ALSA: hda - DP MST audio for Jack support

RFC PATCH In review [email protected] 3 participants 13 messages Nov 3, 2015 – Dec 22, 2015 Sound

lore

Share & tools

lore

[email protected]

author

2015-11-03 08:44 UTC #0 lore raw

cover ALSA: hda - DP MST audio for Jack support

From: Libin Yang <[email protected]>

This is the patch set for Jack support for DP MST audio

Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin

 sound/pci/hda/hda_jack.c   | 186 ++++++++++++++++++++++++++++++++++++++++-----
 sound/pci/hda/hda_jack.h   |  32 +++++++-
 sound/pci/hda/patch_hdmi.c | 116 ++++++++++++++++++++++------
 3 files changed, 288 insertions(+), 46 deletions(-)

Signature · 1 line

1.9.1

[email protected]

author

2015-11-03 08:44 UTC #1 lore raw

RFC PATCH 1/2 ALSA: hda - jack support DP MST audio

From: Libin Yang <[email protected]>

Add jack support for DP MST audio in hda_jack.c

Signed-off-by Libin Yang <[email protected]>

---
 sound/pci/hda/hda_jack.c   | 158 +++++++++++++++++++++++++++++++++++++++------
 sound/pci/hda/hda_jack.h   |  28 +++++++-
 sound/pci/hda/patch_hdmi.c |   2 +-
 3 files changed, 166 insertions(+), 22 deletions(-)

sound/pci/hda/hda_jack.c +140 ~~-18~~

index 366efbf..c0a018b 100644
@@ -47,7 +47,8 @@ bool is_jack_detectable(struct hda_codec *codec, hda_nid_t nid)
 EXPORT_SYMBOL_GPL(is_jack_detectable);
 
 /* execute pin sense measurement */
-static u32 read_pin_sense(struct hda_codec *codec, hda_nid_t nid)
+static u32 read_pin_sense(struct hda_codec *codec, hda_nid_t nid,
+					hda_dev_t dev_id)
 {
 	u32 pincap;
 	u32 val;

@@ -59,7 +60,7 @@ static u32 read_pin_sense(struct hda_codec *codec, hda_nid_t nid)
 					AC_VERB_SET_PIN_SENSE, 0);
 	}
 	val = snd_hda_codec_read(codec, nid, 0,
-				  AC_VERB_GET_PIN_SENSE, 0);
+				  AC_VERB_GET_PIN_SENSE, dev_id);
 	if (codec->inv_jack_detect)
 		val ^= AC_PINSENSE_PRESENCE;
 	return val;

@@ -86,6 +87,28 @@ snd_hda_jack_tbl_get(struct hda_codec *codec, hda_nid_t nid)
 EXPORT_SYMBOL_GPL(snd_hda_jack_tbl_get);
 
 /**
+ * snd_hda_jack_tbl_get_mst - query the jack-table entry for the given NID
+ * @codec: the HDA codec
+ * @nid: pin NID to refer to
+ * @dev_id: device entry id
+ */
+struct hda_jack_tbl *
+snd_hda_jack_tbl_get_mst(struct hda_codec *codec, hda_nid_t nid,
+			hda_dev_t dev_id)
+{
+	struct hda_jack_tbl *jack = codec->jacktbl.list;
+	int i;
+
+	if (!nid || !jack)
+		return NULL;
+	for (i = 0; i < codec->jacktbl.used; i++, jack++)
+		if (jack->nid == nid && jack->dev_id == dev_id)
+			return jack;
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(snd_hda_jack_tbl_get_mst);
+
+/**
  * snd_hda_jack_tbl_get_from_tag - query the jack-table entry for the given tag
  * @codec: the HDA codec
  * @tag: tag value to refer to

@@ -109,17 +132,20 @@ EXPORT_SYMBOL_GPL(snd_hda_jack_tbl_get_from_tag);
  * snd_hda_jack_tbl_new - create a jack-table entry for the given NID
  * @codec: the HDA codec
  * @nid: pin NID to assign
+ * @dev_id: device entry id
  */
 static struct hda_jack_tbl *
-snd_hda_jack_tbl_new(struct hda_codec *codec, hda_nid_t nid)
+snd_hda_jack_tbl_new(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id)
 {
-	struct hda_jack_tbl *jack = snd_hda_jack_tbl_get(codec, nid);
+	struct hda_jack_tbl *jack =
+		snd_hda_jack_tbl_get_mst(codec, nid, dev_id);
 	if (jack)
 		return jack;
 	jack = snd_array_new(&codec->jacktbl);
 	if (!jack)
 		return NULL;
 	jack->nid = nid;
+	jack->dev_id = dev_id;
 	jack->jack_dirty = 1;
 	jack->tag = codec->jacktbl.used;
 	return jack;

@@ -157,9 +183,11 @@ static void jack_detect_update(struct hda_codec *codec,
 	if (jack->phantom_jack)
 		jack->pin_sense = AC_PINSENSE_PRESENCE;
 	else
-		jack->pin_sense = read_pin_sense(codec, jack->nid);
+		jack->pin_sense =
+			read_pin_sense(codec, jack->nid, jack->dev_id);
 
 	/* A gating jack indicates the jack is invalid if gating is unplugged */
+	/* fixme: MST audio doesn't using gating and gated jack. */
 	if (jack->gating_jack && !snd_hda_jack_detect(codec, jack->gating_jack))
 		jack->pin_sense &= ~AC_PINSENSE_PRESENCE;

@@ -198,18 +226,20 @@ EXPORT_SYMBOL_GPL(snd_hda_jack_set_dirty_all);
  * snd_hda_pin_sense - execute pin sense measurement
  * @codec: the CODEC to sense
  * @nid: the pin NID to sense
+ * @dev_id: device entry id
  *
  * Execute necessary pin sense measurement and return its Presence Detect,
  * Impedance, ELD Valid etc. status bits.
  */
-u32 snd_hda_pin_sense(struct hda_codec *codec, hda_nid_t nid)
+u32 snd_hda_pin_sense(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id)
 {
-	struct hda_jack_tbl *jack = snd_hda_jack_tbl_get(codec, nid);
+	struct hda_jack_tbl *jack =
+		snd_hda_jack_tbl_get_mst(codec, nid, dev_id);
 	if (jack) {
 		jack_detect_update(codec, jack);
 		return jack->pin_sense;
 	}
-	return read_pin_sense(codec, nid);
+	return read_pin_sense(codec, nid, dev_id);
 }
 EXPORT_SYMBOL_GPL(snd_hda_pin_sense);

@@ -226,7 +256,7 @@ int snd_hda_jack_detect_state(struct hda_codec *codec, hda_nid_t nid)
 	struct hda_jack_tbl *jack = snd_hda_jack_tbl_get(codec, nid);
 	if (jack && jack->phantom_jack)
 		return HDA_JACK_PHANTOM;
-	else if (snd_hda_pin_sense(codec, nid) & AC_PINSENSE_PRESENCE)
+	else if (snd_hda_pin_sense(codec, nid, 0) & AC_PINSENSE_PRESENCE)
 		return HDA_JACK_PRESENT;
 	else
 		return HDA_JACK_NOT_PRESENT;

@@ -234,7 +264,31 @@ int snd_hda_jack_detect_state(struct hda_codec *codec, hda_nid_t nid)
 EXPORT_SYMBOL_GPL(snd_hda_jack_detect_state);
 
 /**
- * snd_hda_jack_detect_enable - enable the jack-detection
+ * snd_hda_jack_detect_state_mst - query pin Presence Detect status in mst mode
+ * @codec: the CODEC to sense
+ * @nid: the pin NID to sense
+ * @dev_id: device entry id
+ *
+ * Query and return the device entry's Presence Detect status, as either
+ * HDA_JACK_NOT_PRESENT, HDA_JACK_PRESENT or HDA_JACK_PHANTOM.
+ */
+int snd_hda_jack_detect_state_mst(struct hda_codec *codec, hda_nid_t nid,
+					hda_dev_t dev_id)
+{
+	struct hda_jack_tbl *jack =
+		snd_hda_jack_tbl_get_mst(codec, nid, dev_id);
+
+	if (jack && jack->phantom_jack)
+		return HDA_JACK_PHANTOM;
+	else if (snd_hda_pin_sense(codec, nid, dev_id) & AC_PINSENSE_PRESENCE)
+		return HDA_JACK_PRESENT;
+	else
+		return HDA_JACK_NOT_PRESENT;
+}
+EXPORT_SYMBOL_GPL(snd_hda_jack_detect_state_mst);
+
+/**
+ * snd_hda_jack_detect_enable_callback - enable the jack-detection
  * @codec: the HDA codec
  * @nid: pin NID to enable
  * @func: callback function to register

@@ -251,7 +305,7 @@ snd_hda_jack_detect_enable_callback(struct hda_codec *codec, hda_nid_t nid,
 	struct hda_jack_callback *callback = NULL;
 	int err;
 
-	jack = snd_hda_jack_tbl_new(codec, nid);
+	jack = snd_hda_jack_tbl_new(codec, nid, 0);
 	if (!jack)
 		return ERR_PTR(-ENOMEM);
 	if (func) {

@@ -279,6 +333,52 @@ snd_hda_jack_detect_enable_callback(struct hda_codec *codec, hda_nid_t nid,
 EXPORT_SYMBOL_GPL(snd_hda_jack_detect_enable_callback);
 
 /**
+ * snd_hda_jack_detect_enable_callback_mst - enable the jack-detection for mst
+ * @codec: the HDA codec
+ * @nid: pin NID to enable
+ * @dev_id: device entry id
+ * @func: callback function to register
+ *
+ * In the case of error, the return value will be a pointer embedded with
+ * errno.  Check and handle the return value appropriately with standard
+ * macros such as @IS_ERR() and @PTR_ERR().
+ */
+struct hda_jack_callback *
+snd_hda_jack_detect_enable_callback_mst(struct hda_codec *codec, hda_nid_t nid,
+			hda_dev_t dev_id, hda_jack_callback_fn func)
+{
+	struct hda_jack_tbl *jack;
+	struct hda_jack_callback *callback = NULL;
+	int err;
+
+	jack = snd_hda_jack_tbl_new(codec, nid, dev_id);
+	if (!jack)
+		return ERR_PTR(-ENOMEM);
+	if (func) {
+		callback = kzalloc(sizeof(*callback), GFP_KERNEL);
+		if (!callback)
+			return ERR_PTR(-ENOMEM);
+		callback->func = func;
+		callback->tbl = jack;
+		callback->next = jack->callback;
+		jack->callback = callback;
+	}
+
+	if (jack->jack_detect)
+		return callback; /* already registered */
+	jack->jack_detect = 1;
+	if (codec->jackpoll_interval > 0)
+		return callback; /* No unsol if we're polling instead */
+	err = snd_hda_codec_write_cache(codec, nid, 0,
+					 AC_VERB_SET_UNSOLICITED_ENABLE,
+					 AC_USRSP_EN | jack->tag);
+	if (err < 0)
+		return ERR_PTR(err);
+	return callback;
+}
+EXPORT_SYMBOL_GPL(snd_hda_jack_detect_enable_callback_mst);
+
+/**
  * snd_hda_jack_detect_enable - Enable the jack detection on the given pin
  * @codec: the HDA codec
  * @nid: pin NID to enable jack detection

@@ -303,8 +403,9 @@ EXPORT_SYMBOL_GPL(snd_hda_jack_detect_enable);
 int snd_hda_jack_set_gating_jack(struct hda_codec *codec, hda_nid_t gated_nid,
 				 hda_nid_t gating_nid)
 {
-	struct hda_jack_tbl *gated = snd_hda_jack_tbl_new(codec, gated_nid);
-	struct hda_jack_tbl *gating = snd_hda_jack_tbl_new(codec, gating_nid);
+	struct hda_jack_tbl *gated = snd_hda_jack_tbl_new(codec, gated_nid, 0);
+	struct hda_jack_tbl *gating =
+		snd_hda_jack_tbl_new(codec, gating_nid, 0);
 
 	if (!gated || !gating)
 		return -EINVAL;

@@ -369,6 +470,8 @@ static int get_input_jack_type(struct hda_codec *codec, hda_nid_t nid)
 static void hda_free_jack_priv(struct snd_jack *jack)
 {
 	struct hda_jack_tbl *jacks = jack->private_data;
+	if (jacks == NULL)
+		return;
 	jacks->nid = 0;
 	jacks->jack = NULL;
 }

@@ -377,6 +480,7 @@ static void hda_free_jack_priv(struct snd_jack *jack)
  * snd_hda_jack_add_kctl - Add a kctl for the given pin
  * @codec: the HDA codec
  * @nid: pin NID to assign
+ * @dev_id: device entry id
  * @name: string name for the jack
  * @phantom_jack: flag to deal as a phantom jack
  *

@@ -384,17 +488,18 @@ static void hda_free_jack_priv(struct snd_jack *jack)
  * will have the given name and index.
  */
 static int __snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
-			  const char *name, bool phantom_jack)
+		hda_dev_t dev_id, const char *name, bool phantom_jack)
 {
 	struct hda_jack_tbl *jack;
 	int err, state, type;
 
-	jack = snd_hda_jack_tbl_new(codec, nid);
+	jack = snd_hda_jack_tbl_new(codec, nid, dev_id);
 	if (!jack)
 		return 0;
 	if (jack->jack)
 		return 0; /* already created */
 
+	/* all device entries use same pincfg */
 	type = get_input_jack_type(codec, nid);
 	err = snd_jack_new(codec->card, name, type,
 			   &jack->jack, true, phantom_jack);

@@ -405,7 +510,7 @@ static int __snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 	jack->type = type;
 	jack->jack->private_data = jack;
 	jack->jack->private_free = hda_free_jack_priv;
-	state = snd_hda_jack_detect(codec, nid);
+	state = snd_hda_jack_detect_mst(codec, nid, dev_id);
 	snd_jack_report(jack->jack, state ? jack->type : 0);
 
 	return 0;

@@ -422,10 +527,27 @@ static int __snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 int snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 			  const char *name)
 {
-	return __snd_hda_jack_add_kctl(codec, nid, name, false);
+	return __snd_hda_jack_add_kctl(codec, nid, 0, name, false);
 }
 EXPORT_SYMBOL_GPL(snd_hda_jack_add_kctl);
 
+/**
+ * snd_hda_jack_add_kctl_mst - Add a jack kctl for the given pin for mst mode
+ * @codec: the HDA codec
+ * @nid: pin NID
+ * @dev_id: device entry id
+ * @name: the name string for the jack ctl
+ *
+ * This is a simple helper calling __snd_hda_jack_add_kctl().
+ */
+int snd_hda_jack_add_kctl_mst(struct hda_codec *codec, hda_nid_t nid,
+			hda_dev_t dev_id, const char *name)
+{
+	return __snd_hda_jack_add_kctl(codec, nid, dev_id, name, false);
+}
+EXPORT_SYMBOL_GPL(snd_hda_jack_add_kctl_mst);
+
+
 static int add_jack_kctl(struct hda_codec *codec, hda_nid_t nid,
 			 const struct auto_pin_cfg *cfg,
 			 const char *base_name)

@@ -451,7 +573,7 @@ static int add_jack_kctl(struct hda_codec *codec, hda_nid_t nid,
 	if (phantom_jack)
 		/* Example final name: "Internal Mic Phantom Jack" */
 		strncat(name, " Phantom", sizeof(name) - strlen(name) - 1);
-	err = __snd_hda_jack_add_kctl(codec, nid, name, phantom_jack);
+	err = __snd_hda_jack_add_kctl(codec, nid, 0, name, phantom_jack);
 	if (err < 0)
 		return err;

sound/pci/hda/hda_jack.h +25 -3

index 387d309..c9d4f8e 100644
@@ -29,6 +29,7 @@ struct hda_jack_callback {
 
 struct hda_jack_tbl {
 	hda_nid_t nid;
+	hda_dev_t dev_id;
 	unsigned char tag;		/* unsol event tag */
 	struct hda_jack_callback *callback;
 	/* jack-detection stuff */

@@ -46,6 +47,9 @@ struct hda_jack_tbl {
 struct hda_jack_tbl *
 snd_hda_jack_tbl_get(struct hda_codec *codec, hda_nid_t nid);
 struct hda_jack_tbl *
+snd_hda_jack_tbl_get_mst(struct hda_codec *codec, hda_nid_t nid,
+				 hda_dev_t dev_id);
+struct hda_jack_tbl *
 snd_hda_jack_tbl_get_from_tag(struct hda_codec *codec, unsigned char tag);
 
 void snd_hda_jack_tbl_clear(struct hda_codec *codec);

@@ -56,11 +60,14 @@ int snd_hda_jack_detect_enable(struct hda_codec *codec, hda_nid_t nid);
 struct hda_jack_callback *
 snd_hda_jack_detect_enable_callback(struct hda_codec *codec, hda_nid_t nid,
 				    hda_jack_callback_fn cb);
+struct hda_jack_callback *
+snd_hda_jack_detect_enable_callback_mst(struct hda_codec *codec, hda_nid_t nid,
+			hda_dev_t dev_id, hda_jack_callback_fn func);
 
 int snd_hda_jack_set_gating_jack(struct hda_codec *codec, hda_nid_t gated_nid,
 				 hda_nid_t gating_nid);
 
-u32 snd_hda_pin_sense(struct hda_codec *codec, hda_nid_t nid);
+u32 snd_hda_pin_sense(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id);
 
 /* the jack state returned from snd_hda_jack_detect_state() */
 enum {

@@ -68,7 +75,8 @@ enum {
 };
 
 int snd_hda_jack_detect_state(struct hda_codec *codec, hda_nid_t nid);
-
+int snd_hda_jack_detect_state_mst(struct hda_codec *codec,
+					hda_nid_t nid, hda_dev_t dev_id);
 /**
  * snd_hda_jack_detect - Detect the jack
  * @codec: the HDA codec

@@ -79,13 +87,27 @@ static inline bool snd_hda_jack_detect(struct hda_codec *codec, hda_nid_t nid)
 	return snd_hda_jack_detect_state(codec, nid) != HDA_JACK_NOT_PRESENT;
 }
 
+/**
+ * snd_hda_jack_detect - Detect the jack
+ * @codec: the HDA codec
+ * @nid: pin NID to check jack detection
+ * @dev_id: device entry id
+ */
+static inline bool snd_hda_jack_detect_mst(struct hda_codec *codec,
+				hda_nid_t nid, hda_dev_t dev_id)
+{
+	return (snd_hda_jack_detect_state_mst(codec, nid, dev_id) !=
+			HDA_JACK_NOT_PRESENT);
+}
+
 bool is_jack_detectable(struct hda_codec *codec, hda_nid_t nid);
 
 int snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 			  const char *name);
 int snd_hda_jack_add_kctls(struct hda_codec *codec,
 			   const struct auto_pin_cfg *cfg);
-
+int snd_hda_jack_add_kctl_mst(struct hda_codec *codec, hda_nid_t nid,
+			   hda_dev_t dev_id, const char *name);
 void snd_hda_jack_report_sync(struct hda_codec *codec);
 
 void snd_hda_jack_unsol_event(struct hda_codec *codec, unsigned int res);

sound/pci/hda/patch_hdmi.c +1 -1

index 8d31366..39d68ba 100644
@@ -1664,7 +1664,7 @@ static bool hdmi_present_sense(struct hdmi_spec_per_pin *per_pin, int repoll)
 	bool ret;
 
 	snd_hda_power_up_pm(codec);
-	present = snd_hda_pin_sense(codec, pin_nid);
+	present = snd_hda_pin_sense(codec, pin_nid, 0);
 
 	mutex_lock(&per_pin->lock);
 	pin_eld->monitor_present = !!(present & AC_PINSENSE_PRESENCE);

Signature · 1 line

1.9.1

[email protected]

author

2015-11-03 08:44 UTC #2 lore raw

RFC PATCH 2/2 ALSA: hda - hdmi audio add dynamically jack binding to pin

In reply to [email protected]

From: Libin Yang <[email protected]>

Jack is attached to PCM statically.

When there is monitor connected, dynamically
bind the Jack to pin based on the PCM assignment.

When monitor is disconnected, unbind the jack from the pin.

Signed-off-by Libin Yang <[email protected]>

---
 sound/pci/hda/hda_jack.c   |  30 +++++++++++-
 sound/pci/hda/hda_jack.h   |   4 ++
 sound/pci/hda/patch_hdmi.c | 111 +++++++++++++++++++++++++++++++++++----------
 3 files changed, 120 insertions(+), 25 deletions(-)

sound/pci/hda/hda_jack.c +28 -2

index c0a018b..a4a5398 100644
@@ -134,7 +134,7 @@ EXPORT_SYMBOL_GPL(snd_hda_jack_tbl_get_from_tag);
  * @nid: pin NID to assign
  * @dev_id: device entry id
  */
-static struct hda_jack_tbl *
+struct hda_jack_tbl *
 snd_hda_jack_tbl_new(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id)
 {
 	struct hda_jack_tbl *jack =

@@ -150,6 +150,7 @@ snd_hda_jack_tbl_new(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id)
 	jack->tag = codec->jacktbl.used;
 	return jack;
 }
+EXPORT_SYMBOL_GPL(snd_hda_jack_tbl_new);
 
 void snd_hda_jack_tbl_clear(struct hda_codec *codec)
 {

@@ -517,6 +518,32 @@ static int __snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 }
 
 /**
+ * snd_hda_jack_new - create a new snd_jack
+ * @codec: the HDA codec
+ * @jack: Used to provide the allocated jack object to the caller.
+ * @type:  a bitmask of enum snd_jack_type values that can be detected by
+ *         this jack
+ * @id:    an identifying string for this jack
+ * @phantom_jack: Don't create a input device for phantom jacks.
+ *
+ * Return: Zero if successful, or a negative error code on failure.
+ * On success @jack will be initialised.
+ */
+int snd_hda_jack_new(struct hda_codec *codec, struct snd_jack **jack, int type,
+					 const char *id, bool phantom_jack)
+{
+	int err;
+
+	err = snd_jack_new(codec->card, id, type,
+					   jack, true, phantom_jack);
+	if (err < 0)
+		return err;
+	(*jack)->private_free = hda_free_jack_priv;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(snd_hda_jack_new);
+
+/**
  * snd_hda_jack_add_kctl - Add a jack kctl for the given pin
  * @codec: the HDA codec
  * @nid: pin NID

@@ -547,7 +574,6 @@ int snd_hda_jack_add_kctl_mst(struct hda_codec *codec, hda_nid_t nid,
 }
 EXPORT_SYMBOL_GPL(snd_hda_jack_add_kctl_mst);
 
-
 static int add_jack_kctl(struct hda_codec *codec, hda_nid_t nid,
 			 const struct auto_pin_cfg *cfg,
 			 const char *base_name)

sound/pci/hda/hda_jack.h +4 -0

index c9d4f8e..cb79703 100644
@@ -51,6 +51,8 @@ snd_hda_jack_tbl_get_mst(struct hda_codec *codec, hda_nid_t nid,
 				 hda_dev_t dev_id);
 struct hda_jack_tbl *
 snd_hda_jack_tbl_get_from_tag(struct hda_codec *codec, unsigned char tag);
+struct hda_jack_tbl *
+snd_hda_jack_tbl_new(struct hda_codec *codec, hda_nid_t nid, hda_dev_t dev_id);
 
 void snd_hda_jack_tbl_clear(struct hda_codec *codec);

@@ -104,6 +106,8 @@ bool is_jack_detectable(struct hda_codec *codec, hda_nid_t nid);
 
 int snd_hda_jack_add_kctl(struct hda_codec *codec, hda_nid_t nid,
 			  const char *name);
+int snd_hda_jack_new(struct hda_codec *codec, struct snd_jack **jack, int type,
+					 const char *id, bool phantom_jack);
 int snd_hda_jack_add_kctls(struct hda_codec *codec,
 			   const struct auto_pin_cfg *cfg);
 int snd_hda_jack_add_kctl_mst(struct hda_codec *codec, hda_nid_t nid,

sound/pci/hda/patch_hdmi.c +88 ~~-23~~

206 lines not on this page · raw message

Signature · 1 line

1.9.1

Takashi Iwai

2015-11-03 16:49 UTC #3 lore raw

In reply to [email protected]

Quoted 8 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:

From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin

I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem trying to assign such a number to the jack object.

We create all jacks statically at probe time. And in MST case, each jack is still supposed to be corresponding to each PCM stream, not to each pin. So I don't see how this fits with the big picture.

A bit more explanation is needed.

thanks

Takashi

Quoted 9 lines

 sound/pci/hda/hda_jack.c   | 186 ++++++++++++++++++++++++++++++++++++++++-----
 sound/pci/hda/hda_jack.h   |  32 +++++++-
 sound/pci/hda/patch_hdmi.c | 116 ++++++++++++++++++++++------
 3 files changed, 288 insertions(+), 46 deletions(-)

Signature · 1 line

1.9.1

Libin Yang

author

2015-11-04 06:24 UTC #4 lore raw

Quoted 7 lines · Takashi Iwai

Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio

Quoted 8 lines

Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem trying to assign such a number to the jack object.

Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.

These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.

The purpose of the patches is try to dynamically attach the Jack.

Quoted 4 lines

We create all jacks statically at probe time. And in MST case, each jack is still supposed to be corresponding to each PCM stream, not to each pin. So I don't see how this fits with the big picture.

In the generic_hdmi_build_jack(), it will create the snd_jack based on the pcm. This is for the userspace. So userspace can see the Jack number is the same as PCM. The first Jack for the first PCM, and so on. And it is not attached to any pin at first.

In generic_hdmi_build_controls(), it will also create hda_jack_tbl based on the pin. The hda_jack_tbl will record the pin information.

When hotplug, in the hdmi_present_sense(), it will call xxx_jack_bind() to bind the snd_jack to hda_jack_tbl

Quoted 2 lines

A bit more explanation is needed.

I will. Thanks.

Best Regards,
Libin

Quoted 15 lines

thanks

Takashi

Quoted 9 lines

  sound/pci/hda/hda_jack.c   | 186 ++++++++++++++++++++++++++++++++++++++++-----
  sound/pci/hda/hda_jack.h   |  32 +++++++-
  sound/pci/hda/patch_hdmi.c | 116 ++++++++++++++++++++++------
  3 files changed, 288 insertions(+), 46 deletions(-)

--
1.9.1

Takashi Iwai

2015-11-09 07:58 UTC #5 lore raw

Quoted 19 lines · Libin Yang

Quoted 7 lines · Takashi Iwai
Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Quoted 8 lines
Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem trying to assign such a number to the jack object.

Quoted 8 lines

Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.
The purpose of the patches is try to dynamically attach the Jack.

Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether the corresponding PCM is available or not. You're trying to change this scheme completely.

So, please convince us why this change is needed and which merit / demerit it'll bring. Let's start from that point.

thanks,

Takashi

Yang, Libin

2015-11-09 15:00 UTC #6 lore raw

Hi Takashi,

Quoted 38 lines

-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected]; Yang, Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 10 lines · Libin Yang
Quoted 7 lines · Takashi Iwai
Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem
trying
Quoted 9 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.

Quoted 7 lines

Quoted 1 line
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether the corresponding PCM is available or not. You're trying to change this scheme completely.

The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:

for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}

Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to PCM.

Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:

for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}

which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).

If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.

For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin 5. These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.

Best Regards,
Libin

Quoted 8 lines

So, please convince us why this change is needed and which merit / demerit it'll bring. Let's start from that point.(c)
thanks,
Takashi

Takashi Iwai

2015-11-09 15:15 UTC #7 lore raw

Quoted 51 lines · Libin

Hi Takashi,
Quoted 38 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected]; Yang, Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 10 lines · Libin Yang
Quoted 7 lines · Takashi Iwai
Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem
trying
Quoted 9 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.
Quoted 7 lines
Quoted 1 line
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether the corresponding PCM is available or not. You're trying to change this scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:

Quoted 7 lines

for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to PCM.

OK, so far, so good.

Quoted 16 lines

Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.

Quoted 6 lines

For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin 5. These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.

Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to check is its PCM assignment. It doesn't have to issue the jack state check via HDA verb at all. This should make things a lot easier.

Takashi

Libin Yang

author

2015-12-22 08:49 UTC #8 lore raw

Hi Takashi,

Quoted 86 lines · Takashi Iwai

Quoted 51 lines · Libin
Hi Takashi,
Quoted 38 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected]; Yang, Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 10 lines · Libin Yang
Quoted 7 lines · Takashi Iwai
Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem
trying
Quoted 9 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.
Quoted 7 lines
Quoted 1 line
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether the corresponding PCM is available or not. You're trying to change this scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:
Quoted 7 lines
for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to PCM.
OK, so far, so good.
Quoted 16 lines
Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.
Quoted 6 lines
For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin 5. These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.

Quoted 4 lines

Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to check is its PCM assignment. It doesn't have to issue the jack state check via HDA verb at all. This should make things a lot easier.

I'm now porting the jack supporting for MST audio. The code on hdmi-jack branch will create jack differently based codec_has_acomp() condition. Does this mean MST Jack binding will handle both situation?

My idea is:
1. Add struct snd_jack *jack[] member in hdmi_spec.
2. Create the jack based on PCM
3. Assign spec->jack[i] to per_pin->acomp_jack when pin is connected.

This should be OK for codec_has_acomp sitation. But what about the other situation? The other situation still uses the hda_jack_tbl and etc. So we still need handle hda jack if we want to reuse the current code.

Best Regards,
Libin

Quoted 8 lines

Takashi
_______________________________________________
Alsa-devel mailing list
[email protected]
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel

Takashi Iwai

2015-12-22 09:00 UTC #9 lore raw

Quoted 90 lines · Libin Yang

Hi Takashi,
Quoted 86 lines · Takashi Iwai
Quoted 51 lines · Libin
Hi Takashi,
Quoted 38 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected]; Yang, Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 10 lines · Libin Yang
Quoted 7 lines · Takashi Iwai
Quoted 5 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean, you cannot know which MST dev# is given beforehand, while you seem
trying
Quoted 9 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device entry 0 so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio driver support will be implemented in the later patches. The patches help to prepare supporting MST audio.
Quoted 7 lines
Quoted 1 line
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether the corresponding PCM is available or not. You're trying to change this scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:
Quoted 7 lines
for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to PCM.
OK, so far, so good.
Quoted 22 lines
Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.
For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin 5. These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.

Quoted 8 lines

Quoted 4 lines
Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to check is its PCM assignment. It doesn't have to issue the jack state check via HDA verb at all. This should make things a lot easier.
I'm now porting the jack supporting for MST audio. The code on hdmi-jack branch will create jack differently based codec_has_acomp() condition. Does this mean MST Jack binding will handle both situation?

Ideally, yes.

Quoted 6 lines

My idea is:
1. Add struct snd_jack *jack[] member in hdmi_spec.
2. Create the jack based on PCM
3. Assign spec->jack[i] to per_pin->acomp_jack when pin is connected.
This should be OK for codec_has_acomp sitation.

Hrm, not really. As already discussed, a jack is associated to each PCM, and notified plug/unplug when it's assigned / unassigned. That is, you don't have to keep extra jack array. When a PCM is assigned or unassigned, simply notify per_pin->acomp_jack of that PCM. That's all. So there shouldn't be too many difference from the current situation.

Quoted 4 lines

But what about the
other situation? The other situation still uses the hda_jack_tbl and etc. So we still need handle hda jack if we want to reuse the current code.

For other, it's similar. The jack is associated with PCM, which is virtually bound with per_pin. And when it's assigned / unassigned, call it with the given per_pin->pin_nid.

Takashi

Yang, Libin

2015-12-22 11:45 UTC #10 lore raw

Quoted 129 lines

-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Tuesday, December 22, 2015 5:00 PM
To: Libin Yang
Cc: Yang, Libin; [email protected]; alsa-devel@alsa- project.org
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 14 lines · Libin Yang
Hi Takashi,
Quoted 10 lines · Takashi Iwai
Quoted 8 lines · Libin
Hi Takashi,
Quoted 5 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected];
Yang,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST
audio for
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Jack support
Quoted 8 lines · Libin Yang
Quoted 5 lines · Takashi Iwai
Quoted 3 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
This is the patch set for Jack support for DP MST audio
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean,
you
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
cannot know which MST dev# is given beforehand, while you
seem
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
trying
Quoted 3 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device
entry 0
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio
driver
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
support will be implemented in the later patches. The patches help
to
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
prepare supporting MST audio.
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether
the
Quoted 12 lines
Quoted 12 lines
Quoted 5 lines
Quoted 2 lines
corresponding PCM is available or not. You're trying to change this scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:
Quoted 7 lines
for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to
PCM.
Quoted 19 lines
Quoted 19 lines
OK, so far, so good.
Quoted 16 lines
Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin
5.
Quoted 5 lines
Quoted 5 lines
Quoted 2 lines
These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.
Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to check
is
Quoted 6 lines
Quoted 2 lines
its PCM assignment. It doesn't have to issue the jack state check via HDA verb at all. This should make things a lot easier.
I'm now porting the jack supporting for MST audio. The code on hdmi-jack branch will create jack differently based codec_has_acomp() condition. Does this mean MST Jack binding will handle both situation?
Ideally, yes.
Quoted 5 lines
My idea is:
1. Add struct snd_jack *jack[] member in hdmi_spec.
2. Create the jack based on PCM
3. Assign spec->jack[i] to per_pin->acomp_jack when pin is connected.

Quoted 8 lines

Quoted 1 line
This should be OK for codec_has_acomp sitation.
Hrm, not really. As already discussed, a jack is associated to each PCM, and notified plug/unplug when it's assigned / unassigned. That is, you don't have to keep extra jack array. When a PCM is assigned or unassigned, simply notify per_pin->acomp_jack of that PCM. That's all. So there shouldn't be too many difference from the current situation.

If so, the snd_jack is dynamically bound to PCM. This means the jack is bound to the pin, and user space must get the pcm-pin mapping to know the jack is bound to which PCM. Then it can decide how to operate on the PCM. Not sure it is compatible with current pulseaudio.

Quoted 12 lines

Quoted 5 lines
But what about the
other situation? The other situation still uses the hda_jack_tbl and etc. So we still need handle hda jack if we want to reuse the current code.
For other, it's similar. The jack is associated with PCM, which is virtually bound with per_pin. And when it's assigned / unassigned, call it with the given per_pin->pin_nid.
Takashi

Takashi Iwai

2015-12-22 12:08 UTC #11 lore raw

Quoted 139 lines · Libin

Quoted 130 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Tuesday, December 22, 2015 5:00 PM
To: Libin Yang
Cc: Yang, Libin; [email protected]; alsa-devel@alsa- project.org
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 14 lines · Libin Yang
Hi Takashi,
Quoted 10 lines · Takashi Iwai
Quoted 8 lines · Libin
Hi Takashi,
Quoted 5 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected];
Yang,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST
audio for
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Jack support
Quoted 8 lines · Libin Yang
Quoted 5 lines · Takashi Iwai
Quoted 3 lines · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
From: Libin Yang <[email protected]>
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
This is the patch set for Jack support for DP MST audio
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to pin
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I mean,
you
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
cannot know which MST dev# is given beforehand, while you
seem
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
trying
Quoted 3 lines
Quoted 1 line
to assign such a number to the jack object.
Oh, sorry I forgot to mention that we only operate on device
entry 0
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio
driver
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
support will be implemented in the later patches. The patches help
to
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
prepare supporting MST audio.
The purpose of the patches is try to dynamically attach the Jack.
Well, only judging from the quick glance, I don't buy this strategy, sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge whether
the
Quoted 12 lines
Quoted 12 lines
Quoted 5 lines
Quoted 2 lines
corresponding PCM is available or not. You're trying to change this scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:
Quoted 7 lines
for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and pin_idx. As a pin is bound statically to the PCM, the jack is statically bound to
PCM.
Quoted 19 lines
Quoted 19 lines
OK, so far, so good.
Quoted 16 lines
Now we are using the dynamic pcm bound. If we use the old code to create the jack, the jack will be bound to pin. So we need the patch, which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM.
So no pin is bound to the Jack. When there is a monitor connected, the pin is bound to a proper PCM. So we should bind the PCM's jack to the pin.
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from pin
5.
Quoted 5 lines
Quoted 5 lines
Quoted 2 lines
These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.
Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to check
is
Quoted 6 lines
Quoted 2 lines
its PCM assignment. It doesn't have to issue the jack state check via HDA verb at all. This should make things a lot easier.
I'm now porting the jack supporting for MST audio. The code on hdmi-jack branch will create jack differently based codec_has_acomp() condition. Does this mean MST Jack binding will handle both situation?
Ideally, yes.
Quoted 5 lines
My idea is:
1. Add struct snd_jack *jack[] member in hdmi_spec.
2. Create the jack based on PCM
3. Assign spec->jack[i] to per_pin->acomp_jack when pin is connected.
Quoted 8 lines
Quoted 1 line
This should be OK for codec_has_acomp sitation.
Hrm, not really. As already discussed, a jack is associated to each PCM, and notified plug/unplug when it's assigned / unassigned. That is, you don't have to keep extra jack array. When a PCM is assigned or unassigned, simply notify per_pin->acomp_jack of that PCM. That's all. So there shouldn't be too many difference from the current situation.

Quoted 5 lines

If so, the snd_jack is dynamically bound to PCM. This means the jack is bound to the pin, and user space must get the pcm-pin mapping to know the jack is bound to which PCM. Then it can decide how to operate on the PCM. Not sure it is compatible with current pulseaudio.

Erm, I was confused. Yes, for dynamic attach/detaching PCM, we need to keep jack object apart from per_pin. I guess your current patchset is broken in this regard?

Though, for having the jack pointer, I prefer avoiding to introduce yet another array but rather extending hdmi_spec.pcm_rec[]. Currently it contains the hda_pcm pointers. But we can define a new object, e.g.

struct hdmi_pcm {
	struct hda_pcm *pcm;
	struct snd_jack *acomp_jack;
	....
};

and keeping this in hdmi_spec.

struct hdmi_spec {
	....
	struct hdmi_pcm pcm_rec[16];
	....
};

get_pcm_rec() macro needs to be adjusted properly, but other than that, there aren't so many places accessing pcm_rec[] directly, so easy to convert.

Maybe we can start from converting this at first, moving acomp_jack to hdmi_spec, then apply MST patchset on its top. This might be a bit cleaner.

In the case of non-component: the question beforehand is how to handle MST PCM assignment at all. Namely, the non-component case assumes that there is 1:1 object matching with the actual pin to per_pin. If this assumption isn't kept, we'd need to use the same trick like the component case, and stop using hda_jack infrastructure.
(So acomp_jack might be better renamed to just "jack" or so.)

thanks,

Takashi

Yang, Libin

2015-12-22 13:33 UTC #12 lore raw

Quoted 201 lines

-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Tuesday, December 22, 2015 8:08 PM
To: Yang, Libin
Cc: Libin Yang; [email protected]; alsa-devel@alsa- project.org
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio for Jack support
Quoted 9 lines · Libin
Quoted 9 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Tuesday, December 22, 2015 5:00 PM
To: Libin Yang
Cc: Yang, Libin; [email protected]; alsa-devel@alsa- project.org
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST audio
for
Quoted 32 lines
Quoted 32 lines
Jack support
Quoted 14 lines · Libin Yang
Hi Takashi,
Quoted 10 lines · Takashi Iwai
Quoted 8 lines · Libin
Hi Takashi,
Quoted 5 lines
-----Original Message-----
From: Takashi Iwai [mailto:[email protected]]
Sent: Monday, November 09, 2015 3:59 PM
To: Libin Yang
Cc: [email protected]; [email protected];
Yang,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Libin
Subject: Re: [alsa-devel] [RFC PATCH 0/2] ALSA: hda - DP MST
audio for
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Jack support
Quoted 6 lines · Libin Yang
Quoted 3 lines · Takashi Iwai
Quoted 1 line · On Tue, 03 Nov 2015 09:42:54 +0100, [email protected] wrote:
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
From: Libin Yang <[email protected]>
This is the patch set for Jack support for DP MST audio
Libin Yang (2):
ALSA: hda - jack support DP MST audio
ALSA: hda - hdmi audio add dynamically jack binding to
pin
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
I'm somehow confused by this patch series.
Is the unsolicited event handling itself changed for MST? I
mean,
Quoted 6 lines
Quoted 6 lines
you
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
cannot know which MST dev# is given beforehand, while you
seem
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
trying
Quoted 2 lines
Quoted 1 line
to assign such a number to the jack object.
Quoted 7 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Oh, sorry I forgot to mention that we only operate on device
entry 0
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
so far in the code. It's misleading.
These patches don't operate on MST audio. The full MST audio
driver
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
support will be implemented in the later patches. The patches
help
Quoted 4 lines
Quoted 4 lines
to
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
prepare supporting MST audio.
The purpose of the patches is try to dynamically attach the
Jack.
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Well, only judging from the quick glance, I don't buy this
strategy,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
sorry. Currently for HDMI/DP, the jack is for PCM, not for pin. Namely, user-space watches the jack notification to judge
whether
Quoted 2 lines
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
corresponding PCM is available or not. You're trying to change
this
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 1 line
scheme completely.
The patch is to make sure the jack is bound to PCM statically. This patch is needed because in old code, it will create jack with:
for (pin_idx = 0; pin_idx < spec->num_pins; pin_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pin_idx);
	...
}
Which means the jack is created based on pin number and
pin_idx.
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
As a pin is bound statically to the PCM, the jack is statically bound
to
Quoted 5 lines
Quoted 5 lines
PCM.
Quoted 4 lines
Quoted 4 lines
OK, so far, so good.
Quoted 1 line
Now we are using the dynamic pcm bound. If we use the old
code to
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
create the jack, the jack will be bound to pin. So we need the
patch,
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
which will use the new code:
for (pcm_idx = 0; pcm_idx < spec-> pcms_used; pcm_idx++) {
	...
	err = generic_hdmi_build_jack(codec, pcm_idx);
	...
}
which makes sure that the jack is bound to PCM (jack[0] is bound to pcm[0], jack[1] is bound to pcm[1] and so on).
If there is no monitor, no pin is bound to the PCM. So no pin is bound to the Jack. When there is a monitor
connected,
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
the pin is bound to a proper PCM. So we should bind the PCM's
jack
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
to the pin.
For example, jack0 is bound to the PCM 3 (the first jack is bound to the first PCM statically). When a monitor is connected to pin 5, PCM 3 will be assigned to the pin. We need bind jack 0 to pin 5. When monitor is disconnected, we must unbind the jack 0 from
pin
Quoted 6 lines
Quoted 6 lines
5.
Quoted 5 lines
Quoted 5 lines
Quoted 2 lines
These operation are done in function
snd_hda_jack_bind/ snd_hda_jack_unbind.
Then we shouldn't use the hda_jack.c code at all for HDMI/DP. For HDMI/DP with the dynamic PCM binding, the only state to
check
Quoted 2 lines
Quoted 2 lines
is
Quoted 1 line
Quoted 1 line
its PCM assignment. It doesn't have to issue the jack state check
via
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 1 line
HDA verb at all. This should make things a lot easier.
I'm now porting the jack supporting for MST audio. The code on hdmi-jack branch will create jack differently based
codec_has_acomp()
Quoted 1 line
Quoted 1 line
Quoted 1 line
condition. Does this mean MST Jack binding will handle both
situation?
Quoted 7 lines
Quoted 7 lines
Ideally, yes.
Quoted 4 lines
My idea is:
1. Add struct snd_jack *jack[] member in hdmi_spec.
2. Create the jack based on PCM
3. Assign spec->jack[i] to per_pin->acomp_jack when pin is
connected.
Quoted 7 lines
Quoted 7 lines
Quoted 2 lines
This should be OK for codec_has_acomp sitation.
Hrm, not really. As already discussed, a jack is associated to each PCM, and notified plug/unplug when it's assigned / unassigned. That is, you don't have to keep extra jack array. When a PCM is assigned or unassigned, simply notify per_pin->acomp_jack of that
PCM.
Quoted 8 lines
Quoted 2 lines
That's all. So there shouldn't be too many difference from the current situation.
If so, the snd_jack is dynamically bound to PCM. This means the jack is bound to the pin, and user space must get the pcm-pin mapping to know the jack is bound to which PCM. Then it can decide how to operate on the PCM. Not sure it is compatible with current pulseaudio.
Erm, I was confused. Yes, for dynamic attach/detaching PCM, we need to keep jack object apart from per_pin. I guess your current patchset is broken in this regard?
Though, for having the jack pointer, I prefer avoiding to introduce yet another array but rather extending hdmi_spec.pcm_rec[]. Currently it contains the hda_pcm pointers. But we can define a new object, e.g.
struct hdmi_pcm {
	struct hda_pcm *pcm;
	struct snd_jack *acomp_jack;
	....
};
and keeping this in hdmi_spec.
struct hdmi_spec {
	....
	struct hdmi_pcm pcm_rec[16];
	....
};

Quoted 7 lines

get_pcm_rec() macro needs to be adjusted properly, but other than that, there aren't so many places accessing pcm_rec[] directly, so easy to convert.
Maybe we can start from converting this at first, moving acomp_jack to hdmi_spec, then apply MST patchset on its top. This might be a bit cleaner.

It's clear now. It should work. I will try it.

Quoted 8 lines

In the case of non-component: the question beforehand is how to handle MST PCM assignment at all. Namely, the non-component case assumes that there is 1:1 object matching with the actual pin to per_pin. If this assumption isn't kept, we'd need to use the same trick like the component case, and stop using hda_jack infrastructure.
(So acomp_jack might be better renamed to just "jack" or so.)

If MST PCM is supported, then per_pin is not 1:1 object matching with the actual pin. It seems we will stop using hda_jack infrastructure.

I will refine the jack patches based on the new method. Thanks for your suggestion.

Regards,
Libin

Quoted 4 lines

thanks,
Takashi