Merging upstream version 6.9.7.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

36 files changed, 933 insertions, 358 deletions

diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c
index 3564a0f63c..90c51b1214 100644
--- a/drivers/mmc/core/block.c
+++ b/drivers/mmc/core/block.c
@@ -144,7 +144,7 @@ struct mmc_blk_data {
 static dev_t mmc_rpmb_devt;
 
 /* Bus type for RPMB character devices */
-static struct bus_type mmc_rpmb_bus_type = {
+static const struct bus_type mmc_rpmb_bus_type = {
 	.name = "mmc_rpmb",
 };
 
@@ -206,7 +206,7 @@ static void mmc_blk_kref_release(struct kref *ref)
 	int devidx;
 
 	devidx = mmc_get_devidx(md->disk);
-	ida_simple_remove(&mmc_blk_ida, devidx);
+	ida_free(&mmc_blk_ida, devidx);
 
 	mutex_lock(&open_lock);
 	md->disk->private_data = NULL;
@@ -2469,7 +2469,7 @@ static struct mmc_blk_data *mmc_blk_alloc_req(struct mmc_card *card,
 	bool cache_enabled = false;
 	bool fua_enabled = false;
 
-	devidx = ida_simple_get(&mmc_blk_ida, 0, max_devices, GFP_KERNEL);
+	devidx = ida_alloc_max(&mmc_blk_ida, max_devices - 1, GFP_KERNEL);
 	if (devidx < 0) {
 		/*
 		 * We get -ENOSPC because there are no more any available
@@ -2579,7 +2579,7 @@ static struct mmc_blk_data *mmc_blk_alloc_req(struct mmc_card *card,
  err_kfree:
 	kfree(md);
  out:
-	ida_simple_remove(&mmc_blk_ida, devidx);
+	ida_free(&mmc_blk_ida, devidx);
 	return ERR_PTR(ret);
 }
 
@@ -2705,7 +2705,7 @@ static void mmc_blk_rpmb_device_release(struct device *dev)
 {
 	struct mmc_rpmb_data *rpmb = dev_get_drvdata(dev);
 
-	ida_simple_remove(&mmc_rpmb_ida, rpmb->id);
+	ida_free(&mmc_rpmb_ida, rpmb->id);
 	kfree(rpmb);
 }
 
@@ -2721,13 +2721,13 @@ static int mmc_blk_alloc_rpmb_part(struct mmc_card *card,
 	struct mmc_rpmb_data *rpmb;
 
 	/* This creates the minor number for the RPMB char device */
-	devidx = ida_simple_get(&mmc_rpmb_ida, 0, max_devices, GFP_KERNEL);
+	devidx = ida_alloc_max(&mmc_rpmb_ida, max_devices - 1, GFP_KERNEL);
 	if (devidx < 0)
 		return devidx;
 
 	rpmb = kzalloc(sizeof(*rpmb), GFP_KERNEL);
 	if (!rpmb) {
-		ida_simple_remove(&mmc_rpmb_ida, devidx);
+		ida_free(&mmc_rpmb_ida, devidx);
 		return -ENOMEM;
 	}
 
diff --git a/drivers/mmc/core/bus.c b/drivers/mmc/core/bus.c
index 0af96548e7..0ddaee0eae 100644
--- a/drivers/mmc/core/bus.c
+++ b/drivers/mmc/core/bus.c
@@ -214,7 +214,7 @@ static const struct dev_pm_ops mmc_bus_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(mmc_bus_suspend, mmc_bus_resume)
 };
 
-static struct bus_type mmc_bus_type = {
+static const struct bus_type mmc_bus_type = {
 	.name		= "mmc",
 	.dev_groups	= mmc_dev_groups,
 	.uevent		= mmc_bus_uevent,
@@ -272,7 +272,7 @@ static void mmc_release_card(struct device *dev)
 /*
  * Allocate and initialise a new MMC card structure.
  */
-struct mmc_card *mmc_alloc_card(struct mmc_host *host, struct device_type *type)
+struct mmc_card *mmc_alloc_card(struct mmc_host *host, const struct device_type *type)
 {
 	struct mmc_card *card;
 
diff --git a/drivers/mmc/core/bus.h b/drivers/mmc/core/bus.h
index 3996b191b6..cfd0d02d34 100644
--- a/drivers/mmc/core/bus.h
+++ b/drivers/mmc/core/bus.h
@@ -23,7 +23,7 @@ static ssize_t mmc_##name##_show (struct device *dev, struct device_attribute *a
 static DEVICE_ATTR(name, S_IRUGO, mmc_##name##_show, NULL)
 
 struct mmc_card *mmc_alloc_card(struct mmc_host *host,
-	struct device_type *type);
+				const struct device_type *type);
 int mmc_add_card(struct mmc_card *card);
 void mmc_remove_card(struct mmc_card *card);
 
diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
index cf396e8f34..8f8781d6c2 100644
--- a/drivers/mmc/core/host.c
+++ b/drivers/mmc/core/host.c
@@ -76,7 +76,7 @@ static void mmc_host_classdev_release(struct device *dev)
 	struct mmc_host *host = cls_dev_to_mmc_host(dev);
 	wakeup_source_unregister(host->ws);
 	if (of_alias_get_id(host->parent->of_node, "mmc") < 0)
-		ida_simple_remove(&mmc_host_ida, host->index);
+		ida_free(&mmc_host_ida, host->index);
 	kfree(host);
 }
 
@@ -88,7 +88,7 @@ static int mmc_host_classdev_shutdown(struct device *dev)
 	return 0;
 }
 
-static struct class mmc_host_class = {
+static const struct class mmc_host_class = {
 	.name		= "mmc_host",
 	.dev_release	= mmc_host_classdev_release,
 	.shutdown_pre	= mmc_host_classdev_shutdown,
@@ -234,10 +234,8 @@ static void mmc_of_parse_timing_phase(struct device *dev, const char *prop,
 }
 
 void
-mmc_of_parse_clk_phase(struct mmc_host *host, struct mmc_clk_phase_map *map)
+mmc_of_parse_clk_phase(struct device *dev, struct mmc_clk_phase_map *map)
 {
-	struct device *dev = host->parent;
-
 	mmc_of_parse_timing_phase(dev, "clk-phase-legacy",
 				  &map->phase[MMC_TIMING_LEGACY]);
 	mmc_of_parse_timing_phase(dev, "clk-phase-mmc-hs",
@@ -538,7 +536,8 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
 		min_idx = mmc_first_nonreserved_index();
 		max_idx = 0;
 
-		index = ida_simple_get(&mmc_host_ida, min_idx, max_idx, GFP_KERNEL);
+		index = ida_alloc_range(&mmc_host_ida, min_idx, max_idx - 1,
+					GFP_KERNEL);
 		if (index < 0) {
 			kfree(host);
 			return NULL;
diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c
index 58ed7193a3..5b2f7c2854 100644
--- a/drivers/mmc/core/mmc.c
+++ b/drivers/mmc/core/mmc.c
@@ -883,7 +883,7 @@ static struct attribute *mmc_std_attrs[] = {
 };
 ATTRIBUTE_GROUPS(mmc_std);
 
-static struct device_type mmc_type = {
+static const struct device_type mmc_type = {
 	.groups = mmc_std_groups,
 };
 
diff --git a/drivers/mmc/core/queue.c b/drivers/mmc/core/queue.c
index a0a2412f62..241cdc2b2a 100644
--- a/drivers/mmc/core/queue.c
+++ b/drivers/mmc/core/queue.c
@@ -174,8 +174,8 @@ static struct scatterlist *mmc_alloc_sg(unsigned short sg_len, gfp_t gfp)
 	return sg;
 }
 
-static void mmc_queue_setup_discard(struct request_queue *q,
-				    struct mmc_card *card)
+static void mmc_queue_setup_discard(struct mmc_card *card,
+		struct queue_limits *lim)
 {
 	unsigned max_discard;
 
@@ -183,15 +183,17 @@ static void mmc_queue_setup_discard(struct request_queue *q,
 	if (!max_discard)
 		return;
 
-	blk_queue_max_discard_sectors(q, max_discard);
-	q->limits.discard_granularity = card->pref_erase << 9;
-	/* granularity must not be greater than max. discard */
-	if (card->pref_erase > max_discard)
-		q->limits.discard_granularity = SECTOR_SIZE;
+	lim->max_hw_discard_sectors = max_discard;
 	if (mmc_can_secure_erase_trim(card))
-		blk_queue_max_secure_erase_sectors(q, max_discard);
+		lim->max_secure_erase_sectors = max_discard;
 	if (mmc_can_trim(card) && card->erased_byte == 0)
-		blk_queue_max_write_zeroes_sectors(q, max_discard);
+		lim->max_write_zeroes_sectors = max_discard;
+
+	/* granularity must not be greater than max. discard */
+	if (card->pref_erase > max_discard)
+		lim->discard_granularity = SECTOR_SIZE;
+	else
+		lim->discard_granularity = card->pref_erase << 9;
 }
 
 static unsigned short mmc_get_max_segments(struct mmc_host *host)
@@ -341,40 +343,50 @@ static const struct blk_mq_ops mmc_mq_ops = {
 	.timeout	= mmc_mq_timed_out,
 };
 
-static void mmc_setup_queue(struct mmc_queue *mq, struct mmc_card *card)
+static struct gendisk *mmc_alloc_disk(struct mmc_queue *mq,
+		struct mmc_card *card)
 {
 	struct mmc_host *host = card->host;
-	unsigned block_size = 512;
+	struct queue_limits lim = { };
+	struct gendisk *disk;
 
-	blk_queue_flag_set(QUEUE_FLAG_NONROT, mq->queue);
-	blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, mq->queue);
 	if (mmc_can_erase(card))
-		mmc_queue_setup_discard(mq->queue, card);
-
-	if (!mmc_dev(host)->dma_mask || !*mmc_dev(host)->dma_mask)
-		blk_queue_bounce_limit(mq->queue, BLK_BOUNCE_HIGH);
-	blk_queue_max_hw_sectors(mq->queue,
-		min(host->max_blk_count, host->max_req_size / 512));
-	if (host->can_dma_map_merge)
-		WARN(!blk_queue_can_use_dma_map_merging(mq->queue,
-							mmc_dev(host)),
-		     "merging was advertised but not possible");
-	blk_queue_max_segments(mq->queue, mmc_get_max_segments(host));
-
-	if (mmc_card_mmc(card) && card->ext_csd.data_sector_size) {
-		block_size = card->ext_csd.data_sector_size;
-		WARN_ON(block_size != 512 && block_size != 4096);
-	}
+		mmc_queue_setup_discard(card, &lim);
+
+	lim.max_hw_sectors = min(host->max_blk_count, host->max_req_size / 512);
+
+	if (mmc_card_mmc(card) && card->ext_csd.data_sector_size)
+		lim.logical_block_size = card->ext_csd.data_sector_size;
+	else
+		lim.logical_block_size = 512;
+
+	WARN_ON_ONCE(lim.logical_block_size != 512 &&
+		     lim.logical_block_size != 4096);
 
-	blk_queue_logical_block_size(mq->queue, block_size);
 	/*
-	 * After blk_queue_can_use_dma_map_merging() was called with succeed,
-	 * since it calls blk_queue_virt_boundary(), the mmc should not call
-	 * both blk_queue_max_segment_size().
+	 * Setting a virt_boundary implicity sets a max_segment_size, so try
+	 * to set the hardware one here.
 	 */
-	if (!host->can_dma_map_merge)
-		blk_queue_max_segment_size(mq->queue,
-			round_down(host->max_seg_size, block_size));
+	if (host->can_dma_map_merge) {
+		lim.virt_boundary_mask = dma_get_merge_boundary(mmc_dev(host));
+		lim.max_segments = MMC_DMA_MAP_MERGE_SEGMENTS;
+	} else {
+		lim.max_segment_size =
+			round_down(host->max_seg_size, lim.logical_block_size);
+		lim.max_segments = host->max_segs;
+	}
+
+	disk = blk_mq_alloc_disk(&mq->tag_set, &lim, mq);
+	if (IS_ERR(disk))
+		return disk;
+	mq->queue = disk->queue;
+
+	if (mmc_host_is_spi(host) && host->use_spi_crc)
+		blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, mq->queue);
+	blk_queue_rq_timeout(mq->queue, 60 * HZ);
+
+	blk_queue_flag_set(QUEUE_FLAG_NONROT, mq->queue);
+	blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, mq->queue);
 
 	dma_set_max_seg_size(mmc_dev(host), queue_max_segment_size(mq->queue));
 
@@ -386,6 +398,7 @@ static void mmc_setup_queue(struct mmc_queue *mq, struct mmc_card *card)
 	init_waitqueue_head(&mq->wait);
 
 	mmc_crypto_setup_queue(mq->queue, host);
+	return disk;
 }
 
 static inline bool mmc_merge_capable(struct mmc_host *host)
@@ -447,18 +460,9 @@ struct gendisk *mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card)
 		return ERR_PTR(ret);
 		
 
-	disk = blk_mq_alloc_disk(&mq->tag_set, mq);
-	if (IS_ERR(disk)) {
+	disk = mmc_alloc_disk(mq, card);
+	if (IS_ERR(disk))
 		blk_mq_free_tag_set(&mq->tag_set);
-		return disk;
-	}
-	mq->queue = disk->queue;
-
-	if (mmc_host_is_spi(host) && host->use_spi_crc)
-		blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, mq->queue);
-	blk_queue_rq_timeout(mq->queue, 60 * HZ);
-
-	mmc_setup_queue(mq, card);
 	return disk;
 }
 
diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
index c3e554344c..1c8148cdda 100644
--- a/drivers/mmc/core/sd.c
+++ b/drivers/mmc/core/sd.c
@@ -805,7 +805,7 @@ static const struct attribute_group sd_std_group = {
 };
 __ATTRIBUTE_GROUPS(sd_std);
 
-struct device_type sd_type = {
+const struct device_type sd_type = {
 	.groups = sd_std_groups,
 };
 
diff --git a/drivers/mmc/core/sd.h b/drivers/mmc/core/sd.h
index 1af5a038ba..fe6dd46927 100644
--- a/drivers/mmc/core/sd.h
+++ b/drivers/mmc/core/sd.h
@@ -4,7 +4,7 @@
 
 #include <linux/types.h>
 
-extern struct device_type sd_type;
+extern const struct device_type sd_type;
 
 struct mmc_host;
 struct mmc_card;
diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c
index 5914516df2..4fb247fde5 100644
--- a/drivers/mmc/core/sdio.c
+++ b/drivers/mmc/core/sdio.c
@@ -66,7 +66,7 @@ static struct attribute *sdio_std_attrs[] = {
 };
 ATTRIBUTE_GROUPS(sdio_std);
 
-static struct device_type sdio_type = {
+static const struct device_type sdio_type = {
 	.groups = sdio_std_groups,
 };
 
diff --git a/drivers/mmc/core/sdio_bus.c b/drivers/mmc/core/sdio_bus.c
index 47a48e902a..71d885fbc2 100644
--- a/drivers/mmc/core/sdio_bus.c
+++ b/drivers/mmc/core/sdio_bus.c
@@ -244,7 +244,7 @@ static const struct dev_pm_ops sdio_bus_pm_ops = {
 	)
 };
 
-static struct bus_type sdio_bus_type = {
+static const struct bus_type sdio_bus_type = {
 	.name		= "sdio",
 	.dev_groups	= sdio_dev_groups,
 	.match		= sdio_bus_match,
diff --git a/drivers/mmc/core/slot-gpio.c b/drivers/mmc/core/slot-gpio.c
index 39f45c2b6d..8791656e9e 100644
--- a/drivers/mmc/core/slot-gpio.c
+++ b/drivers/mmc/core/slot-gpio.c
@@ -221,6 +221,26 @@ int mmc_gpiod_request_cd(struct mmc_host *host, const char *con_id,
 }
 EXPORT_SYMBOL(mmc_gpiod_request_cd);
 
+/**
+ * mmc_gpiod_set_cd_config - set config for card-detection GPIO
+ * @host: mmc host
+ * @config: Generic pinconf config (from pinconf_to_config_packed())
+ *
+ * This can be used by mmc host drivers to fixup a card-detection GPIO's config
+ * (e.g. set PIN_CONFIG_BIAS_PULL_UP) after acquiring the GPIO descriptor
+ * through mmc_gpiod_request_cd().
+ *
+ * Returns:
+ * 0 on success, or a negative errno value on error.
+ */
+int mmc_gpiod_set_cd_config(struct mmc_host *host, unsigned long config)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+
+	return gpiod_set_config(ctx->cd_gpio, config);
+}
+EXPORT_SYMBOL(mmc_gpiod_set_cd_config);
+
 bool mmc_can_gpio_cd(struct mmc_host *host)
 {
 	struct mmc_gpio *ctx = host->slot.handler_priv;
diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index 81f2c4e052..aebc587f77 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -798,6 +798,15 @@ config MMC_DW_HI3798CV200
 	  Synopsys DesignWare Memory Card Interface driver. Select this option
 	  for platforms based on HiSilicon Hi3798CV200 SoC.
 
+config MMC_DW_HI3798MV200
+	tristate "Hi3798MV200 specific extensions for Synopsys DW Memory Card Interface"
+	depends on MMC_DW
+	select MMC_DW_PLTFM
+	help
+	  This selects support for HiSilicon Hi3798MV200 SoC specific extensions to the
+	  Synopsys DesignWare Memory Card Interface driver. Select this option
+	  for platforms based on HiSilicon Hi3798MV200 SoC.
+
 config MMC_DW_K3
 	tristate "K3 specific extensions for Synopsys DW Memory Card Interface"
 	depends on MMC_DW
diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
index d0be4465f3..f53f86d200 100644
--- a/drivers/mmc/host/Makefile
+++ b/drivers/mmc/host/Makefile
@@ -51,6 +51,7 @@ obj-$(CONFIG_MMC_DW_PLTFM)	+= dw_mmc-pltfm.o
 obj-$(CONFIG_MMC_DW_BLUEFIELD)	+= dw_mmc-bluefield.o
 obj-$(CONFIG_MMC_DW_EXYNOS)	+= dw_mmc-exynos.o
 obj-$(CONFIG_MMC_DW_HI3798CV200) += dw_mmc-hi3798cv200.o
+obj-$(CONFIG_MMC_DW_HI3798MV200) += dw_mmc-hi3798mv200.o
 obj-$(CONFIG_MMC_DW_K3)		+= dw_mmc-k3.o
 obj-$(CONFIG_MMC_DW_PCI)	+= dw_mmc-pci.o
 obj-$(CONFIG_MMC_DW_ROCKCHIP)	+= dw_mmc-rockchip.o
diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index ee3b1a4e08..d7427894e0 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -180,12 +180,6 @@ struct mmc_davinci_host {
 #define DAVINCI_MMC_DATADIR_WRITE	2
 	unsigned char data_dir;
 
-	/* buffer is used during PIO of one scatterlist segment, and
-	 * is updated along with buffer_bytes_left.  bytes_left applies
-	 * to all N blocks of the PIO transfer.
-	 */
-	u8 *buffer;
-	u32 buffer_bytes_left;
 	u32 bytes_left;
 
 	struct dma_chan *dma_tx;
@@ -196,8 +190,8 @@ struct mmc_davinci_host {
 	bool active_request;
 
 	/* For PIO we walk scatterlists one segment at a time. */
+	struct sg_mapping_iter sg_miter;
 	unsigned int		sg_len;
-	struct scatterlist *sg;
 
 	/* Version of the MMC/SD controller */
 	u8 version;
@@ -213,30 +207,22 @@ struct mmc_davinci_host {
 static irqreturn_t mmc_davinci_irq(int irq, void *dev_id);
 
 /* PIO only */
-static void mmc_davinci_sg_to_buf(struct mmc_davinci_host *host)
-{
-	host->buffer_bytes_left = sg_dma_len(host->sg);
-	host->buffer = sg_virt(host->sg);
-	if (host->buffer_bytes_left > host->bytes_left)
-		host->buffer_bytes_left = host->bytes_left;
-}
-
 static void davinci_fifo_data_trans(struct mmc_davinci_host *host,
 					unsigned int n)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	u8 *p;
 	unsigned int i;
 
-	if (host->buffer_bytes_left == 0) {
-		host->sg = sg_next(host->data->sg);
-		mmc_davinci_sg_to_buf(host);
+	/*
+	 * By adjusting sgm->consumed this will give a pointer to the
+	 * current index into the sgm.
+	 */
+	if (!sg_miter_next(sgm)) {
+		dev_err(mmc_dev(host->mmc), "ran out of sglist prematurely\n");
+		return;
 	}
-
-	p = host->buffer;
-	if (n > host->buffer_bytes_left)
-		n = host->buffer_bytes_left;
-	host->buffer_bytes_left -= n;
-	host->bytes_left -= n;
+	p = sgm->addr;
 
 	/* NOTE:  we never transfer more than rw_threshold bytes
 	 * to/from the fifo here; there's no I/O overlap.
@@ -261,7 +247,9 @@ static void davinci_fifo_data_trans(struct mmc_davinci_host *host,
 			p = p + (n & 3);
 		}
 	}
-	host->buffer = p;
+
+	sgm->consumed = n;
+	host->bytes_left -= n;
 }
 
 static void mmc_davinci_start_command(struct mmc_davinci_host *host,
@@ -517,6 +505,7 @@ mmc_davinci_prepare_data(struct mmc_davinci_host *host, struct mmc_request *req)
 	int fifo_lev = (rw_threshold == 32) ? MMCFIFOCTL_FIFOLEV : 0;
 	int timeout;
 	struct mmc_data *data = req->data;
+	unsigned int flags = SG_MITER_ATOMIC; /* Used from IRQ */
 
 	if (host->version == MMC_CTLR_VERSION_2)
 		fifo_lev = (rw_threshold == 64) ? MMCFIFOCTL_FIFOLEV : 0;
@@ -545,12 +534,14 @@ mmc_davinci_prepare_data(struct mmc_davinci_host *host, struct mmc_request *req)
 
 	/* Configure the FIFO */
 	if (data->flags & MMC_DATA_WRITE) {
+		flags |= SG_MITER_FROM_SG;
 		host->data_dir = DAVINCI_MMC_DATADIR_WRITE;
 		writel(fifo_lev | MMCFIFOCTL_FIFODIR_WR | MMCFIFOCTL_FIFORST,
 			host->base + DAVINCI_MMCFIFOCTL);
 		writel(fifo_lev | MMCFIFOCTL_FIFODIR_WR,
 			host->base + DAVINCI_MMCFIFOCTL);
 	} else {
+		flags |= SG_MITER_TO_SG;
 		host->data_dir = DAVINCI_MMC_DATADIR_READ;
 		writel(fifo_lev | MMCFIFOCTL_FIFODIR_RD | MMCFIFOCTL_FIFORST,
 			host->base + DAVINCI_MMCFIFOCTL);
@@ -558,7 +549,6 @@ mmc_davinci_prepare_data(struct mmc_davinci_host *host, struct mmc_request *req)
 			host->base + DAVINCI_MMCFIFOCTL);
 	}
 
-	host->buffer = NULL;
 	host->bytes_left = data->blocks * data->blksz;
 
 	/* For now we try to use DMA whenever we won't need partial FIFO
@@ -576,8 +566,7 @@ mmc_davinci_prepare_data(struct mmc_davinci_host *host, struct mmc_request *req)
 	} else {
 		/* Revert to CPU Copy */
 		host->sg_len = data->sg_len;
-		host->sg = host->data->sg;
-		mmc_davinci_sg_to_buf(host);
+		sg_miter_start(&host->sg_miter, data->sg, data->sg_len, flags);
 	}
 }
 
@@ -843,6 +832,8 @@ davinci_abort_data(struct mmc_davinci_host *host, struct mmc_data *data)
 {
 	mmc_davinci_reset_ctrl(host, 1);
 	mmc_davinci_reset_ctrl(host, 0);
+	if (!host->do_dma)
+		sg_miter_stop(&host->sg_miter);
 }
 
 static irqreturn_t mmc_davinci_sdio_irq(int irq, void *dev_id)
@@ -919,11 +910,13 @@ static irqreturn_t mmc_davinci_irq(int irq, void *dev_id)
 	if (qstatus & MMCST0_DATDNE) {
 		/* All blocks sent/received, and CRC checks passed */
 		if (data != NULL) {
-			if ((host->do_dma == 0) && (host->bytes_left > 0)) {
-				/* if datasize < rw_threshold
-				 * no RX ints are generated
-				 */
-				davinci_fifo_data_trans(host, host->bytes_left);
+			if (!host->do_dma) {
+				if (host->bytes_left > 0)
+					/* if datasize < rw_threshold
+					 * no RX ints are generated
+					 */
+					davinci_fifo_data_trans(host, host->bytes_left);
+				sg_miter_stop(&host->sg_miter);
 			}
 			end_transfer = 1;
 			data->bytes_xfered = data->blocks * data->blksz;
@@ -1344,7 +1337,7 @@ ioremap_fail:
 	return ret;
 }
 
-static void __exit davinci_mmcsd_remove(struct platform_device *pdev)
+static void davinci_mmcsd_remove(struct platform_device *pdev)
 {
 	struct mmc_davinci_host *host = platform_get_drvdata(pdev);
 
@@ -1399,7 +1392,7 @@ static struct platform_driver davinci_mmcsd_driver = {
 		.of_match_table = davinci_mmc_dt_ids,
 	},
 	.probe		= davinci_mmcsd_probe,
-	.remove_new	= __exit_p(davinci_mmcsd_remove),
+	.remove_new	= davinci_mmcsd_remove,
 	.id_table	= davinci_mmc_devtype,
 };
 
diff --git a/drivers/mmc/host/dw_mmc-exynos.c b/drivers/mmc/host/dw_mmc-exynos.c
index 698408e8ba..6dc057718d 100644
--- a/drivers/mmc/host/dw_mmc-exynos.c
+++ b/drivers/mmc/host/dw_mmc-exynos.c
@@ -11,7 +11,6 @@
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
 #include <linux/of.h>
-#include <linux/of_gpio.h>
 #include <linux/pm_runtime.h>
 #include <linux/slab.h>
 
diff --git a/drivers/mmc/host/dw_mmc-hi3798cv200.c b/drivers/mmc/host/dw_mmc-hi3798cv200.c
index e9470c50a3..61923a5183 100644
--- a/drivers/mmc/host/dw_mmc-hi3798cv200.c
+++ b/drivers/mmc/host/dw_mmc-hi3798cv200.c
@@ -201,4 +201,3 @@ module_platform_driver(dw_mci_hi3798cv200_driver);
 
 MODULE_DESCRIPTION("HiSilicon Hi3798CV200 Specific DW-MSHC Driver Extension");
 MODULE_LICENSE("GPL v2");
-MODULE_ALIAS("platform:dwmmc_hi3798cv200");
diff --git a/drivers/mmc/host/dw_mmc-hi3798mv200.c b/drivers/mmc/host/dw_mmc-hi3798mv200.c
new file mode 100644
index 0000000000..989ae8dda7
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-hi3798mv200.c
@@ -0,0 +1,251 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Modified from dw_mmc-hi3798cv200.c
+ *
+ * Copyright (c) 2024 Yang Xiwen <forbidden405@outlook.com>
+ * Copyright (c) 2018 HiSilicon Technologies Co., Ltd.
+ */
+
+#include <linux/clk.h>
+#include <linux/mfd/syscon.h>
+#include <linux/mmc/host.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+
+#include "dw_mmc.h"
+#include "dw_mmc-pltfm.h"
+
+#define SDMMC_TUNING_CTRL	0x118
+#define SDMMC_TUNING_FIND_EDGE	BIT(5)
+
+#define ALL_INT_CLR		0x1ffff
+
+/* DLL ctrl reg */
+#define SAP_DLL_CTRL_DLLMODE	BIT(16)
+
+struct dw_mci_hi3798mv200_priv {
+	struct clk *sample_clk;
+	struct clk *drive_clk;
+	struct regmap *crg_reg;
+	u32 sap_dll_offset;
+	struct mmc_clk_phase_map phase_map;
+};
+
+static void dw_mci_hi3798mv200_set_ios(struct dw_mci *host, struct mmc_ios *ios)
+{
+	struct dw_mci_hi3798mv200_priv *priv = host->priv;
+	struct mmc_clk_phase phase = priv->phase_map.phase[ios->timing];
+	u32 val;
+
+	val = mci_readl(host, ENABLE_SHIFT);
+	if (ios->timing == MMC_TIMING_MMC_DDR52
+	    || ios->timing == MMC_TIMING_UHS_DDR50)
+		val |= SDMMC_ENABLE_PHASE;
+	else
+		val &= ~SDMMC_ENABLE_PHASE;
+	mci_writel(host, ENABLE_SHIFT, val);
+
+	val = mci_readl(host, DDR_REG);
+	if (ios->timing == MMC_TIMING_MMC_HS400)
+		val |= SDMMC_DDR_HS400;
+	else
+		val &= ~SDMMC_DDR_HS400;
+	mci_writel(host, DDR_REG, val);
+
+	if (clk_set_rate(host->ciu_clk, ios->clock))
+		dev_warn(host->dev, "Failed to set rate to %u\n", ios->clock);
+	else
+		/*
+		 * CLK_MUX_ROUND_NEAREST is enabled for this clock
+		 * The actual clock rate is not what we set, but a rounded value
+		 * so we should get the rate once again
+		 */
+		host->bus_hz = clk_get_rate(host->ciu_clk);
+
+	if (phase.valid) {
+		clk_set_phase(priv->drive_clk, phase.out_deg);
+		clk_set_phase(priv->sample_clk, phase.in_deg);
+	} else {
+		dev_warn(host->dev,
+			 "The phase entry for timing mode %d is missing in device tree.\n",
+			 ios->timing);
+	}
+}
+
+static inline int dw_mci_hi3798mv200_enable_tuning(struct dw_mci_slot *slot)
+{
+	struct dw_mci_hi3798mv200_priv *priv = slot->host->priv;
+
+	return regmap_clear_bits(priv->crg_reg, priv->sap_dll_offset, SAP_DLL_CTRL_DLLMODE);
+}
+
+static inline int dw_mci_hi3798mv200_disable_tuning(struct dw_mci_slot *slot)
+{
+	struct dw_mci_hi3798mv200_priv *priv = slot->host->priv;
+
+	return regmap_set_bits(priv->crg_reg, priv->sap_dll_offset, SAP_DLL_CTRL_DLLMODE);
+}
+
+static int dw_mci_hi3798mv200_execute_tuning_mix_mode(struct dw_mci_slot *slot,
+					     u32 opcode)
+{
+	static const int degrees[] = { 0, 45, 90, 135, 180, 225, 270, 315 };
+	struct dw_mci *host = slot->host;
+	struct dw_mci_hi3798mv200_priv *priv = host->priv;
+	int raise_point = -1, fall_point = -1, mid;
+	int err, prev_err = -1;
+	int found = 0;
+	int regval;
+	int i;
+	int ret;
+
+	ret = dw_mci_hi3798mv200_enable_tuning(slot);
+	if (ret < 0)
+		return ret;
+
+	for (i = 0; i < ARRAY_SIZE(degrees); i++) {
+		clk_set_phase(priv->sample_clk, degrees[i]);
+		mci_writel(host, RINTSTS, ALL_INT_CLR);
+
+		/*
+		 * HiSilicon implemented a tuning mechanism.
+		 * It needs special interaction with the DLL.
+		 *
+		 * Treat edge(flip) found as an error too.
+		 */
+		err = mmc_send_tuning(slot->mmc, opcode, NULL);
+		regval = mci_readl(host, TUNING_CTRL);
+		if (err || (regval & SDMMC_TUNING_FIND_EDGE))
+			err = 1;
+		else
+			found = 1;
+
+		if (i > 0) {
+			if (err && !prev_err)
+				fall_point = i - 1;
+			if (!err && prev_err)
+				raise_point = i;
+		}
+
+		if (raise_point != -1 && fall_point != -1)
+			goto tuning_out;
+
+		prev_err = err;
+		err = 0;
+	}
+
+tuning_out:
+	ret = dw_mci_hi3798mv200_disable_tuning(slot);
+	if (ret < 0)
+		return ret;
+
+	if (found) {
+		if (raise_point == -1)
+			raise_point = 0;
+		if (fall_point == -1)
+			fall_point = ARRAY_SIZE(degrees) - 1;
+		if (fall_point < raise_point) {
+			if ((raise_point + fall_point) >
+			    (ARRAY_SIZE(degrees) - 1))
+				mid = fall_point / 2;
+			else
+				mid = (raise_point + ARRAY_SIZE(degrees) - 1) / 2;
+		} else {
+			mid = (raise_point + fall_point) / 2;
+		}
+
+		/*
+		 * We don't care what timing we are tuning for,
+		 * simply use the same phase for all timing needs tuning.
+		 */
+		priv->phase_map.phase[MMC_TIMING_MMC_HS200].in_deg = degrees[mid];
+		priv->phase_map.phase[MMC_TIMING_MMC_HS400].in_deg = degrees[mid];
+		priv->phase_map.phase[MMC_TIMING_UHS_SDR104].in_deg = degrees[mid];
+
+		clk_set_phase(priv->sample_clk, degrees[mid]);
+		dev_dbg(host->dev, "Tuning clk_sample[%d, %d], set[%d]\n",
+			raise_point, fall_point, degrees[mid]);
+		ret = 0;
+	} else {
+		dev_err(host->dev, "No valid clk_sample shift!\n");
+		ret = -EINVAL;
+	}
+
+	mci_writel(host, RINTSTS, ALL_INT_CLR);
+
+	return ret;
+}
+
+static int dw_mci_hi3798mv200_init(struct dw_mci *host)
+{
+	struct dw_mci_hi3798mv200_priv *priv;
+	struct device_node *np = host->dev->of_node;
+	int ret;
+
+	priv = devm_kzalloc(host->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	mmc_of_parse_clk_phase(host->dev, &priv->phase_map);
+
+	priv->sample_clk = devm_clk_get_enabled(host->dev, "ciu-sample");
+	if (IS_ERR(priv->sample_clk))
+		return dev_err_probe(host->dev, PTR_ERR(priv->sample_clk),
+				     "failed to get enabled ciu-sample clock\n");
+
+	priv->drive_clk = devm_clk_get_enabled(host->dev, "ciu-drive");
+	if (IS_ERR(priv->drive_clk))
+		return dev_err_probe(host->dev, PTR_ERR(priv->drive_clk),
+				     "failed to get enabled ciu-drive clock\n");
+
+	priv->crg_reg = syscon_regmap_lookup_by_phandle(np, "hisilicon,sap-dll-reg");
+	if (IS_ERR(priv->crg_reg))
+		return dev_err_probe(host->dev, PTR_ERR(priv->crg_reg),
+				     "failed to get CRG reg\n");
+
+	ret = of_property_read_u32_index(np, "hisilicon,sap-dll-reg", 1, &priv->sap_dll_offset);
+	if (ret)
+		return dev_err_probe(host->dev, ret, "failed to get sample DLL register offset\n");
+
+	host->priv = priv;
+	return 0;
+}
+
+static const struct dw_mci_drv_data hi3798mv200_data = {
+	.common_caps = MMC_CAP_CMD23,
+	.init = dw_mci_hi3798mv200_init,
+	.set_ios = dw_mci_hi3798mv200_set_ios,
+	.execute_tuning = dw_mci_hi3798mv200_execute_tuning_mix_mode,
+};
+
+static const struct of_device_id dw_mci_hi3798mv200_match[] = {
+	{ .compatible = "hisilicon,hi3798mv200-dw-mshc" },
+	{},
+};
+
+static int dw_mci_hi3798mv200_probe(struct platform_device *pdev)
+{
+	return dw_mci_pltfm_register(pdev, &hi3798mv200_data);
+}
+
+static void dw_mci_hi3798mv200_remove(struct platform_device *pdev)
+{
+	dw_mci_pltfm_remove(pdev);
+}
+
+MODULE_DEVICE_TABLE(of, dw_mci_hi3798mv200_match);
+static struct platform_driver dw_mci_hi3798mv200_driver = {
+	.probe = dw_mci_hi3798mv200_probe,
+	.remove_new = dw_mci_hi3798mv200_remove,
+	.driver = {
+		.name = "dwmmc_hi3798mv200",
+		.probe_type = PROBE_PREFER_ASYNCHRONOUS,
+		.of_match_table = dw_mci_hi3798mv200_match,
+	},
+};
+module_platform_driver(dw_mci_hi3798mv200_driver);
+
+MODULE_DESCRIPTION("HiSilicon Hi3798MV200 Specific DW-MSHC Driver Extension");
+MODULE_LICENSE("GPL");
diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
index 829af2c98a..8e2d676b92 100644
--- a/drivers/mmc/host/dw_mmc.c
+++ b/drivers/mmc/host/dw_mmc.c
@@ -35,7 +35,6 @@
 #include <linux/bitops.h>
 #include <linux/regulator/consumer.h>
 #include <linux/of.h>
-#include <linux/of_gpio.h>
 #include <linux/mmc/slot-gpio.h>
 
 #include "dw_mmc.h"
diff --git a/drivers/mmc/host/meson-mx-sdhc-clkc.c b/drivers/mmc/host/meson-mx-sdhc-clkc.c
index 19200b7079..cbd17a596c 100644
--- a/drivers/mmc/host/meson-mx-sdhc-clkc.c
+++ b/drivers/mmc/host/meson-mx-sdhc-clkc.c
@@ -71,12 +71,23 @@ static int meson_mx_sdhc_clk_hw_register(struct device *dev,
 static int meson_mx_sdhc_gate_clk_hw_register(struct device *dev,
 					      const char *name_suffix,
 					      struct clk_hw *parent,
-					      struct clk_hw *hw)
+					      struct clk_hw *hw,
+					      struct clk_bulk_data *clk_bulk_data,
+					      u8 bulk_index)
 {
 	struct clk_parent_data parent_data = { .hw = parent };
+	int ret;
+
+	ret = meson_mx_sdhc_clk_hw_register(dev, name_suffix, &parent_data, 1,
+					    &clk_gate_ops, hw);
+	if (ret)
+		return ret;
+
+	clk_bulk_data[bulk_index].clk = devm_clk_hw_get_clk(dev, hw, name_suffix);
+	if (IS_ERR(clk_bulk_data[bulk_index].clk))
+		return PTR_ERR(clk_bulk_data[bulk_index].clk);
 
-	return meson_mx_sdhc_clk_hw_register(dev, name_suffix, &parent_data, 1,
-					     &clk_gate_ops, hw);
+	return 0;
 }
 
 int meson_mx_sdhc_register_clkc(struct device *dev, void __iomem *base,
@@ -115,7 +126,8 @@ int meson_mx_sdhc_register_clkc(struct device *dev, void __iomem *base,
 	clkc_data->mod_clk_en.bit_idx = 15;
 	ret = meson_mx_sdhc_gate_clk_hw_register(dev, "mod_clk_on",
 						 &clkc_data->div.hw,
-						 &clkc_data->mod_clk_en.hw);
+						 &clkc_data->mod_clk_en.hw,
+						 clk_bulk_data, 0);
 	if (ret)
 		return ret;
 
@@ -123,7 +135,8 @@ int meson_mx_sdhc_register_clkc(struct device *dev, void __iomem *base,
 	clkc_data->tx_clk_en.bit_idx = 14;
 	ret = meson_mx_sdhc_gate_clk_hw_register(dev, "tx_clk_on",
 						 &clkc_data->div.hw,
-						 &clkc_data->tx_clk_en.hw);
+						 &clkc_data->tx_clk_en.hw,
+						 clk_bulk_data, 1);
 	if (ret)
 		return ret;
 
@@ -131,7 +144,8 @@ int meson_mx_sdhc_register_clkc(struct device *dev, void __iomem *base,
 	clkc_data->rx_clk_en.bit_idx = 13;
 	ret = meson_mx_sdhc_gate_clk_hw_register(dev, "rx_clk_on",
 						 &clkc_data->div.hw,
-						 &clkc_data->rx_clk_en.hw);
+						 &clkc_data->rx_clk_en.hw,
+						 clk_bulk_data, 2);
 	if (ret)
 		return ret;
 
@@ -139,18 +153,7 @@ int meson_mx_sdhc_register_clkc(struct device *dev, void __iomem *base,
 	clkc_data->sd_clk_en.bit_idx = 12;
 	ret = meson_mx_sdhc_gate_clk_hw_register(dev, "sd_clk_on",
 						 &clkc_data->div.hw,
-						 &clkc_data->sd_clk_en.hw);
-	if (ret)
-		return ret;
-
-	/*
-	 * TODO: Replace clk_hw.clk with devm_clk_hw_get_clk() once that is
-	 * available.
-	 */
-	clk_bulk_data[0].clk = clkc_data->mod_clk_en.hw.clk;
-	clk_bulk_data[1].clk = clkc_data->sd_clk_en.hw.clk;
-	clk_bulk_data[2].clk = clkc_data->tx_clk_en.hw.clk;
-	clk_bulk_data[3].clk = clkc_data->rx_clk_en.hw.clk;
-
-	return 0;
+						 &clkc_data->sd_clk_en.hw,
+						 clk_bulk_data, 3);
+	return ret;
 }
diff --git a/drivers/mmc/host/meson-mx-sdhc-mmc.c b/drivers/mmc/host/meson-mx-sdhc-mmc.c
index 1ed9731e77..31f750301d 100644
--- a/drivers/mmc/host/meson-mx-sdhc-mmc.c
+++ b/drivers/mmc/host/meson-mx-sdhc-mmc.c
@@ -65,10 +65,8 @@ static const struct regmap_config meson_mx_sdhc_regmap_config = {
 	.max_register = MESON_SDHC_CLK2,
 };
 
-static void meson_mx_sdhc_hw_reset(struct mmc_host *mmc)
+static void meson_mx_sdhc_reset(struct meson_mx_sdhc_host *host)
 {
-	struct meson_mx_sdhc_host *host = mmc_priv(mmc);
-
 	regmap_write(host->regmap, MESON_SDHC_SRST, MESON_SDHC_SRST_MAIN_CTRL |
 		     MESON_SDHC_SRST_RXFIFO | MESON_SDHC_SRST_TXFIFO |
 		     MESON_SDHC_SRST_DPHY_RX | MESON_SDHC_SRST_DPHY_TX |
@@ -116,7 +114,7 @@ static void meson_mx_sdhc_wait_cmd_ready(struct mmc_host *mmc)
 		dev_warn(mmc_dev(mmc),
 			 "Failed to poll for CMD_BUSY while processing CMD%d\n",
 			 host->cmd->opcode);
-		meson_mx_sdhc_hw_reset(mmc);
+		meson_mx_sdhc_reset(host);
 	}
 
 	ret = regmap_read_poll_timeout(host->regmap, MESON_SDHC_ESTA, esta,
@@ -127,7 +125,7 @@ static void meson_mx_sdhc_wait_cmd_ready(struct mmc_host *mmc)
 		dev_warn(mmc_dev(mmc),
 			 "Failed to poll for ESTA[13:11] while processing CMD%d\n",
 			 host->cmd->opcode);
-		meson_mx_sdhc_hw_reset(mmc);
+		meson_mx_sdhc_reset(host);
 	}
 }
 
@@ -495,7 +493,6 @@ static int meson_mx_sdhc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 }
 
 static const struct mmc_host_ops meson_mx_sdhc_ops = {
-	.card_hw_reset			= meson_mx_sdhc_hw_reset,
 	.request			= meson_mx_sdhc_request,
 	.set_ios			= meson_mx_sdhc_set_ios,
 	.card_busy			= meson_mx_sdhc_card_busy,
@@ -618,7 +615,7 @@ static irqreturn_t meson_mx_sdhc_irq_thread(int irq, void *irq_data)
 	}
 
 	if (cmd->error == -EIO || cmd->error == -ETIMEDOUT)
-		meson_mx_sdhc_hw_reset(host->mmc);
+		meson_mx_sdhc_reset(host);
 	else if (cmd->data)
 		/*
 		 * Clear the FIFOs after completing data transfers to prevent
@@ -728,7 +725,7 @@ static void meson_mx_sdhc_init_hw(struct mmc_host *mmc)
 {
 	struct meson_mx_sdhc_host *host = mmc_priv(mmc);
 
-	meson_mx_sdhc_hw_reset(mmc);
+	meson_mx_sdhc_reset(host);
 
 	regmap_write(host->regmap, MESON_SDHC_CTRL,
 		     FIELD_PREP(MESON_SDHC_CTRL_RX_PERIOD, 0xf) |
diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c
index b8dda8160c..09d7a6a0dc 100644
--- a/drivers/mmc/host/mmc_spi.c
+++ b/drivers/mmc/host/mmc_spi.c
@@ -15,7 +15,6 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/bio.h>
-#include <linux/dma-direction.h>
 #include <linux/crc7.h>
 #include <linux/crc-itu-t.h>
 #include <linux/scatterlist.h>
@@ -510,10 +509,7 @@ mmc_spi_command_send(struct mmc_spi_host *host,
  * so we explicitly initialize it to all ones on RX paths.
  */
 static void
-mmc_spi_setup_data_message(
-	struct mmc_spi_host	*host,
-	bool			multiple,
-	enum dma_data_direction	direction)
+mmc_spi_setup_data_message(struct mmc_spi_host *host, bool multiple, bool write)
 {
 	struct spi_transfer	*t;
 	struct scratch		*scratch = host->data;
@@ -523,7 +519,7 @@ mmc_spi_setup_data_message(
 	/* for reads, readblock() skips 0xff bytes before finding
 	 * the token; for writes, this transfer issues that token.
 	 */
-	if (direction == DMA_TO_DEVICE) {
+	if (write) {
 		t = &host->token;
 		memset(t, 0, sizeof(*t));
 		t->len = 1;
@@ -547,7 +543,7 @@ mmc_spi_setup_data_message(
 	t = &host->crc;
 	memset(t, 0, sizeof(*t));
 	t->len = 2;
-	if (direction == DMA_TO_DEVICE) {
+	if (write) {
 		/* the actual CRC may get written later */
 		t->tx_buf = &scratch->crc_val;
 	} else {
@@ -570,10 +566,10 @@ mmc_spi_setup_data_message(
 	 * the next token (next data block, or STOP_TRAN).  We can try to
 	 * minimize I/O ops by using a single read to collect end-of-busy.
 	 */
-	if (multiple || direction == DMA_TO_DEVICE) {
+	if (multiple || write) {
 		t = &host->early_status;
 		memset(t, 0, sizeof(*t));
-		t->len = (direction == DMA_TO_DEVICE) ? sizeof(scratch->status) : 1;
+		t->len = write ? sizeof(scratch->status) : 1;
 		t->tx_buf = host->ones;
 		t->rx_buf = scratch->status;
 		t->cs_change = 1;
@@ -777,15 +773,15 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 {
 	struct spi_device	*spi = host->spi;
 	struct spi_transfer	*t;
-	enum dma_data_direction	direction = mmc_get_dma_dir(data);
 	struct scatterlist	*sg;
 	unsigned		n_sg;
 	bool			multiple = (data->blocks > 1);
-	const char		*write_or_read = (direction == DMA_TO_DEVICE) ? "write" : "read";
+	bool			write = (data->flags & MMC_DATA_WRITE);
+	const char		*write_or_read = write ? "write" : "read";
 	u32			clock_rate;
 	unsigned long		timeout;
 
-	mmc_spi_setup_data_message(host, multiple, direction);
+	mmc_spi_setup_data_message(host, multiple, write);
 	t = &host->t;
 
 	if (t->speed_hz)
@@ -807,7 +803,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 
 		/* allow pio too; we don't allow highmem */
 		kmap_addr = kmap(sg_page(sg));
-		if (direction == DMA_TO_DEVICE)
+		if (write)
 			t->tx_buf = kmap_addr + sg->offset;
 		else
 			t->rx_buf = kmap_addr + sg->offset;
@@ -818,7 +814,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 
 			dev_dbg(&spi->dev, "    %s block, %d bytes\n", write_or_read, t->len);
 
-			if (direction == DMA_TO_DEVICE)
+			if (write)
 				status = mmc_spi_writeblock(host, t, timeout);
 			else
 				status = mmc_spi_readblock(host, t, timeout);
@@ -833,7 +829,9 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 		}
 
 		/* discard mappings */
-		if (direction == DMA_FROM_DEVICE)
+		if (write)
+			/* nothing to do */;
+		else
 			flush_dcache_page(sg_page(sg));
 		kunmap(sg_page(sg));
 
@@ -850,7 +848,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 	 * that can affect the STOP_TRAN logic.   Complete (and current)
 	 * MMC specs should sort that out before Linux starts using CMD23.
 	 */
-	if (direction == DMA_TO_DEVICE && multiple) {
+	if (write && multiple) {
 		struct scratch	*scratch = host->data;
 		int		tmp;
 		const unsigned	statlen = sizeof(scratch->status);
@@ -935,7 +933,7 @@ static void mmc_spi_request(struct mmc_host *mmc, struct mmc_request *mrq)
 #endif
 
 	/* request exclusive bus access */
-	spi_bus_lock(host->spi->master);
+	spi_bus_lock(host->spi->controller);
 
 crc_recover:
 	/* issue command; then optionally data and stop */
@@ -967,7 +965,7 @@ crc_recover:
 	}
 
 	/* release the bus */
-	spi_bus_unlock(host->spi->master);
+	spi_bus_unlock(host->spi->controller);
 
 	mmc_request_done(host->mmc, mrq);
 }
@@ -1157,7 +1155,7 @@ static int mmc_spi_probe(struct spi_device *spi)
 	/* We rely on full duplex transfers, mostly to reduce
 	 * per-transfer overheads (by making fewer transfers).
 	 */
-	if (spi->master->flags & SPI_CONTROLLER_HALF_DUPLEX)
+	if (spi->controller->flags & SPI_CONTROLLER_HALF_DUPLEX)
 		return -EINVAL;
 
 	/* MMC and SD specs only seem to care that sampling is on the
diff --git a/drivers/mmc/host/moxart-mmc.c b/drivers/mmc/host/moxart-mmc.c
index 5cfdd3a86e..9a5f75163a 100644
--- a/drivers/mmc/host/moxart-mmc.c
+++ b/drivers/mmc/host/moxart-mmc.c
@@ -131,12 +131,10 @@ struct moxart_host {
 	struct dma_async_tx_descriptor	*tx_desc;
 	struct mmc_host			*mmc;
 	struct mmc_request		*mrq;
-	struct scatterlist		*cur_sg;
 	struct completion		dma_complete;
 	struct completion		pio_complete;
 
-	u32				num_sg;
-	u32				data_remain;
+	struct sg_mapping_iter		sg_miter;
 	u32				data_len;
 	u32				fifo_width;
 	u32				timeout;
@@ -148,35 +146,6 @@ struct moxart_host {
 	bool				is_removed;
 };
 
-static inline void moxart_init_sg(struct moxart_host *host,
-				  struct mmc_data *data)
-{
-	host->cur_sg = data->sg;
-	host->num_sg = data->sg_len;
-	host->data_remain = host->cur_sg->length;
-
-	if (host->data_remain > host->data_len)
-		host->data_remain = host->data_len;
-}
-
-static inline int moxart_next_sg(struct moxart_host *host)
-{
-	int remain;
-	struct mmc_data *data = host->mrq->cmd->data;
-
-	host->cur_sg++;
-	host->num_sg--;
-
-	if (host->num_sg > 0) {
-		host->data_remain = host->cur_sg->length;
-		remain = host->data_len - data->bytes_xfered;
-		if (remain > 0 && remain < host->data_remain)
-			host->data_remain = remain;
-	}
-
-	return host->num_sg;
-}
-
 static int moxart_wait_for_status(struct moxart_host *host,
 				  u32 mask, u32 *status)
 {
@@ -254,6 +223,11 @@ static void moxart_dma_complete(void *param)
 	complete(&host->dma_complete);
 }
 
+static bool moxart_use_dma(struct moxart_host *host)
+{
+	return (host->data_len > host->fifo_width) && host->have_dma;
+}
+
 static void moxart_transfer_dma(struct mmc_data *data, struct moxart_host *host)
 {
 	u32 len, dir_slave;
@@ -291,11 +265,11 @@ static void moxart_transfer_dma(struct mmc_data *data, struct moxart_host *host)
 		dma_async_issue_pending(dma_chan);
 	}
 
-	data->bytes_xfered += host->data_remain;
-
 	wait_for_completion_interruptible_timeout(&host->dma_complete,
 						  host->timeout);
 
+	data->bytes_xfered = host->data_len;
+
 	dma_unmap_sg(dma_chan->device->dev,
 		     data->sg, data->sg_len,
 		     mmc_get_dma_dir(data));
@@ -304,14 +278,29 @@ static void moxart_transfer_dma(struct mmc_data *data, struct moxart_host *host)
 
 static void moxart_transfer_pio(struct moxart_host *host)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct mmc_data *data = host->mrq->cmd->data;
 	u32 *sgp, len = 0, remain, status;
 
 	if (host->data_len == data->bytes_xfered)
 		return;
 
-	sgp = sg_virt(host->cur_sg);
-	remain = host->data_remain;
+	/*
+	 * By updating sgm->consumes this will get a proper pointer into the
+	 * buffer at any time.
+	 */
+	if (!sg_miter_next(sgm)) {
+		/* This shold not happen */
+		dev_err(mmc_dev(host->mmc), "ran out of scatterlist prematurely\n");
+		data->error = -EINVAL;
+		complete(&host->pio_complete);
+		return;
+	}
+	sgp = sgm->addr;
+	remain = sgm->length;
+	if (remain > host->data_len)
+		remain = host->data_len;
+	sgm->consumed = 0;
 
 	if (data->flags & MMC_DATA_WRITE) {
 		while (remain > 0) {
@@ -326,6 +315,7 @@ static void moxart_transfer_pio(struct moxart_host *host)
 				sgp++;
 				len += 4;
 			}
+			sgm->consumed += len;
 			remain -= len;
 		}
 
@@ -342,22 +332,22 @@ static void moxart_transfer_pio(struct moxart_host *host)
 				sgp++;
 				len += 4;
 			}
+			sgm->consumed += len;
 			remain -= len;
 		}
 	}
 
-	data->bytes_xfered += host->data_remain - remain;
-	host->data_remain = remain;
-
-	if (host->data_len != data->bytes_xfered)
-		moxart_next_sg(host);
-	else
+	data->bytes_xfered += sgm->consumed;
+	if (host->data_len == data->bytes_xfered) {
 		complete(&host->pio_complete);
+		return;
+	}
 }
 
 static void moxart_prepare_data(struct moxart_host *host)
 {
 	struct mmc_data *data = host->mrq->cmd->data;
+	unsigned int flags = SG_MITER_ATOMIC; /* Used from IRQ */
 	u32 datactrl;
 	int blksz_bits;
 
@@ -368,15 +358,19 @@ static void moxart_prepare_data(struct moxart_host *host)
 	blksz_bits = ffs(data->blksz) - 1;
 	BUG_ON(1 << blksz_bits != data->blksz);
 
-	moxart_init_sg(host, data);
-
 	datactrl = DCR_DATA_EN | (blksz_bits & DCR_BLK_SIZE);
 
-	if (data->flags & MMC_DATA_WRITE)
+	if (data->flags & MMC_DATA_WRITE) {
+		flags |= SG_MITER_FROM_SG;
 		datactrl |= DCR_DATA_WRITE;
+	} else {
+		flags |= SG_MITER_TO_SG;
+	}
 
-	if ((host->data_len > host->fifo_width) && host->have_dma)
+	if (moxart_use_dma(host))
 		datactrl |= DCR_DMA_EN;
+	else
+		sg_miter_start(&host->sg_miter, data->sg, data->sg_len, flags);
 
 	writel(DCR_DATA_FIFO_RESET, host->base + REG_DATA_CONTROL);
 	writel(MASK_DATA | FIFO_URUN | FIFO_ORUN, host->base + REG_CLEAR);
@@ -407,7 +401,7 @@ static void moxart_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	moxart_send_command(host, host->mrq->cmd);
 
 	if (mrq->cmd->data) {
-		if ((host->data_len > host->fifo_width) && host->have_dma) {
+		if (moxart_use_dma(host)) {
 
 			writel(CARD_CHANGE, host->base + REG_INTERRUPT_MASK);
 
@@ -449,6 +443,9 @@ static void moxart_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	}
 
 request_done:
+	if (!moxart_use_dma(host))
+		sg_miter_stop(&host->sg_miter);
+
 	spin_unlock_irqrestore(&host->lock, flags);
 	mmc_request_done(host->mmc, mrq);
 }
diff --git a/drivers/mmc/host/mvsdio.c b/drivers/mmc/host/mvsdio.c
index ca01b7d204..af7f21888e 100644
--- a/drivers/mmc/host/mvsdio.c
+++ b/drivers/mmc/host/mvsdio.c
@@ -38,8 +38,9 @@ struct mvsd_host {
 	unsigned int xfer_mode;
 	unsigned int intr_en;
 	unsigned int ctrl;
+	bool use_pio;
+	struct sg_mapping_iter sg_miter;
 	unsigned int pio_size;
-	void *pio_ptr;
 	unsigned int sg_frags;
 	unsigned int ns_per_clk;
 	unsigned int clock;
@@ -114,11 +115,18 @@ static int mvsd_setup_data(struct mvsd_host *host, struct mmc_data *data)
 		 * data when the buffer is not aligned on a 64 byte
 		 * boundary.
 		 */
+		unsigned int miter_flags = SG_MITER_ATOMIC; /* Used from IRQ */
+
+		if (data->flags & MMC_DATA_READ)
+			miter_flags |= SG_MITER_TO_SG;
+		else
+			miter_flags |= SG_MITER_FROM_SG;
+
 		host->pio_size = data->blocks * data->blksz;
-		host->pio_ptr = sg_virt(data->sg);
+		sg_miter_start(&host->sg_miter, data->sg, data->sg_len, miter_flags);
 		if (!nodma)
-			dev_dbg(host->dev, "fallback to PIO for data at 0x%p size %d\n",
-				host->pio_ptr, host->pio_size);
+			dev_dbg(host->dev, "fallback to PIO for data\n");
+		host->use_pio = true;
 		return 1;
 	} else {
 		dma_addr_t phys_addr;
@@ -129,6 +137,7 @@ static int mvsd_setup_data(struct mvsd_host *host, struct mmc_data *data)
 		phys_addr = sg_dma_address(data->sg);
 		mvsd_write(MVSD_SYS_ADDR_LOW, (u32)phys_addr & 0xffff);
 		mvsd_write(MVSD_SYS_ADDR_HI,  (u32)phys_addr >> 16);
+		host->use_pio = false;
 		return 0;
 	}
 }
@@ -288,8 +297,8 @@ static u32 mvsd_finish_data(struct mvsd_host *host, struct mmc_data *data,
 {
 	void __iomem *iobase = host->base;
 
-	if (host->pio_ptr) {
-		host->pio_ptr = NULL;
+	if (host->use_pio) {
+		sg_miter_stop(&host->sg_miter);
 		host->pio_size = 0;
 	} else {
 		dma_unmap_sg(mmc_dev(host->mmc), data->sg, host->sg_frags,
@@ -344,9 +353,12 @@ static u32 mvsd_finish_data(struct mvsd_host *host, struct mmc_data *data,
 static irqreturn_t mvsd_irq(int irq, void *dev)
 {
 	struct mvsd_host *host = dev;
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	void __iomem *iobase = host->base;
 	u32 intr_status, intr_done_mask;
 	int irq_handled = 0;
+	u16 *p;
+	int s;
 
 	intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 	dev_dbg(host->dev, "intr 0x%04x intr_en 0x%04x hw_state 0x%04x\n",
@@ -370,15 +382,36 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 	spin_lock(&host->lock);
 
 	/* PIO handling, if needed. Messy business... */
-	if (host->pio_size &&
+	if (host->use_pio) {
+		/*
+		 * As we set sgm->consumed this always gives a valid buffer
+		 * position.
+		 */
+		if (!sg_miter_next(sgm)) {
+			/* This should not happen */
+			dev_err(host->dev, "ran out of scatter segments\n");
+			spin_unlock(&host->lock);
+			host->intr_en &=
+				~(MVSD_NOR_RX_READY | MVSD_NOR_RX_FIFO_8W |
+				  MVSD_NOR_TX_AVAIL | MVSD_NOR_TX_FIFO_8W);
+			mvsd_write(MVSD_NOR_INTR_EN, host->intr_en);
+			return IRQ_HANDLED;
+		}
+		p = sgm->addr;
+		s = sgm->length;
+		if (s > host->pio_size)
+			s = host->pio_size;
+	}
+
+	if (host->use_pio &&
 	    (intr_status & host->intr_en &
 	     (MVSD_NOR_RX_READY | MVSD_NOR_RX_FIFO_8W))) {
-		u16 *p = host->pio_ptr;
-		int s = host->pio_size;
+
 		while (s >= 32 && (intr_status & MVSD_NOR_RX_FIFO_8W)) {
 			readsw(iobase + MVSD_FIFO, p, 16);
 			p += 16;
 			s -= 32;
+			sgm->consumed += 32;
 			intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 		}
 		/*
@@ -391,6 +424,7 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 				put_unaligned(mvsd_read(MVSD_FIFO), p++);
 				put_unaligned(mvsd_read(MVSD_FIFO), p++);
 				s -= 4;
+				sgm->consumed += 4;
 				intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 			}
 			if (s && s < 4 && (intr_status & MVSD_NOR_RX_READY)) {
@@ -398,10 +432,13 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 				val[0] = mvsd_read(MVSD_FIFO);
 				val[1] = mvsd_read(MVSD_FIFO);
 				memcpy(p, ((void *)&val) + 4 - s, s);
+				sgm->consumed += s;
 				s = 0;
 				intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 			}
-			if (s == 0) {
+			/* PIO transfer done */
+			host->pio_size -= sgm->consumed;
+			if (host->pio_size == 0) {
 				host->intr_en &=
 				     ~(MVSD_NOR_RX_READY | MVSD_NOR_RX_FIFO_8W);
 				mvsd_write(MVSD_NOR_INTR_EN, host->intr_en);
@@ -413,14 +450,10 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 		}
 		dev_dbg(host->dev, "pio %d intr 0x%04x hw_state 0x%04x\n",
 			s, intr_status, mvsd_read(MVSD_HW_STATE));
-		host->pio_ptr = p;
-		host->pio_size = s;
 		irq_handled = 1;
-	} else if (host->pio_size &&
+	} else if (host->use_pio &&
 		   (intr_status & host->intr_en &
 		    (MVSD_NOR_TX_AVAIL | MVSD_NOR_TX_FIFO_8W))) {
-		u16 *p = host->pio_ptr;
-		int s = host->pio_size;
 		/*
 		 * The TX_FIFO_8W bit is unreliable. When set, bursting
 		 * 16 halfwords all at once in the FIFO drops data. Actually
@@ -431,6 +464,7 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 			mvsd_write(MVSD_FIFO, get_unaligned(p++));
 			mvsd_write(MVSD_FIFO, get_unaligned(p++));
 			s -= 4;
+			sgm->consumed += 4;
 			intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 		}
 		if (s < 4) {
@@ -439,10 +473,13 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 				memcpy(((void *)&val) + 4 - s, p, s);
 				mvsd_write(MVSD_FIFO, val[0]);
 				mvsd_write(MVSD_FIFO, val[1]);
+				sgm->consumed += s;
 				s = 0;
 				intr_status = mvsd_read(MVSD_NOR_INTR_STATUS);
 			}
-			if (s == 0) {
+			/* PIO transfer done */
+			host->pio_size -= sgm->consumed;
+			if (host->pio_size == 0) {
 				host->intr_en &=
 				     ~(MVSD_NOR_TX_AVAIL | MVSD_NOR_TX_FIFO_8W);
 				mvsd_write(MVSD_NOR_INTR_EN, host->intr_en);
@@ -450,8 +487,6 @@ static irqreturn_t mvsd_irq(int irq, void *dev)
 		}
 		dev_dbg(host->dev, "pio %d intr 0x%04x hw_state 0x%04x\n",
 			s, intr_status, mvsd_read(MVSD_HW_STATE));
-		host->pio_ptr = p;
-		host->pio_size = s;
 		irq_handled = 1;
 	}
 
diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c
index 5b3ab0e205..1edf652913 100644
--- a/drivers/mmc/host/mxcmmc.c
+++ b/drivers/mmc/host/mxcmmc.c
@@ -266,11 +266,18 @@ static inline void buffer_swap32(u32 *buf, int len)
 
 static void mxcmci_swap_buffers(struct mmc_data *data)
 {
-	struct scatterlist *sg;
-	int i;
+	struct sg_mapping_iter sgm;
+	u32 *buf;
+
+	sg_miter_start(&sgm, data->sg, data->sg_len,
+		       SG_MITER_TO_SG | SG_MITER_FROM_SG);
+
+	while (sg_miter_next(&sgm)) {
+		buf = sgm.addr;
+		buffer_swap32(buf, sgm.length);
+	}
 
-	for_each_sg(data->sg, sg, data->sg_len, i)
-		buffer_swap32(sg_virt(sg), sg->length);
+	sg_miter_stop(&sgm);
 }
 #else
 static inline void mxcmci_swap_buffers(struct mmc_data *data) {}
@@ -526,10 +533,9 @@ static int mxcmci_poll_status(struct mxcmci_host *host, u32 mask)
 	} while (1);
 }
 
-static int mxcmci_pull(struct mxcmci_host *host, void *_buf, int bytes)
+static int mxcmci_pull(struct mxcmci_host *host, u32 *buf, int bytes)
 {
 	unsigned int stat;
-	u32 *buf = _buf;
 
 	while (bytes > 3) {
 		stat = mxcmci_poll_status(host,
@@ -555,10 +561,9 @@ static int mxcmci_pull(struct mxcmci_host *host, void *_buf, int bytes)
 	return 0;
 }
 
-static int mxcmci_push(struct mxcmci_host *host, void *_buf, int bytes)
+static int mxcmci_push(struct mxcmci_host *host, u32 *buf, int bytes)
 {
 	unsigned int stat;
-	u32 *buf = _buf;
 
 	while (bytes > 3) {
 		stat = mxcmci_poll_status(host, STATUS_BUF_WRITE_RDY);
@@ -586,31 +591,39 @@ static int mxcmci_push(struct mxcmci_host *host, void *_buf, int bytes)
 static int mxcmci_transfer_data(struct mxcmci_host *host)
 {
 	struct mmc_data *data = host->req->data;
-	struct scatterlist *sg;
-	int stat, i;
+	struct sg_mapping_iter sgm;
+	int stat;
+	u32 *buf;
 
 	host->data = data;
 	host->datasize = 0;
+	sg_miter_start(&sgm, data->sg, data->sg_len,
+		       (data->flags & MMC_DATA_READ) ? SG_MITER_TO_SG : SG_MITER_FROM_SG);
 
 	if (data->flags & MMC_DATA_READ) {
-		for_each_sg(data->sg, sg, data->sg_len, i) {
-			stat = mxcmci_pull(host, sg_virt(sg), sg->length);
+		while (sg_miter_next(&sgm)) {
+			buf = sgm.addr;
+			stat = mxcmci_pull(host, buf, sgm.length);
 			if (stat)
-				return stat;
-			host->datasize += sg->length;
+				goto transfer_error;
+			host->datasize += sgm.length;
 		}
 	} else {
-		for_each_sg(data->sg, sg, data->sg_len, i) {
-			stat = mxcmci_push(host, sg_virt(sg), sg->length);
+		while (sg_miter_next(&sgm)) {
+			buf = sgm.addr;
+			stat = mxcmci_push(host, buf, sgm.length);
 			if (stat)
-				return stat;
-			host->datasize += sg->length;
+				goto transfer_error;
+			host->datasize += sgm.length;
 		}
 		stat = mxcmci_poll_status(host, STATUS_WRITE_OP_DONE);
 		if (stat)
-			return stat;
+			goto transfer_error;
 	}
-	return 0;
+
+transfer_error:
+	sg_miter_stop(&sgm);
+	return stat;
 }
 
 static void mxcmci_datawork(struct work_struct *work)
diff --git a/drivers/mmc/host/omap.c b/drivers/mmc/host/omap.c
index 13fa8588e3..a8ee0df471 100644
--- a/drivers/mmc/host/omap.c
+++ b/drivers/mmc/host/omap.c
@@ -148,10 +148,8 @@ struct mmc_omap_host {
 	struct work_struct      send_stop_work;
 	struct mmc_data		*stop_data;
 
+	struct sg_mapping_iter	sg_miter;
 	unsigned int		sg_len;
-	int			sg_idx;
-	u16 *			buffer;
-	u32			buffer_bytes_left;
 	u32			total_bytes_left;
 
 	unsigned		features;
@@ -456,6 +454,8 @@ mmc_omap_xfer_done(struct mmc_omap_host *host, struct mmc_data *data)
 {
 	if (host->dma_in_use)
 		mmc_omap_release_dma(host, data, data->error);
+	else
+		sg_miter_stop(&host->sg_miter);
 
 	host->data = NULL;
 	host->sg_len = 0;
@@ -651,19 +651,6 @@ mmc_omap_cmd_timer(struct timer_list *t)
 	spin_unlock_irqrestore(&host->slot_lock, flags);
 }
 
-/* PIO only */
-static void
-mmc_omap_sg_to_buf(struct mmc_omap_host *host)
-{
-	struct scatterlist *sg;
-
-	sg = host->data->sg + host->sg_idx;
-	host->buffer_bytes_left = sg->length;
-	host->buffer = sg_virt(sg);
-	if (host->buffer_bytes_left > host->total_bytes_left)
-		host->buffer_bytes_left = host->total_bytes_left;
-}
-
 static void
 mmc_omap_clk_timer(struct timer_list *t)
 {
@@ -676,33 +663,37 @@ mmc_omap_clk_timer(struct timer_list *t)
 static void
 mmc_omap_xfer_data(struct mmc_omap_host *host, int write)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	int n, nwords;
+	u16 *buffer;
 
-	if (host->buffer_bytes_left == 0) {
-		host->sg_idx++;
-		BUG_ON(host->sg_idx == host->sg_len);
-		mmc_omap_sg_to_buf(host);
+	if (!sg_miter_next(sgm)) {
+		/* This should not happen */
+		dev_err(mmc_dev(host->mmc), "ran out of scatterlist prematurely\n");
+		return;
 	}
+	buffer = sgm->addr;
+
 	n = 64;
-	if (n > host->buffer_bytes_left)
-		n = host->buffer_bytes_left;
+	if (n > sgm->length)
+		n = sgm->length;
+	if (n > host->total_bytes_left)
+		n = host->total_bytes_left;
 
 	/* Round up to handle odd number of bytes to transfer */
 	nwords = DIV_ROUND_UP(n, 2);
 
-	host->buffer_bytes_left -= n;
+	sgm->consumed = n;
 	host->total_bytes_left -= n;
 	host->data->bytes_xfered += n;
 
 	if (write) {
 		__raw_writesw(host->virt_base + OMAP_MMC_REG(host, DATA),
-			      host->buffer, nwords);
+			      buffer, nwords);
 	} else {
 		__raw_readsw(host->virt_base + OMAP_MMC_REG(host, DATA),
-			     host->buffer, nwords);
+			     buffer, nwords);
 	}
-
-	host->buffer += nwords;
 }
 
 #ifdef CONFIG_MMC_DEBUG
@@ -956,6 +947,7 @@ static inline void set_data_timeout(struct mmc_omap_host *host, struct mmc_reque
 static void
 mmc_omap_prepare_data(struct mmc_omap_host *host, struct mmc_request *req)
 {
+	unsigned int miter_flags = SG_MITER_ATOMIC; /* Used from IRQ */
 	struct mmc_data *data = req->data;
 	int i, use_dma = 1, block_size;
 	struct scatterlist *sg;
@@ -990,7 +982,6 @@ mmc_omap_prepare_data(struct mmc_omap_host *host, struct mmc_request *req)
 		}
 	}
 
-	host->sg_idx = 0;
 	if (use_dma) {
 		enum dma_data_direction dma_data_dir;
 		struct dma_async_tx_descriptor *tx;
@@ -1071,7 +1062,11 @@ mmc_omap_prepare_data(struct mmc_omap_host *host, struct mmc_request *req)
 	OMAP_MMC_WRITE(host, BUF, 0x1f1f);
 	host->total_bytes_left = data->blocks * block_size;
 	host->sg_len = sg_len;
-	mmc_omap_sg_to_buf(host);
+	if (data->flags & MMC_DATA_READ)
+		miter_flags |= SG_MITER_TO_SG;
+	else
+		miter_flags |= SG_MITER_FROM_SG;
+	sg_miter_start(&host->sg_miter, data->sg, data->sg_len, miter_flags);
 	host->dma_in_use = 0;
 }
 
diff --git a/drivers/mmc/host/renesas_sdhi.h b/drivers/mmc/host/renesas_sdhi.h
index c1fb9740ea..586f94d4db 100644
--- a/drivers/mmc/host/renesas_sdhi.h
+++ b/drivers/mmc/host/renesas_sdhi.h
@@ -9,6 +9,7 @@
 #ifndef RENESAS_SDHI_H
 #define RENESAS_SDHI_H
 
+#include <linux/dmaengine.h>
 #include <linux/platform_device.h>
 #include "tmio_mmc.h"
 
@@ -63,7 +64,7 @@ struct renesas_sdhi_of_data_with_quirks {
 struct renesas_sdhi_dma {
 	unsigned long end_flags;
 	enum dma_slave_buswidth dma_buswidth;
-	bool (*filter)(struct dma_chan *chan, void *arg);
+	dma_filter_fn filter;
 	void (*enable)(struct tmio_mmc_host *host, bool enable);
 	struct completion dma_dataend;
 	struct tasklet_struct dma_complete;
diff --git a/drivers/mmc/host/sdhci-acpi.c b/drivers/mmc/host/sdhci-acpi.c
index acf5fc3ad7..eb8f427f97 100644
--- a/drivers/mmc/host/sdhci-acpi.c
+++ b/drivers/mmc/host/sdhci-acpi.c
@@ -10,6 +10,7 @@
 #include <linux/export.h>
 #include <linux/module.h>
 #include <linux/device.h>
+#include <linux/pinctrl/pinconf-generic.h>
 #include <linux/platform_device.h>
 #include <linux/ioport.h>
 #include <linux/io.h>
@@ -80,6 +81,8 @@ struct sdhci_acpi_host {
 enum {
 	DMI_QUIRK_RESET_SD_SIGNAL_VOLT_ON_SUSP			= BIT(0),
 	DMI_QUIRK_SD_NO_WRITE_PROTECT				= BIT(1),
+	DMI_QUIRK_SD_CD_ACTIVE_HIGH				= BIT(2),
+	DMI_QUIRK_SD_CD_ENABLE_PULL_UP				= BIT(3),
 };
 
 static inline void *sdhci_acpi_priv(struct sdhci_acpi_host *c)
@@ -719,9 +722,30 @@ static const struct acpi_device_id sdhci_acpi_ids[] = {
 };
 MODULE_DEVICE_TABLE(acpi, sdhci_acpi_ids);
 
+/* Please keep this list sorted alphabetically */
 static const struct dmi_system_id sdhci_acpi_quirks[] = {
 	{
 		/*
+		 * The Acer Aspire Switch 10 (SW5-012) microSD slot always
+		 * reports the card being write-protected even though microSD
+		 * cards do not have a write-protect switch at all.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Aspire SW5-012"),
+		},
+		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
+	},
+	{
+		/* Asus T100TA, needs pull-up for cd but DSDT GpioInt has NoPull set */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "T100TA"),
+		},
+		.driver_data = (void *)DMI_QUIRK_SD_CD_ENABLE_PULL_UP,
+	},
+	{
+		/*
 		 * The Lenovo Miix 320-10ICR has a bug in the _PS0 method of
 		 * the SHC1 ACPI device, this bug causes it to reprogram the
 		 * wrong LDO (DLDO3) to 1.8V if 1.8V modes are used and the
@@ -736,15 +760,23 @@ static const struct dmi_system_id sdhci_acpi_quirks[] = {
 	},
 	{
 		/*
-		 * The Acer Aspire Switch 10 (SW5-012) microSD slot always
-		 * reports the card being write-protected even though microSD
-		 * cards do not have a write-protect switch at all.
+		 * Lenovo Yoga Tablet 2 Pro 1380F/L (13" Android version) this
+		 * has broken WP reporting and an inverted CD signal.
+		 * Note this has more or less the same BIOS as the Lenovo Yoga
+		 * Tablet 2 830F/L or 1050F/L (8" and 10" Android), but unlike
+		 * the 830 / 1050 models which share the same mainboard this
+		 * model has a different mainboard and the inverted CD and
+		 * broken WP are unique to this board.
 		 */
 		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-			DMI_MATCH(DMI_PRODUCT_NAME, "Aspire SW5-012"),
+			DMI_MATCH(DMI_SYS_VENDOR, "Intel Corp."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "VALLEYVIEW C0 PLATFORM"),
+			DMI_MATCH(DMI_BOARD_NAME, "BYT-T FFD8"),
+			/* Full match so as to NOT match the 830/1050 BIOS */
+			DMI_MATCH(DMI_BIOS_VERSION, "BLADE_21.X64.0005.R00.1504101516"),
 		},
-		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
+		.driver_data = (void *)(DMI_QUIRK_SD_NO_WRITE_PROTECT |
+					DMI_QUIRK_SD_CD_ACTIVE_HIGH),
 	},
 	{
 		/*
@@ -757,6 +789,17 @@ static const struct dmi_system_id sdhci_acpi_quirks[] = {
 		},
 		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
 	},
+	{
+		/*
+		 * The Toshiba WT10-A's microSD slot always reports the card being
+		 * write-protected.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "TOSHIBA"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "TOSHIBA WT10-A"),
+		},
+		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
+	},
 	{} /* Terminating entry */
 };
 
@@ -866,12 +909,18 @@ static int sdhci_acpi_probe(struct platform_device *pdev)
 	if (sdhci_acpi_flag(c, SDHCI_ACPI_SD_CD)) {
 		bool v = sdhci_acpi_flag(c, SDHCI_ACPI_SD_CD_OVERRIDE_LEVEL);
 
+		if (quirks & DMI_QUIRK_SD_CD_ACTIVE_HIGH)
+			host->mmc->caps2 |= MMC_CAP2_CD_ACTIVE_HIGH;
+
 		err = mmc_gpiod_request_cd(host->mmc, NULL, 0, v, 0);
 		if (err) {
 			if (err == -EPROBE_DEFER)
 				goto err_free;
 			dev_warn(dev, "failed to setup card detect gpio\n");
 			c->use_runtime_pm = false;
+		} else if (quirks & DMI_QUIRK_SD_CD_ENABLE_PULL_UP) {
+			mmc_gpiod_set_cd_config(host->mmc,
+						PIN_CONF_PACKED(PIN_CONFIG_BIAS_PULL_UP, 20000));
 		}
 
 		if (quirks & DMI_QUIRK_RESET_SD_SIGNAL_VOLT_ON_SUSP)
diff --git a/drivers/mmc/host/sdhci-esdhc-mcf.c b/drivers/mmc/host/sdhci-esdhc-mcf.c
index a07f8333cd..c97363e2d8 100644
--- a/drivers/mmc/host/sdhci-esdhc-mcf.c
+++ b/drivers/mmc/host/sdhci-esdhc-mcf.c
@@ -299,9 +299,8 @@ static void esdhc_mcf_pltfm_set_bus_width(struct sdhci_host *host, int width)
 static void esdhc_mcf_request_done(struct sdhci_host *host,
 				   struct mmc_request *mrq)
 {
-	struct scatterlist *sg;
+	struct sg_mapping_iter sgm;
 	u32 *buffer;
-	int i;
 
 	if (!mrq->data || !mrq->data->bytes_xfered)
 		goto exit_done;
@@ -313,10 +312,13 @@ static void esdhc_mcf_request_done(struct sdhci_host *host,
 	 * On mcf5441x there is no hw sdma option/flag to select the dma
 	 * transfer endiannes. A swap after the transfer is needed.
 	 */
-	for_each_sg(mrq->data->sg, sg, mrq->data->sg_len, i) {
-		buffer = (u32 *)sg_virt(sg);
-		esdhc_mcf_buffer_swap32(buffer, sg->length);
+	sg_miter_start(&sgm, mrq->data->sg, mrq->data->sg_len,
+		       SG_MITER_ATOMIC | SG_MITER_TO_SG | SG_MITER_FROM_SG);
+	while (sg_miter_next(&sgm)) {
+		buffer = sgm.addr;
+		esdhc_mcf_buffer_swap32(buffer, sgm.length);
 	}
+	sg_miter_stop(&sgm);
 
 exit_done:
 	mmc_request_done(host->mmc, mrq);
diff --git a/drivers/mmc/host/sdhci-of-aspeed.c b/drivers/mmc/host/sdhci-of-aspeed.c
index 42d54532ca..430c1f9003 100644
--- a/drivers/mmc/host/sdhci-of-aspeed.c
+++ b/drivers/mmc/host/sdhci-of-aspeed.c
@@ -435,7 +435,7 @@ static int aspeed_sdhci_probe(struct platform_device *pdev)
 		goto err_sdhci_add;
 
 	if (dev->phase_desc)
-		mmc_of_parse_clk_phase(host->mmc, &dev->phase_map);
+		mmc_of_parse_clk_phase(&pdev->dev, &dev->phase_map);
 
 	ret = sdhci_add_host(host);
 	if (ret)
diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
index 60b42126a4..f2e4a93ed1 100644
--- a/drivers/mmc/host/sdhci-of-dwcmshc.c
+++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
@@ -52,6 +52,20 @@
 #define AT_CTRL_SWIN_TH_VAL_MASK	GENMASK(31, 24) /* bits [31:24] */
 #define AT_CTRL_SWIN_TH_VAL		0x9  /* sampling window threshold */
 
+/* Sophgo CV18XX specific Registers */
+#define CV18XX_SDHCI_MSHC_CTRL			0x00
+#define  CV18XX_EMMC_FUNC_EN			BIT(0)
+#define  CV18XX_LATANCY_1T			BIT(1)
+#define CV18XX_SDHCI_PHY_TX_RX_DLY		0x40
+#define  CV18XX_PHY_TX_DLY_MSK			GENMASK(6, 0)
+#define  CV18XX_PHY_TX_SRC_MSK			GENMASK(9, 8)
+#define  CV18XX_PHY_TX_SRC_INVERT_CLK_TX	0x1
+#define  CV18XX_PHY_RX_DLY_MSK			GENMASK(22, 16)
+#define  CV18XX_PHY_RX_SRC_MSK			GENMASK(25, 24)
+#define  CV18XX_PHY_RX_SRC_INVERT_RX_CLK	0x1
+#define CV18XX_SDHCI_PHY_CONFIG			0x4c
+#define  CV18XX_PHY_TX_BPS			BIT(0)
+
 /* Rockchip specific Registers */
 #define DWCMSHC_EMMC_DLL_CTRL		0x800
 #define DWCMSHC_EMMC_DLL_RXCLK		0x804
@@ -643,6 +657,35 @@ static void th1520_sdhci_reset(struct sdhci_host *host, u8 mask)
 	}
 }
 
+static void cv18xx_sdhci_reset(struct sdhci_host *host, u8 mask)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
+	u32 val, emmc_caps = MMC_CAP2_NO_SD | MMC_CAP2_NO_SDIO;
+
+	sdhci_reset(host, mask);
+
+	if ((host->mmc->caps2 & emmc_caps) == emmc_caps) {
+		val = sdhci_readl(host, priv->vendor_specific_area1 + CV18XX_SDHCI_MSHC_CTRL);
+		val |= CV18XX_EMMC_FUNC_EN;
+		sdhci_writel(host, val, priv->vendor_specific_area1 + CV18XX_SDHCI_MSHC_CTRL);
+	}
+
+	val = sdhci_readl(host, priv->vendor_specific_area1 + CV18XX_SDHCI_MSHC_CTRL);
+	val |= CV18XX_LATANCY_1T;
+	sdhci_writel(host, val, priv->vendor_specific_area1 + CV18XX_SDHCI_MSHC_CTRL);
+
+	val = sdhci_readl(host, priv->vendor_specific_area1 + CV18XX_SDHCI_PHY_CONFIG);
+	val |= CV18XX_PHY_TX_BPS;
+	sdhci_writel(host, val, priv->vendor_specific_area1 + CV18XX_SDHCI_PHY_CONFIG);
+
+	val =  (FIELD_PREP(CV18XX_PHY_TX_DLY_MSK, 0) |
+		FIELD_PREP(CV18XX_PHY_TX_SRC_MSK, CV18XX_PHY_TX_SRC_INVERT_CLK_TX) |
+		FIELD_PREP(CV18XX_PHY_RX_DLY_MSK, 0) |
+		FIELD_PREP(CV18XX_PHY_RX_SRC_MSK, CV18XX_PHY_RX_SRC_INVERT_RX_CLK));
+	sdhci_writel(host, val, priv->vendor_specific_area1 + CV18XX_SDHCI_PHY_TX_RX_DLY);
+}
+
 static const struct sdhci_ops sdhci_dwcmshc_ops = {
 	.set_clock		= sdhci_set_clock,
 	.set_bus_width		= sdhci_set_bus_width,
@@ -672,6 +715,15 @@ static const struct sdhci_ops sdhci_dwcmshc_th1520_ops = {
 	.platform_execute_tuning = &th1520_execute_tuning,
 };
 
+static const struct sdhci_ops sdhci_dwcmshc_cv18xx_ops = {
+	.set_clock		= sdhci_set_clock,
+	.set_bus_width		= sdhci_set_bus_width,
+	.set_uhs_signaling	= dwcmshc_set_uhs_signaling,
+	.get_max_clock		= dwcmshc_get_max_clock,
+	.reset			= cv18xx_sdhci_reset,
+	.adma_write_desc	= dwcmshc_adma_write_desc,
+};
+
 static const struct sdhci_pltfm_data sdhci_dwcmshc_pdata = {
 	.ops = &sdhci_dwcmshc_ops,
 	.quirks = SDHCI_QUIRK_CAP_CLOCK_BASE_BROKEN,
@@ -701,6 +753,12 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_th1520_pdata = {
 	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN,
 };
 
+static const struct sdhci_pltfm_data sdhci_dwcmshc_cv18xx_pdata = {
+	.ops = &sdhci_dwcmshc_cv18xx_ops,
+	.quirks = SDHCI_QUIRK_CAP_CLOCK_BASE_BROKEN,
+	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN,
+};
+
 static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
 {
 	int err;
@@ -770,6 +828,14 @@ static const struct of_device_id sdhci_dwcmshc_dt_ids[] = {
 		.data = &sdhci_dwcmshc_pdata,
 	},
 	{
+		.compatible = "sophgo,cv1800b-dwcmshc",
+		.data = &sdhci_dwcmshc_cv18xx_pdata,
+	},
+	{
+		.compatible = "sophgo,sg2002-dwcmshc",
+		.data = &sdhci_dwcmshc_cv18xx_pdata,
+	},
+	{
 		.compatible = "thead,th1520-dwcmshc",
 		.data = &sdhci_dwcmshc_th1520_pdata,
 	},
diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index c79f734599..746f4cf7ab 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -3439,12 +3439,18 @@ static void sdhci_data_irq(struct sdhci_host *host, u32 intmask)
 		host->data->error = -EILSEQ;
 		if (!mmc_op_tuning(SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND))))
 			sdhci_err_stats_inc(host, DAT_CRC);
-	} else if ((intmask & SDHCI_INT_DATA_CRC) &&
+	} else if ((intmask & (SDHCI_INT_DATA_CRC | SDHCI_INT_TUNING_ERROR)) &&
 		SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND))
 			!= MMC_BUS_TEST_R) {
 		host->data->error = -EILSEQ;
 		if (!mmc_op_tuning(SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND))))
 			sdhci_err_stats_inc(host, DAT_CRC);
+		if (intmask & SDHCI_INT_TUNING_ERROR) {
+			u16 ctrl2 = sdhci_readw(host, SDHCI_HOST_CONTROL2);
+
+			ctrl2 &= ~SDHCI_CTRL_TUNED_CLK;
+			sdhci_writew(host, ctrl2, SDHCI_HOST_CONTROL2);
+		}
 	} else if (intmask & SDHCI_INT_ADMA_ERROR) {
 		pr_err("%s: ADMA error: 0x%08x\n", mmc_hostname(host->mmc),
 		       intmask);
@@ -3979,7 +3985,7 @@ bool sdhci_cqe_irq(struct sdhci_host *host, u32 intmask, int *cmd_error,
 	} else
 		*cmd_error = 0;
 
-	if (intmask & (SDHCI_INT_DATA_END_BIT | SDHCI_INT_DATA_CRC)) {
+	if (intmask & (SDHCI_INT_DATA_END_BIT | SDHCI_INT_DATA_CRC | SDHCI_INT_TUNING_ERROR)) {
 		*data_error = -EILSEQ;
 		if (!mmc_op_tuning(SDHCI_GET_CMD(sdhci_readw(host, SDHCI_COMMAND))))
 			sdhci_err_stats_inc(host, DAT_CRC);
diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h
index a20864fc06..957c7a917f 100644
--- a/drivers/mmc/host/sdhci.h
+++ b/drivers/mmc/host/sdhci.h
@@ -158,6 +158,7 @@
 #define  SDHCI_INT_BUS_POWER	0x00800000
 #define  SDHCI_INT_AUTO_CMD_ERR	0x01000000
 #define  SDHCI_INT_ADMA_ERROR	0x02000000
+#define  SDHCI_INT_TUNING_ERROR	0x04000000
 
 #define  SDHCI_INT_NORMAL_MASK	0x00007FFF
 #define  SDHCI_INT_ERROR_MASK	0xFFFF8000
@@ -169,7 +170,7 @@
 		SDHCI_INT_DATA_AVAIL | SDHCI_INT_SPACE_AVAIL | \
 		SDHCI_INT_DATA_TIMEOUT | SDHCI_INT_DATA_CRC | \
 		SDHCI_INT_DATA_END_BIT | SDHCI_INT_ADMA_ERROR | \
-		SDHCI_INT_BLK_GAP)
+		SDHCI_INT_BLK_GAP | SDHCI_INT_TUNING_ERROR)
 #define SDHCI_INT_ALL_MASK	((unsigned int)-1)
 
 #define SDHCI_CQE_INT_ERR_MASK ( \
diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c
index d659c59422..562034af65 100644
--- a/drivers/mmc/host/sdhci_am654.c
+++ b/drivers/mmc/host/sdhci_am654.c
@@ -143,16 +143,24 @@ struct sdhci_am654_data {
 	struct regmap *base;
 	int otap_del_sel[ARRAY_SIZE(td)];
 	int itap_del_sel[ARRAY_SIZE(td)];
+	u32 itap_del_ena[ARRAY_SIZE(td)];
 	int clkbuf_sel;
 	int trm_icp;
 	int drv_strength;
 	int strb_sel;
 	u32 flags;
 	u32 quirks;
+	bool dll_enable;
 
 #define SDHCI_AM654_QUIRK_FORCE_CDTEST BIT(0)
 };
 
+struct window {
+	u8 start;
+	u8 end;
+	u8 length;
+};
+
 struct sdhci_am654_driver_data {
 	const struct sdhci_pltfm_data *pdata;
 	u32 flags;
@@ -232,11 +240,13 @@ static void sdhci_am654_setup_dll(struct sdhci_host *host, unsigned int clock)
 }
 
 static void sdhci_am654_write_itapdly(struct sdhci_am654_data *sdhci_am654,
-				      u32 itapdly)
+				      u32 itapdly, u32 enable)
 {
 	/* Set ITAPCHGWIN before writing to ITAPDLY */
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPCHGWIN_MASK,
 			   1 << ITAPCHGWIN_SHIFT);
+	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPDLYENA_MASK,
+			   enable << ITAPDLYENA_SHIFT);
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPDLYSEL_MASK,
 			   itapdly << ITAPDLYSEL_SHIFT);
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPCHGWIN_MASK, 0);
@@ -253,8 +263,8 @@ static void sdhci_am654_setup_delay_chain(struct sdhci_am654_data *sdhci_am654,
 	mask = SELDLYTXCLK_MASK | SELDLYRXCLK_MASK;
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL5, mask, val);
 
-	sdhci_am654_write_itapdly(sdhci_am654,
-				  sdhci_am654->itap_del_sel[timing]);
+	sdhci_am654_write_itapdly(sdhci_am654, sdhci_am654->itap_del_sel[timing],
+				  sdhci_am654->itap_del_ena[timing]);
 }
 
 static void sdhci_am654_set_clock(struct sdhci_host *host, unsigned int clock)
@@ -263,7 +273,6 @@ static void sdhci_am654_set_clock(struct sdhci_host *host, unsigned int clock)
 	struct sdhci_am654_data *sdhci_am654 = sdhci_pltfm_priv(pltfm_host);
 	unsigned char timing = host->mmc->ios.timing;
 	u32 otap_del_sel;
-	u32 otap_del_ena;
 	u32 mask, val;
 
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL1, ENDLL_MASK, 0);
@@ -272,10 +281,9 @@ static void sdhci_am654_set_clock(struct sdhci_host *host, unsigned int clock)
 
 	/* Setup DLL Output TAP delay */
 	otap_del_sel = sdhci_am654->otap_del_sel[timing];
-	otap_del_ena = (timing > MMC_TIMING_UHS_SDR25) ? 1 : 0;
 
 	mask = OTAPDLYENA_MASK | OTAPDLYSEL_MASK;
-	val = (otap_del_ena << OTAPDLYENA_SHIFT) |
+	val = (0x1 << OTAPDLYENA_SHIFT) |
 	      (otap_del_sel << OTAPDLYSEL_SHIFT);
 
 	/* Write to STRBSEL for HS400 speed mode */
@@ -290,10 +298,21 @@ static void sdhci_am654_set_clock(struct sdhci_host *host, unsigned int clock)
 
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, mask, val);
 
-	if (timing > MMC_TIMING_UHS_SDR25 && clock >= CLOCK_TOO_SLOW_HZ)
+	if (timing > MMC_TIMING_UHS_SDR25 && clock >= CLOCK_TOO_SLOW_HZ) {
 		sdhci_am654_setup_dll(host, clock);
-	else
+		sdhci_am654->dll_enable = true;
+
+		if (timing == MMC_TIMING_MMC_HS400) {
+			sdhci_am654->itap_del_ena[timing] = 0x1;
+			sdhci_am654->itap_del_sel[timing] = sdhci_am654->itap_del_sel[timing - 1];
+		}
+
+		sdhci_am654_write_itapdly(sdhci_am654, sdhci_am654->itap_del_sel[timing],
+					  sdhci_am654->itap_del_ena[timing]);
+	} else {
 		sdhci_am654_setup_delay_chain(sdhci_am654, timing);
+		sdhci_am654->dll_enable = false;
+	}
 
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL5, CLKBUFSEL_MASK,
 			   sdhci_am654->clkbuf_sel);
@@ -306,6 +325,8 @@ static void sdhci_j721e_4bit_set_clock(struct sdhci_host *host,
 	struct sdhci_am654_data *sdhci_am654 = sdhci_pltfm_priv(pltfm_host);
 	unsigned char timing = host->mmc->ios.timing;
 	u32 otap_del_sel;
+	u32 itap_del_ena;
+	u32 itap_del_sel;
 	u32 mask, val;
 
 	/* Setup DLL Output TAP delay */
@@ -314,8 +335,19 @@ static void sdhci_j721e_4bit_set_clock(struct sdhci_host *host,
 	mask = OTAPDLYENA_MASK | OTAPDLYSEL_MASK;
 	val = (0x1 << OTAPDLYENA_SHIFT) |
 	      (otap_del_sel << OTAPDLYSEL_SHIFT);
-	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, mask, val);
 
+	/* Setup Input TAP delay */
+	itap_del_ena = sdhci_am654->itap_del_ena[timing];
+	itap_del_sel = sdhci_am654->itap_del_sel[timing];
+
+	mask |= ITAPDLYENA_MASK | ITAPDLYSEL_MASK;
+	val |= (itap_del_ena << ITAPDLYENA_SHIFT) |
+	       (itap_del_sel << ITAPDLYSEL_SHIFT);
+
+	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPCHGWIN_MASK,
+			   1 << ITAPCHGWIN_SHIFT);
+	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, mask, val);
+	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPCHGWIN_MASK, 0);
 	regmap_update_bits(sdhci_am654->base, PHY_CTRL5, CLKBUFSEL_MASK,
 			   sdhci_am654->clkbuf_sel);
 
@@ -408,40 +440,105 @@ static u32 sdhci_am654_cqhci_irq(struct sdhci_host *host, u32 intmask)
 	return 0;
 }
 
-#define ITAP_MAX	32
+#define ITAPDLY_LENGTH 32
+#define ITAPDLY_LAST_INDEX (ITAPDLY_LENGTH - 1)
+
+static u32 sdhci_am654_calculate_itap(struct sdhci_host *host, struct window
+			  *fail_window, u8 num_fails, bool circular_buffer)
+{
+	u8 itap = 0, start_fail = 0, end_fail = 0, pass_length = 0;
+	u8 first_fail_start = 0, last_fail_end = 0;
+	struct device *dev = mmc_dev(host->mmc);
+	struct window pass_window = {0, 0, 0};
+	int prev_fail_end = -1;
+	u8 i;
+
+	if (!num_fails)
+		return ITAPDLY_LAST_INDEX >> 1;
+
+	if (fail_window->length == ITAPDLY_LENGTH) {
+		dev_err(dev, "No passing ITAPDLY, return 0\n");
+		return 0;
+	}
+
+	first_fail_start = fail_window->start;
+	last_fail_end = fail_window[num_fails - 1].end;
+
+	for (i = 0; i < num_fails; i++) {
+		start_fail = fail_window[i].start;
+		end_fail = fail_window[i].end;
+		pass_length = start_fail - (prev_fail_end + 1);
+
+		if (pass_length > pass_window.length) {
+			pass_window.start = prev_fail_end + 1;
+			pass_window.length = pass_length;
+		}
+		prev_fail_end = end_fail;
+	}
+
+	if (!circular_buffer)
+		pass_length = ITAPDLY_LAST_INDEX - last_fail_end;
+	else
+		pass_length = ITAPDLY_LAST_INDEX - last_fail_end + first_fail_start;
+
+	if (pass_length > pass_window.length) {
+		pass_window.start = last_fail_end + 1;
+		pass_window.length = pass_length;
+	}
+
+	if (!circular_buffer)
+		itap = pass_window.start + (pass_window.length >> 1);
+	else
+		itap = (pass_window.start + (pass_window.length >> 1)) % ITAPDLY_LENGTH;
+
+	return (itap > ITAPDLY_LAST_INDEX) ? ITAPDLY_LAST_INDEX >> 1 : itap;
+}
+
 static int sdhci_am654_platform_execute_tuning(struct sdhci_host *host,
 					       u32 opcode)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	struct sdhci_am654_data *sdhci_am654 = sdhci_pltfm_priv(pltfm_host);
-	int cur_val, prev_val = 1, fail_len = 0, pass_window = 0, pass_len;
-	u32 itap;
+	unsigned char timing = host->mmc->ios.timing;
+	struct window fail_window[ITAPDLY_LENGTH];
+	u8 curr_pass, itap;
+	u8 fail_index = 0;
+	u8 prev_pass = 1;
+
+	memset(fail_window, 0, sizeof(fail_window));
 
 	/* Enable ITAPDLY */
-	regmap_update_bits(sdhci_am654->base, PHY_CTRL4, ITAPDLYENA_MASK,
-			   1 << ITAPDLYENA_SHIFT);
+	sdhci_am654->itap_del_ena[timing] = 0x1;
+
+	for (itap = 0; itap < ITAPDLY_LENGTH; itap++) {
+		sdhci_am654_write_itapdly(sdhci_am654, itap, sdhci_am654->itap_del_ena[timing]);
 
-	for (itap = 0; itap < ITAP_MAX; itap++) {
-		sdhci_am654_write_itapdly(sdhci_am654, itap);
+		curr_pass = !mmc_send_tuning(host->mmc, opcode, NULL);
 
-		cur_val = !mmc_send_tuning(host->mmc, opcode, NULL);
-		if (cur_val && !prev_val)
-			pass_window = itap;
+		if (!curr_pass && prev_pass)
+			fail_window[fail_index].start = itap;
 
-		if (!cur_val)
-			fail_len++;
+		if (!curr_pass) {
+			fail_window[fail_index].end = itap;
+			fail_window[fail_index].length++;
+		}
+
+		if (curr_pass && !prev_pass)
+			fail_index++;
 
-		prev_val = cur_val;
+		prev_pass = curr_pass;
 	}
-	/*
-	 * Having determined the length of the failing window and start of
-	 * the passing window calculate the length of the passing window and
-	 * set the final value halfway through it considering the range as a
-	 * circular buffer
-	 */
-	pass_len = ITAP_MAX - fail_len;
-	itap = (pass_window + (pass_len >> 1)) % ITAP_MAX;
-	sdhci_am654_write_itapdly(sdhci_am654, itap);
+
+	if (fail_window[fail_index].length != 0)
+		fail_index++;
+
+	itap = sdhci_am654_calculate_itap(host, fail_window, fail_index,
+					  sdhci_am654->dll_enable);
+
+	sdhci_am654_write_itapdly(sdhci_am654, itap, sdhci_am654->itap_del_ena[timing]);
+
+	/* Save ITAPDLY */
+	sdhci_am654->itap_del_sel[timing] = itap;
 
 	return 0;
 }
@@ -590,9 +687,12 @@ static int sdhci_am654_get_otap_delay(struct sdhci_host *host,
 				host->mmc->caps2 &= ~td[i].capability;
 		}
 
-		if (td[i].itap_binding)
-			device_property_read_u32(dev, td[i].itap_binding,
-						 &sdhci_am654->itap_del_sel[i]);
+		if (td[i].itap_binding) {
+			ret = device_property_read_u32(dev, td[i].itap_binding,
+						       &sdhci_am654->itap_del_sel[i]);
+			if (!ret)
+				sdhci_am654->itap_del_ena[i] = 0x1;
+		}
 	}
 
 	return 0;
diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c
index 077d711e96..08b4312af9 100644
--- a/drivers/mmc/host/sh_mmcif.c
+++ b/drivers/mmc/host/sh_mmcif.c
@@ -227,14 +227,12 @@ struct sh_mmcif_host {
 	bool dying;
 	long timeout;
 	void __iomem *addr;
-	u32 *pio_ptr;
 	spinlock_t lock;		/* protect sh_mmcif_host::state */
 	enum sh_mmcif_state state;
 	enum sh_mmcif_wait_for wait_for;
 	struct delayed_work timeout_work;
 	size_t blocksize;
-	int sg_idx;
-	int sg_blkidx;
+	struct sg_mapping_iter sg_miter;
 	bool power;
 	bool ccs_enable;		/* Command Completion Signal support */
 	bool clk_ctrl2_enable;
@@ -600,32 +598,17 @@ static int sh_mmcif_error_manage(struct sh_mmcif_host *host)
 	return ret;
 }
 
-static bool sh_mmcif_next_block(struct sh_mmcif_host *host, u32 *p)
-{
-	struct mmc_data *data = host->mrq->data;
-
-	host->sg_blkidx += host->blocksize;
-
-	/* data->sg->length must be a multiple of host->blocksize? */
-	BUG_ON(host->sg_blkidx > data->sg->length);
-
-	if (host->sg_blkidx == data->sg->length) {
-		host->sg_blkidx = 0;
-		if (++host->sg_idx < data->sg_len)
-			host->pio_ptr = sg_virt(++data->sg);
-	} else {
-		host->pio_ptr = p;
-	}
-
-	return host->sg_idx != data->sg_len;
-}
-
 static void sh_mmcif_single_read(struct sh_mmcif_host *host,
 				 struct mmc_request *mrq)
 {
+	struct mmc_data *data = mrq->data;
+
 	host->blocksize = (sh_mmcif_readl(host->addr, MMCIF_CE_BLOCK_SET) &
 			   BLOCK_SIZE_MASK) + 3;
 
+	sg_miter_start(&host->sg_miter, data->sg, data->sg_len,
+		       SG_MITER_TO_SG);
+
 	host->wait_for = MMCIF_WAIT_FOR_READ;
 
 	/* buf read enable */
@@ -634,20 +617,32 @@ static void sh_mmcif_single_read(struct sh_mmcif_host *host,
 
 static bool sh_mmcif_read_block(struct sh_mmcif_host *host)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct device *dev = sh_mmcif_host_to_dev(host);
 	struct mmc_data *data = host->mrq->data;
-	u32 *p = sg_virt(data->sg);
+	u32 *p;
 	int i;
 
 	if (host->sd_error) {
+		sg_miter_stop(sgm);
 		data->error = sh_mmcif_error_manage(host);
 		dev_dbg(dev, "%s(): %d\n", __func__, data->error);
 		return false;
 	}
 
+	if (!sg_miter_next(sgm)) {
+		/* This should not happen on single blocks */
+		sg_miter_stop(sgm);
+		return false;
+	}
+
+	p = sgm->addr;
+
 	for (i = 0; i < host->blocksize / 4; i++)
 		*p++ = sh_mmcif_readl(host->addr, MMCIF_CE_DATA);
 
+	sg_miter_stop(&host->sg_miter);
+
 	/* buffer read end */
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MBUFRE);
 	host->wait_for = MMCIF_WAIT_FOR_READ_END;
@@ -658,6 +653,7 @@ static bool sh_mmcif_read_block(struct sh_mmcif_host *host)
 static void sh_mmcif_multi_read(struct sh_mmcif_host *host,
 				struct mmc_request *mrq)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct mmc_data *data = mrq->data;
 
 	if (!data->sg_len || !data->sg->length)
@@ -666,46 +662,63 @@ static void sh_mmcif_multi_read(struct sh_mmcif_host *host,
 	host->blocksize = sh_mmcif_readl(host->addr, MMCIF_CE_BLOCK_SET) &
 		BLOCK_SIZE_MASK;
 
+	sg_miter_start(sgm, data->sg, data->sg_len,
+		       SG_MITER_TO_SG);
+
+	/* Advance to the first sglist entry */
+	if (!sg_miter_next(sgm)) {
+		sg_miter_stop(sgm);
+		return;
+	}
+
 	host->wait_for = MMCIF_WAIT_FOR_MREAD;
-	host->sg_idx = 0;
-	host->sg_blkidx = 0;
-	host->pio_ptr = sg_virt(data->sg);
 
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MBUFREN);
 }
 
 static bool sh_mmcif_mread_block(struct sh_mmcif_host *host)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct device *dev = sh_mmcif_host_to_dev(host);
 	struct mmc_data *data = host->mrq->data;
-	u32 *p = host->pio_ptr;
+	u32 *p;
 	int i;
 
 	if (host->sd_error) {
+		sg_miter_stop(sgm);
 		data->error = sh_mmcif_error_manage(host);
 		dev_dbg(dev, "%s(): %d\n", __func__, data->error);
 		return false;
 	}
 
-	BUG_ON(!data->sg->length);
+	p = sgm->addr;
 
 	for (i = 0; i < host->blocksize / 4; i++)
 		*p++ = sh_mmcif_readl(host->addr, MMCIF_CE_DATA);
 
-	if (!sh_mmcif_next_block(host, p))
-		return false;
+	sgm->consumed = host->blocksize;
 
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MBUFREN);
 
+	if (!sg_miter_next(sgm)) {
+		sg_miter_stop(sgm);
+		return false;
+	}
+
 	return true;
 }
 
 static void sh_mmcif_single_write(struct sh_mmcif_host *host,
 					struct mmc_request *mrq)
 {
+	struct mmc_data *data = mrq->data;
+
 	host->blocksize = (sh_mmcif_readl(host->addr, MMCIF_CE_BLOCK_SET) &
 			   BLOCK_SIZE_MASK) + 3;
 
+	sg_miter_start(&host->sg_miter, data->sg, data->sg_len,
+		       SG_MITER_FROM_SG);
+
 	host->wait_for = MMCIF_WAIT_FOR_WRITE;
 
 	/* buf write enable */
@@ -714,20 +727,32 @@ static void sh_mmcif_single_write(struct sh_mmcif_host *host,
 
 static bool sh_mmcif_write_block(struct sh_mmcif_host *host)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct device *dev = sh_mmcif_host_to_dev(host);
 	struct mmc_data *data = host->mrq->data;
-	u32 *p = sg_virt(data->sg);
+	u32 *p;
 	int i;
 
 	if (host->sd_error) {
+		sg_miter_stop(sgm);
 		data->error = sh_mmcif_error_manage(host);
 		dev_dbg(dev, "%s(): %d\n", __func__, data->error);
 		return false;
 	}
 
+	if (!sg_miter_next(sgm)) {
+		/* This should not happen on single blocks */
+		sg_miter_stop(sgm);
+		return false;
+	}
+
+	p = sgm->addr;
+
 	for (i = 0; i < host->blocksize / 4; i++)
 		sh_mmcif_writel(host->addr, MMCIF_CE_DATA, *p++);
 
+	sg_miter_stop(&host->sg_miter);
+
 	/* buffer write end */
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MDTRANE);
 	host->wait_for = MMCIF_WAIT_FOR_WRITE_END;
@@ -738,6 +763,7 @@ static bool sh_mmcif_write_block(struct sh_mmcif_host *host)
 static void sh_mmcif_multi_write(struct sh_mmcif_host *host,
 				struct mmc_request *mrq)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct mmc_data *data = mrq->data;
 
 	if (!data->sg_len || !data->sg->length)
@@ -746,34 +772,46 @@ static void sh_mmcif_multi_write(struct sh_mmcif_host *host,
 	host->blocksize = sh_mmcif_readl(host->addr, MMCIF_CE_BLOCK_SET) &
 		BLOCK_SIZE_MASK;
 
+	sg_miter_start(sgm, data->sg, data->sg_len,
+		       SG_MITER_FROM_SG);
+
+	/* Advance to the first sglist entry */
+	if (!sg_miter_next(sgm)) {
+		sg_miter_stop(sgm);
+		return;
+	}
+
 	host->wait_for = MMCIF_WAIT_FOR_MWRITE;
-	host->sg_idx = 0;
-	host->sg_blkidx = 0;
-	host->pio_ptr = sg_virt(data->sg);
 
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MBUFWEN);
 }
 
 static bool sh_mmcif_mwrite_block(struct sh_mmcif_host *host)
 {
+	struct sg_mapping_iter *sgm = &host->sg_miter;
 	struct device *dev = sh_mmcif_host_to_dev(host);
 	struct mmc_data *data = host->mrq->data;
-	u32 *p = host->pio_ptr;
+	u32 *p;
 	int i;
 
 	if (host->sd_error) {
+		sg_miter_stop(sgm);
 		data->error = sh_mmcif_error_manage(host);
 		dev_dbg(dev, "%s(): %d\n", __func__, data->error);
 		return false;
 	}
 
-	BUG_ON(!data->sg->length);
+	p = sgm->addr;
 
 	for (i = 0; i < host->blocksize / 4; i++)
 		sh_mmcif_writel(host->addr, MMCIF_CE_DATA, *p++);
 
-	if (!sh_mmcif_next_block(host, p))
+	sgm->consumed = host->blocksize;
+
+	if (!sg_miter_next(sgm)) {
+		sg_miter_stop(sgm);
 		return false;
+	}
 
 	sh_mmcif_bitset(host, MMCIF_CE_INT_MASK, MASK_MBUFWEN);
 
diff --git a/drivers/mmc/host/tmio_mmc_core.c b/drivers/mmc/host/tmio_mmc_core.c
index c253d176db..93e912afd3 100644
--- a/drivers/mmc/host/tmio_mmc_core.c
+++ b/drivers/mmc/host/tmio_mmc_core.c
@@ -972,6 +972,7 @@ static void tmio_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		return;
 	}
 
+	/* Disallow new mrqs and work handlers to run */
 	host->mrq = ERR_PTR(-EBUSY);
 
 	spin_unlock_irqrestore(&host->lock, flags);
@@ -1006,8 +1007,9 @@ static void tmio_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 			"%s.%d: IOS interrupted: clk %u, mode %u",
 			current->comm, task_pid_nr(current),
 			ios->clock, ios->power_mode);
-	host->mrq = NULL;
 
+	/* Ready for new mrqs */
+	host->mrq = NULL;
 	host->clk_cache = ios->clock;
 
 	mutex_unlock(&host->ios_lock);
diff --git a/drivers/mmc/host/wbsd.c b/drivers/mmc/host/wbsd.c
index 001a468bc1..f0562f712d 100644
--- a/drivers/mmc/host/wbsd.c
+++ b/drivers/mmc/host/wbsd.c
@@ -1284,8 +1284,6 @@ static int wbsd_scan(struct wbsd_host *host)
 			continue;
 
 		for (j = 0; j < ARRAY_SIZE(unlock_codes); j++) {
-			id = 0xFFFF;
-
 			host->config = config_ports[i];
 			host->unlock_code = unlock_codes[j];