsigned long offset); int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, struct vm_area_struct *vma); +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream); int snd_soc_pcm_component_new(struct snd_pcm *pcm); void snd_soc_pcm_component_free(struct snd_pcm *pcm); diff --git a/sound/soc/soc-component.c b/sound/soc/soc-component.c index 79ffc2820ba9..b4be91cb9fa0 100644 --- a/sound/soc/soc-component.c +++ b/sound/soc/soc-component.c @@ -526,6 +526,24 @@ int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, return -EINVAL; } +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream) +{ + struct snd_soc_pcm_runtime *rtd = substream->private_data; + struct snd_soc_rtdcom_list *rtdcom; + struct snd_soc_component *component; + + for_each_rtdcom(rtd, rtdcom) { + component = rtdcom->component; + + /* FIXME. it returns 1st ack now */ + if (component->driver->ops && + component->driver->ops->ack) + return component->driver->ops->ack(substream); + } + + return -EINVAL; +} + int snd_soc_pcm_component_new(struct snd_pcm *pcm) { struct snd_soc_pcm_runtime *rtd = pcm->private_data; diff --git a/sound/soc/soc-pcm.c b/sound/soc/soc-pcm.c index b600d3eaaf5c..c8a5a9a9d35f 100644 --- a/sound/soc/soc-pcm.c +++ b/sound/soc/soc-pcm.c @@ -2941,6 +2941,8 @@ int soc_new_pcm(struct snd_soc_pcm_runtime *rtd, int num) if (!ops) continue; + if (ops->ack) + rtd->ops.ack = snd_soc_pcm_component_ack; if (ops->copy_user) rtd->ops.copy_user = snd_soc_pcm_component_copy_user; if (ops->page) base-commit: ef8e14794308a428b194f8b06ad9ae06b43466e4 -- 2.21.0

1 0

[PATCH 1/1] ASoC: soc-component: add snd_soc_component_ack()
by Gyeongtaek Lee 03 Apr '20

03 Apr '20

Current ALSA SoC is directly using component->driver->ops->xxx, thus, it is deep nested, and makes code difficult to read, and is not good for encapsulation. This patch adds new snd_soc_component_ack() and use it. Ack callback is revived because some ASoC driver is using it. Signed-off-by: Gyeongtaek Lee <gt82.lee(a)samsung.com> --- include/sound/soc-component.h | 3 +++ sound/soc/soc-component.c | 14 ++++++++++++++ sound/soc/soc-pcm.c | 2 ++ 3 files changed, 19 insertions(+) diff --git a/include/sound/soc-component.h b/include/sound/soc-component.h index 154d02fbbfed..a8c60efac8f7 100644 --- a/include/sound/soc-component.h +++ b/include/sound/soc-component.h @@ -104,6 +104,8 @@ struct snd_soc_component_driver { struct page *(*page)(struct snd_soc_component *component, struct snd_pcm_substream *substream, unsigned long offset); + int (*ack) (struct snd_soc_component *component, + struct snd_pcm_substream *substream); int (*mmap)(struct snd_soc_component *component, struct snd_pcm_substream *substream, struct vm_area_struct *vma); @@ -415,6 +417,7 @@ struct page *snd_soc_pcm_component_page(struct snd_pcm_substream *substream, unsigned long offset); int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, struct vm_area_struct *vma); +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream); int snd_soc_pcm_component_new(struct snd_soc_pcm_runtime *rtd); void snd_soc_pcm_component_free(struct snd_soc_pcm_runtime *rtd); diff --git a/sound/soc/soc-component.c b/sound/soc/soc-component.c index 785a0385cc7f..b4b3245ef016 100644 --- a/sound/soc/soc-component.c +++ b/sound/soc/soc-component.c @@ -516,6 +516,20 @@ int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, return -EINVAL; } +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream) +{ + struct snd_soc_pcm_runtime *rtd = substream->private_data; + struct snd_soc_component *component; + int i; + + /* FIXME. it returns 1st ack now */ + for_each_rtd_components(rtd, i, component) + if (component->driver->ack) + return component->driver->ack(component, substream); + + return -EINVAL; +} + int snd_soc_pcm_component_new(struct snd_soc_pcm_runtime *rtd) { struct snd_soc_component *component; diff --git a/sound/soc/soc-pcm.c b/sound/soc/soc-pcm.c index 289aebc15529..c6fddd1cb5aa 100644 --- a/sound/soc/soc-pcm.c +++ b/sound/soc/soc-pcm.c @@ -3027,6 +3027,8 @@ int soc_new_pcm(struct snd_soc_pcm_runtime *rtd, int num) rtd->ops.ioctl = snd_soc_pcm_component_ioctl; if (drv->sync_stop) rtd->ops.sync_stop = snd_soc_pcm_component_sync_stop; + if (drv->ack) + rtd->ops.ack = snd_soc_pcm_component_ack; if (drv->copy_user) rtd->ops.copy_user = snd_soc_pcm_component_copy_user; if (drv->page) -- 2.21.0

1 0

[PATCH 1/1] ASoC: soc-component: add snd_soc_component_ack()
by Gyeongtaek Lee 03 Apr '20

03 Apr '20

Current ALSA SoC is directly using component->driver->ops->xxx, thus, it is deep nested, and makes code difficult to read, and is not good for encapsulation. This patch adds new snd_soc_component_ack() and use it. Ack callback is revived because some ASoC driver is using it. Change-Id: I85573cb43f3ca050c47ca9e3af3cfa9ec1fae3d0 Signed-off-by: Gyeongtaek Lee <gt82.lee(a)samsung.com> --- include/sound/soc-component.h | 1 + sound/soc/soc-component.c | 18 ++++++++++++++++++ sound/soc/soc-pcm.c | 2 ++ 3 files changed, 21 insertions(+) diff --git a/include/sound/soc-component.h b/include/sound/soc-component.h index 5d80b2eef525..50b079272b47 100644 --- a/include/sound/soc-component.h +++ b/include/sound/soc-component.h @@ -381,6 +381,7 @@ struct page *snd_soc_pcm_component_page(struct snd_pcm_substream *substream, unsigned long offset); int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, struct vm_area_struct *vma); +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream); int snd_soc_pcm_component_new(struct snd_pcm *pcm); void snd_soc_pcm_component_free(struct snd_pcm *pcm); diff --git a/sound/soc/soc-component.c b/sound/soc/soc-component.c index 79ffc2820ba9..b4be91cb9fa0 100644 --- a/sound/soc/soc-component.c +++ b/sound/soc/soc-component.c @@ -526,6 +526,24 @@ int snd_soc_pcm_component_mmap(struct snd_pcm_substream *substream, return -EINVAL; } +int snd_soc_pcm_component_ack(struct snd_pcm_substream *substream) +{ + struct snd_soc_pcm_runtime *rtd = substream->private_data; + struct snd_soc_rtdcom_list *rtdcom; + struct snd_soc_component *component; + + for_each_rtdcom(rtd, rtdcom) { + component = rtdcom->component; + + /* FIXME. it returns 1st ack now */ + if (component->driver->ops && + component->driver->ops->ack) + return component->driver->ops->ack(substream); + } + + return -EINVAL; +} + int snd_soc_pcm_component_new(struct snd_pcm *pcm) { struct snd_soc_pcm_runtime *rtd = pcm->private_data; diff --git a/sound/soc/soc-pcm.c b/sound/soc/soc-pcm.c index b600d3eaaf5c..c8a5a9a9d35f 100644 --- a/sound/soc/soc-pcm.c +++ b/sound/soc/soc-pcm.c @@ -2941,6 +2941,8 @@ int soc_new_pcm(struct snd_soc_pcm_runtime *rtd, int num) if (!ops) continue; + if (ops->ack) + rtd->ops.ack = snd_soc_pcm_component_ack; if (ops->copy_user) rtd->ops.copy_user = snd_soc_pcm_component_copy_user; if (ops->page) -- 2.21.0

1 0

[PATCH 0/1] reviving ack callback for 5.4
by Gyeongtaek Lee 03 Apr '20

03 Apr '20

ACK callback is useful to notify application pointer in DMA buffer to DSP, because it is called after application pointer is updated in ALSA. DSP can use the pointer to control preprocessing, detect underrun and so on. I'd like to revive it from 5.4 if it is possible. Gyeongtaek Lee (1): ASoC: soc-component: add snd_soc_component_ack() include/sound/soc-component.h | 1 + sound/soc/soc-component.c | 18 ++++++++++++++++++ sound/soc/soc-pcm.c | 2 ++ 3 files changed, 21 insertions(+) base-commit: ef8e14794308a428b194f8b06ad9ae06b43466e4 -- 2.21.0

1 0