bcm27xx: import latest patches from the RPi foundation

bcm2708: boot tested on RPi B+ v1.2
bcm2709: boot tested on RPi 3B v1.2 and RPi 4B v1.1 4G
bcm2710: boot tested on RPi 3B v1.2
bcm2711: boot tested on RPi 4B v1.1 4G

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>

1 files changed, 801 insertions, 0 deletions

diff --git a/target/linux/bcm27xx/patches-5.4/950-0691-bcm2835-dma-Add-proper-40-bit-DMA-support.patch b/target/linux/bcm27xx/patches-5.4/950-0691-bcm2835-dma-Add-proper-40-bit-DMA-support.patch
new file mode 100644
index 0000000000..3fb47589ee
--- /dev/null
+++ b/target/linux/bcm27xx/patches-5.4/950-0691-bcm2835-dma-Add-proper-40-bit-DMA-support.patch
@@ -0,0 +1,801 @@
+From a2f673d1aa39752608e5f4838ed9656b38cbc4b9 Mon Sep 17 00:00:00 2001
+From: Phil Elwell <phil@raspberrypi.org>
+Date: Thu, 4 Apr 2019 13:33:47 +0100
+Subject: [PATCH] bcm2835-dma: Add proper 40-bit DMA support
+
+BCM2711 has 4 DMA channels with a 40-bit address range, allowing them
+to access the full 4GB of memory on a Pi 4.
+
+Signed-off-by: Phil Elwell <phil@raspberrypi.org>
+---
+ drivers/dma/bcm2835-dma.c | 485 ++++++++++++++++++++++++++++++++------
+ 1 file changed, 412 insertions(+), 73 deletions(-)
+
+--- a/drivers/dma/bcm2835-dma.c
++++ b/drivers/dma/bcm2835-dma.c
+@@ -38,6 +38,11 @@
+ #define BCM2835_DMA_MAX_DMA_CHAN_SUPPORTED 14
+ #define BCM2835_DMA_CHAN_NAME_SIZE 8
+ #define BCM2835_DMA_BULK_MASK  BIT(0)
++#define BCM2711_DMA_MEMCPY_CHAN 14
++
++struct bcm2835_dma_cfg_data {
++	u32	chan_40bit_mask;
++};
+ 
+ /**
+  * struct bcm2835_dmadev - BCM2835 DMA controller
+@@ -52,6 +57,7 @@ struct bcm2835_dmadev {
+ 	void __iomem *base;
+ 	struct device_dma_parameters dma_parms;
+ 	dma_addr_t zero_page;
++	const struct bcm2835_dma_cfg_data *cfg_data;
+ };
+ 
+ struct bcm2835_dma_cb {
+@@ -64,6 +70,17 @@ struct bcm2835_dma_cb {
+ 	uint32_t pad[2];
+ };
+ 
++struct bcm2711_dma40_scb {
++	uint32_t ti;
++	uint32_t src;
++	uint32_t srci;
++	uint32_t dst;
++	uint32_t dsti;
++	uint32_t len;
++	uint32_t next_cb;
++	uint32_t rsvd;
++};
++
+ struct bcm2835_cb_entry {
+ 	struct bcm2835_dma_cb *cb;
+ 	dma_addr_t paddr;
+@@ -84,6 +101,7 @@ struct bcm2835_chan {
+ 	unsigned int irq_flags;
+ 
+ 	bool is_lite_channel;
++	bool is_40bit_channel;
+ };
+ 
+ struct bcm2835_desc {
+@@ -173,13 +191,118 @@ struct bcm2835_desc {
+ #define BCM2835_DMA_DATA_TYPE_S128	16
+ 
+ /* Valid only for channels 0 - 14, 15 has its own base address */
+-#define BCM2835_DMA_CHAN(n)	((n) << 8) /* Base address */
++#define BCM2835_DMA_CHAN_SIZE	0x100
++#define BCM2835_DMA_CHAN(n)	((n) * BCM2835_DMA_CHAN_SIZE) /* Base address */
+ #define BCM2835_DMA_CHANIO(base, n) ((base) + BCM2835_DMA_CHAN(n))
+ 
+ /* the max dma length for different channels */
+ #define MAX_DMA_LEN SZ_1G
+ #define MAX_LITE_DMA_LEN (SZ_64K - 4)
+ 
++/* 40-bit DMA support */
++#define BCM2711_DMA40_CS	0x00
++#define BCM2711_DMA40_CB	0x04
++#define BCM2711_DMA40_DEBUG	0x0c
++#define BCM2711_DMA40_TI	0x10
++#define BCM2711_DMA40_SRC	0x14
++#define BCM2711_DMA40_SRCI	0x18
++#define BCM2711_DMA40_DEST	0x1c
++#define BCM2711_DMA40_DESTI	0x20
++#define BCM2711_DMA40_LEN	0x24
++#define BCM2711_DMA40_NEXT_CB	0x28
++#define BCM2711_DMA40_DEBUG2	0x2c
++
++#define BCM2711_DMA40_ACTIVE		BIT(0)
++#define BCM2711_DMA40_END		BIT(1)
++#define BCM2711_DMA40_INT		BIT(2)
++#define BCM2711_DMA40_DREQ		BIT(3)  /* DREQ state */
++#define BCM2711_DMA40_RD_PAUSED		BIT(4)  /* Reading is paused */
++#define BCM2711_DMA40_WR_PAUSED		BIT(5)  /* Writing is paused */
++#define BCM2711_DMA40_DREQ_PAUSED	BIT(6)  /* Is paused by DREQ flow control */
++#define BCM2711_DMA40_WAITING_FOR_WRITES BIT(7)  /* Waiting for last write */
++#define BCM2711_DMA40_ERR		BIT(10)
++#define BCM2711_DMA40_QOS(x)		(((x) & 0x1f) << 16)
++#define BCM2711_DMA40_PANIC_QOS(x)	(((x) & 0x1f) << 20)
++#define BCM2711_DMA40_WAIT_FOR_WRITES	BIT(28)
++#define BCM2711_DMA40_DISDEBUG		BIT(29)
++#define BCM2711_DMA40_ABORT		BIT(30)
++#define BCM2711_DMA40_HALT		BIT(31)
++#define BCM2711_DMA40_CS_FLAGS(x) (x & (BCM2711_DMA40_QOS(15) | \
++					BCM2711_DMA40_PANIC_QOS(15) | \
++					BCM2711_DMA40_WAIT_FOR_WRITES |	\
++					BCM2711_DMA40_DISDEBUG))
++
++/* Transfer information bits */
++#define BCM2711_DMA40_INTEN		BIT(0)
++#define BCM2711_DMA40_TDMODE		BIT(1) /* 2D-Mode */
++#define BCM2711_DMA40_WAIT_RESP		BIT(2) /* wait for AXI write to be acked */
++#define BCM2711_DMA40_WAIT_RD_RESP	BIT(3) /* wait for AXI read to complete */
++#define BCM2711_DMA40_PER_MAP(x)	((x & 31) << 9) /* REQ source */
++#define BCM2711_DMA40_S_DREQ		BIT(14) /* enable SREQ for source */
++#define BCM2711_DMA40_D_DREQ		BIT(15) /* enable DREQ for destination */
++#define BCM2711_DMA40_S_WAIT(x)		((x & 0xff) << 16) /* add DMA read-wait cycles */
++#define BCM2711_DMA40_D_WAIT(x)		((x & 0xff) << 24) /* add DMA write-wait cycles */
++
++/* debug register bits */
++#define BCM2711_DMA40_DEBUG_WRITE_ERR		BIT(0)
++#define BCM2711_DMA40_DEBUG_FIFO_ERR		BIT(1)
++#define BCM2711_DMA40_DEBUG_READ_ERR		BIT(2)
++#define BCM2711_DMA40_DEBUG_READ_CB_ERR		BIT(3)
++#define BCM2711_DMA40_DEBUG_IN_ON_ERR		BIT(8)
++#define BCM2711_DMA40_DEBUG_ABORT_ON_ERR	BIT(9)
++#define BCM2711_DMA40_DEBUG_HALT_ON_ERR		BIT(10)
++#define BCM2711_DMA40_DEBUG_DISABLE_CLK_GATE	BIT(11)
++#define BCM2711_DMA40_DEBUG_RSTATE_SHIFT	14
++#define BCM2711_DMA40_DEBUG_RSTATE_BITS		4
++#define BCM2711_DMA40_DEBUG_WSTATE_SHIFT	18
++#define BCM2711_DMA40_DEBUG_WSTATE_BITS		4
++#define BCM2711_DMA40_DEBUG_RESET		BIT(23)
++#define BCM2711_DMA40_DEBUG_ID_SHIFT		24
++#define BCM2711_DMA40_DEBUG_ID_BITS		4
++#define BCM2711_DMA40_DEBUG_VERSION_SHIFT	28
++#define BCM2711_DMA40_DEBUG_VERSION_BITS	4
++
++/* Valid only for channels 0 - 3 (11 - 14) */
++#define BCM2711_DMA40_CHAN(n)	(((n) + 11) << 8) /* Base address */
++#define BCM2711_DMA40_CHANIO(base, n) ((base) + BCM2711_DMA_CHAN(n))
++
++/* the max dma length for different channels */
++#define MAX_DMA40_LEN SZ_1G
++
++#define BCM2711_DMA40_BURST_LEN(x)	((min(x,16) - 1) << 8)
++#define BCM2711_DMA40_INC		BIT(12)
++#define BCM2711_DMA40_SIZE_32		(0 << 13)
++#define BCM2711_DMA40_SIZE_64		(1 << 13)
++#define BCM2711_DMA40_SIZE_128		(2 << 13)
++#define BCM2711_DMA40_SIZE_256		(3 << 13)
++#define BCM2711_DMA40_IGNORE		BIT(15)
++#define BCM2711_DMA40_STRIDE(x)		((x) << 16) /* For 2D mode */
++
++#define BCM2711_DMA40_MEMCPY_FLAGS \
++	(BCM2711_DMA40_QOS(0) | \
++	 BCM2711_DMA40_PANIC_QOS(0) | \
++	 BCM2711_DMA40_WAIT_FOR_WRITES | \
++	 BCM2711_DMA40_DISDEBUG)
++
++#define BCM2711_DMA40_MEMCPY_XFER_INFO \
++	(BCM2711_DMA40_SIZE_128 | \
++	 BCM2711_DMA40_INC | \
++	 BCM2711_DMA40_BURST_LEN(16))
++
++struct bcm2835_dmadev *memcpy_parent;
++static void __iomem *memcpy_chan;
++static struct bcm2711_dma40_scb *memcpy_scb;
++static dma_addr_t memcpy_scb_dma;
++DEFINE_SPINLOCK(memcpy_lock);
++
++static const struct bcm2835_dma_cfg_data bcm2835_dma_cfg = {
++	.chan_40bit_mask = 0,
++};
++
++static const struct bcm2835_dma_cfg_data bcm2711_dma_cfg = {
++	.chan_40bit_mask = BIT(11) | BIT(12) | BIT(13) | BIT(14),
++};
++
+ static inline size_t bcm2835_dma_max_frame_length(struct bcm2835_chan *c)
+ {
+ 	/* lite and normal channels have different max frame length */
+@@ -209,6 +332,32 @@ static inline struct bcm2835_desc *to_bc
+ 	return container_of(t, struct bcm2835_desc, vd.tx);
+ }
+ 
++static inline uint32_t to_bcm2711_ti(uint32_t info)
++{
++	return ((info & BCM2835_DMA_INT_EN) ? BCM2711_DMA40_INTEN : 0) |
++		((info & BCM2835_DMA_WAIT_RESP) ? BCM2711_DMA40_WAIT_RESP : 0) |
++		((info & BCM2835_DMA_S_DREQ) ?
++		 (BCM2711_DMA40_S_DREQ | BCM2711_DMA40_WAIT_RD_RESP) : 0) |
++		((info & BCM2835_DMA_D_DREQ) ? BCM2711_DMA40_D_DREQ : 0) |
++		BCM2711_DMA40_PER_MAP((info >> 16) & 0x1f);
++}
++
++static inline uint32_t to_bcm2711_srci(uint32_t info)
++{
++	return ((info & BCM2835_DMA_S_INC) ? BCM2711_DMA40_INC : 0);
++}
++
++static inline uint32_t to_bcm2711_dsti(uint32_t info)
++{
++	return ((info & BCM2835_DMA_D_INC) ? BCM2711_DMA40_INC : 0);
++}
++
++static inline uint32_t to_bcm2711_cbaddr(dma_addr_t addr)
++{
++	BUG_ON(addr & 0x1f);
++	return (addr >> 5);
++}
++
+ static void bcm2835_dma_free_cb_chain(struct bcm2835_desc *desc)
+ {
+ 	size_t i;
+@@ -227,45 +376,53 @@ static void bcm2835_dma_desc_free(struct
+ }
+ 
+ static void bcm2835_dma_create_cb_set_length(
+-	struct bcm2835_chan *chan,
++	struct bcm2835_chan *c,
+ 	struct bcm2835_dma_cb *control_block,
+ 	size_t len,
+ 	size_t period_len,
+ 	size_t *total_len,
+ 	u32 finalextrainfo)
+ {
+-	size_t max_len = bcm2835_dma_max_frame_length(chan);
++	size_t max_len = bcm2835_dma_max_frame_length(c);
++	uint32_t cb_len;
+ 
+ 	/* set the length taking lite-channel limitations into account */
+-	control_block->length = min_t(u32, len, max_len);
++	cb_len = min_t(u32, len, max_len);
+ 
+-	/* finished if we have no period_length */
+-	if (!period_len)
+-		return;
++	if (period_len) {
++		/*
++		 * period_len means: that we need to generate
++		 * transfers that are terminating at every
++		 * multiple of period_len - this is typically
++		 * used to set the interrupt flag in info
++		 * which is required during cyclic transfers
++		 */
+ 
+-	/*
+-	 * period_len means: that we need to generate
+-	 * transfers that are terminating at every
+-	 * multiple of period_len - this is typically
+-	 * used to set the interrupt flag in info
+-	 * which is required during cyclic transfers
+-	 */
++		/* have we filled in period_length yet? */
++		if (*total_len + cb_len < period_len) {
++			/* update number of bytes in this period so far */
++			*total_len += cb_len;
++		} else {
++			/* calculate the length that remains to reach period_len */
++			cb_len = period_len - *total_len;
+ 
+-	/* have we filled in period_length yet? */
+-	if (*total_len + control_block->length < period_len) {
+-		/* update number of bytes in this period so far */
+-		*total_len += control_block->length;
+-		return;
++			/* reset total_length for next period */
++			*total_len = 0;
++		}
+ 	}
+ 
+-	/* calculate the length that remains to reach period_length */
+-	control_block->length = period_len - *total_len;
+-
+-	/* reset total_length for next period */
+-	*total_len = 0;
+-
+-	/* add extrainfo bits in info */
+-	control_block->info |= finalextrainfo;
++	if (c->is_40bit_channel) {
++		struct bcm2711_dma40_scb *scb =
++			(struct bcm2711_dma40_scb *)control_block;
++
++		scb->len = cb_len;
++		/* add extrainfo bits to ti */
++		scb->ti |= to_bcm2711_ti(finalextrainfo);
++	} else {
++		control_block->length = cb_len;
++		/* add extrainfo bits to info */
++		control_block->info |= finalextrainfo;
++	}
+ }
+ 
+ static inline size_t bcm2835_dma_count_frames_for_sg(
+@@ -288,7 +445,7 @@ static inline size_t bcm2835_dma_count_f
+ /**
+  * bcm2835_dma_create_cb_chain - create a control block and fills data in
+  *
+- * @chan:           the @dma_chan for which we run this
++ * @c:              the @bcm2835_chan for which we run this
+  * @direction:      the direction in which we transfer
+  * @cyclic:         it is a cyclic transfer
+  * @info:           the default info bits to apply per controlblock
+@@ -306,12 +463,11 @@ static inline size_t bcm2835_dma_count_f
+  * @gfp:            the GFP flag to use for allocation
+  */
+ static struct bcm2835_desc *bcm2835_dma_create_cb_chain(
+-	struct dma_chan *chan, enum dma_transfer_direction direction,
++	struct bcm2835_chan *c, enum dma_transfer_direction direction,
+ 	bool cyclic, u32 info, u32 finalextrainfo, size_t frames,
+ 	dma_addr_t src, dma_addr_t dst, size_t buf_len,
+ 	size_t period_len, gfp_t gfp)
+ {
+-	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
+ 	size_t len = buf_len, total_len;
+ 	size_t frame;
+ 	struct bcm2835_desc *d;
+@@ -343,11 +499,23 @@ static struct bcm2835_desc *bcm2835_dma_
+ 
+ 		/* fill in the control block */
+ 		control_block = cb_entry->cb;
+-		control_block->info = info;
+-		control_block->src = src;
+-		control_block->dst = dst;
+-		control_block->stride = 0;
+-		control_block->next = 0;
++		if (c->is_40bit_channel) {
++			struct bcm2711_dma40_scb *scb =
++				(struct bcm2711_dma40_scb *)control_block;
++			scb->ti = to_bcm2711_ti(info);
++			scb->src = lower_32_bits(src);
++			scb->srci= upper_32_bits(src) | to_bcm2711_srci(info);
++			scb->dst = lower_32_bits(dst);
++			scb->dsti = upper_32_bits(dst) | to_bcm2711_dsti(info);
++			scb->next_cb = 0;
++		} else {
++			control_block->info = info;
++			control_block->src = src;
++			control_block->dst = dst;
++			control_block->stride = 0;
++			control_block->next = 0;
++		}
++
+ 		/* set up length in control_block if requested */
+ 		if (buf_len) {
+ 			/* calculate length honoring period_length */
+@@ -361,7 +529,11 @@ static struct bcm2835_desc *bcm2835_dma_
+ 		}
+ 
+ 		/* link this the last controlblock */
+-		if (frame)
++		if (frame && c->is_40bit_channel)
++			((struct bcm2711_dma40_scb *)
++			 d->cb_list[frame - 1].cb)->next_cb =
++				to_bcm2711_cbaddr(cb_entry->paddr);
++		if (frame && !c->is_40bit_channel)
+ 			d->cb_list[frame - 1].cb->next = cb_entry->paddr;
+ 
+ 		/* update src and dst and length */
+@@ -371,11 +543,21 @@ static struct bcm2835_desc *bcm2835_dma_
+ 			dst += control_block->length;
+ 
+ 		/* Length of total transfer */
+-		d->size += control_block->length;
++		if (c->is_40bit_channel)
++			d->size += ((struct bcm2711_dma40_scb *)control_block)->len;
++		else
++			d->size += control_block->length;
+ 	}
+ 
+ 	/* the last frame requires extra flags */
+-	d->cb_list[d->frames - 1].cb->info |= finalextrainfo;
++	if (c->is_40bit_channel) {
++		struct bcm2711_dma40_scb *scb =
++			(struct bcm2711_dma40_scb *)d->cb_list[d->frames-1].cb;
++
++		scb->ti |= to_bcm2711_ti(finalextrainfo);
++	} else {
++		d->cb_list[d->frames - 1].cb->info |= finalextrainfo;
++	}
+ 
+ 	/* detect a size missmatch */
+ 	if (buf_len && (d->size != buf_len))
+@@ -389,13 +571,12 @@ error_cb:
+ }
+ 
+ static void bcm2835_dma_fill_cb_chain_with_sg(
+-	struct dma_chan *chan,
++	struct bcm2835_chan *c,
+ 	enum dma_transfer_direction direction,
+ 	struct bcm2835_cb_entry *cb,
+ 	struct scatterlist *sgl,
+ 	unsigned int sg_len)
+ {
+-	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
+ 	size_t len, max_len;
+ 	unsigned int i;
+ 	dma_addr_t addr;
+@@ -403,14 +584,35 @@ static void bcm2835_dma_fill_cb_chain_wi
+ 
+ 	max_len = bcm2835_dma_max_frame_length(c);
+ 	for_each_sg(sgl, sgent, sg_len, i) {
+-		for (addr = sg_dma_address(sgent), len = sg_dma_len(sgent);
+-		     len > 0;
+-		     addr += cb->cb->length, len -= cb->cb->length, cb++) {
+-			if (direction == DMA_DEV_TO_MEM)
+-				cb->cb->dst = addr;
+-			else
+-				cb->cb->src = addr;
+-			cb->cb->length = min(len, max_len);
++		if (c->is_40bit_channel) {
++			struct bcm2711_dma40_scb *scb;
++
++			for (addr = sg_dma_address(sgent),
++				     len = sg_dma_len(sgent);
++				     len > 0;
++			     addr += scb->len, len -= scb->len, cb++) {
++				scb = (struct bcm2711_dma40_scb *)cb->cb;
++				if (direction == DMA_DEV_TO_MEM) {
++					scb->dst = lower_32_bits(addr);
++					scb->dsti = upper_32_bits(addr) | BCM2711_DMA40_INC;
++				} else {
++					scb->src = lower_32_bits(addr);
++					scb->srci = upper_32_bits(addr) | BCM2711_DMA40_INC;
++				}
++				scb->len = min(len, max_len);
++			}
++		} else {
++			for (addr = sg_dma_address(sgent),
++				     len = sg_dma_len(sgent);
++			     len > 0;
++			     addr += cb->cb->length, len -= cb->cb->length,
++			     cb++) {
++				if (direction == DMA_DEV_TO_MEM)
++					cb->cb->dst = addr;
++				else
++					cb->cb->src = addr;
++				cb->cb->length = min(len, max_len);
++			}
+ 		}
+ 	}
+ }
+@@ -419,6 +621,10 @@ static void bcm2835_dma_abort(struct bcm
+ {
+ 	void __iomem *chan_base = c->chan_base;
+ 	long int timeout = 10000;
++	u32 wait_mask = BCM2835_DMA_WAITING_FOR_WRITES;
++
++	if (c->is_40bit_channel)
++		wait_mask = BCM2711_DMA40_WAITING_FOR_WRITES;
+ 
+ 	/*
+ 	 * A zero control block address means the channel is idle.
+@@ -431,8 +637,7 @@ static void bcm2835_dma_abort(struct bcm
+ 	writel(0, chan_base + BCM2835_DMA_CS);
+ 
+ 	/* Wait for any current AXI transfer to complete */
+-	while ((readl(chan_base + BCM2835_DMA_CS) &
+-		BCM2835_DMA_WAITING_FOR_WRITES) && --timeout)
++	while ((readl(chan_base + BCM2835_DMA_CS) & wait_mask) && --timeout)
+ 		cpu_relax();
+ 
+ 	/* Peripheral might be stuck and fail to signal AXI write responses */
+@@ -457,9 +662,16 @@ static void bcm2835_dma_start_desc(struc
+ 
+ 	c->desc = d = to_bcm2835_dma_desc(&vd->tx);
+ 
+-	writel(d->cb_list[0].paddr, c->chan_base + BCM2835_DMA_ADDR);
+-	writel(BCM2835_DMA_ACTIVE | BCM2835_DMA_CS_FLAGS(c->dreq),
+-	       c->chan_base + BCM2835_DMA_CS);
++	if (c->is_40bit_channel) {
++		writel(to_bcm2711_cbaddr(d->cb_list[0].paddr),
++		       c->chan_base + BCM2711_DMA40_CB);
++		writel(BCM2711_DMA40_ACTIVE | BCM2711_DMA40_CS_FLAGS(c->dreq),
++		       c->chan_base + BCM2711_DMA40_CS);
++	} else {
++		writel(d->cb_list[0].paddr, c->chan_base + BCM2835_DMA_ADDR);
++		writel(BCM2835_DMA_ACTIVE | BCM2835_DMA_CS_FLAGS(c->dreq),
++		       c->chan_base + BCM2835_DMA_CS);
++	}
+ }
+ 
+ static irqreturn_t bcm2835_dma_callback(int irq, void *data)
+@@ -486,8 +698,7 @@ static irqreturn_t bcm2835_dma_callback(
+ 	 * if this IRQ handler is threaded.) If the channel is finished, it
+ 	 * will remain idle despite the ACTIVE flag being set.
+ 	 */
+-	writel(BCM2835_DMA_INT | BCM2835_DMA_ACTIVE |
+-	       BCM2835_DMA_CS_FLAGS(c->dreq),
++	writel(BCM2835_DMA_INT | BCM2835_DMA_ACTIVE,
+ 	       c->chan_base + BCM2835_DMA_CS);
+ 
+ 	d = c->desc;
+@@ -590,9 +801,17 @@ static enum dma_status bcm2835_dma_tx_st
+ 		struct bcm2835_desc *d = c->desc;
+ 		dma_addr_t pos;
+ 
+-		if (d->dir == DMA_MEM_TO_DEV)
++		if (d->dir == DMA_MEM_TO_DEV && c->is_40bit_channel)
++			pos = readl(c->chan_base + BCM2711_DMA40_SRC) +
++				((readl(c->chan_base + BCM2711_DMA40_SRCI) &
++				  0xff) << 8);
++		else if (d->dir == DMA_MEM_TO_DEV && !c->is_40bit_channel)
+ 			pos = readl(c->chan_base + BCM2835_DMA_SOURCE_AD);
+-		else if (d->dir == DMA_DEV_TO_MEM)
++		else if (d->dir == DMA_DEV_TO_MEM && c->is_40bit_channel)
++			pos = readl(c->chan_base + BCM2711_DMA40_DEST) +
++				((readl(c->chan_base + BCM2711_DMA40_DESTI) &
++				  0xff) << 8);
++		else if (d->dir == DMA_DEV_TO_MEM && !c->is_40bit_channel)
+ 			pos = readl(c->chan_base + BCM2835_DMA_DEST_AD);
+ 		else
+ 			pos = 0;
+@@ -638,7 +857,7 @@ static struct dma_async_tx_descriptor *b
+ 	frames = bcm2835_dma_frames_for_length(len, max_len);
+ 
+ 	/* allocate the CB chain - this also fills in the pointers */
+-	d = bcm2835_dma_create_cb_chain(chan, DMA_MEM_TO_MEM, false,
++	d = bcm2835_dma_create_cb_chain(c, DMA_MEM_TO_MEM, false,
+ 					info, extra, frames,
+ 					src, dst, len, 0, GFP_KERNEL);
+ 	if (!d)
+@@ -673,11 +892,21 @@ static struct dma_async_tx_descriptor *b
+ 		if (c->cfg.src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES)
+ 			return NULL;
+ 		src = c->cfg.src_addr;
++		/*
++		 * One would think it ought to be possible to get the physical
++		 * to dma address mapping information from the dma-ranges DT
++		 * property, but I've not found a way yet that doesn't involve
++		 * open-coding the whole thing.
++		 */
++		if (c->is_40bit_channel)
++		    src |= 0x400000000ull;
+ 		info |= BCM2835_DMA_S_DREQ | BCM2835_DMA_D_INC;
+ 	} else {
+ 		if (c->cfg.dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES)
+ 			return NULL;
+ 		dst = c->cfg.dst_addr;
++		if (c->is_40bit_channel)
++		    dst |= 0x400000000ull;
+ 		info |= BCM2835_DMA_D_DREQ | BCM2835_DMA_S_INC;
+ 	}
+ 
+@@ -685,7 +914,7 @@ static struct dma_async_tx_descriptor *b
+ 	frames = bcm2835_dma_count_frames_for_sg(c, sgl, sg_len);
+ 
+ 	/* allocate the CB chain */
+-	d = bcm2835_dma_create_cb_chain(chan, direction, false,
++	d = bcm2835_dma_create_cb_chain(c, direction, false,
+ 					info, extra,
+ 					frames, src, dst, 0, 0,
+ 					GFP_NOWAIT);
+@@ -693,7 +922,7 @@ static struct dma_async_tx_descriptor *b
+ 		return NULL;
+ 
+ 	/* fill in frames with scatterlist pointers */
+-	bcm2835_dma_fill_cb_chain_with_sg(chan, direction, d->cb_list,
++	bcm2835_dma_fill_cb_chain_with_sg(c, direction, d->cb_list,
+ 					  sgl, sg_len);
+ 
+ 	return vchan_tx_prep(&c->vc, &d->vd, flags);
+@@ -747,12 +976,16 @@ static struct dma_async_tx_descriptor *b
+ 		if (c->cfg.src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES)
+ 			return NULL;
+ 		src = c->cfg.src_addr;
++		if (c->is_40bit_channel)
++		    src |= 0x400000000ull;
+ 		dst = buf_addr;
+ 		info |= BCM2835_DMA_S_DREQ | BCM2835_DMA_D_INC;
+ 	} else {
+ 		if (c->cfg.dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES)
+ 			return NULL;
+ 		dst = c->cfg.dst_addr;
++		if (c->is_40bit_channel)
++		    dst |= 0x400000000ull;
+ 		src = buf_addr;
+ 		info |= BCM2835_DMA_D_DREQ | BCM2835_DMA_S_INC;
+ 
+@@ -772,7 +1005,7 @@ static struct dma_async_tx_descriptor *b
+ 	 * note that we need to use GFP_NOWAIT, as the ALSA i2s dmaengine
+ 	 * implementation calls prep_dma_cyclic with interrupts disabled.
+ 	 */
+-	d = bcm2835_dma_create_cb_chain(chan, direction, true,
++	d = bcm2835_dma_create_cb_chain(c, direction, true,
+ 					info, extra,
+ 					frames, src, dst, buf_len,
+ 					period_len, GFP_NOWAIT);
+@@ -780,7 +1013,12 @@ static struct dma_async_tx_descriptor *b
+ 		return NULL;
+ 
+ 	/* wrap around into a loop */
+-	d->cb_list[d->frames - 1].cb->next = d->cb_list[0].paddr;
++	if (c->is_40bit_channel)
++		((struct bcm2711_dma40_scb *)
++		 d->cb_list[frames - 1].cb)->next_cb =
++			to_bcm2711_cbaddr(d->cb_list[0].paddr);
++	else
++		d->cb_list[d->frames - 1].cb->next = d->cb_list[0].paddr;
+ 
+ 	return vchan_tx_prep(&c->vc, &d->vd, flags);
+ }
+@@ -844,9 +1082,11 @@ static int bcm2835_dma_chan_init(struct
+ 	c->irq_number = irq;
+ 	c->irq_flags = irq_flags;
+ 
+-	/* check in DEBUG register if this is a LITE channel */
+-	if (readl(c->chan_base + BCM2835_DMA_DEBUG) &
+-		BCM2835_DMA_DEBUG_LITE)
++	/* check for 40bit and lite channels */
++	if (d->cfg_data->chan_40bit_mask & BIT(chan_id))
++		c->is_40bit_channel = true;
++	else if (readl(c->chan_base + BCM2835_DMA_DEBUG) &
++		 BCM2835_DMA_DEBUG_LITE)
+ 		c->is_lite_channel = true;
+ 
+ 	return 0;
+@@ -866,8 +1106,58 @@ static void bcm2835_dma_free(struct bcm2
+ 			     DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+ }
+ 
++int bcm2711_dma40_memcpy_init(void)
++{
++	if (!memcpy_parent)
++		return -EPROBE_DEFER;
++
++	if (!memcpy_chan)
++		return -EINVAL;
++
++	if (!memcpy_scb)
++		return -ENOMEM;
++
++	return 0;
++}
++EXPORT_SYMBOL(bcm2711_dma40_memcpy_init);
++
++void bcm2711_dma40_memcpy(dma_addr_t dst, dma_addr_t src, size_t size)
++{
++	struct bcm2711_dma40_scb *scb = memcpy_scb;
++	unsigned long flags;
++
++	if (!scb) {
++		pr_err("bcm2711_dma40_memcpy not initialised!\n");
++		return;
++	}
++
++	spin_lock_irqsave(&memcpy_lock, flags);
++
++	scb->ti = 0;
++	scb->src = lower_32_bits(src);
++	scb->srci = upper_32_bits(src) | BCM2711_DMA40_MEMCPY_XFER_INFO;
++	scb->dst = lower_32_bits(dst);
++	scb->dsti = upper_32_bits(dst) | BCM2711_DMA40_MEMCPY_XFER_INFO;
++	scb->len = size;
++	scb->next_cb = 0;
++
++	writel((u32)(memcpy_scb_dma >> 5), memcpy_chan + BCM2711_DMA40_CB);
++	writel(BCM2711_DMA40_MEMCPY_FLAGS + BCM2711_DMA40_ACTIVE,
++	       memcpy_chan + BCM2711_DMA40_CS);
++
++	/* Poll for completion */
++	while (!(readl(memcpy_chan + BCM2711_DMA40_CS) & BCM2711_DMA40_END))
++		cpu_relax();
++
++	writel(BCM2711_DMA40_END, memcpy_chan + BCM2711_DMA40_CS);
++
++	spin_unlock_irqrestore(&memcpy_lock, flags);
++}
++EXPORT_SYMBOL(bcm2711_dma40_memcpy);
++
+ static const struct of_device_id bcm2835_dma_of_match[] = {
+-	{ .compatible = "brcm,bcm2835-dma", },
++	{ .compatible = "brcm,bcm2835-dma", .data = &bcm2835_dma_cfg },
++	{ .compatible = "brcm,bcm2711-dma", .data = &bcm2711_dma_cfg },
+ 	{},
+ };
+ MODULE_DEVICE_TABLE(of, bcm2835_dma_of_match);
+@@ -899,6 +1189,8 @@ static int bcm2835_dma_probe(struct plat
+ 	int irq_flags;
+ 	uint32_t chans_available;
+ 	char chan_name[BCM2835_DMA_CHAN_NAME_SIZE];
++	const struct of_device_id *of_id;
++	int chan_count, chan_start, chan_end;
+ 
+ 	if (!pdev->dev.dma_mask)
+ 		pdev->dev.dma_mask = &pdev->dev.coherent_dma_mask;
+@@ -920,9 +1212,13 @@ static int bcm2835_dma_probe(struct plat
+ 	base = devm_ioremap_resource(&pdev->dev, res);
+ 	if (IS_ERR(base))
+ 		return PTR_ERR(base);
+-	rc = bcm_dmaman_probe(pdev, base, BCM2835_DMA_BULK_MASK);
+-	if (rc)
+-		dev_err(&pdev->dev, "Failed to initialize the legacy API\n");
++
++	/* The set of channels can be split across multiple instances. */
++	chan_start = ((u32)(uintptr_t)base / BCM2835_DMA_CHAN_SIZE) & 0xf;
++	base -= BCM2835_DMA_CHAN(chan_start);
++	chan_count = resource_size(res) / BCM2835_DMA_CHAN_SIZE;
++	chan_end = min(chan_start + chan_count,
++			 BCM2835_DMA_MAX_DMA_CHAN_SUPPORTED + 1);
+ 
+ 	od->base = base;
+ 
+@@ -959,6 +1255,14 @@ static int bcm2835_dma_probe(struct plat
+ 		return -ENOMEM;
+ 	}
+ 
++	of_id = of_match_node(bcm2835_dma_of_match, pdev->dev.of_node);
++	if (!of_id) {
++		dev_err(&pdev->dev, "Failed to match compatible string\n");
++		return -EINVAL;
++	}
++
++	od->cfg_data = of_id->data;
++
+ 	/* Request DMA channel mask from device tree */
+ 	if (of_property_read_u32(pdev->dev.of_node,
+ 			"brcm,dma-channel-mask",
+@@ -968,11 +1272,34 @@ static int bcm2835_dma_probe(struct plat
+ 		goto err_no_dma;
+ 	}
+ 
+-	/* Channel 0 is used by the legacy API */
+-	chans_available &= ~BCM2835_DMA_BULK_MASK;
++	/* One channel is reserved for the legacy API */
++	if (chans_available & BCM2835_DMA_BULK_MASK) {
++		rc = bcm_dmaman_probe(pdev, base,
++				      chans_available & BCM2835_DMA_BULK_MASK);
++		if (rc)
++			dev_err(&pdev->dev,
++				"Failed to initialize the legacy API\n");
++
++		chans_available &= ~BCM2835_DMA_BULK_MASK;
++	}
++
++	/* And possibly one for the 40-bit DMA memcpy API */
++	if (chans_available & od->cfg_data->chan_40bit_mask &
++	    BIT(BCM2711_DMA_MEMCPY_CHAN)) {
++		memcpy_parent = od;
++		memcpy_chan = BCM2835_DMA_CHANIO(base, BCM2711_DMA_MEMCPY_CHAN);
++		memcpy_scb = dma_alloc_coherent(memcpy_parent->ddev.dev,
++						sizeof(*memcpy_scb),
++						&memcpy_scb_dma, GFP_KERNEL);
++		if (!memcpy_scb)
++			dev_warn(&pdev->dev,
++				 "Failed to allocated memcpy scb\n");
++
++		chans_available &= ~BIT(BCM2711_DMA_MEMCPY_CHAN);
++	}
+ 
+ 	/* get irqs for each channel that we support */
+-	for (i = 0; i <= BCM2835_DMA_MAX_DMA_CHAN_SUPPORTED; i++) {
++	for (i = chan_start; i < chan_end; i++) {
+ 		/* skip masked out channels */
+ 		if (!(chans_available & (1 << i))) {
+ 			irq[i] = -1;
+@@ -995,13 +1322,17 @@ static int bcm2835_dma_probe(struct plat
+ 		irq[i] = platform_get_irq(pdev, i < 11 ? i : 11);
+ 	}
+ 
++	chan_count = 0;
++
+ 	/* get irqs for each channel */
+-	for (i = 0; i <= BCM2835_DMA_MAX_DMA_CHAN_SUPPORTED; i++) {
++	for (i = chan_start; i < chan_end; i++) {
+ 		/* skip channels without irq */
+ 		if (irq[i] < 0)
+ 			continue;
+ 
+ 		/* check if there are other channels that also use this irq */
++		/* FIXME: This will fail if interrupts are shared across
++		   instances */
+ 		irq_flags = 0;
+ 		for (j = 0; j <= BCM2835_DMA_MAX_DMA_CHAN_SUPPORTED; j++)
+ 			if ((i != j) && (irq[j] == irq[i])) {
+@@ -1013,9 +1344,10 @@ static int bcm2835_dma_probe(struct plat
+ 		rc = bcm2835_dma_chan_init(od, i, irq[i], irq_flags);
+ 		if (rc)
+ 			goto err_no_dma;
++		chan_count++;
+ 	}
+ 
+-	dev_dbg(&pdev->dev, "Initialized %i DMA channels\n", i);
++	dev_dbg(&pdev->dev, "Initialized %i DMA channels\n", chan_count);
+ 
+ 	/* Device-tree DMA controller registration */
+ 	rc = of_dma_controller_register(pdev->dev.of_node,
+@@ -1047,6 +1379,13 @@ static int bcm2835_dma_remove(struct pla
+ 
+ 	bcm_dmaman_remove(pdev);
+ 	dma_async_device_unregister(&od->ddev);
++	if (memcpy_parent == od) {
++		dma_free_coherent(&pdev->dev, sizeof(*memcpy_scb), memcpy_scb,
++				  memcpy_scb_dma);
++		memcpy_parent = NULL;
++		memcpy_scb = NULL;
++		memcpy_chan = NULL;
++	}
+ 	bcm2835_dma_free(od);
+ 
+ 	return 0;