Merge branch 'for-next' into for-linus

41 files changed, 2273 insertions, 1629 deletions

diff --git a/sound/firewire/Kconfig b/sound/firewire/Kconfig
index 9897bd26a438..fd109bea4c53 100644
--- a/sound/firewire/Kconfig
+++ b/sound/firewire/Kconfig
@@ -18,8 +18,25 @@ config SND_DICE
 	select SND_HWDEP
 	select SND_FIREWIRE_LIB
 	help
-	  Say Y here to include support for many DACs based on the DICE
-	  chip family (DICE-II/Jr/Mini) which TC Applied Technologies produces.
+	  Say Y here to include support for devices based on the DICE chip family
+	  (DICE-II/TCD2210(Mini)/TCD2220(Jr.)) which TC Applied Technologies (TCAT) produced.
+	   * Allen and Heath Zed R16
+	   * Alesis iO 14/26 FireWire, MasterControl, MultiMix 8/12/16 FireWire
+	   * Avid Mbox 3 Pro
+	   * FlexRadio Systems FLEX-3000, FLEX-5000
+	   * Focusrite Liquid Saffire 56
+	   * Focusrite Saffire Pro 14, Pro 24, Pro 24 DSP, Pro 26, Pro 40(TCD2220)
+	   * Harman Music Group Lexicon I-ONIX FW810S
+	   * Loud Technologies Mackie Onyx Blackbird, Onyx 820i/1220i/1620i/1640i (latter models)
+	   * M-Audio ProFire 610/2626
+	   * Mytek Stereo192-DSD DAC
+	   * Midas Klark Teknik VeniceF series
+	   * PreSonus FireStudio, FireStudio Mobile, FireStudio Project, FireStudio Tube
+	   * PreSonus StudioLive 16.4.2, 16.0.2, 24.4.2, 32.4.2
+	   * Solid State Logic Duende Classic, Duende Mini
+	   * TC Electronic Studio Konnekt 48, Konnekt 24D, Konnekt Live, Impact Twin
+	   * TC Electronic Digital Konnekt x32, Desktop Konnekt 6
+	   * Weiss Engineering ADC2, Vesta, Minerva, AFI1, DAC1, INT202, DAC202
 
 	  To compile this driver as a module, choose M here: the module
 	  will be called snd-dice.
@@ -38,7 +55,7 @@ config SND_OXFW
 	   * Mackie(Loud) Onyx 1640i (former model)
 	   * Mackie(Loud) Onyx Satellite
 	   * Mackie(Loud) Tapco Link.Firewire
-	   * Mackie(Loud) d.4 pro
+	   * Mackie(Loud) d.2 pro/d.4 pro (built-in FireWire card with OXFW971 ASIC)
 	   * Mackie(Loud) U.420/U.420d
 	   * TASCAM FireOne
 	   * Stanton Controllers & Systems 1 Deck/Mixer
@@ -84,7 +101,7 @@ config SND_BEBOB
 	  * PreSonus FIREBOX/FIREPOD/FP10/Inspire1394
 	  * BridgeCo RDAudio1/Audio5
 	  * Mackie Onyx 1220/1620/1640 (FireWire I/O Card)
-	  * Mackie d.2 (FireWire Option) and d.2 Pro
+	  * Mackie d.2 (optional FireWire card with DM1000 ASIC)
 	  * Stanton FinalScratch 2 (ScratchAmp)
 	  * Tascam IF-FW/DM
 	  * Behringer XENIX UFX 1204/1604
@@ -110,6 +127,7 @@ config SND_BEBOB
 	  * M-Audio Ozonic/NRV10/ProfireLightBridge
 	  * M-Audio FireWire 1814/ProjectMix IO
 	  * Digidesign Mbox 2 Pro
+	  * ToneWeal FW66
 
 	  To compile this driver as a module, choose M here: the module
 	  will be called snd-bebob.
@@ -148,12 +166,16 @@ config SND_FIREWIRE_MOTU
 	select SND_HWDEP
 	help
 	 Say Y here to enable support for FireWire devices which MOTU produced:
+	  * 828
+	  * 896
 	  * 828mk2
 	  * Traveler
 	  * Ultralite
 	  * 8pre
 	  * 828mk3 (FireWire only)
 	  * 828mk3 (Hybrid)
+	  * Ultralite mk3 (FireWire only)
+	  * Ultralite mk3 (Hybrid)
 	  * Audio Express
 	  * 4pre
 
diff --git a/sound/firewire/amdtp-am824.c b/sound/firewire/amdtp-am824.c
index fea92e148790..d9c700f652bb 100644
--- a/sound/firewire/amdtp-am824.c
+++ b/sound/firewire/amdtp-am824.c
@@ -410,10 +410,10 @@ static unsigned int process_ir_ctx_payloads(struct amdtp_stream *s,
  * @s: the AMDTP stream to initialize
  * @unit: the target of the stream
  * @dir: the direction of stream
- * @flags: the packet transmission method to use
+ * @flags: the details of the streaming protocol consist of cip_flags enumeration-constants.
  */
 int amdtp_am824_init(struct amdtp_stream *s, struct fw_unit *unit,
-		     enum amdtp_stream_direction dir, enum cip_flags flags)
+		     enum amdtp_stream_direction dir, unsigned int flags)
 {
 	amdtp_stream_process_ctx_payloads_t process_ctx_payloads;
 
diff --git a/sound/firewire/amdtp-am824.h b/sound/firewire/amdtp-am824.h
index 06d280783581..2b092b1061ba 100644
--- a/sound/firewire/amdtp-am824.h
+++ b/sound/firewire/amdtp-am824.h
@@ -45,5 +45,5 @@ void amdtp_am824_midi_trigger(struct amdtp_stream *s, unsigned int port,
 			      struct snd_rawmidi_substream *midi);
 
 int amdtp_am824_init(struct amdtp_stream *s, struct fw_unit *unit,
-		     enum amdtp_stream_direction dir, enum cip_flags flags);
+		     enum amdtp_stream_direction dir, unsigned int flags);
 #endif
diff --git a/sound/firewire/amdtp-stream-trace.h b/sound/firewire/amdtp-stream-trace.h
index aa53c13b89d3..5fd2aeccdfc2 100644
--- a/sound/firewire/amdtp-stream-trace.h
+++ b/sound/firewire/amdtp-stream-trace.h
@@ -49,7 +49,7 @@ TRACE_EVENT(amdtp_packet,
 		__entry->data_blocks = data_blocks;
 		__entry->data_block_counter = data_block_counter,
 		__entry->packet_index = packet_index;
-		__entry->irq = !!in_interrupt();
+		__entry->irq = !!in_softirq();
 		__entry->index = index;
 	),
 	TP_printk(
diff --git a/sound/firewire/amdtp-stream.c b/sound/firewire/amdtp-stream.c
index 5805c5de39fb..9be2260e4ca2 100644
--- a/sound/firewire/amdtp-stream.c
+++ b/sound/firewire/amdtp-stream.c
@@ -20,7 +20,7 @@
 #define CYCLES_PER_SECOND	8000
 #define TICKS_PER_SECOND	(TICKS_PER_CYCLE * CYCLES_PER_SECOND)
 
-#define OHCI_MAX_SECOND		8
+#define OHCI_SECOND_MODULUS		8
 
 /* Always support Linux tracing subsystem. */
 #define CREATE_TRACE_POINTS
@@ -33,7 +33,8 @@
 #define TAG_NO_CIP_HEADER	0
 #define TAG_CIP			1
 
-/* common isochronous packet header parameters */
+// Common Isochronous Packet (CIP) header parameters. Use two quadlets CIP header when supported.
+#define CIP_HEADER_QUADLETS	2
 #define CIP_EOH_SHIFT		31
 #define CIP_EOH			(1u << CIP_EOH_SHIFT)
 #define CIP_EOH_MASK		0x80000000
@@ -48,36 +49,46 @@
 #define CIP_FMT_MASK		0x3f000000
 #define CIP_FDF_MASK		0x00ff0000
 #define CIP_FDF_SHIFT		16
+#define CIP_FDF_NO_DATA		0xff
 #define CIP_SYT_MASK		0x0000ffff
 #define CIP_SYT_NO_INFO		0xffff
+#define CIP_SYT_CYCLE_MODULUS	16
+#define CIP_NO_DATA		((CIP_FDF_NO_DATA << CIP_FDF_SHIFT) | CIP_SYT_NO_INFO)
+
+#define CIP_HEADER_SIZE		(sizeof(__be32) * CIP_HEADER_QUADLETS)
 
 /* Audio and Music transfer protocol specific parameters */
 #define CIP_FMT_AM		0x10
 #define AMDTP_FDF_NO_DATA	0xff
 
-// For iso header, tstamp and 2 CIP header.
-#define IR_CTX_HEADER_SIZE_CIP		16
 // For iso header and tstamp.
-#define IR_CTX_HEADER_SIZE_NO_CIP	8
+#define IR_CTX_HEADER_DEFAULT_QUADLETS	2
+// Add nothing.
+#define IR_CTX_HEADER_SIZE_NO_CIP	(sizeof(__be32) * IR_CTX_HEADER_DEFAULT_QUADLETS)
+// Add two quadlets CIP header.
+#define IR_CTX_HEADER_SIZE_CIP		(IR_CTX_HEADER_SIZE_NO_CIP + CIP_HEADER_SIZE)
 #define HEADER_TSTAMP_MASK	0x0000ffff
 
-#define IT_PKT_HEADER_SIZE_CIP		8 // For 2 CIP header.
+#define IT_PKT_HEADER_SIZE_CIP		CIP_HEADER_SIZE
 #define IT_PKT_HEADER_SIZE_NO_CIP	0 // Nothing.
 
-static void pcm_period_work(struct work_struct *work);
+// The initial firmware of OXFW970 can postpone transmission of packet during finishing
+// asynchronous transaction. This module accepts 5 cycles to skip as maximum to avoid buffer
+// overrun. Actual device can skip more, then this module stops the packet streaming.
+#define IR_JUMBO_PAYLOAD_MAX_SKIP_CYCLES	5
 
 /**
  * amdtp_stream_init - initialize an AMDTP stream structure
  * @s: the AMDTP stream to initialize
  * @unit: the target of the stream
  * @dir: the direction of stream
- * @flags: the packet transmission method to use
+ * @flags: the details of the streaming protocol consist of cip_flags enumeration-constants.
  * @fmt: the value of fmt field in CIP header
  * @process_ctx_payloads: callback handler to process payloads of isoc context
  * @protocol_size: the size to allocate newly for protocol
  */
 int amdtp_stream_init(struct amdtp_stream *s, struct fw_unit *unit,
-		      enum amdtp_stream_direction dir, enum cip_flags flags,
+		      enum amdtp_stream_direction dir, unsigned int flags,
 		      unsigned int fmt,
 		      amdtp_stream_process_ctx_payloads_t process_ctx_payloads,
 		      unsigned int protocol_size)
@@ -94,18 +105,13 @@ int amdtp_stream_init(struct amdtp_stream *s, struct fw_unit *unit,
 	s->flags = flags;
 	s->context = ERR_PTR(-1);
 	mutex_init(&s->mutex);
-	INIT_WORK(&s->period_work, pcm_period_work);
 	s->packet_index = 0;
 
-	init_waitqueue_head(&s->callback_wait);
-	s->callbacked = false;
+	init_waitqueue_head(&s->ready_wait);
 
 	s->fmt = fmt;
 	s->process_ctx_payloads = process_ctx_payloads;
 
-	if (dir == AMDTP_OUT_STREAM)
-		s->ctx_data.rx.syt_override = -1;
-
 	return 0;
 }
 EXPORT_SYMBOL(amdtp_stream_init);
@@ -183,14 +189,13 @@ int amdtp_stream_add_pcm_hw_constraints(struct amdtp_stream *s,
 	unsigned int maximum_usec_per_period;
 	int err;
 
-	hw->info = SNDRV_PCM_INFO_BATCH |
-		   SNDRV_PCM_INFO_BLOCK_TRANSFER |
+	hw->info = SNDRV_PCM_INFO_BLOCK_TRANSFER |
 		   SNDRV_PCM_INFO_INTERLEAVED |
 		   SNDRV_PCM_INFO_JOINT_DUPLEX |
 		   SNDRV_PCM_INFO_MMAP |
-		   SNDRV_PCM_INFO_MMAP_VALID;
+		   SNDRV_PCM_INFO_MMAP_VALID |
+		   SNDRV_PCM_INFO_NO_PERIOD_WAKEUP;
 
-	/* SNDRV_PCM_INFO_BATCH */
 	hw->periods_min = 2;
 	hw->periods_max = UINT_MAX;
 
@@ -287,22 +292,29 @@ int amdtp_stream_set_parameters(struct amdtp_stream *s, unsigned int rate,
 	s->syt_interval = amdtp_syt_intervals[sfc];
 
 	// default buffering in the device.
-	if (s->direction == AMDTP_OUT_STREAM) {
-		s->ctx_data.rx.transfer_delay =
-					TRANSFER_DELAY_TICKS - TICKS_PER_CYCLE;
-
-		if (s->flags & CIP_BLOCKING) {
-			// additional buffering needed to adjust for no-data
-			// packets.
-			s->ctx_data.rx.transfer_delay +=
-				TICKS_PER_SECOND * s->syt_interval / rate;
-		}
-	}
+	s->transfer_delay = TRANSFER_DELAY_TICKS - TICKS_PER_CYCLE;
+
+	// additional buffering needed to adjust for no-data packets.
+	if (s->flags & CIP_BLOCKING)
+		s->transfer_delay += TICKS_PER_SECOND * s->syt_interval / rate;
 
 	return 0;
 }
 EXPORT_SYMBOL(amdtp_stream_set_parameters);
 
+// The CIP header is processed in context header apart from context payload.
+static int amdtp_stream_get_max_ctx_payload_size(struct amdtp_stream *s)
+{
+	unsigned int multiplier;
+
+	if (s->flags & CIP_JUMBO_PAYLOAD)
+		multiplier = IR_JUMBO_PAYLOAD_MAX_SKIP_CYCLES;
+	else
+		multiplier = 1;
+
+	return s->syt_interval * s->data_block_quadlets * sizeof(__be32) * multiplier;
+}
+
 /**
  * amdtp_stream_get_max_payload - get the stream's packet size
  * @s: the AMDTP stream
@@ -312,16 +324,14 @@ EXPORT_SYMBOL(amdtp_stream_set_parameters);
  */
 unsigned int amdtp_stream_get_max_payload(struct amdtp_stream *s)
 {
-	unsigned int multiplier = 1;
-	unsigned int cip_header_size = 0;
+	unsigned int cip_header_size;
 
-	if (s->flags & CIP_JUMBO_PAYLOAD)
-		multiplier = 5;
 	if (!(s->flags & CIP_NO_HEADER))
-		cip_header_size = sizeof(__be32) * 2;
+		cip_header_size = CIP_HEADER_SIZE;
+	else
+		cip_header_size = 0;
 
-	return cip_header_size +
-		s->syt_interval * s->data_block_quadlets * sizeof(__be32) * multiplier;
+	return cip_header_size + amdtp_stream_get_max_ctx_payload_size(s);
 }
 EXPORT_SYMBOL(amdtp_stream_get_max_payload);
 
@@ -333,32 +343,46 @@ EXPORT_SYMBOL(amdtp_stream_get_max_payload);
  */
 void amdtp_stream_pcm_prepare(struct amdtp_stream *s)
 {
-	cancel_work_sync(&s->period_work);
 	s->pcm_buffer_pointer = 0;
 	s->pcm_period_pointer = 0;
 }
 EXPORT_SYMBOL(amdtp_stream_pcm_prepare);
 
-static unsigned int calculate_data_blocks(unsigned int *data_block_state,
-				bool is_blocking, bool is_no_info,
-				unsigned int syt_interval, enum cip_sfc sfc)
+static void pool_blocking_data_blocks(struct amdtp_stream *s, struct seq_desc *descs,
+				      const unsigned int seq_size, unsigned int seq_tail,
+				      unsigned int count)
 {
-	unsigned int data_blocks;
+	const unsigned int syt_interval = s->syt_interval;
+	int i;
+
+	for (i = 0; i < count; ++i) {
+		struct seq_desc *desc = descs + seq_tail;
 
-	/* Blocking mode. */
-	if (is_blocking) {
-		/* This module generate empty packet for 'no data'. */
-		if (is_no_info)
-			data_blocks = 0;
+		if (desc->syt_offset != CIP_SYT_NO_INFO)
+			desc->data_blocks = syt_interval;
 		else
-			data_blocks = syt_interval;
-	/* Non-blocking mode. */
-	} else {
+			desc->data_blocks = 0;
+
+		seq_tail = (seq_tail + 1) % seq_size;
+	}
+}
+
+static void pool_ideal_nonblocking_data_blocks(struct amdtp_stream *s, struct seq_desc *descs,
+					       const unsigned int seq_size, unsigned int seq_tail,
+					       unsigned int count)
+{
+	const enum cip_sfc sfc = s->sfc;
+	unsigned int state = s->ctx_data.rx.data_block_state;
+	int i;
+
+	for (i = 0; i < count; ++i) {
+		struct seq_desc *desc = descs + seq_tail;
+
 		if (!cip_sfc_is_base_44100(sfc)) {
 			// Sample_rate / 8000 is an integer, and precomputed.
-			data_blocks = *data_block_state;
+			desc->data_blocks = state;
 		} else {
-			unsigned int phase = *data_block_state;
+			unsigned int phase = state;
 
 		/*
 		 * This calculates the number of data blocks per packet so that
@@ -370,18 +394,19 @@ static unsigned int calculate_data_blocks(unsigned int *data_block_state,
 		 */
 			if (sfc == CIP_SFC_44100)
 				/* 6 6 5 6 5 6 5 ... */
-				data_blocks = 5 + ((phase & 1) ^
-						   (phase == 0 || phase >= 40));
+				desc->data_blocks = 5 + ((phase & 1) ^ (phase == 0 || phase >= 40));
 			else
 				/* 12 11 11 11 11 ... or 23 22 22 22 22 ... */
-				data_blocks = 11 * (sfc >> 1) + (phase == 0);
+				desc->data_blocks = 11 * (sfc >> 1) + (phase == 0);
 			if (++phase >= (80 >> (sfc >> 1)))
 				phase = 0;
-			*data_block_state = phase;
+			state = phase;
 		}
+
+		seq_tail = (seq_tail + 1) % seq_size;
 	}
 
-	return data_blocks;
+	s->ctx_data.rx.data_block_state = state;
 }
 
 static unsigned int calculate_syt_offset(unsigned int *last_syt_offset,
@@ -423,6 +448,149 @@ static unsigned int calculate_syt_offset(unsigned int *last_syt_offset,
 	return syt_offset;
 }
 
+static void pool_ideal_syt_offsets(struct amdtp_stream *s, struct seq_desc *descs,
+				   const unsigned int seq_size, unsigned int seq_tail,
+				   unsigned int count)
+{
+	const enum cip_sfc sfc = s->sfc;
+	unsigned int last = s->ctx_data.rx.last_syt_offset;
+	unsigned int state = s->ctx_data.rx.syt_offset_state;
+	int i;
+
+	for (i = 0; i < count; ++i) {
+		struct seq_desc *desc = descs + seq_tail;
+
+		desc->syt_offset = calculate_syt_offset(&last, &state, sfc);
+
+		seq_tail = (seq_tail + 1) % seq_size;
+	}
+
+	s->ctx_data.rx.last_syt_offset = last;
+	s->ctx_data.rx.syt_offset_state = state;
+}
+
+static unsigned int compute_syt_offset(unsigned int syt, unsigned int cycle,
+				       unsigned int transfer_delay)
+{
+	unsigned int cycle_lo = (cycle % CYCLES_PER_SECOND) & 0x0f;
+	unsigned int syt_cycle_lo = (syt & 0xf000) >> 12;
+	unsigned int syt_offset;
+
+	// Round up.
+	if (syt_cycle_lo < cycle_lo)
+		syt_cycle_lo += CIP_SYT_CYCLE_MODULUS;
+	syt_cycle_lo -= cycle_lo;
+
+	// Subtract transfer delay so that the synchronization offset is not so large
+	// at transmission.
+	syt_offset = syt_cycle_lo * TICKS_PER_CYCLE + (syt & 0x0fff);
+	if (syt_offset < transfer_delay)
+		syt_offset += CIP_SYT_CYCLE_MODULUS * TICKS_PER_CYCLE;
+
+	return syt_offset - transfer_delay;
+}
+
+// Both of the producer and consumer of the queue runs in the same clock of IEEE 1394 bus.
+// Additionally, the sequence of tx packets is severely checked against any discontinuity
+// before filling entries in the queue. The calculation is safe even if it looks fragile by
+// overrun.
+static unsigned int calculate_cached_cycle_count(struct amdtp_stream *s, unsigned int head)
+{
+	const unsigned int cache_size = s->ctx_data.tx.cache.size;
+	unsigned int cycles = s->ctx_data.tx.cache.tail;
+
+	if (cycles < head)
+		cycles += cache_size;
+	cycles -= head;
+
+	return cycles;
+}
+
+static void cache_seq(struct amdtp_stream *s, const struct pkt_desc *descs, unsigned int desc_count)
+{
+	const unsigned int transfer_delay = s->transfer_delay;
+	const unsigned int cache_size = s->ctx_data.tx.cache.size;
+	struct seq_desc *cache = s->ctx_data.tx.cache.descs;
+	unsigned int cache_tail = s->ctx_data.tx.cache.tail;
+	bool aware_syt = !(s->flags & CIP_UNAWARE_SYT);
+	int i;
+
+	for (i = 0; i < desc_count; ++i) {
+		struct seq_desc *dst = cache + cache_tail;
+		const struct pkt_desc *src = descs + i;
+
+		if (aware_syt && src->syt != CIP_SYT_NO_INFO)
+			dst->syt_offset = compute_syt_offset(src->syt, src->cycle, transfer_delay);
+		else
+			dst->syt_offset = CIP_SYT_NO_INFO;
+		dst->data_blocks = src->data_blocks;
+
+		cache_tail = (cache_tail + 1) % cache_size;
+	}
+
+	s->ctx_data.tx.cache.tail = cache_tail;
+}
+
+static void pool_ideal_seq_descs(struct amdtp_stream *s, unsigned int count)
+{
+	struct seq_desc *descs = s->ctx_data.rx.seq.descs;
+	unsigned int seq_tail = s->ctx_data.rx.seq.tail;
+	const unsigned int seq_size = s->ctx_data.rx.seq.size;
+
+	pool_ideal_syt_offsets(s, descs, seq_size, seq_tail, count);
+
+	if (s->flags & CIP_BLOCKING)
+		pool_blocking_data_blocks(s, descs, seq_size, seq_tail, count);
+	else
+		pool_ideal_nonblocking_data_blocks(s, descs, seq_size, seq_tail, count);
+
+	s->ctx_data.rx.seq.tail = (seq_tail + count) % seq_size;
+}
+
+static void pool_replayed_seq(struct amdtp_stream *s, unsigned int count)
+{
+	struct amdtp_stream *target = s->ctx_data.rx.replay_target;
+	const struct seq_desc *cache = target->ctx_data.tx.cache.descs;
+	const unsigned int cache_size = target->ctx_data.tx.cache.size;
+	unsigned int cache_head = s->ctx_data.rx.cache_head;
+	struct seq_desc *descs = s->ctx_data.rx.seq.descs;
+	const unsigned int seq_size = s->ctx_data.rx.seq.size;
+	unsigned int seq_tail = s->ctx_data.rx.seq.tail;
+	int i;
+
+	for (i = 0; i < count; ++i) {
+		descs[seq_tail] = cache[cache_head];
+		seq_tail = (seq_tail + 1) % seq_size;
+		cache_head = (cache_head + 1) % cache_size;
+	}
+
+	s->ctx_data.rx.seq.tail = seq_tail;
+	s->ctx_data.rx.cache_head = cache_head;
+}
+
+static void pool_seq_descs(struct amdtp_stream *s, unsigned int count)
+{
+	struct amdtp_domain *d = s->domain;
+
+	if (!d->replay.enable || !s->ctx_data.rx.replay_target) {
+		pool_ideal_seq_descs(s, count);
+	} else {
+		if (!d->replay.on_the_fly) {
+			pool_replayed_seq(s, count);
+		} else {
+			struct amdtp_stream *tx = s->ctx_data.rx.replay_target;
+			const unsigned int cache_size = tx->ctx_data.tx.cache.size;
+			const unsigned int cache_head = s->ctx_data.rx.cache_head;
+			unsigned int cached_cycles = calculate_cached_cycle_count(tx, cache_head);
+
+			if (cached_cycles > count && cached_cycles > cache_size / 2)
+				pool_replayed_seq(s, count);
+			else
+				pool_ideal_seq_descs(s, count);
+		}
+	}
+}
+
 static void update_pcm_pointers(struct amdtp_stream *s,
 				struct snd_pcm_substream *pcm,
 				unsigned int frames)
@@ -437,18 +605,21 @@ static void update_pcm_pointers(struct amdtp_stream *s,
 	s->pcm_period_pointer += frames;
 	if (s->pcm_period_pointer >= pcm->runtime->period_size) {
 		s->pcm_period_pointer -= pcm->runtime->period_size;
-		queue_work(system_highpri_wq, &s->period_work);
-	}
-}
 
-static void pcm_period_work(struct work_struct *work)
-{
-	struct amdtp_stream *s = container_of(work, struct amdtp_stream,
-					      period_work);
-	struct snd_pcm_substream *pcm = READ_ONCE(s->pcm);
-
-	if (pcm)
-		snd_pcm_period_elapsed(pcm);
+		// The program in user process should periodically check the status of intermediate
+		// buffer associated to PCM substream to process PCM frames in the buffer, instead
+		// of receiving notification of period elapsed by poll wait.
+		if (!pcm->runtime->no_period_wakeup) {
+			if (in_softirq()) {
+				// In software IRQ context for 1394 OHCI.
+				snd_pcm_period_elapsed(pcm);
+			} else {
+				// In process context of ALSA PCM application under acquired lock of
+				// PCM substream.
+				snd_pcm_period_elapsed_under_stream_lock(pcm);
+			}
+		}
+	}
 }
 
 static int queue_packet(struct amdtp_stream *s, struct fw_iso_packet *params,
@@ -505,7 +676,7 @@ static void generate_cip_header(struct amdtp_stream *s, __be32 cip_header[2],
 }
 
 static void build_it_pkt_header(struct amdtp_stream *s, unsigned int cycle,
-				struct fw_iso_packet *params,
+				struct fw_iso_packet *params, unsigned int header_length,
 				unsigned int data_blocks,
 				unsigned int data_block_counter,
 				unsigned int syt, unsigned int index)
@@ -516,16 +687,15 @@ static void build_it_pkt_header(struct amdtp_stream *s, unsigned int cycle,
 	payload_length = data_blocks * sizeof(__be32) * s->data_block_quadlets;
 	params->payload_length = payload_length;
 
-	if (!(s->flags & CIP_NO_HEADER)) {
+	if (header_length > 0) {
 		cip_header = (__be32 *)params->header;
 		generate_cip_header(s, cip_header, data_block_counter, syt);
-		params->header_length = 2 * sizeof(__be32);
-		payload_length += params->header_length;
+		params->header_length = header_length;
 	} else {
 		cip_header = NULL;
 	}
 
-	trace_amdtp_packet(s, cycle, cip_header, payload_length, data_blocks,
+	trace_amdtp_packet(s, cycle, cip_header, payload_length + header_length, data_blocks,
 			   data_block_counter, s->packet_index, index);
 }
 
@@ -569,8 +739,7 @@ static int check_cip_header(struct amdtp_stream *s, const __be32 *buf,
 
 	/* Calculate data blocks */
 	fdf = (cip_header[1] & CIP_FDF_MASK) >> CIP_FDF_SHIFT;
-	if (payload_length < sizeof(__be32) * 2 ||
-	    (fmt == CIP_FMT_AM && fdf == AMDTP_FDF_NO_DATA)) {
+	if (payload_length == 0 || (fmt == CIP_FMT_AM && fdf == AMDTP_FDF_NO_DATA)) {
 		*data_blocks = 0;
 	} else {
 		unsigned int data_block_quadlets =
@@ -585,8 +754,7 @@ static int check_cip_header(struct amdtp_stream *s, const __be32 *buf,
 		if (s->flags & CIP_WRONG_DBS)
 			data_block_quadlets = s->data_block_quadlets;
 
-		*data_blocks = (payload_length / sizeof(__be32) - 2) /
-							data_block_quadlets;
+		*data_blocks = payload_length / sizeof(__be32) / data_block_quadlets;
 	}
 
 	/* Check data block counter continuity */
@@ -620,109 +788,163 @@ static int check_cip_header(struct amdtp_stream *s, const __be32 *buf,
 
 	*data_block_counter = dbc;
 
-	*syt = cip_header[1] & CIP_SYT_MASK;
+	if (!(s->flags & CIP_UNAWARE_SYT))
+		*syt = cip_header[1] & CIP_SYT_MASK;
 
 	return 0;
 }
 
 static int parse_ir_ctx_header(struct amdtp_stream *s, unsigned int cycle,
 			       const __be32 *ctx_header,
-			       unsigned int *payload_length,
 			       unsigned int *data_blocks,
 			       unsigned int *data_block_counter,
 			       unsigned int *syt, unsigned int packet_index, unsigned int index)
 {
+	unsigned int payload_length;
 	const __be32 *cip_header;
 	unsigned int cip_header_size;
-	int err;
 
-	*payload_length = be32_to_cpu(ctx_header[0]) >> ISO_DATA_LENGTH_SHIFT;
+	payload_length = be32_to_cpu(ctx_header[0]) >> ISO_DATA_LENGTH_SHIFT;
 
 	if (!(s->flags & CIP_NO_HEADER))
-		cip_header_size = 8;
+		cip_header_size = CIP_HEADER_SIZE;
 	else
 		cip_header_size = 0;
 
-	if (*payload_length > cip_header_size + s->ctx_data.tx.max_ctx_payload_length) {
+	if (payload_length > cip_header_size + s->ctx_data.tx.max_ctx_payload_length) {
 		dev_err(&s->unit->device,
 			"Detect jumbo payload: %04x %04x\n",
-			*payload_length, cip_header_size + s->ctx_data.tx.max_ctx_payload_length);
+			payload_length, cip_header_size + s->ctx_data.tx.max_ctx_payload_length);
 		return -EIO;
 	}
 
 	if (cip_header_size > 0) {
-		cip_header = ctx_header + 2;
-		err = check_cip_header(s, cip_header, *payload_length,
-				       data_blocks, data_block_counter, syt);
-		if (err < 0)
-			return err;
+		if (payload_length >= cip_header_size) {
+			int err;
+
+			cip_header = ctx_header + IR_CTX_HEADER_DEFAULT_QUADLETS;
+			err = check_cip_header(s, cip_header, payload_length - cip_header_size,
+					       data_blocks, data_block_counter, syt);
+			if (err < 0)
+				return err;
+		} else {
+			// Handle the cycle so that empty packet arrives.
+			cip_header = NULL;
+			*data_blocks = 0;
+			*syt = 0;
+		}
 	} else {
 		cip_header = NULL;
-		err = 0;
-		*data_blocks = *payload_length / sizeof(__be32) /
-			       s->data_block_quadlets;
+		*data_blocks = payload_length / sizeof(__be32) / s->data_block_quadlets;
 		*syt = 0;
 
 		if (*data_block_counter == UINT_MAX)
 			*data_block_counter = 0;
 	}
 
-	trace_amdtp_packet(s, cycle, cip_header, *payload_length, *data_blocks,
+	trace_amdtp_packet(s, cycle, cip_header, payload_length, *data_blocks,
 			   *data_block_counter, packet_index, index);
 
-	return err;
+	return 0;
 }
 
 // In CYCLE_TIMER register of IEEE 1394, 7 bits are used to represent second. On
 // the other hand, in DMA descriptors of 1394 OHCI, 3 bits are used to represent
 // it. Thus, via Linux firewire subsystem, we can get the 3 bits for second.
-static inline u32 compute_cycle_count(__be32 ctx_header_tstamp)
+static inline u32 compute_ohci_cycle_count(__be32 ctx_header_tstamp)
 {
 	u32 tstamp = be32_to_cpu(ctx_header_tstamp) & HEADER_TSTAMP_MASK;
 	return (((tstamp >> 13) & 0x07) * 8000) + (tstamp & 0x1fff);
 }
 
-static inline u32 increment_cycle_count(u32 cycle, unsigned int addend)
+static inline u32 increment_ohci_cycle_count(u32 cycle, unsigned int addend)
 {
 	cycle += addend;
-	if (cycle >= OHCI_MAX_SECOND * CYCLES_PER_SECOND)
-		cycle -= OHCI_MAX_SECOND * CYCLES_PER_SECOND;
+	if (cycle >= OHCI_SECOND_MODULUS * CYCLES_PER_SECOND)
+		cycle -= OHCI_SECOND_MODULUS * CYCLES_PER_SECOND;
 	return cycle;
 }
 
+static int compare_ohci_cycle_count(u32 lval, u32 rval)
+{
+	if (lval == rval)
+		return 0;
+	else if (lval < rval && rval - lval < OHCI_SECOND_MODULUS * CYCLES_PER_SECOND / 2)
+		return -1;
+	else
+		return 1;
+}
+
 // Align to actual cycle count for the packet which is going to be scheduled.
 // This module queued the same number of isochronous cycle as the size of queue
 // to kip isochronous cycle, therefore it's OK to just increment the cycle by
 // the size of queue for scheduled cycle.
-static inline u32 compute_it_cycle(const __be32 ctx_header_tstamp,
-				   unsigned int queue_size)
+static inline u32 compute_ohci_it_cycle(const __be32 ctx_header_tstamp,
+					unsigned int queue_size)
 {
-	u32 cycle = compute_cycle_count(ctx_header_tstamp);
-	return increment_cycle_count(cycle, queue_size);
+	u32 cycle = compute_ohci_cycle_count(ctx_header_tstamp);
+	return increment_ohci_cycle_count(cycle, queue_size);
 }
 
 static int generate_device_pkt_descs(struct amdtp_stream *s,
 				     struct pkt_desc *descs,
 				     const __be32 *ctx_header,
-				     unsigned int packets)
+				     unsigned int packets,
+				     unsigned int *desc_count)
 {
+	unsigned int next_cycle = s->next_cycle;
 	unsigned int dbc = s->data_block_counter;
 	unsigned int packet_index = s->packet_index;
 	unsigned int queue_size = s->queue_size;
 	int i;
 	int err;
 
+	*desc_count = 0;
 	for (i = 0; i < packets; ++i) {
-		struct pkt_desc *desc = descs + i;
+		struct pkt_desc *desc = descs + *desc_count;
 		unsigned int cycle;
-		unsigned int payload_length;
+		bool lost;
 		unsigned int data_blocks;
 		unsigned int syt;
 
-		cycle = compute_cycle_count(ctx_header[1]);
+		cycle = compute_ohci_cycle_count(ctx_header[1]);
+		lost = (next_cycle != cycle);
+		if (lost) {
+			if (s->flags & CIP_NO_HEADER) {
+				// Fireface skips transmission just for an isoc cycle corresponding
+				// to empty packet.
+				unsigned int prev_cycle = next_cycle;
+
+				next_cycle = increment_ohci_cycle_count(next_cycle, 1);
+				lost = (next_cycle != cycle);
+				if (!lost) {
+					// Prepare a description for the skipped cycle for
+					// sequence replay.
+					desc->cycle = prev_cycle;
+					desc->syt = 0;
+					desc->data_blocks = 0;
+					desc->data_block_counter = dbc;
+					desc->ctx_payload = NULL;
+					++desc;
+					++(*desc_count);
+				}
+			} else if (s->flags & CIP_JUMBO_PAYLOAD) {
+				// OXFW970 skips transmission for several isoc cycles during
+				// asynchronous transaction. The sequence replay is impossible due
+				// to the reason.
+				unsigned int safe_cycle = increment_ohci_cycle_count(next_cycle,
+								IR_JUMBO_PAYLOAD_MAX_SKIP_CYCLES);
+				lost = (compare_ohci_cycle_count(safe_cycle, cycle) > 0);
+			}
+			if (lost) {
+				dev_err(&s->unit->device, "Detect discontinuity of cycle: %d %d\n",
+					next_cycle, cycle);
+				return -EIO;
+			}
+		}
 
-		err = parse_ir_ctx_header(s, cycle, ctx_header, &payload_length,
-					  &data_blocks, &dbc, &syt, packet_index, i);
+		err = parse_ir_ctx_header(s, cycle, ctx_header, &data_blocks, &dbc, &syt,
+					  packet_index, i);
 		if (err < 0)
 			return err;
 
@@ -735,12 +957,13 @@ static int generate_device_pkt_descs(struct amdtp_stream *s,
 		if (!(s->flags & CIP_DBC_IS_END_EVENT))
 			dbc = (dbc + desc->data_blocks) & 0xff;
 
-		ctx_header +=
-			s->ctx_data.tx.ctx_header_size / sizeof(*ctx_header);
-
+		next_cycle = increment_ohci_cycle_count(next_cycle, 1);
+		++(*desc_count);
+		ctx_header += s->ctx_data.tx.ctx_header_size / sizeof(*ctx_header);
 		packet_index = (packet_index + 1) % queue_size;
 	}
 
+	s->next_cycle = next_cycle;
 	s->data_block_counter = dbc;
 
 	return 0;
@@ -757,29 +980,28 @@ static unsigned int compute_syt(unsigned int syt_offset, unsigned int cycle,
 	return syt & CIP_SYT_MASK;
 }
 
-static void generate_pkt_descs(struct amdtp_stream *s, struct pkt_desc *descs,
-			       const __be32 *ctx_header, unsigned int packets,
-			       const struct seq_desc *seq_descs,
-			       unsigned int seq_size)
+static void generate_pkt_descs(struct amdtp_stream *s, const __be32 *ctx_header, unsigned int packets)
 {
+	struct pkt_desc *descs = s->pkt_descs;
+	const struct seq_desc *seq_descs = s->ctx_data.rx.seq.descs;
+	const unsigned int seq_size = s->ctx_data.rx.seq.size;
 	unsigned int dbc = s->data_block_counter;
-	unsigned int seq_index = s->ctx_data.rx.seq_index;
+	unsigned int seq_head = s->ctx_data.rx.seq.head;
+	bool aware_syt = !(s->flags & CIP_UNAWARE_SYT);
 	int i;
 
 	for (i = 0; i < packets; ++i) {
 		struct pkt_desc *desc = descs + i;
 		unsigned int index = (s->packet_index + i) % s->queue_size;
-		const struct seq_desc *seq = seq_descs + seq_index;
-		unsigned int syt;
+		const struct seq_desc *seq = seq_descs + seq_head;
 
-		desc->cycle = compute_it_cycle(*ctx_header, s->queue_size);
+		desc->cycle = compute_ohci_it_cycle(*ctx_header, s->queue_size);
+
+		if (aware_syt && seq->syt_offset != CIP_SYT_NO_INFO)
+			desc->syt = compute_syt(seq->syt_offset, desc->cycle, s->transfer_delay);
+		else
+			desc->syt = CIP_SYT_NO_INFO;
 
-		syt = seq->syt_offset;
-		if (syt != CIP_SYT_NO_INFO) {
-			syt = compute_syt(syt, desc->cycle,
-					  s->ctx_data.rx.transfer_delay);
-		}
-		desc->syt = syt;
 		desc->data_blocks = seq->data_blocks;
 
 		if (s->flags & CIP_DBC_IS_END_EVENT)
@@ -792,19 +1014,19 @@ static void generate_pkt_descs(struct amdtp_stream *s, struct pkt_desc *descs,
 
 		desc->ctx_payload = s->buffer.packets[index].buffer;
 
-		seq_index = (seq_index + 1) % seq_size;
+		seq_head = (seq_head + 1) % seq_size;
 
 		++ctx_header;
 	}
 
 	s->data_block_counter = dbc;
-	s->ctx_data.rx.seq_index = seq_index;
+	s->ctx_data.rx.seq.head = seq_head;
 }
 
 static inline void cancel_stream(struct amdtp_stream *s)
 {
 	s->packet_index = -1;
-	if (in_interrupt())
+	if (in_softirq())
 		amdtp_stream_pcm_abort(s);
 	WRITE_ONCE(s->pcm_buffer_pointer, SNDRV_PCM_POS_XRUN);
 }
@@ -822,16 +1044,17 @@ static void process_ctx_payloads(struct amdtp_stream *s,
 		update_pcm_pointers(s, pcm, pcm_frames);
 }
 
-static void out_stream_callback(struct fw_iso_context *context, u32 tstamp,
-				size_t header_length, void *header,
-				void *private_data)
+static void process_rx_packets(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+			       void *header, void *private_data)
 {
 	struct amdtp_stream *s = private_data;
 	const struct amdtp_domain *d = s->domain;
 	const __be32 *ctx_header = header;
-	unsigned int events_per_period = s->ctx_data.rx.events_per_period;
+	const unsigned int events_per_period = d->events_per_period;
 	unsigned int event_count = s->ctx_data.rx.event_count;
+	unsigned int pkt_header_length;
 	unsigned int packets;
+	bool need_hw_irq;
 	int i;
 
 	if (s->packet_index < 0)
@@ -840,34 +1063,44 @@ static void out_stream_callback(struct fw_iso_context *context, u32 tstamp,
 	// Calculate the number of packets in buffer and check XRUN.
 	packets = header_length / sizeof(*ctx_header);
 
-	generate_pkt_descs(s, s->pkt_descs, ctx_header, packets, d->seq_descs,
-			   d->seq_size);
+	pool_seq_descs(s, packets);
+
+	generate_pkt_descs(s, ctx_header, packets);
 
 	process_ctx_payloads(s, s->pkt_descs, packets);
 
+	if (!(s->flags & CIP_NO_HEADER))
+		pkt_header_length = IT_PKT_HEADER_SIZE_CIP;
+	else
+		pkt_header_length = 0;
+
+	if (s == d->irq_target) {
+		// At NO_PERIOD_WAKEUP mode, the packets for all IT/IR contexts are processed by
+		// the tasks of user process operating ALSA PCM character device by calling ioctl(2)
+		// with some requests, instead of scheduled hardware IRQ of an IT context.
+		struct snd_pcm_substream *pcm = READ_ONCE(s->pcm);
+		need_hw_irq = !pcm || !pcm->runtime->no_period_wakeup;
+	} else {
+		need_hw_irq = false;
+	}
+
 	for (i = 0; i < packets; ++i) {
 		const struct pkt_desc *desc = s->pkt_descs + i;
-		unsigned int syt;
 		struct {
 			struct fw_iso_packet params;
-			__be32 header[IT_PKT_HEADER_SIZE_CIP / sizeof(__be32)];
+			__be32 header[CIP_HEADER_QUADLETS];
 		} template = { {0}, {0} };
 		bool sched_irq = false;
 
-		if (s->ctx_data.rx.syt_override < 0)
-			syt = desc->syt;
-		else
-			syt = s->ctx_data.rx.syt_override;
-
-		build_it_pkt_header(s, desc->cycle, &template.params,
+		build_it_pkt_header(s, desc->cycle, &template.params, pkt_header_length,
 				    desc->data_blocks, desc->data_block_counter,
-				    syt, i);
+				    desc->syt, i);
 
 		if (s == s->domain->irq_target) {
 			event_count += desc->data_blocks;
 			if (event_count >= events_per_period) {
 				event_count -= events_per_period;
-				sched_irq = true;
+				sched_irq = need_hw_irq;
 			}
 		}
 
@@ -880,13 +1113,99 @@ static void out_stream_callback(struct fw_iso_context *context, u32 tstamp,
 	s->ctx_data.rx.event_count = event_count;
 }
 
-static void in_stream_callback(struct fw_iso_context *context, u32 tstamp,
-			       size_t header_length, void *header,
-			       void *private_data)
+static void skip_rx_packets(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+			    void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+	const __be32 *ctx_header = header;
+	unsigned int packets;
+	unsigned int cycle;
+	int i;
+
+	if (s->packet_index < 0)
+		return;
+
+	packets = header_length / sizeof(*ctx_header);
+
+	cycle = compute_ohci_it_cycle(ctx_header[packets - 1], s->queue_size);
+	s->next_cycle = increment_ohci_cycle_count(cycle, 1);
+
+	for (i = 0; i < packets; ++i) {
+		struct fw_iso_packet params = {
+			.header_length = 0,
+			.payload_length = 0,
+		};
+		bool sched_irq = (s == d->irq_target && i == packets - 1);
+
+		if (queue_out_packet(s, &params, sched_irq) < 0) {
+			cancel_stream(s);
+			return;
+		}
+	}
+}
+
+static void irq_target_callback(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+				void *header, void *private_data);
+
+static void process_rx_packets_intermediately(struct fw_iso_context *context, u32 tstamp,
+					size_t header_length, void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+	__be32 *ctx_header = header;
+	const unsigned int queue_size = s->queue_size;
+	unsigned int packets;
+	unsigned int offset;
+
+	if (s->packet_index < 0)
+		return;
+
+	packets = header_length / sizeof(*ctx_header);
+
+	offset = 0;
+	while (offset < packets) {
+		unsigned int cycle = compute_ohci_it_cycle(ctx_header[offset], queue_size);
+
+		if (compare_ohci_cycle_count(cycle, d->processing_cycle.rx_start) >= 0)
+			break;
+
+		++offset;
+	}
+
+	if (offset > 0) {
+		unsigned int length = sizeof(*ctx_header) * offset;
+
+		skip_rx_packets(context, tstamp, length, ctx_header, private_data);
+		if (amdtp_streaming_error(s))
+			return;
+
+		ctx_header += offset;
+		header_length -= length;
+	}
+
+	if (offset < packets) {
+		s->ready_processing = true;
+		wake_up(&s->ready_wait);
+
+		process_rx_packets(context, tstamp, header_length, ctx_header, private_data);
+		if (amdtp_streaming_error(s))
+			return;
+
+		if (s == d->irq_target)
+			s->context->callback.sc = irq_target_callback;
+		else
+			s->context->callback.sc = process_rx_packets;
+	}
+}
+
+static void process_tx_packets(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+			       void *header, void *private_data)
 {
 	struct amdtp_stream *s = private_data;
 	__be32 *ctx_header = header;
 	unsigned int packets;
+	unsigned int desc_count;
 	int i;
 	int err;
 
@@ -896,14 +1215,20 @@ static void in_stream_callback(struct fw_iso_context *context, u32 tstamp,
 	// Calculate the number of packets in buffer and check XRUN.
 	packets = header_length / s->ctx_data.tx.ctx_header_size;
 
-	err = generate_device_pkt_descs(s, s->pkt_descs, ctx_header, packets);
+	desc_count = 0;
+	err = generate_device_pkt_descs(s, s->pkt_descs, ctx_header, packets, &desc_count);
 	if (err < 0) {
 		if (err != -EAGAIN) {
 			cancel_stream(s);
 			return;
 		}
 	} else {
-		process_ctx_payloads(s, s->pkt_descs, packets);
+		struct amdtp_domain *d = s->domain;
+
+		process_ctx_payloads(s, s->pkt_descs, desc_count);
+
+		if (d->replay.enable)
+			cache_seq(s, s->pkt_descs, desc_count);
 	}
 
 	for (i = 0; i < packets; ++i) {
@@ -916,79 +1241,194 @@ static void in_stream_callback(struct fw_iso_context *context, u32 tstamp,
 	}
 }
 
-static void pool_ideal_seq_descs(struct amdtp_domain *d, unsigned int packets)
+static void drop_tx_packets(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+			    void *header, void *private_data)
 {
-	struct amdtp_stream *irq_target = d->irq_target;
-	unsigned int seq_tail = d->seq_tail;
-	unsigned int seq_size = d->seq_size;
-	unsigned int min_avail;
-	struct amdtp_stream *s;
+	struct amdtp_stream *s = private_data;
+	const __be32 *ctx_header = header;
+	unsigned int packets;
+	unsigned int cycle;
+	int i;
 
-	min_avail = d->seq_size;
-	list_for_each_entry(s, &d->streams, list) {
-		unsigned int seq_index;
-		unsigned int avail;
+	if (s->packet_index < 0)
+		return;
+
+	packets = header_length / s->ctx_data.tx.ctx_header_size;
 
-		if (s->direction == AMDTP_IN_STREAM)
-			continue;
+	ctx_header += (packets - 1) * s->ctx_data.tx.ctx_header_size / sizeof(*ctx_header);
+	cycle = compute_ohci_cycle_count(ctx_header[1]);
+	s->next_cycle = increment_ohci_cycle_count(cycle, 1);
 
-		seq_index = s->ctx_data.rx.seq_index;
-		avail = d->seq_tail;
-		if (seq_index > avail)
-			avail += d->seq_size;
-		avail -= seq_index;
+	for (i = 0; i < packets; ++i) {
+		struct fw_iso_packet params = {0};
 
-		if (avail < min_avail)
-			min_avail = avail;
+		if (queue_in_packet(s, &params) < 0) {
+			cancel_stream(s);
+			return;
+		}
 	}
+}
 
-	while (min_avail < packets) {
-		struct seq_desc *desc = d->seq_descs + seq_tail;
+static void process_tx_packets_intermediately(struct fw_iso_context *context, u32 tstamp,
+					size_t header_length, void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+	__be32 *ctx_header;
+	unsigned int packets;
+	unsigned int offset;
 
-		desc->syt_offset = calculate_syt_offset(&d->last_syt_offset,
-					&d->syt_offset_state, irq_target->sfc);
-		desc->data_blocks = calculate_data_blocks(&d->data_block_state,
-				!!(irq_target->flags & CIP_BLOCKING),
-				desc->syt_offset == CIP_SYT_NO_INFO,
-				irq_target->syt_interval, irq_target->sfc);
+	if (s->packet_index < 0)
+		return;
 
-		++seq_tail;
-		seq_tail %= seq_size;
+	packets = header_length / s->ctx_data.tx.ctx_header_size;
 
-		++min_avail;
+	offset = 0;
+	ctx_header = header;
+	while (offset < packets) {
+		unsigned int cycle = compute_ohci_cycle_count(ctx_header[1]);
+
+		if (compare_ohci_cycle_count(cycle, d->processing_cycle.tx_start) >= 0)
+			break;
+
+		ctx_header += s->ctx_data.tx.ctx_header_size / sizeof(__be32);
+		++offset;
 	}
 
-	d->seq_tail = seq_tail;
+	ctx_header = header;
+
+	if (offset > 0) {
+		size_t length = s->ctx_data.tx.ctx_header_size * offset;
+
+		drop_tx_packets(context, tstamp, length, ctx_header, s);
+		if (amdtp_streaming_error(s))
+			return;
+
+		ctx_header += length / sizeof(*ctx_header);
+		header_length -= length;
+	}
+
+	if (offset < packets) {
+		s->ready_processing = true;
+		wake_up(&s->ready_wait);
+
+		process_tx_packets(context, tstamp, header_length, ctx_header, s);
+		if (amdtp_streaming_error(s))
+			return;
+
+		context->callback.sc = process_tx_packets;
+	}
 }
 
-static void irq_target_callback(struct fw_iso_context *context, u32 tstamp,
-				size_t header_length, void *header,
-				void *private_data)
+static void drop_tx_packets_initially(struct fw_iso_context *context, u32 tstamp,
+				      size_t header_length, void *header, void *private_data)
 {
-	struct amdtp_stream *irq_target = private_data;
-	struct amdtp_domain *d = irq_target->domain;
-	unsigned int packets = header_length / sizeof(__be32);
-	struct amdtp_stream *s;
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+	__be32 *ctx_header;
+	unsigned int count;
+	unsigned int events;
+	int i;
+
+	if (s->packet_index < 0)
+		return;
+
+	count = header_length / s->ctx_data.tx.ctx_header_size;
+
+	// Attempt to detect any event in the batch of packets.
+	events = 0;
+	ctx_header = header;
+	for (i = 0; i < count; ++i) {
+		unsigned int payload_quads =
+			(be32_to_cpu(*ctx_header) >> ISO_DATA_LENGTH_SHIFT) / sizeof(__be32);
+		unsigned int data_blocks;
+
+		if (s->flags & CIP_NO_HEADER) {
+			data_blocks = payload_quads / s->data_block_quadlets;
+		} else {
+			__be32 *cip_headers = ctx_header + IR_CTX_HEADER_DEFAULT_QUADLETS;
+
+			if (payload_quads < CIP_HEADER_QUADLETS) {
+				data_blocks = 0;
+			} else {
+				payload_quads -= CIP_HEADER_QUADLETS;
+
+				if (s->flags & CIP_UNAWARE_SYT) {
+					data_blocks = payload_quads / s->data_block_quadlets;
+				} else {
+					u32 cip1 = be32_to_cpu(cip_headers[1]);
+
+					// NODATA packet can includes any data blocks but they are
+					// not available as event.
+					if ((cip1 & CIP_NO_DATA) == CIP_NO_DATA)
+						data_blocks = 0;
+					else
+						data_blocks = payload_quads / s->data_block_quadlets;
+				}
+			}
+		}
 
-	// Record enough entries with extra 3 cycles at least.
-	pool_ideal_seq_descs(d, packets + 3);
+		events += data_blocks;
 
-	out_stream_callback(context, tstamp, header_length, header, irq_target);
-	if (amdtp_streaming_error(irq_target))
-		goto error;
+		ctx_header += s->ctx_data.tx.ctx_header_size / sizeof(__be32);
+	}
+
+	drop_tx_packets(context, tstamp, header_length, header, s);
+
+	if (events > 0)
+		s->ctx_data.tx.event_starts = true;
+
+	// Decide the cycle count to begin processing content of packet in IR contexts.
+	{
+		unsigned int stream_count = 0;
+		unsigned int event_starts_count = 0;
+		unsigned int cycle = UINT_MAX;
+
+		list_for_each_entry(s, &d->streams, list) {
+			if (s->direction == AMDTP_IN_STREAM) {
+				++stream_count;
+				if (s->ctx_data.tx.event_starts)
+					++event_starts_count;
+			}
+		}
+
+		if (stream_count == event_starts_count) {
+			unsigned int next_cycle;
+
+			list_for_each_entry(s, &d->streams, list) {
+				if (s->direction != AMDTP_IN_STREAM)
+					continue;
+
+				next_cycle = increment_ohci_cycle_count(s->next_cycle,
+								d->processing_cycle.tx_init_skip);
+				if (cycle == UINT_MAX ||
+				    compare_ohci_cycle_count(next_cycle, cycle) > 0)
+					cycle = next_cycle;
+
+				s->context->callback.sc = process_tx_packets_intermediately;
+			}
+
+			d->processing_cycle.tx_start = cycle;
+		}
+	}
+}
+
+static void process_ctxs_in_domain(struct amdtp_domain *d)
+{
+	struct amdtp_stream *s;
 
 	list_for_each_entry(s, &d->streams, list) {
-		if (s != irq_target && amdtp_stream_running(s)) {
+		if (s != d->irq_target && amdtp_stream_running(s))
 			fw_iso_context_flush_completions(s->context);
-			if (amdtp_streaming_error(s))
-				goto error;
-		}
+
+		if (amdtp_streaming_error(s))
+			goto error;
 	}
 
 	return;
 error:
-	if (amdtp_stream_running(irq_target))
-		cancel_stream(irq_target);
+	if (amdtp_stream_running(d->irq_target))
+		cancel_stream(d->irq_target);
 
 	list_for_each_entry(s, &d->streams, list) {
 		if (amdtp_stream_running(s))
@@ -996,37 +1436,99 @@ error:
 	}
 }
 
-// this is executed one time.
+static void irq_target_callback(struct fw_iso_context *context, u32 tstamp, size_t header_length,
+				void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+
+	process_rx_packets(context, tstamp, header_length, header, private_data);
+	process_ctxs_in_domain(d);
+}
+
+static void irq_target_callback_intermediately(struct fw_iso_context *context, u32 tstamp,
+					size_t header_length, void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+
+	process_rx_packets_intermediately(context, tstamp, header_length, header, private_data);
+	process_ctxs_in_domain(d);
+}
+
+static void irq_target_callback_skip(struct fw_iso_context *context, u32 tstamp,
+				     size_t header_length, void *header, void *private_data)
+{
+	struct amdtp_stream *s = private_data;
+	struct amdtp_domain *d = s->domain;
+	bool ready_to_start;
+
+	skip_rx_packets(context, tstamp, header_length, header, private_data);
+	process_ctxs_in_domain(d);
+
+	if (d->replay.enable && !d->replay.on_the_fly) {
+		unsigned int rx_count = 0;
+		unsigned int rx_ready_count = 0;
+		struct amdtp_stream *rx;
+
+		list_for_each_entry(rx, &d->streams, list) {
+			struct amdtp_stream *tx;
+			unsigned int cached_cycles;
+
+			if (rx->direction != AMDTP_OUT_STREAM)
+				continue;
+			++rx_count;
+
+			tx = rx->ctx_data.rx.replay_target;
+			cached_cycles = calculate_cached_cycle_count(tx, 0);
+			if (cached_cycles > tx->ctx_data.tx.cache.size / 2)
+				++rx_ready_count;
+		}
+
+		ready_to_start = (rx_count == rx_ready_count);
+	} else {
+		ready_to_start = true;
+	}
+
+	// Decide the cycle count to begin processing content of packet in IT contexts. All of IT
+	// contexts are expected to start and get callback when reaching here.
+	if (ready_to_start) {
+		unsigned int cycle = s->next_cycle;
+		list_for_each_entry(s, &d->streams, list) {
+			if (s->direction != AMDTP_OUT_STREAM)
+				continue;
+
+			if (compare_ohci_cycle_count(s->next_cycle, cycle) > 0)
+				cycle = s->next_cycle;
+
+			if (s == d->irq_target)
+				s->context->callback.sc = irq_target_callback_intermediately;
+			else
+				s->context->callback.sc = process_rx_packets_intermediately;
+		}
+
+		d->processing_cycle.rx_start = cycle;
+	}
+}
+
+// This is executed one time. For in-stream, first packet has come. For out-stream, prepared to
+// transmit first packet.
 static void amdtp_stream_first_callback(struct fw_iso_context *context,
 					u32 tstamp, size_t header_length,
 					void *header, void *private_data)
 {
 	struct amdtp_stream *s = private_data;
-	const __be32 *ctx_header = header;
-	u32 cycle;
-
-	/*
-	 * For in-stream, first packet has come.
-	 * For out-stream, prepared to transmit first packet
-	 */
-	s->callbacked = true;
-	wake_up(&s->callback_wait);
+	struct amdtp_domain *d = s->domain;
 
 	if (s->direction == AMDTP_IN_STREAM) {
-		cycle = compute_cycle_count(ctx_header[1]);
-
-		context->callback.sc = in_stream_callback;
+		context->callback.sc = drop_tx_packets_initially;
 	} else {
-		cycle = compute_it_cycle(*ctx_header, s->queue_size);
-
-		if (s == s->domain->irq_target)
-			context->callback.sc = irq_target_callback;
+		if (s == d->irq_target)
+			context->callback.sc = irq_target_callback_skip;
 		else
-			context->callback.sc = out_stream_callback;
+			context->callback.sc = skip_rx_packets;
 	}
 
-	s->start_cycle = cycle;
-
 	context->callback.sc(context, tstamp, header_length, header, s);
 }
 
@@ -1035,8 +1537,6 @@ static void amdtp_stream_first_callback(struct fw_iso_context *context,
  * @s: the AMDTP stream to start
  * @channel: the isochronous channel on the bus
  * @speed: firewire speed code
- * @start_cycle: the isochronous cycle to start the context. Start immediately
- *		 if negative value is given.
  * @queue_size: The number of packets in the queue.
  * @idle_irq_interval: the interval to queue packet during initial state.
  *
@@ -1045,8 +1545,7 @@ static void amdtp_stream_first_callback(struct fw_iso_context *context,
  * device can be started.
  */
 static int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed,
-			      int start_cycle, unsigned int queue_size,
-			      unsigned int idle_irq_interval)
+			      unsigned int queue_size, unsigned int idle_irq_interval)
 {
 	bool is_irq_target = (s == s->domain->irq_target);
 	unsigned int ctx_header_size;
@@ -1075,27 +1574,21 @@ static int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed,
 	}
 
 	// initialize packet buffer.
-	max_ctx_payload_size = amdtp_stream_get_max_payload(s);
 	if (s->direction == AMDTP_IN_STREAM) {
 		dir = DMA_FROM_DEVICE;
 		type = FW_ISO_CONTEXT_RECEIVE;
-		if (!(s->flags & CIP_NO_HEADER)) {
-			max_ctx_payload_size -= 8;
+		if (!(s->flags & CIP_NO_HEADER))
 			ctx_header_size = IR_CTX_HEADER_SIZE_CIP;
-		} else {
+		else
 			ctx_header_size = IR_CTX_HEADER_SIZE_NO_CIP;
-		}
 	} else {
 		dir = DMA_TO_DEVICE;
 		type = FW_ISO_CONTEXT_TRANSMIT;
 		ctx_header_size = 0;	// No effect for IT context.
-
-		if (!(s->flags & CIP_NO_HEADER))
-			max_ctx_payload_size -= IT_PKT_HEADER_SIZE_CIP;
 	}
+	max_ctx_payload_size = amdtp_stream_get_max_ctx_payload_size(s);
 
-	err = iso_packets_buffer_init(&s->buffer, s->unit, queue_size,
-				      max_ctx_payload_size, dir);
+	err = iso_packets_buffer_init(&s->buffer, s->unit, queue_size, max_ctx_payload_size, dir);
 	if (err < 0)
 		goto err_unlock;
 	s->queue_size = queue_size;
@@ -1116,6 +1609,50 @@ static int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed,
 	if (s->direction == AMDTP_IN_STREAM) {
 		s->ctx_data.tx.max_ctx_payload_length = max_ctx_payload_size;
 		s->ctx_data.tx.ctx_header_size = ctx_header_size;
+		s->ctx_data.tx.event_starts = false;
+
+		if (s->domain->replay.enable) {
+			// struct fw_iso_context.drop_overflow_headers is false therefore it's
+			// possible to cache much unexpectedly.
+			s->ctx_data.tx.cache.size = max_t(unsigned int, s->syt_interval * 2,
+							  queue_size * 3 / 2);
+			s->ctx_data.tx.cache.tail = 0;
+			s->ctx_data.tx.cache.descs = kcalloc(s->ctx_data.tx.cache.size,
+						sizeof(*s->ctx_data.tx.cache.descs), GFP_KERNEL);
+			if (!s->ctx_data.tx.cache.descs) {
+				err = -ENOMEM;
+				goto err_context;
+			}
+		}
+	} else {
+		static const struct {
+			unsigned int data_block;
+			unsigned int syt_offset;
+		} *entry, initial_state[] = {
+			[CIP_SFC_32000]  = {  4, 3072 },
+			[CIP_SFC_48000]  = {  6, 1024 },
+			[CIP_SFC_96000]  = { 12, 1024 },
+			[CIP_SFC_192000] = { 24, 1024 },
+			[CIP_SFC_44100]  = {  0,   67 },
+			[CIP_SFC_88200]  = {  0,   67 },
+			[CIP_SFC_176400] = {  0,   67 },
+		};
+
+		s->ctx_data.rx.seq.descs = kcalloc(queue_size, sizeof(*s->ctx_data.rx.seq.descs), GFP_KERNEL);
+		if (!s->ctx_data.rx.seq.descs) {
+			err = -ENOMEM;
+			goto err_context;
+		}
+		s->ctx_data.rx.seq.size = queue_size;
+		s->ctx_data.rx.seq.tail = 0;
+		s->ctx_data.rx.seq.head = 0;
+
+		entry = &initial_state[s->sfc];
+		s->ctx_data.rx.data_block_state = entry->data_block;
+		s->ctx_data.rx.syt_offset_state = entry->syt_offset;
+		s->ctx_data.rx.last_syt_offset = TICKS_PER_CYCLE;
+
+		s->ctx_data.rx.event_count = 0;
 	}
 
 	if (s->flags & CIP_NO_HEADER)
@@ -1158,8 +1695,8 @@ static int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed,
 	if ((s->flags & CIP_EMPTY_WITH_TAG0) || (s->flags & CIP_NO_HEADER))
 		tag |= FW_ISO_CONTEXT_MATCH_TAG0;
 
-	s->callbacked = false;
-	err = fw_iso_context_start(s->context, start_cycle, 0, tag);
+	s->ready_processing = false;
+	err = fw_iso_context_start(s->context, -1, 0, tag);
 	if (err < 0)
 		goto err_pkt_descs;
 
@@ -1169,6 +1706,12 @@ static int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed,
 err_pkt_descs:
 	kfree(s->pkt_descs);
 err_context:
+	if (s->direction == AMDTP_OUT_STREAM) {
+		kfree(s->ctx_data.rx.seq.descs);
+	} else {
+		if (s->domain->replay.enable)
+			kfree(s->ctx_data.tx.cache.descs);
+	}
 	fw_iso_context_destroy(s->context);
 	s->context = ERR_PTR(-1);
 err_buffer:
@@ -1191,28 +1734,12 @@ unsigned long amdtp_domain_stream_pcm_pointer(struct amdtp_domain *d,
 {
 	struct amdtp_stream *irq_target = d->irq_target;
 
+	// Process isochronous packets queued till recent isochronous cycle to handle PCM frames.
 	if (irq_target && amdtp_stream_running(irq_target)) {
-		// This function is called in software IRQ context of
-		// period_work or process context.
-		//
-		// When the software IRQ context was scheduled by software IRQ
-		// context of IT contexts, queued packets were already handled.
-		// Therefore, no need to flush the queue in buffer furthermore.
-		//
-		// When the process context reach here, some packets will be
-		// already queued in the buffer. These packets should be handled
-		// immediately to keep better granularity of PCM pointer.
-		//
-		// Later, the process context will sometimes schedules software
-		// IRQ context of the period_work. Then, no need to flush the
-		// queue by the same reason as described in the above
-		if (current_work() != &s->period_work) {
-			// Queued packet should be processed without any kernel
-			// preemption to keep latency against bus cycle.
-			preempt_disable();
+		// In software IRQ context, the call causes dead-lock to disable the tasklet
+		// synchronously.
+		if (!in_softirq())
 			fw_iso_context_flush_completions(irq_target->context);
-			preempt_enable();
-		}
 	}
 
 	return READ_ONCE(s->pcm_buffer_pointer);
@@ -1232,13 +1759,8 @@ int amdtp_domain_stream_pcm_ack(struct amdtp_domain *d, struct amdtp_stream *s)
 
 	// Process isochronous packets for recent isochronous cycle to handle
 	// queued PCM frames.
-	if (irq_target && amdtp_stream_running(irq_target)) {
-		// Queued packet should be processed without any kernel
-		// preemption to keep latency against bus cycle.
-		preempt_disable();
+	if (irq_target && amdtp_stream_running(irq_target))
 		fw_iso_context_flush_completions(irq_target->context);
-		preempt_enable();
-	}
 
 	return 0;
 }
@@ -1272,14 +1794,18 @@ static void amdtp_stream_stop(struct amdtp_stream *s)
 		return;
 	}
 
-	cancel_work_sync(&s->period_work);
 	fw_iso_context_stop(s->context);
 	fw_iso_context_destroy(s->context);
 	s->context = ERR_PTR(-1);
 	iso_packets_buffer_destroy(&s->buffer, s->unit);
 	kfree(s->pkt_descs);
 
-	s->callbacked = false;
+	if (s->direction == AMDTP_OUT_STREAM) {
+		kfree(s->ctx_data.rx.seq.descs);
+	} else {
+		if (s->domain->replay.enable)
+			kfree(s->ctx_data.tx.cache.descs);
+	}
 
 	mutex_unlock(&s->mutex);
 }
@@ -1311,8 +1837,6 @@ int amdtp_domain_init(struct amdtp_domain *d)
 
 	d->events_per_period = 0;
 
-	d->seq_descs = NULL;
-
 	return 0;
 }
 EXPORT_SYMBOL_GPL(amdtp_domain_init);
@@ -1355,26 +1879,49 @@ int amdtp_domain_add_stream(struct amdtp_domain *d, struct amdtp_stream *s,
 }
 EXPORT_SYMBOL_GPL(amdtp_domain_add_stream);
 
-static int get_current_cycle_time(struct fw_card *fw_card, int *cur_cycle)
+// Make the reference from rx stream to tx stream for sequence replay. When the number of tx streams
+// is less than the number of rx streams, the first tx stream is selected.
+static int make_association(struct amdtp_domain *d)
 {
-	int generation;
-	int rcode;
-	__be32 reg;
-	u32 data;
-
-	// This is a request to local 1394 OHCI controller and expected to
-	// complete without any event waiting.
-	generation = fw_card->generation;
-	smp_rmb();	// node_id vs. generation.
-	rcode = fw_run_transaction(fw_card, TCODE_READ_QUADLET_REQUEST,
-				   fw_card->node_id, generation, SCODE_100,
-				   CSR_REGISTER_BASE + CSR_CYCLE_TIME,
-				   &reg, sizeof(reg));
-	if (rcode != RCODE_COMPLETE)
-		return -EIO;
+	unsigned int dst_index = 0;
+	struct amdtp_stream *rx;
+
+	// Make association to replay target.
+	list_for_each_entry(rx, &d->streams, list) {
+		if (rx->direction == AMDTP_OUT_STREAM) {
+			unsigned int src_index = 0;
+			struct amdtp_stream *tx = NULL;
+			struct amdtp_stream *s;
+
+			list_for_each_entry(s, &d->streams, list) {
+				if (s->direction == AMDTP_IN_STREAM) {
+					if (dst_index == src_index) {
+						tx = s;
+						break;
+					}
+
+					++src_index;
+				}
+			}
+			if (!tx) {
+				// Select the first entry.
+				list_for_each_entry(s, &d->streams, list) {
+					if (s->direction == AMDTP_IN_STREAM) {
+						tx = s;
+						break;
+					}
+				}
+				// No target is available to replay sequence.
+				if (!tx)
+					return -EINVAL;
+			}
 
-	data = be32_to_cpu(reg);
-	*cur_cycle = data >> 12;
+			rx->ctx_data.rx.replay_target = tx;
+			rx->ctx_data.rx.cache_head = 0;
+
+			++dst_index;
+		}
+	}
 
 	return 0;
 }
@@ -1382,39 +1929,44 @@ static int get_current_cycle_time(struct fw_card *fw_card, int *cur_cycle)
 /**
  * amdtp_domain_start - start sending packets for isoc context in the domain.
  * @d: the AMDTP domain.
- * @ir_delay_cycle: the cycle delay to start all IR contexts.
+ * @tx_init_skip_cycles: the number of cycles to skip processing packets at initial stage of IR
+ *			 contexts.
+ * @replay_seq: whether to replay the sequence of packet in IR context for the sequence of packet in
+ *		IT context.
+ * @replay_on_the_fly: transfer rx packets according to nominal frequency, then begin to replay
+ *		       according to arrival of events in tx packets.
  */
-int amdtp_domain_start(struct amdtp_domain *d, unsigned int ir_delay_cycle)
+int amdtp_domain_start(struct amdtp_domain *d, unsigned int tx_init_skip_cycles, bool replay_seq,
+		       bool replay_on_the_fly)
 {
-	static const struct {
-		unsigned int data_block;
-		unsigned int syt_offset;
-	} *entry, initial_state[] = {
-		[CIP_SFC_32000]  = {  4, 3072 },
-		[CIP_SFC_48000]  = {  6, 1024 },
-		[CIP_SFC_96000]  = { 12, 1024 },
-		[CIP_SFC_192000] = { 24, 1024 },
-		[CIP_SFC_44100]  = {  0,   67 },
-		[CIP_SFC_88200]  = {  0,   67 },
-		[CIP_SFC_176400] = {  0,   67 },
-	};
 	unsigned int events_per_buffer = d->events_per_buffer;
 	unsigned int events_per_period = d->events_per_period;
-	unsigned int idle_irq_interval;
 	unsigned int queue_size;
 	struct amdtp_stream *s;
-	int cycle;
+	bool found = false;
 	int err;
 
+	if (replay_seq) {
+		err = make_association(d);
+		if (err < 0)
+			return err;
+	}
+	d->replay.enable = replay_seq;
+	d->replay.on_the_fly = replay_on_the_fly;
+
 	// Select an IT context as IRQ target.
 	list_for_each_entry(s, &d->streams, list) {
-		if (s->direction == AMDTP_OUT_STREAM)
+		if (s->direction == AMDTP_OUT_STREAM) {
+			found = true;
 			break;
+		}
 	}
-	if (!s)
+	if (!found)
 		return -ENXIO;
 	d->irq_target = s;
 
+	d->processing_cycle.tx_init_skip = tx_init_skip_cycles;
+
 	// This is a case that AMDTP streams in domain run just for MIDI
 	// substream. Use the number of events equivalent to 10 msec as
 	// interval of hardware IRQ.
@@ -1426,82 +1978,24 @@ int amdtp_domain_start(struct amdtp_domain *d, unsigned int ir_delay_cycle)
 	queue_size = DIV_ROUND_UP(CYCLES_PER_SECOND * events_per_buffer,
 				  amdtp_rate_table[d->irq_target->sfc]);
 
-	d->seq_descs = kcalloc(queue_size, sizeof(*d->seq_descs), GFP_KERNEL);
-	if (!d->seq_descs)
-		return -ENOMEM;
-	d->seq_size = queue_size;
-	d->seq_tail = 0;
-
-	entry = &initial_state[s->sfc];
-	d->data_block_state = entry->data_block;
-	d->syt_offset_state = entry->syt_offset;
-	d->last_syt_offset = TICKS_PER_CYCLE;
-
-	if (ir_delay_cycle > 0) {
-		struct fw_card *fw_card = fw_parent_device(s->unit)->card;
-
-		err = get_current_cycle_time(fw_card, &cycle);
-		if (err < 0)
-			goto error;
-
-		// No need to care overflow in cycle field because of enough
-		// width.
-		cycle += ir_delay_cycle;
-
-		// Round up to sec field.
-		if ((cycle & 0x00001fff) >= CYCLES_PER_SECOND) {
-			unsigned int sec;
-
-			// The sec field can overflow.
-			sec = (cycle & 0xffffe000) >> 13;
-			cycle = (++sec << 13) |
-				((cycle & 0x00001fff) / CYCLES_PER_SECOND);
-		}
-
-		// In OHCI 1394 specification, lower 2 bits are available for
-		// sec field.
-		cycle &= 0x00007fff;
-	} else {
-		cycle = -1;
-	}
-
 	list_for_each_entry(s, &d->streams, list) {
-		int cycle_match;
+		unsigned int idle_irq_interval = 0;
 
-		if (s->direction == AMDTP_IN_STREAM) {
-			cycle_match = cycle;
-		} else {
-			// IT context starts immediately.
-			cycle_match = -1;
-			s->ctx_data.rx.seq_index = 0;
+		if (s->direction == AMDTP_OUT_STREAM && s == d->irq_target) {
+			idle_irq_interval = DIV_ROUND_UP(CYCLES_PER_SECOND * events_per_period,
+							 amdtp_rate_table[d->irq_target->sfc]);
 		}
 
-		if (s != d->irq_target) {
-			err = amdtp_stream_start(s, s->channel, s->speed,
-						 cycle_match, queue_size, 0);
-			if (err < 0)
-				goto error;
-		}
+		// Starts immediately but actually DMA context starts several hundred cycles later.
+		err = amdtp_stream_start(s, s->channel, s->speed, queue_size, idle_irq_interval);
+		if (err < 0)
+			goto error;
 	}
 
-	s = d->irq_target;
-	s->ctx_data.rx.events_per_period = events_per_period;
-	s->ctx_data.rx.event_count = 0;
-	s->ctx_data.rx.seq_index = 0;
-
-	idle_irq_interval = DIV_ROUND_UP(CYCLES_PER_SECOND * events_per_period,
-					 amdtp_rate_table[d->irq_target->sfc]);
-	err = amdtp_stream_start(s, s->channel, s->speed, -1, queue_size,
-				 idle_irq_interval);
-	if (err < 0)
-		goto error;
-
 	return 0;
 error:
 	list_for_each_entry(s, &d->streams, list)
 		amdtp_stream_stop(s);
-	kfree(d->seq_descs);
-	d->seq_descs = NULL;
 	return err;
 }
 EXPORT_SYMBOL_GPL(amdtp_domain_start);
@@ -1526,8 +2020,5 @@ void amdtp_domain_stop(struct amdtp_domain *d)
 
 	d->events_per_period = 0;
 	d->irq_target = NULL;
-
-	kfree(d->seq_descs);
-	d->seq_descs = NULL;
 }
 EXPORT_SYMBOL_GPL(amdtp_domain_stop);
diff --git a/sound/firewire/amdtp-stream.h b/sound/firewire/amdtp-stream.h
index a3daa1f2c1c4..1f957c946c95 100644
--- a/sound/firewire/amdtp-stream.h
+++ b/sound/firewire/amdtp-stream.h
@@ -35,6 +35,8 @@
  * @CIP_NO_HEADERS: a lack of headers in packets
  * @CIP_UNALIGHED_DBC: Only for in-stream. The value of dbc is not alighed to
  *	the value of current SYT_INTERVAL; e.g. initial value is not zero.
+ * @CIP_UNAWARE_SYT: For outgoing packet, the value in SYT field of CIP is 0xffff.
+ *	For incoming packet, the value in SYT field of CIP is not handled.
  */
 enum cip_flags {
 	CIP_NONBLOCKING		= 0x00,
@@ -48,6 +50,7 @@ enum cip_flags {
 	CIP_HEADER_WITHOUT_EOH	= 0x80,
 	CIP_NO_HEADER		= 0x100,
 	CIP_UNALIGHED_DBC	= 0x200,
+	CIP_UNAWARE_SYT		= 0x400,
 };
 
 /**
@@ -112,7 +115,8 @@ typedef unsigned int (*amdtp_stream_process_ctx_payloads_t)(
 struct amdtp_domain;
 struct amdtp_stream {
 	struct fw_unit *unit;
-	enum cip_flags flags;
+	// The combination of cip_flags enumeration-constants.
+	unsigned int flags;
 	enum amdtp_stream_direction direction;
 	struct mutex mutex;
 
@@ -134,19 +138,37 @@ struct amdtp_stream {
 			// Fixed interval of dbc between previos/current
 			// packets.
 			unsigned int dbc_interval;
+
+			// The device starts multiplexing events to the packet.
+			bool event_starts;
+
+			struct {
+				struct seq_desc *descs;
+				unsigned int size;
+				unsigned int tail;
+			} cache;
 		} tx;
 		struct {
-			// To calculate CIP data blocks and tstamp.
-			unsigned int transfer_delay;
-			unsigned int seq_index;
-
 			// To generate CIP header.
 			unsigned int fdf;
-			int syt_override;
 
 			// To generate constant hardware IRQ.
 			unsigned int event_count;
-			unsigned int events_per_period;
+
+			// To calculate CIP data blocks and tstamp.
+			struct {
+				struct seq_desc *descs;
+				unsigned int size;
+				unsigned int tail;
+				unsigned int head;
+			} seq;
+
+			unsigned int data_block_state;
+			unsigned int syt_offset_state;
+			unsigned int last_syt_offset;
+
+			struct amdtp_stream *replay_target;
+			unsigned int cache_head;
 		} rx;
 	} ctx_data;
 
@@ -157,20 +179,21 @@ struct amdtp_stream {
 	unsigned int sph;
 	unsigned int fmt;
 
-	/* Internal flags. */
+	// Internal flags.
+	unsigned int transfer_delay;
 	enum cip_sfc sfc;
 	unsigned int syt_interval;
 
 	/* For a PCM substream processing. */
 	struct snd_pcm_substream *pcm;
-	struct work_struct period_work;
 	snd_pcm_uframes_t pcm_buffer_pointer;
 	unsigned int pcm_period_pointer;
 
-	/* To wait for first packet. */
-	bool callbacked;
-	wait_queue_head_t callback_wait;
-	u32 start_cycle;
+	// To start processing content of packets at the same cycle in several contexts for
+	// each direction.
+	bool ready_processing;
+	wait_queue_head_t ready_wait;
+	unsigned int next_cycle;
 
 	/* For backends to process data blocks. */
 	void *protocol;
@@ -184,7 +207,7 @@ struct amdtp_stream {
 };
 
 int amdtp_stream_init(struct amdtp_stream *s, struct fw_unit *unit,
-		      enum amdtp_stream_direction dir, enum cip_flags flags,
+		      enum amdtp_stream_direction dir, unsigned int flags,
 		      unsigned int fmt,
 		      amdtp_stream_process_ctx_payloads_t process_ctx_payloads,
 		      unsigned int protocol_size);
@@ -259,21 +282,6 @@ static inline bool cip_sfc_is_base_44100(enum cip_sfc sfc)
 	return sfc & 1;
 }
 
-/**
- * amdtp_stream_wait_callback - sleep till callbacked or timeout
- * @s: the AMDTP stream
- * @timeout: msec till timeout
- *
- * If this function return false, the AMDTP stream should be stopped.
- */
-static inline bool amdtp_stream_wait_callback(struct amdtp_stream *s,
-					      unsigned int timeout)
-{
-	return wait_event_timeout(s->callback_wait,
-				  s->callbacked,
-				  msecs_to_jiffies(timeout)) > 0;
-}
-
 struct seq_desc {
 	unsigned int syt_offset;
 	unsigned int data_blocks;
@@ -287,13 +295,16 @@ struct amdtp_domain {
 
 	struct amdtp_stream *irq_target;
 
-	struct seq_desc *seq_descs;
-	unsigned int seq_size;
-	unsigned int seq_tail;
+	struct {
+		unsigned int tx_init_skip;
+		unsigned int tx_start;
+		unsigned int rx_start;
+	} processing_cycle;
 
-	unsigned int data_block_state;
-	unsigned int syt_offset_state;
-	unsigned int last_syt_offset;
+	struct {
+		bool enable:1;
+		bool on_the_fly:1;
+	} replay;
 };
 
 int amdtp_domain_init(struct amdtp_domain *d);
@@ -302,7 +313,8 @@ void amdtp_domain_destroy(struct amdtp_domain *d);
 int amdtp_domain_add_stream(struct amdtp_domain *d, struct amdtp_stream *s,
 			    int channel, int speed);
 
-int amdtp_domain_start(struct amdtp_domain *d, unsigned int ir_delay_cycle);
+int amdtp_domain_start(struct amdtp_domain *d, unsigned int tx_init_skip_cycles, bool replay_seq,
+		       bool replay_on_the_fly);
 void amdtp_domain_stop(struct amdtp_domain *d);
 
 static inline int amdtp_domain_set_events_per_period(struct amdtp_domain *d,
@@ -319,4 +331,25 @@ unsigned long amdtp_domain_stream_pcm_pointer(struct amdtp_domain *d,
 					      struct amdtp_stream *s);
 int amdtp_domain_stream_pcm_ack(struct amdtp_domain *d, struct amdtp_stream *s);
 
+/**
+ * amdtp_domain_wait_ready - sleep till being ready to process packets or timeout
+ * @d: the AMDTP domain
+ * @timeout_ms: msec till timeout
+ *
+ * If this function return false, the AMDTP domain should be stopped.
+ */
+static inline bool amdtp_domain_wait_ready(struct amdtp_domain *d, unsigned int timeout_ms)
+{
+	struct amdtp_stream *s;
+
+	list_for_each_entry(s, &d->streams, list) {
+		unsigned int j = msecs_to_jiffies(timeout_ms);
+
+		if (wait_event_interruptible_timeout(s->ready_wait, s->ready_processing, j) <= 0)
+			return false;
+	}
+
+	return true;
+}
+
 #endif
diff --git a/sound/firewire/bebob/bebob.c b/sound/firewire/bebob/bebob.c
index daeecfa8b9aa..42980da45fbf 100644
--- a/sound/firewire/bebob/bebob.c
+++ b/sound/firewire/bebob/bebob.c
@@ -40,14 +40,12 @@ static DECLARE_BITMAP(devices_used, SNDRV_CARDS);
 #define VEN_EDIROL	0x000040ab
 #define VEN_PRESONUS	0x00000a92
 #define VEN_BRIDGECO	0x000007f5
-#define VEN_MACKIE1	0x0000000f
-#define VEN_MACKIE2	0x00000ff2
+#define VEN_MACKIE	0x00000ff2
 #define VEN_STANTON	0x00001260
 #define VEN_TASCAM	0x0000022e
 #define VEN_BEHRINGER	0x00001564
 #define VEN_APOGEE	0x000003db
 #define VEN_ESI		0x00000f1b
-#define VEN_ACOUSTIC	0x00000002
 #define VEN_CME		0x0000000a
 #define VEN_PHONIC	0x00001496
 #define VEN_LYNX	0x000019e5
@@ -56,14 +54,15 @@ static DECLARE_BITMAP(devices_used, SNDRV_CARDS);
 #define VEN_TERRATEC	0x00000aac
 #define VEN_YAMAHA	0x0000a0de
 #define VEN_FOCUSRITE	0x0000130e
-#define VEN_MAUDIO1	0x00000d6c
-#define VEN_MAUDIO2	0x000007f5
+#define VEN_MAUDIO	0x00000d6c
 #define VEN_DIGIDESIGN	0x00a07e
+#define OUI_SHOUYO	0x002327
 
 #define MODEL_FOCUSRITE_SAFFIRE_BOTH	0x00000000
 #define MODEL_MAUDIO_AUDIOPHILE_BOTH	0x00010060
 #define MODEL_MAUDIO_FW1814		0x00010071
 #define MODEL_MAUDIO_PROJECTMIX		0x00010091
+#define MODEL_MAUDIO_PROFIRELIGHTBRIDGE	0x000100a1
 
 static int
 name_device(struct snd_bebob *bebob)
@@ -74,7 +73,6 @@ name_device(struct snd_bebob *bebob)
 	u32 hw_id;
 	u32 data[2] = {0};
 	u32 revision;
-	u32 version;
 	int err;
 
 	/* get vendor name from root directory */
@@ -107,12 +105,6 @@ name_device(struct snd_bebob *bebob)
 	if (err < 0)
 		goto end;
 
-	err = snd_bebob_read_quad(bebob->unit, INFO_OFFSET_BEBOB_VERSION,
-				  &version);
-	if (err < 0)
-		goto end;
-	bebob->version = version;
-
 	strcpy(bebob->card->driver, "BeBoB");
 	strcpy(bebob->card->shortname, model);
 	strcpy(bebob->card->mixername, model);
@@ -135,6 +127,9 @@ bebob_card_free(struct snd_card *card)
 	mutex_unlock(&devices_mutex);
 
 	snd_bebob_stream_destroy_duplex(bebob);
+
+	mutex_destroy(&bebob->mutex);
+	fw_unit_put(bebob->unit);
 }
 
 static const struct snd_bebob_spec *
@@ -162,16 +157,55 @@ check_audiophile_booted(struct fw_unit *unit)
 	return strncmp(name, "FW Audiophile Bootloader", 24) != 0;
 }
 
-static void
-do_registration(struct work_struct *work)
+static int detect_quirks(struct snd_bebob *bebob, const struct ieee1394_device_id *entry)
+{
+	if (entry->vendor_id == VEN_MAUDIO) {
+		switch (entry->model_id) {
+		case MODEL_MAUDIO_PROFIRELIGHTBRIDGE:
+			// M-Audio ProFire Lightbridge has a quirk to transfer packets with
+			// discontinuous cycle or data block counter in early stage of packet
+			// streaming. The cycle span from the first packet with event is variable.
+			bebob->quirks |= SND_BEBOB_QUIRK_INITIAL_DISCONTINUOUS_DBC;
+			break;
+		case MODEL_MAUDIO_FW1814:
+		case MODEL_MAUDIO_PROJECTMIX:
+			// At high sampling rate, M-Audio special firmware transmits empty packet
+			// with the value of dbc incremented by 8.
+			bebob->quirks |= SND_BEBOB_QUIRK_WRONG_DBC;
+			break;
+		default:
+			break;
+		}
+	}
+
+	return 0;
+}
+
+static int bebob_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_bebob *bebob =
-			container_of(work, struct snd_bebob, dwork.work);
 	unsigned int card_index;
+	struct snd_card *card;
+	struct snd_bebob *bebob;
+	const struct snd_bebob_spec *spec;
 	int err;
 
-	if (bebob->registered)
-		return;
+	if (entry->vendor_id == VEN_FOCUSRITE &&
+	    entry->model_id == MODEL_FOCUSRITE_SAFFIRE_BOTH)
+		spec = get_saffire_spec(unit);
+	else if (entry->vendor_id == VEN_MAUDIO &&
+		 entry->model_id == MODEL_MAUDIO_AUDIOPHILE_BOTH &&
+		 !check_audiophile_booted(unit))
+		spec = NULL;
+	else
+		spec = (const struct snd_bebob_spec *)entry->driver_data;
+
+	if (spec == NULL) {
+		// To boot up M-Audio models.
+		if (entry->vendor_id == VEN_MAUDIO || entry->vendor_id == VEN_BRIDGECO)
+			return snd_bebob_maudio_load_firmware(unit);
+		else
+			return -ENODEV;
+	}
 
 	mutex_lock(&devices_mutex);
 	for (card_index = 0; card_index < SNDRV_CARDS; card_index++) {
@@ -180,27 +214,40 @@ do_registration(struct work_struct *work)
 	}
 	if (card_index >= SNDRV_CARDS) {
 		mutex_unlock(&devices_mutex);
-		return;
+		return -ENOENT;
 	}
 
-	err = snd_card_new(&bebob->unit->device, index[card_index],
-			   id[card_index], THIS_MODULE, 0, &bebob->card);
+	err = snd_card_new(&unit->device, index[card_index], id[card_index], THIS_MODULE,
+			   sizeof(*bebob), &card);
 	if (err < 0) {
 		mutex_unlock(&devices_mutex);
-		return;
+		return err;
 	}
+	card->private_free = bebob_card_free;
 	set_bit(card_index, devices_used);
 	mutex_unlock(&devices_mutex);
 
-	bebob->card->private_free = bebob_card_free;
-	bebob->card->private_data = bebob;
+	bebob = card->private_data;
+	bebob->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, bebob);
+	bebob->card = card;
+	bebob->card_index = card_index;
+
+	bebob->spec = spec;
+	mutex_init(&bebob->mutex);
+	spin_lock_init(&bebob->lock);
+	init_waitqueue_head(&bebob->hwdep_wait);
 
 	err = name_device(bebob);
 	if (err < 0)
 		goto error;
 
+	err = detect_quirks(bebob, entry);
+	if (err < 0)
+		goto error;
+
 	if (bebob->spec == &maudio_special_spec) {
-		if (bebob->entry->model_id == MODEL_MAUDIO_FW1814)
+		if (entry->model_id == MODEL_MAUDIO_FW1814)
 			err = snd_bebob_maudio_special_discover(bebob, true);
 		else
 			err = snd_bebob_maudio_special_discover(bebob, false);
@@ -230,80 +277,26 @@ do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(bebob->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	bebob->registered = true;
-
-	return;
-error:
-	snd_card_free(bebob->card);
-	dev_info(&bebob->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int
-bebob_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
-{
-	struct snd_bebob *bebob;
-	const struct snd_bebob_spec *spec;
-
-	if (entry->vendor_id == VEN_FOCUSRITE &&
-	    entry->model_id == MODEL_FOCUSRITE_SAFFIRE_BOTH)
-		spec = get_saffire_spec(unit);
-	else if (entry->vendor_id == VEN_MAUDIO1 &&
-		 entry->model_id == MODEL_MAUDIO_AUDIOPHILE_BOTH &&
-		 !check_audiophile_booted(unit))
-		spec = NULL;
-	else
-		spec = (const struct snd_bebob_spec *)entry->driver_data;
-
-	if (spec == NULL) {
-		if (entry->vendor_id == VEN_MAUDIO1 ||
-		    entry->vendor_id == VEN_MAUDIO2)
-			return snd_bebob_maudio_load_firmware(unit);
-		else
-			return -ENODEV;
-	}
-
-	/* Allocate this independent of sound card instance. */
-	bebob = devm_kzalloc(&unit->device, sizeof(struct snd_bebob),
-			     GFP_KERNEL);
-	if (!bebob)
-		return -ENOMEM;
-	bebob->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, bebob);
-
-	bebob->entry = entry;
-	bebob->spec = spec;
-	mutex_init(&bebob->mutex);
-	spin_lock_init(&bebob->lock);
-	init_waitqueue_head(&bebob->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&bebob->dwork, do_registration);
-
-	if (entry->vendor_id != VEN_MAUDIO1 ||
-	    (entry->model_id != MODEL_MAUDIO_FW1814 &&
-	     entry->model_id != MODEL_MAUDIO_PROJECTMIX)) {
-		snd_fw_schedule_registration(unit, &bebob->dwork);
-	} else {
-		/*
-		 * This is a workaround. This bus reset seems to have an effect
-		 * to make devices correctly handling transactions. Without
-		 * this, the devices have gap_count mismatch. This causes much
-		 * failure of transaction.
-		 *
-		 * Just after registration, user-land application receive
-		 * signals from dbus and starts I/Os. To avoid I/Os till the
-		 * future bus reset, registration is done in next update().
-		 */
-		fw_schedule_bus_reset(fw_parent_device(bebob->unit)->card,
-				      false, true);
+	if (entry->vendor_id == VEN_MAUDIO &&
+	    (entry->model_id == MODEL_MAUDIO_FW1814 || entry->model_id == MODEL_MAUDIO_PROJECTMIX)) {
+		// This is a workaround. This bus reset seems to have an effect to make devices
+		// correctly handling transactions. Without this, the devices have gap_count
+		// mismatch. This causes much failure of transaction.
+		//
+		// Just after registration, user-land application receive signals from dbus and
+		// starts I/Os. To avoid I/Os till the future bus reset, registration is done in
+		// next update().
+		fw_schedule_bus_reset(fw_parent_device(bebob->unit)->card, false, true);
 	}
 
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 /*
@@ -330,11 +323,7 @@ bebob_update(struct fw_unit *unit)
 	if (bebob == NULL)
 		return;
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!bebob->registered)
-		snd_fw_schedule_registration(unit, &bebob->dwork);
-	else
-		fcp_bus_reset(bebob->unit);
+	fcp_bus_reset(bebob->unit);
 }
 
 static void bebob_remove(struct fw_unit *unit)
@@ -344,20 +333,8 @@ static void bebob_remove(struct fw_unit *unit)
 	if (bebob == NULL)
 		return;
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&bebob->dwork);
-
-	if (bebob->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(bebob->card);
-	}
-
-	mutex_destroy(&bebob->mutex);
-	fw_unit_put(bebob->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(bebob->card);
 }
 
 static const struct snd_bebob_rate_spec normal_rate_spec = {
@@ -370,6 +347,22 @@ static const struct snd_bebob_spec spec_normal = {
 	.meter	= NULL
 };
 
+#define SPECIFIER_1394TA	0x00a02d
+
+// The immediate entry for version in unit directory differs depending on models:
+//  * 0x010001
+//  * 0x014001
+#define SND_BEBOB_DEV_ENTRY(vendor, model, data) \
+{ \
+	.match_flags	= IEEE1394_MATCH_VENDOR_ID | \
+			  IEEE1394_MATCH_MODEL_ID | \
+			  IEEE1394_MATCH_SPECIFIER_ID, \
+	.vendor_id	= vendor, \
+	.model_id	= model, \
+	.specifier_id	= SPECIFIER_1394TA, \
+	.driver_data	= (kernel_ulong_t)data \
+}
+
 static const struct ieee1394_device_id bebob_id_table[] = {
 	/* Edirol, FA-66 */
 	SND_BEBOB_DEV_ENTRY(VEN_EDIROL, 0x00010049, &spec_normal),
@@ -386,9 +379,9 @@ static const struct ieee1394_device_id bebob_id_table[] = {
 	/* BridgeCo, Audio5 */
 	SND_BEBOB_DEV_ENTRY(VEN_BRIDGECO, 0x00010049, &spec_normal),
 	/* Mackie, Onyx 1220/1620/1640 (Firewire I/O Card) */
-	SND_BEBOB_DEV_ENTRY(VEN_MACKIE2, 0x00010065, &spec_normal),
-	// Mackie, d.2 (Firewire option card) and d.2 Pro (the card is built-in).
-	SND_BEBOB_DEV_ENTRY(VEN_MACKIE1, 0x00010067, &spec_normal),
+	SND_BEBOB_DEV_ENTRY(VEN_MACKIE, 0x00010065, &spec_normal),
+	// Mackie, d.2 (optional Firewire card with DM1000).
+	SND_BEBOB_DEV_ENTRY(VEN_MACKIE, 0x00010067, &spec_normal),
 	/* Stanton, ScratchAmp */
 	SND_BEBOB_DEV_ENTRY(VEN_STANTON, 0x00000001, &spec_normal),
 	/* Tascam, IF-FW DM */
@@ -410,17 +403,20 @@ static const struct ieee1394_device_id bebob_id_table[] = {
 	SND_BEBOB_DEV_ENTRY(VEN_APOGEE, 0x01eeee, &spec_normal),
 	/* ESI, Quatafire610 */
 	SND_BEBOB_DEV_ENTRY(VEN_ESI, 0x00010064, &spec_normal),
-	/* AcousticReality, eARMasterOne */
-	SND_BEBOB_DEV_ENTRY(VEN_ACOUSTIC, 0x00000002, &spec_normal),
+	// AcousticReality, eARMasterOne. Terratec OEM.
+	SND_BEBOB_DEV_ENTRY(VEN_TERRATEC, 0x00000002, &spec_normal),
 	/* CME, MatrixKFW */
 	SND_BEBOB_DEV_ENTRY(VEN_CME, 0x00030000, &spec_normal),
-	/* Phonic, Helix Board 12 MkII */
+	// Phonic Helix Board 12 FireWire MkII.
 	SND_BEBOB_DEV_ENTRY(VEN_PHONIC, 0x00050000, &spec_normal),
-	/* Phonic, Helix Board 18 MkII */
+	// Phonic Helix Board 18 FireWire MkII.
 	SND_BEBOB_DEV_ENTRY(VEN_PHONIC, 0x00060000, &spec_normal),
-	/* Phonic, Helix Board 24 MkII */
+	// Phonic Helix Board 24 FireWire MkII.
 	SND_BEBOB_DEV_ENTRY(VEN_PHONIC, 0x00070000, &spec_normal),
-	/* Phonic, Helix Board 12 Universal/18 Universal/24 Universal */
+	// Phonic FireFly 808 FireWire.
+	SND_BEBOB_DEV_ENTRY(VEN_PHONIC, 0x00080000, &spec_normal),
+	// Phonic FireFly 202, 302, 808 Universal.
+	// Phinic Helix Board 12/18/24 FireWire, 12/18/24 Universal
 	SND_BEBOB_DEV_ENTRY(VEN_PHONIC, 0x00000000, &spec_normal),
 	/* Lynx, Aurora 8/16 (LT-FW) */
 	SND_BEBOB_DEV_ENTRY(VEN_LYNX, 0x00000001, &spec_normal),
@@ -447,45 +443,35 @@ static const struct ieee1394_device_id bebob_id_table[] = {
 	/* Focusrite, SaffirePro 26 I/O */
 	SND_BEBOB_DEV_ENTRY(VEN_FOCUSRITE, 0x00000003, &saffirepro_26_spec),
 	/* Focusrite, SaffirePro 10 I/O */
-	{
-		// The combination of vendor_id and model_id is the same as the
-		// same as the one of Liquid Saffire 56.
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID |
-				  IEEE1394_MATCH_SPECIFIER_ID |
-				  IEEE1394_MATCH_VERSION,
-		.vendor_id	= VEN_FOCUSRITE,
-		.model_id	= 0x000006,
-		.specifier_id	= 0x00a02d,
-		.version	= 0x010001,
-		.driver_data	= (kernel_ulong_t)&saffirepro_10_spec,
-	},
+	SND_BEBOB_DEV_ENTRY(VEN_FOCUSRITE, 0x000006, &saffirepro_10_spec),
 	/* Focusrite, Saffire(no label and LE) */
 	SND_BEBOB_DEV_ENTRY(VEN_FOCUSRITE, MODEL_FOCUSRITE_SAFFIRE_BOTH,
 			    &saffire_spec),
-	/* M-Audio, Firewire 410 */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO2, 0x00010058, NULL),	/* bootloader */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO2, 0x00010046, &maudio_fw410_spec),
+	// M-Audio, Firewire 410. The vendor field is left as BridgeCo. AG.
+	SND_BEBOB_DEV_ENTRY(VEN_BRIDGECO, 0x00010058, NULL),
+	SND_BEBOB_DEV_ENTRY(VEN_BRIDGECO, 0x00010046, &maudio_fw410_spec),
 	/* M-Audio, Firewire Audiophile */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, MODEL_MAUDIO_AUDIOPHILE_BOTH,
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, MODEL_MAUDIO_AUDIOPHILE_BOTH,
 			    &maudio_audiophile_spec),
 	/* M-Audio, Firewire Solo */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, 0x00010062, &maudio_solo_spec),
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, 0x00010062, &maudio_solo_spec),
 	/* M-Audio, Ozonic */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, 0x0000000a, &maudio_ozonic_spec),
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, 0x0000000a, &maudio_ozonic_spec),
 	/* M-Audio NRV10 */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, 0x00010081, &maudio_nrv10_spec),
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, 0x00010081, &maudio_nrv10_spec),
 	/* M-Audio, ProFireLightbridge */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, 0x000100a1, &spec_normal),
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, MODEL_MAUDIO_PROFIRELIGHTBRIDGE, &spec_normal),
 	/* Firewire 1814 */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, 0x00010070, NULL),	/* bootloader */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, MODEL_MAUDIO_FW1814,
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, 0x00010070, NULL),	/* bootloader */
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, MODEL_MAUDIO_FW1814,
 			    &maudio_special_spec),
 	/* M-Audio ProjectMix */
-	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO1, MODEL_MAUDIO_PROJECTMIX,
+	SND_BEBOB_DEV_ENTRY(VEN_MAUDIO, MODEL_MAUDIO_PROJECTMIX,
 			    &maudio_special_spec),
 	/* Digidesign Mbox 2 Pro */
 	SND_BEBOB_DEV_ENTRY(VEN_DIGIDESIGN, 0x0000a9, &spec_normal),
+	// Toneweal FW66.
+	SND_BEBOB_DEV_ENTRY(OUI_SHOUYO, 0x020002, &spec_normal),
 	/* IDs are unknown but able to be supported */
 	/*  Apogee, Mini-ME Firewire */
 	/*  Apogee, Mini-DAC Firewire */
@@ -496,11 +482,6 @@ static const struct ieee1394_device_id bebob_id_table[] = {
 	/*  Infrasonic, Windy6 */
 	/*  Mackie, Digital X Bus x.200 */
 	/*  Mackie, Digital X Bus x.400 */
-	/*  Phonic, HB 12 */
-	/*  Phonic, HB 24 */
-	/*  Phonic, HB 18 */
-	/*  Phonic, FireFly 202 */
-	/*  Phonic, FireFly 302 */
 	/*  Rolf Spuler, Firewire Guitar */
 	{}
 };
diff --git a/sound/firewire/bebob/bebob.h b/sound/firewire/bebob/bebob.h
index 4e0ed84adbee..4d73ecb30d79 100644
--- a/sound/firewire/bebob/bebob.h
+++ b/sound/firewire/bebob/bebob.h
@@ -75,6 +75,11 @@ struct snd_bebob_spec {
 	const struct snd_bebob_meter_spec *meter;
 };
 
+enum snd_bebob_quirk {
+	SND_BEBOB_QUIRK_INITIAL_DISCONTINUOUS_DBC = (1 << 0),
+	SND_BEBOB_QUIRK_WRONG_DBC		  = (1 << 1),
+};
+
 struct snd_bebob {
 	struct snd_card *card;
 	struct fw_unit *unit;
@@ -83,11 +88,8 @@ struct snd_bebob {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
-	const struct ieee1394_device_id *entry;
 	const struct snd_bebob_spec *spec;
+	unsigned int quirks;	// Combination of snd_bebob_quirk enumerations.
 
 	unsigned int midi_input_ports;
 	unsigned int midi_output_ports;
@@ -113,9 +115,6 @@ struct snd_bebob {
 	/* for M-Audio special devices */
 	void *maudio_special_quirk;
 
-	/* For BeBoB version quirk. */
-	unsigned int version;
-
 	struct amdtp_domain domain;
 };
 
@@ -254,13 +253,4 @@ extern const struct snd_bebob_spec maudio_special_spec;
 int snd_bebob_maudio_special_discover(struct snd_bebob *bebob, bool is1814);
 int snd_bebob_maudio_load_firmware(struct fw_unit *unit);
 
-#define SND_BEBOB_DEV_ENTRY(vendor, model, data) \
-{ \
-	.match_flags	= IEEE1394_MATCH_VENDOR_ID | \
-			  IEEE1394_MATCH_MODEL_ID, \
-	.vendor_id	= vendor, \
-	.model_id	= model, \
-	.driver_data	= (kernel_ulong_t)data \
-}
-
 #endif
diff --git a/sound/firewire/bebob/bebob_stream.c b/sound/firewire/bebob/bebob_stream.c
index b612ee3e33b6..8629b14ded76 100644
--- a/sound/firewire/bebob/bebob_stream.c
+++ b/sound/firewire/bebob/bebob_stream.c
@@ -7,8 +7,7 @@
 
 #include "./bebob.h"
 
-#define CALLBACK_TIMEOUT	2500
-#define FW_ISO_RESOURCE_DELAY	1000
+#define READY_TIMEOUT_MS	4000
 
 /*
  * NOTE;
@@ -402,12 +401,6 @@ static void break_both_connections(struct snd_bebob *bebob)
 {
 	cmp_connection_break(&bebob->in_conn);
 	cmp_connection_break(&bebob->out_conn);
-
-	// These models seem to be in transition state for a longer time. When
-	// accessing in the state, any transactions is corrupted. In the worst
-	// case, the device is going to reboot.
-	if (bebob->version < 2)
-		msleep(600);
 }
 
 static int start_stream(struct snd_bebob *bebob, struct amdtp_stream *stream)
@@ -437,6 +430,7 @@ static int start_stream(struct snd_bebob *bebob, struct amdtp_stream *stream)
 
 static int init_stream(struct snd_bebob *bebob, struct amdtp_stream *stream)
 {
+	unsigned int flags = CIP_BLOCKING;
 	enum amdtp_stream_direction dir_stream;
 	struct cmp_connection *conn;
 	enum cmp_direction dir_conn;
@@ -452,32 +446,21 @@ static int init_stream(struct snd_bebob *bebob, struct amdtp_stream *stream)
 		dir_conn = CMP_INPUT;
 	}
 
+	if (stream == &bebob->tx_stream) {
+		if (bebob->quirks & SND_BEBOB_QUIRK_WRONG_DBC)
+			flags |= CIP_EMPTY_HAS_WRONG_DBC;
+	}
+
 	err = cmp_connection_init(conn, bebob->unit, dir_conn, 0);
 	if (err < 0)
 		return err;
 
-	err = amdtp_am824_init(stream, bebob->unit, dir_stream, CIP_BLOCKING);
+	err = amdtp_am824_init(stream, bebob->unit, dir_stream, flags);
 	if (err < 0) {
 		cmp_connection_destroy(conn);
 		return err;
 	}
 
-	if (stream == &bebob->tx_stream) {
-		// BeBoB v3 transfers packets with these qurks:
-		//  - In the beginning of streaming, the value of dbc is
-		//    incremented even if no data blocks are transferred.
-		//  - The value of dbc is reset suddenly.
-		if (bebob->version > 2)
-			bebob->tx_stream.flags |= CIP_EMPTY_HAS_WRONG_DBC |
-						  CIP_SKIP_DBC_ZERO_CHECK;
-
-		// At high sampling rate, M-Audio special firmware transmits
-		// empty packet with the value of dbc incremented by 8 but the
-		// others are valid to IEC 61883-1.
-		if (bebob->maudio_special_quirk)
-			bebob->tx_stream.flags |= CIP_EMPTY_HAS_WRONG_DBC;
-	}
-
 	return 0;
 }
 
@@ -624,9 +607,8 @@ int snd_bebob_stream_start_duplex(struct snd_bebob *bebob)
 
 	if (!amdtp_stream_running(&bebob->rx_stream)) {
 		enum snd_bebob_clock_type src;
-		struct amdtp_stream *master, *slave;
 		unsigned int curr_rate;
-		unsigned int ir_delay_cycle;
+		unsigned int tx_init_skip_cycles;
 
 		if (bebob->maudio_special_quirk) {
 			err = bebob->spec->rate->get(bebob, &curr_rate);
@@ -638,36 +620,28 @@ int snd_bebob_stream_start_duplex(struct snd_bebob *bebob)
 		if (err < 0)
 			return err;
 
-		if (src != SND_BEBOB_CLOCK_TYPE_SYT) {
-			master = &bebob->tx_stream;
-			slave = &bebob->rx_stream;
-		} else {
-			master = &bebob->rx_stream;
-			slave = &bebob->tx_stream;
-		}
-
-		err = start_stream(bebob, master);
+		err = start_stream(bebob, &bebob->rx_stream);
 		if (err < 0)
 			goto error;
 
-		err = start_stream(bebob, slave);
+		err = start_stream(bebob, &bebob->tx_stream);
 		if (err < 0)
 			goto error;
 
-		// The device postpones start of transmission mostly for 1 sec
-		// after receives packets firstly. For safe, IR context starts
-		// 0.4 sec (=3200 cycles) later to version 1 or 2 firmware,
-		// 2.0 sec (=16000 cycles) for version 3 firmware. This is
-		// within 2.5 sec (=CALLBACK_TIMEOUT).
-		// Furthermore, some devices transfer isoc packets with
-		// discontinuous counter in the beginning of packet streaming.
-		// The delay has an effect to avoid detection of this
-		// discontinuity.
-		if (bebob->version < 2)
-			ir_delay_cycle = 3200;
+		if (!(bebob->quirks & SND_BEBOB_QUIRK_INITIAL_DISCONTINUOUS_DBC))
+			tx_init_skip_cycles = 0;
 		else
-			ir_delay_cycle = 16000;
-		err = amdtp_domain_start(&bebob->domain, ir_delay_cycle);
+			tx_init_skip_cycles = 16000;
+
+		// MEMO: Some devices start packet transmission long enough after establishment of
+		// CMP connection. In the early stage of packet streaming, any device transfers
+		// NODATA packets. After several hundred cycles, it begins to multiplex event into
+		// the packet with adequate value of syt field in CIP header. Some devices are
+		// strictly to generate any discontinuity in the sequence of tx packet when they
+		// receives inadequate sequence of value in syt field of CIP header. In the case,
+		// the request to break CMP connection is often corrupted, then any transaction
+		// results in unrecoverable error, sometimes generate bus-reset.
+		err = amdtp_domain_start(&bebob->domain, tx_init_skip_cycles, true, false);
 		if (err < 0)
 			goto error;
 
@@ -684,10 +658,9 @@ int snd_bebob_stream_start_duplex(struct snd_bebob *bebob)
 			}
 		}
 
-		if (!amdtp_stream_wait_callback(&bebob->rx_stream,
-						CALLBACK_TIMEOUT) ||
-		    !amdtp_stream_wait_callback(&bebob->tx_stream,
-						CALLBACK_TIMEOUT)) {
+		// Some devices postpone start of transmission mostly for 1 sec after receives
+		// packets firstly.
+		if (!amdtp_domain_wait_ready(&bebob->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
@@ -883,6 +856,11 @@ static int detect_midi_ports(struct snd_bebob *bebob,
 		err = avc_bridgeco_get_plug_ch_count(bebob->unit, addr, &ch_count);
 		if (err < 0)
 			break;
+		// Yamaha GO44, GO46, Terratec Phase 24, Phase x24 reports 0 for the number of
+		// channels in external output plug 3 (MIDI type) even if it has a pair of physical
+		// MIDI jacks. As a workaround, assume it as one.
+		if (ch_count == 0)
+			ch_count = 1;
 		*midi_ports += ch_count;
 	}
 
@@ -961,12 +939,12 @@ int snd_bebob_stream_discover(struct snd_bebob *bebob)
 	if (err < 0)
 		goto end;
 
-	err = detect_midi_ports(bebob, bebob->rx_stream_formations, addr, AVC_BRIDGECO_PLUG_DIR_IN,
+	err = detect_midi_ports(bebob, bebob->tx_stream_formations, addr, AVC_BRIDGECO_PLUG_DIR_IN,
 				plugs[2], &bebob->midi_input_ports);
 	if (err < 0)
 		goto end;
 
-	err = detect_midi_ports(bebob, bebob->tx_stream_formations, addr, AVC_BRIDGECO_PLUG_DIR_OUT,
+	err = detect_midi_ports(bebob, bebob->rx_stream_formations, addr, AVC_BRIDGECO_PLUG_DIR_OUT,
 				plugs[3], &bebob->midi_output_ports);
 	if (err < 0)
 		goto end;
diff --git a/sound/firewire/dice/dice-stream.c b/sound/firewire/dice/dice-stream.c
index c4dfe76500c2..f99e00083141 100644
--- a/sound/firewire/dice/dice-stream.c
+++ b/sound/firewire/dice/dice-stream.c
@@ -8,8 +8,8 @@
 
 #include "dice.h"
 
-#define	CALLBACK_TIMEOUT	200
-#define NOTIFICATION_TIMEOUT_MS	(2 * MSEC_PER_SEC)
+#define	READY_TIMEOUT_MS	200
+#define NOTIFICATION_TIMEOUT_MS	100
 
 struct reg_params {
 	unsigned int count;
@@ -57,13 +57,9 @@ int snd_dice_stream_get_rate_mode(struct snd_dice *dice, unsigned int rate,
 	return -EINVAL;
 }
 
-/*
- * This operation has an effect to synchronize GLOBAL_STATUS/GLOBAL_SAMPLE_RATE
- * to GLOBAL_STATUS. Especially, just after powering on, these are different.
- */
-static int ensure_phase_lock(struct snd_dice *dice, unsigned int rate)
+static int select_clock(struct snd_dice *dice, unsigned int rate)
 {
-	__be32 reg, nominal;
+	__be32 reg;
 	u32 data;
 	int i;
 	int err;
@@ -94,19 +90,8 @@ static int ensure_phase_lock(struct snd_dice *dice, unsigned int rate)
 		return err;
 
 	if (wait_for_completion_timeout(&dice->clock_accepted,
-			msecs_to_jiffies(NOTIFICATION_TIMEOUT_MS)) == 0) {
-		/*
-		 * Old versions of Dice firmware transfer no notification when
-		 * the same clock status as current one is set. In this case,
-		 * just check current clock status.
-		 */
-		err = snd_dice_transaction_read_global(dice, GLOBAL_STATUS,
-						&nominal, sizeof(nominal));
-		if (err < 0)
-			return err;
-		if (!(be32_to_cpu(nominal) & STATUS_SOURCE_LOCKED))
-			return -ETIMEDOUT;
-	}
+			msecs_to_jiffies(NOTIFICATION_TIMEOUT_MS)) == 0)
+		return -ETIMEDOUT;
 
 	return 0;
 }
@@ -304,7 +289,7 @@ int snd_dice_stream_reserve_duplex(struct snd_dice *dice, unsigned int rate,
 		// Just after owning the unit (GLOBAL_OWNER), the unit can
 		// return invalid stream formats. Selecting clock parameters
 		// have an effect for the unit to refine it.
-		err = ensure_phase_lock(dice, rate);
+		err = select_clock(dice, rate);
 		if (err < 0)
 			return err;
 
@@ -459,20 +444,17 @@ int snd_dice_stream_start_duplex(struct snd_dice *dice)
 			goto error;
 		}
 
-		err = amdtp_domain_start(&dice->domain, 0);
+		// MEMO: The device immediately starts packet transmission when enabled. Some
+		// devices are strictly to generate any discontinuity in the sequence of tx packet
+		// when they receives invalid sequence of presentation time in CIP header. The
+		// sequence replay for media clock recovery can suppress the behaviour.
+		err = amdtp_domain_start(&dice->domain, 0, true, false);
 		if (err < 0)
 			goto error;
 
-		for (i = 0; i < MAX_STREAMS; i++) {
-			if ((i < tx_params.count &&
-			    !amdtp_stream_wait_callback(&dice->tx_stream[i],
-							CALLBACK_TIMEOUT)) ||
-			    (i < rx_params.count &&
-			     !amdtp_stream_wait_callback(&dice->rx_stream[i],
-							 CALLBACK_TIMEOUT))) {
-				err = -ETIMEDOUT;
-				goto error;
-			}
+		if (!amdtp_domain_wait_ready(&dice->domain, READY_TIMEOUT_MS)) {
+			err = -ETIMEDOUT;
+			goto error;
 		}
 	}
 
@@ -653,7 +635,7 @@ int snd_dice_stream_detect_current_formats(struct snd_dice *dice)
 	 * invalid stream formats. Selecting clock parameters have an effect
 	 * for the unit to refine it.
 	 */
-	err = ensure_phase_lock(dice, rate);
+	err = select_clock(dice, rate);
 	if (err < 0)
 		return err;
 
diff --git a/sound/firewire/dice/dice-transaction.c b/sound/firewire/dice/dice-transaction.c
index 2c0dde29a024..92941ef83cd5 100644
--- a/sound/firewire/dice/dice-transaction.c
+++ b/sound/firewire/dice/dice-transaction.c
@@ -155,7 +155,7 @@ static void dice_notification(struct fw_card *card, struct fw_request *request,
 
 	fw_send_response(card, request, RCODE_COMPLETE);
 
-	if (bits & NOTIFY_LOCK_CHG)
+	if (bits & NOTIFY_CLOCK_ACCEPTED)
 		complete(&dice->clock_accepted);
 	wake_up(&dice->hwdep_wait);
 }
diff --git a/sound/firewire/dice/dice.c b/sound/firewire/dice/dice.c
index 239d164b0eea..f75902bc8e74 100644
--- a/sound/firewire/dice/dice.c
+++ b/sound/firewire/dice/dice.c
@@ -135,22 +135,51 @@ static void dice_card_free(struct snd_card *card)
 
 	snd_dice_stream_destroy_duplex(dice);
 	snd_dice_transaction_destroy(dice);
+
+	mutex_destroy(&dice->mutex);
+	fw_unit_put(dice->unit);
 }
 
-static void do_registration(struct work_struct *work)
+static int dice_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_dice *dice = container_of(work, struct snd_dice, dwork.work);
+	struct snd_card *card;
+	struct snd_dice *dice;
+	snd_dice_detect_formats_t detect_formats;
 	int err;
 
-	if (dice->registered)
-		return;
+	if (!entry->driver_data && entry->vendor_id != OUI_SSL) {
+		err = check_dice_category(unit);
+		if (err < 0)
+			return -ENODEV;
+	}
 
-	err = snd_card_new(&dice->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &dice->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*dice), &card);
 	if (err < 0)
-		return;
-	dice->card->private_free = dice_card_free;
-	dice->card->private_data = dice;
+		return err;
+	card->private_free = dice_card_free;
+
+	dice = card->private_data;
+	dice->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, dice);
+	dice->card = card;
+
+	if (!entry->driver_data)
+		detect_formats = snd_dice_stream_detect_current_formats;
+	else
+		detect_formats = (snd_dice_detect_formats_t)entry->driver_data;
+
+	// Below models are compliant to IEC 61883-1/6 and have no quirk at high sampling transfer
+	// frequency.
+	// * Avid M-Box 3 Pro
+	// * M-Audio Profire 610
+	// * M-Audio Profire 2626
+	if (entry->vendor_id == OUI_MAUDIO || entry->vendor_id == OUI_AVID)
+		dice->disable_double_pcm_frames = true;
+
+	spin_lock_init(&dice->lock);
+	mutex_init(&dice->mutex);
+	init_completion(&dice->clock_accepted);
+	init_waitqueue_head(&dice->hwdep_wait);
 
 	err = snd_dice_transaction_init(dice);
 	if (err < 0)
@@ -162,7 +191,7 @@ static void do_registration(struct work_struct *work)
 
 	dice_card_strings(dice);
 
-	err = dice->detect_formats(dice);
+	err = detect_formats(dice);
 	if (err < 0)
 		goto error;
 
@@ -184,105 +213,34 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(dice->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	dice->registered = true;
-
-	return;
-error:
-	snd_card_free(dice->card);
-	dev_info(&dice->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int dice_probe(struct fw_unit *unit,
-		      const struct ieee1394_device_id *entry)
-{
-	struct snd_dice *dice;
-	int err;
-
-	if (!entry->driver_data && entry->vendor_id != OUI_SSL) {
-		err = check_dice_category(unit);
-		if (err < 0)
-			return -ENODEV;
-	}
-
-	/* Allocate this independent of sound card instance. */
-	dice = devm_kzalloc(&unit->device, sizeof(struct snd_dice), GFP_KERNEL);
-	if (!dice)
-		return -ENOMEM;
-	dice->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, dice);
-
-	if (!entry->driver_data) {
-		dice->detect_formats = snd_dice_stream_detect_current_formats;
-	} else {
-		dice->detect_formats =
-				(snd_dice_detect_formats_t)entry->driver_data;
-	}
-
-	// Below models are compliant to IEC 61883-1/6 and have no quirk at high sampling transfer
-	// frequency.
-	// * Avid M-Box 3 Pro
-	// * M-Audio Profire 610
-	// * M-Audio Profire 2626
-	if (entry->vendor_id == OUI_MAUDIO || entry->vendor_id == OUI_AVID)
-		dice->disable_double_pcm_frames = true;
-
-	spin_lock_init(&dice->lock);
-	mutex_init(&dice->mutex);
-	init_completion(&dice->clock_accepted);
-	init_waitqueue_head(&dice->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&dice->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &dice->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void dice_remove(struct fw_unit *unit)
 {
 	struct snd_dice *dice = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&dice->dwork);
-
-	if (dice->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(dice->card);
-	}
-
-	mutex_destroy(&dice->mutex);
-	fw_unit_put(dice->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(dice->card);
 }
 
 static void dice_bus_reset(struct fw_unit *unit)
 {
 	struct snd_dice *dice = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!dice->registered)
-		snd_fw_schedule_registration(unit, &dice->dwork);
-
 	/* The handler address register becomes initialized. */
 	snd_dice_transaction_reinit(dice);
 
-	/*
-	 * After registration, userspace can start packet streaming, then this
-	 * code block works fine.
-	 */
-	if (dice->registered) {
-		mutex_lock(&dice->mutex);
-		snd_dice_stream_update_duplex(dice);
-		mutex_unlock(&dice->mutex);
-	}
+	mutex_lock(&dice->mutex);
+	snd_dice_stream_update_duplex(dice);
+	mutex_unlock(&dice->mutex);
 }
 
 #define DICE_INTERFACE	0x000001
diff --git a/sound/firewire/dice/dice.h b/sound/firewire/dice/dice.h
index 3c967d1b3605..fd440cc625f9 100644
--- a/sound/firewire/dice/dice.h
+++ b/sound/firewire/dice/dice.h
@@ -78,9 +78,6 @@ struct snd_dice {
 	spinlock_t lock;
 	struct mutex mutex;
 
-	bool registered;
-	struct delayed_work dwork;
-
 	/* Offsets for sub-addresses */
 	unsigned int global_offset;
 	unsigned int rx_offset;
@@ -93,7 +90,6 @@ struct snd_dice {
 	unsigned int rx_pcm_chs[MAX_STREAMS][SND_DICE_RATE_MODE_COUNT];
 	unsigned int tx_midi_ports[MAX_STREAMS];
 	unsigned int rx_midi_ports[MAX_STREAMS];
-	snd_dice_detect_formats_t detect_formats;
 
 	struct fw_address_handler notification_handler;
 	int owner_generation;
diff --git a/sound/firewire/digi00x/amdtp-dot.c b/sound/firewire/digi00x/amdtp-dot.c
index d613642a2ce3..59b86c8d89e1 100644
--- a/sound/firewire/digi00x/amdtp-dot.c
+++ b/sound/firewire/digi00x/amdtp-dot.c
@@ -396,16 +396,13 @@ int amdtp_dot_init(struct amdtp_stream *s, struct fw_unit *unit,
 		 enum amdtp_stream_direction dir)
 {
 	amdtp_stream_process_ctx_payloads_t process_ctx_payloads;
-	enum cip_flags flags;
+	unsigned int flags = CIP_NONBLOCKING | CIP_UNAWARE_SYT;
 
 	// Use different mode between incoming/outgoing.
-	if (dir == AMDTP_IN_STREAM) {
-		flags = CIP_NONBLOCKING;
+	if (dir == AMDTP_IN_STREAM)
 		process_ctx_payloads = process_ir_ctx_payloads;
-	} else {
-		flags = CIP_BLOCKING;
+	else
 		process_ctx_payloads = process_it_ctx_payloads;
-	}
 
 	return amdtp_stream_init(s, unit, dir, flags, CIP_FMT_AM,
 				process_ctx_payloads, sizeof(struct amdtp_dot));
diff --git a/sound/firewire/digi00x/digi00x-stream.c b/sound/firewire/digi00x/digi00x-stream.c
index 405d6903bfbc..a15f55b0dce3 100644
--- a/sound/firewire/digi00x/digi00x-stream.c
+++ b/sound/firewire/digi00x/digi00x-stream.c
@@ -7,7 +7,7 @@
 
 #include "digi00x.h"
 
-#define CALLBACK_TIMEOUT 500
+#define READY_TIMEOUT_MS	200
 
 const unsigned int snd_dg00x_stream_rates[SND_DG00X_RATE_COUNT] = {
 	[SND_DG00X_RATE_44100] = 44100,
@@ -375,14 +375,15 @@ int snd_dg00x_stream_start_duplex(struct snd_dg00x *dg00x)
 		if (err < 0)
 			goto error;
 
-		err = amdtp_domain_start(&dg00x->domain, 0);
+		// NOTE: The device doesn't start packet transmission till receiving any packet.
+		// It ignores presentation time expressed by the value of syt field of CIP header
+		// in received packets. The sequence of the number of data blocks per packet is
+		// important for media clock recovery.
+		err = amdtp_domain_start(&dg00x->domain, 0, true, true);
 		if (err < 0)
 			goto error;
 
-		if (!amdtp_stream_wait_callback(&dg00x->rx_stream,
-						CALLBACK_TIMEOUT) ||
-		    !amdtp_stream_wait_callback(&dg00x->tx_stream,
-						CALLBACK_TIMEOUT)) {
+		if (!amdtp_domain_wait_ready(&dg00x->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
diff --git a/sound/firewire/digi00x/digi00x.c b/sound/firewire/digi00x/digi00x.c
index ab8408966ec3..995302808c27 100644
--- a/sound/firewire/digi00x/digi00x.c
+++ b/sound/firewire/digi00x/digi00x.c
@@ -47,23 +47,32 @@ static void dg00x_card_free(struct snd_card *card)
 
 	snd_dg00x_stream_destroy_duplex(dg00x);
 	snd_dg00x_transaction_unregister(dg00x);
+
+	mutex_destroy(&dg00x->mutex);
+	fw_unit_put(dg00x->unit);
 }
 
-static void do_registration(struct work_struct *work)
+static int snd_dg00x_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_dg00x *dg00x =
-			container_of(work, struct snd_dg00x, dwork.work);
+	struct snd_card *card;
+	struct snd_dg00x *dg00x;
 	int err;
 
-	if (dg00x->registered)
-		return;
-
-	err = snd_card_new(&dg00x->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &dg00x->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*dg00x), &card);
 	if (err < 0)
-		return;
-	dg00x->card->private_free = dg00x_card_free;
-	dg00x->card->private_data = dg00x;
+		return err;
+	card->private_free = dg00x_card_free;
+
+	dg00x = card->private_data;
+	dg00x->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, dg00x);
+	dg00x->card = card;
+
+	mutex_init(&dg00x->mutex);
+	spin_lock_init(&dg00x->lock);
+	init_waitqueue_head(&dg00x->hwdep_wait);
+
+	dg00x->is_console = entry->model_id == MODEL_CONSOLE;
 
 	err = name_card(dg00x);
 	if (err < 0)
@@ -91,85 +100,33 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(dg00x->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	dg00x->registered = true;
-
-	return;
-error:
-	snd_card_free(dg00x->card);
-	dev_info(&dg00x->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int snd_dg00x_probe(struct fw_unit *unit,
-			   const struct ieee1394_device_id *entry)
-{
-	struct snd_dg00x *dg00x;
-
-	/* Allocate this independent of sound card instance. */
-	dg00x = devm_kzalloc(&unit->device, sizeof(struct snd_dg00x),
-			     GFP_KERNEL);
-	if (!dg00x)
-		return -ENOMEM;
-
-	dg00x->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, dg00x);
-
-	mutex_init(&dg00x->mutex);
-	spin_lock_init(&dg00x->lock);
-	init_waitqueue_head(&dg00x->hwdep_wait);
-
-	dg00x->is_console = entry->model_id == MODEL_CONSOLE;
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&dg00x->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &dg00x->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void snd_dg00x_update(struct fw_unit *unit)
 {
 	struct snd_dg00x *dg00x = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!dg00x->registered)
-		snd_fw_schedule_registration(unit, &dg00x->dwork);
-
 	snd_dg00x_transaction_reregister(dg00x);
 
-	/*
-	 * After registration, userspace can start packet streaming, then this
-	 * code block works fine.
-	 */
-	if (dg00x->registered) {
-		mutex_lock(&dg00x->mutex);
-		snd_dg00x_stream_update_duplex(dg00x);
-		mutex_unlock(&dg00x->mutex);
-	}
+	mutex_lock(&dg00x->mutex);
+	snd_dg00x_stream_update_duplex(dg00x);
+	mutex_unlock(&dg00x->mutex);
 }
 
 static void snd_dg00x_remove(struct fw_unit *unit)
 {
 	struct snd_dg00x *dg00x = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&dg00x->dwork);
-
-	if (dg00x->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(dg00x->card);
-	}
-
-	mutex_destroy(&dg00x->mutex);
-	fw_unit_put(dg00x->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(dg00x->card);
 }
 
 static const struct ieee1394_device_id snd_dg00x_id_table[] = {
diff --git a/sound/firewire/digi00x/digi00x.h b/sound/firewire/digi00x/digi00x.h
index 129de8edd5ea..82b647d383c5 100644
--- a/sound/firewire/digi00x/digi00x.h
+++ b/sound/firewire/digi00x/digi00x.h
@@ -37,9 +37,6 @@ struct snd_dg00x {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
 	struct amdtp_stream tx_stream;
 	struct fw_iso_resources tx_resources;
 
diff --git a/sound/firewire/fireface/amdtp-ff.c b/sound/firewire/fireface/amdtp-ff.c
index 119c0076b17a..98177b0666d3 100644
--- a/sound/firewire/fireface/amdtp-ff.c
+++ b/sound/firewire/fireface/amdtp-ff.c
@@ -168,6 +168,6 @@ int amdtp_ff_init(struct amdtp_stream *s, struct fw_unit *unit,
 	else
 		process_ctx_payloads = process_it_ctx_payloads;
 
-	return amdtp_stream_init(s, unit, dir, CIP_NO_HEADER, 0,
+	return amdtp_stream_init(s, unit, dir, CIP_BLOCKING | CIP_UNAWARE_SYT | CIP_NO_HEADER, 0,
 				 process_ctx_payloads, sizeof(struct amdtp_ff));
 }
diff --git a/sound/firewire/fireface/ff-stream.c b/sound/firewire/fireface/ff-stream.c
index 5452115c0ef9..95bf405adb3d 100644
--- a/sound/firewire/fireface/ff-stream.c
+++ b/sound/firewire/fireface/ff-stream.c
@@ -7,7 +7,7 @@
 
 #include "ff.h"
 
-#define CALLBACK_TIMEOUT_MS	200
+#define READY_TIMEOUT_MS	200
 
 int snd_ff_stream_get_multiplier_mode(enum cip_sfc sfc,
 				      enum snd_ff_stream_mode *mode)
@@ -199,14 +199,15 @@ int snd_ff_stream_start_duplex(struct snd_ff *ff, unsigned int rate)
 		if (err < 0)
 			goto error;
 
-		err = amdtp_domain_start(&ff->domain, 0);
+		// NOTE: The device doesn't transfer packets unless receiving any packet. The
+		// sequence of tx packets includes cycle skip corresponding to empty packet or
+		// NODATA packet in IEC 61883-1/6. The sequence of the number of data blocks per
+		// packet is important for media clock recovery.
+		err = amdtp_domain_start(&ff->domain, 0, true, true);
 		if (err < 0)
 			goto error;
 
-		if (!amdtp_stream_wait_callback(&ff->rx_stream,
-						CALLBACK_TIMEOUT_MS) ||
-		    !amdtp_stream_wait_callback(&ff->tx_stream,
-						CALLBACK_TIMEOUT_MS)) {
+		if (!amdtp_domain_wait_ready(&ff->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
diff --git a/sound/firewire/fireface/ff.c b/sound/firewire/fireface/ff.c
index bc39269415d2..7bf51d062021 100644
--- a/sound/firewire/fireface/ff.c
+++ b/sound/firewire/fireface/ff.c
@@ -42,22 +42,33 @@ static void ff_card_free(struct snd_card *card)
 
 	snd_ff_stream_destroy_duplex(ff);
 	snd_ff_transaction_unregister(ff);
+
+	mutex_destroy(&ff->mutex);
+	fw_unit_put(ff->unit);
 }
 
-static void do_registration(struct work_struct *work)
+static int snd_ff_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_ff *ff = container_of(work, struct snd_ff, dwork.work);
+	struct snd_card *card;
+	struct snd_ff *ff;
 	int err;
 
-	if (ff->registered)
-		return;
-
-	err = snd_card_new(&ff->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &ff->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*ff), &card);
 	if (err < 0)
-		return;
-	ff->card->private_free = ff_card_free;
-	ff->card->private_data = ff;
+		return err;
+	card->private_free = ff_card_free;
+
+	ff = card->private_data;
+	ff->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, ff);
+	ff->card = card;
+
+	mutex_init(&ff->mutex);
+	spin_lock_init(&ff->lock);
+	init_waitqueue_head(&ff->hwdep_wait);
+
+	ff->unit_version = entry->version;
+	ff->spec = (const struct snd_ff_spec *)entry->driver_data;
 
 	err = snd_ff_transaction_register(ff);
 	if (err < 0)
@@ -83,76 +94,31 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(ff->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	ff->registered = true;
-
-	return;
-error:
-	snd_card_free(ff->card);
-	dev_info(&ff->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int snd_ff_probe(struct fw_unit *unit,
-			   const struct ieee1394_device_id *entry)
-{
-	struct snd_ff *ff;
-
-	ff = devm_kzalloc(&unit->device, sizeof(struct snd_ff), GFP_KERNEL);
-	if (!ff)
-		return -ENOMEM;
-	ff->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, ff);
-
-	mutex_init(&ff->mutex);
-	spin_lock_init(&ff->lock);
-	init_waitqueue_head(&ff->hwdep_wait);
-
-	ff->unit_version = entry->version;
-	ff->spec = (const struct snd_ff_spec *)entry->driver_data;
-
-	/* Register this sound card later. */
-	INIT_DEFERRABLE_WORK(&ff->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &ff->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void snd_ff_update(struct fw_unit *unit)
 {
 	struct snd_ff *ff = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!ff->registered)
-		snd_fw_schedule_registration(unit, &ff->dwork);
-
 	snd_ff_transaction_reregister(ff);
 
-	if (ff->registered)
-		snd_ff_stream_update_duplex(ff);
+	snd_ff_stream_update_duplex(ff);
 }
 
 static void snd_ff_remove(struct fw_unit *unit)
 {
 	struct snd_ff *ff = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_work_sync(&ff->dwork.work);
-
-	if (ff->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(ff->card);
-	}
-
-	mutex_destroy(&ff->mutex);
-	fw_unit_put(ff->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(ff->card);
 }
 
 static const struct snd_ff_spec spec_ff800 = {
diff --git a/sound/firewire/fireface/ff.h b/sound/firewire/fireface/ff.h
index 705e7df4f929..0535f0b58b67 100644
--- a/sound/firewire/fireface/ff.h
+++ b/sound/firewire/fireface/ff.h
@@ -69,9 +69,6 @@ struct snd_ff {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
 	enum snd_ff_unit_version unit_version;
 	const struct snd_ff_spec *spec;
 
diff --git a/sound/firewire/fireworks/fireworks.c b/sound/firewire/fireworks/fireworks.c
index b1cc013a3540..ffb6dd796243 100644
--- a/sound/firewire/fireworks/fireworks.c
+++ b/sound/firewire/fireworks/fireworks.c
@@ -194,19 +194,19 @@ efw_card_free(struct snd_card *card)
 
 	snd_efw_stream_destroy_duplex(efw);
 	snd_efw_transaction_remove_instance(efw);
+
+	mutex_destroy(&efw->mutex);
+	fw_unit_put(efw->unit);
 }
 
-static void
-do_registration(struct work_struct *work)
+static int efw_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_efw *efw = container_of(work, struct snd_efw, dwork.work);
 	unsigned int card_index;
+	struct snd_card *card;
+	struct snd_efw *efw;
 	int err;
 
-	if (efw->registered)
-		return;
-
-	/* check registered cards */
+	// check registered cards.
 	mutex_lock(&devices_mutex);
 	for (card_index = 0; card_index < SNDRV_CARDS; ++card_index) {
 		if (!test_bit(card_index, devices_used) && enable[card_index])
@@ -214,26 +214,32 @@ do_registration(struct work_struct *work)
 	}
 	if (card_index >= SNDRV_CARDS) {
 		mutex_unlock(&devices_mutex);
-		return;
+		return -ENOENT;
 	}
 
-	err = snd_card_new(&efw->unit->device, index[card_index],
-			   id[card_index], THIS_MODULE, 0, &efw->card);
+	err = snd_card_new(&unit->device, index[card_index], id[card_index], THIS_MODULE,
+			   sizeof(*efw), &card);
 	if (err < 0) {
 		mutex_unlock(&devices_mutex);
-		return;
+		return err;
 	}
+	card->private_free = efw_card_free;
 	set_bit(card_index, devices_used);
 	mutex_unlock(&devices_mutex);
 
-	efw->card->private_free = efw_card_free;
-	efw->card->private_data = efw;
+	efw = card->private_data;
+	efw->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, efw);
+	efw->card = card;
+	efw->card_index = card_index;
+
+	mutex_init(&efw->mutex);
+	spin_lock_init(&efw->lock);
+	init_waitqueue_head(&efw->hwdep_wait);
 
-	/* prepare response buffer */
-	snd_efw_resp_buf_size = clamp(snd_efw_resp_buf_size,
-				      SND_EFW_RESPONSE_MAXIMUM_BYTES, 4096U);
-	efw->resp_buf = devm_kzalloc(&efw->card->card_dev,
-				     snd_efw_resp_buf_size, GFP_KERNEL);
+	// prepare response buffer.
+	snd_efw_resp_buf_size = clamp(snd_efw_resp_buf_size, SND_EFW_RESPONSE_MAXIMUM_BYTES, 4096U);
+	efw->resp_buf = devm_kzalloc(&card->card_dev, snd_efw_resp_buf_size, GFP_KERNEL);
 	if (!efw->resp_buf) {
 		err = -ENOMEM;
 		goto error;
@@ -265,80 +271,48 @@ do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(efw->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	efw->registered = true;
-
-	return;
-error:
-	snd_card_free(efw->card);
-	dev_info(&efw->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int
-efw_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
-{
-	struct snd_efw *efw;
-
-	efw = devm_kzalloc(&unit->device, sizeof(struct snd_efw), GFP_KERNEL);
-	if (efw == NULL)
-		return -ENOMEM;
-	efw->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, efw);
-
-	mutex_init(&efw->mutex);
-	spin_lock_init(&efw->lock);
-	init_waitqueue_head(&efw->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&efw->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &efw->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void efw_update(struct fw_unit *unit)
 {
 	struct snd_efw *efw = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!efw->registered)
-		snd_fw_schedule_registration(unit, &efw->dwork);
-
 	snd_efw_transaction_bus_reset(efw->unit);
 
-	/*
-	 * After registration, userspace can start packet streaming, then this
-	 * code block works fine.
-	 */
-	if (efw->registered) {
-		mutex_lock(&efw->mutex);
-		snd_efw_stream_update_duplex(efw);
-		mutex_unlock(&efw->mutex);
-	}
+	mutex_lock(&efw->mutex);
+	snd_efw_stream_update_duplex(efw);
+	mutex_unlock(&efw->mutex);
 }
 
 static void efw_remove(struct fw_unit *unit)
 {
 	struct snd_efw *efw = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&efw->dwork);
-
-	if (efw->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(efw->card);
-	}
+	// Block till all of ALSA character devices are released.
+	snd_card_free(efw->card);
+}
 
-	mutex_destroy(&efw->mutex);
-	fw_unit_put(efw->unit);
+#define SPECIFIER_1394TA	0x00a02d
+#define VERSION_EFW		0x010000
+
+#define SND_EFW_DEV_ENTRY(vendor, model) \
+{ \
+	.match_flags	= IEEE1394_MATCH_VENDOR_ID | \
+			  IEEE1394_MATCH_MODEL_ID | \
+			  IEEE1394_MATCH_SPECIFIER_ID | \
+			  IEEE1394_MATCH_VERSION, \
+	.vendor_id	= vendor,\
+	.model_id	= model, \
+	.specifier_id	= SPECIFIER_1394TA, \
+	.version	= VERSION_EFW, \
 }
 
 static const struct ieee1394_device_id efw_id_table[] = {
diff --git a/sound/firewire/fireworks/fireworks.h b/sound/firewire/fireworks/fireworks.h
index 654e28a6669f..c8d5879efe28 100644
--- a/sound/firewire/fireworks/fireworks.h
+++ b/sound/firewire/fireworks/fireworks.h
@@ -65,9 +65,6 @@ struct snd_efw {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
 	/* for transaction */
 	u32 seqnum;
 	bool resp_addr_changable;
@@ -181,7 +178,7 @@ struct snd_efw_phys_meters {
 } __packed;
 enum snd_efw_clock_source {
 	SND_EFW_CLOCK_SOURCE_INTERNAL	= 0,
-	SND_EFW_CLOCK_SOURCE_SYTMATCH	= 1,
+	// Unused.
 	SND_EFW_CLOCK_SOURCE_WORDCLOCK	= 2,
 	SND_EFW_CLOCK_SOURCE_SPDIF	= 3,
 	SND_EFW_CLOCK_SOURCE_ADAT_1	= 4,
@@ -227,12 +224,4 @@ int snd_efw_get_multiplier_mode(unsigned int sampling_rate, unsigned int *mode);
 
 int snd_efw_create_hwdep_device(struct snd_efw *efw);
 
-#define SND_EFW_DEV_ENTRY(vendor, model) \
-{ \
-	.match_flags	= IEEE1394_MATCH_VENDOR_ID | \
-			  IEEE1394_MATCH_MODEL_ID, \
-	.vendor_id	= vendor,\
-	.model_id	= model \
-}
-
 #endif
diff --git a/sound/firewire/fireworks/fireworks_stream.c b/sound/firewire/fireworks/fireworks_stream.c
index 2206af0fef42..ac66f08acd6b 100644
--- a/sound/firewire/fireworks/fireworks_stream.c
+++ b/sound/firewire/fireworks/fireworks_stream.c
@@ -6,7 +6,7 @@
  */
 #include "./fireworks.h"
 
-#define CALLBACK_TIMEOUT	100
+#define READY_TIMEOUT_MS	1000
 
 static int init_stream(struct snd_efw *efw, struct amdtp_stream *stream)
 {
@@ -29,7 +29,7 @@ static int init_stream(struct snd_efw *efw, struct amdtp_stream *stream)
 	if (err < 0)
 		return err;
 
-	err = amdtp_am824_init(stream, efw->unit, s_dir, CIP_BLOCKING);
+	err = amdtp_am824_init(stream, efw->unit, s_dir, CIP_BLOCKING | CIP_UNAWARE_SYT);
 	if (err < 0) {
 		amdtp_stream_destroy(stream);
 		cmp_connection_destroy(conn);
@@ -264,6 +264,15 @@ int snd_efw_stream_start_duplex(struct snd_efw *efw)
 		return err;
 
 	if (!amdtp_stream_running(&efw->rx_stream)) {
+		unsigned int tx_init_skip_cycles;
+
+		// Audiofire 2/4 skip an isochronous cycle several thousands after starting
+		// packet transmission.
+		if (efw->is_fireworks3 && !efw->is_af9)
+			tx_init_skip_cycles = 6000;
+		else
+			tx_init_skip_cycles = 0;
+
 		err = start_stream(efw, &efw->rx_stream, rate);
 		if (err < 0)
 			goto error;
@@ -272,15 +281,14 @@ int snd_efw_stream_start_duplex(struct snd_efw *efw)
 		if (err < 0)
 			goto error;
 
-		err = amdtp_domain_start(&efw->domain, 0);
+		// NOTE: The device ignores presentation time expressed by the value of syt field
+		// of CIP header in received packets. The sequence of the number of data blocks per
+		// packet is important for media clock recovery.
+		err = amdtp_domain_start(&efw->domain, tx_init_skip_cycles, true, false);
 		if (err < 0)
 			goto error;
 
-		// Wait first callback.
-		if (!amdtp_stream_wait_callback(&efw->rx_stream,
-						CALLBACK_TIMEOUT) ||
-		    !amdtp_stream_wait_callback(&efw->tx_stream,
-						CALLBACK_TIMEOUT)) {
+		if (!amdtp_domain_wait_ready(&efw->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
diff --git a/sound/firewire/lib.c b/sound/firewire/lib.c
index 85c4f4477c7f..e0a2337e8f27 100644
--- a/sound/firewire/lib.c
+++ b/sound/firewire/lib.c
@@ -67,38 +67,6 @@ int snd_fw_transaction(struct fw_unit *unit, int tcode,
 }
 EXPORT_SYMBOL(snd_fw_transaction);
 
-#define PROBE_DELAY_MS		(2 * MSEC_PER_SEC)
-
-/**
- * snd_fw_schedule_registration - schedule work for sound card registration
- * @unit: an instance for unit on IEEE 1394 bus
- * @dwork: delayed work with callback function
- *
- * This function is not designed for general purposes. When new unit is
- * connected to IEEE 1394 bus, the bus is under bus-reset state because of
- * topological change. In this state, units tend to fail both of asynchronous
- * and isochronous communication. To avoid this problem, this function is used
- * to postpone sound card registration after the state. The callers must
- * set up instance of delayed work in advance.
- */
-void snd_fw_schedule_registration(struct fw_unit *unit,
-				  struct delayed_work *dwork)
-{
-	u64 now, delay;
-
-	now = get_jiffies_64();
-	delay = fw_parent_device(unit)->card->reset_jiffies
-					+ msecs_to_jiffies(PROBE_DELAY_MS);
-
-	if (time_after64(delay, now))
-		delay -= now;
-	else
-		delay = 0;
-
-	mod_delayed_work(system_wq, dwork, delay);
-}
-EXPORT_SYMBOL(snd_fw_schedule_registration);
-
 MODULE_DESCRIPTION("FireWire audio helper functions");
 MODULE_AUTHOR("Clemens Ladisch <clemens@ladisch.de>");
 MODULE_LICENSE("GPL v2");
diff --git a/sound/firewire/lib.h b/sound/firewire/lib.h
index dc815dc3933e..664dfdb9e58d 100644
--- a/sound/firewire/lib.h
+++ b/sound/firewire/lib.h
@@ -23,7 +23,4 @@ static inline bool rcode_is_permanent_error(int rcode)
 	return rcode == RCODE_TYPE_ERROR || rcode == RCODE_ADDRESS_ERROR;
 }
 
-void snd_fw_schedule_registration(struct fw_unit *unit,
-				  struct delayed_work *dwork);
-
 #endif
diff --git a/sound/firewire/motu/Makefile b/sound/firewire/motu/Makefile
index 7c502d35103c..acdf66564fb0 100644
--- a/sound/firewire/motu/Makefile
+++ b/sound/firewire/motu/Makefile
@@ -3,5 +3,6 @@ CFLAGS_amdtp-motu.o	:= -I$(src)
 
 snd-firewire-motu-objs := motu.o amdtp-motu.o motu-transaction.o motu-stream.o \
 			  motu-proc.o motu-pcm.o motu-midi.o motu-hwdep.o \
-			  motu-protocol-v2.o motu-protocol-v3.o
+			  motu-protocol-v2.o motu-protocol-v3.o \
+			  motu-protocol-v1.o
 obj-$(CONFIG_SND_FIREWIRE_MOTU) += snd-firewire-motu.o
diff --git a/sound/firewire/motu/amdtp-motu.c b/sound/firewire/motu/amdtp-motu.c
index edb31ac26868..5388b85fb60e 100644
--- a/sound/firewire/motu/amdtp-motu.c
+++ b/sound/firewire/motu/amdtp-motu.c
@@ -16,6 +16,14 @@
 #define CIP_FMT_MOTU_TX_V3	0x22
 #define MOTU_FDF_AM824		0x22
 
+#define TICKS_PER_CYCLE		3072
+#define CYCLES_PER_SECOND	8000
+#define TICKS_PER_SECOND	(TICKS_PER_CYCLE * CYCLES_PER_SECOND)
+
+#define CIP_SPH_CYCLE_SHIFT	12
+#define CIP_SPH_CYCLE_MASK	0x01fff000
+#define CIP_SPH_OFFSET_MASK	0x00000fff
+
 /*
  * Nominally 3125 bytes/second, but the MIDI port's clock might be
  * 1% too slow, and the bus clock 100 ppm too fast.
@@ -23,14 +31,6 @@
 #define MIDI_BYTES_PER_SECOND	3093
 
 struct amdtp_motu {
-	/* For timestamp processing.  */
-	unsigned int quotient_ticks_per_event;
-	unsigned int remainder_ticks_per_event;
-	unsigned int next_ticks;
-	unsigned int next_accumulated;
-	unsigned int next_cycles;
-	unsigned int next_seconds;
-
 	unsigned int pcm_chunks;
 	unsigned int pcm_byte_offset;
 
@@ -41,26 +41,16 @@ struct amdtp_motu {
 
 	int midi_db_count;
 	unsigned int midi_db_interval;
+
+	struct amdtp_motu_cache *cache;
 };
 
 int amdtp_motu_set_parameters(struct amdtp_stream *s, unsigned int rate,
 			      unsigned int midi_ports,
 			      struct snd_motu_packet_format *formats)
 {
-	static const struct {
-		unsigned int quotient_ticks_per_event;
-		unsigned int remainder_ticks_per_event;
-	} params[] = {
-		[CIP_SFC_44100]  = { 557, 123 },
-		[CIP_SFC_48000]  = { 512,   0 },
-		[CIP_SFC_88200]  = { 278, 282 },
-		[CIP_SFC_96000]  = { 256,   0 },
-		[CIP_SFC_176400] = { 139, 141 },
-		[CIP_SFC_192000] = { 128,   0 },
-	};
 	struct amdtp_motu *p = s->protocol;
 	unsigned int pcm_chunks, data_chunks, data_block_quadlets;
-	unsigned int delay;
 	unsigned int mode;
 	int i, err;
 
@@ -97,19 +87,6 @@ int amdtp_motu_set_parameters(struct amdtp_stream *s, unsigned int rate,
 	p->midi_db_count = 0;
 	p->midi_db_interval = rate / MIDI_BYTES_PER_SECOND;
 
-	/* IEEE 1394 bus requires. */
-	delay = 0x2e00;
-
-	/* For no-data or empty packets to adjust PCM sampling frequency. */
-	delay += 8000 * 3072 * s->syt_interval / rate;
-
-	p->next_seconds = 0;
-	p->next_cycles = delay / 3072;
-	p->quotient_ticks_per_event = params[s->sfc].quotient_ticks_per_event;
-	p->remainder_ticks_per_event = params[s->sfc].remainder_ticks_per_event;
-	p->next_ticks = delay % 3072;
-	p->next_accumulated = 0;
-
 	return 0;
 }
 
@@ -322,6 +299,34 @@ static void probe_tracepoints_events(struct amdtp_stream *s,
 	}
 }
 
+static void cache_event_offsets(struct amdtp_motu_cache *cache, const __be32 *buf,
+				unsigned int data_blocks, unsigned int data_block_quadlets)
+{
+	unsigned int *event_offsets = cache->event_offsets;
+	const unsigned int cache_size = cache->size;
+	unsigned int cache_tail = cache->tail;
+	unsigned int base_tick = cache->tx_cycle_count * TICKS_PER_CYCLE;
+	int i;
+
+	for (i = 0; i < data_blocks; ++i) {
+		u32 sph = be32_to_cpu(*buf);
+		unsigned int tick;
+
+		tick = ((sph & CIP_SPH_CYCLE_MASK) >> CIP_SPH_CYCLE_SHIFT) * TICKS_PER_CYCLE +
+		       (sph & CIP_SPH_OFFSET_MASK);
+
+		if (tick < base_tick)
+			tick += TICKS_PER_SECOND;
+		event_offsets[cache_tail] = tick - base_tick;
+
+		cache_tail = (cache_tail + 1) % cache_size;
+		buf += data_block_quadlets;
+	}
+
+	cache->tail = cache_tail;
+	cache->tx_cycle_count = (cache->tx_cycle_count + 1) % CYCLES_PER_SECOND;
+}
+
 static unsigned int process_ir_ctx_payloads(struct amdtp_stream *s,
 					    const struct pkt_desc *descs,
 					    unsigned int packets,
@@ -331,12 +336,17 @@ static unsigned int process_ir_ctx_payloads(struct amdtp_stream *s,
 	unsigned int pcm_frames = 0;
 	int i;
 
+	if (p->cache->tx_cycle_count == UINT_MAX)
+		p->cache->tx_cycle_count = (s->domain->processing_cycle.tx_start % CYCLES_PER_SECOND);
+
 	// For data block processing.
 	for (i = 0; i < packets; ++i) {
 		const struct pkt_desc *desc = descs + i;
 		__be32 *buf = desc->ctx_payload;
 		unsigned int data_blocks = desc->data_blocks;
 
+		cache_event_offsets(p->cache, buf, data_blocks, s->data_block_quadlets);
+
 		if (pcm) {
 			read_pcm_s32(s, pcm, buf, data_blocks, pcm_frames);
 			pcm_frames += data_blocks;
@@ -354,46 +364,26 @@ static unsigned int process_ir_ctx_payloads(struct amdtp_stream *s,
 	return pcm_frames;
 }
 
-static inline void compute_next_elapse_from_start(struct amdtp_motu *p)
-{
-	p->next_accumulated += p->remainder_ticks_per_event;
-	if (p->next_accumulated >= 441) {
-		p->next_accumulated -= 441;
-		p->next_ticks++;
-	}
-
-	p->next_ticks += p->quotient_ticks_per_event;
-	if (p->next_ticks >= 3072) {
-		p->next_ticks -= 3072;
-		p->next_cycles++;
-	}
-
-	if (p->next_cycles >= 8000) {
-		p->next_cycles -= 8000;
-		p->next_seconds++;
-	}
-
-	if (p->next_seconds >= 128)
-		p->next_seconds -= 128;
-}
-
-static void write_sph(struct amdtp_stream *s, __be32 *buffer,
-		      unsigned int data_blocks)
+static void write_sph(struct amdtp_motu_cache *cache, __be32 *buffer, unsigned int data_blocks,
+		      unsigned int data_block_quadlets)
 {
-	struct amdtp_motu *p = s->protocol;
-	unsigned int next_cycles;
-	unsigned int i;
-	u32 sph;
+	unsigned int *event_offsets = cache->event_offsets;
+	const unsigned int cache_size = cache->size;
+	unsigned int cache_head = cache->head;
+	unsigned int base_tick = cache->rx_cycle_count * TICKS_PER_CYCLE;
+	int i;
 
 	for (i = 0; i < data_blocks; i++) {
-		next_cycles = (s->start_cycle + p->next_cycles) % 8000;
-		sph = ((next_cycles << 12) | p->next_ticks) & 0x01ffffff;
+		unsigned int tick = (base_tick + event_offsets[cache_head]) % TICKS_PER_SECOND;
+		u32 sph = ((tick / TICKS_PER_CYCLE) << CIP_SPH_CYCLE_SHIFT) | (tick % TICKS_PER_CYCLE);
 		*buffer = cpu_to_be32(sph);
 
-		compute_next_elapse_from_start(p);
-
-		buffer += s->data_block_quadlets;
+		cache_head = (cache_head + 1) % cache_size;
+		buffer += data_block_quadlets;
 	}
+
+	cache->head = cache_head;
+	cache->rx_cycle_count = (cache->rx_cycle_count + 1) % CYCLES_PER_SECOND;
 }
 
 static unsigned int process_it_ctx_payloads(struct amdtp_stream *s,
@@ -405,6 +395,9 @@ static unsigned int process_it_ctx_payloads(struct amdtp_stream *s,
 	unsigned int pcm_frames = 0;
 	int i;
 
+	if (p->cache->rx_cycle_count == UINT_MAX)
+		p->cache->rx_cycle_count = (s->domain->processing_cycle.rx_start % CYCLES_PER_SECOND);
+
 	// For data block processing.
 	for (i = 0; i < packets; ++i) {
 		const struct pkt_desc *desc = descs + i;
@@ -423,7 +416,7 @@ static unsigned int process_it_ctx_payloads(struct amdtp_stream *s,
 
 		// TODO: how to interact control messages between userspace?
 
-		write_sph(s, buf, data_blocks);
+		write_sph(p->cache, buf, data_blocks, s->data_block_quadlets);
 	}
 
 	// For tracepoints.
@@ -436,11 +429,12 @@ static unsigned int process_it_ctx_payloads(struct amdtp_stream *s,
 
 int amdtp_motu_init(struct amdtp_stream *s, struct fw_unit *unit,
 		    enum amdtp_stream_direction dir,
-		    const struct snd_motu_spec *spec)
+		    const struct snd_motu_spec *spec, struct amdtp_motu_cache *cache)
 {
 	amdtp_stream_process_ctx_payloads_t process_ctx_payloads;
 	int fmt = CIP_FMT_MOTU;
-	int flags = CIP_BLOCKING;
+	unsigned int flags = CIP_BLOCKING | CIP_UNAWARE_SYT;
+	struct amdtp_motu *p;
 	int err;
 
 	if (dir == AMDTP_IN_STREAM) {
@@ -478,9 +472,10 @@ int amdtp_motu_init(struct amdtp_stream *s, struct fw_unit *unit,
 	if (dir == AMDTP_OUT_STREAM) {
 		// Use fixed value for FDF field.
 		s->ctx_data.rx.fdf = MOTU_FDF_AM824;
-		// Not used.
-		s->ctx_data.rx.syt_override = 0xffff;
 	}
 
+	p = s->protocol;
+	p->cache = cache;
+
 	return 0;
 }
diff --git a/sound/firewire/motu/motu-protocol-v1.c b/sound/firewire/motu/motu-protocol-v1.c
new file mode 100644
index 000000000000..f1d6a326dc07
--- /dev/null
+++ b/sound/firewire/motu/motu-protocol-v1.c
@@ -0,0 +1,470 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+// motu-protocol-v1.c - a part of driver for MOTU FireWire series
+//
+// Copyright (c) 2021 Takashi Sakamoto <o-takashi@sakamocchi.jp>
+//
+// Licensed under the terms of the GNU General Public License, version 2.
+
+#include "motu.h"
+
+#include <linux/delay.h>
+
+// Status register for MOTU 828 (0x'ffff'f000'0b00).
+//
+// 0xffff0000: ISOC_COMM_CONTROL_MASK in motu-stream.c.
+// 0x00008000: mode of optical input interface.
+//   0x00008000: for S/PDIF signal.
+//   0x00000000: disabled or for ADAT signal.
+// 0x00004000: mode of optical output interface.
+//   0x00004000: for S/PDIF signal.
+//   0x00000000: disabled or for ADAT signal.
+// 0x00003f00: monitor input mode.
+//   0x00000800: analog-1/2
+//   0x00001a00: analog-3/4
+//   0x00002c00: analog-5/6
+//   0x00003e00: analog-7/8
+//   0x00000000: analog-1
+//   0x00000900: analog-2
+//   0x00001200: analog-3
+//   0x00001b00: analog-4
+//   0x00002400: analog-5
+//   0x00002d00: analog-6
+//   0x00003600: analog-7
+//   0x00003f00: analog-8
+// 0x00000080: enable stream input.
+// 0x00000040: disable monitor input.
+// 0x00000008: enable main out.
+// 0x00000004: rate of sampling clock.
+//   0x00000004: 48.0 kHz
+//   0x00000000: 44.1 kHz
+// 0x00000023: source of sampling clock.
+//   0x00000003: source packet header (SPH)
+//   0x00000002: S/PDIF on optical/coaxial interface.
+//   0x00000021: ADAT on optical interface
+//   0x00000001: ADAT on Dsub 9pin
+//   0x00000000: internal
+
+#define CLK_828_STATUS_OFFSET				0x0b00
+#define  CLK_828_STATUS_MASK				0x0000ffff
+#define  CLK_828_STATUS_FLAG_OPT_IN_IFACE_IS_SPDIF	0x00008000
+#define  CLK_828_STATUS_FLAG_OPT_OUT_IFACE_IS_SPDIF	0x00004000
+#define  CLK_828_STATUS_FLAG_FETCH_PCM_FRAMES		0x00000080
+#define  CLK_828_STATUS_FLAG_ENABLE_OUTPUT		0x00000008
+#define  CLK_828_STATUS_FLAG_RATE_48000			0x00000004
+#define  CLK_828_STATUS_MASK_SRC			0x00000023
+#define   CLK_828_STATUS_FLAG_SRC_ADAT_ON_OPT		0x00000021
+#define   CLK_828_STATUS_FLAG_SRC_SPH			0x00000003
+#define   CLK_828_STATUS_FLAG_SRC_SPDIF			0x00000002
+#define   CLK_828_STATUS_FLAG_SRC_ADAT_ON_DSUB		0x00000001
+#define   CLK_828_STATUS_FLAG_SRC_INTERNAL		0x00000000
+
+// Status register for MOTU 896 (0x'ffff'f000'0b14).
+//
+// 0xf0000000: enable physical and stream input to DAC.
+//   0x80000000: disable
+//   0x40000000: disable
+//   0x20000000: enable (prior to the other bits)
+//   0x10000000: disable
+//   0x00000000: disable
+// 0x08000000: speed of word clock signal output on BNC interface.
+//   0x00000000: force to low rate (44.1/48.0 kHz).
+//   0x08000000: follow to system clock.
+// 0x04000000: something relevant to clock.
+// 0x03000000: enable output.
+//  0x02000000: enabled irreversibly once standing unless the device voluntarily disables it.
+//  0x01000000: enabled irreversibly once standing unless the device voluntarily disables it.
+// 0x00ffff00: monitor input mode.
+//   0x00000000: disabled
+//   0x00004800: analog-1/2
+//   0x00005a00: analog-3/4
+//   0x00006c00: analog-5/6
+//   0x00007e00: analog-7/8
+//   0x00104800: AES/EBU-1/2
+//   0x00004000: analog-1
+//   0x00004900: analog-2
+//   0x00005200: analog-3
+//   0x00005b00: analog-4
+//   0x00006400: analog-5
+//   0x00006d00: analog-6
+//   0x00007600: analog-7
+//   0x00007f00: analog-8
+//   0x00104000: AES/EBU-1
+//   0x00104900: AES/EBU-2
+// 0x00000060: sample rate conversion for AES/EBU input/output.
+//   0x00000000: None
+//   0x00000020: input signal is converted to system rate
+//   0x00000040: output is slave to input, ignoring system rate
+//   0x00000060: output is double rate than system rate
+// 0x00000018: nominal rate of sampling clock.
+//   0x00000000: 44.1 kHz
+//   0x00000008: 48.0 kHz
+//   0x00000010: 88.2 kHz
+//   0x00000018: 96.0 kHz
+// 0x00000007: source of sampling clock.
+//   0x00000000: internal
+//   0x00000001: ADAT on optical interface
+//   0x00000002: AES/EBU on XLR
+//   0x00000003: source packet header (SPH)
+//   0x00000004: word clock on BNC
+//   0x00000005: ADAT on Dsub 9pin
+
+#define CLK_896_STATUS_OFFSET			0x0b14
+#define  CLK_896_STATUS_FLAG_FETCH_ENABLE	0x20000000
+#define  CLK_896_STATUS_FLAG_OUTPUT_ON		0x03000000
+#define  CLK_896_STATUS_MASK_SRC		0x00000007
+#define   CLK_896_STATUS_FLAG_SRC_INTERNAL	0x00000000
+#define   CLK_896_STATUS_FLAG_SRC_ADAT_ON_OPT	0x00000001
+#define   CLK_896_STATUS_FLAG_SRC_AESEBU	0x00000002
+#define   CLK_896_STATUS_FLAG_SRC_SPH		0x00000003
+#define   CLK_896_STATUS_FLAG_SRC_WORD		0x00000004
+#define   CLK_896_STATUS_FLAG_SRC_ADAT_ON_DSUB	0x00000005
+#define  CLK_896_STATUS_MASK_RATE		0x00000018
+#define   CLK_896_STATUS_FLAG_RATE_44100	0x00000000
+#define   CLK_896_STATUS_FLAG_RATE_48000	0x00000008
+#define   CLK_896_STATUS_FLAG_RATE_88200	0x00000010
+#define   CLK_896_STATUS_FLAG_RATE_96000	0x00000018
+
+static void parse_clock_rate_828(u32 data, unsigned int *rate)
+{
+	if (data & CLK_828_STATUS_FLAG_RATE_48000)
+		*rate = 48000;
+	else
+		*rate = 44100;
+}
+
+static int get_clock_rate_828(struct snd_motu *motu, unsigned int *rate)
+{
+	__be32 reg;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	parse_clock_rate_828(be32_to_cpu(reg), rate);
+
+	return 0;
+}
+
+static int parse_clock_rate_896(u32 data, unsigned int *rate)
+{
+	switch (data & CLK_896_STATUS_MASK_RATE) {
+	case CLK_896_STATUS_FLAG_RATE_44100:
+		*rate = 44100;
+		break;
+	case CLK_896_STATUS_FLAG_RATE_48000:
+		*rate = 48000;
+		break;
+	case CLK_896_STATUS_FLAG_RATE_88200:
+		*rate = 88200;
+		break;
+	case CLK_896_STATUS_FLAG_RATE_96000:
+		*rate = 96000;
+		break;
+	default:
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static int get_clock_rate_896(struct snd_motu *motu, unsigned int *rate)
+{
+	__be32 reg;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	return parse_clock_rate_896(be32_to_cpu(reg), rate);
+}
+
+int snd_motu_protocol_v1_get_clock_rate(struct snd_motu *motu, unsigned int *rate)
+{
+	if (motu->spec == &snd_motu_spec_828)
+		return get_clock_rate_828(motu, rate);
+	else if (motu->spec == &snd_motu_spec_896)
+		return get_clock_rate_896(motu, rate);
+	else
+		return -ENXIO;
+}
+
+static int set_clock_rate_828(struct snd_motu *motu, unsigned int rate)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg) & CLK_828_STATUS_MASK;
+
+	data &= ~CLK_828_STATUS_FLAG_RATE_48000;
+	if (rate == 48000)
+		data |= CLK_828_STATUS_FLAG_RATE_48000;
+
+	reg = cpu_to_be32(data);
+	return snd_motu_transaction_write(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+}
+
+static int set_clock_rate_896(struct snd_motu *motu, unsigned int rate)
+{
+	unsigned int flag;
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg);
+
+	switch (rate) {
+	case 44100:
+		flag = CLK_896_STATUS_FLAG_RATE_44100;
+		break;
+	case 48000:
+		flag = CLK_896_STATUS_FLAG_RATE_48000;
+		break;
+	case 88200:
+		flag = CLK_896_STATUS_FLAG_RATE_88200;
+		break;
+	case 96000:
+		flag = CLK_896_STATUS_FLAG_RATE_96000;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	data &= ~CLK_896_STATUS_MASK_RATE;
+	data |= flag;
+
+	reg = cpu_to_be32(data);
+	return snd_motu_transaction_write(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+}
+
+int snd_motu_protocol_v1_set_clock_rate(struct snd_motu *motu, unsigned int rate)
+{
+	if (motu->spec == &snd_motu_spec_828)
+		return set_clock_rate_828(motu, rate);
+	else if (motu->spec == &snd_motu_spec_896)
+		return set_clock_rate_896(motu, rate);
+	else
+		return -ENXIO;
+}
+
+static int get_clock_source_828(struct snd_motu *motu, enum snd_motu_clock_source *src)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg) & CLK_828_STATUS_MASK;
+
+	switch (data & CLK_828_STATUS_MASK_SRC) {
+	case CLK_828_STATUS_FLAG_SRC_ADAT_ON_OPT:
+		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_OPT;
+		break;
+	case CLK_828_STATUS_FLAG_SRC_SPH:
+		*src = SND_MOTU_CLOCK_SOURCE_SPH;
+		break;
+	case CLK_828_STATUS_FLAG_SRC_SPDIF:
+	{
+		if (data & CLK_828_STATUS_FLAG_OPT_IN_IFACE_IS_SPDIF)
+			*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
+		else
+			*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_OPT;
+		break;
+	}
+	case CLK_828_STATUS_FLAG_SRC_ADAT_ON_DSUB:
+		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_DSUB;
+		break;
+	case CLK_828_STATUS_FLAG_SRC_INTERNAL:
+		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
+		break;
+	default:
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static int get_clock_source_896(struct snd_motu *motu, enum snd_motu_clock_source *src)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg);
+
+	switch (data & CLK_896_STATUS_MASK_SRC) {
+	case CLK_896_STATUS_FLAG_SRC_INTERNAL:
+		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
+		break;
+	case CLK_896_STATUS_FLAG_SRC_ADAT_ON_OPT:
+		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_OPT;
+		break;
+	case CLK_896_STATUS_FLAG_SRC_AESEBU:
+		*src = SND_MOTU_CLOCK_SOURCE_AESEBU_ON_XLR;
+		break;
+	case CLK_896_STATUS_FLAG_SRC_SPH:
+		*src = SND_MOTU_CLOCK_SOURCE_SPH;
+		break;
+	case CLK_896_STATUS_FLAG_SRC_WORD:
+		*src = SND_MOTU_CLOCK_SOURCE_WORD_ON_BNC;
+		break;
+	case CLK_896_STATUS_FLAG_SRC_ADAT_ON_DSUB:
+		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_DSUB;
+		break;
+	default:
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+int snd_motu_protocol_v1_get_clock_source(struct snd_motu *motu, enum snd_motu_clock_source *src)
+{
+	if (motu->spec == &snd_motu_spec_828)
+		return get_clock_source_828(motu, src);
+	else if (motu->spec == &snd_motu_spec_896)
+		return get_clock_source_896(motu, src);
+	else
+		return -ENXIO;
+}
+
+static int switch_fetching_mode_828(struct snd_motu *motu, bool enable)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg) & CLK_828_STATUS_MASK;
+
+	data &= ~(CLK_828_STATUS_FLAG_FETCH_PCM_FRAMES | CLK_828_STATUS_FLAG_ENABLE_OUTPUT);
+	if (enable) {
+		// This transaction should be initiated after the device receives batch of packets
+		// since the device voluntarily mutes outputs. As a workaround, yield processor over
+		// 100 msec.
+		msleep(100);
+		data |= CLK_828_STATUS_FLAG_FETCH_PCM_FRAMES | CLK_828_STATUS_FLAG_ENABLE_OUTPUT;
+	}
+
+	reg = cpu_to_be32(data);
+	return snd_motu_transaction_write(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+}
+
+static int switch_fetching_mode_896(struct snd_motu *motu, bool enable)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg);
+
+	data &= ~CLK_896_STATUS_FLAG_FETCH_ENABLE;
+	if (enable)
+		data |= CLK_896_STATUS_FLAG_FETCH_ENABLE | CLK_896_STATUS_FLAG_OUTPUT_ON;
+
+	reg = cpu_to_be32(data);
+	return snd_motu_transaction_write(motu, CLK_896_STATUS_OFFSET, &reg, sizeof(reg));
+}
+
+int snd_motu_protocol_v1_switch_fetching_mode(struct snd_motu *motu, bool enable)
+{
+	if (motu->spec == &snd_motu_spec_828)
+		return switch_fetching_mode_828(motu, enable);
+	else if (motu->spec == &snd_motu_spec_896)
+		return switch_fetching_mode_896(motu, enable);
+	else
+		return -ENXIO;
+}
+
+static int detect_packet_formats_828(struct snd_motu *motu)
+{
+	__be32 reg;
+	u32 data;
+	int err;
+
+	motu->tx_packet_formats.pcm_byte_offset = 4;
+	motu->tx_packet_formats.msg_chunks = 2;
+
+	motu->rx_packet_formats.pcm_byte_offset = 4;
+	motu->rx_packet_formats.msg_chunks = 0;
+
+	err = snd_motu_transaction_read(motu, CLK_828_STATUS_OFFSET, &reg, sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg) & CLK_828_STATUS_MASK;
+
+	// The number of chunks is just reduced when SPDIF is activated.
+	if (!(data & CLK_828_STATUS_FLAG_OPT_IN_IFACE_IS_SPDIF))
+		motu->tx_packet_formats.pcm_chunks[0] += 8;
+
+	if (!(data & CLK_828_STATUS_FLAG_OPT_OUT_IFACE_IS_SPDIF))
+		motu->rx_packet_formats.pcm_chunks[0] += 8;
+
+	return 0;
+}
+
+static int detect_packet_formats_896(struct snd_motu *motu)
+{
+	// 24bit PCM frames follow to source packet header without message chunk.
+	motu->tx_packet_formats.pcm_byte_offset = 4;
+	motu->rx_packet_formats.pcm_byte_offset = 4;
+
+	// No message chunk in data block.
+	motu->tx_packet_formats.msg_chunks = 0;
+	motu->rx_packet_formats.msg_chunks = 0;
+
+	// Always enable optical interface for ADAT signal since the device have no registers
+	// to refer to current configuration.
+	motu->tx_packet_formats.pcm_chunks[0] += 8;
+	motu->tx_packet_formats.pcm_chunks[1] += 8;
+
+	motu->rx_packet_formats.pcm_chunks[0] += 8;
+	motu->rx_packet_formats.pcm_chunks[1] += 8;
+
+	return 0;
+}
+
+int snd_motu_protocol_v1_cache_packet_formats(struct snd_motu *motu)
+{
+	memcpy(motu->tx_packet_formats.pcm_chunks, motu->spec->tx_fixed_pcm_chunks,
+	       sizeof(motu->tx_packet_formats.pcm_chunks));
+	memcpy(motu->rx_packet_formats.pcm_chunks, motu->spec->rx_fixed_pcm_chunks,
+	       sizeof(motu->rx_packet_formats.pcm_chunks));
+
+	if (motu->spec == &snd_motu_spec_828)
+		return detect_packet_formats_828(motu);
+	else if (motu->spec == &snd_motu_spec_896)
+		return detect_packet_formats_896(motu);
+	else
+		return 0;
+}
+
+const struct snd_motu_spec snd_motu_spec_828 = {
+	.name = "828",
+	.protocol_version = SND_MOTU_PROTOCOL_V1,
+	.tx_fixed_pcm_chunks = {10, 0, 0},
+	.rx_fixed_pcm_chunks = {10, 0, 0},
+};
+
+const struct snd_motu_spec snd_motu_spec_896 = {
+	.name = "896",
+	.tx_fixed_pcm_chunks = {10, 10, 0},
+	.rx_fixed_pcm_chunks = {10, 10, 0},
+};
diff --git a/sound/firewire/motu/motu-protocol-v2.c b/sound/firewire/motu/motu-protocol-v2.c
index e59e69ab1538..93d5df1ae550 100644
--- a/sound/firewire/motu/motu-protocol-v2.c
+++ b/sound/firewire/motu/motu-protocol-v2.c
@@ -12,6 +12,13 @@
 #define  V2_CLOCK_RATE_SHIFT			3
 #define  V2_CLOCK_SRC_MASK			0x00000007
 #define  V2_CLOCK_SRC_SHIFT			0
+#define   V2_CLOCK_SRC_AESEBU_ON_XLR		0x07
+#define   V2_CLOCK_SRC_ADAT_ON_DSUB		0x05
+#define   V2_CLOCK_SRC_WORD_ON_BNC		0x04
+#define   V2_CLOCK_SRC_SPH			0x03
+#define   V2_CLOCK_SRC_SPDIF			0x02	// on either coaxial or optical
+#define   V2_CLOCK_SRC_ADAT_ON_OPT		0x01
+#define   V2_CLOCK_SRC_INTERNAL			0x00
 #define  V2_CLOCK_FETCH_ENABLE			0x02000000
 #define  V2_CLOCK_MODEL_SPECIFIC		0x04000000
 
@@ -78,64 +85,51 @@ int snd_motu_protocol_v2_set_clock_rate(struct snd_motu *motu,
 					  sizeof(reg));
 }
 
-static int detect_clock_source_optical_model(struct snd_motu *motu, u32 data,
-					     enum snd_motu_clock_source *src)
+static int get_clock_source(struct snd_motu *motu, u32 data,
+			    enum snd_motu_clock_source *src)
 {
-	switch (data) {
-	case 0:
+	switch (data & V2_CLOCK_SRC_MASK) {
+	case V2_CLOCK_SRC_INTERNAL:
 		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
 		break;
-	case 1:
+	case V2_CLOCK_SRC_ADAT_ON_OPT:
+		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_OPT;
+		break;
+	case V2_CLOCK_SRC_SPDIF:
 	{
-		__be32 reg;
-
-		// To check the configuration of optical interface.
-		int err = snd_motu_transaction_read(motu, V2_IN_OUT_CONF_OFFSET,
-						    &reg, sizeof(reg));
-		if (err < 0)
-			return err;
-
-		if (be32_to_cpu(reg) & 0x00000200)
-			*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_OPT;
-		else
-			*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_OPT;
+		bool support_iec60958_on_opt = (motu->spec == &snd_motu_spec_828mk2 ||
+						motu->spec == &snd_motu_spec_traveler);
+
+		if (!support_iec60958_on_opt) {
+			*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
+		} else {
+			__be32 reg;
+
+			// To check the configuration of optical interface.
+			int err = snd_motu_transaction_read(motu, V2_IN_OUT_CONF_OFFSET, &reg,
+							    sizeof(reg));
+			if (err < 0)
+				return err;
+
+			if (((data & V2_OPT_IN_IFACE_MASK) >> V2_OPT_IN_IFACE_SHIFT) ==
+			    V2_OPT_IFACE_MODE_SPDIF)
+				*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_OPT;
+			else
+				*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
+		}
 		break;
 	}
-	case 2:
-		*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
-		break;
-	case 3:
+	case V2_CLOCK_SRC_SPH:
 		*src = SND_MOTU_CLOCK_SOURCE_SPH;
 		break;
-	case 4:
+	case V2_CLOCK_SRC_WORD_ON_BNC:
 		*src = SND_MOTU_CLOCK_SOURCE_WORD_ON_BNC;
 		break;
-	case 5:
+	case V2_CLOCK_SRC_ADAT_ON_DSUB:
 		*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_DSUB;
 		break;
-	default:
-		*src = SND_MOTU_CLOCK_SOURCE_UNKNOWN;
-		break;
-	}
-
-	return 0;
-}
-
-static int v2_detect_clock_source(struct snd_motu *motu, u32 data,
-				  enum snd_motu_clock_source *src)
-{
-	switch (data) {
-	case 0:
-		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
-		break;
-	case 2:
-		*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
-		break;
-	case 3:
-		*src = SND_MOTU_CLOCK_SOURCE_SPH;
-		break;
-	case 4:
-		*src = SND_MOTU_CLOCK_SOURCE_WORD_ON_BNC;
+	case V2_CLOCK_SRC_AESEBU_ON_XLR:
+		*src = SND_MOTU_CLOCK_SOURCE_AESEBU_ON_XLR;
 		break;
 	default:
 		*src = SND_MOTU_CLOCK_SOURCE_UNKNOWN;
@@ -145,17 +139,6 @@ static int v2_detect_clock_source(struct snd_motu *motu, u32 data,
 	return 0;
 }
 
-static int get_clock_source(struct snd_motu *motu, u32 data,
-			    enum snd_motu_clock_source *src)
-{
-	data &= V2_CLOCK_SRC_MASK;
-	if (motu->spec == &snd_motu_spec_828mk2 ||
-	    motu->spec == &snd_motu_spec_traveler)
-		return detect_clock_source_optical_model(motu, data, src);
-	else
-		return v2_detect_clock_source(motu, data, src);
-}
-
 int snd_motu_protocol_v2_get_clock_source(struct snd_motu *motu,
 					  enum snd_motu_clock_source *src)
 {
@@ -235,59 +218,9 @@ int snd_motu_protocol_v2_switch_fetching_mode(struct snd_motu *motu,
 	}
 }
 
-static int detect_packet_formats_828mk2(struct snd_motu *motu, u32 data)
-{
-	if (((data & V2_OPT_IN_IFACE_MASK) >> V2_OPT_IN_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->tx_packet_formats.pcm_chunks[0] += 8;
-		motu->tx_packet_formats.pcm_chunks[1] += 4;
-	}
-
-	if (((data & V2_OPT_OUT_IFACE_MASK) >> V2_OPT_OUT_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->rx_packet_formats.pcm_chunks[0] += 8;
-		motu->rx_packet_formats.pcm_chunks[1] += 4;
-	}
-
-	return 0;
-}
-
-static int detect_packet_formats_traveler(struct snd_motu *motu, u32 data)
-{
-	if (((data & V2_OPT_IN_IFACE_MASK) >> V2_OPT_IN_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->tx_packet_formats.pcm_chunks[0] += 8;
-		motu->tx_packet_formats.pcm_chunks[1] += 4;
-	}
-
-	if (((data & V2_OPT_OUT_IFACE_MASK) >> V2_OPT_OUT_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->rx_packet_formats.pcm_chunks[0] += 8;
-		motu->rx_packet_formats.pcm_chunks[1] += 4;
-	}
-
-	return 0;
-}
-
-static int detect_packet_formats_8pre(struct snd_motu *motu, u32 data)
-{
-	if (((data & V2_OPT_IN_IFACE_MASK) >> V2_OPT_IN_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->tx_packet_formats.pcm_chunks[0] += 8;
-		motu->tx_packet_formats.pcm_chunks[1] += 8;
-	}
-
-	if (((data & V2_OPT_OUT_IFACE_MASK) >> V2_OPT_OUT_IFACE_SHIFT) ==
-	    V2_OPT_IFACE_MODE_ADAT) {
-		motu->rx_packet_formats.pcm_chunks[0] += 8;
-		motu->rx_packet_formats.pcm_chunks[1] += 8;
-	}
-
-	return 0;
-}
-
 int snd_motu_protocol_v2_cache_packet_formats(struct snd_motu *motu)
 {
+	bool has_two_opt_ifaces = (motu->spec == &snd_motu_spec_8pre);
 	__be32 reg;
 	u32 data;
 	int err;
@@ -311,14 +244,25 @@ int snd_motu_protocol_v2_cache_packet_formats(struct snd_motu *motu)
 	       motu->spec->rx_fixed_pcm_chunks,
 	       sizeof(motu->rx_packet_formats.pcm_chunks));
 
-	if (motu->spec == &snd_motu_spec_828mk2)
-		return detect_packet_formats_828mk2(motu, data);
-	else if (motu->spec == &snd_motu_spec_traveler)
-		return detect_packet_formats_traveler(motu, data);
-	else if (motu->spec == &snd_motu_spec_8pre)
-		return detect_packet_formats_8pre(motu, data);
-	else
-		return 0;
+	if (((data & V2_OPT_IN_IFACE_MASK) >> V2_OPT_IN_IFACE_SHIFT) == V2_OPT_IFACE_MODE_ADAT) {
+		motu->tx_packet_formats.pcm_chunks[0] += 8;
+
+		if (!has_two_opt_ifaces)
+			motu->tx_packet_formats.pcm_chunks[1] += 4;
+		else
+			motu->tx_packet_formats.pcm_chunks[1] += 8;
+	}
+
+	if (((data & V2_OPT_OUT_IFACE_MASK) >> V2_OPT_OUT_IFACE_SHIFT) == V2_OPT_IFACE_MODE_ADAT) {
+		motu->rx_packet_formats.pcm_chunks[0] += 8;
+
+		if (!has_two_opt_ifaces)
+			motu->rx_packet_formats.pcm_chunks[1] += 4;
+		else
+			motu->rx_packet_formats.pcm_chunks[1] += 8;
+	}
+
+	return 0;
 }
 
 const struct snd_motu_spec snd_motu_spec_828mk2 = {
@@ -353,6 +297,7 @@ const struct snd_motu_spec snd_motu_spec_8pre = {
 	.protocol_version = SND_MOTU_PROTOCOL_V2,
 	.flags = SND_MOTU_SPEC_RX_MIDI_2ND_Q |
 		 SND_MOTU_SPEC_TX_MIDI_2ND_Q,
-	.tx_fixed_pcm_chunks = {10, 6, 0},
-	.rx_fixed_pcm_chunks = {10, 6, 0},
+	// Two dummy chunks always in the end of data block.
+	.tx_fixed_pcm_chunks = {10, 10, 0},
+	.rx_fixed_pcm_chunks = {6, 6, 0},
 };
diff --git a/sound/firewire/motu/motu-protocol-v3.c b/sound/firewire/motu/motu-protocol-v3.c
index 4e6b0e449ee4..56e4504e7ec9 100644
--- a/sound/firewire/motu/motu-protocol-v3.c
+++ b/sound/firewire/motu/motu-protocol-v3.c
@@ -13,6 +13,12 @@
 #define  V3_CLOCK_RATE_MASK		0x0000ff00
 #define  V3_CLOCK_RATE_SHIFT		8
 #define  V3_CLOCK_SOURCE_MASK		0x000000ff
+#define   V3_CLOCK_SRC_INTERNAL		0x00
+#define   V3_CLOCK_SRC_WORD_ON_BNC	0x01
+#define   V3_CLOCK_SRC_SPH		0x02
+#define   V3_CLOCK_SRC_SPDIF_ON_COAX	0x10
+#define   V3_CLOCK_SRC_OPT_IFACE_A	0x18
+#define   V3_CLOCK_SRC_OPT_IFACE_B	0x19
 
 #define V3_OPT_IFACE_MODE_OFFSET	0x0c94
 #define  V3_ENABLE_OPT_IN_IFACE_A	0x00000001
@@ -97,28 +103,37 @@ int snd_motu_protocol_v3_set_clock_rate(struct snd_motu *motu,
 	return 0;
 }
 
-static int detect_clock_source_828mk3(struct snd_motu *motu, u32 data,
-				      enum snd_motu_clock_source *src)
+int snd_motu_protocol_v3_get_clock_source(struct snd_motu *motu,
+					  enum snd_motu_clock_source *src)
 {
+	__be32 reg;
+	u32 data;
+	int err;
+
+	err = snd_motu_transaction_read(motu, V3_CLOCK_STATUS_OFFSET, &reg,
+					sizeof(reg));
+	if (err < 0)
+		return err;
+	data = be32_to_cpu(reg) & V3_CLOCK_SOURCE_MASK;
+
 	switch (data) {
-	case 0x00:
+	case V3_CLOCK_SRC_INTERNAL:
 		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
 		break;
-	case 0x01:
+	case V3_CLOCK_SRC_WORD_ON_BNC:
 		*src = SND_MOTU_CLOCK_SOURCE_WORD_ON_BNC;
 		break;
-	case 0x02:
+	case V3_CLOCK_SRC_SPH:
 		*src = SND_MOTU_CLOCK_SOURCE_SPH;
 		break;
-	case 0x10:
+	case V3_CLOCK_SRC_SPDIF_ON_COAX:
 		*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
 		break;
-	case 0x18:
-	case 0x19:
+	case V3_CLOCK_SRC_OPT_IFACE_A:
+	case V3_CLOCK_SRC_OPT_IFACE_B:
 	{
 		__be32 reg;
 		u32 options;
-		int err;
 
 		err = snd_motu_transaction_read(motu,
 				V3_OPT_IFACE_MODE_OFFSET, &reg, sizeof(reg));
@@ -126,7 +141,7 @@ static int detect_clock_source_828mk3(struct snd_motu *motu, u32 data,
 			return err;
 		options = be32_to_cpu(reg);
 
-		if (data == 0x18) {
+		if (data == V3_CLOCK_SRC_OPT_IFACE_A) {
 			if (options & V3_NO_ADAT_OPT_IN_IFACE_A)
 				*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_OPT_A;
 			else
@@ -137,7 +152,6 @@ static int detect_clock_source_828mk3(struct snd_motu *motu, u32 data,
 			else
 				*src = SND_MOTU_CLOCK_SOURCE_ADAT_ON_OPT_B;
 		}
-
 		break;
 	}
 	default:
@@ -148,49 +162,6 @@ static int detect_clock_source_828mk3(struct snd_motu *motu, u32 data,
 	return 0;
 }
 
-static int v3_detect_clock_source(struct snd_motu *motu, u32 data,
-				  enum snd_motu_clock_source *src)
-{
-	switch (data) {
-	case 0x00:
-		*src = SND_MOTU_CLOCK_SOURCE_INTERNAL;
-		break;
-	case 0x01:
-		*src = SND_MOTU_CLOCK_SOURCE_WORD_ON_BNC;
-		break;
-	case 0x02:
-		*src = SND_MOTU_CLOCK_SOURCE_SPH;
-		break;
-	case 0x10:
-		*src = SND_MOTU_CLOCK_SOURCE_SPDIF_ON_COAX;
-		break;
-	default:
-		*src = SND_MOTU_CLOCK_SOURCE_UNKNOWN;
-		break;
-	}
-
-	return 0;
-}
-
-int snd_motu_protocol_v3_get_clock_source(struct snd_motu *motu,
-					  enum snd_motu_clock_source *src)
-{
-	__be32 reg;
-	u32 data;
-	int err;
-
-	err = snd_motu_transaction_read(motu, V3_CLOCK_STATUS_OFFSET, &reg,
-					sizeof(reg));
-	if (err < 0)
-		return err;
-	data = be32_to_cpu(reg) & V3_CLOCK_SOURCE_MASK;
-
-	if (motu->spec == &snd_motu_spec_828mk3)
-		return detect_clock_source_828mk3(motu, data, src);
-	else
-		return v3_detect_clock_source(motu, data, src);
-}
-
 int snd_motu_protocol_v3_switch_fetching_mode(struct snd_motu *motu,
 					      bool enable)
 {
@@ -284,14 +255,14 @@ int snd_motu_protocol_v3_cache_packet_formats(struct snd_motu *motu)
 	       motu->spec->rx_fixed_pcm_chunks,
 	       sizeof(motu->rx_packet_formats.pcm_chunks));
 
-	if (motu->spec == &snd_motu_spec_828mk3)
+	if (motu->spec == &snd_motu_spec_828mk3_fw || motu->spec == &snd_motu_spec_828mk3_hybrid)
 		return detect_packet_formats_828mk3(motu, data);
 	else
 		return 0;
 }
 
 
-const struct snd_motu_spec snd_motu_spec_828mk3 = {
+const struct snd_motu_spec snd_motu_spec_828mk3_fw = {
 	.name = "828mk3",
 	.protocol_version = SND_MOTU_PROTOCOL_V3,
 	.flags = SND_MOTU_SPEC_RX_MIDI_3RD_Q |
@@ -300,6 +271,15 @@ const struct snd_motu_spec snd_motu_spec_828mk3 = {
 	.rx_fixed_pcm_chunks = {14, 14, 10},
 };
 
+const struct snd_motu_spec snd_motu_spec_828mk3_hybrid = {
+	.name = "828mk3",
+	.protocol_version = SND_MOTU_PROTOCOL_V3,
+	.flags = SND_MOTU_SPEC_RX_MIDI_3RD_Q |
+		 SND_MOTU_SPEC_TX_MIDI_3RD_Q,
+	.tx_fixed_pcm_chunks = {18, 18, 14},
+	.rx_fixed_pcm_chunks = {14, 14, 14},	// Additional 4 dummy chunks at higher rate.
+};
+
 const struct snd_motu_spec snd_motu_spec_ultralite_mk3 = {
 	.name = "UltraLiteMk3",
 	.protocol_version = SND_MOTU_PROTOCOL_V3,
diff --git a/sound/firewire/motu/motu-stream.c b/sound/firewire/motu/motu-stream.c
index 2028c5419f6f..9e6ca39ebd7f 100644
--- a/sound/firewire/motu/motu-stream.c
+++ b/sound/firewire/motu/motu-stream.c
@@ -7,7 +7,7 @@
 
 #include "motu.h"
 
-#define	CALLBACK_TIMEOUT	200
+#define	READY_TIMEOUT_MS	200
 
 #define ISOC_COMM_CONTROL_OFFSET		0x0b00
 #define  ISOC_COMM_CONTROL_MASK			0xffff0000
@@ -153,6 +153,9 @@ int snd_motu_stream_reserve_duplex(struct snd_motu *motu, unsigned int rate,
 		fw_iso_resources_free(&motu->tx_resources);
 		fw_iso_resources_free(&motu->rx_resources);
 
+		kfree(motu->cache.event_offsets);
+		motu->cache.event_offsets = NULL;
+
 		err = snd_motu_protocol_set_clock_rate(motu, rate);
 		if (err < 0) {
 			dev_err(&motu->unit->device,
@@ -181,6 +184,15 @@ int snd_motu_stream_reserve_duplex(struct snd_motu *motu, unsigned int rate,
 			fw_iso_resources_free(&motu->rx_resources);
 			return err;
 		}
+
+		motu->cache.size = motu->tx_stream.syt_interval * frames_per_buffer;
+		motu->cache.event_offsets = kcalloc(motu->cache.size, sizeof(*motu->cache.event_offsets),
+						  GFP_KERNEL);
+		if (!motu->cache.event_offsets) {
+			fw_iso_resources_free(&motu->tx_resources);
+			fw_iso_resources_free(&motu->rx_resources);
+			return -ENOMEM;
+		}
 	}
 
 	return 0;
@@ -260,14 +272,19 @@ int snd_motu_stream_start_duplex(struct snd_motu *motu)
 		if (err < 0)
 			goto stop_streams;
 
-		err = amdtp_domain_start(&motu->domain, 0);
+		motu->cache.tail = 0;
+		motu->cache.tx_cycle_count = UINT_MAX;
+		motu->cache.head = 0;
+		motu->cache.rx_cycle_count = UINT_MAX;
+
+		// NOTE: The device requires both of replay; the sequence of the number of data
+		// blocks per packet, and the sequence of source packet header per data block as
+		// presentation time.
+		err = amdtp_domain_start(&motu->domain, 0, true, false);
 		if (err < 0)
 			goto stop_streams;
 
-		if (!amdtp_stream_wait_callback(&motu->tx_stream,
-						CALLBACK_TIMEOUT) ||
-		    !amdtp_stream_wait_callback(&motu->rx_stream,
-						CALLBACK_TIMEOUT)) {
+		if (!amdtp_domain_wait_ready(&motu->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto stop_streams;
 		}
@@ -296,6 +313,9 @@ void snd_motu_stream_stop_duplex(struct snd_motu *motu)
 
 		fw_iso_resources_free(&motu->tx_resources);
 		fw_iso_resources_free(&motu->rx_resources);
+
+		kfree(motu->cache.event_offsets);
+		motu->cache.event_offsets = NULL;
 	}
 }
 
@@ -317,7 +337,7 @@ static int init_stream(struct snd_motu *motu, struct amdtp_stream *s)
 	if (err < 0)
 		return err;
 
-	err = amdtp_motu_init(s, motu->unit, dir, motu->spec);
+	err = amdtp_motu_init(s, motu->unit, dir, motu->spec, &motu->cache);
 	if (err < 0)
 		fw_iso_resources_destroy(resources);
 
diff --git a/sound/firewire/motu/motu.c b/sound/firewire/motu/motu.c
index a4929c1302dc..543136578c70 100644
--- a/sound/firewire/motu/motu.c
+++ b/sound/firewire/motu/motu.c
@@ -57,22 +57,31 @@ static void motu_card_free(struct snd_card *card)
 
 	snd_motu_transaction_unregister(motu);
 	snd_motu_stream_destroy_duplex(motu);
+
+	mutex_destroy(&motu->mutex);
+	fw_unit_put(motu->unit);
 }
 
-static void do_registration(struct work_struct *work)
+static int motu_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_motu *motu = container_of(work, struct snd_motu, dwork.work);
+	struct snd_card *card;
+	struct snd_motu *motu;
 	int err;
 
-	if (motu->registered)
-		return;
-
-	err = snd_card_new(&motu->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &motu->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*motu), &card);
 	if (err < 0)
-		return;
-	motu->card->private_free = motu_card_free;
-	motu->card->private_data = motu;
+		return err;
+	card->private_free = motu_card_free;
+
+	motu = card->private_data;
+	motu->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, motu);
+	motu->card = card;
+
+	motu->spec = (const struct snd_motu_spec *)entry->driver_data;
+	mutex_init(&motu->mutex);
+	spin_lock_init(&motu->lock);
+	init_waitqueue_head(&motu->hwdep_wait);
 
 	name_card(motu);
 
@@ -103,71 +112,28 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(motu->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	motu->registered = true;
-
-	return;
-error:
-	snd_card_free(motu->card);
-	dev_info(&motu->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int motu_probe(struct fw_unit *unit,
-		      const struct ieee1394_device_id *entry)
-{
-	struct snd_motu *motu;
-
-	/* Allocate this independently of sound card instance. */
-	motu = devm_kzalloc(&unit->device, sizeof(struct snd_motu), GFP_KERNEL);
-	if (!motu)
-		return -ENOMEM;
-	motu->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, motu);
-
-	motu->spec = (const struct snd_motu_spec *)entry->driver_data;
-	mutex_init(&motu->mutex);
-	spin_lock_init(&motu->lock);
-	init_waitqueue_head(&motu->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&motu->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &motu->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void motu_remove(struct fw_unit *unit)
 {
 	struct snd_motu *motu = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&motu->dwork);
-
-	if (motu->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(motu->card);
-	}
-
-	mutex_destroy(&motu->mutex);
-	fw_unit_put(motu->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(motu->card);
 }
 
 static void motu_bus_update(struct fw_unit *unit)
 {
 	struct snd_motu *motu = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!motu->registered)
-		snd_fw_schedule_registration(unit, &motu->dwork);
-
 	/* The handler address register becomes initialized. */
 	snd_motu_transaction_reregister(motu);
 }
@@ -184,13 +150,16 @@ static void motu_bus_update(struct fw_unit *unit)
 }
 
 static const struct ieee1394_device_id motu_id_table[] = {
+	SND_MOTU_DEV_ENTRY(0x000001, &snd_motu_spec_828),
+	SND_MOTU_DEV_ENTRY(0x000002, &snd_motu_spec_896),
 	SND_MOTU_DEV_ENTRY(0x000003, &snd_motu_spec_828mk2),
 	SND_MOTU_DEV_ENTRY(0x000009, &snd_motu_spec_traveler),
 	SND_MOTU_DEV_ENTRY(0x00000d, &snd_motu_spec_ultralite),
 	SND_MOTU_DEV_ENTRY(0x00000f, &snd_motu_spec_8pre),
-	SND_MOTU_DEV_ENTRY(0x000015, &snd_motu_spec_828mk3), // FireWire only.
+	SND_MOTU_DEV_ENTRY(0x000015, &snd_motu_spec_828mk3_fw), // FireWire only.
 	SND_MOTU_DEV_ENTRY(0x000019, &snd_motu_spec_ultralite_mk3), // FireWire only.
-	SND_MOTU_DEV_ENTRY(0x000035, &snd_motu_spec_828mk3), // Hybrid.
+	SND_MOTU_DEV_ENTRY(0x000030, &snd_motu_spec_ultralite_mk3), // Hybrid.
+	SND_MOTU_DEV_ENTRY(0x000035, &snd_motu_spec_828mk3_hybrid), // Hybrid.
 	SND_MOTU_DEV_ENTRY(0x000033, &snd_motu_spec_audio_express),
 	SND_MOTU_DEV_ENTRY(0x000045, &snd_motu_spec_4pre),
 	{ }
diff --git a/sound/firewire/motu/motu.h b/sound/firewire/motu/motu.h
index 3d0236ee6716..73f36d1be515 100644
--- a/sound/firewire/motu/motu.h
+++ b/sound/firewire/motu/motu.h
@@ -39,15 +39,21 @@ struct snd_motu_packet_format {
 	unsigned char pcm_chunks[3];
 };
 
+struct amdtp_motu_cache {
+	unsigned int *event_offsets;
+	unsigned int size;
+	unsigned int tail;
+	unsigned int tx_cycle_count;
+	unsigned int head;
+	unsigned int rx_cycle_count;
+};
+
 struct snd_motu {
 	struct snd_card *card;
 	struct fw_unit *unit;
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
 	/* Model dependent information. */
 	const struct snd_motu_spec *spec;
 
@@ -70,6 +76,8 @@ struct snd_motu {
 	wait_queue_head_t hwdep_wait;
 
 	struct amdtp_domain domain;
+
+	struct amdtp_motu_cache cache;
 };
 
 enum snd_motu_spec_flags {
@@ -99,6 +107,7 @@ enum snd_motu_clock_source {
 };
 
 enum snd_motu_protocol_version {
+	SND_MOTU_PROTOCOL_V1,
 	SND_MOTU_PROTOCOL_V2,
 	SND_MOTU_PROTOCOL_V3,
 };
@@ -106,25 +115,31 @@ enum snd_motu_protocol_version {
 struct snd_motu_spec {
 	const char *const name;
 	enum snd_motu_protocol_version protocol_version;
-	enum snd_motu_spec_flags flags;
+	// The combination of snd_motu_spec_flags enumeration-constants.
+	unsigned int flags;
 
 	unsigned char tx_fixed_pcm_chunks[3];
 	unsigned char rx_fixed_pcm_chunks[3];
 };
 
+extern const struct snd_motu_spec snd_motu_spec_828;
+extern const struct snd_motu_spec snd_motu_spec_896;
+
 extern const struct snd_motu_spec snd_motu_spec_828mk2;
 extern const struct snd_motu_spec snd_motu_spec_traveler;
 extern const struct snd_motu_spec snd_motu_spec_ultralite;
 extern const struct snd_motu_spec snd_motu_spec_8pre;
 
-extern const struct snd_motu_spec snd_motu_spec_828mk3;
+extern const struct snd_motu_spec snd_motu_spec_828mk3_fw;
+extern const struct snd_motu_spec snd_motu_spec_828mk3_hybrid;
 extern const struct snd_motu_spec snd_motu_spec_ultralite_mk3;
 extern const struct snd_motu_spec snd_motu_spec_audio_express;
 extern const struct snd_motu_spec snd_motu_spec_4pre;
 
 int amdtp_motu_init(struct amdtp_stream *s, struct fw_unit *unit,
 		    enum amdtp_stream_direction dir,
-		    const struct snd_motu_spec *spec);
+		    const struct snd_motu_spec *spec,
+		    struct amdtp_motu_cache *cache);
 int amdtp_motu_set_parameters(struct amdtp_stream *s, unsigned int rate,
 			      unsigned int midi_ports,
 			      struct snd_motu_packet_format *formats);
@@ -160,6 +175,16 @@ int snd_motu_create_midi_devices(struct snd_motu *motu);
 
 int snd_motu_create_hwdep_device(struct snd_motu *motu);
 
+int snd_motu_protocol_v1_get_clock_rate(struct snd_motu *motu,
+					unsigned int *rate);
+int snd_motu_protocol_v1_set_clock_rate(struct snd_motu *motu,
+					unsigned int rate);
+int snd_motu_protocol_v1_get_clock_source(struct snd_motu *motu,
+					  enum snd_motu_clock_source *src);
+int snd_motu_protocol_v1_switch_fetching_mode(struct snd_motu *motu,
+					      bool enable);
+int snd_motu_protocol_v1_cache_packet_formats(struct snd_motu *motu);
+
 int snd_motu_protocol_v2_get_clock_rate(struct snd_motu *motu,
 					unsigned int *rate);
 int snd_motu_protocol_v2_set_clock_rate(struct snd_motu *motu,
@@ -187,6 +212,8 @@ static inline int snd_motu_protocol_get_clock_rate(struct snd_motu *motu,
 		return snd_motu_protocol_v2_get_clock_rate(motu, rate);
 	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V3)
 		return snd_motu_protocol_v3_get_clock_rate(motu, rate);
+	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V1)
+		return snd_motu_protocol_v1_get_clock_rate(motu, rate);
 	else
 		return -ENXIO;
 }
@@ -198,6 +225,8 @@ static inline int snd_motu_protocol_set_clock_rate(struct snd_motu *motu,
 		return snd_motu_protocol_v2_set_clock_rate(motu, rate);
 	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V3)
 		return snd_motu_protocol_v3_set_clock_rate(motu, rate);
+	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V1)
+		return snd_motu_protocol_v1_set_clock_rate(motu, rate);
 	else
 		return -ENXIO;
 }
@@ -209,6 +238,8 @@ static inline int snd_motu_protocol_get_clock_source(struct snd_motu *motu,
 		return snd_motu_protocol_v2_get_clock_source(motu, source);
 	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V3)
 		return snd_motu_protocol_v3_get_clock_source(motu, source);
+	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V1)
+		return snd_motu_protocol_v1_get_clock_source(motu, source);
 	else
 		return -ENXIO;
 }
@@ -220,6 +251,8 @@ static inline int snd_motu_protocol_switch_fetching_mode(struct snd_motu *motu,
 		return snd_motu_protocol_v2_switch_fetching_mode(motu, enable);
 	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V3)
 		return snd_motu_protocol_v3_switch_fetching_mode(motu, enable);
+	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V1)
+		return snd_motu_protocol_v1_switch_fetching_mode(motu, enable);
 	else
 		return -ENXIO;
 }
@@ -230,6 +263,8 @@ static inline int snd_motu_protocol_cache_packet_formats(struct snd_motu *motu)
 		return snd_motu_protocol_v2_cache_packet_formats(motu);
 	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V3)
 		return snd_motu_protocol_v3_cache_packet_formats(motu);
+	else if (motu->spec->protocol_version == SND_MOTU_PROTOCOL_V1)
+		return snd_motu_protocol_v1_cache_packet_formats(motu);
 	else
 		return -ENXIO;
 }
diff --git a/sound/firewire/oxfw/oxfw-stream.c b/sound/firewire/oxfw/oxfw-stream.c
index 80c9dc13f1b5..0ef242fdd3bc 100644
--- a/sound/firewire/oxfw/oxfw-stream.c
+++ b/sound/firewire/oxfw/oxfw-stream.c
@@ -9,7 +9,7 @@
 #include <linux/delay.h>
 
 #define AVC_GENERIC_FRAME_MAXIMUM_BYTES	512
-#define CALLBACK_TIMEOUT	200
+#define READY_TIMEOUT_MS	200
 
 /*
  * According to datasheet of Oxford Semiconductor:
@@ -153,12 +153,23 @@ static int init_stream(struct snd_oxfw *oxfw, struct amdtp_stream *stream)
 	struct cmp_connection *conn;
 	enum cmp_direction c_dir;
 	enum amdtp_stream_direction s_dir;
+	unsigned int flags = CIP_UNAWARE_SYT;
 	int err;
 
+	if (!(oxfw->quirks & SND_OXFW_QUIRK_BLOCKING_TRANSMISSION))
+		flags |= CIP_NONBLOCKING;
+	else
+		flags |= CIP_BLOCKING;
+
 	if (stream == &oxfw->tx_stream) {
 		conn = &oxfw->out_conn;
 		c_dir = CMP_OUTPUT;
 		s_dir = AMDTP_IN_STREAM;
+
+		if (oxfw->quirks & SND_OXFW_QUIRK_JUMBO_PAYLOAD)
+			flags |= CIP_JUMBO_PAYLOAD;
+		if (oxfw->quirks & SND_OXFW_QUIRK_WRONG_DBS)
+			flags |= CIP_WRONG_DBS;
 	} else {
 		conn = &oxfw->in_conn;
 		c_dir = CMP_INPUT;
@@ -169,24 +180,12 @@ static int init_stream(struct snd_oxfw *oxfw, struct amdtp_stream *stream)
 	if (err < 0)
 		return err;
 
-	err = amdtp_am824_init(stream, oxfw->unit, s_dir, CIP_NONBLOCKING);
+	err = amdtp_am824_init(stream, oxfw->unit, s_dir, flags);
 	if (err < 0) {
 		cmp_connection_destroy(conn);
 		return err;
 	}
 
-	/*
-	 * OXFW starts to transmit packets with non-zero dbc.
-	 * OXFW postpone transferring packets till handling any asynchronous
-	 * packets. As a result, next isochronous packet includes more data
-	 * blocks than IEC 61883-6 defines.
-	 */
-	if (stream == &oxfw->tx_stream) {
-		oxfw->tx_stream.flags |= CIP_JUMBO_PAYLOAD;
-		if (oxfw->wrong_dbs)
-			oxfw->tx_stream.flags |= CIP_WRONG_DBS;
-	}
-
 	return 0;
 }
 
@@ -338,6 +337,9 @@ int snd_oxfw_stream_start_duplex(struct snd_oxfw *oxfw)
 	}
 
 	if (!amdtp_stream_running(&oxfw->rx_stream)) {
+		unsigned int tx_init_skip_cycles = 0;
+		bool replay_seq = false;
+
 		err = start_stream(oxfw, &oxfw->rx_stream);
 		if (err < 0) {
 			dev_err(&oxfw->unit->device,
@@ -353,26 +355,27 @@ int snd_oxfw_stream_start_duplex(struct snd_oxfw *oxfw)
 					"fail to prepare tx stream: %d\n", err);
 				goto error;
 			}
+
+			if (oxfw->quirks & SND_OXFW_QUIRK_JUMBO_PAYLOAD) {
+				// Just after changing sampling transfer frequency, many cycles are
+				// skipped for packet transmission.
+				tx_init_skip_cycles = 400;
+			} else {
+				replay_seq = true;
+			}
 		}
 
-		err = amdtp_domain_start(&oxfw->domain, 0);
+		// NOTE: The device ignores presentation time expressed by the value of syt field
+		// of CIP header in received packets. The sequence of the number of data blocks per
+		// packet is important for media clock recovery.
+		err = amdtp_domain_start(&oxfw->domain, tx_init_skip_cycles, replay_seq, false);
 		if (err < 0)
 			goto error;
 
-		// Wait first packet.
-		if (!amdtp_stream_wait_callback(&oxfw->rx_stream,
-						CALLBACK_TIMEOUT)) {
+		if (!amdtp_domain_wait_ready(&oxfw->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
-
-		if (oxfw->has_output) {
-			if (!amdtp_stream_wait_callback(&oxfw->tx_stream,
-							CALLBACK_TIMEOUT)) {
-				err = -ETIMEDOUT;
-				goto error;
-			}
-		}
 	}
 
 	return 0;
diff --git a/sound/firewire/oxfw/oxfw.c b/sound/firewire/oxfw/oxfw.c
index 9eea25c46dc7..84971d78d152 100644
--- a/sound/firewire/oxfw/oxfw.c
+++ b/sound/firewire/oxfw/oxfw.c
@@ -23,6 +23,8 @@
 #define OUI_APOGEE		0x0003db
 
 #define MODEL_SATELLITE		0x00200f
+#define MODEL_SCS1M		0x001000
+#define MODEL_DUET_FW		0x01dddd
 
 #define SPECIFIER_1394TA	0x00a02d
 #define VERSION_AVC		0x010001
@@ -46,8 +48,6 @@ static bool detect_loud_models(struct fw_unit *unit)
 		"Onyx-i",
 		"Onyx 1640i",
 		"d.Pro",
-		"Mackie Onyx Satellite",
-		"Tapco LINK.firewire 4x6",
 		"U.420"};
 	char model[32];
 	int err;
@@ -60,7 +60,7 @@ static bool detect_loud_models(struct fw_unit *unit)
 	return match_string(models, ARRAY_SIZE(models), model) >= 0;
 }
 
-static int name_card(struct snd_oxfw *oxfw)
+static int name_card(struct snd_oxfw *oxfw, const struct ieee1394_device_id *entry)
 {
 	struct fw_device *fw_dev = fw_parent_device(oxfw->unit);
 	const struct compat_info *info;
@@ -88,10 +88,12 @@ static int name_card(struct snd_oxfw *oxfw)
 		goto end;
 	be32_to_cpus(&firmware);
 
+	if (firmware >> 20 == 0x970)
+		oxfw->quirks |= SND_OXFW_QUIRK_JUMBO_PAYLOAD;
+
 	/* to apply card definitions */
-	if (oxfw->entry->vendor_id == VENDOR_GRIFFIN ||
-	    oxfw->entry->vendor_id == VENDOR_LACIE) {
-		info = (const struct compat_info *)oxfw->entry->driver_data;
+	if (entry->vendor_id == VENDOR_GRIFFIN || entry->vendor_id == VENDOR_LACIE) {
+		info = (const struct compat_info *)entry->driver_data;
 		d = info->driver_name;
 		v = info->vendor_name;
 		m = info->model_name;
@@ -120,9 +122,12 @@ static void oxfw_card_free(struct snd_card *card)
 
 	if (oxfw->has_output || oxfw->has_input)
 		snd_oxfw_stream_destroy_duplex(oxfw);
+
+	mutex_destroy(&oxfw->mutex);
+	fw_unit_put(oxfw->unit);
 }
 
-static int detect_quirks(struct snd_oxfw *oxfw)
+static int detect_quirks(struct snd_oxfw *oxfw, const struct ieee1394_device_id *entry)
 {
 	struct fw_device *fw_dev = fw_parent_device(oxfw->unit);
 	struct fw_csr_iterator it;
@@ -133,28 +138,35 @@ static int detect_quirks(struct snd_oxfw *oxfw)
 	 * Add ALSA control elements for two models to keep compatibility to
 	 * old firewire-speaker module.
 	 */
-	if (oxfw->entry->vendor_id == VENDOR_GRIFFIN)
+	if (entry->vendor_id == VENDOR_GRIFFIN)
 		return snd_oxfw_add_spkr(oxfw, false);
-	if (oxfw->entry->vendor_id == VENDOR_LACIE)
+	if (entry->vendor_id == VENDOR_LACIE)
 		return snd_oxfw_add_spkr(oxfw, true);
 
 	/*
 	 * Stanton models supports asynchronous transactions for unique MIDI
 	 * messages.
 	 */
-	if (oxfw->entry->vendor_id == OUI_STANTON) {
-		/* No physical MIDI ports. */
+	if (entry->vendor_id == OUI_STANTON) {
+		oxfw->quirks |= SND_OXFW_QUIRK_SCS_TRANSACTION;
+		if (entry->model_id == MODEL_SCS1M)
+			oxfw->quirks |= SND_OXFW_QUIRK_BLOCKING_TRANSMISSION;
+
+		// No physical MIDI ports.
 		oxfw->midi_input_ports = 0;
 		oxfw->midi_output_ports = 0;
 
 		return snd_oxfw_scs1x_add(oxfw);
 	}
 
+	if (entry->vendor_id == OUI_APOGEE && entry->model_id == MODEL_DUET_FW)
+		oxfw->quirks |= SND_OXFW_QUIRK_BLOCKING_TRANSMISSION;
+
 	/*
 	 * TASCAM FireOne has physical control and requires a pair of additional
 	 * MIDI ports.
 	 */
-	if (oxfw->entry->vendor_id == VENDOR_TASCAM) {
+	if (entry->vendor_id == VENDOR_TASCAM) {
 		oxfw->midi_input_ports++;
 		oxfw->midi_output_ports++;
 		return 0;
@@ -175,27 +187,35 @@ static int detect_quirks(struct snd_oxfw *oxfw)
 	 * value in 'dbs' field of CIP header against its format information.
 	 */
 	if (vendor == VENDOR_LOUD && model == MODEL_SATELLITE)
-		oxfw->wrong_dbs = true;
+		oxfw->quirks |= SND_OXFW_QUIRK_WRONG_DBS;
 
 	return 0;
 }
 
-static void do_registration(struct work_struct *work)
+static int oxfw_probe(struct fw_unit *unit, const struct ieee1394_device_id *entry)
 {
-	struct snd_oxfw *oxfw = container_of(work, struct snd_oxfw, dwork.work);
+	struct snd_card *card;
+	struct snd_oxfw *oxfw;
 	int err;
 
-	if (oxfw->registered)
-		return;
+	if (entry->vendor_id == VENDOR_LOUD && entry->model_id == 0 && !detect_loud_models(unit))
+		return -ENODEV;
 
-	err = snd_card_new(&oxfw->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &oxfw->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*oxfw), &card);
 	if (err < 0)
-		return;
-	oxfw->card->private_free = oxfw_card_free;
-	oxfw->card->private_data = oxfw;
+		return err;
+	card->private_free = oxfw_card_free;
+
+	oxfw = card->private_data;
+	oxfw->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, oxfw);
+	oxfw->card = card;
 
-	err = name_card(oxfw);
+	mutex_init(&oxfw->mutex);
+	spin_lock_init(&oxfw->lock);
+	init_waitqueue_head(&oxfw->hwdep_wait);
+
+	err = name_card(oxfw, entry);
 	if (err < 0)
 		goto error;
 
@@ -203,7 +223,7 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = detect_quirks(oxfw);
+	err = detect_quirks(oxfw, entry);
 	if (err < 0)
 		goto error;
 
@@ -227,85 +247,38 @@ static void do_registration(struct work_struct *work)
 			goto error;
 	}
 
-	err = snd_card_register(oxfw->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	oxfw->registered = true;
-
-	return;
-error:
-	snd_card_free(oxfw->card);
-	dev_info(&oxfw->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int oxfw_probe(struct fw_unit *unit,
-		      const struct ieee1394_device_id *entry)
-{
-	struct snd_oxfw *oxfw;
-
-	if (entry->vendor_id == VENDOR_LOUD && !detect_loud_models(unit))
-		return -ENODEV;
-
-	/* Allocate this independent of sound card instance. */
-	oxfw = devm_kzalloc(&unit->device, sizeof(struct snd_oxfw), GFP_KERNEL);
-	if (!oxfw)
-		return -ENOMEM;
-	oxfw->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, oxfw);
-
-	oxfw->entry = entry;
-	mutex_init(&oxfw->mutex);
-	spin_lock_init(&oxfw->lock);
-	init_waitqueue_head(&oxfw->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&oxfw->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &oxfw->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void oxfw_bus_reset(struct fw_unit *unit)
 {
 	struct snd_oxfw *oxfw = dev_get_drvdata(&unit->device);
 
-	if (!oxfw->registered)
-		snd_fw_schedule_registration(unit, &oxfw->dwork);
-
 	fcp_bus_reset(oxfw->unit);
 
-	if (oxfw->registered) {
-		if (oxfw->has_output || oxfw->has_input) {
-			mutex_lock(&oxfw->mutex);
-			snd_oxfw_stream_update_duplex(oxfw);
-			mutex_unlock(&oxfw->mutex);
-		}
-
-		if (oxfw->entry->vendor_id == OUI_STANTON)
-			snd_oxfw_scs1x_update(oxfw);
+	if (oxfw->has_output || oxfw->has_input) {
+		mutex_lock(&oxfw->mutex);
+		snd_oxfw_stream_update_duplex(oxfw);
+		mutex_unlock(&oxfw->mutex);
 	}
+
+	if (oxfw->quirks & SND_OXFW_QUIRK_SCS_TRANSACTION)
+		snd_oxfw_scs1x_update(oxfw);
 }
 
 static void oxfw_remove(struct fw_unit *unit)
 {
 	struct snd_oxfw *oxfw = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&oxfw->dwork);
-
-	if (oxfw->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(oxfw->card);
-	}
-
-	mutex_destroy(&oxfw->mutex);
-	fw_unit_put(oxfw->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(oxfw->card);
 }
 
 static const struct compat_info griffin_firewave = {
@@ -320,81 +293,67 @@ static const struct compat_info lacie_speakers = {
 	.model_name = "FireWire Speakers",
 };
 
+#define OXFW_DEV_ENTRY(vendor, model, data) \
+{ \
+	.match_flags  = IEEE1394_MATCH_VENDOR_ID | \
+			IEEE1394_MATCH_MODEL_ID | \
+			IEEE1394_MATCH_SPECIFIER_ID | \
+			IEEE1394_MATCH_VERSION, \
+	.vendor_id    = vendor, \
+	.model_id     = model, \
+	.specifier_id = SPECIFIER_1394TA, \
+	.version      = VERSION_AVC, \
+	.driver_data  = (kernel_ulong_t)data, \
+}
+
 static const struct ieee1394_device_id oxfw_id_table[] = {
-	{
-		.match_flags  = IEEE1394_MATCH_VENDOR_ID |
-				IEEE1394_MATCH_MODEL_ID |
-				IEEE1394_MATCH_SPECIFIER_ID |
-				IEEE1394_MATCH_VERSION,
-		.vendor_id    = VENDOR_GRIFFIN,
-		.model_id     = 0x00f970,
-		.specifier_id = SPECIFIER_1394TA,
-		.version      = VERSION_AVC,
-		.driver_data  = (kernel_ulong_t)&griffin_firewave,
-	},
-	{
-		.match_flags  = IEEE1394_MATCH_VENDOR_ID |
-				IEEE1394_MATCH_MODEL_ID |
-				IEEE1394_MATCH_SPECIFIER_ID |
-				IEEE1394_MATCH_VERSION,
-		.vendor_id    = VENDOR_LACIE,
-		.model_id     = 0x00f970,
-		.specifier_id = SPECIFIER_1394TA,
-		.version      = VERSION_AVC,
-		.driver_data  = (kernel_ulong_t)&lacie_speakers,
-	},
-	/* Behringer,F-Control Audio 202 */
-	{
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID,
-		.vendor_id	= VENDOR_BEHRINGER,
-		.model_id	= 0x00fc22,
-	},
-	/*
-	 * Any Mackie(Loud) models (name string/model id):
-	 *  Onyx-i series (former models):	0x081216
-	 *  Mackie Onyx Satellite:		0x00200f
-	 *  Tapco LINK.firewire 4x6:		0x000460
-	 *  d.4 pro:				Unknown
-	 *  U.420:				Unknown
-	 *  U.420d:				Unknown
-	 */
+	//
+	// OXFW970 devices:
+	// Initial firmware has a quirk to postpone isoc packet transmission during finishing async
+	// transaction. As a result, several isochronous cycles are skipped to transfer the packets
+	// and the audio data frames which should have been transferred during the cycles are put
+	// into packet at the first isoc cycle after the postpone. Furthermore, the value of SYT
+	// field in CIP header is not reliable as synchronization timing,
+	//
+	OXFW_DEV_ENTRY(VENDOR_GRIFFIN, 0x00f970, &griffin_firewave),
+	OXFW_DEV_ENTRY(VENDOR_LACIE, 0x00f970, &lacie_speakers),
+	// Behringer,F-Control Audio 202. The value of SYT field is not reliable at all.
+	OXFW_DEV_ENTRY(VENDOR_BEHRINGER, 0x00fc22, NULL),
+	// Loud Technologies, Tapco Link.FireWire 4x6. The value of SYT field is always 0xffff.
+	OXFW_DEV_ENTRY(VENDOR_LOUD, 0x000460, NULL),
+	// Loud Technologies, Mackie Onyx Satellite. Although revised version of firmware is
+	// installed to avoid the postpone, the value of SYT field is always 0xffff.
+	OXFW_DEV_ENTRY(VENDOR_LOUD, MODEL_SATELLITE, NULL),
+	// Miglia HarmonyAudio. Not yet identified.
+
+	//
+	// OXFW971 devices:
+	// The value of SYT field in CIP header is enough reliable. Both of blocking and non-blocking
+	// transmission methods are available.
+	//
+	// Any Mackie(Loud) models (name string/model id):
+	//  Onyx-i series (former models):	0x081216
+	//  Onyx 1640i:				0x001640
+	//  d.2 pro/d.4 pro (built-in card):	Unknown
+	//  U.420:				Unknown
+	//  U.420d:				Unknown
 	{
 		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
 				  IEEE1394_MATCH_SPECIFIER_ID |
 				  IEEE1394_MATCH_VERSION,
 		.vendor_id	= VENDOR_LOUD,
+		.model_id	= 0,
 		.specifier_id	= SPECIFIER_1394TA,
 		.version	= VERSION_AVC,
 	},
-	/* TASCAM, FireOne */
-	{
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID,
-		.vendor_id	= VENDOR_TASCAM,
-		.model_id	= 0x800007,
-	},
-	/* Stanton, Stanton Controllers & Systems 1 Mixer (SCS.1m) */
-	{
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID,
-		.vendor_id	= OUI_STANTON,
-		.model_id	= 0x001000,
-	},
-	/* Stanton, Stanton Controllers & Systems 1 Deck (SCS.1d) */
-	{
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID,
-		.vendor_id	= OUI_STANTON,
-		.model_id	= 0x002000,
-	},
-	// APOGEE, duet FireWire
-	{
-		.match_flags	= IEEE1394_MATCH_VENDOR_ID |
-				  IEEE1394_MATCH_MODEL_ID,
-		.vendor_id	= OUI_APOGEE,
-		.model_id	= 0x01dddd,
-	},
+	// TASCAM, FireOne.
+	OXFW_DEV_ENTRY(VENDOR_TASCAM, 0x800007, NULL),
+	// Stanton, Stanton Controllers & Systems 1 Mixer (SCS.1m).
+	OXFW_DEV_ENTRY(OUI_STANTON, MODEL_SCS1M, NULL),
+	// Stanton, Stanton Controllers & Systems 1 Deck (SCS.1d).
+	OXFW_DEV_ENTRY(OUI_STANTON, 0x002000, NULL),
+	// APOGEE, duet FireWire.
+	OXFW_DEV_ENTRY(OUI_APOGEE, MODEL_DUET_FW, NULL),
 	{ }
 };
 MODULE_DEVICE_TABLE(ieee1394, oxfw_id_table);
diff --git a/sound/firewire/oxfw/oxfw.h b/sound/firewire/oxfw/oxfw.h
index fa2d7f9e2dc3..ee47abcb0c90 100644
--- a/sound/firewire/oxfw/oxfw.h
+++ b/sound/firewire/oxfw/oxfw.h
@@ -32,6 +32,18 @@
 #include "../amdtp-am824.h"
 #include "../cmp.h"
 
+enum snd_oxfw_quirk {
+	// Postpone transferring packets during handling asynchronous transaction. As a result,
+	// next isochronous packet includes more events than one packet can include.
+	SND_OXFW_QUIRK_JUMBO_PAYLOAD = 0x01,
+	// The dbs field of CIP header in tx packet is wrong.
+	SND_OXFW_QUIRK_WRONG_DBS = 0x02,
+	// Blocking transmission mode is used.
+	SND_OXFW_QUIRK_BLOCKING_TRANSMISSION = 0x04,
+	// Stanton SCS1.d and SCS1.m support unique transaction.
+	SND_OXFW_QUIRK_SCS_TRANSACTION = 0x08,
+};
+
 /* This is an arbitrary number for convinience. */
 #define	SND_OXFW_STREAM_FORMAT_ENTRIES	10
 struct snd_oxfw {
@@ -40,10 +52,8 @@ struct snd_oxfw {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
-
-	bool wrong_dbs;
+	// The combination of snd_oxfw_quirk enumeration-constants.
+	unsigned int quirks;
 	bool has_output;
 	bool has_input;
 	u8 *tx_stream_formats[SND_OXFW_STREAM_FORMAT_ENTRIES];
@@ -62,7 +72,6 @@ struct snd_oxfw {
 	bool dev_lock_changed;
 	wait_queue_head_t hwdep_wait;
 
-	const struct ieee1394_device_id *entry;
 	void *spec;
 
 	struct amdtp_domain domain;
diff --git a/sound/firewire/tascam/amdtp-tascam.c b/sound/firewire/tascam/amdtp-tascam.c
index f823a2ab3544..64d66a802545 100644
--- a/sound/firewire/tascam/amdtp-tascam.c
+++ b/sound/firewire/tascam/amdtp-tascam.c
@@ -228,6 +228,7 @@ int amdtp_tscm_init(struct amdtp_stream *s, struct fw_unit *unit,
 		    enum amdtp_stream_direction dir, unsigned int pcm_channels)
 {
 	amdtp_stream_process_ctx_payloads_t process_ctx_payloads;
+	unsigned int flags = CIP_NONBLOCKING | CIP_SKIP_DBC_ZERO_CHECK | CIP_UNAWARE_SYT;
 	struct amdtp_tscm *p;
 	unsigned int fmt;
 	int err;
@@ -240,8 +241,7 @@ int amdtp_tscm_init(struct amdtp_stream *s, struct fw_unit *unit,
 		process_ctx_payloads = process_it_ctx_payloads;
 	}
 
-	err = amdtp_stream_init(s, unit, dir,
-			CIP_NONBLOCKING | CIP_SKIP_DBC_ZERO_CHECK, fmt,
+	err = amdtp_stream_init(s, unit, dir, flags, fmt,
 			process_ctx_payloads, sizeof(struct amdtp_tscm));
 	if (err < 0)
 		return 0;
@@ -249,8 +249,6 @@ int amdtp_tscm_init(struct amdtp_stream *s, struct fw_unit *unit,
 	if (dir == AMDTP_OUT_STREAM) {
 		// Use fixed value for FDF field.
 		s->ctx_data.rx.fdf = 0x00;
-		// Not used.
-		s->ctx_data.rx.syt_override = 0x0000;
 	}
 
 	/* This protocol uses fixed number of data channels for PCM samples. */
diff --git a/sound/firewire/tascam/tascam-stream.c b/sound/firewire/tascam/tascam-stream.c
index eb07e1decf9b..53e094cc411f 100644
--- a/sound/firewire/tascam/tascam-stream.c
+++ b/sound/firewire/tascam/tascam-stream.c
@@ -11,7 +11,7 @@
 #define CLOCK_STATUS_MASK      0xffff0000
 #define CLOCK_CONFIG_MASK      0x0000ffff
 
-#define CALLBACK_TIMEOUT 500
+#define READY_TIMEOUT_MS	4000
 
 static int get_clock(struct snd_tscm *tscm, u32 *data)
 {
@@ -423,6 +423,8 @@ int snd_tscm_stream_reserve_duplex(struct snd_tscm *tscm, unsigned int rate,
 			fw_iso_resources_free(&tscm->rx_resources);
 			return err;
 		}
+
+		tscm->need_long_tx_init_skip = (rate != curr_rate);
 	}
 
 	return 0;
@@ -454,6 +456,7 @@ int snd_tscm_stream_start_duplex(struct snd_tscm *tscm, unsigned int rate)
 
 	if (!amdtp_stream_running(&tscm->rx_stream)) {
 		int spd = fw_parent_device(tscm->unit)->max_speed;
+		unsigned int tx_init_skip_cycles;
 
 		err = set_stream_formats(tscm, rate);
 		if (err < 0)
@@ -473,14 +476,23 @@ int snd_tscm_stream_start_duplex(struct snd_tscm *tscm, unsigned int rate)
 		if (err < 0)
 			goto error;
 
-		err = amdtp_domain_start(&tscm->domain, 0);
+		if (tscm->need_long_tx_init_skip)
+			tx_init_skip_cycles = 16000;
+		else
+			tx_init_skip_cycles = 0;
+
+		// MEMO: Just after starting packet streaming, it transfers packets without any
+		// event. Enough after receiving the sequence of packets, it multiplexes events into
+		// the packet. However, just after changing sampling transfer frequency, it stops
+		// multiplexing during packet transmission. Enough after, it restarts multiplexing
+		// again. The device ignores presentation time expressed by the value of syt field
+		// of CIP header in received packets. The sequence of the number of data blocks per
+		// packet is important for media clock recovery.
+		err = amdtp_domain_start(&tscm->domain, tx_init_skip_cycles, true, true);
 		if (err < 0)
 			return err;
 
-		if (!amdtp_stream_wait_callback(&tscm->rx_stream,
-						CALLBACK_TIMEOUT) ||
-		    !amdtp_stream_wait_callback(&tscm->tx_stream,
-						CALLBACK_TIMEOUT)) {
+		if (!amdtp_domain_wait_ready(&tscm->domain, READY_TIMEOUT_MS)) {
 			err = -ETIMEDOUT;
 			goto error;
 		}
@@ -502,6 +514,8 @@ void snd_tscm_stream_stop_duplex(struct snd_tscm *tscm)
 
 		fw_iso_resources_free(&tscm->tx_resources);
 		fw_iso_resources_free(&tscm->rx_resources);
+
+		tscm->need_long_tx_init_skip = false;
 	}
 }
 
diff --git a/sound/firewire/tascam/tascam.c b/sound/firewire/tascam/tascam.c
index 75f2edd8e78f..eb58d3fcf087 100644
--- a/sound/firewire/tascam/tascam.c
+++ b/sound/firewire/tascam/tascam.c
@@ -90,19 +90,31 @@ static void tscm_card_free(struct snd_card *card)
 
 	snd_tscm_transaction_unregister(tscm);
 	snd_tscm_stream_destroy_duplex(tscm);
+
+	mutex_destroy(&tscm->mutex);
+	fw_unit_put(tscm->unit);
 }
 
-static void do_registration(struct work_struct *work)
+static int snd_tscm_probe(struct fw_unit *unit,
+			   const struct ieee1394_device_id *entry)
 {
-	struct snd_tscm *tscm = container_of(work, struct snd_tscm, dwork.work);
+	struct snd_card *card;
+	struct snd_tscm *tscm;
 	int err;
 
-	err = snd_card_new(&tscm->unit->device, -1, NULL, THIS_MODULE, 0,
-			   &tscm->card);
+	err = snd_card_new(&unit->device, -1, NULL, THIS_MODULE, sizeof(*tscm), &card);
 	if (err < 0)
-		return;
-	tscm->card->private_free = tscm_card_free;
-	tscm->card->private_data = tscm;
+		return err;
+	card->private_free = tscm_card_free;
+
+	tscm = card->private_data;
+	tscm->unit = fw_unit_get(unit);
+	dev_set_drvdata(&unit->device, tscm);
+	tscm->card = card;
+
+	mutex_init(&tscm->mutex);
+	spin_lock_init(&tscm->lock);
+	init_waitqueue_head(&tscm->hwdep_wait);
 
 	err = identify_model(tscm);
 	if (err < 0)
@@ -130,81 +142,33 @@ static void do_registration(struct work_struct *work)
 	if (err < 0)
 		goto error;
 
-	err = snd_card_register(tscm->card);
+	err = snd_card_register(card);
 	if (err < 0)
 		goto error;
 
-	tscm->registered = true;
-
-	return;
-error:
-	snd_card_free(tscm->card);
-	dev_info(&tscm->unit->device,
-		 "Sound card registration failed: %d\n", err);
-}
-
-static int snd_tscm_probe(struct fw_unit *unit,
-			   const struct ieee1394_device_id *entry)
-{
-	struct snd_tscm *tscm;
-
-	/* Allocate this independent of sound card instance. */
-	tscm = devm_kzalloc(&unit->device, sizeof(struct snd_tscm), GFP_KERNEL);
-	if (!tscm)
-		return -ENOMEM;
-	tscm->unit = fw_unit_get(unit);
-	dev_set_drvdata(&unit->device, tscm);
-
-	mutex_init(&tscm->mutex);
-	spin_lock_init(&tscm->lock);
-	init_waitqueue_head(&tscm->hwdep_wait);
-
-	/* Allocate and register this sound card later. */
-	INIT_DEFERRABLE_WORK(&tscm->dwork, do_registration);
-	snd_fw_schedule_registration(unit, &tscm->dwork);
-
 	return 0;
+error:
+	snd_card_free(card);
+	return err;
 }
 
 static void snd_tscm_update(struct fw_unit *unit)
 {
 	struct snd_tscm *tscm = dev_get_drvdata(&unit->device);
 
-	/* Postpone a workqueue for deferred registration. */
-	if (!tscm->registered)
-		snd_fw_schedule_registration(unit, &tscm->dwork);
-
 	snd_tscm_transaction_reregister(tscm);
 
-	/*
-	 * After registration, userspace can start packet streaming, then this
-	 * code block works fine.
-	 */
-	if (tscm->registered) {
-		mutex_lock(&tscm->mutex);
-		snd_tscm_stream_update_duplex(tscm);
-		mutex_unlock(&tscm->mutex);
-	}
+	mutex_lock(&tscm->mutex);
+	snd_tscm_stream_update_duplex(tscm);
+	mutex_unlock(&tscm->mutex);
 }
 
 static void snd_tscm_remove(struct fw_unit *unit)
 {
 	struct snd_tscm *tscm = dev_get_drvdata(&unit->device);
 
-	/*
-	 * Confirm to stop the work for registration before the sound card is
-	 * going to be released. The work is not scheduled again because bus
-	 * reset handler is not called anymore.
-	 */
-	cancel_delayed_work_sync(&tscm->dwork);
-
-	if (tscm->registered) {
-		// Block till all of ALSA character devices are released.
-		snd_card_free(tscm->card);
-	}
-
-	mutex_destroy(&tscm->mutex);
-	fw_unit_put(tscm->unit);
+	// Block till all of ALSA character devices are released.
+	snd_card_free(tscm->card);
 }
 
 static const struct ieee1394_device_id snd_tscm_id_table[] = {
diff --git a/sound/firewire/tascam/tascam.h b/sound/firewire/tascam/tascam.h
index 78b7a08986a1..d07ffcb27be6 100644
--- a/sound/firewire/tascam/tascam.h
+++ b/sound/firewire/tascam/tascam.h
@@ -70,8 +70,6 @@ struct snd_tscm {
 	struct mutex mutex;
 	spinlock_t lock;
 
-	bool registered;
-	struct delayed_work dwork;
 	const struct snd_tscm_spec *spec;
 
 	struct fw_iso_resources tx_resources;
@@ -99,6 +97,7 @@ struct snd_tscm {
 	unsigned int push_pos;
 
 	struct amdtp_domain domain;
+	bool need_long_tx_init_skip;
 };
 
 #define TSCM_ADDR_BASE			0xffff00000000ull