1 files changed, 11 insertions, 11 deletions
diff --git a/drivers/net/ethernet/hp/hp100.c b/drivers/net/ethernet/hp/hp100.c
index 3f4391bede8..37860096f74 100644
--- a/drivers/net/ethernet/hp/hp100.c
+++ b/drivers/net/ethernet/hp/hp100.c
@@ -308,7 +308,7 @@ static void wait(void)
  * Read board id and convert to string.
  * Effectively same code as decode_eisa_sig
  */
-static __devinit const char *hp100_read_id(int ioaddr)
+static const char *hp100_read_id(int ioaddr)
 {
 	int i;
 	static char str[HP100_SIG_LEN];
@@ -447,8 +447,8 @@ static const struct net_device_ops hp100_netdev_ops = {
 	.ndo_validate_addr	= eth_validate_addr,
 };
 
-static int __devinit hp100_probe1(struct net_device *dev, int ioaddr,
-				  u_char bus, struct pci_dev *pci_dev)
+static int hp100_probe1(struct net_device *dev, int ioaddr, u_char bus,
+			struct pci_dev *pci_dev)
 {
 	int i;
 	int err = -ENODEV;
@@ -1097,8 +1097,8 @@ static int hp100_open(struct net_device *dev)
 	/* New: if bus is PCI or EISA, interrupts might be shared interrupts */
 	if (request_irq(dev->irq, hp100_interrupt,
 			lp->bus == HP100_BUS_PCI || lp->bus ==
-			HP100_BUS_EISA ? IRQF_SHARED : IRQF_DISABLED,
-			"hp100", dev)) {
+			HP100_BUS_EISA ? IRQF_SHARED : 0,
+			dev->name, dev)) {
 		printk("hp100: %s: unable to get IRQ %d\n", dev->name, dev->irq);
 		return -EAGAIN;
 	}
@@ -2866,7 +2866,7 @@ static int __init hp100_eisa_probe (struct device *gendev)
 	return err;
 }
 
-static int __devexit hp100_eisa_remove (struct device *gendev)
+static int hp100_eisa_remove(struct device *gendev)
 {
 	struct net_device *dev = dev_get_drvdata(gendev);
 	cleanup_dev(dev);
@@ -2878,14 +2878,14 @@ static struct eisa_driver hp100_eisa_driver = {
         .driver   = {
                 .name    = "hp100",
                 .probe   = hp100_eisa_probe,
-                .remove  = __devexit_p (hp100_eisa_remove),
+		.remove  = hp100_eisa_remove,
         }
 };
 #endif
 
 #ifdef CONFIG_PCI
-static int __devinit hp100_pci_probe (struct pci_dev *pdev,
-				     const struct pci_device_id *ent)
+static int hp100_pci_probe(struct pci_dev *pdev,
+			   const struct pci_device_id *ent)
 {
 	struct net_device *dev;
 	int ioaddr;
@@ -2937,7 +2937,7 @@ static int __devinit hp100_pci_probe (struct pci_dev *pdev,
 	return err;
 }
 
-static void __devexit hp100_pci_remove (struct pci_dev *pdev)
+static void hp100_pci_remove(struct pci_dev *pdev)
 {
 	struct net_device *dev = pci_get_drvdata(pdev);
 
@@ -2950,7 +2950,7 @@ static struct pci_driver hp100_pci_driver = {
 	.name		= "hp100",
 	.id_table	= hp100_pci_tbl,
 	.probe		= hp100_pci_probe,
-	.remove		= __devexit_p(hp100_pci_remove),
+	.remove		= hp100_pci_remove,
 };
 #endif
 
diff --git a/drivers/mmc/Makefile b/drivers/mmc/Makefile
index 9979f5e9765..400756ec7c4 100644
--- a/drivers/mmc/Makefile
+++ b/drivers/mmc/Makefile
@@ -2,11 +2,8 @@
 # Makefile for the kernel mmc device drivers.
 #
 
-ifeq ($(CONFIG_MMC_DEBUG),y)
-	EXTRA_CFLAGS		+= -DDEBUG
-endif
+subdir-ccflags-$(CONFIG_MMC_DEBUG) := -DDEBUG
 
 obj-$(CONFIG_MMC)		+= core/
 obj-$(CONFIG_MMC)		+= card/
-obj-$(CONFIG_MMC)		+= host/
-
+obj-$(subst m,y,$(CONFIG_MMC))	+= host/
diff --git a/drivers/mmc/card/Kconfig b/drivers/mmc/card/Kconfig
index 3f2a912659a..5562308699b 100644
--- a/drivers/mmc/card/Kconfig
+++ b/drivers/mmc/card/Kconfig
@@ -14,6 +14,24 @@ config MMC_BLOCK
 	  mount the filesystem. Almost everyone wishing MMC support
 	  should say Y or M here.
 
+config MMC_BLOCK_MINORS
+	int "Number of minors per block device"
+	depends on MMC_BLOCK
+	range 4 256
+	default 8
+	help
+	  Number of minors per block device. One is needed for every
+	  partition on the disk (plus one for the whole disk).
+
+	  Number of total MMC minors available is 256, so your number
+	  of supported block devices will be limited to 256 divided
+	  by this number.
+
+	  Default is 8 to be backwards compatible with previous
+	  hardwired device numbering.
+
+	  If unsure, say 8 here.
+
 config MMC_BLOCK_BOUNCE
 	bool "Use bounce buffer for simple hosts"
 	depends on MMC_BLOCK
@@ -34,18 +52,18 @@ config MMC_BLOCK_BOUNCE
 
 config SDIO_UART
 	tristate "SDIO UART/GPS class support"
+	depends on TTY
 	help
 	  SDIO function driver for SDIO cards that implements the UART
 	  class, as well as the GPS class which appears like a UART.
 
 config MMC_TEST
 	tristate "MMC host test driver"
-	default n
 	help
 	  Development driver that performs a series of reads and writes
 	  to a memory card in order to expose certain well known bugs
 	  in host controllers. The tests are executed by writing to the
-	  "test" file in sysfs under each card. Note that whatever is
+	  "test" file in debugfs under each card. Note that whatever is
 	  on your card will be overwritten by these tests.
 
 	  This driver is only of interest to those developing or
diff --git a/drivers/mmc/card/Makefile b/drivers/mmc/card/Makefile
index 0d407514f67..c73b406a06c 100644
--- a/drivers/mmc/card/Makefile
+++ b/drivers/mmc/card/Makefile
@@ -2,10 +2,6 @@
 # Makefile for MMC/SD card drivers
 #
 
-ifeq ($(CONFIG_MMC_DEBUG),y)
-	EXTRA_CFLAGS		+= -DDEBUG
-endif
-
 obj-$(CONFIG_MMC_BLOCK)		+= mmc_block.o
 mmc_block-objs			:= block.o queue.o
 obj-$(CONFIG_MMC_TEST)		+= mmc_test.o
diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c
index fe8041e619e..452782bffeb 100644
--- a/drivers/mmc/card/block.c
+++ b/drivers/mmc/card/block.c
@@ -23,6 +23,7 @@
 
 #include <linux/kernel.h>
 #include <linux/fs.h>
+#include <linux/slab.h>
 #include <linux/errno.h>
 #include <linux/hdreg.h>
 #include <linux/kdev_t.h>
@@ -30,26 +31,60 @@
 #include <linux/mutex.h>
 #include <linux/scatterlist.h>
 #include <linux/string_helpers.h>
+#include <linux/delay.h>
+#include <linux/capability.h>
+#include <linux/compat.h>
+#include <linux/pm_runtime.h>
 
+#include <linux/mmc/ioctl.h>
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
 #include <linux/mmc/sd.h>
 
-#include <asm/system.h>
 #include <asm/uaccess.h>
 
 #include "queue.h"
 
 MODULE_ALIAS("mmc:block");
+#ifdef MODULE_PARAM_PREFIX
+#undef MODULE_PARAM_PREFIX
+#endif
+#define MODULE_PARAM_PREFIX "mmcblk."
+
+#define INAND_CMD38_ARG_EXT_CSD  113
+#define INAND_CMD38_ARG_ERASE    0x00
+#define INAND_CMD38_ARG_TRIM     0x01
+#define INAND_CMD38_ARG_SECERASE 0x80
+#define INAND_CMD38_ARG_SECTRIM1 0x81
+#define INAND_CMD38_ARG_SECTRIM2 0x88
+#define MMC_BLK_TIMEOUT_MS  (10 * 60 * 1000)        /* 10 minute timeout */
+#define MMC_SANITIZE_REQ_TIMEOUT 240000
+#define MMC_EXTRACT_INDEX_FROM_ARG(x) ((x & 0x00FF0000) >> 16)
+
+#define mmc_req_rel_wr(req)	(((req->cmd_flags & REQ_FUA) || \
+				  (req->cmd_flags & REQ_META)) && \
+				  (rq_data_dir(req) == WRITE))
+#define PACKED_CMD_VER	0x01
+#define PACKED_CMD_WR	0x02
+
+static DEFINE_MUTEX(block_mutex);
+
+/*
+ * The defaults come from config options but can be overriden by module
+ * or bootarg options.
+ */
+static int perdev_minors = CONFIG_MMC_BLOCK_MINORS;
 
 /*
- * max 8 partitions per card
+ * We've only got one major, so number of mmcblk devices is
+ * limited to 256 / number of minors per device.
  */
-#define MMC_SHIFT	3
-#define MMC_NUM_MINORS	(256 >> MMC_SHIFT)
+static int max_devices;
 
-static DECLARE_BITMAP(dev_use, MMC_NUM_MINORS);
+/* 256 minors, so at most 256 separate devices */
+static DECLARE_BITMAP(dev_use, 256);
+static DECLARE_BITMAP(name_use, 256);
 
 /*
  * There is one mmc_blk_data per slot.
@@ -58,13 +93,62 @@ struct mmc_blk_data {
 	spinlock_t	lock;
 	struct gendisk	*disk;
 	struct mmc_queue queue;
+	struct list_head part;
+
+	unsigned int	flags;
+#define MMC_BLK_CMD23	(1 << 0)	/* Can do SET_BLOCK_COUNT for multiblock */
+#define MMC_BLK_REL_WR	(1 << 1)	/* MMC Reliable write support */
+#define MMC_BLK_PACKED_CMD	(1 << 2)	/* MMC packed command support */
 
 	unsigned int	usage;
 	unsigned int	read_only;
+	unsigned int	part_type;
+	unsigned int	name_idx;
+	unsigned int	reset_done;
+#define MMC_BLK_READ		BIT(0)
+#define MMC_BLK_WRITE		BIT(1)
+#define MMC_BLK_DISCARD		BIT(2)
+#define MMC_BLK_SECDISCARD	BIT(3)
+
+	/*
+	 * Only set in main mmc_blk_data associated
+	 * with mmc_card with mmc_set_drvdata, and keeps
+	 * track of the current selected device partition.
+	 */
+	unsigned int	part_curr;
+	struct device_attribute force_ro;
+	struct device_attribute power_ro_lock;
+	int	area_type;
 };
 
 static DEFINE_MUTEX(open_lock);
 
+enum {
+	MMC_PACKED_NR_IDX = -1,
+	MMC_PACKED_NR_ZERO,
+	MMC_PACKED_NR_SINGLE,
+};
+
+module_param(perdev_minors, int, 0444);
+MODULE_PARM_DESC(perdev_minors, "Minors numbers to allocate per device");
+
+static inline int mmc_blk_part_switch(struct mmc_card *card,
+				      struct mmc_blk_data *md);
+static int get_card_status(struct mmc_card *card, u32 *status, int retries);
+
+static inline void mmc_blk_clear_packed(struct mmc_queue_req *mqrq)
+{
+	struct mmc_packed *packed = mqrq->packed;
+
+	BUG_ON(!packed);
+
+	mqrq->cmd_type = MMC_PACKED_NONE;
+	packed->nr_entries = MMC_PACKED_NR_ZERO;
+	packed->idx_failure = MMC_PACKED_NR_IDX;
+	packed->retries = 0;
+	packed->blocks = 0;
+}
+
 static struct mmc_blk_data *mmc_blk_get(struct gendisk *disk)
 {
 	struct mmc_blk_data *md;
@@ -80,12 +164,24 @@ static struct mmc_blk_data *mmc_blk_get(struct gendisk *disk)
 	return md;
 }
 
+static inline int mmc_get_devidx(struct gendisk *disk)
+{
+	int devmaj = MAJOR(disk_devt(disk));
+	int devidx = MINOR(disk_devt(disk)) / perdev_minors;
+
+	if (!devmaj)
+		devidx = disk->first_minor / perdev_minors;
+	return devidx;
+}
+
 static void mmc_blk_put(struct mmc_blk_data *md)
 {
 	mutex_lock(&open_lock);
 	md->usage--;
 	if (md->usage == 0) {
-		int devidx = MINOR(disk_devt(md->disk)) >> MMC_SHIFT;
+		int devidx = mmc_get_devidx(md->disk);
+		blk_cleanup_queue(md->queue.queue);
+
 		__clear_bit(devidx, dev_use);
 
 		put_disk(md->disk);
@@ -94,11 +190,108 @@ static void mmc_blk_put(struct mmc_blk_data *md)
 	mutex_unlock(&open_lock);
 }
 
+static ssize_t power_ro_lock_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	int ret;
+	struct mmc_blk_data *md = mmc_blk_get(dev_to_disk(dev));
+	struct mmc_card *card = md->queue.card;
+	int locked = 0;
+
+	if (card->ext_csd.boot_ro_lock & EXT_CSD_BOOT_WP_B_PERM_WP_EN)
+		locked = 2;
+	else if (card->ext_csd.boot_ro_lock & EXT_CSD_BOOT_WP_B_PWR_WP_EN)
+		locked = 1;
+
+	ret = snprintf(buf, PAGE_SIZE, "%d\n", locked);
+
+	return ret;
+}
+
+static ssize_t power_ro_lock_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t count)
+{
+	int ret;
+	struct mmc_blk_data *md, *part_md;
+	struct mmc_card *card;
+	unsigned long set;
+
+	if (kstrtoul(buf, 0, &set))
+		return -EINVAL;
+
+	if (set != 1)
+		return count;
+
+	md = mmc_blk_get(dev_to_disk(dev));
+	card = md->queue.card;
+
+	mmc_get_card(card);
+
+	ret = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_BOOT_WP,
+				card->ext_csd.boot_ro_lock |
+				EXT_CSD_BOOT_WP_B_PWR_WP_EN,
+				card->ext_csd.part_time);
+	if (ret)
+		pr_err("%s: Locking boot partition ro until next power on failed: %d\n", md->disk->disk_name, ret);
+	else
+		card->ext_csd.boot_ro_lock |= EXT_CSD_BOOT_WP_B_PWR_WP_EN;
+
+	mmc_put_card(card);
+
+	if (!ret) {
+		pr_info("%s: Locking boot partition ro until next power on\n",
+			md->disk->disk_name);
+		set_disk_ro(md->disk, 1);
+
+		list_for_each_entry(part_md, &md->part, part)
+			if (part_md->area_type == MMC_BLK_DATA_AREA_BOOT) {
+				pr_info("%s: Locking boot partition ro until next power on\n", part_md->disk->disk_name);
+				set_disk_ro(part_md->disk, 1);
+			}
+	}
+
+	mmc_blk_put(md);
+	return count;
+}
+
+static ssize_t force_ro_show(struct device *dev, struct device_attribute *attr,
+			     char *buf)
+{
+	int ret;
+	struct mmc_blk_data *md = mmc_blk_get(dev_to_disk(dev));
+
+	ret = snprintf(buf, PAGE_SIZE, "%d",
+		       get_disk_ro(dev_to_disk(dev)) ^
+		       md->read_only);
+	mmc_blk_put(md);
+	return ret;
+}
+
+static ssize_t force_ro_store(struct device *dev, struct device_attribute *attr,
+			      const char *buf, size_t count)
+{
+	int ret;
+	char *end;
+	struct mmc_blk_data *md = mmc_blk_get(dev_to_disk(dev));
+	unsigned long set = simple_strtoul(buf, &end, 0);
+	if (end == buf) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	set_disk_ro(dev_to_disk(dev), set || md->read_only);
+	ret = count;
+out:
+	mmc_blk_put(md);
+	return ret;
+}
+
 static int mmc_blk_open(struct block_device *bdev, fmode_t mode)
 {
 	struct mmc_blk_data *md = mmc_blk_get(bdev->bd_disk);
 	int ret = -ENXIO;
 
+	mutex_lock(&block_mutex);
 	if (md) {
 		if (md->usage == 2)
 			check_disk_change(bdev);
@@ -109,16 +302,18 @@ static int mmc_blk_open(struct block_device *bdev, fmode_t mode)
 			ret = -EROFS;
 		}
 	}
+	mutex_unlock(&block_mutex);
 
 	return ret;
 }
 
-static int mmc_blk_release(struct gendisk *disk, fmode_t mode)
+static void mmc_blk_release(struct gendisk *disk, fmode_t mode)
 {
 	struct mmc_blk_data *md = disk->private_data;
 
+	mutex_lock(&block_mutex);
 	mmc_blk_put(md);
-	return 0;
+	mutex_unlock(&block_mutex);
 }
 
 static int
@@ -130,34 +325,359 @@ mmc_blk_getgeo(struct block_device *bdev, struct hd_geometry *geo)
 	return 0;
 }
 
-static struct block_device_operations mmc_bdops = {
+struct mmc_blk_ioc_data {
+	struct mmc_ioc_cmd ic;
+	unsigned char *buf;
+	u64 buf_bytes;
+};
+
+static struct mmc_blk_ioc_data *mmc_blk_ioctl_copy_from_user(
+	struct mmc_ioc_cmd __user *user)
+{
+	struct mmc_blk_ioc_data *idata;
+	int err;
+
+	idata = kzalloc(sizeof(*idata), GFP_KERNEL);
+	if (!idata) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	if (copy_from_user(&idata->ic, user, sizeof(idata->ic))) {
+		err = -EFAULT;
+		goto idata_err;
+	}
+
+	idata->buf_bytes = (u64) idata->ic.blksz * idata->ic.blocks;
+	if (idata->buf_bytes > MMC_IOC_MAX_BYTES) {
+		err = -EOVERFLOW;
+		goto idata_err;
+	}
+
+	if (!idata->buf_bytes)
+		return idata;
+
+	idata->buf = kzalloc(idata->buf_bytes, GFP_KERNEL);
+	if (!idata->buf) {
+		err = -ENOMEM;
+		goto idata_err;
+	}
+
+	if (copy_from_user(idata->buf, (void __user *)(unsigned long)
+					idata->ic.data_ptr, idata->buf_bytes)) {
+		err = -EFAULT;
+		goto copy_err;
+	}
+
+	return idata;
+
+copy_err:
+	kfree(idata->buf);
+idata_err:
+	kfree(idata);
+out:
+	return ERR_PTR(err);
+}
+
+static int ioctl_rpmb_card_status_poll(struct mmc_card *card, u32 *status,
+				       u32 retries_max)
+{
+	int err;
+	u32 retry_count = 0;
+
+	if (!status || !retries_max)
+		return -EINVAL;
+
+	do {
+		err = get_card_status(card, status, 5);
+		if (err)
+			break;
+
+		if (!R1_STATUS(*status) &&
+				(R1_CURRENT_STATE(*status) != R1_STATE_PRG))
+			break; /* RPMB programming operation complete */
+
+		/*
+		 * Rechedule to give the MMC device a chance to continue
+		 * processing the previous command without being polled too
+		 * frequently.
+		 */
+		usleep_range(1000, 5000);
+	} while (++retry_count < retries_max);
+
+	if (retry_count == retries_max)
+		err = -EPERM;
+
+	return err;
+}
+
+static int ioctl_do_sanitize(struct mmc_card *card)
+{
+	int err;
+
+	if (!mmc_can_sanitize(card)) {
+			pr_warn("%s: %s - SANITIZE is not supported\n",
+				mmc_hostname(card->host), __func__);
+			err = -EOPNOTSUPP;
+			goto out;
+	}
+
+	pr_debug("%s: %s - SANITIZE IN PROGRESS...\n",
+		mmc_hostname(card->host), __func__);
+
+	err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+					EXT_CSD_SANITIZE_START, 1,
+					MMC_SANITIZE_REQ_TIMEOUT);
+
+	if (err)
+		pr_err("%s: %s - EXT_CSD_SANITIZE_START failed. err=%d\n",
+		       mmc_hostname(card->host), __func__, err);
+
+	pr_debug("%s: %s - SANITIZE COMPLETED\n", mmc_hostname(card->host),
+					     __func__);
+out:
+	return err;
+}
+
+static int mmc_blk_ioctl_cmd(struct block_device *bdev,
+	struct mmc_ioc_cmd __user *ic_ptr)
+{
+	struct mmc_blk_ioc_data *idata;
+	struct mmc_blk_data *md;
+	struct mmc_card *card;
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
+	struct mmc_request mrq = {NULL};
+	struct scatterlist sg;
+	int err;
+	int is_rpmb = false;
+	u32 status = 0;
+
+	/*
+	 * The caller must have CAP_SYS_RAWIO, and must be calling this on the
+	 * whole block device, not on a partition.  This prevents overspray
+	 * between sibling partitions.
+	 */
+	if ((!capable(CAP_SYS_RAWIO)) || (bdev != bdev->bd_contains))
+		return -EPERM;
+
+	idata = mmc_blk_ioctl_copy_from_user(ic_ptr);
+	if (IS_ERR(idata))
+		return PTR_ERR(idata);
+
+	md = mmc_blk_get(bdev->bd_disk);
+	if (!md) {
+		err = -EINVAL;
+		goto cmd_err;
+	}
+
+	if (md->area_type & MMC_BLK_DATA_AREA_RPMB)
+		is_rpmb = true;
+
+	card = md->queue.card;
+	if (IS_ERR(card)) {
+		err = PTR_ERR(card);
+		goto cmd_done;
+	}
+
+	cmd.opcode = idata->ic.opcode;
+	cmd.arg = idata->ic.arg;
+	cmd.flags = idata->ic.flags;
+
+	if (idata->buf_bytes) {
+		data.sg = &sg;
+		data.sg_len = 1;
+		data.blksz = idata->ic.blksz;
+		data.blocks = idata->ic.blocks;
+
+		sg_init_one(data.sg, idata->buf, idata->buf_bytes);
+
+		if (idata->ic.write_flag)
+			data.flags = MMC_DATA_WRITE;
+		else
+			data.flags = MMC_DATA_READ;
+
+		/* data.flags must already be set before doing this. */
+		mmc_set_data_timeout(&data, card);
+
+		/* Allow overriding the timeout_ns for empirical tuning. */
+		if (idata->ic.data_timeout_ns)
+			data.timeout_ns = idata->ic.data_timeout_ns;
+
+		if ((cmd.flags & MMC_RSP_R1B) == MMC_RSP_R1B) {
+			/*
+			 * Pretend this is a data transfer and rely on the
+			 * host driver to compute timeout.  When all host
+			 * drivers support cmd.cmd_timeout for R1B, this
+			 * can be changed to:
+			 *
+			 *     mrq.data = NULL;
+			 *     cmd.cmd_timeout = idata->ic.cmd_timeout_ms;
+			 */
+			data.timeout_ns = idata->ic.cmd_timeout_ms * 1000000;
+		}
+
+		mrq.data = &data;
+	}
+
+	mrq.cmd = &cmd;
+
+	mmc_get_card(card);
+
+	err = mmc_blk_part_switch(card, md);
+	if (err)
+		goto cmd_rel_host;
+
+	if (idata->ic.is_acmd) {
+		err = mmc_app_cmd(card->host, card);
+		if (err)
+			goto cmd_rel_host;
+	}
+
+	if (is_rpmb) {
+		err = mmc_set_blockcount(card, data.blocks,
+			idata->ic.write_flag & (1 << 31));
+		if (err)
+			goto cmd_rel_host;
+	}
+
+	if ((MMC_EXTRACT_INDEX_FROM_ARG(cmd.arg) == EXT_CSD_SANITIZE_START) &&
+	    (cmd.opcode == MMC_SWITCH)) {
+		err = ioctl_do_sanitize(card);
+
+		if (err)
+			pr_err("%s: ioctl_do_sanitize() failed. err = %d",
+			       __func__, err);
+
+		goto cmd_rel_host;
+	}
+
+	mmc_wait_for_req(card->host, &mrq);
+
+	if (cmd.error) {
+		dev_err(mmc_dev(card->host), "%s: cmd error %d\n",
+						__func__, cmd.error);
+		err = cmd.error;
+		goto cmd_rel_host;
+	}
+	if (data.error) {
+		dev_err(mmc_dev(card->host), "%s: data error %d\n",
+						__func__, data.error);
+		err = data.error;
+		goto cmd_rel_host;
+	}
+
+	/*
+	 * According to the SD specs, some commands require a delay after
+	 * issuing the command.
+	 */
+	if (idata->ic.postsleep_min_us)
+		usleep_range(idata->ic.postsleep_min_us, idata->ic.postsleep_max_us);
+
+	if (copy_to_user(&(ic_ptr->response), cmd.resp, sizeof(cmd.resp))) {
+		err = -EFAULT;
+		goto cmd_rel_host;
+	}
+
+	if (!idata->ic.write_flag) {
+		if (copy_to_user((void __user *)(unsigned long) idata->ic.data_ptr,
+						idata->buf, idata->buf_bytes)) {
+			err = -EFAULT;
+			goto cmd_rel_host;
+		}
+	}
+
+	if (is_rpmb) {
+		/*
+		 * Ensure RPMB command has completed by polling CMD13
+		 * "Send Status".
+		 */
+		err = ioctl_rpmb_card_status_poll(card, &status, 5);
+		if (err)
+			dev_err(mmc_dev(card->host),
+					"%s: Card Status=0x%08X, error %d\n",
+					__func__, status, err);
+	}
+
+cmd_rel_host:
+	mmc_put_card(card);
+
+cmd_done:
+	mmc_blk_put(md);
+cmd_err:
+	kfree(idata->buf);
+	kfree(idata);
+	return err;
+}
+
+static int mmc_blk_ioctl(struct block_device *bdev, fmode_t mode,
+	unsigned int cmd, unsigned long arg)
+{
+	int ret = -EINVAL;
+	if (cmd == MMC_IOC_CMD)
+		ret = mmc_blk_ioctl_cmd(bdev, (struct mmc_ioc_cmd __user *)arg);
+	return ret;
+}
+
+#ifdef CONFIG_COMPAT
+static int mmc_blk_compat_ioctl(struct block_device *bdev, fmode_t mode,
+	unsigned int cmd, unsigned long arg)
+{
+	return mmc_blk_ioctl(bdev, mode, cmd, (unsigned long) compat_ptr(arg));
+}
+#endif
+
+static const struct block_device_operations mmc_bdops = {
 	.open			= mmc_blk_open,
 	.release		= mmc_blk_release,
 	.getgeo			= mmc_blk_getgeo,
 	.owner			= THIS_MODULE,
+	.ioctl			= mmc_blk_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl		= mmc_blk_compat_ioctl,
+#endif
 };
 
-struct mmc_blk_request {
-	struct mmc_request	mrq;
-	struct mmc_command	cmd;
-	struct mmc_command	stop;
-	struct mmc_data		data;
-};
+static inline int mmc_blk_part_switch(struct mmc_card *card,
+				      struct mmc_blk_data *md)
+{
+	int ret;
+	struct mmc_blk_data *main_md = mmc_get_drvdata(card);
+
+	if (main_md->part_curr == md->part_type)
+		return 0;
+
+	if (mmc_card_mmc(card)) {
+		u8 part_config = card->ext_csd.part_config;
+
+		part_config &= ~EXT_CSD_PART_CONFIG_ACC_MASK;
+		part_config |= md->part_type;
+
+		ret = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 EXT_CSD_PART_CONFIG, part_config,
+				 card->ext_csd.part_time);
+		if (ret)
+			return ret;
+
+		card->ext_csd.part_config = part_config;
+	}
+
+	main_md->part_curr = md->part_type;
+	return 0;
+}
 
 static u32 mmc_sd_num_wr_blocks(struct mmc_card *card)
 {
 	int err;
-	__be32 blocks;
+	u32 result;
+	__be32 *blocks;
 
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_data data;
-	unsigned int timeout_us;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
 
 	struct scatterlist sg;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_APP_CMD;
 	cmd.arg = card->rca << 16;
 	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
@@ -174,249 +694,1035 @@ static u32 mmc_sd_num_wr_blocks(struct mmc_card *card)
 	cmd.arg = 0;
 	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
 
-	memset(&data, 0, sizeof(struct mmc_data));
-
-	data.timeout_ns = card->csd.tacc_ns * 100;
-	data.timeout_clks = card->csd.tacc_clks * 100;
-
-	timeout_us = data.timeout_ns / 1000;
-	timeout_us += data.timeout_clks * 1000 /
-		(card->host->ios.clock / 1000);
-
-	if (timeout_us > 100000) {
-		data.timeout_ns = 100000000;
-		data.timeout_clks = 0;
-	}
-
 	data.blksz = 4;
 	data.blocks = 1;
 	data.flags = MMC_DATA_READ;
 	data.sg = &sg;
 	data.sg_len = 1;
-
-	memset(&mrq, 0, sizeof(struct mmc_request));
+	mmc_set_data_timeout(&data, card);
 
 	mrq.cmd = &cmd;
 	mrq.data = &data;
 
-	sg_init_one(&sg, &blocks, 4);
+	blocks = kmalloc(4, GFP_KERNEL);
+	if (!blocks)
+		return (u32)-1;
+
+	sg_init_one(&sg, blocks, 4);
 
 	mmc_wait_for_req(card->host, &mrq);
 
+	result = ntohl(*blocks);
+	kfree(blocks);
+
 	if (cmd.error || data.error)
-		return (u32)-1;
+		result = (u32)-1;
 
-	return ntohl(blocks);
+	return result;
 }
 
-static u32 get_card_status(struct mmc_card *card, struct request *req)
+static int get_card_status(struct mmc_card *card, u32 *status, int retries)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int err;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
 	cmd.opcode = MMC_SEND_STATUS;
 	if (!mmc_host_is_spi(card->host))
 		cmd.arg = card->rca << 16;
 	cmd.flags = MMC_RSP_SPI_R2 | MMC_RSP_R1 | MMC_CMD_AC;
-	err = mmc_wait_for_cmd(card->host, &cmd, 0);
-	if (err)
-		printk(KERN_ERR "%s: error %d sending status comand",
-		       req->rq_disk->disk_name, err);
-	return cmd.resp[0];
+	err = mmc_wait_for_cmd(card->host, &cmd, retries);
+	if (err == 0)
+		*status = cmd.resp[0];
+	return err;
 }
 
-static int mmc_blk_issue_rq(struct mmc_queue *mq, struct request *req)
+static int card_busy_detect(struct mmc_card *card, unsigned int timeout_ms,
+		bool hw_busy_detect, struct request *req, int *gen_err)
 {
-	struct mmc_blk_data *md = mq->data;
-	struct mmc_card *card = md->queue.card;
-	struct mmc_blk_request brq;
-	int ret = 1, disable_multi = 0;
-
-	mmc_claim_host(card->host);
+	unsigned long timeout = jiffies + msecs_to_jiffies(timeout_ms);
+	int err = 0;
+	u32 status;
 
 	do {
-		struct mmc_command cmd;
-		u32 readcmd, writecmd, status = 0;
-
-		memset(&brq, 0, sizeof(struct mmc_blk_request));
-		brq.mrq.cmd = &brq.cmd;
-		brq.mrq.data = &brq.data;
-
-		brq.cmd.arg = req->sector;
-		if (!mmc_card_blockaddr(card))
-			brq.cmd.arg <<= 9;
-		brq.cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
-		brq.data.blksz = 512;
-		brq.stop.opcode = MMC_STOP_TRANSMISSION;
-		brq.stop.arg = 0;
-		brq.stop.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
-		brq.data.blocks = req->nr_sectors;
+		err = get_card_status(card, &status, 5);
+		if (err) {
+			pr_err("%s: error %d requesting status\n",
+			       req->rq_disk->disk_name, err);
+			return err;
+		}
+
+		if (status & R1_ERROR) {
+			pr_err("%s: %s: error sending status cmd, status %#x\n",
+				req->rq_disk->disk_name, __func__, status);
+			*gen_err = 1;
+		}
+
+		/* We may rely on the host hw to handle busy detection.*/
+		if ((card->host->caps & MMC_CAP_WAIT_WHILE_BUSY) &&
+			hw_busy_detect)
+			break;
+
+		/*
+		 * Timeout if the device never becomes ready for data and never
+		 * leaves the program state.
+		 */
+		if (time_after(jiffies, timeout)) {
+			pr_err("%s: Card stuck in programming state! %s %s\n",
+				mmc_hostname(card->host),
+				req->rq_disk->disk_name, __func__);
+			return -ETIMEDOUT;
+		}
 
 		/*
-		 * After a read error, we redo the request one sector at a time
-		 * in order to accurately determine which sectors can be read
-		 * successfully.
+		 * Some cards mishandle the status bits,
+		 * so make sure to check both the busy
+		 * indication and the card state.
 		 */
-		if (disable_multi && brq.data.blocks > 1)
-			brq.data.blocks = 1;
+	} while (!(status & R1_READY_FOR_DATA) ||
+		 (R1_CURRENT_STATE(status) == R1_STATE_PRG));
+
+	return err;
+}
+
+static int send_stop(struct mmc_card *card, unsigned int timeout_ms,
+		struct request *req, int *gen_err, u32 *stop_status)
+{
+	struct mmc_host *host = card->host;
+	struct mmc_command cmd = {0};
+	int err;
+	bool use_r1b_resp = rq_data_dir(req) == WRITE;
+
+	/*
+	 * Normally we use R1B responses for WRITE, but in cases where the host
+	 * has specified a max_busy_timeout we need to validate it. A failure
+	 * means we need to prevent the host from doing hw busy detection, which
+	 * is done by converting to a R1 response instead.
+	 */
+	if (host->max_busy_timeout && (timeout_ms > host->max_busy_timeout))
+		use_r1b_resp = false;
+
+	cmd.opcode = MMC_STOP_TRANSMISSION;
+	if (use_r1b_resp) {
+		cmd.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
+		cmd.busy_timeout = timeout_ms;
+	} else {
+		cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
+	}
+
+	err = mmc_wait_for_cmd(host, &cmd, 5);
+	if (err)
+		return err;
+
+	*stop_status = cmd.resp[0];
+
+	/* No need to check card status in case of READ. */
+	if (rq_data_dir(req) == READ)
+		return 0;
+
+	if (!mmc_host_is_spi(host) &&
+		(*stop_status & R1_ERROR)) {
+		pr_err("%s: %s: general error sending stop command, resp %#x\n",
+			req->rq_disk->disk_name, __func__, *stop_status);
+		*gen_err = 1;
+	}
+
+	return card_busy_detect(card, timeout_ms, use_r1b_resp, req, gen_err);
+}
 
-		if (brq.data.blocks > 1) {
-			/* SPI multiblock writes terminate using a special
-			 * token, not a STOP_TRANSMISSION request.
+#define ERR_NOMEDIUM	3
+#define ERR_RETRY	2
+#define ERR_ABORT	1
+#define ERR_CONTINUE	0
+
+static int mmc_blk_cmd_error(struct request *req, const char *name, int error,
+	bool status_valid, u32 status)
+{
+	switch (error) {
+	case -EILSEQ:
+		/* response crc error, retry the r/w cmd */
+		pr_err("%s: %s sending %s command, card status %#x\n",
+			req->rq_disk->disk_name, "response CRC error",
+			name, status);
+		return ERR_RETRY;
+
+	case -ETIMEDOUT:
+		pr_err("%s: %s sending %s command, card status %#x\n",
+			req->rq_disk->disk_name, "timed out", name, status);
+
+		/* If the status cmd initially failed, retry the r/w cmd */
+		if (!status_valid)
+			return ERR_RETRY;
+
+		/*
+		 * If it was a r/w cmd crc error, or illegal command
+		 * (eg, issued in wrong state) then retry - we should
+		 * have corrected the state problem above.
+		 */
+		if (status & (R1_COM_CRC_ERROR | R1_ILLEGAL_COMMAND))
+			return ERR_RETRY;
+
+		/* Otherwise abort the command */
+		return ERR_ABORT;
+
+	default:
+		/* We don't understand the error code the driver gave us */
+		pr_err("%s: unknown error %d sending read/write command, card status %#x\n",
+		       req->rq_disk->disk_name, error, status);
+		return ERR_ABORT;
+	}
+}
+
+/*
+ * Initial r/w and stop cmd error recovery.
+ * We don't know whether the card received the r/w cmd or not, so try to
+ * restore things back to a sane state.  Essentially, we do this as follows:
+ * - Obtain card status.  If the first attempt to obtain card status fails,
+ *   the status word will reflect the failed status cmd, not the failed
+ *   r/w cmd.  If we fail to obtain card status, it suggests we can no
+ *   longer communicate with the card.
+ * - Check the card state.  If the card received the cmd but there was a
+ *   transient problem with the response, it might still be in a data transfer
+ *   mode.  Try to send it a stop command.  If this fails, we can't recover.
+ * - If the r/w cmd failed due to a response CRC error, it was probably
+ *   transient, so retry the cmd.
+ * - If the r/w cmd timed out, but we didn't get the r/w cmd status, retry.
+ * - If the r/w cmd timed out, and the r/w cmd failed due to CRC error or
+ *   illegal cmd, retry.
+ * Otherwise we don't understand what happened, so abort.
+ */
+static int mmc_blk_cmd_recovery(struct mmc_card *card, struct request *req,
+	struct mmc_blk_request *brq, int *ecc_err, int *gen_err)
+{
+	bool prev_cmd_status_valid = true;
+	u32 status, stop_status = 0;
+	int err, retry;
+
+	if (mmc_card_removed(card))
+		return ERR_NOMEDIUM;
+
+	/*
+	 * Try to get card status which indicates both the card state
+	 * and why there was no response.  If the first attempt fails,
+	 * we can't be sure the returned status is for the r/w command.
+	 */
+	for (retry = 2; retry >= 0; retry--) {
+		err = get_card_status(card, &status, 0);
+		if (!err)
+			break;
+
+		prev_cmd_status_valid = false;
+		pr_err("%s: error %d sending status command, %sing\n",
+		       req->rq_disk->disk_name, err, retry ? "retry" : "abort");
+	}
+
+	/* We couldn't get a response from the card.  Give up. */
+	if (err) {
+		/* Check if the card is removed */
+		if (mmc_detect_card_removed(card->host))
+			return ERR_NOMEDIUM;
+		return ERR_ABORT;
+	}
+
+	/* Flag ECC errors */
+	if ((status & R1_CARD_ECC_FAILED) ||
+	    (brq->stop.resp[0] & R1_CARD_ECC_FAILED) ||
+	    (brq->cmd.resp[0] & R1_CARD_ECC_FAILED))
+		*ecc_err = 1;
+
+	/* Flag General errors */
+	if (!mmc_host_is_spi(card->host) && rq_data_dir(req) != READ)
+		if ((status & R1_ERROR) ||
+			(brq->stop.resp[0] & R1_ERROR)) {
+			pr_err("%s: %s: general error sending stop or status command, stop cmd response %#x, card status %#x\n",
+			       req->rq_disk->disk_name, __func__,
+			       brq->stop.resp[0], status);
+			*gen_err = 1;
+		}
+
+	/*
+	 * Check the current card state.  If it is in some data transfer
+	 * mode, tell it to stop (and hopefully transition back to TRAN.)
+	 */
+	if (R1_CURRENT_STATE(status) == R1_STATE_DATA ||
+	    R1_CURRENT_STATE(status) == R1_STATE_RCV) {
+		err = send_stop(card,
+			DIV_ROUND_UP(brq->data.timeout_ns, 1000000),
+			req, gen_err, &stop_status);
+		if (err) {
+			pr_err("%s: error %d sending stop command\n",
+			       req->rq_disk->disk_name, err);
+			/*
+			 * If the stop cmd also timed out, the card is probably
+			 * not present, so abort. Other errors are bad news too.
 			 */
-			if (!mmc_host_is_spi(card->host)
-					|| rq_data_dir(req) == READ)
-				brq.mrq.stop = &brq.stop;
-			readcmd = MMC_READ_MULTIPLE_BLOCK;
-			writecmd = MMC_WRITE_MULTIPLE_BLOCK;
-		} else {
-			brq.mrq.stop = NULL;
-			readcmd = MMC_READ_SINGLE_BLOCK;
-			writecmd = MMC_WRITE_BLOCK;
+			return ERR_ABORT;
+		}
+
+		if (stop_status & R1_CARD_ECC_FAILED)
+			*ecc_err = 1;
+	}
+
+	/* Check for set block count errors */
+	if (brq->sbc.error)
+		return mmc_blk_cmd_error(req, "SET_BLOCK_COUNT", brq->sbc.error,
+				prev_cmd_status_valid, status);
+
+	/* Check for r/w command errors */
+	if (brq->cmd.error)
+		return mmc_blk_cmd_error(req, "r/w cmd", brq->cmd.error,
+				prev_cmd_status_valid, status);
+
+	/* Data errors */
+	if (!brq->stop.error)
+		return ERR_CONTINUE;
+
+	/* Now for stop errors.  These aren't fatal to the transfer. */
+	pr_err("%s: error %d sending stop command, original cmd response %#x, card status %#x\n",
+	       req->rq_disk->disk_name, brq->stop.error,
+	       brq->cmd.resp[0], status);
+
+	/*
+	 * Subsitute in our own stop status as this will give the error
+	 * state which happened during the execution of the r/w command.
+	 */
+	if (stop_status) {
+		brq->stop.resp[0] = stop_status;
+		brq->stop.error = 0;
+	}
+	return ERR_CONTINUE;
+}
+
+static int mmc_blk_reset(struct mmc_blk_data *md, struct mmc_host *host,
+			 int type)
+{
+	int err;
+
+	if (md->reset_done & type)
+		return -EEXIST;
+
+	md->reset_done |= type;
+	err = mmc_hw_reset(host);
+	/* Ensure we switch back to the correct partition */
+	if (err != -EOPNOTSUPP) {
+		struct mmc_blk_data *main_md = mmc_get_drvdata(host->card);
+		int part_err;
+
+		main_md->part_curr = main_md->part_type;
+		part_err = mmc_blk_part_switch(host->card, md);
+		if (part_err) {
+			/*
+			 * We have failed to get back into the correct
+			 * partition, so we need to abort the whole request.
+			 */
+			return -ENODEV;
+		}
+	}
+	return err;
+}
+
+static inline void mmc_blk_reset_success(struct mmc_blk_data *md, int type)
+{
+	md->reset_done &= ~type;
+}
+
+static int mmc_blk_issue_discard_rq(struct mmc_queue *mq, struct request *req)
+{
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_card *card = md->queue.card;
+	unsigned int from, nr, arg;
+	int err = 0, type = MMC_BLK_DISCARD;
+
+	if (!mmc_can_erase(card)) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	from = blk_rq_pos(req);
+	nr = blk_rq_sectors(req);
+
+	if (mmc_can_discard(card))
+		arg = MMC_DISCARD_ARG;
+	else if (mmc_can_trim(card))
+		arg = MMC_TRIM_ARG;
+	else
+		arg = MMC_ERASE_ARG;
+retry:
+	if (card->quirks & MMC_QUIRK_INAND_CMD38) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 INAND_CMD38_ARG_EXT_CSD,
+				 arg == MMC_TRIM_ARG ?
+				 INAND_CMD38_ARG_TRIM :
+				 INAND_CMD38_ARG_ERASE,
+				 0);
+		if (err)
+			goto out;
+	}
+	err = mmc_erase(card, from, nr, arg);
+out:
+	if (err == -EIO && !mmc_blk_reset(md, card->host, type))
+		goto retry;
+	if (!err)
+		mmc_blk_reset_success(md, type);
+	blk_end_request(req, err, blk_rq_bytes(req));
+
+	return err ? 0 : 1;
+}
+
+static int mmc_blk_issue_secdiscard_rq(struct mmc_queue *mq,
+				       struct request *req)
+{
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_card *card = md->queue.card;
+	unsigned int from, nr, arg;
+	int err = 0, type = MMC_BLK_SECDISCARD;
+
+	if (!(mmc_can_secure_erase_trim(card))) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	from = blk_rq_pos(req);
+	nr = blk_rq_sectors(req);
+
+	if (mmc_can_trim(card) && !mmc_erase_group_aligned(card, from, nr))
+		arg = MMC_SECURE_TRIM1_ARG;
+	else
+		arg = MMC_SECURE_ERASE_ARG;
+
+retry:
+	if (card->quirks & MMC_QUIRK_INAND_CMD38) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 INAND_CMD38_ARG_EXT_CSD,
+				 arg == MMC_SECURE_TRIM1_ARG ?
+				 INAND_CMD38_ARG_SECTRIM1 :
+				 INAND_CMD38_ARG_SECERASE,
+				 0);
+		if (err)
+			goto out_retry;
+	}
+
+	err = mmc_erase(card, from, nr, arg);
+	if (err == -EIO)
+		goto out_retry;
+	if (err)
+		goto out;
+
+	if (arg == MMC_SECURE_TRIM1_ARG) {
+		if (card->quirks & MMC_QUIRK_INAND_CMD38) {
+			err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+					 INAND_CMD38_ARG_EXT_CSD,
+					 INAND_CMD38_ARG_SECTRIM2,
+					 0);
+			if (err)
+				goto out_retry;
+		}
+
+		err = mmc_erase(card, from, nr, MMC_SECURE_TRIM2_ARG);
+		if (err == -EIO)
+			goto out_retry;
+		if (err)
+			goto out;
+	}
+
+out_retry:
+	if (err && !mmc_blk_reset(md, card->host, type))
+		goto retry;
+	if (!err)
+		mmc_blk_reset_success(md, type);
+out:
+	blk_end_request(req, err, blk_rq_bytes(req));
+
+	return err ? 0 : 1;
+}
+
+static int mmc_blk_issue_flush(struct mmc_queue *mq, struct request *req)
+{
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_card *card = md->queue.card;
+	int ret = 0;
+
+	ret = mmc_flush_cache(card);
+	if (ret)
+		ret = -EIO;
+
+	blk_end_request_all(req, ret);
+
+	return ret ? 0 : 1;
+}
+
+/*
+ * Reformat current write as a reliable write, supporting
+ * both legacy and the enhanced reliable write MMC cards.
+ * In each transfer we'll handle only as much as a single
+ * reliable write can handle, thus finish the request in
+ * partial completions.
+ */
+static inline void mmc_apply_rel_rw(struct mmc_blk_request *brq,
+				    struct mmc_card *card,
+				    struct request *req)
+{
+	if (!(card->ext_csd.rel_param & EXT_CSD_WR_REL_PARAM_EN)) {
+		/* Legacy mode imposes restrictions on transfers. */
+		if (!IS_ALIGNED(brq->cmd.arg, card->ext_csd.rel_sectors))
+			brq->data.blocks = 1;
+
+		if (brq->data.blocks > card->ext_csd.rel_sectors)
+			brq->data.blocks = card->ext_csd.rel_sectors;
+		else if (brq->data.blocks < card->ext_csd.rel_sectors)
+			brq->data.blocks = 1;
+	}
+}
+
+#define CMD_ERRORS							\
+	(R1_OUT_OF_RANGE |	/* Command argument out of range */	\
+	 R1_ADDRESS_ERROR |	/* Misaligned address */		\
+	 R1_BLOCK_LEN_ERROR |	/* Transferred block length incorrect */\
+	 R1_WP_VIOLATION |	/* Tried to write to protected block */	\
+	 R1_CC_ERROR |		/* Card controller error */		\
+	 R1_ERROR)		/* General/unknown error */
+
+static int mmc_blk_err_check(struct mmc_card *card,
+			     struct mmc_async_req *areq)
+{
+	struct mmc_queue_req *mq_mrq = container_of(areq, struct mmc_queue_req,
+						    mmc_active);
+	struct mmc_blk_request *brq = &mq_mrq->brq;
+	struct request *req = mq_mrq->req;
+	int ecc_err = 0, gen_err = 0;
+
+	/*
+	 * sbc.error indicates a problem with the set block count
+	 * command.  No data will have been transferred.
+	 *
+	 * cmd.error indicates a problem with the r/w command.  No
+	 * data will have been transferred.
+	 *
+	 * stop.error indicates a problem with the stop command.  Data
+	 * may have been transferred, or may still be transferring.
+	 */
+	if (brq->sbc.error || brq->cmd.error || brq->stop.error ||
+	    brq->data.error) {
+		switch (mmc_blk_cmd_recovery(card, req, brq, &ecc_err, &gen_err)) {
+		case ERR_RETRY:
+			return MMC_BLK_RETRY;
+		case ERR_ABORT:
+			return MMC_BLK_ABORT;
+		case ERR_NOMEDIUM:
+			return MMC_BLK_NOMEDIUM;
+		case ERR_CONTINUE:
+			break;
+		}
+	}
+
+	/*
+	 * Check for errors relating to the execution of the
+	 * initial command - such as address errors.  No data
+	 * has been transferred.
+	 */
+	if (brq->cmd.resp[0] & CMD_ERRORS) {
+		pr_err("%s: r/w command failed, status = %#x\n",
+		       req->rq_disk->disk_name, brq->cmd.resp[0]);
+		return MMC_BLK_ABORT;
+	}
+
+	/*
+	 * Everything else is either success, or a data error of some
+	 * kind.  If it was a write, we may have transitioned to
+	 * program mode, which we have to wait for it to complete.
+	 */
+	if (!mmc_host_is_spi(card->host) && rq_data_dir(req) != READ) {
+		int err;
+
+		/* Check stop command response */
+		if (brq->stop.resp[0] & R1_ERROR) {
+			pr_err("%s: %s: general error sending stop command, stop cmd response %#x\n",
+			       req->rq_disk->disk_name, __func__,
+			       brq->stop.resp[0]);
+			gen_err = 1;
 		}
 
+		err = card_busy_detect(card, MMC_BLK_TIMEOUT_MS, false, req,
+					&gen_err);
+		if (err)
+			return MMC_BLK_CMD_ERR;
+	}
+
+	/* if general error occurs, retry the write operation. */
+	if (gen_err) {
+		pr_warn("%s: retrying write for general error\n",
+				req->rq_disk->disk_name);
+		return MMC_BLK_RETRY;
+	}
+
+	if (brq->data.error) {
+		pr_err("%s: error %d transferring data, sector %u, nr %u, cmd response %#x, card status %#x\n",
+		       req->rq_disk->disk_name, brq->data.error,
+		       (unsigned)blk_rq_pos(req),
+		       (unsigned)blk_rq_sectors(req),
+		       brq->cmd.resp[0], brq->stop.resp[0]);
+
 		if (rq_data_dir(req) == READ) {
-			brq.cmd.opcode = readcmd;
-			brq.data.flags |= MMC_DATA_READ;
+			if (ecc_err)
+				return MMC_BLK_ECC_ERR;
+			return MMC_BLK_DATA_ERR;
 		} else {
-			brq.cmd.opcode = writecmd;
-			brq.data.flags |= MMC_DATA_WRITE;
+			return MMC_BLK_CMD_ERR;
 		}
+	}
 
-		mmc_set_data_timeout(&brq.data, card);
+	if (!brq->data.bytes_xfered)
+		return MMC_BLK_RETRY;
 
-		brq.data.sg = mq->sg;
-		brq.data.sg_len = mmc_queue_map_sg(mq);
+	if (mmc_packed_cmd(mq_mrq->cmd_type)) {
+		if (unlikely(brq->data.blocks << 9 != brq->data.bytes_xfered))
+			return MMC_BLK_PARTIAL;
+		else
+			return MMC_BLK_SUCCESS;
+	}
 
-		/*
-		 * Adjust the sg list so it is the same size as the
-		 * request.
-		 */
-		if (brq.data.blocks != req->nr_sectors) {
-			int i, data_size = brq.data.blocks << 9;
-			struct scatterlist *sg;
-
-			for_each_sg(brq.data.sg, sg, brq.data.sg_len, i) {
-				data_size -= sg->length;
-				if (data_size <= 0) {
-					sg->length += data_size;
-					i++;
-					break;
-				}
+	if (blk_rq_bytes(req) != brq->data.bytes_xfered)
+		return MMC_BLK_PARTIAL;
+
+	return MMC_BLK_SUCCESS;
+}
+
+static int mmc_blk_packed_err_check(struct mmc_card *card,
+				    struct mmc_async_req *areq)
+{
+	struct mmc_queue_req *mq_rq = container_of(areq, struct mmc_queue_req,
+			mmc_active);
+	struct request *req = mq_rq->req;
+	struct mmc_packed *packed = mq_rq->packed;
+	int err, check, status;
+	u8 *ext_csd;
+
+	BUG_ON(!packed);
+
+	packed->retries--;
+	check = mmc_blk_err_check(card, areq);
+	err = get_card_status(card, &status, 0);
+	if (err) {
+		pr_err("%s: error %d sending status command\n",
+		       req->rq_disk->disk_name, err);
+		return MMC_BLK_ABORT;
+	}
+
+	if (status & R1_EXCEPTION_EVENT) {
+		ext_csd = kzalloc(512, GFP_KERNEL);
+		if (!ext_csd) {
+			pr_err("%s: unable to allocate buffer for ext_csd\n",
+			       req->rq_disk->disk_name);
+			return -ENOMEM;
+		}
+
+		err = mmc_send_ext_csd(card, ext_csd);
+		if (err) {
+			pr_err("%s: error %d sending ext_csd\n",
+			       req->rq_disk->disk_name, err);
+			check = MMC_BLK_ABORT;
+			goto free;
+		}
+
+		if ((ext_csd[EXT_CSD_EXP_EVENTS_STATUS] &
+		     EXT_CSD_PACKED_FAILURE) &&
+		    (ext_csd[EXT_CSD_PACKED_CMD_STATUS] &
+		     EXT_CSD_PACKED_GENERIC_ERROR)) {
+			if (ext_csd[EXT_CSD_PACKED_CMD_STATUS] &
+			    EXT_CSD_PACKED_INDEXED_ERROR) {
+				packed->idx_failure =
+				  ext_csd[EXT_CSD_PACKED_FAILURE_INDEX] - 1;
+				check = MMC_BLK_PARTIAL;
 			}
-			brq.data.sg_len = i;
+			pr_err("%s: packed cmd failed, nr %u, sectors %u, "
+			       "failure index: %d\n",
+			       req->rq_disk->disk_name, packed->nr_entries,
+			       packed->blocks, packed->idx_failure);
 		}
+free:
+		kfree(ext_csd);
+	}
 
-		mmc_queue_bounce_pre(mq);
+	return check;
+}
 
-		mmc_wait_for_req(card->host, &brq.mrq);
+static void mmc_blk_rw_rq_prep(struct mmc_queue_req *mqrq,
+			       struct mmc_card *card,
+			       int disable_multi,
+			       struct mmc_queue *mq)
+{
+	u32 readcmd, writecmd;
+	struct mmc_blk_request *brq = &mqrq->brq;
+	struct request *req = mqrq->req;
+	struct mmc_blk_data *md = mq->data;
+	bool do_data_tag;
 
-		mmc_queue_bounce_post(mq);
+	/*
+	 * Reliable writes are used to implement Forced Unit Access and
+	 * REQ_META accesses, and are supported only on MMCs.
+	 *
+	 * XXX: this really needs a good explanation of why REQ_META
+	 * is treated special.
+	 */
+	bool do_rel_wr = ((req->cmd_flags & REQ_FUA) ||
+			  (req->cmd_flags & REQ_META)) &&
+		(rq_data_dir(req) == WRITE) &&
+		(md->flags & MMC_BLK_REL_WR);
+
+	memset(brq, 0, sizeof(struct mmc_blk_request));
+	brq->mrq.cmd = &brq->cmd;
+	brq->mrq.data = &brq->data;
+
+	brq->cmd.arg = blk_rq_pos(req);
+	if (!mmc_card_blockaddr(card))
+		brq->cmd.arg <<= 9;
+	brq->cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
+	brq->data.blksz = 512;
+	brq->stop.opcode = MMC_STOP_TRANSMISSION;
+	brq->stop.arg = 0;
+	brq->data.blocks = blk_rq_sectors(req);
 
+	/*
+	 * The block layer doesn't support all sector count
+	 * restrictions, so we need to be prepared for too big
+	 * requests.
+	 */
+	if (brq->data.blocks > card->host->max_blk_count)
+		brq->data.blocks = card->host->max_blk_count;
+
+	if (brq->data.blocks > 1) {
 		/*
-		 * Check for errors here, but don't jump to cmd_err
-		 * until later as we need to wait for the card to leave
-		 * programming mode even when things go wrong.
+		 * After a read error, we redo the request one sector
+		 * at a time in order to accurately determine which
+		 * sectors can be read successfully.
 		 */
-		if (brq.cmd.error || brq.data.error || brq.stop.error) {
-			if (brq.data.blocks > 1 && rq_data_dir(req) == READ) {
-				/* Redo read one sector at a time */
-				printk(KERN_WARNING "%s: retrying using single "
-				       "block read\n", req->rq_disk->disk_name);
-				disable_multi = 1;
-				continue;
+		if (disable_multi)
+			brq->data.blocks = 1;
+
+		/* Some controllers can't do multiblock reads due to hw bugs */
+		if (card->host->caps2 & MMC_CAP2_NO_MULTI_READ &&
+		    rq_data_dir(req) == READ)
+			brq->data.blocks = 1;
+	}
+
+	if (brq->data.blocks > 1 || do_rel_wr) {
+		/* SPI multiblock writes terminate using a special
+		 * token, not a STOP_TRANSMISSION request.
+		 */
+		if (!mmc_host_is_spi(card->host) ||
+		    rq_data_dir(req) == READ)
+			brq->mrq.stop = &brq->stop;
+		readcmd = MMC_READ_MULTIPLE_BLOCK;
+		writecmd = MMC_WRITE_MULTIPLE_BLOCK;
+	} else {
+		brq->mrq.stop = NULL;
+		readcmd = MMC_READ_SINGLE_BLOCK;
+		writecmd = MMC_WRITE_BLOCK;
+	}
+	if (rq_data_dir(req) == READ) {
+		brq->cmd.opcode = readcmd;
+		brq->data.flags |= MMC_DATA_READ;
+		if (brq->mrq.stop)
+			brq->stop.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 |
+					MMC_CMD_AC;
+	} else {
+		brq->cmd.opcode = writecmd;
+		brq->data.flags |= MMC_DATA_WRITE;
+		if (brq->mrq.stop)
+			brq->stop.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B |
+					MMC_CMD_AC;
+	}
+
+	if (do_rel_wr)
+		mmc_apply_rel_rw(brq, card, req);
+
+	/*
+	 * Data tag is used only during writing meta data to speed
+	 * up write and any subsequent read of this meta data
+	 */
+	do_data_tag = (card->ext_csd.data_tag_unit_size) &&
+		(req->cmd_flags & REQ_META) &&
+		(rq_data_dir(req) == WRITE) &&
+		((brq->data.blocks * brq->data.blksz) >=
+		 card->ext_csd.data_tag_unit_size);
+
+	/*
+	 * Pre-defined multi-block transfers are preferable to
+	 * open ended-ones (and necessary for reliable writes).
+	 * However, it is not sufficient to just send CMD23,
+	 * and avoid the final CMD12, as on an error condition
+	 * CMD12 (stop) needs to be sent anyway. This, coupled
+	 * with Auto-CMD23 enhancements provided by some
+	 * hosts, means that the complexity of dealing
+	 * with this is best left to the host. If CMD23 is
+	 * supported by card and host, we'll fill sbc in and let
+	 * the host deal with handling it correctly. This means
+	 * that for hosts that don't expose MMC_CAP_CMD23, no
+	 * change of behavior will be observed.
+	 *
+	 * N.B: Some MMC cards experience perf degradation.
+	 * We'll avoid using CMD23-bounded multiblock writes for
+	 * these, while retaining features like reliable writes.
+	 */
+	if ((md->flags & MMC_BLK_CMD23) && mmc_op_multi(brq->cmd.opcode) &&
+	    (do_rel_wr || !(card->quirks & MMC_QUIRK_BLK_NO_CMD23) ||
+	     do_data_tag)) {
+		brq->sbc.opcode = MMC_SET_BLOCK_COUNT;
+		brq->sbc.arg = brq->data.blocks |
+			(do_rel_wr ? (1 << 31) : 0) |
+			(do_data_tag ? (1 << 29) : 0);
+		brq->sbc.flags = MMC_RSP_R1 | MMC_CMD_AC;
+		brq->mrq.sbc = &brq->sbc;
+	}
+
+	mmc_set_data_timeout(&brq->data, card);
+
+	brq->data.sg = mqrq->sg;
+	brq->data.sg_len = mmc_queue_map_sg(mq, mqrq);
+
+	/*
+	 * Adjust the sg list so it is the same size as the
+	 * request.
+	 */
+	if (brq->data.blocks != blk_rq_sectors(req)) {
+		int i, data_size = brq->data.blocks << 9;
+		struct scatterlist *sg;
+
+		for_each_sg(brq->data.sg, sg, brq->data.sg_len, i) {
+			data_size -= sg->length;
+			if (data_size <= 0) {
+				sg->length += data_size;
+				i++;
+				break;
 			}
-			status = get_card_status(card, req);
-		}
-
-		if (brq.cmd.error) {
-			printk(KERN_ERR "%s: error %d sending read/write "
-			       "command, response %#x, card status %#x\n",
-			       req->rq_disk->disk_name, brq.cmd.error,
-			       brq.cmd.resp[0], status);
-		}
-
-		if (brq.data.error) {
-			if (brq.data.error == -ETIMEDOUT && brq.mrq.stop)
-				/* 'Stop' response contains card status */
-				status = brq.mrq.stop->resp[0];
-			printk(KERN_ERR "%s: error %d transferring data,"
-			       " sector %u, nr %u, card status %#x\n",
-			       req->rq_disk->disk_name, brq.data.error,
-			       (unsigned)req->sector,
-			       (unsigned)req->nr_sectors, status);
-		}
-
-		if (brq.stop.error) {
-			printk(KERN_ERR "%s: error %d sending stop command, "
-			       "response %#x, card status %#x\n",
-			       req->rq_disk->disk_name, brq.stop.error,
-			       brq.stop.resp[0], status);
-		}
-
-		if (!mmc_host_is_spi(card->host) && rq_data_dir(req) != READ) {
-			do {
-				int err;
-
-				cmd.opcode = MMC_SEND_STATUS;
-				cmd.arg = card->rca << 16;
-				cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
-				err = mmc_wait_for_cmd(card->host, &cmd, 5);
-				if (err) {
-					printk(KERN_ERR "%s: error %d requesting status\n",
-					       req->rq_disk->disk_name, err);
-					goto cmd_err;
-				}
-				/*
-				 * Some cards mishandle the status bits,
-				 * so make sure to check both the busy
-				 * indication and the card state.
-				 */
-			} while (!(cmd.resp[0] & R1_READY_FOR_DATA) ||
-				(R1_CURRENT_STATE(cmd.resp[0]) == 7));
-
-#if 0
-			if (cmd.resp[0] & ~0x00000900)
-				printk(KERN_ERR "%s: status = %08x\n",
-				       req->rq_disk->disk_name, cmd.resp[0]);
-			if (mmc_decode_status(cmd.resp))
-				goto cmd_err;
-#endif
 		}
+		brq->data.sg_len = i;
+	}
 
-		if (brq.cmd.error || brq.stop.error || brq.data.error) {
-			if (rq_data_dir(req) == READ) {
-				/*
-				 * After an error, we redo I/O one sector at a
-				 * time, so we only reach here after trying to
-				 * read a single sector.
-				 */
-				spin_lock_irq(&md->lock);
-				ret = __blk_end_request(req, -EIO, brq.data.blksz);
-				spin_unlock_irq(&md->lock);
-				continue;
-			}
-			goto cmd_err;
+	mqrq->mmc_active.mrq = &brq->mrq;
+	mqrq->mmc_active.err_check = mmc_blk_err_check;
+
+	mmc_queue_bounce_pre(mqrq);
+}
+
+static inline u8 mmc_calc_packed_hdr_segs(struct request_queue *q,
+					  struct mmc_card *card)
+{
+	unsigned int hdr_sz = mmc_large_sector(card) ? 4096 : 512;
+	unsigned int max_seg_sz = queue_max_segment_size(q);
+	unsigned int len, nr_segs = 0;
+
+	do {
+		len = min(hdr_sz, max_seg_sz);
+		hdr_sz -= len;
+		nr_segs++;
+	} while (hdr_sz);
+
+	return nr_segs;
+}
+
+static u8 mmc_blk_prep_packed_list(struct mmc_queue *mq, struct request *req)
+{
+	struct request_queue *q = mq->queue;
+	struct mmc_card *card = mq->card;
+	struct request *cur = req, *next = NULL;
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_queue_req *mqrq = mq->mqrq_cur;
+	bool en_rel_wr = card->ext_csd.rel_param & EXT_CSD_WR_REL_PARAM_EN;
+	unsigned int req_sectors = 0, phys_segments = 0;
+	unsigned int max_blk_count, max_phys_segs;
+	bool put_back = true;
+	u8 max_packed_rw = 0;
+	u8 reqs = 0;
+
+	if (!(md->flags & MMC_BLK_PACKED_CMD))
+		goto no_packed;
+
+	if ((rq_data_dir(cur) == WRITE) &&
+	    mmc_host_packed_wr(card->host))
+		max_packed_rw = card->ext_csd.max_packed_writes;
+
+	if (max_packed_rw == 0)
+		goto no_packed;
+
+	if (mmc_req_rel_wr(cur) &&
+	    (md->flags & MMC_BLK_REL_WR) && !en_rel_wr)
+		goto no_packed;
+
+	if (mmc_large_sector(card) &&
+	    !IS_ALIGNED(blk_rq_sectors(cur), 8))
+		goto no_packed;
+
+	mmc_blk_clear_packed(mqrq);
+
+	max_blk_count = min(card->host->max_blk_count,
+			    card->host->max_req_size >> 9);
+	if (unlikely(max_blk_count > 0xffff))
+		max_blk_count = 0xffff;
+
+	max_phys_segs = queue_max_segments(q);
+	req_sectors += blk_rq_sectors(cur);
+	phys_segments += cur->nr_phys_segments;
+
+	if (rq_data_dir(cur) == WRITE) {
+		req_sectors += mmc_large_sector(card) ? 8 : 1;
+		phys_segments += mmc_calc_packed_hdr_segs(q, card);
+	}
+
+	do {
+		if (reqs >= max_packed_rw - 1) {
+			put_back = false;
+			break;
 		}
 
-		/*
-		 * A block was successfully transferred.
-		 */
-		spin_lock_irq(&md->lock);
-		ret = __blk_end_request(req, 0, brq.data.bytes_xfered);
-		spin_unlock_irq(&md->lock);
-	} while (ret);
+		spin_lock_irq(q->queue_lock);
+		next = blk_fetch_request(q);
+		spin_unlock_irq(q->queue_lock);
+		if (!next) {
+			put_back = false;
+			break;
+		}
 
-	mmc_release_host(card->host);
+		if (mmc_large_sector(card) &&
+		    !IS_ALIGNED(blk_rq_sectors(next), 8))
+			break;
 
-	return 1;
+		if (next->cmd_flags & REQ_DISCARD ||
+		    next->cmd_flags & REQ_FLUSH)
+			break;
+
+		if (rq_data_dir(cur) != rq_data_dir(next))
+			break;
+
+		if (mmc_req_rel_wr(next) &&
+		    (md->flags & MMC_BLK_REL_WR) && !en_rel_wr)
+			break;
+
+		req_sectors += blk_rq_sectors(next);
+		if (req_sectors > max_blk_count)
+			break;
+
+		phys_segments +=  next->nr_phys_segments;
+		if (phys_segments > max_phys_segs)
+			break;
+
+		list_add_tail(&next->queuelist, &mqrq->packed->list);
+		cur = next;
+		reqs++;
+	} while (1);
+
+	if (put_back) {
+		spin_lock_irq(q->queue_lock);
+		blk_requeue_request(q, next);
+		spin_unlock_irq(q->queue_lock);
+	}
+
+	if (reqs > 0) {
+		list_add(&req->queuelist, &mqrq->packed->list);
+		mqrq->packed->nr_entries = ++reqs;
+		mqrq->packed->retries = reqs;
+		return reqs;
+	}
 
- cmd_err:
- 	/*
- 	 * If this is an SD card and we're writing, we can first
- 	 * mark the known good sectors as ok.
- 	 *
+no_packed:
+	mqrq->cmd_type = MMC_PACKED_NONE;
+	return 0;
+}
+
+static void mmc_blk_packed_hdr_wrq_prep(struct mmc_queue_req *mqrq,
+					struct mmc_card *card,
+					struct mmc_queue *mq)
+{
+	struct mmc_blk_request *brq = &mqrq->brq;
+	struct request *req = mqrq->req;
+	struct request *prq;
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_packed *packed = mqrq->packed;
+	bool do_rel_wr, do_data_tag;
+	u32 *packed_cmd_hdr;
+	u8 hdr_blocks;
+	u8 i = 1;
+
+	BUG_ON(!packed);
+
+	mqrq->cmd_type = MMC_PACKED_WRITE;
+	packed->blocks = 0;
+	packed->idx_failure = MMC_PACKED_NR_IDX;
+
+	packed_cmd_hdr = packed->cmd_hdr;
+	memset(packed_cmd_hdr, 0, sizeof(packed->cmd_hdr));
+	packed_cmd_hdr[0] = (packed->nr_entries << 16) |
+		(PACKED_CMD_WR << 8) | PACKED_CMD_VER;
+	hdr_blocks = mmc_large_sector(card) ? 8 : 1;
+
+	/*
+	 * Argument for each entry of packed group
+	 */
+	list_for_each_entry(prq, &packed->list, queuelist) {
+		do_rel_wr = mmc_req_rel_wr(prq) && (md->flags & MMC_BLK_REL_WR);
+		do_data_tag = (card->ext_csd.data_tag_unit_size) &&
+			(prq->cmd_flags & REQ_META) &&
+			(rq_data_dir(prq) == WRITE) &&
+			((brq->data.blocks * brq->data.blksz) >=
+			 card->ext_csd.data_tag_unit_size);
+		/* Argument of CMD23 */
+		packed_cmd_hdr[(i * 2)] =
+			(do_rel_wr ? MMC_CMD23_ARG_REL_WR : 0) |
+			(do_data_tag ? MMC_CMD23_ARG_TAG_REQ : 0) |
+			blk_rq_sectors(prq);
+		/* Argument of CMD18 or CMD25 */
+		packed_cmd_hdr[((i * 2)) + 1] =
+			mmc_card_blockaddr(card) ?
+			blk_rq_pos(prq) : blk_rq_pos(prq) << 9;
+		packed->blocks += blk_rq_sectors(prq);
+		i++;
+	}
+
+	memset(brq, 0, sizeof(struct mmc_blk_request));
+	brq->mrq.cmd = &brq->cmd;
+	brq->mrq.data = &brq->data;
+	brq->mrq.sbc = &brq->sbc;
+	brq->mrq.stop = &brq->stop;
+
+	brq->sbc.opcode = MMC_SET_BLOCK_COUNT;
+	brq->sbc.arg = MMC_CMD23_ARG_PACKED | (packed->blocks + hdr_blocks);
+	brq->sbc.flags = MMC_RSP_R1 | MMC_CMD_AC;
+
+	brq->cmd.opcode = MMC_WRITE_MULTIPLE_BLOCK;
+	brq->cmd.arg = blk_rq_pos(req);
+	if (!mmc_card_blockaddr(card))
+		brq->cmd.arg <<= 9;
+	brq->cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
+
+	brq->data.blksz = 512;
+	brq->data.blocks = packed->blocks + hdr_blocks;
+	brq->data.flags |= MMC_DATA_WRITE;
+
+	brq->stop.opcode = MMC_STOP_TRANSMISSION;
+	brq->stop.arg = 0;
+	brq->stop.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
+
+	mmc_set_data_timeout(&brq->data, card);
+
+	brq->data.sg = mqrq->sg;
+	brq->data.sg_len = mmc_queue_map_sg(mq, mqrq);
+
+	mqrq->mmc_active.mrq = &brq->mrq;
+	mqrq->mmc_active.err_check = mmc_blk_packed_err_check;
+
+	mmc_queue_bounce_pre(mqrq);
+}
+
+static int mmc_blk_cmd_err(struct mmc_blk_data *md, struct mmc_card *card,
+			   struct mmc_blk_request *brq, struct request *req,
+			   int ret)
+{
+	struct mmc_queue_req *mq_rq;
+	mq_rq = container_of(brq, struct mmc_queue_req, brq);
+
+	/*
+	 * If this is an SD card and we're writing, we can first
+	 * mark the known good sectors as ok.
+	 *
 	 * If the card is not SD, we can still ok written sectors
 	 * as reported by the controller (which might be less than
 	 * the real number of written sectors, but never more).
@@ -426,26 +1732,333 @@ static int mmc_blk_issue_rq(struct mmc_queue *mq, struct request *req)
 
 		blocks = mmc_sd_num_wr_blocks(card);
 		if (blocks != (u32)-1) {
-			spin_lock_irq(&md->lock);
-			ret = __blk_end_request(req, 0, blocks << 9);
-			spin_unlock_irq(&md->lock);
+			ret = blk_end_request(req, 0, blocks << 9);
 		}
 	} else {
-		spin_lock_irq(&md->lock);
-		ret = __blk_end_request(req, 0, brq.data.bytes_xfered);
-		spin_unlock_irq(&md->lock);
+		if (!mmc_packed_cmd(mq_rq->cmd_type))
+			ret = blk_end_request(req, 0, brq->data.bytes_xfered);
 	}
+	return ret;
+}
 
-	mmc_release_host(card->host);
+static int mmc_blk_end_packed_req(struct mmc_queue_req *mq_rq)
+{
+	struct request *prq;
+	struct mmc_packed *packed = mq_rq->packed;
+	int idx = packed->idx_failure, i = 0;
+	int ret = 0;
+
+	BUG_ON(!packed);
+
+	while (!list_empty(&packed->list)) {
+		prq = list_entry_rq(packed->list.next);
+		if (idx == i) {
+			/* retry from error index */
+			packed->nr_entries -= idx;
+			mq_rq->req = prq;
+			ret = 1;
+
+			if (packed->nr_entries == MMC_PACKED_NR_SINGLE) {
+				list_del_init(&prq->queuelist);
+				mmc_blk_clear_packed(mq_rq);
+			}
+			return ret;
+		}
+		list_del_init(&prq->queuelist);
+		blk_end_request(prq, 0, blk_rq_bytes(prq));
+		i++;
+	}
+
+	mmc_blk_clear_packed(mq_rq);
+	return ret;
+}
+
+static void mmc_blk_abort_packed_req(struct mmc_queue_req *mq_rq)
+{
+	struct request *prq;
+	struct mmc_packed *packed = mq_rq->packed;
+
+	BUG_ON(!packed);
+
+	while (!list_empty(&packed->list)) {
+		prq = list_entry_rq(packed->list.next);
+		list_del_init(&prq->queuelist);
+		blk_end_request(prq, -EIO, blk_rq_bytes(prq));
+	}
+
+	mmc_blk_clear_packed(mq_rq);
+}
+
+static void mmc_blk_revert_packed_req(struct mmc_queue *mq,
+				      struct mmc_queue_req *mq_rq)
+{
+	struct request *prq;
+	struct request_queue *q = mq->queue;
+	struct mmc_packed *packed = mq_rq->packed;
+
+	BUG_ON(!packed);
+
+	while (!list_empty(&packed->list)) {
+		prq = list_entry_rq(packed->list.prev);
+		if (prq->queuelist.prev != &packed->list) {
+			list_del_init(&prq->queuelist);
+			spin_lock_irq(q->queue_lock);
+			blk_requeue_request(mq->queue, prq);
+			spin_unlock_irq(q->queue_lock);
+		} else {
+			list_del_init(&prq->queuelist);
+		}
+	}
+
+	mmc_blk_clear_packed(mq_rq);
+}
+
+static int mmc_blk_issue_rw_rq(struct mmc_queue *mq, struct request *rqc)
+{
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_card *card = md->queue.card;
+	struct mmc_blk_request *brq = &mq->mqrq_cur->brq;
+	int ret = 1, disable_multi = 0, retry = 0, type;
+	enum mmc_blk_status status;
+	struct mmc_queue_req *mq_rq;
+	struct request *req = rqc;
+	struct mmc_async_req *areq;
+	const u8 packed_nr = 2;
+	u8 reqs = 0;
+
+	if (!rqc && !mq->mqrq_prev->req)
+		return 0;
+
+	if (rqc)
+		reqs = mmc_blk_prep_packed_list(mq, rqc);
+
+	do {
+		if (rqc) {
+			/*
+			 * When 4KB native sector is enabled, only 8 blocks
+			 * multiple read or write is allowed
+			 */
+			if ((brq->data.blocks & 0x07) &&
+			    (card->ext_csd.data_sector_size == 4096)) {
+				pr_err("%s: Transfer size is not 4KB sector size aligned\n",
+					req->rq_disk->disk_name);
+				mq_rq = mq->mqrq_cur;
+				goto cmd_abort;
+			}
+
+			if (reqs >= packed_nr)
+				mmc_blk_packed_hdr_wrq_prep(mq->mqrq_cur,
+							    card, mq);
+			else
+				mmc_blk_rw_rq_prep(mq->mqrq_cur, card, 0, mq);
+			areq = &mq->mqrq_cur->mmc_active;
+		} else
+			areq = NULL;
+		areq = mmc_start_req(card->host, areq, (int *) &status);
+		if (!areq) {
+			if (status == MMC_BLK_NEW_REQUEST)
+				mq->flags |= MMC_QUEUE_NEW_REQUEST;
+			return 0;
+		}
+
+		mq_rq = container_of(areq, struct mmc_queue_req, mmc_active);
+		brq = &mq_rq->brq;
+		req = mq_rq->req;
+		type = rq_data_dir(req) == READ ? MMC_BLK_READ : MMC_BLK_WRITE;
+		mmc_queue_bounce_post(mq_rq);
+
+		switch (status) {
+		case MMC_BLK_SUCCESS:
+		case MMC_BLK_PARTIAL:
+			/*
+			 * A block was successfully transferred.
+			 */
+			mmc_blk_reset_success(md, type);
+
+			if (mmc_packed_cmd(mq_rq->cmd_type)) {
+				ret = mmc_blk_end_packed_req(mq_rq);
+				break;
+			} else {
+				ret = blk_end_request(req, 0,
+						brq->data.bytes_xfered);
+			}
+
+			/*
+			 * If the blk_end_request function returns non-zero even
+			 * though all data has been transferred and no errors
+			 * were returned by the host controller, it's a bug.
+			 */
+			if (status == MMC_BLK_SUCCESS && ret) {
+				pr_err("%s BUG rq_tot %d d_xfer %d\n",
+				       __func__, blk_rq_bytes(req),
+				       brq->data.bytes_xfered);
+				rqc = NULL;
+				goto cmd_abort;
+			}
+			break;
+		case MMC_BLK_CMD_ERR:
+			ret = mmc_blk_cmd_err(md, card, brq, req, ret);
+			if (!mmc_blk_reset(md, card->host, type))
+				break;
+			goto cmd_abort;
+		case MMC_BLK_RETRY:
+			if (retry++ < 5)
+				break;
+			/* Fall through */
+		case MMC_BLK_ABORT:
+			if (!mmc_blk_reset(md, card->host, type))
+				break;
+			goto cmd_abort;
+		case MMC_BLK_DATA_ERR: {
+			int err;
+
+			err = mmc_blk_reset(md, card->host, type);
+			if (!err)
+				break;
+			if (err == -ENODEV ||
+				mmc_packed_cmd(mq_rq->cmd_type))
+				goto cmd_abort;
+			/* Fall through */
+		}
+		case MMC_BLK_ECC_ERR:
+			if (brq->data.blocks > 1) {
+				/* Redo read one sector at a time */
+				pr_warning("%s: retrying using single block read\n",
+					   req->rq_disk->disk_name);
+				disable_multi = 1;
+				break;
+			}
+			/*
+			 * After an error, we redo I/O one sector at a
+			 * time, so we only reach here after trying to
+			 * read a single sector.
+			 */
+			ret = blk_end_request(req, -EIO,
+						brq->data.blksz);
+			if (!ret)
+				goto start_new_req;
+			break;
+		case MMC_BLK_NOMEDIUM:
+			goto cmd_abort;
+		default:
+			pr_err("%s: Unhandled return value (%d)",
+					req->rq_disk->disk_name, status);
+			goto cmd_abort;
+		}
+
+		if (ret) {
+			if (mmc_packed_cmd(mq_rq->cmd_type)) {
+				if (!mq_rq->packed->retries)
+					goto cmd_abort;
+				mmc_blk_packed_hdr_wrq_prep(mq_rq, card, mq);
+				mmc_start_req(card->host,
+					      &mq_rq->mmc_active, NULL);
+			} else {
+
+				/*
+				 * In case of a incomplete request
+				 * prepare it again and resend.
+				 */
+				mmc_blk_rw_rq_prep(mq_rq, card,
+						disable_multi, mq);
+				mmc_start_req(card->host,
+						&mq_rq->mmc_active, NULL);
+			}
+		}
+	} while (ret);
+
+	return 1;
+
+ cmd_abort:
+	if (mmc_packed_cmd(mq_rq->cmd_type)) {
+		mmc_blk_abort_packed_req(mq_rq);
+	} else {
+		if (mmc_card_removed(card))
+			req->cmd_flags |= REQ_QUIET;
+		while (ret)
+			ret = blk_end_request(req, -EIO,
+					blk_rq_cur_bytes(req));
+	}
 
-	spin_lock_irq(&md->lock);
-	while (ret)
-		ret = __blk_end_request(req, -EIO, blk_rq_cur_bytes(req));
-	spin_unlock_irq(&md->lock);
+ start_new_req:
+	if (rqc) {
+		if (mmc_card_removed(card)) {
+			rqc->cmd_flags |= REQ_QUIET;
+			blk_end_request_all(rqc, -EIO);
+		} else {
+			/*
+			 * If current request is packed, it needs to put back.
+			 */
+			if (mmc_packed_cmd(mq->mqrq_cur->cmd_type))
+				mmc_blk_revert_packed_req(mq, mq->mqrq_cur);
+
+			mmc_blk_rw_rq_prep(mq->mqrq_cur, card, 0, mq);
+			mmc_start_req(card->host,
+				      &mq->mqrq_cur->mmc_active, NULL);
+		}
+	}
 
 	return 0;
 }
 
+static int mmc_blk_issue_rq(struct mmc_queue *mq, struct request *req)
+{
+	int ret;
+	struct mmc_blk_data *md = mq->data;
+	struct mmc_card *card = md->queue.card;
+	struct mmc_host *host = card->host;
+	unsigned long flags;
+	unsigned int cmd_flags = req ? req->cmd_flags : 0;
+
+	if (req && !mq->mqrq_prev->req)
+		/* claim host only for the first request */
+		mmc_get_card(card);
+
+	ret = mmc_blk_part_switch(card, md);
+	if (ret) {
+		if (req) {
+			blk_end_request_all(req, -EIO);
+		}
+		ret = 0;
+		goto out;
+	}
+
+	mq->flags &= ~MMC_QUEUE_NEW_REQUEST;
+	if (cmd_flags & REQ_DISCARD) {
+		/* complete ongoing async transfer before issuing discard */
+		if (card->host->areq)
+			mmc_blk_issue_rw_rq(mq, NULL);
+		if (req->cmd_flags & REQ_SECURE &&
+			!(card->quirks & MMC_QUIRK_SEC_ERASE_TRIM_BROKEN))
+			ret = mmc_blk_issue_secdiscard_rq(mq, req);
+		else
+			ret = mmc_blk_issue_discard_rq(mq, req);
+	} else if (cmd_flags & REQ_FLUSH) {
+		/* complete ongoing async transfer before issuing flush */
+		if (card->host->areq)
+			mmc_blk_issue_rw_rq(mq, NULL);
+		ret = mmc_blk_issue_flush(mq, req);
+	} else {
+		if (!req && host->areq) {
+			spin_lock_irqsave(&host->context_info.lock, flags);
+			host->context_info.is_waiting_last_req = true;
+			spin_unlock_irqrestore(&host->context_info.lock, flags);
+		}
+		ret = mmc_blk_issue_rw_rq(mq, req);
+	}
+
+out:
+	if ((!req && !(mq->flags & MMC_QUEUE_NEW_REQUEST)) ||
+	     (cmd_flags & MMC_REQ_SPECIAL_MASK))
+		/*
+		 * Release host when there are no more requests
+		 * and after special request(discard, flush) is done.
+		 * In case sepecial request, there is no reentry to
+		 * the 'mmc_blk_issue_rq' with 'mqrq_prev->req'.
+		 */
+		mmc_put_card(card);
+	return ret;
+}
 
 static inline int mmc_blk_readonly(struct mmc_card *card)
 {
@@ -453,13 +2066,18 @@ static inline int mmc_blk_readonly(struct mmc_card *card)
 	       !(card->csd.cmdclass & CCC_BLOCK_WRITE);
 }
 
-static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
+static struct mmc_blk_data *mmc_blk_alloc_req(struct mmc_card *card,
+					      struct device *parent,
+					      sector_t size,
+					      bool default_ro,
+					      const char *subname,
+					      int area_type)
 {
 	struct mmc_blk_data *md;
 	int devidx, ret;
 
-	devidx = find_first_zero_bit(dev_use, MMC_NUM_MINORS);
-	if (devidx >= MMC_NUM_MINORS)
+	devidx = find_first_zero_bit(dev_use, max_devices);
+	if (devidx >= max_devices)
 		return ERR_PTR(-ENOSPC);
 	__set_bit(devidx, dev_use);
 
@@ -469,6 +2087,20 @@ static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
 		goto out;
 	}
 
+	/*
+	 * !subname implies we are creating main mmc_blk_data that will be
+	 * associated with mmc_card with mmc_set_drvdata. Due to device
+	 * partitions, devidx will not coincide with a per-physical card
+	 * index anymore so we keep track of a name index.
+	 */
+	if (!subname) {
+		md->name_idx = find_first_zero_bit(name_use, max_devices);
+		__set_bit(md->name_idx, name_use);
+	} else
+		md->name_idx = ((struct mmc_blk_data *)
+				dev_to_disk(parent)->private_data)->name_idx;
+
+	md->area_type = area_type;
 
 	/*
 	 * Set the read-only status based on the supported commands
@@ -476,16 +2108,17 @@ static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
 	 */
 	md->read_only = mmc_blk_readonly(card);
 
-	md->disk = alloc_disk(1 << MMC_SHIFT);
+	md->disk = alloc_disk(perdev_minors);
 	if (md->disk == NULL) {
 		ret = -ENOMEM;
 		goto err_kfree;
 	}
 
 	spin_lock_init(&md->lock);
+	INIT_LIST_HEAD(&md->part);
 	md->usage = 1;
 
-	ret = mmc_init_queue(&md->queue, card, &md->lock);
+	ret = mmc_init_queue(&md->queue, card, &md->lock, subname);
 	if (ret)
 		goto err_putdisk;
 
@@ -493,11 +2126,14 @@ static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
 	md->queue.data = md;
 
 	md->disk->major	= MMC_BLOCK_MAJOR;
-	md->disk->first_minor = devidx << MMC_SHIFT;
+	md->disk->first_minor = devidx * perdev_minors;
 	md->disk->fops = &mmc_bdops;
 	md->disk->private_data = md;
 	md->disk->queue = md->queue.queue;
-	md->disk->driverfs_dev = &card->dev;
+	md->disk->driverfs_dev = parent;
+	set_disk_ro(md->disk, md->read_only || default_ro);
+	if (area_type & MMC_BLK_DATA_AREA_RPMB)
+		md->disk->flags |= GENHD_FL_NO_PART_SCAN;
 
 	/*
 	 * As discussed on lkml, GENHD_FL_REMOVABLE should:
@@ -511,65 +2147,283 @@ static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
 	 * messages to tell when the card is present.
 	 */
 
-	sprintf(md->disk->disk_name, "mmcblk%d", devidx);
+	snprintf(md->disk->disk_name, sizeof(md->disk->disk_name),
+		 "mmcblk%d%s", md->name_idx, subname ? subname : "");
+
+	if (mmc_card_mmc(card))
+		blk_queue_logical_block_size(md->queue.queue,
+					     card->ext_csd.data_sector_size);
+	else
+		blk_queue_logical_block_size(md->queue.queue, 512);
+
+	set_capacity(md->disk, size);
+
+	if (mmc_host_cmd23(card->host)) {
+		if (mmc_card_mmc(card) ||
+		    (mmc_card_sd(card) &&
+		     card->scr.cmds & SD_SCR_CMD23_SUPPORT))
+			md->flags |= MMC_BLK_CMD23;
+	}
+
+	if (mmc_card_mmc(card) &&
+	    md->flags & MMC_BLK_CMD23 &&
+	    ((card->ext_csd.rel_param & EXT_CSD_WR_REL_PARAM_EN) ||
+	     card->ext_csd.rel_sectors)) {
+		md->flags |= MMC_BLK_REL_WR;
+		blk_queue_flush(md->queue.queue, REQ_FLUSH | REQ_FUA);
+	}
+
+	if (mmc_card_mmc(card) &&
+	    (area_type == MMC_BLK_DATA_AREA_MAIN) &&
+	    (md->flags & MMC_BLK_CMD23) &&
+	    card->ext_csd.packed_event_en) {
+		if (!mmc_packed_init(&md->queue, card))
+			md->flags |= MMC_BLK_PACKED_CMD;
+	}
+
+	return md;
+
+ err_putdisk:
+	put_disk(md->disk);
+ err_kfree:
+	kfree(md);
+ out:
+	return ERR_PTR(ret);
+}
 
-	blk_queue_hardsect_size(md->queue.queue, 512);
+static struct mmc_blk_data *mmc_blk_alloc(struct mmc_card *card)
+{
+	sector_t size;
+	struct mmc_blk_data *md;
 
 	if (!mmc_card_sd(card) && mmc_card_blockaddr(card)) {
 		/*
 		 * The EXT_CSD sector count is in number or 512 byte
 		 * sectors.
 		 */
-		set_capacity(md->disk, card->ext_csd.sectors);
+		size = card->ext_csd.sectors;
 	} else {
 		/*
 		 * The CSD capacity field is in units of read_blkbits.
 		 * set_capacity takes units of 512 bytes.
 		 */
-		set_capacity(md->disk,
-			card->csd.capacity << (card->csd.read_blkbits - 9));
+		size = card->csd.capacity << (card->csd.read_blkbits - 9);
 	}
+
+	md = mmc_blk_alloc_req(card, &card->dev, size, false, NULL,
+					MMC_BLK_DATA_AREA_MAIN);
 	return md;
+}
 
- err_putdisk:
-	put_disk(md->disk);
- err_kfree:
-	kfree(md);
- out:
-	return ERR_PTR(ret);
+static int mmc_blk_alloc_part(struct mmc_card *card,
+			      struct mmc_blk_data *md,
+			      unsigned int part_type,
+			      sector_t size,
+			      bool default_ro,
+			      const char *subname,
+			      int area_type)
+{
+	char cap_str[10];
+	struct mmc_blk_data *part_md;
+
+	part_md = mmc_blk_alloc_req(card, disk_to_dev(md->disk), size, default_ro,
+				    subname, area_type);
+	if (IS_ERR(part_md))
+		return PTR_ERR(part_md);
+	part_md->part_type = part_type;
+	list_add(&part_md->part, &md->part);
+
+	string_get_size((u64)get_capacity(part_md->disk) << 9, STRING_UNITS_2,
+			cap_str, sizeof(cap_str));
+	pr_info("%s: %s %s partition %u %s\n",
+	       part_md->disk->disk_name, mmc_card_id(card),
+	       mmc_card_name(card), part_md->part_type, cap_str);
+	return 0;
 }
 
-static int
-mmc_blk_set_blksize(struct mmc_blk_data *md, struct mmc_card *card)
+/* MMC Physical partitions consist of two boot partitions and
+ * up to four general purpose partitions.
+ * For each partition enabled in EXT_CSD a block device will be allocatedi
+ * to provide access to the partition.
+ */
+
+static int mmc_blk_alloc_parts(struct mmc_card *card, struct mmc_blk_data *md)
 {
-	struct mmc_command cmd;
-	int err;
+	int idx, ret = 0;
 
-	/* Block-addressed cards ignore MMC_SET_BLOCKLEN. */
-	if (mmc_card_blockaddr(card))
+	if (!mmc_card_mmc(card))
 		return 0;
 
-	mmc_claim_host(card->host);
-	cmd.opcode = MMC_SET_BLOCKLEN;
-	cmd.arg = 512;
-	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
-	err = mmc_wait_for_cmd(card->host, &cmd, 5);
-	mmc_release_host(card->host);
+	for (idx = 0; idx < card->nr_parts; idx++) {
+		if (card->part[idx].size) {
+			ret = mmc_blk_alloc_part(card, md,
+				card->part[idx].part_cfg,
+				card->part[idx].size >> 9,
+				card->part[idx].force_ro,
+				card->part[idx].name,
+				card->part[idx].area_type);
+			if (ret)
+				return ret;
+		}
+	}
 
-	if (err) {
-		printk(KERN_ERR "%s: unable to set block size to %d: %d\n",
-			md->disk->disk_name, cmd.arg, err);
-		return -EINVAL;
+	return ret;
+}
+
+static void mmc_blk_remove_req(struct mmc_blk_data *md)
+{
+	struct mmc_card *card;
+
+	if (md) {
+		/*
+		 * Flush remaining requests and free queues. It
+		 * is freeing the queue that stops new requests
+		 * from being accepted.
+		 */
+		card = md->queue.card;
+		mmc_cleanup_queue(&md->queue);
+		if (md->flags & MMC_BLK_PACKED_CMD)
+			mmc_packed_clean(&md->queue);
+		if (md->disk->flags & GENHD_FL_UP) {
+			device_remove_file(disk_to_dev(md->disk), &md->force_ro);
+			if ((md->area_type & MMC_BLK_DATA_AREA_BOOT) &&
+					card->ext_csd.boot_ro_lockable)
+				device_remove_file(disk_to_dev(md->disk),
+					&md->power_ro_lock);
+
+			del_gendisk(md->disk);
+		}
+		mmc_blk_put(md);
 	}
+}
 
-	return 0;
+static void mmc_blk_remove_parts(struct mmc_card *card,
+				 struct mmc_blk_data *md)
+{
+	struct list_head *pos, *q;
+	struct mmc_blk_data *part_md;
+
+	__clear_bit(md->name_idx, name_use);
+	list_for_each_safe(pos, q, &md->part) {
+		part_md = list_entry(pos, struct mmc_blk_data, part);
+		list_del(pos);
+		mmc_blk_remove_req(part_md);
+	}
 }
 
-static int mmc_blk_probe(struct mmc_card *card)
+static int mmc_add_disk(struct mmc_blk_data *md)
 {
-	struct mmc_blk_data *md;
-	int err;
+	int ret;
+	struct mmc_card *card = md->queue.card;
 
+	add_disk(md->disk);
+	md->force_ro.show = force_ro_show;
+	md->force_ro.store = force_ro_store;
+	sysfs_attr_init(&md->force_ro.attr);
+	md->force_ro.attr.name = "force_ro";
+	md->force_ro.attr.mode = S_IRUGO | S_IWUSR;
+	ret = device_create_file(disk_to_dev(md->disk), &md->force_ro);
+	if (ret)
+		goto force_ro_fail;
+
+	if ((md->area_type & MMC_BLK_DATA_AREA_BOOT) &&
+	     card->ext_csd.boot_ro_lockable) {
+		umode_t mode;
+
+		if (card->ext_csd.boot_ro_lock & EXT_CSD_BOOT_WP_B_PWR_WP_DIS)
+			mode = S_IRUGO;
+		else
+			mode = S_IRUGO | S_IWUSR;
+
+		md->power_ro_lock.show = power_ro_lock_show;
+		md->power_ro_lock.store = power_ro_lock_store;
+		sysfs_attr_init(&md->power_ro_lock.attr);
+		md->power_ro_lock.attr.mode = mode;
+		md->power_ro_lock.attr.name =
+					"ro_lock_until_next_power_on";
+		ret = device_create_file(disk_to_dev(md->disk),
+				&md->power_ro_lock);
+		if (ret)
+			goto power_ro_lock_fail;
+	}
+	return ret;
+
+power_ro_lock_fail:
+	device_remove_file(disk_to_dev(md->disk), &md->force_ro);
+force_ro_fail:
+	del_gendisk(md->disk);
+
+	return ret;
+}
+
+#define CID_MANFID_SANDISK	0x2
+#define CID_MANFID_TOSHIBA	0x11
+#define CID_MANFID_MICRON	0x13
+#define CID_MANFID_SAMSUNG	0x15
+
+static const struct mmc_fixup blk_fixups[] =
+{
+	MMC_FIXUP("SEM02G", CID_MANFID_SANDISK, 0x100, add_quirk,
+		  MMC_QUIRK_INAND_CMD38),
+	MMC_FIXUP("SEM04G", CID_MANFID_SANDISK, 0x100, add_quirk,
+		  MMC_QUIRK_INAND_CMD38),
+	MMC_FIXUP("SEM08G", CID_MANFID_SANDISK, 0x100, add_quirk,
+		  MMC_QUIRK_INAND_CMD38),
+	MMC_FIXUP("SEM16G", CID_MANFID_SANDISK, 0x100, add_quirk,
+		  MMC_QUIRK_INAND_CMD38),
+	MMC_FIXUP("SEM32G", CID_MANFID_SANDISK, 0x100, add_quirk,
+		  MMC_QUIRK_INAND_CMD38),
+
+	/*
+	 * Some MMC cards experience performance degradation with CMD23
+	 * instead of CMD12-bounded multiblock transfers. For now we'll
+	 * black list what's bad...
+	 * - Certain Toshiba cards.
+	 *
+	 * N.B. This doesn't affect SD cards.
+	 */
+	MMC_FIXUP("MMC08G", CID_MANFID_TOSHIBA, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_BLK_NO_CMD23),
+	MMC_FIXUP("MMC16G", CID_MANFID_TOSHIBA, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_BLK_NO_CMD23),
+	MMC_FIXUP("MMC32G", CID_MANFID_TOSHIBA, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_BLK_NO_CMD23),
+
+	/*
+	 * Some Micron MMC cards needs longer data read timeout than
+	 * indicated in CSD.
+	 */
+	MMC_FIXUP(CID_NAME_ANY, CID_MANFID_MICRON, 0x200, add_quirk_mmc,
+		  MMC_QUIRK_LONG_READ_TIME),
+
+	/*
+	 * On these Samsung MoviNAND parts, performing secure erase or
+	 * secure trim can result in unrecoverable corruption due to a
+	 * firmware bug.
+	 */
+	MMC_FIXUP("M8G2FA", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("MAG4FA", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("MBG8FA", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("MCGAFA", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("VAL00M", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("VYL00M", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("KYL00M", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+	MMC_FIXUP("VZL00M", CID_MANFID_SAMSUNG, CID_OEMID_ANY, add_quirk_mmc,
+		  MMC_QUIRK_SEC_ERASE_TRIM_BROKEN),
+
+	END_FIXUP
+};
+
+static int mmc_blk_probe(struct mmc_card *card)
+{
+	struct mmc_blk_data *md, *part_md;
 	char cap_str[10];
 
 	/*
@@ -582,60 +2436,102 @@ static int mmc_blk_probe(struct mmc_card *card)
 	if (IS_ERR(md))
 		return PTR_ERR(md);
 
-	err = mmc_blk_set_blksize(md, card);
-	if (err)
-		goto out;
-
 	string_get_size((u64)get_capacity(md->disk) << 9, STRING_UNITS_2,
 			cap_str, sizeof(cap_str));
-	printk(KERN_INFO "%s: %s %s %s %s\n",
+	pr_info("%s: %s %s %s %s\n",
 		md->disk->disk_name, mmc_card_id(card), mmc_card_name(card),
 		cap_str, md->read_only ? "(ro)" : "");
 
+	if (mmc_blk_alloc_parts(card, md))
+		goto out;
+
 	mmc_set_drvdata(card, md);
-	add_disk(md->disk);
+	mmc_fixup_device(card, blk_fixups);
+
+	if (mmc_add_disk(md))
+		goto out;
+
+	list_for_each_entry(part_md, &md->part, part) {
+		if (mmc_add_disk(part_md))
+			goto out;
+	}
+
+	pm_runtime_set_autosuspend_delay(&card->dev, 3000);
+	pm_runtime_use_autosuspend(&card->dev);
+
+	/*
+	 * Don't enable runtime PM for SD-combo cards here. Leave that
+	 * decision to be taken during the SDIO init sequence instead.
+	 */
+	if (card->type != MMC_TYPE_SD_COMBO) {
+		pm_runtime_set_active(&card->dev);
+		pm_runtime_enable(&card->dev);
+	}
+
 	return 0;
 
  out:
-	mmc_blk_put(md);
-
-	return err;
+	mmc_blk_remove_parts(card, md);
+	mmc_blk_remove_req(md);
+	return 0;
 }
 
 static void mmc_blk_remove(struct mmc_card *card)
 {
 	struct mmc_blk_data *md = mmc_get_drvdata(card);
 
-	if (md) {
-		/* Stop new requests from getting into the queue */
-		del_gendisk(md->disk);
-
-		/* Then flush out any already in there */
-		mmc_cleanup_queue(&md->queue);
-
-		mmc_blk_put(md);
-	}
+	mmc_blk_remove_parts(card, md);
+	pm_runtime_get_sync(&card->dev);
+	mmc_claim_host(card->host);
+	mmc_blk_part_switch(card, md);
+	mmc_release_host(card->host);
+	if (card->type != MMC_TYPE_SD_COMBO)
+		pm_runtime_disable(&card->dev);
+	pm_runtime_put_noidle(&card->dev);
+	mmc_blk_remove_req(md);
 	mmc_set_drvdata(card, NULL);
 }
 
-#ifdef CONFIG_PM
-static int mmc_blk_suspend(struct mmc_card *card, pm_message_t state)
+static int _mmc_blk_suspend(struct mmc_card *card)
 {
+	struct mmc_blk_data *part_md;
 	struct mmc_blk_data *md = mmc_get_drvdata(card);
 
 	if (md) {
 		mmc_queue_suspend(&md->queue);
+		list_for_each_entry(part_md, &md->part, part) {
+			mmc_queue_suspend(&part_md->queue);
+		}
 	}
 	return 0;
 }
 
+static void mmc_blk_shutdown(struct mmc_card *card)
+{
+	_mmc_blk_suspend(card);
+}
+
+#ifdef CONFIG_PM
+static int mmc_blk_suspend(struct mmc_card *card)
+{
+	return _mmc_blk_suspend(card);
+}
+
 static int mmc_blk_resume(struct mmc_card *card)
 {
+	struct mmc_blk_data *part_md;
 	struct mmc_blk_data *md = mmc_get_drvdata(card);
 
 	if (md) {
-		mmc_blk_set_blksize(md, card);
+		/*
+		 * Resume involves the card going into idle state,
+		 * so current partition is always the main one.
+		 */
+		md->part_curr = md->part_type;
 		mmc_queue_resume(&md->queue);
+		list_for_each_entry(part_md, &md->part, part) {
+			mmc_queue_resume(&part_md->queue);
+		}
 	}
 	return 0;
 }
@@ -652,12 +2548,18 @@ static struct mmc_driver mmc_driver = {
 	.remove		= mmc_blk_remove,
 	.suspend	= mmc_blk_suspend,
 	.resume		= mmc_blk_resume,
+	.shutdown	= mmc_blk_shutdown,
 };
 
 static int __init mmc_blk_init(void)
 {
 	int res;
 
+	if (perdev_minors != CONFIG_MMC_BLOCK_MINORS)
+		pr_info("mmcblk: using %d minors per device\n", perdev_minors);
+
+	max_devices = 256 / perdev_minors;
+
 	res = register_blkdev(MMC_BLOCK_MAJOR, "mmc");
 	if (res)
 		goto out;
diff --git a/drivers/mmc/card/mmc_test.c b/drivers/mmc/card/mmc_test.c
index b9f1e84897c..0c0fc52d42c 100644
--- a/drivers/mmc/card/mmc_test.c
+++ b/drivers/mmc/card/mmc_test.c
@@ -13,8 +13,16 @@
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
+#include <linux/slab.h>
 
 #include <linux/scatterlist.h>
+#include <linux/swap.h>		/* For nr_free_buffer_pages() */
+#include <linux/list.h>
+
+#include <linux/debugfs.h>
+#include <linux/uaccess.h>
+#include <linux/seq_file.h>
+#include <linux/module.h>
 
 #define RESULT_OK		0
 #define RESULT_FAIL		1
@@ -24,6 +32,111 @@
 #define BUFFER_ORDER		2
 #define BUFFER_SIZE		(PAGE_SIZE << BUFFER_ORDER)
 
+/*
+ * Limit the test area size to the maximum MMC HC erase group size.  Note that
+ * the maximum SD allocation unit size is just 4MiB.
+ */
+#define TEST_AREA_MAX_SIZE (128 * 1024 * 1024)
+
+/**
+ * struct mmc_test_pages - pages allocated by 'alloc_pages()'.
+ * @page: first page in the allocation
+ * @order: order of the number of pages allocated
+ */
+struct mmc_test_pages {
+	struct page *page;
+	unsigned int order;
+};
+
+/**
+ * struct mmc_test_mem - allocated memory.
+ * @arr: array of allocations
+ * @cnt: number of allocations
+ */
+struct mmc_test_mem {
+	struct mmc_test_pages *arr;
+	unsigned int cnt;
+};
+
+/**
+ * struct mmc_test_area - information for performance tests.
+ * @max_sz: test area size (in bytes)
+ * @dev_addr: address on card at which to do performance tests
+ * @max_tfr: maximum transfer size allowed by driver (in bytes)
+ * @max_segs: maximum segments allowed by driver in scatterlist @sg
+ * @max_seg_sz: maximum segment size allowed by driver
+ * @blocks: number of (512 byte) blocks currently mapped by @sg
+ * @sg_len: length of currently mapped scatterlist @sg
+ * @mem: allocated memory
+ * @sg: scatterlist
+ */
+struct mmc_test_area {
+	unsigned long max_sz;
+	unsigned int dev_addr;
+	unsigned int max_tfr;
+	unsigned int max_segs;
+	unsigned int max_seg_sz;
+	unsigned int blocks;
+	unsigned int sg_len;
+	struct mmc_test_mem *mem;
+	struct scatterlist *sg;
+};
+
+/**
+ * struct mmc_test_transfer_result - transfer results for performance tests.
+ * @link: double-linked list
+ * @count: amount of group of sectors to check
+ * @sectors: amount of sectors to check in one group
+ * @ts: time values of transfer
+ * @rate: calculated transfer rate
+ * @iops: I/O operations per second (times 100)
+ */
+struct mmc_test_transfer_result {
+	struct list_head link;
+	unsigned int count;
+	unsigned int sectors;
+	struct timespec ts;
+	unsigned int rate;
+	unsigned int iops;
+};
+
+/**
+ * struct mmc_test_general_result - results for tests.
+ * @link: double-linked list
+ * @card: card under test
+ * @testcase: number of test case
+ * @result: result of test run
+ * @tr_lst: transfer measurements if any as mmc_test_transfer_result
+ */
+struct mmc_test_general_result {
+	struct list_head link;
+	struct mmc_card *card;
+	int testcase;
+	int result;
+	struct list_head tr_lst;
+};
+
+/**
+ * struct mmc_test_dbgfs_file - debugfs related file.
+ * @link: double-linked list
+ * @card: card under test
+ * @file: file created under debugfs
+ */
+struct mmc_test_dbgfs_file {
+	struct list_head link;
+	struct mmc_card *card;
+	struct dentry *file;
+};
+
+/**
+ * struct mmc_test_card - test information.
+ * @card: card under test
+ * @scratch: transfer buffer
+ * @buffer: transfer buffer
+ * @highmem: buffer for highmem tests
+ * @area: information for performance tests
+ * @gr: pointer to results of current testcase
+ */
 struct mmc_test_card {
 	struct mmc_card	*card;
 
@@ -32,6 +145,29 @@ struct mmc_test_card {
 #ifdef CONFIG_HIGHMEM
 	struct page	*highmem;
 #endif
+	struct mmc_test_area		area;
+	struct mmc_test_general_result	*gr;
+};
+
+enum mmc_test_prep_media {
+	MMC_TEST_PREP_NONE = 0,
+	MMC_TEST_PREP_WRITE_FULL = 1 << 0,
+	MMC_TEST_PREP_ERASE = 1 << 1,
+};
+
+struct mmc_test_multiple_rw {
+	unsigned int *sg_len;
+	unsigned int *bs;
+	unsigned int len;
+	unsigned int size;
+	bool do_write;
+	bool do_nonblock_req;
+	enum mmc_test_prep_media prepare;
+};
+
+struct mmc_test_async_req {
+	struct mmc_async_req areq;
+	struct mmc_test_card *test;
 };
 
 /*******************************************************************/
@@ -43,17 +179,7 @@ struct mmc_test_card {
  */
 static int mmc_test_set_blksize(struct mmc_test_card *test, unsigned size)
 {
-	struct mmc_command cmd;
-	int ret;
-
-	cmd.opcode = MMC_SET_BLOCKLEN;
-	cmd.arg = size;
-	cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
-	ret = mmc_wait_for_cmd(test->card->host, &cmd, 0);
-	if (ret)
-		return ret;
-
-	return 0;
+	return mmc_set_blocklen(test->card, size);
 }
 
 /*
@@ -74,6 +200,9 @@ static void mmc_test_prepare_mrq(struct mmc_test_card *test,
 	}
 
 	mrq->cmd->arg = dev_addr;
+	if (!mmc_card_blockaddr(test->card))
+		mrq->cmd->arg <<= 9;
+
 	mrq->cmd->flags = MMC_RSP_R1 | MMC_CMD_ADTC;
 
 	if (blocks == 1)
@@ -93,13 +222,19 @@ static void mmc_test_prepare_mrq(struct mmc_test_card *test,
 	mmc_set_data_timeout(mrq->data, test->card);
 }
 
+static int mmc_test_busy(struct mmc_command *cmd)
+{
+	return !(cmd->resp[0] & R1_READY_FOR_DATA) ||
+		(R1_CURRENT_STATE(cmd->resp[0]) == R1_STATE_PRG);
+}
+
 /*
  * Wait for the card to finish the busy state
  */
 static int mmc_test_wait_busy(struct mmc_test_card *test)
 {
 	int ret, busy;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	busy = 0;
 	do {
@@ -113,13 +248,14 @@ static int mmc_test_wait_busy(struct mmc_test_card *test)
 		if (ret)
 			break;
 
-		if (!busy && !(cmd.resp[0] & R1_READY_FOR_DATA)) {
+		if (!busy && mmc_test_busy(&cmd)) {
 			busy = 1;
-			printk(KERN_INFO "%s: Warning: Host did not "
-				"wait for busy state to end.\n",
-				mmc_hostname(test->card->host));
+			if (test->card->host->caps & MMC_CAP_WAIT_WHILE_BUSY)
+				pr_info("%s: Warning: Host did not "
+					"wait for busy state to end.\n",
+					mmc_hostname(test->card->host));
 		}
-	} while (!(cmd.resp[0] & R1_READY_FOR_DATA));
+	} while (mmc_test_busy(&cmd));
 
 	return ret;
 }
@@ -132,18 +268,13 @@ static int mmc_test_buffer_transfer(struct mmc_test_card *test,
 {
 	int ret;
 
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_command stop;
-	struct mmc_data data;
+	struct mmc_request mrq = {0};
+	struct mmc_command cmd = {0};
+	struct mmc_command stop = {0};
+	struct mmc_data data = {0};
 
 	struct scatterlist sg;
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
-	memset(&stop, 0, sizeof(struct mmc_command));
-
 	mrq.cmd = &cmd;
 	mrq.data = &data;
 	mrq.stop = &stop;
@@ -166,6 +297,311 @@ static int mmc_test_buffer_transfer(struct mmc_test_card *test,
 	return 0;
 }
 
+static void mmc_test_free_mem(struct mmc_test_mem *mem)
+{
+	if (!mem)
+		return;
+	while (mem->cnt--)
+		__free_pages(mem->arr[mem->cnt].page,
+			     mem->arr[mem->cnt].order);
+	kfree(mem->arr);
+	kfree(mem);
+}
+
+/*
+ * Allocate a lot of memory, preferably max_sz but at least min_sz.  In case
+ * there isn't much memory do not exceed 1/16th total lowmem pages.  Also do
+ * not exceed a maximum number of segments and try not to make segments much
+ * bigger than maximum segment size.
+ */
+static struct mmc_test_mem *mmc_test_alloc_mem(unsigned long min_sz,
+					       unsigned long max_sz,
+					       unsigned int max_segs,
+					       unsigned int max_seg_sz)
+{
+	unsigned long max_page_cnt = DIV_ROUND_UP(max_sz, PAGE_SIZE);
+	unsigned long min_page_cnt = DIV_ROUND_UP(min_sz, PAGE_SIZE);
+	unsigned long max_seg_page_cnt = DIV_ROUND_UP(max_seg_sz, PAGE_SIZE);
+	unsigned long page_cnt = 0;
+	unsigned long limit = nr_free_buffer_pages() >> 4;
+	struct mmc_test_mem *mem;
+
+	if (max_page_cnt > limit)
+		max_page_cnt = limit;
+	if (min_page_cnt > max_page_cnt)
+		min_page_cnt = max_page_cnt;
+
+	if (max_seg_page_cnt > max_page_cnt)
+		max_seg_page_cnt = max_page_cnt;
+
+	if (max_segs > max_page_cnt)
+		max_segs = max_page_cnt;
+
+	mem = kzalloc(sizeof(struct mmc_test_mem), GFP_KERNEL);
+	if (!mem)
+		return NULL;
+
+	mem->arr = kzalloc(sizeof(struct mmc_test_pages) * max_segs,
+			   GFP_KERNEL);
+	if (!mem->arr)
+		goto out_free;
+
+	while (max_page_cnt) {
+		struct page *page;
+		unsigned int order;
+		gfp_t flags = GFP_KERNEL | GFP_DMA | __GFP_NOWARN |
+				__GFP_NORETRY;
+
+		order = get_order(max_seg_page_cnt << PAGE_SHIFT);
+		while (1) {
+			page = alloc_pages(flags, order);
+			if (page || !order)
+				break;
+			order -= 1;
+		}
+		if (!page) {
+			if (page_cnt < min_page_cnt)
+				goto out_free;
+			break;
+		}
+		mem->arr[mem->cnt].page = page;
+		mem->arr[mem->cnt].order = order;
+		mem->cnt += 1;
+		if (max_page_cnt <= (1UL << order))
+			break;
+		max_page_cnt -= 1UL << order;
+		page_cnt += 1UL << order;
+		if (mem->cnt >= max_segs) {
+			if (page_cnt < min_page_cnt)
+				goto out_free;
+			break;
+		}
+	}
+
+	return mem;
+
+out_free:
+	mmc_test_free_mem(mem);
+	return NULL;
+}
+
+/*
+ * Map memory into a scatterlist.  Optionally allow the same memory to be
+ * mapped more than once.
+ */
+static int mmc_test_map_sg(struct mmc_test_mem *mem, unsigned long size,
+			   struct scatterlist *sglist, int repeat,
+			   unsigned int max_segs, unsigned int max_seg_sz,
+			   unsigned int *sg_len, int min_sg_len)
+{
+	struct scatterlist *sg = NULL;
+	unsigned int i;
+	unsigned long sz = size;
+
+	sg_init_table(sglist, max_segs);
+	if (min_sg_len > max_segs)
+		min_sg_len = max_segs;
+
+	*sg_len = 0;
+	do {
+		for (i = 0; i < mem->cnt; i++) {
+			unsigned long len = PAGE_SIZE << mem->arr[i].order;
+
+			if (min_sg_len && (size / min_sg_len < len))
+				len = ALIGN(size / min_sg_len, 512);
+			if (len > sz)
+				len = sz;
+			if (len > max_seg_sz)
+				len = max_seg_sz;
+			if (sg)
+				sg = sg_next(sg);
+			else
+				sg = sglist;
+			if (!sg)
+				return -EINVAL;
+			sg_set_page(sg, mem->arr[i].page, len, 0);
+			sz -= len;
+			*sg_len += 1;
+			if (!sz)
+				break;
+		}
+	} while (sz && repeat);
+
+	if (sz)
+		return -EINVAL;
+
+	if (sg)
+		sg_mark_end(sg);
+
+	return 0;
+}
+
+/*
+ * Map memory into a scatterlist so that no pages are contiguous.  Allow the
+ * same memory to be mapped more than once.
+ */
+static int mmc_test_map_sg_max_scatter(struct mmc_test_mem *mem,
+				       unsigned long sz,
+				       struct scatterlist *sglist,
+				       unsigned int max_segs,
+				       unsigned int max_seg_sz,
+				       unsigned int *sg_len)
+{
+	struct scatterlist *sg = NULL;
+	unsigned int i = mem->cnt, cnt;
+	unsigned long len;
+	void *base, *addr, *last_addr = NULL;
+
+	sg_init_table(sglist, max_segs);
+
+	*sg_len = 0;
+	while (sz) {
+		base = page_address(mem->arr[--i].page);
+		cnt = 1 << mem->arr[i].order;
+		while (sz && cnt) {
+			addr = base + PAGE_SIZE * --cnt;
+			if (last_addr && last_addr + PAGE_SIZE == addr)
+				continue;
+			last_addr = addr;
+			len = PAGE_SIZE;
+			if (len > max_seg_sz)
+				len = max_seg_sz;
+			if (len > sz)
+				len = sz;
+			if (sg)
+				sg = sg_next(sg);
+			else
+				sg = sglist;
+			if (!sg)
+				return -EINVAL;
+			sg_set_page(sg, virt_to_page(addr), len, 0);
+			sz -= len;
+			*sg_len += 1;
+		}
+		if (i == 0)
+			i = mem->cnt;
+	}
+
+	if (sg)
+		sg_mark_end(sg);
+
+	return 0;
+}
+
+/*
+ * Calculate transfer rate in bytes per second.
+ */
+static unsigned int mmc_test_rate(uint64_t bytes, struct timespec *ts)
+{
+	uint64_t ns;
+
+	ns = ts->tv_sec;
+	ns *= 1000000000;
+	ns += ts->tv_nsec;
+
+	bytes *= 1000000000;
+
+	while (ns > UINT_MAX) {
+		bytes >>= 1;
+		ns >>= 1;
+	}
+
+	if (!ns)
+		return 0;
+
+	do_div(bytes, (uint32_t)ns);
+
+	return bytes;
+}
+
+/*
+ * Save transfer results for future usage
+ */
+static void mmc_test_save_transfer_result(struct mmc_test_card *test,
+	unsigned int count, unsigned int sectors, struct timespec ts,
+	unsigned int rate, unsigned int iops)
+{
+	struct mmc_test_transfer_result *tr;
+
+	if (!test->gr)
+		return;
+
+	tr = kmalloc(sizeof(struct mmc_test_transfer_result), GFP_KERNEL);
+	if (!tr)
+		return;
+
+	tr->count = count;
+	tr->sectors = sectors;
+	tr->ts = ts;
+	tr->rate = rate;
+	tr->iops = iops;
+
+	list_add_tail(&tr->link, &test->gr->tr_lst);
+}
+
+/*
+ * Print the transfer rate.
+ */
+static void mmc_test_print_rate(struct mmc_test_card *test, uint64_t bytes,
+				struct timespec *ts1, struct timespec *ts2)
+{
+	unsigned int rate, iops, sectors = bytes >> 9;
+	struct timespec ts;
+
+	ts = timespec_sub(*ts2, *ts1);
+
+	rate = mmc_test_rate(bytes, &ts);
+	iops = mmc_test_rate(100, &ts); /* I/O ops per sec x 100 */
+
+	pr_info("%s: Transfer of %u sectors (%u%s KiB) took %lu.%09lu "
+			 "seconds (%u kB/s, %u KiB/s, %u.%02u IOPS)\n",
+			 mmc_hostname(test->card->host), sectors, sectors >> 1,
+			 (sectors & 1 ? ".5" : ""), (unsigned long)ts.tv_sec,
+			 (unsigned long)ts.tv_nsec, rate / 1000, rate / 1024,
+			 iops / 100, iops % 100);
+
+	mmc_test_save_transfer_result(test, 1, sectors, ts, rate, iops);
+}
+
+/*
+ * Print the average transfer rate.
+ */
+static void mmc_test_print_avg_rate(struct mmc_test_card *test, uint64_t bytes,
+				    unsigned int count, struct timespec *ts1,
+				    struct timespec *ts2)
+{
+	unsigned int rate, iops, sectors = bytes >> 9;
+	uint64_t tot = bytes * count;
+	struct timespec ts;
+
+	ts = timespec_sub(*ts2, *ts1);
+
+	rate = mmc_test_rate(tot, &ts);
+	iops = mmc_test_rate(count * 100, &ts); /* I/O ops per sec x 100 */
+
+	pr_info("%s: Transfer of %u x %u sectors (%u x %u%s KiB) took "
+			 "%lu.%09lu seconds (%u kB/s, %u KiB/s, "
+			 "%u.%02u IOPS, sg_len %d)\n",
+			 mmc_hostname(test->card->host), count, sectors, count,
+			 sectors >> 1, (sectors & 1 ? ".5" : ""),
+			 (unsigned long)ts.tv_sec, (unsigned long)ts.tv_nsec,
+			 rate / 1000, rate / 1024, iops / 100, iops % 100,
+			 test->area.sg_len);
+
+	mmc_test_save_transfer_result(test, count, sectors, ts, rate, iops);
+}
+
+/*
+ * Return the card size in sectors.
+ */
+static unsigned int mmc_test_capacity(struct mmc_card *card)
+{
+	if (!mmc_card_sd(card) && mmc_card_blockaddr(card))
+		return card->ext_csd.sectors;
+	else
+		return card->csd.capacity << (card->csd.read_blkbits - 9);
+}
+
 /*******************************************************************/
 /*  Test preparation and cleanup                                   */
 /*******************************************************************/
@@ -190,7 +626,7 @@ static int __mmc_test_prepare(struct mmc_test_card *test, int write)
 	}
 
 	for (i = 0;i < BUFFER_SIZE / 512;i++) {
-		ret = mmc_test_buffer_transfer(test, test->buffer, i * 512, 512, 1);
+		ret = mmc_test_buffer_transfer(test, test->buffer, i, 512, 1);
 		if (ret)
 			return ret;
 	}
@@ -219,7 +655,7 @@ static int mmc_test_cleanup(struct mmc_test_card *test)
 	memset(test->buffer, 0, 512);
 
 	for (i = 0;i < BUFFER_SIZE / 512;i++) {
-		ret = mmc_test_buffer_transfer(test, test->buffer, i * 512, 512, 1);
+		ret = mmc_test_buffer_transfer(test, test->buffer, i, 512, 1);
 		if (ret)
 			return ret;
 	}
@@ -253,7 +689,7 @@ static void mmc_test_prepare_broken_mrq(struct mmc_test_card *test,
  * Checks that a normal transfer didn't have any errors
  */
 static int mmc_test_check_result(struct mmc_test_card *test,
-	struct mmc_request *mrq)
+				 struct mmc_request *mrq)
 {
 	int ret;
 
@@ -277,6 +713,17 @@ static int mmc_test_check_result(struct mmc_test_card *test,
 	return ret;
 }
 
+static int mmc_test_check_result_async(struct mmc_card *card,
+				       struct mmc_async_req *areq)
+{
+	struct mmc_test_async_req *test_async =
+		container_of(areq, struct mmc_test_async_req, areq);
+
+	mmc_test_wait_busy(test_async->test);
+
+	return mmc_test_check_result(test_async->test, areq->mrq);
+}
+
 /*
  * Checks that a "short transfer" behaved as expected
  */
@@ -312,21 +759,95 @@ static int mmc_test_check_broken_result(struct mmc_test_card *test,
 }
 
 /*
+ * Tests nonblock transfer with certain parameters
+ */
+static void mmc_test_nonblock_reset(struct mmc_request *mrq,
+				    struct mmc_command *cmd,
+				    struct mmc_command *stop,
+				    struct mmc_data *data)
+{
+	memset(mrq, 0, sizeof(struct mmc_request));
+	memset(cmd, 0, sizeof(struct mmc_command));
+	memset(data, 0, sizeof(struct mmc_data));
+	memset(stop, 0, sizeof(struct mmc_command));
+
+	mrq->cmd = cmd;
+	mrq->data = data;
+	mrq->stop = stop;
+}
+static int mmc_test_nonblock_transfer(struct mmc_test_card *test,
+				      struct scatterlist *sg, unsigned sg_len,
+				      unsigned dev_addr, unsigned blocks,
+				      unsigned blksz, int write, int count)
+{
+	struct mmc_request mrq1;
+	struct mmc_command cmd1;
+	struct mmc_command stop1;
+	struct mmc_data data1;
+
+	struct mmc_request mrq2;
+	struct mmc_command cmd2;
+	struct mmc_command stop2;
+	struct mmc_data data2;
+
+	struct mmc_test_async_req test_areq[2];
+	struct mmc_async_req *done_areq;
+	struct mmc_async_req *cur_areq = &test_areq[0].areq;
+	struct mmc_async_req *other_areq = &test_areq[1].areq;
+	int i;
+	int ret;
+
+	test_areq[0].test = test;
+	test_areq[1].test = test;
+
+	mmc_test_nonblock_reset(&mrq1, &cmd1, &stop1, &data1);
+	mmc_test_nonblock_reset(&mrq2, &cmd2, &stop2, &data2);
+
+	cur_areq->mrq = &mrq1;
+	cur_areq->err_check = mmc_test_check_result_async;
+	other_areq->mrq = &mrq2;
+	other_areq->err_check = mmc_test_check_result_async;
+
+	for (i = 0; i < count; i++) {
+		mmc_test_prepare_mrq(test, cur_areq->mrq, sg, sg_len, dev_addr,
+				     blocks, blksz, write);
+		done_areq = mmc_start_req(test->card->host, cur_areq, &ret);
+
+		if (ret || (!done_areq && i > 0))
+			goto err;
+
+		if (done_areq) {
+			if (done_areq->mrq == &mrq2)
+				mmc_test_nonblock_reset(&mrq2, &cmd2,
+							&stop2, &data2);
+			else
+				mmc_test_nonblock_reset(&mrq1, &cmd1,
+							&stop1, &data1);
+		}
+		done_areq = cur_areq;
+		cur_areq = other_areq;
+		other_areq = done_areq;
+		dev_addr += blocks;
+	}
+
+	done_areq = mmc_start_req(test->card->host, NULL, &ret);
+
+	return ret;
+err:
+	return ret;
+}
+
+/*
  * Tests a basic transfer with certain parameters
  */
 static int mmc_test_simple_transfer(struct mmc_test_card *test,
 	struct scatterlist *sg, unsigned sg_len, unsigned dev_addr,
 	unsigned blocks, unsigned blksz, int write)
 {
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_command stop;
-	struct mmc_data data;
-
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
-	memset(&stop, 0, sizeof(struct mmc_command));
+	struct mmc_request mrq = {0};
+	struct mmc_command cmd = {0};
+	struct mmc_command stop = {0};
+	struct mmc_data data = {0};
 
 	mrq.cmd = &cmd;
 	mrq.data = &data;
@@ -348,18 +869,13 @@ static int mmc_test_simple_transfer(struct mmc_test_card *test,
 static int mmc_test_broken_transfer(struct mmc_test_card *test,
 	unsigned blocks, unsigned blksz, int write)
 {
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_command stop;
-	struct mmc_data data;
+	struct mmc_request mrq = {0};
+	struct mmc_command cmd = {0};
+	struct mmc_command stop = {0};
+	struct mmc_data data = {0};
 
 	struct scatterlist sg;
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
-	memset(&stop, 0, sizeof(struct mmc_command));
-
 	mrq.cmd = &cmd;
 	mrq.data = &data;
 	mrq.stop = &stop;
@@ -426,7 +942,7 @@ static int mmc_test_transfer(struct mmc_test_card *test,
 		for (i = 0;i < sectors;i++) {
 			ret = mmc_test_buffer_transfer(test,
 				test->buffer + i * 512,
-				dev_addr + i * 512, 512, 0);
+				dev_addr + i, 512, 0);
 			if (ret)
 				return ret;
 		}
@@ -889,8 +1405,956 @@ static int mmc_test_multi_read_high(struct mmc_test_card *test)
 	return 0;
 }
 
+#else
+
+static int mmc_test_no_highmem(struct mmc_test_card *test)
+{
+	pr_info("%s: Highmem not configured - test skipped\n",
+	       mmc_hostname(test->card->host));
+	return 0;
+}
+
 #endif /* CONFIG_HIGHMEM */
 
+/*
+ * Map sz bytes so that it can be transferred.
+ */
+static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz,
+			     int max_scatter, int min_sg_len)
+{
+	struct mmc_test_area *t = &test->area;
+	int err;
+
+	t->blocks = sz >> 9;
+
+	if (max_scatter) {
+		err = mmc_test_map_sg_max_scatter(t->mem, sz, t->sg,
+						  t->max_segs, t->max_seg_sz,
+				       &t->sg_len);
+	} else {
+		err = mmc_test_map_sg(t->mem, sz, t->sg, 1, t->max_segs,
+				      t->max_seg_sz, &t->sg_len, min_sg_len);
+	}
+	if (err)
+		pr_info("%s: Failed to map sg list\n",
+		       mmc_hostname(test->card->host));
+	return err;
+}
+
+/*
+ * Transfer bytes mapped by mmc_test_area_map().
+ */
+static int mmc_test_area_transfer(struct mmc_test_card *test,
+				  unsigned int dev_addr, int write)
+{
+	struct mmc_test_area *t = &test->area;
+
+	return mmc_test_simple_transfer(test, t->sg, t->sg_len, dev_addr,
+					t->blocks, 512, write);
+}
+
+/*
+ * Map and transfer bytes for multiple transfers.
+ */
+static int mmc_test_area_io_seq(struct mmc_test_card *test, unsigned long sz,
+				unsigned int dev_addr, int write,
+				int max_scatter, int timed, int count,
+				bool nonblock, int min_sg_len)
+{
+	struct timespec ts1, ts2;
+	int ret = 0;
+	int i;
+	struct mmc_test_area *t = &test->area;
+
+	/*
+	 * In the case of a maximally scattered transfer, the maximum transfer
+	 * size is further limited by using PAGE_SIZE segments.
+	 */
+	if (max_scatter) {
+		struct mmc_test_area *t = &test->area;
+		unsigned long max_tfr;
+
+		if (t->max_seg_sz >= PAGE_SIZE)
+			max_tfr = t->max_segs * PAGE_SIZE;
+		else
+			max_tfr = t->max_segs * t->max_seg_sz;
+		if (sz > max_tfr)
+			sz = max_tfr;
+	}
+
+	ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len);
+	if (ret)
+		return ret;
+
+	if (timed)
+		getnstimeofday(&ts1);
+	if (nonblock)
+		ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_len,
+				 dev_addr, t->blocks, 512, write, count);
+	else
+		for (i = 0; i < count && ret == 0; i++) {
+			ret = mmc_test_area_transfer(test, dev_addr, write);
+			dev_addr += sz >> 9;
+		}
+
+	if (ret)
+		return ret;
+
+	if (timed)
+		getnstimeofday(&ts2);
+
+	if (timed)
+		mmc_test_print_avg_rate(test, sz, count, &ts1, &ts2);
+
+	return 0;
+}
+
+static int mmc_test_area_io(struct mmc_test_card *test, unsigned long sz,
+			    unsigned int dev_addr, int write, int max_scatter,
+			    int timed)
+{
+	return mmc_test_area_io_seq(test, sz, dev_addr, write, max_scatter,
+				    timed, 1, false, 0);
+}
+
+/*
+ * Write the test area entirely.
+ */
+static int mmc_test_area_fill(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+
+	return mmc_test_area_io(test, t->max_tfr, t->dev_addr, 1, 0, 0);
+}
+
+/*
+ * Erase the test area entirely.
+ */
+static int mmc_test_area_erase(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+
+	if (!mmc_can_erase(test->card))
+		return 0;
+
+	return mmc_erase(test->card, t->dev_addr, t->max_sz >> 9,
+			 MMC_ERASE_ARG);
+}
+
+/*
+ * Cleanup struct mmc_test_area.
+ */
+static int mmc_test_area_cleanup(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+
+	kfree(t->sg);
+	mmc_test_free_mem(t->mem);
+
+	return 0;
+}
+
+/*
+ * Initialize an area for testing large transfers.  The test area is set to the
+ * middle of the card because cards may have different charateristics at the
+ * front (for FAT file system optimization).  Optionally, the area is erased
+ * (if the card supports it) which may improve write performance.  Optionally,
+ * the area is filled with data for subsequent read tests.
+ */
+static int mmc_test_area_init(struct mmc_test_card *test, int erase, int fill)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long min_sz = 64 * 1024, sz;
+	int ret;
+
+	ret = mmc_test_set_blksize(test, 512);
+	if (ret)
+		return ret;
+
+	/* Make the test area size about 4MiB */
+	sz = (unsigned long)test->card->pref_erase << 9;
+	t->max_sz = sz;
+	while (t->max_sz < 4 * 1024 * 1024)
+		t->max_sz += sz;
+	while (t->max_sz > TEST_AREA_MAX_SIZE && t->max_sz > sz)
+		t->max_sz -= sz;
+
+	t->max_segs = test->card->host->max_segs;
+	t->max_seg_sz = test->card->host->max_seg_size;
+	t->max_seg_sz -= t->max_seg_sz % 512;
+
+	t->max_tfr = t->max_sz;
+	if (t->max_tfr >> 9 > test->card->host->max_blk_count)
+		t->max_tfr = test->card->host->max_blk_count << 9;
+	if (t->max_tfr > test->card->host->max_req_size)
+		t->max_tfr = test->card->host->max_req_size;
+	if (t->max_tfr / t->max_seg_sz > t->max_segs)
+		t->max_tfr = t->max_segs * t->max_seg_sz;
+
+	/*
+	 * Try to allocate enough memory for a max. sized transfer.  Less is OK
+	 * because the same memory can be mapped into the scatterlist more than
+	 * once.  Also, take into account the limits imposed on scatterlist
+	 * segments by the host driver.
+	 */
+	t->mem = mmc_test_alloc_mem(min_sz, t->max_tfr, t->max_segs,
+				    t->max_seg_sz);
+	if (!t->mem)
+		return -ENOMEM;
+
+	t->sg = kmalloc(sizeof(struct scatterlist) * t->max_segs, GFP_KERNEL);
+	if (!t->sg) {
+		ret = -ENOMEM;
+		goto out_free;
+	}
+
+	t->dev_addr = mmc_test_capacity(test->card) / 2;
+	t->dev_addr -= t->dev_addr % (t->max_sz >> 9);
+
+	if (erase) {
+		ret = mmc_test_area_erase(test);
+		if (ret)
+			goto out_free;
+	}
+
+	if (fill) {
+		ret = mmc_test_area_fill(test);
+		if (ret)
+			goto out_free;
+	}
+
+	return 0;
+
+out_free:
+	mmc_test_area_cleanup(test);
+	return ret;
+}
+
+/*
+ * Prepare for large transfers.  Do not erase the test area.
+ */
+static int mmc_test_area_prepare(struct mmc_test_card *test)
+{
+	return mmc_test_area_init(test, 0, 0);
+}
+
+/*
+ * Prepare for large transfers.  Do erase the test area.
+ */
+static int mmc_test_area_prepare_erase(struct mmc_test_card *test)
+{
+	return mmc_test_area_init(test, 1, 0);
+}
+
+/*
+ * Prepare for large transfers.  Erase and fill the test area.
+ */
+static int mmc_test_area_prepare_fill(struct mmc_test_card *test)
+{
+	return mmc_test_area_init(test, 1, 1);
+}
+
+/*
+ * Test best-case performance.  Best-case performance is expected from
+ * a single large transfer.
+ *
+ * An additional option (max_scatter) allows the measurement of the same
+ * transfer but with no contiguous pages in the scatter list.  This tests
+ * the efficiency of DMA to handle scattered pages.
+ */
+static int mmc_test_best_performance(struct mmc_test_card *test, int write,
+				     int max_scatter)
+{
+	struct mmc_test_area *t = &test->area;
+
+	return mmc_test_area_io(test, t->max_tfr, t->dev_addr, write,
+				max_scatter, 1);
+}
+
+/*
+ * Best-case read performance.
+ */
+static int mmc_test_best_read_performance(struct mmc_test_card *test)
+{
+	return mmc_test_best_performance(test, 0, 0);
+}
+
+/*
+ * Best-case write performance.
+ */
+static int mmc_test_best_write_performance(struct mmc_test_card *test)
+{
+	return mmc_test_best_performance(test, 1, 0);
+}
+
+/*
+ * Best-case read performance into scattered pages.
+ */
+static int mmc_test_best_read_perf_max_scatter(struct mmc_test_card *test)
+{
+	return mmc_test_best_performance(test, 0, 1);
+}
+
+/*
+ * Best-case write performance from scattered pages.
+ */
+static int mmc_test_best_write_perf_max_scatter(struct mmc_test_card *test)
+{
+	return mmc_test_best_performance(test, 1, 1);
+}
+
+/*
+ * Single read performance by transfer size.
+ */
+static int mmc_test_profile_read_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	unsigned int dev_addr;
+	int ret;
+
+	for (sz = 512; sz < t->max_tfr; sz <<= 1) {
+		dev_addr = t->dev_addr + (sz >> 9);
+		ret = mmc_test_area_io(test, sz, dev_addr, 0, 0, 1);
+		if (ret)
+			return ret;
+	}
+	sz = t->max_tfr;
+	dev_addr = t->dev_addr;
+	return mmc_test_area_io(test, sz, dev_addr, 0, 0, 1);
+}
+
+/*
+ * Single write performance by transfer size.
+ */
+static int mmc_test_profile_write_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	unsigned int dev_addr;
+	int ret;
+
+	ret = mmc_test_area_erase(test);
+	if (ret)
+		return ret;
+	for (sz = 512; sz < t->max_tfr; sz <<= 1) {
+		dev_addr = t->dev_addr + (sz >> 9);
+		ret = mmc_test_area_io(test, sz, dev_addr, 1, 0, 1);
+		if (ret)
+			return ret;
+	}
+	ret = mmc_test_area_erase(test);
+	if (ret)
+		return ret;
+	sz = t->max_tfr;
+	dev_addr = t->dev_addr;
+	return mmc_test_area_io(test, sz, dev_addr, 1, 0, 1);
+}
+
+/*
+ * Single trim performance by transfer size.
+ */
+static int mmc_test_profile_trim_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	unsigned int dev_addr;
+	struct timespec ts1, ts2;
+	int ret;
+
+	if (!mmc_can_trim(test->card))
+		return RESULT_UNSUP_CARD;
+
+	if (!mmc_can_erase(test->card))
+		return RESULT_UNSUP_HOST;
+
+	for (sz = 512; sz < t->max_sz; sz <<= 1) {
+		dev_addr = t->dev_addr + (sz >> 9);
+		getnstimeofday(&ts1);
+		ret = mmc_erase(test->card, dev_addr, sz >> 9, MMC_TRIM_ARG);
+		if (ret)
+			return ret;
+		getnstimeofday(&ts2);
+		mmc_test_print_rate(test, sz, &ts1, &ts2);
+	}
+	dev_addr = t->dev_addr;
+	getnstimeofday(&ts1);
+	ret = mmc_erase(test->card, dev_addr, sz >> 9, MMC_TRIM_ARG);
+	if (ret)
+		return ret;
+	getnstimeofday(&ts2);
+	mmc_test_print_rate(test, sz, &ts1, &ts2);
+	return 0;
+}
+
+static int mmc_test_seq_read_perf(struct mmc_test_card *test, unsigned long sz)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned int dev_addr, i, cnt;
+	struct timespec ts1, ts2;
+	int ret;
+
+	cnt = t->max_sz / sz;
+	dev_addr = t->dev_addr;
+	getnstimeofday(&ts1);
+	for (i = 0; i < cnt; i++) {
+		ret = mmc_test_area_io(test, sz, dev_addr, 0, 0, 0);
+		if (ret)
+			return ret;
+		dev_addr += (sz >> 9);
+	}
+	getnstimeofday(&ts2);
+	mmc_test_print_avg_rate(test, sz, cnt, &ts1, &ts2);
+	return 0;
+}
+
+/*
+ * Consecutive read performance by transfer size.
+ */
+static int mmc_test_profile_seq_read_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	int ret;
+
+	for (sz = 512; sz < t->max_tfr; sz <<= 1) {
+		ret = mmc_test_seq_read_perf(test, sz);
+		if (ret)
+			return ret;
+	}
+	sz = t->max_tfr;
+	return mmc_test_seq_read_perf(test, sz);
+}
+
+static int mmc_test_seq_write_perf(struct mmc_test_card *test, unsigned long sz)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned int dev_addr, i, cnt;
+	struct timespec ts1, ts2;
+	int ret;
+
+	ret = mmc_test_area_erase(test);
+	if (ret)
+		return ret;
+	cnt = t->max_sz / sz;
+	dev_addr = t->dev_addr;
+	getnstimeofday(&ts1);
+	for (i = 0; i < cnt; i++) {
+		ret = mmc_test_area_io(test, sz, dev_addr, 1, 0, 0);
+		if (ret)
+			return ret;
+		dev_addr += (sz >> 9);
+	}
+	getnstimeofday(&ts2);
+	mmc_test_print_avg_rate(test, sz, cnt, &ts1, &ts2);
+	return 0;
+}
+
+/*
+ * Consecutive write performance by transfer size.
+ */
+static int mmc_test_profile_seq_write_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	int ret;
+
+	for (sz = 512; sz < t->max_tfr; sz <<= 1) {
+		ret = mmc_test_seq_write_perf(test, sz);
+		if (ret)
+			return ret;
+	}
+	sz = t->max_tfr;
+	return mmc_test_seq_write_perf(test, sz);
+}
+
+/*
+ * Consecutive trim performance by transfer size.
+ */
+static int mmc_test_profile_seq_trim_perf(struct mmc_test_card *test)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned long sz;
+	unsigned int dev_addr, i, cnt;
+	struct timespec ts1, ts2;
+	int ret;
+
+	if (!mmc_can_trim(test->card))
+		return RESULT_UNSUP_CARD;
+
+	if (!mmc_can_erase(test->card))
+		return RESULT_UNSUP_HOST;
+
+	for (sz = 512; sz <= t->max_sz; sz <<= 1) {
+		ret = mmc_test_area_erase(test);
+		if (ret)
+			return ret;
+		ret = mmc_test_area_fill(test);
+		if (ret)
+			return ret;
+		cnt = t->max_sz / sz;
+		dev_addr = t->dev_addr;
+		getnstimeofday(&ts1);
+		for (i = 0; i < cnt; i++) {
+			ret = mmc_erase(test->card, dev_addr, sz >> 9,
+					MMC_TRIM_ARG);
+			if (ret)
+				return ret;
+			dev_addr += (sz >> 9);
+		}
+		getnstimeofday(&ts2);
+		mmc_test_print_avg_rate(test, sz, cnt, &ts1, &ts2);
+	}
+	return 0;
+}
+
+static unsigned int rnd_next = 1;
+
+static unsigned int mmc_test_rnd_num(unsigned int rnd_cnt)
+{
+	uint64_t r;
+
+	rnd_next = rnd_next * 1103515245 + 12345;
+	r = (rnd_next >> 16) & 0x7fff;
+	return (r * rnd_cnt) >> 15;
+}
+
+static int mmc_test_rnd_perf(struct mmc_test_card *test, int write, int print,
+			     unsigned long sz)
+{
+	unsigned int dev_addr, cnt, rnd_addr, range1, range2, last_ea = 0, ea;
+	unsigned int ssz;
+	struct timespec ts1, ts2, ts;
+	int ret;
+
+	ssz = sz >> 9;
+
+	rnd_addr = mmc_test_capacity(test->card) / 4;
+	range1 = rnd_addr / test->card->pref_erase;
+	range2 = range1 / ssz;
+
+	getnstimeofday(&ts1);
+	for (cnt = 0; cnt < UINT_MAX; cnt++) {
+		getnstimeofday(&ts2);
+		ts = timespec_sub(ts2, ts1);
+		if (ts.tv_sec >= 10)
+			break;
+		ea = mmc_test_rnd_num(range1);
+		if (ea == last_ea)
+			ea -= 1;
+		last_ea = ea;
+		dev_addr = rnd_addr + test->card->pref_erase * ea +
+			   ssz * mmc_test_rnd_num(range2);
+		ret = mmc_test_area_io(test, sz, dev_addr, write, 0, 0);
+		if (ret)
+			return ret;
+	}
+	if (print)
+		mmc_test_print_avg_rate(test, sz, cnt, &ts1, &ts2);
+	return 0;
+}
+
+static int mmc_test_random_perf(struct mmc_test_card *test, int write)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned int next;
+	unsigned long sz;
+	int ret;
+
+	for (sz = 512; sz < t->max_tfr; sz <<= 1) {
+		/*
+		 * When writing, try to get more consistent results by running
+		 * the test twice with exactly the same I/O but outputting the
+		 * results only for the 2nd run.
+		 */
+		if (write) {
+			next = rnd_next;
+			ret = mmc_test_rnd_perf(test, write, 0, sz);
+			if (ret)
+				return ret;
+			rnd_next = next;
+		}
+		ret = mmc_test_rnd_perf(test, write, 1, sz);
+		if (ret)
+			return ret;
+	}
+	sz = t->max_tfr;
+	if (write) {
+		next = rnd_next;
+		ret = mmc_test_rnd_perf(test, write, 0, sz);
+		if (ret)
+			return ret;
+		rnd_next = next;
+	}
+	return mmc_test_rnd_perf(test, write, 1, sz);
+}
+
+/*
+ * Random read performance by transfer size.
+ */
+static int mmc_test_random_read_perf(struct mmc_test_card *test)
+{
+	return mmc_test_random_perf(test, 0);
+}
+
+/*
+ * Random write performance by transfer size.
+ */
+static int mmc_test_random_write_perf(struct mmc_test_card *test)
+{
+	return mmc_test_random_perf(test, 1);
+}
+
+static int mmc_test_seq_perf(struct mmc_test_card *test, int write,
+			     unsigned int tot_sz, int max_scatter)
+{
+	struct mmc_test_area *t = &test->area;
+	unsigned int dev_addr, i, cnt, sz, ssz;
+	struct timespec ts1, ts2;
+	int ret;
+
+	sz = t->max_tfr;
+
+	/*
+	 * In the case of a maximally scattered transfer, the maximum transfer
+	 * size is further limited by using PAGE_SIZE segments.
+	 */
+	if (max_scatter) {
+		unsigned long max_tfr;
+
+		if (t->max_seg_sz >= PAGE_SIZE)
+			max_tfr = t->max_segs * PAGE_SIZE;
+		else
+			max_tfr = t->max_segs * t->max_seg_sz;
+		if (sz > max_tfr)
+			sz = max_tfr;
+	}
+
+	ssz = sz >> 9;
+	dev_addr = mmc_test_capacity(test->card) / 4;
+	if (tot_sz > dev_addr << 9)
+		tot_sz = dev_addr << 9;
+	cnt = tot_sz / sz;
+	dev_addr &= 0xffff0000; /* Round to 64MiB boundary */
+
+	getnstimeofday(&ts1);
+	for (i = 0; i < cnt; i++) {
+		ret = mmc_test_area_io(test, sz, dev_addr, write,
+				       max_scatter, 0);
+		if (ret)
+			return ret;
+		dev_addr += ssz;
+	}
+	getnstimeofday(&ts2);
+
+	mmc_test_print_avg_rate(test, sz, cnt, &ts1, &ts2);
+
+	return 0;
+}
+
+static int mmc_test_large_seq_perf(struct mmc_test_card *test, int write)
+{
+	int ret, i;
+
+	for (i = 0; i < 10; i++) {
+		ret = mmc_test_seq_perf(test, write, 10 * 1024 * 1024, 1);
+		if (ret)
+			return ret;
+	}
+	for (i = 0; i < 5; i++) {
+		ret = mmc_test_seq_perf(test, write, 100 * 1024 * 1024, 1);
+		if (ret)
+			return ret;
+	}
+	for (i = 0; i < 3; i++) {
+		ret = mmc_test_seq_perf(test, write, 1000 * 1024 * 1024, 1);
+		if (ret)
+			return ret;
+	}
+
+	return ret;
+}
+
+/*
+ * Large sequential read performance.
+ */
+static int mmc_test_large_seq_read_perf(struct mmc_test_card *test)
+{
+	return mmc_test_large_seq_perf(test, 0);
+}
+
+/*
+ * Large sequential write performance.
+ */
+static int mmc_test_large_seq_write_perf(struct mmc_test_card *test)
+{
+	return mmc_test_large_seq_perf(test, 1);
+}
+
+static int mmc_test_rw_multiple(struct mmc_test_card *test,
+				struct mmc_test_multiple_rw *tdata,
+				unsigned int reqsize, unsigned int size,
+				int min_sg_len)
+{
+	unsigned int dev_addr;
+	struct mmc_test_area *t = &test->area;
+	int ret = 0;
+
+	/* Set up test area */
+	if (size > mmc_test_capacity(test->card) / 2 * 512)
+		size = mmc_test_capacity(test->card) / 2 * 512;
+	if (reqsize > t->max_tfr)
+		reqsize = t->max_tfr;
+	dev_addr = mmc_test_capacity(test->card) / 4;
+	if ((dev_addr & 0xffff0000))
+		dev_addr &= 0xffff0000; /* Round to 64MiB boundary */
+	else
+		dev_addr &= 0xfffff800; /* Round to 1MiB boundary */
+	if (!dev_addr)
+		goto err;
+
+	if (reqsize > size)
+		return 0;
+
+	/* prepare test area */
+	if (mmc_can_erase(test->card) &&
+	    tdata->prepare & MMC_TEST_PREP_ERASE) {
+		ret = mmc_erase(test->card, dev_addr,
+				size / 512, MMC_SECURE_ERASE_ARG);
+		if (ret)
+			ret = mmc_erase(test->card, dev_addr,
+					size / 512, MMC_ERASE_ARG);
+		if (ret)
+			goto err;
+	}
+
+	/* Run test */
+	ret = mmc_test_area_io_seq(test, reqsize, dev_addr,
+				   tdata->do_write, 0, 1, size / reqsize,
+				   tdata->do_nonblock_req, min_sg_len);
+	if (ret)
+		goto err;
+
+	return ret;
+ err:
+	pr_info("[%s] error\n", __func__);
+	return ret;
+}
+
+static int mmc_test_rw_multiple_size(struct mmc_test_card *test,
+				     struct mmc_test_multiple_rw *rw)
+{
+	int ret = 0;
+	int i;
+	void *pre_req = test->card->host->ops->pre_req;
+	void *post_req = test->card->host->ops->post_req;
+
+	if (rw->do_nonblock_req &&
+	    ((!pre_req && post_req) || (pre_req && !post_req))) {
+		pr_info("error: only one of pre/post is defined\n");
+		return -EINVAL;
+	}
+
+	for (i = 0 ; i < rw->len && ret == 0; i++) {
+		ret = mmc_test_rw_multiple(test, rw, rw->bs[i], rw->size, 0);
+		if (ret)
+			break;
+	}
+	return ret;
+}
+
+static int mmc_test_rw_multiple_sg_len(struct mmc_test_card *test,
+				       struct mmc_test_multiple_rw *rw)
+{
+	int ret = 0;
+	int i;
+
+	for (i = 0 ; i < rw->len && ret == 0; i++) {
+		ret = mmc_test_rw_multiple(test, rw, 512*1024, rw->size,
+					   rw->sg_len[i]);
+		if (ret)
+			break;
+	}
+	return ret;
+}
+
+/*
+ * Multiple blocking write 4k to 4 MB chunks
+ */
+static int mmc_test_profile_mult_write_blocking_perf(struct mmc_test_card *test)
+{
+	unsigned int bs[] = {1 << 12, 1 << 13, 1 << 14, 1 << 15, 1 << 16,
+			     1 << 17, 1 << 18, 1 << 19, 1 << 20, 1 << 22};
+	struct mmc_test_multiple_rw test_data = {
+		.bs = bs,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(bs),
+		.do_write = true,
+		.do_nonblock_req = false,
+		.prepare = MMC_TEST_PREP_ERASE,
+	};
+
+	return mmc_test_rw_multiple_size(test, &test_data);
+};
+
+/*
+ * Multiple non-blocking write 4k to 4 MB chunks
+ */
+static int mmc_test_profile_mult_write_nonblock_perf(struct mmc_test_card *test)
+{
+	unsigned int bs[] = {1 << 12, 1 << 13, 1 << 14, 1 << 15, 1 << 16,
+			     1 << 17, 1 << 18, 1 << 19, 1 << 20, 1 << 22};
+	struct mmc_test_multiple_rw test_data = {
+		.bs = bs,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(bs),
+		.do_write = true,
+		.do_nonblock_req = true,
+		.prepare = MMC_TEST_PREP_ERASE,
+	};
+
+	return mmc_test_rw_multiple_size(test, &test_data);
+}
+
+/*
+ * Multiple blocking read 4k to 4 MB chunks
+ */
+static int mmc_test_profile_mult_read_blocking_perf(struct mmc_test_card *test)
+{
+	unsigned int bs[] = {1 << 12, 1 << 13, 1 << 14, 1 << 15, 1 << 16,
+			     1 << 17, 1 << 18, 1 << 19, 1 << 20, 1 << 22};
+	struct mmc_test_multiple_rw test_data = {
+		.bs = bs,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(bs),
+		.do_write = false,
+		.do_nonblock_req = false,
+		.prepare = MMC_TEST_PREP_NONE,
+	};
+
+	return mmc_test_rw_multiple_size(test, &test_data);
+}
+
+/*
+ * Multiple non-blocking read 4k to 4 MB chunks
+ */
+static int mmc_test_profile_mult_read_nonblock_perf(struct mmc_test_card *test)
+{
+	unsigned int bs[] = {1 << 12, 1 << 13, 1 << 14, 1 << 15, 1 << 16,
+			     1 << 17, 1 << 18, 1 << 19, 1 << 20, 1 << 22};
+	struct mmc_test_multiple_rw test_data = {
+		.bs = bs,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(bs),
+		.do_write = false,
+		.do_nonblock_req = true,
+		.prepare = MMC_TEST_PREP_NONE,
+	};
+
+	return mmc_test_rw_multiple_size(test, &test_data);
+}
+
+/*
+ * Multiple blocking write 1 to 512 sg elements
+ */
+static int mmc_test_profile_sglen_wr_blocking_perf(struct mmc_test_card *test)
+{
+	unsigned int sg_len[] = {1, 1 << 3, 1 << 4, 1 << 5, 1 << 6,
+				 1 << 7, 1 << 8, 1 << 9};
+	struct mmc_test_multiple_rw test_data = {
+		.sg_len = sg_len,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(sg_len),
+		.do_write = true,
+		.do_nonblock_req = false,
+		.prepare = MMC_TEST_PREP_ERASE,
+	};
+
+	return mmc_test_rw_multiple_sg_len(test, &test_data);
+};
+
+/*
+ * Multiple non-blocking write 1 to 512 sg elements
+ */
+static int mmc_test_profile_sglen_wr_nonblock_perf(struct mmc_test_card *test)
+{
+	unsigned int sg_len[] = {1, 1 << 3, 1 << 4, 1 << 5, 1 << 6,
+				 1 << 7, 1 << 8, 1 << 9};
+	struct mmc_test_multiple_rw test_data = {
+		.sg_len = sg_len,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(sg_len),
+		.do_write = true,
+		.do_nonblock_req = true,
+		.prepare = MMC_TEST_PREP_ERASE,
+	};
+
+	return mmc_test_rw_multiple_sg_len(test, &test_data);
+}
+
+/*
+ * Multiple blocking read 1 to 512 sg elements
+ */
+static int mmc_test_profile_sglen_r_blocking_perf(struct mmc_test_card *test)
+{
+	unsigned int sg_len[] = {1, 1 << 3, 1 << 4, 1 << 5, 1 << 6,
+				 1 << 7, 1 << 8, 1 << 9};
+	struct mmc_test_multiple_rw test_data = {
+		.sg_len = sg_len,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(sg_len),
+		.do_write = false,
+		.do_nonblock_req = false,
+		.prepare = MMC_TEST_PREP_NONE,
+	};
+
+	return mmc_test_rw_multiple_sg_len(test, &test_data);
+}
+
+/*
+ * Multiple non-blocking read 1 to 512 sg elements
+ */
+static int mmc_test_profile_sglen_r_nonblock_perf(struct mmc_test_card *test)
+{
+	unsigned int sg_len[] = {1, 1 << 3, 1 << 4, 1 << 5, 1 << 6,
+				 1 << 7, 1 << 8, 1 << 9};
+	struct mmc_test_multiple_rw test_data = {
+		.sg_len = sg_len,
+		.size = TEST_AREA_MAX_SIZE,
+		.len = ARRAY_SIZE(sg_len),
+		.do_write = false,
+		.do_nonblock_req = true,
+		.prepare = MMC_TEST_PREP_NONE,
+	};
+
+	return mmc_test_rw_multiple_sg_len(test, &test_data);
+}
+
+/*
+ * eMMC hardware reset.
+ */
+static int mmc_test_hw_reset(struct mmc_test_card *test)
+{
+	struct mmc_card *card = test->card;
+	struct mmc_host *host = card->host;
+	int err;
+
+	err = mmc_hw_reset_check(host);
+	if (!err)
+		return RESULT_OK;
+
+	if (err == -ENOSYS)
+		return RESULT_FAIL;
+
+	if (err != -EOPNOTSUPP)
+		return err;
+
+	if (!mmc_can_reset(card))
+		return RESULT_UNSUP_CARD;
+
+	return RESULT_UNSUP_HOST;
+}
+
 static const struct mmc_test_case mmc_test_cases[] = {
 	{
 		.name = "Basic write (no data verification)",
@@ -1036,33 +2500,217 @@ static const struct mmc_test_case mmc_test_cases[] = {
 		.cleanup = mmc_test_cleanup,
 	},
 
+#else
+
+	{
+		.name = "Highmem write",
+		.run = mmc_test_no_highmem,
+	},
+
+	{
+		.name = "Highmem read",
+		.run = mmc_test_no_highmem,
+	},
+
+	{
+		.name = "Multi-block highmem write",
+		.run = mmc_test_no_highmem,
+	},
+
+	{
+		.name = "Multi-block highmem read",
+		.run = mmc_test_no_highmem,
+	},
+
 #endif /* CONFIG_HIGHMEM */
 
+	{
+		.name = "Best-case read performance",
+		.prepare = mmc_test_area_prepare_fill,
+		.run = mmc_test_best_read_performance,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Best-case write performance",
+		.prepare = mmc_test_area_prepare_erase,
+		.run = mmc_test_best_write_performance,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Best-case read performance into scattered pages",
+		.prepare = mmc_test_area_prepare_fill,
+		.run = mmc_test_best_read_perf_max_scatter,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Best-case write performance from scattered pages",
+		.prepare = mmc_test_area_prepare_erase,
+		.run = mmc_test_best_write_perf_max_scatter,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Single read performance by transfer size",
+		.prepare = mmc_test_area_prepare_fill,
+		.run = mmc_test_profile_read_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Single write performance by transfer size",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_write_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Single trim performance by transfer size",
+		.prepare = mmc_test_area_prepare_fill,
+		.run = mmc_test_profile_trim_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Consecutive read performance by transfer size",
+		.prepare = mmc_test_area_prepare_fill,
+		.run = mmc_test_profile_seq_read_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Consecutive write performance by transfer size",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_seq_write_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Consecutive trim performance by transfer size",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_seq_trim_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Random read performance by transfer size",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_random_read_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Random write performance by transfer size",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_random_write_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Large sequential read into scattered pages",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_large_seq_read_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Large sequential write from scattered pages",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_large_seq_write_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Write performance with blocking req 4k to 4MB",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_mult_write_blocking_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Write performance with non-blocking req 4k to 4MB",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_mult_write_nonblock_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Read performance with blocking req 4k to 4MB",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_mult_read_blocking_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Read performance with non-blocking req 4k to 4MB",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_mult_read_nonblock_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Write performance blocking req 1 to 512 sg elems",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_sglen_wr_blocking_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Write performance non-blocking req 1 to 512 sg elems",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_sglen_wr_nonblock_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Read performance blocking req 1 to 512 sg elems",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_sglen_r_blocking_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "Read performance non-blocking req 1 to 512 sg elems",
+		.prepare = mmc_test_area_prepare,
+		.run = mmc_test_profile_sglen_r_nonblock_perf,
+		.cleanup = mmc_test_area_cleanup,
+	},
+
+	{
+		.name = "eMMC hardware reset",
+		.run = mmc_test_hw_reset,
+	},
 };
 
 static DEFINE_MUTEX(mmc_test_lock);
 
+static LIST_HEAD(mmc_test_result);
+
 static void mmc_test_run(struct mmc_test_card *test, int testcase)
 {
 	int i, ret;
 
-	printk(KERN_INFO "%s: Starting tests of card %s...\n",
+	pr_info("%s: Starting tests of card %s...\n",
 		mmc_hostname(test->card->host), mmc_card_id(test->card));
 
 	mmc_claim_host(test->card->host);
 
 	for (i = 0;i < ARRAY_SIZE(mmc_test_cases);i++) {
+		struct mmc_test_general_result *gr;
+
 		if (testcase && ((i + 1) != testcase))
 			continue;
 
-		printk(KERN_INFO "%s: Test case %d. %s...\n",
+		pr_info("%s: Test case %d. %s...\n",
 			mmc_hostname(test->card->host), i + 1,
 			mmc_test_cases[i].name);
 
 		if (mmc_test_cases[i].prepare) {
 			ret = mmc_test_cases[i].prepare(test);
 			if (ret) {
-				printk(KERN_INFO "%s: Result: Prepare "
+				pr_info("%s: Result: Prepare "
 					"stage failed! (%d)\n",
 					mmc_hostname(test->card->host),
 					ret);
@@ -1070,35 +2718,58 @@ static void mmc_test_run(struct mmc_test_card *test, int testcase)
 			}
 		}
 
+		gr = kzalloc(sizeof(struct mmc_test_general_result),
+			GFP_KERNEL);
+		if (gr) {
+			INIT_LIST_HEAD(&gr->tr_lst);
+
+			/* Assign data what we know already */
+			gr->card = test->card;
+			gr->testcase = i;
+
+			/* Append container to global one */
+			list_add_tail(&gr->link, &mmc_test_result);
+
+			/*
+			 * Save the pointer to created container in our private
+			 * structure.
+			 */
+			test->gr = gr;
+		}
+
 		ret = mmc_test_cases[i].run(test);
 		switch (ret) {
 		case RESULT_OK:
-			printk(KERN_INFO "%s: Result: OK\n",
+			pr_info("%s: Result: OK\n",
 				mmc_hostname(test->card->host));
 			break;
 		case RESULT_FAIL:
-			printk(KERN_INFO "%s: Result: FAILED\n",
+			pr_info("%s: Result: FAILED\n",
 				mmc_hostname(test->card->host));
 			break;
 		case RESULT_UNSUP_HOST:
-			printk(KERN_INFO "%s: Result: UNSUPPORTED "
+			pr_info("%s: Result: UNSUPPORTED "
 				"(by host)\n",
 				mmc_hostname(test->card->host));
 			break;
 		case RESULT_UNSUP_CARD:
-			printk(KERN_INFO "%s: Result: UNSUPPORTED "
+			pr_info("%s: Result: UNSUPPORTED "
 				"(by card)\n",
 				mmc_hostname(test->card->host));
 			break;
 		default:
-			printk(KERN_INFO "%s: Result: ERROR (%d)\n",
+			pr_info("%s: Result: ERROR (%d)\n",
 				mmc_hostname(test->card->host), ret);
 		}
 
+		/* Save the result */
+		if (gr)
+			gr->result = ret;
+
 		if (mmc_test_cases[i].cleanup) {
 			ret = mmc_test_cases[i].cleanup(test);
 			if (ret) {
-				printk(KERN_INFO "%s: Warning: Cleanup "
+				pr_info("%s: Warning: Cleanup "
 					"stage failed! (%d)\n",
 					mmc_hostname(test->card->host),
 					ret);
@@ -1108,34 +2779,93 @@ static void mmc_test_run(struct mmc_test_card *test, int testcase)
 
 	mmc_release_host(test->card->host);
 
-	printk(KERN_INFO "%s: Tests completed.\n",
+	pr_info("%s: Tests completed.\n",
 		mmc_hostname(test->card->host));
 }
 
-static ssize_t mmc_test_show(struct device *dev,
-	struct device_attribute *attr, char *buf)
+static void mmc_test_free_result(struct mmc_card *card)
 {
+	struct mmc_test_general_result *gr, *grs;
+
 	mutex_lock(&mmc_test_lock);
+
+	list_for_each_entry_safe(gr, grs, &mmc_test_result, link) {
+		struct mmc_test_transfer_result *tr, *trs;
+
+		if (card && gr->card != card)
+			continue;
+
+		list_for_each_entry_safe(tr, trs, &gr->tr_lst, link) {
+			list_del(&tr->link);
+			kfree(tr);
+		}
+
+		list_del(&gr->link);
+		kfree(gr);
+	}
+
+	mutex_unlock(&mmc_test_lock);
+}
+
+static LIST_HEAD(mmc_test_file_test);
+
+static int mtf_test_show(struct seq_file *sf, void *data)
+{
+	struct mmc_card *card = (struct mmc_card *)sf->private;
+	struct mmc_test_general_result *gr;
+
+	mutex_lock(&mmc_test_lock);
+
+	list_for_each_entry(gr, &mmc_test_result, link) {
+		struct mmc_test_transfer_result *tr;
+
+		if (gr->card != card)
+			continue;
+
+		seq_printf(sf, "Test %d: %d\n", gr->testcase + 1, gr->result);
+
+		list_for_each_entry(tr, &gr->tr_lst, link) {
+			seq_printf(sf, "%u %d %lu.%09lu %u %u.%02u\n",
+				tr->count, tr->sectors,
+				(unsigned long)tr->ts.tv_sec,
+				(unsigned long)tr->ts.tv_nsec,
+				tr->rate, tr->iops / 100, tr->iops % 100);
+		}
+	}
+
 	mutex_unlock(&mmc_test_lock);
 
 	return 0;
 }
 
-static ssize_t mmc_test_store(struct device *dev,
-	struct device_attribute *attr, const char *buf, size_t count)
+static int mtf_test_open(struct inode *inode, struct file *file)
 {
-	struct mmc_card *card;
-	struct mmc_test_card *test;
-	int testcase;
+	return single_open(file, mtf_test_show, inode->i_private);
+}
 
-	card = container_of(dev, struct mmc_card, dev);
+static ssize_t mtf_test_write(struct file *file, const char __user *buf,
+	size_t count, loff_t *pos)
+{
+	struct seq_file *sf = (struct seq_file *)file->private_data;
+	struct mmc_card *card = (struct mmc_card *)sf->private;
+	struct mmc_test_card *test;
+	long testcase;
+	int ret;
 
-	testcase = simple_strtol(buf, NULL, 10);
+	ret = kstrtol_from_user(buf, count, 10, &testcase);
+	if (ret)
+		return ret;
 
 	test = kzalloc(sizeof(struct mmc_test_card), GFP_KERNEL);
 	if (!test)
 		return -ENOMEM;
 
+	/*
+	 * Remove all test cases associated with given card. Thus we have only
+	 * actual data of the last run.
+	 */
+	mmc_test_free_result(card);
+
 	test->card = card;
 
 	test->buffer = kzalloc(BUFFER_SIZE, GFP_KERNEL);
@@ -1162,16 +2892,119 @@ static ssize_t mmc_test_store(struct device *dev,
 	return count;
 }
 
-static DEVICE_ATTR(test, S_IWUSR | S_IRUGO, mmc_test_show, mmc_test_store);
+static const struct file_operations mmc_test_fops_test = {
+	.open		= mtf_test_open,
+	.read		= seq_read,
+	.write		= mtf_test_write,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static int mtf_testlist_show(struct seq_file *sf, void *data)
+{
+	int i;
+
+	mutex_lock(&mmc_test_lock);
+
+	for (i = 0; i < ARRAY_SIZE(mmc_test_cases); i++)
+		seq_printf(sf, "%d:\t%s\n", i+1, mmc_test_cases[i].name);
+
+	mutex_unlock(&mmc_test_lock);
+
+	return 0;
+}
+
+static int mtf_testlist_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, mtf_testlist_show, inode->i_private);
+}
+
+static const struct file_operations mmc_test_fops_testlist = {
+	.open		= mtf_testlist_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static void mmc_test_free_dbgfs_file(struct mmc_card *card)
+{
+	struct mmc_test_dbgfs_file *df, *dfs;
+
+	mutex_lock(&mmc_test_lock);
+
+	list_for_each_entry_safe(df, dfs, &mmc_test_file_test, link) {
+		if (card && df->card != card)
+			continue;
+		debugfs_remove(df->file);
+		list_del(&df->link);
+		kfree(df);
+	}
+
+	mutex_unlock(&mmc_test_lock);
+}
+
+static int __mmc_test_register_dbgfs_file(struct mmc_card *card,
+	const char *name, umode_t mode, const struct file_operations *fops)
+{
+	struct dentry *file = NULL;
+	struct mmc_test_dbgfs_file *df;
+
+	if (card->debugfs_root)
+		file = debugfs_create_file(name, mode, card->debugfs_root,
+			card, fops);
+
+	if (IS_ERR_OR_NULL(file)) {
+		dev_err(&card->dev,
+			"Can't create %s. Perhaps debugfs is disabled.\n",
+			name);
+		return -ENODEV;
+	}
+
+	df = kmalloc(sizeof(struct mmc_test_dbgfs_file), GFP_KERNEL);
+	if (!df) {
+		debugfs_remove(file);
+		dev_err(&card->dev,
+			"Can't allocate memory for internal usage.\n");
+		return -ENOMEM;
+	}
+
+	df->card = card;
+	df->file = file;
+
+	list_add(&df->link, &mmc_test_file_test);
+	return 0;
+}
+
+static int mmc_test_register_dbgfs_file(struct mmc_card *card)
+{
+	int ret;
+
+	mutex_lock(&mmc_test_lock);
+
+	ret = __mmc_test_register_dbgfs_file(card, "test", S_IWUSR | S_IRUGO,
+		&mmc_test_fops_test);
+	if (ret)
+		goto err;
+
+	ret = __mmc_test_register_dbgfs_file(card, "testlist", S_IRUGO,
+		&mmc_test_fops_testlist);
+	if (ret)
+		goto err;
+
+err:
+	mutex_unlock(&mmc_test_lock);
+
+	return ret;
+}
 
 static int mmc_test_probe(struct mmc_card *card)
 {
 	int ret;
 
-	if ((card->type != MMC_TYPE_MMC) && (card->type != MMC_TYPE_SD))
+	if (!mmc_card_mmc(card) && !mmc_card_sd(card))
 		return -ENODEV;
 
-	ret = device_create_file(&card->dev, &dev_attr_test);
+	ret = mmc_test_register_dbgfs_file(card);
 	if (ret)
 		return ret;
 
@@ -1182,7 +3015,12 @@ static int mmc_test_probe(struct mmc_card *card)
 
 static void mmc_test_remove(struct mmc_card *card)
 {
-	device_remove_file(&card->dev, &dev_attr_test);
+	mmc_test_free_result(card);
+	mmc_test_free_dbgfs_file(card);
+}
+
+static void mmc_test_shutdown(struct mmc_card *card)
+{
 }
 
 static struct mmc_driver mmc_driver = {
@@ -1191,6 +3029,7 @@ static struct mmc_driver mmc_driver = {
 	},
 	.probe		= mmc_test_probe,
 	.remove		= mmc_test_remove,
+	.shutdown	= mmc_test_shutdown,
 };
 
 static int __init mmc_test_init(void)
@@ -1200,6 +3039,10 @@ static int __init mmc_test_init(void)
 
 static void __exit mmc_test_exit(void)
 {
+	/* Clear stalled data if card is still plugged */
+	mmc_test_free_result(NULL);
+	mmc_test_free_dbgfs_file(NULL);
+
 	mmc_unregister_driver(&mmc_driver);
 }
 
diff --git a/drivers/mmc/card/queue.c b/drivers/mmc/card/queue.c
index 7a72e75d5c6..3e049c13429 100644
--- a/drivers/mmc/card/queue.c
+++ b/drivers/mmc/card/queue.c
@@ -9,11 +9,13 @@
  * published by the Free Software Foundation.
  *
  */
+#include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/blkdev.h>
 #include <linux/freezer.h>
 #include <linux/kthread.h>
 #include <linux/scatterlist.h>
+#include <linux/dma-mapping.h>
 
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
@@ -21,21 +23,24 @@
 
 #define MMC_QUEUE_BOUNCESZ	65536
 
-#define MMC_QUEUE_SUSPENDED	(1 << 0)
-
 /*
  * Prepare a MMC request. This just filters out odd stuff.
  */
 static int mmc_prep_request(struct request_queue *q, struct request *req)
 {
+	struct mmc_queue *mq = q->queuedata;
+
 	/*
-	 * We only like normal block requests.
+	 * We only like normal block requests and discards.
 	 */
-	if (!blk_fs_request(req)) {
+	if (req->cmd_type != REQ_TYPE_FS && !(req->cmd_flags & REQ_DISCARD)) {
 		blk_dump_rq_flags(req, "MMC bad request");
 		return BLKPREP_KILL;
 	}
 
+	if (mq && mmc_card_removed(mq->card))
+		return BLKPREP_KILL;
+
 	req->cmd_flags |= REQ_DONTPREP;
 
 	return BLKPREP_OK;
@@ -51,15 +56,40 @@ static int mmc_queue_thread(void *d)
 	down(&mq->thread_sem);
 	do {
 		struct request *req = NULL;
+		struct mmc_queue_req *tmp;
+		unsigned int cmd_flags = 0;
 
 		spin_lock_irq(q->queue_lock);
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (!blk_queue_plugged(q))
-			req = elv_next_request(q);
-		mq->req = req;
+		req = blk_fetch_request(q);
+		mq->mqrq_cur->req = req;
 		spin_unlock_irq(q->queue_lock);
 
-		if (!req) {
+		if (req || mq->mqrq_prev->req) {
+			set_current_state(TASK_RUNNING);
+			cmd_flags = req ? req->cmd_flags : 0;
+			mq->issue_fn(mq, req);
+			if (mq->flags & MMC_QUEUE_NEW_REQUEST) {
+				mq->flags &= ~MMC_QUEUE_NEW_REQUEST;
+				continue; /* fetch again */
+			}
+
+			/*
+			 * Current request becomes previous request
+			 * and vice versa.
+			 * In case of special requests, current request
+			 * has been finished. Do not assign it to previous
+			 * request.
+			 */
+			if (cmd_flags & MMC_REQ_SPECIAL_MASK)
+				mq->mqrq_cur->req = NULL;
+
+			mq->mqrq_prev->brq.mrq.data = NULL;
+			mq->mqrq_prev->req = NULL;
+			tmp = mq->mqrq_prev;
+			mq->mqrq_prev = mq->mqrq_cur;
+			mq->mqrq_cur = tmp;
+		} else {
 			if (kthread_should_stop()) {
 				set_current_state(TASK_RUNNING);
 				break;
@@ -67,11 +97,7 @@ static int mmc_queue_thread(void *d)
 			up(&mq->thread_sem);
 			schedule();
 			down(&mq->thread_sem);
-			continue;
 		}
-		set_current_state(TASK_RUNNING);
-
-		mq->issue_fn(mq, req);
 	} while (1);
 	up(&mq->thread_sem);
 
@@ -84,58 +110,111 @@ static int mmc_queue_thread(void *d)
  * on any queue on this host, and attempt to issue it.  This may
  * not be the queue we were asked to process.
  */
-static void mmc_request(struct request_queue *q)
+static void mmc_request_fn(struct request_queue *q)
 {
 	struct mmc_queue *mq = q->queuedata;
 	struct request *req;
-	int ret;
+	unsigned long flags;
+	struct mmc_context_info *cntx;
 
 	if (!mq) {
-		printk(KERN_ERR "MMC: killing requests for dead queue\n");
-		while ((req = elv_next_request(q)) != NULL) {
-			do {
-				ret = __blk_end_request(req, -EIO,
-							blk_rq_cur_bytes(req));
-			} while (ret);
+		while ((req = blk_fetch_request(q)) != NULL) {
+			req->cmd_flags |= REQ_QUIET;
+			__blk_end_request_all(req, -EIO);
 		}
 		return;
 	}
 
-	if (!mq->req)
+	cntx = &mq->card->host->context_info;
+	if (!mq->mqrq_cur->req && mq->mqrq_prev->req) {
+		/*
+		 * New MMC request arrived when MMC thread may be
+		 * blocked on the previous request to be complete
+		 * with no current request fetched
+		 */
+		spin_lock_irqsave(&cntx->lock, flags);
+		if (cntx->is_waiting_last_req) {
+			cntx->is_new_req = true;
+			wake_up_interruptible(&cntx->wait);
+		}
+		spin_unlock_irqrestore(&cntx->lock, flags);
+	} else if (!mq->mqrq_cur->req && !mq->mqrq_prev->req)
 		wake_up_process(mq->thread);
 }
 
+static struct scatterlist *mmc_alloc_sg(int sg_len, int *err)
+{
+	struct scatterlist *sg;
+
+	sg = kmalloc(sizeof(struct scatterlist)*sg_len, GFP_KERNEL);
+	if (!sg)
+		*err = -ENOMEM;
+	else {
+		*err = 0;
+		sg_init_table(sg, sg_len);
+	}
+
+	return sg;
+}
+
+static void mmc_queue_setup_discard(struct request_queue *q,
+				    struct mmc_card *card)
+{
+	unsigned max_discard;
+
+	max_discard = mmc_calc_max_discard(card);
+	if (!max_discard)
+		return;
+
+	queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, q);
+	q->limits.max_discard_sectors = max_discard;
+	if (card->erased_byte == 0 && !mmc_can_discard(card))
+		q->limits.discard_zeroes_data = 1;
+	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 = 0;
+	if (mmc_can_secure_erase_trim(card))
+		queue_flag_set_unlocked(QUEUE_FLAG_SECDISCARD, q);
+}
+
 /**
  * mmc_init_queue - initialise a queue structure.
  * @mq: mmc queue
  * @card: mmc card to attach this queue
  * @lock: queue lock
+ * @subname: partition subname
  *
  * Initialise a MMC card request queue.
  */
-int mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card, spinlock_t *lock)
+int mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card,
+		   spinlock_t *lock, const char *subname)
 {
 	struct mmc_host *host = card->host;
 	u64 limit = BLK_BOUNCE_HIGH;
 	int ret;
+	struct mmc_queue_req *mqrq_cur = &mq->mqrq[0];
+	struct mmc_queue_req *mqrq_prev = &mq->mqrq[1];
 
 	if (mmc_dev(host)->dma_mask && *mmc_dev(host)->dma_mask)
-		limit = *mmc_dev(host)->dma_mask;
+		limit = (u64)dma_max_pfn(mmc_dev(host)) << PAGE_SHIFT;
 
 	mq->card = card;
-	mq->queue = blk_init_queue(mmc_request, lock);
+	mq->queue = blk_init_queue(mmc_request_fn, lock);
 	if (!mq->queue)
 		return -ENOMEM;
 
+	mq->mqrq_cur = mqrq_cur;
+	mq->mqrq_prev = mqrq_prev;
 	mq->queue->queuedata = mq;
-	mq->req = NULL;
 
 	blk_queue_prep_rq(mq->queue, mmc_prep_request);
-	blk_queue_ordered(mq->queue, QUEUE_ORDERED_DRAIN, NULL);
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, mq->queue);
+	if (mmc_can_erase(card))
+		mmc_queue_setup_discard(mq->queue, card);
 
 #ifdef CONFIG_MMC_BLOCK_BOUNCE
-	if (host->max_hw_segs == 1) {
+	if (host->max_segs == 1) {
 		unsigned int bouncesz;
 
 		bouncesz = MMC_QUEUE_BOUNCESZ;
@@ -148,60 +227,71 @@ int mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card, spinlock_t *lock
 			bouncesz = host->max_blk_count * 512;
 
 		if (bouncesz > 512) {
-			mq->bounce_buf = kmalloc(bouncesz, GFP_KERNEL);
-			if (!mq->bounce_buf) {
-				printk(KERN_WARNING "%s: unable to "
-					"allocate bounce buffer\n",
+			mqrq_cur->bounce_buf = kmalloc(bouncesz, GFP_KERNEL);
+			if (!mqrq_cur->bounce_buf) {
+				pr_warning("%s: unable to "
+					"allocate bounce cur buffer\n",
+					mmc_card_name(card));
+			}
+			mqrq_prev->bounce_buf = kmalloc(bouncesz, GFP_KERNEL);
+			if (!mqrq_prev->bounce_buf) {
+				pr_warning("%s: unable to "
+					"allocate bounce prev buffer\n",
 					mmc_card_name(card));
+				kfree(mqrq_cur->bounce_buf);
+				mqrq_cur->bounce_buf = NULL;
 			}
 		}
 
-		if (mq->bounce_buf) {
+		if (mqrq_cur->bounce_buf && mqrq_prev->bounce_buf) {
 			blk_queue_bounce_limit(mq->queue, BLK_BOUNCE_ANY);
-			blk_queue_max_sectors(mq->queue, bouncesz / 512);
-			blk_queue_max_phys_segments(mq->queue, bouncesz / 512);
-			blk_queue_max_hw_segments(mq->queue, bouncesz / 512);
+			blk_queue_max_hw_sectors(mq->queue, bouncesz / 512);
+			blk_queue_max_segments(mq->queue, bouncesz / 512);
 			blk_queue_max_segment_size(mq->queue, bouncesz);
 
-			mq->sg = kmalloc(sizeof(struct scatterlist),
-				GFP_KERNEL);
-			if (!mq->sg) {
-				ret = -ENOMEM;
+			mqrq_cur->sg = mmc_alloc_sg(1, &ret);
+			if (ret)
 				goto cleanup_queue;
-			}
-			sg_init_table(mq->sg, 1);
 
-			mq->bounce_sg = kmalloc(sizeof(struct scatterlist) *
-				bouncesz / 512, GFP_KERNEL);
-			if (!mq->bounce_sg) {
-				ret = -ENOMEM;
+			mqrq_cur->bounce_sg =
+				mmc_alloc_sg(bouncesz / 512, &ret);
+			if (ret)
+				goto cleanup_queue;
+
+			mqrq_prev->sg = mmc_alloc_sg(1, &ret);
+			if (ret)
+				goto cleanup_queue;
+
+			mqrq_prev->bounce_sg =
+				mmc_alloc_sg(bouncesz / 512, &ret);
+			if (ret)
 				goto cleanup_queue;
-			}
-			sg_init_table(mq->bounce_sg, bouncesz / 512);
 		}
 	}
 #endif
 
-	if (!mq->bounce_buf) {
+	if (!mqrq_cur->bounce_buf && !mqrq_prev->bounce_buf) {
 		blk_queue_bounce_limit(mq->queue, limit);
-		blk_queue_max_sectors(mq->queue,
+		blk_queue_max_hw_sectors(mq->queue,
 			min(host->max_blk_count, host->max_req_size / 512));
-		blk_queue_max_phys_segments(mq->queue, host->max_phys_segs);
-		blk_queue_max_hw_segments(mq->queue, host->max_hw_segs);
+		blk_queue_max_segments(mq->queue, host->max_segs);
 		blk_queue_max_segment_size(mq->queue, host->max_seg_size);
 
-		mq->sg = kmalloc(sizeof(struct scatterlist) *
-			host->max_phys_segs, GFP_KERNEL);
-		if (!mq->sg) {
-			ret = -ENOMEM;
+		mqrq_cur->sg = mmc_alloc_sg(host->max_segs, &ret);
+		if (ret)
+			goto cleanup_queue;
+
+
+		mqrq_prev->sg = mmc_alloc_sg(host->max_segs, &ret);
+		if (ret)
 			goto cleanup_queue;
-		}
-		sg_init_table(mq->sg, host->max_phys_segs);
 	}
 
-	init_MUTEX(&mq->thread_sem);
+	sema_init(&mq->thread_sem, 1);
+
+	mq->thread = kthread_run(mmc_queue_thread, mq, "mmcqd/%d%s",
+		host->index, subname ? subname : "");
 
-	mq->thread = kthread_run(mmc_queue_thread, mq, "mmcqd");
 	if (IS_ERR(mq->thread)) {
 		ret = PTR_ERR(mq->thread);
 		goto free_bounce_sg;
@@ -209,16 +299,22 @@ int mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card, spinlock_t *lock
 
 	return 0;
  free_bounce_sg:
- 	if (mq->bounce_sg)
- 		kfree(mq->bounce_sg);
- 	mq->bounce_sg = NULL;
+	kfree(mqrq_cur->bounce_sg);
+	mqrq_cur->bounce_sg = NULL;
+	kfree(mqrq_prev->bounce_sg);
+	mqrq_prev->bounce_sg = NULL;
+
  cleanup_queue:
- 	if (mq->sg)
-		kfree(mq->sg);
-	mq->sg = NULL;
-	if (mq->bounce_buf)
-		kfree(mq->bounce_buf);
-	mq->bounce_buf = NULL;
+	kfree(mqrq_cur->sg);
+	mqrq_cur->sg = NULL;
+	kfree(mqrq_cur->bounce_buf);
+	mqrq_cur->bounce_buf = NULL;
+
+	kfree(mqrq_prev->sg);
+	mqrq_prev->sg = NULL;
+	kfree(mqrq_prev->bounce_buf);
+	mqrq_prev->bounce_buf = NULL;
+
 	blk_cleanup_queue(mq->queue);
 	return ret;
 }
@@ -227,11 +323,8 @@ void mmc_cleanup_queue(struct mmc_queue *mq)
 {
 	struct request_queue *q = mq->queue;
 	unsigned long flags;
-
-	/* Mark that we should start throwing out stragglers */
-	spin_lock_irqsave(q->queue_lock, flags);
-	q->queuedata = NULL;
-	spin_unlock_irqrestore(q->queue_lock, flags);
+	struct mmc_queue_req *mqrq_cur = mq->mqrq_cur;
+	struct mmc_queue_req *mqrq_prev = mq->mqrq_prev;
 
 	/* Make sure the queue isn't suspended, as that will deadlock */
 	mmc_queue_resume(mq);
@@ -239,23 +332,77 @@ void mmc_cleanup_queue(struct mmc_queue *mq)
 	/* Then terminate our worker thread */
 	kthread_stop(mq->thread);
 
- 	if (mq->bounce_sg)
- 		kfree(mq->bounce_sg);
- 	mq->bounce_sg = NULL;
+	/* Empty the queue */
+	spin_lock_irqsave(q->queue_lock, flags);
+	q->queuedata = NULL;
+	blk_start_queue(q);
+	spin_unlock_irqrestore(q->queue_lock, flags);
+
+	kfree(mqrq_cur->bounce_sg);
+	mqrq_cur->bounce_sg = NULL;
 
-	kfree(mq->sg);
-	mq->sg = NULL;
+	kfree(mqrq_cur->sg);
+	mqrq_cur->sg = NULL;
 
-	if (mq->bounce_buf)
-		kfree(mq->bounce_buf);
-	mq->bounce_buf = NULL;
+	kfree(mqrq_cur->bounce_buf);
+	mqrq_cur->bounce_buf = NULL;
 
-	blk_cleanup_queue(mq->queue);
+	kfree(mqrq_prev->bounce_sg);
+	mqrq_prev->bounce_sg = NULL;
+
+	kfree(mqrq_prev->sg);
+	mqrq_prev->sg = NULL;
+
+	kfree(mqrq_prev->bounce_buf);
+	mqrq_prev->bounce_buf = NULL;
 
 	mq->card = NULL;
 }
 EXPORT_SYMBOL(mmc_cleanup_queue);
 
+int mmc_packed_init(struct mmc_queue *mq, struct mmc_card *card)
+{
+	struct mmc_queue_req *mqrq_cur = &mq->mqrq[0];
+	struct mmc_queue_req *mqrq_prev = &mq->mqrq[1];
+	int ret = 0;
+
+
+	mqrq_cur->packed = kzalloc(sizeof(struct mmc_packed), GFP_KERNEL);
+	if (!mqrq_cur->packed) {
+		pr_warn("%s: unable to allocate packed cmd for mqrq_cur\n",
+			mmc_card_name(card));
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	mqrq_prev->packed = kzalloc(sizeof(struct mmc_packed), GFP_KERNEL);
+	if (!mqrq_prev->packed) {
+		pr_warn("%s: unable to allocate packed cmd for mqrq_prev\n",
+			mmc_card_name(card));
+		kfree(mqrq_cur->packed);
+		mqrq_cur->packed = NULL;
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	INIT_LIST_HEAD(&mqrq_cur->packed->list);
+	INIT_LIST_HEAD(&mqrq_prev->packed->list);
+
+out:
+	return ret;
+}
+
+void mmc_packed_clean(struct mmc_queue *mq)
+{
+	struct mmc_queue_req *mqrq_cur = &mq->mqrq[0];
+	struct mmc_queue_req *mqrq_prev = &mq->mqrq[1];
+
+	kfree(mqrq_cur->packed);
+	mqrq_cur->packed = NULL;
+	kfree(mqrq_prev->packed);
+	mqrq_prev->packed = NULL;
+}
+
 /**
  * mmc_queue_suspend - suspend a MMC request queue
  * @mq: MMC queue to suspend
@@ -300,30 +447,77 @@ void mmc_queue_resume(struct mmc_queue *mq)
 	}
 }
 
+static unsigned int mmc_queue_packed_map_sg(struct mmc_queue *mq,
+					    struct mmc_packed *packed,
+					    struct scatterlist *sg,
+					    enum mmc_packed_type cmd_type)
+{
+	struct scatterlist *__sg = sg;
+	unsigned int sg_len = 0;
+	struct request *req;
+
+	if (mmc_packed_wr(cmd_type)) {
+		unsigned int hdr_sz = mmc_large_sector(mq->card) ? 4096 : 512;
+		unsigned int max_seg_sz = queue_max_segment_size(mq->queue);
+		unsigned int len, remain, offset = 0;
+		u8 *buf = (u8 *)packed->cmd_hdr;
+
+		remain = hdr_sz;
+		do {
+			len = min(remain, max_seg_sz);
+			sg_set_buf(__sg, buf + offset, len);
+			offset += len;
+			remain -= len;
+			(__sg++)->page_link &= ~0x02;
+			sg_len++;
+		} while (remain);
+	}
+
+	list_for_each_entry(req, &packed->list, queuelist) {
+		sg_len += blk_rq_map_sg(mq->queue, req, __sg);
+		__sg = sg + (sg_len - 1);
+		(__sg++)->page_link &= ~0x02;
+	}
+	sg_mark_end(sg + (sg_len - 1));
+	return sg_len;
+}
+
 /*
  * Prepare the sg list(s) to be handed of to the host driver
  */
-unsigned int mmc_queue_map_sg(struct mmc_queue *mq)
+unsigned int mmc_queue_map_sg(struct mmc_queue *mq, struct mmc_queue_req *mqrq)
 {
 	unsigned int sg_len;
 	size_t buflen;
 	struct scatterlist *sg;
+	enum mmc_packed_type cmd_type;
 	int i;
 
-	if (!mq->bounce_buf)
-		return blk_rq_map_sg(mq->queue, mq->req, mq->sg);
+	cmd_type = mqrq->cmd_type;
+
+	if (!mqrq->bounce_buf) {
+		if (mmc_packed_cmd(cmd_type))
+			return mmc_queue_packed_map_sg(mq, mqrq->packed,
+						       mqrq->sg, cmd_type);
+		else
+			return blk_rq_map_sg(mq->queue, mqrq->req, mqrq->sg);
+	}
 
-	BUG_ON(!mq->bounce_sg);
+	BUG_ON(!mqrq->bounce_sg);
 
-	sg_len = blk_rq_map_sg(mq->queue, mq->req, mq->bounce_sg);
+	if (mmc_packed_cmd(cmd_type))
+		sg_len = mmc_queue_packed_map_sg(mq, mqrq->packed,
+						 mqrq->bounce_sg, cmd_type);
+	else
+		sg_len = blk_rq_map_sg(mq->queue, mqrq->req, mqrq->bounce_sg);
 
-	mq->bounce_sg_len = sg_len;
+	mqrq->bounce_sg_len = sg_len;
 
 	buflen = 0;
-	for_each_sg(mq->bounce_sg, sg, sg_len, i)
+	for_each_sg(mqrq->bounce_sg, sg, sg_len, i)
 		buflen += sg->length;
 
-	sg_init_one(mq->sg, mq->bounce_buf, buflen);
+	sg_init_one(mqrq->sg, mqrq->bounce_buf, buflen);
 
 	return 1;
 }
@@ -332,39 +526,30 @@ unsigned int mmc_queue_map_sg(struct mmc_queue *mq)
  * If writing, bounce the data to the buffer before the request
  * is sent to the host driver
  */
-void mmc_queue_bounce_pre(struct mmc_queue *mq)
+void mmc_queue_bounce_pre(struct mmc_queue_req *mqrq)
 {
-	unsigned long flags;
-
-	if (!mq->bounce_buf)
+	if (!mqrq->bounce_buf)
 		return;
 
-	if (rq_data_dir(mq->req) != WRITE)
+	if (rq_data_dir(mqrq->req) != WRITE)
 		return;
 
-	local_irq_save(flags);
-	sg_copy_to_buffer(mq->bounce_sg, mq->bounce_sg_len,
-		mq->bounce_buf, mq->sg[0].length);
-	local_irq_restore(flags);
+	sg_copy_to_buffer(mqrq->bounce_sg, mqrq->bounce_sg_len,
+		mqrq->bounce_buf, mqrq->sg[0].length);
 }
 
 /*
  * If reading, bounce the data from the buffer after the request
  * has been handled by the host driver
  */
-void mmc_queue_bounce_post(struct mmc_queue *mq)
+void mmc_queue_bounce_post(struct mmc_queue_req *mqrq)
 {
-	unsigned long flags;
-
-	if (!mq->bounce_buf)
+	if (!mqrq->bounce_buf)
 		return;
 
-	if (rq_data_dir(mq->req) != READ)
+	if (rq_data_dir(mqrq->req) != READ)
 		return;
 
-	local_irq_save(flags);
-	sg_copy_from_buffer(mq->bounce_sg, mq->bounce_sg_len,
-		mq->bounce_buf, mq->sg[0].length);
-	local_irq_restore(flags);
+	sg_copy_from_buffer(mqrq->bounce_sg, mqrq->bounce_sg_len,
+		mqrq->bounce_buf, mqrq->sg[0].length);
 }
-
diff --git a/drivers/mmc/card/queue.h b/drivers/mmc/card/queue.h
index 64e66e0d499..5752d50049a 100644
--- a/drivers/mmc/card/queue.h
+++ b/drivers/mmc/card/queue.h
@@ -1,31 +1,76 @@
 #ifndef MMC_QUEUE_H
 #define MMC_QUEUE_H
 
+#define MMC_REQ_SPECIAL_MASK	(REQ_DISCARD | REQ_FLUSH)
+
 struct request;
 struct task_struct;
 
+struct mmc_blk_request {
+	struct mmc_request	mrq;
+	struct mmc_command	sbc;
+	struct mmc_command	cmd;
+	struct mmc_command	stop;
+	struct mmc_data		data;
+};
+
+enum mmc_packed_type {
+	MMC_PACKED_NONE = 0,
+	MMC_PACKED_WRITE,
+};
+
+#define mmc_packed_cmd(type)	((type) != MMC_PACKED_NONE)
+#define mmc_packed_wr(type)	((type) == MMC_PACKED_WRITE)
+
+struct mmc_packed {
+	struct list_head	list;
+	u32			cmd_hdr[1024];
+	unsigned int		blocks;
+	u8			nr_entries;
+	u8			retries;
+	s16			idx_failure;
+};
+
+struct mmc_queue_req {
+	struct request		*req;
+	struct mmc_blk_request	brq;
+	struct scatterlist	*sg;
+	char			*bounce_buf;
+	struct scatterlist	*bounce_sg;
+	unsigned int		bounce_sg_len;
+	struct mmc_async_req	mmc_active;
+	enum mmc_packed_type	cmd_type;
+	struct mmc_packed	*packed;
+};
+
 struct mmc_queue {
 	struct mmc_card		*card;
 	struct task_struct	*thread;
 	struct semaphore	thread_sem;
 	unsigned int		flags;
-	struct request		*req;
+#define MMC_QUEUE_SUSPENDED	(1 << 0)
+#define MMC_QUEUE_NEW_REQUEST	(1 << 1)
+
 	int			(*issue_fn)(struct mmc_queue *, struct request *);
 	void			*data;
 	struct request_queue	*queue;
-	struct scatterlist	*sg;
-	char			*bounce_buf;
-	struct scatterlist	*bounce_sg;
-	unsigned int		bounce_sg_len;
+	struct mmc_queue_req	mqrq[2];
+	struct mmc_queue_req	*mqrq_cur;
+	struct mmc_queue_req	*mqrq_prev;
 };
 
-extern int mmc_init_queue(struct mmc_queue *, struct mmc_card *, spinlock_t *);
+extern int mmc_init_queue(struct mmc_queue *, struct mmc_card *, spinlock_t *,
+			  const char *);
 extern void mmc_cleanup_queue(struct mmc_queue *);
 extern void mmc_queue_suspend(struct mmc_queue *);
 extern void mmc_queue_resume(struct mmc_queue *);
 
-extern unsigned int mmc_queue_map_sg(struct mmc_queue *);
-extern void mmc_queue_bounce_pre(struct mmc_queue *);
-extern void mmc_queue_bounce_post(struct mmc_queue *);
+extern unsigned int mmc_queue_map_sg(struct mmc_queue *,
+				     struct mmc_queue_req *);
+extern void mmc_queue_bounce_pre(struct mmc_queue_req *);
+extern void mmc_queue_bounce_post(struct mmc_queue_req *);
+
+extern int mmc_packed_init(struct mmc_queue *, struct mmc_card *);
+extern void mmc_packed_clean(struct mmc_queue *);
 
 #endif
diff --git a/drivers/mmc/card/sdio_uart.c b/drivers/mmc/card/sdio_uart.c
index 36a8d53ad2a..f093cea0d06 100644
--- a/drivers/mmc/card/sdio_uart.c
+++ b/drivers/mmc/card/sdio_uart.c
@@ -29,13 +29,15 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
+#include <linux/sched.h>
 #include <linux/mutex.h>
 #include <linux/seq_file.h>
 #include <linux/serial_reg.h>
 #include <linux/circ_buf.h>
-#include <linux/gfp.h>
 #include <linux/tty.h>
 #include <linux/tty_flip.h>
+#include <linux/kfifo.h>
+#include <linux/slab.h>
 
 #include <linux/mmc/core.h>
 #include <linux/mmc/card.h>
@@ -46,19 +48,9 @@
 #define UART_NR		8	/* Number of UARTs this driver can handle */
 
 
-#define UART_XMIT_SIZE	PAGE_SIZE
+#define FIFO_SIZE	PAGE_SIZE
 #define WAKEUP_CHARS	256
 
-#define circ_empty(circ)	((circ)->head == (circ)->tail)
-#define circ_clear(circ)	((circ)->head = (circ)->tail = 0)
-
-#define circ_chars_pending(circ) \
-		(CIRC_CNT((circ)->head, (circ)->tail, UART_XMIT_SIZE))
-
-#define circ_chars_free(circ) \
-		(CIRC_SPACE((circ)->head, (circ)->tail, UART_XMIT_SIZE))
-
-
 struct uart_icount {
 	__u32	cts;
 	__u32	dsr;
@@ -73,20 +65,18 @@ struct uart_icount {
 };
 
 struct sdio_uart_port {
-	struct kref		kref;
-	struct tty_struct	*tty;
+	struct tty_port		port;
 	unsigned int		index;
-	unsigned int		opened;
-	struct mutex		open_lock;
 	struct sdio_func	*func;
 	struct mutex		func_lock;
 	struct task_struct	*in_sdio_uart_irq;
 	unsigned int		regs_offset;
-	struct circ_buf		xmit;
+	struct kfifo		xmit_fifo;
 	spinlock_t		write_lock;
 	struct uart_icount	icount;
 	unsigned int		uartclk;
 	unsigned int		mctrl;
+	unsigned int		rx_mctrl;
 	unsigned int		read_status_mask;
 	unsigned int		ignore_status_mask;
 	unsigned char		x_char;
@@ -101,10 +91,10 @@ static int sdio_uart_add_port(struct sdio_uart_port *port)
 {
 	int index, ret = -EBUSY;
 
-	kref_init(&port->kref);
-	mutex_init(&port->open_lock);
 	mutex_init(&port->func_lock);
 	spin_lock_init(&port->write_lock);
+	if (kfifo_alloc(&port->xmit_fifo, FIFO_SIZE, GFP_KERNEL))
+		return -ENOMEM;
 
 	spin_lock(&sdio_uart_table_lock);
 	for (index = 0; index < UART_NR; index++) {
@@ -130,22 +120,15 @@ static struct sdio_uart_port *sdio_uart_port_get(unsigned index)
 	spin_lock(&sdio_uart_table_lock);
 	port = sdio_uart_table[index];
 	if (port)
-		kref_get(&port->kref);
+		tty_port_get(&port->port);
 	spin_unlock(&sdio_uart_table_lock);
 
 	return port;
 }
 
-static void sdio_uart_port_destroy(struct kref *kref)
-{
-	struct sdio_uart_port *port =
-		container_of(kref, struct sdio_uart_port, kref);
-	kfree(port);
-}
-
 static void sdio_uart_port_put(struct sdio_uart_port *port)
 {
-	kref_put(&port->kref, sdio_uart_port_destroy);
+	tty_port_put(&port->port);
 }
 
 static void sdio_uart_port_remove(struct sdio_uart_port *port)
@@ -165,15 +148,15 @@ static void sdio_uart_port_remove(struct sdio_uart_port *port)
 	 * give up on that port ASAP.
 	 * Beware: the lock ordering is critical.
 	 */
-	mutex_lock(&port->open_lock);
+	mutex_lock(&port->port.mutex);
 	mutex_lock(&port->func_lock);
 	func = port->func;
 	sdio_claim_host(func);
 	port->func = NULL;
 	mutex_unlock(&port->func_lock);
-	if (port->opened)
-		tty_hangup(port->tty);
-	mutex_unlock(&port->open_lock);
+	/* tty_hangup is async so is this safe as is ?? */
+	tty_port_tty_hangup(&port->port, false);
+	mutex_unlock(&port->port.mutex);
 	sdio_release_irq(func);
 	sdio_disable_func(func);
 	sdio_release_host(func);
@@ -217,6 +200,8 @@ static unsigned int sdio_uart_get_mctrl(struct sdio_uart_port *port)
 	unsigned char status;
 	unsigned int ret;
 
+	/* FIXME: What stops this losing the delta bits and breaking
+	   sdio_uart_check_modem_status ? */
 	status = sdio_in(port, UART_MSR);
 
 	ret = 0;
@@ -231,7 +216,8 @@ static unsigned int sdio_uart_get_mctrl(struct sdio_uart_port *port)
 	return ret;
 }
 
-static void sdio_uart_write_mctrl(struct sdio_uart_port *port, unsigned int mctrl)
+static void sdio_uart_write_mctrl(struct sdio_uart_port *port,
+				  unsigned int mctrl)
 {
 	unsigned char mcr = 0;
 
@@ -387,9 +373,9 @@ static void sdio_uart_stop_rx(struct sdio_uart_port *port)
 	sdio_out(port, UART_IER, port->ier);
 }
 
-static void sdio_uart_receive_chars(struct sdio_uart_port *port, unsigned int *status)
+static void sdio_uart_receive_chars(struct sdio_uart_port *port,
+				    unsigned int *status)
 {
-	struct tty_struct *tty = port->tty;
 	unsigned int ch, flag;
 	int max_count = 256;
 
@@ -399,7 +385,7 @@ static void sdio_uart_receive_chars(struct sdio_uart_port *port, unsigned int *s
 		port->icount.rx++;
 
 		if (unlikely(*status & (UART_LSR_BI | UART_LSR_PE |
-				        UART_LSR_FE | UART_LSR_OE))) {
+					UART_LSR_FE | UART_LSR_OE))) {
 			/*
 			 * For statistics only
 			 */
@@ -417,33 +403,37 @@ static void sdio_uart_receive_chars(struct sdio_uart_port *port, unsigned int *s
 			 * Mask off conditions which should be ignored.
 			 */
 			*status &= port->read_status_mask;
-			if (*status & UART_LSR_BI) {
+			if (*status & UART_LSR_BI)
 				flag = TTY_BREAK;
-			} else if (*status & UART_LSR_PE)
+			else if (*status & UART_LSR_PE)
 				flag = TTY_PARITY;
 			else if (*status & UART_LSR_FE)
 				flag = TTY_FRAME;
 		}
 
 		if ((*status & port->ignore_status_mask & ~UART_LSR_OE) == 0)
-			tty_insert_flip_char(tty, ch, flag);
+			tty_insert_flip_char(&port->port, ch, flag);
 
 		/*
 		 * Overrun is special.  Since it's reported immediately,
 		 * it doesn't affect the current character.
 		 */
 		if (*status & ~port->ignore_status_mask & UART_LSR_OE)
-			tty_insert_flip_char(tty, 0, TTY_OVERRUN);
+			tty_insert_flip_char(&port->port, 0, TTY_OVERRUN);
 
 		*status = sdio_in(port, UART_LSR);
 	} while ((*status & UART_LSR_DR) && (max_count-- > 0));
-	tty_flip_buffer_push(tty);
+
+	tty_flip_buffer_push(&port->port);
 }
 
 static void sdio_uart_transmit_chars(struct sdio_uart_port *port)
 {
-	struct circ_buf *xmit = &port->xmit;
+	struct kfifo *xmit = &port->xmit_fifo;
 	int count;
+	struct tty_struct *tty;
+	u8 iobuf[16];
+	int len;
 
 	if (port->x_char) {
 		sdio_out(port, UART_TX, port->x_char);
@@ -451,30 +441,35 @@ static void sdio_uart_transmit_chars(struct sdio_uart_port *port)
 		port->x_char = 0;
 		return;
 	}
-	if (circ_empty(xmit) || port->tty->stopped || port->tty->hw_stopped) {
+
+	tty = tty_port_tty_get(&port->port);
+
+	if (tty == NULL || !kfifo_len(xmit) ||
+				tty->stopped || tty->hw_stopped) {
 		sdio_uart_stop_tx(port);
+		tty_kref_put(tty);
 		return;
 	}
 
-	count = 16;
-	do {
-		sdio_out(port, UART_TX, xmit->buf[xmit->tail]);
-		xmit->tail = (xmit->tail + 1) & (UART_XMIT_SIZE - 1);
+	len = kfifo_out_locked(xmit, iobuf, 16, &port->write_lock);
+	for (count = 0; count < len; count++) {
+		sdio_out(port, UART_TX, iobuf[count]);
 		port->icount.tx++;
-		if (circ_empty(xmit))
-			break;
-	} while (--count > 0);
-
-	if (circ_chars_pending(xmit) < WAKEUP_CHARS)
-		tty_wakeup(port->tty);
+	}
 
-	if (circ_empty(xmit))
-		sdio_uart_stop_tx(port);
+	len = kfifo_len(xmit);
+	if (len < WAKEUP_CHARS) {
+		tty_wakeup(tty);
+		if (len == 0)
+			sdio_uart_stop_tx(port);
+	}
+	tty_kref_put(tty);
 }
 
 static void sdio_uart_check_modem_status(struct sdio_uart_port *port)
 {
 	int status;
+	struct tty_struct *tty;
 
 	status = sdio_in(port, UART_MSR);
 
@@ -485,25 +480,35 @@ static void sdio_uart_check_modem_status(struct sdio_uart_port *port)
 		port->icount.rng++;
 	if (status & UART_MSR_DDSR)
 		port->icount.dsr++;
-	if (status & UART_MSR_DDCD)
+	if (status & UART_MSR_DDCD) {
 		port->icount.dcd++;
+		/* DCD raise - wake for open */
+		if (status & UART_MSR_DCD)
+			wake_up_interruptible(&port->port.open_wait);
+		else {
+			/* DCD drop - hang up if tty attached */
+			tty_port_tty_hangup(&port->port, false);
+		}
+	}
 	if (status & UART_MSR_DCTS) {
 		port->icount.cts++;
-		if (port->tty->termios->c_cflag & CRTSCTS) {
+		tty = tty_port_tty_get(&port->port);
+		if (tty && (tty->termios.c_cflag & CRTSCTS)) {
 			int cts = (status & UART_MSR_CTS);
-			if (port->tty->hw_stopped) {
+			if (tty->hw_stopped) {
 				if (cts) {
-					port->tty->hw_stopped = 0;
+					tty->hw_stopped = 0;
 					sdio_uart_start_tx(port);
-					tty_wakeup(port->tty);
+					tty_wakeup(tty);
 				}
 			} else {
 				if (!cts) {
-					port->tty->hw_stopped = 1;
+					tty->hw_stopped = 1;
 					sdio_uart_stop_tx(port);
 				}
 			}
 		}
+		tty_kref_put(tty);
 	}
 }
 
@@ -540,33 +545,81 @@ static void sdio_uart_irq(struct sdio_func *func)
 	port->in_sdio_uart_irq = NULL;
 }
 
-static int sdio_uart_startup(struct sdio_uart_port *port)
+static int uart_carrier_raised(struct tty_port *tport)
 {
-	unsigned long page;
+	struct sdio_uart_port *port =
+			container_of(tport, struct sdio_uart_port, port);
+	unsigned int ret = sdio_uart_claim_func(port);
+	if (ret)	/* Missing hardware shouldn't block for carrier */
+		return 1;
+	ret = sdio_uart_get_mctrl(port);
+	sdio_uart_release_func(port);
+	if (ret & TIOCM_CAR)
+		return 1;
+	return 0;
+}
+
+/**
+ *	uart_dtr_rts		-	 port helper to set uart signals
+ *	@tport: tty port to be updated
+ *	@onoff: set to turn on DTR/RTS
+ *
+ *	Called by the tty port helpers when the modem signals need to be
+ *	adjusted during an open, close and hangup.
+ */
+
+static void uart_dtr_rts(struct tty_port *tport, int onoff)
+{
+	struct sdio_uart_port *port =
+			container_of(tport, struct sdio_uart_port, port);
+	int ret = sdio_uart_claim_func(port);
+	if (ret)
+		return;
+	if (onoff == 0)
+		sdio_uart_clear_mctrl(port, TIOCM_DTR | TIOCM_RTS);
+	else
+		sdio_uart_set_mctrl(port, TIOCM_DTR | TIOCM_RTS);
+	sdio_uart_release_func(port);
+}
+
+/**
+ *	sdio_uart_activate	-	start up hardware
+ *	@tport: tty port to activate
+ *	@tty: tty bound to this port
+ *
+ *	Activate a tty port. The port locking guarantees us this will be
+ *	run exactly once per set of opens, and if successful will see the
+ *	shutdown method run exactly once to match. Start up and shutdown are
+ *	protected from each other by the internal locking and will not run
+ *	at the same time even during a hangup event.
+ *
+ *	If we successfully start up the port we take an extra kref as we
+ *	will keep it around until shutdown when the kref is dropped.
+ */
+
+static int sdio_uart_activate(struct tty_port *tport, struct tty_struct *tty)
+{
+	struct sdio_uart_port *port =
+			container_of(tport, struct sdio_uart_port, port);
 	int ret;
 
 	/*
 	 * Set the TTY IO error marker - we will only clear this
 	 * once we have successfully opened the port.
 	 */
-	set_bit(TTY_IO_ERROR, &port->tty->flags);
+	set_bit(TTY_IO_ERROR, &tty->flags);
 
-	/* Initialise and allocate the transmit buffer. */
-	page = __get_free_page(GFP_KERNEL);
-	if (!page)
-		return -ENOMEM;
-	port->xmit.buf = (unsigned char *)page;
-	circ_clear(&port->xmit);
+	kfifo_reset(&port->xmit_fifo);
 
 	ret = sdio_uart_claim_func(port);
 	if (ret)
-		goto err1;
+		return ret;
 	ret = sdio_enable_func(port->func);
 	if (ret)
-		goto err2;
+		goto err1;
 	ret = sdio_claim_irq(port->func, sdio_uart_irq);
 	if (ret)
-		goto err3;
+		goto err2;
 
 	/*
 	 * Clear the FIFO buffers and disable them.
@@ -574,7 +627,7 @@ static int sdio_uart_startup(struct sdio_uart_port *port)
 	 */
 	sdio_out(port, UART_FCR, UART_FCR_ENABLE_FIFO);
 	sdio_out(port, UART_FCR, UART_FCR_ENABLE_FIFO |
-			UART_FCR_CLEAR_RCVR | UART_FCR_CLEAR_XMIT);
+		       UART_FCR_CLEAR_RCVR | UART_FCR_CLEAR_XMIT);
 	sdio_out(port, UART_FCR, 0);
 
 	/*
@@ -590,19 +643,19 @@ static int sdio_uart_startup(struct sdio_uart_port *port)
 	 */
 	sdio_out(port, UART_LCR, UART_LCR_WLEN8);
 
-	port->ier = UART_IER_RLSI | UART_IER_RDI | UART_IER_RTOIE | UART_IER_UUE;
+	port->ier = UART_IER_RLSI|UART_IER_RDI|UART_IER_RTOIE|UART_IER_UUE;
 	port->mctrl = TIOCM_OUT2;
 
-	sdio_uart_change_speed(port, port->tty->termios, NULL);
+	sdio_uart_change_speed(port, &tty->termios, NULL);
 
-	if (port->tty->termios->c_cflag & CBAUD)
+	if (tty->termios.c_cflag & CBAUD)
 		sdio_uart_set_mctrl(port, TIOCM_RTS | TIOCM_DTR);
 
-	if (port->tty->termios->c_cflag & CRTSCTS)
+	if (tty->termios.c_cflag & CRTSCTS)
 		if (!(sdio_uart_get_mctrl(port) & TIOCM_CTS))
-			port->tty->hw_stopped = 1;
+			tty->hw_stopped = 1;
 
-	clear_bit(TTY_IO_ERROR, &port->tty->flags);
+	clear_bit(TTY_IO_ERROR, &tty->flags);
 
 	/* Kick the IRQ handler once while we're still holding the host lock */
 	sdio_uart_irq(port->func);
@@ -610,32 +663,36 @@ static int sdio_uart_startup(struct sdio_uart_port *port)
 	sdio_uart_release_func(port);
 	return 0;
 
-err3:
-	sdio_disable_func(port->func);
 err2:
-	sdio_uart_release_func(port);
+	sdio_disable_func(port->func);
 err1:
-	free_page((unsigned long)port->xmit.buf);
+	sdio_uart_release_func(port);
 	return ret;
 }
 
-static void sdio_uart_shutdown(struct sdio_uart_port *port)
+/**
+ *	sdio_uart_shutdown	-	stop hardware
+ *	@tport: tty port to shut down
+ *
+ *	Deactivate a tty port. The port locking guarantees us this will be
+ *	run only if a successful matching activate already ran. The two are
+ *	protected from each other by the internal locking and will not run
+ *	at the same time even during a hangup event.
+ */
+
+static void sdio_uart_shutdown(struct tty_port *tport)
 {
+	struct sdio_uart_port *port =
+			container_of(tport, struct sdio_uart_port, port);
 	int ret;
 
 	ret = sdio_uart_claim_func(port);
 	if (ret)
-		goto skip;
+		return;
 
 	sdio_uart_stop_rx(port);
 
-	/* TODO: wait here for TX FIFO to drain */
-
-	/* Turn off DTR and RTS early. */
-	if (port->tty->termios->c_cflag & HUPCL)
-		sdio_uart_clear_mctrl(port, TIOCM_DTR | TIOCM_RTS);
-
-	 /* Disable interrupts from this port */
+	/* Disable interrupts from this port */
 	sdio_release_irq(port->func);
 	port->ier = 0;
 	sdio_out(port, UART_IER, 0);
@@ -653,108 +710,87 @@ static void sdio_uart_shutdown(struct sdio_uart_port *port)
 	sdio_disable_func(port->func);
 
 	sdio_uart_release_func(port);
-
-skip:
-	/* Free the transmit buffer page. */
-	free_page((unsigned long)port->xmit.buf);
 }
 
-static int sdio_uart_open (struct tty_struct *tty, struct file * filp)
+static void sdio_uart_port_destroy(struct tty_port *tport)
 {
-	struct sdio_uart_port *port;
-	int ret;
-
-	port = sdio_uart_port_get(tty->index);
-	if (!port)
-		return -ENODEV;
+	struct sdio_uart_port *port =
+		container_of(tport, struct sdio_uart_port, port);
+	kfifo_free(&port->xmit_fifo);
+	kfree(port);
+}
 
-	mutex_lock(&port->open_lock);
+/**
+ *	sdio_uart_install	-	install method
+ *	@driver: the driver in use (sdio_uart in our case)
+ *	@tty: the tty being bound
+ *
+ *	Look up and bind the tty and the driver together. Initialize
+ *	any needed private data (in our case the termios)
+ */
 
-	/*
-	 * Make sure not to mess up with a dead port
-	 * which has not been closed yet.
-	 */
-	if (tty->driver_data && tty->driver_data != port) {
-		mutex_unlock(&port->open_lock);
-		sdio_uart_port_put(port);
-		return -EBUSY;
-	}
+static int sdio_uart_install(struct tty_driver *driver, struct tty_struct *tty)
+{
+	int idx = tty->index;
+	struct sdio_uart_port *port = sdio_uart_port_get(idx);
+	int ret = tty_standard_install(driver, tty);
 
-	if (!port->opened) {
+	if (ret == 0)
+		/* This is the ref sdio_uart_port get provided */
 		tty->driver_data = port;
-		port->tty = tty;
-		ret = sdio_uart_startup(port);
-		if (ret) {
-			tty->driver_data = NULL;
-			port->tty = NULL;
-			mutex_unlock(&port->open_lock);
-			sdio_uart_port_put(port);
-			return ret;
-		}
-	}
-	port->opened++;
-	mutex_unlock(&port->open_lock);
-	return 0;
+	else
+		sdio_uart_port_put(port);
+	return ret;
 }
 
-static void sdio_uart_close(struct tty_struct *tty, struct file * filp)
+/**
+ *	sdio_uart_cleanup	-	called on the last tty kref drop
+ *	@tty: the tty being destroyed
+ *
+ *	Called asynchronously when the last reference to the tty is dropped.
+ *	We cannot destroy the tty->driver_data port kref until this point
+ */
+
+static void sdio_uart_cleanup(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
+	tty->driver_data = NULL;	/* Bug trap */
+	sdio_uart_port_put(port);
+}
 
-	if (!port)
-		return;
+/*
+ *	Open/close/hangup is now entirely boilerplate
+ */
 
-	mutex_lock(&port->open_lock);
-	BUG_ON(!port->opened);
+static int sdio_uart_open(struct tty_struct *tty, struct file *filp)
+{
+	struct sdio_uart_port *port = tty->driver_data;
+	return tty_port_open(&port->port, tty, filp);
+}
 
-	/*
-	 * This is messy.  The tty layer calls us even when open()
-	 * returned an error.  Ignore this close request if tty->count
-	 * is larger than port->count.
-	 */
-	if (tty->count > port->opened) {
-		mutex_unlock(&port->open_lock);
-		return;
-	}
+static void sdio_uart_close(struct tty_struct *tty, struct file * filp)
+{
+	struct sdio_uart_port *port = tty->driver_data;
+	tty_port_close(&port->port, tty, filp);
+}
 
-	if (--port->opened == 0) {
-		tty->closing = 1;
-		sdio_uart_shutdown(port);
-		tty_ldisc_flush(tty);
-		port->tty = NULL;
-		tty->driver_data = NULL;
-		tty->closing = 0;
-	}
-	mutex_unlock(&port->open_lock);
-	sdio_uart_port_put(port);
+static void sdio_uart_hangup(struct tty_struct *tty)
+{
+	struct sdio_uart_port *port = tty->driver_data;
+	tty_port_hangup(&port->port);
 }
 
-static int sdio_uart_write(struct tty_struct * tty, const unsigned char *buf,
+static int sdio_uart_write(struct tty_struct *tty, const unsigned char *buf,
 			   int count)
 {
 	struct sdio_uart_port *port = tty->driver_data;
-	struct circ_buf *circ = &port->xmit;
-	int c, ret = 0;
+	int ret;
 
 	if (!port->func)
 		return -ENODEV;
 
-	spin_lock(&port->write_lock);
-	while (1) {
-		c = CIRC_SPACE_TO_END(circ->head, circ->tail, UART_XMIT_SIZE);
-		if (count < c)
-			c = count;
-		if (c <= 0)
-			break;
-		memcpy(circ->buf + circ->head, buf, c);
-		circ->head = (circ->head + c) & (UART_XMIT_SIZE - 1);
-		buf += c;
-		count -= c;
-		ret += c;
-	}
-	spin_unlock(&port->write_lock);
-
-	if ( !(port->ier & UART_IER_THRI)) {
+	ret = kfifo_in_locked(&port->xmit_fifo, buf, count, &port->write_lock);
+	if (!(port->ier & UART_IER_THRI)) {
 		int err = sdio_uart_claim_func(port);
 		if (!err) {
 			sdio_uart_start_tx(port);
@@ -770,13 +806,13 @@ static int sdio_uart_write(struct tty_struct * tty, const unsigned char *buf,
 static int sdio_uart_write_room(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
-	return port ? circ_chars_free(&port->xmit) : 0;
+	return FIFO_SIZE - kfifo_len(&port->xmit_fifo);
 }
 
 static int sdio_uart_chars_in_buffer(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
-	return port ? circ_chars_pending(&port->xmit) : 0;
+	return kfifo_len(&port->xmit_fifo);
 }
 
 static void sdio_uart_send_xchar(struct tty_struct *tty, char ch)
@@ -797,7 +833,7 @@ static void sdio_uart_throttle(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
 
-	if (!I_IXOFF(tty) && !(tty->termios->c_cflag & CRTSCTS))
+	if (!I_IXOFF(tty) && !(tty->termios.c_cflag & CRTSCTS))
 		return;
 
 	if (sdio_uart_claim_func(port) != 0)
@@ -808,7 +844,7 @@ static void sdio_uart_throttle(struct tty_struct *tty)
 		sdio_uart_start_tx(port);
 	}
 
-	if (tty->termios->c_cflag & CRTSCTS)
+	if (tty->termios.c_cflag & CRTSCTS)
 		sdio_uart_clear_mctrl(port, TIOCM_RTS);
 
 	sdio_uart_irq(port->func);
@@ -819,7 +855,7 @@ static void sdio_uart_unthrottle(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
 
-	if (!I_IXOFF(tty) && !(tty->termios->c_cflag & CRTSCTS))
+	if (!I_IXOFF(tty) && !(tty->termios.c_cflag & CRTSCTS))
 		return;
 
 	if (sdio_uart_claim_func(port) != 0)
@@ -834,28 +870,23 @@ static void sdio_uart_unthrottle(struct tty_struct *tty)
 		}
 	}
 
-	if (tty->termios->c_cflag & CRTSCTS)
+	if (tty->termios.c_cflag & CRTSCTS)
 		sdio_uart_set_mctrl(port, TIOCM_RTS);
 
 	sdio_uart_irq(port->func);
 	sdio_uart_release_func(port);
 }
 
-static void sdio_uart_set_termios(struct tty_struct *tty, struct ktermios *old_termios)
+static void sdio_uart_set_termios(struct tty_struct *tty,
+						struct ktermios *old_termios)
 {
 	struct sdio_uart_port *port = tty->driver_data;
-	unsigned int cflag = tty->termios->c_cflag;
-
-#define RELEVANT_IFLAG(iflag)   ((iflag) & (IGNBRK|BRKINT|IGNPAR|PARMRK|INPCK))
-
-	if ((cflag ^ old_termios->c_cflag) == 0 &&
-	    RELEVANT_IFLAG(tty->termios->c_iflag ^ old_termios->c_iflag) == 0)
-		return;
+	unsigned int cflag = tty->termios.c_cflag;
 
 	if (sdio_uart_claim_func(port) != 0)
 		return;
 
-	sdio_uart_change_speed(port, tty->termios, old_termios);
+	sdio_uart_change_speed(port, &tty->termios, old_termios);
 
 	/* Handle transition to B0 status */
 	if ((old_termios->c_cflag & CBAUD) && !(cflag & CBAUD))
@@ -905,7 +936,7 @@ static int sdio_uart_break_ctl(struct tty_struct *tty, int break_state)
 	return 0;
 }
 
-static int sdio_uart_tiocmget(struct tty_struct *tty, struct file *file)
+static int sdio_uart_tiocmget(struct tty_struct *tty)
 {
 	struct sdio_uart_port *port = tty->driver_data;
 	int result;
@@ -919,14 +950,14 @@ static int sdio_uart_tiocmget(struct tty_struct *tty, struct file *file)
 	return result;
 }
 
-static int sdio_uart_tiocmset(struct tty_struct *tty, struct file *file,
+static int sdio_uart_tiocmset(struct tty_struct *tty,
 			      unsigned int set, unsigned int clear)
 {
 	struct sdio_uart_port *port = tty->driver_data;
 	int result;
 
-	result =sdio_uart_claim_func(port);
-	if(!result) {
+	result = sdio_uart_claim_func(port);
+	if (!result) {
 		sdio_uart_update_mctrl(port, set, clear);
 		sdio_uart_release_func(port);
 	}
@@ -944,33 +975,33 @@ static int sdio_uart_proc_show(struct seq_file *m, void *v)
 		struct sdio_uart_port *port = sdio_uart_port_get(i);
 		if (port) {
 			seq_printf(m, "%d: uart:SDIO", i);
-			if(capable(CAP_SYS_ADMIN)) {
+			if (capable(CAP_SYS_ADMIN)) {
 				seq_printf(m, " tx:%d rx:%d",
-					       port->icount.tx, port->icount.rx);
+					      port->icount.tx, port->icount.rx);
 				if (port->icount.frame)
 					seq_printf(m, " fe:%d",
-						       port->icount.frame);
+						      port->icount.frame);
 				if (port->icount.parity)
 					seq_printf(m, " pe:%d",
-						       port->icount.parity);
+						      port->icount.parity);
 				if (port->icount.brk)
 					seq_printf(m, " brk:%d",
-						       port->icount.brk);
+						      port->icount.brk);
 				if (port->icount.overrun)
 					seq_printf(m, " oe:%d",
-						       port->icount.overrun);
+						      port->icount.overrun);
 				if (port->icount.cts)
 					seq_printf(m, " cts:%d",
-						       port->icount.cts);
+						      port->icount.cts);
 				if (port->icount.dsr)
 					seq_printf(m, " dsr:%d",
-						       port->icount.dsr);
+						      port->icount.dsr);
 				if (port->icount.rng)
 					seq_printf(m, " rng:%d",
-						       port->icount.rng);
+						      port->icount.rng);
 				if (port->icount.dcd)
 					seq_printf(m, " dcd:%d",
-						       port->icount.dcd);
+						      port->icount.dcd);
 			}
 			sdio_uart_port_put(port);
 			seq_putc(m, '\n');
@@ -992,6 +1023,14 @@ static const struct file_operations sdio_uart_proc_fops = {
 	.release	= single_release,
 };
 
+static const struct tty_port_operations sdio_uart_port_ops = {
+	.dtr_rts = uart_dtr_rts,
+	.carrier_raised = uart_carrier_raised,
+	.shutdown = sdio_uart_shutdown,
+	.activate = sdio_uart_activate,
+	.destruct = sdio_uart_port_destroy,
+};
+
 static const struct tty_operations sdio_uart_ops = {
 	.open			= sdio_uart_open,
 	.close			= sdio_uart_close,
@@ -1002,9 +1041,12 @@ static const struct tty_operations sdio_uart_ops = {
 	.throttle		= sdio_uart_throttle,
 	.unthrottle		= sdio_uart_unthrottle,
 	.set_termios		= sdio_uart_set_termios,
+	.hangup			= sdio_uart_hangup,
 	.break_ctl		= sdio_uart_break_ctl,
 	.tiocmget		= sdio_uart_tiocmget,
 	.tiocmset		= sdio_uart_tiocmset,
+	.install		= sdio_uart_install,
+	.cleanup		= sdio_uart_cleanup,
 	.proc_fops		= &sdio_uart_proc_fops,
 };
 
@@ -1021,7 +1063,7 @@ static int sdio_uart_probe(struct sdio_func *func,
 		return -ENOMEM;
 
 	if (func->class == SDIO_CLASS_UART) {
-		printk(KERN_WARNING "%s: need info on UART class basic setup\n",
+		pr_warning("%s: need info on UART class basic setup\n",
 		       sdio_func_id(func));
 		kfree(port);
 		return -ENOSYS;
@@ -1040,23 +1082,23 @@ static int sdio_uart_probe(struct sdio_func *func,
 				break;
 		}
 		if (!tpl) {
-			printk(KERN_WARNING
-			       "%s: can't find tuple 0x91 subtuple 0 (SUBTPL_SIOREG) for GPS class\n",
+			pr_warning(
+       "%s: can't find tuple 0x91 subtuple 0 (SUBTPL_SIOREG) for GPS class\n",
 			       sdio_func_id(func));
 			kfree(port);
 			return -EINVAL;
 		}
-		printk(KERN_DEBUG "%s: Register ID = 0x%02x, Exp ID = 0x%02x\n",
+		pr_debug("%s: Register ID = 0x%02x, Exp ID = 0x%02x\n",
 		       sdio_func_id(func), tpl->data[2], tpl->data[3]);
 		port->regs_offset = (tpl->data[4] << 0) |
 				    (tpl->data[5] << 8) |
 				    (tpl->data[6] << 16);
-		printk(KERN_DEBUG "%s: regs offset = 0x%x\n",
+		pr_debug("%s: regs offset = 0x%x\n",
 		       sdio_func_id(func), port->regs_offset);
 		port->uartclk = tpl->data[7] * 115200;
 		if (port->uartclk == 0)
 			port->uartclk = 115200;
-		printk(KERN_DEBUG "%s: clk %d baudcode %u 4800-div %u\n",
+		pr_debug("%s: clk %d baudcode %u 4800-div %u\n",
 		       sdio_func_id(func), port->uartclk,
 		       tpl->data[7], tpl->data[8] | (tpl->data[9] << 8));
 	} else {
@@ -1066,13 +1108,16 @@ static int sdio_uart_probe(struct sdio_func *func,
 
 	port->func = func;
 	sdio_set_drvdata(func, port);
+	tty_port_init(&port->port);
+	port->port.ops = &sdio_uart_port_ops;
 
 	ret = sdio_uart_add_port(port);
 	if (ret) {
 		kfree(port);
 	} else {
 		struct device *dev;
-		dev = tty_register_device(sdio_uart_tty_driver, port->index, &func->dev);
+		dev = tty_port_register_device(&port->port,
+				sdio_uart_tty_driver, port->index, &func->dev);
 		if (IS_ERR(dev)) {
 			sdio_uart_port_remove(port);
 			ret = PTR_ERR(dev);
@@ -1114,7 +1159,6 @@ static int __init sdio_uart_init(void)
 	if (!tty_drv)
 		return -ENOMEM;
 
-	tty_drv->owner = THIS_MODULE;
 	tty_drv->driver_name = "sdio_uart";
 	tty_drv->name =   "ttySDIO";
 	tty_drv->major = 0;  /* dynamically allocated */
diff --git a/drivers/mmc/core/Kconfig b/drivers/mmc/core/Kconfig
index ab37a6d9d32..9ebee72d9c3 100644
--- a/drivers/mmc/core/Kconfig
+++ b/drivers/mmc/core/Kconfig
@@ -2,15 +2,12 @@
 # MMC core configuration
 #
 
-config MMC_UNSAFE_RESUME
-	bool "Allow unsafe resume (DANGEROUS)"
+config MMC_CLKGATE
+	bool "MMC host clock gating"
 	help
-	  If you say Y here, the MMC layer will assume that all cards
-	  stayed in their respective slots during the suspend. The
-	  normal behaviour is to remove them at suspend and
-	  redetecting them at resume. Breaking this assumption will
-	  in most cases result in data corruption.
-
-	  This option is usually just for embedded systems which use
-	  a MMC/SD card for rootfs. Most people should say N here.
+	  This will attempt to aggressively gate the clock to the MMC card.
+	  This is done to save power due to gating off the logic and bus
+	  noise when the MMC card is not in use. Your host driver has to
+	  support handling this in order for it to be of any use.
 
+	  If unsure, say N.
diff --git a/drivers/mmc/core/Makefile b/drivers/mmc/core/Makefile
index 889e5f898f6..38ed210ce2f 100644
--- a/drivers/mmc/core/Makefile
+++ b/drivers/mmc/core/Makefile
@@ -2,14 +2,11 @@
 # Makefile for the kernel mmc core.
 #
 
-ifeq ($(CONFIG_MMC_DEBUG),y)
-	EXTRA_CFLAGS		+= -DDEBUG
-endif
-
 obj-$(CONFIG_MMC)		+= mmc_core.o
 mmc_core-y			:= core.o bus.o host.o \
 				   mmc.o mmc_ops.o sd.o sd_ops.o \
 				   sdio.o sdio_ops.o sdio_bus.o \
-				   sdio_cis.o sdio_io.o sdio_irq.o
+				   sdio_cis.o sdio_io.o sdio_irq.o \
+				   quirks.o slot-gpio.o
 
 mmc_core-$(CONFIG_DEBUG_FS)	+= debugfs.o
diff --git a/drivers/mmc/core/bus.c b/drivers/mmc/core/bus.c
index bdb165f9304..d2dbf02022b 100644
--- a/drivers/mmc/core/bus.c
+++ b/drivers/mmc/core/bus.c
@@ -11,8 +11,12 @@
  *  MMC card bus driver model
  */
 
+#include <linux/export.h>
 #include <linux/device.h>
 #include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/pm_runtime.h>
 
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
@@ -21,13 +25,12 @@
 #include "sdio_cis.h"
 #include "bus.h"
 
-#define dev_to_mmc_card(d)	container_of(d, struct mmc_card, dev)
 #define to_mmc_driver(d)	container_of(d, struct mmc_driver, drv)
 
-static ssize_t mmc_type_show(struct device *dev,
+static ssize_t type_show(struct device *dev,
 	struct device_attribute *attr, char *buf)
 {
-	struct mmc_card *card = dev_to_mmc_card(dev);
+	struct mmc_card *card = mmc_dev_to_card(dev);
 
 	switch (card->type) {
 	case MMC_TYPE_MMC:
@@ -36,15 +39,19 @@ static ssize_t mmc_type_show(struct device *dev,
 		return sprintf(buf, "SD\n");
 	case MMC_TYPE_SDIO:
 		return sprintf(buf, "SDIO\n");
+	case MMC_TYPE_SD_COMBO:
+		return sprintf(buf, "SDcombo\n");
 	default:
 		return -EFAULT;
 	}
 }
+static DEVICE_ATTR_RO(type);
 
-static struct device_attribute mmc_dev_attrs[] = {
-	__ATTR(type, S_IRUGO, mmc_type_show, NULL),
-	__ATTR_NULL,
+static struct attribute *mmc_dev_attrs[] = {
+	&dev_attr_type.attr,
+	NULL,
 };
+ATTRIBUTE_GROUPS(mmc_dev);
 
 /*
  * This currently matches any MMC driver to any MMC card - drivers
@@ -59,7 +66,7 @@ static int mmc_bus_match(struct device *dev, struct device_driver *drv)
 static int
 mmc_bus_uevent(struct device *dev, struct kobj_uevent_env *env)
 {
-	struct mmc_card *card = dev_to_mmc_card(dev);
+	struct mmc_card *card = mmc_dev_to_card(dev);
 	const char *type;
 	int retval = 0;
 
@@ -73,6 +80,9 @@ mmc_bus_uevent(struct device *dev, struct kobj_uevent_env *env)
 	case MMC_TYPE_SDIO:
 		type = "SDIO";
 		break;
+	case MMC_TYPE_SD_COMBO:
+		type = "SDcombo";
+		break;
 	default:
 		type = NULL;
 	}
@@ -99,7 +109,7 @@ mmc_bus_uevent(struct device *dev, struct kobj_uevent_env *env)
 static int mmc_bus_probe(struct device *dev)
 {
 	struct mmc_driver *drv = to_mmc_driver(dev->driver);
-	struct mmc_card *card = dev_to_mmc_card(dev);
+	struct mmc_card *card = mmc_dev_to_card(dev);
 
 	return drv->probe(card);
 }
@@ -107,44 +117,108 @@ static int mmc_bus_probe(struct device *dev)
 static int mmc_bus_remove(struct device *dev)
 {
 	struct mmc_driver *drv = to_mmc_driver(dev->driver);
-	struct mmc_card *card = dev_to_mmc_card(dev);
+	struct mmc_card *card = mmc_dev_to_card(dev);
 
 	drv->remove(card);
 
 	return 0;
 }
 
-static int mmc_bus_suspend(struct device *dev, pm_message_t state)
+static void mmc_bus_shutdown(struct device *dev)
+{
+	struct mmc_driver *drv = to_mmc_driver(dev->driver);
+	struct mmc_card *card = mmc_dev_to_card(dev);
+	struct mmc_host *host = card->host;
+	int ret;
+
+	if (dev->driver && drv->shutdown)
+		drv->shutdown(card);
+
+	if (host->bus_ops->shutdown) {
+		ret = host->bus_ops->shutdown(host);
+		if (ret)
+			pr_warn("%s: error %d during shutdown\n",
+				mmc_hostname(host), ret);
+	}
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int mmc_bus_suspend(struct device *dev)
 {
 	struct mmc_driver *drv = to_mmc_driver(dev->driver);
-	struct mmc_card *card = dev_to_mmc_card(dev);
-	int ret = 0;
+	struct mmc_card *card = mmc_dev_to_card(dev);
+	struct mmc_host *host = card->host;
+	int ret;
+
+	if (dev->driver && drv->suspend) {
+		ret = drv->suspend(card);
+		if (ret)
+			return ret;
+	}
 
-	if (dev->driver && drv->suspend)
-		ret = drv->suspend(card, state);
+	ret = host->bus_ops->suspend(host);
 	return ret;
 }
 
 static int mmc_bus_resume(struct device *dev)
 {
 	struct mmc_driver *drv = to_mmc_driver(dev->driver);
-	struct mmc_card *card = dev_to_mmc_card(dev);
-	int ret = 0;
+	struct mmc_card *card = mmc_dev_to_card(dev);
+	struct mmc_host *host = card->host;
+	int ret;
+
+	ret = host->bus_ops->resume(host);
+	if (ret)
+		pr_warn("%s: error %d during resume (card was removed?)\n",
+			mmc_hostname(host), ret);
 
 	if (dev->driver && drv->resume)
 		ret = drv->resume(card);
+
 	return ret;
 }
+#endif
+
+#ifdef CONFIG_PM_RUNTIME
+
+static int mmc_runtime_suspend(struct device *dev)
+{
+	struct mmc_card *card = mmc_dev_to_card(dev);
+	struct mmc_host *host = card->host;
+
+	return host->bus_ops->runtime_suspend(host);
+}
+
+static int mmc_runtime_resume(struct device *dev)
+{
+	struct mmc_card *card = mmc_dev_to_card(dev);
+	struct mmc_host *host = card->host;
+
+	return host->bus_ops->runtime_resume(host);
+}
+
+static int mmc_runtime_idle(struct device *dev)
+{
+	return 0;
+}
+
+#endif /* !CONFIG_PM_RUNTIME */
+
+static const struct dev_pm_ops mmc_bus_pm_ops = {
+	SET_RUNTIME_PM_OPS(mmc_runtime_suspend, mmc_runtime_resume,
+			mmc_runtime_idle)
+	SET_SYSTEM_SLEEP_PM_OPS(mmc_bus_suspend, mmc_bus_resume)
+};
 
 static struct bus_type mmc_bus_type = {
 	.name		= "mmc",
-	.dev_attrs	= mmc_dev_attrs,
+	.dev_groups	= mmc_dev_groups,
 	.match		= mmc_bus_match,
 	.uevent		= mmc_bus_uevent,
 	.probe		= mmc_bus_probe,
 	.remove		= mmc_bus_remove,
-	.suspend	= mmc_bus_suspend,
-	.resume		= mmc_bus_resume,
+	.shutdown	= mmc_bus_shutdown,
+	.pm		= &mmc_bus_pm_ops,
 };
 
 int mmc_register_bus(void)
@@ -183,12 +257,11 @@ EXPORT_SYMBOL(mmc_unregister_driver);
 
 static void mmc_release_card(struct device *dev)
 {
-	struct mmc_card *card = dev_to_mmc_card(dev);
+	struct mmc_card *card = mmc_dev_to_card(dev);
 
 	sdio_free_common_cis(card);
 
-	if (card->info)
-		kfree(card->info);
+	kfree(card->info);
 
 	kfree(card);
 }
@@ -223,6 +296,15 @@ int mmc_add_card(struct mmc_card *card)
 {
 	int ret;
 	const char *type;
+	const char *uhs_bus_speed_mode = "";
+	static const char *const uhs_speeds[] = {
+		[UHS_SDR12_BUS_SPEED] = "SDR12 ",
+		[UHS_SDR25_BUS_SPEED] = "SDR25 ",
+		[UHS_SDR50_BUS_SPEED] = "SDR50 ",
+		[UHS_SDR104_BUS_SPEED] = "SDR104 ",
+		[UHS_DDR50_BUS_SPEED] = "DDR50 ",
+	};
+
 
 	dev_set_name(&card->dev, "%s:%04x", mmc_hostname(card->host), card->rca);
 
@@ -232,36 +314,55 @@ int mmc_add_card(struct mmc_card *card)
 		break;
 	case MMC_TYPE_SD:
 		type = "SD";
-		if (mmc_card_blockaddr(card))
-			type = "SDHC";
+		if (mmc_card_blockaddr(card)) {
+			if (mmc_card_ext_capacity(card))
+				type = "SDXC";
+			else
+				type = "SDHC";
+		}
 		break;
 	case MMC_TYPE_SDIO:
 		type = "SDIO";
 		break;
+	case MMC_TYPE_SD_COMBO:
+		type = "SD-combo";
+		if (mmc_card_blockaddr(card))
+			type = "SDHC-combo";
+		break;
 	default:
 		type = "?";
 		break;
 	}
 
+	if (mmc_card_uhs(card) &&
+		(card->sd_bus_speed < ARRAY_SIZE(uhs_speeds)))
+		uhs_bus_speed_mode = uhs_speeds[card->sd_bus_speed];
+
 	if (mmc_host_is_spi(card->host)) {
-		printk(KERN_INFO "%s: new %s%s card on SPI\n",
+		pr_info("%s: new %s%s%s card on SPI\n",
 			mmc_hostname(card->host),
-			mmc_card_highspeed(card) ? "high speed " : "",
+			mmc_card_hs(card) ? "high speed " : "",
+			mmc_card_ddr52(card) ? "DDR " : "",
 			type);
 	} else {
-		printk(KERN_INFO "%s: new %s%s card at address %04x\n",
+		pr_info("%s: new %s%s%s%s%s card at address %04x\n",
 			mmc_hostname(card->host),
-			mmc_card_highspeed(card) ? "high speed " : "",
-			type, card->rca);
+			mmc_card_uhs(card) ? "ultra high speed " :
+			(mmc_card_hs(card) ? "high speed " : ""),
+			mmc_card_hs400(card) ? "HS400 " :
+			(mmc_card_hs200(card) ? "HS200 " : ""),
+			mmc_card_ddr52(card) ? "DDR " : "",
+			uhs_bus_speed_mode, type, card->rca);
 	}
 
-	ret = device_add(&card->dev);
-	if (ret)
-		return ret;
-
 #ifdef CONFIG_DEBUG_FS
 	mmc_add_card_debugfs(card);
 #endif
+	mmc_init_context_info(card->host);
+
+	ret = device_add(&card->dev);
+	if (ret)
+		return ret;
 
 	mmc_card_set_present(card);
 
@@ -280,10 +381,10 @@ void mmc_remove_card(struct mmc_card *card)
 
 	if (mmc_card_present(card)) {
 		if (mmc_host_is_spi(card->host)) {
-			printk(KERN_INFO "%s: SPI card removed\n",
+			pr_info("%s: SPI card removed\n",
 				mmc_hostname(card->host));
 		} else {
-			printk(KERN_INFO "%s: card %04x removed\n",
+			pr_info("%s: card %04x removed\n",
 				mmc_hostname(card->host), card->rca);
 		}
 		device_del(&card->dev);
diff --git a/drivers/mmc/core/bus.h b/drivers/mmc/core/bus.h
index 18178766ab4..00a19710b6b 100644
--- a/drivers/mmc/core/bus.h
+++ b/drivers/mmc/core/bus.h
@@ -14,7 +14,7 @@
 #define MMC_DEV_ATTR(name, fmt, args...)					\
 static ssize_t mmc_##name##_show (struct device *dev, struct device_attribute *attr, char *buf)	\
 {										\
-	struct mmc_card *card = container_of(dev, struct mmc_card, dev);	\
+	struct mmc_card *card = mmc_dev_to_card(dev);				\
 	return sprintf(buf, fmt, args);						\
 }										\
 static DEVICE_ATTR(name, S_IRUGO, mmc_##name##_show, NULL)
diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
index fa073ab3fa3..7dc0c85fdb6 100644
--- a/drivers/mmc/core/core.c
+++ b/drivers/mmc/core/core.c
@@ -22,11 +22,19 @@
 #include <linux/scatterlist.h>
 #include <linux/log2.h>
 #include <linux/regulator/consumer.h>
+#include <linux/pm_runtime.h>
+#include <linux/pm_wakeup.h>
+#include <linux/suspend.h>
+#include <linux/fault-inject.h>
+#include <linux/random.h>
+#include <linux/slab.h>
+#include <linux/of.h>
 
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
 #include <linux/mmc/sd.h>
+#include <linux/mmc/slot-gpio.h>
 
 #include "core.h"
 #include "bus.h"
@@ -37,14 +45,24 @@
 #include "sd_ops.h"
 #include "sdio_ops.h"
 
+/* If the device is not responding */
+#define MMC_CORE_TIMEOUT_MS	(10 * 60 * 1000) /* 10 minute timeout */
+
+/*
+ * Background operations can take a long time, depending on the housekeeping
+ * operations the card has to perform.
+ */
+#define MMC_BKOPS_MAX_TIMEOUT	(4 * 60 * 1000) /* max time to wait in ms */
+
 static struct workqueue_struct *workqueue;
+static const unsigned freqs[] = { 400000, 300000, 200000, 100000 };
 
 /*
  * Enabling software CRCs on the data blocks can be a significant (30%)
  * performance cost, and for other reasons may not always be desired.
  * So we allow it it to be disabled.
  */
-int use_spi_crc = 1;
+bool use_spi_crc = 1;
 module_param(use_spi_crc, bool, 0);
 
 /*
@@ -64,6 +82,43 @@ static void mmc_flush_scheduled_work(void)
 	flush_workqueue(workqueue);
 }
 
+#ifdef CONFIG_FAIL_MMC_REQUEST
+
+/*
+ * Internal function. Inject random data errors.
+ * If mmc_data is NULL no errors are injected.
+ */
+static void mmc_should_fail_request(struct mmc_host *host,
+				    struct mmc_request *mrq)
+{
+	struct mmc_command *cmd = mrq->cmd;
+	struct mmc_data *data = mrq->data;
+	static const int data_errors[] = {
+		-ETIMEDOUT,
+		-EILSEQ,
+		-EIO,
+	};
+
+	if (!data)
+		return;
+
+	if (cmd->error || data->error ||
+	    !should_fail(&host->fail_mmc_request, data->blksz * data->blocks))
+		return;
+
+	data->error = data_errors[prandom_u32() % ARRAY_SIZE(data_errors)];
+	data->bytes_xfered = (prandom_u32() % (data->bytes_xfered >> 9)) << 9;
+}
+
+#else /* CONFIG_FAIL_MMC_REQUEST */
+
+static inline void mmc_should_fail_request(struct mmc_host *host,
+					   struct mmc_request *mrq)
+{
+}
+
+#endif /* CONFIG_FAIL_MMC_REQUEST */
+
 /**
  *	mmc_request_done - finish processing an MMC request
  *	@host: MMC host which completed request
@@ -82,14 +137,16 @@ void mmc_request_done(struct mmc_host *host, struct mmc_request *mrq)
 			cmd->retries = 0;
 	}
 
-	if (err && cmd->retries) {
-		pr_debug("%s: req failed (CMD%u): %d, retrying...\n",
-			mmc_hostname(host), cmd->opcode, err);
-
-		cmd->retries--;
-		cmd->error = 0;
-		host->ops->request(host, mrq);
+	if (err && cmd->retries && !mmc_card_removed(host->card)) {
+		/*
+		 * Request starter must handle retries - see
+		 * mmc_wait_for_req_done().
+		 */
+		if (mrq->done)
+			mrq->done(mrq);
 	} else {
+		mmc_should_fail_request(host, mrq);
+
 		led_trigger_event(host->led, LED_OFF);
 
 		pr_debug("%s: req done (CMD%u): %d: %08x %08x %08x %08x\n",
@@ -113,6 +170,8 @@ void mmc_request_done(struct mmc_host *host, struct mmc_request *mrq)
 
 		if (mrq->done)
 			mrq->done(mrq);
+
+		mmc_host_clk_release(host);
 	}
 }
 
@@ -126,6 +185,12 @@ mmc_start_request(struct mmc_host *host, struct mmc_request *mrq)
 	struct scatterlist *sg;
 #endif
 
+	if (mrq->sbc) {
+		pr_debug("<%s: starting CMD%u arg %08x flags %08x>\n",
+			 mmc_hostname(host), mrq->sbc->opcode,
+			 mrq->sbc->arg, mrq->sbc->flags);
+	}
+
 	pr_debug("%s: starting CMD%u arg %08x flags %08x\n",
 		 mmc_hostname(host), mrq->cmd->opcode,
 		 mrq->cmd->arg, mrq->cmd->flags);
@@ -147,8 +212,6 @@ mmc_start_request(struct mmc_host *host, struct mmc_request *mrq)
 
 	WARN_ON(!host->claimed);
 
-	led_trigger_event(host->led, LED_FULL);
-
 	mrq->cmd->error = 0;
 	mrq->cmd->mrq = mrq;
 	if (mrq->data) {
@@ -173,15 +236,334 @@ mmc_start_request(struct mmc_host *host, struct mmc_request *mrq)
 			mrq->stop->mrq = mrq;
 		}
 	}
+	mmc_host_clk_hold(host);
+	led_trigger_event(host->led, LED_FULL);
 	host->ops->request(host, mrq);
 }
 
+/**
+ *	mmc_start_bkops - start BKOPS for supported cards
+ *	@card: MMC card to start BKOPS
+ *	@form_exception: A flag to indicate if this function was
+ *			 called due to an exception raised by the card
+ *
+ *	Start background operations whenever requested.
+ *	When the urgent BKOPS bit is set in a R1 command response
+ *	then background operations should be started immediately.
+*/
+void mmc_start_bkops(struct mmc_card *card, bool from_exception)
+{
+	int err;
+	int timeout;
+	bool use_busy_signal;
+
+	BUG_ON(!card);
+
+	if (!card->ext_csd.bkops_en || mmc_card_doing_bkops(card))
+		return;
+
+	err = mmc_read_bkops_status(card);
+	if (err) {
+		pr_err("%s: Failed to read bkops status: %d\n",
+		       mmc_hostname(card->host), err);
+		return;
+	}
+
+	if (!card->ext_csd.raw_bkops_status)
+		return;
+
+	if (card->ext_csd.raw_bkops_status < EXT_CSD_BKOPS_LEVEL_2 &&
+	    from_exception)
+		return;
+
+	mmc_claim_host(card->host);
+	if (card->ext_csd.raw_bkops_status >= EXT_CSD_BKOPS_LEVEL_2) {
+		timeout = MMC_BKOPS_MAX_TIMEOUT;
+		use_busy_signal = true;
+	} else {
+		timeout = 0;
+		use_busy_signal = false;
+	}
+
+	err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			EXT_CSD_BKOPS_START, 1, timeout,
+			use_busy_signal, true, false);
+	if (err) {
+		pr_warn("%s: Error %d starting bkops\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	/*
+	 * For urgent bkops status (LEVEL_2 and more)
+	 * bkops executed synchronously, otherwise
+	 * the operation is in progress
+	 */
+	if (!use_busy_signal)
+		mmc_card_set_doing_bkops(card);
+out:
+	mmc_release_host(card->host);
+}
+EXPORT_SYMBOL(mmc_start_bkops);
+
+/*
+ * mmc_wait_data_done() - done callback for data request
+ * @mrq: done data request
+ *
+ * Wakes up mmc context, passed as a callback to host controller driver
+ */
+static void mmc_wait_data_done(struct mmc_request *mrq)
+{
+	mrq->host->context_info.is_done_rcv = true;
+	wake_up_interruptible(&mrq->host->context_info.wait);
+}
+
 static void mmc_wait_done(struct mmc_request *mrq)
 {
-	complete(mrq->done_data);
+	complete(&mrq->completion);
+}
+
+/*
+ *__mmc_start_data_req() - starts data request
+ * @host: MMC host to start the request
+ * @mrq: data request to start
+ *
+ * Sets the done callback to be called when request is completed by the card.
+ * Starts data mmc request execution
+ */
+static int __mmc_start_data_req(struct mmc_host *host, struct mmc_request *mrq)
+{
+	mrq->done = mmc_wait_data_done;
+	mrq->host = host;
+	if (mmc_card_removed(host->card)) {
+		mrq->cmd->error = -ENOMEDIUM;
+		mmc_wait_data_done(mrq);
+		return -ENOMEDIUM;
+	}
+	mmc_start_request(host, mrq);
+
+	return 0;
+}
+
+static int __mmc_start_req(struct mmc_host *host, struct mmc_request *mrq)
+{
+	init_completion(&mrq->completion);
+	mrq->done = mmc_wait_done;
+	if (mmc_card_removed(host->card)) {
+		mrq->cmd->error = -ENOMEDIUM;
+		complete(&mrq->completion);
+		return -ENOMEDIUM;
+	}
+	mmc_start_request(host, mrq);
+	return 0;
+}
+
+/*
+ * mmc_wait_for_data_req_done() - wait for request completed
+ * @host: MMC host to prepare the command.
+ * @mrq: MMC request to wait for
+ *
+ * Blocks MMC context till host controller will ack end of data request
+ * execution or new request notification arrives from the block layer.
+ * Handles command retries.
+ *
+ * Returns enum mmc_blk_status after checking errors.
+ */
+static int mmc_wait_for_data_req_done(struct mmc_host *host,
+				      struct mmc_request *mrq,
+				      struct mmc_async_req *next_req)
+{
+	struct mmc_command *cmd;
+	struct mmc_context_info *context_info = &host->context_info;
+	int err;
+	unsigned long flags;
+
+	while (1) {
+		wait_event_interruptible(context_info->wait,
+				(context_info->is_done_rcv ||
+				 context_info->is_new_req));
+		spin_lock_irqsave(&context_info->lock, flags);
+		context_info->is_waiting_last_req = false;
+		spin_unlock_irqrestore(&context_info->lock, flags);
+		if (context_info->is_done_rcv) {
+			context_info->is_done_rcv = false;
+			context_info->is_new_req = false;
+			cmd = mrq->cmd;
+
+			if (!cmd->error || !cmd->retries ||
+			    mmc_card_removed(host->card)) {
+				err = host->areq->err_check(host->card,
+							    host->areq);
+				break; /* return err */
+			} else {
+				pr_info("%s: req failed (CMD%u): %d, retrying...\n",
+					mmc_hostname(host),
+					cmd->opcode, cmd->error);
+				cmd->retries--;
+				cmd->error = 0;
+				host->ops->request(host, mrq);
+				continue; /* wait for done/new event again */
+			}
+		} else if (context_info->is_new_req) {
+			context_info->is_new_req = false;
+			if (!next_req) {
+				err = MMC_BLK_NEW_REQUEST;
+				break; /* return err */
+			}
+		}
+	}
+	return err;
+}
+
+static void mmc_wait_for_req_done(struct mmc_host *host,
+				  struct mmc_request *mrq)
+{
+	struct mmc_command *cmd;
+
+	while (1) {
+		wait_for_completion(&mrq->completion);
+
+		cmd = mrq->cmd;
+
+		/*
+		 * If host has timed out waiting for the sanitize
+		 * to complete, card might be still in programming state
+		 * so let's try to bring the card out of programming
+		 * state.
+		 */
+		if (cmd->sanitize_busy && cmd->error == -ETIMEDOUT) {
+			if (!mmc_interrupt_hpi(host->card)) {
+				pr_warning("%s: %s: Interrupted sanitize\n",
+					   mmc_hostname(host), __func__);
+				cmd->error = 0;
+				break;
+			} else {
+				pr_err("%s: %s: Failed to interrupt sanitize\n",
+				       mmc_hostname(host), __func__);
+			}
+		}
+		if (!cmd->error || !cmd->retries ||
+		    mmc_card_removed(host->card))
+			break;
+
+		pr_debug("%s: req failed (CMD%u): %d, retrying...\n",
+			 mmc_hostname(host), cmd->opcode, cmd->error);
+		cmd->retries--;
+		cmd->error = 0;
+		host->ops->request(host, mrq);
+	}
+}
+
+/**
+ *	mmc_pre_req - Prepare for a new request
+ *	@host: MMC host to prepare command
+ *	@mrq: MMC request to prepare for
+ *	@is_first_req: true if there is no previous started request
+ *                     that may run in parellel to this call, otherwise false
+ *
+ *	mmc_pre_req() is called in prior to mmc_start_req() to let
+ *	host prepare for the new request. Preparation of a request may be
+ *	performed while another request is running on the host.
+ */
+static void mmc_pre_req(struct mmc_host *host, struct mmc_request *mrq,
+		 bool is_first_req)
+{
+	if (host->ops->pre_req) {
+		mmc_host_clk_hold(host);
+		host->ops->pre_req(host, mrq, is_first_req);
+		mmc_host_clk_release(host);
+	}
+}
+
+/**
+ *	mmc_post_req - Post process a completed request
+ *	@host: MMC host to post process command
+ *	@mrq: MMC request to post process for
+ *	@err: Error, if non zero, clean up any resources made in pre_req
+ *
+ *	Let the host post process a completed request. Post processing of
+ *	a request may be performed while another reuqest is running.
+ */
+static void mmc_post_req(struct mmc_host *host, struct mmc_request *mrq,
+			 int err)
+{
+	if (host->ops->post_req) {
+		mmc_host_clk_hold(host);
+		host->ops->post_req(host, mrq, err);
+		mmc_host_clk_release(host);
+	}
 }
 
 /**
+ *	mmc_start_req - start a non-blocking request
+ *	@host: MMC host to start command
+ *	@areq: async request to start
+ *	@error: out parameter returns 0 for success, otherwise non zero
+ *
+ *	Start a new MMC custom command request for a host.
+ *	If there is on ongoing async request wait for completion
+ *	of that request and start the new one and return.
+ *	Does not wait for the new request to complete.
+ *
+ *      Returns the completed request, NULL in case of none completed.
+ *	Wait for the an ongoing request (previoulsy started) to complete and
+ *	return the completed request. If there is no ongoing request, NULL
+ *	is returned without waiting. NULL is not an error condition.
+ */
+struct mmc_async_req *mmc_start_req(struct mmc_host *host,
+				    struct mmc_async_req *areq, int *error)
+{
+	int err = 0;
+	int start_err = 0;
+	struct mmc_async_req *data = host->areq;
+
+	/* Prepare a new request */
+	if (areq)
+		mmc_pre_req(host, areq->mrq, !host->areq);
+
+	if (host->areq) {
+		err = mmc_wait_for_data_req_done(host, host->areq->mrq,	areq);
+		if (err == MMC_BLK_NEW_REQUEST) {
+			if (error)
+				*error = err;
+			/*
+			 * The previous request was not completed,
+			 * nothing to return
+			 */
+			return NULL;
+		}
+		/*
+		 * Check BKOPS urgency for each R1 response
+		 */
+		if (host->card && mmc_card_mmc(host->card) &&
+		    ((mmc_resp_type(host->areq->mrq->cmd) == MMC_RSP_R1) ||
+		     (mmc_resp_type(host->areq->mrq->cmd) == MMC_RSP_R1B)) &&
+		    (host->areq->mrq->cmd->resp[0] & R1_EXCEPTION_EVENT))
+			mmc_start_bkops(host->card, true);
+	}
+
+	if (!err && areq)
+		start_err = __mmc_start_data_req(host, areq->mrq);
+
+	if (host->areq)
+		mmc_post_req(host, host->areq->mrq, 0);
+
+	 /* Cancel a prepared request if it was not started. */
+	if ((err || start_err) && areq)
+		mmc_post_req(host, areq->mrq, -EINVAL);
+
+	if (err)
+		host->areq = NULL;
+	else
+		host->areq = areq;
+
+	if (error)
+		*error = err;
+	return data;
+}
+EXPORT_SYMBOL(mmc_start_req);
+
+/**
  *	mmc_wait_for_req - start a request and wait for completion
  *	@host: MMC host to start command
  *	@mrq: MMC request to start
@@ -192,17 +574,77 @@ static void mmc_wait_done(struct mmc_request *mrq)
  */
 void mmc_wait_for_req(struct mmc_host *host, struct mmc_request *mrq)
 {
-	DECLARE_COMPLETION_ONSTACK(complete);
+	__mmc_start_req(host, mrq);
+	mmc_wait_for_req_done(host, mrq);
+}
+EXPORT_SYMBOL(mmc_wait_for_req);
 
-	mrq->done_data = &complete;
-	mrq->done = mmc_wait_done;
+/**
+ *	mmc_interrupt_hpi - Issue for High priority Interrupt
+ *	@card: the MMC card associated with the HPI transfer
+ *
+ *	Issued High Priority Interrupt, and check for card status
+ *	until out-of prg-state.
+ */
+int mmc_interrupt_hpi(struct mmc_card *card)
+{
+	int err;
+	u32 status;
+	unsigned long prg_wait;
 
-	mmc_start_request(host, mrq);
+	BUG_ON(!card);
 
-	wait_for_completion(&complete);
-}
+	if (!card->ext_csd.hpi_en) {
+		pr_info("%s: HPI enable bit unset\n", mmc_hostname(card->host));
+		return 1;
+	}
 
-EXPORT_SYMBOL(mmc_wait_for_req);
+	mmc_claim_host(card->host);
+	err = mmc_send_status(card, &status);
+	if (err) {
+		pr_err("%s: Get card status fail\n", mmc_hostname(card->host));
+		goto out;
+	}
+
+	switch (R1_CURRENT_STATE(status)) {
+	case R1_STATE_IDLE:
+	case R1_STATE_READY:
+	case R1_STATE_STBY:
+	case R1_STATE_TRAN:
+		/*
+		 * In idle and transfer states, HPI is not needed and the caller
+		 * can issue the next intended command immediately
+		 */
+		goto out;
+	case R1_STATE_PRG:
+		break;
+	default:
+		/* In all other states, it's illegal to issue HPI */
+		pr_debug("%s: HPI cannot be sent. Card state=%d\n",
+			mmc_hostname(card->host), R1_CURRENT_STATE(status));
+		err = -EINVAL;
+		goto out;
+	}
+
+	err = mmc_send_hpi_cmd(card, &status);
+	if (err)
+		goto out;
+
+	prg_wait = jiffies + msecs_to_jiffies(card->ext_csd.out_of_int_time);
+	do {
+		err = mmc_send_status(card, &status);
+
+		if (!err && R1_CURRENT_STATE(status) == R1_STATE_TRAN)
+			break;
+		if (time_after(jiffies, prg_wait))
+			err = -ETIMEDOUT;
+	} while (!err);
+
+out:
+	mmc_release_host(card->host);
+	return err;
+}
+EXPORT_SYMBOL(mmc_interrupt_hpi);
 
 /**
  *	mmc_wait_for_cmd - start a command and wait for completion
@@ -216,12 +658,10 @@ EXPORT_SYMBOL(mmc_wait_for_req);
  */
 int mmc_wait_for_cmd(struct mmc_host *host, struct mmc_command *cmd, int retries)
 {
-	struct mmc_request mrq;
+	struct mmc_request mrq = {NULL};
 
 	WARN_ON(!host->claimed);
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-
 	memset(cmd->resp, 0, sizeof(cmd->resp));
 	cmd->retries = retries;
 
@@ -236,6 +676,64 @@ int mmc_wait_for_cmd(struct mmc_host *host, struct mmc_command *cmd, int retries
 EXPORT_SYMBOL(mmc_wait_for_cmd);
 
 /**
+ *	mmc_stop_bkops - stop ongoing BKOPS
+ *	@card: MMC card to check BKOPS
+ *
+ *	Send HPI command to stop ongoing background operations to
+ *	allow rapid servicing of foreground operations, e.g. read/
+ *	writes. Wait until the card comes out of the programming state
+ *	to avoid errors in servicing read/write requests.
+ */
+int mmc_stop_bkops(struct mmc_card *card)
+{
+	int err = 0;
+
+	BUG_ON(!card);
+	err = mmc_interrupt_hpi(card);
+
+	/*
+	 * If err is EINVAL, we can't issue an HPI.
+	 * It should complete the BKOPS.
+	 */
+	if (!err || (err == -EINVAL)) {
+		mmc_card_clr_doing_bkops(card);
+		err = 0;
+	}
+
+	return err;
+}
+EXPORT_SYMBOL(mmc_stop_bkops);
+
+int mmc_read_bkops_status(struct mmc_card *card)
+{
+	int err;
+	u8 *ext_csd;
+
+	/*
+	 * In future work, we should consider storing the entire ext_csd.
+	 */
+	ext_csd = kmalloc(512, GFP_KERNEL);
+	if (!ext_csd) {
+		pr_err("%s: could not allocate buffer to receive the ext_csd.\n",
+		       mmc_hostname(card->host));
+		return -ENOMEM;
+	}
+
+	mmc_claim_host(card->host);
+	err = mmc_send_ext_csd(card, ext_csd);
+	mmc_release_host(card->host);
+	if (err)
+		goto out;
+
+	card->ext_csd.raw_bkops_status = ext_csd[EXT_CSD_BKOPS_STATUS];
+	card->ext_csd.raw_exception_status = ext_csd[EXT_CSD_EXP_EVENTS_STATUS];
+out:
+	kfree(ext_csd);
+	return err;
+}
+EXPORT_SYMBOL(mmc_read_bkops_status);
+
+/**
  *	mmc_set_data_timeout - set the timeout for a data command
  *	@data: data phase for command
  *	@card: the MMC card associated with the data transfer
@@ -278,15 +776,20 @@ void mmc_set_data_timeout(struct mmc_data *data, const struct mmc_card *card)
 		unsigned int timeout_us, limit_us;
 
 		timeout_us = data->timeout_ns / 1000;
-		timeout_us += data->timeout_clks * 1000 /
-			(card->host->ios.clock / 1000);
+		if (mmc_host_clk_rate(card->host))
+			timeout_us += data->timeout_clks * 1000 /
+				(mmc_host_clk_rate(card->host) / 1000);
 
 		if (data->flags & MMC_DATA_WRITE)
 			/*
-			 * The limit is really 250 ms, but that is
-			 * insufficient for some crappy cards.
+			 * The MMC spec "It is strongly recommended
+			 * for hosts to implement more than 500ms
+			 * timeout value even if the card indicates
+			 * the 250ms maximum busy length."  Even the
+			 * previous value of 300ms is known to be
+			 * insufficient for some cards.
 			 */
-			limit_us = 300000;
+			limit_us = 3000000;
 		else
 			limit_us = 100000;
 
@@ -297,7 +800,23 @@ void mmc_set_data_timeout(struct mmc_data *data, const struct mmc_card *card)
 			data->timeout_ns = limit_us * 1000;
 			data->timeout_clks = 0;
 		}
+
+		/* assign limit value if invalid */
+		if (timeout_us == 0)
+			data->timeout_ns = limit_us * 1000;
+	}
+
+	/*
+	 * Some cards require longer data read timeout than indicated in CSD.
+	 * Address this by setting the read timeout to a "reasonably high"
+	 * value. For the cards tested, 300ms has proven enough. If necessary,
+	 * this value can be increased if other problematic cards require this.
+	 */
+	if (mmc_card_long_read_time(card) && data->flags & MMC_DATA_READ) {
+		data->timeout_ns = 300000000;
+		data->timeout_clks = 0;
 	}
+
 	/*
 	 * Some cards need very high timeouts if driven in SPI mode.
 	 * The worst observed timeout was 900ms after writing a
@@ -366,19 +885,23 @@ int __mmc_claim_host(struct mmc_host *host, atomic_t *abort)
 	while (1) {
 		set_current_state(TASK_UNINTERRUPTIBLE);
 		stop = abort ? atomic_read(abort) : 0;
-		if (stop || !host->claimed)
+		if (stop || !host->claimed || host->claimer == current)
 			break;
 		spin_unlock_irqrestore(&host->lock, flags);
 		schedule();
 		spin_lock_irqsave(&host->lock, flags);
 	}
 	set_current_state(TASK_RUNNING);
-	if (!stop)
+	if (!stop) {
 		host->claimed = 1;
-	else
+		host->claimer = current;
+		host->claim_cnt += 1;
+	} else
 		wake_up(&host->wq);
 	spin_unlock_irqrestore(&host->lock, flags);
 	remove_wait_queue(&host->wq, &wait);
+	if (host->ops->enable && !stop && host->claim_cnt == 1)
+		host->ops->enable(host);
 	return stop;
 }
 
@@ -397,14 +920,44 @@ void mmc_release_host(struct mmc_host *host)
 
 	WARN_ON(!host->claimed);
 
+	if (host->ops->disable && host->claim_cnt == 1)
+		host->ops->disable(host);
+
 	spin_lock_irqsave(&host->lock, flags);
-	host->claimed = 0;
-	spin_unlock_irqrestore(&host->lock, flags);
+	if (--host->claim_cnt) {
+		/* Release for nested claim */
+		spin_unlock_irqrestore(&host->lock, flags);
+	} else {
+		host->claimed = 0;
+		host->claimer = NULL;
+		spin_unlock_irqrestore(&host->lock, flags);
+		wake_up(&host->wq);
+	}
+}
+EXPORT_SYMBOL(mmc_release_host);
 
-	wake_up(&host->wq);
+/*
+ * This is a helper function, which fetches a runtime pm reference for the
+ * card device and also claims the host.
+ */
+void mmc_get_card(struct mmc_card *card)
+{
+	pm_runtime_get_sync(&card->dev);
+	mmc_claim_host(card->host);
 }
+EXPORT_SYMBOL(mmc_get_card);
 
-EXPORT_SYMBOL(mmc_release_host);
+/*
+ * This is a helper function, which releases the host and drops the runtime
+ * pm reference for the card device.
+ */
+void mmc_put_card(struct mmc_card *card)
+{
+	mmc_release_host(card->host);
+	pm_runtime_mark_last_busy(&card->dev);
+	pm_runtime_put_autosuspend(&card->dev);
+}
+EXPORT_SYMBOL(mmc_put_card);
 
 /*
  * Internal function that does the actual ios call to the host driver,
@@ -420,6 +973,8 @@ static inline void mmc_set_ios(struct mmc_host *host)
 		 ios->power_mode, ios->chip_select, ios->vdd,
 		 ios->bus_width, ios->timing);
 
+	if (ios->clock > 0)
+		mmc_set_ungated(host);
 	host->ops->set_ios(host, ios);
 }
 
@@ -428,15 +983,17 @@ static inline void mmc_set_ios(struct mmc_host *host)
  */
 void mmc_set_chip_select(struct mmc_host *host, int mode)
 {
+	mmc_host_clk_hold(host);
 	host->ios.chip_select = mode;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
  * Sets the host clock to the highest possible frequency that
  * is below "hz".
  */
-void mmc_set_clock(struct mmc_host *host, unsigned int hz)
+static void __mmc_set_clock(struct mmc_host *host, unsigned int hz)
 {
 	WARN_ON(hz < host->f_min);
 
@@ -447,13 +1004,77 @@ void mmc_set_clock(struct mmc_host *host, unsigned int hz)
 	mmc_set_ios(host);
 }
 
+void mmc_set_clock(struct mmc_host *host, unsigned int hz)
+{
+	mmc_host_clk_hold(host);
+	__mmc_set_clock(host, hz);
+	mmc_host_clk_release(host);
+}
+
+#ifdef CONFIG_MMC_CLKGATE
+/*
+ * This gates the clock by setting it to 0 Hz.
+ */
+void mmc_gate_clock(struct mmc_host *host)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->clk_lock, flags);
+	host->clk_old = host->ios.clock;
+	host->ios.clock = 0;
+	host->clk_gated = true;
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+	mmc_set_ios(host);
+}
+
+/*
+ * This restores the clock from gating by using the cached
+ * clock value.
+ */
+void mmc_ungate_clock(struct mmc_host *host)
+{
+	/*
+	 * We should previously have gated the clock, so the clock shall
+	 * be 0 here! The clock may however be 0 during initialization,
+	 * when some request operations are performed before setting
+	 * the frequency. When ungate is requested in that situation
+	 * we just ignore the call.
+	 */
+	if (host->clk_old) {
+		BUG_ON(host->ios.clock);
+		/* This call will also set host->clk_gated to false */
+		__mmc_set_clock(host, host->clk_old);
+	}
+}
+
+void mmc_set_ungated(struct mmc_host *host)
+{
+	unsigned long flags;
+
+	/*
+	 * We've been given a new frequency while the clock is gated,
+	 * so make sure we regard this as ungating it.
+	 */
+	spin_lock_irqsave(&host->clk_lock, flags);
+	host->clk_gated = false;
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+}
+
+#else
+void mmc_set_ungated(struct mmc_host *host)
+{
+}
+#endif
+
 /*
  * Change the bus mode (open drain/push-pull) of a host.
  */
 void mmc_set_bus_mode(struct mmc_host *host, unsigned int mode)
 {
+	mmc_host_clk_hold(host);
 	host->ios.bus_mode = mode;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
@@ -461,8 +1082,10 @@ void mmc_set_bus_mode(struct mmc_host *host, unsigned int mode)
  */
 void mmc_set_bus_width(struct mmc_host *host, unsigned int width)
 {
+	mmc_host_clk_hold(host);
 	host->ios.bus_width = width;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /**
@@ -539,6 +1162,49 @@ u32 mmc_vddrange_to_ocrmask(int vdd_min, int vdd_max)
 }
 EXPORT_SYMBOL(mmc_vddrange_to_ocrmask);
 
+#ifdef CONFIG_OF
+
+/**
+ * mmc_of_parse_voltage - return mask of supported voltages
+ * @np: The device node need to be parsed.
+ * @mask: mask of voltages available for MMC/SD/SDIO
+ *
+ * 1. Return zero on success.
+ * 2. Return negative errno: voltage-range is invalid.
+ */
+int mmc_of_parse_voltage(struct device_node *np, u32 *mask)
+{
+	const u32 *voltage_ranges;
+	int num_ranges, i;
+
+	voltage_ranges = of_get_property(np, "voltage-ranges", &num_ranges);
+	num_ranges = num_ranges / sizeof(*voltage_ranges) / 2;
+	if (!voltage_ranges || !num_ranges) {
+		pr_info("%s: voltage-ranges unspecified\n", np->full_name);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < num_ranges; i++) {
+		const int j = i * 2;
+		u32 ocr_mask;
+
+		ocr_mask = mmc_vddrange_to_ocrmask(
+				be32_to_cpu(voltage_ranges[j]),
+				be32_to_cpu(voltage_ranges[j + 1]));
+		if (!ocr_mask) {
+			pr_err("%s: voltage-range #%d is invalid\n",
+				np->full_name, i);
+			return -EINVAL;
+		}
+		*mask |= ocr_mask;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(mmc_of_parse_voltage);
+
+#endif /* CONFIG_OF */
+
 #ifdef CONFIG_REGULATOR
 
 /**
@@ -574,12 +1240,13 @@ int mmc_regulator_get_ocrmask(struct regulator *supply)
 
 	return result;
 }
-EXPORT_SYMBOL(mmc_regulator_get_ocrmask);
+EXPORT_SYMBOL_GPL(mmc_regulator_get_ocrmask);
 
 /**
  * mmc_regulator_set_ocr - set regulator to match host->ios voltage
- * @vdd_bit: zero for power off, else a bit number (host->ios.vdd)
+ * @mmc: the host to regulate
  * @supply: regulator to use
+ * @vdd_bit: zero for power off, else a bit number (host->ios.vdd)
  *
  * Returns zero on success, else negative errno.
  *
@@ -587,21 +1254,19 @@ EXPORT_SYMBOL(mmc_regulator_get_ocrmask);
  * a particular supply voltage.  This would normally be called from the
  * set_ios() method.
  */
-int mmc_regulator_set_ocr(struct regulator *supply, unsigned short vdd_bit)
+int mmc_regulator_set_ocr(struct mmc_host *mmc,
+			struct regulator *supply,
+			unsigned short vdd_bit)
 {
 	int			result = 0;
 	int			min_uV, max_uV;
-	int			enabled;
-
-	enabled = regulator_is_enabled(supply);
-	if (enabled < 0)
-		return enabled;
 
 	if (vdd_bit) {
 		int		tmp;
 		int		voltage;
 
-		/* REVISIT mmc_vddrange_to_ocrmask() may have set some
+		/*
+		 * REVISIT mmc_vddrange_to_ocrmask() may have set some
 		 * bits this regulator doesn't quite support ... don't
 		 * be too picky, most cards and regulators are OK with
 		 * a 0.1V range goof (it's a small error percentage).
@@ -615,10 +1280,15 @@ int mmc_regulator_set_ocr(struct regulator *supply, unsigned short vdd_bit)
 			max_uV = min_uV + 100 * 1000;
 		}
 
-		/* avoid needless changes to this voltage; the regulator
-		 * might not allow this operation
+		/*
+		 * If we're using a fixed/static regulator, don't call
+		 * regulator_set_voltage; it would fail.
 		 */
 		voltage = regulator_get_voltage(supply);
+
+		if (!regulator_can_change_voltage(supply))
+			min_uV = max_uV = voltage;
+
 		if (voltage < 0)
 			result = voltage;
 		else if (voltage < min_uV || voltage > max_uV)
@@ -626,17 +1296,55 @@ int mmc_regulator_set_ocr(struct regulator *supply, unsigned short vdd_bit)
 		else
 			result = 0;
 
-		if (result == 0 && !enabled)
+		if (result == 0 && !mmc->regulator_enabled) {
 			result = regulator_enable(supply);
-	} else if (enabled) {
+			if (!result)
+				mmc->regulator_enabled = true;
+		}
+	} else if (mmc->regulator_enabled) {
 		result = regulator_disable(supply);
+		if (result == 0)
+			mmc->regulator_enabled = false;
 	}
 
+	if (result)
+		dev_err(mmc_dev(mmc),
+			"could not set regulator OCR (%d)\n", result);
 	return result;
 }
-EXPORT_SYMBOL(mmc_regulator_set_ocr);
+EXPORT_SYMBOL_GPL(mmc_regulator_set_ocr);
 
-#endif
+#endif /* CONFIG_REGULATOR */
+
+int mmc_regulator_get_supply(struct mmc_host *mmc)
+{
+	struct device *dev = mmc_dev(mmc);
+	int ret;
+
+	mmc->supply.vmmc = devm_regulator_get_optional(dev, "vmmc");
+	mmc->supply.vqmmc = devm_regulator_get_optional(dev, "vqmmc");
+
+	if (IS_ERR(mmc->supply.vmmc)) {
+		if (PTR_ERR(mmc->supply.vmmc) == -EPROBE_DEFER)
+			return -EPROBE_DEFER;
+		dev_info(dev, "No vmmc regulator found\n");
+	} else {
+		ret = mmc_regulator_get_ocrmask(mmc->supply.vmmc);
+		if (ret > 0)
+			mmc->ocr_avail = ret;
+		else
+			dev_warn(dev, "Failed getting OCR mask: %d\n", ret);
+	}
+
+	if (IS_ERR(mmc->supply.vqmmc)) {
+		if (PTR_ERR(mmc->supply.vqmmc) == -EPROBE_DEFER)
+			return -EPROBE_DEFER;
+		dev_info(dev, "No vqmmc regulator found\n");
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mmc_regulator_get_supply);
 
 /*
  * Mask off any voltages we don't support and select
@@ -646,32 +1354,165 @@ u32 mmc_select_voltage(struct mmc_host *host, u32 ocr)
 {
 	int bit;
 
-	ocr &= host->ocr_avail;
+	/*
+	 * Sanity check the voltages that the card claims to
+	 * support.
+	 */
+	if (ocr & 0x7F) {
+		dev_warn(mmc_dev(host),
+		"card claims to support voltages below defined range\n");
+		ocr &= ~0x7F;
+	}
 
-	bit = ffs(ocr);
-	if (bit) {
-		bit -= 1;
+	ocr &= host->ocr_avail;
+	if (!ocr) {
+		dev_warn(mmc_dev(host), "no support for card's volts\n");
+		return 0;
+	}
 
+	if (host->caps2 & MMC_CAP2_FULL_PWR_CYCLE) {
+		bit = ffs(ocr) - 1;
 		ocr &= 3 << bit;
-
-		host->ios.vdd = bit;
-		mmc_set_ios(host);
+		mmc_power_cycle(host, ocr);
 	} else {
-		pr_warning("%s: host doesn't support card's voltages\n",
-				mmc_hostname(host));
-		ocr = 0;
+		bit = fls(ocr) - 1;
+		ocr &= 3 << bit;
+		if (bit != host->ios.vdd)
+			dev_warn(mmc_dev(host), "exceeding card's volts\n");
 	}
 
 	return ocr;
 }
 
+int __mmc_set_signal_voltage(struct mmc_host *host, int signal_voltage)
+{
+	int err = 0;
+	int old_signal_voltage = host->ios.signal_voltage;
+
+	host->ios.signal_voltage = signal_voltage;
+	if (host->ops->start_signal_voltage_switch) {
+		mmc_host_clk_hold(host);
+		err = host->ops->start_signal_voltage_switch(host, &host->ios);
+		mmc_host_clk_release(host);
+	}
+
+	if (err)
+		host->ios.signal_voltage = old_signal_voltage;
+
+	return err;
+
+}
+
+int mmc_set_signal_voltage(struct mmc_host *host, int signal_voltage, u32 ocr)
+{
+	struct mmc_command cmd = {0};
+	int err = 0;
+	u32 clock;
+
+	BUG_ON(!host);
+
+	/*
+	 * Send CMD11 only if the request is to switch the card to
+	 * 1.8V signalling.
+	 */
+	if (signal_voltage == MMC_SIGNAL_VOLTAGE_330)
+		return __mmc_set_signal_voltage(host, signal_voltage);
+
+	/*
+	 * If we cannot switch voltages, return failure so the caller
+	 * can continue without UHS mode
+	 */
+	if (!host->ops->start_signal_voltage_switch)
+		return -EPERM;
+	if (!host->ops->card_busy)
+		pr_warning("%s: cannot verify signal voltage switch\n",
+				mmc_hostname(host));
+
+	cmd.opcode = SD_SWITCH_VOLTAGE;
+	cmd.arg = 0;
+	cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
+
+	err = mmc_wait_for_cmd(host, &cmd, 0);
+	if (err)
+		return err;
+
+	if (!mmc_host_is_spi(host) && (cmd.resp[0] & R1_ERROR))
+		return -EIO;
+
+	mmc_host_clk_hold(host);
+	/*
+	 * The card should drive cmd and dat[0:3] low immediately
+	 * after the response of cmd11, but wait 1 ms to be sure
+	 */
+	mmc_delay(1);
+	if (host->ops->card_busy && !host->ops->card_busy(host)) {
+		err = -EAGAIN;
+		goto power_cycle;
+	}
+	/*
+	 * During a signal voltage level switch, the clock must be gated
+	 * for 5 ms according to the SD spec
+	 */
+	clock = host->ios.clock;
+	host->ios.clock = 0;
+	mmc_set_ios(host);
+
+	if (__mmc_set_signal_voltage(host, signal_voltage)) {
+		/*
+		 * Voltages may not have been switched, but we've already
+		 * sent CMD11, so a power cycle is required anyway
+		 */
+		err = -EAGAIN;
+		goto power_cycle;
+	}
+
+	/* Keep clock gated for at least 5 ms */
+	mmc_delay(5);
+	host->ios.clock = clock;
+	mmc_set_ios(host);
+
+	/* Wait for at least 1 ms according to spec */
+	mmc_delay(1);
+
+	/*
+	 * Failure to switch is indicated by the card holding
+	 * dat[0:3] low
+	 */
+	if (host->ops->card_busy && host->ops->card_busy(host))
+		err = -EAGAIN;
+
+power_cycle:
+	if (err) {
+		pr_debug("%s: Signal voltage switch failed, "
+			"power cycling card\n", mmc_hostname(host));
+		mmc_power_cycle(host, ocr);
+	}
+
+	mmc_host_clk_release(host);
+
+	return err;
+}
+
 /*
  * Select timing parameters for host.
  */
 void mmc_set_timing(struct mmc_host *host, unsigned int timing)
 {
+	mmc_host_clk_hold(host);
 	host->ios.timing = timing;
 	mmc_set_ios(host);
+	mmc_host_clk_release(host);
+}
+
+/*
+ * Select appropriate driver type for host.
+ */
+void mmc_set_driver_type(struct mmc_host *host, unsigned int drv_type)
+{
+	mmc_host_clk_hold(host);
+	host->ios.drv_type = drv_type;
+	mmc_set_ios(host);
+	mmc_host_clk_release(host);
 }
 
 /*
@@ -685,30 +1526,40 @@ void mmc_set_timing(struct mmc_host *host, unsigned int timing)
  * If a host does all the power sequencing itself, ignore the
  * initial MMC_POWER_UP stage.
  */
-static void mmc_power_up(struct mmc_host *host)
+void mmc_power_up(struct mmc_host *host, u32 ocr)
 {
-	int bit = fls(host->ocr_avail) - 1;
+	if (host->ios.power_mode == MMC_POWER_ON)
+		return;
 
-	host->ios.vdd = bit;
-	if (mmc_host_is_spi(host)) {
+	mmc_host_clk_hold(host);
+
+	host->ios.vdd = fls(ocr) - 1;
+	if (mmc_host_is_spi(host))
 		host->ios.chip_select = MMC_CS_HIGH;
-		host->ios.bus_mode = MMC_BUSMODE_PUSHPULL;
-	} else {
+	else
 		host->ios.chip_select = MMC_CS_DONTCARE;
-		host->ios.bus_mode = MMC_BUSMODE_OPENDRAIN;
-	}
+	host->ios.bus_mode = MMC_BUSMODE_PUSHPULL;
 	host->ios.power_mode = MMC_POWER_UP;
 	host->ios.bus_width = MMC_BUS_WIDTH_1;
 	host->ios.timing = MMC_TIMING_LEGACY;
 	mmc_set_ios(host);
 
+	/* Try to set signal voltage to 3.3V but fall back to 1.8v or 1.2v */
+	if (__mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_330) == 0)
+		dev_dbg(mmc_dev(host), "Initial signal voltage of 3.3v\n");
+	else if (__mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_180) == 0)
+		dev_dbg(mmc_dev(host), "Initial signal voltage of 1.8v\n");
+	else if (__mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_120) == 0)
+		dev_dbg(mmc_dev(host), "Initial signal voltage of 1.2v\n");
+
 	/*
 	 * This delay should be sufficient to allow the power supply
 	 * to reach the minimum voltage.
 	 */
-	mmc_delay(2);
+	mmc_delay(10);
+
+	host->ios.clock = host->f_init;
 
-	host->ios.clock = host->f_min;
 	host->ios.power_mode = MMC_POWER_ON;
 	mmc_set_ios(host);
 
@@ -716,13 +1567,21 @@ static void mmc_power_up(struct mmc_host *host)
 	 * This delay must be at least 74 clock sizes, or 1 ms, or the
 	 * time required to reach a stable voltage.
 	 */
-	mmc_delay(2);
+	mmc_delay(10);
+
+	mmc_host_clk_release(host);
 }
 
-static void mmc_power_off(struct mmc_host *host)
+void mmc_power_off(struct mmc_host *host)
 {
+	if (host->ios.power_mode == MMC_POWER_OFF)
+		return;
+
+	mmc_host_clk_hold(host);
+
 	host->ios.clock = 0;
 	host->ios.vdd = 0;
+
 	if (!mmc_host_is_spi(host)) {
 		host->ios.bus_mode = MMC_BUSMODE_OPENDRAIN;
 		host->ios.chip_select = MMC_CS_DONTCARE;
@@ -731,6 +1590,23 @@ static void mmc_power_off(struct mmc_host *host)
 	host->ios.bus_width = MMC_BUS_WIDTH_1;
 	host->ios.timing = MMC_TIMING_LEGACY;
 	mmc_set_ios(host);
+
+	/*
+	 * Some configurations, such as the 802.11 SDIO card in the OLPC
+	 * XO-1.5, require a short delay after poweroff before the card
+	 * can be successfully turned on again.
+	 */
+	mmc_delay(1);
+
+	mmc_host_clk_release(host);
+}
+
+void mmc_power_cycle(struct mmc_host *host, u32 ocr)
+{
+	mmc_power_off(host);
+	/* Wait at least 1 ms according to SD spec */
+	mmc_delay(1);
+	mmc_power_up(host, ocr);
 }
 
 /*
@@ -798,8 +1674,7 @@ void mmc_attach_bus(struct mmc_host *host, const struct mmc_bus_ops *ops)
 }
 
 /*
- * Remove the current bus handler from a host. Assumes that there are
- * no interesting cards left, so the bus is powered down.
+ * Remove the current bus handler from a host.
  */
 void mmc_detach_bus(struct mmc_host *host)
 {
@@ -816,11 +1691,31 @@ void mmc_detach_bus(struct mmc_host *host)
 
 	spin_unlock_irqrestore(&host->lock, flags);
 
-	mmc_power_off(host);
-
 	mmc_bus_put(host);
 }
 
+static void _mmc_detect_change(struct mmc_host *host, unsigned long delay,
+				bool cd_irq)
+{
+#ifdef CONFIG_MMC_DEBUG
+	unsigned long flags;
+	spin_lock_irqsave(&host->lock, flags);
+	WARN_ON(host->removed);
+	spin_unlock_irqrestore(&host->lock, flags);
+#endif
+
+	/*
+	 * If the device is configured as wakeup, we prevent a new sleep for
+	 * 5 s to give provision for user space to consume the event.
+	 */
+	if (cd_irq && !(host->caps & MMC_CAP_NEEDS_POLL) &&
+		device_can_wakeup(mmc_dev(host)))
+		pm_wakeup_event(mmc_dev(host), 5000);
+
+	host->detect_change = 1;
+	mmc_schedule_delayed_work(&host->detect, delay);
+}
+
 /**
  *	mmc_detect_change - process change of state on a MMC socket
  *	@host: host which changed state.
@@ -833,92 +1728,772 @@ void mmc_detach_bus(struct mmc_host *host)
  */
 void mmc_detect_change(struct mmc_host *host, unsigned long delay)
 {
-#ifdef CONFIG_MMC_DEBUG
-	unsigned long flags;
-	spin_lock_irqsave(&host->lock, flags);
-	WARN_ON(host->removed);
-	spin_unlock_irqrestore(&host->lock, flags);
-#endif
-
-	mmc_schedule_delayed_work(&host->detect, delay);
+	_mmc_detect_change(host, delay, true);
 }
-
 EXPORT_SYMBOL(mmc_detect_change);
 
+void mmc_init_erase(struct mmc_card *card)
+{
+	unsigned int sz;
 
-void mmc_rescan(struct work_struct *work)
+	if (is_power_of_2(card->erase_size))
+		card->erase_shift = ffs(card->erase_size) - 1;
+	else
+		card->erase_shift = 0;
+
+	/*
+	 * It is possible to erase an arbitrarily large area of an SD or MMC
+	 * card.  That is not desirable because it can take a long time
+	 * (minutes) potentially delaying more important I/O, and also the
+	 * timeout calculations become increasingly hugely over-estimated.
+	 * Consequently, 'pref_erase' is defined as a guide to limit erases
+	 * to that size and alignment.
+	 *
+	 * For SD cards that define Allocation Unit size, limit erases to one
+	 * Allocation Unit at a time.  For MMC cards that define High Capacity
+	 * Erase Size, whether it is switched on or not, limit to that size.
+	 * Otherwise just have a stab at a good value.  For modern cards it
+	 * will end up being 4MiB.  Note that if the value is too small, it
+	 * can end up taking longer to erase.
+	 */
+	if (mmc_card_sd(card) && card->ssr.au) {
+		card->pref_erase = card->ssr.au;
+		card->erase_shift = ffs(card->ssr.au) - 1;
+	} else if (card->ext_csd.hc_erase_size) {
+		card->pref_erase = card->ext_csd.hc_erase_size;
+	} else {
+		sz = (card->csd.capacity << (card->csd.read_blkbits - 9)) >> 11;
+		if (sz < 128)
+			card->pref_erase = 512 * 1024 / 512;
+		else if (sz < 512)
+			card->pref_erase = 1024 * 1024 / 512;
+		else if (sz < 1024)
+			card->pref_erase = 2 * 1024 * 1024 / 512;
+		else
+			card->pref_erase = 4 * 1024 * 1024 / 512;
+		if (card->pref_erase < card->erase_size)
+			card->pref_erase = card->erase_size;
+		else {
+			sz = card->pref_erase % card->erase_size;
+			if (sz)
+				card->pref_erase += card->erase_size - sz;
+		}
+	}
+}
+
+static unsigned int mmc_mmc_erase_timeout(struct mmc_card *card,
+				          unsigned int arg, unsigned int qty)
 {
-	struct mmc_host *host =
-		container_of(work, struct mmc_host, detect.work);
-	u32 ocr;
-	int err;
+	unsigned int erase_timeout;
+
+	if (arg == MMC_DISCARD_ARG ||
+	    (arg == MMC_TRIM_ARG && card->ext_csd.rev >= 6)) {
+		erase_timeout = card->ext_csd.trim_timeout;
+	} else if (card->ext_csd.erase_group_def & 1) {
+		/* High Capacity Erase Group Size uses HC timeouts */
+		if (arg == MMC_TRIM_ARG)
+			erase_timeout = card->ext_csd.trim_timeout;
+		else
+			erase_timeout = card->ext_csd.hc_erase_timeout;
+	} else {
+		/* CSD Erase Group Size uses write timeout */
+		unsigned int mult = (10 << card->csd.r2w_factor);
+		unsigned int timeout_clks = card->csd.tacc_clks * mult;
+		unsigned int timeout_us;
+
+		/* Avoid overflow: e.g. tacc_ns=80000000 mult=1280 */
+		if (card->csd.tacc_ns < 1000000)
+			timeout_us = (card->csd.tacc_ns * mult) / 1000;
+		else
+			timeout_us = (card->csd.tacc_ns / 1000) * mult;
 
-	mmc_bus_get(host);
+		/*
+		 * ios.clock is only a target.  The real clock rate might be
+		 * less but not that much less, so fudge it by multiplying by 2.
+		 */
+		timeout_clks <<= 1;
+		timeout_us += (timeout_clks * 1000) /
+			      (mmc_host_clk_rate(card->host) / 1000);
+
+		erase_timeout = timeout_us / 1000;
 
-	if (host->bus_ops == NULL) {
 		/*
-		 * Only we can add a new handler, so it's safe to
-		 * release the lock here.
+		 * Theoretically, the calculation could underflow so round up
+		 * to 1ms in that case.
 		 */
-		mmc_bus_put(host);
+		if (!erase_timeout)
+			erase_timeout = 1;
+	}
 
-		if (host->ops->get_cd && host->ops->get_cd(host) == 0)
-			goto out;
+	/* Multiplier for secure operations */
+	if (arg & MMC_SECURE_ARGS) {
+		if (arg == MMC_SECURE_ERASE_ARG)
+			erase_timeout *= card->ext_csd.sec_erase_mult;
+		else
+			erase_timeout *= card->ext_csd.sec_trim_mult;
+	}
 
-		mmc_claim_host(host);
+	erase_timeout *= qty;
+
+	/*
+	 * Ensure at least a 1 second timeout for SPI as per
+	 * 'mmc_set_data_timeout()'
+	 */
+	if (mmc_host_is_spi(card->host) && erase_timeout < 1000)
+		erase_timeout = 1000;
 
-		mmc_power_up(host);
-		mmc_go_idle(host);
+	return erase_timeout;
+}
 
-		mmc_send_if_cond(host, host->ocr_avail);
+static unsigned int mmc_sd_erase_timeout(struct mmc_card *card,
+					 unsigned int arg,
+					 unsigned int qty)
+{
+	unsigned int erase_timeout;
 
+	if (card->ssr.erase_timeout) {
+		/* Erase timeout specified in SD Status Register (SSR) */
+		erase_timeout = card->ssr.erase_timeout * qty +
+				card->ssr.erase_offset;
+	} else {
 		/*
-		 * First we search for SDIO...
+		 * Erase timeout not specified in SD Status Register (SSR) so
+		 * use 250ms per write block.
 		 */
-		err = mmc_send_io_op_cond(host, 0, &ocr);
-		if (!err) {
-			if (mmc_attach_sdio(host, ocr))
-				mmc_power_off(host);
+		erase_timeout = 250 * qty;
+	}
+
+	/* Must not be less than 1 second */
+	if (erase_timeout < 1000)
+		erase_timeout = 1000;
+
+	return erase_timeout;
+}
+
+static unsigned int mmc_erase_timeout(struct mmc_card *card,
+				      unsigned int arg,
+				      unsigned int qty)
+{
+	if (mmc_card_sd(card))
+		return mmc_sd_erase_timeout(card, arg, qty);
+	else
+		return mmc_mmc_erase_timeout(card, arg, qty);
+}
+
+static int mmc_do_erase(struct mmc_card *card, unsigned int from,
+			unsigned int to, unsigned int arg)
+{
+	struct mmc_command cmd = {0};
+	unsigned int qty = 0;
+	unsigned long timeout;
+	int err;
+
+	/*
+	 * qty is used to calculate the erase timeout which depends on how many
+	 * erase groups (or allocation units in SD terminology) are affected.
+	 * We count erasing part of an erase group as one erase group.
+	 * For SD, the allocation units are always a power of 2.  For MMC, the
+	 * erase group size is almost certainly also power of 2, but it does not
+	 * seem to insist on that in the JEDEC standard, so we fall back to
+	 * division in that case.  SD may not specify an allocation unit size,
+	 * in which case the timeout is based on the number of write blocks.
+	 *
+	 * Note that the timeout for secure trim 2 will only be correct if the
+	 * number of erase groups specified is the same as the total of all
+	 * preceding secure trim 1 commands.  Since the power may have been
+	 * lost since the secure trim 1 commands occurred, it is generally
+	 * impossible to calculate the secure trim 2 timeout correctly.
+	 */
+	if (card->erase_shift)
+		qty += ((to >> card->erase_shift) -
+			(from >> card->erase_shift)) + 1;
+	else if (mmc_card_sd(card))
+		qty += to - from + 1;
+	else
+		qty += ((to / card->erase_size) -
+			(from / card->erase_size)) + 1;
+
+	if (!mmc_card_blockaddr(card)) {
+		from <<= 9;
+		to <<= 9;
+	}
+
+	if (mmc_card_sd(card))
+		cmd.opcode = SD_ERASE_WR_BLK_START;
+	else
+		cmd.opcode = MMC_ERASE_GROUP_START;
+	cmd.arg = from;
+	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
+	err = mmc_wait_for_cmd(card->host, &cmd, 0);
+	if (err) {
+		pr_err("mmc_erase: group start error %d, "
+		       "status %#x\n", err, cmd.resp[0]);
+		err = -EIO;
+		goto out;
+	}
+
+	memset(&cmd, 0, sizeof(struct mmc_command));
+	if (mmc_card_sd(card))
+		cmd.opcode = SD_ERASE_WR_BLK_END;
+	else
+		cmd.opcode = MMC_ERASE_GROUP_END;
+	cmd.arg = to;
+	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
+	err = mmc_wait_for_cmd(card->host, &cmd, 0);
+	if (err) {
+		pr_err("mmc_erase: group end error %d, status %#x\n",
+		       err, cmd.resp[0]);
+		err = -EIO;
+		goto out;
+	}
+
+	memset(&cmd, 0, sizeof(struct mmc_command));
+	cmd.opcode = MMC_ERASE;
+	cmd.arg = arg;
+	cmd.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
+	cmd.busy_timeout = mmc_erase_timeout(card, arg, qty);
+	err = mmc_wait_for_cmd(card->host, &cmd, 0);
+	if (err) {
+		pr_err("mmc_erase: erase error %d, status %#x\n",
+		       err, cmd.resp[0]);
+		err = -EIO;
+		goto out;
+	}
+
+	if (mmc_host_is_spi(card->host))
+		goto out;
+
+	timeout = jiffies + msecs_to_jiffies(MMC_CORE_TIMEOUT_MS);
+	do {
+		memset(&cmd, 0, sizeof(struct mmc_command));
+		cmd.opcode = MMC_SEND_STATUS;
+		cmd.arg = card->rca << 16;
+		cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
+		/* Do not retry else we can't see errors */
+		err = mmc_wait_for_cmd(card->host, &cmd, 0);
+		if (err || (cmd.resp[0] & 0xFDF92000)) {
+			pr_err("error %d requesting status %#x\n",
+				err, cmd.resp[0]);
+			err = -EIO;
 			goto out;
 		}
 
-		/*
-		 * ...then normal SD...
+		/* Timeout if the device never becomes ready for data and
+		 * never leaves the program state.
 		 */
-		err = mmc_send_app_op_cond(host, 0, &ocr);
-		if (!err) {
-			if (mmc_attach_sd(host, ocr))
-				mmc_power_off(host);
+		if (time_after(jiffies, timeout)) {
+			pr_err("%s: Card stuck in programming state! %s\n",
+				mmc_hostname(card->host), __func__);
+			err =  -EIO;
 			goto out;
 		}
 
-		/*
-		 * ...and finally MMC.
-		 */
-		err = mmc_send_op_cond(host, 0, &ocr);
+	} while (!(cmd.resp[0] & R1_READY_FOR_DATA) ||
+		 (R1_CURRENT_STATE(cmd.resp[0]) == R1_STATE_PRG));
+out:
+	return err;
+}
+
+/**
+ * mmc_erase - erase sectors.
+ * @card: card to erase
+ * @from: first sector to erase
+ * @nr: number of sectors to erase
+ * @arg: erase command argument (SD supports only %MMC_ERASE_ARG)
+ *
+ * Caller must claim host before calling this function.
+ */
+int mmc_erase(struct mmc_card *card, unsigned int from, unsigned int nr,
+	      unsigned int arg)
+{
+	unsigned int rem, to = from + nr;
+
+	if (!(card->host->caps & MMC_CAP_ERASE) ||
+	    !(card->csd.cmdclass & CCC_ERASE))
+		return -EOPNOTSUPP;
+
+	if (!card->erase_size)
+		return -EOPNOTSUPP;
+
+	if (mmc_card_sd(card) && arg != MMC_ERASE_ARG)
+		return -EOPNOTSUPP;
+
+	if ((arg & MMC_SECURE_ARGS) &&
+	    !(card->ext_csd.sec_feature_support & EXT_CSD_SEC_ER_EN))
+		return -EOPNOTSUPP;
+
+	if ((arg & MMC_TRIM_ARGS) &&
+	    !(card->ext_csd.sec_feature_support & EXT_CSD_SEC_GB_CL_EN))
+		return -EOPNOTSUPP;
+
+	if (arg == MMC_SECURE_ERASE_ARG) {
+		if (from % card->erase_size || nr % card->erase_size)
+			return -EINVAL;
+	}
+
+	if (arg == MMC_ERASE_ARG) {
+		rem = from % card->erase_size;
+		if (rem) {
+			rem = card->erase_size - rem;
+			from += rem;
+			if (nr > rem)
+				nr -= rem;
+			else
+				return 0;
+		}
+		rem = nr % card->erase_size;
+		if (rem)
+			nr -= rem;
+	}
+
+	if (nr == 0)
+		return 0;
+
+	to = from + nr;
+
+	if (to <= from)
+		return -EINVAL;
+
+	/* 'from' and 'to' are inclusive */
+	to -= 1;
+
+	return mmc_do_erase(card, from, to, arg);
+}
+EXPORT_SYMBOL(mmc_erase);
+
+int mmc_can_erase(struct mmc_card *card)
+{
+	if ((card->host->caps & MMC_CAP_ERASE) &&
+	    (card->csd.cmdclass & CCC_ERASE) && card->erase_size)
+		return 1;
+	return 0;
+}
+EXPORT_SYMBOL(mmc_can_erase);
+
+int mmc_can_trim(struct mmc_card *card)
+{
+	if (card->ext_csd.sec_feature_support & EXT_CSD_SEC_GB_CL_EN)
+		return 1;
+	return 0;
+}
+EXPORT_SYMBOL(mmc_can_trim);
+
+int mmc_can_discard(struct mmc_card *card)
+{
+	/*
+	 * As there's no way to detect the discard support bit at v4.5
+	 * use the s/w feature support filed.
+	 */
+	if (card->ext_csd.feature_support & MMC_DISCARD_FEATURE)
+		return 1;
+	return 0;
+}
+EXPORT_SYMBOL(mmc_can_discard);
+
+int mmc_can_sanitize(struct mmc_card *card)
+{
+	if (!mmc_can_trim(card) && !mmc_can_erase(card))
+		return 0;
+	if (card->ext_csd.sec_feature_support & EXT_CSD_SEC_SANITIZE)
+		return 1;
+	return 0;
+}
+EXPORT_SYMBOL(mmc_can_sanitize);
+
+int mmc_can_secure_erase_trim(struct mmc_card *card)
+{
+	if (card->ext_csd.sec_feature_support & EXT_CSD_SEC_ER_EN)
+		return 1;
+	return 0;
+}
+EXPORT_SYMBOL(mmc_can_secure_erase_trim);
+
+int mmc_erase_group_aligned(struct mmc_card *card, unsigned int from,
+			    unsigned int nr)
+{
+	if (!card->erase_size)
+		return 0;
+	if (from % card->erase_size || nr % card->erase_size)
+		return 0;
+	return 1;
+}
+EXPORT_SYMBOL(mmc_erase_group_aligned);
+
+static unsigned int mmc_do_calc_max_discard(struct mmc_card *card,
+					    unsigned int arg)
+{
+	struct mmc_host *host = card->host;
+	unsigned int max_discard, x, y, qty = 0, max_qty, timeout;
+	unsigned int last_timeout = 0;
+
+	if (card->erase_shift)
+		max_qty = UINT_MAX >> card->erase_shift;
+	else if (mmc_card_sd(card))
+		max_qty = UINT_MAX;
+	else
+		max_qty = UINT_MAX / card->erase_size;
+
+	/* Find the largest qty with an OK timeout */
+	do {
+		y = 0;
+		for (x = 1; x && x <= max_qty && max_qty - x >= qty; x <<= 1) {
+			timeout = mmc_erase_timeout(card, arg, qty + x);
+			if (timeout > host->max_busy_timeout)
+				break;
+			if (timeout < last_timeout)
+				break;
+			last_timeout = timeout;
+			y = x;
+		}
+		qty += y;
+	} while (y);
+
+	if (!qty)
+		return 0;
+
+	if (qty == 1)
+		return 1;
+
+	/* Convert qty to sectors */
+	if (card->erase_shift)
+		max_discard = --qty << card->erase_shift;
+	else if (mmc_card_sd(card))
+		max_discard = qty;
+	else
+		max_discard = --qty * card->erase_size;
+
+	return max_discard;
+}
+
+unsigned int mmc_calc_max_discard(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	unsigned int max_discard, max_trim;
+
+	if (!host->max_busy_timeout)
+		return UINT_MAX;
+
+	/*
+	 * Without erase_group_def set, MMC erase timeout depends on clock
+	 * frequence which can change.  In that case, the best choice is
+	 * just the preferred erase size.
+	 */
+	if (mmc_card_mmc(card) && !(card->ext_csd.erase_group_def & 1))
+		return card->pref_erase;
+
+	max_discard = mmc_do_calc_max_discard(card, MMC_ERASE_ARG);
+	if (mmc_can_trim(card)) {
+		max_trim = mmc_do_calc_max_discard(card, MMC_TRIM_ARG);
+		if (max_trim < max_discard)
+			max_discard = max_trim;
+	} else if (max_discard < card->erase_size) {
+		max_discard = 0;
+	}
+	pr_debug("%s: calculated max. discard sectors %u for timeout %u ms\n",
+		 mmc_hostname(host), max_discard, host->max_busy_timeout);
+	return max_discard;
+}
+EXPORT_SYMBOL(mmc_calc_max_discard);
+
+int mmc_set_blocklen(struct mmc_card *card, unsigned int blocklen)
+{
+	struct mmc_command cmd = {0};
+
+	if (mmc_card_blockaddr(card) || mmc_card_ddr52(card))
+		return 0;
+
+	cmd.opcode = MMC_SET_BLOCKLEN;
+	cmd.arg = blocklen;
+	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
+	return mmc_wait_for_cmd(card->host, &cmd, 5);
+}
+EXPORT_SYMBOL(mmc_set_blocklen);
+
+int mmc_set_blockcount(struct mmc_card *card, unsigned int blockcount,
+			bool is_rel_write)
+{
+	struct mmc_command cmd = {0};
+
+	cmd.opcode = MMC_SET_BLOCK_COUNT;
+	cmd.arg = blockcount & 0x0000FFFF;
+	if (is_rel_write)
+		cmd.arg |= 1 << 31;
+	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
+	return mmc_wait_for_cmd(card->host, &cmd, 5);
+}
+EXPORT_SYMBOL(mmc_set_blockcount);
+
+static void mmc_hw_reset_for_init(struct mmc_host *host)
+{
+	if (!(host->caps & MMC_CAP_HW_RESET) || !host->ops->hw_reset)
+		return;
+	mmc_host_clk_hold(host);
+	host->ops->hw_reset(host);
+	mmc_host_clk_release(host);
+}
+
+int mmc_can_reset(struct mmc_card *card)
+{
+	u8 rst_n_function;
+
+	if (!mmc_card_mmc(card))
+		return 0;
+	rst_n_function = card->ext_csd.rst_n_function;
+	if ((rst_n_function & EXT_CSD_RST_N_EN_MASK) != EXT_CSD_RST_N_ENABLED)
+		return 0;
+	return 1;
+}
+EXPORT_SYMBOL(mmc_can_reset);
+
+static int mmc_do_hw_reset(struct mmc_host *host, int check)
+{
+	struct mmc_card *card = host->card;
+
+	if (!(host->caps & MMC_CAP_HW_RESET) || !host->ops->hw_reset)
+		return -EOPNOTSUPP;
+
+	if (!card)
+		return -EINVAL;
+
+	if (!mmc_can_reset(card))
+		return -EOPNOTSUPP;
+
+	mmc_host_clk_hold(host);
+	mmc_set_clock(host, host->f_init);
+
+	host->ops->hw_reset(host);
+
+	/* If the reset has happened, then a status command will fail */
+	if (check) {
+		struct mmc_command cmd = {0};
+		int err;
+
+		cmd.opcode = MMC_SEND_STATUS;
+		if (!mmc_host_is_spi(card->host))
+			cmd.arg = card->rca << 16;
+		cmd.flags = MMC_RSP_SPI_R2 | MMC_RSP_R1 | MMC_CMD_AC;
+		err = mmc_wait_for_cmd(card->host, &cmd, 0);
 		if (!err) {
-			if (mmc_attach_mmc(host, ocr))
-				mmc_power_off(host);
-			goto out;
+			mmc_host_clk_release(host);
+			return -ENOSYS;
 		}
+	}
 
-		mmc_release_host(host);
-		mmc_power_off(host);
+	if (mmc_host_is_spi(host)) {
+		host->ios.chip_select = MMC_CS_HIGH;
+		host->ios.bus_mode = MMC_BUSMODE_PUSHPULL;
 	} else {
-		if (host->bus_ops->detect && !host->bus_dead)
-			host->bus_ops->detect(host);
+		host->ios.chip_select = MMC_CS_DONTCARE;
+		host->ios.bus_mode = MMC_BUSMODE_OPENDRAIN;
+	}
+	host->ios.bus_width = MMC_BUS_WIDTH_1;
+	host->ios.timing = MMC_TIMING_LEGACY;
+	mmc_set_ios(host);
+
+	mmc_host_clk_release(host);
+
+	return host->bus_ops->power_restore(host);
+}
+
+int mmc_hw_reset(struct mmc_host *host)
+{
+	return mmc_do_hw_reset(host, 0);
+}
+EXPORT_SYMBOL(mmc_hw_reset);
+
+int mmc_hw_reset_check(struct mmc_host *host)
+{
+	return mmc_do_hw_reset(host, 1);
+}
+EXPORT_SYMBOL(mmc_hw_reset_check);
+
+static int mmc_rescan_try_freq(struct mmc_host *host, unsigned freq)
+{
+	host->f_init = freq;
+
+#ifdef CONFIG_MMC_DEBUG
+	pr_info("%s: %s: trying to init card at %u Hz\n",
+		mmc_hostname(host), __func__, host->f_init);
+#endif
+	mmc_power_up(host, host->ocr_avail);
+
+	/*
+	 * Some eMMCs (with VCCQ always on) may not be reset after power up, so
+	 * do a hardware reset if possible.
+	 */
+	mmc_hw_reset_for_init(host);
+
+	/*
+	 * sdio_reset sends CMD52 to reset card.  Since we do not know
+	 * if the card is being re-initialized, just send it.  CMD52
+	 * should be ignored by SD/eMMC cards.
+	 */
+	sdio_reset(host);
+	mmc_go_idle(host);
+
+	mmc_send_if_cond(host, host->ocr_avail);
+
+	/* Order's important: probe SDIO, then SD, then MMC */
+	if (!mmc_attach_sdio(host))
+		return 0;
+	if (!mmc_attach_sd(host))
+		return 0;
+	if (!mmc_attach_mmc(host))
+		return 0;
+
+	mmc_power_off(host);
+	return -EIO;
+}
+
+int _mmc_detect_card_removed(struct mmc_host *host)
+{
+	int ret;
+
+	if (host->caps & MMC_CAP_NONREMOVABLE)
+		return 0;
+
+	if (!host->card || mmc_card_removed(host->card))
+		return 1;
+
+	ret = host->bus_ops->alive(host);
+
+	/*
+	 * Card detect status and alive check may be out of sync if card is
+	 * removed slowly, when card detect switch changes while card/slot
+	 * pads are still contacted in hardware (refer to "SD Card Mechanical
+	 * Addendum, Appendix C: Card Detection Switch"). So reschedule a
+	 * detect work 200ms later for this case.
+	 */
+	if (!ret && host->ops->get_cd && !host->ops->get_cd(host)) {
+		mmc_detect_change(host, msecs_to_jiffies(200));
+		pr_debug("%s: card removed too slowly\n", mmc_hostname(host));
+	}
+
+	if (ret) {
+		mmc_card_set_removed(host->card);
+		pr_debug("%s: card remove detected\n", mmc_hostname(host));
+	}
+
+	return ret;
+}
+
+int mmc_detect_card_removed(struct mmc_host *host)
+{
+	struct mmc_card *card = host->card;
+	int ret;
+
+	WARN_ON(!host->claimed);
+
+	if (!card)
+		return 1;
+
+	ret = mmc_card_removed(card);
+	/*
+	 * The card will be considered unchanged unless we have been asked to
+	 * detect a change or host requires polling to provide card detection.
+	 */
+	if (!host->detect_change && !(host->caps & MMC_CAP_NEEDS_POLL))
+		return ret;
+
+	host->detect_change = 0;
+	if (!ret) {
+		ret = _mmc_detect_card_removed(host);
+		if (ret && (host->caps & MMC_CAP_NEEDS_POLL)) {
+			/*
+			 * Schedule a detect work as soon as possible to let a
+			 * rescan handle the card removal.
+			 */
+			cancel_delayed_work(&host->detect);
+			_mmc_detect_change(host, 0, false);
+		}
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(mmc_detect_card_removed);
+
+void mmc_rescan(struct work_struct *work)
+{
+	struct mmc_host *host =
+		container_of(work, struct mmc_host, detect.work);
+	int i;
+
+	if (host->trigger_card_event && host->ops->card_event) {
+		host->ops->card_event(host);
+		host->trigger_card_event = false;
+	}
+
+	if (host->rescan_disable)
+		return;
+
+	/* If there is a non-removable card registered, only scan once */
+	if ((host->caps & MMC_CAP_NONREMOVABLE) && host->rescan_entered)
+		return;
+	host->rescan_entered = 1;
+
+	mmc_bus_get(host);
+
+	/*
+	 * if there is a _removable_ card registered, check whether it is
+	 * still present
+	 */
+	if (host->bus_ops && !host->bus_dead
+	    && !(host->caps & MMC_CAP_NONREMOVABLE))
+		host->bus_ops->detect(host);
+
+	host->detect_change = 0;
+
+	/*
+	 * Let mmc_bus_put() free the bus/bus_ops if we've found that
+	 * the card is no longer present.
+	 */
+	mmc_bus_put(host);
+	mmc_bus_get(host);
 
+	/* if there still is a card present, stop here */
+	if (host->bus_ops != NULL) {
 		mmc_bus_put(host);
+		goto out;
 	}
-out:
+
+	/*
+	 * Only we can add a new handler, so it's safe to
+	 * release the lock here.
+	 */
+	mmc_bus_put(host);
+
+	if (!(host->caps & MMC_CAP_NONREMOVABLE) && host->ops->get_cd &&
+			host->ops->get_cd(host) == 0) {
+		mmc_claim_host(host);
+		mmc_power_off(host);
+		mmc_release_host(host);
+		goto out;
+	}
+
+	mmc_claim_host(host);
+	for (i = 0; i < ARRAY_SIZE(freqs); i++) {
+		if (!mmc_rescan_try_freq(host, max(freqs[i], host->f_min)))
+			break;
+		if (freqs[i] <= host->f_min)
+			break;
+	}
+	mmc_release_host(host);
+
+ out:
 	if (host->caps & MMC_CAP_NEEDS_POLL)
 		mmc_schedule_delayed_work(&host->detect, HZ);
 }
 
 void mmc_start_host(struct mmc_host *host)
 {
-	mmc_power_off(host);
-	mmc_detect_change(host, 0);
+	host->f_init = max(freqs[0], host->f_min);
+	host->rescan_disable = 0;
+	if (host->caps2 & MMC_CAP2_NO_PRESCAN_POWERUP)
+		mmc_power_off(host);
+	else
+		mmc_power_up(host, host->ocr_avail);
+	mmc_gpiod_request_cd_irq(host);
+	_mmc_detect_change(host, 0, false);
 }
 
 void mmc_stop_host(struct mmc_host *host)
@@ -929,18 +2504,26 @@ void mmc_stop_host(struct mmc_host *host)
 	host->removed = 1;
 	spin_unlock_irqrestore(&host->lock, flags);
 #endif
+	if (host->slot.cd_irq >= 0)
+		disable_irq(host->slot.cd_irq);
 
-	cancel_delayed_work(&host->detect);
+	host->rescan_disable = 1;
+	cancel_delayed_work_sync(&host->detect);
 	mmc_flush_scheduled_work();
 
+	/* clear pm flags now and let card drivers set them as needed */
+	host->pm_flags = 0;
+
 	mmc_bus_get(host);
 	if (host->bus_ops && !host->bus_dead) {
-		if (host->bus_ops->remove)
-			host->bus_ops->remove(host);
-
+		/* Calling bus_ops->remove() with a claimed host can deadlock */
+		host->bus_ops->remove(host);
 		mmc_claim_host(host);
 		mmc_detach_bus(host);
+		mmc_power_off(host);
 		mmc_release_host(host);
+		mmc_bus_put(host);
+		return;
 	}
 	mmc_bus_put(host);
 
@@ -949,73 +2532,154 @@ void mmc_stop_host(struct mmc_host *host)
 	mmc_power_off(host);
 }
 
-#ifdef CONFIG_PM
-
-/**
- *	mmc_suspend_host - suspend a host
- *	@host: mmc host
- *	@state: suspend mode (PM_SUSPEND_xxx)
- */
-int mmc_suspend_host(struct mmc_host *host, pm_message_t state)
+int mmc_power_save_host(struct mmc_host *host)
 {
-	cancel_delayed_work(&host->detect);
-	mmc_flush_scheduled_work();
+	int ret = 0;
+
+#ifdef CONFIG_MMC_DEBUG
+	pr_info("%s: %s: powering down\n", mmc_hostname(host), __func__);
+#endif
 
 	mmc_bus_get(host);
-	if (host->bus_ops && !host->bus_dead) {
-		if (host->bus_ops->suspend)
-			host->bus_ops->suspend(host);
-		if (!host->bus_ops->resume) {
-			if (host->bus_ops->remove)
-				host->bus_ops->remove(host);
-
-			mmc_claim_host(host);
-			mmc_detach_bus(host);
-			mmc_release_host(host);
-		}
+
+	if (!host->bus_ops || host->bus_dead) {
+		mmc_bus_put(host);
+		return -EINVAL;
 	}
+
+	if (host->bus_ops->power_save)
+		ret = host->bus_ops->power_save(host);
+
 	mmc_bus_put(host);
 
 	mmc_power_off(host);
 
-	return 0;
+	return ret;
 }
+EXPORT_SYMBOL(mmc_power_save_host);
 
-EXPORT_SYMBOL(mmc_suspend_host);
-
-/**
- *	mmc_resume_host - resume a previously suspended host
- *	@host: mmc host
- */
-int mmc_resume_host(struct mmc_host *host)
+int mmc_power_restore_host(struct mmc_host *host)
 {
+	int ret;
+
+#ifdef CONFIG_MMC_DEBUG
+	pr_info("%s: %s: powering up\n", mmc_hostname(host), __func__);
+#endif
+
 	mmc_bus_get(host);
-	if (host->bus_ops && !host->bus_dead) {
-		mmc_power_up(host);
-		mmc_select_voltage(host, host->ocr);
-		BUG_ON(!host->bus_ops->resume);
-		host->bus_ops->resume(host);
+
+	if (!host->bus_ops || host->bus_dead) {
+		mmc_bus_put(host);
+		return -EINVAL;
 	}
+
+	mmc_power_up(host, host->card->ocr);
+	ret = host->bus_ops->power_restore(host);
+
 	mmc_bus_put(host);
 
-	/*
-	 * We add a slight delay here so that resume can progress
-	 * in parallel.
-	 */
-	mmc_detect_change(host, 1);
+	return ret;
+}
+EXPORT_SYMBOL(mmc_power_restore_host);
 
-	return 0;
+/*
+ * Flush the cache to the non-volatile storage.
+ */
+int mmc_flush_cache(struct mmc_card *card)
+{
+	int err = 0;
+
+	if (mmc_card_mmc(card) &&
+			(card->ext_csd.cache_size > 0) &&
+			(card->ext_csd.cache_ctrl & 1)) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				EXT_CSD_FLUSH_CACHE, 1, 0);
+		if (err)
+			pr_err("%s: cache flush error %d\n",
+					mmc_hostname(card->host), err);
+	}
+
+	return err;
 }
+EXPORT_SYMBOL(mmc_flush_cache);
+
+#ifdef CONFIG_PM
+
+/* Do the card removal on suspend if card is assumed removeable
+ * Do that in pm notifier while userspace isn't yet frozen, so we will be able
+   to sync the card.
+*/
+int mmc_pm_notify(struct notifier_block *notify_block,
+					unsigned long mode, void *unused)
+{
+	struct mmc_host *host = container_of(
+		notify_block, struct mmc_host, pm_notify);
+	unsigned long flags;
+	int err = 0;
 
-EXPORT_SYMBOL(mmc_resume_host);
+	switch (mode) {
+	case PM_HIBERNATION_PREPARE:
+	case PM_SUSPEND_PREPARE:
+		spin_lock_irqsave(&host->lock, flags);
+		host->rescan_disable = 1;
+		spin_unlock_irqrestore(&host->lock, flags);
+		cancel_delayed_work_sync(&host->detect);
 
+		if (!host->bus_ops)
+			break;
+
+		/* Validate prerequisites for suspend */
+		if (host->bus_ops->pre_suspend)
+			err = host->bus_ops->pre_suspend(host);
+		if (!err)
+			break;
+
+		/* Calling bus_ops->remove() with a claimed host can deadlock */
+		host->bus_ops->remove(host);
+		mmc_claim_host(host);
+		mmc_detach_bus(host);
+		mmc_power_off(host);
+		mmc_release_host(host);
+		host->pm_flags = 0;
+		break;
+
+	case PM_POST_SUSPEND:
+	case PM_POST_HIBERNATION:
+	case PM_POST_RESTORE:
+
+		spin_lock_irqsave(&host->lock, flags);
+		host->rescan_disable = 0;
+		spin_unlock_irqrestore(&host->lock, flags);
+		_mmc_detect_change(host, 0, false);
+
+	}
+
+	return 0;
+}
 #endif
 
+/**
+ * mmc_init_context_info() - init synchronization context
+ * @host: mmc host
+ *
+ * Init struct context_info needed to implement asynchronous
+ * request mechanism, used by mmc core, host driver and mmc requests
+ * supplier.
+ */
+void mmc_init_context_info(struct mmc_host *host)
+{
+	spin_lock_init(&host->context_info.lock);
+	host->context_info.is_new_req = false;
+	host->context_info.is_done_rcv = false;
+	host->context_info.is_waiting_last_req = false;
+	init_waitqueue_head(&host->context_info.wait);
+}
+
 static int __init mmc_init(void)
 {
 	int ret;
 
-	workqueue = create_singlethread_workqueue("kmmcd");
+	workqueue = alloc_ordered_workqueue("kmmcd", 0);
 	if (!workqueue)
 		return -ENOMEM;
 
diff --git a/drivers/mmc/core/core.h b/drivers/mmc/core/core.h
index c819effa103..443a584660f 100644
--- a/drivers/mmc/core/core.h
+++ b/drivers/mmc/core/core.h
@@ -18,19 +18,37 @@
 struct mmc_bus_ops {
 	void (*remove)(struct mmc_host *);
 	void (*detect)(struct mmc_host *);
-	void (*suspend)(struct mmc_host *);
-	void (*resume)(struct mmc_host *);
+	int (*pre_suspend)(struct mmc_host *);
+	int (*suspend)(struct mmc_host *);
+	int (*resume)(struct mmc_host *);
+	int (*runtime_suspend)(struct mmc_host *);
+	int (*runtime_resume)(struct mmc_host *);
+	int (*power_save)(struct mmc_host *);
+	int (*power_restore)(struct mmc_host *);
+	int (*alive)(struct mmc_host *);
+	int (*shutdown)(struct mmc_host *);
 };
 
 void mmc_attach_bus(struct mmc_host *host, const struct mmc_bus_ops *ops);
 void mmc_detach_bus(struct mmc_host *host);
 
+void mmc_init_erase(struct mmc_card *card);
+
 void mmc_set_chip_select(struct mmc_host *host, int mode);
 void mmc_set_clock(struct mmc_host *host, unsigned int hz);
+void mmc_gate_clock(struct mmc_host *host);
+void mmc_ungate_clock(struct mmc_host *host);
+void mmc_set_ungated(struct mmc_host *host);
 void mmc_set_bus_mode(struct mmc_host *host, unsigned int mode);
 void mmc_set_bus_width(struct mmc_host *host, unsigned int width);
 u32 mmc_select_voltage(struct mmc_host *host, u32 ocr);
+int mmc_set_signal_voltage(struct mmc_host *host, int signal_voltage, u32 ocr);
+int __mmc_set_signal_voltage(struct mmc_host *host, int signal_voltage);
 void mmc_set_timing(struct mmc_host *host, unsigned int timing);
+void mmc_set_driver_type(struct mmc_host *host, unsigned int drv_type);
+void mmc_power_up(struct mmc_host *host, u32 ocr);
+void mmc_power_off(struct mmc_host *host);
+void mmc_power_cycle(struct mmc_host *host, u32 ocr);
 
 static inline void mmc_delay(unsigned int ms)
 {
@@ -46,11 +64,14 @@ void mmc_rescan(struct work_struct *work);
 void mmc_start_host(struct mmc_host *host);
 void mmc_stop_host(struct mmc_host *host);
 
-int mmc_attach_mmc(struct mmc_host *host, u32 ocr);
-int mmc_attach_sd(struct mmc_host *host, u32 ocr);
-int mmc_attach_sdio(struct mmc_host *host, u32 ocr);
+int _mmc_detect_card_removed(struct mmc_host *host);
+
+int mmc_attach_mmc(struct mmc_host *host);
+int mmc_attach_sd(struct mmc_host *host);
+int mmc_attach_sdio(struct mmc_host *host);
 
-extern int use_spi_crc;
+/* Module parameters */
+extern bool use_spi_crc;
 
 /* Debugfs information for hosts and cards */
 void mmc_add_host_debugfs(struct mmc_host *host);
@@ -59,5 +80,6 @@ void mmc_remove_host_debugfs(struct mmc_host *host);
 void mmc_add_card_debugfs(struct mmc_card *card);
 void mmc_remove_card_debugfs(struct mmc_card *card);
 
+void mmc_init_context_info(struct mmc_host *host);
 #endif
 
diff --git a/drivers/mmc/core/debugfs.c b/drivers/mmc/core/debugfs.c
index 610dbd1fcc8..91eb1622324 100644
--- a/drivers/mmc/core/debugfs.c
+++ b/drivers/mmc/core/debugfs.c
@@ -7,10 +7,14 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
+#include <linux/moduleparam.h>
+#include <linux/export.h>
 #include <linux/debugfs.h>
 #include <linux/fs.h>
 #include <linux/seq_file.h>
+#include <linux/slab.h>
 #include <linux/stat.h>
+#include <linux/fault-inject.h>
 
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
@@ -18,6 +22,14 @@
 #include "core.h"
 #include "mmc_ops.h"
 
+#ifdef CONFIG_FAIL_MMC_REQUEST
+
+static DECLARE_FAULT_ATTR(fail_default_attr);
+static char *fail_request;
+module_param(fail_request, charp, 0);
+
+#endif /* CONFIG_FAIL_MMC_REQUEST */
+
 /* The debugfs functions are optimized away when CONFIG_DEBUG_FS isn't set. */
 static int mmc_ios_show(struct seq_file *s, void *data)
 {
@@ -45,6 +57,8 @@ static int mmc_ios_show(struct seq_file *s, void *data)
 	const char *str;
 
 	seq_printf(s, "clock:\t\t%u Hz\n", ios->clock);
+	if (host->actual_clock)
+		seq_printf(s, "actual clock:\t%u Hz\n", host->actual_clock);
 	seq_printf(s, "vdd:\t\t%u ", ios->vdd);
 	if ((1 << ios->vdd) & MMC_VDD_165_195)
 		seq_printf(s, "(1.65 - 1.95 V)\n");
@@ -112,12 +126,46 @@ static int mmc_ios_show(struct seq_file *s, void *data)
 	case MMC_TIMING_SD_HS:
 		str = "sd high-speed";
 		break;
+	case MMC_TIMING_UHS_SDR50:
+		str = "sd uhs SDR50";
+		break;
+	case MMC_TIMING_UHS_SDR104:
+		str = "sd uhs SDR104";
+		break;
+	case MMC_TIMING_UHS_DDR50:
+		str = "sd uhs DDR50";
+		break;
+	case MMC_TIMING_MMC_DDR52:
+		str = "mmc DDR52";
+		break;
+	case MMC_TIMING_MMC_HS200:
+		str = "mmc HS200";
+		break;
+	case MMC_TIMING_MMC_HS400:
+		str = "mmc HS400";
+		break;
 	default:
 		str = "invalid";
 		break;
 	}
 	seq_printf(s, "timing spec:\t%u (%s)\n", ios->timing, str);
 
+	switch (ios->signal_voltage) {
+	case MMC_SIGNAL_VOLTAGE_330:
+		str = "3.30 V";
+		break;
+	case MMC_SIGNAL_VOLTAGE_180:
+		str = "1.80 V";
+		break;
+	case MMC_SIGNAL_VOLTAGE_120:
+		str = "1.20 V";
+		break;
+	default:
+		str = "invalid";
+		break;
+	}
+	seq_printf(s, "signal voltage:\t%u (%s)\n", ios->chip_select, str);
+
 	return 0;
 }
 
@@ -133,6 +181,33 @@ static const struct file_operations mmc_ios_fops = {
 	.release	= single_release,
 };
 
+static int mmc_clock_opt_get(void *data, u64 *val)
+{
+	struct mmc_host *host = data;
+
+	*val = host->ios.clock;
+
+	return 0;
+}
+
+static int mmc_clock_opt_set(void *data, u64 val)
+{
+	struct mmc_host *host = data;
+
+	/* We need this check due to input value is u64 */
+	if (val > host->f_max)
+		return -EINVAL;
+
+	mmc_claim_host(host);
+	mmc_set_clock(host, (unsigned int) val);
+	mmc_release_host(host);
+
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(mmc_clock_fops, mmc_clock_opt_get, mmc_clock_opt_set,
+	"%llu\n");
+
 void mmc_add_host_debugfs(struct mmc_host *host)
 {
 	struct dentry *root;
@@ -149,11 +224,29 @@ void mmc_add_host_debugfs(struct mmc_host *host)
 	host->debugfs_root = root;
 
 	if (!debugfs_create_file("ios", S_IRUSR, root, host, &mmc_ios_fops))
-		goto err_ios;
-
+		goto err_node;
+
+	if (!debugfs_create_file("clock", S_IRUSR | S_IWUSR, root, host,
+			&mmc_clock_fops))
+		goto err_node;
+
+#ifdef CONFIG_MMC_CLKGATE
+	if (!debugfs_create_u32("clk_delay", (S_IRUSR | S_IWUSR),
+				root, &host->clk_delay))
+		goto err_node;
+#endif
+#ifdef CONFIG_FAIL_MMC_REQUEST
+	if (fail_request)
+		setup_fault_attr(&fail_default_attr, fail_request);
+	host->fail_mmc_request = fail_default_attr;
+	if (IS_ERR(fault_create_debugfs_attr("fail_mmc_request",
+					     root,
+					     &host->fail_mmc_request)))
+		goto err_node;
+#endif
 	return;
 
-err_ios:
+err_node:
 	debugfs_remove_recursive(root);
 	host->debugfs_root = NULL;
 err_root:
@@ -171,13 +264,13 @@ static int mmc_dbg_card_status_get(void *data, u64 *val)
 	u32		status;
 	int		ret;
 
-	mmc_claim_host(card->host);
+	mmc_get_card(card);
 
 	ret = mmc_send_status(data, &status);
 	if (!ret)
 		*val = status;
 
-	mmc_release_host(card->host);
+	mmc_put_card(card);
 
 	return ret;
 }
@@ -204,13 +297,13 @@ static int mmc_ext_csd_open(struct inode *inode, struct file *filp)
 		goto out_free;
 	}
 
-	mmc_claim_host(card->host);
+	mmc_get_card(card);
 	err = mmc_send_ext_csd(card, ext_csd);
-	mmc_release_host(card->host);
+	mmc_put_card(card);
 	if (err)
 		goto out_free;
 
-	for (i = 511; i >= 0; i--)
+	for (i = 0; i < 512; i++)
 		n += sprintf(buf + n, "%02x", ext_csd[i]);
 	n += sprintf(buf + n, "\n");
 	BUG_ON(n != EXT_CSD_STR_LEN);
@@ -240,10 +333,11 @@ static int mmc_ext_csd_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static struct file_operations mmc_dbg_ext_csd_fops = {
+static const struct file_operations mmc_dbg_ext_csd_fops = {
 	.open		= mmc_ext_csd_open,
 	.read		= mmc_ext_csd_read,
 	.release	= mmc_ext_csd_release,
+	.llseek		= default_llseek,
 };
 
 void mmc_add_card_debugfs(struct mmc_card *card)
diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
index 5e945e64ead..95cceae9694 100644
--- a/drivers/mmc/core/host.c
+++ b/drivers/mmc/core/host.c
@@ -3,6 +3,7 @@
  *
  *  Copyright (C) 2003 Russell King, All Rights Reserved.
  *  Copyright (C) 2007-2008 Pierre Ossman
+ *  Copyright (C) 2010 Linus Walleij
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -14,10 +15,17 @@
 #include <linux/device.h>
 #include <linux/err.h>
 #include <linux/idr.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
 #include <linux/pagemap.h>
+#include <linux/export.h>
 #include <linux/leds.h>
+#include <linux/slab.h>
+#include <linux/suspend.h>
 
 #include <linux/mmc/host.h>
+#include <linux/mmc/card.h>
+#include <linux/mmc/slot-gpio.h>
 
 #include "core.h"
 #include "host.h"
@@ -27,6 +35,7 @@
 static void mmc_host_classdev_release(struct device *dev)
 {
 	struct mmc_host *host = cls_dev_to_mmc_host(dev);
+	mutex_destroy(&host->slot.lock);
 	kfree(host);
 }
 
@@ -48,6 +57,410 @@ void mmc_unregister_host_class(void)
 static DEFINE_IDR(mmc_host_idr);
 static DEFINE_SPINLOCK(mmc_host_lock);
 
+#ifdef CONFIG_MMC_CLKGATE
+static ssize_t clkgate_delay_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct mmc_host *host = cls_dev_to_mmc_host(dev);
+	return snprintf(buf, PAGE_SIZE, "%lu\n", host->clkgate_delay);
+}
+
+static ssize_t clkgate_delay_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t count)
+{
+	struct mmc_host *host = cls_dev_to_mmc_host(dev);
+	unsigned long flags, value;
+
+	if (kstrtoul(buf, 0, &value))
+		return -EINVAL;
+
+	spin_lock_irqsave(&host->clk_lock, flags);
+	host->clkgate_delay = value;
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+	return count;
+}
+
+/*
+ * Enabling clock gating will make the core call out to the host
+ * once up and once down when it performs a request or card operation
+ * intermingled in any fashion. The driver will see this through
+ * set_ios() operations with ios.clock field set to 0 to gate (disable)
+ * the block clock, and to the old frequency to enable it again.
+ */
+static void mmc_host_clk_gate_delayed(struct mmc_host *host)
+{
+	unsigned long tick_ns;
+	unsigned long freq = host->ios.clock;
+	unsigned long flags;
+
+	if (!freq) {
+		pr_debug("%s: frequency set to 0 in disable function, "
+			 "this means the clock is already disabled.\n",
+			 mmc_hostname(host));
+		return;
+	}
+	/*
+	 * New requests may have appeared while we were scheduling,
+	 * then there is no reason to delay the check before
+	 * clk_disable().
+	 */
+	spin_lock_irqsave(&host->clk_lock, flags);
+
+	/*
+	 * Delay n bus cycles (at least 8 from MMC spec) before attempting
+	 * to disable the MCI block clock. The reference count may have
+	 * gone up again after this delay due to rescheduling!
+	 */
+	if (!host->clk_requests) {
+		spin_unlock_irqrestore(&host->clk_lock, flags);
+		tick_ns = DIV_ROUND_UP(1000000000, freq);
+		ndelay(host->clk_delay * tick_ns);
+	} else {
+		/* New users appeared while waiting for this work */
+		spin_unlock_irqrestore(&host->clk_lock, flags);
+		return;
+	}
+	mutex_lock(&host->clk_gate_mutex);
+	spin_lock_irqsave(&host->clk_lock, flags);
+	if (!host->clk_requests) {
+		spin_unlock_irqrestore(&host->clk_lock, flags);
+		/* This will set host->ios.clock to 0 */
+		mmc_gate_clock(host);
+		spin_lock_irqsave(&host->clk_lock, flags);
+		pr_debug("%s: gated MCI clock\n", mmc_hostname(host));
+	}
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+	mutex_unlock(&host->clk_gate_mutex);
+}
+
+/*
+ * Internal work. Work to disable the clock at some later point.
+ */
+static void mmc_host_clk_gate_work(struct work_struct *work)
+{
+	struct mmc_host *host = container_of(work, struct mmc_host,
+					      clk_gate_work.work);
+
+	mmc_host_clk_gate_delayed(host);
+}
+
+/**
+ *	mmc_host_clk_hold - ungate hardware MCI clocks
+ *	@host: host to ungate.
+ *
+ *	Makes sure the host ios.clock is restored to a non-zero value
+ *	past this call.	Increase clock reference count and ungate clock
+ *	if we're the first user.
+ */
+void mmc_host_clk_hold(struct mmc_host *host)
+{
+	unsigned long flags;
+
+	/* cancel any clock gating work scheduled by mmc_host_clk_release() */
+	cancel_delayed_work_sync(&host->clk_gate_work);
+	mutex_lock(&host->clk_gate_mutex);
+	spin_lock_irqsave(&host->clk_lock, flags);
+	if (host->clk_gated) {
+		spin_unlock_irqrestore(&host->clk_lock, flags);
+		mmc_ungate_clock(host);
+		spin_lock_irqsave(&host->clk_lock, flags);
+		pr_debug("%s: ungated MCI clock\n", mmc_hostname(host));
+	}
+	host->clk_requests++;
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+	mutex_unlock(&host->clk_gate_mutex);
+}
+
+/**
+ *	mmc_host_may_gate_card - check if this card may be gated
+ *	@card: card to check.
+ */
+static bool mmc_host_may_gate_card(struct mmc_card *card)
+{
+	/* If there is no card we may gate it */
+	if (!card)
+		return true;
+	/*
+	 * Don't gate SDIO cards! These need to be clocked at all times
+	 * since they may be independent systems generating interrupts
+	 * and other events. The clock requests counter from the core will
+	 * go down to zero since the core does not need it, but we will not
+	 * gate the clock, because there is somebody out there that may still
+	 * be using it.
+	 */
+	return !(card->quirks & MMC_QUIRK_BROKEN_CLK_GATING);
+}
+
+/**
+ *	mmc_host_clk_release - gate off hardware MCI clocks
+ *	@host: host to gate.
+ *
+ *	Calls the host driver with ios.clock set to zero as often as possible
+ *	in order to gate off hardware MCI clocks. Decrease clock reference
+ *	count and schedule disabling of clock.
+ */
+void mmc_host_clk_release(struct mmc_host *host)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->clk_lock, flags);
+	host->clk_requests--;
+	if (mmc_host_may_gate_card(host->card) &&
+	    !host->clk_requests)
+		schedule_delayed_work(&host->clk_gate_work,
+				      msecs_to_jiffies(host->clkgate_delay));
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+}
+
+/**
+ *	mmc_host_clk_rate - get current clock frequency setting
+ *	@host: host to get the clock frequency for.
+ *
+ *	Returns current clock frequency regardless of gating.
+ */
+unsigned int mmc_host_clk_rate(struct mmc_host *host)
+{
+	unsigned long freq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->clk_lock, flags);
+	if (host->clk_gated)
+		freq = host->clk_old;
+	else
+		freq = host->ios.clock;
+	spin_unlock_irqrestore(&host->clk_lock, flags);
+	return freq;
+}
+
+/**
+ *	mmc_host_clk_init - set up clock gating code
+ *	@host: host with potential clock to control
+ */
+static inline void mmc_host_clk_init(struct mmc_host *host)
+{
+	host->clk_requests = 0;
+	/* Hold MCI clock for 8 cycles by default */
+	host->clk_delay = 8;
+	/*
+	 * Default clock gating delay is 0ms to avoid wasting power.
+	 * This value can be tuned by writing into sysfs entry.
+	 */
+	host->clkgate_delay = 0;
+	host->clk_gated = false;
+	INIT_DELAYED_WORK(&host->clk_gate_work, mmc_host_clk_gate_work);
+	spin_lock_init(&host->clk_lock);
+	mutex_init(&host->clk_gate_mutex);
+}
+
+/**
+ *	mmc_host_clk_exit - shut down clock gating code
+ *	@host: host with potential clock to control
+ */
+static inline void mmc_host_clk_exit(struct mmc_host *host)
+{
+	/*
+	 * Wait for any outstanding gate and then make sure we're
+	 * ungated before exiting.
+	 */
+	if (cancel_delayed_work_sync(&host->clk_gate_work))
+		mmc_host_clk_gate_delayed(host);
+	if (host->clk_gated)
+		mmc_host_clk_hold(host);
+	/* There should be only one user now */
+	WARN_ON(host->clk_requests > 1);
+}
+
+static inline void mmc_host_clk_sysfs_init(struct mmc_host *host)
+{
+	host->clkgate_delay_attr.show = clkgate_delay_show;
+	host->clkgate_delay_attr.store = clkgate_delay_store;
+	sysfs_attr_init(&host->clkgate_delay_attr.attr);
+	host->clkgate_delay_attr.attr.name = "clkgate_delay";
+	host->clkgate_delay_attr.attr.mode = S_IRUGO | S_IWUSR;
+	if (device_create_file(&host->class_dev, &host->clkgate_delay_attr))
+		pr_err("%s: Failed to create clkgate_delay sysfs entry\n",
+				mmc_hostname(host));
+}
+#else
+
+static inline void mmc_host_clk_init(struct mmc_host *host)
+{
+}
+
+static inline void mmc_host_clk_exit(struct mmc_host *host)
+{
+}
+
+static inline void mmc_host_clk_sysfs_init(struct mmc_host *host)
+{
+}
+
+#endif
+
+/**
+ *	mmc_of_parse() - parse host's device-tree node
+ *	@host: host whose node should be parsed.
+ *
+ * To keep the rest of the MMC subsystem unaware of whether DT has been
+ * used to to instantiate and configure this host instance or not, we
+ * parse the properties and set respective generic mmc-host flags and
+ * parameters.
+ */
+int mmc_of_parse(struct mmc_host *host)
+{
+	struct device_node *np;
+	u32 bus_width;
+	bool explicit_inv_wp, gpio_inv_wp = false;
+	enum of_gpio_flags flags;
+	int len, ret, gpio;
+
+	if (!host->parent || !host->parent->of_node)
+		return 0;
+
+	np = host->parent->of_node;
+
+	/* "bus-width" is translated to MMC_CAP_*_BIT_DATA flags */
+	if (of_property_read_u32(np, "bus-width", &bus_width) < 0) {
+		dev_dbg(host->parent,
+			"\"bus-width\" property is missing, assuming 1 bit.\n");
+		bus_width = 1;
+	}
+
+	switch (bus_width) {
+	case 8:
+		host->caps |= MMC_CAP_8_BIT_DATA;
+		/* Hosts capable of 8-bit transfers can also do 4 bits */
+	case 4:
+		host->caps |= MMC_CAP_4_BIT_DATA;
+		break;
+	case 1:
+		break;
+	default:
+		dev_err(host->parent,
+			"Invalid \"bus-width\" value %u!\n", bus_width);
+		return -EINVAL;
+	}
+
+	/* f_max is obtained from the optional "max-frequency" property */
+	of_property_read_u32(np, "max-frequency", &host->f_max);
+
+	/*
+	 * Configure CD and WP pins. They are both by default active low to
+	 * match the SDHCI spec. If GPIOs are provided for CD and / or WP, the
+	 * mmc-gpio helpers are used to attach, configure and use them. If
+	 * polarity inversion is specified in DT, one of MMC_CAP2_CD_ACTIVE_HIGH
+	 * and MMC_CAP2_RO_ACTIVE_HIGH capability-2 flags is set. If the
+	 * "broken-cd" property is provided, the MMC_CAP_NEEDS_POLL capability
+	 * is set. If the "non-removable" property is found, the
+	 * MMC_CAP_NONREMOVABLE capability is set and no card-detection
+	 * configuration is performed.
+	 */
+
+	/* Parse Card Detection */
+	if (of_find_property(np, "non-removable", &len)) {
+		host->caps |= MMC_CAP_NONREMOVABLE;
+	} else {
+		bool explicit_inv_cd, gpio_inv_cd = false;
+
+		explicit_inv_cd = of_property_read_bool(np, "cd-inverted");
+
+		if (of_find_property(np, "broken-cd", &len))
+			host->caps |= MMC_CAP_NEEDS_POLL;
+
+		gpio = of_get_named_gpio_flags(np, "cd-gpios", 0, &flags);
+		if (gpio == -EPROBE_DEFER)
+			return gpio;
+		if (gpio_is_valid(gpio)) {
+			if (!(flags & OF_GPIO_ACTIVE_LOW))
+				gpio_inv_cd = true;
+
+			ret = mmc_gpio_request_cd(host, gpio, 0);
+			if (ret < 0) {
+				dev_err(host->parent,
+					"Failed to request CD GPIO #%d: %d!\n",
+					gpio, ret);
+				return ret;
+			} else {
+				dev_info(host->parent, "Got CD GPIO #%d.\n",
+					 gpio);
+			}
+		}
+
+		if (explicit_inv_cd ^ gpio_inv_cd)
+			host->caps2 |= MMC_CAP2_CD_ACTIVE_HIGH;
+	}
+
+	/* Parse Write Protection */
+	explicit_inv_wp = of_property_read_bool(np, "wp-inverted");
+
+	gpio = of_get_named_gpio_flags(np, "wp-gpios", 0, &flags);
+	if (gpio == -EPROBE_DEFER) {
+		ret = -EPROBE_DEFER;
+		goto out;
+	}
+	if (gpio_is_valid(gpio)) {
+		if (!(flags & OF_GPIO_ACTIVE_LOW))
+			gpio_inv_wp = true;
+
+		ret = mmc_gpio_request_ro(host, gpio);
+		if (ret < 0) {
+			dev_err(host->parent,
+				"Failed to request WP GPIO: %d!\n", ret);
+			goto out;
+		} else {
+				dev_info(host->parent, "Got WP GPIO #%d.\n",
+					 gpio);
+		}
+	}
+	if (explicit_inv_wp ^ gpio_inv_wp)
+		host->caps2 |= MMC_CAP2_RO_ACTIVE_HIGH;
+
+	if (of_find_property(np, "cap-sd-highspeed", &len))
+		host->caps |= MMC_CAP_SD_HIGHSPEED;
+	if (of_find_property(np, "cap-mmc-highspeed", &len))
+		host->caps |= MMC_CAP_MMC_HIGHSPEED;
+	if (of_find_property(np, "sd-uhs-sdr12", &len))
+		host->caps |= MMC_CAP_UHS_SDR12;
+	if (of_find_property(np, "sd-uhs-sdr25", &len))
+		host->caps |= MMC_CAP_UHS_SDR25;
+	if (of_find_property(np, "sd-uhs-sdr50", &len))
+		host->caps |= MMC_CAP_UHS_SDR50;
+	if (of_find_property(np, "sd-uhs-sdr104", &len))
+		host->caps |= MMC_CAP_UHS_SDR104;
+	if (of_find_property(np, "sd-uhs-ddr50", &len))
+		host->caps |= MMC_CAP_UHS_DDR50;
+	if (of_find_property(np, "cap-power-off-card", &len))
+		host->caps |= MMC_CAP_POWER_OFF_CARD;
+	if (of_find_property(np, "cap-sdio-irq", &len))
+		host->caps |= MMC_CAP_SDIO_IRQ;
+	if (of_find_property(np, "full-pwr-cycle", &len))
+		host->caps2 |= MMC_CAP2_FULL_PWR_CYCLE;
+	if (of_find_property(np, "keep-power-in-suspend", &len))
+		host->pm_caps |= MMC_PM_KEEP_POWER;
+	if (of_find_property(np, "enable-sdio-wakeup", &len))
+		host->pm_caps |= MMC_PM_WAKE_SDIO_IRQ;
+	if (of_find_property(np, "mmc-ddr-1_8v", &len))
+		host->caps |= MMC_CAP_1_8V_DDR;
+	if (of_find_property(np, "mmc-ddr-1_2v", &len))
+		host->caps |= MMC_CAP_1_2V_DDR;
+	if (of_find_property(np, "mmc-hs200-1_8v", &len))
+		host->caps2 |= MMC_CAP2_HS200_1_8V_SDR;
+	if (of_find_property(np, "mmc-hs200-1_2v", &len))
+		host->caps2 |= MMC_CAP2_HS200_1_2V_SDR;
+	if (of_find_property(np, "mmc-hs400-1_8v", &len))
+		host->caps2 |= MMC_CAP2_HS400_1_8V | MMC_CAP2_HS200_1_8V_SDR;
+	if (of_find_property(np, "mmc-hs400-1_2v", &len))
+		host->caps2 |= MMC_CAP2_HS400_1_2V | MMC_CAP2_HS200_1_2V_SDR;
+
+	return 0;
+
+out:
+	mmc_gpio_free_cd(host);
+	return ret;
+}
+
+EXPORT_SYMBOL(mmc_of_parse);
+
 /**
  *	mmc_alloc_host - initialise the per-host structure.
  *	@extra: sizeof private data structure
@@ -60,17 +473,20 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
 	int err;
 	struct mmc_host *host;
 
-	if (!idr_pre_get(&mmc_host_idr, GFP_KERNEL))
-		return NULL;
-
 	host = kzalloc(sizeof(struct mmc_host) + extra, GFP_KERNEL);
 	if (!host)
 		return NULL;
 
+	/* scanning will be enabled when we're ready */
+	host->rescan_disable = 1;
+	idr_preload(GFP_KERNEL);
 	spin_lock(&mmc_host_lock);
-	err = idr_get_new(&mmc_host_idr, host, &host->index);
+	err = idr_alloc(&mmc_host_idr, host, 0, 0, GFP_NOWAIT);
+	if (err >= 0)
+		host->index = err;
 	spin_unlock(&mmc_host_lock);
-	if (err)
+	idr_preload_end();
+	if (err < 0)
 		goto free;
 
 	dev_set_name(&host->class_dev, "mmc%d", host->index);
@@ -80,16 +496,23 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
 	host->class_dev.class = &mmc_host_class;
 	device_initialize(&host->class_dev);
 
+	mmc_host_clk_init(host);
+
+	mutex_init(&host->slot.lock);
+	host->slot.cd_irq = -EINVAL;
+
 	spin_lock_init(&host->lock);
 	init_waitqueue_head(&host->wq);
 	INIT_DELAYED_WORK(&host->detect, mmc_rescan);
+#ifdef CONFIG_PM
+	host->pm_notify.notifier_call = mmc_pm_notify;
+#endif
 
 	/*
 	 * By default, hosts do not support SGIO or large requests.
 	 * They have to set these according to their abilities.
 	 */
-	host->max_hw_segs = 1;
-	host->max_phys_segs = 1;
+	host->max_segs = 1;
 	host->max_seg_size = PAGE_CACHE_SIZE;
 
 	host->max_req_size = PAGE_CACHE_SIZE;
@@ -120,17 +543,19 @@ int mmc_add_host(struct mmc_host *host)
 	WARN_ON((host->caps & MMC_CAP_SDIO_IRQ) &&
 		!host->ops->enable_sdio_irq);
 
-	led_trigger_register_simple(dev_name(&host->class_dev), &host->led);
-
 	err = device_add(&host->class_dev);
 	if (err)
 		return err;
 
+	led_trigger_register_simple(dev_name(&host->class_dev), &host->led);
+
 #ifdef CONFIG_DEBUG_FS
 	mmc_add_host_debugfs(host);
 #endif
+	mmc_host_clk_sysfs_init(host);
 
 	mmc_start_host(host);
+	register_pm_notifier(&host->pm_notify);
 
 	return 0;
 }
@@ -147,6 +572,7 @@ EXPORT_SYMBOL(mmc_add_host);
  */
 void mmc_remove_host(struct mmc_host *host)
 {
+	unregister_pm_notifier(&host->pm_notify);
 	mmc_stop_host(host);
 
 #ifdef CONFIG_DEBUG_FS
@@ -156,6 +582,8 @@ void mmc_remove_host(struct mmc_host *host)
 	device_del(&host->class_dev);
 
 	led_trigger_unregister_simple(host->led);
+
+	mmc_host_clk_exit(host);
 }
 
 EXPORT_SYMBOL(mmc_remove_host);
@@ -176,4 +604,3 @@ void mmc_free_host(struct mmc_host *host)
 }
 
 EXPORT_SYMBOL(mmc_free_host);
-
diff --git a/drivers/mmc/core/host.h b/drivers/mmc/core/host.h
index c2dc3d2d9f9..f2ab9e57812 100644
--- a/drivers/mmc/core/host.h
+++ b/drivers/mmc/core/host.h
@@ -10,6 +10,7 @@
  */
 #ifndef _MMC_CORE_HOST_H
 #define _MMC_CORE_HOST_H
+#include <linux/mmc/host.h>
 
 int mmc_register_host_class(void);
 void mmc_unregister_host_class(void);
diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c
index 06084dbf127..793c6f7ddb0 100644
--- a/drivers/mmc/core/mmc.c
+++ b/drivers/mmc/core/mmc.c
@@ -11,6 +11,9 @@
  */
 
 #include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/pm_runtime.h>
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/card.h>
@@ -19,6 +22,7 @@
 #include "core.h"
 #include "bus.h"
 #include "mmc_ops.h"
+#include "sd_ops.h"
 
 static const unsigned int tran_exp[] = {
 	10000,		100000,		1000000,	10000000,
@@ -93,13 +97,14 @@ static int mmc_decode_cid(struct mmc_card *card)
 		card->cid.prod_name[3]	= UNSTUFF_BITS(resp, 72, 8);
 		card->cid.prod_name[4]	= UNSTUFF_BITS(resp, 64, 8);
 		card->cid.prod_name[5]	= UNSTUFF_BITS(resp, 56, 8);
+		card->cid.prv		= UNSTUFF_BITS(resp, 48, 8);
 		card->cid.serial	= UNSTUFF_BITS(resp, 16, 32);
 		card->cid.month		= UNSTUFF_BITS(resp, 12, 4);
 		card->cid.year		= UNSTUFF_BITS(resp, 8, 4) + 1997;
 		break;
 
 	default:
-		printk(KERN_ERR "%s: card has unknown MMCA version %d\n",
+		pr_err("%s: card has unknown MMCA version %d\n",
 			mmc_hostname(card->host), card->csd.mmca_vsn);
 		return -EINVAL;
 	}
@@ -107,23 +112,34 @@ static int mmc_decode_cid(struct mmc_card *card)
 	return 0;
 }
 
+static void mmc_set_erase_size(struct mmc_card *card)
+{
+	if (card->ext_csd.erase_group_def & 1)
+		card->erase_size = card->ext_csd.hc_erase_size;
+	else
+		card->erase_size = card->csd.erase_size;
+
+	mmc_init_erase(card);
+}
+
 /*
  * Given a 128-bit response, decode to our card CSD structure.
  */
 static int mmc_decode_csd(struct mmc_card *card)
 {
 	struct mmc_csd *csd = &card->csd;
-	unsigned int e, m, csd_struct;
+	unsigned int e, m, a, b;
 	u32 *resp = card->raw_csd;
 
 	/*
 	 * We only understand CSD structure v1.1 and v1.2.
 	 * v1.2 has extra information in bits 15, 11 and 10.
+	 * We also support eMMC v4.4 & v4.41.
 	 */
-	csd_struct = UNSTUFF_BITS(resp, 126, 2);
-	if (csd_struct != 1 && csd_struct != 2) {
-		printk(KERN_ERR "%s: unrecognised CSD structure version %d\n",
-			mmc_hostname(card->host), csd_struct);
+	csd->structure = UNSTUFF_BITS(resp, 126, 2);
+	if (csd->structure == 0) {
+		pr_err("%s: unrecognised CSD structure version %d\n",
+			mmc_hostname(card->host), csd->structure);
 		return -EINVAL;
 	}
 
@@ -150,19 +166,28 @@ static int mmc_decode_csd(struct mmc_card *card)
 	csd->write_blkbits = UNSTUFF_BITS(resp, 22, 4);
 	csd->write_partial = UNSTUFF_BITS(resp, 21, 1);
 
+	if (csd->write_blkbits >= 9) {
+		a = UNSTUFF_BITS(resp, 42, 5);
+		b = UNSTUFF_BITS(resp, 37, 5);
+		csd->erase_size = (a + 1) * (b + 1);
+		csd->erase_size <<= csd->write_blkbits - 9;
+	}
+
 	return 0;
 }
 
 /*
- * Read and decode extended CSD.
+ * Read extended CSD.
  */
-static int mmc_read_ext_csd(struct mmc_card *card)
+static int mmc_get_ext_csd(struct mmc_card *card, u8 **new_ext_csd)
 {
 	int err;
 	u8 *ext_csd;
-	unsigned int ext_csd_struct;
 
 	BUG_ON(!card);
+	BUG_ON(!new_ext_csd);
+
+	*new_ext_csd = NULL;
 
 	if (card->csd.mmca_vsn < CSD_SPEC_VER_4)
 		return 0;
@@ -173,77 +198,496 @@ static int mmc_read_ext_csd(struct mmc_card *card)
 	 */
 	ext_csd = kmalloc(512, GFP_KERNEL);
 	if (!ext_csd) {
-		printk(KERN_ERR "%s: could not allocate a buffer to "
+		pr_err("%s: could not allocate a buffer to "
 			"receive the ext_csd.\n", mmc_hostname(card->host));
 		return -ENOMEM;
 	}
 
 	err = mmc_send_ext_csd(card, ext_csd);
 	if (err) {
-		/*
-		 * We all hosts that cannot perform the command
-		 * to fail more gracefully
-		 */
-		if (err != -EINVAL)
-			goto out;
+		kfree(ext_csd);
+		*new_ext_csd = NULL;
+
+		/* If the host or the card can't do the switch,
+		 * fail more gracefully. */
+		if ((err != -EINVAL)
+		 && (err != -ENOSYS)
+		 && (err != -EFAULT))
+			return err;
 
 		/*
 		 * High capacity cards should have this "magic" size
 		 * stored in their CSD.
 		 */
 		if (card->csd.capacity == (4096 * 512)) {
-			printk(KERN_ERR "%s: unable to read EXT_CSD "
+			pr_err("%s: unable to read EXT_CSD "
 				"on a possible high capacity card. "
 				"Card will be ignored.\n",
 				mmc_hostname(card->host));
 		} else {
-			printk(KERN_WARNING "%s: unable to read "
+			pr_warning("%s: unable to read "
 				"EXT_CSD, performance might "
 				"suffer.\n",
 				mmc_hostname(card->host));
 			err = 0;
 		}
+	} else
+		*new_ext_csd = ext_csd;
 
-		goto out;
+	return err;
+}
+
+static void mmc_select_card_type(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	u8 card_type = card->ext_csd.raw_card_type;
+	u32 caps = host->caps, caps2 = host->caps2;
+	unsigned int hs_max_dtr = 0, hs200_max_dtr = 0;
+	unsigned int avail_type = 0;
+
+	if (caps & MMC_CAP_MMC_HIGHSPEED &&
+	    card_type & EXT_CSD_CARD_TYPE_HS_26) {
+		hs_max_dtr = MMC_HIGH_26_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS_26;
+	}
+
+	if (caps & MMC_CAP_MMC_HIGHSPEED &&
+	    card_type & EXT_CSD_CARD_TYPE_HS_52) {
+		hs_max_dtr = MMC_HIGH_52_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS_52;
+	}
+
+	if (caps & MMC_CAP_1_8V_DDR &&
+	    card_type & EXT_CSD_CARD_TYPE_DDR_1_8V) {
+		hs_max_dtr = MMC_HIGH_DDR_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_DDR_1_8V;
+	}
+
+	if (caps & MMC_CAP_1_2V_DDR &&
+	    card_type & EXT_CSD_CARD_TYPE_DDR_1_2V) {
+		hs_max_dtr = MMC_HIGH_DDR_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_DDR_1_2V;
+	}
+
+	if (caps2 & MMC_CAP2_HS200_1_8V_SDR &&
+	    card_type & EXT_CSD_CARD_TYPE_HS200_1_8V) {
+		hs200_max_dtr = MMC_HS200_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS200_1_8V;
+	}
+
+	if (caps2 & MMC_CAP2_HS200_1_2V_SDR &&
+	    card_type & EXT_CSD_CARD_TYPE_HS200_1_2V) {
+		hs200_max_dtr = MMC_HS200_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS200_1_2V;
+	}
+
+	if (caps2 & MMC_CAP2_HS400_1_8V &&
+	    card_type & EXT_CSD_CARD_TYPE_HS400_1_8V) {
+		hs200_max_dtr = MMC_HS200_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS400_1_8V;
+	}
+
+	if (caps2 & MMC_CAP2_HS400_1_2V &&
+	    card_type & EXT_CSD_CARD_TYPE_HS400_1_2V) {
+		hs200_max_dtr = MMC_HS200_MAX_DTR;
+		avail_type |= EXT_CSD_CARD_TYPE_HS400_1_2V;
+	}
+
+	card->ext_csd.hs_max_dtr = hs_max_dtr;
+	card->ext_csd.hs200_max_dtr = hs200_max_dtr;
+	card->mmc_avail_type = avail_type;
+}
+
+/*
+ * Decode extended CSD.
+ */
+static int mmc_read_ext_csd(struct mmc_card *card, u8 *ext_csd)
+{
+	int err = 0, idx;
+	unsigned int part_size;
+	u8 hc_erase_grp_sz = 0, hc_wp_grp_sz = 0;
+
+	BUG_ON(!card);
+
+	if (!ext_csd)
+		return 0;
+
+	/* Version is coded in the CSD_STRUCTURE byte in the EXT_CSD register */
+	card->ext_csd.raw_ext_csd_structure = ext_csd[EXT_CSD_STRUCTURE];
+	if (card->csd.structure == 3) {
+		if (card->ext_csd.raw_ext_csd_structure > 2) {
+			pr_err("%s: unrecognised EXT_CSD structure "
+				"version %d\n", mmc_hostname(card->host),
+					card->ext_csd.raw_ext_csd_structure);
+			err = -EINVAL;
+			goto out;
+		}
 	}
 
-	ext_csd_struct = ext_csd[EXT_CSD_REV];
-	if (ext_csd_struct > 3) {
-		printk(KERN_ERR "%s: unrecognised EXT_CSD structure "
-			"version %d\n", mmc_hostname(card->host),
-			ext_csd_struct);
+	card->ext_csd.rev = ext_csd[EXT_CSD_REV];
+	if (card->ext_csd.rev > 7) {
+		pr_err("%s: unrecognised EXT_CSD revision %d\n",
+			mmc_hostname(card->host), card->ext_csd.rev);
 		err = -EINVAL;
 		goto out;
 	}
 
-	if (ext_csd_struct >= 2) {
+	card->ext_csd.raw_sectors[0] = ext_csd[EXT_CSD_SEC_CNT + 0];
+	card->ext_csd.raw_sectors[1] = ext_csd[EXT_CSD_SEC_CNT + 1];
+	card->ext_csd.raw_sectors[2] = ext_csd[EXT_CSD_SEC_CNT + 2];
+	card->ext_csd.raw_sectors[3] = ext_csd[EXT_CSD_SEC_CNT + 3];
+	if (card->ext_csd.rev >= 2) {
 		card->ext_csd.sectors =
 			ext_csd[EXT_CSD_SEC_CNT + 0] << 0 |
 			ext_csd[EXT_CSD_SEC_CNT + 1] << 8 |
 			ext_csd[EXT_CSD_SEC_CNT + 2] << 16 |
 			ext_csd[EXT_CSD_SEC_CNT + 3] << 24;
-		if (card->ext_csd.sectors)
+
+		/* Cards with density > 2GiB are sector addressed */
+		if (card->ext_csd.sectors > (2u * 1024 * 1024 * 1024) / 512)
 			mmc_card_set_blockaddr(card);
 	}
 
-	switch (ext_csd[EXT_CSD_CARD_TYPE]) {
-	case EXT_CSD_CARD_TYPE_52 | EXT_CSD_CARD_TYPE_26:
-		card->ext_csd.hs_max_dtr = 52000000;
-		break;
-	case EXT_CSD_CARD_TYPE_26:
-		card->ext_csd.hs_max_dtr = 26000000;
-		break;
-	default:
-		/* MMC v4 spec says this cannot happen */
-		printk(KERN_WARNING "%s: card is mmc v4 but doesn't "
-			"support any high-speed modes.\n",
-			mmc_hostname(card->host));
-		goto out;
+	card->ext_csd.raw_card_type = ext_csd[EXT_CSD_CARD_TYPE];
+	mmc_select_card_type(card);
+
+	card->ext_csd.raw_s_a_timeout = ext_csd[EXT_CSD_S_A_TIMEOUT];
+	card->ext_csd.raw_erase_timeout_mult =
+		ext_csd[EXT_CSD_ERASE_TIMEOUT_MULT];
+	card->ext_csd.raw_hc_erase_grp_size =
+		ext_csd[EXT_CSD_HC_ERASE_GRP_SIZE];
+	if (card->ext_csd.rev >= 3) {
+		u8 sa_shift = ext_csd[EXT_CSD_S_A_TIMEOUT];
+		card->ext_csd.part_config = ext_csd[EXT_CSD_PART_CONFIG];
+
+		/* EXT_CSD value is in units of 10ms, but we store in ms */
+		card->ext_csd.part_time = 10 * ext_csd[EXT_CSD_PART_SWITCH_TIME];
+
+		/* Sleep / awake timeout in 100ns units */
+		if (sa_shift > 0 && sa_shift <= 0x17)
+			card->ext_csd.sa_timeout =
+					1 << ext_csd[EXT_CSD_S_A_TIMEOUT];
+		card->ext_csd.erase_group_def =
+			ext_csd[EXT_CSD_ERASE_GROUP_DEF];
+		card->ext_csd.hc_erase_timeout = 300 *
+			ext_csd[EXT_CSD_ERASE_TIMEOUT_MULT];
+		card->ext_csd.hc_erase_size =
+			ext_csd[EXT_CSD_HC_ERASE_GRP_SIZE] << 10;
+
+		card->ext_csd.rel_sectors = ext_csd[EXT_CSD_REL_WR_SEC_C];
+
+		/*
+		 * There are two boot regions of equal size, defined in
+		 * multiples of 128K.
+		 */
+		if (ext_csd[EXT_CSD_BOOT_MULT] && mmc_boot_partition_access(card->host)) {
+			for (idx = 0; idx < MMC_NUM_BOOT_PARTITION; idx++) {
+				part_size = ext_csd[EXT_CSD_BOOT_MULT] << 17;
+				mmc_part_add(card, part_size,
+					EXT_CSD_PART_CONFIG_ACC_BOOT0 + idx,
+					"boot%d", idx, true,
+					MMC_BLK_DATA_AREA_BOOT);
+			}
+		}
+	}
+
+	card->ext_csd.raw_hc_erase_gap_size =
+		ext_csd[EXT_CSD_HC_WP_GRP_SIZE];
+	card->ext_csd.raw_sec_trim_mult =
+		ext_csd[EXT_CSD_SEC_TRIM_MULT];
+	card->ext_csd.raw_sec_erase_mult =
+		ext_csd[EXT_CSD_SEC_ERASE_MULT];
+	card->ext_csd.raw_sec_feature_support =
+		ext_csd[EXT_CSD_SEC_FEATURE_SUPPORT];
+	card->ext_csd.raw_trim_mult =
+		ext_csd[EXT_CSD_TRIM_MULT];
+	card->ext_csd.raw_partition_support = ext_csd[EXT_CSD_PARTITION_SUPPORT];
+	if (card->ext_csd.rev >= 4) {
+		/*
+		 * Enhanced area feature support -- check whether the eMMC
+		 * card has the Enhanced area enabled.  If so, export enhanced
+		 * area offset and size to user by adding sysfs interface.
+		 */
+		if ((ext_csd[EXT_CSD_PARTITION_SUPPORT] & 0x2) &&
+		    (ext_csd[EXT_CSD_PARTITION_ATTRIBUTE] & 0x1)) {
+			hc_erase_grp_sz =
+				ext_csd[EXT_CSD_HC_ERASE_GRP_SIZE];
+			hc_wp_grp_sz =
+				ext_csd[EXT_CSD_HC_WP_GRP_SIZE];
+
+			card->ext_csd.enhanced_area_en = 1;
+			/*
+			 * calculate the enhanced data area offset, in bytes
+			 */
+			card->ext_csd.enhanced_area_offset =
+				(ext_csd[139] << 24) + (ext_csd[138] << 16) +
+				(ext_csd[137] << 8) + ext_csd[136];
+			if (mmc_card_blockaddr(card))
+				card->ext_csd.enhanced_area_offset <<= 9;
+			/*
+			 * calculate the enhanced data area size, in kilobytes
+			 */
+			card->ext_csd.enhanced_area_size =
+				(ext_csd[142] << 16) + (ext_csd[141] << 8) +
+				ext_csd[140];
+			card->ext_csd.enhanced_area_size *=
+				(size_t)(hc_erase_grp_sz * hc_wp_grp_sz);
+			card->ext_csd.enhanced_area_size <<= 9;
+		} else {
+			/*
+			 * If the enhanced area is not enabled, disable these
+			 * device attributes.
+			 */
+			card->ext_csd.enhanced_area_offset = -EINVAL;
+			card->ext_csd.enhanced_area_size = -EINVAL;
+		}
+
+		/*
+		 * General purpose partition feature support --
+		 * If ext_csd has the size of general purpose partitions,
+		 * set size, part_cfg, partition name in mmc_part.
+		 */
+		if (ext_csd[EXT_CSD_PARTITION_SUPPORT] &
+			EXT_CSD_PART_SUPPORT_PART_EN) {
+			if (card->ext_csd.enhanced_area_en != 1) {
+				hc_erase_grp_sz =
+					ext_csd[EXT_CSD_HC_ERASE_GRP_SIZE];
+				hc_wp_grp_sz =
+					ext_csd[EXT_CSD_HC_WP_GRP_SIZE];
+
+				card->ext_csd.enhanced_area_en = 1;
+			}
+
+			for (idx = 0; idx < MMC_NUM_GP_PARTITION; idx++) {
+				if (!ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3] &&
+				!ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3 + 1] &&
+				!ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3 + 2])
+					continue;
+				part_size =
+				(ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3 + 2]
+					<< 16) +
+				(ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3 + 1]
+					<< 8) +
+				ext_csd[EXT_CSD_GP_SIZE_MULT + idx * 3];
+				part_size *= (size_t)(hc_erase_grp_sz *
+					hc_wp_grp_sz);
+				mmc_part_add(card, part_size << 19,
+					EXT_CSD_PART_CONFIG_ACC_GP0 + idx,
+					"gp%d", idx, false,
+					MMC_BLK_DATA_AREA_GP);
+			}
+		}
+		card->ext_csd.sec_trim_mult =
+			ext_csd[EXT_CSD_SEC_TRIM_MULT];
+		card->ext_csd.sec_erase_mult =
+			ext_csd[EXT_CSD_SEC_ERASE_MULT];
+		card->ext_csd.sec_feature_support =
+			ext_csd[EXT_CSD_SEC_FEATURE_SUPPORT];
+		card->ext_csd.trim_timeout = 300 *
+			ext_csd[EXT_CSD_TRIM_MULT];
+
+		/*
+		 * Note that the call to mmc_part_add above defaults to read
+		 * only. If this default assumption is changed, the call must
+		 * take into account the value of boot_locked below.
+		 */
+		card->ext_csd.boot_ro_lock = ext_csd[EXT_CSD_BOOT_WP];
+		card->ext_csd.boot_ro_lockable = true;
+
+		/* Save power class values */
+		card->ext_csd.raw_pwr_cl_52_195 =
+			ext_csd[EXT_CSD_PWR_CL_52_195];
+		card->ext_csd.raw_pwr_cl_26_195 =
+			ext_csd[EXT_CSD_PWR_CL_26_195];
+		card->ext_csd.raw_pwr_cl_52_360 =
+			ext_csd[EXT_CSD_PWR_CL_52_360];
+		card->ext_csd.raw_pwr_cl_26_360 =
+			ext_csd[EXT_CSD_PWR_CL_26_360];
+		card->ext_csd.raw_pwr_cl_200_195 =
+			ext_csd[EXT_CSD_PWR_CL_200_195];
+		card->ext_csd.raw_pwr_cl_200_360 =
+			ext_csd[EXT_CSD_PWR_CL_200_360];
+		card->ext_csd.raw_pwr_cl_ddr_52_195 =
+			ext_csd[EXT_CSD_PWR_CL_DDR_52_195];
+		card->ext_csd.raw_pwr_cl_ddr_52_360 =
+			ext_csd[EXT_CSD_PWR_CL_DDR_52_360];
+		card->ext_csd.raw_pwr_cl_ddr_200_360 =
+			ext_csd[EXT_CSD_PWR_CL_DDR_200_360];
+	}
+
+	if (card->ext_csd.rev >= 5) {
+		/* Adjust production date as per JEDEC JESD84-B451 */
+		if (card->cid.year < 2010)
+			card->cid.year += 16;
+
+		/* check whether the eMMC card supports BKOPS */
+		if (ext_csd[EXT_CSD_BKOPS_SUPPORT] & 0x1) {
+			card->ext_csd.bkops = 1;
+			card->ext_csd.bkops_en = ext_csd[EXT_CSD_BKOPS_EN];
+			card->ext_csd.raw_bkops_status =
+				ext_csd[EXT_CSD_BKOPS_STATUS];
+			if (!card->ext_csd.bkops_en)
+				pr_info("%s: BKOPS_EN bit is not set\n",
+					mmc_hostname(card->host));
+		}
+
+		/* check whether the eMMC card supports HPI */
+		if (ext_csd[EXT_CSD_HPI_FEATURES] & 0x1) {
+			card->ext_csd.hpi = 1;
+			if (ext_csd[EXT_CSD_HPI_FEATURES] & 0x2)
+				card->ext_csd.hpi_cmd =	MMC_STOP_TRANSMISSION;
+			else
+				card->ext_csd.hpi_cmd = MMC_SEND_STATUS;
+			/*
+			 * Indicate the maximum timeout to close
+			 * a command interrupted by HPI
+			 */
+			card->ext_csd.out_of_int_time =
+				ext_csd[EXT_CSD_OUT_OF_INTERRUPT_TIME] * 10;
+		}
+
+		card->ext_csd.rel_param = ext_csd[EXT_CSD_WR_REL_PARAM];
+		card->ext_csd.rst_n_function = ext_csd[EXT_CSD_RST_N_FUNCTION];
+
+		/*
+		 * RPMB regions are defined in multiples of 128K.
+		 */
+		card->ext_csd.raw_rpmb_size_mult = ext_csd[EXT_CSD_RPMB_MULT];
+		if (ext_csd[EXT_CSD_RPMB_MULT] && mmc_host_cmd23(card->host)) {
+			mmc_part_add(card, ext_csd[EXT_CSD_RPMB_MULT] << 17,
+				EXT_CSD_PART_CONFIG_ACC_RPMB,
+				"rpmb", 0, false,
+				MMC_BLK_DATA_AREA_RPMB);
+		}
+	}
+
+	card->ext_csd.raw_erased_mem_count = ext_csd[EXT_CSD_ERASED_MEM_CONT];
+	if (ext_csd[EXT_CSD_ERASED_MEM_CONT])
+		card->erased_byte = 0xFF;
+	else
+		card->erased_byte = 0x0;
+
+	/* eMMC v4.5 or later */
+	if (card->ext_csd.rev >= 6) {
+		card->ext_csd.feature_support |= MMC_DISCARD_FEATURE;
+
+		card->ext_csd.generic_cmd6_time = 10 *
+			ext_csd[EXT_CSD_GENERIC_CMD6_TIME];
+		card->ext_csd.power_off_longtime = 10 *
+			ext_csd[EXT_CSD_POWER_OFF_LONG_TIME];
+
+		card->ext_csd.cache_size =
+			ext_csd[EXT_CSD_CACHE_SIZE + 0] << 0 |
+			ext_csd[EXT_CSD_CACHE_SIZE + 1] << 8 |
+			ext_csd[EXT_CSD_CACHE_SIZE + 2] << 16 |
+			ext_csd[EXT_CSD_CACHE_SIZE + 3] << 24;
+
+		if (ext_csd[EXT_CSD_DATA_SECTOR_SIZE] == 1)
+			card->ext_csd.data_sector_size = 4096;
+		else
+			card->ext_csd.data_sector_size = 512;
+
+		if ((ext_csd[EXT_CSD_DATA_TAG_SUPPORT] & 1) &&
+		    (ext_csd[EXT_CSD_TAG_UNIT_SIZE] <= 8)) {
+			card->ext_csd.data_tag_unit_size =
+			((unsigned int) 1 << ext_csd[EXT_CSD_TAG_UNIT_SIZE]) *
+			(card->ext_csd.data_sector_size);
+		} else {
+			card->ext_csd.data_tag_unit_size = 0;
+		}
+
+		card->ext_csd.max_packed_writes =
+			ext_csd[EXT_CSD_MAX_PACKED_WRITES];
+		card->ext_csd.max_packed_reads =
+			ext_csd[EXT_CSD_MAX_PACKED_READS];
+	} else {
+		card->ext_csd.data_sector_size = 512;
 	}
 
 out:
+	return err;
+}
+
+static inline void mmc_free_ext_csd(u8 *ext_csd)
+{
 	kfree(ext_csd);
+}
 
+
+static int mmc_compare_ext_csds(struct mmc_card *card, unsigned bus_width)
+{
+	u8 *bw_ext_csd;
+	int err;
+
+	if (bus_width == MMC_BUS_WIDTH_1)
+		return 0;
+
+	err = mmc_get_ext_csd(card, &bw_ext_csd);
+
+	if (err || bw_ext_csd == NULL) {
+		err = -EINVAL;
+		goto out;
+	}
+
+	/* only compare read only fields */
+	err = !((card->ext_csd.raw_partition_support ==
+			bw_ext_csd[EXT_CSD_PARTITION_SUPPORT]) &&
+		(card->ext_csd.raw_erased_mem_count ==
+			bw_ext_csd[EXT_CSD_ERASED_MEM_CONT]) &&
+		(card->ext_csd.rev ==
+			bw_ext_csd[EXT_CSD_REV]) &&
+		(card->ext_csd.raw_ext_csd_structure ==
+			bw_ext_csd[EXT_CSD_STRUCTURE]) &&
+		(card->ext_csd.raw_card_type ==
+			bw_ext_csd[EXT_CSD_CARD_TYPE]) &&
+		(card->ext_csd.raw_s_a_timeout ==
+			bw_ext_csd[EXT_CSD_S_A_TIMEOUT]) &&
+		(card->ext_csd.raw_hc_erase_gap_size ==
+			bw_ext_csd[EXT_CSD_HC_WP_GRP_SIZE]) &&
+		(card->ext_csd.raw_erase_timeout_mult ==
+			bw_ext_csd[EXT_CSD_ERASE_TIMEOUT_MULT]) &&
+		(card->ext_csd.raw_hc_erase_grp_size ==
+			bw_ext_csd[EXT_CSD_HC_ERASE_GRP_SIZE]) &&
+		(card->ext_csd.raw_sec_trim_mult ==
+			bw_ext_csd[EXT_CSD_SEC_TRIM_MULT]) &&
+		(card->ext_csd.raw_sec_erase_mult ==
+			bw_ext_csd[EXT_CSD_SEC_ERASE_MULT]) &&
+		(card->ext_csd.raw_sec_feature_support ==
+			bw_ext_csd[EXT_CSD_SEC_FEATURE_SUPPORT]) &&
+		(card->ext_csd.raw_trim_mult ==
+			bw_ext_csd[EXT_CSD_TRIM_MULT]) &&
+		(card->ext_csd.raw_sectors[0] ==
+			bw_ext_csd[EXT_CSD_SEC_CNT + 0]) &&
+		(card->ext_csd.raw_sectors[1] ==
+			bw_ext_csd[EXT_CSD_SEC_CNT + 1]) &&
+		(card->ext_csd.raw_sectors[2] ==
+			bw_ext_csd[EXT_CSD_SEC_CNT + 2]) &&
+		(card->ext_csd.raw_sectors[3] ==
+			bw_ext_csd[EXT_CSD_SEC_CNT + 3]) &&
+		(card->ext_csd.raw_pwr_cl_52_195 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_52_195]) &&
+		(card->ext_csd.raw_pwr_cl_26_195 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_26_195]) &&
+		(card->ext_csd.raw_pwr_cl_52_360 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_52_360]) &&
+		(card->ext_csd.raw_pwr_cl_26_360 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_26_360]) &&
+		(card->ext_csd.raw_pwr_cl_200_195 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_200_195]) &&
+		(card->ext_csd.raw_pwr_cl_200_360 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_200_360]) &&
+		(card->ext_csd.raw_pwr_cl_ddr_52_195 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_DDR_52_195]) &&
+		(card->ext_csd.raw_pwr_cl_ddr_52_360 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_DDR_52_360]) &&
+		(card->ext_csd.raw_pwr_cl_ddr_200_360 ==
+			bw_ext_csd[EXT_CSD_PWR_CL_DDR_200_360]));
+
+	if (err)
+		err = -EINVAL;
+
+out:
+	mmc_free_ext_csd(bw_ext_csd);
 	return err;
 }
 
@@ -252,38 +696,476 @@ MMC_DEV_ATTR(cid, "%08x%08x%08x%08x\n", card->raw_cid[0], card->raw_cid[1],
 MMC_DEV_ATTR(csd, "%08x%08x%08x%08x\n", card->raw_csd[0], card->raw_csd[1],
 	card->raw_csd[2], card->raw_csd[3]);
 MMC_DEV_ATTR(date, "%02d/%04d\n", card->cid.month, card->cid.year);
+MMC_DEV_ATTR(erase_size, "%u\n", card->erase_size << 9);
+MMC_DEV_ATTR(preferred_erase_size, "%u\n", card->pref_erase << 9);
 MMC_DEV_ATTR(fwrev, "0x%x\n", card->cid.fwrev);
 MMC_DEV_ATTR(hwrev, "0x%x\n", card->cid.hwrev);
 MMC_DEV_ATTR(manfid, "0x%06x\n", card->cid.manfid);
 MMC_DEV_ATTR(name, "%s\n", card->cid.prod_name);
 MMC_DEV_ATTR(oemid, "0x%04x\n", card->cid.oemid);
+MMC_DEV_ATTR(prv, "0x%x\n", card->cid.prv);
 MMC_DEV_ATTR(serial, "0x%08x\n", card->cid.serial);
+MMC_DEV_ATTR(enhanced_area_offset, "%llu\n",
+		card->ext_csd.enhanced_area_offset);
+MMC_DEV_ATTR(enhanced_area_size, "%u\n", card->ext_csd.enhanced_area_size);
+MMC_DEV_ATTR(raw_rpmb_size_mult, "%#x\n", card->ext_csd.raw_rpmb_size_mult);
+MMC_DEV_ATTR(rel_sectors, "%#x\n", card->ext_csd.rel_sectors);
 
 static struct attribute *mmc_std_attrs[] = {
 	&dev_attr_cid.attr,
 	&dev_attr_csd.attr,
 	&dev_attr_date.attr,
+	&dev_attr_erase_size.attr,
+	&dev_attr_preferred_erase_size.attr,
 	&dev_attr_fwrev.attr,
 	&dev_attr_hwrev.attr,
 	&dev_attr_manfid.attr,
 	&dev_attr_name.attr,
 	&dev_attr_oemid.attr,
+	&dev_attr_prv.attr,
 	&dev_attr_serial.attr,
+	&dev_attr_enhanced_area_offset.attr,
+	&dev_attr_enhanced_area_size.attr,
+	&dev_attr_raw_rpmb_size_mult.attr,
+	&dev_attr_rel_sectors.attr,
 	NULL,
 };
+ATTRIBUTE_GROUPS(mmc_std);
 
-static struct attribute_group mmc_std_attr_group = {
-	.attrs = mmc_std_attrs,
+static struct device_type mmc_type = {
+	.groups = mmc_std_groups,
 };
 
-static struct attribute_group *mmc_attr_groups[] = {
-	&mmc_std_attr_group,
-	NULL,
-};
+/*
+ * Select the PowerClass for the current bus width
+ * If power class is defined for 4/8 bit bus in the
+ * extended CSD register, select it by executing the
+ * mmc_switch command.
+ */
+static int __mmc_select_powerclass(struct mmc_card *card,
+				   unsigned int bus_width)
+{
+	struct mmc_host *host = card->host;
+	struct mmc_ext_csd *ext_csd = &card->ext_csd;
+	unsigned int pwrclass_val = 0;
+	int err = 0;
 
-static struct device_type mmc_type = {
-	.groups = mmc_attr_groups,
-};
+	/* Power class selection is supported for versions >= 4.0 */
+	if (card->csd.mmca_vsn < CSD_SPEC_VER_4)
+		return 0;
+
+	/* Power class values are defined only for 4/8 bit bus */
+	if (bus_width == EXT_CSD_BUS_WIDTH_1)
+		return 0;
+
+	switch (1 << host->ios.vdd) {
+	case MMC_VDD_165_195:
+		if (host->ios.clock <= MMC_HIGH_26_MAX_DTR)
+			pwrclass_val = ext_csd->raw_pwr_cl_26_195;
+		else if (host->ios.clock <= MMC_HIGH_52_MAX_DTR)
+			pwrclass_val = (bus_width <= EXT_CSD_BUS_WIDTH_8) ?
+				ext_csd->raw_pwr_cl_52_195 :
+				ext_csd->raw_pwr_cl_ddr_52_195;
+		else if (host->ios.clock <= MMC_HS200_MAX_DTR)
+			pwrclass_val = ext_csd->raw_pwr_cl_200_195;
+		break;
+	case MMC_VDD_27_28:
+	case MMC_VDD_28_29:
+	case MMC_VDD_29_30:
+	case MMC_VDD_30_31:
+	case MMC_VDD_31_32:
+	case MMC_VDD_32_33:
+	case MMC_VDD_33_34:
+	case MMC_VDD_34_35:
+	case MMC_VDD_35_36:
+		if (host->ios.clock <= MMC_HIGH_26_MAX_DTR)
+			pwrclass_val = ext_csd->raw_pwr_cl_26_360;
+		else if (host->ios.clock <= MMC_HIGH_52_MAX_DTR)
+			pwrclass_val = (bus_width <= EXT_CSD_BUS_WIDTH_8) ?
+				ext_csd->raw_pwr_cl_52_360 :
+				ext_csd->raw_pwr_cl_ddr_52_360;
+		else if (host->ios.clock <= MMC_HS200_MAX_DTR)
+			pwrclass_val = (bus_width == EXT_CSD_DDR_BUS_WIDTH_8) ?
+				ext_csd->raw_pwr_cl_ddr_200_360 :
+				ext_csd->raw_pwr_cl_200_360;
+		break;
+	default:
+		pr_warning("%s: Voltage range not supported "
+			   "for power class.\n", mmc_hostname(host));
+		return -EINVAL;
+	}
+
+	if (bus_width & (EXT_CSD_BUS_WIDTH_8 | EXT_CSD_DDR_BUS_WIDTH_8))
+		pwrclass_val = (pwrclass_val & EXT_CSD_PWR_CL_8BIT_MASK) >>
+				EXT_CSD_PWR_CL_8BIT_SHIFT;
+	else
+		pwrclass_val = (pwrclass_val & EXT_CSD_PWR_CL_4BIT_MASK) >>
+				EXT_CSD_PWR_CL_4BIT_SHIFT;
+
+	/* If the power class is different from the default value */
+	if (pwrclass_val > 0) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 EXT_CSD_POWER_CLASS,
+				 pwrclass_val,
+				 card->ext_csd.generic_cmd6_time);
+	}
+
+	return err;
+}
+
+static int mmc_select_powerclass(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	u32 bus_width, ext_csd_bits;
+	int err, ddr;
+
+	/* Power class selection is supported for versions >= 4.0 */
+	if (card->csd.mmca_vsn < CSD_SPEC_VER_4)
+		return 0;
+
+	bus_width = host->ios.bus_width;
+	/* Power class values are defined only for 4/8 bit bus */
+	if (bus_width == MMC_BUS_WIDTH_1)
+		return 0;
+
+	ddr = card->mmc_avail_type & EXT_CSD_CARD_TYPE_DDR_52;
+	if (ddr)
+		ext_csd_bits = (bus_width == MMC_BUS_WIDTH_8) ?
+			EXT_CSD_DDR_BUS_WIDTH_8 : EXT_CSD_DDR_BUS_WIDTH_4;
+	else
+		ext_csd_bits = (bus_width == MMC_BUS_WIDTH_8) ?
+			EXT_CSD_BUS_WIDTH_8 :  EXT_CSD_BUS_WIDTH_4;
+
+	err = __mmc_select_powerclass(card, ext_csd_bits);
+	if (err)
+		pr_warn("%s: power class selection to bus width %d ddr %d failed\n",
+			mmc_hostname(host), 1 << bus_width, ddr);
+
+	return err;
+}
+
+/*
+ * Set the bus speed for the selected speed mode.
+ */
+static void mmc_set_bus_speed(struct mmc_card *card)
+{
+	unsigned int max_dtr = (unsigned int)-1;
+
+	if ((mmc_card_hs200(card) || mmc_card_hs400(card)) &&
+	     max_dtr > card->ext_csd.hs200_max_dtr)
+		max_dtr = card->ext_csd.hs200_max_dtr;
+	else if (mmc_card_hs(card) && max_dtr > card->ext_csd.hs_max_dtr)
+		max_dtr = card->ext_csd.hs_max_dtr;
+	else if (max_dtr > card->csd.max_dtr)
+		max_dtr = card->csd.max_dtr;
+
+	mmc_set_clock(card->host, max_dtr);
+}
+
+/*
+ * Select the bus width amoung 4-bit and 8-bit(SDR).
+ * If the bus width is changed successfully, return the selected width value.
+ * Zero is returned instead of error value if the wide width is not supported.
+ */
+static int mmc_select_bus_width(struct mmc_card *card)
+{
+	static unsigned ext_csd_bits[] = {
+		EXT_CSD_BUS_WIDTH_8,
+		EXT_CSD_BUS_WIDTH_4,
+	};
+	static unsigned bus_widths[] = {
+		MMC_BUS_WIDTH_8,
+		MMC_BUS_WIDTH_4,
+	};
+	struct mmc_host *host = card->host;
+	unsigned idx, bus_width = 0;
+	int err = 0;
+
+	if ((card->csd.mmca_vsn < CSD_SPEC_VER_4) &&
+	    !(host->caps & (MMC_CAP_4_BIT_DATA | MMC_CAP_8_BIT_DATA)))
+		return 0;
+
+	idx = (host->caps & MMC_CAP_8_BIT_DATA) ? 0 : 1;
+
+	/*
+	 * Unlike SD, MMC cards dont have a configuration register to notify
+	 * supported bus width. So bus test command should be run to identify
+	 * the supported bus width or compare the ext csd values of current
+	 * bus width and ext csd values of 1 bit mode read earlier.
+	 */
+	for (; idx < ARRAY_SIZE(bus_widths); idx++) {
+		/*
+		 * Host is capable of 8bit transfer, then switch
+		 * the device to work in 8bit transfer mode. If the
+		 * mmc switch command returns error then switch to
+		 * 4bit transfer mode. On success set the corresponding
+		 * bus width on the host.
+		 */
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 EXT_CSD_BUS_WIDTH,
+				 ext_csd_bits[idx],
+				 card->ext_csd.generic_cmd6_time);
+		if (err)
+			continue;
+
+		bus_width = bus_widths[idx];
+		mmc_set_bus_width(host, bus_width);
+
+		/*
+		 * If controller can't handle bus width test,
+		 * compare ext_csd previously read in 1 bit mode
+		 * against ext_csd at new bus width
+		 */
+		if (!(host->caps & MMC_CAP_BUS_WIDTH_TEST))
+			err = mmc_compare_ext_csds(card, bus_width);
+		else
+			err = mmc_bus_test(card, bus_width);
+
+		if (!err) {
+			err = bus_width;
+			break;
+		} else {
+			pr_warn("%s: switch to bus width %d failed\n",
+				mmc_hostname(host), ext_csd_bits[idx]);
+		}
+	}
+
+	return err;
+}
+
+/*
+ * Switch to the high-speed mode
+ */
+static int mmc_select_hs(struct mmc_card *card)
+{
+	int err;
+
+	err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			   EXT_CSD_HS_TIMING, EXT_CSD_TIMING_HS,
+			   card->ext_csd.generic_cmd6_time,
+			   true, true, true);
+	if (!err)
+		mmc_set_timing(card->host, MMC_TIMING_MMC_HS);
+
+	return err;
+}
+
+/*
+ * Activate wide bus and DDR if supported.
+ */
+static int mmc_select_hs_ddr(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	u32 bus_width, ext_csd_bits;
+	int err = 0;
+
+	if (!(card->mmc_avail_type & EXT_CSD_CARD_TYPE_DDR_52))
+		return 0;
+
+	bus_width = host->ios.bus_width;
+	if (bus_width == MMC_BUS_WIDTH_1)
+		return 0;
+
+	ext_csd_bits = (bus_width == MMC_BUS_WIDTH_8) ?
+		EXT_CSD_DDR_BUS_WIDTH_8 : EXT_CSD_DDR_BUS_WIDTH_4;
+
+	err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			EXT_CSD_BUS_WIDTH,
+			ext_csd_bits,
+			card->ext_csd.generic_cmd6_time);
+	if (err) {
+		pr_warn("%s: switch to bus width %d ddr failed\n",
+			mmc_hostname(host), 1 << bus_width);
+		return err;
+	}
+
+	/*
+	 * eMMC cards can support 3.3V to 1.2V i/o (vccq)
+	 * signaling.
+	 *
+	 * EXT_CSD_CARD_TYPE_DDR_1_8V means 3.3V or 1.8V vccq.
+	 *
+	 * 1.8V vccq at 3.3V core voltage (vcc) is not required
+	 * in the JEDEC spec for DDR.
+	 *
+	 * Do not force change in vccq since we are obviously
+	 * working and no change to vccq is needed.
+	 *
+	 * WARNING: eMMC rules are NOT the same as SD DDR
+	 */
+	if (card->mmc_avail_type & EXT_CSD_CARD_TYPE_DDR_1_2V) {
+		err = __mmc_set_signal_voltage(host,
+				MMC_SIGNAL_VOLTAGE_120);
+		if (err)
+			return err;
+	}
+
+	mmc_set_timing(host, MMC_TIMING_MMC_DDR52);
+
+	return err;
+}
+
+static int mmc_select_hs400(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	int err = 0;
+
+	/*
+	 * HS400 mode requires 8-bit bus width
+	 */
+	if (!(card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS400 &&
+	      host->ios.bus_width == MMC_BUS_WIDTH_8))
+		return 0;
+
+	/*
+	 * Before switching to dual data rate operation for HS400,
+	 * it is required to convert from HS200 mode to HS mode.
+	 */
+	mmc_set_timing(card->host, MMC_TIMING_MMC_HS);
+	mmc_set_bus_speed(card);
+
+	err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			   EXT_CSD_HS_TIMING, EXT_CSD_TIMING_HS,
+			   card->ext_csd.generic_cmd6_time,
+			   true, true, true);
+	if (err) {
+		pr_warn("%s: switch to high-speed from hs200 failed, err:%d\n",
+			mmc_hostname(host), err);
+		return err;
+	}
+
+	err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			 EXT_CSD_BUS_WIDTH,
+			 EXT_CSD_DDR_BUS_WIDTH_8,
+			 card->ext_csd.generic_cmd6_time);
+	if (err) {
+		pr_warn("%s: switch to bus width for hs400 failed, err:%d\n",
+			mmc_hostname(host), err);
+		return err;
+	}
+
+	err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			   EXT_CSD_HS_TIMING, EXT_CSD_TIMING_HS400,
+			   card->ext_csd.generic_cmd6_time,
+			   true, true, true);
+	if (err) {
+		pr_warn("%s: switch to hs400 failed, err:%d\n",
+			 mmc_hostname(host), err);
+		return err;
+	}
+
+	mmc_set_timing(host, MMC_TIMING_MMC_HS400);
+	mmc_set_bus_speed(card);
+
+	return 0;
+}
+
+/*
+ * For device supporting HS200 mode, the following sequence
+ * should be done before executing the tuning process.
+ * 1. set the desired bus width(4-bit or 8-bit, 1-bit is not supported)
+ * 2. switch to HS200 mode
+ * 3. set the clock to > 52Mhz and <=200MHz
+ */
+static int mmc_select_hs200(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	int err = -EINVAL;
+
+	if (card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS200_1_2V)
+		err = __mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_120);
+
+	if (err && card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS200_1_8V)
+		err = __mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_180);
+
+	/* If fails try again during next card power cycle */
+	if (err)
+		goto err;
+
+	/*
+	 * Set the bus width(4 or 8) with host's support and
+	 * switch to HS200 mode if bus width is set successfully.
+	 */
+	err = mmc_select_bus_width(card);
+	if (!IS_ERR_VALUE(err)) {
+		err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				   EXT_CSD_HS_TIMING, EXT_CSD_TIMING_HS200,
+				   card->ext_csd.generic_cmd6_time,
+				   true, true, true);
+		if (!err)
+			mmc_set_timing(host, MMC_TIMING_MMC_HS200);
+	}
+err:
+	return err;
+}
+
+/*
+ * Activate High Speed or HS200 mode if supported.
+ */
+static int mmc_select_timing(struct mmc_card *card)
+{
+	int err = 0;
+
+	if ((card->csd.mmca_vsn < CSD_SPEC_VER_4 &&
+	     card->ext_csd.hs_max_dtr == 0))
+		goto bus_speed;
+
+	if (card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS200)
+		err = mmc_select_hs200(card);
+	else if (card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS)
+		err = mmc_select_hs(card);
+
+	if (err && err != -EBADMSG)
+		return err;
+
+	if (err) {
+		pr_warn("%s: switch to %s failed\n",
+			mmc_card_hs(card) ? "high-speed" :
+			(mmc_card_hs200(card) ? "hs200" : ""),
+			mmc_hostname(card->host));
+		err = 0;
+	}
+
+bus_speed:
+	/*
+	 * Set the bus speed to the selected bus timing.
+	 * If timing is not selected, backward compatible is the default.
+	 */
+	mmc_set_bus_speed(card);
+	return err;
+}
+
+/*
+ * Execute tuning sequence to seek the proper bus operating
+ * conditions for HS200 and HS400, which sends CMD21 to the device.
+ */
+static int mmc_hs200_tuning(struct mmc_card *card)
+{
+	struct mmc_host *host = card->host;
+	int err = 0;
+
+	/*
+	 * Timing should be adjusted to the HS400 target
+	 * operation frequency for tuning process
+	 */
+	if (card->mmc_avail_type & EXT_CSD_CARD_TYPE_HS400 &&
+	    host->ios.bus_width == MMC_BUS_WIDTH_8)
+		if (host->ops->prepare_hs400_tuning)
+			host->ops->prepare_hs400_tuning(host, &host->ios);
+
+	if (host->ops->execute_tuning) {
+		mmc_host_clk_hold(host);
+		err = host->ops->execute_tuning(host,
+				MMC_SEND_TUNING_BLOCK_HS200);
+		mmc_host_clk_release(host);
+
+		if (err)
+			pr_warn("%s: tuning execution failed\n",
+				mmc_hostname(host));
+	}
+
+	return err;
+}
 
 /*
  * Handle the detection and initialisation of a card.
@@ -297,21 +1179,27 @@ static int mmc_init_card(struct mmc_host *host, u32 ocr,
 	struct mmc_card *card;
 	int err;
 	u32 cid[4];
-	unsigned int max_dtr;
+	u32 rocr;
+	u8 *ext_csd = NULL;
 
 	BUG_ON(!host);
 	WARN_ON(!host->claimed);
 
+	/* Set correct bus mode for MMC before attempting init */
+	if (!mmc_host_is_spi(host))
+		mmc_set_bus_mode(host, MMC_BUSMODE_OPENDRAIN);
+
 	/*
 	 * Since we're changing the OCR value, we seem to
 	 * need to tell some cards to go back to the idle
 	 * state.  We wait 1ms to give cards time to
 	 * respond.
+	 * mmc_go_idle is needed for eMMC that are asleep
 	 */
 	mmc_go_idle(host);
 
 	/* The extra bit indicates that we support high capacity */
-	err = mmc_send_op_cond(host, ocr | (1 << 30), NULL);
+	err = mmc_send_op_cond(host, ocr | (1 << 30), &rocr);
 	if (err)
 		goto err;
 
@@ -351,6 +1239,7 @@ static int mmc_init_card(struct mmc_host *host, u32 ocr,
 			goto err;
 		}
 
+		card->ocr = ocr;
 		card->type = MMC_TYPE_MMC;
 		card->rca = 1;
 		memcpy(card->raw_cid, cid, sizeof(card->raw_cid));
@@ -396,73 +1285,274 @@ static int mmc_init_card(struct mmc_host *host, u32 ocr,
 		/*
 		 * Fetch and process extended CSD.
 		 */
-		err = mmc_read_ext_csd(card);
+
+		err = mmc_get_ext_csd(card, &ext_csd);
+		if (err)
+			goto free_card;
+		err = mmc_read_ext_csd(card, ext_csd);
 		if (err)
 			goto free_card;
+
+		/* If doing byte addressing, check if required to do sector
+		 * addressing.  Handle the case of <2GB cards needing sector
+		 * addressing.  See section 8.1 JEDEC Standard JED84-A441;
+		 * ocr register has bit 30 set for sector addressing.
+		 */
+		if (!(mmc_card_blockaddr(card)) && (rocr & (1<<30)))
+			mmc_card_set_blockaddr(card);
+
+		/* Erase size depends on CSD and Extended CSD */
+		mmc_set_erase_size(card);
 	}
 
 	/*
-	 * Activate high speed (if supported)
+	 * If enhanced_area_en is TRUE, host needs to enable ERASE_GRP_DEF
+	 * bit.  This bit will be lost every time after a reset or power off.
 	 */
-	if ((card->ext_csd.hs_max_dtr != 0) &&
-		(host->caps & MMC_CAP_MMC_HIGHSPEED)) {
+	if (card->ext_csd.enhanced_area_en ||
+	    (card->ext_csd.rev >= 3 && (host->caps2 & MMC_CAP2_HC_ERASE_SZ))) {
 		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
-			EXT_CSD_HS_TIMING, 1);
-		if (err)
+				 EXT_CSD_ERASE_GROUP_DEF, 1,
+				 card->ext_csd.generic_cmd6_time);
+
+		if (err && err != -EBADMSG)
 			goto free_card;
 
-		mmc_card_set_highspeed(card);
+		if (err) {
+			err = 0;
+			/*
+			 * Just disable enhanced area off & sz
+			 * will try to enable ERASE_GROUP_DEF
+			 * during next time reinit
+			 */
+			card->ext_csd.enhanced_area_offset = -EINVAL;
+			card->ext_csd.enhanced_area_size = -EINVAL;
+		} else {
+			card->ext_csd.erase_group_def = 1;
+			/*
+			 * enable ERASE_GRP_DEF successfully.
+			 * This will affect the erase size, so
+			 * here need to reset erase size
+			 */
+			mmc_set_erase_size(card);
+		}
+	}
 
-		mmc_set_timing(card->host, MMC_TIMING_MMC_HS);
+	/*
+	 * Ensure eMMC user default partition is enabled
+	 */
+	if (card->ext_csd.part_config & EXT_CSD_PART_CONFIG_ACC_MASK) {
+		card->ext_csd.part_config &= ~EXT_CSD_PART_CONFIG_ACC_MASK;
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_PART_CONFIG,
+				 card->ext_csd.part_config,
+				 card->ext_csd.part_time);
+		if (err && err != -EBADMSG)
+			goto free_card;
 	}
 
 	/*
-	 * Compute bus speed.
+	 * Enable power_off_notification byte in the ext_csd register
 	 */
-	max_dtr = (unsigned int)-1;
+	if (card->ext_csd.rev >= 6) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				 EXT_CSD_POWER_OFF_NOTIFICATION,
+				 EXT_CSD_POWER_ON,
+				 card->ext_csd.generic_cmd6_time);
+		if (err && err != -EBADMSG)
+			goto free_card;
 
-	if (mmc_card_highspeed(card)) {
-		if (max_dtr > card->ext_csd.hs_max_dtr)
-			max_dtr = card->ext_csd.hs_max_dtr;
-	} else if (max_dtr > card->csd.max_dtr) {
-		max_dtr = card->csd.max_dtr;
+		/*
+		 * The err can be -EBADMSG or 0,
+		 * so check for success and update the flag
+		 */
+		if (!err)
+			card->ext_csd.power_off_notification = EXT_CSD_POWER_ON;
 	}
 
-	mmc_set_clock(host, max_dtr);
-
 	/*
-	 * Activate wide bus (if supported).
+	 * Select timing interface
 	 */
-	if ((card->csd.mmca_vsn >= CSD_SPEC_VER_4) &&
-	    (host->caps & (MMC_CAP_4_BIT_DATA | MMC_CAP_8_BIT_DATA))) {
-		unsigned ext_csd_bit, bus_width;
+	err = mmc_select_timing(card);
+	if (err)
+		goto free_card;
 
-		if (host->caps & MMC_CAP_8_BIT_DATA) {
-			ext_csd_bit = EXT_CSD_BUS_WIDTH_8;
-			bus_width = MMC_BUS_WIDTH_8;
-		} else {
-			ext_csd_bit = EXT_CSD_BUS_WIDTH_4;
-			bus_width = MMC_BUS_WIDTH_4;
+	if (mmc_card_hs200(card)) {
+		err = mmc_hs200_tuning(card);
+		if (err)
+			goto err;
+
+		err = mmc_select_hs400(card);
+		if (err)
+			goto err;
+	} else if (mmc_card_hs(card)) {
+		/* Select the desired bus width optionally */
+		err = mmc_select_bus_width(card);
+		if (!IS_ERR_VALUE(err)) {
+			err = mmc_select_hs_ddr(card);
+			if (err)
+				goto err;
 		}
+	}
 
+	/*
+	 * Choose the power class with selected bus interface
+	 */
+	mmc_select_powerclass(card);
+
+	/*
+	 * Enable HPI feature (if supported)
+	 */
+	if (card->ext_csd.hpi) {
 		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
-				 EXT_CSD_BUS_WIDTH, ext_csd_bit);
+				EXT_CSD_HPI_MGMT, 1,
+				card->ext_csd.generic_cmd6_time);
+		if (err && err != -EBADMSG)
+			goto free_card;
+		if (err) {
+			pr_warning("%s: Enabling HPI failed\n",
+				   mmc_hostname(card->host));
+			err = 0;
+		} else
+			card->ext_csd.hpi_en = 1;
+	}
 
-		if (err)
+	/*
+	 * If cache size is higher than 0, this indicates
+	 * the existence of cache and it can be turned on.
+	 */
+	if (card->ext_csd.cache_size > 0) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				EXT_CSD_CACHE_CTRL, 1,
+				card->ext_csd.generic_cmd6_time);
+		if (err && err != -EBADMSG)
 			goto free_card;
 
-		mmc_set_bus_width(card->host, bus_width);
+		/*
+		 * Only if no error, cache is turned on successfully.
+		 */
+		if (err) {
+			pr_warning("%s: Cache is supported, "
+					"but failed to turn on (%d)\n",
+					mmc_hostname(card->host), err);
+			card->ext_csd.cache_ctrl = 0;
+			err = 0;
+		} else {
+			card->ext_csd.cache_ctrl = 1;
+		}
+	}
+
+	/*
+	 * The mandatory minimum values are defined for packed command.
+	 * read: 5, write: 3
+	 */
+	if (card->ext_csd.max_packed_writes >= 3 &&
+	    card->ext_csd.max_packed_reads >= 5 &&
+	    host->caps2 & MMC_CAP2_PACKED_CMD) {
+		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+				EXT_CSD_EXP_EVENTS_CTRL,
+				EXT_CSD_PACKED_EVENT_EN,
+				card->ext_csd.generic_cmd6_time);
+		if (err && err != -EBADMSG)
+			goto free_card;
+		if (err) {
+			pr_warn("%s: Enabling packed event failed\n",
+				mmc_hostname(card->host));
+			card->ext_csd.packed_event_en = 0;
+			err = 0;
+		} else {
+			card->ext_csd.packed_event_en = 1;
+		}
 	}
 
 	if (!oldcard)
 		host->card = card;
 
+	mmc_free_ext_csd(ext_csd);
 	return 0;
 
 free_card:
 	if (!oldcard)
 		mmc_remove_card(card);
 err:
+	mmc_free_ext_csd(ext_csd);
+
+	return err;
+}
+
+static int mmc_can_sleep(struct mmc_card *card)
+{
+	return (card && card->ext_csd.rev >= 3);
+}
+
+static int mmc_sleep(struct mmc_host *host)
+{
+	struct mmc_command cmd = {0};
+	struct mmc_card *card = host->card;
+	unsigned int timeout_ms = DIV_ROUND_UP(card->ext_csd.sa_timeout, 10000);
+	int err;
+
+	err = mmc_deselect_cards(host);
+	if (err)
+		return err;
+
+	cmd.opcode = MMC_SLEEP_AWAKE;
+	cmd.arg = card->rca << 16;
+	cmd.arg |= 1 << 15;
+
+	/*
+	 * If the max_busy_timeout of the host is specified, validate it against
+	 * the sleep cmd timeout. A failure means we need to prevent the host
+	 * from doing hw busy detection, which is done by converting to a R1
+	 * response instead of a R1B.
+	 */
+	if (host->max_busy_timeout && (timeout_ms > host->max_busy_timeout)) {
+		cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
+	} else {
+		cmd.flags = MMC_RSP_R1B | MMC_CMD_AC;
+		cmd.busy_timeout = timeout_ms;
+	}
+
+	err = mmc_wait_for_cmd(host, &cmd, 0);
+	if (err)
+		return err;
+
+	/*
+	 * If the host does not wait while the card signals busy, then we will
+	 * will have to wait the sleep/awake timeout.  Note, we cannot use the
+	 * SEND_STATUS command to poll the status because that command (and most
+	 * others) is invalid while the card sleeps.
+	 */
+	if (!cmd.busy_timeout || !(host->caps & MMC_CAP_WAIT_WHILE_BUSY))
+		mmc_delay(timeout_ms);
+
+	return err;
+}
+
+static int mmc_can_poweroff_notify(const struct mmc_card *card)
+{
+	return card &&
+		mmc_card_mmc(card) &&
+		(card->ext_csd.power_off_notification == EXT_CSD_POWER_ON);
+}
+
+static int mmc_poweroff_notify(struct mmc_card *card, unsigned int notify_type)
+{
+	unsigned int timeout = card->ext_csd.generic_cmd6_time;
+	int err;
+
+	/* Use EXT_CSD_POWER_OFF_SHORT as default notification type. */
+	if (notify_type == EXT_CSD_POWER_OFF_LONG)
+		timeout = card->ext_csd.power_off_longtime;
+
+	err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
+			EXT_CSD_POWER_OFF_NOTIFICATION,
+			notify_type, timeout, true, false, false);
+	if (err)
+		pr_err("%s: Power Off Notification timed out, %u\n",
+		       mmc_hostname(card->host), timeout);
+
+	/* Disable the power off notification after the switch operation. */
+	card->ext_csd.power_off_notification = EXT_CSD_NO_POWER_NOTIFICATION;
 
 	return err;
 }
@@ -480,6 +1570,14 @@ static void mmc_remove(struct mmc_host *host)
 }
 
 /*
+ * Card detection - card is alive.
+ */
+static int mmc_alive(struct mmc_host *host)
+{
+	return mmc_send_status(host->card, NULL);
+}
+
+/*
  * Card detection callback from host.
  */
 static void mmc_detect(struct mmc_host *host)
@@ -489,93 +1587,226 @@ static void mmc_detect(struct mmc_host *host)
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
-	mmc_claim_host(host);
+	mmc_get_card(host->card);
 
 	/*
 	 * Just check if our card has been removed.
 	 */
-	err = mmc_send_status(host->card, NULL);
+	err = _mmc_detect_card_removed(host);
 
-	mmc_release_host(host);
+	mmc_put_card(host->card);
 
 	if (err) {
 		mmc_remove(host);
 
 		mmc_claim_host(host);
 		mmc_detach_bus(host);
+		mmc_power_off(host);
 		mmc_release_host(host);
 	}
 }
 
-#ifdef CONFIG_MMC_UNSAFE_RESUME
-
-/*
- * Suspend callback from host.
- */
-static void mmc_suspend(struct mmc_host *host)
+static int _mmc_suspend(struct mmc_host *host, bool is_suspend)
 {
+	int err = 0;
+	unsigned int notify_type = is_suspend ? EXT_CSD_POWER_OFF_SHORT :
+					EXT_CSD_POWER_OFF_LONG;
+
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
 	mmc_claim_host(host);
-	if (!mmc_host_is_spi(host))
-		mmc_deselect_cards(host);
-	host->card->state &= ~MMC_STATE_HIGHSPEED;
+
+	if (mmc_card_suspended(host->card))
+		goto out;
+
+	if (mmc_card_doing_bkops(host->card)) {
+		err = mmc_stop_bkops(host->card);
+		if (err)
+			goto out;
+	}
+
+	err = mmc_flush_cache(host->card);
+	if (err)
+		goto out;
+
+	if (mmc_can_poweroff_notify(host->card) &&
+		((host->caps2 & MMC_CAP2_FULL_PWR_CYCLE) || !is_suspend))
+		err = mmc_poweroff_notify(host->card, notify_type);
+	else if (mmc_can_sleep(host->card))
+		err = mmc_sleep(host);
+	else if (!mmc_host_is_spi(host))
+		err = mmc_deselect_cards(host);
+
+	if (!err) {
+		mmc_power_off(host);
+		mmc_card_set_suspended(host->card);
+	}
+out:
 	mmc_release_host(host);
+	return err;
+}
+
+/*
+ * Suspend callback
+ */
+static int mmc_suspend(struct mmc_host *host)
+{
+	int err;
+
+	err = _mmc_suspend(host, true);
+	if (!err) {
+		pm_runtime_disable(&host->card->dev);
+		pm_runtime_set_suspended(&host->card->dev);
+	}
+
+	return err;
 }
 
 /*
- * Resume callback from host.
- *
  * This function tries to determine if the same card is still present
  * and, if so, restore all state to it.
  */
-static void mmc_resume(struct mmc_host *host)
+static int _mmc_resume(struct mmc_host *host)
 {
-	int err;
+	int err = 0;
 
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
 	mmc_claim_host(host);
-	err = mmc_init_card(host, host->ocr, host->card);
+
+	if (!mmc_card_suspended(host->card))
+		goto out;
+
+	mmc_power_up(host, host->card->ocr);
+	err = mmc_init_card(host, host->card->ocr, host->card);
+	mmc_card_clr_suspended(host->card);
+
+out:
 	mmc_release_host(host);
+	return err;
+}
 
-	if (err) {
-		mmc_remove(host);
+/*
+ * Shutdown callback
+ */
+static int mmc_shutdown(struct mmc_host *host)
+{
+	int err = 0;
 
-		mmc_claim_host(host);
-		mmc_detach_bus(host);
-		mmc_release_host(host);
+	/*
+	 * In a specific case for poweroff notify, we need to resume the card
+	 * before we can shutdown it properly.
+	 */
+	if (mmc_can_poweroff_notify(host->card) &&
+		!(host->caps2 & MMC_CAP2_FULL_PWR_CYCLE))
+		err = _mmc_resume(host);
+
+	if (!err)
+		err = _mmc_suspend(host, false);
+
+	return err;
+}
+
+/*
+ * Callback for resume.
+ */
+static int mmc_resume(struct mmc_host *host)
+{
+	int err = 0;
+
+	if (!(host->caps & MMC_CAP_RUNTIME_RESUME)) {
+		err = _mmc_resume(host);
+		pm_runtime_set_active(&host->card->dev);
+		pm_runtime_mark_last_busy(&host->card->dev);
 	}
+	pm_runtime_enable(&host->card->dev);
 
+	return err;
 }
 
-#else
+/*
+ * Callback for runtime_suspend.
+ */
+static int mmc_runtime_suspend(struct mmc_host *host)
+{
+	int err;
 
-#define mmc_suspend NULL
-#define mmc_resume NULL
+	if (!(host->caps & MMC_CAP_AGGRESSIVE_PM))
+		return 0;
 
-#endif
+	err = _mmc_suspend(host, true);
+	if (err)
+		pr_err("%s: error %d doing aggessive suspend\n",
+			mmc_hostname(host), err);
+
+	return err;
+}
+
+/*
+ * Callback for runtime_resume.
+ */
+static int mmc_runtime_resume(struct mmc_host *host)
+{
+	int err;
+
+	if (!(host->caps & (MMC_CAP_AGGRESSIVE_PM | MMC_CAP_RUNTIME_RESUME)))
+		return 0;
+
+	err = _mmc_resume(host);
+	if (err)
+		pr_err("%s: error %d doing aggessive resume\n",
+			mmc_hostname(host), err);
+
+	return 0;
+}
+
+static int mmc_power_restore(struct mmc_host *host)
+{
+	int ret;
+
+	mmc_claim_host(host);
+	ret = mmc_init_card(host, host->card->ocr, host->card);
+	mmc_release_host(host);
+
+	return ret;
+}
 
 static const struct mmc_bus_ops mmc_ops = {
 	.remove = mmc_remove,
 	.detect = mmc_detect,
 	.suspend = mmc_suspend,
 	.resume = mmc_resume,
+	.runtime_suspend = mmc_runtime_suspend,
+	.runtime_resume = mmc_runtime_resume,
+	.power_restore = mmc_power_restore,
+	.alive = mmc_alive,
+	.shutdown = mmc_shutdown,
 };
 
 /*
  * Starting point for MMC card init.
  */
-int mmc_attach_mmc(struct mmc_host *host, u32 ocr)
+int mmc_attach_mmc(struct mmc_host *host)
 {
 	int err;
+	u32 ocr, rocr;
 
 	BUG_ON(!host);
 	WARN_ON(!host->claimed);
 
+	/* Set correct bus mode for MMC before attempting attach */
+	if (!mmc_host_is_spi(host))
+		mmc_set_bus_mode(host, MMC_BUSMODE_OPENDRAIN);
+
+	err = mmc_send_op_cond(host, 0, &ocr);
+	if (err)
+		return err;
+
 	mmc_attach_bus(host, &mmc_ops);
+	if (host->ocr_avail_mmc)
+		host->ocr_avail = host->ocr_avail_mmc;
 
 	/*
 	 * We need to get OCR a different way for SPI.
@@ -586,23 +1817,12 @@ int mmc_attach_mmc(struct mmc_host *host, u32 ocr)
 			goto err;
 	}
 
-	/*
-	 * Sanity check the voltages that the card claims to
-	 * support.
-	 */
-	if (ocr & 0x7F) {
-		printk(KERN_WARNING "%s: card claims to support voltages "
-		       "below the defined range. These will be ignored.\n",
-		       mmc_hostname(host));
-		ocr &= ~0x7F;
-	}
-
-	host->ocr = mmc_select_voltage(host, ocr);
+	rocr = mmc_select_voltage(host, ocr);
 
 	/*
 	 * Can we support the voltage of the card?
 	 */
-	if (!host->ocr) {
+	if (!rocr) {
 		err = -EINVAL;
 		goto err;
 	}
@@ -610,27 +1830,27 @@ int mmc_attach_mmc(struct mmc_host *host, u32 ocr)
 	/*
 	 * Detect and init the card.
 	 */
-	err = mmc_init_card(host, host->ocr, NULL);
+	err = mmc_init_card(host, rocr, NULL);
 	if (err)
 		goto err;
 
 	mmc_release_host(host);
-
 	err = mmc_add_card(host->card);
+	mmc_claim_host(host);
 	if (err)
 		goto remove_card;
 
 	return 0;
 
 remove_card:
+	mmc_release_host(host);
 	mmc_remove_card(host->card);
-	host->card = NULL;
 	mmc_claim_host(host);
+	host->card = NULL;
 err:
 	mmc_detach_bus(host);
-	mmc_release_host(host);
 
-	printk(KERN_ERR "%s: error %d whilst initialising MMC card\n",
+	pr_err("%s: error %d whilst initialising MMC card\n",
 		mmc_hostname(host), err);
 
 	return err;
diff --git a/drivers/mmc/core/mmc_ops.c b/drivers/mmc/core/mmc_ops.c
index 34ce2703d29..f51b5ba3bbe 100644
--- a/drivers/mmc/core/mmc_ops.c
+++ b/drivers/mmc/core/mmc_ops.c
@@ -9,6 +9,8 @@
  * your option) any later version.
  */
 
+#include <linux/slab.h>
+#include <linux/export.h>
 #include <linux/types.h>
 #include <linux/scatterlist.h>
 
@@ -19,15 +21,49 @@
 #include "core.h"
 #include "mmc_ops.h"
 
+#define MMC_OPS_TIMEOUT_MS	(10 * 60 * 1000) /* 10 minute timeout */
+
+static inline int __mmc_send_status(struct mmc_card *card, u32 *status,
+				    bool ignore_crc)
+{
+	int err;
+	struct mmc_command cmd = {0};
+
+	BUG_ON(!card);
+	BUG_ON(!card->host);
+
+	cmd.opcode = MMC_SEND_STATUS;
+	if (!mmc_host_is_spi(card->host))
+		cmd.arg = card->rca << 16;
+	cmd.flags = MMC_RSP_SPI_R2 | MMC_RSP_R1 | MMC_CMD_AC;
+	if (ignore_crc)
+		cmd.flags &= ~MMC_RSP_CRC;
+
+	err = mmc_wait_for_cmd(card->host, &cmd, MMC_CMD_RETRIES);
+	if (err)
+		return err;
+
+	/* NOTE: callers are required to understand the difference
+	 * between "native" and SPI format status words!
+	 */
+	if (status)
+		*status = cmd.resp[0];
+
+	return 0;
+}
+
+int mmc_send_status(struct mmc_card *card, u32 *status)
+{
+	return __mmc_send_status(card, status, false);
+}
+
 static int _mmc_select_card(struct mmc_host *host, struct mmc_card *card)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_SELECT_CARD;
 
 	if (card) {
@@ -60,7 +96,7 @@ int mmc_deselect_cards(struct mmc_host *host)
 int mmc_go_idle(struct mmc_host *host)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	/*
 	 * Non-SPI hosts need to prevent chipselect going active during
@@ -68,7 +104,7 @@ int mmc_go_idle(struct mmc_host *host)
 	 * that in case of hardware that won't pull up DAT3/nCS otherwise.
 	 *
 	 * SPI hosts ignore ios.chip_select; it's managed according to
-	 * rules that must accomodate non-MMC slaves which this layer
+	 * rules that must accommodate non-MMC slaves which this layer
 	 * won't even know about.
 	 */
 	if (!mmc_host_is_spi(host)) {
@@ -76,8 +112,6 @@ int mmc_go_idle(struct mmc_host *host)
 		mmc_delay(1);
 	}
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_GO_IDLE_STATE;
 	cmd.arg = 0;
 	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_NONE | MMC_CMD_BC;
@@ -98,13 +132,11 @@ int mmc_go_idle(struct mmc_host *host)
 
 int mmc_send_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int i, err = 0;
 
 	BUG_ON(!host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_SEND_OP_COND;
 	cmd.arg = mmc_host_is_spi(host) ? 0 : ocr;
 	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R3 | MMC_CMD_BCR;
@@ -141,13 +173,11 @@ int mmc_send_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 int mmc_all_send_cid(struct mmc_host *host, u32 *cid)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!host);
 	BUG_ON(!cid);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_ALL_SEND_CID;
 	cmd.arg = 0;
 	cmd.flags = MMC_RSP_R2 | MMC_CMD_BCR;
@@ -164,13 +194,11 @@ int mmc_all_send_cid(struct mmc_host *host, u32 *cid)
 int mmc_set_relative_addr(struct mmc_card *card)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!card);
 	BUG_ON(!card->host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_SET_RELATIVE_ADDR;
 	cmd.arg = card->rca << 16;
 	cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
@@ -186,13 +214,11 @@ static int
 mmc_send_cxd_native(struct mmc_host *host, u32 arg, u32 *cxd, int opcode)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!host);
 	BUG_ON(!cxd);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = opcode;
 	cmd.arg = arg;
 	cmd.flags = MMC_RSP_R2 | MMC_CMD_AC;
@@ -206,26 +232,32 @@ mmc_send_cxd_native(struct mmc_host *host, u32 arg, u32 *cxd, int opcode)
 	return 0;
 }
 
+/*
+ * NOTE: void *buf, caller for the buf is required to use DMA-capable
+ * buffer or on-stack buffer (with some overhead in callee).
+ */
 static int
 mmc_send_cxd_data(struct mmc_card *card, struct mmc_host *host,
 		u32 opcode, void *buf, unsigned len)
 {
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_data data;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
 	struct scatterlist sg;
 	void *data_buf;
+	int is_on_stack;
 
-	/* dma onto stack is unsafe/nonportable, but callers to this
-	 * routine normally provide temporary on-stack buffers ...
-	 */
-	data_buf = kmalloc(len, GFP_KERNEL);
-	if (data_buf == NULL)
-		return -ENOMEM;
-
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
+	is_on_stack = object_is_on_stack(buf);
+	if (is_on_stack) {
+		/*
+		 * dma onto stack is unsafe/nonportable, but callers to this
+		 * routine normally provide temporary on-stack buffers ...
+		 */
+		data_buf = kmalloc(len, GFP_KERNEL);
+		if (!data_buf)
+			return -ENOMEM;
+	} else
+		data_buf = buf;
 
 	mrq.cmd = &cmd;
 	mrq.data = &data;
@@ -260,8 +292,10 @@ mmc_send_cxd_data(struct mmc_card *card, struct mmc_host *host,
 
 	mmc_wait_for_req(host, &mrq);
 
-	memcpy(buf, data_buf, len);
-	kfree(data_buf);
+	if (is_on_stack) {
+		memcpy(buf, data_buf, len);
+		kfree(data_buf);
+	}
 
 	if (cmd.error)
 		return cmd.error;
@@ -274,24 +308,32 @@ mmc_send_cxd_data(struct mmc_card *card, struct mmc_host *host,
 int mmc_send_csd(struct mmc_card *card, u32 *csd)
 {
 	int ret, i;
+	u32 *csd_tmp;
 
 	if (!mmc_host_is_spi(card->host))
 		return mmc_send_cxd_native(card->host, card->rca << 16,
 				csd, MMC_SEND_CSD);
 
-	ret = mmc_send_cxd_data(card, card->host, MMC_SEND_CSD, csd, 16);
+	csd_tmp = kmalloc(16, GFP_KERNEL);
+	if (!csd_tmp)
+		return -ENOMEM;
+
+	ret = mmc_send_cxd_data(card, card->host, MMC_SEND_CSD, csd_tmp, 16);
 	if (ret)
-		return ret;
+		goto err;
 
 	for (i = 0;i < 4;i++)
-		csd[i] = be32_to_cpu(csd[i]);
+		csd[i] = be32_to_cpu(csd_tmp[i]);
 
-	return 0;
+err:
+	kfree(csd_tmp);
+	return ret;
 }
 
 int mmc_send_cid(struct mmc_host *host, u32 *cid)
 {
 	int ret, i;
+	u32 *cid_tmp;
 
 	if (!mmc_host_is_spi(host)) {
 		if (!host->card)
@@ -300,14 +342,20 @@ int mmc_send_cid(struct mmc_host *host, u32 *cid)
 				cid, MMC_SEND_CID);
 	}
 
-	ret = mmc_send_cxd_data(NULL, host, MMC_SEND_CID, cid, 16);
+	cid_tmp = kmalloc(16, GFP_KERNEL);
+	if (!cid_tmp)
+		return -ENOMEM;
+
+	ret = mmc_send_cxd_data(NULL, host, MMC_SEND_CID, cid_tmp, 16);
 	if (ret)
-		return ret;
+		goto err;
 
 	for (i = 0;i < 4;i++)
-		cid[i] = be32_to_cpu(cid[i]);
+		cid[i] = be32_to_cpu(cid_tmp[i]);
 
-	return 0;
+err:
+	kfree(cid_tmp);
+	return ret;
 }
 
 int mmc_send_ext_csd(struct mmc_card *card, u8 *ext_csd)
@@ -315,14 +363,13 @@ int mmc_send_ext_csd(struct mmc_card *card, u8 *ext_csd)
 	return mmc_send_cxd_data(card, card->host, MMC_SEND_EXT_CSD,
 			ext_csd, 512);
 }
+EXPORT_SYMBOL_GPL(mmc_send_ext_csd);
 
 int mmc_spi_read_ocr(struct mmc_host *host, int highcap, u32 *ocrp)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int err;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_SPI_READ_OCR;
 	cmd.arg = highcap ? (1 << 30) : 0;
 	cmd.flags = MMC_RSP_SPI_R3;
@@ -335,11 +382,9 @@ int mmc_spi_read_ocr(struct mmc_host *host, int highcap, u32 *ocrp)
 
 int mmc_spi_set_crc(struct mmc_host *host, int use_crc)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int err;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = MMC_SPI_CRC_ON_OFF;
 	cmd.flags = MMC_RSP_SPI_R1;
 	cmd.arg = use_crc;
@@ -350,55 +395,263 @@ int mmc_spi_set_crc(struct mmc_host *host, int use_crc)
 	return err;
 }
 
-int mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value)
+/**
+ *	__mmc_switch - modify EXT_CSD register
+ *	@card: the MMC card associated with the data transfer
+ *	@set: cmd set values
+ *	@index: EXT_CSD register index
+ *	@value: value to program into EXT_CSD register
+ *	@timeout_ms: timeout (ms) for operation performed by register write,
+ *                   timeout of zero implies maximum possible timeout
+ *	@use_busy_signal: use the busy signal as response type
+ *	@send_status: send status cmd to poll for busy
+ *	@ignore_crc: ignore CRC errors when sending status cmd to poll for busy
+ *
+ *	Modifies the EXT_CSD register for selected card.
+ */
+int __mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
+		unsigned int timeout_ms, bool use_busy_signal, bool send_status,
+		bool ignore_crc)
 {
+	struct mmc_host *host = card->host;
 	int err;
-	struct mmc_command cmd;
-
-	BUG_ON(!card);
-	BUG_ON(!card->host);
+	struct mmc_command cmd = {0};
+	unsigned long timeout;
+	u32 status = 0;
+	bool use_r1b_resp = use_busy_signal;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
+	/*
+	 * If the cmd timeout and the max_busy_timeout of the host are both
+	 * specified, let's validate them. A failure means we need to prevent
+	 * the host from doing hw busy detection, which is done by converting
+	 * to a R1 response instead of a R1B.
+	 */
+	if (timeout_ms && host->max_busy_timeout &&
+		(timeout_ms > host->max_busy_timeout))
+		use_r1b_resp = false;
 
 	cmd.opcode = MMC_SWITCH;
 	cmd.arg = (MMC_SWITCH_MODE_WRITE_BYTE << 24) |
 		  (index << 16) |
 		  (value << 8) |
 		  set;
-	cmd.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
+	cmd.flags = MMC_CMD_AC;
+	if (use_r1b_resp) {
+		cmd.flags |= MMC_RSP_SPI_R1B | MMC_RSP_R1B;
+		/*
+		 * A busy_timeout of zero means the host can decide to use
+		 * whatever value it finds suitable.
+		 */
+		cmd.busy_timeout = timeout_ms;
+	} else {
+		cmd.flags |= MMC_RSP_SPI_R1 | MMC_RSP_R1;
+	}
 
-	err = mmc_wait_for_cmd(card->host, &cmd, MMC_CMD_RETRIES);
+	if (index == EXT_CSD_SANITIZE_START)
+		cmd.sanitize_busy = true;
+
+	err = mmc_wait_for_cmd(host, &cmd, MMC_CMD_RETRIES);
 	if (err)
 		return err;
 
+	/* No need to check card status in case of unblocking command */
+	if (!use_busy_signal)
+		return 0;
+
+	/*
+	 * CRC errors shall only be ignored in cases were CMD13 is used to poll
+	 * to detect busy completion.
+	 */
+	if ((host->caps & MMC_CAP_WAIT_WHILE_BUSY) && use_r1b_resp)
+		ignore_crc = false;
+
+	/* We have an unspecified cmd timeout, use the fallback value. */
+	if (!timeout_ms)
+		timeout_ms = MMC_OPS_TIMEOUT_MS;
+
+	/* Must check status to be sure of no errors. */
+	timeout = jiffies + msecs_to_jiffies(timeout_ms);
+	do {
+		if (send_status) {
+			err = __mmc_send_status(card, &status, ignore_crc);
+			if (err)
+				return err;
+		}
+		if ((host->caps & MMC_CAP_WAIT_WHILE_BUSY) && use_r1b_resp)
+			break;
+		if (mmc_host_is_spi(host))
+			break;
+
+		/*
+		 * We are not allowed to issue a status command and the host
+		 * does'nt support MMC_CAP_WAIT_WHILE_BUSY, then we can only
+		 * rely on waiting for the stated timeout to be sufficient.
+		 */
+		if (!send_status) {
+			mmc_delay(timeout_ms);
+			return 0;
+		}
+
+		/* Timeout if the device never leaves the program state. */
+		if (time_after(jiffies, timeout)) {
+			pr_err("%s: Card stuck in programming state! %s\n",
+				mmc_hostname(host), __func__);
+			return -ETIMEDOUT;
+		}
+	} while (R1_CURRENT_STATE(status) == R1_STATE_PRG);
+
+	if (mmc_host_is_spi(host)) {
+		if (status & R1_SPI_ILLEGAL_COMMAND)
+			return -EBADMSG;
+	} else {
+		if (status & 0xFDFFA000)
+			pr_warn("%s: unexpected status %#x after switch\n",
+				mmc_hostname(host), status);
+		if (status & R1_SWITCH_ERROR)
+			return -EBADMSG;
+	}
+
 	return 0;
 }
+EXPORT_SYMBOL_GPL(__mmc_switch);
 
-int mmc_send_status(struct mmc_card *card, u32 *status)
+int mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
+		unsigned int timeout_ms)
 {
-	int err;
-	struct mmc_command cmd;
+	return __mmc_switch(card, set, index, value, timeout_ms, true, true,
+				false);
+}
+EXPORT_SYMBOL_GPL(mmc_switch);
 
-	BUG_ON(!card);
-	BUG_ON(!card->host);
+static int
+mmc_send_bus_test(struct mmc_card *card, struct mmc_host *host, u8 opcode,
+		  u8 len)
+{
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
+	struct scatterlist sg;
+	u8 *data_buf;
+	u8 *test_buf;
+	int i, err;
+	static u8 testdata_8bit[8] = { 0x55, 0xaa, 0, 0, 0, 0, 0, 0 };
+	static u8 testdata_4bit[4] = { 0x5a, 0, 0, 0 };
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
+	/* dma onto stack is unsafe/nonportable, but callers to this
+	 * routine normally provide temporary on-stack buffers ...
+	 */
+	data_buf = kmalloc(len, GFP_KERNEL);
+	if (!data_buf)
+		return -ENOMEM;
 
-	cmd.opcode = MMC_SEND_STATUS;
-	if (!mmc_host_is_spi(card->host))
-		cmd.arg = card->rca << 16;
-	cmd.flags = MMC_RSP_SPI_R2 | MMC_RSP_R1 | MMC_CMD_AC;
+	if (len == 8)
+		test_buf = testdata_8bit;
+	else if (len == 4)
+		test_buf = testdata_4bit;
+	else {
+		pr_err("%s: Invalid bus_width %d\n",
+		       mmc_hostname(host), len);
+		kfree(data_buf);
+		return -EINVAL;
+	}
 
-	err = mmc_wait_for_cmd(card->host, &cmd, MMC_CMD_RETRIES);
-	if (err)
-		return err;
+	if (opcode == MMC_BUS_TEST_W)
+		memcpy(data_buf, test_buf, len);
 
-	/* NOTE: callers are required to understand the difference
-	 * between "native" and SPI format status words!
+	mrq.cmd = &cmd;
+	mrq.data = &data;
+	cmd.opcode = opcode;
+	cmd.arg = 0;
+
+	/* NOTE HACK:  the MMC_RSP_SPI_R1 is always correct here, but we
+	 * rely on callers to never use this with "native" calls for reading
+	 * CSD or CID.  Native versions of those commands use the R2 type,
+	 * not R1 plus a data block.
 	 */
+	cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
+
+	data.blksz = len;
+	data.blocks = 1;
+	if (opcode == MMC_BUS_TEST_R)
+		data.flags = MMC_DATA_READ;
+	else
+		data.flags = MMC_DATA_WRITE;
+
+	data.sg = &sg;
+	data.sg_len = 1;
+	mmc_set_data_timeout(&data, card);
+	sg_init_one(&sg, data_buf, len);
+	mmc_wait_for_req(host, &mrq);
+	err = 0;
+	if (opcode == MMC_BUS_TEST_R) {
+		for (i = 0; i < len / 4; i++)
+			if ((test_buf[i] ^ data_buf[i]) != 0xff) {
+				err = -EIO;
+				break;
+			}
+	}
+	kfree(data_buf);
+
+	if (cmd.error)
+		return cmd.error;
+	if (data.error)
+		return data.error;
+
+	return err;
+}
+
+int mmc_bus_test(struct mmc_card *card, u8 bus_width)
+{
+	int err, width;
+
+	if (bus_width == MMC_BUS_WIDTH_8)
+		width = 8;
+	else if (bus_width == MMC_BUS_WIDTH_4)
+		width = 4;
+	else if (bus_width == MMC_BUS_WIDTH_1)
+		return 0; /* no need for test */
+	else
+		return -EINVAL;
+
+	/*
+	 * Ignore errors from BUS_TEST_W.  BUS_TEST_R will fail if there
+	 * is a problem.  This improves chances that the test will work.
+	 */
+	mmc_send_bus_test(card, card->host, MMC_BUS_TEST_W, width);
+	err = mmc_send_bus_test(card, card->host, MMC_BUS_TEST_R, width);
+	return err;
+}
+
+int mmc_send_hpi_cmd(struct mmc_card *card, u32 *status)
+{
+	struct mmc_command cmd = {0};
+	unsigned int opcode;
+	int err;
+
+	if (!card->ext_csd.hpi) {
+		pr_warning("%s: Card didn't support HPI command\n",
+			   mmc_hostname(card->host));
+		return -EINVAL;
+	}
+
+	opcode = card->ext_csd.hpi_cmd;
+	if (opcode == MMC_STOP_TRANSMISSION)
+		cmd.flags = MMC_RSP_R1B | MMC_CMD_AC;
+	else if (opcode == MMC_SEND_STATUS)
+		cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
+
+	cmd.opcode = opcode;
+	cmd.arg = card->rca << 16 | 1;
+
+	err = mmc_wait_for_cmd(card->host, &cmd, 0);
+	if (err) {
+		pr_warn("%s: error %d interrupting operation. "
+			"HPI command response %#x\n", mmc_hostname(card->host),
+			err, cmd.resp[0]);
+		return err;
+	}
 	if (status)
 		*status = cmd.resp[0];
 
 	return 0;
 }
-
diff --git a/drivers/mmc/core/mmc_ops.h b/drivers/mmc/core/mmc_ops.h
index 17854bf7cf0..80ae9f4e029 100644
--- a/drivers/mmc/core/mmc_ops.h
+++ b/drivers/mmc/core/mmc_ops.h
@@ -20,11 +20,12 @@ int mmc_all_send_cid(struct mmc_host *host, u32 *cid);
 int mmc_set_relative_addr(struct mmc_card *card);
 int mmc_send_csd(struct mmc_card *card, u32 *csd);
 int mmc_send_ext_csd(struct mmc_card *card, u8 *ext_csd);
-int mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value);
 int mmc_send_status(struct mmc_card *card, u32 *status);
 int mmc_send_cid(struct mmc_host *host, u32 *cid);
 int mmc_spi_read_ocr(struct mmc_host *host, int highcap, u32 *ocrp);
 int mmc_spi_set_crc(struct mmc_host *host, int use_crc);
+int mmc_bus_test(struct mmc_card *card, u8 bus_width);
+int mmc_send_hpi_cmd(struct mmc_card *card, u32 *status);
 
 #endif
 
diff --git a/drivers/mmc/core/quirks.c b/drivers/mmc/core/quirks.c
new file mode 100644
index 00000000000..6c36fccaa1e
--- /dev/null
+++ b/drivers/mmc/core/quirks.c
@@ -0,0 +1,99 @@
+/*
+ *  This file contains work-arounds for many known SD/MMC
+ *  and SDIO hardware bugs.
+ *
+ *  Copyright (c) 2011 Andrei Warkentin <andreiw@motorola.com>
+ *  Copyright (c) 2011 Pierre Tardy <tardyp@gmail.com>
+ *  Inspired from pci fixup code:
+ *  Copyright (c) 1999 Martin Mares <mj@ucw.cz>
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/export.h>
+#include <linux/mmc/card.h>
+#include <linux/mmc/sdio_ids.h>
+
+#ifndef SDIO_VENDOR_ID_TI
+#define SDIO_VENDOR_ID_TI		0x0097
+#endif
+
+#ifndef SDIO_DEVICE_ID_TI_WL1271
+#define SDIO_DEVICE_ID_TI_WL1271	0x4076
+#endif
+
+#ifndef SDIO_VENDOR_ID_STE
+#define SDIO_VENDOR_ID_STE		0x0020
+#endif
+
+#ifndef SDIO_DEVICE_ID_STE_CW1200
+#define SDIO_DEVICE_ID_STE_CW1200	0x2280
+#endif
+
+#ifndef SDIO_DEVICE_ID_MARVELL_8797_F0
+#define SDIO_DEVICE_ID_MARVELL_8797_F0	0x9128
+#endif
+
+/*
+ * This hook just adds a quirk for all sdio devices
+ */
+static void add_quirk_for_sdio_devices(struct mmc_card *card, int data)
+{
+	if (mmc_card_sdio(card))
+		card->quirks |= data;
+}
+
+static const struct mmc_fixup mmc_fixup_methods[] = {
+	/* by default sdio devices are considered CLK_GATING broken */
+	/* good cards will be whitelisted as they are tested */
+	SDIO_FIXUP(SDIO_ANY_ID, SDIO_ANY_ID,
+		   add_quirk_for_sdio_devices,
+		   MMC_QUIRK_BROKEN_CLK_GATING),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1271,
+		   remove_quirk, MMC_QUIRK_BROKEN_CLK_GATING),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1271,
+		   add_quirk, MMC_QUIRK_NONSTD_FUNC_IF),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1271,
+		   add_quirk, MMC_QUIRK_DISABLE_CD),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_STE, SDIO_DEVICE_ID_STE_CW1200,
+		   add_quirk, MMC_QUIRK_BROKEN_BYTE_MODE_512),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_MARVELL, SDIO_DEVICE_ID_MARVELL_8797_F0,
+		   add_quirk, MMC_QUIRK_BROKEN_IRQ_POLLING),
+
+	END_FIXUP
+};
+
+void mmc_fixup_device(struct mmc_card *card, const struct mmc_fixup *table)
+{
+	const struct mmc_fixup *f;
+	u64 rev = cid_rev_card(card);
+
+	/* Non-core specific workarounds. */
+	if (!table)
+		table = mmc_fixup_methods;
+
+	for (f = table; f->vendor_fixup; f++) {
+		if ((f->manfid == CID_MANFID_ANY ||
+		     f->manfid == card->cid.manfid) &&
+		    (f->oemid == CID_OEMID_ANY ||
+		     f->oemid == card->cid.oemid) &&
+		    (f->name == CID_NAME_ANY ||
+		     !strncmp(f->name, card->cid.prod_name,
+			      sizeof(card->cid.prod_name))) &&
+		    (f->cis_vendor == card->cis.vendor ||
+		     f->cis_vendor == (u16) SDIO_ANY_ID) &&
+		    (f->cis_device == card->cis.device ||
+		     f->cis_device == (u16) SDIO_ANY_ID) &&
+		    rev >= f->rev_start && rev <= f->rev_end) {
+			dev_dbg(&card->dev, "calling %pF\n", f->vendor_fixup);
+			f->vendor_fixup(card, f->data);
+		}
+	}
+}
+EXPORT_SYMBOL(mmc_fixup_device);
diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
index cd81c395e16..0c44510bf71 100644
--- a/drivers/mmc/core/sd.c
+++ b/drivers/mmc/core/sd.c
@@ -11,6 +11,10 @@
  */
 
 #include <linux/err.h>
+#include <linux/sizes.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/pm_runtime.h>
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/card.h>
@@ -20,6 +24,7 @@
 #include "core.h"
 #include "bus.h"
 #include "mmc_ops.h"
+#include "sd.h"
 #include "sd_ops.h"
 
 static const unsigned int tran_exp[] = {
@@ -41,6 +46,13 @@ static const unsigned int tacc_mant[] = {
 	35,	40,	45,	50,	55,	60,	70,	80,
 };
 
+static const unsigned int sd_au_size[] = {
+	0,		SZ_16K / 512,		SZ_32K / 512,	SZ_64K / 512,
+	SZ_128K / 512,	SZ_256K / 512,		SZ_512K / 512,	SZ_1M / 512,
+	SZ_2M / 512,	SZ_4M / 512,		SZ_8M / 512,	(SZ_8M + SZ_4M) / 512,
+	SZ_16M / 512,	(SZ_16M + SZ_8M) / 512,	SZ_32M / 512,	SZ_64M / 512,
+};
+
 #define UNSTUFF_BITS(resp,start,size)					\
 	({								\
 		const int __size = size;				\
@@ -58,7 +70,7 @@ static const unsigned int tacc_mant[] = {
 /*
  * Given the decoded CSD structure, decode the raw CID to our CID structure.
  */
-static void mmc_decode_cid(struct mmc_card *card)
+void mmc_decode_cid(struct mmc_card *card)
 {
 	u32 *resp = card->raw_cid;
 
@@ -118,10 +130,17 @@ static int mmc_decode_csd(struct mmc_card *card)
 		csd->r2w_factor = UNSTUFF_BITS(resp, 26, 3);
 		csd->write_blkbits = UNSTUFF_BITS(resp, 22, 4);
 		csd->write_partial = UNSTUFF_BITS(resp, 21, 1);
+
+		if (UNSTUFF_BITS(resp, 46, 1)) {
+			csd->erase_size = 1;
+		} else if (csd->write_blkbits >= 9) {
+			csd->erase_size = UNSTUFF_BITS(resp, 39, 7) + 1;
+			csd->erase_size <<= csd->write_blkbits - 9;
+		}
 		break;
 	case 1:
 		/*
-		 * This is a block-addressed SDHC card. Most
+		 * This is a block-addressed SDHC or SDXC card. Most
 		 * interesting fields are unused and have fixed
 		 * values. To avoid getting tripped by buggy cards,
 		 * we assume those fixed values ourselves.
@@ -135,6 +154,11 @@ static int mmc_decode_csd(struct mmc_card *card)
 		e = UNSTUFF_BITS(resp, 96, 3);
 		csd->max_dtr	  = tran_exp[e] * tran_mant[m];
 		csd->cmdclass	  = UNSTUFF_BITS(resp, 84, 12);
+		csd->c_size	  = UNSTUFF_BITS(resp, 48, 22);
+
+		/* SDXC cards have a minimum C_SIZE of 0x00FFFF */
+		if (csd->c_size >= 0xFFFF)
+			mmc_card_set_ext_capacity(card);
 
 		m = UNSTUFF_BITS(resp, 48, 22);
 		csd->capacity     = (1 + m) << 10;
@@ -146,13 +170,16 @@ static int mmc_decode_csd(struct mmc_card *card)
 		csd->r2w_factor = 4; /* Unused */
 		csd->write_blkbits = 9;
 		csd->write_partial = 0;
+		csd->erase_size = 1;
 		break;
 	default:
-		printk(KERN_ERR "%s: unrecognised CSD structure version %d\n",
+		pr_err("%s: unrecognised CSD structure version %d\n",
 			mmc_hostname(card->host), csd_struct);
 		return -EINVAL;
 	}
 
+	card->erase_size = csd->erase_size;
+
 	return 0;
 }
 
@@ -170,18 +197,83 @@ static int mmc_decode_scr(struct mmc_card *card)
 
 	scr_struct = UNSTUFF_BITS(resp, 60, 4);
 	if (scr_struct != 0) {
-		printk(KERN_ERR "%s: unrecognised SCR structure version %d\n",
+		pr_err("%s: unrecognised SCR structure version %d\n",
 			mmc_hostname(card->host), scr_struct);
 		return -EINVAL;
 	}
 
 	scr->sda_vsn = UNSTUFF_BITS(resp, 56, 4);
 	scr->bus_widths = UNSTUFF_BITS(resp, 48, 4);
+	if (scr->sda_vsn == SCR_SPEC_VER_2)
+		/* Check if Physical Layer Spec v3.0 is supported */
+		scr->sda_spec3 = UNSTUFF_BITS(resp, 47, 1);
+
+	if (UNSTUFF_BITS(resp, 55, 1))
+		card->erased_byte = 0xFF;
+	else
+		card->erased_byte = 0x0;
 
+	if (scr->sda_spec3)
+		scr->cmds = UNSTUFF_BITS(resp, 32, 2);
 	return 0;
 }
 
 /*
+ * Fetch and process SD Status register.
+ */
+static int mmc_read_ssr(struct mmc_card *card)
+{
+	unsigned int au, es, et, eo;
+	int err, i;
+	u32 *ssr;
+
+	if (!(card->csd.cmdclass & CCC_APP_SPEC)) {
+		pr_warning("%s: card lacks mandatory SD Status "
+			"function.\n", mmc_hostname(card->host));
+		return 0;
+	}
+
+	ssr = kmalloc(64, GFP_KERNEL);
+	if (!ssr)
+		return -ENOMEM;
+
+	err = mmc_app_sd_status(card, ssr);
+	if (err) {
+		pr_warning("%s: problem reading SD Status "
+			"register.\n", mmc_hostname(card->host));
+		err = 0;
+		goto out;
+	}
+
+	for (i = 0; i < 16; i++)
+		ssr[i] = be32_to_cpu(ssr[i]);
+
+	/*
+	 * UNSTUFF_BITS only works with four u32s so we have to offset the
+	 * bitfield positions accordingly.
+	 */
+	au = UNSTUFF_BITS(ssr, 428 - 384, 4);
+	if (au) {
+		if (au <= 9 || card->scr.sda_spec3) {
+			card->ssr.au = sd_au_size[au];
+			es = UNSTUFF_BITS(ssr, 408 - 384, 16);
+			et = UNSTUFF_BITS(ssr, 402 - 384, 6);
+			if (es && et) {
+				eo = UNSTUFF_BITS(ssr, 400 - 384, 2);
+				card->ssr.erase_timeout = (et * 1000) / es;
+				card->ssr.erase_offset = eo * 1000;
+			}
+		} else {
+			pr_warning("%s: SD Status: Invalid Allocation Unit size.\n",
+				   mmc_hostname(card->host));
+		}
+	}
+out:
+	kfree(ssr);
+	return err;
+}
+
+/*
  * Fetches and decodes switch information
  */
 static int mmc_read_switch(struct mmc_card *card)
@@ -193,7 +285,7 @@ static int mmc_read_switch(struct mmc_card *card)
 		return 0;
 
 	if (!(card->csd.cmdclass & CCC_SWITCH)) {
-		printk(KERN_WARNING "%s: card lacks mandatory switch "
+		pr_warning("%s: card lacks mandatory switch "
 			"function, performance might suffer.\n",
 			mmc_hostname(card->host));
 		return 0;
@@ -203,30 +295,41 @@ static int mmc_read_switch(struct mmc_card *card)
 
 	status = kmalloc(64, GFP_KERNEL);
 	if (!status) {
-		printk(KERN_ERR "%s: could not allocate a buffer for "
-			"switch capabilities.\n", mmc_hostname(card->host));
+		pr_err("%s: could not allocate a buffer for "
+			"switch capabilities.\n",
+			mmc_hostname(card->host));
 		return -ENOMEM;
 	}
 
-	err = mmc_sd_switch(card, 0, 0, 1, status);
+	/*
+	 * Find out the card's support bits with a mode 0 operation.
+	 * The argument does not matter, as the support bits do not
+	 * change with the arguments.
+	 */
+	err = mmc_sd_switch(card, 0, 0, 0, status);
 	if (err) {
 		/*
-		 * We all hosts that cannot perform the command
-		 * to fail more gracefully
+		 * If the host or the card can't do the switch,
+		 * fail more gracefully.
 		 */
-		if (err != -EINVAL)
+		if (err != -EINVAL && err != -ENOSYS && err != -EFAULT)
 			goto out;
 
-		printk(KERN_WARNING "%s: problem reading switch "
-			"capabilities, performance might suffer.\n",
+		pr_warning("%s: problem reading Bus Speed modes.\n",
 			mmc_hostname(card->host));
 		err = 0;
 
 		goto out;
 	}
 
-	if (status[13] & 0x02)
-		card->sw_caps.hs_max_dtr = 50000000;
+	if (status[13] & SD_MODE_HIGH_SPEED)
+		card->sw_caps.hs_max_dtr = HIGH_SPEED_MAX_DTR;
+
+	if (card->scr.sda_spec3) {
+		card->sw_caps.sd3_bus_mode = status[13];
+		/* Driver Strengths supported by the card */
+		card->sw_caps.sd3_drv_type = status[9];
+	}
 
 out:
 	kfree(status);
@@ -237,7 +340,7 @@ out:
 /*
  * Test if the card supports high-speed mode and, if so, switch to it.
  */
-static int mmc_switch_hs(struct mmc_card *card)
+int mmc_sd_switch_hs(struct mmc_card *card)
 {
 	int err;
 	u8 *status;
@@ -258,7 +361,7 @@ static int mmc_switch_hs(struct mmc_card *card)
 
 	status = kmalloc(64, GFP_KERNEL);
 	if (!status) {
-		printk(KERN_ERR "%s: could not allocate a buffer for "
+		pr_err("%s: could not allocate a buffer for "
 			"switch capabilities.\n", mmc_hostname(card->host));
 		return -ENOMEM;
 	}
@@ -268,12 +371,303 @@ static int mmc_switch_hs(struct mmc_card *card)
 		goto out;
 
 	if ((status[16] & 0xF) != 1) {
-		printk(KERN_WARNING "%s: Problem switching card "
+		pr_warning("%s: Problem switching card "
 			"into high-speed mode!\n",
 			mmc_hostname(card->host));
+		err = 0;
 	} else {
-		mmc_card_set_highspeed(card);
-		mmc_set_timing(card->host, MMC_TIMING_SD_HS);
+		err = 1;
+	}
+
+out:
+	kfree(status);
+
+	return err;
+}
+
+static int sd_select_driver_type(struct mmc_card *card, u8 *status)
+{
+	int host_drv_type = SD_DRIVER_TYPE_B;
+	int card_drv_type = SD_DRIVER_TYPE_B;
+	int drive_strength;
+	int err;
+
+	/*
+	 * If the host doesn't support any of the Driver Types A,C or D,
+	 * or there is no board specific handler then default Driver
+	 * Type B is used.
+	 */
+	if (!(card->host->caps & (MMC_CAP_DRIVER_TYPE_A | MMC_CAP_DRIVER_TYPE_C
+	    | MMC_CAP_DRIVER_TYPE_D)))
+		return 0;
+
+	if (!card->host->ops->select_drive_strength)
+		return 0;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_A)
+		host_drv_type |= SD_DRIVER_TYPE_A;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_C)
+		host_drv_type |= SD_DRIVER_TYPE_C;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_D)
+		host_drv_type |= SD_DRIVER_TYPE_D;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_A)
+		card_drv_type |= SD_DRIVER_TYPE_A;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_C)
+		card_drv_type |= SD_DRIVER_TYPE_C;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_D)
+		card_drv_type |= SD_DRIVER_TYPE_D;
+
+	/*
+	 * The drive strength that the hardware can support
+	 * depends on the board design.  Pass the appropriate
+	 * information and let the hardware specific code
+	 * return what is possible given the options
+	 */
+	mmc_host_clk_hold(card->host);
+	drive_strength = card->host->ops->select_drive_strength(
+		card->sw_caps.uhs_max_dtr,
+		host_drv_type, card_drv_type);
+	mmc_host_clk_release(card->host);
+
+	err = mmc_sd_switch(card, 1, 2, drive_strength, status);
+	if (err)
+		return err;
+
+	if ((status[15] & 0xF) != drive_strength) {
+		pr_warning("%s: Problem setting drive strength!\n",
+			mmc_hostname(card->host));
+		return 0;
+	}
+
+	mmc_set_driver_type(card->host, drive_strength);
+
+	return 0;
+}
+
+static void sd_update_bus_speed_mode(struct mmc_card *card)
+{
+	/*
+	 * If the host doesn't support any of the UHS-I modes, fallback on
+	 * default speed.
+	 */
+	if (!mmc_host_uhs(card->host)) {
+		card->sd_bus_speed = 0;
+		return;
+	}
+
+	if ((card->host->caps & MMC_CAP_UHS_SDR104) &&
+	    (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR104)) {
+			card->sd_bus_speed = UHS_SDR104_BUS_SPEED;
+	} else if ((card->host->caps & MMC_CAP_UHS_DDR50) &&
+		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_DDR50)) {
+			card->sd_bus_speed = UHS_DDR50_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50)) && (card->sw_caps.sd3_bus_mode &
+		    SD_MODE_UHS_SDR50)) {
+			card->sd_bus_speed = UHS_SDR50_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25)) &&
+		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR25)) {
+			card->sd_bus_speed = UHS_SDR25_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25 |
+		    MMC_CAP_UHS_SDR12)) && (card->sw_caps.sd3_bus_mode &
+		    SD_MODE_UHS_SDR12)) {
+			card->sd_bus_speed = UHS_SDR12_BUS_SPEED;
+	}
+}
+
+static int sd_set_bus_speed_mode(struct mmc_card *card, u8 *status)
+{
+	int err;
+	unsigned int timing = 0;
+
+	switch (card->sd_bus_speed) {
+	case UHS_SDR104_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR104;
+		card->sw_caps.uhs_max_dtr = UHS_SDR104_MAX_DTR;
+		break;
+	case UHS_DDR50_BUS_SPEED:
+		timing = MMC_TIMING_UHS_DDR50;
+		card->sw_caps.uhs_max_dtr = UHS_DDR50_MAX_DTR;
+		break;
+	case UHS_SDR50_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR50;
+		card->sw_caps.uhs_max_dtr = UHS_SDR50_MAX_DTR;
+		break;
+	case UHS_SDR25_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR25;
+		card->sw_caps.uhs_max_dtr = UHS_SDR25_MAX_DTR;
+		break;
+	case UHS_SDR12_BUS_SPEED:
+		timing = MMC_TIMING_UHS_SDR12;
+		card->sw_caps.uhs_max_dtr = UHS_SDR12_MAX_DTR;
+		break;
+	default:
+		return 0;
+	}
+
+	err = mmc_sd_switch(card, 1, 0, card->sd_bus_speed, status);
+	if (err)
+		return err;
+
+	if ((status[16] & 0xF) != card->sd_bus_speed)
+		pr_warning("%s: Problem setting bus speed mode!\n",
+			mmc_hostname(card->host));
+	else {
+		mmc_set_timing(card->host, timing);
+		mmc_set_clock(card->host, card->sw_caps.uhs_max_dtr);
+	}
+
+	return 0;
+}
+
+/* Get host's max current setting at its current voltage */
+static u32 sd_get_host_max_current(struct mmc_host *host)
+{
+	u32 voltage, max_current;
+
+	voltage = 1 << host->ios.vdd;
+	switch (voltage) {
+	case MMC_VDD_165_195:
+		max_current = host->max_current_180;
+		break;
+	case MMC_VDD_29_30:
+	case MMC_VDD_30_31:
+		max_current = host->max_current_300;
+		break;
+	case MMC_VDD_32_33:
+	case MMC_VDD_33_34:
+		max_current = host->max_current_330;
+		break;
+	default:
+		max_current = 0;
+	}
+
+	return max_current;
+}
+
+static int sd_set_current_limit(struct mmc_card *card, u8 *status)
+{
+	int current_limit = SD_SET_CURRENT_NO_CHANGE;
+	int err;
+	u32 max_current;
+
+	/*
+	 * Current limit switch is only defined for SDR50, SDR104, and DDR50
+	 * bus speed modes. For other bus speed modes, we do not change the
+	 * current limit.
+	 */
+	if ((card->sd_bus_speed != UHS_SDR50_BUS_SPEED) &&
+	    (card->sd_bus_speed != UHS_SDR104_BUS_SPEED) &&
+	    (card->sd_bus_speed != UHS_DDR50_BUS_SPEED))
+		return 0;
+
+	/*
+	 * Host has different current capabilities when operating at
+	 * different voltages, so find out its max current first.
+	 */
+	max_current = sd_get_host_max_current(card->host);
+
+	/*
+	 * We only check host's capability here, if we set a limit that is
+	 * higher than the card's maximum current, the card will be using its
+	 * maximum current, e.g. if the card's maximum current is 300ma, and
+	 * when we set current limit to 200ma, the card will draw 200ma, and
+	 * when we set current limit to 400/600/800ma, the card will draw its
+	 * maximum 300ma from the host.
+	 */
+	if (max_current >= 800)
+		current_limit = SD_SET_CURRENT_LIMIT_800;
+	else if (max_current >= 600)
+		current_limit = SD_SET_CURRENT_LIMIT_600;
+	else if (max_current >= 400)
+		current_limit = SD_SET_CURRENT_LIMIT_400;
+	else if (max_current >= 200)
+		current_limit = SD_SET_CURRENT_LIMIT_200;
+
+	if (current_limit != SD_SET_CURRENT_NO_CHANGE) {
+		err = mmc_sd_switch(card, 1, 3, current_limit, status);
+		if (err)
+			return err;
+
+		if (((status[15] >> 4) & 0x0F) != current_limit)
+			pr_warning("%s: Problem setting current limit!\n",
+				mmc_hostname(card->host));
+
+	}
+
+	return 0;
+}
+
+/*
+ * UHS-I specific initialization procedure
+ */
+static int mmc_sd_init_uhs_card(struct mmc_card *card)
+{
+	int err;
+	u8 *status;
+
+	if (!card->scr.sda_spec3)
+		return 0;
+
+	if (!(card->csd.cmdclass & CCC_SWITCH))
+		return 0;
+
+	status = kmalloc(64, GFP_KERNEL);
+	if (!status) {
+		pr_err("%s: could not allocate a buffer for "
+			"switch capabilities.\n", mmc_hostname(card->host));
+		return -ENOMEM;
+	}
+
+	/* Set 4-bit bus width */
+	if ((card->host->caps & MMC_CAP_4_BIT_DATA) &&
+	    (card->scr.bus_widths & SD_SCR_BUS_WIDTH_4)) {
+		err = mmc_app_set_bus_width(card, MMC_BUS_WIDTH_4);
+		if (err)
+			goto out;
+
+		mmc_set_bus_width(card->host, MMC_BUS_WIDTH_4);
+	}
+
+	/*
+	 * Select the bus speed mode depending on host
+	 * and card capability.
+	 */
+	sd_update_bus_speed_mode(card);
+
+	/* Set the driver strength for the card */
+	err = sd_select_driver_type(card, status);
+	if (err)
+		goto out;
+
+	/* Set current limit for the card */
+	err = sd_set_current_limit(card, status);
+	if (err)
+		goto out;
+
+	/* Set bus speed mode of the card */
+	err = sd_set_bus_speed_mode(card, status);
+	if (err)
+		goto out;
+
+	/*
+	 * SPI mode doesn't define CMD19 and tuning is only valid for SDR50 and
+	 * SDR104 mode SD-cards. Note that tuning is mandatory for SDR104.
+	 */
+	if (!mmc_host_is_spi(card->host) && card->host->ops->execute_tuning &&
+			(card->sd_bus_speed == UHS_SDR50_BUS_SPEED ||
+			 card->sd_bus_speed == UHS_SDR104_BUS_SPEED)) {
+		mmc_host_clk_hold(card->host);
+		err = card->host->ops->execute_tuning(card->host,
+						      MMC_SEND_TUNING_BLOCK);
+		mmc_host_clk_release(card->host);
 	}
 
 out:
@@ -288,6 +682,8 @@ MMC_DEV_ATTR(csd, "%08x%08x%08x%08x\n", card->raw_csd[0], card->raw_csd[1],
 	card->raw_csd[2], card->raw_csd[3]);
 MMC_DEV_ATTR(scr, "%08x%08x\n", card->raw_scr[0], card->raw_scr[1]);
 MMC_DEV_ATTR(date, "%02d/%04d\n", card->cid.month, card->cid.year);
+MMC_DEV_ATTR(erase_size, "%u\n", card->erase_size << 9);
+MMC_DEV_ATTR(preferred_erase_size, "%u\n", card->pref_erase << 9);
 MMC_DEV_ATTR(fwrev, "0x%x\n", card->cid.fwrev);
 MMC_DEV_ATTR(hwrev, "0x%x\n", card->cid.hwrev);
 MMC_DEV_ATTR(manfid, "0x%06x\n", card->cid.manfid);
@@ -301,6 +697,8 @@ static struct attribute *sd_std_attrs[] = {
 	&dev_attr_csd.attr,
 	&dev_attr_scr.attr,
 	&dev_attr_date.attr,
+	&dev_attr_erase_size.attr,
+	&dev_attr_preferred_erase_size.attr,
 	&dev_attr_fwrev.attr,
 	&dev_attr_hwrev.attr,
 	&dev_attr_manfid.attr,
@@ -309,36 +707,28 @@ static struct attribute *sd_std_attrs[] = {
 	&dev_attr_serial.attr,
 	NULL,
 };
+ATTRIBUTE_GROUPS(sd_std);
 
-static struct attribute_group sd_std_attr_group = {
-	.attrs = sd_std_attrs,
-};
-
-static struct attribute_group *sd_attr_groups[] = {
-	&sd_std_attr_group,
-	NULL,
-};
-
-static struct device_type sd_type = {
-	.groups = sd_attr_groups,
+struct device_type sd_type = {
+	.groups = sd_std_groups,
 };
 
 /*
- * Handle the detection and initialisation of a card.
- *
- * In the case of a resume, "oldcard" will contain the card
- * we're trying to reinitialise.
+ * Fetch CID from card.
  */
-static int mmc_sd_init_card(struct mmc_host *host, u32 ocr,
-	struct mmc_card *oldcard)
+int mmc_sd_get_cid(struct mmc_host *host, u32 ocr, u32 *cid, u32 *rocr)
 {
-	struct mmc_card *card;
 	int err;
-	u32 cid[4];
-	unsigned int max_dtr;
-
-	BUG_ON(!host);
-	WARN_ON(!host->claimed);
+	u32 max_current;
+	int retries = 10;
+	u32 pocr = ocr;
+
+try_again:
+	if (!retries) {
+		ocr &= ~SD_OCR_S18R;
+		pr_warning("%s: Skipping voltage switch\n",
+			mmc_hostname(host));
+	}
 
 	/*
 	 * Since we're changing the OCR value, we seem to
@@ -356,96 +746,104 @@ static int mmc_sd_init_card(struct mmc_host *host, u32 ocr,
 	 */
 	err = mmc_send_if_cond(host, ocr);
 	if (!err)
-		ocr |= 1 << 30;
+		ocr |= SD_OCR_CCS;
+
+	/*
+	 * If the host supports one of UHS-I modes, request the card
+	 * to switch to 1.8V signaling level. If the card has failed
+	 * repeatedly to switch however, skip this.
+	 */
+	if (retries && mmc_host_uhs(host))
+		ocr |= SD_OCR_S18R;
+
+	/*
+	 * If the host can supply more than 150mA at current voltage,
+	 * XPC should be set to 1.
+	 */
+	max_current = sd_get_host_max_current(host);
+	if (max_current > 150)
+		ocr |= SD_OCR_XPC;
 
-	err = mmc_send_app_op_cond(host, ocr, NULL);
+	err = mmc_send_app_op_cond(host, ocr, rocr);
 	if (err)
-		goto err;
+		return err;
 
 	/*
-	 * Fetch CID from card.
+	 * In case CCS and S18A in the response is set, start Signal Voltage
+	 * Switch procedure. SPI mode doesn't support CMD11.
 	 */
+	if (!mmc_host_is_spi(host) && rocr &&
+	   ((*rocr & 0x41000000) == 0x41000000)) {
+		err = mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_180,
+					pocr);
+		if (err == -EAGAIN) {
+			retries--;
+			goto try_again;
+		} else if (err) {
+			retries = 0;
+			goto try_again;
+		}
+	}
+
 	if (mmc_host_is_spi(host))
 		err = mmc_send_cid(host, cid);
 	else
 		err = mmc_all_send_cid(host, cid);
-	if (err)
-		goto err;
 
-	if (oldcard) {
-		if (memcmp(cid, oldcard->raw_cid, sizeof(cid)) != 0) {
-			err = -ENOENT;
-			goto err;
-		}
-
-		card = oldcard;
-	} else {
-		/*
-		 * Allocate card structure.
-		 */
-		card = mmc_alloc_card(host, &sd_type);
-		if (IS_ERR(card)) {
-			err = PTR_ERR(card);
-			goto err;
-		}
+	return err;
+}
 
-		card->type = MMC_TYPE_SD;
-		memcpy(card->raw_cid, cid, sizeof(card->raw_cid));
-	}
+int mmc_sd_get_csd(struct mmc_host *host, struct mmc_card *card)
+{
+	int err;
 
 	/*
-	 * For native busses:  get card RCA and quit open drain mode.
+	 * Fetch CSD from card.
 	 */
-	if (!mmc_host_is_spi(host)) {
-		err = mmc_send_relative_addr(host, &card->rca);
-		if (err)
-			goto free_card;
+	err = mmc_send_csd(card, card->raw_csd);
+	if (err)
+		return err;
 
-		mmc_set_bus_mode(host, MMC_BUSMODE_PUSHPULL);
-	}
+	err = mmc_decode_csd(card);
+	if (err)
+		return err;
 
-	if (!oldcard) {
+	return 0;
+}
+
+int mmc_sd_setup_card(struct mmc_host *host, struct mmc_card *card,
+	bool reinit)
+{
+	int err;
+
+	if (!reinit) {
 		/*
-		 * Fetch CSD from card.
+		 * Fetch SCR from card.
 		 */
-		err = mmc_send_csd(card, card->raw_csd);
-		if (err)
-			goto free_card;
-
-		err = mmc_decode_csd(card);
+		err = mmc_app_send_scr(card, card->raw_scr);
 		if (err)
-			goto free_card;
+			return err;
 
-		mmc_decode_cid(card);
-	}
-
-	/*
-	 * Select card, as all following commands rely on that.
-	 */
-	if (!mmc_host_is_spi(host)) {
-		err = mmc_select_card(card);
+		err = mmc_decode_scr(card);
 		if (err)
-			goto free_card;
-	}
+			return err;
 
-	if (!oldcard) {
 		/*
-		 * Fetch SCR from card.
+		 * Fetch and process SD Status register.
 		 */
-		err = mmc_app_send_scr(card, card->raw_scr);
+		err = mmc_read_ssr(card);
 		if (err)
-			goto free_card;
+			return err;
 
-		err = mmc_decode_scr(card);
-		if (err < 0)
-			goto free_card;
+		/* Erase init depends on CSD and SSR */
+		mmc_init_erase(card);
 
 		/*
 		 * Fetch switch information from card.
 		 */
 		err = mmc_read_switch(card);
 		if (err)
-			goto free_card;
+			return err;
 	}
 
 	/*
@@ -457,66 +855,156 @@ static int mmc_sd_init_card(struct mmc_host *host, u32 ocr,
 	if (mmc_host_is_spi(host)) {
 		err = mmc_spi_set_crc(host, use_spi_crc);
 		if (err)
-			goto free_card;
+			return err;
 	}
 
 	/*
-	 * Attempt to change to high-speed (if supported)
+	 * Check if read-only switch is active.
 	 */
-	err = mmc_switch_hs(card);
-	if (err)
-		goto free_card;
+	if (!reinit) {
+		int ro = -1;
 
-	/*
-	 * Compute bus speed.
-	 */
-	max_dtr = (unsigned int)-1;
+		if (host->ops->get_ro) {
+			mmc_host_clk_hold(card->host);
+			ro = host->ops->get_ro(host);
+			mmc_host_clk_release(card->host);
+		}
+
+		if (ro < 0) {
+			pr_warning("%s: host does not "
+				"support reading read-only "
+				"switch. assuming write-enable.\n",
+				mmc_hostname(host));
+		} else if (ro > 0) {
+			mmc_card_set_readonly(card);
+		}
+	}
+
+	return 0;
+}
 
-	if (mmc_card_highspeed(card)) {
+unsigned mmc_sd_get_max_clock(struct mmc_card *card)
+{
+	unsigned max_dtr = (unsigned int)-1;
+
+	if (mmc_card_hs(card)) {
 		if (max_dtr > card->sw_caps.hs_max_dtr)
 			max_dtr = card->sw_caps.hs_max_dtr;
 	} else if (max_dtr > card->csd.max_dtr) {
 		max_dtr = card->csd.max_dtr;
 	}
 
-	mmc_set_clock(host, max_dtr);
+	return max_dtr;
+}
+
+/*
+ * Handle the detection and initialisation of a card.
+ *
+ * In the case of a resume, "oldcard" will contain the card
+ * we're trying to reinitialise.
+ */
+static int mmc_sd_init_card(struct mmc_host *host, u32 ocr,
+	struct mmc_card *oldcard)
+{
+	struct mmc_card *card;
+	int err;
+	u32 cid[4];
+	u32 rocr = 0;
+
+	BUG_ON(!host);
+	WARN_ON(!host->claimed);
+
+	err = mmc_sd_get_cid(host, ocr, cid, &rocr);
+	if (err)
+		return err;
+
+	if (oldcard) {
+		if (memcmp(cid, oldcard->raw_cid, sizeof(cid)) != 0)
+			return -ENOENT;
+
+		card = oldcard;
+	} else {
+		/*
+		 * Allocate card structure.
+		 */
+		card = mmc_alloc_card(host, &sd_type);
+		if (IS_ERR(card))
+			return PTR_ERR(card);
+
+		card->ocr = ocr;
+		card->type = MMC_TYPE_SD;
+		memcpy(card->raw_cid, cid, sizeof(card->raw_cid));
+	}
 
 	/*
-	 * Switch to wider bus (if supported).
+	 * For native busses:  get card RCA and quit open drain mode.
 	 */
-	if ((host->caps & MMC_CAP_4_BIT_DATA) &&
-		(card->scr.bus_widths & SD_SCR_BUS_WIDTH_4)) {
-		err = mmc_app_set_bus_width(card, MMC_BUS_WIDTH_4);
+	if (!mmc_host_is_spi(host)) {
+		err = mmc_send_relative_addr(host, &card->rca);
 		if (err)
 			goto free_card;
+	}
 
-		mmc_set_bus_width(host, MMC_BUS_WIDTH_4);
+	if (!oldcard) {
+		err = mmc_sd_get_csd(host, card);
+		if (err)
+			goto free_card;
+
+		mmc_decode_cid(card);
 	}
 
 	/*
-	 * Check if read-only switch is active.
+	 * Select card, as all following commands rely on that.
 	 */
-	if (!oldcard) {
-		if (!host->ops->get_ro || host->ops->get_ro(host) < 0) {
-			printk(KERN_WARNING "%s: host does not "
-				"support reading read-only "
-				"switch. assuming write-enable.\n",
-				mmc_hostname(host));
-		} else {
-			if (host->ops->get_ro(host) > 0)
-				mmc_card_set_readonly(card);
-		}
+	if (!mmc_host_is_spi(host)) {
+		err = mmc_select_card(card);
+		if (err)
+			goto free_card;
 	}
 
-	if (!oldcard)
-		host->card = card;
+	err = mmc_sd_setup_card(host, card, oldcard != NULL);
+	if (err)
+		goto free_card;
+
+	/* Initialization sequence for UHS-I cards */
+	if (rocr & SD_ROCR_S18A) {
+		err = mmc_sd_init_uhs_card(card);
+		if (err)
+			goto free_card;
+	} else {
+		/*
+		 * Attempt to change to high-speed (if supported)
+		 */
+		err = mmc_sd_switch_hs(card);
+		if (err > 0)
+			mmc_set_timing(card->host, MMC_TIMING_SD_HS);
+		else if (err)
+			goto free_card;
+
+		/*
+		 * Set bus speed.
+		 */
+		mmc_set_clock(host, mmc_sd_get_max_clock(card));
+
+		/*
+		 * Switch to wider bus (if supported).
+		 */
+		if ((host->caps & MMC_CAP_4_BIT_DATA) &&
+			(card->scr.bus_widths & SD_SCR_BUS_WIDTH_4)) {
+			err = mmc_app_set_bus_width(card, MMC_BUS_WIDTH_4);
+			if (err)
+				goto free_card;
+
+			mmc_set_bus_width(host, MMC_BUS_WIDTH_4);
+		}
+	}
 
+	host->card = card;
 	return 0;
 
 free_card:
 	if (!oldcard)
 		mmc_remove_card(card);
-err:
 
 	return err;
 }
@@ -534,6 +1022,14 @@ static void mmc_sd_remove(struct mmc_host *host)
 }
 
 /*
+ * Card detection - card is alive.
+ */
+static int mmc_sd_alive(struct mmc_host *host)
+{
+	return mmc_send_status(host->card, NULL);
+}
+
+/*
  * Card detection callback from host.
  */
 static void mmc_sd_detect(struct mmc_host *host)
@@ -543,93 +1039,185 @@ static void mmc_sd_detect(struct mmc_host *host)
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
-	mmc_claim_host(host);
+	mmc_get_card(host->card);
 
 	/*
 	 * Just check if our card has been removed.
 	 */
-	err = mmc_send_status(host->card, NULL);
+	err = _mmc_detect_card_removed(host);
 
-	mmc_release_host(host);
+	mmc_put_card(host->card);
 
 	if (err) {
 		mmc_sd_remove(host);
 
 		mmc_claim_host(host);
 		mmc_detach_bus(host);
+		mmc_power_off(host);
 		mmc_release_host(host);
 	}
 }
 
-#ifdef CONFIG_MMC_UNSAFE_RESUME
-
-/*
- * Suspend callback from host.
- */
-static void mmc_sd_suspend(struct mmc_host *host)
+static int _mmc_sd_suspend(struct mmc_host *host)
 {
+	int err = 0;
+
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
 	mmc_claim_host(host);
+
+	if (mmc_card_suspended(host->card))
+		goto out;
+
 	if (!mmc_host_is_spi(host))
-		mmc_deselect_cards(host);
-	host->card->state &= ~MMC_STATE_HIGHSPEED;
+		err = mmc_deselect_cards(host);
+
+	if (!err) {
+		mmc_power_off(host);
+		mmc_card_set_suspended(host->card);
+	}
+
+out:
 	mmc_release_host(host);
+	return err;
+}
+
+/*
+ * Callback for suspend
+ */
+static int mmc_sd_suspend(struct mmc_host *host)
+{
+	int err;
+
+	err = _mmc_sd_suspend(host);
+	if (!err) {
+		pm_runtime_disable(&host->card->dev);
+		pm_runtime_set_suspended(&host->card->dev);
+	}
+
+	return err;
 }
 
 /*
- * Resume callback from host.
- *
  * This function tries to determine if the same card is still present
  * and, if so, restore all state to it.
  */
-static void mmc_sd_resume(struct mmc_host *host)
+static int _mmc_sd_resume(struct mmc_host *host)
 {
-	int err;
+	int err = 0;
 
 	BUG_ON(!host);
 	BUG_ON(!host->card);
 
 	mmc_claim_host(host);
-	err = mmc_sd_init_card(host, host->ocr, host->card);
+
+	if (!mmc_card_suspended(host->card))
+		goto out;
+
+	mmc_power_up(host, host->card->ocr);
+	err = mmc_sd_init_card(host, host->card->ocr, host->card);
+	mmc_card_clr_suspended(host->card);
+
+out:
 	mmc_release_host(host);
+	return err;
+}
 
-	if (err) {
-		mmc_sd_remove(host);
+/*
+ * Callback for resume
+ */
+static int mmc_sd_resume(struct mmc_host *host)
+{
+	int err = 0;
 
-		mmc_claim_host(host);
-		mmc_detach_bus(host);
-		mmc_release_host(host);
+	if (!(host->caps & MMC_CAP_RUNTIME_RESUME)) {
+		err = _mmc_sd_resume(host);
+		pm_runtime_set_active(&host->card->dev);
+		pm_runtime_mark_last_busy(&host->card->dev);
 	}
+	pm_runtime_enable(&host->card->dev);
+
+	return err;
+}
+
+/*
+ * Callback for runtime_suspend.
+ */
+static int mmc_sd_runtime_suspend(struct mmc_host *host)
+{
+	int err;
+
+	if (!(host->caps & MMC_CAP_AGGRESSIVE_PM))
+		return 0;
+
+	err = _mmc_sd_suspend(host);
+	if (err)
+		pr_err("%s: error %d doing aggessive suspend\n",
+			mmc_hostname(host), err);
+
+	return err;
+}
+
+/*
+ * Callback for runtime_resume.
+ */
+static int mmc_sd_runtime_resume(struct mmc_host *host)
+{
+	int err;
+
+	if (!(host->caps & (MMC_CAP_AGGRESSIVE_PM | MMC_CAP_RUNTIME_RESUME)))
+		return 0;
+
+	err = _mmc_sd_resume(host);
+	if (err)
+		pr_err("%s: error %d doing aggessive resume\n",
+			mmc_hostname(host), err);
 
+	return 0;
 }
 
-#else
+static int mmc_sd_power_restore(struct mmc_host *host)
+{
+	int ret;
 
-#define mmc_sd_suspend NULL
-#define mmc_sd_resume NULL
+	mmc_claim_host(host);
+	ret = mmc_sd_init_card(host, host->card->ocr, host->card);
+	mmc_release_host(host);
 
-#endif
+	return ret;
+}
 
 static const struct mmc_bus_ops mmc_sd_ops = {
 	.remove = mmc_sd_remove,
 	.detect = mmc_sd_detect,
+	.runtime_suspend = mmc_sd_runtime_suspend,
+	.runtime_resume = mmc_sd_runtime_resume,
 	.suspend = mmc_sd_suspend,
 	.resume = mmc_sd_resume,
+	.power_restore = mmc_sd_power_restore,
+	.alive = mmc_sd_alive,
+	.shutdown = mmc_sd_suspend,
 };
 
 /*
  * Starting point for SD card init.
  */
-int mmc_attach_sd(struct mmc_host *host, u32 ocr)
+int mmc_attach_sd(struct mmc_host *host)
 {
 	int err;
+	u32 ocr, rocr;
 
 	BUG_ON(!host);
 	WARN_ON(!host->claimed);
 
+	err = mmc_send_app_op_cond(host, 0, &ocr);
+	if (err)
+		return err;
+
 	mmc_attach_bus(host, &mmc_sd_ops);
+	if (host->ocr_avail_sd)
+		host->ocr_avail = host->ocr_avail_sd;
 
 	/*
 	 * We need to get OCR a different way for SPI.
@@ -642,30 +1230,12 @@ int mmc_attach_sd(struct mmc_host *host, u32 ocr)
 			goto err;
 	}
 
-	/*
-	 * Sanity check the voltages that the card claims to
-	 * support.
-	 */
-	if (ocr & 0x7F) {
-		printk(KERN_WARNING "%s: card claims to support voltages "
-		       "below the defined range. These will be ignored.\n",
-		       mmc_hostname(host));
-		ocr &= ~0x7F;
-	}
-
-	if (ocr & MMC_VDD_165_195) {
-		printk(KERN_WARNING "%s: SD card claims to support the "
-		       "incompletely defined 'low voltage range'. This "
-		       "will be ignored.\n", mmc_hostname(host));
-		ocr &= ~MMC_VDD_165_195;
-	}
-
-	host->ocr = mmc_select_voltage(host, ocr);
+	rocr = mmc_select_voltage(host, ocr);
 
 	/*
 	 * Can we support the voltage(s) of the card(s)?
 	 */
-	if (!host->ocr) {
+	if (!rocr) {
 		err = -EINVAL;
 		goto err;
 	}
@@ -673,27 +1243,27 @@ int mmc_attach_sd(struct mmc_host *host, u32 ocr)
 	/*
 	 * Detect and init the card.
 	 */
-	err = mmc_sd_init_card(host, host->ocr, NULL);
+	err = mmc_sd_init_card(host, rocr, NULL);
 	if (err)
 		goto err;
 
 	mmc_release_host(host);
-
 	err = mmc_add_card(host->card);
+	mmc_claim_host(host);
 	if (err)
 		goto remove_card;
 
 	return 0;
 
 remove_card:
+	mmc_release_host(host);
 	mmc_remove_card(host->card);
 	host->card = NULL;
 	mmc_claim_host(host);
 err:
 	mmc_detach_bus(host);
-	mmc_release_host(host);
 
-	printk(KERN_ERR "%s: error %d whilst initialising SD card\n",
+	pr_err("%s: error %d whilst initialising SD card\n",
 		mmc_hostname(host), err);
 
 	return err;
diff --git a/drivers/mmc/core/sd.h b/drivers/mmc/core/sd.h
new file mode 100644
index 00000000000..aab824a9a7f
--- /dev/null
+++ b/drivers/mmc/core/sd.h
@@ -0,0 +1,16 @@
+#ifndef _MMC_CORE_SD_H
+#define _MMC_CORE_SD_H
+
+#include <linux/mmc/card.h>
+
+extern struct device_type sd_type;
+
+int mmc_sd_get_cid(struct mmc_host *host, u32 ocr, u32 *cid, u32 *rocr);
+int mmc_sd_get_csd(struct mmc_host *host, struct mmc_card *card);
+void mmc_decode_cid(struct mmc_card *card);
+int mmc_sd_setup_card(struct mmc_host *host, struct mmc_card *card,
+	bool reinit);
+unsigned mmc_sd_get_max_clock(struct mmc_card *card);
+int mmc_sd_switch_hs(struct mmc_card *card);
+
+#endif
diff --git a/drivers/mmc/core/sd_ops.c b/drivers/mmc/core/sd_ops.c
index 0d96080d44b..274ef00b446 100644
--- a/drivers/mmc/core/sd_ops.c
+++ b/drivers/mmc/core/sd_ops.c
@@ -9,7 +9,9 @@
  * your option) any later version.
  */
 
+#include <linux/slab.h>
 #include <linux/types.h>
+#include <linux/export.h>
 #include <linux/scatterlist.h>
 
 #include <linux/mmc/host.h>
@@ -20,10 +22,10 @@
 #include "core.h"
 #include "sd_ops.h"
 
-static int mmc_app_cmd(struct mmc_host *host, struct mmc_card *card)
+int mmc_app_cmd(struct mmc_host *host, struct mmc_card *card)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!host);
 	BUG_ON(card && (card->host != host));
@@ -48,6 +50,7 @@ static int mmc_app_cmd(struct mmc_host *host, struct mmc_card *card)
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(mmc_app_cmd);
 
 /**
  *	mmc_wait_for_app_cmd - start an application command and wait for
@@ -65,7 +68,7 @@ static int mmc_app_cmd(struct mmc_host *host, struct mmc_card *card)
 int mmc_wait_for_app_cmd(struct mmc_host *host, struct mmc_card *card,
 	struct mmc_command *cmd, int retries)
 {
-	struct mmc_request mrq;
+	struct mmc_request mrq = {NULL};
 
 	int i, err;
 
@@ -79,8 +82,6 @@ int mmc_wait_for_app_cmd(struct mmc_host *host, struct mmc_card *card,
 	 * we cannot use the retries field in mmc_command.
 	 */
 	for (i = 0;i <= retries;i++) {
-		memset(&mrq, 0, sizeof(struct mmc_request));
-
 		err = mmc_app_cmd(host, card);
 		if (err) {
 			/* no point in retrying; no APP commands allowed */
@@ -120,13 +121,11 @@ EXPORT_SYMBOL(mmc_wait_for_app_cmd);
 int mmc_app_set_bus_width(struct mmc_card *card, int width)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!card);
 	BUG_ON(!card->host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = SD_APP_SET_BUS_WIDTH;
 	cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
 
@@ -150,13 +149,11 @@ int mmc_app_set_bus_width(struct mmc_card *card, int width)
 
 int mmc_send_app_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int i, err = 0;
 
 	BUG_ON(!host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = SD_APP_OP_COND;
 	if (mmc_host_is_spi(host))
 		cmd.arg = ocr & (1 << 30); /* SPI only defines one bit */
@@ -195,7 +192,7 @@ int mmc_send_app_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 
 int mmc_send_if_cond(struct mmc_host *host, u32 ocr)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int err;
 	static const u8 test_pattern = 0xAA;
 	u8 result_pattern;
@@ -227,13 +224,11 @@ int mmc_send_if_cond(struct mmc_host *host, u32 ocr)
 int mmc_send_relative_addr(struct mmc_host *host, unsigned int *rca)
 {
 	int err;
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 
 	BUG_ON(!host);
 	BUG_ON(!rca);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = SD_SEND_RELATIVE_ADDR;
 	cmd.arg = 0;
 	cmd.flags = MMC_RSP_R6 | MMC_CMD_BCR;
@@ -250,10 +245,11 @@ int mmc_send_relative_addr(struct mmc_host *host, unsigned int *rca)
 int mmc_app_send_scr(struct mmc_card *card, u32 *scr)
 {
 	int err;
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_data data;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
 	struct scatterlist sg;
+	void *data_buf;
 
 	BUG_ON(!card);
 	BUG_ON(!card->host);
@@ -265,9 +261,12 @@ int mmc_app_send_scr(struct mmc_card *card, u32 *scr)
 	if (err)
 		return err;
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
+	/* dma onto stack is unsafe/nonportable, but callers to this
+	 * routine normally provide temporary on-stack buffers ...
+	 */
+	data_buf = kmalloc(sizeof(card->raw_scr), GFP_KERNEL);
+	if (data_buf == NULL)
+		return -ENOMEM;
 
 	mrq.cmd = &cmd;
 	mrq.data = &data;
@@ -282,12 +281,15 @@ int mmc_app_send_scr(struct mmc_card *card, u32 *scr)
 	data.sg = &sg;
 	data.sg_len = 1;
 
-	sg_init_one(&sg, scr, 8);
+	sg_init_one(&sg, data_buf, 8);
 
 	mmc_set_data_timeout(&data, card);
 
 	mmc_wait_for_req(card->host, &mrq);
 
+	memcpy(scr, data_buf, sizeof(card->raw_scr));
+	kfree(data_buf);
+
 	if (cmd.error)
 		return cmd.error;
 	if (data.error)
@@ -302,9 +304,9 @@ int mmc_app_send_scr(struct mmc_card *card, u32 *scr)
 int mmc_sd_switch(struct mmc_card *card, int mode, int group,
 	u8 value, u8 *resp)
 {
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_data data;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
 	struct scatterlist sg;
 
 	BUG_ON(!card);
@@ -315,10 +317,6 @@ int mmc_sd_switch(struct mmc_card *card, int mode, int group,
 	mode = !!mode;
 	value &= 0xF;
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
-
 	mrq.cmd = &cmd;
 	mrq.data = &data;
 
@@ -348,3 +346,47 @@ int mmc_sd_switch(struct mmc_card *card, int mode, int group,
 	return 0;
 }
 
+int mmc_app_sd_status(struct mmc_card *card, void *ssr)
+{
+	int err;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
+	struct scatterlist sg;
+
+	BUG_ON(!card);
+	BUG_ON(!card->host);
+	BUG_ON(!ssr);
+
+	/* NOTE: caller guarantees ssr is heap-allocated */
+
+	err = mmc_app_cmd(card->host, card);
+	if (err)
+		return err;
+
+	mrq.cmd = &cmd;
+	mrq.data = &data;
+
+	cmd.opcode = SD_APP_SD_STATUS;
+	cmd.arg = 0;
+	cmd.flags = MMC_RSP_SPI_R2 | MMC_RSP_R1 | MMC_CMD_ADTC;
+
+	data.blksz = 64;
+	data.blocks = 1;
+	data.flags = MMC_DATA_READ;
+	data.sg = &sg;
+	data.sg_len = 1;
+
+	sg_init_one(&sg, ssr, 64);
+
+	mmc_set_data_timeout(&data, card);
+
+	mmc_wait_for_req(card->host, &mrq);
+
+	if (cmd.error)
+		return cmd.error;
+	if (data.error)
+		return data.error;
+
+	return 0;
+}
diff --git a/drivers/mmc/core/sd_ops.h b/drivers/mmc/core/sd_ops.h
index 9742d8a3066..ffc2305d905 100644
--- a/drivers/mmc/core/sd_ops.h
+++ b/drivers/mmc/core/sd_ops.h
@@ -19,6 +19,7 @@ int mmc_send_relative_addr(struct mmc_host *host, unsigned int *rca);
 int mmc_app_send_scr(struct mmc_card *card, u32 *scr);
 int mmc_sd_switch(struct mmc_card *card, int mode, int group,
 	u8 value, u8 *resp);
+int mmc_app_sd_status(struct mmc_card *card, void *ssr);
 
 #endif
 
diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c
index fb99ccff908..e636d9e99e4 100644
--- a/drivers/mmc/core/sdio.c
+++ b/drivers/mmc/core/sdio.c
@@ -10,14 +10,18 @@
  */
 
 #include <linux/err.h>
+#include <linux/pm_runtime.h>
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/card.h>
+#include <linux/mmc/mmc.h>
 #include <linux/mmc/sdio.h>
 #include <linux/mmc/sdio_func.h>
+#include <linux/mmc/sdio_ids.h>
 
 #include "core.h"
 #include "bus.h"
+#include "sd.h"
 #include "sdio_bus.h"
 #include "mmc_ops.h"
 #include "sd_ops.h"
@@ -29,6 +33,11 @@ static int sdio_read_fbr(struct sdio_func *func)
 	int ret;
 	unsigned char data;
 
+	if (mmc_card_nonstd_func_interface(func->card)) {
+		func->class = SDIO_CLASS_NONE;
+		return 0;
+	}
+
 	ret = mmc_io_rw_direct(func->card, 0, 0,
 		SDIO_FBR_BASE(func->num) + SDIO_FBR_STD_IF, 0, &data);
 	if (ret)
@@ -62,13 +71,19 @@ static int sdio_init_func(struct mmc_card *card, unsigned int fn)
 
 	func->num = fn;
 
-	ret = sdio_read_fbr(func);
-	if (ret)
-		goto fail;
+	if (!(card->quirks & MMC_QUIRK_NONSTD_SDIO)) {
+		ret = sdio_read_fbr(func);
+		if (ret)
+			goto fail;
 
-	ret = sdio_read_func_cis(func);
-	if (ret)
-		goto fail;
+		ret = sdio_read_func_cis(func);
+		if (ret)
+			goto fail;
+	} else {
+		func->vendor = func->card->cis.vendor;
+		func->device = func->card->cis.device;
+		func->max_blksize = func->card->cis.blksize;
+	}
 
 	card->sdio_func[fn - 1] = func;
 
@@ -83,11 +98,13 @@ fail:
 	return ret;
 }
 
-static int sdio_read_cccr(struct mmc_card *card)
+static int sdio_read_cccr(struct mmc_card *card, u32 ocr)
 {
 	int ret;
 	int cccr_vsn;
+	int uhs = ocr & R4_18V_PRESENT;
 	unsigned char data;
+	unsigned char speed;
 
 	memset(&card->cccr, 0, sizeof(struct sdio_cccr));
 
@@ -97,8 +114,8 @@ static int sdio_read_cccr(struct mmc_card *card)
 
 	cccr_vsn = data & 0x0f;
 
-	if (cccr_vsn > SDIO_CCCR_REV_1_20) {
-		printk(KERN_ERR "%s: unrecognised CCCR structure version %d\n",
+	if (cccr_vsn > SDIO_CCCR_REV_3_00) {
+		pr_err("%s: unrecognised CCCR structure version %d\n",
 			mmc_hostname(card->host), cccr_vsn);
 		return -EINVAL;
 	}
@@ -126,12 +143,57 @@ static int sdio_read_cccr(struct mmc_card *card)
 	}
 
 	if (cccr_vsn >= SDIO_CCCR_REV_1_20) {
-		ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_SPEED, 0, &data);
+		ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_SPEED, 0, &speed);
 		if (ret)
 			goto out;
 
-		if (data & SDIO_SPEED_SHS)
-			card->cccr.high_speed = 1;
+		card->scr.sda_spec3 = 0;
+		card->sw_caps.sd3_bus_mode = 0;
+		card->sw_caps.sd3_drv_type = 0;
+		if (cccr_vsn >= SDIO_CCCR_REV_3_00 && uhs) {
+			card->scr.sda_spec3 = 1;
+			ret = mmc_io_rw_direct(card, 0, 0,
+				SDIO_CCCR_UHS, 0, &data);
+			if (ret)
+				goto out;
+
+			if (mmc_host_uhs(card->host)) {
+				if (data & SDIO_UHS_DDR50)
+					card->sw_caps.sd3_bus_mode
+						|= SD_MODE_UHS_DDR50;
+
+				if (data & SDIO_UHS_SDR50)
+					card->sw_caps.sd3_bus_mode
+						|= SD_MODE_UHS_SDR50;
+
+				if (data & SDIO_UHS_SDR104)
+					card->sw_caps.sd3_bus_mode
+						|= SD_MODE_UHS_SDR104;
+			}
+
+			ret = mmc_io_rw_direct(card, 0, 0,
+				SDIO_CCCR_DRIVE_STRENGTH, 0, &data);
+			if (ret)
+				goto out;
+
+			if (data & SDIO_DRIVE_SDTA)
+				card->sw_caps.sd3_drv_type |= SD_DRIVER_TYPE_A;
+			if (data & SDIO_DRIVE_SDTC)
+				card->sw_caps.sd3_drv_type |= SD_DRIVER_TYPE_C;
+			if (data & SDIO_DRIVE_SDTD)
+				card->sw_caps.sd3_drv_type |= SD_DRIVER_TYPE_D;
+		}
+
+		/* if no uhs mode ensure we check for high speed */
+		if (!card->sw_caps.sd3_bus_mode) {
+			if (speed & SDIO_SPEED_SHS) {
+				card->cccr.high_speed = 1;
+				card->sw_caps.hs_max_dtr = 50000000;
+			} else {
+				card->cccr.high_speed = 0;
+				card->sw_caps.hs_max_dtr = 25000000;
+			}
+		}
 	}
 
 out:
@@ -153,21 +215,106 @@ static int sdio_enable_wide(struct mmc_card *card)
 	if (ret)
 		return ret;
 
+	if ((ctrl & SDIO_BUS_WIDTH_MASK) == SDIO_BUS_WIDTH_RESERVED)
+		pr_warning("%s: SDIO_CCCR_IF is invalid: 0x%02x\n",
+			   mmc_hostname(card->host), ctrl);
+
+	/* set as 4-bit bus width */
+	ctrl &= ~SDIO_BUS_WIDTH_MASK;
 	ctrl |= SDIO_BUS_WIDTH_4BIT;
 
 	ret = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_IF, ctrl, NULL);
 	if (ret)
 		return ret;
 
-	mmc_set_bus_width(card->host, MMC_BUS_WIDTH_4);
+	return 1;
+}
+
+/*
+ * If desired, disconnect the pull-up resistor on CD/DAT[3] (pin 1)
+ * of the card. This may be required on certain setups of boards,
+ * controllers and embedded sdio device which do not need the card's
+ * pull-up. As a result, card detection is disabled and power is saved.
+ */
+static int sdio_disable_cd(struct mmc_card *card)
+{
+	int ret;
+	u8 ctrl;
+
+	if (!mmc_card_disable_cd(card))
+		return 0;
+
+	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_IF, 0, &ctrl);
+	if (ret)
+		return ret;
+
+	ctrl |= SDIO_BUS_CD_DISABLE;
+
+	return mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_IF, ctrl, NULL);
+}
+
+/*
+ * Devices that remain active during a system suspend are
+ * put back into 1-bit mode.
+ */
+static int sdio_disable_wide(struct mmc_card *card)
+{
+	int ret;
+	u8 ctrl;
+
+	if (!(card->host->caps & MMC_CAP_4_BIT_DATA))
+		return 0;
+
+	if (card->cccr.low_speed && !card->cccr.wide_bus)
+		return 0;
+
+	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_IF, 0, &ctrl);
+	if (ret)
+		return ret;
+
+	if (!(ctrl & SDIO_BUS_WIDTH_4BIT))
+		return 0;
+
+	ctrl &= ~SDIO_BUS_WIDTH_4BIT;
+	ctrl |= SDIO_BUS_ASYNC_INT;
+
+	ret = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_IF, ctrl, NULL);
+	if (ret)
+		return ret;
+
+	mmc_set_bus_width(card->host, MMC_BUS_WIDTH_1);
 
 	return 0;
 }
 
+
+static int sdio_enable_4bit_bus(struct mmc_card *card)
+{
+	int err;
+
+	if (card->type == MMC_TYPE_SDIO)
+		return sdio_enable_wide(card);
+
+	if ((card->host->caps & MMC_CAP_4_BIT_DATA) &&
+		(card->scr.bus_widths & SD_SCR_BUS_WIDTH_4)) {
+		err = mmc_app_set_bus_width(card, MMC_BUS_WIDTH_4);
+		if (err)
+			return err;
+	} else
+		return 0;
+
+	err = sdio_enable_wide(card);
+	if (err <= 0)
+		mmc_app_set_bus_width(card, MMC_BUS_WIDTH_1);
+
+	return err;
+}
+
+
 /*
  * Test if the card supports high-speed mode and, if so, switch to it.
  */
-static int sdio_enable_hs(struct mmc_card *card)
+static int mmc_sdio_switch_hs(struct mmc_card *card, int enable)
 {
 	int ret;
 	u8 speed;
@@ -182,122 +329,295 @@ static int sdio_enable_hs(struct mmc_card *card)
 	if (ret)
 		return ret;
 
-	speed |= SDIO_SPEED_EHS;
+	if (enable)
+		speed |= SDIO_SPEED_EHS;
+	else
+		speed &= ~SDIO_SPEED_EHS;
 
 	ret = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_SPEED, speed, NULL);
 	if (ret)
 		return ret;
 
-	mmc_card_set_highspeed(card);
-	mmc_set_timing(card->host, MMC_TIMING_SD_HS);
-
-	return 0;
+	return 1;
 }
 
 /*
- * Host is being removed. Free up the current card.
+ * Enable SDIO/combo card's high-speed mode. Return 0/1 if [not]supported.
  */
-static void mmc_sdio_remove(struct mmc_host *host)
+static int sdio_enable_hs(struct mmc_card *card)
 {
-	int i;
+	int ret;
 
-	BUG_ON(!host);
-	BUG_ON(!host->card);
+	ret = mmc_sdio_switch_hs(card, true);
+	if (ret <= 0 || card->type == MMC_TYPE_SDIO)
+		return ret;
 
-	for (i = 0;i < host->card->sdio_funcs;i++) {
-		if (host->card->sdio_func[i]) {
-			sdio_remove_func(host->card->sdio_func[i]);
-			host->card->sdio_func[i] = NULL;
-		}
+	ret = mmc_sd_switch_hs(card);
+	if (ret <= 0)
+		mmc_sdio_switch_hs(card, false);
+
+	return ret;
+}
+
+static unsigned mmc_sdio_get_max_clock(struct mmc_card *card)
+{
+	unsigned max_dtr;
+
+	if (mmc_card_hs(card)) {
+		/*
+		 * The SDIO specification doesn't mention how
+		 * the CIS transfer speed register relates to
+		 * high-speed, but it seems that 50 MHz is
+		 * mandatory.
+		 */
+		max_dtr = 50000000;
+	} else {
+		max_dtr = card->cis.max_dtr;
 	}
 
-	mmc_remove_card(host->card);
-	host->card = NULL;
+	if (card->type == MMC_TYPE_SD_COMBO)
+		max_dtr = min(max_dtr, mmc_sd_get_max_clock(card));
+
+	return max_dtr;
 }
 
-/*
- * Card detection callback from host.
- */
-static void mmc_sdio_detect(struct mmc_host *host)
+static unsigned char host_drive_to_sdio_drive(int host_strength)
 {
+	switch (host_strength) {
+	case MMC_SET_DRIVER_TYPE_A:
+		return SDIO_DTSx_SET_TYPE_A;
+	case MMC_SET_DRIVER_TYPE_B:
+		return SDIO_DTSx_SET_TYPE_B;
+	case MMC_SET_DRIVER_TYPE_C:
+		return SDIO_DTSx_SET_TYPE_C;
+	case MMC_SET_DRIVER_TYPE_D:
+		return SDIO_DTSx_SET_TYPE_D;
+	default:
+		return SDIO_DTSx_SET_TYPE_B;
+	}
+}
+
+static void sdio_select_driver_type(struct mmc_card *card)
+{
+	int host_drv_type = SD_DRIVER_TYPE_B;
+	int card_drv_type = SD_DRIVER_TYPE_B;
+	int drive_strength;
+	unsigned char card_strength;
 	int err;
 
-	BUG_ON(!host);
-	BUG_ON(!host->card);
+	/*
+	 * If the host doesn't support any of the Driver Types A,C or D,
+	 * or there is no board specific handler then default Driver
+	 * Type B is used.
+	 */
+	if (!(card->host->caps &
+		(MMC_CAP_DRIVER_TYPE_A |
+		 MMC_CAP_DRIVER_TYPE_C |
+		 MMC_CAP_DRIVER_TYPE_D)))
+		return;
 
-	mmc_claim_host(host);
+	if (!card->host->ops->select_drive_strength)
+		return;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_A)
+		host_drv_type |= SD_DRIVER_TYPE_A;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_C)
+		host_drv_type |= SD_DRIVER_TYPE_C;
+
+	if (card->host->caps & MMC_CAP_DRIVER_TYPE_D)
+		host_drv_type |= SD_DRIVER_TYPE_D;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_A)
+		card_drv_type |= SD_DRIVER_TYPE_A;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_C)
+		card_drv_type |= SD_DRIVER_TYPE_C;
+
+	if (card->sw_caps.sd3_drv_type & SD_DRIVER_TYPE_D)
+		card_drv_type |= SD_DRIVER_TYPE_D;
 
 	/*
-	 * Just check if our card has been removed.
+	 * The drive strength that the hardware can support
+	 * depends on the board design.  Pass the appropriate
+	 * information and let the hardware specific code
+	 * return what is possible given the options
 	 */
-	err = mmc_select_card(host->card);
+	drive_strength = card->host->ops->select_drive_strength(
+		card->sw_caps.uhs_max_dtr,
+		host_drv_type, card_drv_type);
 
-	mmc_release_host(host);
+	/* if error just use default for drive strength B */
+	err = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_DRIVE_STRENGTH, 0,
+		&card_strength);
+	if (err)
+		return;
 
-	if (err) {
-		mmc_sdio_remove(host);
+	card_strength &= ~(SDIO_DRIVE_DTSx_MASK<<SDIO_DRIVE_DTSx_SHIFT);
+	card_strength |= host_drive_to_sdio_drive(drive_strength);
 
-		mmc_claim_host(host);
-		mmc_detach_bus(host);
-		mmc_release_host(host);
-	}
+	err = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_DRIVE_STRENGTH,
+		card_strength, NULL);
+
+	/* if error default to drive strength B */
+	if (!err)
+		mmc_set_driver_type(card->host, drive_strength);
 }
 
 
-static const struct mmc_bus_ops mmc_sdio_ops = {
-	.remove = mmc_sdio_remove,
-	.detect = mmc_sdio_detect,
-};
+static int sdio_set_bus_speed_mode(struct mmc_card *card)
+{
+	unsigned int bus_speed, timing;
+	int err;
+	unsigned char speed;
 
+	/*
+	 * If the host doesn't support any of the UHS-I modes, fallback on
+	 * default speed.
+	 */
+	if (!mmc_host_uhs(card->host))
+		return 0;
+
+	bus_speed = SDIO_SPEED_SDR12;
+	timing = MMC_TIMING_UHS_SDR12;
+	if ((card->host->caps & MMC_CAP_UHS_SDR104) &&
+	    (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR104)) {
+			bus_speed = SDIO_SPEED_SDR104;
+			timing = MMC_TIMING_UHS_SDR104;
+			card->sw_caps.uhs_max_dtr = UHS_SDR104_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR104_BUS_SPEED;
+	} else if ((card->host->caps & MMC_CAP_UHS_DDR50) &&
+		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_DDR50)) {
+			bus_speed = SDIO_SPEED_DDR50;
+			timing = MMC_TIMING_UHS_DDR50;
+			card->sw_caps.uhs_max_dtr = UHS_DDR50_MAX_DTR;
+			card->sd_bus_speed = UHS_DDR50_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50)) && (card->sw_caps.sd3_bus_mode &
+		    SD_MODE_UHS_SDR50)) {
+			bus_speed = SDIO_SPEED_SDR50;
+			timing = MMC_TIMING_UHS_SDR50;
+			card->sw_caps.uhs_max_dtr = UHS_SDR50_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR50_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25)) &&
+		   (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR25)) {
+			bus_speed = SDIO_SPEED_SDR25;
+			timing = MMC_TIMING_UHS_SDR25;
+			card->sw_caps.uhs_max_dtr = UHS_SDR25_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR25_BUS_SPEED;
+	} else if ((card->host->caps & (MMC_CAP_UHS_SDR104 |
+		    MMC_CAP_UHS_SDR50 | MMC_CAP_UHS_SDR25 |
+		    MMC_CAP_UHS_SDR12)) && (card->sw_caps.sd3_bus_mode &
+		    SD_MODE_UHS_SDR12)) {
+			bus_speed = SDIO_SPEED_SDR12;
+			timing = MMC_TIMING_UHS_SDR12;
+			card->sw_caps.uhs_max_dtr = UHS_SDR12_MAX_DTR;
+			card->sd_bus_speed = UHS_SDR12_BUS_SPEED;
+	}
+
+	err = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_SPEED, 0, &speed);
+	if (err)
+		return err;
+
+	speed &= ~SDIO_SPEED_BSS_MASK;
+	speed |= bus_speed;
+	err = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_SPEED, speed, NULL);
+	if (err)
+		return err;
+
+	if (bus_speed) {
+		mmc_set_timing(card->host, timing);
+		mmc_set_clock(card->host, card->sw_caps.uhs_max_dtr);
+	}
+
+	return 0;
+}
 
 /*
- * Starting point for SDIO card init.
+ * UHS-I specific initialization procedure
  */
-int mmc_attach_sdio(struct mmc_host *host, u32 ocr)
+static int mmc_sdio_init_uhs_card(struct mmc_card *card)
 {
 	int err;
-	int i, funcs;
-	struct mmc_card *card;
 
-	BUG_ON(!host);
-	WARN_ON(!host->claimed);
-
-	mmc_attach_bus(host, &mmc_sdio_ops);
+	if (!card->scr.sda_spec3)
+		return 0;
 
 	/*
-	 * Sanity check the voltages that the card claims to
-	 * support.
+	 * Switch to wider bus (if supported).
 	 */
-	if (ocr & 0x7F) {
-		printk(KERN_WARNING "%s: card claims to support voltages "
-		       "below the defined range. These will be ignored.\n",
-		       mmc_hostname(host));
-		ocr &= ~0x7F;
+	if (card->host->caps & MMC_CAP_4_BIT_DATA) {
+		err = sdio_enable_4bit_bus(card);
+		if (err > 0) {
+			mmc_set_bus_width(card->host, MMC_BUS_WIDTH_4);
+			err = 0;
+		}
 	}
 
-	if (ocr & MMC_VDD_165_195) {
-		printk(KERN_WARNING "%s: SDIO card claims to support the "
-		       "incompletely defined 'low voltage range'. This "
-		       "will be ignored.\n", mmc_hostname(host));
-		ocr &= ~MMC_VDD_165_195;
-	}
+	/* Set the driver strength for the card */
+	sdio_select_driver_type(card);
 
-	host->ocr = mmc_select_voltage(host, ocr);
+	/* Set bus speed mode of the card */
+	err = sdio_set_bus_speed_mode(card);
+	if (err)
+		goto out;
 
 	/*
-	 * Can we support the voltage(s) of the card(s)?
+	 * SPI mode doesn't define CMD19 and tuning is only valid for SDR50 and
+	 * SDR104 mode SD-cards. Note that tuning is mandatory for SDR104.
 	 */
-	if (!host->ocr) {
-		err = -EINVAL;
-		goto err;
+	if (!mmc_host_is_spi(card->host) && card->host->ops->execute_tuning &&
+			((card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR50) ||
+			 (card->sw_caps.sd3_bus_mode & SD_MODE_UHS_SDR104))) {
+		mmc_host_clk_hold(card->host);
+		err = card->host->ops->execute_tuning(card->host,
+						      MMC_SEND_TUNING_BLOCK);
+		mmc_host_clk_release(card->host);
+	}
+
+out:
+
+	return err;
+}
+
+/*
+ * Handle the detection and initialisation of a card.
+ *
+ * In the case of a resume, "oldcard" will contain the card
+ * we're trying to reinitialise.
+ */
+static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr,
+			      struct mmc_card *oldcard, int powered_resume)
+{
+	struct mmc_card *card;
+	int err;
+	int retries = 10;
+	u32 rocr = 0;
+	u32 ocr_card = ocr;
+
+	BUG_ON(!host);
+	WARN_ON(!host->claimed);
+
+	/* to query card if 1.8V signalling is supported */
+	if (mmc_host_uhs(host))
+		ocr |= R4_18V_PRESENT;
+
+try_again:
+	if (!retries) {
+		pr_warning("%s: Skipping voltage switch\n",
+				mmc_hostname(host));
+		ocr &= ~R4_18V_PRESENT;
 	}
 
 	/*
 	 * Inform the card of the voltage
 	 */
-	err = mmc_send_io_op_cond(host, host->ocr, &ocr);
-	if (err)
-		goto err;
+	if (!powered_resume) {
+		err = mmc_send_io_op_cond(host, ocr, &rocr);
+		if (err)
+			goto err;
+	}
 
 	/*
 	 * For SPI, enable CRC as appropriate.
@@ -309,12 +629,6 @@ int mmc_attach_sdio(struct mmc_host *host, u32 ocr)
 	}
 
 	/*
-	 * The number of functions on the card is encoded inside
-	 * the ocr.
-	 */
-	funcs = (ocr & 0x70000000) >> 28;
-
-	/*
 	 * Allocate card structure.
 	 */
 	card = mmc_alloc_card(host, NULL);
@@ -323,35 +637,112 @@ int mmc_attach_sdio(struct mmc_host *host, u32 ocr)
 		goto err;
 	}
 
-	card->type = MMC_TYPE_SDIO;
-	card->sdio_funcs = funcs;
+	if ((rocr & R4_MEMORY_PRESENT) &&
+	    mmc_sd_get_cid(host, ocr & rocr, card->raw_cid, NULL) == 0) {
+		card->type = MMC_TYPE_SD_COMBO;
 
-	host->card = card;
+		if (oldcard && (oldcard->type != MMC_TYPE_SD_COMBO ||
+		    memcmp(card->raw_cid, oldcard->raw_cid, sizeof(card->raw_cid)) != 0)) {
+			mmc_remove_card(card);
+			return -ENOENT;
+		}
+	} else {
+		card->type = MMC_TYPE_SDIO;
+
+		if (oldcard && oldcard->type != MMC_TYPE_SDIO) {
+			mmc_remove_card(card);
+			return -ENOENT;
+		}
+	}
+
+	/*
+	 * Call the optional HC's init_card function to handle quirks.
+	 */
+	if (host->ops->init_card)
+		host->ops->init_card(host, card);
+
+	/*
+	 * If the host and card support UHS-I mode request the card
+	 * to switch to 1.8V signaling level.  No 1.8v signalling if
+	 * UHS mode is not enabled to maintain compatibility and some
+	 * systems that claim 1.8v signalling in fact do not support
+	 * it.
+	 */
+	if (!powered_resume && (rocr & ocr & R4_18V_PRESENT)) {
+		err = mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_180,
+					ocr);
+		if (err == -EAGAIN) {
+			sdio_reset(host);
+			mmc_go_idle(host);
+			mmc_send_if_cond(host, host->ocr_avail);
+			mmc_remove_card(card);
+			retries--;
+			goto try_again;
+		} else if (err) {
+			ocr &= ~R4_18V_PRESENT;
+		}
+		err = 0;
+	} else {
+		ocr &= ~R4_18V_PRESENT;
+	}
 
 	/*
 	 * For native busses:  set card RCA and quit open drain mode.
 	 */
-	if (!mmc_host_is_spi(host)) {
+	if (!powered_resume && !mmc_host_is_spi(host)) {
 		err = mmc_send_relative_addr(host, &card->rca);
 		if (err)
 			goto remove;
 
-		mmc_set_bus_mode(host, MMC_BUSMODE_PUSHPULL);
+		/*
+		 * Update oldcard with the new RCA received from the SDIO
+		 * device -- we're doing this so that it's updated in the
+		 * "card" struct when oldcard overwrites that later.
+		 */
+		if (oldcard)
+			oldcard->rca = card->rca;
+	}
+
+	/*
+	 * Read CSD, before selecting the card
+	 */
+	if (!oldcard && card->type == MMC_TYPE_SD_COMBO) {
+		err = mmc_sd_get_csd(host, card);
+		if (err)
+			return err;
+
+		mmc_decode_cid(card);
 	}
 
 	/*
 	 * Select card, as all following commands rely on that.
 	 */
-	if (!mmc_host_is_spi(host)) {
+	if (!powered_resume && !mmc_host_is_spi(host)) {
 		err = mmc_select_card(card);
 		if (err)
 			goto remove;
 	}
 
+	if (card->quirks & MMC_QUIRK_NONSTD_SDIO) {
+		/*
+		 * This is non-standard SDIO device, meaning it doesn't
+		 * have any CIA (Common I/O area) registers present.
+		 * It's host's responsibility to fill cccr and cis
+		 * structures in init_card().
+		 */
+		mmc_set_clock(host, card->cis.max_dtr);
+
+		if (card->cccr.high_speed) {
+			mmc_set_timing(card->host, MMC_TIMING_SD_HS);
+		}
+
+		goto finish;
+	}
+
 	/*
 	 * Read the common registers.
 	 */
-	err = sdio_read_cccr(card);
+	err = sdio_read_cccr(card, ocr);
 	if (err)
 		goto remove;
 
@@ -362,49 +753,408 @@ int mmc_attach_sdio(struct mmc_host *host, u32 ocr)
 	if (err)
 		goto remove;
 
+	if (oldcard) {
+		int same = (card->cis.vendor == oldcard->cis.vendor &&
+			    card->cis.device == oldcard->cis.device);
+		mmc_remove_card(card);
+		if (!same)
+			return -ENOENT;
+
+		card = oldcard;
+	}
+	card->ocr = ocr_card;
+	mmc_fixup_device(card, NULL);
+
+	if (card->type == MMC_TYPE_SD_COMBO) {
+		err = mmc_sd_setup_card(host, card, oldcard != NULL);
+		/* handle as SDIO-only card if memory init failed */
+		if (err) {
+			mmc_go_idle(host);
+			if (mmc_host_is_spi(host))
+				/* should not fail, as it worked previously */
+				mmc_spi_set_crc(host, use_spi_crc);
+			card->type = MMC_TYPE_SDIO;
+		} else
+			card->dev.type = &sd_type;
+	}
+
 	/*
-	 * Switch to high-speed (if supported).
+	 * If needed, disconnect card detection pull-up resistor.
 	 */
-	err = sdio_enable_hs(card);
+	err = sdio_disable_cd(card);
 	if (err)
 		goto remove;
 
+	/* Initialization sequence for UHS-I cards */
+	/* Only if card supports 1.8v and UHS signaling */
+	if ((ocr & R4_18V_PRESENT) && card->sw_caps.sd3_bus_mode) {
+		err = mmc_sdio_init_uhs_card(card);
+		if (err)
+			goto remove;
+	} else {
+		/*
+		 * Switch to high-speed (if supported).
+		 */
+		err = sdio_enable_hs(card);
+		if (err > 0)
+			mmc_set_timing(card->host, MMC_TIMING_SD_HS);
+		else if (err)
+			goto remove;
+
+		/*
+		 * Change to the card's maximum speed.
+		 */
+		mmc_set_clock(host, mmc_sdio_get_max_clock(card));
+
+		/*
+		 * Switch to wider bus (if supported).
+		 */
+		err = sdio_enable_4bit_bus(card);
+		if (err > 0)
+			mmc_set_bus_width(card->host, MMC_BUS_WIDTH_4);
+		else if (err)
+			goto remove;
+	}
+finish:
+	if (!oldcard)
+		host->card = card;
+	return 0;
+
+remove:
+	if (!oldcard)
+		mmc_remove_card(card);
+
+err:
+	return err;
+}
+
+/*
+ * Host is being removed. Free up the current card.
+ */
+static void mmc_sdio_remove(struct mmc_host *host)
+{
+	int i;
+
+	BUG_ON(!host);
+	BUG_ON(!host->card);
+
+	for (i = 0;i < host->card->sdio_funcs;i++) {
+		if (host->card->sdio_func[i]) {
+			sdio_remove_func(host->card->sdio_func[i]);
+			host->card->sdio_func[i] = NULL;
+		}
+	}
+
+	mmc_remove_card(host->card);
+	host->card = NULL;
+}
+
+/*
+ * Card detection - card is alive.
+ */
+static int mmc_sdio_alive(struct mmc_host *host)
+{
+	return mmc_select_card(host->card);
+}
+
+/*
+ * Card detection callback from host.
+ */
+static void mmc_sdio_detect(struct mmc_host *host)
+{
+	int err;
+
+	BUG_ON(!host);
+	BUG_ON(!host->card);
+
+	/* Make sure card is powered before detecting it */
+	if (host->caps & MMC_CAP_POWER_OFF_CARD) {
+		err = pm_runtime_get_sync(&host->card->dev);
+		if (err < 0) {
+			pm_runtime_put_noidle(&host->card->dev);
+			goto out;
+		}
+	}
+
+	mmc_claim_host(host);
+
+	/*
+	 * Just check if our card has been removed.
+	 */
+	err = _mmc_detect_card_removed(host);
+
+	mmc_release_host(host);
+
 	/*
-	 * Change to the card's maximum speed.
+	 * Tell PM core it's OK to power off the card now.
+	 *
+	 * The _sync variant is used in order to ensure that the card
+	 * is left powered off in case an error occurred, and the card
+	 * is going to be removed.
+	 *
+	 * Since there is no specific reason to believe a new user
+	 * is about to show up at this point, the _sync variant is
+	 * desirable anyway.
 	 */
-	if (mmc_card_highspeed(card)) {
+	if (host->caps & MMC_CAP_POWER_OFF_CARD)
+		pm_runtime_put_sync(&host->card->dev);
+
+out:
+	if (err) {
+		mmc_sdio_remove(host);
+
+		mmc_claim_host(host);
+		mmc_detach_bus(host);
+		mmc_power_off(host);
+		mmc_release_host(host);
+	}
+}
+
+/*
+ * SDIO pre_suspend.  We need to suspend all functions separately.
+ * Therefore all registered functions must have drivers with suspend
+ * and resume methods.  Failing that we simply remove the whole card.
+ */
+static int mmc_sdio_pre_suspend(struct mmc_host *host)
+{
+	int i, err = 0;
+
+	for (i = 0; i < host->card->sdio_funcs; i++) {
+		struct sdio_func *func = host->card->sdio_func[i];
+		if (func && sdio_func_present(func) && func->dev.driver) {
+			const struct dev_pm_ops *pmops = func->dev.driver->pm;
+			if (!pmops || !pmops->suspend || !pmops->resume) {
+				/* force removal of entire card in that case */
+				err = -ENOSYS;
+				break;
+			}
+		}
+	}
+
+	return err;
+}
+
+/*
+ * SDIO suspend.  Suspend all functions separately.
+ */
+static int mmc_sdio_suspend(struct mmc_host *host)
+{
+	if (mmc_card_keep_power(host) && mmc_card_wake_sdio_irq(host)) {
+		mmc_claim_host(host);
+		sdio_disable_wide(host->card);
+		mmc_release_host(host);
+	}
+
+	if (!mmc_card_keep_power(host))
+		mmc_power_off(host);
+
+	return 0;
+}
+
+static int mmc_sdio_resume(struct mmc_host *host)
+{
+	int err = 0;
+
+	BUG_ON(!host);
+	BUG_ON(!host->card);
+
+	/* Basic card reinitialization. */
+	mmc_claim_host(host);
+
+	/* Restore power if needed */
+	if (!mmc_card_keep_power(host)) {
+		mmc_power_up(host, host->card->ocr);
 		/*
-		 * The SDIO specification doesn't mention how
-		 * the CIS transfer speed register relates to
-		 * high-speed, but it seems that 50 MHz is
-		 * mandatory.
+		 * Tell runtime PM core we just powered up the card,
+		 * since it still believes the card is powered off.
+		 * Note that currently runtime PM is only enabled
+		 * for SDIO cards that are MMC_CAP_POWER_OFF_CARD
 		 */
-		mmc_set_clock(host, 50000000);
-	} else {
-		mmc_set_clock(host, card->cis.max_dtr);
+		if (host->caps & MMC_CAP_POWER_OFF_CARD) {
+			pm_runtime_disable(&host->card->dev);
+			pm_runtime_set_active(&host->card->dev);
+			pm_runtime_enable(&host->card->dev);
+		}
+	}
+
+	/* No need to reinitialize powered-resumed nonremovable cards */
+	if (mmc_card_is_removable(host) || !mmc_card_keep_power(host)) {
+		sdio_reset(host);
+		mmc_go_idle(host);
+		err = mmc_sdio_init_card(host, host->card->ocr, host->card,
+					mmc_card_keep_power(host));
+	} else if (mmc_card_keep_power(host) && mmc_card_wake_sdio_irq(host)) {
+		/* We may have switched to 1-bit mode during suspend */
+		err = sdio_enable_4bit_bus(host->card);
+		if (err > 0) {
+			mmc_set_bus_width(host, MMC_BUS_WIDTH_4);
+			err = 0;
+		}
 	}
 
+	if (!err && host->sdio_irqs)
+		wake_up_process(host->sdio_irq_thread);
+	mmc_release_host(host);
+
+	host->pm_flags &= ~MMC_PM_KEEP_POWER;
+	return err;
+}
+
+static int mmc_sdio_power_restore(struct mmc_host *host)
+{
+	int ret;
+
+	BUG_ON(!host);
+	BUG_ON(!host->card);
+
+	mmc_claim_host(host);
+
 	/*
-	 * Switch to wider bus (if supported).
+	 * Reset the card by performing the same steps that are taken by
+	 * mmc_rescan_try_freq() and mmc_attach_sdio() during a "normal" probe.
+	 *
+	 * sdio_reset() is technically not needed. Having just powered up the
+	 * hardware, it should already be in reset state. However, some
+	 * platforms (such as SD8686 on OLPC) do not instantly cut power,
+	 * meaning that a reset is required when restoring power soon after
+	 * powering off. It is harmless in other cases.
+	 *
+	 * The CMD5 reset (mmc_send_io_op_cond()), according to the SDIO spec,
+	 * is not necessary for non-removable cards. However, it is required
+	 * for OLPC SD8686 (which expects a [CMD5,5,3,7] init sequence), and
+	 * harmless in other situations.
+	 *
 	 */
-	err = sdio_enable_wide(card);
+
+	sdio_reset(host);
+	mmc_go_idle(host);
+	mmc_send_if_cond(host, host->ocr_avail);
+
+	ret = mmc_send_io_op_cond(host, 0, NULL);
+	if (ret)
+		goto out;
+
+	ret = mmc_sdio_init_card(host, host->card->ocr, host->card,
+				mmc_card_keep_power(host));
+	if (!ret && host->sdio_irqs)
+		mmc_signal_sdio_irq(host);
+
+out:
+	mmc_release_host(host);
+
+	return ret;
+}
+
+static int mmc_sdio_runtime_suspend(struct mmc_host *host)
+{
+	/* No references to the card, cut the power to it. */
+	mmc_power_off(host);
+	return 0;
+}
+
+static int mmc_sdio_runtime_resume(struct mmc_host *host)
+{
+	/* Restore power and re-initialize. */
+	mmc_power_up(host, host->card->ocr);
+	return mmc_sdio_power_restore(host);
+}
+
+static const struct mmc_bus_ops mmc_sdio_ops = {
+	.remove = mmc_sdio_remove,
+	.detect = mmc_sdio_detect,
+	.pre_suspend = mmc_sdio_pre_suspend,
+	.suspend = mmc_sdio_suspend,
+	.resume = mmc_sdio_resume,
+	.runtime_suspend = mmc_sdio_runtime_suspend,
+	.runtime_resume = mmc_sdio_runtime_resume,
+	.power_restore = mmc_sdio_power_restore,
+	.alive = mmc_sdio_alive,
+};
+
+
+/*
+ * Starting point for SDIO card init.
+ */
+int mmc_attach_sdio(struct mmc_host *host)
+{
+	int err, i, funcs;
+	u32 ocr, rocr;
+	struct mmc_card *card;
+
+	BUG_ON(!host);
+	WARN_ON(!host->claimed);
+
+	err = mmc_send_io_op_cond(host, 0, &ocr);
 	if (err)
-		goto remove;
+		return err;
+
+	mmc_attach_bus(host, &mmc_sdio_ops);
+	if (host->ocr_avail_sdio)
+		host->ocr_avail = host->ocr_avail_sdio;
+
+
+	rocr = mmc_select_voltage(host, ocr);
+
+	/*
+	 * Can we support the voltage(s) of the card(s)?
+	 */
+	if (!rocr) {
+		err = -EINVAL;
+		goto err;
+	}
+
+	/*
+	 * Detect and init the card.
+	 */
+	err = mmc_sdio_init_card(host, rocr, NULL, 0);
+	if (err)
+		goto err;
+
+	card = host->card;
+
+	/*
+	 * Enable runtime PM only if supported by host+card+board
+	 */
+	if (host->caps & MMC_CAP_POWER_OFF_CARD) {
+		/*
+		 * Let runtime PM core know our card is active
+		 */
+		err = pm_runtime_set_active(&card->dev);
+		if (err)
+			goto remove;
+
+		/*
+		 * Enable runtime PM for this card
+		 */
+		pm_runtime_enable(&card->dev);
+	}
+
+	/*
+	 * The number of functions on the card is encoded inside
+	 * the ocr.
+	 */
+	funcs = (ocr & 0x70000000) >> 28;
+	card->sdio_funcs = 0;
 
 	/*
 	 * Initialize (but don't add) all present functions.
 	 */
-	for (i = 0;i < funcs;i++) {
+	for (i = 0; i < funcs; i++, card->sdio_funcs++) {
 		err = sdio_init_func(host->card, i + 1);
 		if (err)
 			goto remove;
-	}
 
-	mmc_release_host(host);
+		/*
+		 * Enable Runtime PM for this func (if supported)
+		 */
+		if (host->caps & MMC_CAP_POWER_OFF_CARD)
+			pm_runtime_enable(&card->sdio_func[i]->dev);
+	}
 
 	/*
 	 * First add the card to the driver model...
 	 */
+	mmc_release_host(host);
 	err = mmc_add_card(host->card);
 	if (err)
 		goto remove_added;
@@ -418,6 +1168,7 @@ int mmc_attach_sdio(struct mmc_host *host, u32 ocr)
 			goto remove_added;
 	}
 
+	mmc_claim_host(host);
 	return 0;
 
 
@@ -427,13 +1178,14 @@ remove_added:
 	mmc_claim_host(host);
 remove:
 	/* And with lock if it hasn't been added. */
+	mmc_release_host(host);
 	if (host->card)
 		mmc_sdio_remove(host);
+	mmc_claim_host(host);
 err:
 	mmc_detach_bus(host);
-	mmc_release_host(host);
 
-	printk(KERN_ERR "%s: error %d whilst initialising SDIO card\n",
+	pr_err("%s: error %d whilst initialising SDIO card\n",
 		mmc_hostname(host), err);
 
 	return err;
diff --git a/drivers/mmc/core/sdio_bus.c b/drivers/mmc/core/sdio_bus.c
index 46284b52739..4fa8fef9147 100644
--- a/drivers/mmc/core/sdio_bus.c
+++ b/drivers/mmc/core/sdio_bus.c
@@ -13,16 +13,18 @@
 
 #include <linux/device.h>
 #include <linux/err.h>
+#include <linux/export.h>
+#include <linux/slab.h>
+#include <linux/pm_runtime.h>
+#include <linux/acpi.h>
 
 #include <linux/mmc/card.h>
+#include <linux/mmc/host.h>
 #include <linux/mmc/sdio_func.h>
 
 #include "sdio_cis.h"
 #include "sdio_bus.h"
 
-#define dev_to_sdio_func(d)	container_of(d, struct sdio_func, dev)
-#define to_sdio_driver(d)      container_of(d, struct sdio_driver, drv)
-
 /* show configuration fields */
 #define sdio_config_attr(field, format_string)				\
 static ssize_t								\
@@ -32,7 +34,8 @@ field##_show(struct device *dev, struct device_attribute *attr, char *buf)				\
 									\
 	func = dev_to_sdio_func (dev);					\
 	return sprintf (buf, format_string, func->field);		\
-}
+}									\
+static DEVICE_ATTR_RO(field)
 
 sdio_config_attr(class, "0x%02x\n");
 sdio_config_attr(vendor, "0x%04x\n");
@@ -45,14 +48,16 @@ static ssize_t modalias_show(struct device *dev, struct device_attribute *attr,
 	return sprintf(buf, "sdio:c%02Xv%04Xd%04X\n",
 			func->class, func->vendor, func->device);
 }
-
-static struct device_attribute sdio_dev_attrs[] = {
-	__ATTR_RO(class),
-	__ATTR_RO(vendor),
-	__ATTR_RO(device),
-	__ATTR_RO(modalias),
-	__ATTR_NULL,
+static DEVICE_ATTR_RO(modalias);
+
+static struct attribute *sdio_dev_attrs[] = {
+	&dev_attr_class.attr,
+	&dev_attr_vendor.attr,
+	&dev_attr_device.attr,
+	&dev_attr_modalias.attr,
+	NULL,
 };
+ATTRIBUTE_GROUPS(sdio_dev);
 
 static const struct sdio_device_id *sdio_match_one(struct sdio_func *func,
 	const struct sdio_device_id *id)
@@ -127,42 +132,96 @@ static int sdio_bus_probe(struct device *dev)
 	if (!id)
 		return -ENODEV;
 
+	/* Unbound SDIO functions are always suspended.
+	 * During probe, the function is set active and the usage count
+	 * is incremented.  If the driver supports runtime PM,
+	 * it should call pm_runtime_put_noidle() in its probe routine and
+	 * pm_runtime_get_noresume() in its remove routine.
+	 */
+	if (func->card->host->caps & MMC_CAP_POWER_OFF_CARD) {
+		ret = pm_runtime_get_sync(dev);
+		if (ret < 0)
+			goto disable_runtimepm;
+	}
+
 	/* Set the default block size so the driver is sure it's something
 	 * sensible. */
 	sdio_claim_host(func);
 	ret = sdio_set_block_size(func, 0);
 	sdio_release_host(func);
 	if (ret)
-		return ret;
+		goto disable_runtimepm;
+
+	ret = drv->probe(func, id);
+	if (ret)
+		goto disable_runtimepm;
 
-	return drv->probe(func, id);
+	return 0;
+
+disable_runtimepm:
+	if (func->card->host->caps & MMC_CAP_POWER_OFF_CARD)
+		pm_runtime_put_noidle(dev);
+	return ret;
 }
 
 static int sdio_bus_remove(struct device *dev)
 {
 	struct sdio_driver *drv = to_sdio_driver(dev->driver);
 	struct sdio_func *func = dev_to_sdio_func(dev);
+	int ret = 0;
+
+	/* Make sure card is powered before invoking ->remove() */
+	if (func->card->host->caps & MMC_CAP_POWER_OFF_CARD)
+		pm_runtime_get_sync(dev);
 
 	drv->remove(func);
 
 	if (func->irq_handler) {
-		printk(KERN_WARNING "WARNING: driver %s did not remove "
+		pr_warning("WARNING: driver %s did not remove "
 			"its interrupt handler!\n", drv->name);
 		sdio_claim_host(func);
 		sdio_release_irq(func);
 		sdio_release_host(func);
 	}
 
-	return 0;
+	/* First, undo the increment made directly above */
+	if (func->card->host->caps & MMC_CAP_POWER_OFF_CARD)
+		pm_runtime_put_noidle(dev);
+
+	/* Then undo the runtime PM settings in sdio_bus_probe() */
+	if (func->card->host->caps & MMC_CAP_POWER_OFF_CARD)
+		pm_runtime_put_sync(dev);
+
+	return ret;
 }
 
+#ifdef CONFIG_PM
+
+static const struct dev_pm_ops sdio_bus_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pm_generic_suspend, pm_generic_resume)
+	SET_RUNTIME_PM_OPS(
+		pm_generic_runtime_suspend,
+		pm_generic_runtime_resume,
+		NULL
+	)
+};
+
+#define SDIO_PM_OPS_PTR	(&sdio_bus_pm_ops)
+
+#else /* !CONFIG_PM */
+
+#define SDIO_PM_OPS_PTR	NULL
+
+#endif /* !CONFIG_PM */
+
 static struct bus_type sdio_bus_type = {
 	.name		= "sdio",
-	.dev_attrs	= sdio_dev_attrs,
+	.dev_groups	= sdio_dev_groups,
 	.match		= sdio_bus_match,
 	.uevent		= sdio_bus_uevent,
 	.probe		= sdio_bus_probe,
 	.remove		= sdio_bus_remove,
+	.pm		= SDIO_PM_OPS_PTR,
 };
 
 int sdio_register_bus(void)
@@ -204,8 +263,7 @@ static void sdio_release_func(struct device *dev)
 
 	sdio_free_func_cis(func);
 
-	if (func->info)
-		kfree(func->info);
+	kfree(func->info);
 
 	kfree(func);
 }
@@ -232,6 +290,18 @@ struct sdio_func *sdio_alloc_func(struct mmc_card *card)
 	return func;
 }
 
+#ifdef CONFIG_ACPI
+static void sdio_acpi_set_handle(struct sdio_func *func)
+{
+	struct mmc_host *host = func->card->host;
+	u64 addr = (host->slotno << 16) | func->num;
+
+	acpi_preset_companion(&func->dev, ACPI_COMPANION(host->parent), addr);
+}
+#else
+static inline void sdio_acpi_set_handle(struct sdio_func *func) {}
+#endif
+
 /*
  * Register a new SDIO function with the driver model.
  */
@@ -241,9 +311,12 @@ int sdio_add_func(struct sdio_func *func)
 
 	dev_set_name(&func->dev, "%s:%d", mmc_card_id(func->card), func->num);
 
+	sdio_acpi_set_handle(func);
 	ret = device_add(&func->dev);
-	if (ret == 0)
+	if (ret == 0) {
 		sdio_func_set_present(func);
+		acpi_dev_pm_attach(&func->dev, false);
+	}
 
 	return ret;
 }
@@ -251,12 +324,16 @@ int sdio_add_func(struct sdio_func *func)
 /*
  * Unregister a SDIO function with the driver model, and
  * (eventually) free it.
+ * This function can be called through error paths where sdio_add_func() was
+ * never executed (because a failure occurred at an earlier point).
  */
 void sdio_remove_func(struct sdio_func *func)
 {
-	if (sdio_func_present(func))
-		device_del(&func->dev);
+	if (!sdio_func_present(func))
+		return;
 
+	acpi_dev_pm_detach(&func->dev, false);
+	device_del(&func->dev);
 	put_device(&func->dev);
 }
 
diff --git a/drivers/mmc/core/sdio_cis.c b/drivers/mmc/core/sdio_cis.c
index 963f2937c5e..8e94e555b78 100644
--- a/drivers/mmc/core/sdio_cis.c
+++ b/drivers/mmc/core/sdio_cis.c
@@ -14,6 +14,7 @@
  */
 
 #include <linux/kernel.h>
+#include <linux/slab.h>
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/card.h>
@@ -29,6 +30,8 @@ static int cistpl_vers_1(struct mmc_card *card, struct sdio_func *func,
 	unsigned i, nr_strings;
 	char **buffer, *string;
 
+	/* Find all null-terminated (including zero length) strings in
+	   the TPLLV1_INFO field. Trailing garbage is ignored. */
 	buf += 2;
 	size -= 2;
 
@@ -39,11 +42,8 @@ static int cistpl_vers_1(struct mmc_card *card, struct sdio_func *func,
 		if (buf[i] == 0)
 			nr_strings++;
 	}
-
-	if (buf[i-1] != '\0') {
-		printk(KERN_WARNING "SDIO: ignoring broken CISTPL_VERS_1\n");
+	if (nr_strings == 0)
 		return 0;
-	}
 
 	size = i;
 
@@ -98,10 +98,56 @@ static const unsigned char speed_val[16] =
 static const unsigned int speed_unit[8] =
 	{ 10000, 100000, 1000000, 10000000, 0, 0, 0, 0 };
 
-static int cistpl_funce_common(struct mmc_card *card,
+
+typedef int (tpl_parse_t)(struct mmc_card *, struct sdio_func *,
+			   const unsigned char *, unsigned);
+
+struct cis_tpl {
+	unsigned char code;
+	unsigned char min_size;
+	tpl_parse_t *parse;
+};
+
+static int cis_tpl_parse(struct mmc_card *card, struct sdio_func *func,
+			 const char *tpl_descr,
+			 const struct cis_tpl *tpl, int tpl_count,
+			 unsigned char code,
+			 const unsigned char *buf, unsigned size)
+{
+	int i, ret;
+
+	/* look for a matching code in the table */
+	for (i = 0; i < tpl_count; i++, tpl++) {
+		if (tpl->code == code)
+			break;
+	}
+	if (i < tpl_count) {
+		if (size >= tpl->min_size) {
+			if (tpl->parse)
+				ret = tpl->parse(card, func, buf, size);
+			else
+				ret = -EILSEQ;	/* known tuple, not parsed */
+		} else {
+			/* invalid tuple */
+			ret = -EINVAL;
+		}
+		if (ret && ret != -EILSEQ && ret != -ENOENT) {
+			pr_err("%s: bad %s tuple 0x%02x (%u bytes)\n",
+			       mmc_hostname(card->host), tpl_descr, code, size);
+		}
+	} else {
+		/* unknown tuple */
+		ret = -ENOENT;
+	}
+
+	return ret;
+}
+
+static int cistpl_funce_common(struct mmc_card *card, struct sdio_func *func,
 			       const unsigned char *buf, unsigned size)
 {
-	if (size < 0x04 || buf[0] != 0)
+	/* Only valid for the common CIS (function 0) */
+	if (func)
 		return -EINVAL;
 
 	/* TPLFE_FN0_BLK_SIZE */
@@ -114,16 +160,24 @@ static int cistpl_funce_common(struct mmc_card *card,
 	return 0;
 }
 
-static int cistpl_funce_func(struct sdio_func *func,
+static int cistpl_funce_func(struct mmc_card *card, struct sdio_func *func,
 			     const unsigned char *buf, unsigned size)
 {
 	unsigned vsn;
 	unsigned min_size;
 
+	/* Only valid for the individual function's CIS (1-7) */
+	if (!func)
+		return -EINVAL;
+
+	/*
+	 * This tuple has a different length depending on the SDIO spec
+	 * version.
+	 */
 	vsn = func->card->cccr.sdio_vsn;
 	min_size = (vsn == SDIO_SDIO_REV_1_00) ? 28 : 42;
 
-	if (size < min_size || buf[0] != 1)
+	if (size < min_size)
 		return -EINVAL;
 
 	/* TPLFE_MAX_BLK_SIZE */
@@ -138,40 +192,32 @@ static int cistpl_funce_func(struct sdio_func *func,
 	return 0;
 }
 
+/*
+ * Known TPLFE_TYPEs table for CISTPL_FUNCE tuples.
+ *
+ * Note that, unlike PCMCIA, CISTPL_FUNCE tuples are not parsed depending
+ * on the TPLFID_FUNCTION value of the previous CISTPL_FUNCID as on SDIO
+ * TPLFID_FUNCTION is always hardcoded to 0x0C.
+ */
+static const struct cis_tpl cis_tpl_funce_list[] = {
+	{	0x00,	4,	cistpl_funce_common		},
+	{	0x01,	0,	cistpl_funce_func		},
+	{	0x04,	1+1+6,	/* CISTPL_FUNCE_LAN_NODE_ID */	},
+};
+
 static int cistpl_funce(struct mmc_card *card, struct sdio_func *func,
 			const unsigned char *buf, unsigned size)
 {
-	int ret;
-
-	/*
-	 * There should be two versions of the CISTPL_FUNCE tuple,
-	 * one for the common CIS (function 0) and a version used by
-	 * the individual function's CIS (1-7). Yet, the later has a
-	 * different length depending on the SDIO spec version.
-	 */
-	if (func)
-		ret = cistpl_funce_func(func, buf, size);
-	else
-		ret = cistpl_funce_common(card, buf, size);
-
-	if (ret) {
-		printk(KERN_ERR "%s: bad CISTPL_FUNCE size %u "
-		       "type %u\n", mmc_hostname(card->host), size, buf[0]);
-		return ret;
-	}
+	if (size < 1)
+		return -EINVAL;
 
-	return 0;
+	return cis_tpl_parse(card, func, "CISTPL_FUNCE",
+			     cis_tpl_funce_list,
+			     ARRAY_SIZE(cis_tpl_funce_list),
+			     buf[0], buf, size);
 }
 
-typedef int (tpl_parse_t)(struct mmc_card *, struct sdio_func *,
-			   const unsigned char *, unsigned);
-
-struct cis_tpl {
-	unsigned char code;
-	unsigned char min_size;
-	tpl_parse_t *parse;
-};
-
+/* Known TPL_CODEs table for CIS tuples */
 static const struct cis_tpl cis_tpl_list[] = {
 	{	0x15,	3,	cistpl_vers_1		},
 	{	0x20,	4,	cistpl_manfid		},
@@ -250,31 +296,37 @@ static int sdio_read_cis(struct mmc_card *card, struct sdio_func *func)
 			break;
 		}
 
-		for (i = 0; i < ARRAY_SIZE(cis_tpl_list); i++)
-			if (cis_tpl_list[i].code == tpl_code)
-				break;
-		if (i >= ARRAY_SIZE(cis_tpl_list)) {
-			/* this tuple is unknown to the core */
+		/* Try to parse the CIS tuple */
+		ret = cis_tpl_parse(card, func, "CIS",
+				    cis_tpl_list, ARRAY_SIZE(cis_tpl_list),
+				    tpl_code, this->data, tpl_link);
+		if (ret == -EILSEQ || ret == -ENOENT) {
+			/*
+			 * The tuple is unknown or known but not parsed.
+			 * Queue the tuple for the function driver.
+			 */
 			this->next = NULL;
 			this->code = tpl_code;
 			this->size = tpl_link;
 			*prev = this;
 			prev = &this->next;
-			printk(KERN_DEBUG
-			       "%s: queuing CIS tuple 0x%02x length %u\n",
-			       mmc_hostname(card->host), tpl_code, tpl_link);
-		} else {
-			const struct cis_tpl *tpl = cis_tpl_list + i;
-			if (tpl_link < tpl->min_size) {
-				printk(KERN_ERR
-				       "%s: bad CIS tuple 0x%02x (length = %u, expected >= %u)\n",
+
+			if (ret == -ENOENT) {
+				/* warn about unknown tuples */
+				pr_warn_ratelimited("%s: queuing unknown"
+				       " CIS tuple 0x%02x (%u bytes)\n",
 				       mmc_hostname(card->host),
-				       tpl_code, tpl_link, tpl->min_size);
-				ret = -EINVAL;
-			} else if (tpl->parse) {
-				ret = tpl->parse(card, func,
-						 this->data, tpl_link);
+				       tpl_code, tpl_link);
 			}
+
+			/* keep on analyzing tuples */
+			ret = 0;
+		} else {
+			/*
+			 * We don't need the tuple anymore if it was
+			 * successfully parsed by the SDIO core or if it is
+			 * not going to be queued for a driver.
+			 */
 			kfree(this);
 		}
 
diff --git a/drivers/mmc/core/sdio_io.c b/drivers/mmc/core/sdio_io.c
index f61fc2d4cd0..78cb4d5d9d5 100644
--- a/drivers/mmc/core/sdio_io.c
+++ b/drivers/mmc/core/sdio_io.c
@@ -9,6 +9,7 @@
  * your option) any later version.
  */
 
+#include <linux/export.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/card.h>
 #include <linux/mmc/sdio.h>
@@ -187,9 +188,16 @@ EXPORT_SYMBOL_GPL(sdio_set_block_size);
  */
 static inline unsigned int sdio_max_byte_size(struct sdio_func *func)
 {
-	unsigned mval =	min(func->card->host->max_seg_size,
-			    func->card->host->max_blk_size);
-	mval = min(mval, func->max_blksize);
+	unsigned mval =	func->card->host->max_blk_size;
+
+	if (mmc_blksz_for_byte_mode(func->card))
+		mval = min(mval, func->cur_blksize);
+	else
+		mval = min(mval, func->max_blksize);
+
+	if (mmc_card_broken_byte_mode_512(func->card))
+		return min(mval, 511u);
+
 	return min(mval, 512u); /* maximum size for byte mode */
 }
 
@@ -302,13 +310,10 @@ static int sdio_io_rw_ext_helper(struct sdio_func *func, int write,
 	/* Do the bulk of the transfer using block mode (if supported). */
 	if (func->card->cccr.multi_block && (size > sdio_max_byte_size(func))) {
 		/* Blocks per command is limited by host count, host transfer
-		 * size (we only use a single sg entry) and the maximum for
-		 * IO_RW_EXTENDED of 511 blocks. */
-		max_blocks = min(func->card->host->max_blk_count,
-			func->card->host->max_seg_size / func->cur_blksize);
-		max_blocks = min(max_blocks, 511u);
+		 * size and the maximum for IO_RW_EXTENDED of 511 blocks. */
+		max_blocks = min(func->card->host->max_blk_count, 511u);
 
-		while (remainder > func->cur_blksize) {
+		while (remainder >= func->cur_blksize) {
 			unsigned blocks;
 
 			blocks = remainder / func->cur_blksize;
@@ -333,8 +338,9 @@ static int sdio_io_rw_ext_helper(struct sdio_func *func, int write,
 	while (remainder > 0) {
 		size = min(remainder, sdio_max_byte_size(func));
 
+		/* Indicate byte mode by setting "blocks" = 0 */
 		ret = mmc_io_rw_extended(func->card, write, func->num, addr,
-			 incr_addr, buf, 1, size);
+			 incr_addr, buf, 0, size);
 		if (ret)
 			return ret;
 
@@ -401,6 +407,36 @@ void sdio_writeb(struct sdio_func *func, u8 b, unsigned int addr, int *err_ret)
 EXPORT_SYMBOL_GPL(sdio_writeb);
 
 /**
+ *	sdio_writeb_readb - write and read a byte from SDIO function
+ *	@func: SDIO function to access
+ *	@write_byte: byte to write
+ *	@addr: address to write to
+ *	@err_ret: optional status value from transfer
+ *
+ *	Performs a RAW (Read after Write) operation as defined by SDIO spec -
+ *	single byte is written to address space of a given SDIO function and
+ *	response is read back from the same address, both using single request.
+ *	If there is a problem with the operation, 0xff is returned and
+ *	@err_ret will contain the error code.
+ */
+u8 sdio_writeb_readb(struct sdio_func *func, u8 write_byte,
+	unsigned int addr, int *err_ret)
+{
+	int ret;
+	u8 val;
+
+	ret = mmc_io_rw_direct(func->card, 1, func->num, addr,
+			write_byte, &val);
+	if (err_ret)
+		*err_ret = ret;
+	if (ret)
+		val = 0xff;
+
+	return val;
+}
+EXPORT_SYMBOL_GPL(sdio_writeb_readb);
+
+/**
  *	sdio_memcpy_fromio - read a chunk of memory from a SDIO function
  *	@func: SDIO function to access
  *	@dst: buffer to store the data
@@ -624,7 +660,7 @@ void sdio_f0_writeb(struct sdio_func *func, unsigned char b, unsigned int addr,
 
 	BUG_ON(!func);
 
-	if (addr < 0xF0 || addr > 0xFF) {
+	if ((addr < 0xF0 || addr > 0xFF) && (!mmc_card_lenient_fn0(func->card))) {
 		if (err_ret)
 			*err_ret = -EINVAL;
 		return;
@@ -635,3 +671,52 @@ void sdio_f0_writeb(struct sdio_func *func, unsigned char b, unsigned int addr,
 		*err_ret = ret;
 }
 EXPORT_SYMBOL_GPL(sdio_f0_writeb);
+
+/**
+ *	sdio_get_host_pm_caps - get host power management capabilities
+ *	@func: SDIO function attached to host
+ *
+ *	Returns a capability bitmask corresponding to power management
+ *	features supported by the host controller that the card function
+ *	might rely upon during a system suspend.  The host doesn't need
+ *	to be claimed, nor the function active, for this information to be
+ *	obtained.
+ */
+mmc_pm_flag_t sdio_get_host_pm_caps(struct sdio_func *func)
+{
+	BUG_ON(!func);
+	BUG_ON(!func->card);
+
+	return func->card->host->pm_caps;
+}
+EXPORT_SYMBOL_GPL(sdio_get_host_pm_caps);
+
+/**
+ *	sdio_set_host_pm_flags - set wanted host power management capabilities
+ *	@func: SDIO function attached to host
+ *
+ *	Set a capability bitmask corresponding to wanted host controller
+ *	power management features for the upcoming suspend state.
+ *	This must be called, if needed, each time the suspend method of
+ *	the function driver is called, and must contain only bits that
+ *	were returned by sdio_get_host_pm_caps().
+ *	The host doesn't need to be claimed, nor the function active,
+ *	for this information to be set.
+ */
+int sdio_set_host_pm_flags(struct sdio_func *func, mmc_pm_flag_t flags)
+{
+	struct mmc_host *host;
+
+	BUG_ON(!func);
+	BUG_ON(!func->card);
+
+	host = func->card->host;
+
+	if (flags & ~host->pm_caps)
+		return -EINVAL;
+
+	/* function suspend methods are serialized, hence no lock needed */
+	host->pm_flags |= flags;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(sdio_set_host_pm_flags);
diff --git a/drivers/mmc/core/sdio_irq.c b/drivers/mmc/core/sdio_irq.c
index bb192f90e8e..5cc13c8d35b 100644
--- a/drivers/mmc/core/sdio_irq.c
+++ b/drivers/mmc/core/sdio_irq.c
@@ -16,6 +16,7 @@
 #include <linux/kernel.h>
 #include <linux/sched.h>
 #include <linux/kthread.h>
+#include <linux/export.h>
 #include <linux/wait.h>
 #include <linux/delay.h>
 
@@ -27,32 +28,56 @@
 
 #include "sdio_ops.h"
 
-static int process_sdio_pending_irqs(struct mmc_card *card)
+static int process_sdio_pending_irqs(struct mmc_host *host)
 {
+	struct mmc_card *card = host->card;
 	int i, ret, count;
 	unsigned char pending;
+	struct sdio_func *func;
+
+	/*
+	 * Optimization, if there is only 1 function interrupt registered
+	 * and we know an IRQ was signaled then call irq handler directly.
+	 * Otherwise do the full probe.
+	 */
+	func = card->sdio_single_irq;
+	if (func && host->sdio_irq_pending) {
+		func->irq_handler(func);
+		return 1;
+	}
 
 	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTx, 0, &pending);
 	if (ret) {
-		printk(KERN_DEBUG "%s: error %d reading SDIO_CCCR_INTx\n",
+		pr_debug("%s: error %d reading SDIO_CCCR_INTx\n",
 		       mmc_card_id(card), ret);
 		return ret;
 	}
 
+	if (pending && mmc_card_broken_irq_polling(card) &&
+	    !(host->caps & MMC_CAP_SDIO_IRQ)) {
+		unsigned char dummy;
+
+		/* A fake interrupt could be created when we poll SDIO_CCCR_INTx
+		 * register with a Marvell SD8797 card. A dummy CMD52 read to
+		 * function 0 register 0xff can avoid this.
+		 */
+		mmc_io_rw_direct(card, 0, 0, 0xff, 0, &dummy);
+	}
+
 	count = 0;
 	for (i = 1; i <= 7; i++) {
 		if (pending & (1 << i)) {
-			struct sdio_func *func = card->sdio_func[i - 1];
+			func = card->sdio_func[i - 1];
 			if (!func) {
-				printk(KERN_WARNING "%s: pending IRQ for "
-					"non-existant function\n",
+				pr_warning("%s: pending IRQ for "
+					"non-existent function\n",
 					mmc_card_id(card));
 				ret = -EINVAL;
 			} else if (func->irq_handler) {
 				func->irq_handler(func);
 				count++;
 			} else {
-				printk(KERN_WARNING "%s: pending IRQ with no handler\n",
+				pr_warning("%s: pending IRQ with no handler\n",
 				       sdio_func_id(func));
 				ret = -EINVAL;
 			}
@@ -65,6 +90,15 @@ static int process_sdio_pending_irqs(struct mmc_card *card)
 	return ret;
 }
 
+void sdio_run_irqs(struct mmc_host *host)
+{
+	mmc_claim_host(host);
+	host->sdio_irq_pending = true;
+	process_sdio_pending_irqs(host);
+	mmc_release_host(host);
+}
+EXPORT_SYMBOL_GPL(sdio_run_irqs);
+
 static int sdio_irq_thread(void *_host)
 {
 	struct mmc_host *host = _host;
@@ -104,7 +138,8 @@ static int sdio_irq_thread(void *_host)
 		ret = __mmc_claim_host(host, &host->sdio_irq_thread_abort);
 		if (ret)
 			break;
-		ret = process_sdio_pending_irqs(host->card);
+		ret = process_sdio_pending_irqs(host);
+		host->sdio_irq_pending = false;
 		mmc_release_host(host);
 
 		/*
@@ -134,15 +169,21 @@ static int sdio_irq_thread(void *_host)
 		}
 
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (host->caps & MMC_CAP_SDIO_IRQ)
+		if (host->caps & MMC_CAP_SDIO_IRQ) {
+			mmc_host_clk_hold(host);
 			host->ops->enable_sdio_irq(host, 1);
+			mmc_host_clk_release(host);
+		}
 		if (!kthread_should_stop())
 			schedule_timeout(period);
 		set_current_state(TASK_RUNNING);
 	} while (!kthread_should_stop());
 
-	if (host->caps & MMC_CAP_SDIO_IRQ)
+	if (host->caps & MMC_CAP_SDIO_IRQ) {
+		mmc_host_clk_hold(host);
 		host->ops->enable_sdio_irq(host, 0);
+		mmc_host_clk_release(host);
+	}
 
 	pr_debug("%s: IRQ thread exiting with code %d\n",
 		 mmc_hostname(host), ret);
@@ -157,14 +198,20 @@ static int sdio_card_irq_get(struct mmc_card *card)
 	WARN_ON(!host->claimed);
 
 	if (!host->sdio_irqs++) {
-		atomic_set(&host->sdio_irq_thread_abort, 0);
-		host->sdio_irq_thread =
-			kthread_run(sdio_irq_thread, host, "ksdioirqd/%s",
-				mmc_hostname(host));
-		if (IS_ERR(host->sdio_irq_thread)) {
-			int err = PTR_ERR(host->sdio_irq_thread);
-			host->sdio_irqs--;
-			return err;
+		if (!(host->caps2 & MMC_CAP2_SDIO_IRQ_NOTHREAD)) {
+			atomic_set(&host->sdio_irq_thread_abort, 0);
+			host->sdio_irq_thread =
+				kthread_run(sdio_irq_thread, host,
+					    "ksdioirqd/%s", mmc_hostname(host));
+			if (IS_ERR(host->sdio_irq_thread)) {
+				int err = PTR_ERR(host->sdio_irq_thread);
+				host->sdio_irqs--;
+				return err;
+			}
+		} else {
+			mmc_host_clk_hold(host);
+			host->ops->enable_sdio_irq(host, 1);
+			mmc_host_clk_release(host);
 		}
 	}
 
@@ -179,13 +226,37 @@ static int sdio_card_irq_put(struct mmc_card *card)
 	BUG_ON(host->sdio_irqs < 1);
 
 	if (!--host->sdio_irqs) {
-		atomic_set(&host->sdio_irq_thread_abort, 1);
-		kthread_stop(host->sdio_irq_thread);
+		if (!(host->caps2 & MMC_CAP2_SDIO_IRQ_NOTHREAD)) {
+			atomic_set(&host->sdio_irq_thread_abort, 1);
+			kthread_stop(host->sdio_irq_thread);
+		} else {
+			mmc_host_clk_hold(host);
+			host->ops->enable_sdio_irq(host, 0);
+			mmc_host_clk_release(host);
+		}
 	}
 
 	return 0;
 }
 
+/* If there is only 1 function registered set sdio_single_irq */
+static void sdio_single_irq_set(struct mmc_card *card)
+{
+	struct sdio_func *func;
+	int i;
+
+	card->sdio_single_irq = NULL;
+	if ((card->host->caps & MMC_CAP_SDIO_IRQ) &&
+	    card->host->sdio_irqs == 1)
+		for (i = 0; i < card->sdio_funcs; i++) {
+		       func = card->sdio_func[i];
+		       if (func && func->irq_handler) {
+			       card->sdio_single_irq = func;
+			       break;
+		       }
+	       }
+}
+
 /**
  *	sdio_claim_irq - claim the IRQ for a SDIO function
  *	@func: SDIO function
@@ -227,6 +298,7 @@ int sdio_claim_irq(struct sdio_func *func, sdio_irq_handler_t *handler)
 	ret = sdio_card_irq_get(func->card);
 	if (ret)
 		func->irq_handler = NULL;
+	sdio_single_irq_set(func->card);
 
 	return ret;
 }
@@ -251,6 +323,7 @@ int sdio_release_irq(struct sdio_func *func)
 	if (func->irq_handler) {
 		func->irq_handler = NULL;
 		sdio_card_irq_put(func->card);
+		sdio_single_irq_set(func->card);
 	}
 
 	ret = mmc_io_rw_direct(func->card, 0, 0, SDIO_CCCR_IENx, 0, &reg);
diff --git a/drivers/mmc/core/sdio_ops.c b/drivers/mmc/core/sdio_ops.c
index 4eb7825fd1a..62508b457c4 100644
--- a/drivers/mmc/core/sdio_ops.c
+++ b/drivers/mmc/core/sdio_ops.c
@@ -21,13 +21,11 @@
 
 int mmc_send_io_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int i, err = 0;
 
 	BUG_ON(!host);
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = SD_IO_SEND_OP_COND;
 	cmd.arg = ocr;
 	cmd.flags = MMC_RSP_SPI_R4 | MMC_RSP_R4 | MMC_CMD_BCR;
@@ -67,21 +65,19 @@ int mmc_send_io_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr)
 	return err;
 }
 
-int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
-	unsigned addr, u8 in, u8* out)
+static int mmc_io_rw_direct_host(struct mmc_host *host, int write, unsigned fn,
+	unsigned addr, u8 in, u8 *out)
 {
-	struct mmc_command cmd;
+	struct mmc_command cmd = {0};
 	int err;
 
-	BUG_ON(!card);
+	BUG_ON(!host);
 	BUG_ON(fn > 7);
 
 	/* sanity check */
 	if (addr & ~0x1FFFF)
 		return -EINVAL;
 
-	memset(&cmd, 0, sizeof(struct mmc_command));
-
 	cmd.opcode = SD_IO_RW_DIRECT;
 	cmd.arg = write ? 0x80000000 : 0x00000000;
 	cmd.arg |= fn << 28;
@@ -90,11 +86,11 @@ int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
 	cmd.arg |= in;
 	cmd.flags = MMC_RSP_SPI_R5 | MMC_RSP_R5 | MMC_CMD_AC;
 
-	err = mmc_wait_for_cmd(card->host, &cmd, 0);
+	err = mmc_wait_for_cmd(host, &cmd, 0);
 	if (err)
 		return err;
 
-	if (mmc_host_is_spi(card->host)) {
+	if (mmc_host_is_spi(host)) {
 		/* host driver already reported errors */
 	} else {
 		if (cmd.resp[0] & R5_ERROR)
@@ -106,7 +102,7 @@ int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
 	}
 
 	if (out) {
-		if (mmc_host_is_spi(card->host))
+		if (mmc_host_is_spi(host))
 			*out = (cmd.resp[0] >> 8) & 0xFF;
 		else
 			*out = cmd.resp[0] & 0xFF;
@@ -115,28 +111,32 @@ int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
 	return 0;
 }
 
+int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
+	unsigned addr, u8 in, u8 *out)
+{
+	BUG_ON(!card);
+	return mmc_io_rw_direct_host(card->host, write, fn, addr, in, out);
+}
+
 int mmc_io_rw_extended(struct mmc_card *card, int write, unsigned fn,
 	unsigned addr, int incr_addr, u8 *buf, unsigned blocks, unsigned blksz)
 {
-	struct mmc_request mrq;
-	struct mmc_command cmd;
-	struct mmc_data data;
-	struct scatterlist sg;
+	struct mmc_request mrq = {NULL};
+	struct mmc_command cmd = {0};
+	struct mmc_data data = {0};
+	struct scatterlist sg, *sg_ptr;
+	struct sg_table sgtable;
+	unsigned int nents, left_size, i;
+	unsigned int seg_size = card->host->max_seg_size;
 
 	BUG_ON(!card);
 	BUG_ON(fn > 7);
-	BUG_ON(blocks == 1 && blksz > 512);
-	WARN_ON(blocks == 0);
 	WARN_ON(blksz == 0);
 
 	/* sanity check */
 	if (addr & ~0x1FFFF)
 		return -EINVAL;
 
-	memset(&mrq, 0, sizeof(struct mmc_request));
-	memset(&cmd, 0, sizeof(struct mmc_command));
-	memset(&data, 0, sizeof(struct mmc_data));
-
 	mrq.cmd = &cmd;
 	mrq.data = &data;
 
@@ -145,24 +145,46 @@ int mmc_io_rw_extended(struct mmc_card *card, int write, unsigned fn,
 	cmd.arg |= fn << 28;
 	cmd.arg |= incr_addr ? 0x04000000 : 0x00000000;
 	cmd.arg |= addr << 9;
-	if (blocks == 1 && blksz <= 512)
+	if (blocks == 0)
 		cmd.arg |= (blksz == 512) ? 0 : blksz;	/* byte mode */
 	else
 		cmd.arg |= 0x08000000 | blocks;		/* block mode */
 	cmd.flags = MMC_RSP_SPI_R5 | MMC_RSP_R5 | MMC_CMD_ADTC;
 
 	data.blksz = blksz;
-	data.blocks = blocks;
+	/* Code in host drivers/fwk assumes that "blocks" always is >=1 */
+	data.blocks = blocks ? blocks : 1;
 	data.flags = write ? MMC_DATA_WRITE : MMC_DATA_READ;
-	data.sg = &sg;
-	data.sg_len = 1;
 
-	sg_init_one(&sg, buf, blksz * blocks);
+	left_size = data.blksz * data.blocks;
+	nents = (left_size - 1) / seg_size + 1;
+	if (nents > 1) {
+		if (sg_alloc_table(&sgtable, nents, GFP_KERNEL))
+			return -ENOMEM;
+
+		data.sg = sgtable.sgl;
+		data.sg_len = nents;
+
+		for_each_sg(data.sg, sg_ptr, data.sg_len, i) {
+			sg_set_page(sg_ptr, virt_to_page(buf + (i * seg_size)),
+					min(seg_size, left_size),
+					offset_in_page(buf + (i * seg_size)));
+			left_size = left_size - seg_size;
+		}
+	} else {
+		data.sg = &sg;
+		data.sg_len = 1;
+
+		sg_init_one(&sg, buf, left_size);
+	}
 
 	mmc_set_data_timeout(&data, card);
 
 	mmc_wait_for_req(card->host, &mrq);
 
+	if (nents > 1)
+		sg_free_table(&sgtable);
+
 	if (cmd.error)
 		return cmd.error;
 	if (data.error)
@@ -182,3 +204,20 @@ int mmc_io_rw_extended(struct mmc_card *card, int write, unsigned fn,
 	return 0;
 }
 
+int sdio_reset(struct mmc_host *host)
+{
+	int ret;
+	u8 abort;
+
+	/* SDIO Simplified Specification V2.0, 4.4 Reset for SDIO */
+
+	ret = mmc_io_rw_direct_host(host, 0, 0, SDIO_CCCR_ABORT, 0, &abort);
+	if (ret)
+		abort = 0x08;
+	else
+		abort |= 0x08;
+
+	ret = mmc_io_rw_direct_host(host, 1, 0, SDIO_CCCR_ABORT, abort, NULL);
+	return ret;
+}
+
diff --git a/drivers/mmc/core/sdio_ops.h b/drivers/mmc/core/sdio_ops.h
index e2e74b0d17d..12a4d3ab174 100644
--- a/drivers/mmc/core/sdio_ops.h
+++ b/drivers/mmc/core/sdio_ops.h
@@ -17,6 +17,7 @@ int mmc_io_rw_direct(struct mmc_card *card, int write, unsigned fn,
 	unsigned addr, u8 in, u8* out);
 int mmc_io_rw_extended(struct mmc_card *card, int write, unsigned fn,
 	unsigned addr, int incr_addr, u8 *buf, unsigned blocks, unsigned blksz);
+int sdio_reset(struct mmc_host *host);
 
 #endif
 
diff --git a/drivers/mmc/core/slot-gpio.c b/drivers/mmc/core/slot-gpio.c
new file mode 100644
index 00000000000..5f89cb83d5f
--- /dev/null
+++ b/drivers/mmc/core/slot-gpio.c
@@ -0,0 +1,355 @@
+/*
+ * Generic GPIO card-detect helper
+ *
+ * Copyright (C) 2011, Guennadi Liakhovetski <g.liakhovetski@gmx.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/interrupt.h>
+#include <linux/jiffies.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/slot-gpio.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+struct mmc_gpio {
+	struct gpio_desc *ro_gpio;
+	struct gpio_desc *cd_gpio;
+	bool override_ro_active_level;
+	bool override_cd_active_level;
+	char *ro_label;
+	char cd_label[0];
+};
+
+static irqreturn_t mmc_gpio_cd_irqt(int irq, void *dev_id)
+{
+	/* Schedule a card detection after a debounce timeout */
+	struct mmc_host *host = dev_id;
+
+	host->trigger_card_event = true;
+	mmc_detect_change(host, msecs_to_jiffies(200));
+
+	return IRQ_HANDLED;
+}
+
+static int mmc_gpio_alloc(struct mmc_host *host)
+{
+	size_t len = strlen(dev_name(host->parent)) + 4;
+	struct mmc_gpio *ctx;
+
+	mutex_lock(&host->slot.lock);
+
+	ctx = host->slot.handler_priv;
+	if (!ctx) {
+		/*
+		 * devm_kzalloc() can be called after device_initialize(), even
+		 * before device_add(), i.e., between mmc_alloc_host() and
+		 * mmc_add_host()
+		 */
+		ctx = devm_kzalloc(&host->class_dev, sizeof(*ctx) + 2 * len,
+				   GFP_KERNEL);
+		if (ctx) {
+			ctx->ro_label = ctx->cd_label + len;
+			snprintf(ctx->cd_label, len, "%s cd", dev_name(host->parent));
+			snprintf(ctx->ro_label, len, "%s ro", dev_name(host->parent));
+			host->slot.handler_priv = ctx;
+		}
+	}
+
+	mutex_unlock(&host->slot.lock);
+
+	return ctx ? 0 : -ENOMEM;
+}
+
+int mmc_gpio_get_ro(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+
+	if (!ctx || !ctx->ro_gpio)
+		return -ENOSYS;
+
+	if (ctx->override_ro_active_level)
+		return !gpiod_get_raw_value_cansleep(ctx->ro_gpio) ^
+			!!(host->caps2 & MMC_CAP2_RO_ACTIVE_HIGH);
+
+	return gpiod_get_value_cansleep(ctx->ro_gpio);
+}
+EXPORT_SYMBOL(mmc_gpio_get_ro);
+
+int mmc_gpio_get_cd(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+
+	if (!ctx || !ctx->cd_gpio)
+		return -ENOSYS;
+
+	if (ctx->override_cd_active_level)
+		return !gpiod_get_raw_value_cansleep(ctx->cd_gpio) ^
+			!!(host->caps2 & MMC_CAP2_CD_ACTIVE_HIGH);
+
+	return gpiod_get_value_cansleep(ctx->cd_gpio);
+}
+EXPORT_SYMBOL(mmc_gpio_get_cd);
+
+/**
+ * mmc_gpio_request_ro - request a gpio for write-protection
+ * @host: mmc host
+ * @gpio: gpio number requested
+ *
+ * As devm_* managed functions are used in mmc_gpio_request_ro(), client
+ * drivers do not need to explicitly call mmc_gpio_free_ro() for freeing up,
+ * if the requesting and freeing are only needed at probing and unbinding time
+ * for once.  However, if client drivers do something special like runtime
+ * switching for write-protection, they are responsible for calling
+ * mmc_gpio_request_ro() and mmc_gpio_free_ro() as a pair on their own.
+ *
+ * Returns zero on success, else an error.
+ */
+int mmc_gpio_request_ro(struct mmc_host *host, unsigned int gpio)
+{
+	struct mmc_gpio *ctx;
+	int ret;
+
+	if (!gpio_is_valid(gpio))
+		return -EINVAL;
+
+	ret = mmc_gpio_alloc(host);
+	if (ret < 0)
+		return ret;
+
+	ctx = host->slot.handler_priv;
+
+	ret = devm_gpio_request_one(&host->class_dev, gpio, GPIOF_DIR_IN,
+				    ctx->ro_label);
+	if (ret < 0)
+		return ret;
+
+	ctx->override_ro_active_level = true;
+	ctx->ro_gpio = gpio_to_desc(gpio);
+
+	return 0;
+}
+EXPORT_SYMBOL(mmc_gpio_request_ro);
+
+void mmc_gpiod_request_cd_irq(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+	int ret, irq;
+
+	if (host->slot.cd_irq >= 0 || !ctx || !ctx->cd_gpio)
+		return;
+
+	irq = gpiod_to_irq(ctx->cd_gpio);
+
+	/*
+	 * Even if gpiod_to_irq() returns a valid IRQ number, the platform might
+	 * still prefer to poll, e.g., because that IRQ number is already used
+	 * by another unit and cannot be shared.
+	 */
+	if (irq >= 0 && host->caps & MMC_CAP_NEEDS_POLL)
+		irq = -EINVAL;
+
+	if (irq >= 0) {
+		ret = devm_request_threaded_irq(&host->class_dev, irq,
+			NULL, mmc_gpio_cd_irqt,
+			IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
+			ctx->cd_label, host);
+		if (ret < 0)
+			irq = ret;
+	}
+
+	host->slot.cd_irq = irq;
+
+	if (irq < 0)
+		host->caps |= MMC_CAP_NEEDS_POLL;
+}
+EXPORT_SYMBOL(mmc_gpiod_request_cd_irq);
+
+/**
+ * mmc_gpio_request_cd - request a gpio for card-detection
+ * @host: mmc host
+ * @gpio: gpio number requested
+ * @debounce: debounce time in microseconds
+ *
+ * As devm_* managed functions are used in mmc_gpio_request_cd(), client
+ * drivers do not need to explicitly call mmc_gpio_free_cd() for freeing up,
+ * if the requesting and freeing are only needed at probing and unbinding time
+ * for once.  However, if client drivers do something special like runtime
+ * switching for card-detection, they are responsible for calling
+ * mmc_gpio_request_cd() and mmc_gpio_free_cd() as a pair on their own.
+ *
+ * If GPIO debouncing is desired, set the debounce parameter to a non-zero
+ * value. The caller is responsible for ensuring that the GPIO driver associated
+ * with the GPIO supports debouncing, otherwise an error will be returned.
+ *
+ * Returns zero on success, else an error.
+ */
+int mmc_gpio_request_cd(struct mmc_host *host, unsigned int gpio,
+			unsigned int debounce)
+{
+	struct mmc_gpio *ctx;
+	int ret;
+
+	ret = mmc_gpio_alloc(host);
+	if (ret < 0)
+		return ret;
+
+	ctx = host->slot.handler_priv;
+
+	ret = devm_gpio_request_one(&host->class_dev, gpio, GPIOF_DIR_IN,
+				    ctx->cd_label);
+	if (ret < 0)
+		/*
+		 * don't bother freeing memory. It might still get used by other
+		 * slot functions, in any case it will be freed, when the device
+		 * is destroyed.
+		 */
+		return ret;
+
+	if (debounce) {
+		ret = gpio_set_debounce(gpio, debounce);
+		if (ret < 0)
+			return ret;
+	}
+
+	ctx->override_cd_active_level = true;
+	ctx->cd_gpio = gpio_to_desc(gpio);
+
+	mmc_gpiod_request_cd_irq(host);
+
+	return 0;
+}
+EXPORT_SYMBOL(mmc_gpio_request_cd);
+
+/**
+ * mmc_gpio_free_ro - free the write-protection gpio
+ * @host: mmc host
+ *
+ * It's provided only for cases that client drivers need to manually free
+ * up the write-protection gpio requested by mmc_gpio_request_ro().
+ */
+void mmc_gpio_free_ro(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+	int gpio;
+
+	if (!ctx || !ctx->ro_gpio)
+		return;
+
+	gpio = desc_to_gpio(ctx->ro_gpio);
+	ctx->ro_gpio = NULL;
+
+	devm_gpio_free(&host->class_dev, gpio);
+}
+EXPORT_SYMBOL(mmc_gpio_free_ro);
+
+/**
+ * mmc_gpio_free_cd - free the card-detection gpio
+ * @host: mmc host
+ *
+ * It's provided only for cases that client drivers need to manually free
+ * up the card-detection gpio requested by mmc_gpio_request_cd().
+ */
+void mmc_gpio_free_cd(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+	int gpio;
+
+	if (!ctx || !ctx->cd_gpio)
+		return;
+
+	if (host->slot.cd_irq >= 0) {
+		devm_free_irq(&host->class_dev, host->slot.cd_irq, host);
+		host->slot.cd_irq = -EINVAL;
+	}
+
+	gpio = desc_to_gpio(ctx->cd_gpio);
+	ctx->cd_gpio = NULL;
+
+	devm_gpio_free(&host->class_dev, gpio);
+}
+EXPORT_SYMBOL(mmc_gpio_free_cd);
+
+/**
+ * mmc_gpiod_request_cd - request a gpio descriptor for card-detection
+ * @host: mmc host
+ * @con_id: function within the GPIO consumer
+ * @idx: index of the GPIO to obtain in the consumer
+ * @override_active_level: ignore %GPIO_ACTIVE_LOW flag
+ * @debounce: debounce time in microseconds
+ *
+ * Use this function in place of mmc_gpio_request_cd() to use the GPIO
+ * descriptor API.  Note that it is paired with mmc_gpiod_free_cd() not
+ * mmc_gpio_free_cd().  Note also that it must be called prior to mmc_add_host()
+ * otherwise the caller must also call mmc_gpiod_request_cd_irq().
+ *
+ * Returns zero on success, else an error.
+ */
+int mmc_gpiod_request_cd(struct mmc_host *host, const char *con_id,
+			 unsigned int idx, bool override_active_level,
+			 unsigned int debounce)
+{
+	struct mmc_gpio *ctx;
+	struct gpio_desc *desc;
+	int ret;
+
+	ret = mmc_gpio_alloc(host);
+	if (ret < 0)
+		return ret;
+
+	ctx = host->slot.handler_priv;
+
+	if (!con_id)
+		con_id = ctx->cd_label;
+
+	desc = devm_gpiod_get_index(host->parent, con_id, idx);
+	if (IS_ERR(desc))
+		return PTR_ERR(desc);
+
+	ret = gpiod_direction_input(desc);
+	if (ret < 0)
+		return ret;
+
+	if (debounce) {
+		ret = gpiod_set_debounce(desc, debounce);
+		if (ret < 0)
+			return ret;
+	}
+
+	ctx->override_cd_active_level = override_active_level;
+	ctx->cd_gpio = desc;
+
+	return 0;
+}
+EXPORT_SYMBOL(mmc_gpiod_request_cd);
+
+/**
+ * mmc_gpiod_free_cd - free the card-detection gpio descriptor
+ * @host: mmc host
+ *
+ * It's provided only for cases that client drivers need to manually free
+ * up the card-detection gpio requested by mmc_gpiod_request_cd().
+ */
+void mmc_gpiod_free_cd(struct mmc_host *host)
+{
+	struct mmc_gpio *ctx = host->slot.handler_priv;
+
+	if (!ctx || !ctx->cd_gpio)
+		return;
+
+	if (host->slot.cd_irq >= 0) {
+		devm_free_irq(&host->class_dev, host->slot.cd_irq, host);
+		host->slot.cd_irq = -EINVAL;
+	}
+
+	devm_gpiod_put(&host->class_dev, ctx->cd_gpio);
+
+	ctx->cd_gpio = NULL;
+}
+EXPORT_SYMBOL(mmc_gpiod_free_cd);
diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index b4cf691f3f6..a5652548230 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -44,6 +44,19 @@ config MMC_SDHCI_IO_ACCESSORS
 	  This is silent Kconfig symbol that is selected by the drivers that
 	  need to overwrite SDHCI IO memory accessors.
 
+config MMC_SDHCI_BIG_ENDIAN_32BIT_BYTE_SWAPPER
+	bool
+	select MMC_SDHCI_IO_ACCESSORS
+	help
+	  This option is selected by drivers running on big endian hosts
+	  and performing I/O to a SDHCI controller through a bus that
+	  implements a hardware byte swapper using a 32-bit datum.
+	  This endian mapping mode is called "data invariance" and
+	  has the effect of scrambling the addresses and formats of data
+	  accessed in sizes other than the datum size.
+
+	  This is the case for the Freescale eSDHC and Nintendo Wii SDHCI.
+
 config MMC_SDHCI_PCI
 	tristate "SDHCI support on PCI bus"
 	depends on MMC_SDHCI && PCI
@@ -56,37 +69,231 @@ config MMC_SDHCI_PCI
 	  If unsure, say N.
 
 config MMC_RICOH_MMC
-	tristate "Ricoh MMC Controller Disabler  (EXPERIMENTAL)"
+	bool "Ricoh MMC Controller Disabler"
 	depends on MMC_SDHCI_PCI
 	help
-	  This selects the disabler for the Ricoh MMC Controller. This
+	  This adds a pci quirk to disable Ricoh MMC Controller. This
 	  proprietary controller is unnecessary because the SDHCI driver
 	  supports MMC cards on the SD controller, but if it is not
 	  disabled, it will steal the MMC cards away - rendering them
-	  useless. It is safe to select this driver even if you don't
+	  useless. It is safe to select this even if you don't
 	  have a Ricoh based card reader.
 
+	  If unsure, say Y.
+
+config MMC_SDHCI_ACPI
+	tristate "SDHCI support for ACPI enumerated SDHCI controllers"
+	depends on MMC_SDHCI && ACPI
+	help
+	  This selects support for ACPI enumerated SDHCI controllers,
+	  identified by ACPI Compatibility ID PNP0D40 or specific
+	  ACPI Hardware IDs.
+
+	  If you have a controller with this interface, say Y or M here.
 
-	  To compile this driver as a module, choose M here:
-	  the module will be called ricoh_mmc.
+	  If unsure, say N.
 
-	  If unsure, say Y.
+config MMC_SDHCI_PLTFM
+	tristate "SDHCI platform and OF driver helper"
+	depends on MMC_SDHCI
+	help
+	  This selects the common helper functions support for Secure Digital
+	  Host Controller Interface based platform and OF drivers.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_OF_ARASAN
+	tristate "SDHCI OF support for the Arasan SDHCI controllers"
+	depends on MMC_SDHCI_PLTFM
+	depends on OF
+	help
+	  This selects the Arasan Secure Digital Host Controller Interface
+	  (SDHCI). This hardware is found e.g. in Xilinx' Zynq SoC.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_OF_ESDHC
+	tristate "SDHCI OF support for the Freescale eSDHC controller"
+	depends on MMC_SDHCI_PLTFM
+	depends on PPC_OF
+	select MMC_SDHCI_BIG_ENDIAN_32BIT_BYTE_SWAPPER
+	help
+	  This selects the Freescale eSDHC controller support.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_OF_HLWD
+	tristate "SDHCI OF support for the Nintendo Wii SDHCI controllers"
+	depends on MMC_SDHCI_PLTFM
+	depends on PPC_OF
+	select MMC_SDHCI_BIG_ENDIAN_32BIT_BYTE_SWAPPER
+	help
+	  This selects the Secure Digital Host Controller Interface (SDHCI)
+	  found in the "Hollywood" chipset of the Nintendo Wii video game
+	  console.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_CNS3XXX
+	tristate "SDHCI support on the Cavium Networks CNS3xxx SoC"
+	depends on ARCH_CNS3XXX
+	depends on MMC_SDHCI_PLTFM
+	help
+	  This selects the SDHCI support for CNS3xxx System-on-Chip devices.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_ESDHC_IMX
+	tristate "SDHCI support for the Freescale eSDHC/uSDHC i.MX controller"
+	depends on ARCH_MXC
+	depends on MMC_SDHCI_PLTFM
+	select MMC_SDHCI_IO_ACCESSORS
+	help
+	  This selects the Freescale eSDHC/uSDHC controller support
+	  found on i.MX25, i.MX35 i.MX5x and i.MX6x.
+
+	  If you have a controller with this interface, say Y or M here.
 
-config MMC_SDHCI_OF
-	tristate "SDHCI support on OpenFirmware platforms"
-	depends on MMC_SDHCI && PPC_OF
+	  If unsure, say N.
+
+config MMC_SDHCI_DOVE
+	tristate "SDHCI support on Marvell's Dove SoC"
+	depends on ARCH_DOVE || MACH_DOVE
+	depends on MMC_SDHCI_PLTFM
+	select MMC_SDHCI_IO_ACCESSORS
+	help
+	  This selects the Secure Digital Host Controller Interface in
+	  Marvell's Dove SoC.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_TEGRA
+	tristate "SDHCI platform support for the Tegra SD/MMC Controller"
+	depends on ARCH_TEGRA
+	depends on MMC_SDHCI_PLTFM
+	select MMC_SDHCI_IO_ACCESSORS
+	help
+	  This selects the Tegra SD/MMC controller. If you have a Tegra
+	  platform with SD or MMC devices, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_S3C
+	tristate "SDHCI support on Samsung S3C SoC"
+	depends on MMC_SDHCI && PLAT_SAMSUNG
+	help
+	  This selects the Secure Digital Host Controller Interface (SDHCI)
+	  often referrered to as the HSMMC block in some of the Samsung S3C
+	  range of SoC.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_SIRF
+	tristate "SDHCI support on CSR SiRFprimaII and SiRFmarco SoCs"
+	depends on ARCH_SIRF
+	depends on MMC_SDHCI_PLTFM
+	help
+	  This selects the SDHCI support for SiRF System-on-Chip devices.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_PXAV3
+	tristate "Marvell MMP2 SD Host Controller support (PXAV3)"
+	depends on CLKDEV_LOOKUP
+	depends on MMC_SDHCI_PLTFM
+	default CPU_MMP2
+	help
+	  This selects the Marvell(R) PXAV3 SD Host Controller.
+	  If you have a MMP2 platform with SD Host Controller
+	  and a card slot, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_PXAV2
+	tristate "Marvell PXA9XX SD Host Controller support (PXAV2)"
+	depends on CLKDEV_LOOKUP
+	depends on MMC_SDHCI_PLTFM
+	default CPU_PXA910
+	help
+	  This selects the Marvell(R) PXAV2 SD Host Controller.
+	  If you have a PXA9XX platform with SD Host Controller
+	  and a card slot, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_SPEAR
+	tristate "SDHCI support on ST SPEAr platform"
+	depends on MMC_SDHCI && PLAT_SPEAR
+	help
+	  This selects the Secure Digital Host Controller Interface (SDHCI)
+	  often referrered to as the HSMMC block in some of the ST SPEAR range
+	  of SoC
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SDHCI_S3C_DMA
+	bool "DMA support on S3C SDHCI"
+	depends on MMC_SDHCI_S3C
+	help
+	  Enable DMA support on the Samsung S3C SDHCI glue. The DMA
+	  has proved to be problematic if the controller encounters
+	  certain errors, and thus should be treated with care.
+
+	  YMMV.
+
+config MMC_SDHCI_BCM_KONA
+	tristate "SDHCI support on Broadcom KONA platform"
+	depends on ARCH_BCM_MOBILE
+	depends on MMC_SDHCI_PLTFM
+	help
+	  This selects the Broadcom Kona Secure Digital Host Controller
+	  Interface(SDHCI) support.
+	  This is used in Broadcom mobile SoCs.
+
+	  If you have a controller with this interface, say Y or M here.
+
+config MMC_SDHCI_BCM2835
+	tristate "SDHCI platform support for the BCM2835 SD/MMC Controller"
+	depends on ARCH_BCM2835
+	depends on MMC_SDHCI_PLTFM
 	select MMC_SDHCI_IO_ACCESSORS
 	help
-	  This selects the OF support for Secure Digital Host Controller
-	  Interfaces. So far, only the Freescale eSDHC controller is known
-	  to exist on OF platforms.
+	  This selects the BCM2835 SD/MMC controller. If you have a BCM2835
+	  platform with SD or MMC devices, say Y or M here.
 
 	  If unsure, say N.
 
+config MMC_MOXART
+	tristate "MOXART SD/MMC Host Controller support"
+	depends on ARCH_MOXART && MMC
+	help
+	  This selects support for the MOXART SD/MMC Host Controller.
+	  MOXA provides one multi-functional card reader which can
+	  be found on some embedded hardware such as UC-7112-LX.
+	  If you have a controller with this interface, say Y here.
+
 config MMC_OMAP
 	tristate "TI OMAP Multimedia Card Interface support"
 	depends on ARCH_OMAP
-	select TPS65010 if MACH_OMAP_H2
+	depends on TPS65010 || !MACH_OMAP_H2
 	help
 	  This selects the TI OMAP Multimedia card Interface.
 	  If you have an OMAP board with a Multimedia Card slot,
@@ -96,10 +303,10 @@ config MMC_OMAP
 
 config MMC_OMAP_HS
 	tristate "TI OMAP High Speed Multimedia Card Interface support"
-	depends on ARCH_OMAP2430 || ARCH_OMAP3
+	depends on ARCH_OMAP2PLUS || COMPILE_TEST
 	help
 	  This selects the TI OMAP High Speed Multimedia card Interface.
-	  If you have an OMAP2430 or OMAP3 board with a Multimedia Card slot,
+	  If you have an omap2plus board with a Multimedia Card slot,
 	  say Y or M here.
 
 	  If unsure, say N.
@@ -117,65 +324,66 @@ config MMC_WBSD
 
 config MMC_AU1X
 	tristate "Alchemy AU1XX0 MMC Card Interface support"
-	depends on SOC_AU1200
+	depends on MIPS_ALCHEMY
 	help
 	  This selects the AMD Alchemy(R) Multimedia card interface.
 	  If you have a Alchemy platform with a MMC slot, say Y or M here.
 
 	  If unsure, say N.
 
-config MMC_AT91
-	tristate "AT91 SD/MMC Card Interface support"
-	depends on ARCH_AT91
-	help
-	  This selects the AT91 MCI controller.
-
-	  If unsure, say N.
-
 config MMC_ATMELMCI
-	tristate "Atmel Multimedia Card Interface support"
-	depends on AVR32
+	tristate "Atmel SD/MMC Driver (Multimedia Card Interface)"
+	depends on AVR32 || ARCH_AT91
 	help
 	  This selects the Atmel Multimedia Card Interface driver. If
-	  you have an AT32 (AVR32) platform with a Multimedia Card
-	  slot, say Y or M here.
+	  you have an AT32 (AVR32) or AT91 platform with a Multimedia
+	  Card slot, say Y or M here.
 
 	  If unsure, say N.
 
-config MMC_ATMELMCI_DMA
-	bool "Atmel MCI DMA support (EXPERIMENTAL)"
-	depends on MMC_ATMELMCI && DMA_ENGINE && EXPERIMENTAL
+config MMC_SDHCI_MSM
+	tristate "Qualcomm SDHCI Controller Support"
+	depends on ARCH_QCOM
+	depends on MMC_SDHCI_PLTFM
 	help
-	  Say Y here to have the Atmel MCI driver use a DMA engine to
-	  do data transfers and thus increase the throughput and
-	  reduce the CPU utilization. Note that this is highly
-	  experimental and may cause the driver to lock up.
+	  This selects the Secure Digital Host Controller Interface (SDHCI)
+	  support present in Qualcomm SOCs. The controller supports
+	  SD/MMC/SDIO devices.
+
+	  If you have a controller with this interface, say Y or M here.
 
 	  If unsure, say N.
 
-config MMC_IMX
-	tristate "Motorola i.MX Multimedia Card Interface support"
-	depends on ARCH_IMX
+config MMC_MSM
+	tristate "Qualcomm SDCC Controller Support"
+	depends on MMC && (ARCH_MSM7X00A || ARCH_MSM7X30 || ARCH_QSD8X50)
 	help
-	  This selects the Motorola i.MX Multimedia card Interface.
-	  If you have a i.MX platform with a Multimedia Card slot,
-	  say Y or M here.
+	  This provides support for the SD/MMC cell found in the
+	  MSM and QSD SOCs from Qualcomm. The controller also has
+	  support for SDIO devices.
+
+config MMC_MXC
+	tristate "Freescale i.MX21/27/31 or MPC512x Multimedia Card support"
+	depends on ARCH_MXC || PPC_MPC512x
+	help
+	  This selects the Freescale i.MX21, i.MX27, i.MX31 or MPC512x
+	  Multimedia Card Interface. If you have an i.MX or MPC512x platform
+	  with a Multimedia Card slot, say Y or M here.
 
 	  If unsure, say N.
 
-config MMC_MXC
-	tristate "Freescale i.MX2/3 Multimedia Card Interface support"
-	depends on ARCH_MXC
+config MMC_MXS
+	tristate "Freescale MXS Multimedia Card Interface support"
+	depends on ARCH_MXS && MXS_DMA
 	help
-	  This selects the Freescale i.MX2/3 Multimedia card Interface.
-	  If you have a i.MX platform with a Multimedia Card slot,
-	  say Y or M here.
+	  This selects the Freescale SSP MMC controller found on MXS based
+	  platforms like mx23/28.
 
 	  If unsure, say N.
 
 config MMC_TIFM_SD
-	tristate "TI Flash Media MMC/SD Interface support  (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PCI
+	tristate "TI Flash Media MMC/SD Interface support"
+	depends on PCI
 	select TIFM_CORE
 	help
 	  Say Y here if you want to be able to access MMC/SD cards with
@@ -200,6 +408,21 @@ config MMC_MVSDIO
 	  To compile this driver as a module, choose M here: the
 	  module will be called mvsdio.
 
+config MMC_DAVINCI
+        tristate "TI DAVINCI Multimedia Card Interface support"
+        depends on ARCH_DAVINCI
+        help
+          This selects the TI DAVINCI Multimedia card Interface.
+          If you have an DAVINCI board with a Multimedia Card slot,
+          say Y or M here.  If unsure, say N.
+
+config MMC_GOLDFISH
+	tristate "goldfish qemu Multimedia Card Interface support"
+	depends on GOLDFISH
+	help
+	  This selects the Goldfish Multimedia card Interface emulation
+	  found on the Goldfish Android virtual device emulation.
+
 config MMC_SPI
 	tristate "MMC/SD/SDIO over SPI"
 	depends on SPI_MASTER && !HIGHMEM && HAS_DMA
@@ -216,7 +439,7 @@ config MMC_SPI
 
 config MMC_S3C
 	tristate "Samsung S3C SD/MMC Card Interface support"
-	depends on ARCH_S3C2410
+	depends on ARCH_S3C24XX
 	help
 	  This selects a driver for the MCI interface found in
           Samsung's S3C2410, S3C2412, S3C2440, S3C2442 CPUs.
@@ -225,9 +448,49 @@ config MMC_S3C
 
 	  If unsure, say N.
 
+config MMC_S3C_HW_SDIO_IRQ
+       bool "Hardware support for SDIO IRQ"
+       depends on MMC_S3C
+       help
+         Enable the hardware support for SDIO interrupts instead of using
+	 the generic polling code.
+
+choice
+	prompt "Samsung S3C SD/MMC transfer code"
+	depends on MMC_S3C
+
+config MMC_S3C_PIO
+	bool "Use PIO transfers only"
+	help
+	  Use PIO to transfer data between memory and the hardware.
+
+	  PIO is slower than DMA as it requires CPU instructions to
+	  move the data. This has been the traditional default for
+	  the S3C MCI driver.
+
+config MMC_S3C_DMA
+	bool "Use DMA transfers only"
+	help
+	  Use DMA to transfer data between memory and the hardare.
+
+	  Currently, the DMA support in this driver seems to not be
+	  working properly and needs to be debugged before this
+	  option is useful.
+
+config MMC_S3C_PIODMA
+	bool "Support for both PIO and DMA"
+	help
+	  Compile both the PIO and DMA transfer routines into the
+	  driver and let the platform select at run-time which one
+	  is best.
+
+	  See notes for the DMA option.
+
+endchoice
+
 config MMC_SDRICOH_CS
-	tristate "MMC/SD driver for Ricoh Bay1Controllers (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PCI && PCMCIA
+	tristate "MMC/SD driver for Ricoh Bay1Controllers"
+	depends on PCI && PCMCIA
 	help
 	  Say Y here if your Notebook reports a Ricoh Bay1Controller PCMCIA
 	  card whenever you insert a MMC or SD card into the card slot.
@@ -235,9 +498,226 @@ config MMC_SDRICOH_CS
 	  To compile this driver as a module, choose M here: the
 	  module will be called sdricoh_cs.
 
+config MMC_TMIO_CORE
+	tristate
+
 config MMC_TMIO
 	tristate "Toshiba Mobile IO Controller (TMIO) MMC/SD function support"
-	depends on MFD_TMIO
+	depends on MFD_TMIO || MFD_ASIC3
+	select MMC_TMIO_CORE
 	help
 	  This provides support for the SD/MMC cell found in TC6393XB,
-	  T7L66XB and also ipaq ASIC3
+	  T7L66XB and also HTC ASIC3
+
+config MMC_SDHI
+	tristate "SH-Mobile SDHI SD/SDIO controller support"
+	depends on SUPERH || ARM
+	depends on SUPERH || ARCH_SHMOBILE || COMPILE_TEST
+	select MMC_TMIO_CORE
+	help
+	  This provides support for the SDHI SD/SDIO controller found in
+	  SuperH and ARM SH-Mobile SoCs
+
+config MMC_CB710
+	tristate "ENE CB710 MMC/SD Interface support"
+	depends on PCI
+	select CB710_CORE
+	help
+	  This option enables support for MMC/SD part of ENE CB710/720 Flash
+	  memory card reader found in some laptops (ie. some versions of
+	  HP Compaq nx9500).
+
+	  This driver can also be built as a module. If so, the module
+	  will be called cb710-mmc.
+
+config MMC_VIA_SDMMC
+	tristate "VIA SD/MMC Card Reader Driver"
+	depends on PCI
+	help
+	  This selects the VIA SD/MMC Card Reader driver, say Y or M here.
+	  VIA provides one multi-functional card reader which integrated into
+	  some motherboards manufactured by VIA. This card reader supports
+	  SD/MMC/SDHC.
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config SDH_BFIN
+	tristate "Blackfin Secure Digital Host support"
+	depends on (BF54x && !BF544) || (BF51x && !BF512)
+	help
+	  If you say yes here you will get support for the Blackfin on-chip
+	  Secure Digital Host interface.  This includes support for MMC and
+	  SD cards.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called bfin_sdh.
+
+	  If unsure, say N.
+
+config SDH_BFIN_MISSING_CMD_PULLUP_WORKAROUND
+	bool "Blackfin EZkit Missing SDH_CMD Pull Up Resistor Workaround"
+	depends on SDH_BFIN
+	help
+	  If you say yes here SD-Cards may work on the EZkit.
+
+config MMC_DW
+	tristate "Synopsys DesignWare Memory Card Interface"
+	depends on ARC || ARM
+	help
+	  This selects support for the Synopsys DesignWare Mobile Storage IP
+	  block, this provides host support for SD and MMC interfaces, in both
+	  PIO and external DMA modes.
+
+config MMC_DW_IDMAC
+	bool "Internal DMAC interface"
+	depends on MMC_DW
+	help
+	  This selects support for the internal DMAC block within the Synopsys
+	  Designware Mobile Storage IP block. This disables the external DMA
+	  interface.
+
+config MMC_DW_PLTFM
+	tristate "Synopsys Designware MCI Support as platform device"
+	depends on MMC_DW
+	default y
+	help
+	  This selects the common helper functions support for Host Controller
+	  Interface based platform driver. Please select this option if the IP
+	  is present as a platform device. This is the common interface for the
+	  Synopsys Designware IP.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say Y.
+
+config MMC_DW_EXYNOS
+	tristate "Exynos specific extensions for Synopsys DW Memory Card Interface"
+	depends on MMC_DW
+	select MMC_DW_PLTFM
+	help
+	  This selects support for Samsung Exynos SoC specific extensions to the
+	  Synopsys DesignWare Memory Card Interface driver. Select this option
+	  for platforms based on Exynos4 and Exynos5 SoC's.
+
+config MMC_DW_K3
+	tristate "K3 specific extensions for Synopsys DW Memory Card Interface"
+	depends on MMC_DW
+	select MMC_DW_PLTFM
+	select MMC_DW_IDMAC
+	help
+	  This selects support for Hisilicon K3 SoC specific extensions to the
+	  Synopsys DesignWare Memory Card Interface driver. Select this option
+	  for platforms based on Hisilicon K3 SoC's.
+
+config MMC_DW_PCI
+	tristate "Synopsys Designware MCI support on PCI bus"
+	depends on MMC_DW && PCI
+	help
+	  This selects the PCI bus for the Synopsys Designware Mobile Storage IP.
+	  Select this option if the IP is present on PCI platform.
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
+config MMC_SH_MMCIF
+	tristate "SuperH Internal MMCIF support"
+	depends on MMC_BLOCK
+	depends on SUPERH || ARCH_SHMOBILE || COMPILE_TEST
+	help
+	  This selects the MMC Host Interface controller (MMCIF).
+
+	  This driver supports MMCIF in sh7724/sh7757/sh7372.
+
+config MMC_JZ4740
+	tristate "JZ4740 SD/Multimedia Card Interface support"
+	depends on MACH_JZ4740
+	help
+	  This selects support for the SD/MMC controller on Ingenic JZ4740
+	  SoCs.
+	  If you have a board based on such a SoC and with a SD/MMC slot,
+	  say Y or M here.
+
+config MMC_VUB300
+	tristate "VUB300 USB to SDIO/SD/MMC Host Controller support"
+	depends on USB
+	help
+	  This selects support for Elan Digital Systems' VUB300 chip.
+
+	  The VUB300 is a USB-SDIO Host Controller Interface chip
+	  that enables the host computer to use SDIO/SD/MMC cards
+	  via a USB 2.0 or USB 1.1 host.
+
+	  The VUB300 chip will be found in both physically separate
+	  USB to SDIO/SD/MMC adapters and embedded on some motherboards.
+
+	  The VUB300 chip supports SD and MMC memory cards in addition
+	  to single and multifunction SDIO cards.
+
+	  Some SDIO cards will need a firmware file to be loaded and
+	  sent to VUB300 chip in order to achieve better data throughput.
+	  Download these "Offload Pseudocode" from Elan Digital Systems'
+	  web-site http://www.elandigitalsystems.com/support/downloads.php
+	  and put them in /lib/firmware. Note that without these additional
+	  firmware files the VUB300 chip will still function, but not at
+	  the best obtainable data rate.
+
+	  To compile this mmc host controller driver as a module,
+	  choose M here: the module will be called vub300.
+
+	  If you have a computer with an embedded VUB300 chip
+	  or if you intend connecting a USB adapter based on a
+	  VUB300 chip say Y or M here.
+
+config MMC_USHC
+	tristate "USB SD Host Controller (USHC) support"
+	depends on USB
+	help
+	  This selects support for USB SD Host Controllers based on
+	  the Cypress Astoria chip with firmware compliant with CSR's
+	  USB SD Host Controller specification (CS-118793-SP).
+
+	  CSR boards with this device include: USB<>SDIO (M1985v2),
+	  and Ultrasira.
+
+	  Note: These controllers only support SDIO cards and do not
+	  support MMC or SD memory cards.
+
+config MMC_WMT
+	tristate "Wondermedia SD/MMC Host Controller support"
+	depends on ARCH_VT8500
+	default y
+	help
+	  This selects support for the SD/MMC Host Controller on
+	  Wondermedia WM8505/WM8650 based SoCs.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called wmt-sdmmc.
+
+config MMC_USDHI6ROL0
+	tristate "Renesas USDHI6ROL0 SD/SDIO Host Controller support"
+	help
+	  This selects support for the Renesas USDHI6ROL0 SD/SDIO
+	  Host Controller
+
+config MMC_REALTEK_PCI
+	tristate "Realtek PCI-E SD/MMC Card Interface Driver"
+	depends on MFD_RTSX_PCI
+	help
+	  Say Y here to include driver code to support SD/MMC card interface
+	  of Realtek PCI-E card reader
+
+config MMC_REALTEK_USB
+	tristate "Realtek USB SD/MMC Card Interface Driver"
+	depends on MFD_RTSX_USB
+	help
+	  Say Y here to include driver code to support SD/MMC card interface
+	  of Realtek RTS5129/39 series card reader
+
+config MMC_SUNXI
+	tristate "Allwinner sunxi SD/MMC Host Controller support"
+	depends on ARCH_SUNXI
+	help
+	  This selects support for the SD/MMC Host Controller on
+	  Allwinner sunxi SoCs.
diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
index 970a997620e..7f81ddf1dd2 100644
--- a/drivers/mmc/host/Makefile
+++ b/drivers/mmc/host/Makefile
@@ -2,26 +2,30 @@
 # Makefile for MMC/SD host controller drivers
 #
 
-ifeq ($(CONFIG_MMC_DEBUG),y)
-	EXTRA_CFLAGS		+= -DDEBUG
-endif
-
 obj-$(CONFIG_MMC_ARMMMCI)	+= mmci.o
 obj-$(CONFIG_MMC_PXA)		+= pxamci.o
-obj-$(CONFIG_MMC_IMX)		+= imxmmc.o
 obj-$(CONFIG_MMC_MXC)		+= mxcmmc.o
+obj-$(CONFIG_MMC_MXS)		+= mxs-mmc.o
 obj-$(CONFIG_MMC_SDHCI)		+= sdhci.o
 obj-$(CONFIG_MMC_SDHCI_PCI)	+= sdhci-pci.o
-obj-$(CONFIG_MMC_RICOH_MMC)	+= ricoh_mmc.o
-obj-$(CONFIG_MMC_SDHCI_OF)	+= sdhci-of.o
+obj-$(subst m,y,$(CONFIG_MMC_SDHCI_PCI))	+= sdhci-pci-data.o
+obj-$(subst m,y,$(CONFIG_MMC_SDHCI_PCI))	+= sdhci-pci-o2micro.o
+obj-$(CONFIG_MMC_SDHCI_ACPI)	+= sdhci-acpi.o
+obj-$(CONFIG_MMC_SDHCI_PXAV3)	+= sdhci-pxav3.o
+obj-$(CONFIG_MMC_SDHCI_PXAV2)	+= sdhci-pxav2.o
+obj-$(CONFIG_MMC_SDHCI_S3C)	+= sdhci-s3c.o
+obj-$(CONFIG_MMC_SDHCI_SIRF)   	+= sdhci-sirf.o
+obj-$(CONFIG_MMC_SDHCI_SPEAR)	+= sdhci-spear.o
 obj-$(CONFIG_MMC_WBSD)		+= wbsd.o
 obj-$(CONFIG_MMC_AU1X)		+= au1xmmc.o
 obj-$(CONFIG_MMC_OMAP)		+= omap.o
 obj-$(CONFIG_MMC_OMAP_HS)	+= omap_hsmmc.o
-obj-$(CONFIG_MMC_AT91)		+= at91_mci.o
 obj-$(CONFIG_MMC_ATMELMCI)	+= atmel-mci.o
 obj-$(CONFIG_MMC_TIFM_SD)	+= tifm_sd.o
+obj-$(CONFIG_MMC_MSM)		+= msm_sdcc.o
 obj-$(CONFIG_MMC_MVSDIO)	+= mvsdio.o
+obj-$(CONFIG_MMC_DAVINCI)       += davinci_mmc.o
+obj-$(CONFIG_MMC_GOLDFISH)	+= android-goldfish.o
 obj-$(CONFIG_MMC_SPI)		+= mmc_spi.o
 ifeq ($(CONFIG_OF),y)
 obj-$(CONFIG_MMC_SPI)		+= of_mmc_spi.o
@@ -29,4 +33,42 @@ endif
 obj-$(CONFIG_MMC_S3C)   	+= s3cmci.o
 obj-$(CONFIG_MMC_SDRICOH_CS)	+= sdricoh_cs.o
 obj-$(CONFIG_MMC_TMIO)		+= tmio_mmc.o
+obj-$(CONFIG_MMC_TMIO_CORE)	+= tmio_mmc_core.o
+tmio_mmc_core-y			:= tmio_mmc_pio.o
+tmio_mmc_core-$(subst m,y,$(CONFIG_MMC_SDHI))	+= tmio_mmc_dma.o
+obj-$(CONFIG_MMC_SDHI)		+= sh_mobile_sdhi.o
+obj-$(CONFIG_MMC_CB710)		+= cb710-mmc.o
+obj-$(CONFIG_MMC_VIA_SDMMC)	+= via-sdmmc.o
+obj-$(CONFIG_SDH_BFIN)		+= bfin_sdh.o
+obj-$(CONFIG_MMC_DW)		+= dw_mmc.o
+obj-$(CONFIG_MMC_DW_PLTFM)	+= dw_mmc-pltfm.o
+obj-$(CONFIG_MMC_DW_EXYNOS)	+= dw_mmc-exynos.o
+obj-$(CONFIG_MMC_DW_K3)		+= dw_mmc-k3.o
+obj-$(CONFIG_MMC_DW_PCI)	+= dw_mmc-pci.o
+obj-$(CONFIG_MMC_SH_MMCIF)	+= sh_mmcif.o
+obj-$(CONFIG_MMC_JZ4740)	+= jz4740_mmc.o
+obj-$(CONFIG_MMC_VUB300)	+= vub300.o
+obj-$(CONFIG_MMC_USHC)		+= ushc.o
+obj-$(CONFIG_MMC_WMT)		+= wmt-sdmmc.o
+obj-$(CONFIG_MMC_MOXART)	+= moxart-mmc.o
+obj-$(CONFIG_MMC_SUNXI)		+= sunxi-mmc.o
+obj-$(CONFIG_MMC_USDHI6ROL0)	+= usdhi6rol0.o
+
+obj-$(CONFIG_MMC_REALTEK_PCI)	+= rtsx_pci_sdmmc.o
+obj-$(CONFIG_MMC_REALTEK_USB)	+= rtsx_usb_sdmmc.o
 
+obj-$(CONFIG_MMC_SDHCI_PLTFM)		+= sdhci-pltfm.o
+obj-$(CONFIG_MMC_SDHCI_CNS3XXX)		+= sdhci-cns3xxx.o
+obj-$(CONFIG_MMC_SDHCI_ESDHC_IMX)	+= sdhci-esdhc-imx.o
+obj-$(CONFIG_MMC_SDHCI_DOVE)		+= sdhci-dove.o
+obj-$(CONFIG_MMC_SDHCI_TEGRA)		+= sdhci-tegra.o
+obj-$(CONFIG_MMC_SDHCI_OF_ARASAN)	+= sdhci-of-arasan.o
+obj-$(CONFIG_MMC_SDHCI_OF_ESDHC)	+= sdhci-of-esdhc.o
+obj-$(CONFIG_MMC_SDHCI_OF_HLWD)		+= sdhci-of-hlwd.o
+obj-$(CONFIG_MMC_SDHCI_BCM_KONA)	+= sdhci-bcm-kona.o
+obj-$(CONFIG_MMC_SDHCI_BCM2835)		+= sdhci-bcm2835.o
+obj-$(CONFIG_MMC_SDHCI_MSM)		+= sdhci-msm.o
+
+ifeq ($(CONFIG_CB710_DEBUG),y)
+	CFLAGS-cb710-mmc	+= -DDEBUG
+endif
diff --git a/drivers/mmc/host/android-goldfish.c b/drivers/mmc/host/android-goldfish.c
new file mode 100644
index 00000000000..8b4e20a3f16
--- /dev/null
+++ b/drivers/mmc/host/android-goldfish.c
@@ -0,0 +1,568 @@
+/*
+ *  Copyright 2007, Google Inc.
+ *  Copyright 2012, Intel Inc.
+ *
+ *  based on omap.c driver, which was
+ *  Copyright (C) 2004 Nokia Corporation
+ *  Written by Tuukka Tikkanen and Juha Yrjölä <juha.yrjola@nokia.com>
+ *  Misc hacks here and there by Tony Lindgren <tony@atomide.com>
+ *  Other hacks (DMA, SD, etc) by David Brownell
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/major.h>
+
+#include <linux/types.h>
+#include <linux/pci.h>
+#include <linux/interrupt.h>
+
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/errno.h>
+#include <linux/hdreg.h>
+#include <linux/kdev_t.h>
+#include <linux/blkdev.h>
+#include <linux/mutex.h>
+#include <linux/scatterlist.h>
+#include <linux/mmc/mmc.h>
+#include <linux/mmc/sdio.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/card.h>
+
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/ioport.h>
+#include <linux/dma-mapping.h>
+#include <linux/delay.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/clk.h>
+
+#include <asm/io.h>
+#include <asm/irq.h>
+#include <asm/scatterlist.h>
+
+#include <asm/types.h>
+#include <asm/io.h>
+#include <asm/uaccess.h>
+
+#define DRIVER_NAME "goldfish_mmc"
+
+#define BUFFER_SIZE   16384
+
+#define GOLDFISH_MMC_READ(host, addr)   (readl(host->reg_base + addr))
+#define GOLDFISH_MMC_WRITE(host, addr, x)   (writel(x, host->reg_base + addr))
+
+enum {
+	/* status register */
+	MMC_INT_STATUS	        = 0x00,
+	/* set this to enable IRQ */
+	MMC_INT_ENABLE	        = 0x04,
+	/* set this to specify buffer address */
+	MMC_SET_BUFFER          = 0x08,
+
+	/* MMC command number */
+	MMC_CMD	                = 0x0C,
+
+	/* MMC argument */
+	MMC_ARG	                = 0x10,
+
+	/* MMC response (or R2 bits 0 - 31) */
+	MMC_RESP_0		        = 0x14,
+
+	/* MMC R2 response bits 32 - 63 */
+	MMC_RESP_1		        = 0x18,
+
+	/* MMC R2 response bits 64 - 95 */
+	MMC_RESP_2		        = 0x1C,
+
+	/* MMC R2 response bits 96 - 127 */
+	MMC_RESP_3		        = 0x20,
+
+	MMC_BLOCK_LENGTH        = 0x24,
+	MMC_BLOCK_COUNT         = 0x28,
+
+	/* MMC state flags */
+	MMC_STATE               = 0x2C,
+
+	/* MMC_INT_STATUS bits */
+
+	MMC_STAT_END_OF_CMD     = 1U << 0,
+	MMC_STAT_END_OF_DATA    = 1U << 1,
+	MMC_STAT_STATE_CHANGE   = 1U << 2,
+	MMC_STAT_CMD_TIMEOUT    = 1U << 3,
+
+	/* MMC_STATE bits */
+	MMC_STATE_INSERTED     = 1U << 0,
+	MMC_STATE_READ_ONLY    = 1U << 1,
+};
+
+/*
+ * Command types
+ */
+#define OMAP_MMC_CMDTYPE_BC	0
+#define OMAP_MMC_CMDTYPE_BCR	1
+#define OMAP_MMC_CMDTYPE_AC	2
+#define OMAP_MMC_CMDTYPE_ADTC	3
+
+
+struct goldfish_mmc_host {
+	struct mmc_request	*mrq;
+	struct mmc_command	*cmd;
+	struct mmc_data		*data;
+	struct mmc_host		*mmc;
+	struct device		*dev;
+	unsigned char		id; /* 16xx chips have 2 MMC blocks */
+	void __iomem		*virt_base;
+	unsigned int		phys_base;
+	int			irq;
+	unsigned char		bus_mode;
+	unsigned char		hw_bus_mode;
+
+	unsigned int		sg_len;
+	unsigned		dma_done:1;
+	unsigned		dma_in_use:1;
+
+	void __iomem		*reg_base;
+};
+
+static inline int
+goldfish_mmc_cover_is_open(struct goldfish_mmc_host *host)
+{
+	return 0;
+}
+
+static ssize_t
+goldfish_mmc_show_cover_switch(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	struct goldfish_mmc_host *host = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%s\n", goldfish_mmc_cover_is_open(host) ? "open" :
+		       "closed");
+}
+
+static DEVICE_ATTR(cover_switch, S_IRUGO, goldfish_mmc_show_cover_switch, NULL);
+
+static void
+goldfish_mmc_start_command(struct goldfish_mmc_host *host, struct mmc_command *cmd)
+{
+	u32 cmdreg;
+	u32 resptype;
+	u32 cmdtype;
+
+	host->cmd = cmd;
+
+	resptype = 0;
+	cmdtype = 0;
+
+	/* Our hardware needs to know exact type */
+	switch (mmc_resp_type(cmd)) {
+	case MMC_RSP_NONE:
+		break;
+	case MMC_RSP_R1:
+	case MMC_RSP_R1B:
+		/* resp 1, 1b, 6, 7 */
+		resptype = 1;
+		break;
+	case MMC_RSP_R2:
+		resptype = 2;
+		break;
+	case MMC_RSP_R3:
+		resptype = 3;
+		break;
+	default:
+		dev_err(mmc_dev(host->mmc),
+			"Invalid response type: %04x\n", mmc_resp_type(cmd));
+		break;
+	}
+
+	if (mmc_cmd_type(cmd) == MMC_CMD_ADTC)
+		cmdtype = OMAP_MMC_CMDTYPE_ADTC;
+	else if (mmc_cmd_type(cmd) == MMC_CMD_BC)
+		cmdtype = OMAP_MMC_CMDTYPE_BC;
+	else if (mmc_cmd_type(cmd) == MMC_CMD_BCR)
+		cmdtype = OMAP_MMC_CMDTYPE_BCR;
+	else
+		cmdtype = OMAP_MMC_CMDTYPE_AC;
+
+	cmdreg = cmd->opcode | (resptype << 8) | (cmdtype << 12);
+
+	if (host->bus_mode == MMC_BUSMODE_OPENDRAIN)
+		cmdreg |= 1 << 6;
+
+	if (cmd->flags & MMC_RSP_BUSY)
+		cmdreg |= 1 << 11;
+
+	if (host->data && !(host->data->flags & MMC_DATA_WRITE))
+		cmdreg |= 1 << 15;
+
+	GOLDFISH_MMC_WRITE(host, MMC_ARG, cmd->arg);
+	GOLDFISH_MMC_WRITE(host, MMC_CMD, cmdreg);
+}
+
+static void goldfish_mmc_xfer_done(struct goldfish_mmc_host *host,
+				   struct mmc_data *data)
+{
+	if (host->dma_in_use) {
+		enum dma_data_direction dma_data_dir;
+
+		if (data->flags & MMC_DATA_WRITE)
+			dma_data_dir = DMA_TO_DEVICE;
+		else
+			dma_data_dir = DMA_FROM_DEVICE;
+
+		if (dma_data_dir == DMA_FROM_DEVICE) {
+			/*
+			 * We don't really have DMA, so we need
+			 * to copy from our platform driver buffer
+			 */
+			uint8_t *dest = (uint8_t *)sg_virt(data->sg);
+			memcpy(dest, host->virt_base, data->sg->length);
+		}
+		host->data->bytes_xfered += data->sg->length;
+		dma_unmap_sg(mmc_dev(host->mmc), data->sg, host->sg_len,
+			     dma_data_dir);
+	}
+
+	host->data = NULL;
+	host->sg_len = 0;
+
+	/*
+	 * NOTE:  MMC layer will sometimes poll-wait CMD13 next, issuing
+	 * dozens of requests until the card finishes writing data.
+	 * It'd be cheaper to just wait till an EOFB interrupt arrives...
+	 */
+
+	if (!data->stop) {
+		host->mrq = NULL;
+		mmc_request_done(host->mmc, data->mrq);
+		return;
+	}
+
+	goldfish_mmc_start_command(host, data->stop);
+}
+
+static void goldfish_mmc_end_of_data(struct goldfish_mmc_host *host,
+				     struct mmc_data *data)
+{
+	if (!host->dma_in_use) {
+		goldfish_mmc_xfer_done(host, data);
+		return;
+	}
+	if (host->dma_done)
+		goldfish_mmc_xfer_done(host, data);
+}
+
+static void goldfish_mmc_cmd_done(struct goldfish_mmc_host *host,
+				  struct mmc_command *cmd)
+{
+	host->cmd = NULL;
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		if (cmd->flags & MMC_RSP_136) {
+			/* response type 2 */
+			cmd->resp[3] =
+				GOLDFISH_MMC_READ(host, MMC_RESP_0);
+			cmd->resp[2] =
+				GOLDFISH_MMC_READ(host, MMC_RESP_1);
+			cmd->resp[1] =
+				GOLDFISH_MMC_READ(host, MMC_RESP_2);
+			cmd->resp[0] =
+				GOLDFISH_MMC_READ(host, MMC_RESP_3);
+		} else {
+			/* response types 1, 1b, 3, 4, 5, 6 */
+			cmd->resp[0] =
+				GOLDFISH_MMC_READ(host, MMC_RESP_0);
+		}
+	}
+
+	if (host->data == NULL || cmd->error) {
+		host->mrq = NULL;
+		mmc_request_done(host->mmc, cmd->mrq);
+	}
+}
+
+static irqreturn_t goldfish_mmc_irq(int irq, void *dev_id)
+{
+	struct goldfish_mmc_host *host = (struct goldfish_mmc_host *)dev_id;
+	u16 status;
+	int end_command = 0;
+	int end_transfer = 0;
+	int transfer_error = 0;
+	int state_changed = 0;
+	int cmd_timeout = 0;
+
+	while ((status = GOLDFISH_MMC_READ(host, MMC_INT_STATUS)) != 0) {
+		GOLDFISH_MMC_WRITE(host, MMC_INT_STATUS, status);
+
+		if (status & MMC_STAT_END_OF_CMD)
+			end_command = 1;
+
+		if (status & MMC_STAT_END_OF_DATA)
+			end_transfer = 1;
+
+		if (status & MMC_STAT_STATE_CHANGE)
+			state_changed = 1;
+
+                if (status & MMC_STAT_CMD_TIMEOUT) {
+			end_command = 0;
+			cmd_timeout = 1;
+                }
+	}
+
+	if (cmd_timeout) {
+		struct mmc_request *mrq = host->mrq;
+		mrq->cmd->error = -ETIMEDOUT;
+		host->mrq = NULL;
+		mmc_request_done(host->mmc, mrq);
+	}
+
+	if (end_command)
+		goldfish_mmc_cmd_done(host, host->cmd);
+
+	if (transfer_error)
+		goldfish_mmc_xfer_done(host, host->data);
+	else if (end_transfer) {
+		host->dma_done = 1;
+		goldfish_mmc_end_of_data(host, host->data);
+	} else if (host->data != NULL) {
+		/*
+		 * WORKAROUND -- after porting this driver from 2.6 to 3.4,
+		 * during device initialization, cases where host->data is
+		 * non-null but end_transfer is false would occur. Doing
+		 * nothing in such cases results in no further interrupts,
+		 * and initialization failure.
+		 * TODO -- find the real cause.
+		 */
+		host->dma_done = 1;
+		goldfish_mmc_end_of_data(host, host->data);
+	}
+
+	if (state_changed) {
+		u32 state = GOLDFISH_MMC_READ(host, MMC_STATE);
+		pr_info("%s: Card detect now %d\n", __func__,
+			(state & MMC_STATE_INSERTED));
+		mmc_detect_change(host->mmc, 0);
+	}
+
+	if (!end_command && !end_transfer &&
+	    !transfer_error && !state_changed && !cmd_timeout) {
+		status = GOLDFISH_MMC_READ(host, MMC_INT_STATUS);
+		dev_info(mmc_dev(host->mmc),"spurious irq 0x%04x\n", status);
+		if (status != 0) {
+			GOLDFISH_MMC_WRITE(host, MMC_INT_STATUS, status);
+			GOLDFISH_MMC_WRITE(host, MMC_INT_ENABLE, 0);
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
+static void goldfish_mmc_prepare_data(struct goldfish_mmc_host *host,
+				      struct mmc_request *req)
+{
+	struct mmc_data *data = req->data;
+	int block_size;
+	unsigned sg_len;
+	enum dma_data_direction dma_data_dir;
+
+	host->data = data;
+	if (data == NULL) {
+		GOLDFISH_MMC_WRITE(host, MMC_BLOCK_LENGTH, 0);
+		GOLDFISH_MMC_WRITE(host, MMC_BLOCK_COUNT, 0);
+		host->dma_in_use = 0;
+		return;
+	}
+
+	block_size = data->blksz;
+
+	GOLDFISH_MMC_WRITE(host, MMC_BLOCK_COUNT, data->blocks - 1);
+	GOLDFISH_MMC_WRITE(host, MMC_BLOCK_LENGTH, block_size - 1);
+
+	/*
+	 * Cope with calling layer confusion; it issues "single
+	 * block" writes using multi-block scatterlists.
+	 */
+	sg_len = (data->blocks == 1) ? 1 : data->sg_len;
+
+	if (data->flags & MMC_DATA_WRITE)
+		dma_data_dir = DMA_TO_DEVICE;
+	else
+		dma_data_dir = DMA_FROM_DEVICE;
+
+	host->sg_len = dma_map_sg(mmc_dev(host->mmc), data->sg,
+				  sg_len, dma_data_dir);
+	host->dma_done = 0;
+	host->dma_in_use = 1;
+
+	if (dma_data_dir == DMA_TO_DEVICE) {
+		/*
+		 * We don't really have DMA, so we need to copy to our
+		 * platform driver buffer
+		 */
+		const uint8_t *src = (uint8_t *)sg_virt(data->sg);
+		memcpy(host->virt_base, src, data->sg->length);
+	}
+}
+
+static void goldfish_mmc_request(struct mmc_host *mmc, struct mmc_request *req)
+{
+	struct goldfish_mmc_host *host = mmc_priv(mmc);
+
+	WARN_ON(host->mrq != NULL);
+
+	host->mrq = req;
+	goldfish_mmc_prepare_data(host, req);
+	goldfish_mmc_start_command(host, req->cmd);
+
+	/*
+	 * This is to avoid accidentally being detected as an SDIO card
+	 * in mmc_attach_sdio().
+	 */
+	if (req->cmd->opcode == SD_IO_SEND_OP_COND &&
+	    req->cmd->flags == (MMC_RSP_SPI_R4 | MMC_RSP_R4 | MMC_CMD_BCR))
+		req->cmd->error = -EINVAL;
+}
+
+static void goldfish_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct goldfish_mmc_host *host = mmc_priv(mmc);
+
+	host->bus_mode = ios->bus_mode;
+	host->hw_bus_mode = host->bus_mode;
+}
+
+static int goldfish_mmc_get_ro(struct mmc_host *mmc)
+{
+	uint32_t state;
+	struct goldfish_mmc_host *host = mmc_priv(mmc);
+
+	state = GOLDFISH_MMC_READ(host, MMC_STATE);
+	return ((state & MMC_STATE_READ_ONLY) != 0);
+}
+
+static const struct mmc_host_ops goldfish_mmc_ops = {
+	.request	= goldfish_mmc_request,
+	.set_ios	= goldfish_mmc_set_ios,
+	.get_ro		= goldfish_mmc_get_ro,
+};
+
+static int goldfish_mmc_probe(struct platform_device *pdev)
+{
+	struct mmc_host *mmc;
+	struct goldfish_mmc_host *host = NULL;
+	struct resource *res;
+	int ret = 0;
+	int irq;
+	dma_addr_t buf_addr;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	irq = platform_get_irq(pdev, 0);
+	if (res == NULL || irq < 0)
+		return -ENXIO;
+
+	mmc = mmc_alloc_host(sizeof(struct goldfish_mmc_host), &pdev->dev);
+	if (mmc == NULL) {
+		ret = -ENOMEM;
+		goto err_alloc_host_failed;
+	}
+
+	host = mmc_priv(mmc);
+	host->mmc = mmc;
+
+	pr_err("mmc: Mapping %lX to %lX\n", (long)res->start, (long)res->end);
+	host->reg_base = ioremap(res->start, resource_size(res));
+	if (host->reg_base == NULL) {
+		ret = -ENOMEM;
+		goto ioremap_failed;
+	}
+	host->virt_base = dma_alloc_coherent(&pdev->dev, BUFFER_SIZE,
+					     &buf_addr, GFP_KERNEL);
+
+	if (host->virt_base == 0) {
+		ret = -ENOMEM;
+		goto dma_alloc_failed;
+	}
+	host->phys_base = buf_addr;
+
+	host->id = pdev->id;
+	host->irq = irq;
+
+	mmc->ops = &goldfish_mmc_ops;
+	mmc->f_min = 400000;
+	mmc->f_max = 24000000;
+	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+	mmc->caps = MMC_CAP_4_BIT_DATA;
+
+	/* Use scatterlist DMA to reduce per-transfer costs.
+	 * NOTE max_seg_size assumption that small blocks aren't
+	 * normally used (except e.g. for reading SD registers).
+	 */
+	mmc->max_segs = 32;
+	mmc->max_blk_size = 2048;	/* MMC_BLOCK_LENGTH is 11 bits (+1) */
+	mmc->max_blk_count = 2048;	/* MMC_BLOCK_COUNT is 11 bits (+1) */
+	mmc->max_req_size = BUFFER_SIZE;
+	mmc->max_seg_size = mmc->max_req_size;
+
+	ret = request_irq(host->irq, goldfish_mmc_irq, 0, DRIVER_NAME, host);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed IRQ Adding goldfish MMC\n");
+		goto err_request_irq_failed;
+	}
+
+	host->dev = &pdev->dev;
+	platform_set_drvdata(pdev, host);
+
+	ret = device_create_file(&pdev->dev, &dev_attr_cover_switch);
+	if (ret)
+		dev_warn(mmc_dev(host->mmc),
+			 "Unable to create sysfs attributes\n");
+
+	GOLDFISH_MMC_WRITE(host, MMC_SET_BUFFER, host->phys_base);
+	GOLDFISH_MMC_WRITE(host, MMC_INT_ENABLE,
+			   MMC_STAT_END_OF_CMD | MMC_STAT_END_OF_DATA |
+			   MMC_STAT_STATE_CHANGE | MMC_STAT_CMD_TIMEOUT);
+
+	mmc_add_host(mmc);
+	return 0;
+
+err_request_irq_failed:
+	dma_free_coherent(&pdev->dev, BUFFER_SIZE, host->virt_base,
+			  host->phys_base);
+dma_alloc_failed:
+	iounmap(host->reg_base);
+ioremap_failed:
+	mmc_free_host(host->mmc);
+err_alloc_host_failed:
+	return ret;
+}
+
+static int goldfish_mmc_remove(struct platform_device *pdev)
+{
+	struct goldfish_mmc_host *host = platform_get_drvdata(pdev);
+
+	BUG_ON(host == NULL);
+
+	mmc_remove_host(host->mmc);
+	free_irq(host->irq, host);
+	dma_free_coherent(&pdev->dev, BUFFER_SIZE, host->virt_base, host->phys_base);
+	iounmap(host->reg_base);
+	mmc_free_host(host->mmc);
+	return 0;
+}
+
+static struct platform_driver goldfish_mmc_driver = {
+	.probe		= goldfish_mmc_probe,
+	.remove		= goldfish_mmc_remove,
+	.driver		= {
+		.name	= DRIVER_NAME,
+	},
+};
+
+module_platform_driver(goldfish_mmc_driver);
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/at91_mci.c b/drivers/mmc/host/at91_mci.c
deleted file mode 100644
index e556d42cc45..00000000000
--- a/drivers/mmc/host/at91_mci.c
+++ /dev/null
@@ -1,1250 +0,0 @@
-/*
- *  linux/drivers/mmc/host/at91_mci.c - ATMEL AT91 MCI Driver
- *
- *  Copyright (C) 2005 Cougar Creek Computing Devices Ltd, All Rights Reserved
- *
- *  Copyright (C) 2006 Malcolm Noyes
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-/*
-   This is the AT91 MCI driver that has been tested with both MMC cards
-   and SD-cards.  Boards that support write protect are now supported.
-   The CCAT91SBC001 board does not support SD cards.
-
-   The three entry points are at91_mci_request, at91_mci_set_ios
-   and at91_mci_get_ro.
-
-   SET IOS
-     This configures the device to put it into the correct mode and clock speed
-     required.
-
-   MCI REQUEST
-     MCI request processes the commands sent in the mmc_request structure. This
-     can consist of a processing command and a stop command in the case of
-     multiple block transfers.
-
-     There are three main types of request, commands, reads and writes.
-
-     Commands are straight forward. The command is submitted to the controller and
-     the request function returns. When the controller generates an interrupt to indicate
-     the command is finished, the response to the command are read and the mmc_request_done
-     function called to end the request.
-
-     Reads and writes work in a similar manner to normal commands but involve the PDC (DMA)
-     controller to manage the transfers.
-
-     A read is done from the controller directly to the scatterlist passed in from the request.
-     Due to a bug in the AT91RM9200 controller, when a read is completed, all the words are byte
-     swapped in the scatterlist buffers.  AT91SAM926x are not affected by this bug.
-
-     The sequence of read interrupts is: ENDRX, RXBUFF, CMDRDY
-
-     A write is slightly different in that the bytes to write are read from the scatterlist
-     into a dma memory buffer (this is in case the source buffer should be read only). The
-     entire write buffer is then done from this single dma memory buffer.
-
-     The sequence of write interrupts is: ENDTX, TXBUFE, NOTBUSY, CMDRDY
-
-   GET RO
-     Gets the status of the write protect pin, if available.
-*/
-
-#include <linux/module.h>
-#include <linux/moduleparam.h>
-#include <linux/init.h>
-#include <linux/ioport.h>
-#include <linux/platform_device.h>
-#include <linux/interrupt.h>
-#include <linux/blkdev.h>
-#include <linux/delay.h>
-#include <linux/err.h>
-#include <linux/dma-mapping.h>
-#include <linux/clk.h>
-#include <linux/atmel_pdc.h>
-
-#include <linux/mmc/host.h>
-
-#include <asm/io.h>
-#include <asm/irq.h>
-#include <asm/gpio.h>
-
-#include <asm/mach/mmc.h>
-#include <mach/board.h>
-#include <mach/cpu.h>
-#include <mach/at91_mci.h>
-
-#define DRIVER_NAME "at91_mci"
-
-#define FL_SENT_COMMAND	(1 << 0)
-#define FL_SENT_STOP	(1 << 1)
-
-#define AT91_MCI_ERRORS	(AT91_MCI_RINDE | AT91_MCI_RDIRE | AT91_MCI_RCRCE	\
-		| AT91_MCI_RENDE | AT91_MCI_RTOE | AT91_MCI_DCRCE		\
-		| AT91_MCI_DTOE | AT91_MCI_OVRE | AT91_MCI_UNRE)
-
-#define at91_mci_read(host, reg)	__raw_readl((host)->baseaddr + (reg))
-#define at91_mci_write(host, reg, val)	__raw_writel((val), (host)->baseaddr + (reg))
-
-
-/*
- * Low level type for this driver
- */
-struct at91mci_host
-{
-	struct mmc_host *mmc;
-	struct mmc_command *cmd;
-	struct mmc_request *request;
-
-	void __iomem *baseaddr;
-	int irq;
-
-	struct at91_mmc_data *board;
-	int present;
-
-	struct clk *mci_clk;
-
-	/*
-	 * Flag indicating when the command has been sent. This is used to
-	 * work out whether or not to send the stop
-	 */
-	unsigned int flags;
-	/* flag for current bus settings */
-	u32 bus_mode;
-
-	/* DMA buffer used for transmitting */
-	unsigned int* buffer;
-	dma_addr_t physical_address;
-	unsigned int total_length;
-
-	/* Latest in the scatterlist that has been enabled for transfer, but not freed */
-	int in_use_index;
-
-	/* Latest in the scatterlist that has been enabled for transfer */
-	int transfer_index;
-
-	/* Timer for timeouts */
-	struct timer_list timer;
-};
-
-/*
- * Reset the controller and restore most of the state
- */
-static void at91_reset_host(struct at91mci_host *host)
-{
-	unsigned long flags;
-	u32 mr;
-	u32 sdcr;
-	u32 dtor;
-	u32 imr;
-
-	local_irq_save(flags);
-	imr = at91_mci_read(host, AT91_MCI_IMR);
-
-	at91_mci_write(host, AT91_MCI_IDR, 0xffffffff);
-
-	/* save current state */
-	mr = at91_mci_read(host, AT91_MCI_MR) & 0x7fff;
-	sdcr = at91_mci_read(host, AT91_MCI_SDCR);
-	dtor = at91_mci_read(host, AT91_MCI_DTOR);
-
-	/* reset the controller */
-	at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIDIS | AT91_MCI_SWRST);
-
-	/* restore state */
-	at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIEN);
-	at91_mci_write(host, AT91_MCI_MR, mr);
-	at91_mci_write(host, AT91_MCI_SDCR, sdcr);
-	at91_mci_write(host, AT91_MCI_DTOR, dtor);
-	at91_mci_write(host, AT91_MCI_IER, imr);
-
-	/* make sure sdio interrupts will fire */
-	at91_mci_read(host, AT91_MCI_SR);
-
-	local_irq_restore(flags);
-}
-
-static void at91_timeout_timer(unsigned long data)
-{
-	struct at91mci_host *host;
-
-	host = (struct at91mci_host *)data;
-
-	if (host->request) {
-		dev_err(host->mmc->parent, "Timeout waiting end of packet\n");
-
-		if (host->cmd && host->cmd->data) {
-			host->cmd->data->error = -ETIMEDOUT;
-		} else {
-			if (host->cmd)
-				host->cmd->error = -ETIMEDOUT;
-			else
-				host->request->cmd->error = -ETIMEDOUT;
-		}
-
-		at91_reset_host(host);
-		mmc_request_done(host->mmc, host->request);
-	}
-}
-
-/*
- * Copy from sg to a dma block - used for transfers
- */
-static inline void at91_mci_sg_to_dma(struct at91mci_host *host, struct mmc_data *data)
-{
-	unsigned int len, i, size;
-	unsigned *dmabuf = host->buffer;
-
-	size = data->blksz * data->blocks;
-	len = data->sg_len;
-
-	/* AT91SAM926[0/3] Data Write Operation and number of bytes erratum */
-	if (cpu_is_at91sam9260() || cpu_is_at91sam9263())
-		if (host->total_length == 12)
-			memset(dmabuf, 0, 12);
-
-	/*
-	 * Just loop through all entries. Size might not
-	 * be the entire list though so make sure that
-	 * we do not transfer too much.
-	 */
-	for (i = 0; i < len; i++) {
-		struct scatterlist *sg;
-		int amount;
-		unsigned int *sgbuffer;
-
-		sg = &data->sg[i];
-
-		sgbuffer = kmap_atomic(sg_page(sg), KM_BIO_SRC_IRQ) + sg->offset;
-		amount = min(size, sg->length);
-		size -= amount;
-
-		if (cpu_is_at91rm9200()) {	/* AT91RM9200 errata */
-			int index;
-
-			for (index = 0; index < (amount / 4); index++)
-				*dmabuf++ = swab32(sgbuffer[index]);
-		} else {
-			memcpy(dmabuf, sgbuffer, amount);
-			dmabuf += amount;
-		}
-
-		kunmap_atomic(sgbuffer, KM_BIO_SRC_IRQ);
-
-		if (size == 0)
-			break;
-	}
-
-	/*
-	 * Check that we didn't get a request to transfer
-	 * more data than can fit into the SG list.
-	 */
-	BUG_ON(size != 0);
-}
-
-/*
- * Prepare a dma read
- */
-static void at91_mci_pre_dma_read(struct at91mci_host *host)
-{
-	int i;
-	struct scatterlist *sg;
-	struct mmc_command *cmd;
-	struct mmc_data *data;
-
-	pr_debug("pre dma read\n");
-
-	cmd = host->cmd;
-	if (!cmd) {
-		pr_debug("no command\n");
-		return;
-	}
-
-	data = cmd->data;
-	if (!data) {
-		pr_debug("no data\n");
-		return;
-	}
-
-	for (i = 0; i < 2; i++) {
-		/* nothing left to transfer */
-		if (host->transfer_index >= data->sg_len) {
-			pr_debug("Nothing left to transfer (index = %d)\n", host->transfer_index);
-			break;
-		}
-
-		/* Check to see if this needs filling */
-		if (i == 0) {
-			if (at91_mci_read(host, ATMEL_PDC_RCR) != 0) {
-				pr_debug("Transfer active in current\n");
-				continue;
-			}
-		}
-		else {
-			if (at91_mci_read(host, ATMEL_PDC_RNCR) != 0) {
-				pr_debug("Transfer active in next\n");
-				continue;
-			}
-		}
-
-		/* Setup the next transfer */
-		pr_debug("Using transfer index %d\n", host->transfer_index);
-
-		sg = &data->sg[host->transfer_index++];
-		pr_debug("sg = %p\n", sg);
-
-		sg->dma_address = dma_map_page(NULL, sg_page(sg), sg->offset, sg->length, DMA_FROM_DEVICE);
-
-		pr_debug("dma address = %08X, length = %d\n", sg->dma_address, sg->length);
-
-		if (i == 0) {
-			at91_mci_write(host, ATMEL_PDC_RPR, sg->dma_address);
-			at91_mci_write(host, ATMEL_PDC_RCR, (data->blksz & 0x3) ? sg->length : sg->length / 4);
-		}
-		else {
-			at91_mci_write(host, ATMEL_PDC_RNPR, sg->dma_address);
-			at91_mci_write(host, ATMEL_PDC_RNCR, (data->blksz & 0x3) ? sg->length : sg->length / 4);
-		}
-	}
-
-	pr_debug("pre dma read done\n");
-}
-
-/*
- * Handle after a dma read
- */
-static void at91_mci_post_dma_read(struct at91mci_host *host)
-{
-	struct mmc_command *cmd;
-	struct mmc_data *data;
-
-	pr_debug("post dma read\n");
-
-	cmd = host->cmd;
-	if (!cmd) {
-		pr_debug("no command\n");
-		return;
-	}
-
-	data = cmd->data;
-	if (!data) {
-		pr_debug("no data\n");
-		return;
-	}
-
-	while (host->in_use_index < host->transfer_index) {
-		struct scatterlist *sg;
-
-		pr_debug("finishing index %d\n", host->in_use_index);
-
-		sg = &data->sg[host->in_use_index++];
-
-		pr_debug("Unmapping page %08X\n", sg->dma_address);
-
-		dma_unmap_page(NULL, sg->dma_address, sg->length, DMA_FROM_DEVICE);
-
-		if (cpu_is_at91rm9200()) {	/* AT91RM9200 errata */
-			unsigned int *buffer;
-			int index;
-
-			/* Swap the contents of the buffer */
-			buffer = kmap_atomic(sg_page(sg), KM_BIO_SRC_IRQ) + sg->offset;
-			pr_debug("buffer = %p, length = %d\n", buffer, sg->length);
-
-			for (index = 0; index < (sg->length / 4); index++)
-				buffer[index] = swab32(buffer[index]);
-
-			kunmap_atomic(buffer, KM_BIO_SRC_IRQ);
-		}
-
-		flush_dcache_page(sg_page(sg));
-
-		data->bytes_xfered += sg->length;
-	}
-
-	/* Is there another transfer to trigger? */
-	if (host->transfer_index < data->sg_len)
-		at91_mci_pre_dma_read(host);
-	else {
-		at91_mci_write(host, AT91_MCI_IDR, AT91_MCI_ENDRX);
-		at91_mci_write(host, AT91_MCI_IER, AT91_MCI_RXBUFF);
-	}
-
-	pr_debug("post dma read done\n");
-}
-
-/*
- * Handle transmitted data
- */
-static void at91_mci_handle_transmitted(struct at91mci_host *host)
-{
-	struct mmc_command *cmd;
-	struct mmc_data *data;
-
-	pr_debug("Handling the transmit\n");
-
-	/* Disable the transfer */
-	at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTDIS | ATMEL_PDC_TXTDIS);
-
-	/* Now wait for cmd ready */
-	at91_mci_write(host, AT91_MCI_IDR, AT91_MCI_TXBUFE);
-
-	cmd = host->cmd;
-	if (!cmd) return;
-
-	data = cmd->data;
-	if (!data) return;
-
-	if (cmd->data->blocks > 1) {
-		pr_debug("multiple write : wait for BLKE...\n");
-		at91_mci_write(host, AT91_MCI_IER, AT91_MCI_BLKE);
-	} else
-		at91_mci_write(host, AT91_MCI_IER, AT91_MCI_NOTBUSY);
-}
-
-/*
- * Update bytes tranfered count during a write operation
- */
-static void at91_mci_update_bytes_xfered(struct at91mci_host *host)
-{
-	struct mmc_data *data;
-
-	/* always deal with the effective request (and not the current cmd) */
-
-	if (host->request->cmd && host->request->cmd->error != 0)
-		return;
-
-	if (host->request->data) {
-		data = host->request->data;
-		if (data->flags & MMC_DATA_WRITE) {
-			/* card is in IDLE mode now */
-			pr_debug("-> bytes_xfered %d, total_length = %d\n",
-				data->bytes_xfered, host->total_length);
-			data->bytes_xfered = data->blksz * data->blocks;
-		}
-	}
-}
-
-
-/*Handle after command sent ready*/
-static int at91_mci_handle_cmdrdy(struct at91mci_host *host)
-{
-	if (!host->cmd)
-		return 1;
-	else if (!host->cmd->data) {
-		if (host->flags & FL_SENT_STOP) {
-			/*After multi block write, we must wait for NOTBUSY*/
-			at91_mci_write(host, AT91_MCI_IER, AT91_MCI_NOTBUSY);
-		} else return 1;
-	} else if (host->cmd->data->flags & MMC_DATA_WRITE) {
-		/*After sendding multi-block-write command, start DMA transfer*/
-		at91_mci_write(host, AT91_MCI_IER, AT91_MCI_TXBUFE | AT91_MCI_BLKE);
-		at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_TXTEN);
-	}
-
-	/* command not completed, have to wait */
-	return 0;
-}
-
-
-/*
- * Enable the controller
- */
-static void at91_mci_enable(struct at91mci_host *host)
-{
-	unsigned int mr;
-
-	at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIEN);
-	at91_mci_write(host, AT91_MCI_IDR, 0xffffffff);
-	at91_mci_write(host, AT91_MCI_DTOR, AT91_MCI_DTOMUL_1M | AT91_MCI_DTOCYC);
-	mr = AT91_MCI_PDCMODE | 0x34a;
-
-	if (cpu_is_at91sam9260() || cpu_is_at91sam9263())
-		mr |= AT91_MCI_RDPROOF | AT91_MCI_WRPROOF;
-
-	at91_mci_write(host, AT91_MCI_MR, mr);
-
-	/* use Slot A or B (only one at same time) */
-	at91_mci_write(host, AT91_MCI_SDCR, host->board->slot_b);
-}
-
-/*
- * Disable the controller
- */
-static void at91_mci_disable(struct at91mci_host *host)
-{
-	at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIDIS | AT91_MCI_SWRST);
-}
-
-/*
- * Send a command
- */
-static void at91_mci_send_command(struct at91mci_host *host, struct mmc_command *cmd)
-{
-	unsigned int cmdr, mr;
-	unsigned int block_length;
-	struct mmc_data *data = cmd->data;
-
-	unsigned int blocks;
-	unsigned int ier = 0;
-
-	host->cmd = cmd;
-
-	/* Needed for leaving busy state before CMD1 */
-	if ((at91_mci_read(host, AT91_MCI_SR) & AT91_MCI_RTOE) && (cmd->opcode == 1)) {
-		pr_debug("Clearing timeout\n");
-		at91_mci_write(host, AT91_MCI_ARGR, 0);
-		at91_mci_write(host, AT91_MCI_CMDR, AT91_MCI_OPDCMD);
-		while (!(at91_mci_read(host, AT91_MCI_SR) & AT91_MCI_CMDRDY)) {
-			/* spin */
-			pr_debug("Clearing: SR = %08X\n", at91_mci_read(host, AT91_MCI_SR));
-		}
-	}
-
-	cmdr = cmd->opcode;
-
-	if (mmc_resp_type(cmd) == MMC_RSP_NONE)
-		cmdr |= AT91_MCI_RSPTYP_NONE;
-	else {
-		/* if a response is expected then allow maximum response latancy */
-		cmdr |= AT91_MCI_MAXLAT;
-		/* set 136 bit response for R2, 48 bit response otherwise */
-		if (mmc_resp_type(cmd) == MMC_RSP_R2)
-			cmdr |= AT91_MCI_RSPTYP_136;
-		else
-			cmdr |= AT91_MCI_RSPTYP_48;
-	}
-
-	if (data) {
-
-		if (cpu_is_at91rm9200() || cpu_is_at91sam9261()) {
-			if (data->blksz & 0x3) {
-				pr_debug("Unsupported block size\n");
-				cmd->error = -EINVAL;
-				mmc_request_done(host->mmc, host->request);
-				return;
-			}
-			if (data->flags & MMC_DATA_STREAM) {
-				pr_debug("Stream commands not supported\n");
-				cmd->error = -EINVAL;
-				mmc_request_done(host->mmc, host->request);
-				return;
-			}
-		}
-
-		block_length = data->blksz;
-		blocks = data->blocks;
-
-		/* always set data start - also set direction flag for read */
-		if (data->flags & MMC_DATA_READ)
-			cmdr |= (AT91_MCI_TRDIR | AT91_MCI_TRCMD_START);
-		else if (data->flags & MMC_DATA_WRITE)
-			cmdr |= AT91_MCI_TRCMD_START;
-
-		if (data->flags & MMC_DATA_STREAM)
-			cmdr |= AT91_MCI_TRTYP_STREAM;
-		if (data->blocks > 1)
-			cmdr |= AT91_MCI_TRTYP_MULTIPLE;
-	}
-	else {
-		block_length = 0;
-		blocks = 0;
-	}
-
-	if (host->flags & FL_SENT_STOP)
-		cmdr |= AT91_MCI_TRCMD_STOP;
-
-	if (host->bus_mode == MMC_BUSMODE_OPENDRAIN)
-		cmdr |= AT91_MCI_OPDCMD;
-
-	/*
-	 * Set the arguments and send the command
-	 */
-	pr_debug("Sending command %d as %08X, arg = %08X, blocks = %d, length = %d (MR = %08X)\n",
-		cmd->opcode, cmdr, cmd->arg, blocks, block_length, at91_mci_read(host, AT91_MCI_MR));
-
-	if (!data) {
-		at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_TXTDIS | ATMEL_PDC_RXTDIS);
-		at91_mci_write(host, ATMEL_PDC_RPR, 0);
-		at91_mci_write(host, ATMEL_PDC_RCR, 0);
-		at91_mci_write(host, ATMEL_PDC_RNPR, 0);
-		at91_mci_write(host, ATMEL_PDC_RNCR, 0);
-		at91_mci_write(host, ATMEL_PDC_TPR, 0);
-		at91_mci_write(host, ATMEL_PDC_TCR, 0);
-		at91_mci_write(host, ATMEL_PDC_TNPR, 0);
-		at91_mci_write(host, ATMEL_PDC_TNCR, 0);
-		ier = AT91_MCI_CMDRDY;
-	} else {
-		/* zero block length and PDC mode */
-		mr = at91_mci_read(host, AT91_MCI_MR) & 0x5fff;
-		mr |= (data->blksz & 0x3) ? AT91_MCI_PDCFBYTE : 0;
-		mr |= (block_length << 16);
-		mr |= AT91_MCI_PDCMODE;
-		at91_mci_write(host, AT91_MCI_MR, mr);
-
-		if (!(cpu_is_at91rm9200() || cpu_is_at91sam9261()))
-			at91_mci_write(host, AT91_MCI_BLKR,
-				AT91_MCI_BLKR_BCNT(blocks) |
-				AT91_MCI_BLKR_BLKLEN(block_length));
-
-		/*
-		 * Disable the PDC controller
-		 */
-		at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTDIS | ATMEL_PDC_TXTDIS);
-
-		if (cmdr & AT91_MCI_TRCMD_START) {
-			data->bytes_xfered = 0;
-			host->transfer_index = 0;
-			host->in_use_index = 0;
-			if (cmdr & AT91_MCI_TRDIR) {
-				/*
-				 * Handle a read
-				 */
-				host->buffer = NULL;
-				host->total_length = 0;
-
-				at91_mci_pre_dma_read(host);
-				ier = AT91_MCI_ENDRX /* | AT91_MCI_RXBUFF */;
-			}
-			else {
-				/*
-				 * Handle a write
-				 */
-				host->total_length = block_length * blocks;
-				/*
-				 * AT91SAM926[0/3] Data Write Operation and
-				 * number of bytes erratum
-				 */
-				if (cpu_is_at91sam9260 () || cpu_is_at91sam9263())
-					if (host->total_length < 12)
-						host->total_length = 12;
-
-				host->buffer = kmalloc(host->total_length, GFP_KERNEL);
-				if (!host->buffer) {
-					pr_debug("Can't alloc tx buffer\n");
-					cmd->error = -ENOMEM;
-					mmc_request_done(host->mmc, host->request);
-					return;
-				}
-
-				at91_mci_sg_to_dma(host, data);
-
-				host->physical_address = dma_map_single(NULL,
-						host->buffer, host->total_length,
-						DMA_TO_DEVICE);
-
-				pr_debug("Transmitting %d bytes\n", host->total_length);
-
-				at91_mci_write(host, ATMEL_PDC_TPR, host->physical_address);
-				at91_mci_write(host, ATMEL_PDC_TCR, (data->blksz & 0x3) ?
-						host->total_length : host->total_length / 4);
-
-				ier = AT91_MCI_CMDRDY;
-			}
-		}
-	}
-
-	/*
-	 * Send the command and then enable the PDC - not the other way round as
-	 * the data sheet says
-	 */
-
-	at91_mci_write(host, AT91_MCI_ARGR, cmd->arg);
-	at91_mci_write(host, AT91_MCI_CMDR, cmdr);
-
-	if (cmdr & AT91_MCI_TRCMD_START) {
-		if (cmdr & AT91_MCI_TRDIR)
-			at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTEN);
-	}
-
-	/* Enable selected interrupts */
-	at91_mci_write(host, AT91_MCI_IER, AT91_MCI_ERRORS | ier);
-}
-
-/*
- * Process the next step in the request
- */
-static void at91_mci_process_next(struct at91mci_host *host)
-{
-	if (!(host->flags & FL_SENT_COMMAND)) {
-		host->flags |= FL_SENT_COMMAND;
-		at91_mci_send_command(host, host->request->cmd);
-	}
-	else if ((!(host->flags & FL_SENT_STOP)) && host->request->stop) {
-		host->flags |= FL_SENT_STOP;
-		at91_mci_send_command(host, host->request->stop);
-	} else {
-		del_timer(&host->timer);
-		/* the at91rm9200 mci controller hangs after some transfers,
-		 * and the workaround is to reset it after each transfer.
-		 */
-		if (cpu_is_at91rm9200())
-			at91_reset_host(host);
-		mmc_request_done(host->mmc, host->request);
-	}
-}
-
-/*
- * Handle a command that has been completed
- */
-static void at91_mci_completed_command(struct at91mci_host *host, unsigned int status)
-{
-	struct mmc_command *cmd = host->cmd;
-	struct mmc_data *data = cmd->data;
-
-	at91_mci_write(host, AT91_MCI_IDR, 0xffffffff & ~(AT91_MCI_SDIOIRQA | AT91_MCI_SDIOIRQB));
-
-	cmd->resp[0] = at91_mci_read(host, AT91_MCI_RSPR(0));
-	cmd->resp[1] = at91_mci_read(host, AT91_MCI_RSPR(1));
-	cmd->resp[2] = at91_mci_read(host, AT91_MCI_RSPR(2));
-	cmd->resp[3] = at91_mci_read(host, AT91_MCI_RSPR(3));
-
-	if (host->buffer) {
-		dma_unmap_single(NULL,
-				host->physical_address, host->total_length,
-				DMA_TO_DEVICE);
-		kfree(host->buffer);
-		host->buffer = NULL;
-	}
-
-	pr_debug("Status = %08X/%08x [%08X %08X %08X %08X]\n",
-		 status, at91_mci_read(host, AT91_MCI_SR),
-		 cmd->resp[0], cmd->resp[1], cmd->resp[2], cmd->resp[3]);
-
-	if (status & AT91_MCI_ERRORS) {
-		if ((status & AT91_MCI_RCRCE) && !(mmc_resp_type(cmd) & MMC_RSP_CRC)) {
-			cmd->error = 0;
-		}
-		else {
-			if (status & (AT91_MCI_DTOE | AT91_MCI_DCRCE)) {
-				if (data) {
-					if (status & AT91_MCI_DTOE)
-						data->error = -ETIMEDOUT;
-					else if (status & AT91_MCI_DCRCE)
-						data->error = -EILSEQ;
-				}
-			} else {
-				if (status & AT91_MCI_RTOE)
-					cmd->error = -ETIMEDOUT;
-				else if (status & AT91_MCI_RCRCE)
-					cmd->error = -EILSEQ;
-				else
-					cmd->error = -EIO;
-			}
-
-			pr_debug("Error detected and set to %d/%d (cmd = %d, retries = %d)\n",
-				cmd->error, data ? data->error : 0,
-				 cmd->opcode, cmd->retries);
-		}
-	}
-	else
-		cmd->error = 0;
-
-	at91_mci_process_next(host);
-}
-
-/*
- * Handle an MMC request
- */
-static void at91_mci_request(struct mmc_host *mmc, struct mmc_request *mrq)
-{
-	struct at91mci_host *host = mmc_priv(mmc);
-	host->request = mrq;
-	host->flags = 0;
-
-	mod_timer(&host->timer, jiffies +  HZ);
-
-	at91_mci_process_next(host);
-}
-
-/*
- * Set the IOS
- */
-static void at91_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
-{
-	int clkdiv;
-	struct at91mci_host *host = mmc_priv(mmc);
-	unsigned long at91_master_clock = clk_get_rate(host->mci_clk);
-
-	host->bus_mode = ios->bus_mode;
-
-	if (ios->clock == 0) {
-		/* Disable the MCI controller */
-		at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIDIS);
-		clkdiv = 0;
-	}
-	else {
-		/* Enable the MCI controller */
-		at91_mci_write(host, AT91_MCI_CR, AT91_MCI_MCIEN);
-
-		if ((at91_master_clock % (ios->clock * 2)) == 0)
-			clkdiv = ((at91_master_clock / ios->clock) / 2) - 1;
-		else
-			clkdiv = (at91_master_clock / ios->clock) / 2;
-
-		pr_debug("clkdiv = %d. mcck = %ld\n", clkdiv,
-			at91_master_clock / (2 * (clkdiv + 1)));
-	}
-	if (ios->bus_width == MMC_BUS_WIDTH_4 && host->board->wire4) {
-		pr_debug("MMC: Setting controller bus width to 4\n");
-		at91_mci_write(host, AT91_MCI_SDCR, at91_mci_read(host, AT91_MCI_SDCR) | AT91_MCI_SDCBUS);
-	}
-	else {
-		pr_debug("MMC: Setting controller bus width to 1\n");
-		at91_mci_write(host, AT91_MCI_SDCR, at91_mci_read(host, AT91_MCI_SDCR) & ~AT91_MCI_SDCBUS);
-	}
-
-	/* Set the clock divider */
-	at91_mci_write(host, AT91_MCI_MR, (at91_mci_read(host, AT91_MCI_MR) & ~AT91_MCI_CLKDIV) | clkdiv);
-
-	/* maybe switch power to the card */
-	if (host->board->vcc_pin) {
-		switch (ios->power_mode) {
-			case MMC_POWER_OFF:
-				gpio_set_value(host->board->vcc_pin, 0);
-				break;
-			case MMC_POWER_UP:
-				gpio_set_value(host->board->vcc_pin, 1);
-				break;
-			case MMC_POWER_ON:
-				break;
-			default:
-				WARN_ON(1);
-		}
-	}
-}
-
-/*
- * Handle an interrupt
- */
-static irqreturn_t at91_mci_irq(int irq, void *devid)
-{
-	struct at91mci_host *host = devid;
-	int completed = 0;
-	unsigned int int_status, int_mask;
-
-	int_status = at91_mci_read(host, AT91_MCI_SR);
-	int_mask = at91_mci_read(host, AT91_MCI_IMR);
-
-	pr_debug("MCI irq: status = %08X, %08X, %08X\n", int_status, int_mask,
-		int_status & int_mask);
-
-	int_status = int_status & int_mask;
-
-	if (int_status & AT91_MCI_ERRORS) {
-		completed = 1;
-
-		if (int_status & AT91_MCI_UNRE)
-			pr_debug("MMC: Underrun error\n");
-		if (int_status & AT91_MCI_OVRE)
-			pr_debug("MMC: Overrun error\n");
-		if (int_status & AT91_MCI_DTOE)
-			pr_debug("MMC: Data timeout\n");
-		if (int_status & AT91_MCI_DCRCE)
-			pr_debug("MMC: CRC error in data\n");
-		if (int_status & AT91_MCI_RTOE)
-			pr_debug("MMC: Response timeout\n");
-		if (int_status & AT91_MCI_RENDE)
-			pr_debug("MMC: Response end bit error\n");
-		if (int_status & AT91_MCI_RCRCE)
-			pr_debug("MMC: Response CRC error\n");
-		if (int_status & AT91_MCI_RDIRE)
-			pr_debug("MMC: Response direction error\n");
-		if (int_status & AT91_MCI_RINDE)
-			pr_debug("MMC: Response index error\n");
-	} else {
-		/* Only continue processing if no errors */
-
-		if (int_status & AT91_MCI_TXBUFE) {
-			pr_debug("TX buffer empty\n");
-			at91_mci_handle_transmitted(host);
-		}
-
-		if (int_status & AT91_MCI_ENDRX) {
-			pr_debug("ENDRX\n");
-			at91_mci_post_dma_read(host);
-		}
-
-		if (int_status & AT91_MCI_RXBUFF) {
-			pr_debug("RX buffer full\n");
-			at91_mci_write(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTDIS | ATMEL_PDC_TXTDIS);
-			at91_mci_write(host, AT91_MCI_IDR, AT91_MCI_RXBUFF | AT91_MCI_ENDRX);
-			completed = 1;
-		}
-
-		if (int_status & AT91_MCI_ENDTX)
-			pr_debug("Transmit has ended\n");
-
-		if (int_status & AT91_MCI_NOTBUSY) {
-			pr_debug("Card is ready\n");
-			at91_mci_update_bytes_xfered(host);
-			completed = 1;
-		}
-
-		if (int_status & AT91_MCI_DTIP)
-			pr_debug("Data transfer in progress\n");
-
-		if (int_status & AT91_MCI_BLKE) {
-			pr_debug("Block transfer has ended\n");
-			if (host->request->data && host->request->data->blocks > 1) {
-				/* multi block write : complete multi write
-				 * command and send stop */
-				completed = 1;
-			} else {
-				at91_mci_write(host, AT91_MCI_IER, AT91_MCI_NOTBUSY);
-			}
-		}
-
-		if (int_status & AT91_MCI_SDIOIRQA)
-			mmc_signal_sdio_irq(host->mmc);
-
-		if (int_status & AT91_MCI_SDIOIRQB)
-			mmc_signal_sdio_irq(host->mmc);
-
-		if (int_status & AT91_MCI_TXRDY)
-			pr_debug("Ready to transmit\n");
-
-		if (int_status & AT91_MCI_RXRDY)
-			pr_debug("Ready to receive\n");
-
-		if (int_status & AT91_MCI_CMDRDY) {
-			pr_debug("Command ready\n");
-			completed = at91_mci_handle_cmdrdy(host);
-		}
-	}
-
-	if (completed) {
-		pr_debug("Completed command\n");
-		at91_mci_write(host, AT91_MCI_IDR, 0xffffffff & ~(AT91_MCI_SDIOIRQA | AT91_MCI_SDIOIRQB));
-		at91_mci_completed_command(host, int_status);
-	} else
-		at91_mci_write(host, AT91_MCI_IDR, int_status & ~(AT91_MCI_SDIOIRQA | AT91_MCI_SDIOIRQB));
-
-	return IRQ_HANDLED;
-}
-
-static irqreturn_t at91_mmc_det_irq(int irq, void *_host)
-{
-	struct at91mci_host *host = _host;
-	int present = !gpio_get_value(irq_to_gpio(irq));
-
-	/*
-	 * we expect this irq on both insert and remove,
-	 * and use a short delay to debounce.
-	 */
-	if (present != host->present) {
-		host->present = present;
-		pr_debug("%s: card %s\n", mmc_hostname(host->mmc),
-			present ? "insert" : "remove");
-		if (!present) {
-			pr_debug("****** Resetting SD-card bus width ******\n");
-			at91_mci_write(host, AT91_MCI_SDCR, at91_mci_read(host, AT91_MCI_SDCR) & ~AT91_MCI_SDCBUS);
-		}
-		mmc_detect_change(host->mmc, msecs_to_jiffies(100));
-	}
-	return IRQ_HANDLED;
-}
-
-static int at91_mci_get_ro(struct mmc_host *mmc)
-{
-	struct at91mci_host *host = mmc_priv(mmc);
-
-	if (host->board->wp_pin)
-		return !!gpio_get_value(host->board->wp_pin);
-	/*
-	 * Board doesn't support read only detection; let the mmc core
-	 * decide what to do.
-	 */
-	return -ENOSYS;
-}
-
-static void at91_mci_enable_sdio_irq(struct mmc_host *mmc, int enable)
-{
-	struct at91mci_host *host = mmc_priv(mmc);
-
-	pr_debug("%s: sdio_irq %c : %s\n", mmc_hostname(host->mmc),
-		host->board->slot_b ? 'B':'A', enable ? "enable" : "disable");
-	at91_mci_write(host, enable ? AT91_MCI_IER : AT91_MCI_IDR,
-		host->board->slot_b ? AT91_MCI_SDIOIRQB : AT91_MCI_SDIOIRQA);
-
-}
-
-static const struct mmc_host_ops at91_mci_ops = {
-	.request	= at91_mci_request,
-	.set_ios	= at91_mci_set_ios,
-	.get_ro		= at91_mci_get_ro,
-	.enable_sdio_irq = at91_mci_enable_sdio_irq,
-};
-
-/*
- * Probe for the device
- */
-static int __init at91_mci_probe(struct platform_device *pdev)
-{
-	struct mmc_host *mmc;
-	struct at91mci_host *host;
-	struct resource *res;
-	int ret;
-
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!res)
-		return -ENXIO;
-
-	if (!request_mem_region(res->start, res->end - res->start + 1, DRIVER_NAME))
-		return -EBUSY;
-
-	mmc = mmc_alloc_host(sizeof(struct at91mci_host), &pdev->dev);
-	if (!mmc) {
-		ret = -ENOMEM;
-		dev_dbg(&pdev->dev, "couldn't allocate mmc host\n");
-		goto fail6;
-	}
-
-	mmc->ops = &at91_mci_ops;
-	mmc->f_min = 375000;
-	mmc->f_max = 25000000;
-	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
-	mmc->caps = MMC_CAP_SDIO_IRQ;
-
-	mmc->max_blk_size = 4095;
-	mmc->max_blk_count = mmc->max_req_size;
-
-	host = mmc_priv(mmc);
-	host->mmc = mmc;
-	host->buffer = NULL;
-	host->bus_mode = 0;
-	host->board = pdev->dev.platform_data;
-	if (host->board->wire4) {
-		if (cpu_is_at91sam9260() || cpu_is_at91sam9263())
-			mmc->caps |= MMC_CAP_4_BIT_DATA;
-		else
-			dev_warn(&pdev->dev, "4 wire bus mode not supported"
-				" - using 1 wire\n");
-	}
-
-	/*
-	 * Reserve GPIOs ... board init code makes sure these pins are set
-	 * up as GPIOs with the right direction (input, except for vcc)
-	 */
-	if (host->board->det_pin) {
-		ret = gpio_request(host->board->det_pin, "mmc_detect");
-		if (ret < 0) {
-			dev_dbg(&pdev->dev, "couldn't claim card detect pin\n");
-			goto fail5;
-		}
-	}
-	if (host->board->wp_pin) {
-		ret = gpio_request(host->board->wp_pin, "mmc_wp");
-		if (ret < 0) {
-			dev_dbg(&pdev->dev, "couldn't claim wp sense pin\n");
-			goto fail4;
-		}
-	}
-	if (host->board->vcc_pin) {
-		ret = gpio_request(host->board->vcc_pin, "mmc_vcc");
-		if (ret < 0) {
-			dev_dbg(&pdev->dev, "couldn't claim vcc switch pin\n");
-			goto fail3;
-		}
-	}
-
-	/*
-	 * Get Clock
-	 */
-	host->mci_clk = clk_get(&pdev->dev, "mci_clk");
-	if (IS_ERR(host->mci_clk)) {
-		ret = -ENODEV;
-		dev_dbg(&pdev->dev, "no mci_clk?\n");
-		goto fail2;
-	}
-
-	/*
-	 * Map I/O region
-	 */
-	host->baseaddr = ioremap(res->start, res->end - res->start + 1);
-	if (!host->baseaddr) {
-		ret = -ENOMEM;
-		goto fail1;
-	}
-
-	/*
-	 * Reset hardware
-	 */
-	clk_enable(host->mci_clk);		/* Enable the peripheral clock */
-	at91_mci_disable(host);
-	at91_mci_enable(host);
-
-	/*
-	 * Allocate the MCI interrupt
-	 */
-	host->irq = platform_get_irq(pdev, 0);
-	ret = request_irq(host->irq, at91_mci_irq, IRQF_SHARED,
-			mmc_hostname(mmc), host);
-	if (ret) {
-		dev_dbg(&pdev->dev, "request MCI interrupt failed\n");
-		goto fail0;
-	}
-
-	setup_timer(&host->timer, at91_timeout_timer, (unsigned long)host);
-
-	platform_set_drvdata(pdev, mmc);
-
-	/*
-	 * Add host to MMC layer
-	 */
-	if (host->board->det_pin) {
-		host->present = !gpio_get_value(host->board->det_pin);
-	}
-	else
-		host->present = -1;
-
-	mmc_add_host(mmc);
-
-	/*
-	 * monitor card insertion/removal if we can
-	 */
-	if (host->board->det_pin) {
-		ret = request_irq(gpio_to_irq(host->board->det_pin),
-				at91_mmc_det_irq, 0, mmc_hostname(mmc), host);
-		if (ret)
-			dev_warn(&pdev->dev, "request MMC detect irq failed\n");
-		else
-			device_init_wakeup(&pdev->dev, 1);
-	}
-
-	pr_debug("Added MCI driver\n");
-
-	return 0;
-
-fail0:
-	clk_disable(host->mci_clk);
-	iounmap(host->baseaddr);
-fail1:
-	clk_put(host->mci_clk);
-fail2:
-	if (host->board->vcc_pin)
-		gpio_free(host->board->vcc_pin);
-fail3:
-	if (host->board->wp_pin)
-		gpio_free(host->board->wp_pin);
-fail4:
-	if (host->board->det_pin)
-		gpio_free(host->board->det_pin);
-fail5:
-	mmc_free_host(mmc);
-fail6:
-	release_mem_region(res->start, res->end - res->start + 1);
-	dev_err(&pdev->dev, "probe failed, err %d\n", ret);
-	return ret;
-}
-
-/*
- * Remove a device
- */
-static int __exit at91_mci_remove(struct platform_device *pdev)
-{
-	struct mmc_host *mmc = platform_get_drvdata(pdev);
-	struct at91mci_host *host;
-	struct resource *res;
-
-	if (!mmc)
-		return -1;
-
-	host = mmc_priv(mmc);
-
-	if (host->board->det_pin) {
-		if (device_can_wakeup(&pdev->dev))
-			free_irq(gpio_to_irq(host->board->det_pin), host);
-		device_init_wakeup(&pdev->dev, 0);
-		gpio_free(host->board->det_pin);
-	}
-
-	at91_mci_disable(host);
-	del_timer_sync(&host->timer);
-	mmc_remove_host(mmc);
-	free_irq(host->irq, host);
-
-	clk_disable(host->mci_clk);			/* Disable the peripheral clock */
-	clk_put(host->mci_clk);
-
-	if (host->board->vcc_pin)
-		gpio_free(host->board->vcc_pin);
-	if (host->board->wp_pin)
-		gpio_free(host->board->wp_pin);
-
-	iounmap(host->baseaddr);
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	release_mem_region(res->start, res->end - res->start + 1);
-
-	mmc_free_host(mmc);
-	platform_set_drvdata(pdev, NULL);
-	pr_debug("MCI Removed\n");
-
-	return 0;
-}
-
-#ifdef CONFIG_PM
-static int at91_mci_suspend(struct platform_device *pdev, pm_message_t state)
-{
-	struct mmc_host *mmc = platform_get_drvdata(pdev);
-	struct at91mci_host *host = mmc_priv(mmc);
-	int ret = 0;
-
-	if (host->board->det_pin && device_may_wakeup(&pdev->dev))
-		enable_irq_wake(host->board->det_pin);
-
-	if (mmc)
-		ret = mmc_suspend_host(mmc, state);
-
-	return ret;
-}
-
-static int at91_mci_resume(struct platform_device *pdev)
-{
-	struct mmc_host *mmc = platform_get_drvdata(pdev);
-	struct at91mci_host *host = mmc_priv(mmc);
-	int ret = 0;
-
-	if (host->board->det_pin && device_may_wakeup(&pdev->dev))
-		disable_irq_wake(host->board->det_pin);
-
-	if (mmc)
-		ret = mmc_resume_host(mmc);
-
-	return ret;
-}
-#else
-#define at91_mci_suspend	NULL
-#define at91_mci_resume		NULL
-#endif
-
-static struct platform_driver at91_mci_driver = {
-	.remove		= __exit_p(at91_mci_remove),
-	.suspend	= at91_mci_suspend,
-	.resume		= at91_mci_resume,
-	.driver		= {
-		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
-	},
-};
-
-static int __init at91_mci_init(void)
-{
-	return platform_driver_probe(&at91_mci_driver, at91_mci_probe);
-}
-
-static void __exit at91_mci_exit(void)
-{
-	platform_driver_unregister(&at91_mci_driver);
-}
-
-module_init(at91_mci_init);
-module_exit(at91_mci_exit);
-
-MODULE_DESCRIPTION("AT91 Multimedia Card Interface driver");
-MODULE_AUTHOR("Nick Randell");
-MODULE_LICENSE("GPL");
-MODULE_ALIAS("platform:at91_mci");
diff --git a/drivers/mmc/host/atmel-mci-regs.h b/drivers/mmc/host/atmel-mci-regs.h
index b58364ed6bb..c97001e1522 100644
--- a/drivers/mmc/host/atmel-mci-regs.h
+++ b/drivers/mmc/host/atmel-mci-regs.h
@@ -7,89 +7,158 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
+
+/*
+ * Superset of MCI IP registers integrated in Atmel AVR32 and AT91 Processors
+ * Registers and bitfields marked with [2] are only available in MCI2
+ */
+
 #ifndef __DRIVERS_MMC_ATMEL_MCI_H__
 #define __DRIVERS_MMC_ATMEL_MCI_H__
 
 /* MCI Register Definitions */
-#define MCI_CR			0x0000	/* Control */
-# define MCI_CR_MCIEN		(  1 <<  0)	/* MCI Enable */
-# define MCI_CR_MCIDIS		(  1 <<  1)	/* MCI Disable */
-# define MCI_CR_SWRST		(  1 <<  7)	/* Software Reset */
-#define MCI_MR			0x0004	/* Mode */
-# define MCI_MR_CLKDIV(x)	((x) <<  0)	/* Clock Divider */
-# define MCI_MR_RDPROOF		(  1 << 11)	/* Read Proof */
-# define MCI_MR_WRPROOF		(  1 << 12)	/* Write Proof */
-#define MCI_DTOR		0x0008	/* Data Timeout */
-# define MCI_DTOCYC(x)		((x) <<  0)	/* Data Timeout Cycles */
-# define MCI_DTOMUL(x)		((x) <<  4)	/* Data Timeout Multiplier */
-#define MCI_SDCR		0x000c	/* SD Card / SDIO */
-# define MCI_SDCSEL_SLOT_A	(  0 <<  0)	/* Select SD slot A */
-# define MCI_SDCSEL_SLOT_B	(  1 <<  0)	/* Select SD slot A */
-# define MCI_SDCSEL_MASK	(  3 <<  0)
-# define MCI_SDCBUS_1BIT	(  0 <<  6)	/* 1-bit data bus */
-# define MCI_SDCBUS_4BIT	(  2 <<  6)	/* 4-bit data bus */
-# define MCI_SDCBUS_MASK	(  3 <<  6)
-#define MCI_ARGR		0x0010	/* Command Argument */
-#define MCI_CMDR		0x0014	/* Command */
-# define MCI_CMDR_CMDNB(x)	((x) <<  0)	/* Command Opcode */
-# define MCI_CMDR_RSPTYP_NONE	(  0 <<  6)	/* No response */
-# define MCI_CMDR_RSPTYP_48BIT	(  1 <<  6)	/* 48-bit response */
-# define MCI_CMDR_RSPTYP_136BIT	(  2 <<  6)	/* 136-bit response */
-# define MCI_CMDR_SPCMD_INIT	(  1 <<  8)	/* Initialization command */
-# define MCI_CMDR_SPCMD_SYNC	(  2 <<  8)	/* Synchronized command */
-# define MCI_CMDR_SPCMD_INT	(  4 <<  8)	/* Interrupt command */
-# define MCI_CMDR_SPCMD_INTRESP	(  5 <<  8)	/* Interrupt response */
-# define MCI_CMDR_OPDCMD	(  1 << 11)	/* Open Drain */
-# define MCI_CMDR_MAXLAT_5CYC	(  0 << 12)	/* Max latency 5 cycles */
-# define MCI_CMDR_MAXLAT_64CYC	(  1 << 12)	/* Max latency 64 cycles */
-# define MCI_CMDR_START_XFER	(  1 << 16)	/* Start data transfer */
-# define MCI_CMDR_STOP_XFER	(  2 << 16)	/* Stop data transfer */
-# define MCI_CMDR_TRDIR_WRITE	(  0 << 18)	/* Write data */
-# define MCI_CMDR_TRDIR_READ	(  1 << 18)	/* Read data */
-# define MCI_CMDR_BLOCK		(  0 << 19)	/* Single-block transfer */
-# define MCI_CMDR_MULTI_BLOCK	(  1 << 19)	/* Multi-block transfer */
-# define MCI_CMDR_STREAM	(  2 << 19)	/* MMC Stream transfer */
-# define MCI_CMDR_SDIO_BYTE	(  4 << 19)	/* SDIO Byte transfer */
-# define MCI_CMDR_SDIO_BLOCK	(  5 << 19)	/* SDIO Block transfer */
-# define MCI_CMDR_SDIO_SUSPEND	(  1 << 24)	/* SDIO Suspend Command */
-# define MCI_CMDR_SDIO_RESUME	(  2 << 24)	/* SDIO Resume Command */
-#define MCI_BLKR		0x0018	/* Block */
-# define MCI_BCNT(x)		((x) <<  0)	/* Data Block Count */
-# define MCI_BLKLEN(x)		((x) << 16)	/* Data Block Length */
-#define MCI_RSPR		0x0020	/* Response 0 */
-#define MCI_RSPR1		0x0024	/* Response 1 */
-#define MCI_RSPR2		0x0028	/* Response 2 */
-#define MCI_RSPR3		0x002c	/* Response 3 */
-#define MCI_RDR			0x0030	/* Receive Data */
-#define MCI_TDR			0x0034	/* Transmit Data */
-#define MCI_SR			0x0040	/* Status */
-#define MCI_IER			0x0044	/* Interrupt Enable */
-#define MCI_IDR			0x0048	/* Interrupt Disable */
-#define MCI_IMR			0x004c	/* Interrupt Mask */
-# define MCI_CMDRDY		(  1 <<   0)	/* Command Ready */
-# define MCI_RXRDY		(  1 <<   1)	/* Receiver Ready */
-# define MCI_TXRDY		(  1 <<   2)	/* Transmitter Ready */
-# define MCI_BLKE		(  1 <<   3)	/* Data Block Ended */
-# define MCI_DTIP		(  1 <<   4)	/* Data Transfer In Progress */
-# define MCI_NOTBUSY		(  1 <<   5)	/* Data Not Busy */
-# define MCI_SDIOIRQA		(  1 <<   8)	/* SDIO IRQ in slot A */
-# define MCI_SDIOIRQB		(  1 <<   9)	/* SDIO IRQ in slot B */
-# define MCI_RINDE		(  1 <<  16)	/* Response Index Error */
-# define MCI_RDIRE		(  1 <<  17)	/* Response Direction Error */
-# define MCI_RCRCE		(  1 <<  18)	/* Response CRC Error */
-# define MCI_RENDE		(  1 <<  19)	/* Response End Bit Error */
-# define MCI_RTOE		(  1 <<  20)	/* Response Time-Out Error */
-# define MCI_DCRCE		(  1 <<  21)	/* Data CRC Error */
-# define MCI_DTOE		(  1 <<  22)	/* Data Time-Out Error */
-# define MCI_OVRE		(  1 <<  30)	/* RX Overrun Error */
-# define MCI_UNRE		(  1 <<  31)	/* TX Underrun Error */
+#define ATMCI_CR			0x0000	/* Control */
+# define ATMCI_CR_MCIEN			(  1 <<  0)	/* MCI Enable */
+# define ATMCI_CR_MCIDIS		(  1 <<  1)	/* MCI Disable */
+# define ATMCI_CR_PWSEN			(  1 <<  2)	/* Power Save Enable */
+# define ATMCI_CR_PWSDIS		(  1 <<  3)	/* Power Save Disable */
+# define ATMCI_CR_SWRST			(  1 <<  7)	/* Software Reset */
+#define ATMCI_MR			0x0004	/* Mode */
+# define ATMCI_MR_CLKDIV(x)		((x) <<  0)	/* Clock Divider */
+# define ATMCI_MR_PWSDIV(x)		((x) <<  8)	/* Power Saving Divider */
+# define ATMCI_MR_RDPROOF		(  1 << 11)	/* Read Proof */
+# define ATMCI_MR_WRPROOF		(  1 << 12)	/* Write Proof */
+# define ATMCI_MR_PDCFBYTE		(  1 << 13)	/* Force Byte Transfer */
+# define ATMCI_MR_PDCPADV		(  1 << 14)	/* Padding Value */
+# define ATMCI_MR_PDCMODE		(  1 << 15)	/* PDC-oriented Mode */
+# define ATMCI_MR_CLKODD(x)		((x) << 16)	/* LSB of Clock Divider */
+#define ATMCI_DTOR			0x0008	/* Data Timeout */
+# define ATMCI_DTOCYC(x)		((x) <<  0)	/* Data Timeout Cycles */
+# define ATMCI_DTOMUL(x)		((x) <<  4)	/* Data Timeout Multiplier */
+#define ATMCI_SDCR			0x000c	/* SD Card / SDIO */
+# define ATMCI_SDCSEL_SLOT_A		(  0 <<  0)	/* Select SD slot A */
+# define ATMCI_SDCSEL_SLOT_B		(  1 <<  0)	/* Select SD slot A */
+# define ATMCI_SDCSEL_MASK		(  3 <<  0)
+# define ATMCI_SDCBUS_1BIT		(  0 <<  6)	/* 1-bit data bus */
+# define ATMCI_SDCBUS_4BIT		(  2 <<  6)	/* 4-bit data bus */
+# define ATMCI_SDCBUS_8BIT		(  3 <<  6)	/* 8-bit data bus[2] */
+# define ATMCI_SDCBUS_MASK		(  3 <<  6)
+#define ATMCI_ARGR			0x0010	/* Command Argument */
+#define ATMCI_CMDR			0x0014	/* Command */
+# define ATMCI_CMDR_CMDNB(x)		((x) <<  0)	/* Command Opcode */
+# define ATMCI_CMDR_RSPTYP_NONE		(  0 <<  6)	/* No response */
+# define ATMCI_CMDR_RSPTYP_48BIT	(  1 <<  6)	/* 48-bit response */
+# define ATMCI_CMDR_RSPTYP_136BIT	(  2 <<  6)	/* 136-bit response */
+# define ATMCI_CMDR_SPCMD_INIT		(  1 <<  8)	/* Initialization command */
+# define ATMCI_CMDR_SPCMD_SYNC		(  2 <<  8)	/* Synchronized command */
+# define ATMCI_CMDR_SPCMD_INT		(  4 <<  8)	/* Interrupt command */
+# define ATMCI_CMDR_SPCMD_INTRESP	(  5 <<  8)	/* Interrupt response */
+# define ATMCI_CMDR_OPDCMD		(  1 << 11)	/* Open Drain */
+# define ATMCI_CMDR_MAXLAT_5CYC		(  0 << 12)	/* Max latency 5 cycles */
+# define ATMCI_CMDR_MAXLAT_64CYC	(  1 << 12)	/* Max latency 64 cycles */
+# define ATMCI_CMDR_START_XFER		(  1 << 16)	/* Start data transfer */
+# define ATMCI_CMDR_STOP_XFER		(  2 << 16)	/* Stop data transfer */
+# define ATMCI_CMDR_TRDIR_WRITE		(  0 << 18)	/* Write data */
+# define ATMCI_CMDR_TRDIR_READ		(  1 << 18)	/* Read data */
+# define ATMCI_CMDR_BLOCK		(  0 << 19)	/* Single-block transfer */
+# define ATMCI_CMDR_MULTI_BLOCK		(  1 << 19)	/* Multi-block transfer */
+# define ATMCI_CMDR_STREAM		(  2 << 19)	/* MMC Stream transfer */
+# define ATMCI_CMDR_SDIO_BYTE		(  4 << 19)	/* SDIO Byte transfer */
+# define ATMCI_CMDR_SDIO_BLOCK		(  5 << 19)	/* SDIO Block transfer */
+# define ATMCI_CMDR_SDIO_SUSPEND	(  1 << 24)	/* SDIO Suspend Command */
+# define ATMCI_CMDR_SDIO_RESUME		(  2 << 24)	/* SDIO Resume Command */
+#define ATMCI_BLKR			0x0018	/* Block */
+# define ATMCI_BCNT(x)			((x) <<  0)	/* Data Block Count */
+# define ATMCI_BLKLEN(x)		((x) << 16)	/* Data Block Length */
+#define ATMCI_CSTOR			0x001c	/* Completion Signal Timeout[2] */
+# define ATMCI_CSTOCYC(x)		((x) <<  0)	/* CST cycles */
+# define ATMCI_CSTOMUL(x)		((x) <<  4)	/* CST multiplier */
+#define ATMCI_RSPR			0x0020	/* Response 0 */
+#define ATMCI_RSPR1			0x0024	/* Response 1 */
+#define ATMCI_RSPR2			0x0028	/* Response 2 */
+#define ATMCI_RSPR3			0x002c	/* Response 3 */
+#define ATMCI_RDR			0x0030	/* Receive Data */
+#define ATMCI_TDR			0x0034	/* Transmit Data */
+#define ATMCI_SR			0x0040	/* Status */
+#define ATMCI_IER			0x0044	/* Interrupt Enable */
+#define ATMCI_IDR			0x0048	/* Interrupt Disable */
+#define ATMCI_IMR			0x004c	/* Interrupt Mask */
+# define ATMCI_CMDRDY			(  1 <<   0)	/* Command Ready */
+# define ATMCI_RXRDY			(  1 <<   1)	/* Receiver Ready */
+# define ATMCI_TXRDY			(  1 <<   2)	/* Transmitter Ready */
+# define ATMCI_BLKE			(  1 <<   3)	/* Data Block Ended */
+# define ATMCI_DTIP			(  1 <<   4)	/* Data Transfer In Progress */
+# define ATMCI_NOTBUSY			(  1 <<   5)	/* Data Not Busy */
+# define ATMCI_ENDRX			(  1 <<   6)    /* End of RX Buffer */
+# define ATMCI_ENDTX			(  1 <<   7)    /* End of TX Buffer */
+# define ATMCI_SDIOIRQA			(  1 <<   8)	/* SDIO IRQ in slot A */
+# define ATMCI_SDIOIRQB			(  1 <<   9)	/* SDIO IRQ in slot B */
+# define ATMCI_SDIOWAIT			(  1 <<  12)    /* SDIO Read Wait Operation Status */
+# define ATMCI_CSRCV			(  1 <<  13)    /* CE-ATA Completion Signal Received */
+# define ATMCI_RXBUFF			(  1 <<  14)    /* RX Buffer Full */
+# define ATMCI_TXBUFE			(  1 <<  15)    /* TX Buffer Empty */
+# define ATMCI_RINDE			(  1 <<  16)	/* Response Index Error */
+# define ATMCI_RDIRE			(  1 <<  17)	/* Response Direction Error */
+# define ATMCI_RCRCE			(  1 <<  18)	/* Response CRC Error */
+# define ATMCI_RENDE			(  1 <<  19)	/* Response End Bit Error */
+# define ATMCI_RTOE			(  1 <<  20)	/* Response Time-Out Error */
+# define ATMCI_DCRCE			(  1 <<  21)	/* Data CRC Error */
+# define ATMCI_DTOE			(  1 <<  22)	/* Data Time-Out Error */
+# define ATMCI_CSTOE			(  1 <<  23)    /* Completion Signal Time-out Error */
+# define ATMCI_BLKOVRE			(  1 <<  24)    /* DMA Block Overrun Error */
+# define ATMCI_DMADONE			(  1 <<  25)    /* DMA Transfer Done */
+# define ATMCI_FIFOEMPTY		(  1 <<  26)    /* FIFO Empty Flag */
+# define ATMCI_XFRDONE			(  1 <<  27)    /* Transfer Done Flag */
+# define ATMCI_ACKRCV			(  1 <<  28)    /* Boot Operation Acknowledge Received */
+# define ATMCI_ACKRCVE			(  1 <<  29)    /* Boot Operation Acknowledge Error */
+# define ATMCI_OVRE			(  1 <<  30)	/* RX Overrun Error */
+# define ATMCI_UNRE			(  1 <<  31)	/* TX Underrun Error */
+#define ATMCI_DMA			0x0050	/* DMA Configuration[2] */
+# define ATMCI_DMA_OFFSET(x)		((x) <<  0)	/* DMA Write Buffer Offset */
+# define ATMCI_DMA_CHKSIZE(x)		((x) <<  4)	/* DMA Channel Read and Write Chunk Size */
+# define ATMCI_DMAEN			(  1 <<  8)	/* DMA Hardware Handshaking Enable */
+#define ATMCI_CFG			0x0054	/* Configuration[2] */
+# define ATMCI_CFG_FIFOMODE_1DATA	(  1 <<  0)	/* MCI Internal FIFO control mode */
+# define ATMCI_CFG_FERRCTRL_COR		(  1 <<  4)	/* Flow Error flag reset control mode */
+# define ATMCI_CFG_HSMODE		(  1 <<  8)	/* High Speed Mode */
+# define ATMCI_CFG_LSYNC		(  1 << 12)	/* Synchronize on the last block */
+#define ATMCI_WPMR			0x00e4	/* Write Protection Mode[2] */
+# define ATMCI_WP_EN			(  1 <<  0)	/* WP Enable */
+# define ATMCI_WP_KEY			(0x4d4349 << 8)	/* WP Key */
+#define ATMCI_WPSR			0x00e8	/* Write Protection Status[2] */
+# define ATMCI_GET_WP_VS(x)		((x) & 0x0f)
+# define ATMCI_GET_WP_VSRC(x)		(((x) >> 8) & 0xffff)
+#define ATMCI_VERSION			0x00FC  /* Version */
+#define ATMCI_FIFO_APERTURE		0x0200	/* FIFO Aperture[2] */
 
-#define MCI_REGS_SIZE		0x100
+/* This is not including the FIFO Aperture on MCI2 */
+#define ATMCI_REGS_SIZE		0x100
 
 /* Register access macros */
-#define mci_readl(port,reg)				\
-	__raw_readl((port)->regs + MCI_##reg)
-#define mci_writel(port,reg,value)			\
-	__raw_writel((value), (port)->regs + MCI_##reg)
+#define atmci_readl(port,reg)				\
+	__raw_readl((port)->regs + reg)
+#define atmci_writel(port,reg,value)			\
+	__raw_writel((value), (port)->regs + reg)
+
+/* On AVR chips the Peripheral DMA Controller is not connected to MCI. */
+#ifdef CONFIG_AVR32
+#	define ATMCI_PDC_CONNECTED	0
+#else
+#	define ATMCI_PDC_CONNECTED	1
+#endif
+
+/*
+ * Fix sconfig's burst size according to atmel MCI. We need to convert them as:
+ * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3.
+ *
+ * This can be done by finding most significant bit set.
+ */
+static inline unsigned int atmci_convert_chksize(unsigned int maxburst)
+{
+	if (maxburst > 1)
+		return fls(maxburst) - 2;
+	else
+		return 0;
+}
 
 #endif /* __DRIVERS_MMC_ATMEL_MCI_H__ */
diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c
index cf6a100bb38..bb585d94090 100644
--- a/drivers/mmc/host/atmel-mci.c
+++ b/drivers/mmc/host/atmel-mci.c
@@ -19,59 +19,96 @@
 #include <linux/interrupt.h>
 #include <linux/ioport.h>
 #include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
 #include <linux/platform_device.h>
 #include <linux/scatterlist.h>
 #include <linux/seq_file.h>
+#include <linux/slab.h>
 #include <linux/stat.h>
+#include <linux/types.h>
+#include <linux/platform_data/atmel.h>
 
 #include <linux/mmc/host.h>
+#include <linux/mmc/sdio.h>
+
+#include <mach/atmel-mci.h>
 #include <linux/atmel-mci.h>
+#include <linux/atmel_pdc.h>
 
+#include <asm/cacheflush.h>
 #include <asm/io.h>
 #include <asm/unaligned.h>
 
-#include <mach/board.h>
-
 #include "atmel-mci-regs.h"
 
-#define ATMCI_DATA_ERROR_FLAGS	(MCI_DCRCE | MCI_DTOE | MCI_OVRE | MCI_UNRE)
+#define ATMCI_DATA_ERROR_FLAGS	(ATMCI_DCRCE | ATMCI_DTOE | ATMCI_OVRE | ATMCI_UNRE)
 #define ATMCI_DMA_THRESHOLD	16
 
 enum {
-	EVENT_CMD_COMPLETE = 0,
+	EVENT_CMD_RDY = 0,
 	EVENT_XFER_COMPLETE,
-	EVENT_DATA_COMPLETE,
+	EVENT_NOTBUSY,
 	EVENT_DATA_ERROR,
 };
 
 enum atmel_mci_state {
 	STATE_IDLE = 0,
 	STATE_SENDING_CMD,
-	STATE_SENDING_DATA,
-	STATE_DATA_BUSY,
+	STATE_DATA_XFER,
+	STATE_WAITING_NOTBUSY,
 	STATE_SENDING_STOP,
-	STATE_DATA_ERROR,
+	STATE_END_REQUEST,
+};
+
+enum atmci_xfer_dir {
+	XFER_RECEIVE = 0,
+	XFER_TRANSMIT,
+};
+
+enum atmci_pdc_buf {
+	PDC_FIRST_BUF = 0,
+	PDC_SECOND_BUF,
+};
+
+struct atmel_mci_caps {
+	bool    has_dma_conf_reg;
+	bool    has_pdc;
+	bool    has_cfg_reg;
+	bool    has_cstor_reg;
+	bool    has_highspeed;
+	bool    has_rwproof;
+	bool	has_odd_clk_div;
+	bool	has_bad_data_ordering;
+	bool	need_reset_after_xfer;
+	bool	need_blksz_mul_4;
+	bool	need_notbusy_for_read_ops;
 };
 
 struct atmel_mci_dma {
-#ifdef CONFIG_MMC_ATMELMCI_DMA
 	struct dma_chan			*chan;
 	struct dma_async_tx_descriptor	*data_desc;
-#endif
 };
 
 /**
  * struct atmel_mci - MMC controller state shared between all slots
  * @lock: Spinlock protecting the queue and associated data.
  * @regs: Pointer to MMIO registers.
- * @sg: Scatterlist entry currently being processed by PIO code, if any.
+ * @sg: Scatterlist entry currently being processed by PIO or PDC code.
  * @pio_offset: Offset into the current scatterlist entry.
+ * @buffer: Buffer used if we don't have the r/w proof capability. We
+ *      don't have the time to switch pdc buffers so we have to use only
+ *      one buffer for the full transaction.
+ * @buf_size: size of the buffer.
+ * @phys_buf_addr: buffer address needed for pdc.
  * @cur_slot: The slot which is currently using the controller.
  * @mrq: The request currently being processed on @cur_slot,
  *	or NULL if the controller is idle.
  * @cmd: The command currently being sent to the card, or NULL.
  * @data: The data currently being transferred, or NULL if no data
  *	transfer is in progress.
+ * @data_size: just data->blocks * data->blksz.
  * @dma: DMA client state.
  * @data_chan: DMA channel being used for the current data transfer.
  * @cmd_status: Snapshot of SR taken upon completion of the current
@@ -90,13 +127,22 @@ struct atmel_mci_dma {
  * @queue: List of slots waiting for access to the controller.
  * @need_clock_update: Update the clock rate before the next request.
  * @need_reset: Reset controller before next request.
+ * @timer: Timer to balance the data timeout error flag which cannot rise.
  * @mode_reg: Value of the MR register.
+ * @cfg_reg: Value of the CFG register.
  * @bus_hz: The rate of @mck in Hz. This forms the basis for MMC bus
  *	rate and timeout calculations.
  * @mapbase: Physical address of the MMIO registers.
  * @mck: The peripheral bus clock hooked up to the MMC controller.
  * @pdev: Platform device associated with the MMC controller.
  * @slot: Slots sharing this MMC controller.
+ * @caps: MCI capabilities depending on MCI version.
+ * @prepare_data: function to setup MCI before data transfer which
+ * depends on MCI capabilities.
+ * @submit_data: function to start data transfer which depends on MCI
+ * capabilities.
+ * @stop_transfer: function to stop data transfer which depends on MCI
+ * capabilities.
  *
  * Locking
  * =======
@@ -121,7 +167,7 @@ struct atmel_mci_dma {
  * EVENT_DATA_COMPLETE is set in @pending_events, all data-related
  * interrupts must be disabled and @data_status updated with a
  * snapshot of SR. Similarly, before EVENT_CMD_COMPLETE is set, the
- * CMDRDY interupt must be disabled and @cmd_status updated with a
+ * CMDRDY interrupt must be disabled and @cmd_status updated with a
  * snapshot of SR, and before EVENT_XFER_COMPLETE can be set, the
  * bytes_xfered field of @data must be written. This is ensured by
  * using barriers.
@@ -131,15 +177,21 @@ struct atmel_mci {
 	void __iomem		*regs;
 
 	struct scatterlist	*sg;
+	unsigned int		sg_len;
 	unsigned int		pio_offset;
+	unsigned int		*buffer;
+	unsigned int		buf_size;
+	dma_addr_t		buf_phys_addr;
 
 	struct atmel_mci_slot	*cur_slot;
 	struct mmc_request	*mrq;
 	struct mmc_command	*cmd;
 	struct mmc_data		*data;
+	unsigned int		data_size;
 
 	struct atmel_mci_dma	dma;
 	struct dma_chan		*data_chan;
+	struct dma_slave_config	dma_conf;
 
 	u32			cmd_status;
 	u32			data_status;
@@ -153,13 +205,21 @@ struct atmel_mci {
 
 	bool			need_clock_update;
 	bool			need_reset;
+	struct timer_list	timer;
 	u32			mode_reg;
+	u32			cfg_reg;
 	unsigned long		bus_hz;
 	unsigned long		mapbase;
 	struct clk		*mck;
 	struct platform_device	*pdev;
 
-	struct atmel_mci_slot	*slot[ATMEL_MCI_MAX_NR_SLOTS];
+	struct atmel_mci_slot	*slot[ATMCI_MAX_NR_SLOTS];
+
+	struct atmel_mci_caps   caps;
+
+	u32 (*prepare_data)(struct atmel_mci *host, struct mmc_data *data);
+	void (*submit_data)(struct atmel_mci *host, struct mmc_data *data);
+	void (*stop_transfer)(struct atmel_mci *host);
 };
 
 /**
@@ -167,6 +227,7 @@ struct atmel_mci {
  * @mmc: The mmc_host representing this slot.
  * @host: The MMC controller this slot is using.
  * @sdc_reg: Value of SDCR to be written before using this slot.
+ * @sdio_irq: SDIO irq mask for this slot.
  * @mrq: mmc_request currently being processed or waiting to be
  *	processed, or NULL when the slot is idle.
  * @queue_node: List node for placing this node in the @queue list of
@@ -177,6 +238,7 @@ struct atmel_mci {
  *	available.
  * @wp_pin: GPIO pin used for card write protect sending, or negative
  *	if not available.
+ * @detect_is_active_high: The state of the detect pin when it is active.
  * @detect_timer: Timer used for debouncing @detect_pin interrupts.
  */
 struct atmel_mci_slot {
@@ -184,6 +246,7 @@ struct atmel_mci_slot {
 	struct atmel_mci	*host;
 
 	u32			sdc_reg;
+	u32			sdio_irq;
 
 	struct mmc_request	*mrq;
 	struct list_head	queue_node;
@@ -196,6 +259,7 @@ struct atmel_mci_slot {
 
 	int			detect_pin;
 	int			wp_pin;
+	bool			detect_is_active_high;
 
 	struct timer_list	detect_timer;
 };
@@ -233,7 +297,7 @@ static int atmci_req_show(struct seq_file *s, void *v)
 				"CMD%u(0x%x) flg %x rsp %x %x %x %x err %d\n",
 				cmd->opcode, cmd->arg, cmd->flags,
 				cmd->resp[0], cmd->resp[1], cmd->resp[2],
-				cmd->resp[2], cmd->error);
+				cmd->resp[3], cmd->error);
 		if (data)
 			seq_printf(s, "DATA %u / %u * %u flg %x err %d\n",
 				data->bytes_xfered, data->blocks,
@@ -243,7 +307,7 @@ static int atmci_req_show(struct seq_file *s, void *v)
 				"CMD%u(0x%x) flg %x rsp %x %x %x %x err %d\n",
 				stop->opcode, stop->arg, stop->flags,
 				stop->resp[0], stop->resp[1], stop->resp[2],
-				stop->resp[2], stop->error);
+				stop->resp[3], stop->error);
 	}
 
 	spin_unlock_bh(&slot->host->lock);
@@ -274,8 +338,13 @@ static void atmci_show_status_reg(struct seq_file *s,
 		[3]	= "BLKE",
 		[4]	= "DTIP",
 		[5]	= "NOTBUSY",
+		[6]	= "ENDRX",
+		[7]	= "ENDTX",
 		[8]	= "SDIOIRQA",
 		[9]	= "SDIOIRQB",
+		[12]	= "SDIOWAIT",
+		[14]	= "RXBUFF",
+		[15]	= "TXBUFE",
 		[16]	= "RINDE",
 		[17]	= "RDIRE",
 		[18]	= "RCRCE",
@@ -283,6 +352,11 @@ static void atmci_show_status_reg(struct seq_file *s,
 		[20]	= "RTOE",
 		[21]	= "DCRCE",
 		[22]	= "DTOE",
+		[23]	= "CSTOE",
+		[24]	= "BLKOVRE",
+		[25]	= "DMADONE",
+		[26]	= "FIFOEMPTY",
+		[27]	= "XFRDONE",
 		[30]	= "OVRE",
 		[31]	= "UNRE",
 	};
@@ -304,8 +378,10 @@ static int atmci_regs_show(struct seq_file *s, void *v)
 {
 	struct atmel_mci	*host = s->private;
 	u32			*buf;
+	int			ret = 0;
+
 
-	buf = kmalloc(MCI_REGS_SIZE, GFP_KERNEL);
+	buf = kmalloc(ATMCI_REGS_SIZE, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
 
@@ -314,33 +390,68 @@ static int atmci_regs_show(struct seq_file *s, void *v)
 	 * not disabling interrupts, so IMR and SR may not be
 	 * consistent.
 	 */
+	ret = clk_prepare_enable(host->mck);
+	if (ret)
+		goto out;
+
 	spin_lock_bh(&host->lock);
-	clk_enable(host->mck);
-	memcpy_fromio(buf, host->regs, MCI_REGS_SIZE);
-	clk_disable(host->mck);
+	memcpy_fromio(buf, host->regs, ATMCI_REGS_SIZE);
 	spin_unlock_bh(&host->lock);
 
-	seq_printf(s, "MR:\t0x%08x%s%s CLKDIV=%u\n",
-			buf[MCI_MR / 4],
-			buf[MCI_MR / 4] & MCI_MR_RDPROOF ? " RDPROOF" : "",
-			buf[MCI_MR / 4] & MCI_MR_WRPROOF ? " WRPROOF" : "",
-			buf[MCI_MR / 4] & 0xff);
-	seq_printf(s, "DTOR:\t0x%08x\n", buf[MCI_DTOR / 4]);
-	seq_printf(s, "SDCR:\t0x%08x\n", buf[MCI_SDCR / 4]);
-	seq_printf(s, "ARGR:\t0x%08x\n", buf[MCI_ARGR / 4]);
+	clk_disable_unprepare(host->mck);
+
+	seq_printf(s, "MR:\t0x%08x%s%s ",
+			buf[ATMCI_MR / 4],
+			buf[ATMCI_MR / 4] & ATMCI_MR_RDPROOF ? " RDPROOF" : "",
+			buf[ATMCI_MR / 4] & ATMCI_MR_WRPROOF ? " WRPROOF" : "");
+	if (host->caps.has_odd_clk_div)
+		seq_printf(s, "{CLKDIV,CLKODD}=%u\n",
+				((buf[ATMCI_MR / 4] & 0xff) << 1)
+				| ((buf[ATMCI_MR / 4] >> 16) & 1));
+	else
+		seq_printf(s, "CLKDIV=%u\n",
+				(buf[ATMCI_MR / 4] & 0xff));
+	seq_printf(s, "DTOR:\t0x%08x\n", buf[ATMCI_DTOR / 4]);
+	seq_printf(s, "SDCR:\t0x%08x\n", buf[ATMCI_SDCR / 4]);
+	seq_printf(s, "ARGR:\t0x%08x\n", buf[ATMCI_ARGR / 4]);
 	seq_printf(s, "BLKR:\t0x%08x BCNT=%u BLKLEN=%u\n",
-			buf[MCI_BLKR / 4],
-			buf[MCI_BLKR / 4] & 0xffff,
-			(buf[MCI_BLKR / 4] >> 16) & 0xffff);
+			buf[ATMCI_BLKR / 4],
+			buf[ATMCI_BLKR / 4] & 0xffff,
+			(buf[ATMCI_BLKR / 4] >> 16) & 0xffff);
+	if (host->caps.has_cstor_reg)
+		seq_printf(s, "CSTOR:\t0x%08x\n", buf[ATMCI_CSTOR / 4]);
 
 	/* Don't read RSPR and RDR; it will consume the data there */
 
-	atmci_show_status_reg(s, "SR", buf[MCI_SR / 4]);
-	atmci_show_status_reg(s, "IMR", buf[MCI_IMR / 4]);
+	atmci_show_status_reg(s, "SR", buf[ATMCI_SR / 4]);
+	atmci_show_status_reg(s, "IMR", buf[ATMCI_IMR / 4]);
+
+	if (host->caps.has_dma_conf_reg) {
+		u32 val;
+
+		val = buf[ATMCI_DMA / 4];
+		seq_printf(s, "DMA:\t0x%08x OFFSET=%u CHKSIZE=%u%s\n",
+				val, val & 3,
+				((val >> 4) & 3) ?
+					1 << (((val >> 4) & 3) + 1) : 1,
+				val & ATMCI_DMAEN ? " DMAEN" : "");
+	}
+	if (host->caps.has_cfg_reg) {
+		u32 val;
+
+		val = buf[ATMCI_CFG / 4];
+		seq_printf(s, "CFG:\t0x%08x%s%s%s%s\n",
+				val,
+				val & ATMCI_CFG_FIFOMODE_1DATA ? " FIFOMODE_ONE_DATA" : "",
+				val & ATMCI_CFG_FERRCTRL_COR ? " FERRCTRL_CLEAR_ON_READ" : "",
+				val & ATMCI_CFG_HSMODE ? " HSMODE" : "",
+				val & ATMCI_CFG_LSYNC ? " LSYNC" : "");
+	}
 
+out:
 	kfree(buf);
 
-	return 0;
+	return ret;
 }
 
 static int atmci_regs_open(struct inode *inode, struct file *file)
@@ -398,10 +509,114 @@ err:
 	dev_err(&mmc->class_dev, "failed to initialize debugfs for slot\n");
 }
 
-static inline unsigned int ns_to_clocks(struct atmel_mci *host,
+#if defined(CONFIG_OF)
+static const struct of_device_id atmci_dt_ids[] = {
+	{ .compatible = "atmel,hsmci" },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(of, atmci_dt_ids);
+
+static struct mci_platform_data*
+atmci_of_init(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct device_node *cnp;
+	struct mci_platform_data *pdata;
+	u32 slot_id;
+
+	if (!np) {
+		dev_err(&pdev->dev, "device node not found\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata) {
+		dev_err(&pdev->dev, "could not allocate memory for pdata\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	for_each_child_of_node(np, cnp) {
+		if (of_property_read_u32(cnp, "reg", &slot_id)) {
+			dev_warn(&pdev->dev, "reg property is missing for %s\n",
+				 cnp->full_name);
+			continue;
+		}
+
+		if (slot_id >= ATMCI_MAX_NR_SLOTS) {
+			dev_warn(&pdev->dev, "can't have more than %d slots\n",
+			         ATMCI_MAX_NR_SLOTS);
+			break;
+		}
+
+		if (of_property_read_u32(cnp, "bus-width",
+		                         &pdata->slot[slot_id].bus_width))
+			pdata->slot[slot_id].bus_width = 1;
+
+		pdata->slot[slot_id].detect_pin =
+			of_get_named_gpio(cnp, "cd-gpios", 0);
+
+		pdata->slot[slot_id].detect_is_active_high =
+			of_property_read_bool(cnp, "cd-inverted");
+
+		pdata->slot[slot_id].wp_pin =
+			of_get_named_gpio(cnp, "wp-gpios", 0);
+	}
+
+	return pdata;
+}
+#else /* CONFIG_OF */
+static inline struct mci_platform_data*
+atmci_of_init(struct platform_device *dev)
+{
+	return ERR_PTR(-EINVAL);
+}
+#endif
+
+static inline unsigned int atmci_get_version(struct atmel_mci *host)
+{
+	return atmci_readl(host, ATMCI_VERSION) & 0x00000fff;
+}
+
+static void atmci_timeout_timer(unsigned long data)
+{
+	struct atmel_mci *host;
+
+	host = (struct atmel_mci *)data;
+
+	dev_dbg(&host->pdev->dev, "software timeout\n");
+
+	if (host->mrq->cmd->data) {
+		host->mrq->cmd->data->error = -ETIMEDOUT;
+		host->data = NULL;
+		/*
+		 * With some SDIO modules, sometimes DMA transfer hangs. If
+		 * stop_transfer() is not called then the DMA request is not
+		 * removed, following ones are queued and never computed.
+		 */
+		if (host->state == STATE_DATA_XFER)
+			host->stop_transfer(host);
+	} else {
+		host->mrq->cmd->error = -ETIMEDOUT;
+		host->cmd = NULL;
+	}
+	host->need_reset = 1;
+	host->state = STATE_END_REQUEST;
+	smp_wmb();
+	tasklet_schedule(&host->tasklet);
+}
+
+static inline unsigned int atmci_ns_to_clocks(struct atmel_mci *host,
 					unsigned int ns)
 {
-	return (ns * (host->bus_hz / 1000000) + 999) / 1000;
+	/*
+	 * It is easier here to use us instead of ns for the timeout,
+	 * it prevents from overflows during calculation.
+	 */
+	unsigned int us = DIV_ROUND_UP(ns, 1000);
+
+	/* Maximum clock frequency is host->bus_hz/2 */
+	return us * (DIV_ROUND_UP(host->bus_hz, 2000000));
 }
 
 static void atmci_set_timeout(struct atmel_mci *host,
@@ -414,7 +629,8 @@ static void atmci_set_timeout(struct atmel_mci *host,
 	unsigned	dtocyc;
 	unsigned	dtomul;
 
-	timeout = ns_to_clocks(host, data->timeout_ns) + data->timeout_clks;
+	timeout = atmci_ns_to_clocks(host, data->timeout_ns)
+		+ data->timeout_clks;
 
 	for (dtomul = 0; dtomul < 8; dtomul++) {
 		unsigned shift = dtomul_to_shift[dtomul];
@@ -430,7 +646,7 @@ static void atmci_set_timeout(struct atmel_mci *host,
 
 	dev_vdbg(&slot->mmc->class_dev, "setting timeout to %u cycles\n",
 			dtocyc << dtomul_to_shift[dtomul]);
-	mci_writel(host, DTOR, (MCI_DTOMUL(dtomul) | MCI_DTOCYC(dtocyc)));
+	atmci_writel(host, ATMCI_DTOR, (ATMCI_DTOMUL(dtomul) | ATMCI_DTOCYC(dtocyc)));
 }
 
 /*
@@ -444,13 +660,13 @@ static u32 atmci_prepare_command(struct mmc_host *mmc,
 
 	cmd->error = -EINPROGRESS;
 
-	cmdr = MCI_CMDR_CMDNB(cmd->opcode);
+	cmdr = ATMCI_CMDR_CMDNB(cmd->opcode);
 
 	if (cmd->flags & MMC_RSP_PRESENT) {
 		if (cmd->flags & MMC_RSP_136)
-			cmdr |= MCI_CMDR_RSPTYP_136BIT;
+			cmdr |= ATMCI_CMDR_RSPTYP_136BIT;
 		else
-			cmdr |= MCI_CMDR_RSPTYP_48BIT;
+			cmdr |= ATMCI_CMDR_RSPTYP_48BIT;
 	}
 
 	/*
@@ -458,29 +674,34 @@ static u32 atmci_prepare_command(struct mmc_host *mmc,
 	 * it's too difficult to determine whether this is an ACMD or
 	 * not. Better make it 64.
 	 */
-	cmdr |= MCI_CMDR_MAXLAT_64CYC;
+	cmdr |= ATMCI_CMDR_MAXLAT_64CYC;
 
 	if (mmc->ios.bus_mode == MMC_BUSMODE_OPENDRAIN)
-		cmdr |= MCI_CMDR_OPDCMD;
+		cmdr |= ATMCI_CMDR_OPDCMD;
 
 	data = cmd->data;
 	if (data) {
-		cmdr |= MCI_CMDR_START_XFER;
-		if (data->flags & MMC_DATA_STREAM)
-			cmdr |= MCI_CMDR_STREAM;
-		else if (data->blocks > 1)
-			cmdr |= MCI_CMDR_MULTI_BLOCK;
-		else
-			cmdr |= MCI_CMDR_BLOCK;
+		cmdr |= ATMCI_CMDR_START_XFER;
+
+		if (cmd->opcode == SD_IO_RW_EXTENDED) {
+			cmdr |= ATMCI_CMDR_SDIO_BLOCK;
+		} else {
+			if (data->flags & MMC_DATA_STREAM)
+				cmdr |= ATMCI_CMDR_STREAM;
+			else if (data->blocks > 1)
+				cmdr |= ATMCI_CMDR_MULTI_BLOCK;
+			else
+				cmdr |= ATMCI_CMDR_BLOCK;
+		}
 
 		if (data->flags & MMC_DATA_READ)
-			cmdr |= MCI_CMDR_TRDIR_READ;
+			cmdr |= ATMCI_CMDR_TRDIR_READ;
 	}
 
 	return cmdr;
 }
 
-static void atmci_start_command(struct atmel_mci *host,
+static void atmci_send_command(struct atmel_mci *host,
 		struct mmc_command *cmd, u32 cmd_flags)
 {
 	WARN_ON(host->cmd);
@@ -490,41 +711,135 @@ static void atmci_start_command(struct atmel_mci *host,
 			"start command: ARGR=0x%08x CMDR=0x%08x\n",
 			cmd->arg, cmd_flags);
 
-	mci_writel(host, ARGR, cmd->arg);
-	mci_writel(host, CMDR, cmd_flags);
+	atmci_writel(host, ATMCI_ARGR, cmd->arg);
+	atmci_writel(host, ATMCI_CMDR, cmd_flags);
 }
 
-static void send_stop_cmd(struct atmel_mci *host, struct mmc_data *data)
+static void atmci_send_stop_cmd(struct atmel_mci *host, struct mmc_data *data)
 {
-	atmci_start_command(host, data->stop, host->stop_cmdr);
-	mci_writel(host, IER, MCI_CMDRDY);
+	dev_dbg(&host->pdev->dev, "send stop command\n");
+	atmci_send_command(host, data->stop, host->stop_cmdr);
+	atmci_writel(host, ATMCI_IER, ATMCI_CMDRDY);
 }
 
-#ifdef CONFIG_MMC_ATMELMCI_DMA
-static void atmci_dma_cleanup(struct atmel_mci *host)
+/*
+ * Configure given PDC buffer taking care of alignement issues.
+ * Update host->data_size and host->sg.
+ */
+static void atmci_pdc_set_single_buf(struct atmel_mci *host,
+	enum atmci_xfer_dir dir, enum atmci_pdc_buf buf_nb)
 {
-	struct mmc_data			*data = host->data;
+	u32 pointer_reg, counter_reg;
+	unsigned int buf_size;
 
-	dma_unmap_sg(&host->pdev->dev, data->sg, data->sg_len,
-		     ((data->flags & MMC_DATA_WRITE)
-		      ? DMA_TO_DEVICE : DMA_FROM_DEVICE));
+	if (dir == XFER_RECEIVE) {
+		pointer_reg = ATMEL_PDC_RPR;
+		counter_reg = ATMEL_PDC_RCR;
+	} else {
+		pointer_reg = ATMEL_PDC_TPR;
+		counter_reg = ATMEL_PDC_TCR;
+	}
+
+	if (buf_nb == PDC_SECOND_BUF) {
+		pointer_reg += ATMEL_PDC_SCND_BUF_OFF;
+		counter_reg += ATMEL_PDC_SCND_BUF_OFF;
+	}
+
+	if (!host->caps.has_rwproof) {
+		buf_size = host->buf_size;
+		atmci_writel(host, pointer_reg, host->buf_phys_addr);
+	} else {
+		buf_size = sg_dma_len(host->sg);
+		atmci_writel(host, pointer_reg, sg_dma_address(host->sg));
+	}
+
+	if (host->data_size <= buf_size) {
+		if (host->data_size & 0x3) {
+			/* If size is different from modulo 4, transfer bytes */
+			atmci_writel(host, counter_reg, host->data_size);
+			atmci_writel(host, ATMCI_MR, host->mode_reg | ATMCI_MR_PDCFBYTE);
+		} else {
+			/* Else transfer 32-bits words */
+			atmci_writel(host, counter_reg, host->data_size / 4);
+		}
+		host->data_size = 0;
+	} else {
+		/* We assume the size of a page is 32-bits aligned */
+		atmci_writel(host, counter_reg, sg_dma_len(host->sg) / 4);
+		host->data_size -= sg_dma_len(host->sg);
+		if (host->data_size)
+			host->sg = sg_next(host->sg);
+	}
 }
 
-static void atmci_stop_dma(struct atmel_mci *host)
+/*
+ * Configure PDC buffer according to the data size ie configuring one or two
+ * buffers. Don't use this function if you want to configure only the second
+ * buffer. In this case, use atmci_pdc_set_single_buf.
+ */
+static void atmci_pdc_set_both_buf(struct atmel_mci *host, int dir)
 {
-	struct dma_chan *chan = host->data_chan;
+	atmci_pdc_set_single_buf(host, dir, PDC_FIRST_BUF);
+	if (host->data_size)
+		atmci_pdc_set_single_buf(host, dir, PDC_SECOND_BUF);
+}
 
-	if (chan) {
-		chan->device->device_terminate_all(chan);
-		atmci_dma_cleanup(host);
-	} else {
-		/* Data transfer was stopped by the interrupt handler */
-		atmci_set_pending(host, EVENT_XFER_COMPLETE);
-		mci_writel(host, IER, MCI_NOTBUSY);
+/*
+ * Unmap sg lists, called when transfer is finished.
+ */
+static void atmci_pdc_cleanup(struct atmel_mci *host)
+{
+	struct mmc_data         *data = host->data;
+
+	if (data)
+		dma_unmap_sg(&host->pdev->dev,
+				data->sg, data->sg_len,
+				((data->flags & MMC_DATA_WRITE)
+				 ? DMA_TO_DEVICE : DMA_FROM_DEVICE));
+}
+
+/*
+ * Disable PDC transfers. Update pending flags to EVENT_XFER_COMPLETE after
+ * having received ATMCI_TXBUFE or ATMCI_RXBUFF interrupt. Enable ATMCI_NOTBUSY
+ * interrupt needed for both transfer directions.
+ */
+static void atmci_pdc_complete(struct atmel_mci *host)
+{
+	int transfer_size = host->data->blocks * host->data->blksz;
+	int i;
+
+	atmci_writel(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTDIS | ATMEL_PDC_TXTDIS);
+
+	if ((!host->caps.has_rwproof)
+	    && (host->data->flags & MMC_DATA_READ)) {
+		if (host->caps.has_bad_data_ordering)
+			for (i = 0; i < transfer_size; i++)
+				host->buffer[i] = swab32(host->buffer[i]);
+		sg_copy_from_buffer(host->data->sg, host->data->sg_len,
+		                    host->buffer, transfer_size);
 	}
+
+	atmci_pdc_cleanup(host);
+
+	dev_dbg(&host->pdev->dev, "(%s) set pending xfer complete\n", __func__);
+	atmci_set_pending(host, EVENT_XFER_COMPLETE);
+	tasklet_schedule(&host->tasklet);
+}
+
+static void atmci_dma_cleanup(struct atmel_mci *host)
+{
+	struct mmc_data                 *data = host->data;
+
+	if (data)
+		dma_unmap_sg(host->dma.chan->device->dev,
+				data->sg, data->sg_len,
+				((data->flags & MMC_DATA_WRITE)
+				 ? DMA_TO_DEVICE : DMA_FROM_DEVICE));
 }
 
-/* This function is called by the DMA driver from tasklet context. */
+/*
+ * This function is called by the DMA driver from tasklet context.
+ */
 static void atmci_dma_complete(void *arg)
 {
 	struct atmel_mci	*host = arg;
@@ -532,6 +847,10 @@ static void atmci_dma_complete(void *arg)
 
 	dev_vdbg(&host->pdev->dev, "DMA complete\n");
 
+	if (host->caps.has_dma_conf_reg)
+		/* Disable DMA hardware handshaking on MCI */
+		atmci_writel(host, ATMCI_DMA, atmci_readl(host, ATMCI_DMA) & ~ATMCI_DMAEN);
+
 	atmci_dma_cleanup(host);
 
 	/*
@@ -539,6 +858,8 @@ static void atmci_dma_complete(void *arg)
 	 * to send the stop command or waiting for NBUSY in this case.
 	 */
 	if (data) {
+		dev_dbg(&host->pdev->dev,
+		        "(%s) set pending xfer complete\n", __func__);
 		atmci_set_pending(host, EVENT_XFER_COMPLETE);
 		tasklet_schedule(&host->tasklet);
 
@@ -562,18 +883,124 @@ static void atmci_dma_complete(void *arg)
 		 * completion callback" rule of the dma engine
 		 * framework.
 		 */
-		mci_writel(host, IER, MCI_NOTBUSY);
+		atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
 	}
 }
 
-static int
-atmci_submit_data_dma(struct atmel_mci *host, struct mmc_data *data)
+/*
+ * Returns a mask of interrupt flags to be enabled after the whole
+ * request has been prepared.
+ */
+static u32 atmci_prepare_data(struct atmel_mci *host, struct mmc_data *data)
+{
+	u32 iflags;
+
+	data->error = -EINPROGRESS;
+
+	host->sg = data->sg;
+	host->sg_len = data->sg_len;
+	host->data = data;
+	host->data_chan = NULL;
+
+	iflags = ATMCI_DATA_ERROR_FLAGS;
+
+	/*
+	 * Errata: MMC data write operation with less than 12
+	 * bytes is impossible.
+	 *
+	 * Errata: MCI Transmit Data Register (TDR) FIFO
+	 * corruption when length is not multiple of 4.
+	 */
+	if (data->blocks * data->blksz < 12
+			|| (data->blocks * data->blksz) & 3)
+		host->need_reset = true;
+
+	host->pio_offset = 0;
+	if (data->flags & MMC_DATA_READ)
+		iflags |= ATMCI_RXRDY;
+	else
+		iflags |= ATMCI_TXRDY;
+
+	return iflags;
+}
+
+/*
+ * Set interrupt flags and set block length into the MCI mode register even
+ * if this value is also accessible in the MCI block register. It seems to be
+ * necessary before the High Speed MCI version. It also map sg and configure
+ * PDC registers.
+ */
+static u32
+atmci_prepare_data_pdc(struct atmel_mci *host, struct mmc_data *data)
+{
+	u32 iflags, tmp;
+	unsigned int sg_len;
+	enum dma_data_direction dir;
+	int i;
+
+	data->error = -EINPROGRESS;
+
+	host->data = data;
+	host->sg = data->sg;
+	iflags = ATMCI_DATA_ERROR_FLAGS;
+
+	/* Enable pdc mode */
+	atmci_writel(host, ATMCI_MR, host->mode_reg | ATMCI_MR_PDCMODE);
+
+	if (data->flags & MMC_DATA_READ) {
+		dir = DMA_FROM_DEVICE;
+		iflags |= ATMCI_ENDRX | ATMCI_RXBUFF;
+	} else {
+		dir = DMA_TO_DEVICE;
+		iflags |= ATMCI_ENDTX | ATMCI_TXBUFE | ATMCI_BLKE;
+	}
+
+	/* Set BLKLEN */
+	tmp = atmci_readl(host, ATMCI_MR);
+	tmp &= 0x0000ffff;
+	tmp |= ATMCI_BLKLEN(data->blksz);
+	atmci_writel(host, ATMCI_MR, tmp);
+
+	/* Configure PDC */
+	host->data_size = data->blocks * data->blksz;
+	sg_len = dma_map_sg(&host->pdev->dev, data->sg, data->sg_len, dir);
+
+	if ((!host->caps.has_rwproof)
+	    && (host->data->flags & MMC_DATA_WRITE)) {
+		sg_copy_to_buffer(host->data->sg, host->data->sg_len,
+		                  host->buffer, host->data_size);
+		if (host->caps.has_bad_data_ordering)
+			for (i = 0; i < host->data_size; i++)
+				host->buffer[i] = swab32(host->buffer[i]);
+	}
+
+	if (host->data_size)
+		atmci_pdc_set_both_buf(host,
+			((dir == DMA_FROM_DEVICE) ? XFER_RECEIVE : XFER_TRANSMIT));
+
+	return iflags;
+}
+
+static u32
+atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data)
 {
 	struct dma_chan			*chan;
 	struct dma_async_tx_descriptor	*desc;
 	struct scatterlist		*sg;
 	unsigned int			i;
 	enum dma_data_direction		direction;
+	enum dma_transfer_direction	slave_dirn;
+	unsigned int			sglen;
+	u32				maxburst;
+	u32 iflags;
+
+	data->error = -EINPROGRESS;
+
+	WARN_ON(host->data);
+	host->sg = NULL;
+	host->data = data;
+
+	iflags = ATMCI_DATA_ERROR_FLAGS;
 
 	/*
 	 * We don't do DMA on "complex" transfers, i.e. with
@@ -581,13 +1008,13 @@ atmci_submit_data_dma(struct atmel_mci *host, struct mmc_data *data)
 	 * with all the DMA setup overhead for short transfers.
 	 */
 	if (data->blocks * data->blksz < ATMCI_DMA_THRESHOLD)
-		return -EINVAL;
+		return atmci_prepare_data(host, data);
 	if (data->blksz & 3)
-		return -EINVAL;
+		return atmci_prepare_data(host, data);
 
 	for_each_sg(data->sg, sg, data->sg_len, i) {
 		if (sg->offset & 3 || sg->length & 3)
-			return -EINVAL;
+			return atmci_prepare_data(host, data);
 	}
 
 	/* If we don't have a channel, we can't do DMA */
@@ -598,84 +1025,106 @@ atmci_submit_data_dma(struct atmel_mci *host, struct mmc_data *data)
 	if (!chan)
 		return -ENODEV;
 
-	if (data->flags & MMC_DATA_READ)
+	if (data->flags & MMC_DATA_READ) {
 		direction = DMA_FROM_DEVICE;
-	else
+		host->dma_conf.direction = slave_dirn = DMA_DEV_TO_MEM;
+		maxburst = atmci_convert_chksize(host->dma_conf.src_maxburst);
+	} else {
 		direction = DMA_TO_DEVICE;
+		host->dma_conf.direction = slave_dirn = DMA_MEM_TO_DEV;
+		maxburst = atmci_convert_chksize(host->dma_conf.dst_maxburst);
+	}
+
+	if (host->caps.has_dma_conf_reg)
+		atmci_writel(host, ATMCI_DMA, ATMCI_DMA_CHKSIZE(maxburst) |
+			ATMCI_DMAEN);
+
+	sglen = dma_map_sg(chan->device->dev, data->sg,
+			data->sg_len, direction);
 
-	desc = chan->device->device_prep_slave_sg(chan,
-			data->sg, data->sg_len, direction,
+	dmaengine_slave_config(chan, &host->dma_conf);
+	desc = dmaengine_prep_slave_sg(chan,
+			data->sg, sglen, slave_dirn,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc)
-		return -ENOMEM;
+		goto unmap_exit;
 
 	host->dma.data_desc = desc;
 	desc->callback = atmci_dma_complete;
 	desc->callback_param = host;
-	desc->tx_submit(desc);
 
-	/* Go! */
-	chan->device->device_issue_pending(chan);
+	return iflags;
+unmap_exit:
+	dma_unmap_sg(chan->device->dev, data->sg, data->sg_len, direction);
+	return -ENOMEM;
+}
 
-	return 0;
+static void
+atmci_submit_data(struct atmel_mci *host, struct mmc_data *data)
+{
+	return;
 }
 
-#else /* CONFIG_MMC_ATMELMCI_DMA */
+/*
+ * Start PDC according to transfer direction.
+ */
+static void
+atmci_submit_data_pdc(struct atmel_mci *host, struct mmc_data *data)
+{
+	if (data->flags & MMC_DATA_READ)
+		atmci_writel(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTEN);
+	else
+		atmci_writel(host, ATMEL_PDC_PTCR, ATMEL_PDC_TXTEN);
+}
 
-static int atmci_submit_data_dma(struct atmel_mci *host, struct mmc_data *data)
+static void
+atmci_submit_data_dma(struct atmel_mci *host, struct mmc_data *data)
 {
-	return -ENOSYS;
+	struct dma_chan			*chan = host->data_chan;
+	struct dma_async_tx_descriptor	*desc = host->dma.data_desc;
+
+	if (chan) {
+		dmaengine_submit(desc);
+		dma_async_issue_pending(chan);
+	}
 }
 
-static void atmci_stop_dma(struct atmel_mci *host)
+static void atmci_stop_transfer(struct atmel_mci *host)
 {
-	/* Data transfer was stopped by the interrupt handler */
+	dev_dbg(&host->pdev->dev,
+	        "(%s) set pending xfer complete\n", __func__);
 	atmci_set_pending(host, EVENT_XFER_COMPLETE);
-	mci_writel(host, IER, MCI_NOTBUSY);
+	atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
 }
 
-#endif /* CONFIG_MMC_ATMELMCI_DMA */
-
 /*
- * Returns a mask of interrupt flags to be enabled after the whole
- * request has been prepared.
+ * Stop data transfer because error(s) occurred.
  */
-static u32 atmci_submit_data(struct atmel_mci *host, struct mmc_data *data)
+static void atmci_stop_transfer_pdc(struct atmel_mci *host)
 {
-	u32 iflags;
-
-	data->error = -EINPROGRESS;
-
-	WARN_ON(host->data);
-	host->sg = NULL;
-	host->data = data;
-
-	iflags = ATMCI_DATA_ERROR_FLAGS;
-	if (atmci_submit_data_dma(host, data)) {
-		host->data_chan = NULL;
+	atmci_writel(host, ATMEL_PDC_PTCR, ATMEL_PDC_RXTDIS | ATMEL_PDC_TXTDIS);
+}
 
-		/*
-		 * Errata: MMC data write operation with less than 12
-		 * bytes is impossible.
-		 *
-		 * Errata: MCI Transmit Data Register (TDR) FIFO
-		 * corruption when length is not multiple of 4.
-		 */
-		if (data->blocks * data->blksz < 12
-				|| (data->blocks * data->blksz) & 3)
-			host->need_reset = true;
+static void atmci_stop_transfer_dma(struct atmel_mci *host)
+{
+	struct dma_chan *chan = host->data_chan;
 
-		host->sg = data->sg;
-		host->pio_offset = 0;
-		if (data->flags & MMC_DATA_READ)
-			iflags |= MCI_RXRDY;
-		else
-			iflags |= MCI_TXRDY;
+	if (chan) {
+		dmaengine_terminate_all(chan);
+		atmci_dma_cleanup(host);
+	} else {
+		/* Data transfer was stopped by the interrupt handler */
+		dev_dbg(&host->pdev->dev,
+		        "(%s) set pending xfer complete\n", __func__);
+		atmci_set_pending(host, EVENT_XFER_COMPLETE);
+		atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
 	}
-
-	return iflags;
 }
 
+/*
+ * Start a request: prepare data if needed, prepare the command and activate
+ * interrupts.
+ */
 static void atmci_start_request(struct atmel_mci *host,
 		struct atmel_mci_slot *slot)
 {
@@ -691,55 +1140,77 @@ static void atmci_start_request(struct atmel_mci *host,
 
 	host->pending_events = 0;
 	host->completed_events = 0;
+	host->cmd_status = 0;
 	host->data_status = 0;
 
-	if (host->need_reset) {
-		mci_writel(host, CR, MCI_CR_SWRST);
-		mci_writel(host, CR, MCI_CR_MCIEN);
-		mci_writel(host, MR, host->mode_reg);
+	dev_dbg(&host->pdev->dev, "start request: cmd %u\n", mrq->cmd->opcode);
+
+	if (host->need_reset || host->caps.need_reset_after_xfer) {
+		iflags = atmci_readl(host, ATMCI_IMR);
+		iflags &= (ATMCI_SDIOIRQA | ATMCI_SDIOIRQB);
+		atmci_writel(host, ATMCI_CR, ATMCI_CR_SWRST);
+		atmci_writel(host, ATMCI_CR, ATMCI_CR_MCIEN);
+		atmci_writel(host, ATMCI_MR, host->mode_reg);
+		if (host->caps.has_cfg_reg)
+			atmci_writel(host, ATMCI_CFG, host->cfg_reg);
+		atmci_writel(host, ATMCI_IER, iflags);
 		host->need_reset = false;
 	}
-	mci_writel(host, SDCR, slot->sdc_reg);
+	atmci_writel(host, ATMCI_SDCR, slot->sdc_reg);
 
-	iflags = mci_readl(host, IMR);
-	if (iflags)
-		dev_warn(&slot->mmc->class_dev, "WARNING: IMR=0x%08x\n",
+	iflags = atmci_readl(host, ATMCI_IMR);
+	if (iflags & ~(ATMCI_SDIOIRQA | ATMCI_SDIOIRQB))
+		dev_dbg(&slot->mmc->class_dev, "WARNING: IMR=0x%08x\n",
 				iflags);
 
 	if (unlikely(test_and_clear_bit(ATMCI_CARD_NEED_INIT, &slot->flags))) {
 		/* Send init sequence (74 clock cycles) */
-		mci_writel(host, CMDR, MCI_CMDR_SPCMD_INIT);
-		while (!(mci_readl(host, SR) & MCI_CMDRDY))
+		atmci_writel(host, ATMCI_CMDR, ATMCI_CMDR_SPCMD_INIT);
+		while (!(atmci_readl(host, ATMCI_SR) & ATMCI_CMDRDY))
 			cpu_relax();
 	}
+	iflags = 0;
 	data = mrq->data;
 	if (data) {
 		atmci_set_timeout(host, slot, data);
 
 		/* Must set block count/size before sending command */
-		mci_writel(host, BLKR, MCI_BCNT(data->blocks)
-				| MCI_BLKLEN(data->blksz));
+		atmci_writel(host, ATMCI_BLKR, ATMCI_BCNT(data->blocks)
+				| ATMCI_BLKLEN(data->blksz));
 		dev_vdbg(&slot->mmc->class_dev, "BLKR=0x%08x\n",
-			MCI_BCNT(data->blocks) | MCI_BLKLEN(data->blksz));
+			ATMCI_BCNT(data->blocks) | ATMCI_BLKLEN(data->blksz));
+
+		iflags |= host->prepare_data(host, data);
 	}
 
-	iflags = MCI_CMDRDY;
+	iflags |= ATMCI_CMDRDY;
 	cmd = mrq->cmd;
 	cmdflags = atmci_prepare_command(slot->mmc, cmd);
-	atmci_start_command(host, cmd, cmdflags);
+
+	/*
+	 * DMA transfer should be started before sending the command to avoid
+	 * unexpected errors especially for read operations in SDIO mode.
+	 * Unfortunately, in PDC mode, command has to be sent before starting
+	 * the transfer.
+	 */
+	if (host->submit_data != &atmci_submit_data_dma)
+		atmci_send_command(host, cmd, cmdflags);
 
 	if (data)
-		iflags |= atmci_submit_data(host, data);
+		host->submit_data(host, data);
+
+	if (host->submit_data == &atmci_submit_data_dma)
+		atmci_send_command(host, cmd, cmdflags);
 
 	if (mrq->stop) {
 		host->stop_cmdr = atmci_prepare_command(slot->mmc, mrq->stop);
-		host->stop_cmdr |= MCI_CMDR_STOP_XFER;
+		host->stop_cmdr |= ATMCI_CMDR_STOP_XFER;
 		if (!(data->flags & MMC_DATA_WRITE))
-			host->stop_cmdr |= MCI_CMDR_TRDIR_READ;
+			host->stop_cmdr |= ATMCI_CMDR_TRDIR_READ;
 		if (data->flags & MMC_DATA_STREAM)
-			host->stop_cmdr |= MCI_CMDR_STREAM;
+			host->stop_cmdr |= ATMCI_CMDR_STREAM;
 		else
-			host->stop_cmdr |= MCI_CMDR_MULTI_BLOCK;
+			host->stop_cmdr |= ATMCI_CMDR_MULTI_BLOCK;
 	}
 
 	/*
@@ -748,7 +1219,9 @@ static void atmci_start_request(struct atmel_mci *host,
 	 * conditions (e.g. command and data complete, but stop not
 	 * prepared yet.)
 	 */
-	mci_writel(host, IER, iflags);
+	atmci_writel(host, ATMCI_IER, iflags);
+
+	mod_timer(&host->timer, jiffies +  msecs_to_jiffies(2000));
 }
 
 static void atmci_queue_request(struct atmel_mci *host,
@@ -763,6 +1236,7 @@ static void atmci_queue_request(struct atmel_mci *host,
 		host->state = STATE_SENDING_CMD;
 		atmci_start_request(host, slot);
 	} else {
+		dev_dbg(&host->pdev->dev, "queue request\n");
 		list_add_tail(&slot->queue_node, &host->queue);
 	}
 	spin_unlock_bh(&host->lock);
@@ -775,6 +1249,7 @@ static void atmci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	struct mmc_data		*data;
 
 	WARN_ON(slot->mrq);
+	dev_dbg(&host->pdev->dev, "MRQ: cmd %u\n", mrq->cmd->opcode);
 
 	/*
 	 * We may "know" the card is gone even though there's still an
@@ -805,14 +1280,15 @@ static void atmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	struct atmel_mci_slot	*slot = mmc_priv(mmc);
 	struct atmel_mci	*host = slot->host;
 	unsigned int		i;
+	bool			unprepare_clk;
 
-	slot->sdc_reg &= ~MCI_SDCBUS_MASK;
+	slot->sdc_reg &= ~ATMCI_SDCBUS_MASK;
 	switch (ios->bus_width) {
 	case MMC_BUS_WIDTH_1:
-		slot->sdc_reg |= MCI_SDCBUS_1BIT;
+		slot->sdc_reg |= ATMCI_SDCBUS_1BIT;
 		break;
 	case MMC_BUS_WIDTH_4:
-		slot->sdc_reg |= MCI_SDCBUS_4BIT;
+		slot->sdc_reg |= ATMCI_SDCBUS_4BIT;
 		break;
 	}
 
@@ -820,11 +1296,17 @@ static void atmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		unsigned int clock_min = ~0U;
 		u32 clkdiv;
 
+		clk_prepare(host->mck);
+		unprepare_clk = true;
+
 		spin_lock_bh(&host->lock);
 		if (!host->mode_reg) {
 			clk_enable(host->mck);
-			mci_writel(host, CR, MCI_CR_SWRST);
-			mci_writel(host, CR, MCI_CR_MCIEN);
+			unprepare_clk = false;
+			atmci_writel(host, ATMCI_CR, ATMCI_CR_SWRST);
+			atmci_writel(host, ATMCI_CR, ATMCI_CR_MCIEN);
+			if (host->caps.has_cfg_reg)
+				atmci_writel(host, ATMCI_CFG, host->cfg_reg);
 		}
 
 		/*
@@ -832,19 +1314,32 @@ static void atmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		 * core ios update when finding the minimum.
 		 */
 		slot->clock = ios->clock;
-		for (i = 0; i < ATMEL_MCI_MAX_NR_SLOTS; i++) {
+		for (i = 0; i < ATMCI_MAX_NR_SLOTS; i++) {
 			if (host->slot[i] && host->slot[i]->clock
 					&& host->slot[i]->clock < clock_min)
 				clock_min = host->slot[i]->clock;
 		}
 
 		/* Calculate clock divider */
-		clkdiv = DIV_ROUND_UP(host->bus_hz, 2 * clock_min) - 1;
-		if (clkdiv > 255) {
-			dev_warn(&mmc->class_dev,
-				"clock %u too slow; using %lu\n",
-				clock_min, host->bus_hz / (2 * 256));
-			clkdiv = 255;
+		if (host->caps.has_odd_clk_div) {
+			clkdiv = DIV_ROUND_UP(host->bus_hz, clock_min) - 2;
+			if (clkdiv > 511) {
+				dev_warn(&mmc->class_dev,
+				         "clock %u too slow; using %lu\n",
+				         clock_min, host->bus_hz / (511 + 2));
+				clkdiv = 511;
+			}
+			host->mode_reg = ATMCI_MR_CLKDIV(clkdiv >> 1)
+			                 | ATMCI_MR_CLKODD(clkdiv & 1);
+		} else {
+			clkdiv = DIV_ROUND_UP(host->bus_hz, 2 * clock_min) - 1;
+			if (clkdiv > 255) {
+				dev_warn(&mmc->class_dev,
+				         "clock %u too slow; using %lu\n",
+				         clock_min, host->bus_hz / (2 * 256));
+				clkdiv = 255;
+			}
+			host->mode_reg = ATMCI_MR_CLKDIV(clkdiv);
 		}
 
 		/*
@@ -852,40 +1347,63 @@ static void atmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		 * stopping the clock when the FIFO is full/empty.
 		 * This state is not expected to last for long.
 		 */
-		host->mode_reg = MCI_MR_CLKDIV(clkdiv) | MCI_MR_WRPROOF
-					| MCI_MR_RDPROOF;
+		if (host->caps.has_rwproof)
+			host->mode_reg |= (ATMCI_MR_WRPROOF | ATMCI_MR_RDPROOF);
 
-		if (list_empty(&host->queue))
-			mci_writel(host, MR, host->mode_reg);
-		else
+		if (host->caps.has_cfg_reg) {
+			/* setup High Speed mode in relation with card capacity */
+			if (ios->timing == MMC_TIMING_SD_HS)
+				host->cfg_reg |= ATMCI_CFG_HSMODE;
+			else
+				host->cfg_reg &= ~ATMCI_CFG_HSMODE;
+		}
+
+		if (list_empty(&host->queue)) {
+			atmci_writel(host, ATMCI_MR, host->mode_reg);
+			if (host->caps.has_cfg_reg)
+				atmci_writel(host, ATMCI_CFG, host->cfg_reg);
+		} else {
 			host->need_clock_update = true;
+		}
 
 		spin_unlock_bh(&host->lock);
 	} else {
 		bool any_slot_active = false;
 
+		unprepare_clk = false;
+
 		spin_lock_bh(&host->lock);
 		slot->clock = 0;
-		for (i = 0; i < ATMEL_MCI_MAX_NR_SLOTS; i++) {
+		for (i = 0; i < ATMCI_MAX_NR_SLOTS; i++) {
 			if (host->slot[i] && host->slot[i]->clock) {
 				any_slot_active = true;
 				break;
 			}
 		}
 		if (!any_slot_active) {
-			mci_writel(host, CR, MCI_CR_MCIDIS);
+			atmci_writel(host, ATMCI_CR, ATMCI_CR_MCIDIS);
 			if (host->mode_reg) {
-				mci_readl(host, MR);
+				atmci_readl(host, ATMCI_MR);
 				clk_disable(host->mck);
+				unprepare_clk = true;
 			}
 			host->mode_reg = 0;
 		}
 		spin_unlock_bh(&host->lock);
 	}
 
+	if (unprepare_clk)
+		clk_unprepare(host->mck);
+
 	switch (ios->power_mode) {
+	case MMC_POWER_OFF:
+		if (!IS_ERR(mmc->supply.vmmc))
+			mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, 0);
+		break;
 	case MMC_POWER_UP:
 		set_bit(ATMCI_CARD_NEED_INIT, &slot->flags);
+		if (!IS_ERR(mmc->supply.vmmc))
+			mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, ios->vdd);
 		break;
 	default:
 		/*
@@ -924,7 +1442,8 @@ static int atmci_get_cd(struct mmc_host *mmc)
 	struct atmel_mci_slot	*slot = mmc_priv(mmc);
 
 	if (gpio_is_valid(slot->detect_pin)) {
-		present = !gpio_get_value(slot->detect_pin);
+		present = !(gpio_get_value(slot->detect_pin) ^
+			    slot->detect_is_active_high);
 		dev_dbg(&mmc->class_dev, "card is %spresent\n",
 				present ? "" : "not ");
 	}
@@ -932,11 +1451,23 @@ static int atmci_get_cd(struct mmc_host *mmc)
 	return present;
 }
 
+static void atmci_enable_sdio_irq(struct mmc_host *mmc, int enable)
+{
+	struct atmel_mci_slot	*slot = mmc_priv(mmc);
+	struct atmel_mci	*host = slot->host;
+
+	if (enable)
+		atmci_writel(host, ATMCI_IER, slot->sdio_irq);
+	else
+		atmci_writel(host, ATMCI_IDR, slot->sdio_irq);
+}
+
 static const struct mmc_host_ops atmci_ops = {
 	.request	= atmci_request,
 	.set_ios	= atmci_set_ios,
 	.get_ro		= atmci_get_ro,
 	.get_cd		= atmci_get_cd,
+	.enable_sdio_irq = atmci_enable_sdio_irq,
 };
 
 /* Called with host->lock held */
@@ -952,10 +1483,13 @@ static void atmci_request_end(struct atmel_mci *host, struct mmc_request *mrq)
 	/*
 	 * Update the MMC clock rate if necessary. This may be
 	 * necessary if set_ios() is called when a different slot is
-	 * busy transfering data.
+	 * busy transferring data.
 	 */
-	if (host->need_clock_update)
-		mci_writel(host, MR, host->mode_reg);
+	if (host->need_clock_update) {
+		atmci_writel(host, ATMCI_MR, host->mode_reg);
+		if (host->caps.has_cfg_reg)
+			atmci_writel(host, ATMCI_CFG, host->cfg_reg);
+	}
 
 	host->cur_slot->mrq = NULL;
 	host->mrq = NULL;
@@ -972,6 +1506,8 @@ static void atmci_request_end(struct atmel_mci *host, struct mmc_request *mrq)
 		host->state = STATE_IDLE;
 	}
 
+	del_timer(&host->timer);
+
 	spin_unlock(&host->lock);
 	mmc_request_done(prev_mmc, mrq);
 	spin_lock(&host->lock);
@@ -983,32 +1519,24 @@ static void atmci_command_complete(struct atmel_mci *host,
 	u32		status = host->cmd_status;
 
 	/* Read the response from the card (up to 16 bytes) */
-	cmd->resp[0] = mci_readl(host, RSPR);
-	cmd->resp[1] = mci_readl(host, RSPR);
-	cmd->resp[2] = mci_readl(host, RSPR);
-	cmd->resp[3] = mci_readl(host, RSPR);
+	cmd->resp[0] = atmci_readl(host, ATMCI_RSPR);
+	cmd->resp[1] = atmci_readl(host, ATMCI_RSPR);
+	cmd->resp[2] = atmci_readl(host, ATMCI_RSPR);
+	cmd->resp[3] = atmci_readl(host, ATMCI_RSPR);
 
-	if (status & MCI_RTOE)
+	if (status & ATMCI_RTOE)
 		cmd->error = -ETIMEDOUT;
-	else if ((cmd->flags & MMC_RSP_CRC) && (status & MCI_RCRCE))
+	else if ((cmd->flags & MMC_RSP_CRC) && (status & ATMCI_RCRCE))
 		cmd->error = -EILSEQ;
-	else if (status & (MCI_RINDE | MCI_RDIRE | MCI_RENDE))
+	else if (status & (ATMCI_RINDE | ATMCI_RDIRE | ATMCI_RENDE))
 		cmd->error = -EIO;
-	else
-		cmd->error = 0;
-
-	if (cmd->error) {
-		dev_dbg(&host->pdev->dev,
-			"command error: status=0x%08x\n", status);
-
-		if (cmd->data) {
-			host->data = NULL;
-			atmci_stop_dma(host);
-			mci_writel(host, IDR, MCI_NOTBUSY
-					| MCI_TXRDY | MCI_RXRDY
-					| ATMCI_DATA_ERROR_FLAGS);
+	else if (host->mrq->data && (host->mrq->data->blksz & 3)) {
+		if (host->caps.need_blksz_mul_4) {
+			cmd->error = -EINVAL;
+			host->need_reset = 1;
 		}
-	}
+	} else
+		cmd->error = 0;
 }
 
 static void atmci_detect_change(unsigned long data)
@@ -1028,7 +1556,8 @@ static void atmci_detect_change(unsigned long data)
 		return;
 
 	enable_irq(gpio_to_irq(slot->detect_pin));
-	present = !gpio_get_value(slot->detect_pin);
+	present = !(gpio_get_value(slot->detect_pin) ^
+		    slot->detect_is_active_high);
 	present_old = test_bit(ATMCI_CARD_PRESENT, &slot->flags);
 
 	dev_vdbg(&slot->mmc->class_dev, "detect change: %d (was %d)\n",
@@ -1056,9 +1585,11 @@ static void atmci_detect_change(unsigned long data)
 				 * Reset controller to terminate any ongoing
 				 * commands or data transfers.
 				 */
-				mci_writel(host, CR, MCI_CR_SWRST);
-				mci_writel(host, CR, MCI_CR_MCIEN);
-				mci_writel(host, MR, host->mode_reg);
+				atmci_writel(host, ATMCI_CR, ATMCI_CR_SWRST);
+				atmci_writel(host, ATMCI_CR, ATMCI_CR_MCIEN);
+				atmci_writel(host, ATMCI_MR, host->mode_reg);
+				if (host->caps.has_cfg_reg)
+					atmci_writel(host, ATMCI_CFG, host->cfg_reg);
 
 				host->data = NULL;
 				host->cmd = NULL;
@@ -1068,23 +1599,21 @@ static void atmci_detect_change(unsigned long data)
 					break;
 				case STATE_SENDING_CMD:
 					mrq->cmd->error = -ENOMEDIUM;
-					if (!mrq->data)
-						break;
-					/* fall through */
-				case STATE_SENDING_DATA:
+					if (mrq->data)
+						host->stop_transfer(host);
+					break;
+				case STATE_DATA_XFER:
+					mrq->data->error = -ENOMEDIUM;
+					host->stop_transfer(host);
+					break;
+				case STATE_WAITING_NOTBUSY:
 					mrq->data->error = -ENOMEDIUM;
-					atmci_stop_dma(host);
 					break;
-				case STATE_DATA_BUSY:
-				case STATE_DATA_ERROR:
-					if (mrq->data->error == -EINPROGRESS)
-						mrq->data->error = -ENOMEDIUM;
-					if (!mrq->stop)
-						break;
-					/* fall through */
 				case STATE_SENDING_STOP:
 					mrq->stop->error = -ENOMEDIUM;
 					break;
+				case STATE_END_REQUEST:
+					break;
 				}
 
 				atmci_request_end(host, mrq);
@@ -1112,7 +1641,6 @@ static void atmci_tasklet_func(unsigned long priv)
 	struct atmel_mci	*host = (struct atmel_mci *)priv;
 	struct mmc_request	*mrq = host->mrq;
 	struct mmc_data		*data = host->data;
-	struct mmc_command	*cmd = host->cmd;
 	enum atmel_mci_state	state = host->state;
 	enum atmel_mci_state	prev_state;
 	u32			status;
@@ -1124,110 +1652,193 @@ static void atmci_tasklet_func(unsigned long priv)
 	dev_vdbg(&host->pdev->dev,
 		"tasklet: state %u pending/completed/mask %lx/%lx/%x\n",
 		state, host->pending_events, host->completed_events,
-		mci_readl(host, IMR));
+		atmci_readl(host, ATMCI_IMR));
 
 	do {
 		prev_state = state;
+		dev_dbg(&host->pdev->dev, "FSM: state=%d\n", state);
 
 		switch (state) {
 		case STATE_IDLE:
 			break;
 
 		case STATE_SENDING_CMD:
+			/*
+			 * Command has been sent, we are waiting for command
+			 * ready. Then we have three next states possible:
+			 * END_REQUEST by default, WAITING_NOTBUSY if it's a
+			 * command needing it or DATA_XFER if there is data.
+			 */
+			dev_dbg(&host->pdev->dev, "FSM: cmd ready?\n");
 			if (!atmci_test_and_clear_pending(host,
-						EVENT_CMD_COMPLETE))
+						EVENT_CMD_RDY))
 				break;
 
+			dev_dbg(&host->pdev->dev, "set completed cmd ready\n");
 			host->cmd = NULL;
-			atmci_set_completed(host, EVENT_CMD_COMPLETE);
+			atmci_set_completed(host, EVENT_CMD_RDY);
 			atmci_command_complete(host, mrq->cmd);
-			if (!mrq->data || cmd->error) {
-				atmci_request_end(host, host->mrq);
-				goto unlock;
-			}
+			if (mrq->data) {
+				dev_dbg(&host->pdev->dev,
+				        "command with data transfer");
+				/*
+				 * If there is a command error don't start
+				 * data transfer.
+				 */
+				if (mrq->cmd->error) {
+					host->stop_transfer(host);
+					host->data = NULL;
+					atmci_writel(host, ATMCI_IDR,
+					             ATMCI_TXRDY | ATMCI_RXRDY
+					             | ATMCI_DATA_ERROR_FLAGS);
+					state = STATE_END_REQUEST;
+				} else
+					state = STATE_DATA_XFER;
+			} else if ((!mrq->data) && (mrq->cmd->flags & MMC_RSP_BUSY)) {
+				dev_dbg(&host->pdev->dev,
+				        "command response need waiting notbusy");
+				atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
+				state = STATE_WAITING_NOTBUSY;
+			} else
+				state = STATE_END_REQUEST;
 
-			prev_state = state = STATE_SENDING_DATA;
-			/* fall through */
+			break;
 
-		case STATE_SENDING_DATA:
+		case STATE_DATA_XFER:
 			if (atmci_test_and_clear_pending(host,
 						EVENT_DATA_ERROR)) {
-				atmci_stop_dma(host);
-				if (data->stop)
-					send_stop_cmd(host, data);
-				state = STATE_DATA_ERROR;
+				dev_dbg(&host->pdev->dev, "set completed data error\n");
+				atmci_set_completed(host, EVENT_DATA_ERROR);
+				state = STATE_END_REQUEST;
 				break;
 			}
 
+			/*
+			 * A data transfer is in progress. The event expected
+			 * to move to the next state depends of data transfer
+			 * type (PDC or DMA). Once transfer done we can move
+			 * to the next step which is WAITING_NOTBUSY in write
+			 * case and directly SENDING_STOP in read case.
+			 */
+			dev_dbg(&host->pdev->dev, "FSM: xfer complete?\n");
 			if (!atmci_test_and_clear_pending(host,
 						EVENT_XFER_COMPLETE))
 				break;
 
+			dev_dbg(&host->pdev->dev,
+			        "(%s) set completed xfer complete\n",
+				__func__);
 			atmci_set_completed(host, EVENT_XFER_COMPLETE);
-			prev_state = state = STATE_DATA_BUSY;
-			/* fall through */
-
-		case STATE_DATA_BUSY:
-			if (!atmci_test_and_clear_pending(host,
-						EVENT_DATA_COMPLETE))
-				break;
 
-			host->data = NULL;
-			atmci_set_completed(host, EVENT_DATA_COMPLETE);
-			status = host->data_status;
-			if (unlikely(status & ATMCI_DATA_ERROR_FLAGS)) {
-				if (status & MCI_DTOE) {
-					dev_dbg(&host->pdev->dev,
-							"data timeout error\n");
-					data->error = -ETIMEDOUT;
-				} else if (status & MCI_DCRCE) {
-					dev_dbg(&host->pdev->dev,
-							"data CRC error\n");
-					data->error = -EILSEQ;
-				} else {
-					dev_dbg(&host->pdev->dev,
-						"data FIFO error (status=%08x)\n",
-						status);
-					data->error = -EIO;
-				}
+			if (host->caps.need_notbusy_for_read_ops ||
+			   (host->data->flags & MMC_DATA_WRITE)) {
+				atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
+				state = STATE_WAITING_NOTBUSY;
+			} else if (host->mrq->stop) {
+				atmci_writel(host, ATMCI_IER, ATMCI_CMDRDY);
+				atmci_send_stop_cmd(host, data);
+				state = STATE_SENDING_STOP;
 			} else {
+				host->data = NULL;
 				data->bytes_xfered = data->blocks * data->blksz;
 				data->error = 0;
+				state = STATE_END_REQUEST;
 			}
+			break;
 
-			if (!data->stop) {
-				atmci_request_end(host, host->mrq);
-				goto unlock;
-			}
+		case STATE_WAITING_NOTBUSY:
+			/*
+			 * We can be in the state for two reasons: a command
+			 * requiring waiting not busy signal (stop command
+			 * included) or a write operation. In the latest case,
+			 * we need to send a stop command.
+			 */
+			dev_dbg(&host->pdev->dev, "FSM: not busy?\n");
+			if (!atmci_test_and_clear_pending(host,
+						EVENT_NOTBUSY))
+				break;
 
-			prev_state = state = STATE_SENDING_STOP;
-			if (!data->error)
-				send_stop_cmd(host, data);
-			/* fall through */
+			dev_dbg(&host->pdev->dev, "set completed not busy\n");
+			atmci_set_completed(host, EVENT_NOTBUSY);
+
+			if (host->data) {
+				/*
+				 * For some commands such as CMD53, even if
+				 * there is data transfer, there is no stop
+				 * command to send.
+				 */
+				if (host->mrq->stop) {
+					atmci_writel(host, ATMCI_IER,
+					             ATMCI_CMDRDY);
+					atmci_send_stop_cmd(host, data);
+					state = STATE_SENDING_STOP;
+				} else {
+					host->data = NULL;
+					data->bytes_xfered = data->blocks
+					                     * data->blksz;
+					data->error = 0;
+					state = STATE_END_REQUEST;
+				}
+			} else
+				state = STATE_END_REQUEST;
+			break;
 
 		case STATE_SENDING_STOP:
+			/*
+			 * In this state, it is important to set host->data to
+			 * NULL (which is tested in the waiting notbusy state)
+			 * in order to go to the end request state instead of
+			 * sending stop again.
+			 */
+			dev_dbg(&host->pdev->dev, "FSM: cmd ready?\n");
 			if (!atmci_test_and_clear_pending(host,
-						EVENT_CMD_COMPLETE))
+						EVENT_CMD_RDY))
 				break;
 
+			dev_dbg(&host->pdev->dev, "FSM: cmd ready\n");
 			host->cmd = NULL;
+			data->bytes_xfered = data->blocks * data->blksz;
+			data->error = 0;
 			atmci_command_complete(host, mrq->stop);
-			atmci_request_end(host, host->mrq);
-			goto unlock;
+			if (mrq->stop->error) {
+				host->stop_transfer(host);
+				atmci_writel(host, ATMCI_IDR,
+				             ATMCI_TXRDY | ATMCI_RXRDY
+				             | ATMCI_DATA_ERROR_FLAGS);
+				state = STATE_END_REQUEST;
+			} else {
+				atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
+				state = STATE_WAITING_NOTBUSY;
+			}
+			host->data = NULL;
+			break;
 
-		case STATE_DATA_ERROR:
-			if (!atmci_test_and_clear_pending(host,
-						EVENT_XFER_COMPLETE))
-				break;
+		case STATE_END_REQUEST:
+			atmci_writel(host, ATMCI_IDR, ATMCI_TXRDY | ATMCI_RXRDY
+			                   | ATMCI_DATA_ERROR_FLAGS);
+			status = host->data_status;
+			if (unlikely(status)) {
+				host->stop_transfer(host);
+				host->data = NULL;
+				if (data) {
+					if (status & ATMCI_DTOE) {
+						data->error = -ETIMEDOUT;
+					} else if (status & ATMCI_DCRCE) {
+						data->error = -EILSEQ;
+					} else {
+						data->error = -EIO;
+					}
+				}
+			}
 
-			state = STATE_DATA_BUSY;
+			atmci_request_end(host, host->mrq);
+			state = STATE_IDLE;
 			break;
 		}
 	} while (state != prev_state);
 
 	host->state = state;
 
-unlock:
 	spin_unlock(&host->lock);
 }
 
@@ -1242,7 +1853,7 @@ static void atmci_read_data_pio(struct atmel_mci *host)
 	unsigned int		nbytes = 0;
 
 	do {
-		value = mci_readl(host, RDR);
+		value = atmci_readl(host, ATMCI_RDR);
 		if (likely(offset + 4 <= sg->length)) {
 			put_unaligned(value, (u32 *)(buf + offset));
 
@@ -1252,7 +1863,8 @@ static void atmci_read_data_pio(struct atmel_mci *host)
 			if (offset == sg->length) {
 				flush_dcache_page(sg_page(sg));
 				host->sg = sg = sg_next(sg);
-				if (!sg)
+				host->sg_len--;
+				if (!sg || !host->sg_len)
 					goto done;
 
 				offset = 0;
@@ -1265,7 +1877,8 @@ static void atmci_read_data_pio(struct atmel_mci *host)
 
 			flush_dcache_page(sg_page(sg));
 			host->sg = sg = sg_next(sg);
-			if (!sg)
+			host->sg_len--;
+			if (!sg || !host->sg_len)
 				goto done;
 
 			offset = 4 - remaining;
@@ -1274,18 +1887,15 @@ static void atmci_read_data_pio(struct atmel_mci *host)
 			nbytes += offset;
 		}
 
-		status = mci_readl(host, SR);
+		status = atmci_readl(host, ATMCI_SR);
 		if (status & ATMCI_DATA_ERROR_FLAGS) {
-			mci_writel(host, IDR, (MCI_NOTBUSY | MCI_RXRDY
+			atmci_writel(host, ATMCI_IDR, (ATMCI_NOTBUSY | ATMCI_RXRDY
 						| ATMCI_DATA_ERROR_FLAGS));
 			host->data_status = status;
 			data->bytes_xfered += nbytes;
-			smp_wmb();
-			atmci_set_pending(host, EVENT_DATA_ERROR);
-			tasklet_schedule(&host->tasklet);
 			return;
 		}
-	} while (status & MCI_RXRDY);
+	} while (status & ATMCI_RXRDY);
 
 	host->pio_offset = offset;
 	data->bytes_xfered += nbytes;
@@ -1293,8 +1903,8 @@ static void atmci_read_data_pio(struct atmel_mci *host)
 	return;
 
 done:
-	mci_writel(host, IDR, MCI_RXRDY);
-	mci_writel(host, IER, MCI_NOTBUSY);
+	atmci_writel(host, ATMCI_IDR, ATMCI_RXRDY);
+	atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
 	data->bytes_xfered += nbytes;
 	smp_wmb();
 	atmci_set_pending(host, EVENT_XFER_COMPLETE);
@@ -1313,13 +1923,14 @@ static void atmci_write_data_pio(struct atmel_mci *host)
 	do {
 		if (likely(offset + 4 <= sg->length)) {
 			value = get_unaligned((u32 *)(buf + offset));
-			mci_writel(host, TDR, value);
+			atmci_writel(host, ATMCI_TDR, value);
 
 			offset += 4;
 			nbytes += 4;
 			if (offset == sg->length) {
 				host->sg = sg = sg_next(sg);
-				if (!sg)
+				host->sg_len--;
+				if (!sg || !host->sg_len)
 					goto done;
 
 				offset = 0;
@@ -1333,30 +1944,28 @@ static void atmci_write_data_pio(struct atmel_mci *host)
 			nbytes += remaining;
 
 			host->sg = sg = sg_next(sg);
-			if (!sg) {
-				mci_writel(host, TDR, value);
+			host->sg_len--;
+			if (!sg || !host->sg_len) {
+				atmci_writel(host, ATMCI_TDR, value);
 				goto done;
 			}
 
 			offset = 4 - remaining;
 			buf = sg_virt(sg);
 			memcpy((u8 *)&value + remaining, buf, offset);
-			mci_writel(host, TDR, value);
+			atmci_writel(host, ATMCI_TDR, value);
 			nbytes += offset;
 		}
 
-		status = mci_readl(host, SR);
+		status = atmci_readl(host, ATMCI_SR);
 		if (status & ATMCI_DATA_ERROR_FLAGS) {
-			mci_writel(host, IDR, (MCI_NOTBUSY | MCI_TXRDY
+			atmci_writel(host, ATMCI_IDR, (ATMCI_NOTBUSY | ATMCI_TXRDY
 						| ATMCI_DATA_ERROR_FLAGS));
 			host->data_status = status;
 			data->bytes_xfered += nbytes;
-			smp_wmb();
-			atmci_set_pending(host, EVENT_DATA_ERROR);
-			tasklet_schedule(&host->tasklet);
 			return;
 		}
-	} while (status & MCI_TXRDY);
+	} while (status & ATMCI_TXRDY);
 
 	host->pio_offset = offset;
 	data->bytes_xfered += nbytes;
@@ -1364,23 +1973,26 @@ static void atmci_write_data_pio(struct atmel_mci *host)
 	return;
 
 done:
-	mci_writel(host, IDR, MCI_TXRDY);
-	mci_writel(host, IER, MCI_NOTBUSY);
+	atmci_writel(host, ATMCI_IDR, ATMCI_TXRDY);
+	atmci_writel(host, ATMCI_IER, ATMCI_NOTBUSY);
 	data->bytes_xfered += nbytes;
 	smp_wmb();
 	atmci_set_pending(host, EVENT_XFER_COMPLETE);
 }
 
-static void atmci_cmd_interrupt(struct atmel_mci *host, u32 status)
+static void atmci_sdio_interrupt(struct atmel_mci *host, u32 status)
 {
-	mci_writel(host, IDR, MCI_CMDRDY);
+	int	i;
 
-	host->cmd_status = status;
-	smp_wmb();
-	atmci_set_pending(host, EVENT_CMD_COMPLETE);
-	tasklet_schedule(&host->tasklet);
+	for (i = 0; i < ATMCI_MAX_NR_SLOTS; i++) {
+		struct atmel_mci_slot *slot = host->slot[i];
+		if (slot && (status & slot->sdio_irq)) {
+			mmc_signal_sdio_irq(slot->mmc);
+		}
+	}
 }
 
+
 static irqreturn_t atmci_interrupt(int irq, void *dev_id)
 {
 	struct atmel_mci	*host = dev_id;
@@ -1388,38 +2000,122 @@ static irqreturn_t atmci_interrupt(int irq, void *dev_id)
 	unsigned int		pass_count = 0;
 
 	do {
-		status = mci_readl(host, SR);
-		mask = mci_readl(host, IMR);
+		status = atmci_readl(host, ATMCI_SR);
+		mask = atmci_readl(host, ATMCI_IMR);
 		pending = status & mask;
 		if (!pending)
 			break;
 
 		if (pending & ATMCI_DATA_ERROR_FLAGS) {
-			mci_writel(host, IDR, ATMCI_DATA_ERROR_FLAGS
-					| MCI_RXRDY | MCI_TXRDY);
-			pending &= mci_readl(host, IMR);
+			dev_dbg(&host->pdev->dev, "IRQ: data error\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_DATA_ERROR_FLAGS
+					| ATMCI_RXRDY | ATMCI_TXRDY
+					| ATMCI_ENDRX | ATMCI_ENDTX
+					| ATMCI_RXBUFF | ATMCI_TXBUFE);
 
 			host->data_status = status;
+			dev_dbg(&host->pdev->dev, "set pending data error\n");
 			smp_wmb();
 			atmci_set_pending(host, EVENT_DATA_ERROR);
 			tasklet_schedule(&host->tasklet);
 		}
-		if (pending & MCI_NOTBUSY) {
-			mci_writel(host, IDR,
-					ATMCI_DATA_ERROR_FLAGS | MCI_NOTBUSY);
-			if (!host->data_status)
-				host->data_status = status;
+
+		if (pending & ATMCI_TXBUFE) {
+			dev_dbg(&host->pdev->dev, "IRQ: tx buffer empty\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_TXBUFE);
+			atmci_writel(host, ATMCI_IDR, ATMCI_ENDTX);
+			/*
+			 * We can receive this interruption before having configured
+			 * the second pdc buffer, so we need to reconfigure first and
+			 * second buffers again
+			 */
+			if (host->data_size) {
+				atmci_pdc_set_both_buf(host, XFER_TRANSMIT);
+				atmci_writel(host, ATMCI_IER, ATMCI_ENDTX);
+				atmci_writel(host, ATMCI_IER, ATMCI_TXBUFE);
+			} else {
+				atmci_pdc_complete(host);
+			}
+		} else if (pending & ATMCI_ENDTX) {
+			dev_dbg(&host->pdev->dev, "IRQ: end of tx buffer\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_ENDTX);
+
+			if (host->data_size) {
+				atmci_pdc_set_single_buf(host,
+						XFER_TRANSMIT, PDC_SECOND_BUF);
+				atmci_writel(host, ATMCI_IER, ATMCI_ENDTX);
+			}
+		}
+
+		if (pending & ATMCI_RXBUFF) {
+			dev_dbg(&host->pdev->dev, "IRQ: rx buffer full\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_RXBUFF);
+			atmci_writel(host, ATMCI_IDR, ATMCI_ENDRX);
+			/*
+			 * We can receive this interruption before having configured
+			 * the second pdc buffer, so we need to reconfigure first and
+			 * second buffers again
+			 */
+			if (host->data_size) {
+				atmci_pdc_set_both_buf(host, XFER_RECEIVE);
+				atmci_writel(host, ATMCI_IER, ATMCI_ENDRX);
+				atmci_writel(host, ATMCI_IER, ATMCI_RXBUFF);
+			} else {
+				atmci_pdc_complete(host);
+			}
+		} else if (pending & ATMCI_ENDRX) {
+			dev_dbg(&host->pdev->dev, "IRQ: end of rx buffer\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_ENDRX);
+
+			if (host->data_size) {
+				atmci_pdc_set_single_buf(host,
+						XFER_RECEIVE, PDC_SECOND_BUF);
+				atmci_writel(host, ATMCI_IER, ATMCI_ENDRX);
+			}
+		}
+
+		/*
+		 * First mci IPs, so mainly the ones having pdc, have some
+		 * issues with the notbusy signal. You can't get it after
+		 * data transmission if you have not sent a stop command.
+		 * The appropriate workaround is to use the BLKE signal.
+		 */
+		if (pending & ATMCI_BLKE) {
+			dev_dbg(&host->pdev->dev, "IRQ: blke\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_BLKE);
+			smp_wmb();
+			dev_dbg(&host->pdev->dev, "set pending notbusy\n");
+			atmci_set_pending(host, EVENT_NOTBUSY);
+			tasklet_schedule(&host->tasklet);
+		}
+
+		if (pending & ATMCI_NOTBUSY) {
+			dev_dbg(&host->pdev->dev, "IRQ: not_busy\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_NOTBUSY);
 			smp_wmb();
-			atmci_set_pending(host, EVENT_DATA_COMPLETE);
+			dev_dbg(&host->pdev->dev, "set pending notbusy\n");
+			atmci_set_pending(host, EVENT_NOTBUSY);
 			tasklet_schedule(&host->tasklet);
 		}
-		if (pending & MCI_RXRDY)
+
+		if (pending & ATMCI_RXRDY)
 			atmci_read_data_pio(host);
-		if (pending & MCI_TXRDY)
+		if (pending & ATMCI_TXRDY)
 			atmci_write_data_pio(host);
 
-		if (pending & MCI_CMDRDY)
-			atmci_cmd_interrupt(host, status);
+		if (pending & ATMCI_CMDRDY) {
+			dev_dbg(&host->pdev->dev, "IRQ: cmd ready\n");
+			atmci_writel(host, ATMCI_IDR, ATMCI_CMDRDY);
+			host->cmd_status = status;
+			smp_wmb();
+			dev_dbg(&host->pdev->dev, "set pending cmd rdy\n");
+			atmci_set_pending(host, EVENT_CMD_RDY);
+			tasklet_schedule(&host->tasklet);
+		}
+
+		if (pending & (ATMCI_SDIOIRQA | ATMCI_SDIOIRQB))
+			atmci_sdio_interrupt(host, status);
+
 	} while (pass_count++ < 5);
 
 	return pass_count ? IRQ_HANDLED : IRQ_NONE;
@@ -1442,7 +2138,7 @@ static irqreturn_t atmci_detect_interrupt(int irq, void *dev_id)
 
 static int __init atmci_init_slot(struct atmel_mci *host,
 		struct mci_slot_pdata *slot_data, unsigned int id,
-		u32 sdc_reg)
+		u32 sdc_reg, u32 sdio_irq)
 {
 	struct mmc_host			*mmc;
 	struct atmel_mci_slot		*slot;
@@ -1456,20 +2152,45 @@ static int __init atmci_init_slot(struct atmel_mci *host,
 	slot->host = host;
 	slot->detect_pin = slot_data->detect_pin;
 	slot->wp_pin = slot_data->wp_pin;
+	slot->detect_is_active_high = slot_data->detect_is_active_high;
 	slot->sdc_reg = sdc_reg;
+	slot->sdio_irq = sdio_irq;
+
+	dev_dbg(&mmc->class_dev,
+	        "slot[%u]: bus_width=%u, detect_pin=%d, "
+		"detect_is_active_high=%s, wp_pin=%d\n",
+		id, slot_data->bus_width, slot_data->detect_pin,
+		slot_data->detect_is_active_high ? "true" : "false",
+		slot_data->wp_pin);
 
 	mmc->ops = &atmci_ops;
 	mmc->f_min = DIV_ROUND_UP(host->bus_hz, 512);
 	mmc->f_max = host->bus_hz / 2;
 	mmc->ocr_avail	= MMC_VDD_32_33 | MMC_VDD_33_34;
-	if (slot_data->bus_width >= 4)
+	if (sdio_irq)
+		mmc->caps |= MMC_CAP_SDIO_IRQ;
+	if (host->caps.has_highspeed)
+		mmc->caps |= MMC_CAP_SD_HIGHSPEED;
+	/*
+	 * Without the read/write proof capability, it is strongly suggested to
+	 * use only one bit for data to prevent fifo underruns and overruns
+	 * which will corrupt data.
+	 */
+	if ((slot_data->bus_width >= 4) && host->caps.has_rwproof)
 		mmc->caps |= MMC_CAP_4_BIT_DATA;
 
-	mmc->max_hw_segs = 64;
-	mmc->max_phys_segs = 64;
-	mmc->max_req_size = 32768 * 512;
-	mmc->max_blk_size = 32768;
-	mmc->max_blk_count = 512;
+	if (atmci_get_version(host) < 0x200) {
+		mmc->max_segs = 256;
+		mmc->max_blk_size = 4095;
+		mmc->max_blk_count = 256;
+		mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count;
+		mmc->max_seg_size = mmc->max_blk_size * mmc->max_segs;
+	} else {
+		mmc->max_segs = 64;
+		mmc->max_req_size = 32768 * 512;
+		mmc->max_blk_size = 32768;
+		mmc->max_blk_count = 512;
+	}
 
 	/* Assume card is present initially */
 	set_bit(ATMCI_CARD_PRESENT, &slot->flags);
@@ -1477,7 +2198,8 @@ static int __init atmci_init_slot(struct atmel_mci *host,
 		if (gpio_request(slot->detect_pin, "mmc_detect")) {
 			dev_dbg(&mmc->class_dev, "no detect pin available\n");
 			slot->detect_pin = -EBUSY;
-		} else if (gpio_get_value(slot->detect_pin)) {
+		} else if (gpio_get_value(slot->detect_pin) ^
+				slot->detect_is_active_high) {
 			clear_bit(ATMCI_CARD_PRESENT, &slot->flags);
 		}
 	}
@@ -1493,6 +2215,7 @@ static int __init atmci_init_slot(struct atmel_mci *host,
 	}
 
 	host->slot[id] = slot;
+	mmc_regulator_get_supply(mmc);
 	mmc_add_host(mmc);
 
 	if (gpio_is_valid(slot->detect_pin)) {
@@ -1543,18 +2266,109 @@ static void __exit atmci_cleanup_slot(struct atmel_mci_slot *slot,
 	mmc_free_host(slot->mmc);
 }
 
-#ifdef CONFIG_MMC_ATMELMCI_DMA
-static bool filter(struct dma_chan *chan, void *slave)
+static bool atmci_filter(struct dma_chan *chan, void *pdata)
 {
-	struct dw_dma_slave *dws = slave;
+	struct mci_platform_data *sl_pdata = pdata;
+	struct mci_dma_data *sl;
+
+	if (!sl_pdata)
+		return false;
 
-	if (dws->dma_dev == chan->device->dev) {
-		chan->private = dws;
+	sl = sl_pdata->dma_slave;
+	if (sl && find_slave_dev(sl) == chan->device->dev) {
+		chan->private = slave_data_ptr(sl);
 		return true;
-	} else
+	} else {
 		return false;
+	}
+}
+
+static bool atmci_configure_dma(struct atmel_mci *host)
+{
+	struct mci_platform_data	*pdata;
+	dma_cap_mask_t mask;
+
+	if (host == NULL)
+		return false;
+
+	pdata = host->pdev->dev.platform_data;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	host->dma.chan = dma_request_slave_channel_compat(mask, atmci_filter, pdata,
+							  &host->pdev->dev, "rxtx");
+	if (!host->dma.chan) {
+		dev_warn(&host->pdev->dev, "no DMA channel available\n");
+		return false;
+	} else {
+		dev_info(&host->pdev->dev,
+					"using %s for DMA transfers\n",
+					dma_chan_name(host->dma.chan));
+
+		host->dma_conf.src_addr = host->mapbase + ATMCI_RDR;
+		host->dma_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		host->dma_conf.src_maxburst = 1;
+		host->dma_conf.dst_addr = host->mapbase + ATMCI_TDR;
+		host->dma_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		host->dma_conf.dst_maxburst = 1;
+		host->dma_conf.device_fc = false;
+		return true;
+	}
+}
+
+/*
+ * HSMCI (High Speed MCI) module is not fully compatible with MCI module.
+ * HSMCI provides DMA support and a new config register but no more supports
+ * PDC.
+ */
+static void __init atmci_get_cap(struct atmel_mci *host)
+{
+	unsigned int version;
+
+	version = atmci_get_version(host);
+	dev_info(&host->pdev->dev,
+			"version: 0x%x\n", version);
+
+	host->caps.has_dma_conf_reg = 0;
+	host->caps.has_pdc = ATMCI_PDC_CONNECTED;
+	host->caps.has_cfg_reg = 0;
+	host->caps.has_cstor_reg = 0;
+	host->caps.has_highspeed = 0;
+	host->caps.has_rwproof = 0;
+	host->caps.has_odd_clk_div = 0;
+	host->caps.has_bad_data_ordering = 1;
+	host->caps.need_reset_after_xfer = 1;
+	host->caps.need_blksz_mul_4 = 1;
+	host->caps.need_notbusy_for_read_ops = 0;
+
+	/* keep only major version number */
+	switch (version & 0xf00) {
+	case 0x500:
+		host->caps.has_odd_clk_div = 1;
+	case 0x400:
+	case 0x300:
+		host->caps.has_dma_conf_reg = 1;
+		host->caps.has_pdc = 0;
+		host->caps.has_cfg_reg = 1;
+		host->caps.has_cstor_reg = 1;
+		host->caps.has_highspeed = 1;
+	case 0x200:
+		host->caps.has_rwproof = 1;
+		host->caps.need_blksz_mul_4 = 0;
+		host->caps.need_notbusy_for_read_ops = 1;
+	case 0x100:
+		host->caps.has_bad_data_ordering = 0;
+		host->caps.need_reset_after_xfer = 0;
+	case 0x0:
+		break;
+	default:
+		host->caps.has_pdc = 0;
+		dev_warn(&host->pdev->dev,
+				"Unmanaged mci version, set minimum capabilities\n");
+		break;
+	}
 }
-#endif
 
 static int __init atmci_probe(struct platform_device *pdev)
 {
@@ -1569,8 +2383,14 @@ static int __init atmci_probe(struct platform_device *pdev)
 	if (!regs)
 		return -ENXIO;
 	pdata = pdev->dev.platform_data;
-	if (!pdata)
-		return -ENXIO;
+	if (!pdata) {
+		pdata = atmci_of_init(pdev);
+		if (IS_ERR(pdata)) {
+			dev_err(&pdev->dev, "platform data not available\n");
+			return PTR_ERR(pdata);
+		}
+	}
+
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0)
 		return irq;
@@ -1590,14 +2410,16 @@ static int __init atmci_probe(struct platform_device *pdev)
 	}
 
 	ret = -ENOMEM;
-	host->regs = ioremap(regs->start, regs->end - regs->start + 1);
+	host->regs = ioremap(regs->start, resource_size(regs));
 	if (!host->regs)
 		goto err_ioremap;
 
-	clk_enable(host->mck);
-	mci_writel(host, CR, MCI_CR_SWRST);
+	ret = clk_prepare_enable(host->mck);
+	if (ret)
+		goto err_request_irq;
+	atmci_writel(host, ATMCI_CR, ATMCI_CR_SWRST);
 	host->bus_hz = clk_get_rate(host->mck);
-	clk_disable(host->mck);
+	clk_disable_unprepare(host->mck);
 
 	host->mapbase = regs->start;
 
@@ -1607,43 +2429,65 @@ static int __init atmci_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_request_irq;
 
-#ifdef CONFIG_MMC_ATMELMCI_DMA
-	if (pdata->dma_slave.dma_dev) {
-		struct dw_dma_slave *dws = &pdata->dma_slave;
-		dma_cap_mask_t mask;
-
-		dws->tx_reg = regs->start + MCI_TDR;
-		dws->rx_reg = regs->start + MCI_RDR;
-
-		/* Try to grab a DMA channel */
-		dma_cap_zero(mask);
-		dma_cap_set(DMA_SLAVE, mask);
-		host->dma.chan = dma_request_channel(mask, filter, dws);
+	/* Get MCI capabilities and set operations according to it */
+	atmci_get_cap(host);
+	if (atmci_configure_dma(host)) {
+		host->prepare_data = &atmci_prepare_data_dma;
+		host->submit_data = &atmci_submit_data_dma;
+		host->stop_transfer = &atmci_stop_transfer_dma;
+	} else if (host->caps.has_pdc) {
+		dev_info(&pdev->dev, "using PDC\n");
+		host->prepare_data = &atmci_prepare_data_pdc;
+		host->submit_data = &atmci_submit_data_pdc;
+		host->stop_transfer = &atmci_stop_transfer_pdc;
+	} else {
+		dev_info(&pdev->dev, "using PIO\n");
+		host->prepare_data = &atmci_prepare_data;
+		host->submit_data = &atmci_submit_data;
+		host->stop_transfer = &atmci_stop_transfer;
 	}
-	if (!host->dma.chan)
-		dev_notice(&pdev->dev, "DMA not available, using PIO\n");
-#endif /* CONFIG_MMC_ATMELMCI_DMA */
 
 	platform_set_drvdata(pdev, host);
 
+	setup_timer(&host->timer, atmci_timeout_timer, (unsigned long)host);
+
 	/* We need at least one slot to succeed */
 	nr_slots = 0;
 	ret = -ENODEV;
 	if (pdata->slot[0].bus_width) {
 		ret = atmci_init_slot(host, &pdata->slot[0],
-				MCI_SDCSEL_SLOT_A, 0);
-		if (!ret)
+				0, ATMCI_SDCSEL_SLOT_A, ATMCI_SDIOIRQA);
+		if (!ret) {
 			nr_slots++;
+			host->buf_size = host->slot[0]->mmc->max_req_size;
+		}
 	}
 	if (pdata->slot[1].bus_width) {
 		ret = atmci_init_slot(host, &pdata->slot[1],
-				MCI_SDCSEL_SLOT_B, 1);
-		if (!ret)
+				1, ATMCI_SDCSEL_SLOT_B, ATMCI_SDIOIRQB);
+		if (!ret) {
 			nr_slots++;
+			if (host->slot[1]->mmc->max_req_size > host->buf_size)
+				host->buf_size =
+					host->slot[1]->mmc->max_req_size;
+		}
 	}
 
-	if (!nr_slots)
+	if (!nr_slots) {
+		dev_err(&pdev->dev, "init failed: no slot defined\n");
 		goto err_init_slot;
+	}
+
+	if (!host->caps.has_rwproof) {
+		host->buffer = dma_alloc_coherent(&pdev->dev, host->buf_size,
+		                                  &host->buf_phys_addr,
+						  GFP_KERNEL);
+		if (!host->buffer) {
+			ret = -ENOMEM;
+			dev_err(&pdev->dev, "buffer allocation failed\n");
+			goto err_init_slot;
+		}
+	}
 
 	dev_info(&pdev->dev,
 			"Atmel MCI controller at 0x%08lx irq %d, %u slots\n",
@@ -1652,10 +2496,8 @@ static int __init atmci_probe(struct platform_device *pdev)
 	return 0;
 
 err_init_slot:
-#ifdef CONFIG_MMC_ATMELMCI_DMA
 	if (host->dma.chan)
 		dma_release_channel(host->dma.chan);
-#endif
 	free_irq(irq, host);
 err_request_irq:
 	iounmap(host->regs);
@@ -1671,23 +2513,23 @@ static int __exit atmci_remove(struct platform_device *pdev)
 	struct atmel_mci	*host = platform_get_drvdata(pdev);
 	unsigned int		i;
 
-	platform_set_drvdata(pdev, NULL);
+	if (host->buffer)
+		dma_free_coherent(&pdev->dev, host->buf_size,
+		                  host->buffer, host->buf_phys_addr);
 
-	for (i = 0; i < ATMEL_MCI_MAX_NR_SLOTS; i++) {
+	for (i = 0; i < ATMCI_MAX_NR_SLOTS; i++) {
 		if (host->slot[i])
 			atmci_cleanup_slot(host->slot[i], i);
 	}
 
-	clk_enable(host->mck);
-	mci_writel(host, IDR, ~0UL);
-	mci_writel(host, CR, MCI_CR_MCIDIS);
-	mci_readl(host, SR);
-	clk_disable(host->mck);
+	clk_prepare_enable(host->mck);
+	atmci_writel(host, ATMCI_IDR, ~0UL);
+	atmci_writel(host, ATMCI_CR, ATMCI_CR_MCIDIS);
+	atmci_readl(host, ATMCI_SR);
+	clk_disable_unprepare(host->mck);
 
-#ifdef CONFIG_MMC_ATMELMCI_DMA
 	if (host->dma.chan)
 		dma_release_channel(host->dma.chan);
-#endif
 
 	free_irq(platform_get_irq(pdev, 0), host);
 	iounmap(host->regs);
@@ -1702,6 +2544,7 @@ static struct platform_driver atmci_driver = {
 	.remove		= __exit_p(atmci_remove),
 	.driver		= {
 		.name		= "atmel_mci",
+		.of_match_table	= of_match_ptr(atmci_dt_ids),
 	},
 };
 
@@ -1719,5 +2562,5 @@ late_initcall(atmci_init); /* try to load after dma driver when built-in */
 module_exit(atmci_exit);
 
 MODULE_DESCRIPTION("Atmel Multimedia Card Interface driver");
-MODULE_AUTHOR("Haavard Skinnemoen <haavard.skinnemoen@atmel.com>");
+MODULE_AUTHOR("Haavard Skinnemoen (Atmel)");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/au1xmmc.c b/drivers/mmc/host/au1xmmc.c
index d3f55615c09..f5443a6c491 100644
--- a/drivers/mmc/host/au1xmmc.c
+++ b/drivers/mmc/host/au1xmmc.c
@@ -41,6 +41,7 @@
 #include <linux/scatterlist.h>
 #include <linux/leds.h>
 #include <linux/mmc/host.h>
+#include <linux/slab.h>
 
 #include <asm/io.h>
 #include <asm/mach-au1x00/au1000.h>
@@ -54,7 +55,7 @@
 
 #ifdef DEBUG
 #define DBG(fmt, idx, args...)	\
-	printk(KERN_DEBUG "au1xmmc(%d): DEBUG: " fmt, idx, ##args)
+	pr_debug("au1xmmc(%d): DEBUG: " fmt, idx, ##args)
 #else
 #define DBG(fmt, idx, args...) do {} while (0)
 #endif
@@ -63,11 +64,8 @@
 #define AU1XMMC_DESCRIPTOR_COUNT 1
 
 /* max DMA seg size: 64KB on Au1100, 4MB on Au1200 */
-#ifdef CONFIG_SOC_AU1100
-#define AU1XMMC_DESCRIPTOR_SIZE 0x0000ffff
-#else	/* Au1200 */
-#define AU1XMMC_DESCRIPTOR_SIZE 0x003fffff
-#endif
+#define AU1100_MMC_DESCRIPTOR_SIZE 0x0000ffff
+#define AU1200_MMC_DESCRIPTOR_SIZE 0x003fffff
 
 #define AU1XMMC_OCR (MMC_VDD_27_28 | MMC_VDD_28_29 | MMC_VDD_29_30 | \
 		     MMC_VDD_30_31 | MMC_VDD_31_32 | MMC_VDD_32_33 | \
@@ -126,6 +124,7 @@ struct au1xmmc_host {
 #define HOST_F_XMIT	0x0001
 #define HOST_F_RECV	0x0002
 #define HOST_F_DMA	0x0010
+#define HOST_F_DBDMA	0x0020
 #define HOST_F_ACTIVE	0x0100
 #define HOST_F_STOP	0x1000
 
@@ -150,6 +149,17 @@ struct au1xmmc_host {
 #define DMA_CHANNEL(h)	\
 	(((h)->flags & HOST_F_XMIT) ? (h)->tx_chan : (h)->rx_chan)
 
+static inline int has_dbdma(void)
+{
+	switch (alchemy_get_cputype()) {
+	case ALCHEMY_CPU_AU1200:
+	case ALCHEMY_CPU_AU1300:
+		return 1;
+	default:
+		return 0;
+	}
+}
+
 static inline void IRQ_ON(struct au1xmmc_host *host, u32 mask)
 {
 	u32 val = au_readl(HOST_CONFIG(host));
@@ -191,7 +201,7 @@ static inline void SEND_STOP(struct au1xmmc_host *host)
 	au_writel(config2 | SD_CONFIG2_DF, HOST_CONFIG2(host));
 	au_sync();
 
-	/* Send the stop commmand */
+	/* Send the stop command */
 	au_writel(STOP_CMD, HOST_CMD(host));
 }
 
@@ -267,7 +277,7 @@ static int au1xmmc_send_command(struct au1xmmc_host *host, int wait,
 		mmccmd |= SD_CMD_RT_3;
 		break;
 	default:
-		printk(KERN_INFO "au1xmmc: unhandled response type %02x\n",
+		pr_info("au1xmmc: unhandled response type %02x\n",
 			mmc_resp_type(cmd));
 		return -EINVAL;
 	}
@@ -352,14 +362,12 @@ static void au1xmmc_data_complete(struct au1xmmc_host *host, u32 status)
 	data->bytes_xfered = 0;
 
 	if (!data->error) {
-		if (host->flags & HOST_F_DMA) {
-#ifdef CONFIG_SOC_AU1200	/* DBDMA */
+		if (host->flags & (HOST_F_DMA | HOST_F_DBDMA)) {
 			u32 chan = DMA_CHANNEL(host);
 
 			chan_tab_t *c = *((chan_tab_t **)chan);
 			au1x_dma_chan_t *cp = c->chan_ptr;
 			data->bytes_xfered = cp->ddma_bytecnt;
-#endif
 		} else
 			data->bytes_xfered =
 				(data->blocks * data->blksz) - host->pio.len;
@@ -569,11 +577,10 @@ static void au1xmmc_cmd_complete(struct au1xmmc_host *host, u32 status)
 
 	host->status = HOST_S_DATA;
 
-	if (host->flags & HOST_F_DMA) {
-#ifdef CONFIG_SOC_AU1200	/* DBDMA */
+	if ((host->flags & (HOST_F_DMA | HOST_F_DBDMA))) {
 		u32 channel = DMA_CHANNEL(host);
 
-		/* Start the DMA as soon as the buffer gets something in it */
+		/* Start the DBDMA as soon as the buffer gets something in it */
 
 		if (host->flags & HOST_F_RECV) {
 			u32 mask = SD_STATUS_DB | SD_STATUS_NE;
@@ -583,7 +590,6 @@ static void au1xmmc_cmd_complete(struct au1xmmc_host *host, u32 status)
 		}
 
 		au1xxx_dbdma_start(channel);
-#endif
 	}
 }
 
@@ -632,8 +638,7 @@ static int au1xmmc_prepare_data(struct au1xmmc_host *host,
 
 	au_writel(data->blksz - 1, HOST_BLKSIZE(host));
 
-	if (host->flags & HOST_F_DMA) {
-#ifdef CONFIG_SOC_AU1200	/* DBDMA */
+	if (host->flags & (HOST_F_DMA | HOST_F_DBDMA)) {
 		int i;
 		u32 channel = DMA_CHANNEL(host);
 
@@ -650,11 +655,11 @@ static int au1xmmc_prepare_data(struct au1xmmc_host *host,
 				flags = DDMA_FLAGS_IE;
 
 			if (host->flags & HOST_F_XMIT) {
-				ret = au1xxx_dbdma_put_source_flags(channel,
-					(void *)sg_virt(sg), len, flags);
+				ret = au1xxx_dbdma_put_source(channel,
+					sg_phys(sg), len, flags);
 			} else {
-				ret = au1xxx_dbdma_put_dest_flags(channel,
-					(void *)sg_virt(sg), len, flags);
+				ret = au1xxx_dbdma_put_dest(channel,
+					sg_phys(sg), len, flags);
 			}
 
 			if (!ret)
@@ -662,7 +667,6 @@ static int au1xmmc_prepare_data(struct au1xmmc_host *host,
 
 			datalen -= len;
 		}
-#endif
 	} else {
 		host->pio.index = 0;
 		host->pio.offset = 0;
@@ -765,11 +769,15 @@ static void au1xmmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 	config2 = au_readl(HOST_CONFIG2(host));
 	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_8:
+		config2 |= SD_CONFIG2_BB;
+		break;
 	case MMC_BUS_WIDTH_4:
+		config2 &= ~SD_CONFIG2_BB;
 		config2 |= SD_CONFIG2_WB;
 		break;
 	case MMC_BUS_WIDTH_1:
-		config2 &= ~SD_CONFIG2_WB;
+		config2 &= ~(SD_CONFIG2_WB | SD_CONFIG2_BB);
 		break;
 	}
 	au_writel(config2, HOST_CONFIG2(host));
@@ -837,7 +845,6 @@ static irqreturn_t au1xmmc_irq(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
-#ifdef CONFIG_SOC_AU1200
 /* 8bit memory DMA device */
 static dbdev_tab_t au1xmmc_mem_dbdev = {
 	.dev_id		= DSCR_CMD0_ALWAYS,
@@ -904,7 +911,7 @@ static int au1xmmc_dbdma_init(struct au1xmmc_host *host)
 	au1xxx_dbdma_ring_alloc(host->rx_chan, AU1XMMC_DESCRIPTOR_COUNT);
 
 	/* DBDMA is good to go */
-	host->flags |= HOST_F_DMA;
+	host->flags |= HOST_F_DMA | HOST_F_DBDMA;
 
 	return 0;
 }
@@ -917,7 +924,6 @@ static void au1xmmc_dbdma_shutdown(struct au1xmmc_host *host)
 		au1xxx_dbdma_chan_free(host->rx_chan);
 	}
 }
-#endif
 
 static void au1xmmc_enable_sdio_irq(struct mmc_host *mmc, int en)
 {
@@ -937,12 +943,12 @@ static const struct mmc_host_ops au1xmmc_ops = {
 	.enable_sdio_irq = au1xmmc_enable_sdio_irq,
 };
 
-static int __devinit au1xmmc_probe(struct platform_device *pdev)
+static int au1xmmc_probe(struct platform_device *pdev)
 {
 	struct mmc_host *mmc;
 	struct au1xmmc_host *host;
 	struct resource *r;
-	int ret;
+	int ret, iflag;
 
 	mmc = mmc_alloc_host(sizeof(struct au1xmmc_host), &pdev->dev);
 	if (!mmc) {
@@ -963,7 +969,7 @@ static int __devinit au1xmmc_probe(struct platform_device *pdev)
 		goto out1;
 	}
 
-	host->ioarea = request_mem_region(r->start, r->end - r->start + 1,
+	host->ioarea = request_mem_region(r->start, resource_size(r),
 					   pdev->name);
 	if (!host->ioarea) {
 		dev_err(&pdev->dev, "mmio already in use\n");
@@ -981,29 +987,43 @@ static int __devinit au1xmmc_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "no IRQ defined\n");
 		goto out3;
 	}
-
 	host->irq = r->start;
-	/* IRQ is shared among both SD controllers */
-	ret = request_irq(host->irq, au1xmmc_irq, IRQF_SHARED,
-			  DRIVER_NAME, host);
-	if (ret) {
-		dev_err(&pdev->dev, "cannot grab IRQ\n");
-		goto out3;
-	}
 
 	mmc->ops = &au1xmmc_ops;
 
 	mmc->f_min =   450000;
 	mmc->f_max = 24000000;
 
-	mmc->max_seg_size = AU1XMMC_DESCRIPTOR_SIZE;
-	mmc->max_phys_segs = AU1XMMC_DESCRIPTOR_COUNT;
-
 	mmc->max_blk_size = 2048;
 	mmc->max_blk_count = 512;
 
 	mmc->ocr_avail = AU1XMMC_OCR;
 	mmc->caps = MMC_CAP_4_BIT_DATA | MMC_CAP_SDIO_IRQ;
+	mmc->max_segs = AU1XMMC_DESCRIPTOR_COUNT;
+
+	iflag = IRQF_SHARED;	/* Au1100/Au1200: one int for both ctrls */
+
+	switch (alchemy_get_cputype()) {
+	case ALCHEMY_CPU_AU1100:
+		mmc->max_seg_size = AU1100_MMC_DESCRIPTOR_SIZE;
+		break;
+	case ALCHEMY_CPU_AU1200:
+		mmc->max_seg_size = AU1200_MMC_DESCRIPTOR_SIZE;
+		break;
+	case ALCHEMY_CPU_AU1300:
+		iflag = 0;	/* nothing is shared */
+		mmc->max_seg_size = AU1200_MMC_DESCRIPTOR_SIZE;
+		mmc->f_max = 52000000;
+		if (host->ioarea->start == AU1100_SD0_PHYS_ADDR)
+			mmc->caps |= MMC_CAP_8_BIT_DATA;
+		break;
+	}
+
+	ret = request_irq(host->irq, au1xmmc_irq, iflag, DRIVER_NAME, host);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot grab IRQ\n");
+		goto out3;
+	}
 
 	host->status = HOST_S_IDLE;
 
@@ -1017,17 +1037,21 @@ static int __devinit au1xmmc_probe(struct platform_device *pdev)
 	} else
 		mmc->caps |= MMC_CAP_NEEDS_POLL;
 
+	/* platform may not be able to use all advertised caps */
+	if (host->platdata)
+		mmc->caps &= ~(host->platdata->mask_host_caps);
+
 	tasklet_init(&host->data_task, au1xmmc_tasklet_data,
 			(unsigned long)host);
 
 	tasklet_init(&host->finish_task, au1xmmc_tasklet_finish,
 			(unsigned long)host);
 
-#ifdef CONFIG_SOC_AU1200
-	ret = au1xmmc_dbdma_init(host);
-	if (ret)
-		printk(KERN_INFO DRIVER_NAME ": DBDMA init failed; using PIO\n");
-#endif
+	if (has_dbdma()) {
+		ret = au1xmmc_dbdma_init(host);
+		if (ret)
+			pr_info(DRIVER_NAME ": DBDMA init failed; using PIO\n");
+	}
 
 #ifdef CONFIG_LEDS_CLASS
 	if (host->platdata && host->platdata->led) {
@@ -1051,7 +1075,7 @@ static int __devinit au1xmmc_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, host);
 
-	printk(KERN_INFO DRIVER_NAME ": MMC Controller %d set up at %8.8X"
+	pr_info(DRIVER_NAME ": MMC Controller %d set up at %8.8X"
 		" (mode=%s)\n", pdev->id, host->iobase,
 		host->flags & HOST_F_DMA ? "dma" : "pio");
 
@@ -1068,9 +1092,8 @@ out5:
 	au_writel(0, HOST_CONFIG2(host));
 	au_sync();
 
-#ifdef CONFIG_SOC_AU1200
-	au1xmmc_dbdma_shutdown(host);
-#endif
+	if (host->flags & HOST_F_DBDMA)
+		au1xmmc_dbdma_shutdown(host);
 
 	tasklet_kill(&host->data_task);
 	tasklet_kill(&host->finish_task);
@@ -1091,7 +1114,7 @@ out0:
 	return ret;
 }
 
-static int __devexit au1xmmc_remove(struct platform_device *pdev)
+static int au1xmmc_remove(struct platform_device *pdev)
 {
 	struct au1xmmc_host *host = platform_get_drvdata(pdev);
 
@@ -1115,9 +1138,9 @@ static int __devexit au1xmmc_remove(struct platform_device *pdev)
 		tasklet_kill(&host->data_task);
 		tasklet_kill(&host->finish_task);
 
-#ifdef CONFIG_SOC_AU1200
-		au1xmmc_dbdma_shutdown(host);
-#endif
+		if (host->flags & HOST_F_DBDMA)
+			au1xmmc_dbdma_shutdown(host);
+
 		au1xmmc_set_power(host, 0);
 
 		free_irq(host->irq, host);
@@ -1126,7 +1149,6 @@ static int __devexit au1xmmc_remove(struct platform_device *pdev)
 		kfree(host->ioarea);
 
 		mmc_free_host(host->mmc);
-		platform_set_drvdata(pdev, NULL);
 	}
 	return 0;
 }
@@ -1135,11 +1157,6 @@ static int __devexit au1xmmc_remove(struct platform_device *pdev)
 static int au1xmmc_suspend(struct platform_device *pdev, pm_message_t state)
 {
 	struct au1xmmc_host *host = platform_get_drvdata(pdev);
-	int ret;
-
-	ret = mmc_suspend_host(host->mmc, state);
-	if (ret)
-		return ret;
 
 	au_writel(0, HOST_CONFIG2(host));
 	au_writel(0, HOST_CONFIG(host));
@@ -1156,7 +1173,7 @@ static int au1xmmc_resume(struct platform_device *pdev)
 
 	au1xmmc_reset_controller(host);
 
-	return mmc_resume_host(host->mmc);
+	return 0;
 }
 #else
 #define au1xmmc_suspend NULL
@@ -1176,24 +1193,23 @@ static struct platform_driver au1xmmc_driver = {
 
 static int __init au1xmmc_init(void)
 {
-#ifdef CONFIG_SOC_AU1200
-	/* DSCR_CMD0_ALWAYS has a stride of 32 bits, we need a stride
-	 * of 8 bits.  And since devices are shared, we need to create
-	 * our own to avoid freaking out other devices.
-	 */
-	memid = au1xxx_ddma_add_device(&au1xmmc_mem_dbdev);
-	if (!memid)
-		printk(KERN_ERR "au1xmmc: cannot add memory dbdma dev\n");
-#endif
+	if (has_dbdma()) {
+		/* DSCR_CMD0_ALWAYS has a stride of 32 bits, we need a stride
+		* of 8 bits.  And since devices are shared, we need to create
+		* our own to avoid freaking out other devices.
+		*/
+		memid = au1xxx_ddma_add_device(&au1xmmc_mem_dbdev);
+		if (!memid)
+			pr_err("au1xmmc: cannot add memory dbdma\n");
+	}
 	return platform_driver_register(&au1xmmc_driver);
 }
 
 static void __exit au1xmmc_exit(void)
 {
-#ifdef CONFIG_SOC_AU1200
-	if (memid)
+	if (has_dbdma() && memid)
 		au1xxx_ddma_del_device(memid);
-#endif
+
 	platform_driver_unregister(&au1xmmc_driver);
 }
 
diff --git a/drivers/mmc/host/bfin_sdh.c b/drivers/mmc/host/bfin_sdh.c
new file mode 100644
index 00000000000..2b7f37e82ca
--- /dev/null
+++ b/drivers/mmc/host/bfin_sdh.c
@@ -0,0 +1,682 @@
+/*
+ * bfin_sdh.c - Analog Devices Blackfin SDH Controller
+ *
+ * Copyright (C) 2007-2009 Analog Device Inc.
+ *
+ * Licensed under the GPL-2 or later.
+ */
+
+#define DRIVER_NAME	"bfin-sdh"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/ioport.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/dma-mapping.h>
+#include <linux/mmc/host.h>
+#include <linux/proc_fs.h>
+#include <linux/gfp.h>
+
+#include <asm/cacheflush.h>
+#include <asm/dma.h>
+#include <asm/portmux.h>
+#include <asm/bfin_sdh.h>
+
+#if defined(CONFIG_BF51x) || defined(__ADSPBF60x__)
+#define bfin_read_SDH_CLK_CTL		bfin_read_RSI_CLK_CTL
+#define bfin_write_SDH_CLK_CTL		bfin_write_RSI_CLK_CTL
+#define bfin_write_SDH_ARGUMENT		bfin_write_RSI_ARGUMENT
+#define bfin_write_SDH_COMMAND		bfin_write_RSI_COMMAND
+#define bfin_write_SDH_DATA_TIMER	bfin_write_RSI_DATA_TIMER
+#define bfin_read_SDH_RESPONSE0		bfin_read_RSI_RESPONSE0
+#define bfin_read_SDH_RESPONSE1		bfin_read_RSI_RESPONSE1
+#define bfin_read_SDH_RESPONSE2		bfin_read_RSI_RESPONSE2
+#define bfin_read_SDH_RESPONSE3		bfin_read_RSI_RESPONSE3
+#define bfin_write_SDH_DATA_LGTH	bfin_write_RSI_DATA_LGTH
+#define bfin_read_SDH_DATA_CTL		bfin_read_RSI_DATA_CTL
+#define bfin_write_SDH_DATA_CTL		bfin_write_RSI_DATA_CTL
+#define bfin_read_SDH_DATA_CNT		bfin_read_RSI_DATA_CNT
+#define bfin_write_SDH_STATUS_CLR	bfin_write_RSI_STATUS_CLR
+#define bfin_read_SDH_E_STATUS		bfin_read_RSI_E_STATUS
+#define bfin_write_SDH_E_STATUS		bfin_write_RSI_E_STATUS
+#define bfin_read_SDH_STATUS		bfin_read_RSI_STATUS
+#define bfin_write_SDH_MASK0		bfin_write_RSI_MASK0
+#define bfin_write_SDH_E_MASK		bfin_write_RSI_E_MASK
+#define bfin_read_SDH_CFG		bfin_read_RSI_CFG
+#define bfin_write_SDH_CFG		bfin_write_RSI_CFG
+# if defined(__ADSPBF60x__)
+#  define bfin_read_SDH_BLK_SIZE	bfin_read_RSI_BLKSZ
+#  define bfin_write_SDH_BLK_SIZE	bfin_write_RSI_BLKSZ
+# else
+#  define bfin_read_SDH_PWR_CTL		bfin_read_RSI_PWR_CTL
+#  define bfin_write_SDH_PWR_CTL	bfin_write_RSI_PWR_CTL
+# endif
+#endif
+
+struct sdh_host {
+	struct mmc_host		*mmc;
+	spinlock_t		lock;
+	struct resource		*res;
+	void __iomem		*base;
+	int			irq;
+	int			stat_irq;
+	int			dma_ch;
+	int			dma_dir;
+	struct dma_desc_array	*sg_cpu;
+	dma_addr_t		sg_dma;
+	int			dma_len;
+
+	unsigned long		sclk;
+	unsigned int		imask;
+	unsigned int		power_mode;
+	unsigned int		clk_div;
+
+	struct mmc_request	*mrq;
+	struct mmc_command	*cmd;
+	struct mmc_data		*data;
+};
+
+static struct bfin_sd_host *get_sdh_data(struct platform_device *pdev)
+{
+	return pdev->dev.platform_data;
+}
+
+static void sdh_stop_clock(struct sdh_host *host)
+{
+	bfin_write_SDH_CLK_CTL(bfin_read_SDH_CLK_CTL() & ~CLK_E);
+	SSYNC();
+}
+
+static void sdh_enable_stat_irq(struct sdh_host *host, unsigned int mask)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+	host->imask |= mask;
+	bfin_write_SDH_MASK0(mask);
+	SSYNC();
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void sdh_disable_stat_irq(struct sdh_host *host, unsigned int mask)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+	host->imask &= ~mask;
+	bfin_write_SDH_MASK0(host->imask);
+	SSYNC();
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static int sdh_setup_data(struct sdh_host *host, struct mmc_data *data)
+{
+	unsigned int length;
+	unsigned int data_ctl;
+	unsigned int dma_cfg;
+	unsigned int cycle_ns, timeout;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter flags: 0x%x\n", __func__, data->flags);
+	host->data = data;
+	data_ctl = 0;
+	dma_cfg = 0;
+
+	length = data->blksz * data->blocks;
+	bfin_write_SDH_DATA_LGTH(length);
+
+	if (data->flags & MMC_DATA_STREAM)
+		data_ctl |= DTX_MODE;
+
+	if (data->flags & MMC_DATA_READ)
+		data_ctl |= DTX_DIR;
+	/* Only supports power-of-2 block size */
+	if (data->blksz & (data->blksz - 1))
+		return -EINVAL;
+#ifndef RSI_BLKSZ
+	data_ctl |= ((ffs(data->blksz) - 1) << 4);
+#else
+        bfin_write_SDH_BLK_SIZE(data->blksz);
+#endif
+
+	bfin_write_SDH_DATA_CTL(data_ctl);
+	/* the time of a host clock period in ns */
+	cycle_ns = 1000000000 / (host->sclk / (2 * (host->clk_div + 1)));
+	timeout = data->timeout_ns / cycle_ns;
+	timeout += data->timeout_clks;
+	bfin_write_SDH_DATA_TIMER(timeout);
+	SSYNC();
+
+	if (data->flags & MMC_DATA_READ) {
+		host->dma_dir = DMA_FROM_DEVICE;
+		dma_cfg |= WNR;
+	} else
+		host->dma_dir = DMA_TO_DEVICE;
+
+	sdh_enable_stat_irq(host, (DAT_CRC_FAIL | DAT_TIME_OUT | DAT_END));
+	host->dma_len = dma_map_sg(mmc_dev(host->mmc), data->sg, data->sg_len, host->dma_dir);
+#if defined(CONFIG_BF54x) || defined(CONFIG_BF60x)
+	dma_cfg |= DMAFLOW_ARRAY | RESTART | WDSIZE_32 | DMAEN;
+# ifdef RSI_BLKSZ
+	dma_cfg |= PSIZE_32 | NDSIZE_3;
+# else
+	dma_cfg |= NDSIZE_5;
+# endif
+	{
+		struct scatterlist *sg;
+		int i;
+		for_each_sg(data->sg, sg, host->dma_len, i) {
+			host->sg_cpu[i].start_addr = sg_dma_address(sg);
+			host->sg_cpu[i].cfg = dma_cfg;
+			host->sg_cpu[i].x_count = sg_dma_len(sg) / 4;
+			host->sg_cpu[i].x_modify = 4;
+			dev_dbg(mmc_dev(host->mmc), "%d: start_addr:0x%lx, "
+				"cfg:0x%lx, x_count:0x%lx, x_modify:0x%lx\n",
+				i, host->sg_cpu[i].start_addr,
+				host->sg_cpu[i].cfg, host->sg_cpu[i].x_count,
+				host->sg_cpu[i].x_modify);
+		}
+	}
+	flush_dcache_range((unsigned int)host->sg_cpu,
+		(unsigned int)host->sg_cpu +
+			host->dma_len * sizeof(struct dma_desc_array));
+	/* Set the last descriptor to stop mode */
+	host->sg_cpu[host->dma_len - 1].cfg &= ~(DMAFLOW | NDSIZE);
+	host->sg_cpu[host->dma_len - 1].cfg |= DI_EN;
+
+	set_dma_curr_desc_addr(host->dma_ch, (unsigned long *)host->sg_dma);
+	set_dma_x_count(host->dma_ch, 0);
+	set_dma_x_modify(host->dma_ch, 0);
+	SSYNC();
+	set_dma_config(host->dma_ch, dma_cfg);
+#elif defined(CONFIG_BF51x)
+	/* RSI DMA doesn't work in array mode */
+	dma_cfg |= WDSIZE_32 | DMAEN;
+	set_dma_start_addr(host->dma_ch, sg_dma_address(&data->sg[0]));
+	set_dma_x_count(host->dma_ch, length / 4);
+	set_dma_x_modify(host->dma_ch, 4);
+	SSYNC();
+	set_dma_config(host->dma_ch, dma_cfg);
+#endif
+	bfin_write_SDH_DATA_CTL(bfin_read_SDH_DATA_CTL() | DTX_DMA_E | DTX_E);
+
+	SSYNC();
+
+	dev_dbg(mmc_dev(host->mmc), "%s exit\n", __func__);
+	return 0;
+}
+
+static void sdh_start_cmd(struct sdh_host *host, struct mmc_command *cmd)
+{
+	unsigned int sdh_cmd;
+	unsigned int stat_mask;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter cmd: 0x%p\n", __func__, cmd);
+	WARN_ON(host->cmd != NULL);
+	host->cmd = cmd;
+
+	sdh_cmd = 0;
+	stat_mask = 0;
+
+	sdh_cmd |= cmd->opcode;
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		sdh_cmd |= CMD_RSP;
+		stat_mask |= CMD_RESP_END;
+	} else {
+		stat_mask |= CMD_SENT;
+	}
+
+	if (cmd->flags & MMC_RSP_136)
+		sdh_cmd |= CMD_L_RSP;
+
+	stat_mask |= CMD_CRC_FAIL | CMD_TIME_OUT;
+
+	sdh_enable_stat_irq(host, stat_mask);
+
+	bfin_write_SDH_ARGUMENT(cmd->arg);
+	bfin_write_SDH_COMMAND(sdh_cmd | CMD_E);
+	bfin_write_SDH_CLK_CTL(bfin_read_SDH_CLK_CTL() | CLK_E);
+	SSYNC();
+}
+
+static void sdh_finish_request(struct sdh_host *host, struct mmc_request *mrq)
+{
+	dev_dbg(mmc_dev(host->mmc), "%s enter\n", __func__);
+	host->mrq = NULL;
+	host->cmd = NULL;
+	host->data = NULL;
+	mmc_request_done(host->mmc, mrq);
+}
+
+static int sdh_cmd_done(struct sdh_host *host, unsigned int stat)
+{
+	struct mmc_command *cmd = host->cmd;
+	int ret = 0;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter cmd: %p\n", __func__, cmd);
+	if (!cmd)
+		return 0;
+
+	host->cmd = NULL;
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		cmd->resp[0] = bfin_read_SDH_RESPONSE0();
+		if (cmd->flags & MMC_RSP_136) {
+			cmd->resp[1] = bfin_read_SDH_RESPONSE1();
+			cmd->resp[2] = bfin_read_SDH_RESPONSE2();
+			cmd->resp[3] = bfin_read_SDH_RESPONSE3();
+		}
+	}
+	if (stat & CMD_TIME_OUT)
+		cmd->error = -ETIMEDOUT;
+	else if (stat & CMD_CRC_FAIL && cmd->flags & MMC_RSP_CRC)
+		cmd->error = -EILSEQ;
+
+	sdh_disable_stat_irq(host, (CMD_SENT | CMD_RESP_END | CMD_TIME_OUT | CMD_CRC_FAIL));
+
+	if (host->data && !cmd->error) {
+		if (host->data->flags & MMC_DATA_WRITE) {
+			ret = sdh_setup_data(host, host->data);
+			if (ret)
+				return 0;
+		}
+
+		sdh_enable_stat_irq(host, DAT_END | RX_OVERRUN | TX_UNDERRUN | DAT_TIME_OUT);
+	} else
+		sdh_finish_request(host, host->mrq);
+
+	return 1;
+}
+
+static int sdh_data_done(struct sdh_host *host, unsigned int stat)
+{
+	struct mmc_data *data = host->data;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter stat: 0x%x\n", __func__, stat);
+	if (!data)
+		return 0;
+
+	disable_dma(host->dma_ch);
+	dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len,
+		     host->dma_dir);
+
+	if (stat & DAT_TIME_OUT)
+		data->error = -ETIMEDOUT;
+	else if (stat & DAT_CRC_FAIL)
+		data->error = -EILSEQ;
+	else if (stat & (RX_OVERRUN | TX_UNDERRUN))
+		data->error = -EIO;
+
+	if (!data->error)
+		data->bytes_xfered = data->blocks * data->blksz;
+	else
+		data->bytes_xfered = 0;
+
+	bfin_write_SDH_STATUS_CLR(DAT_END_STAT | DAT_TIMEOUT_STAT | \
+			DAT_CRC_FAIL_STAT | DAT_BLK_END_STAT | RX_OVERRUN | TX_UNDERRUN);
+	bfin_write_SDH_DATA_CTL(0);
+	SSYNC();
+
+	host->data = NULL;
+	if (host->mrq->stop) {
+		sdh_stop_clock(host);
+		sdh_start_cmd(host, host->mrq->stop);
+	} else {
+		sdh_finish_request(host, host->mrq);
+	}
+
+	return 1;
+}
+
+static void sdh_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct sdh_host *host = mmc_priv(mmc);
+	int ret = 0;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter, mrp:%p, cmd:%p\n", __func__, mrq, mrq->cmd);
+	WARN_ON(host->mrq != NULL);
+
+	spin_lock(&host->lock);
+	host->mrq = mrq;
+	host->data = mrq->data;
+
+	if (mrq->data && mrq->data->flags & MMC_DATA_READ) {
+		ret = sdh_setup_data(host, mrq->data);
+		if (ret)
+			goto data_err;
+	}
+
+	sdh_start_cmd(host, mrq->cmd);
+data_err:
+	spin_unlock(&host->lock);
+}
+
+static void sdh_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct sdh_host *host;
+	u16 clk_ctl = 0;
+#ifndef RSI_BLKSZ
+	u16 pwr_ctl = 0;
+#endif
+	u16 cfg;
+	host = mmc_priv(mmc);
+
+	spin_lock(&host->lock);
+
+	cfg = bfin_read_SDH_CFG();
+	cfg |= MWE;
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_4:
+#ifndef RSI_BLKSZ
+		cfg &= ~PD_SDDAT3;
+#endif
+		cfg |= PUP_SDDAT3;
+		/* Enable 4 bit SDIO */
+		cfg |= SD4E;
+		clk_ctl |= WIDE_BUS_4;
+		break;
+	case MMC_BUS_WIDTH_8:
+#ifndef RSI_BLKSZ
+		cfg &= ~PD_SDDAT3;
+#endif
+		cfg |= PUP_SDDAT3;
+		/* Disable 4 bit SDIO */
+		cfg &= ~SD4E;
+		clk_ctl |= BYTE_BUS_8;
+		break;
+	default:
+		cfg &= ~PUP_SDDAT3;
+		/* Disable 4 bit SDIO */
+		cfg &= ~SD4E;
+	}
+	bfin_write_SDH_CFG(cfg);
+
+	host->power_mode = ios->power_mode;
+#ifndef RSI_BLKSZ
+	if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN) {
+		pwr_ctl |= ROD_CTL;
+# ifndef CONFIG_SDH_BFIN_MISSING_CMD_PULLUP_WORKAROUND
+		pwr_ctl |= SD_CMD_OD;
+# endif
+	}
+
+	if (ios->power_mode != MMC_POWER_OFF)
+		pwr_ctl |= PWR_ON;
+	else
+		pwr_ctl &= ~PWR_ON;
+
+	bfin_write_SDH_PWR_CTL(pwr_ctl);
+#else
+# ifndef CONFIG_SDH_BFIN_MISSING_CMD_PULLUP_WORKAROUND
+	if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN)
+		cfg |= SD_CMD_OD;
+	else
+		cfg &= ~SD_CMD_OD;
+# endif
+
+	if (ios->power_mode != MMC_POWER_OFF)
+		cfg |= PWR_ON;
+	else
+		cfg &= ~PWR_ON;
+
+	bfin_write_SDH_CFG(cfg);
+#endif
+	SSYNC();
+
+	if (ios->power_mode == MMC_POWER_ON && ios->clock) {
+		unsigned char clk_div;
+		clk_div = (get_sclk() / ios->clock - 1) / 2;
+		clk_div = min_t(unsigned char, clk_div, 0xFF);
+		clk_ctl |= clk_div;
+		clk_ctl |= CLK_E;
+		host->clk_div = clk_div;
+		bfin_write_SDH_CLK_CTL(clk_ctl);
+	} else
+		sdh_stop_clock(host);
+
+	/* set up sdh interrupt mask*/
+	if (ios->power_mode == MMC_POWER_ON)
+		bfin_write_SDH_MASK0(DAT_END | DAT_TIME_OUT | DAT_CRC_FAIL |
+			RX_OVERRUN | TX_UNDERRUN | CMD_SENT | CMD_RESP_END |
+			CMD_TIME_OUT | CMD_CRC_FAIL);
+	else
+		bfin_write_SDH_MASK0(0);
+	SSYNC();
+
+	spin_unlock(&host->lock);
+
+	dev_dbg(mmc_dev(host->mmc), "SDH: clk_div = 0x%x actual clock:%ld expected clock:%d\n",
+		host->clk_div,
+		host->clk_div ? get_sclk() / (2 * (host->clk_div + 1)) : 0,
+		ios->clock);
+}
+
+static const struct mmc_host_ops sdh_ops = {
+	.request	= sdh_request,
+	.set_ios	= sdh_set_ios,
+};
+
+static irqreturn_t sdh_dma_irq(int irq, void *devid)
+{
+	struct sdh_host *host = devid;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter, irq_stat: 0x%04lx\n", __func__,
+		get_dma_curr_irqstat(host->dma_ch));
+	clear_dma_irqstat(host->dma_ch);
+	SSYNC();
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t sdh_stat_irq(int irq, void *devid)
+{
+	struct sdh_host *host = devid;
+	unsigned int status;
+	int handled = 0;
+
+	dev_dbg(mmc_dev(host->mmc), "%s enter\n", __func__);
+
+	spin_lock(&host->lock);
+
+	status = bfin_read_SDH_E_STATUS();
+	if (status & SD_CARD_DET) {
+		mmc_detect_change(host->mmc, 0);
+		bfin_write_SDH_E_STATUS(SD_CARD_DET);
+	}
+	status = bfin_read_SDH_STATUS();
+	if (status & (CMD_SENT | CMD_RESP_END | CMD_TIME_OUT | CMD_CRC_FAIL)) {
+		handled |= sdh_cmd_done(host, status);
+		bfin_write_SDH_STATUS_CLR(CMD_SENT_STAT | CMD_RESP_END_STAT | \
+				CMD_TIMEOUT_STAT | CMD_CRC_FAIL_STAT);
+		SSYNC();
+	}
+
+	status = bfin_read_SDH_STATUS();
+	if (status & (DAT_END | DAT_TIME_OUT | DAT_CRC_FAIL | RX_OVERRUN | TX_UNDERRUN))
+		handled |= sdh_data_done(host, status);
+
+	spin_unlock(&host->lock);
+
+	dev_dbg(mmc_dev(host->mmc), "%s exit\n\n", __func__);
+
+	return IRQ_RETVAL(handled);
+}
+
+static void sdh_reset(void)
+{
+#if defined(CONFIG_BF54x)
+	/* Secure Digital Host shares DMA with Nand controller */
+	bfin_write_DMAC1_PERIMUX(bfin_read_DMAC1_PERIMUX() | 0x1);
+#endif
+
+	bfin_write_SDH_CFG(bfin_read_SDH_CFG() | CLKS_EN);
+	SSYNC();
+
+	/* Disable card inserting detection pin. set MMC_CAP_NEEDS_POLL, and
+	 * mmc stack will do the detection.
+	 */
+	bfin_write_SDH_CFG((bfin_read_SDH_CFG() & 0x1F) | (PUP_SDDAT | PUP_SDDAT3));
+	SSYNC();
+}
+
+static int sdh_probe(struct platform_device *pdev)
+{
+	struct mmc_host *mmc;
+	struct sdh_host *host;
+	struct bfin_sd_host *drv_data = get_sdh_data(pdev);
+	int ret;
+
+	if (!drv_data) {
+		dev_err(&pdev->dev, "missing platform driver data\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	mmc = mmc_alloc_host(sizeof(struct sdh_host), &pdev->dev);
+	if (!mmc) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	mmc->ops = &sdh_ops;
+#if defined(CONFIG_BF51x)
+	mmc->max_segs = 1;
+#else
+	mmc->max_segs = PAGE_SIZE / sizeof(struct dma_desc_array);
+#endif
+#ifdef RSI_BLKSZ
+	mmc->max_seg_size = -1;
+#else
+	mmc->max_seg_size = 1 << 16;
+#endif
+	mmc->max_blk_size = 1 << 11;
+	mmc->max_blk_count = 1 << 11;
+	mmc->max_req_size = PAGE_SIZE;
+	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+	mmc->f_max = get_sclk();
+	mmc->f_min = mmc->f_max >> 9;
+	mmc->caps = MMC_CAP_4_BIT_DATA | MMC_CAP_NEEDS_POLL;
+	host = mmc_priv(mmc);
+	host->mmc = mmc;
+	host->sclk = get_sclk();
+
+	spin_lock_init(&host->lock);
+	host->irq = drv_data->irq_int0;
+	host->dma_ch = drv_data->dma_chan;
+
+	ret = request_dma(host->dma_ch, DRIVER_NAME "DMA");
+	if (ret) {
+		dev_err(&pdev->dev, "unable to request DMA channel\n");
+		goto out1;
+	}
+
+	ret = set_dma_callback(host->dma_ch, sdh_dma_irq, host);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to request DMA irq\n");
+		goto out2;
+	}
+
+	host->sg_cpu = dma_alloc_coherent(&pdev->dev, PAGE_SIZE, &host->sg_dma, GFP_KERNEL);
+	if (host->sg_cpu == NULL) {
+		ret = -ENOMEM;
+		goto out2;
+	}
+
+	platform_set_drvdata(pdev, mmc);
+
+	ret = request_irq(host->irq, sdh_stat_irq, 0, "SDH Status IRQ", host);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to request status irq\n");
+		goto out3;
+	}
+
+	ret = peripheral_request_list(drv_data->pin_req, DRIVER_NAME);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to request peripheral pins\n");
+		goto out4;
+	}
+
+	sdh_reset();
+
+	mmc_add_host(mmc);
+	return 0;
+
+out4:
+	free_irq(host->irq, host);
+out3:
+	mmc_remove_host(mmc);
+	dma_free_coherent(&pdev->dev, PAGE_SIZE, host->sg_cpu, host->sg_dma);
+out2:
+	free_dma(host->dma_ch);
+out1:
+	mmc_free_host(mmc);
+ out:
+	return ret;
+}
+
+static int sdh_remove(struct platform_device *pdev)
+{
+	struct mmc_host *mmc = platform_get_drvdata(pdev);
+
+	if (mmc) {
+		struct sdh_host *host = mmc_priv(mmc);
+
+		mmc_remove_host(mmc);
+
+		sdh_stop_clock(host);
+		free_irq(host->irq, host);
+		free_dma(host->dma_ch);
+		dma_free_coherent(&pdev->dev, PAGE_SIZE, host->sg_cpu, host->sg_dma);
+
+		mmc_free_host(mmc);
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int sdh_suspend(struct platform_device *dev, pm_message_t state)
+{
+	struct bfin_sd_host *drv_data = get_sdh_data(dev);
+
+	peripheral_free_list(drv_data->pin_req);
+
+	return 0;
+}
+
+static int sdh_resume(struct platform_device *dev)
+{
+	struct bfin_sd_host *drv_data = get_sdh_data(dev);
+	int ret = 0;
+
+	ret = peripheral_request_list(drv_data->pin_req, DRIVER_NAME);
+	if (ret) {
+		dev_err(&dev->dev, "unable to request peripheral pins\n");
+		return ret;
+	}
+
+	sdh_reset();
+	return ret;
+}
+#else
+# define sdh_suspend NULL
+# define sdh_resume  NULL
+#endif
+
+static struct platform_driver sdh_driver = {
+	.probe   = sdh_probe,
+	.remove  = sdh_remove,
+	.suspend = sdh_suspend,
+	.resume  = sdh_resume,
+	.driver  = {
+		.name = DRIVER_NAME,
+	},
+};
+
+module_platform_driver(sdh_driver);
+
+MODULE_DESCRIPTION("Blackfin Secure Digital Host Driver");
+MODULE_AUTHOR("Cliff Cai, Roy Huang");
+MODULE_LICENSE("GPL");
diff --git a/drivers/mmc/host/cb710-mmc.c b/drivers/mmc/host/cb710-mmc.c
new file mode 100644
index 00000000000..1087b4c79cd
--- /dev/null
+++ b/drivers/mmc/host/cb710-mmc.c
@@ -0,0 +1,780 @@
+/*
+ *  cb710/mmc.c
+ *
+ *  Copyright by Michał Mirosław, 2008-2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include "cb710-mmc.h"
+
+static const u8 cb710_clock_divider_log2[8] = {
+/*	1, 2, 4, 8, 16, 32, 128, 512 */
+	0, 1, 2, 3,  4,  5,   7,   9
+};
+#define CB710_MAX_DIVIDER_IDX	\
+	(ARRAY_SIZE(cb710_clock_divider_log2) - 1)
+
+static const u8 cb710_src_freq_mhz[16] = {
+	33, 10, 20, 25, 30, 35, 40, 45,
+	50, 55, 60, 65, 70, 75, 80, 85
+};
+
+static void cb710_mmc_select_clock_divider(struct mmc_host *mmc, int hz)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+	struct pci_dev *pdev = cb710_slot_to_chip(slot)->pdev;
+	u32 src_freq_idx;
+	u32 divider_idx;
+	int src_hz;
+
+	/* on CB710 in HP nx9500:
+	 *   src_freq_idx == 0
+	 *   indexes 1-7 work as written in the table
+	 *   indexes 0,8-15 give no clock output
+	 */
+	pci_read_config_dword(pdev, 0x48, &src_freq_idx);
+	src_freq_idx = (src_freq_idx >> 16) & 0xF;
+	src_hz = cb710_src_freq_mhz[src_freq_idx] * 1000000;
+
+	for (divider_idx = 0; divider_idx < CB710_MAX_DIVIDER_IDX; ++divider_idx) {
+		if (hz >= src_hz >> cb710_clock_divider_log2[divider_idx])
+			break;
+	}
+
+	if (src_freq_idx)
+		divider_idx |= 0x8;
+	else if (divider_idx == 0)
+		divider_idx = 1;
+
+	cb710_pci_update_config_reg(pdev, 0x40, ~0xF0000000, divider_idx << 28);
+
+	dev_dbg(cb710_slot_dev(slot),
+		"clock set to %d Hz, wanted %d Hz; src_freq_idx = %d, divider_idx = %d|%d\n",
+		src_hz >> cb710_clock_divider_log2[divider_idx & 7],
+		hz, src_freq_idx, divider_idx & 7, divider_idx & 8);
+}
+
+static void __cb710_mmc_enable_irq(struct cb710_slot *slot,
+	unsigned short enable, unsigned short mask)
+{
+	/* clear global IE
+	 * - it gets set later if any interrupt sources are enabled */
+	mask |= CB710_MMC_IE_IRQ_ENABLE;
+
+	/* look like interrupt is fired whenever
+	 * WORD[0x0C] & WORD[0x10] != 0;
+	 * -> bit 15 port 0x0C seems to be global interrupt enable
+	 */
+
+	enable = (cb710_read_port_16(slot, CB710_MMC_IRQ_ENABLE_PORT)
+		& ~mask) | enable;
+
+	if (enable)
+		enable |= CB710_MMC_IE_IRQ_ENABLE;
+
+	cb710_write_port_16(slot, CB710_MMC_IRQ_ENABLE_PORT, enable);
+}
+
+static void cb710_mmc_enable_irq(struct cb710_slot *slot,
+	unsigned short enable, unsigned short mask)
+{
+	struct cb710_mmc_reader *reader = mmc_priv(cb710_slot_to_mmc(slot));
+	unsigned long flags;
+
+	spin_lock_irqsave(&reader->irq_lock, flags);
+	/* this is the only thing irq_lock protects */
+	__cb710_mmc_enable_irq(slot, enable, mask);
+	spin_unlock_irqrestore(&reader->irq_lock, flags);
+}
+
+static void cb710_mmc_reset_events(struct cb710_slot *slot)
+{
+	cb710_write_port_8(slot, CB710_MMC_STATUS0_PORT, 0xFF);
+	cb710_write_port_8(slot, CB710_MMC_STATUS1_PORT, 0xFF);
+	cb710_write_port_8(slot, CB710_MMC_STATUS2_PORT, 0xFF);
+}
+
+static void cb710_mmc_enable_4bit_data(struct cb710_slot *slot, int enable)
+{
+	if (enable)
+		cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT,
+			CB710_MMC_C1_4BIT_DATA_BUS, 0);
+	else
+		cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT,
+			0, CB710_MMC_C1_4BIT_DATA_BUS);
+}
+
+static int cb710_check_event(struct cb710_slot *slot, u8 what)
+{
+	u16 status;
+
+	status = cb710_read_port_16(slot, CB710_MMC_STATUS_PORT);
+
+	if (status & CB710_MMC_S0_FIFO_UNDERFLOW) {
+		/* it is just a guess, so log it */
+		dev_dbg(cb710_slot_dev(slot),
+			"CHECK : ignoring bit 6 in status %04X\n", status);
+		cb710_write_port_8(slot, CB710_MMC_STATUS0_PORT,
+			CB710_MMC_S0_FIFO_UNDERFLOW);
+		status &= ~CB710_MMC_S0_FIFO_UNDERFLOW;
+	}
+
+	if (status & CB710_MMC_STATUS_ERROR_EVENTS) {
+		dev_dbg(cb710_slot_dev(slot),
+			"CHECK : returning EIO on status %04X\n", status);
+		cb710_write_port_8(slot, CB710_MMC_STATUS0_PORT, status & 0xFF);
+		cb710_write_port_8(slot, CB710_MMC_STATUS1_PORT,
+			CB710_MMC_S1_RESET);
+		return -EIO;
+	}
+
+	/* 'what' is a bit in MMC_STATUS1 */
+	if ((status >> 8) & what) {
+		cb710_write_port_8(slot, CB710_MMC_STATUS1_PORT, what);
+		return 1;
+	}
+
+	return 0;
+}
+
+static int cb710_wait_for_event(struct cb710_slot *slot, u8 what)
+{
+	int err = 0;
+	unsigned limit = 2000000;	/* FIXME: real timeout */
+
+#ifdef CONFIG_CB710_DEBUG
+	u32 e, x;
+	e = cb710_read_port_32(slot, CB710_MMC_STATUS_PORT);
+#endif
+
+	while (!(err = cb710_check_event(slot, what))) {
+		if (!--limit) {
+			cb710_dump_regs(cb710_slot_to_chip(slot),
+				CB710_DUMP_REGS_MMC);
+			err = -ETIMEDOUT;
+			break;
+		}
+		udelay(1);
+	}
+
+#ifdef CONFIG_CB710_DEBUG
+	x = cb710_read_port_32(slot, CB710_MMC_STATUS_PORT);
+
+	limit = 2000000 - limit;
+	if (limit > 100)
+		dev_dbg(cb710_slot_dev(slot),
+			"WAIT10: waited %d loops, what %d, entry val %08X, exit val %08X\n",
+			limit, what, e, x);
+#endif
+	return err < 0 ? err : 0;
+}
+
+
+static int cb710_wait_while_busy(struct cb710_slot *slot, uint8_t mask)
+{
+	unsigned limit = 500000;	/* FIXME: real timeout */
+	int err = 0;
+
+#ifdef CONFIG_CB710_DEBUG
+	u32 e, x;
+	e = cb710_read_port_32(slot, CB710_MMC_STATUS_PORT);
+#endif
+
+	while (cb710_read_port_8(slot, CB710_MMC_STATUS2_PORT) & mask) {
+		if (!--limit) {
+			cb710_dump_regs(cb710_slot_to_chip(slot),
+				CB710_DUMP_REGS_MMC);
+			err = -ETIMEDOUT;
+			break;
+		}
+		udelay(1);
+	}
+
+#ifdef CONFIG_CB710_DEBUG
+	x = cb710_read_port_32(slot, CB710_MMC_STATUS_PORT);
+
+	limit = 500000 - limit;
+	if (limit > 100)
+		dev_dbg(cb710_slot_dev(slot),
+			"WAIT12: waited %d loops, mask %02X, entry val %08X, exit val %08X\n",
+			limit, mask, e, x);
+#endif
+	return err;
+}
+
+static void cb710_mmc_set_transfer_size(struct cb710_slot *slot,
+	size_t count, size_t blocksize)
+{
+	cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	cb710_write_port_32(slot, CB710_MMC_TRANSFER_SIZE_PORT,
+		((count - 1) << 16)|(blocksize - 1));
+
+	dev_vdbg(cb710_slot_dev(slot), "set up for %zu block%s of %zu bytes\n",
+		count, count == 1 ? "" : "s", blocksize);
+}
+
+static void cb710_mmc_fifo_hack(struct cb710_slot *slot)
+{
+	/* without this, received data is prepended with 8-bytes of zeroes */
+	u32 r1, r2;
+	int ok = 0;
+
+	r1 = cb710_read_port_32(slot, CB710_MMC_DATA_PORT);
+	r2 = cb710_read_port_32(slot, CB710_MMC_DATA_PORT);
+	if (cb710_read_port_8(slot, CB710_MMC_STATUS0_PORT)
+	    & CB710_MMC_S0_FIFO_UNDERFLOW) {
+		cb710_write_port_8(slot, CB710_MMC_STATUS0_PORT,
+			CB710_MMC_S0_FIFO_UNDERFLOW);
+		ok = 1;
+	}
+
+	dev_dbg(cb710_slot_dev(slot),
+		"FIFO-read-hack: expected STATUS0 bit was %s\n",
+		ok ? "set." : "NOT SET!");
+	dev_dbg(cb710_slot_dev(slot),
+		"FIFO-read-hack: dwords ignored: %08X %08X - %s\n",
+		r1, r2, (r1|r2) ? "BAD (NOT ZERO)!" : "ok");
+}
+
+static int cb710_mmc_receive_pio(struct cb710_slot *slot,
+	struct sg_mapping_iter *miter, size_t dw_count)
+{
+	if (!(cb710_read_port_8(slot, CB710_MMC_STATUS2_PORT) & CB710_MMC_S2_FIFO_READY)) {
+		int err = cb710_wait_for_event(slot,
+			CB710_MMC_S1_PIO_TRANSFER_DONE);
+		if (err)
+			return err;
+	}
+
+	cb710_sg_dwiter_write_from_io(miter,
+		slot->iobase + CB710_MMC_DATA_PORT, dw_count);
+
+	return 0;
+}
+
+static bool cb710_is_transfer_size_supported(struct mmc_data *data)
+{
+	return !(data->blksz & 15 && (data->blocks != 1 || data->blksz != 8));
+}
+
+static int cb710_mmc_receive(struct cb710_slot *slot, struct mmc_data *data)
+{
+	struct sg_mapping_iter miter;
+	size_t len, blocks = data->blocks;
+	int err = 0;
+
+	/* TODO: I don't know how/if the hardware handles non-16B-boundary blocks
+	 * except single 8B block */
+	if (unlikely(data->blksz & 15 && (data->blocks != 1 || data->blksz != 8)))
+		return -EINVAL;
+
+	sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_TO_SG);
+
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT,
+		15, CB710_MMC_C2_READ_PIO_SIZE_MASK);
+
+	cb710_mmc_fifo_hack(slot);
+
+	while (blocks-- > 0) {
+		len = data->blksz;
+
+		while (len >= 16) {
+			err = cb710_mmc_receive_pio(slot, &miter, 4);
+			if (err)
+				goto out;
+			len -= 16;
+		}
+
+		if (!len)
+			continue;
+
+		cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT,
+			len - 1, CB710_MMC_C2_READ_PIO_SIZE_MASK);
+
+		len = (len >= 8) ? 4 : 2;
+		err = cb710_mmc_receive_pio(slot, &miter, len);
+		if (err)
+			goto out;
+	}
+out:
+	sg_miter_stop(&miter);
+	return err;
+}
+
+static int cb710_mmc_send(struct cb710_slot *slot, struct mmc_data *data)
+{
+	struct sg_mapping_iter miter;
+	size_t len, blocks = data->blocks;
+	int err = 0;
+
+	/* TODO: I don't know how/if the hardware handles multiple
+	 * non-16B-boundary blocks */
+	if (unlikely(data->blocks > 1 && data->blksz & 15))
+		return -EINVAL;
+
+	sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_FROM_SG);
+
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT,
+		0, CB710_MMC_C2_READ_PIO_SIZE_MASK);
+
+	while (blocks-- > 0) {
+		len = (data->blksz + 15) >> 4;
+		do {
+			if (!(cb710_read_port_8(slot, CB710_MMC_STATUS2_PORT)
+			    & CB710_MMC_S2_FIFO_EMPTY)) {
+				err = cb710_wait_for_event(slot,
+					CB710_MMC_S1_PIO_TRANSFER_DONE);
+				if (err)
+					goto out;
+			}
+			cb710_sg_dwiter_read_to_io(&miter,
+				slot->iobase + CB710_MMC_DATA_PORT, 4);
+		} while (--len);
+	}
+out:
+	sg_miter_stop(&miter);
+	return err;
+}
+
+static u16 cb710_encode_cmd_flags(struct cb710_mmc_reader *reader,
+	struct mmc_command *cmd)
+{
+	unsigned int flags = cmd->flags;
+	u16 cb_flags = 0;
+
+	/* Windows driver returned 0 for commands for which no response
+	 * is expected. It happened that there were only two such commands
+	 * used: MMC_GO_IDLE_STATE and MMC_GO_INACTIVE_STATE so it might
+	 * as well be a bug in that driver.
+	 *
+	 * Original driver set bit 14 for MMC/SD application
+	 * commands. There's no difference 'on the wire' and
+	 * it apparently works without it anyway.
+	 */
+
+	switch (flags & MMC_CMD_MASK) {
+	case MMC_CMD_AC:	cb_flags = CB710_MMC_CMD_AC;	break;
+	case MMC_CMD_ADTC:	cb_flags = CB710_MMC_CMD_ADTC;	break;
+	case MMC_CMD_BC:	cb_flags = CB710_MMC_CMD_BC;	break;
+	case MMC_CMD_BCR:	cb_flags = CB710_MMC_CMD_BCR;	break;
+	}
+
+	if (flags & MMC_RSP_BUSY)
+		cb_flags |= CB710_MMC_RSP_BUSY;
+
+	cb_flags |= cmd->opcode << CB710_MMC_CMD_CODE_SHIFT;
+
+	if (cmd->data && (cmd->data->flags & MMC_DATA_READ))
+		cb_flags |= CB710_MMC_DATA_READ;
+
+	if (flags & MMC_RSP_PRESENT) {
+		/* Windows driver set 01 at bits 4,3 except for
+		 * MMC_SET_BLOCKLEN where it set 10. Maybe the
+		 * hardware can do something special about this
+		 * command? The original driver looks buggy/incomplete
+		 * anyway so we ignore this for now.
+		 *
+		 * I assume that 00 here means no response is expected.
+		 */
+		cb_flags |= CB710_MMC_RSP_PRESENT;
+
+		if (flags & MMC_RSP_136)
+			cb_flags |= CB710_MMC_RSP_136;
+		if (!(flags & MMC_RSP_CRC))
+			cb_flags |= CB710_MMC_RSP_NO_CRC;
+	}
+
+	return cb_flags;
+}
+
+static void cb710_receive_response(struct cb710_slot *slot,
+	struct mmc_command *cmd)
+{
+	unsigned rsp_opcode, wanted_opcode;
+
+	/* Looks like final byte with CRC is always stripped (same as SDHCI) */
+	if (cmd->flags & MMC_RSP_136) {
+		u32 resp[4];
+
+		resp[0] = cb710_read_port_32(slot, CB710_MMC_RESPONSE3_PORT);
+		resp[1] = cb710_read_port_32(slot, CB710_MMC_RESPONSE2_PORT);
+		resp[2] = cb710_read_port_32(slot, CB710_MMC_RESPONSE1_PORT);
+		resp[3] = cb710_read_port_32(slot, CB710_MMC_RESPONSE0_PORT);
+		rsp_opcode = resp[0] >> 24;
+
+		cmd->resp[0] = (resp[0] << 8)|(resp[1] >> 24);
+		cmd->resp[1] = (resp[1] << 8)|(resp[2] >> 24);
+		cmd->resp[2] = (resp[2] << 8)|(resp[3] >> 24);
+		cmd->resp[3] = (resp[3] << 8);
+	} else {
+		rsp_opcode = cb710_read_port_32(slot, CB710_MMC_RESPONSE1_PORT) & 0x3F;
+		cmd->resp[0] = cb710_read_port_32(slot, CB710_MMC_RESPONSE0_PORT);
+	}
+
+	wanted_opcode = (cmd->flags & MMC_RSP_OPCODE) ? cmd->opcode : 0x3F;
+	if (rsp_opcode != wanted_opcode)
+		cmd->error = -EILSEQ;
+}
+
+static int cb710_mmc_transfer_data(struct cb710_slot *slot,
+	struct mmc_data *data)
+{
+	int error, to;
+
+	if (data->flags & MMC_DATA_READ)
+		error = cb710_mmc_receive(slot, data);
+	else
+		error = cb710_mmc_send(slot, data);
+
+	to = cb710_wait_for_event(slot, CB710_MMC_S1_DATA_TRANSFER_DONE);
+	if (!error)
+		error = to;
+
+	if (!error)
+		data->bytes_xfered = data->blksz * data->blocks;
+	return error;
+}
+
+static int cb710_mmc_command(struct mmc_host *mmc, struct mmc_command *cmd)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+	struct mmc_data *data = cmd->data;
+
+	u16 cb_cmd = cb710_encode_cmd_flags(reader, cmd);
+	dev_dbg(cb710_slot_dev(slot), "cmd request: 0x%04X\n", cb_cmd);
+
+	if (data) {
+		if (!cb710_is_transfer_size_supported(data)) {
+			data->error = -EINVAL;
+			return -1;
+		}
+		cb710_mmc_set_transfer_size(slot, data->blocks, data->blksz);
+	}
+
+	cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20|CB710_MMC_S2_BUSY_10);
+	cb710_write_port_16(slot, CB710_MMC_CMD_TYPE_PORT, cb_cmd);
+	cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	cb710_write_port_32(slot, CB710_MMC_CMD_PARAM_PORT, cmd->arg);
+	cb710_mmc_reset_events(slot);
+	cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG0_PORT, 0x01, 0);
+
+	cmd->error = cb710_wait_for_event(slot, CB710_MMC_S1_COMMAND_SENT);
+	if (cmd->error)
+		return -1;
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		cb710_receive_response(slot, cmd);
+		if (cmd->error)
+			return -1;
+	}
+
+	if (data)
+		data->error = cb710_mmc_transfer_data(slot, data);
+	return 0;
+}
+
+static void cb710_mmc_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+
+	WARN_ON(reader->mrq != NULL);
+
+	reader->mrq = mrq;
+	cb710_mmc_enable_irq(slot, CB710_MMC_IE_TEST_MASK, 0);
+
+	if (!cb710_mmc_command(mmc, mrq->cmd) && mrq->stop)
+		cb710_mmc_command(mmc, mrq->stop);
+
+	tasklet_schedule(&reader->finish_req_tasklet);
+}
+
+static int cb710_mmc_powerup(struct cb710_slot *slot)
+{
+#ifdef CONFIG_CB710_DEBUG
+	struct cb710_chip *chip = cb710_slot_to_chip(slot);
+#endif
+	int err;
+
+	/* a lot of magic for now */
+	dev_dbg(cb710_slot_dev(slot), "bus powerup\n");
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	err = cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	if (unlikely(err))
+		return err;
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT, 0x80, 0);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG3_PORT, 0x80, 0);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	mdelay(1);
+	dev_dbg(cb710_slot_dev(slot), "after delay 1\n");
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	err = cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	if (unlikely(err))
+		return err;
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT, 0x09, 0);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	mdelay(1);
+	dev_dbg(cb710_slot_dev(slot), "after delay 2\n");
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	err = cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	if (unlikely(err))
+		return err;
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT, 0, 0x08);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	mdelay(2);
+	dev_dbg(cb710_slot_dev(slot), "after delay 3\n");
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG0_PORT, 0x06, 0);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT, 0x70, 0);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT, 0x80, 0);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG3_PORT, 0x03, 0);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	err = cb710_wait_while_busy(slot, CB710_MMC_S2_BUSY_20);
+	if (unlikely(err))
+		return err;
+	/* This port behaves weird: quick byte reads of 0x08,0x09 return
+	 * 0xFF,0x00 after writing 0xFFFF to 0x08; it works correctly when
+	 * read/written from userspace...  What am I missing here?
+	 * (it doesn't depend on write-to-read delay) */
+	cb710_write_port_16(slot, CB710_MMC_CONFIGB_PORT, 0xFFFF);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG0_PORT, 0x06, 0);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+	dev_dbg(cb710_slot_dev(slot), "bus powerup finished\n");
+
+	return cb710_check_event(slot, 0);
+}
+
+static void cb710_mmc_powerdown(struct cb710_slot *slot)
+{
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG1_PORT, 0, 0x81);
+	cb710_modify_port_8(slot, CB710_MMC_CONFIG3_PORT, 0, 0x80);
+}
+
+static void cb710_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+	int err;
+
+	cb710_mmc_select_clock_divider(mmc, ios->clock);
+
+	if (ios->power_mode != reader->last_power_mode)
+	switch (ios->power_mode) {
+	case MMC_POWER_ON:
+		err = cb710_mmc_powerup(slot);
+		if (err) {
+			dev_warn(cb710_slot_dev(slot),
+				"powerup failed (%d)- retrying\n", err);
+			cb710_mmc_powerdown(slot);
+			udelay(1);
+			err = cb710_mmc_powerup(slot);
+			if (err)
+				dev_warn(cb710_slot_dev(slot),
+					"powerup retry failed (%d) - expect errors\n",
+					err);
+		}
+		reader->last_power_mode = MMC_POWER_ON;
+		break;
+	case MMC_POWER_OFF:
+		cb710_mmc_powerdown(slot);
+		reader->last_power_mode = MMC_POWER_OFF;
+		break;
+	case MMC_POWER_UP:
+	default:
+		/* ignore */;
+	}
+
+	cb710_mmc_enable_4bit_data(slot, ios->bus_width != MMC_BUS_WIDTH_1);
+
+	cb710_mmc_enable_irq(slot, CB710_MMC_IE_TEST_MASK, 0);
+}
+
+static int cb710_mmc_get_ro(struct mmc_host *mmc)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+
+	return cb710_read_port_8(slot, CB710_MMC_STATUS3_PORT)
+		& CB710_MMC_S3_WRITE_PROTECTED;
+}
+
+static int cb710_mmc_get_cd(struct mmc_host *mmc)
+{
+	struct cb710_slot *slot = cb710_mmc_to_slot(mmc);
+
+	return cb710_read_port_8(slot, CB710_MMC_STATUS3_PORT)
+		& CB710_MMC_S3_CARD_DETECTED;
+}
+
+static int cb710_mmc_irq_handler(struct cb710_slot *slot)
+{
+	struct mmc_host *mmc = cb710_slot_to_mmc(slot);
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+	u32 status, config1, config2, irqen;
+
+	status = cb710_read_port_32(slot, CB710_MMC_STATUS_PORT);
+	irqen = cb710_read_port_32(slot, CB710_MMC_IRQ_ENABLE_PORT);
+	config2 = cb710_read_port_32(slot, CB710_MMC_CONFIGB_PORT);
+	config1 = cb710_read_port_32(slot, CB710_MMC_CONFIG_PORT);
+
+	dev_dbg(cb710_slot_dev(slot), "interrupt; status: %08X, "
+		"ie: %08X, c2: %08X, c1: %08X\n",
+		status, irqen, config2, config1);
+
+	if (status & (CB710_MMC_S1_CARD_CHANGED << 8)) {
+		/* ack the event */
+		cb710_write_port_8(slot, CB710_MMC_STATUS1_PORT,
+			CB710_MMC_S1_CARD_CHANGED);
+		if ((irqen & CB710_MMC_IE_CISTATUS_MASK)
+		    == CB710_MMC_IE_CISTATUS_MASK)
+			mmc_detect_change(mmc, HZ/5);
+	} else {
+		dev_dbg(cb710_slot_dev(slot), "unknown interrupt (test)\n");
+		spin_lock(&reader->irq_lock);
+		__cb710_mmc_enable_irq(slot, 0, CB710_MMC_IE_TEST_MASK);
+		spin_unlock(&reader->irq_lock);
+	}
+
+	return 1;
+}
+
+static void cb710_mmc_finish_request_tasklet(unsigned long data)
+{
+	struct mmc_host *mmc = (void *)data;
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+	struct mmc_request *mrq = reader->mrq;
+
+	reader->mrq = NULL;
+	mmc_request_done(mmc, mrq);
+}
+
+static const struct mmc_host_ops cb710_mmc_host = {
+	.request = cb710_mmc_request,
+	.set_ios = cb710_mmc_set_ios,
+	.get_ro = cb710_mmc_get_ro,
+	.get_cd = cb710_mmc_get_cd,
+};
+
+#ifdef CONFIG_PM
+
+static int cb710_mmc_suspend(struct platform_device *pdev, pm_message_t state)
+{
+	struct cb710_slot *slot = cb710_pdev_to_slot(pdev);
+
+	cb710_mmc_enable_irq(slot, 0, ~0);
+	return 0;
+}
+
+static int cb710_mmc_resume(struct platform_device *pdev)
+{
+	struct cb710_slot *slot = cb710_pdev_to_slot(pdev);
+
+	cb710_mmc_enable_irq(slot, 0, ~0);
+	return 0;
+}
+
+#endif /* CONFIG_PM */
+
+static int cb710_mmc_init(struct platform_device *pdev)
+{
+	struct cb710_slot *slot = cb710_pdev_to_slot(pdev);
+	struct cb710_chip *chip = cb710_slot_to_chip(slot);
+	struct mmc_host *mmc;
+	struct cb710_mmc_reader *reader;
+	int err;
+	u32 val;
+
+	mmc = mmc_alloc_host(sizeof(*reader), cb710_slot_dev(slot));
+	if (!mmc)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, mmc);
+
+	/* harmless (maybe) magic */
+	pci_read_config_dword(chip->pdev, 0x48, &val);
+	val = cb710_src_freq_mhz[(val >> 16) & 0xF];
+	dev_dbg(cb710_slot_dev(slot), "source frequency: %dMHz\n", val);
+	val *= 1000000;
+
+	mmc->ops = &cb710_mmc_host;
+	mmc->f_max = val;
+	mmc->f_min = val >> cb710_clock_divider_log2[CB710_MAX_DIVIDER_IDX];
+	mmc->ocr_avail = MMC_VDD_32_33|MMC_VDD_33_34;
+	mmc->caps = MMC_CAP_4_BIT_DATA;
+
+	reader = mmc_priv(mmc);
+
+	tasklet_init(&reader->finish_req_tasklet,
+		cb710_mmc_finish_request_tasklet, (unsigned long)mmc);
+	spin_lock_init(&reader->irq_lock);
+	cb710_dump_regs(chip, CB710_DUMP_REGS_MMC);
+
+	cb710_mmc_enable_irq(slot, 0, ~0);
+	cb710_set_irq_handler(slot, cb710_mmc_irq_handler);
+
+	err = mmc_add_host(mmc);
+	if (unlikely(err))
+		goto err_free_mmc;
+
+	dev_dbg(cb710_slot_dev(slot), "mmc_hostname is %s\n",
+		mmc_hostname(mmc));
+
+	cb710_mmc_enable_irq(slot, CB710_MMC_IE_CARD_INSERTION_STATUS, 0);
+
+	return 0;
+
+err_free_mmc:
+	dev_dbg(cb710_slot_dev(slot), "mmc_add_host() failed: %d\n", err);
+
+	cb710_set_irq_handler(slot, NULL);
+	mmc_free_host(mmc);
+	return err;
+}
+
+static int cb710_mmc_exit(struct platform_device *pdev)
+{
+	struct cb710_slot *slot = cb710_pdev_to_slot(pdev);
+	struct mmc_host *mmc = cb710_slot_to_mmc(slot);
+	struct cb710_mmc_reader *reader = mmc_priv(mmc);
+
+	cb710_mmc_enable_irq(slot, 0, CB710_MMC_IE_CARD_INSERTION_STATUS);
+
+	mmc_remove_host(mmc);
+
+	/* IRQs should be disabled now, but let's stay on the safe side */
+	cb710_mmc_enable_irq(slot, 0, ~0);
+	cb710_set_irq_handler(slot, NULL);
+
+	/* clear config ports - just in case */
+	cb710_write_port_32(slot, CB710_MMC_CONFIG_PORT, 0);
+	cb710_write_port_16(slot, CB710_MMC_CONFIGB_PORT, 0);
+
+	tasklet_kill(&reader->finish_req_tasklet);
+
+	mmc_free_host(mmc);
+	return 0;
+}
+
+static struct platform_driver cb710_mmc_driver = {
+	.driver.name = "cb710-mmc",
+	.probe = cb710_mmc_init,
+	.remove = cb710_mmc_exit,
+#ifdef CONFIG_PM
+	.suspend = cb710_mmc_suspend,
+	.resume = cb710_mmc_resume,
+#endif
+};
+
+module_platform_driver(cb710_mmc_driver);
+
+MODULE_AUTHOR("Michał Mirosław <mirq-linux@rere.qmqm.pl>");
+MODULE_DESCRIPTION("ENE CB710 memory card reader driver - MMC/SD part");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:cb710-mmc");
diff --git a/drivers/mmc/host/cb710-mmc.h b/drivers/mmc/host/cb710-mmc.h
new file mode 100644
index 00000000000..8984ec878fc
--- /dev/null
+++ b/drivers/mmc/host/cb710-mmc.h
@@ -0,0 +1,104 @@
+/*
+ *  cb710/cb710-mmc.h
+ *
+ *  Copyright by Michał Mirosław, 2008-2009
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#ifndef LINUX_CB710_MMC_H
+#define LINUX_CB710_MMC_H
+
+#include <linux/cb710.h>
+
+/* per-MMC-reader structure */
+struct cb710_mmc_reader {
+	struct tasklet_struct finish_req_tasklet;
+	struct mmc_request *mrq;
+	spinlock_t irq_lock;
+	unsigned char last_power_mode;
+};
+
+/* some device struct walking */
+
+static inline struct mmc_host *cb710_slot_to_mmc(struct cb710_slot *slot)
+{
+	return platform_get_drvdata(&slot->pdev);
+}
+
+static inline struct cb710_slot *cb710_mmc_to_slot(struct mmc_host *mmc)
+{
+	struct platform_device *pdev = container_of(mmc_dev(mmc),
+		struct platform_device, dev);
+	return cb710_pdev_to_slot(pdev);
+}
+
+/* registers (this might be all wrong ;) */
+
+#define CB710_MMC_DATA_PORT		0x00
+
+#define CB710_MMC_CONFIG_PORT		0x04
+#define CB710_MMC_CONFIG0_PORT		0x04
+#define CB710_MMC_CONFIG1_PORT		0x05
+#define   CB710_MMC_C1_4BIT_DATA_BUS		0x40
+#define CB710_MMC_CONFIG2_PORT		0x06
+#define   CB710_MMC_C2_READ_PIO_SIZE_MASK	0x0F	/* N-1 */
+#define CB710_MMC_CONFIG3_PORT		0x07
+
+#define CB710_MMC_CONFIGB_PORT		0x08
+
+#define CB710_MMC_IRQ_ENABLE_PORT	0x0C
+#define   CB710_MMC_IE_TEST_MASK		0x00BF
+#define   CB710_MMC_IE_CARD_INSERTION_STATUS	0x1000
+#define   CB710_MMC_IE_IRQ_ENABLE		0x8000
+#define   CB710_MMC_IE_CISTATUS_MASK		\
+		(CB710_MMC_IE_CARD_INSERTION_STATUS|CB710_MMC_IE_IRQ_ENABLE)
+
+#define CB710_MMC_STATUS_PORT		0x10
+#define   CB710_MMC_STATUS_ERROR_EVENTS		0x60FF
+#define CB710_MMC_STATUS0_PORT		0x10
+#define   CB710_MMC_S0_FIFO_UNDERFLOW		0x40
+#define CB710_MMC_STATUS1_PORT		0x11
+#define   CB710_MMC_S1_COMMAND_SENT		0x01
+#define   CB710_MMC_S1_DATA_TRANSFER_DONE	0x02
+#define   CB710_MMC_S1_PIO_TRANSFER_DONE	0x04
+#define   CB710_MMC_S1_CARD_CHANGED		0x10
+#define   CB710_MMC_S1_RESET			0x20
+#define CB710_MMC_STATUS2_PORT		0x12
+#define   CB710_MMC_S2_FIFO_READY		0x01
+#define   CB710_MMC_S2_FIFO_EMPTY		0x02
+#define   CB710_MMC_S2_BUSY_10			0x10
+#define   CB710_MMC_S2_BUSY_20			0x20
+#define CB710_MMC_STATUS3_PORT		0x13
+#define   CB710_MMC_S3_CARD_DETECTED		0x02
+#define   CB710_MMC_S3_WRITE_PROTECTED		0x04
+
+#define CB710_MMC_CMD_TYPE_PORT		0x14
+#define   CB710_MMC_RSP_TYPE_MASK		0x0007
+#define     CB710_MMC_RSP_R1			(0)
+#define     CB710_MMC_RSP_136			(5)
+#define     CB710_MMC_RSP_NO_CRC		(2)
+#define   CB710_MMC_RSP_PRESENT_MASK		0x0018
+#define     CB710_MMC_RSP_NONE			(0 << 3)
+#define     CB710_MMC_RSP_PRESENT		(1 << 3)
+#define     CB710_MMC_RSP_PRESENT_X		(2 << 3)
+#define   CB710_MMC_CMD_TYPE_MASK		0x0060
+#define     CB710_MMC_CMD_BC			(0 << 5)
+#define     CB710_MMC_CMD_BCR			(1 << 5)
+#define     CB710_MMC_CMD_AC			(2 << 5)
+#define     CB710_MMC_CMD_ADTC			(3 << 5)
+#define   CB710_MMC_DATA_READ			0x0080
+#define   CB710_MMC_CMD_CODE_MASK		0x3F00
+#define   CB710_MMC_CMD_CODE_SHIFT		8
+#define   CB710_MMC_IS_APP_CMD			0x4000
+#define   CB710_MMC_RSP_BUSY			0x8000
+
+#define CB710_MMC_CMD_PARAM_PORT	0x18
+#define CB710_MMC_TRANSFER_SIZE_PORT	0x1C
+#define CB710_MMC_RESPONSE0_PORT	0x20
+#define CB710_MMC_RESPONSE1_PORT	0x24
+#define CB710_MMC_RESPONSE2_PORT	0x28
+#define CB710_MMC_RESPONSE3_PORT	0x2C
+
+#endif /* LINUX_CB710_MMC_H */
diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
new file mode 100644
index 00000000000..5d4c5e0fba2
--- /dev/null
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -0,0 +1,1483 @@
+/*
+ * davinci_mmc.c - TI DaVinci MMC/SD/SDIO driver
+ *
+ * Copyright (C) 2006 Texas Instruments.
+ *       Original author: Purushotam Kumar
+ * Copyright (C) 2009 David Brownell
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/module.h>
+#include <linux/ioport.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/cpufreq.h>
+#include <linux/mmc/host.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/edma.h>
+#include <linux/mmc/mmc.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+
+#include <linux/platform_data/edma.h>
+#include <linux/platform_data/mmc-davinci.h>
+
+/*
+ * Register Definitions
+ */
+#define DAVINCI_MMCCTL       0x00 /* Control Register                  */
+#define DAVINCI_MMCCLK       0x04 /* Memory Clock Control Register     */
+#define DAVINCI_MMCST0       0x08 /* Status Register 0                 */
+#define DAVINCI_MMCST1       0x0C /* Status Register 1                 */
+#define DAVINCI_MMCIM        0x10 /* Interrupt Mask Register           */
+#define DAVINCI_MMCTOR       0x14 /* Response Time-Out Register        */
+#define DAVINCI_MMCTOD       0x18 /* Data Read Time-Out Register       */
+#define DAVINCI_MMCBLEN      0x1C /* Block Length Register             */
+#define DAVINCI_MMCNBLK      0x20 /* Number of Blocks Register         */
+#define DAVINCI_MMCNBLC      0x24 /* Number of Blocks Counter Register */
+#define DAVINCI_MMCDRR       0x28 /* Data Receive Register             */
+#define DAVINCI_MMCDXR       0x2C /* Data Transmit Register            */
+#define DAVINCI_MMCCMD       0x30 /* Command Register                  */
+#define DAVINCI_MMCARGHL     0x34 /* Argument Register                 */
+#define DAVINCI_MMCRSP01     0x38 /* Response Register 0 and 1         */
+#define DAVINCI_MMCRSP23     0x3C /* Response Register 0 and 1         */
+#define DAVINCI_MMCRSP45     0x40 /* Response Register 0 and 1         */
+#define DAVINCI_MMCRSP67     0x44 /* Response Register 0 and 1         */
+#define DAVINCI_MMCDRSP      0x48 /* Data Response Register            */
+#define DAVINCI_MMCETOK      0x4C
+#define DAVINCI_MMCCIDX      0x50 /* Command Index Register            */
+#define DAVINCI_MMCCKC       0x54
+#define DAVINCI_MMCTORC      0x58
+#define DAVINCI_MMCTODC      0x5C
+#define DAVINCI_MMCBLNC      0x60
+#define DAVINCI_SDIOCTL      0x64
+#define DAVINCI_SDIOST0      0x68
+#define DAVINCI_SDIOIEN      0x6C
+#define DAVINCI_SDIOIST      0x70
+#define DAVINCI_MMCFIFOCTL   0x74 /* FIFO Control Register             */
+
+/* DAVINCI_MMCCTL definitions */
+#define MMCCTL_DATRST         (1 << 0)
+#define MMCCTL_CMDRST         (1 << 1)
+#define MMCCTL_WIDTH_8_BIT    (1 << 8)
+#define MMCCTL_WIDTH_4_BIT    (1 << 2)
+#define MMCCTL_DATEG_DISABLED (0 << 6)
+#define MMCCTL_DATEG_RISING   (1 << 6)
+#define MMCCTL_DATEG_FALLING  (2 << 6)
+#define MMCCTL_DATEG_BOTH     (3 << 6)
+#define MMCCTL_PERMDR_LE      (0 << 9)
+#define MMCCTL_PERMDR_BE      (1 << 9)
+#define MMCCTL_PERMDX_LE      (0 << 10)
+#define MMCCTL_PERMDX_BE      (1 << 10)
+
+/* DAVINCI_MMCCLK definitions */
+#define MMCCLK_CLKEN          (1 << 8)
+#define MMCCLK_CLKRT_MASK     (0xFF << 0)
+
+/* IRQ bit definitions, for DAVINCI_MMCST0 and DAVINCI_MMCIM */
+#define MMCST0_DATDNE         BIT(0)	/* data done */
+#define MMCST0_BSYDNE         BIT(1)	/* busy done */
+#define MMCST0_RSPDNE         BIT(2)	/* command done */
+#define MMCST0_TOUTRD         BIT(3)	/* data read timeout */
+#define MMCST0_TOUTRS         BIT(4)	/* command response timeout */
+#define MMCST0_CRCWR          BIT(5)	/* data write CRC error */
+#define MMCST0_CRCRD          BIT(6)	/* data read CRC error */
+#define MMCST0_CRCRS          BIT(7)	/* command response CRC error */
+#define MMCST0_DXRDY          BIT(9)	/* data transmit ready (fifo empty) */
+#define MMCST0_DRRDY          BIT(10)	/* data receive ready (data in fifo)*/
+#define MMCST0_DATED          BIT(11)	/* DAT3 edge detect */
+#define MMCST0_TRNDNE         BIT(12)	/* transfer done */
+
+/* DAVINCI_MMCST1 definitions */
+#define MMCST1_BUSY           (1 << 0)
+
+/* DAVINCI_MMCCMD definitions */
+#define MMCCMD_CMD_MASK       (0x3F << 0)
+#define MMCCMD_PPLEN          (1 << 7)
+#define MMCCMD_BSYEXP         (1 << 8)
+#define MMCCMD_RSPFMT_MASK    (3 << 9)
+#define MMCCMD_RSPFMT_NONE    (0 << 9)
+#define MMCCMD_RSPFMT_R1456   (1 << 9)
+#define MMCCMD_RSPFMT_R2      (2 << 9)
+#define MMCCMD_RSPFMT_R3      (3 << 9)
+#define MMCCMD_DTRW           (1 << 11)
+#define MMCCMD_STRMTP         (1 << 12)
+#define MMCCMD_WDATX          (1 << 13)
+#define MMCCMD_INITCK         (1 << 14)
+#define MMCCMD_DCLR           (1 << 15)
+#define MMCCMD_DMATRIG        (1 << 16)
+
+/* DAVINCI_MMCFIFOCTL definitions */
+#define MMCFIFOCTL_FIFORST    (1 << 0)
+#define MMCFIFOCTL_FIFODIR_WR (1 << 1)
+#define MMCFIFOCTL_FIFODIR_RD (0 << 1)
+#define MMCFIFOCTL_FIFOLEV    (1 << 2) /* 0 = 128 bits, 1 = 256 bits */
+#define MMCFIFOCTL_ACCWD_4    (0 << 3) /* access width of 4 bytes    */
+#define MMCFIFOCTL_ACCWD_3    (1 << 3) /* access width of 3 bytes    */
+#define MMCFIFOCTL_ACCWD_2    (2 << 3) /* access width of 2 bytes    */
+#define MMCFIFOCTL_ACCWD_1    (3 << 3) /* access width of 1 byte     */
+
+/* DAVINCI_SDIOST0 definitions */
+#define SDIOST0_DAT1_HI       BIT(0)
+
+/* DAVINCI_SDIOIEN definitions */
+#define SDIOIEN_IOINTEN       BIT(0)
+
+/* DAVINCI_SDIOIST definitions */
+#define SDIOIST_IOINT         BIT(0)
+
+/* MMCSD Init clock in Hz in opendrain mode */
+#define MMCSD_INIT_CLOCK		200000
+
+/*
+ * One scatterlist dma "segment" is at most MAX_CCNT rw_threshold units,
+ * and we handle up to MAX_NR_SG segments.  MMC_BLOCK_BOUNCE kicks in only
+ * for drivers with max_segs == 1, making the segments bigger (64KB)
+ * than the page or two that's otherwise typical. nr_sg (passed from
+ * platform data) == 16 gives at least the same throughput boost, using
+ * EDMA transfer linkage instead of spending CPU time copying pages.
+ */
+#define MAX_CCNT	((1 << 16) - 1)
+
+#define MAX_NR_SG	16
+
+static unsigned rw_threshold = 32;
+module_param(rw_threshold, uint, S_IRUGO);
+MODULE_PARM_DESC(rw_threshold,
+		"Read/Write threshold. Default = 32");
+
+static unsigned poll_threshold = 128;
+module_param(poll_threshold, uint, S_IRUGO);
+MODULE_PARM_DESC(poll_threshold,
+		 "Polling transaction size threshold. Default = 128");
+
+static unsigned poll_loopcount = 32;
+module_param(poll_loopcount, uint, S_IRUGO);
+MODULE_PARM_DESC(poll_loopcount,
+		 "Maximum polling loop count. Default = 32");
+
+static unsigned __initdata use_dma = 1;
+module_param(use_dma, uint, 0);
+MODULE_PARM_DESC(use_dma, "Whether to use DMA or not. Default = 1");
+
+struct mmc_davinci_host {
+	struct mmc_command *cmd;
+	struct mmc_data *data;
+	struct mmc_host *mmc;
+	struct clk *clk;
+	unsigned int mmc_input_clk;
+	void __iomem *base;
+	struct resource *mem_res;
+	int mmc_irq, sdio_irq;
+	unsigned char bus_mode;
+
+#define DAVINCI_MMC_DATADIR_NONE	0
+#define DAVINCI_MMC_DATADIR_READ	1
+#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;
+
+	u32 rxdma, txdma;
+	struct dma_chan *dma_tx;
+	struct dma_chan *dma_rx;
+	bool use_dma;
+	bool do_dma;
+	bool sdio_int;
+	bool active_request;
+
+	/* For PIO we walk scatterlists one segment at a time. */
+	unsigned int		sg_len;
+	struct scatterlist *sg;
+
+	/* Version of the MMC/SD controller */
+	u8 version;
+	/* for ns in one cycle calculation */
+	unsigned ns_in_one_cycle;
+	/* Number of sg segments */
+	u8 nr_sg;
+#ifdef CONFIG_CPU_FREQ
+	struct notifier_block	freq_transition;
+#endif
+};
+
+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)
+{
+	u8 *p;
+	unsigned int i;
+
+	if (host->buffer_bytes_left == 0) {
+		host->sg = sg_next(host->data->sg);
+		mmc_davinci_sg_to_buf(host);
+	}
+
+	p = host->buffer;
+	if (n > host->buffer_bytes_left)
+		n = host->buffer_bytes_left;
+	host->buffer_bytes_left -= n;
+	host->bytes_left -= n;
+
+	/* NOTE:  we never transfer more than rw_threshold bytes
+	 * to/from the fifo here; there's no I/O overlap.
+	 * This also assumes that access width( i.e. ACCWD) is 4 bytes
+	 */
+	if (host->data_dir == DAVINCI_MMC_DATADIR_WRITE) {
+		for (i = 0; i < (n >> 2); i++) {
+			writel(*((u32 *)p), host->base + DAVINCI_MMCDXR);
+			p = p + 4;
+		}
+		if (n & 3) {
+			iowrite8_rep(host->base + DAVINCI_MMCDXR, p, (n & 3));
+			p = p + (n & 3);
+		}
+	} else {
+		for (i = 0; i < (n >> 2); i++) {
+			*((u32 *)p) = readl(host->base + DAVINCI_MMCDRR);
+			p  = p + 4;
+		}
+		if (n & 3) {
+			ioread8_rep(host->base + DAVINCI_MMCDRR, p, (n & 3));
+			p = p + (n & 3);
+		}
+	}
+	host->buffer = p;
+}
+
+static void mmc_davinci_start_command(struct mmc_davinci_host *host,
+		struct mmc_command *cmd)
+{
+	u32 cmd_reg = 0;
+	u32 im_val;
+
+	dev_dbg(mmc_dev(host->mmc), "CMD%d, arg 0x%08x%s\n",
+		cmd->opcode, cmd->arg,
+		({ char *s;
+		switch (mmc_resp_type(cmd)) {
+		case MMC_RSP_R1:
+			s = ", R1/R5/R6/R7 response";
+			break;
+		case MMC_RSP_R1B:
+			s = ", R1b response";
+			break;
+		case MMC_RSP_R2:
+			s = ", R2 response";
+			break;
+		case MMC_RSP_R3:
+			s = ", R3/R4 response";
+			break;
+		default:
+			s = ", (R? response)";
+			break;
+		}; s; }));
+	host->cmd = cmd;
+
+	switch (mmc_resp_type(cmd)) {
+	case MMC_RSP_R1B:
+		/* There's some spec confusion about when R1B is
+		 * allowed, but if the card doesn't issue a BUSY
+		 * then it's harmless for us to allow it.
+		 */
+		cmd_reg |= MMCCMD_BSYEXP;
+		/* FALLTHROUGH */
+	case MMC_RSP_R1:		/* 48 bits, CRC */
+		cmd_reg |= MMCCMD_RSPFMT_R1456;
+		break;
+	case MMC_RSP_R2:		/* 136 bits, CRC */
+		cmd_reg |= MMCCMD_RSPFMT_R2;
+		break;
+	case MMC_RSP_R3:		/* 48 bits, no CRC */
+		cmd_reg |= MMCCMD_RSPFMT_R3;
+		break;
+	default:
+		cmd_reg |= MMCCMD_RSPFMT_NONE;
+		dev_dbg(mmc_dev(host->mmc), "unknown resp_type %04x\n",
+			mmc_resp_type(cmd));
+		break;
+	}
+
+	/* Set command index */
+	cmd_reg |= cmd->opcode;
+
+	/* Enable EDMA transfer triggers */
+	if (host->do_dma)
+		cmd_reg |= MMCCMD_DMATRIG;
+
+	if (host->version == MMC_CTLR_VERSION_2 && host->data != NULL &&
+			host->data_dir == DAVINCI_MMC_DATADIR_READ)
+		cmd_reg |= MMCCMD_DMATRIG;
+
+	/* Setting whether command involves data transfer or not */
+	if (cmd->data)
+		cmd_reg |= MMCCMD_WDATX;
+
+	/* Setting whether stream or block transfer */
+	if (cmd->flags & MMC_DATA_STREAM)
+		cmd_reg |= MMCCMD_STRMTP;
+
+	/* Setting whether data read or write */
+	if (host->data_dir == DAVINCI_MMC_DATADIR_WRITE)
+		cmd_reg |= MMCCMD_DTRW;
+
+	if (host->bus_mode == MMC_BUSMODE_PUSHPULL)
+		cmd_reg |= MMCCMD_PPLEN;
+
+	/* set Command timeout */
+	writel(0x1FFF, host->base + DAVINCI_MMCTOR);
+
+	/* Enable interrupt (calculate here, defer until FIFO is stuffed). */
+	im_val =  MMCST0_RSPDNE | MMCST0_CRCRS | MMCST0_TOUTRS;
+	if (host->data_dir == DAVINCI_MMC_DATADIR_WRITE) {
+		im_val |= MMCST0_DATDNE | MMCST0_CRCWR;
+
+		if (!host->do_dma)
+			im_val |= MMCST0_DXRDY;
+	} else if (host->data_dir == DAVINCI_MMC_DATADIR_READ) {
+		im_val |= MMCST0_DATDNE | MMCST0_CRCRD | MMCST0_TOUTRD;
+
+		if (!host->do_dma)
+			im_val |= MMCST0_DRRDY;
+	}
+
+	/*
+	 * Before non-DMA WRITE commands the controller needs priming:
+	 * FIFO should be populated with 32 bytes i.e. whatever is the FIFO size
+	 */
+	if (!host->do_dma && (host->data_dir == DAVINCI_MMC_DATADIR_WRITE))
+		davinci_fifo_data_trans(host, rw_threshold);
+
+	writel(cmd->arg, host->base + DAVINCI_MMCARGHL);
+	writel(cmd_reg,  host->base + DAVINCI_MMCCMD);
+
+	host->active_request = true;
+
+	if (!host->do_dma && host->bytes_left <= poll_threshold) {
+		u32 count = poll_loopcount;
+
+		while (host->active_request && count--) {
+			mmc_davinci_irq(0, host);
+			cpu_relax();
+		}
+	}
+
+	if (host->active_request)
+		writel(im_val, host->base + DAVINCI_MMCIM);
+}
+
+/*----------------------------------------------------------------------*/
+
+/* DMA infrastructure */
+
+static void davinci_abort_dma(struct mmc_davinci_host *host)
+{
+	struct dma_chan *sync_dev;
+
+	if (host->data_dir == DAVINCI_MMC_DATADIR_READ)
+		sync_dev = host->dma_rx;
+	else
+		sync_dev = host->dma_tx;
+
+	dmaengine_terminate_all(sync_dev);
+}
+
+static int mmc_davinci_send_dma_request(struct mmc_davinci_host *host,
+		struct mmc_data *data)
+{
+	struct dma_chan *chan;
+	struct dma_async_tx_descriptor *desc;
+	int ret = 0;
+
+	if (host->data_dir == DAVINCI_MMC_DATADIR_WRITE) {
+		struct dma_slave_config dma_tx_conf = {
+			.direction = DMA_MEM_TO_DEV,
+			.dst_addr = host->mem_res->start + DAVINCI_MMCDXR,
+			.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
+			.dst_maxburst =
+				rw_threshold / DMA_SLAVE_BUSWIDTH_4_BYTES,
+		};
+		chan = host->dma_tx;
+		dmaengine_slave_config(host->dma_tx, &dma_tx_conf);
+
+		desc = dmaengine_prep_slave_sg(host->dma_tx,
+				data->sg,
+				host->sg_len,
+				DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc) {
+			dev_dbg(mmc_dev(host->mmc),
+				"failed to allocate DMA TX descriptor");
+			ret = -1;
+			goto out;
+		}
+	} else {
+		struct dma_slave_config dma_rx_conf = {
+			.direction = DMA_DEV_TO_MEM,
+			.src_addr = host->mem_res->start + DAVINCI_MMCDRR,
+			.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
+			.src_maxburst =
+				rw_threshold / DMA_SLAVE_BUSWIDTH_4_BYTES,
+		};
+		chan = host->dma_rx;
+		dmaengine_slave_config(host->dma_rx, &dma_rx_conf);
+
+		desc = dmaengine_prep_slave_sg(host->dma_rx,
+				data->sg,
+				host->sg_len,
+				DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc) {
+			dev_dbg(mmc_dev(host->mmc),
+				"failed to allocate DMA RX descriptor");
+			ret = -1;
+			goto out;
+		}
+	}
+
+	dmaengine_submit(desc);
+	dma_async_issue_pending(chan);
+
+out:
+	return ret;
+}
+
+static int mmc_davinci_start_dma_transfer(struct mmc_davinci_host *host,
+		struct mmc_data *data)
+{
+	int i;
+	int mask = rw_threshold - 1;
+	int ret = 0;
+
+	host->sg_len = dma_map_sg(mmc_dev(host->mmc), data->sg, data->sg_len,
+				((data->flags & MMC_DATA_WRITE)
+				? DMA_TO_DEVICE
+				: DMA_FROM_DEVICE));
+
+	/* no individual DMA segment should need a partial FIFO */
+	for (i = 0; i < host->sg_len; i++) {
+		if (sg_dma_len(data->sg + i) & mask) {
+			dma_unmap_sg(mmc_dev(host->mmc),
+					data->sg, data->sg_len,
+					(data->flags & MMC_DATA_WRITE)
+					? DMA_TO_DEVICE
+					: DMA_FROM_DEVICE);
+			return -1;
+		}
+	}
+
+	host->do_dma = 1;
+	ret = mmc_davinci_send_dma_request(host, data);
+
+	return ret;
+}
+
+static void __init_or_module
+davinci_release_dma_channels(struct mmc_davinci_host *host)
+{
+	if (!host->use_dma)
+		return;
+
+	dma_release_channel(host->dma_tx);
+	dma_release_channel(host->dma_rx);
+}
+
+static int __init davinci_acquire_dma_channels(struct mmc_davinci_host *host)
+{
+	int r;
+	dma_cap_mask_t mask;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	host->dma_tx =
+		dma_request_slave_channel_compat(mask, edma_filter_fn,
+				&host->txdma, mmc_dev(host->mmc), "tx");
+	if (!host->dma_tx) {
+		dev_err(mmc_dev(host->mmc), "Can't get dma_tx channel\n");
+		return -ENODEV;
+	}
+
+	host->dma_rx =
+		dma_request_slave_channel_compat(mask, edma_filter_fn,
+				&host->rxdma, mmc_dev(host->mmc), "rx");
+	if (!host->dma_rx) {
+		dev_err(mmc_dev(host->mmc), "Can't get dma_rx channel\n");
+		r = -ENODEV;
+		goto free_master_write;
+	}
+
+	return 0;
+
+free_master_write:
+	dma_release_channel(host->dma_tx);
+
+	return r;
+}
+
+/*----------------------------------------------------------------------*/
+
+static void
+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;
+
+	if (host->version == MMC_CTLR_VERSION_2)
+		fifo_lev = (rw_threshold == 64) ? MMCFIFOCTL_FIFOLEV : 0;
+
+	host->data = data;
+	if (data == NULL) {
+		host->data_dir = DAVINCI_MMC_DATADIR_NONE;
+		writel(0, host->base + DAVINCI_MMCBLEN);
+		writel(0, host->base + DAVINCI_MMCNBLK);
+		return;
+	}
+
+	dev_dbg(mmc_dev(host->mmc), "%s %s, %d blocks of %d bytes\n",
+		(data->flags & MMC_DATA_STREAM) ? "stream" : "block",
+		(data->flags & MMC_DATA_WRITE) ? "write" : "read",
+		data->blocks, data->blksz);
+	dev_dbg(mmc_dev(host->mmc), "  DTO %d cycles + %d ns\n",
+		data->timeout_clks, data->timeout_ns);
+	timeout = data->timeout_clks +
+		(data->timeout_ns / host->ns_in_one_cycle);
+	if (timeout > 0xffff)
+		timeout = 0xffff;
+
+	writel(timeout, host->base + DAVINCI_MMCTOD);
+	writel(data->blocks, host->base + DAVINCI_MMCNBLK);
+	writel(data->blksz, host->base + DAVINCI_MMCBLEN);
+
+	/* Configure the FIFO */
+	switch (data->flags & MMC_DATA_WRITE) {
+	case MMC_DATA_WRITE:
+		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);
+		break;
+
+	default:
+		host->data_dir = DAVINCI_MMC_DATADIR_READ;
+		writel(fifo_lev | MMCFIFOCTL_FIFODIR_RD | MMCFIFOCTL_FIFORST,
+			host->base + DAVINCI_MMCFIFOCTL);
+		writel(fifo_lev | MMCFIFOCTL_FIFODIR_RD,
+			host->base + DAVINCI_MMCFIFOCTL);
+		break;
+	}
+
+	host->buffer = NULL;
+	host->bytes_left = data->blocks * data->blksz;
+
+	/* For now we try to use DMA whenever we won't need partial FIFO
+	 * reads or writes, either for the whole transfer (as tested here)
+	 * or for any individual scatterlist segment (tested when we call
+	 * start_dma_transfer).
+	 *
+	 * While we *could* change that, unusual block sizes are rarely
+	 * used.  The occasional fallback to PIO should't hurt.
+	 */
+	if (host->use_dma && (host->bytes_left & (rw_threshold - 1)) == 0
+			&& mmc_davinci_start_dma_transfer(host, data) == 0) {
+		/* zero this to ensure we take no PIO paths */
+		host->bytes_left = 0;
+	} else {
+		/* Revert to CPU Copy */
+		host->sg_len = data->sg_len;
+		host->sg = host->data->sg;
+		mmc_davinci_sg_to_buf(host);
+	}
+}
+
+static void mmc_davinci_request(struct mmc_host *mmc, struct mmc_request *req)
+{
+	struct mmc_davinci_host *host = mmc_priv(mmc);
+	unsigned long timeout = jiffies + msecs_to_jiffies(900);
+	u32 mmcst1 = 0;
+
+	/* Card may still be sending BUSY after a previous operation,
+	 * typically some kind of write.  If so, we can't proceed yet.
+	 */
+	while (time_before(jiffies, timeout)) {
+		mmcst1  = readl(host->base + DAVINCI_MMCST1);
+		if (!(mmcst1 & MMCST1_BUSY))
+			break;
+		cpu_relax();
+	}
+	if (mmcst1 & MMCST1_BUSY) {
+		dev_err(mmc_dev(host->mmc), "still BUSY? bad ... \n");
+		req->cmd->error = -ETIMEDOUT;
+		mmc_request_done(mmc, req);
+		return;
+	}
+
+	host->do_dma = 0;
+	mmc_davinci_prepare_data(host, req);
+	mmc_davinci_start_command(host, req->cmd);
+}
+
+static unsigned int calculate_freq_for_card(struct mmc_davinci_host *host,
+	unsigned int mmc_req_freq)
+{
+	unsigned int mmc_freq = 0, mmc_pclk = 0, mmc_push_pull_divisor = 0;
+
+	mmc_pclk = host->mmc_input_clk;
+	if (mmc_req_freq && mmc_pclk > (2 * mmc_req_freq))
+		mmc_push_pull_divisor = ((unsigned int)mmc_pclk
+				/ (2 * mmc_req_freq)) - 1;
+	else
+		mmc_push_pull_divisor = 0;
+
+	mmc_freq = (unsigned int)mmc_pclk
+		/ (2 * (mmc_push_pull_divisor + 1));
+
+	if (mmc_freq > mmc_req_freq)
+		mmc_push_pull_divisor = mmc_push_pull_divisor + 1;
+	/* Convert ns to clock cycles */
+	if (mmc_req_freq <= 400000)
+		host->ns_in_one_cycle = (1000000) / (((mmc_pclk
+				/ (2 * (mmc_push_pull_divisor + 1)))/1000));
+	else
+		host->ns_in_one_cycle = (1000000) / (((mmc_pclk
+				/ (2 * (mmc_push_pull_divisor + 1)))/1000000));
+
+	return mmc_push_pull_divisor;
+}
+
+static void calculate_clk_divider(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	unsigned int open_drain_freq = 0, mmc_pclk = 0;
+	unsigned int mmc_push_pull_freq = 0;
+	struct mmc_davinci_host *host = mmc_priv(mmc);
+
+	if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN) {
+		u32 temp;
+
+		/* Ignoring the init clock value passed for fixing the inter
+		 * operability with different cards.
+		 */
+		open_drain_freq = ((unsigned int)mmc_pclk
+				/ (2 * MMCSD_INIT_CLOCK)) - 1;
+
+		if (open_drain_freq > 0xFF)
+			open_drain_freq = 0xFF;
+
+		temp = readl(host->base + DAVINCI_MMCCLK) & ~MMCCLK_CLKRT_MASK;
+		temp |= open_drain_freq;
+		writel(temp, host->base + DAVINCI_MMCCLK);
+
+		/* Convert ns to clock cycles */
+		host->ns_in_one_cycle = (1000000) / (MMCSD_INIT_CLOCK/1000);
+	} else {
+		u32 temp;
+		mmc_push_pull_freq = calculate_freq_for_card(host, ios->clock);
+
+		if (mmc_push_pull_freq > 0xFF)
+			mmc_push_pull_freq = 0xFF;
+
+		temp = readl(host->base + DAVINCI_MMCCLK) & ~MMCCLK_CLKEN;
+		writel(temp, host->base + DAVINCI_MMCCLK);
+
+		udelay(10);
+
+		temp = readl(host->base + DAVINCI_MMCCLK) & ~MMCCLK_CLKRT_MASK;
+		temp |= mmc_push_pull_freq;
+		writel(temp, host->base + DAVINCI_MMCCLK);
+
+		writel(temp | MMCCLK_CLKEN, host->base + DAVINCI_MMCCLK);
+
+		udelay(10);
+	}
+}
+
+static void mmc_davinci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct mmc_davinci_host *host = mmc_priv(mmc);
+	struct platform_device *pdev = to_platform_device(mmc->parent);
+	struct davinci_mmc_config *config = pdev->dev.platform_data;
+
+	dev_dbg(mmc_dev(host->mmc),
+		"clock %dHz busmode %d powermode %d Vdd %04x\n",
+		ios->clock, ios->bus_mode, ios->power_mode,
+		ios->vdd);
+
+	switch (ios->power_mode) {
+	case MMC_POWER_OFF:
+		if (config && config->set_power)
+			config->set_power(pdev->id, false);
+		break;
+	case MMC_POWER_UP:
+		if (config && config->set_power)
+			config->set_power(pdev->id, true);
+		break;
+	}
+
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_8:
+		dev_dbg(mmc_dev(host->mmc), "Enabling 8 bit mode\n");
+		writel((readl(host->base + DAVINCI_MMCCTL) &
+			~MMCCTL_WIDTH_4_BIT) | MMCCTL_WIDTH_8_BIT,
+			host->base + DAVINCI_MMCCTL);
+		break;
+	case MMC_BUS_WIDTH_4:
+		dev_dbg(mmc_dev(host->mmc), "Enabling 4 bit mode\n");
+		if (host->version == MMC_CTLR_VERSION_2)
+			writel((readl(host->base + DAVINCI_MMCCTL) &
+				~MMCCTL_WIDTH_8_BIT) | MMCCTL_WIDTH_4_BIT,
+				host->base + DAVINCI_MMCCTL);
+		else
+			writel(readl(host->base + DAVINCI_MMCCTL) |
+				MMCCTL_WIDTH_4_BIT,
+				host->base + DAVINCI_MMCCTL);
+		break;
+	case MMC_BUS_WIDTH_1:
+		dev_dbg(mmc_dev(host->mmc), "Enabling 1 bit mode\n");
+		if (host->version == MMC_CTLR_VERSION_2)
+			writel(readl(host->base + DAVINCI_MMCCTL) &
+				~(MMCCTL_WIDTH_8_BIT | MMCCTL_WIDTH_4_BIT),
+				host->base + DAVINCI_MMCCTL);
+		else
+			writel(readl(host->base + DAVINCI_MMCCTL) &
+				~MMCCTL_WIDTH_4_BIT,
+				host->base + DAVINCI_MMCCTL);
+		break;
+	}
+
+	calculate_clk_divider(mmc, ios);
+
+	host->bus_mode = ios->bus_mode;
+	if (ios->power_mode == MMC_POWER_UP) {
+		unsigned long timeout = jiffies + msecs_to_jiffies(50);
+		bool lose = true;
+
+		/* Send clock cycles, poll completion */
+		writel(0, host->base + DAVINCI_MMCARGHL);
+		writel(MMCCMD_INITCK, host->base + DAVINCI_MMCCMD);
+		while (time_before(jiffies, timeout)) {
+			u32 tmp = readl(host->base + DAVINCI_MMCST0);
+
+			if (tmp & MMCST0_RSPDNE) {
+				lose = false;
+				break;
+			}
+			cpu_relax();
+		}
+		if (lose)
+			dev_warn(mmc_dev(host->mmc), "powerup timeout\n");
+	}
+
+	/* FIXME on power OFF, reset things ... */
+}
+
+static void
+mmc_davinci_xfer_done(struct mmc_davinci_host *host, struct mmc_data *data)
+{
+	host->data = NULL;
+
+	if (host->mmc->caps & MMC_CAP_SDIO_IRQ) {
+		/*
+		 * SDIO Interrupt Detection work-around as suggested by
+		 * Davinci Errata (TMS320DM355 Silicon Revision 1.1 Errata
+		 * 2.1.6): Signal SDIO interrupt only if it is enabled by core
+		 */
+		if (host->sdio_int && !(readl(host->base + DAVINCI_SDIOST0) &
+					SDIOST0_DAT1_HI)) {
+			writel(SDIOIST_IOINT, host->base + DAVINCI_SDIOIST);
+			mmc_signal_sdio_irq(host->mmc);
+		}
+	}
+
+	if (host->do_dma) {
+		davinci_abort_dma(host);
+
+		dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len,
+			     (data->flags & MMC_DATA_WRITE)
+			     ? DMA_TO_DEVICE
+			     : DMA_FROM_DEVICE);
+		host->do_dma = false;
+	}
+	host->data_dir = DAVINCI_MMC_DATADIR_NONE;
+
+	if (!data->stop || (host->cmd && host->cmd->error)) {
+		mmc_request_done(host->mmc, data->mrq);
+		writel(0, host->base + DAVINCI_MMCIM);
+		host->active_request = false;
+	} else
+		mmc_davinci_start_command(host, data->stop);
+}
+
+static void mmc_davinci_cmd_done(struct mmc_davinci_host *host,
+				 struct mmc_command *cmd)
+{
+	host->cmd = NULL;
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		if (cmd->flags & MMC_RSP_136) {
+			/* response type 2 */
+			cmd->resp[3] = readl(host->base + DAVINCI_MMCRSP01);
+			cmd->resp[2] = readl(host->base + DAVINCI_MMCRSP23);
+			cmd->resp[1] = readl(host->base + DAVINCI_MMCRSP45);
+			cmd->resp[0] = readl(host->base + DAVINCI_MMCRSP67);
+		} else {
+			/* response types 1, 1b, 3, 4, 5, 6 */
+			cmd->resp[0] = readl(host->base + DAVINCI_MMCRSP67);
+		}
+	}
+
+	if (host->data == NULL || cmd->error) {
+		if (cmd->error == -ETIMEDOUT)
+			cmd->mrq->cmd->retries = 0;
+		mmc_request_done(host->mmc, cmd->mrq);
+		writel(0, host->base + DAVINCI_MMCIM);
+		host->active_request = false;
+	}
+}
+
+static inline void mmc_davinci_reset_ctrl(struct mmc_davinci_host *host,
+								int val)
+{
+	u32 temp;
+
+	temp = readl(host->base + DAVINCI_MMCCTL);
+	if (val)	/* reset */
+		temp |= MMCCTL_CMDRST | MMCCTL_DATRST;
+	else		/* enable */
+		temp &= ~(MMCCTL_CMDRST | MMCCTL_DATRST);
+
+	writel(temp, host->base + DAVINCI_MMCCTL);
+	udelay(10);
+}
+
+static void
+davinci_abort_data(struct mmc_davinci_host *host, struct mmc_data *data)
+{
+	mmc_davinci_reset_ctrl(host, 1);
+	mmc_davinci_reset_ctrl(host, 0);
+}
+
+static irqreturn_t mmc_davinci_sdio_irq(int irq, void *dev_id)
+{
+	struct mmc_davinci_host *host = dev_id;
+	unsigned int status;
+
+	status = readl(host->base + DAVINCI_SDIOIST);
+	if (status & SDIOIST_IOINT) {
+		dev_dbg(mmc_dev(host->mmc),
+			"SDIO interrupt status %x\n", status);
+		writel(status | SDIOIST_IOINT, host->base + DAVINCI_SDIOIST);
+		mmc_signal_sdio_irq(host->mmc);
+	}
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t mmc_davinci_irq(int irq, void *dev_id)
+{
+	struct mmc_davinci_host *host = (struct mmc_davinci_host *)dev_id;
+	unsigned int status, qstatus;
+	int end_command = 0;
+	int end_transfer = 0;
+	struct mmc_data *data = host->data;
+
+	if (host->cmd == NULL && host->data == NULL) {
+		status = readl(host->base + DAVINCI_MMCST0);
+		dev_dbg(mmc_dev(host->mmc),
+			"Spurious interrupt 0x%04x\n", status);
+		/* Disable the interrupt from mmcsd */
+		writel(0, host->base + DAVINCI_MMCIM);
+		return IRQ_NONE;
+	}
+
+	status = readl(host->base + DAVINCI_MMCST0);
+	qstatus = status;
+
+	/* handle FIFO first when using PIO for data.
+	 * bytes_left will decrease to zero as I/O progress and status will
+	 * read zero over iteration because this controller status
+	 * register(MMCST0) reports any status only once and it is cleared
+	 * by read. So, it is not unbouned loop even in the case of
+	 * non-dma.
+	 */
+	if (host->bytes_left && (status & (MMCST0_DXRDY | MMCST0_DRRDY))) {
+		unsigned long im_val;
+
+		/*
+		 * If interrupts fire during the following loop, they will be
+		 * handled by the handler, but the PIC will still buffer these.
+		 * As a result, the handler will be called again to serve these
+		 * needlessly. In order to avoid these spurious interrupts,
+		 * keep interrupts masked during the loop.
+		 */
+		im_val = readl(host->base + DAVINCI_MMCIM);
+		writel(0, host->base + DAVINCI_MMCIM);
+
+		do {
+			davinci_fifo_data_trans(host, rw_threshold);
+			status = readl(host->base + DAVINCI_MMCST0);
+			qstatus |= status;
+		} while (host->bytes_left &&
+			 (status & (MMCST0_DXRDY | MMCST0_DRRDY)));
+
+		/*
+		 * If an interrupt is pending, it is assumed it will fire when
+		 * it is unmasked. This assumption is also taken when the MMCIM
+		 * is first set. Otherwise, writing to MMCIM after reading the
+		 * status is race-prone.
+		 */
+		writel(im_val, host->base + DAVINCI_MMCIM);
+	}
+
+	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);
+			}
+			end_transfer = 1;
+			data->bytes_xfered = data->blocks * data->blksz;
+		} else {
+			dev_err(mmc_dev(host->mmc),
+					"DATDNE with no host->data\n");
+		}
+	}
+
+	if (qstatus & MMCST0_TOUTRD) {
+		/* Read data timeout */
+		data->error = -ETIMEDOUT;
+		end_transfer = 1;
+
+		dev_dbg(mmc_dev(host->mmc),
+			"read data timeout, status %x\n",
+			qstatus);
+
+		davinci_abort_data(host, data);
+	}
+
+	if (qstatus & (MMCST0_CRCWR | MMCST0_CRCRD)) {
+		/* Data CRC error */
+		data->error = -EILSEQ;
+		end_transfer = 1;
+
+		/* NOTE:  this controller uses CRCWR to report both CRC
+		 * errors and timeouts (on writes).  MMCDRSP values are
+		 * only weakly documented, but 0x9f was clearly a timeout
+		 * case and the two three-bit patterns in various SD specs
+		 * (101, 010) aren't part of it ...
+		 */
+		if (qstatus & MMCST0_CRCWR) {
+			u32 temp = readb(host->base + DAVINCI_MMCDRSP);
+
+			if (temp == 0x9f)
+				data->error = -ETIMEDOUT;
+		}
+		dev_dbg(mmc_dev(host->mmc), "data %s %s error\n",
+			(qstatus & MMCST0_CRCWR) ? "write" : "read",
+			(data->error == -ETIMEDOUT) ? "timeout" : "CRC");
+
+		davinci_abort_data(host, data);
+	}
+
+	if (qstatus & MMCST0_TOUTRS) {
+		/* Command timeout */
+		if (host->cmd) {
+			dev_dbg(mmc_dev(host->mmc),
+				"CMD%d timeout, status %x\n",
+				host->cmd->opcode, qstatus);
+			host->cmd->error = -ETIMEDOUT;
+			if (data) {
+				end_transfer = 1;
+				davinci_abort_data(host, data);
+			} else
+				end_command = 1;
+		}
+	}
+
+	if (qstatus & MMCST0_CRCRS) {
+		/* Command CRC error */
+		dev_dbg(mmc_dev(host->mmc), "Command CRC error\n");
+		if (host->cmd) {
+			host->cmd->error = -EILSEQ;
+			end_command = 1;
+		}
+	}
+
+	if (qstatus & MMCST0_RSPDNE) {
+		/* End of command phase */
+		end_command = (int) host->cmd;
+	}
+
+	if (end_command)
+		mmc_davinci_cmd_done(host, host->cmd);
+	if (end_transfer)
+		mmc_davinci_xfer_done(host, data);
+	return IRQ_HANDLED;
+}
+
+static int mmc_davinci_get_cd(struct mmc_host *mmc)
+{
+	struct platform_device *pdev = to_platform_device(mmc->parent);
+	struct davinci_mmc_config *config = pdev->dev.platform_data;
+
+	if (!config || !config->get_cd)
+		return -ENOSYS;
+	return config->get_cd(pdev->id);
+}
+
+static int mmc_davinci_get_ro(struct mmc_host *mmc)
+{
+	struct platform_device *pdev = to_platform_device(mmc->parent);
+	struct davinci_mmc_config *config = pdev->dev.platform_data;
+
+	if (!config || !config->get_ro)
+		return -ENOSYS;
+	return config->get_ro(pdev->id);
+}
+
+static void mmc_davinci_enable_sdio_irq(struct mmc_host *mmc, int enable)
+{
+	struct mmc_davinci_host *host = mmc_priv(mmc);
+
+	if (enable) {
+		if (!(readl(host->base + DAVINCI_SDIOST0) & SDIOST0_DAT1_HI)) {
+			writel(SDIOIST_IOINT, host->base + DAVINCI_SDIOIST);
+			mmc_signal_sdio_irq(host->mmc);
+		} else {
+			host->sdio_int = true;
+			writel(readl(host->base + DAVINCI_SDIOIEN) |
+			       SDIOIEN_IOINTEN, host->base + DAVINCI_SDIOIEN);
+		}
+	} else {
+		host->sdio_int = false;
+		writel(readl(host->base + DAVINCI_SDIOIEN) & ~SDIOIEN_IOINTEN,
+		       host->base + DAVINCI_SDIOIEN);
+	}
+}
+
+static struct mmc_host_ops mmc_davinci_ops = {
+	.request	= mmc_davinci_request,
+	.set_ios	= mmc_davinci_set_ios,
+	.get_cd		= mmc_davinci_get_cd,
+	.get_ro		= mmc_davinci_get_ro,
+	.enable_sdio_irq = mmc_davinci_enable_sdio_irq,
+};
+
+/*----------------------------------------------------------------------*/
+
+#ifdef CONFIG_CPU_FREQ
+static int mmc_davinci_cpufreq_transition(struct notifier_block *nb,
+				     unsigned long val, void *data)
+{
+	struct mmc_davinci_host *host;
+	unsigned int mmc_pclk;
+	struct mmc_host *mmc;
+	unsigned long flags;
+
+	host = container_of(nb, struct mmc_davinci_host, freq_transition);
+	mmc = host->mmc;
+	mmc_pclk = clk_get_rate(host->clk);
+
+	if (val == CPUFREQ_POSTCHANGE) {
+		spin_lock_irqsave(&mmc->lock, flags);
+		host->mmc_input_clk = mmc_pclk;
+		calculate_clk_divider(mmc, &mmc->ios);
+		spin_unlock_irqrestore(&mmc->lock, flags);
+	}
+
+	return 0;
+}
+
+static inline int mmc_davinci_cpufreq_register(struct mmc_davinci_host *host)
+{
+	host->freq_transition.notifier_call = mmc_davinci_cpufreq_transition;
+
+	return cpufreq_register_notifier(&host->freq_transition,
+					 CPUFREQ_TRANSITION_NOTIFIER);
+}
+
+static inline void mmc_davinci_cpufreq_deregister(struct mmc_davinci_host *host)
+{
+	cpufreq_unregister_notifier(&host->freq_transition,
+				    CPUFREQ_TRANSITION_NOTIFIER);
+}
+#else
+static inline int mmc_davinci_cpufreq_register(struct mmc_davinci_host *host)
+{
+	return 0;
+}
+
+static inline void mmc_davinci_cpufreq_deregister(struct mmc_davinci_host *host)
+{
+}
+#endif
+static void __init init_mmcsd_host(struct mmc_davinci_host *host)
+{
+
+	mmc_davinci_reset_ctrl(host, 1);
+
+	writel(0, host->base + DAVINCI_MMCCLK);
+	writel(MMCCLK_CLKEN, host->base + DAVINCI_MMCCLK);
+
+	writel(0x1FFF, host->base + DAVINCI_MMCTOR);
+	writel(0xFFFF, host->base + DAVINCI_MMCTOD);
+
+	mmc_davinci_reset_ctrl(host, 0);
+}
+
+static struct platform_device_id davinci_mmc_devtype[] = {
+	{
+		.name	= "dm6441-mmc",
+		.driver_data = MMC_CTLR_VERSION_1,
+	}, {
+		.name	= "da830-mmc",
+		.driver_data = MMC_CTLR_VERSION_2,
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(platform, davinci_mmc_devtype);
+
+static const struct of_device_id davinci_mmc_dt_ids[] = {
+	{
+		.compatible = "ti,dm6441-mmc",
+		.data = &davinci_mmc_devtype[MMC_CTLR_VERSION_1],
+	},
+	{
+		.compatible = "ti,da830-mmc",
+		.data = &davinci_mmc_devtype[MMC_CTLR_VERSION_2],
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, davinci_mmc_dt_ids);
+
+static struct davinci_mmc_config
+	*mmc_parse_pdata(struct platform_device *pdev)
+{
+	struct device_node *np;
+	struct davinci_mmc_config *pdata = pdev->dev.platform_data;
+	const struct of_device_id *match =
+		of_match_device(davinci_mmc_dt_ids, &pdev->dev);
+	u32 data;
+
+	np = pdev->dev.of_node;
+	if (!np)
+		return pdata;
+
+	pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata) {
+		dev_err(&pdev->dev, "Failed to allocate memory for struct davinci_mmc_config\n");
+		goto nodata;
+	}
+
+	if (match)
+		pdev->id_entry = match->data;
+
+	if (of_property_read_u32(np, "max-frequency", &pdata->max_freq))
+		dev_info(&pdev->dev, "'max-frequency' property not specified, defaulting to 25MHz\n");
+
+	of_property_read_u32(np, "bus-width", &data);
+	switch (data) {
+	case 1:
+	case 4:
+	case 8:
+		pdata->wires = data;
+		break;
+	default:
+		pdata->wires = 1;
+		dev_info(&pdev->dev, "Unsupported buswidth, defaulting to 1 bit\n");
+	}
+nodata:
+	return pdata;
+}
+
+static int __init davinci_mmcsd_probe(struct platform_device *pdev)
+{
+	struct davinci_mmc_config *pdata = NULL;
+	struct mmc_davinci_host *host = NULL;
+	struct mmc_host *mmc = NULL;
+	struct resource *r, *mem = NULL;
+	int ret = 0, irq = 0;
+	size_t mem_size;
+	const struct platform_device_id *id_entry;
+
+	pdata = mmc_parse_pdata(pdev);
+	if (pdata == NULL) {
+		dev_err(&pdev->dev, "Couldn't get platform data\n");
+		return -ENOENT;
+	}
+
+	ret = -ENODEV;
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	irq = platform_get_irq(pdev, 0);
+	if (!r || irq == NO_IRQ)
+		goto out;
+
+	ret = -EBUSY;
+	mem_size = resource_size(r);
+	mem = request_mem_region(r->start, mem_size, pdev->name);
+	if (!mem)
+		goto out;
+
+	ret = -ENOMEM;
+	mmc = mmc_alloc_host(sizeof(struct mmc_davinci_host), &pdev->dev);
+	if (!mmc)
+		goto out;
+
+	host = mmc_priv(mmc);
+	host->mmc = mmc;	/* Important */
+
+	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
+	if (!r)
+		dev_warn(&pdev->dev, "RX DMA resource not specified\n");
+	else
+		host->rxdma = r->start;
+
+	r = platform_get_resource(pdev, IORESOURCE_DMA, 1);
+	if (!r)
+		dev_warn(&pdev->dev, "TX DMA resource not specified\n");
+	else
+		host->txdma = r->start;
+
+	host->mem_res = mem;
+	host->base = ioremap(mem->start, mem_size);
+	if (!host->base)
+		goto out;
+
+	ret = -ENXIO;
+	host->clk = clk_get(&pdev->dev, "MMCSDCLK");
+	if (IS_ERR(host->clk)) {
+		ret = PTR_ERR(host->clk);
+		goto out;
+	}
+	clk_enable(host->clk);
+	host->mmc_input_clk = clk_get_rate(host->clk);
+
+	init_mmcsd_host(host);
+
+	if (pdata->nr_sg)
+		host->nr_sg = pdata->nr_sg - 1;
+
+	if (host->nr_sg > MAX_NR_SG || !host->nr_sg)
+		host->nr_sg = MAX_NR_SG;
+
+	host->use_dma = use_dma;
+	host->mmc_irq = irq;
+	host->sdio_irq = platform_get_irq(pdev, 1);
+
+	if (host->use_dma && davinci_acquire_dma_channels(host) != 0)
+		host->use_dma = 0;
+
+	/* REVISIT:  someday, support IRQ-driven card detection.  */
+	mmc->caps |= MMC_CAP_NEEDS_POLL;
+	mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
+
+	if (pdata && (pdata->wires == 4 || pdata->wires == 0))
+		mmc->caps |= MMC_CAP_4_BIT_DATA;
+
+	if (pdata && (pdata->wires == 8))
+		mmc->caps |= (MMC_CAP_4_BIT_DATA | MMC_CAP_8_BIT_DATA);
+
+	id_entry = platform_get_device_id(pdev);
+	if (id_entry)
+		host->version = id_entry->driver_data;
+
+	mmc->ops = &mmc_davinci_ops;
+	mmc->f_min = 312500;
+	mmc->f_max = 25000000;
+	if (pdata && pdata->max_freq)
+		mmc->f_max = pdata->max_freq;
+	if (pdata && pdata->caps)
+		mmc->caps |= pdata->caps;
+	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+
+	/* With no iommu coalescing pages, each phys_seg is a hw_seg.
+	 * Each hw_seg uses one EDMA parameter RAM slot, always one
+	 * channel and then usually some linked slots.
+	 */
+	mmc->max_segs		= MAX_NR_SG;
+
+	/* EDMA limit per hw segment (one or two MBytes) */
+	mmc->max_seg_size	= MAX_CCNT * rw_threshold;
+
+	/* MMC/SD controller limits for multiblock requests */
+	mmc->max_blk_size	= 4095;  /* BLEN is 12 bits */
+	mmc->max_blk_count	= 65535; /* NBLK is 16 bits */
+	mmc->max_req_size	= mmc->max_blk_size * mmc->max_blk_count;
+
+	dev_dbg(mmc_dev(host->mmc), "max_segs=%d\n", mmc->max_segs);
+	dev_dbg(mmc_dev(host->mmc), "max_blk_size=%d\n", mmc->max_blk_size);
+	dev_dbg(mmc_dev(host->mmc), "max_req_size=%d\n", mmc->max_req_size);
+	dev_dbg(mmc_dev(host->mmc), "max_seg_size=%d\n", mmc->max_seg_size);
+
+	platform_set_drvdata(pdev, host);
+
+	ret = mmc_davinci_cpufreq_register(host);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register cpufreq\n");
+		goto cpu_freq_fail;
+	}
+
+	ret = mmc_add_host(mmc);
+	if (ret < 0)
+		goto out;
+
+	ret = request_irq(irq, mmc_davinci_irq, 0, mmc_hostname(mmc), host);
+	if (ret)
+		goto out;
+
+	if (host->sdio_irq >= 0) {
+		ret = request_irq(host->sdio_irq, mmc_davinci_sdio_irq, 0,
+				  mmc_hostname(mmc), host);
+		if (!ret)
+			mmc->caps |= MMC_CAP_SDIO_IRQ;
+	}
+
+	rename_region(mem, mmc_hostname(mmc));
+
+	dev_info(mmc_dev(host->mmc), "Using %s, %d-bit mode\n",
+		host->use_dma ? "DMA" : "PIO",
+		(mmc->caps & MMC_CAP_4_BIT_DATA) ? 4 : 1);
+
+	return 0;
+
+out:
+	mmc_davinci_cpufreq_deregister(host);
+cpu_freq_fail:
+	if (host) {
+		davinci_release_dma_channels(host);
+
+		if (host->clk) {
+			clk_disable(host->clk);
+			clk_put(host->clk);
+		}
+
+		if (host->base)
+			iounmap(host->base);
+	}
+
+	if (mmc)
+		mmc_free_host(mmc);
+
+	if (mem)
+		release_resource(mem);
+
+	dev_dbg(&pdev->dev, "probe err %d\n", ret);
+
+	return ret;
+}
+
+static int __exit davinci_mmcsd_remove(struct platform_device *pdev)
+{
+	struct mmc_davinci_host *host = platform_get_drvdata(pdev);
+
+	if (host) {
+		mmc_davinci_cpufreq_deregister(host);
+
+		mmc_remove_host(host->mmc);
+		free_irq(host->mmc_irq, host);
+		if (host->mmc->caps & MMC_CAP_SDIO_IRQ)
+			free_irq(host->sdio_irq, host);
+
+		davinci_release_dma_channels(host);
+
+		clk_disable(host->clk);
+		clk_put(host->clk);
+
+		iounmap(host->base);
+
+		release_resource(host->mem_res);
+
+		mmc_free_host(host->mmc);
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int davinci_mmcsd_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct mmc_davinci_host *host = platform_get_drvdata(pdev);
+
+	writel(0, host->base + DAVINCI_MMCIM);
+	mmc_davinci_reset_ctrl(host, 1);
+	clk_disable(host->clk);
+
+	return 0;
+}
+
+static int davinci_mmcsd_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct mmc_davinci_host *host = platform_get_drvdata(pdev);
+
+	clk_enable(host->clk);
+	mmc_davinci_reset_ctrl(host, 0);
+
+	return 0;
+}
+
+static const struct dev_pm_ops davinci_mmcsd_pm = {
+	.suspend        = davinci_mmcsd_suspend,
+	.resume         = davinci_mmcsd_resume,
+};
+
+#define davinci_mmcsd_pm_ops (&davinci_mmcsd_pm)
+#else
+#define davinci_mmcsd_pm_ops NULL
+#endif
+
+static struct platform_driver davinci_mmcsd_driver = {
+	.driver		= {
+		.name	= "davinci_mmc",
+		.owner	= THIS_MODULE,
+		.pm	= davinci_mmcsd_pm_ops,
+		.of_match_table = davinci_mmc_dt_ids,
+	},
+	.remove		= __exit_p(davinci_mmcsd_remove),
+	.id_table	= davinci_mmc_devtype,
+};
+
+module_platform_driver_probe(davinci_mmcsd_driver, davinci_mmcsd_probe);
+
+MODULE_AUTHOR("Texas Instruments India");
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("MMC/SD driver for Davinci MMC controller");
+MODULE_ALIAS("platform:davinci_mmc");
+
diff --git a/drivers/mmc/host/dw_mmc-exynos.c b/drivers/mmc/host/dw_mmc-exynos.c
new file mode 100644
index 00000000000..0fbc53ac7ea
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-exynos.c
@@ -0,0 +1,450 @@
+/*
+ * Exynos Specific Extensions for Synopsys DW Multimedia Card Interface driver
+ *
+ * Copyright (C) 2012, Samsung Electronics Co., Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/dw_mmc.h>
+#include <linux/mmc/mmc.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/slab.h>
+
+#include "dw_mmc.h"
+#include "dw_mmc-pltfm.h"
+
+#define NUM_PINS(x)			(x + 2)
+
+#define SDMMC_CLKSEL			0x09C
+#define SDMMC_CLKSEL_CCLK_SAMPLE(x)	(((x) & 7) << 0)
+#define SDMMC_CLKSEL_CCLK_DRIVE(x)	(((x) & 7) << 16)
+#define SDMMC_CLKSEL_CCLK_DIVIDER(x)	(((x) & 7) << 24)
+#define SDMMC_CLKSEL_GET_DRV_WD3(x)	(((x) >> 16) & 0x7)
+#define SDMMC_CLKSEL_TIMING(x, y, z)	(SDMMC_CLKSEL_CCLK_SAMPLE(x) |	\
+					SDMMC_CLKSEL_CCLK_DRIVE(y) |	\
+					SDMMC_CLKSEL_CCLK_DIVIDER(z))
+#define SDMMC_CLKSEL_WAKEUP_INT		BIT(11)
+
+#define EXYNOS4210_FIXED_CIU_CLK_DIV	2
+#define EXYNOS4412_FIXED_CIU_CLK_DIV	4
+
+/* Block number in eMMC */
+#define DWMCI_BLOCK_NUM		0xFFFFFFFF
+
+#define SDMMC_EMMCP_BASE	0x1000
+#define SDMMC_MPSECURITY	(SDMMC_EMMCP_BASE + 0x0010)
+#define SDMMC_MPSBEGIN0		(SDMMC_EMMCP_BASE + 0x0200)
+#define SDMMC_MPSEND0		(SDMMC_EMMCP_BASE + 0x0204)
+#define SDMMC_MPSCTRL0		(SDMMC_EMMCP_BASE + 0x020C)
+
+/* SMU control bits */
+#define DWMCI_MPSCTRL_SECURE_READ_BIT		BIT(7)
+#define DWMCI_MPSCTRL_SECURE_WRITE_BIT		BIT(6)
+#define DWMCI_MPSCTRL_NON_SECURE_READ_BIT	BIT(5)
+#define DWMCI_MPSCTRL_NON_SECURE_WRITE_BIT	BIT(4)
+#define DWMCI_MPSCTRL_USE_FUSE_KEY		BIT(3)
+#define DWMCI_MPSCTRL_ECB_MODE			BIT(2)
+#define DWMCI_MPSCTRL_ENCRYPTION		BIT(1)
+#define DWMCI_MPSCTRL_VALID			BIT(0)
+
+#define EXYNOS_CCLKIN_MIN	50000000	/* unit: HZ */
+
+/* Variations in Exynos specific dw-mshc controller */
+enum dw_mci_exynos_type {
+	DW_MCI_TYPE_EXYNOS4210,
+	DW_MCI_TYPE_EXYNOS4412,
+	DW_MCI_TYPE_EXYNOS5250,
+	DW_MCI_TYPE_EXYNOS5420,
+	DW_MCI_TYPE_EXYNOS5420_SMU,
+};
+
+/* Exynos implementation specific driver private data */
+struct dw_mci_exynos_priv_data {
+	enum dw_mci_exynos_type		ctrl_type;
+	u8				ciu_div;
+	u32				sdr_timing;
+	u32				ddr_timing;
+	u32				cur_speed;
+};
+
+static struct dw_mci_exynos_compatible {
+	char				*compatible;
+	enum dw_mci_exynos_type		ctrl_type;
+} exynos_compat[] = {
+	{
+		.compatible	= "samsung,exynos4210-dw-mshc",
+		.ctrl_type	= DW_MCI_TYPE_EXYNOS4210,
+	}, {
+		.compatible	= "samsung,exynos4412-dw-mshc",
+		.ctrl_type	= DW_MCI_TYPE_EXYNOS4412,
+	}, {
+		.compatible	= "samsung,exynos5250-dw-mshc",
+		.ctrl_type	= DW_MCI_TYPE_EXYNOS5250,
+	}, {
+		.compatible	= "samsung,exynos5420-dw-mshc",
+		.ctrl_type	= DW_MCI_TYPE_EXYNOS5420,
+	}, {
+		.compatible	= "samsung,exynos5420-dw-mshc-smu",
+		.ctrl_type	= DW_MCI_TYPE_EXYNOS5420_SMU,
+	},
+};
+
+static int dw_mci_exynos_priv_init(struct dw_mci *host)
+{
+	struct dw_mci_exynos_priv_data *priv = host->priv;
+
+	if (priv->ctrl_type == DW_MCI_TYPE_EXYNOS5420_SMU) {
+		mci_writel(host, MPSBEGIN0, 0);
+		mci_writel(host, MPSEND0, DWMCI_BLOCK_NUM);
+		mci_writel(host, MPSCTRL0, DWMCI_MPSCTRL_SECURE_WRITE_BIT |
+			   DWMCI_MPSCTRL_NON_SECURE_READ_BIT |
+			   DWMCI_MPSCTRL_VALID |
+			   DWMCI_MPSCTRL_NON_SECURE_WRITE_BIT);
+	}
+
+	return 0;
+}
+
+static int dw_mci_exynos_setup_clock(struct dw_mci *host)
+{
+	struct dw_mci_exynos_priv_data *priv = host->priv;
+	unsigned long rate = clk_get_rate(host->ciu_clk);
+
+	host->bus_hz = rate / (priv->ciu_div + 1);
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int dw_mci_exynos_suspend(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+
+	return dw_mci_suspend(host);
+}
+
+static int dw_mci_exynos_resume(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+
+	dw_mci_exynos_priv_init(host);
+	return dw_mci_resume(host);
+}
+
+/**
+ * dw_mci_exynos_resume_noirq - Exynos-specific resume code
+ *
+ * On exynos5420 there is a silicon errata that will sometimes leave the
+ * WAKEUP_INT bit in the CLKSEL register asserted.  This bit is 1 to indicate
+ * that it fired and we can clear it by writing a 1 back.  Clear it to prevent
+ * interrupts from going off constantly.
+ *
+ * We run this code on all exynos variants because it doesn't hurt.
+ */
+
+static int dw_mci_exynos_resume_noirq(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+	u32 clksel;
+
+	clksel = mci_readl(host, CLKSEL);
+	if (clksel & SDMMC_CLKSEL_WAKEUP_INT)
+		mci_writel(host, CLKSEL, clksel);
+
+	return 0;
+}
+#else
+#define dw_mci_exynos_suspend		NULL
+#define dw_mci_exynos_resume		NULL
+#define dw_mci_exynos_resume_noirq	NULL
+#endif /* CONFIG_PM_SLEEP */
+
+static void dw_mci_exynos_prepare_command(struct dw_mci *host, u32 *cmdr)
+{
+	/*
+	 * Exynos4412 and Exynos5250 extends the use of CMD register with the
+	 * use of bit 29 (which is reserved on standard MSHC controllers) for
+	 * optionally bypassing the HOLD register for command and data. The
+	 * HOLD register should be bypassed in case there is no phase shift
+	 * applied on CMD/DATA that is sent to the card.
+	 */
+	if (SDMMC_CLKSEL_GET_DRV_WD3(mci_readl(host, CLKSEL)))
+		*cmdr |= SDMMC_CMD_USE_HOLD_REG;
+}
+
+static void dw_mci_exynos_set_ios(struct dw_mci *host, struct mmc_ios *ios)
+{
+	struct dw_mci_exynos_priv_data *priv = host->priv;
+	unsigned int wanted = ios->clock;
+	unsigned long actual;
+	u8 div = priv->ciu_div + 1;
+
+	if (ios->timing == MMC_TIMING_MMC_DDR52) {
+		mci_writel(host, CLKSEL, priv->ddr_timing);
+		/* Should be double rate for DDR mode */
+		if (ios->bus_width == MMC_BUS_WIDTH_8)
+			wanted <<= 1;
+	} else {
+		mci_writel(host, CLKSEL, priv->sdr_timing);
+	}
+
+	/* Don't care if wanted clock is zero */
+	if (!wanted)
+		return;
+
+	/* Guaranteed minimum frequency for cclkin */
+	if (wanted < EXYNOS_CCLKIN_MIN)
+		wanted = EXYNOS_CCLKIN_MIN;
+
+	if (wanted != priv->cur_speed) {
+		int ret = clk_set_rate(host->ciu_clk, wanted * div);
+		if (ret)
+			dev_warn(host->dev,
+				"failed to set clk-rate %u error: %d\n",
+				 wanted * div, ret);
+		actual = clk_get_rate(host->ciu_clk);
+		host->bus_hz = actual / div;
+		priv->cur_speed = wanted;
+		host->current_speed = 0;
+	}
+}
+
+static int dw_mci_exynos_parse_dt(struct dw_mci *host)
+{
+	struct dw_mci_exynos_priv_data *priv;
+	struct device_node *np = host->dev->of_node;
+	u32 timing[2];
+	u32 div = 0;
+	int idx;
+	int ret;
+
+	priv = devm_kzalloc(host->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv) {
+		dev_err(host->dev, "mem alloc failed for private data\n");
+		return -ENOMEM;
+	}
+
+	for (idx = 0; idx < ARRAY_SIZE(exynos_compat); idx++) {
+		if (of_device_is_compatible(np, exynos_compat[idx].compatible))
+			priv->ctrl_type = exynos_compat[idx].ctrl_type;
+	}
+
+	if (priv->ctrl_type == DW_MCI_TYPE_EXYNOS4412)
+		priv->ciu_div = EXYNOS4412_FIXED_CIU_CLK_DIV - 1;
+	else if (priv->ctrl_type == DW_MCI_TYPE_EXYNOS4210)
+		priv->ciu_div = EXYNOS4210_FIXED_CIU_CLK_DIV - 1;
+	else {
+		of_property_read_u32(np, "samsung,dw-mshc-ciu-div", &div);
+		priv->ciu_div = div;
+	}
+
+	ret = of_property_read_u32_array(np,
+			"samsung,dw-mshc-sdr-timing", timing, 2);
+	if (ret)
+		return ret;
+
+	priv->sdr_timing = SDMMC_CLKSEL_TIMING(timing[0], timing[1], div);
+
+	ret = of_property_read_u32_array(np,
+			"samsung,dw-mshc-ddr-timing", timing, 2);
+	if (ret)
+		return ret;
+
+	priv->ddr_timing = SDMMC_CLKSEL_TIMING(timing[0], timing[1], div);
+	host->priv = priv;
+	return 0;
+}
+
+static inline u8 dw_mci_exynos_get_clksmpl(struct dw_mci *host)
+{
+	return SDMMC_CLKSEL_CCLK_SAMPLE(mci_readl(host, CLKSEL));
+}
+
+static inline void dw_mci_exynos_set_clksmpl(struct dw_mci *host, u8 sample)
+{
+	u32 clksel;
+	clksel = mci_readl(host, CLKSEL);
+	clksel = (clksel & ~0x7) | SDMMC_CLKSEL_CCLK_SAMPLE(sample);
+	mci_writel(host, CLKSEL, clksel);
+}
+
+static inline u8 dw_mci_exynos_move_next_clksmpl(struct dw_mci *host)
+{
+	u32 clksel;
+	u8 sample;
+
+	clksel = mci_readl(host, CLKSEL);
+	sample = (clksel + 1) & 0x7;
+	clksel = (clksel & ~0x7) | sample;
+	mci_writel(host, CLKSEL, clksel);
+	return sample;
+}
+
+static s8 dw_mci_exynos_get_best_clksmpl(u8 candiates)
+{
+	const u8 iter = 8;
+	u8 __c;
+	s8 i, loc = -1;
+
+	for (i = 0; i < iter; i++) {
+		__c = ror8(candiates, i);
+		if ((__c & 0xc7) == 0xc7) {
+			loc = i;
+			goto out;
+		}
+	}
+
+	for (i = 0; i < iter; i++) {
+		__c = ror8(candiates, i);
+		if ((__c & 0x83) == 0x83) {
+			loc = i;
+			goto out;
+		}
+	}
+
+out:
+	return loc;
+}
+
+static int dw_mci_exynos_execute_tuning(struct dw_mci_slot *slot, u32 opcode,
+					struct dw_mci_tuning_data *tuning_data)
+{
+	struct dw_mci *host = slot->host;
+	struct mmc_host *mmc = slot->mmc;
+	const u8 *blk_pattern = tuning_data->blk_pattern;
+	u8 *blk_test;
+	unsigned int blksz = tuning_data->blksz;
+	u8 start_smpl, smpl, candiates = 0;
+	s8 found = -1;
+	int ret = 0;
+
+	blk_test = kmalloc(blksz, GFP_KERNEL);
+	if (!blk_test)
+		return -ENOMEM;
+
+	start_smpl = dw_mci_exynos_get_clksmpl(host);
+
+	do {
+		struct mmc_request mrq = {NULL};
+		struct mmc_command cmd = {0};
+		struct mmc_command stop = {0};
+		struct mmc_data data = {0};
+		struct scatterlist sg;
+
+		cmd.opcode = opcode;
+		cmd.arg = 0;
+		cmd.flags = MMC_RSP_R1 | MMC_CMD_ADTC;
+
+		stop.opcode = MMC_STOP_TRANSMISSION;
+		stop.arg = 0;
+		stop.flags = MMC_RSP_R1B | MMC_CMD_AC;
+
+		data.blksz = blksz;
+		data.blocks = 1;
+		data.flags = MMC_DATA_READ;
+		data.sg = &sg;
+		data.sg_len = 1;
+
+		sg_init_one(&sg, blk_test, blksz);
+		mrq.cmd = &cmd;
+		mrq.stop = &stop;
+		mrq.data = &data;
+		host->mrq = &mrq;
+
+		mci_writel(host, TMOUT, ~0);
+		smpl = dw_mci_exynos_move_next_clksmpl(host);
+
+		mmc_wait_for_req(mmc, &mrq);
+
+		if (!cmd.error && !data.error) {
+			if (!memcmp(blk_pattern, blk_test, blksz))
+				candiates |= (1 << smpl);
+		} else {
+			dev_dbg(host->dev,
+				"Tuning error: cmd.error:%d, data.error:%d\n",
+				cmd.error, data.error);
+		}
+	} while (start_smpl != smpl);
+
+	found = dw_mci_exynos_get_best_clksmpl(candiates);
+	if (found >= 0)
+		dw_mci_exynos_set_clksmpl(host, found);
+	else
+		ret = -EIO;
+
+	kfree(blk_test);
+	return ret;
+}
+
+/* Common capabilities of Exynos4/Exynos5 SoC */
+static unsigned long exynos_dwmmc_caps[4] = {
+	MMC_CAP_1_8V_DDR | MMC_CAP_8_BIT_DATA | MMC_CAP_CMD23,
+	MMC_CAP_CMD23,
+	MMC_CAP_CMD23,
+	MMC_CAP_CMD23,
+};
+
+static const struct dw_mci_drv_data exynos_drv_data = {
+	.caps			= exynos_dwmmc_caps,
+	.init			= dw_mci_exynos_priv_init,
+	.setup_clock		= dw_mci_exynos_setup_clock,
+	.prepare_command	= dw_mci_exynos_prepare_command,
+	.set_ios		= dw_mci_exynos_set_ios,
+	.parse_dt		= dw_mci_exynos_parse_dt,
+	.execute_tuning		= dw_mci_exynos_execute_tuning,
+};
+
+static const struct of_device_id dw_mci_exynos_match[] = {
+	{ .compatible = "samsung,exynos4412-dw-mshc",
+			.data = &exynos_drv_data, },
+	{ .compatible = "samsung,exynos5250-dw-mshc",
+			.data = &exynos_drv_data, },
+	{ .compatible = "samsung,exynos5420-dw-mshc",
+			.data = &exynos_drv_data, },
+	{ .compatible = "samsung,exynos5420-dw-mshc-smu",
+			.data = &exynos_drv_data, },
+	{},
+};
+MODULE_DEVICE_TABLE(of, dw_mci_exynos_match);
+
+static int dw_mci_exynos_probe(struct platform_device *pdev)
+{
+	const struct dw_mci_drv_data *drv_data;
+	const struct of_device_id *match;
+
+	match = of_match_node(dw_mci_exynos_match, pdev->dev.of_node);
+	drv_data = match->data;
+	return dw_mci_pltfm_register(pdev, drv_data);
+}
+
+static const struct dev_pm_ops dw_mci_exynos_pmops = {
+	SET_SYSTEM_SLEEP_PM_OPS(dw_mci_exynos_suspend, dw_mci_exynos_resume)
+	.resume_noirq = dw_mci_exynos_resume_noirq,
+	.thaw_noirq = dw_mci_exynos_resume_noirq,
+	.restore_noirq = dw_mci_exynos_resume_noirq,
+};
+
+static struct platform_driver dw_mci_exynos_pltfm_driver = {
+	.probe		= dw_mci_exynos_probe,
+	.remove		= __exit_p(dw_mci_pltfm_remove),
+	.driver		= {
+		.name		= "dwmmc_exynos",
+		.of_match_table	= dw_mci_exynos_match,
+		.pm		= &dw_mci_exynos_pmops,
+	},
+};
+
+module_platform_driver(dw_mci_exynos_pltfm_driver);
+
+MODULE_DESCRIPTION("Samsung Specific DW-MSHC Driver Extension");
+MODULE_AUTHOR("Thomas Abraham <thomas.ab@samsung.com");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dwmmc-exynos");
diff --git a/drivers/mmc/host/dw_mmc-k3.c b/drivers/mmc/host/dw_mmc-k3.c
new file mode 100644
index 00000000000..650f9cc3f7a
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-k3.c
@@ -0,0 +1,97 @@
+/*
+ * Copyright (c) 2013 Linaro Ltd.
+ * Copyright (c) 2013 Hisilicon Limited.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/dw_mmc.h>
+#include <linux/of_address.h>
+
+#include "dw_mmc.h"
+#include "dw_mmc-pltfm.h"
+
+static void dw_mci_k3_set_ios(struct dw_mci *host, struct mmc_ios *ios)
+{
+	int ret;
+
+	ret = clk_set_rate(host->ciu_clk, ios->clock);
+	if (ret)
+		dev_warn(host->dev, "failed to set rate %uHz\n", ios->clock);
+
+	host->bus_hz = clk_get_rate(host->ciu_clk);
+}
+
+static const struct dw_mci_drv_data k3_drv_data = {
+	.set_ios		= dw_mci_k3_set_ios,
+};
+
+static const struct of_device_id dw_mci_k3_match[] = {
+	{ .compatible = "hisilicon,hi4511-dw-mshc", .data = &k3_drv_data, },
+	{},
+};
+MODULE_DEVICE_TABLE(of, dw_mci_k3_match);
+
+static int dw_mci_k3_probe(struct platform_device *pdev)
+{
+	const struct dw_mci_drv_data *drv_data;
+	const struct of_device_id *match;
+
+	match = of_match_node(dw_mci_k3_match, pdev->dev.of_node);
+	drv_data = match->data;
+
+	return dw_mci_pltfm_register(pdev, drv_data);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int dw_mci_k3_suspend(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+	int ret;
+
+	ret = dw_mci_suspend(host);
+	if (!ret)
+		clk_disable_unprepare(host->ciu_clk);
+
+	return ret;
+}
+
+static int dw_mci_k3_resume(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+	int ret;
+
+	ret = clk_prepare_enable(host->ciu_clk);
+	if (ret) {
+		dev_err(host->dev, "failed to enable ciu clock\n");
+		return ret;
+	}
+
+	return dw_mci_resume(host);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static SIMPLE_DEV_PM_OPS(dw_mci_k3_pmops, dw_mci_k3_suspend, dw_mci_k3_resume);
+
+static struct platform_driver dw_mci_k3_pltfm_driver = {
+	.probe		= dw_mci_k3_probe,
+	.remove		= dw_mci_pltfm_remove,
+	.driver		= {
+		.name		= "dwmmc_k3",
+		.of_match_table	= dw_mci_k3_match,
+		.pm		= &dw_mci_k3_pmops,
+	},
+};
+
+module_platform_driver(dw_mci_k3_pltfm_driver);
+
+MODULE_DESCRIPTION("K3 Specific DW-MSHC Driver Extension");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dwmmc-k3");
diff --git a/drivers/mmc/host/dw_mmc-pci.c b/drivers/mmc/host/dw_mmc-pci.c
new file mode 100644
index 00000000000..f70546a3a7c
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-pci.c
@@ -0,0 +1,125 @@
+/*
+ * Synopsys DesignWare Multimedia Card PCI Interface driver
+ *
+ * Copyright (C) 2012 Vayavya Labs Pvt. Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/mmc.h>
+#include <linux/mmc/dw_mmc.h>
+#include "dw_mmc.h"
+
+#define PCI_BAR_NO 2
+#define SYNOPSYS_DW_MCI_VENDOR_ID 0x700
+#define SYNOPSYS_DW_MCI_DEVICE_ID 0x1107
+/* Defining the Capabilities */
+#define DW_MCI_CAPABILITIES (MMC_CAP_4_BIT_DATA | MMC_CAP_MMC_HIGHSPEED |\
+				MMC_CAP_SD_HIGHSPEED | MMC_CAP_8_BIT_DATA |\
+				MMC_CAP_SDIO_IRQ)
+
+static struct dw_mci_board pci_board_data = {
+	.num_slots			= 1,
+	.caps				= DW_MCI_CAPABILITIES,
+	.bus_hz				= 33 * 1000 * 1000,
+	.detect_delay_ms		= 200,
+	.fifo_depth			= 32,
+};
+
+static int dw_mci_pci_probe(struct pci_dev *pdev,
+			    const struct pci_device_id *entries)
+{
+	struct dw_mci *host;
+	int ret;
+
+	ret = pcim_enable_device(pdev);
+	if (ret)
+		return ret;
+
+	host = devm_kzalloc(&pdev->dev, sizeof(struct dw_mci), GFP_KERNEL);
+	if (!host)
+		return -ENOMEM;
+
+	host->irq = pdev->irq;
+	host->irq_flags = IRQF_SHARED;
+	host->dev = &pdev->dev;
+	host->pdata = &pci_board_data;
+
+	ret = pcim_iomap_regions(pdev, 1 << PCI_BAR_NO, pci_name(pdev));
+	if (ret)
+		return ret;
+
+	host->regs = pcim_iomap_table(pdev)[PCI_BAR_NO];
+
+	pci_set_master(pdev);
+
+	ret = dw_mci_probe(host);
+	if (ret)
+		return ret;
+
+	pci_set_drvdata(pdev, host);
+
+	return 0;
+}
+
+static void dw_mci_pci_remove(struct pci_dev *pdev)
+{
+	struct dw_mci *host = pci_get_drvdata(pdev);
+
+	dw_mci_remove(host);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int dw_mci_pci_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct dw_mci *host = pci_get_drvdata(pdev);
+
+	return dw_mci_suspend(host);
+}
+
+static int dw_mci_pci_resume(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct dw_mci *host = pci_get_drvdata(pdev);
+
+	return dw_mci_resume(host);
+}
+#else
+#define dw_mci_pci_suspend	NULL
+#define dw_mci_pci_resume	NULL
+#endif /* CONFIG_PM_SLEEP */
+
+static SIMPLE_DEV_PM_OPS(dw_mci_pci_pmops, dw_mci_pci_suspend, dw_mci_pci_resume);
+
+static DEFINE_PCI_DEVICE_TABLE(dw_mci_pci_id) = {
+	{ PCI_DEVICE(SYNOPSYS_DW_MCI_VENDOR_ID, SYNOPSYS_DW_MCI_DEVICE_ID) },
+	{}
+};
+MODULE_DEVICE_TABLE(pci, dw_mci_pci_id);
+
+static struct pci_driver dw_mci_pci_driver = {
+	.name		= "dw_mmc_pci",
+	.id_table	= dw_mci_pci_id,
+	.probe		= dw_mci_pci_probe,
+	.remove		= dw_mci_pci_remove,
+	.driver		=	{
+		.pm =   &dw_mci_pci_pmops
+	},
+};
+
+module_pci_driver(dw_mci_pci_driver);
+
+MODULE_DESCRIPTION("DW Multimedia Card PCI Interface driver");
+MODULE_AUTHOR("Shashidhar Hiremath <shashidharh@vayavyalabs.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/dw_mmc-pltfm.c b/drivers/mmc/host/dw_mmc-pltfm.c
new file mode 100644
index 00000000000..d4a47a9f558
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-pltfm.c
@@ -0,0 +1,142 @@
+/*
+ * Synopsys DesignWare Multimedia Card Interface driver
+ *
+ * Copyright (C) 2009 NXP Semiconductors
+ * Copyright (C) 2009, 2010 Imagination Technologies Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/mmc.h>
+#include <linux/mmc/dw_mmc.h>
+#include <linux/of.h>
+
+#include "dw_mmc.h"
+#include "dw_mmc-pltfm.h"
+
+static void dw_mci_pltfm_prepare_command(struct dw_mci *host, u32 *cmdr)
+{
+	*cmdr |= SDMMC_CMD_USE_HOLD_REG;
+}
+
+static const struct dw_mci_drv_data rockchip_drv_data = {
+	.prepare_command	= dw_mci_pltfm_prepare_command,
+};
+
+static const struct dw_mci_drv_data socfpga_drv_data = {
+	.prepare_command	= dw_mci_pltfm_prepare_command,
+};
+
+int dw_mci_pltfm_register(struct platform_device *pdev,
+			  const struct dw_mci_drv_data *drv_data)
+{
+	struct dw_mci *host;
+	struct resource	*regs;
+
+	host = devm_kzalloc(&pdev->dev, sizeof(struct dw_mci), GFP_KERNEL);
+	if (!host)
+		return -ENOMEM;
+
+	host->irq = platform_get_irq(pdev, 0);
+	if (host->irq < 0)
+		return host->irq;
+
+	host->drv_data = drv_data;
+	host->dev = &pdev->dev;
+	host->irq_flags = 0;
+	host->pdata = pdev->dev.platform_data;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	host->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(host->regs))
+		return PTR_ERR(host->regs);
+
+	platform_set_drvdata(pdev, host);
+	return dw_mci_probe(host);
+}
+EXPORT_SYMBOL_GPL(dw_mci_pltfm_register);
+
+#ifdef CONFIG_PM_SLEEP
+/*
+ * TODO: we should probably disable the clock to the card in the suspend path.
+ */
+static int dw_mci_pltfm_suspend(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+
+	return dw_mci_suspend(host);
+}
+
+static int dw_mci_pltfm_resume(struct device *dev)
+{
+	struct dw_mci *host = dev_get_drvdata(dev);
+
+	return dw_mci_resume(host);
+}
+#else
+#define dw_mci_pltfm_suspend	NULL
+#define dw_mci_pltfm_resume	NULL
+#endif /* CONFIG_PM_SLEEP */
+
+SIMPLE_DEV_PM_OPS(dw_mci_pltfm_pmops, dw_mci_pltfm_suspend, dw_mci_pltfm_resume);
+EXPORT_SYMBOL_GPL(dw_mci_pltfm_pmops);
+
+static const struct of_device_id dw_mci_pltfm_match[] = {
+	{ .compatible = "snps,dw-mshc", },
+	{ .compatible = "rockchip,rk2928-dw-mshc",
+		.data = &rockchip_drv_data },
+	{ .compatible = "altr,socfpga-dw-mshc",
+		.data = &socfpga_drv_data },
+	{},
+};
+MODULE_DEVICE_TABLE(of, dw_mci_pltfm_match);
+
+static int dw_mci_pltfm_probe(struct platform_device *pdev)
+{
+	const struct dw_mci_drv_data *drv_data = NULL;
+	const struct of_device_id *match;
+
+	if (pdev->dev.of_node) {
+		match = of_match_node(dw_mci_pltfm_match, pdev->dev.of_node);
+		drv_data = match->data;
+	}
+
+	return dw_mci_pltfm_register(pdev, drv_data);
+}
+
+int dw_mci_pltfm_remove(struct platform_device *pdev)
+{
+	struct dw_mci *host = platform_get_drvdata(pdev);
+
+	dw_mci_remove(host);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dw_mci_pltfm_remove);
+
+static struct platform_driver dw_mci_pltfm_driver = {
+	.probe		= dw_mci_pltfm_probe,
+	.remove		= dw_mci_pltfm_remove,
+	.driver		= {
+		.name		= "dw_mmc",
+		.of_match_table	= dw_mci_pltfm_match,
+		.pm		= &dw_mci_pltfm_pmops,
+	},
+};
+
+module_platform_driver(dw_mci_pltfm_driver);
+
+MODULE_DESCRIPTION("DW Multimedia Card Interface driver");
+MODULE_AUTHOR("NXP Semiconductor VietNam");
+MODULE_AUTHOR("Imagination Technologies Ltd");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/dw_mmc-pltfm.h b/drivers/mmc/host/dw_mmc-pltfm.h
new file mode 100644
index 00000000000..68e7fd2f614
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc-pltfm.h
@@ -0,0 +1,20 @@
+/*
+ * Synopsys DesignWare Multimedia Card Interface Platform driver
+ *
+ * Copyright (C) 2012, Samsung Electronics Co., Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef _DW_MMC_PLTFM_H_
+#define _DW_MMC_PLTFM_H_
+
+extern int dw_mci_pltfm_register(struct platform_device *pdev,
+				const struct dw_mci_drv_data *drv_data);
+extern int dw_mci_pltfm_remove(struct platform_device *pdev);
+extern const struct dev_pm_ops dw_mci_pltfm_pmops;
+
+#endif /* _DW_MMC_PLTFM_H_ */
diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
new file mode 100644
index 00000000000..1ac227c603b
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc.c
@@ -0,0 +1,2669 @@
+/*
+ * Synopsys DesignWare Multimedia Card Interface driver
+ *  (Based on NXP driver for lpc 31xx)
+ *
+ * Copyright (C) 2009 NXP Semiconductors
+ * Copyright (C) 2009, 2010 Imagination Technologies Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/blkdev.h>
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/ioport.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/delay.h>
+#include <linux/irq.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/mmc.h>
+#include <linux/mmc/sdio.h>
+#include <linux/mmc/dw_mmc.h>
+#include <linux/bitops.h>
+#include <linux/regulator/consumer.h>
+#include <linux/workqueue.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/mmc/slot-gpio.h>
+
+#include "dw_mmc.h"
+
+/* Common flag combinations */
+#define DW_MCI_DATA_ERROR_FLAGS	(SDMMC_INT_DRTO | SDMMC_INT_DCRC | \
+				 SDMMC_INT_HTO | SDMMC_INT_SBE  | \
+				 SDMMC_INT_EBE)
+#define DW_MCI_CMD_ERROR_FLAGS	(SDMMC_INT_RTO | SDMMC_INT_RCRC | \
+				 SDMMC_INT_RESP_ERR)
+#define DW_MCI_ERROR_FLAGS	(DW_MCI_DATA_ERROR_FLAGS | \
+				 DW_MCI_CMD_ERROR_FLAGS  | SDMMC_INT_HLE)
+#define DW_MCI_SEND_STATUS	1
+#define DW_MCI_RECV_STATUS	2
+#define DW_MCI_DMA_THRESHOLD	16
+
+#define DW_MCI_FREQ_MAX	200000000	/* unit: HZ */
+#define DW_MCI_FREQ_MIN	400000		/* unit: HZ */
+
+#ifdef CONFIG_MMC_DW_IDMAC
+#define IDMAC_INT_CLR		(SDMMC_IDMAC_INT_AI | SDMMC_IDMAC_INT_NI | \
+				 SDMMC_IDMAC_INT_CES | SDMMC_IDMAC_INT_DU | \
+				 SDMMC_IDMAC_INT_FBE | SDMMC_IDMAC_INT_RI | \
+				 SDMMC_IDMAC_INT_TI)
+
+struct idmac_desc {
+	u32		des0;	/* Control Descriptor */
+#define IDMAC_DES0_DIC	BIT(1)
+#define IDMAC_DES0_LD	BIT(2)
+#define IDMAC_DES0_FD	BIT(3)
+#define IDMAC_DES0_CH	BIT(4)
+#define IDMAC_DES0_ER	BIT(5)
+#define IDMAC_DES0_CES	BIT(30)
+#define IDMAC_DES0_OWN	BIT(31)
+
+	u32		des1;	/* Buffer sizes */
+#define IDMAC_SET_BUFFER1_SIZE(d, s) \
+	((d)->des1 = ((d)->des1 & 0x03ffe000) | ((s) & 0x1fff))
+
+	u32		des2;	/* buffer 1 physical address */
+
+	u32		des3;	/* buffer 2 physical address */
+};
+#endif /* CONFIG_MMC_DW_IDMAC */
+
+static const u8 tuning_blk_pattern_4bit[] = {
+	0xff, 0x0f, 0xff, 0x00, 0xff, 0xcc, 0xc3, 0xcc,
+	0xc3, 0x3c, 0xcc, 0xff, 0xfe, 0xff, 0xfe, 0xef,
+	0xff, 0xdf, 0xff, 0xdd, 0xff, 0xfb, 0xff, 0xfb,
+	0xbf, 0xff, 0x7f, 0xff, 0x77, 0xf7, 0xbd, 0xef,
+	0xff, 0xf0, 0xff, 0xf0, 0x0f, 0xfc, 0xcc, 0x3c,
+	0xcc, 0x33, 0xcc, 0xcf, 0xff, 0xef, 0xff, 0xee,
+	0xff, 0xfd, 0xff, 0xfd, 0xdf, 0xff, 0xbf, 0xff,
+	0xbb, 0xff, 0xf7, 0xff, 0xf7, 0x7f, 0x7b, 0xde,
+};
+
+static const u8 tuning_blk_pattern_8bit[] = {
+	0xff, 0xff, 0x00, 0xff, 0xff, 0xff, 0x00, 0x00,
+	0xff, 0xff, 0xcc, 0xcc, 0xcc, 0x33, 0xcc, 0xcc,
+	0xcc, 0x33, 0x33, 0xcc, 0xcc, 0xcc, 0xff, 0xff,
+	0xff, 0xee, 0xff, 0xff, 0xff, 0xee, 0xee, 0xff,
+	0xff, 0xff, 0xdd, 0xff, 0xff, 0xff, 0xdd, 0xdd,
+	0xff, 0xff, 0xff, 0xbb, 0xff, 0xff, 0xff, 0xbb,
+	0xbb, 0xff, 0xff, 0xff, 0x77, 0xff, 0xff, 0xff,
+	0x77, 0x77, 0xff, 0x77, 0xbb, 0xdd, 0xee, 0xff,
+	0xff, 0xff, 0xff, 0x00, 0xff, 0xff, 0xff, 0x00,
+	0x00, 0xff, 0xff, 0xcc, 0xcc, 0xcc, 0x33, 0xcc,
+	0xcc, 0xcc, 0x33, 0x33, 0xcc, 0xcc, 0xcc, 0xff,
+	0xff, 0xff, 0xee, 0xff, 0xff, 0xff, 0xee, 0xee,
+	0xff, 0xff, 0xff, 0xdd, 0xff, 0xff, 0xff, 0xdd,
+	0xdd, 0xff, 0xff, 0xff, 0xbb, 0xff, 0xff, 0xff,
+	0xbb, 0xbb, 0xff, 0xff, 0xff, 0x77, 0xff, 0xff,
+	0xff, 0x77, 0x77, 0xff, 0x77, 0xbb, 0xdd, 0xee,
+};
+
+static inline bool dw_mci_fifo_reset(struct dw_mci *host);
+static inline bool dw_mci_ctrl_all_reset(struct dw_mci *host);
+
+#if defined(CONFIG_DEBUG_FS)
+static int dw_mci_req_show(struct seq_file *s, void *v)
+{
+	struct dw_mci_slot *slot = s->private;
+	struct mmc_request *mrq;
+	struct mmc_command *cmd;
+	struct mmc_command *stop;
+	struct mmc_data	*data;
+
+	/* Make sure we get a consistent snapshot */
+	spin_lock_bh(&slot->host->lock);
+	mrq = slot->mrq;
+
+	if (mrq) {
+		cmd = mrq->cmd;
+		data = mrq->data;
+		stop = mrq->stop;
+
+		if (cmd)
+			seq_printf(s,
+				   "CMD%u(0x%x) flg %x rsp %x %x %x %x err %d\n",
+				   cmd->opcode, cmd->arg, cmd->flags,
+				   cmd->resp[0], cmd->resp[1], cmd->resp[2],
+				   cmd->resp[2], cmd->error);
+		if (data)
+			seq_printf(s, "DATA %u / %u * %u flg %x err %d\n",
+				   data->bytes_xfered, data->blocks,
+				   data->blksz, data->flags, data->error);
+		if (stop)
+			seq_printf(s,
+				   "CMD%u(0x%x) flg %x rsp %x %x %x %x err %d\n",
+				   stop->opcode, stop->arg, stop->flags,
+				   stop->resp[0], stop->resp[1], stop->resp[2],
+				   stop->resp[2], stop->error);
+	}
+
+	spin_unlock_bh(&slot->host->lock);
+
+	return 0;
+}
+
+static int dw_mci_req_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, dw_mci_req_show, inode->i_private);
+}
+
+static const struct file_operations dw_mci_req_fops = {
+	.owner		= THIS_MODULE,
+	.open		= dw_mci_req_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static int dw_mci_regs_show(struct seq_file *s, void *v)
+{
+	seq_printf(s, "STATUS:\t0x%08x\n", SDMMC_STATUS);
+	seq_printf(s, "RINTSTS:\t0x%08x\n", SDMMC_RINTSTS);
+	seq_printf(s, "CMD:\t0x%08x\n", SDMMC_CMD);
+	seq_printf(s, "CTRL:\t0x%08x\n", SDMMC_CTRL);
+	seq_printf(s, "INTMASK:\t0x%08x\n", SDMMC_INTMASK);
+	seq_printf(s, "CLKENA:\t0x%08x\n", SDMMC_CLKENA);
+
+	return 0;
+}
+
+static int dw_mci_regs_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, dw_mci_regs_show, inode->i_private);
+}
+
+static const struct file_operations dw_mci_regs_fops = {
+	.owner		= THIS_MODULE,
+	.open		= dw_mci_regs_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static void dw_mci_init_debugfs(struct dw_mci_slot *slot)
+{
+	struct mmc_host	*mmc = slot->mmc;
+	struct dw_mci *host = slot->host;
+	struct dentry *root;
+	struct dentry *node;
+
+	root = mmc->debugfs_root;
+	if (!root)
+		return;
+
+	node = debugfs_create_file("regs", S_IRUSR, root, host,
+				   &dw_mci_regs_fops);
+	if (!node)
+		goto err;
+
+	node = debugfs_create_file("req", S_IRUSR, root, slot,
+				   &dw_mci_req_fops);
+	if (!node)
+		goto err;
+
+	node = debugfs_create_u32("state", S_IRUSR, root, (u32 *)&host->state);
+	if (!node)
+		goto err;
+
+	node = debugfs_create_x32("pending_events", S_IRUSR, root,
+				  (u32 *)&host->pending_events);
+	if (!node)
+		goto err;
+
+	node = debugfs_create_x32("completed_events", S_IRUSR, root,
+				  (u32 *)&host->completed_events);
+	if (!node)
+		goto err;
+
+	return;
+
+err:
+	dev_err(&mmc->class_dev, "failed to initialize debugfs for slot\n");
+}
+#endif /* defined(CONFIG_DEBUG_FS) */
+
+static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
+{
+	struct mmc_data	*data;
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
+	u32 cmdr;
+	cmd->error = -EINPROGRESS;
+
+	cmdr = cmd->opcode;
+
+	if (cmd->opcode == MMC_STOP_TRANSMISSION ||
+	    cmd->opcode == MMC_GO_IDLE_STATE ||
+	    cmd->opcode == MMC_GO_INACTIVE_STATE ||
+	    (cmd->opcode == SD_IO_RW_DIRECT &&
+	     ((cmd->arg >> 9) & 0x1FFFF) == SDIO_CCCR_ABORT))
+		cmdr |= SDMMC_CMD_STOP;
+	else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
+		cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		/* We expect a response, so set this bit */
+		cmdr |= SDMMC_CMD_RESP_EXP;
+		if (cmd->flags & MMC_RSP_136)
+			cmdr |= SDMMC_CMD_RESP_LONG;
+	}
+
+	if (cmd->flags & MMC_RSP_CRC)
+		cmdr |= SDMMC_CMD_RESP_CRC;
+
+	data = cmd->data;
+	if (data) {
+		cmdr |= SDMMC_CMD_DAT_EXP;
+		if (data->flags & MMC_DATA_STREAM)
+			cmdr |= SDMMC_CMD_STRM_MODE;
+		if (data->flags & MMC_DATA_WRITE)
+			cmdr |= SDMMC_CMD_DAT_WR;
+	}
+
+	if (drv_data && drv_data->prepare_command)
+		drv_data->prepare_command(slot->host, &cmdr);
+
+	return cmdr;
+}
+
+static u32 dw_mci_prep_stop_abort(struct dw_mci *host, struct mmc_command *cmd)
+{
+	struct mmc_command *stop;
+	u32 cmdr;
+
+	if (!cmd->data)
+		return 0;
+
+	stop = &host->stop_abort;
+	cmdr = cmd->opcode;
+	memset(stop, 0, sizeof(struct mmc_command));
+
+	if (cmdr == MMC_READ_SINGLE_BLOCK ||
+	    cmdr == MMC_READ_MULTIPLE_BLOCK ||
+	    cmdr == MMC_WRITE_BLOCK ||
+	    cmdr == MMC_WRITE_MULTIPLE_BLOCK) {
+		stop->opcode = MMC_STOP_TRANSMISSION;
+		stop->arg = 0;
+		stop->flags = MMC_RSP_R1B | MMC_CMD_AC;
+	} else if (cmdr == SD_IO_RW_EXTENDED) {
+		stop->opcode = SD_IO_RW_DIRECT;
+		stop->arg |= (1 << 31) | (0 << 28) | (SDIO_CCCR_ABORT << 9) |
+			     ((cmd->arg >> 28) & 0x7);
+		stop->flags = MMC_RSP_SPI_R5 | MMC_RSP_R5 | MMC_CMD_AC;
+	} else {
+		return 0;
+	}
+
+	cmdr = stop->opcode | SDMMC_CMD_STOP |
+		SDMMC_CMD_RESP_CRC | SDMMC_CMD_RESP_EXP;
+
+	return cmdr;
+}
+
+static void dw_mci_start_command(struct dw_mci *host,
+				 struct mmc_command *cmd, u32 cmd_flags)
+{
+	host->cmd = cmd;
+	dev_vdbg(host->dev,
+		 "start command: ARGR=0x%08x CMDR=0x%08x\n",
+		 cmd->arg, cmd_flags);
+
+	mci_writel(host, CMDARG, cmd->arg);
+	wmb();
+
+	mci_writel(host, CMD, cmd_flags | SDMMC_CMD_START);
+}
+
+static inline void send_stop_abort(struct dw_mci *host, struct mmc_data *data)
+{
+	struct mmc_command *stop = data->stop ? data->stop : &host->stop_abort;
+	dw_mci_start_command(host, stop, host->stop_cmdr);
+}
+
+/* DMA interface functions */
+static void dw_mci_stop_dma(struct dw_mci *host)
+{
+	if (host->using_dma) {
+		host->dma_ops->stop(host);
+		host->dma_ops->cleanup(host);
+	}
+
+	/* Data transfer was stopped by the interrupt handler */
+	set_bit(EVENT_XFER_COMPLETE, &host->pending_events);
+}
+
+static int dw_mci_get_dma_dir(struct mmc_data *data)
+{
+	if (data->flags & MMC_DATA_WRITE)
+		return DMA_TO_DEVICE;
+	else
+		return DMA_FROM_DEVICE;
+}
+
+#ifdef CONFIG_MMC_DW_IDMAC
+static void dw_mci_dma_cleanup(struct dw_mci *host)
+{
+	struct mmc_data *data = host->data;
+
+	if (data)
+		if (!data->host_cookie)
+			dma_unmap_sg(host->dev,
+				     data->sg,
+				     data->sg_len,
+				     dw_mci_get_dma_dir(data));
+}
+
+static void dw_mci_idmac_reset(struct dw_mci *host)
+{
+	u32 bmod = mci_readl(host, BMOD);
+	/* Software reset of DMA */
+	bmod |= SDMMC_IDMAC_SWRESET;
+	mci_writel(host, BMOD, bmod);
+}
+
+static void dw_mci_idmac_stop_dma(struct dw_mci *host)
+{
+	u32 temp;
+
+	/* Disable and reset the IDMAC interface */
+	temp = mci_readl(host, CTRL);
+	temp &= ~SDMMC_CTRL_USE_IDMAC;
+	temp |= SDMMC_CTRL_DMA_RESET;
+	mci_writel(host, CTRL, temp);
+
+	/* Stop the IDMAC running */
+	temp = mci_readl(host, BMOD);
+	temp &= ~(SDMMC_IDMAC_ENABLE | SDMMC_IDMAC_FB);
+	temp |= SDMMC_IDMAC_SWRESET;
+	mci_writel(host, BMOD, temp);
+}
+
+static void dw_mci_idmac_complete_dma(struct dw_mci *host)
+{
+	struct mmc_data *data = host->data;
+
+	dev_vdbg(host->dev, "DMA complete\n");
+
+	host->dma_ops->cleanup(host);
+
+	/*
+	 * If the card was removed, data will be NULL. No point in trying to
+	 * send the stop command or waiting for NBUSY in this case.
+	 */
+	if (data) {
+		set_bit(EVENT_XFER_COMPLETE, &host->pending_events);
+		tasklet_schedule(&host->tasklet);
+	}
+}
+
+static void dw_mci_translate_sglist(struct dw_mci *host, struct mmc_data *data,
+				    unsigned int sg_len)
+{
+	int i;
+	struct idmac_desc *desc = host->sg_cpu;
+
+	for (i = 0; i < sg_len; i++, desc++) {
+		unsigned int length = sg_dma_len(&data->sg[i]);
+		u32 mem_addr = sg_dma_address(&data->sg[i]);
+
+		/* Set the OWN bit and disable interrupts for this descriptor */
+		desc->des0 = IDMAC_DES0_OWN | IDMAC_DES0_DIC | IDMAC_DES0_CH;
+
+		/* Buffer length */
+		IDMAC_SET_BUFFER1_SIZE(desc, length);
+
+		/* Physical address to DMA to/from */
+		desc->des2 = mem_addr;
+	}
+
+	/* Set first descriptor */
+	desc = host->sg_cpu;
+	desc->des0 |= IDMAC_DES0_FD;
+
+	/* Set last descriptor */
+	desc = host->sg_cpu + (i - 1) * sizeof(struct idmac_desc);
+	desc->des0 &= ~(IDMAC_DES0_CH | IDMAC_DES0_DIC);
+	desc->des0 |= IDMAC_DES0_LD;
+
+	wmb();
+}
+
+static void dw_mci_idmac_start_dma(struct dw_mci *host, unsigned int sg_len)
+{
+	u32 temp;
+
+	dw_mci_translate_sglist(host, host->data, sg_len);
+
+	/* Select IDMAC interface */
+	temp = mci_readl(host, CTRL);
+	temp |= SDMMC_CTRL_USE_IDMAC;
+	mci_writel(host, CTRL, temp);
+
+	wmb();
+
+	/* Enable the IDMAC */
+	temp = mci_readl(host, BMOD);
+	temp |= SDMMC_IDMAC_ENABLE | SDMMC_IDMAC_FB;
+	mci_writel(host, BMOD, temp);
+
+	/* Start it running */
+	mci_writel(host, PLDMND, 1);
+}
+
+static int dw_mci_idmac_init(struct dw_mci *host)
+{
+	struct idmac_desc *p;
+	int i;
+
+	/* Number of descriptors in the ring buffer */
+	host->ring_size = PAGE_SIZE / sizeof(struct idmac_desc);
+
+	/* Forward link the descriptor list */
+	for (i = 0, p = host->sg_cpu; i < host->ring_size - 1; i++, p++)
+		p->des3 = host->sg_dma + (sizeof(struct idmac_desc) * (i + 1));
+
+	/* Set the last descriptor as the end-of-ring descriptor */
+	p->des3 = host->sg_dma;
+	p->des0 = IDMAC_DES0_ER;
+
+	dw_mci_idmac_reset(host);
+
+	/* Mask out interrupts - get Tx & Rx complete only */
+	mci_writel(host, IDSTS, IDMAC_INT_CLR);
+	mci_writel(host, IDINTEN, SDMMC_IDMAC_INT_NI | SDMMC_IDMAC_INT_RI |
+		   SDMMC_IDMAC_INT_TI);
+
+	/* Set the descriptor base address */
+	mci_writel(host, DBADDR, host->sg_dma);
+	return 0;
+}
+
+static const struct dw_mci_dma_ops dw_mci_idmac_ops = {
+	.init = dw_mci_idmac_init,
+	.start = dw_mci_idmac_start_dma,
+	.stop = dw_mci_idmac_stop_dma,
+	.complete = dw_mci_idmac_complete_dma,
+	.cleanup = dw_mci_dma_cleanup,
+};
+#endif /* CONFIG_MMC_DW_IDMAC */
+
+static int dw_mci_pre_dma_transfer(struct dw_mci *host,
+				   struct mmc_data *data,
+				   bool next)
+{
+	struct scatterlist *sg;
+	unsigned int i, sg_len;
+
+	if (!next && data->host_cookie)
+		return data->host_cookie;
+
+	/*
+	 * We don't do DMA on "complex" transfers, i.e. with
+	 * non-word-aligned buffers or lengths. Also, we don't bother
+	 * with all the DMA setup overhead for short transfers.
+	 */
+	if (data->blocks * data->blksz < DW_MCI_DMA_THRESHOLD)
+		return -EINVAL;
+
+	if (data->blksz & 3)
+		return -EINVAL;
+
+	for_each_sg(data->sg, sg, data->sg_len, i) {
+		if (sg->offset & 3 || sg->length & 3)
+			return -EINVAL;
+	}
+
+	sg_len = dma_map_sg(host->dev,
+			    data->sg,
+			    data->sg_len,
+			    dw_mci_get_dma_dir(data));
+	if (sg_len == 0)
+		return -EINVAL;
+
+	if (next)
+		data->host_cookie = sg_len;
+
+	return sg_len;
+}
+
+static void dw_mci_pre_req(struct mmc_host *mmc,
+			   struct mmc_request *mrq,
+			   bool is_first_req)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
+
+	if (!slot->host->use_dma || !data)
+		return;
+
+	if (data->host_cookie) {
+		data->host_cookie = 0;
+		return;
+	}
+
+	if (dw_mci_pre_dma_transfer(slot->host, mrq->data, 1) < 0)
+		data->host_cookie = 0;
+}
+
+static void dw_mci_post_req(struct mmc_host *mmc,
+			    struct mmc_request *mrq,
+			    int err)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
+
+	if (!slot->host->use_dma || !data)
+		return;
+
+	if (data->host_cookie)
+		dma_unmap_sg(slot->host->dev,
+			     data->sg,
+			     data->sg_len,
+			     dw_mci_get_dma_dir(data));
+	data->host_cookie = 0;
+}
+
+static void dw_mci_adjust_fifoth(struct dw_mci *host, struct mmc_data *data)
+{
+#ifdef CONFIG_MMC_DW_IDMAC
+	unsigned int blksz = data->blksz;
+	const u32 mszs[] = {1, 4, 8, 16, 32, 64, 128, 256};
+	u32 fifo_width = 1 << host->data_shift;
+	u32 blksz_depth = blksz / fifo_width, fifoth_val;
+	u32 msize = 0, rx_wmark = 1, tx_wmark, tx_wmark_invers;
+	int idx = (sizeof(mszs) / sizeof(mszs[0])) - 1;
+
+	tx_wmark = (host->fifo_depth) / 2;
+	tx_wmark_invers = host->fifo_depth - tx_wmark;
+
+	/*
+	 * MSIZE is '1',
+	 * if blksz is not a multiple of the FIFO width
+	 */
+	if (blksz % fifo_width) {
+		msize = 0;
+		rx_wmark = 1;
+		goto done;
+	}
+
+	do {
+		if (!((blksz_depth % mszs[idx]) ||
+		     (tx_wmark_invers % mszs[idx]))) {
+			msize = idx;
+			rx_wmark = mszs[idx] - 1;
+			break;
+		}
+	} while (--idx > 0);
+	/*
+	 * If idx is '0', it won't be tried
+	 * Thus, initial values are uesed
+	 */
+done:
+	fifoth_val = SDMMC_SET_FIFOTH(msize, rx_wmark, tx_wmark);
+	mci_writel(host, FIFOTH, fifoth_val);
+#endif
+}
+
+static void dw_mci_ctrl_rd_thld(struct dw_mci *host, struct mmc_data *data)
+{
+	unsigned int blksz = data->blksz;
+	u32 blksz_depth, fifo_depth;
+	u16 thld_size;
+
+	WARN_ON(!(data->flags & MMC_DATA_READ));
+
+	if (host->timing != MMC_TIMING_MMC_HS200 &&
+	    host->timing != MMC_TIMING_UHS_SDR104)
+		goto disable;
+
+	blksz_depth = blksz / (1 << host->data_shift);
+	fifo_depth = host->fifo_depth;
+
+	if (blksz_depth > fifo_depth)
+		goto disable;
+
+	/*
+	 * If (blksz_depth) >= (fifo_depth >> 1), should be 'thld_size <= blksz'
+	 * If (blksz_depth) <  (fifo_depth >> 1), should be thld_size = blksz
+	 * Currently just choose blksz.
+	 */
+	thld_size = blksz;
+	mci_writel(host, CDTHRCTL, SDMMC_SET_RD_THLD(thld_size, 1));
+	return;
+
+disable:
+	mci_writel(host, CDTHRCTL, SDMMC_SET_RD_THLD(0, 0));
+}
+
+static int dw_mci_submit_data_dma(struct dw_mci *host, struct mmc_data *data)
+{
+	int sg_len;
+	u32 temp;
+
+	host->using_dma = 0;
+
+	/* If we don't have a channel, we can't do DMA */
+	if (!host->use_dma)
+		return -ENODEV;
+
+	sg_len = dw_mci_pre_dma_transfer(host, data, 0);
+	if (sg_len < 0) {
+		host->dma_ops->stop(host);
+		return sg_len;
+	}
+
+	host->using_dma = 1;
+
+	dev_vdbg(host->dev,
+		 "sd sg_cpu: %#lx sg_dma: %#lx sg_len: %d\n",
+		 (unsigned long)host->sg_cpu, (unsigned long)host->sg_dma,
+		 sg_len);
+
+	/*
+	 * Decide the MSIZE and RX/TX Watermark.
+	 * If current block size is same with previous size,
+	 * no need to update fifoth.
+	 */
+	if (host->prev_blksz != data->blksz)
+		dw_mci_adjust_fifoth(host, data);
+
+	/* Enable the DMA interface */
+	temp = mci_readl(host, CTRL);
+	temp |= SDMMC_CTRL_DMA_ENABLE;
+	mci_writel(host, CTRL, temp);
+
+	/* Disable RX/TX IRQs, let DMA handle it */
+	temp = mci_readl(host, INTMASK);
+	temp  &= ~(SDMMC_INT_RXDR | SDMMC_INT_TXDR);
+	mci_writel(host, INTMASK, temp);
+
+	host->dma_ops->start(host, sg_len);
+
+	return 0;
+}
+
+static void dw_mci_submit_data(struct dw_mci *host, struct mmc_data *data)
+{
+	u32 temp;
+
+	data->error = -EINPROGRESS;
+
+	WARN_ON(host->data);
+	host->sg = NULL;
+	host->data = data;
+
+	if (data->flags & MMC_DATA_READ) {
+		host->dir_status = DW_MCI_RECV_STATUS;
+		dw_mci_ctrl_rd_thld(host, data);
+	} else {
+		host->dir_status = DW_MCI_SEND_STATUS;
+	}
+
+	if (dw_mci_submit_data_dma(host, data)) {
+		int flags = SG_MITER_ATOMIC;
+		if (host->data->flags & MMC_DATA_READ)
+			flags |= SG_MITER_TO_SG;
+		else
+			flags |= SG_MITER_FROM_SG;
+
+		sg_miter_start(&host->sg_miter, data->sg, data->sg_len, flags);
+		host->sg = data->sg;
+		host->part_buf_start = 0;
+		host->part_buf_count = 0;
+
+		mci_writel(host, RINTSTS, SDMMC_INT_TXDR | SDMMC_INT_RXDR);
+		temp = mci_readl(host, INTMASK);
+		temp |= SDMMC_INT_TXDR | SDMMC_INT_RXDR;
+		mci_writel(host, INTMASK, temp);
+
+		temp = mci_readl(host, CTRL);
+		temp &= ~SDMMC_CTRL_DMA_ENABLE;
+		mci_writel(host, CTRL, temp);
+
+		/*
+		 * Use the initial fifoth_val for PIO mode.
+		 * If next issued data may be transfered by DMA mode,
+		 * prev_blksz should be invalidated.
+		 */
+		mci_writel(host, FIFOTH, host->fifoth_val);
+		host->prev_blksz = 0;
+	} else {
+		/*
+		 * Keep the current block size.
+		 * It will be used to decide whether to update
+		 * fifoth register next time.
+		 */
+		host->prev_blksz = data->blksz;
+	}
+}
+
+static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg)
+{
+	struct dw_mci *host = slot->host;
+	unsigned long timeout = jiffies + msecs_to_jiffies(500);
+	unsigned int cmd_status = 0;
+
+	mci_writel(host, CMDARG, arg);
+	wmb();
+	mci_writel(host, CMD, SDMMC_CMD_START | cmd);
+
+	while (time_before(jiffies, timeout)) {
+		cmd_status = mci_readl(host, CMD);
+		if (!(cmd_status & SDMMC_CMD_START))
+			return;
+	}
+	dev_err(&slot->mmc->class_dev,
+		"Timeout sending command (cmd %#x arg %#x status %#x)\n",
+		cmd, arg, cmd_status);
+}
+
+static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
+{
+	struct dw_mci *host = slot->host;
+	unsigned int clock = slot->clock;
+	u32 div;
+	u32 clk_en_a;
+
+	if (!clock) {
+		mci_writel(host, CLKENA, 0);
+		mci_send_cmd(slot,
+			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+	} else if (clock != host->current_speed || force_clkinit) {
+		div = host->bus_hz / clock;
+		if (host->bus_hz % clock && host->bus_hz > clock)
+			/*
+			 * move the + 1 after the divide to prevent
+			 * over-clocking the card.
+			 */
+			div += 1;
+
+		div = (host->bus_hz != clock) ? DIV_ROUND_UP(div, 2) : 0;
+
+		if ((clock << div) != slot->__clk_old || force_clkinit)
+			dev_info(&slot->mmc->class_dev,
+				 "Bus speed (slot %d) = %dHz (slot req %dHz, actual %dHZ div = %d)\n",
+				 slot->id, host->bus_hz, clock,
+				 div ? ((host->bus_hz / div) >> 1) :
+				 host->bus_hz, div);
+
+		/* disable clock */
+		mci_writel(host, CLKENA, 0);
+		mci_writel(host, CLKSRC, 0);
+
+		/* inform CIU */
+		mci_send_cmd(slot,
+			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+
+		/* set clock to desired speed */
+		mci_writel(host, CLKDIV, div);
+
+		/* inform CIU */
+		mci_send_cmd(slot,
+			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+
+		/* enable clock; only low power if no SDIO */
+		clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
+		if (!(mci_readl(host, INTMASK) & SDMMC_INT_SDIO(slot->id)))
+			clk_en_a |= SDMMC_CLKEN_LOW_PWR << slot->id;
+		mci_writel(host, CLKENA, clk_en_a);
+
+		/* inform CIU */
+		mci_send_cmd(slot,
+			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+
+		/* keep the clock with reflecting clock dividor */
+		slot->__clk_old = clock << div;
+	}
+
+	host->current_speed = clock;
+
+	/* Set the current slot bus width */
+	mci_writel(host, CTYPE, (slot->ctype << slot->id));
+}
+
+static void __dw_mci_start_request(struct dw_mci *host,
+				   struct dw_mci_slot *slot,
+				   struct mmc_command *cmd)
+{
+	struct mmc_request *mrq;
+	struct mmc_data	*data;
+	u32 cmdflags;
+
+	mrq = slot->mrq;
+
+	host->cur_slot = slot;
+	host->mrq = mrq;
+
+	host->pending_events = 0;
+	host->completed_events = 0;
+	host->cmd_status = 0;
+	host->data_status = 0;
+	host->dir_status = 0;
+
+	data = cmd->data;
+	if (data) {
+		mci_writel(host, TMOUT, 0xFFFFFFFF);
+		mci_writel(host, BYTCNT, data->blksz*data->blocks);
+		mci_writel(host, BLKSIZ, data->blksz);
+	}
+
+	cmdflags = dw_mci_prepare_command(slot->mmc, cmd);
+
+	/* this is the first command, send the initialization clock */
+	if (test_and_clear_bit(DW_MMC_CARD_NEED_INIT, &slot->flags))
+		cmdflags |= SDMMC_CMD_INIT;
+
+	if (data) {
+		dw_mci_submit_data(host, data);
+		wmb();
+	}
+
+	dw_mci_start_command(host, cmd, cmdflags);
+
+	if (mrq->stop)
+		host->stop_cmdr = dw_mci_prepare_command(slot->mmc, mrq->stop);
+	else
+		host->stop_cmdr = dw_mci_prep_stop_abort(host, cmd);
+}
+
+static void dw_mci_start_request(struct dw_mci *host,
+				 struct dw_mci_slot *slot)
+{
+	struct mmc_request *mrq = slot->mrq;
+	struct mmc_command *cmd;
+
+	cmd = mrq->sbc ? mrq->sbc : mrq->cmd;
+	__dw_mci_start_request(host, slot, cmd);
+}
+
+/* must be called with host->lock held */
+static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
+				 struct mmc_request *mrq)
+{
+	dev_vdbg(&slot->mmc->class_dev, "queue request: state=%d\n",
+		 host->state);
+
+	slot->mrq = mrq;
+
+	if (host->state == STATE_IDLE) {
+		host->state = STATE_SENDING_CMD;
+		dw_mci_start_request(host, slot);
+	} else {
+		list_add_tail(&slot->queue_node, &host->queue);
+	}
+}
+
+static void dw_mci_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci *host = slot->host;
+
+	WARN_ON(slot->mrq);
+
+	/*
+	 * The check for card presence and queueing of the request must be
+	 * atomic, otherwise the card could be removed in between and the
+	 * request wouldn't fail until another card was inserted.
+	 */
+	spin_lock_bh(&host->lock);
+
+	if (!test_bit(DW_MMC_CARD_PRESENT, &slot->flags)) {
+		spin_unlock_bh(&host->lock);
+		mrq->cmd->error = -ENOMEDIUM;
+		mmc_request_done(mmc, mrq);
+		return;
+	}
+
+	dw_mci_queue_request(host, slot, mrq);
+
+	spin_unlock_bh(&host->lock);
+}
+
+static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
+	u32 regs;
+
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_4:
+		slot->ctype = SDMMC_CTYPE_4BIT;
+		break;
+	case MMC_BUS_WIDTH_8:
+		slot->ctype = SDMMC_CTYPE_8BIT;
+		break;
+	default:
+		/* set default 1 bit mode */
+		slot->ctype = SDMMC_CTYPE_1BIT;
+	}
+
+	regs = mci_readl(slot->host, UHS_REG);
+
+	/* DDR mode set */
+	if (ios->timing == MMC_TIMING_MMC_DDR52)
+		regs |= ((0x1 << slot->id) << 16);
+	else
+		regs &= ~((0x1 << slot->id) << 16);
+
+	mci_writel(slot->host, UHS_REG, regs);
+	slot->host->timing = ios->timing;
+
+	/*
+	 * Use mirror of ios->clock to prevent race with mmc
+	 * core ios update when finding the minimum.
+	 */
+	slot->clock = ios->clock;
+
+	if (drv_data && drv_data->set_ios)
+		drv_data->set_ios(slot->host, ios);
+
+	/* Slot specific timing and width adjustment */
+	dw_mci_setup_bus(slot, false);
+
+	switch (ios->power_mode) {
+	case MMC_POWER_UP:
+		set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
+		regs = mci_readl(slot->host, PWREN);
+		regs |= (1 << slot->id);
+		mci_writel(slot->host, PWREN, regs);
+		break;
+	case MMC_POWER_OFF:
+		regs = mci_readl(slot->host, PWREN);
+		regs &= ~(1 << slot->id);
+		mci_writel(slot->host, PWREN, regs);
+		break;
+	default:
+		break;
+	}
+}
+
+static int dw_mci_get_ro(struct mmc_host *mmc)
+{
+	int read_only;
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	int gpio_ro = mmc_gpio_get_ro(mmc);
+
+	/* Use platform get_ro function, else try on board write protect */
+	if (slot->quirks & DW_MCI_SLOT_QUIRK_NO_WRITE_PROTECT)
+		read_only = 0;
+	else if (!IS_ERR_VALUE(gpio_ro))
+		read_only = gpio_ro;
+	else
+		read_only =
+			mci_readl(slot->host, WRTPRT) & (1 << slot->id) ? 1 : 0;
+
+	dev_dbg(&mmc->class_dev, "card is %s\n",
+		read_only ? "read-only" : "read-write");
+
+	return read_only;
+}
+
+static int dw_mci_get_cd(struct mmc_host *mmc)
+{
+	int present;
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci_board *brd = slot->host->pdata;
+	struct dw_mci *host = slot->host;
+	int gpio_cd = mmc_gpio_get_cd(mmc);
+
+	/* Use platform get_cd function, else try onboard card detect */
+	if (brd->quirks & DW_MCI_QUIRK_BROKEN_CARD_DETECTION)
+		present = 1;
+	else if (!IS_ERR_VALUE(gpio_cd))
+		present = gpio_cd;
+	else
+		present = (mci_readl(slot->host, CDETECT) & (1 << slot->id))
+			== 0 ? 1 : 0;
+
+	spin_lock_bh(&host->lock);
+	if (present) {
+		set_bit(DW_MMC_CARD_PRESENT, &slot->flags);
+		dev_dbg(&mmc->class_dev, "card is present\n");
+	} else {
+		clear_bit(DW_MMC_CARD_PRESENT, &slot->flags);
+		dev_dbg(&mmc->class_dev, "card is not present\n");
+	}
+	spin_unlock_bh(&host->lock);
+
+	return present;
+}
+
+/*
+ * Disable lower power mode.
+ *
+ * Low power mode will stop the card clock when idle.  According to the
+ * description of the CLKENA register we should disable low power mode
+ * for SDIO cards if we need SDIO interrupts to work.
+ *
+ * This function is fast if low power mode is already disabled.
+ */
+static void dw_mci_disable_low_power(struct dw_mci_slot *slot)
+{
+	struct dw_mci *host = slot->host;
+	u32 clk_en_a;
+	const u32 clken_low_pwr = SDMMC_CLKEN_LOW_PWR << slot->id;
+
+	clk_en_a = mci_readl(host, CLKENA);
+
+	if (clk_en_a & clken_low_pwr) {
+		mci_writel(host, CLKENA, clk_en_a & ~clken_low_pwr);
+		mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
+			     SDMMC_CMD_PRV_DAT_WAIT, 0);
+	}
+}
+
+static void dw_mci_enable_sdio_irq(struct mmc_host *mmc, int enb)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci *host = slot->host;
+	u32 int_mask;
+
+	/* Enable/disable Slot Specific SDIO interrupt */
+	int_mask = mci_readl(host, INTMASK);
+	if (enb) {
+		/*
+		 * Turn off low power mode if it was enabled.  This is a bit of
+		 * a heavy operation and we disable / enable IRQs a lot, so
+		 * we'll leave low power mode disabled and it will get
+		 * re-enabled again in dw_mci_setup_bus().
+		 */
+		dw_mci_disable_low_power(slot);
+
+		mci_writel(host, INTMASK,
+			   (int_mask | SDMMC_INT_SDIO(slot->id)));
+	} else {
+		mci_writel(host, INTMASK,
+			   (int_mask & ~SDMMC_INT_SDIO(slot->id)));
+	}
+}
+
+static int dw_mci_execute_tuning(struct mmc_host *mmc, u32 opcode)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci *host = slot->host;
+	const struct dw_mci_drv_data *drv_data = host->drv_data;
+	struct dw_mci_tuning_data tuning_data;
+	int err = -ENOSYS;
+
+	if (opcode == MMC_SEND_TUNING_BLOCK_HS200) {
+		if (mmc->ios.bus_width == MMC_BUS_WIDTH_8) {
+			tuning_data.blk_pattern = tuning_blk_pattern_8bit;
+			tuning_data.blksz = sizeof(tuning_blk_pattern_8bit);
+		} else if (mmc->ios.bus_width == MMC_BUS_WIDTH_4) {
+			tuning_data.blk_pattern = tuning_blk_pattern_4bit;
+			tuning_data.blksz = sizeof(tuning_blk_pattern_4bit);
+		} else {
+			return -EINVAL;
+		}
+	} else if (opcode == MMC_SEND_TUNING_BLOCK) {
+		tuning_data.blk_pattern = tuning_blk_pattern_4bit;
+		tuning_data.blksz = sizeof(tuning_blk_pattern_4bit);
+	} else {
+		dev_err(host->dev,
+			"Undefined command(%d) for tuning\n", opcode);
+		return -EINVAL;
+	}
+
+	if (drv_data && drv_data->execute_tuning)
+		err = drv_data->execute_tuning(slot, opcode, &tuning_data);
+	return err;
+}
+
+static const struct mmc_host_ops dw_mci_ops = {
+	.request		= dw_mci_request,
+	.pre_req		= dw_mci_pre_req,
+	.post_req		= dw_mci_post_req,
+	.set_ios		= dw_mci_set_ios,
+	.get_ro			= dw_mci_get_ro,
+	.get_cd			= dw_mci_get_cd,
+	.enable_sdio_irq	= dw_mci_enable_sdio_irq,
+	.execute_tuning		= dw_mci_execute_tuning,
+};
+
+static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
+	__releases(&host->lock)
+	__acquires(&host->lock)
+{
+	struct dw_mci_slot *slot;
+	struct mmc_host	*prev_mmc = host->cur_slot->mmc;
+
+	WARN_ON(host->cmd || host->data);
+
+	host->cur_slot->mrq = NULL;
+	host->mrq = NULL;
+	if (!list_empty(&host->queue)) {
+		slot = list_entry(host->queue.next,
+				  struct dw_mci_slot, queue_node);
+		list_del(&slot->queue_node);
+		dev_vdbg(host->dev, "list not empty: %s is next\n",
+			 mmc_hostname(slot->mmc));
+		host->state = STATE_SENDING_CMD;
+		dw_mci_start_request(host, slot);
+	} else {
+		dev_vdbg(host->dev, "list empty\n");
+		host->state = STATE_IDLE;
+	}
+
+	spin_unlock(&host->lock);
+	mmc_request_done(prev_mmc, mrq);
+	spin_lock(&host->lock);
+}
+
+static int dw_mci_command_complete(struct dw_mci *host, struct mmc_command *cmd)
+{
+	u32 status = host->cmd_status;
+
+	host->cmd_status = 0;
+
+	/* Read the response from the card (up to 16 bytes) */
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		if (cmd->flags & MMC_RSP_136) {
+			cmd->resp[3] = mci_readl(host, RESP0);
+			cmd->resp[2] = mci_readl(host, RESP1);
+			cmd->resp[1] = mci_readl(host, RESP2);
+			cmd->resp[0] = mci_readl(host, RESP3);
+		} else {
+			cmd->resp[0] = mci_readl(host, RESP0);
+			cmd->resp[1] = 0;
+			cmd->resp[2] = 0;
+			cmd->resp[3] = 0;
+		}
+	}
+
+	if (status & SDMMC_INT_RTO)
+		cmd->error = -ETIMEDOUT;
+	else if ((cmd->flags & MMC_RSP_CRC) && (status & SDMMC_INT_RCRC))
+		cmd->error = -EILSEQ;
+	else if (status & SDMMC_INT_RESP_ERR)
+		cmd->error = -EIO;
+	else
+		cmd->error = 0;
+
+	if (cmd->error) {
+		/* newer ip versions need a delay between retries */
+		if (host->quirks & DW_MCI_QUIRK_RETRY_DELAY)
+			mdelay(20);
+	}
+
+	return cmd->error;
+}
+
+static int dw_mci_data_complete(struct dw_mci *host, struct mmc_data *data)
+{
+	u32 status = host->data_status;
+
+	if (status & DW_MCI_DATA_ERROR_FLAGS) {
+		if (status & SDMMC_INT_DRTO) {
+			data->error = -ETIMEDOUT;
+		} else if (status & SDMMC_INT_DCRC) {
+			data->error = -EILSEQ;
+		} else if (status & SDMMC_INT_EBE) {
+			if (host->dir_status ==
+				DW_MCI_SEND_STATUS) {
+				/*
+				 * No data CRC status was returned.
+				 * The number of bytes transferred
+				 * will be exaggerated in PIO mode.
+				 */
+				data->bytes_xfered = 0;
+				data->error = -ETIMEDOUT;
+			} else if (host->dir_status ==
+					DW_MCI_RECV_STATUS) {
+				data->error = -EIO;
+			}
+		} else {
+			/* SDMMC_INT_SBE is included */
+			data->error = -EIO;
+		}
+
+		dev_dbg(host->dev, "data error, status 0x%08x\n", status);
+
+		/*
+		 * After an error, there may be data lingering
+		 * in the FIFO
+		 */
+		dw_mci_fifo_reset(host);
+	} else {
+		data->bytes_xfered = data->blocks * data->blksz;
+		data->error = 0;
+	}
+
+	return data->error;
+}
+
+static void dw_mci_tasklet_func(unsigned long priv)
+{
+	struct dw_mci *host = (struct dw_mci *)priv;
+	struct mmc_data	*data;
+	struct mmc_command *cmd;
+	struct mmc_request *mrq;
+	enum dw_mci_state state;
+	enum dw_mci_state prev_state;
+	unsigned int err;
+
+	spin_lock(&host->lock);
+
+	state = host->state;
+	data = host->data;
+	mrq = host->mrq;
+
+	do {
+		prev_state = state;
+
+		switch (state) {
+		case STATE_IDLE:
+			break;
+
+		case STATE_SENDING_CMD:
+			if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
+						&host->pending_events))
+				break;
+
+			cmd = host->cmd;
+			host->cmd = NULL;
+			set_bit(EVENT_CMD_COMPLETE, &host->completed_events);
+			err = dw_mci_command_complete(host, cmd);
+			if (cmd == mrq->sbc && !err) {
+				prev_state = state = STATE_SENDING_CMD;
+				__dw_mci_start_request(host, host->cur_slot,
+						       mrq->cmd);
+				goto unlock;
+			}
+
+			if (cmd->data && err) {
+				dw_mci_stop_dma(host);
+				send_stop_abort(host, data);
+				state = STATE_SENDING_STOP;
+				break;
+			}
+
+			if (!cmd->data || err) {
+				dw_mci_request_end(host, mrq);
+				goto unlock;
+			}
+
+			prev_state = state = STATE_SENDING_DATA;
+			/* fall through */
+
+		case STATE_SENDING_DATA:
+			if (test_and_clear_bit(EVENT_DATA_ERROR,
+					       &host->pending_events)) {
+				dw_mci_stop_dma(host);
+				send_stop_abort(host, data);
+				state = STATE_DATA_ERROR;
+				break;
+			}
+
+			if (!test_and_clear_bit(EVENT_XFER_COMPLETE,
+						&host->pending_events))
+				break;
+
+			set_bit(EVENT_XFER_COMPLETE, &host->completed_events);
+			prev_state = state = STATE_DATA_BUSY;
+			/* fall through */
+
+		case STATE_DATA_BUSY:
+			if (!test_and_clear_bit(EVENT_DATA_COMPLETE,
+						&host->pending_events))
+				break;
+
+			host->data = NULL;
+			set_bit(EVENT_DATA_COMPLETE, &host->completed_events);
+			err = dw_mci_data_complete(host, data);
+
+			if (!err) {
+				if (!data->stop || mrq->sbc) {
+					if (mrq->sbc && data->stop)
+						data->stop->error = 0;
+					dw_mci_request_end(host, mrq);
+					goto unlock;
+				}
+
+				/* stop command for open-ended transfer*/
+				if (data->stop)
+					send_stop_abort(host, data);
+			}
+
+			/*
+			 * If err has non-zero,
+			 * stop-abort command has been already issued.
+			 */
+			prev_state = state = STATE_SENDING_STOP;
+
+			/* fall through */
+
+		case STATE_SENDING_STOP:
+			if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
+						&host->pending_events))
+				break;
+
+			/* CMD error in data command */
+			if (mrq->cmd->error && mrq->data)
+				dw_mci_fifo_reset(host);
+
+			host->cmd = NULL;
+			host->data = NULL;
+
+			if (mrq->stop)
+				dw_mci_command_complete(host, mrq->stop);
+			else
+				host->cmd_status = 0;
+
+			dw_mci_request_end(host, mrq);
+			goto unlock;
+
+		case STATE_DATA_ERROR:
+			if (!test_and_clear_bit(EVENT_XFER_COMPLETE,
+						&host->pending_events))
+				break;
+
+			state = STATE_DATA_BUSY;
+			break;
+		}
+	} while (state != prev_state);
+
+	host->state = state;
+unlock:
+	spin_unlock(&host->lock);
+
+}
+
+/* push final bytes to part_buf, only use during push */
+static void dw_mci_set_part_bytes(struct dw_mci *host, void *buf, int cnt)
+{
+	memcpy((void *)&host->part_buf, buf, cnt);
+	host->part_buf_count = cnt;
+}
+
+/* append bytes to part_buf, only use during push */
+static int dw_mci_push_part_bytes(struct dw_mci *host, void *buf, int cnt)
+{
+	cnt = min(cnt, (1 << host->data_shift) - host->part_buf_count);
+	memcpy((void *)&host->part_buf + host->part_buf_count, buf, cnt);
+	host->part_buf_count += cnt;
+	return cnt;
+}
+
+/* pull first bytes from part_buf, only use during pull */
+static int dw_mci_pull_part_bytes(struct dw_mci *host, void *buf, int cnt)
+{
+	cnt = min(cnt, (int)host->part_buf_count);
+	if (cnt) {
+		memcpy(buf, (void *)&host->part_buf + host->part_buf_start,
+		       cnt);
+		host->part_buf_count -= cnt;
+		host->part_buf_start += cnt;
+	}
+	return cnt;
+}
+
+/* pull final bytes from the part_buf, assuming it's just been filled */
+static void dw_mci_pull_final_bytes(struct dw_mci *host, void *buf, int cnt)
+{
+	memcpy(buf, &host->part_buf, cnt);
+	host->part_buf_start = cnt;
+	host->part_buf_count = (1 << host->data_shift) - cnt;
+}
+
+static void dw_mci_push_data16(struct dw_mci *host, void *buf, int cnt)
+{
+	struct mmc_data *data = host->data;
+	int init_cnt = cnt;
+
+	/* try and push anything in the part_buf */
+	if (unlikely(host->part_buf_count)) {
+		int len = dw_mci_push_part_bytes(host, buf, cnt);
+		buf += len;
+		cnt -= len;
+		if (host->part_buf_count == 2) {
+			mci_writew(host, DATA(host->data_offset),
+					host->part_buf16);
+			host->part_buf_count = 0;
+		}
+	}
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x1)) {
+		while (cnt >= 2) {
+			u16 aligned_buf[64];
+			int len = min(cnt & -2, (int)sizeof(aligned_buf));
+			int items = len >> 1;
+			int i;
+			/* memcpy from input buffer into aligned buffer */
+			memcpy(aligned_buf, buf, len);
+			buf += len;
+			cnt -= len;
+			/* push data from aligned buffer into fifo */
+			for (i = 0; i < items; ++i)
+				mci_writew(host, DATA(host->data_offset),
+						aligned_buf[i]);
+		}
+	} else
+#endif
+	{
+		u16 *pdata = buf;
+		for (; cnt >= 2; cnt -= 2)
+			mci_writew(host, DATA(host->data_offset), *pdata++);
+		buf = pdata;
+	}
+	/* put anything remaining in the part_buf */
+	if (cnt) {
+		dw_mci_set_part_bytes(host, buf, cnt);
+		 /* Push data if we have reached the expected data length */
+		if ((data->bytes_xfered + init_cnt) ==
+		    (data->blksz * data->blocks))
+			mci_writew(host, DATA(host->data_offset),
+				   host->part_buf16);
+	}
+}
+
+static void dw_mci_pull_data16(struct dw_mci *host, void *buf, int cnt)
+{
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x1)) {
+		while (cnt >= 2) {
+			/* pull data from fifo into aligned buffer */
+			u16 aligned_buf[64];
+			int len = min(cnt & -2, (int)sizeof(aligned_buf));
+			int items = len >> 1;
+			int i;
+			for (i = 0; i < items; ++i)
+				aligned_buf[i] = mci_readw(host,
+						DATA(host->data_offset));
+			/* memcpy from aligned buffer into output buffer */
+			memcpy(buf, aligned_buf, len);
+			buf += len;
+			cnt -= len;
+		}
+	} else
+#endif
+	{
+		u16 *pdata = buf;
+		for (; cnt >= 2; cnt -= 2)
+			*pdata++ = mci_readw(host, DATA(host->data_offset));
+		buf = pdata;
+	}
+	if (cnt) {
+		host->part_buf16 = mci_readw(host, DATA(host->data_offset));
+		dw_mci_pull_final_bytes(host, buf, cnt);
+	}
+}
+
+static void dw_mci_push_data32(struct dw_mci *host, void *buf, int cnt)
+{
+	struct mmc_data *data = host->data;
+	int init_cnt = cnt;
+
+	/* try and push anything in the part_buf */
+	if (unlikely(host->part_buf_count)) {
+		int len = dw_mci_push_part_bytes(host, buf, cnt);
+		buf += len;
+		cnt -= len;
+		if (host->part_buf_count == 4) {
+			mci_writel(host, DATA(host->data_offset),
+					host->part_buf32);
+			host->part_buf_count = 0;
+		}
+	}
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x3)) {
+		while (cnt >= 4) {
+			u32 aligned_buf[32];
+			int len = min(cnt & -4, (int)sizeof(aligned_buf));
+			int items = len >> 2;
+			int i;
+			/* memcpy from input buffer into aligned buffer */
+			memcpy(aligned_buf, buf, len);
+			buf += len;
+			cnt -= len;
+			/* push data from aligned buffer into fifo */
+			for (i = 0; i < items; ++i)
+				mci_writel(host, DATA(host->data_offset),
+						aligned_buf[i]);
+		}
+	} else
+#endif
+	{
+		u32 *pdata = buf;
+		for (; cnt >= 4; cnt -= 4)
+			mci_writel(host, DATA(host->data_offset), *pdata++);
+		buf = pdata;
+	}
+	/* put anything remaining in the part_buf */
+	if (cnt) {
+		dw_mci_set_part_bytes(host, buf, cnt);
+		 /* Push data if we have reached the expected data length */
+		if ((data->bytes_xfered + init_cnt) ==
+		    (data->blksz * data->blocks))
+			mci_writel(host, DATA(host->data_offset),
+				   host->part_buf32);
+	}
+}
+
+static void dw_mci_pull_data32(struct dw_mci *host, void *buf, int cnt)
+{
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x3)) {
+		while (cnt >= 4) {
+			/* pull data from fifo into aligned buffer */
+			u32 aligned_buf[32];
+			int len = min(cnt & -4, (int)sizeof(aligned_buf));
+			int items = len >> 2;
+			int i;
+			for (i = 0; i < items; ++i)
+				aligned_buf[i] = mci_readl(host,
+						DATA(host->data_offset));
+			/* memcpy from aligned buffer into output buffer */
+			memcpy(buf, aligned_buf, len);
+			buf += len;
+			cnt -= len;
+		}
+	} else
+#endif
+	{
+		u32 *pdata = buf;
+		for (; cnt >= 4; cnt -= 4)
+			*pdata++ = mci_readl(host, DATA(host->data_offset));
+		buf = pdata;
+	}
+	if (cnt) {
+		host->part_buf32 = mci_readl(host, DATA(host->data_offset));
+		dw_mci_pull_final_bytes(host, buf, cnt);
+	}
+}
+
+static void dw_mci_push_data64(struct dw_mci *host, void *buf, int cnt)
+{
+	struct mmc_data *data = host->data;
+	int init_cnt = cnt;
+
+	/* try and push anything in the part_buf */
+	if (unlikely(host->part_buf_count)) {
+		int len = dw_mci_push_part_bytes(host, buf, cnt);
+		buf += len;
+		cnt -= len;
+
+		if (host->part_buf_count == 8) {
+			mci_writeq(host, DATA(host->data_offset),
+					host->part_buf);
+			host->part_buf_count = 0;
+		}
+	}
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x7)) {
+		while (cnt >= 8) {
+			u64 aligned_buf[16];
+			int len = min(cnt & -8, (int)sizeof(aligned_buf));
+			int items = len >> 3;
+			int i;
+			/* memcpy from input buffer into aligned buffer */
+			memcpy(aligned_buf, buf, len);
+			buf += len;
+			cnt -= len;
+			/* push data from aligned buffer into fifo */
+			for (i = 0; i < items; ++i)
+				mci_writeq(host, DATA(host->data_offset),
+						aligned_buf[i]);
+		}
+	} else
+#endif
+	{
+		u64 *pdata = buf;
+		for (; cnt >= 8; cnt -= 8)
+			mci_writeq(host, DATA(host->data_offset), *pdata++);
+		buf = pdata;
+	}
+	/* put anything remaining in the part_buf */
+	if (cnt) {
+		dw_mci_set_part_bytes(host, buf, cnt);
+		/* Push data if we have reached the expected data length */
+		if ((data->bytes_xfered + init_cnt) ==
+		    (data->blksz * data->blocks))
+			mci_writeq(host, DATA(host->data_offset),
+				   host->part_buf);
+	}
+}
+
+static void dw_mci_pull_data64(struct dw_mci *host, void *buf, int cnt)
+{
+#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
+	if (unlikely((unsigned long)buf & 0x7)) {
+		while (cnt >= 8) {
+			/* pull data from fifo into aligned buffer */
+			u64 aligned_buf[16];
+			int len = min(cnt & -8, (int)sizeof(aligned_buf));
+			int items = len >> 3;
+			int i;
+			for (i = 0; i < items; ++i)
+				aligned_buf[i] = mci_readq(host,
+						DATA(host->data_offset));
+			/* memcpy from aligned buffer into output buffer */
+			memcpy(buf, aligned_buf, len);
+			buf += len;
+			cnt -= len;
+		}
+	} else
+#endif
+	{
+		u64 *pdata = buf;
+		for (; cnt >= 8; cnt -= 8)
+			*pdata++ = mci_readq(host, DATA(host->data_offset));
+		buf = pdata;
+	}
+	if (cnt) {
+		host->part_buf = mci_readq(host, DATA(host->data_offset));
+		dw_mci_pull_final_bytes(host, buf, cnt);
+	}
+}
+
+static void dw_mci_pull_data(struct dw_mci *host, void *buf, int cnt)
+{
+	int len;
+
+	/* get remaining partial bytes */
+	len = dw_mci_pull_part_bytes(host, buf, cnt);
+	if (unlikely(len == cnt))
+		return;
+	buf += len;
+	cnt -= len;
+
+	/* get the rest of the data */
+	host->pull_data(host, buf, cnt);
+}
+
+static void dw_mci_read_data_pio(struct dw_mci *host, bool dto)
+{
+	struct sg_mapping_iter *sg_miter = &host->sg_miter;
+	void *buf;
+	unsigned int offset;
+	struct mmc_data	*data = host->data;
+	int shift = host->data_shift;
+	u32 status;
+	unsigned int len;
+	unsigned int remain, fcnt;
+
+	do {
+		if (!sg_miter_next(sg_miter))
+			goto done;
+
+		host->sg = sg_miter->piter.sg;
+		buf = sg_miter->addr;
+		remain = sg_miter->length;
+		offset = 0;
+
+		do {
+			fcnt = (SDMMC_GET_FCNT(mci_readl(host, STATUS))
+					<< shift) + host->part_buf_count;
+			len = min(remain, fcnt);
+			if (!len)
+				break;
+			dw_mci_pull_data(host, (void *)(buf + offset), len);
+			data->bytes_xfered += len;
+			offset += len;
+			remain -= len;
+		} while (remain);
+
+		sg_miter->consumed = offset;
+		status = mci_readl(host, MINTSTS);
+		mci_writel(host, RINTSTS, SDMMC_INT_RXDR);
+	/* if the RXDR is ready read again */
+	} while ((status & SDMMC_INT_RXDR) ||
+		 (dto && SDMMC_GET_FCNT(mci_readl(host, STATUS))));
+
+	if (!remain) {
+		if (!sg_miter_next(sg_miter))
+			goto done;
+		sg_miter->consumed = 0;
+	}
+	sg_miter_stop(sg_miter);
+	return;
+
+done:
+	sg_miter_stop(sg_miter);
+	host->sg = NULL;
+	smp_wmb();
+	set_bit(EVENT_XFER_COMPLETE, &host->pending_events);
+}
+
+static void dw_mci_write_data_pio(struct dw_mci *host)
+{
+	struct sg_mapping_iter *sg_miter = &host->sg_miter;
+	void *buf;
+	unsigned int offset;
+	struct mmc_data	*data = host->data;
+	int shift = host->data_shift;
+	u32 status;
+	unsigned int len;
+	unsigned int fifo_depth = host->fifo_depth;
+	unsigned int remain, fcnt;
+
+	do {
+		if (!sg_miter_next(sg_miter))
+			goto done;
+
+		host->sg = sg_miter->piter.sg;
+		buf = sg_miter->addr;
+		remain = sg_miter->length;
+		offset = 0;
+
+		do {
+			fcnt = ((fifo_depth -
+				 SDMMC_GET_FCNT(mci_readl(host, STATUS)))
+					<< shift) - host->part_buf_count;
+			len = min(remain, fcnt);
+			if (!len)
+				break;
+			host->push_data(host, (void *)(buf + offset), len);
+			data->bytes_xfered += len;
+			offset += len;
+			remain -= len;
+		} while (remain);
+
+		sg_miter->consumed = offset;
+		status = mci_readl(host, MINTSTS);
+		mci_writel(host, RINTSTS, SDMMC_INT_TXDR);
+	} while (status & SDMMC_INT_TXDR); /* if TXDR write again */
+
+	if (!remain) {
+		if (!sg_miter_next(sg_miter))
+			goto done;
+		sg_miter->consumed = 0;
+	}
+	sg_miter_stop(sg_miter);
+	return;
+
+done:
+	sg_miter_stop(sg_miter);
+	host->sg = NULL;
+	smp_wmb();
+	set_bit(EVENT_XFER_COMPLETE, &host->pending_events);
+}
+
+static void dw_mci_cmd_interrupt(struct dw_mci *host, u32 status)
+{
+	if (!host->cmd_status)
+		host->cmd_status = status;
+
+	smp_wmb();
+
+	set_bit(EVENT_CMD_COMPLETE, &host->pending_events);
+	tasklet_schedule(&host->tasklet);
+}
+
+static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
+{
+	struct dw_mci *host = dev_id;
+	u32 pending;
+	int i;
+
+	pending = mci_readl(host, MINTSTS); /* read-only mask reg */
+
+	/*
+	 * DTO fix - version 2.10a and below, and only if internal DMA
+	 * is configured.
+	 */
+	if (host->quirks & DW_MCI_QUIRK_IDMAC_DTO) {
+		if (!pending &&
+		    ((mci_readl(host, STATUS) >> 17) & 0x1fff))
+			pending |= SDMMC_INT_DATA_OVER;
+	}
+
+	if (pending) {
+		if (pending & DW_MCI_CMD_ERROR_FLAGS) {
+			mci_writel(host, RINTSTS, DW_MCI_CMD_ERROR_FLAGS);
+			host->cmd_status = pending;
+			smp_wmb();
+			set_bit(EVENT_CMD_COMPLETE, &host->pending_events);
+		}
+
+		if (pending & DW_MCI_DATA_ERROR_FLAGS) {
+			/* if there is an error report DATA_ERROR */
+			mci_writel(host, RINTSTS, DW_MCI_DATA_ERROR_FLAGS);
+			host->data_status = pending;
+			smp_wmb();
+			set_bit(EVENT_DATA_ERROR, &host->pending_events);
+			tasklet_schedule(&host->tasklet);
+		}
+
+		if (pending & SDMMC_INT_DATA_OVER) {
+			mci_writel(host, RINTSTS, SDMMC_INT_DATA_OVER);
+			if (!host->data_status)
+				host->data_status = pending;
+			smp_wmb();
+			if (host->dir_status == DW_MCI_RECV_STATUS) {
+				if (host->sg != NULL)
+					dw_mci_read_data_pio(host, true);
+			}
+			set_bit(EVENT_DATA_COMPLETE, &host->pending_events);
+			tasklet_schedule(&host->tasklet);
+		}
+
+		if (pending & SDMMC_INT_RXDR) {
+			mci_writel(host, RINTSTS, SDMMC_INT_RXDR);
+			if (host->dir_status == DW_MCI_RECV_STATUS && host->sg)
+				dw_mci_read_data_pio(host, false);
+		}
+
+		if (pending & SDMMC_INT_TXDR) {
+			mci_writel(host, RINTSTS, SDMMC_INT_TXDR);
+			if (host->dir_status == DW_MCI_SEND_STATUS && host->sg)
+				dw_mci_write_data_pio(host);
+		}
+
+		if (pending & SDMMC_INT_CMD_DONE) {
+			mci_writel(host, RINTSTS, SDMMC_INT_CMD_DONE);
+			dw_mci_cmd_interrupt(host, pending);
+		}
+
+		if (pending & SDMMC_INT_CD) {
+			mci_writel(host, RINTSTS, SDMMC_INT_CD);
+			queue_work(host->card_workqueue, &host->card_work);
+		}
+
+		/* Handle SDIO Interrupts */
+		for (i = 0; i < host->num_slots; i++) {
+			struct dw_mci_slot *slot = host->slot[i];
+			if (pending & SDMMC_INT_SDIO(i)) {
+				mci_writel(host, RINTSTS, SDMMC_INT_SDIO(i));
+				mmc_signal_sdio_irq(slot->mmc);
+			}
+		}
+
+	}
+
+#ifdef CONFIG_MMC_DW_IDMAC
+	/* Handle DMA interrupts */
+	pending = mci_readl(host, IDSTS);
+	if (pending & (SDMMC_IDMAC_INT_TI | SDMMC_IDMAC_INT_RI)) {
+		mci_writel(host, IDSTS, SDMMC_IDMAC_INT_TI | SDMMC_IDMAC_INT_RI);
+		mci_writel(host, IDSTS, SDMMC_IDMAC_INT_NI);
+		host->dma_ops->complete(host);
+	}
+#endif
+
+	return IRQ_HANDLED;
+}
+
+static void dw_mci_work_routine_card(struct work_struct *work)
+{
+	struct dw_mci *host = container_of(work, struct dw_mci, card_work);
+	int i;
+
+	for (i = 0; i < host->num_slots; i++) {
+		struct dw_mci_slot *slot = host->slot[i];
+		struct mmc_host *mmc = slot->mmc;
+		struct mmc_request *mrq;
+		int present;
+
+		present = dw_mci_get_cd(mmc);
+		while (present != slot->last_detect_state) {
+			dev_dbg(&slot->mmc->class_dev, "card %s\n",
+				present ? "inserted" : "removed");
+
+			spin_lock_bh(&host->lock);
+
+			/* Card change detected */
+			slot->last_detect_state = present;
+
+			/* Clean up queue if present */
+			mrq = slot->mrq;
+			if (mrq) {
+				if (mrq == host->mrq) {
+					host->data = NULL;
+					host->cmd = NULL;
+
+					switch (host->state) {
+					case STATE_IDLE:
+						break;
+					case STATE_SENDING_CMD:
+						mrq->cmd->error = -ENOMEDIUM;
+						if (!mrq->data)
+							break;
+						/* fall through */
+					case STATE_SENDING_DATA:
+						mrq->data->error = -ENOMEDIUM;
+						dw_mci_stop_dma(host);
+						break;
+					case STATE_DATA_BUSY:
+					case STATE_DATA_ERROR:
+						if (mrq->data->error == -EINPROGRESS)
+							mrq->data->error = -ENOMEDIUM;
+						/* fall through */
+					case STATE_SENDING_STOP:
+						if (mrq->stop)
+							mrq->stop->error = -ENOMEDIUM;
+						break;
+					}
+
+					dw_mci_request_end(host, mrq);
+				} else {
+					list_del(&slot->queue_node);
+					mrq->cmd->error = -ENOMEDIUM;
+					if (mrq->data)
+						mrq->data->error = -ENOMEDIUM;
+					if (mrq->stop)
+						mrq->stop->error = -ENOMEDIUM;
+
+					spin_unlock(&host->lock);
+					mmc_request_done(slot->mmc, mrq);
+					spin_lock(&host->lock);
+				}
+			}
+
+			/* Power down slot */
+			if (present == 0) {
+				/* Clear down the FIFO */
+				dw_mci_fifo_reset(host);
+#ifdef CONFIG_MMC_DW_IDMAC
+				dw_mci_idmac_reset(host);
+#endif
+
+			}
+
+			spin_unlock_bh(&host->lock);
+
+			present = dw_mci_get_cd(mmc);
+		}
+
+		mmc_detect_change(slot->mmc,
+			msecs_to_jiffies(host->pdata->detect_delay_ms));
+	}
+}
+
+#ifdef CONFIG_OF
+/* given a slot id, find out the device node representing that slot */
+static struct device_node *dw_mci_of_find_slot_node(struct device *dev, u8 slot)
+{
+	struct device_node *np;
+	const __be32 *addr;
+	int len;
+
+	if (!dev || !dev->of_node)
+		return NULL;
+
+	for_each_child_of_node(dev->of_node, np) {
+		addr = of_get_property(np, "reg", &len);
+		if (!addr || (len < sizeof(int)))
+			continue;
+		if (be32_to_cpup(addr) == slot)
+			return np;
+	}
+	return NULL;
+}
+
+static struct dw_mci_of_slot_quirks {
+	char *quirk;
+	int id;
+} of_slot_quirks[] = {
+	{
+		.quirk	= "disable-wp",
+		.id	= DW_MCI_SLOT_QUIRK_NO_WRITE_PROTECT,
+	},
+};
+
+static int dw_mci_of_get_slot_quirks(struct device *dev, u8 slot)
+{
+	struct device_node *np = dw_mci_of_find_slot_node(dev, slot);
+	int quirks = 0;
+	int idx;
+
+	/* get quirks */
+	for (idx = 0; idx < ARRAY_SIZE(of_slot_quirks); idx++)
+		if (of_get_property(np, of_slot_quirks[idx].quirk, NULL))
+			quirks |= of_slot_quirks[idx].id;
+
+	return quirks;
+}
+#else /* CONFIG_OF */
+static int dw_mci_of_get_slot_quirks(struct device *dev, u8 slot)
+{
+	return 0;
+}
+static struct device_node *dw_mci_of_find_slot_node(struct device *dev, u8 slot)
+{
+	return NULL;
+}
+#endif /* CONFIG_OF */
+
+static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
+{
+	struct mmc_host *mmc;
+	struct dw_mci_slot *slot;
+	const struct dw_mci_drv_data *drv_data = host->drv_data;
+	int ctrl_id, ret;
+	u32 freq[2];
+
+	mmc = mmc_alloc_host(sizeof(struct dw_mci_slot), host->dev);
+	if (!mmc)
+		return -ENOMEM;
+
+	slot = mmc_priv(mmc);
+	slot->id = id;
+	slot->mmc = mmc;
+	slot->host = host;
+	host->slot[id] = slot;
+
+	slot->quirks = dw_mci_of_get_slot_quirks(host->dev, slot->id);
+
+	mmc->ops = &dw_mci_ops;
+	if (of_property_read_u32_array(host->dev->of_node,
+				       "clock-freq-min-max", freq, 2)) {
+		mmc->f_min = DW_MCI_FREQ_MIN;
+		mmc->f_max = DW_MCI_FREQ_MAX;
+	} else {
+		mmc->f_min = freq[0];
+		mmc->f_max = freq[1];
+	}
+
+	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+
+	if (host->pdata->caps)
+		mmc->caps = host->pdata->caps;
+
+	if (host->pdata->pm_caps)
+		mmc->pm_caps = host->pdata->pm_caps;
+
+	if (host->dev->of_node) {
+		ctrl_id = of_alias_get_id(host->dev->of_node, "mshc");
+		if (ctrl_id < 0)
+			ctrl_id = 0;
+	} else {
+		ctrl_id = to_platform_device(host->dev)->id;
+	}
+	if (drv_data && drv_data->caps)
+		mmc->caps |= drv_data->caps[ctrl_id];
+
+	if (host->pdata->caps2)
+		mmc->caps2 = host->pdata->caps2;
+
+	mmc_of_parse(mmc);
+
+	if (host->pdata->blk_settings) {
+		mmc->max_segs = host->pdata->blk_settings->max_segs;
+		mmc->max_blk_size = host->pdata->blk_settings->max_blk_size;
+		mmc->max_blk_count = host->pdata->blk_settings->max_blk_count;
+		mmc->max_req_size = host->pdata->blk_settings->max_req_size;
+		mmc->max_seg_size = host->pdata->blk_settings->max_seg_size;
+	} else {
+		/* Useful defaults if platform data is unset. */
+#ifdef CONFIG_MMC_DW_IDMAC
+		mmc->max_segs = host->ring_size;
+		mmc->max_blk_size = 65536;
+		mmc->max_blk_count = host->ring_size;
+		mmc->max_seg_size = 0x1000;
+		mmc->max_req_size = mmc->max_seg_size * mmc->max_blk_count;
+#else
+		mmc->max_segs = 64;
+		mmc->max_blk_size = 65536; /* BLKSIZ is 16 bits */
+		mmc->max_blk_count = 512;
+		mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count;
+		mmc->max_seg_size = mmc->max_req_size;
+#endif /* CONFIG_MMC_DW_IDMAC */
+	}
+
+	if (dw_mci_get_cd(mmc))
+		set_bit(DW_MMC_CARD_PRESENT, &slot->flags);
+	else
+		clear_bit(DW_MMC_CARD_PRESENT, &slot->flags);
+
+	ret = mmc_add_host(mmc);
+	if (ret)
+		goto err_setup_bus;
+
+#if defined(CONFIG_DEBUG_FS)
+	dw_mci_init_debugfs(slot);
+#endif
+
+	/* Card initially undetected */
+	slot->last_detect_state = 0;
+
+	return 0;
+
+err_setup_bus:
+	mmc_free_host(mmc);
+	return -EINVAL;
+}
+
+static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
+{
+	/* Debugfs stuff is cleaned up by mmc core */
+	mmc_remove_host(slot->mmc);
+	slot->host->slot[id] = NULL;
+	mmc_free_host(slot->mmc);
+}
+
+static void dw_mci_init_dma(struct dw_mci *host)
+{
+	/* Alloc memory for sg translation */
+	host->sg_cpu = dmam_alloc_coherent(host->dev, PAGE_SIZE,
+					  &host->sg_dma, GFP_KERNEL);
+	if (!host->sg_cpu) {
+		dev_err(host->dev, "%s: could not alloc DMA memory\n",
+			__func__);
+		goto no_dma;
+	}
+
+	/* Determine which DMA interface to use */
+#ifdef CONFIG_MMC_DW_IDMAC
+	host->dma_ops = &dw_mci_idmac_ops;
+	dev_info(host->dev, "Using internal DMA controller.\n");
+#endif
+
+	if (!host->dma_ops)
+		goto no_dma;
+
+	if (host->dma_ops->init && host->dma_ops->start &&
+	    host->dma_ops->stop && host->dma_ops->cleanup) {
+		if (host->dma_ops->init(host)) {
+			dev_err(host->dev, "%s: Unable to initialize "
+				"DMA Controller.\n", __func__);
+			goto no_dma;
+		}
+	} else {
+		dev_err(host->dev, "DMA initialization not found.\n");
+		goto no_dma;
+	}
+
+	host->use_dma = 1;
+	return;
+
+no_dma:
+	dev_info(host->dev, "Using PIO mode.\n");
+	host->use_dma = 0;
+	return;
+}
+
+static bool dw_mci_ctrl_reset(struct dw_mci *host, u32 reset)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(500);
+	u32 ctrl;
+
+	ctrl = mci_readl(host, CTRL);
+	ctrl |= reset;
+	mci_writel(host, CTRL, ctrl);
+
+	/* wait till resets clear */
+	do {
+		ctrl = mci_readl(host, CTRL);
+		if (!(ctrl & reset))
+			return true;
+	} while (time_before(jiffies, timeout));
+
+	dev_err(host->dev,
+		"Timeout resetting block (ctrl reset %#x)\n",
+		ctrl & reset);
+
+	return false;
+}
+
+static inline bool dw_mci_fifo_reset(struct dw_mci *host)
+{
+	/*
+	 * Reseting generates a block interrupt, hence setting
+	 * the scatter-gather pointer to NULL.
+	 */
+	if (host->sg) {
+		sg_miter_stop(&host->sg_miter);
+		host->sg = NULL;
+	}
+
+	return dw_mci_ctrl_reset(host, SDMMC_CTRL_FIFO_RESET);
+}
+
+static inline bool dw_mci_ctrl_all_reset(struct dw_mci *host)
+{
+	return dw_mci_ctrl_reset(host,
+				 SDMMC_CTRL_FIFO_RESET |
+				 SDMMC_CTRL_RESET |
+				 SDMMC_CTRL_DMA_RESET);
+}
+
+#ifdef CONFIG_OF
+static struct dw_mci_of_quirks {
+	char *quirk;
+	int id;
+} of_quirks[] = {
+	{
+		.quirk	= "broken-cd",
+		.id	= DW_MCI_QUIRK_BROKEN_CARD_DETECTION,
+	},
+};
+
+static struct dw_mci_board *dw_mci_parse_dt(struct dw_mci *host)
+{
+	struct dw_mci_board *pdata;
+	struct device *dev = host->dev;
+	struct device_node *np = dev->of_node;
+	const struct dw_mci_drv_data *drv_data = host->drv_data;
+	int idx, ret;
+	u32 clock_frequency;
+
+	pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata) {
+		dev_err(dev, "could not allocate memory for pdata\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/* find out number of slots supported */
+	if (of_property_read_u32(dev->of_node, "num-slots",
+				&pdata->num_slots)) {
+		dev_info(dev, "num-slots property not found, "
+				"assuming 1 slot is available\n");
+		pdata->num_slots = 1;
+	}
+
+	/* get quirks */
+	for (idx = 0; idx < ARRAY_SIZE(of_quirks); idx++)
+		if (of_get_property(np, of_quirks[idx].quirk, NULL))
+			pdata->quirks |= of_quirks[idx].id;
+
+	if (of_property_read_u32(np, "fifo-depth", &pdata->fifo_depth))
+		dev_info(dev, "fifo-depth property not found, using "
+				"value of FIFOTH register as default\n");
+
+	of_property_read_u32(np, "card-detect-delay", &pdata->detect_delay_ms);
+
+	if (!of_property_read_u32(np, "clock-frequency", &clock_frequency))
+		pdata->bus_hz = clock_frequency;
+
+	if (drv_data && drv_data->parse_dt) {
+		ret = drv_data->parse_dt(host);
+		if (ret)
+			return ERR_PTR(ret);
+	}
+
+	if (of_find_property(np, "supports-highspeed", NULL))
+		pdata->caps |= MMC_CAP_SD_HIGHSPEED | MMC_CAP_MMC_HIGHSPEED;
+
+	return pdata;
+}
+
+#else /* CONFIG_OF */
+static struct dw_mci_board *dw_mci_parse_dt(struct dw_mci *host)
+{
+	return ERR_PTR(-EINVAL);
+}
+#endif /* CONFIG_OF */
+
+int dw_mci_probe(struct dw_mci *host)
+{
+	const struct dw_mci_drv_data *drv_data = host->drv_data;
+	int width, i, ret = 0;
+	u32 fifo_size;
+	int init_slots = 0;
+
+	if (!host->pdata) {
+		host->pdata = dw_mci_parse_dt(host);
+		if (IS_ERR(host->pdata)) {
+			dev_err(host->dev, "platform data not available\n");
+			return -EINVAL;
+		}
+	}
+
+	if (host->pdata->num_slots > 1) {
+		dev_err(host->dev,
+			"Platform data must supply num_slots.\n");
+		return -ENODEV;
+	}
+
+	host->biu_clk = devm_clk_get(host->dev, "biu");
+	if (IS_ERR(host->biu_clk)) {
+		dev_dbg(host->dev, "biu clock not available\n");
+	} else {
+		ret = clk_prepare_enable(host->biu_clk);
+		if (ret) {
+			dev_err(host->dev, "failed to enable biu clock\n");
+			return ret;
+		}
+	}
+
+	host->ciu_clk = devm_clk_get(host->dev, "ciu");
+	if (IS_ERR(host->ciu_clk)) {
+		dev_dbg(host->dev, "ciu clock not available\n");
+		host->bus_hz = host->pdata->bus_hz;
+	} else {
+		ret = clk_prepare_enable(host->ciu_clk);
+		if (ret) {
+			dev_err(host->dev, "failed to enable ciu clock\n");
+			goto err_clk_biu;
+		}
+
+		if (host->pdata->bus_hz) {
+			ret = clk_set_rate(host->ciu_clk, host->pdata->bus_hz);
+			if (ret)
+				dev_warn(host->dev,
+					 "Unable to set bus rate to %uHz\n",
+					 host->pdata->bus_hz);
+		}
+		host->bus_hz = clk_get_rate(host->ciu_clk);
+	}
+
+	if (!host->bus_hz) {
+		dev_err(host->dev,
+			"Platform data must supply bus speed\n");
+		ret = -ENODEV;
+		goto err_clk_ciu;
+	}
+
+	if (drv_data && drv_data->init) {
+		ret = drv_data->init(host);
+		if (ret) {
+			dev_err(host->dev,
+				"implementation specific init failed\n");
+			goto err_clk_ciu;
+		}
+	}
+
+	if (drv_data && drv_data->setup_clock) {
+		ret = drv_data->setup_clock(host);
+		if (ret) {
+			dev_err(host->dev,
+				"implementation specific clock setup failed\n");
+			goto err_clk_ciu;
+		}
+	}
+
+	host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
+	if (IS_ERR(host->vmmc)) {
+		ret = PTR_ERR(host->vmmc);
+		if (ret == -EPROBE_DEFER)
+			goto err_clk_ciu;
+
+		dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
+		host->vmmc = NULL;
+	} else {
+		ret = regulator_enable(host->vmmc);
+		if (ret) {
+			if (ret != -EPROBE_DEFER)
+				dev_err(host->dev,
+					"regulator_enable fail: %d\n", ret);
+			goto err_clk_ciu;
+		}
+	}
+
+	host->quirks = host->pdata->quirks;
+
+	spin_lock_init(&host->lock);
+	INIT_LIST_HEAD(&host->queue);
+
+	/*
+	 * Get the host data width - this assumes that HCON has been set with
+	 * the correct values.
+	 */
+	i = (mci_readl(host, HCON) >> 7) & 0x7;
+	if (!i) {
+		host->push_data = dw_mci_push_data16;
+		host->pull_data = dw_mci_pull_data16;
+		width = 16;
+		host->data_shift = 1;
+	} else if (i == 2) {
+		host->push_data = dw_mci_push_data64;
+		host->pull_data = dw_mci_pull_data64;
+		width = 64;
+		host->data_shift = 3;
+	} else {
+		/* Check for a reserved value, and warn if it is */
+		WARN((i != 1),
+		     "HCON reports a reserved host data width!\n"
+		     "Defaulting to 32-bit access.\n");
+		host->push_data = dw_mci_push_data32;
+		host->pull_data = dw_mci_pull_data32;
+		width = 32;
+		host->data_shift = 2;
+	}
+
+	/* Reset all blocks */
+	if (!dw_mci_ctrl_all_reset(host))
+		return -ENODEV;
+
+	host->dma_ops = host->pdata->dma_ops;
+	dw_mci_init_dma(host);
+
+	/* Clear the interrupts for the host controller */
+	mci_writel(host, RINTSTS, 0xFFFFFFFF);
+	mci_writel(host, INTMASK, 0); /* disable all mmc interrupt first */
+
+	/* Put in max timeout */
+	mci_writel(host, TMOUT, 0xFFFFFFFF);
+
+	/*
+	 * FIFO threshold settings  RxMark  = fifo_size / 2 - 1,
+	 *                          Tx Mark = fifo_size / 2 DMA Size = 8
+	 */
+	if (!host->pdata->fifo_depth) {
+		/*
+		 * Power-on value of RX_WMark is FIFO_DEPTH-1, but this may
+		 * have been overwritten by the bootloader, just like we're
+		 * about to do, so if you know the value for your hardware, you
+		 * should put it in the platform data.
+		 */
+		fifo_size = mci_readl(host, FIFOTH);
+		fifo_size = 1 + ((fifo_size >> 16) & 0xfff);
+	} else {
+		fifo_size = host->pdata->fifo_depth;
+	}
+	host->fifo_depth = fifo_size;
+	host->fifoth_val =
+		SDMMC_SET_FIFOTH(0x2, fifo_size / 2 - 1, fifo_size / 2);
+	mci_writel(host, FIFOTH, host->fifoth_val);
+
+	/* disable clock to CIU */
+	mci_writel(host, CLKENA, 0);
+	mci_writel(host, CLKSRC, 0);
+
+	/*
+	 * In 2.40a spec, Data offset is changed.
+	 * Need to check the version-id and set data-offset for DATA register.
+	 */
+	host->verid = SDMMC_GET_VERID(mci_readl(host, VERID));
+	dev_info(host->dev, "Version ID is %04x\n", host->verid);
+
+	if (host->verid < DW_MMC_240A)
+		host->data_offset = DATA_OFFSET;
+	else
+		host->data_offset = DATA_240A_OFFSET;
+
+	tasklet_init(&host->tasklet, dw_mci_tasklet_func, (unsigned long)host);
+	host->card_workqueue = alloc_workqueue("dw-mci-card",
+			WQ_MEM_RECLAIM, 1);
+	if (!host->card_workqueue) {
+		ret = -ENOMEM;
+		goto err_dmaunmap;
+	}
+	INIT_WORK(&host->card_work, dw_mci_work_routine_card);
+	ret = devm_request_irq(host->dev, host->irq, dw_mci_interrupt,
+			       host->irq_flags, "dw-mci", host);
+	if (ret)
+		goto err_workqueue;
+
+	if (host->pdata->num_slots)
+		host->num_slots = host->pdata->num_slots;
+	else
+		host->num_slots = ((mci_readl(host, HCON) >> 1) & 0x1F) + 1;
+
+	/*
+	 * Enable interrupts for command done, data over, data empty, card det,
+	 * receive ready and error such as transmit, receive timeout, crc error
+	 */
+	mci_writel(host, RINTSTS, 0xFFFFFFFF);
+	mci_writel(host, INTMASK, SDMMC_INT_CMD_DONE | SDMMC_INT_DATA_OVER |
+		   SDMMC_INT_TXDR | SDMMC_INT_RXDR |
+		   DW_MCI_ERROR_FLAGS | SDMMC_INT_CD);
+	mci_writel(host, CTRL, SDMMC_CTRL_INT_ENABLE); /* Enable mci interrupt */
+
+	dev_info(host->dev, "DW MMC controller at irq %d, "
+		 "%d bit host data width, "
+		 "%u deep fifo\n",
+		 host->irq, width, fifo_size);
+
+	/* We need at least one slot to succeed */
+	for (i = 0; i < host->num_slots; i++) {
+		ret = dw_mci_init_slot(host, i);
+		if (ret)
+			dev_dbg(host->dev, "slot %d init failed\n", i);
+		else
+			init_slots++;
+	}
+
+	if (init_slots) {
+		dev_info(host->dev, "%d slots initialized\n", init_slots);
+	} else {
+		dev_dbg(host->dev, "attempted to initialize %d slots, "
+					"but failed on all\n", host->num_slots);
+		goto err_workqueue;
+	}
+
+	if (host->quirks & DW_MCI_QUIRK_IDMAC_DTO)
+		dev_info(host->dev, "Internal DMAC interrupt fix enabled.\n");
+
+	return 0;
+
+err_workqueue:
+	destroy_workqueue(host->card_workqueue);
+
+err_dmaunmap:
+	if (host->use_dma && host->dma_ops->exit)
+		host->dma_ops->exit(host);
+	if (host->vmmc)
+		regulator_disable(host->vmmc);
+
+err_clk_ciu:
+	if (!IS_ERR(host->ciu_clk))
+		clk_disable_unprepare(host->ciu_clk);
+
+err_clk_biu:
+	if (!IS_ERR(host->biu_clk))
+		clk_disable_unprepare(host->biu_clk);
+
+	return ret;
+}
+EXPORT_SYMBOL(dw_mci_probe);
+
+void dw_mci_remove(struct dw_mci *host)
+{
+	int i;
+
+	mci_writel(host, RINTSTS, 0xFFFFFFFF);
+	mci_writel(host, INTMASK, 0); /* disable all mmc interrupt first */
+
+	for (i = 0; i < host->num_slots; i++) {
+		dev_dbg(host->dev, "remove slot %d\n", i);
+		if (host->slot[i])
+			dw_mci_cleanup_slot(host->slot[i], i);
+	}
+
+	/* disable clock to CIU */
+	mci_writel(host, CLKENA, 0);
+	mci_writel(host, CLKSRC, 0);
+
+	destroy_workqueue(host->card_workqueue);
+
+	if (host->use_dma && host->dma_ops->exit)
+		host->dma_ops->exit(host);
+
+	if (host->vmmc)
+		regulator_disable(host->vmmc);
+
+	if (!IS_ERR(host->ciu_clk))
+		clk_disable_unprepare(host->ciu_clk);
+
+	if (!IS_ERR(host->biu_clk))
+		clk_disable_unprepare(host->biu_clk);
+}
+EXPORT_SYMBOL(dw_mci_remove);
+
+
+
+#ifdef CONFIG_PM_SLEEP
+/*
+ * TODO: we should probably disable the clock to the card in the suspend path.
+ */
+int dw_mci_suspend(struct dw_mci *host)
+{
+	if (host->vmmc)
+		regulator_disable(host->vmmc);
+
+	return 0;
+}
+EXPORT_SYMBOL(dw_mci_suspend);
+
+int dw_mci_resume(struct dw_mci *host)
+{
+	int i, ret;
+
+	if (host->vmmc) {
+		ret = regulator_enable(host->vmmc);
+		if (ret) {
+			dev_err(host->dev,
+				"failed to enable regulator: %d\n", ret);
+			return ret;
+		}
+	}
+
+	if (!dw_mci_ctrl_all_reset(host)) {
+		ret = -ENODEV;
+		return ret;
+	}
+
+	if (host->use_dma && host->dma_ops->init)
+		host->dma_ops->init(host);
+
+	/*
+	 * Restore the initial value at FIFOTH register
+	 * And Invalidate the prev_blksz with zero
+	 */
+	mci_writel(host, FIFOTH, host->fifoth_val);
+	host->prev_blksz = 0;
+
+	/* Put in max timeout */
+	mci_writel(host, TMOUT, 0xFFFFFFFF);
+
+	mci_writel(host, RINTSTS, 0xFFFFFFFF);
+	mci_writel(host, INTMASK, SDMMC_INT_CMD_DONE | SDMMC_INT_DATA_OVER |
+		   SDMMC_INT_TXDR | SDMMC_INT_RXDR |
+		   DW_MCI_ERROR_FLAGS | SDMMC_INT_CD);
+	mci_writel(host, CTRL, SDMMC_CTRL_INT_ENABLE);
+
+	for (i = 0; i < host->num_slots; i++) {
+		struct dw_mci_slot *slot = host->slot[i];
+		if (!slot)
+			continue;
+		if (slot->mmc->pm_flags & MMC_PM_KEEP_POWER) {
+			dw_mci_set_ios(slot->mmc, &slot->mmc->ios);
+			dw_mci_setup_bus(slot, true);
+		}
+	}
+	return 0;
+}
+EXPORT_SYMBOL(dw_mci_resume);
+#endif /* CONFIG_PM_SLEEP */
+
+static int __init dw_mci_init(void)
+{
+	pr_info("Synopsys Designware Multimedia Card Interface Driver\n");
+	return 0;
+}
+
+static void __exit dw_mci_exit(void)
+{
+}
+
+module_init(dw_mci_init);
+module_exit(dw_mci_exit);
+
+MODULE_DESCRIPTION("DW Multimedia Card Interface driver");
+MODULE_AUTHOR("NXP Semiconductor VietNam");
+MODULE_AUTHOR("Imagination Technologies Ltd");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
new file mode 100644
index 00000000000..738fa241d05
--- /dev/null
+++ b/drivers/mmc/host/dw_mmc.h
@@ -0,0 +1,261 @@
+/*
+ * Synopsys DesignWare Multimedia Card Interface driver
+ *  (Based on NXP driver for lpc 31xx)
+ *
+ * Copyright (C) 2009 NXP Semiconductors
+ * Copyright (C) 2009, 2010 Imagination Technologies Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef _DW_MMC_H_
+#define _DW_MMC_H_
+
+#define DW_MMC_240A		0x240a
+
+#define SDMMC_CTRL		0x000
+#define SDMMC_PWREN		0x004
+#define SDMMC_CLKDIV		0x008
+#define SDMMC_CLKSRC		0x00c
+#define SDMMC_CLKENA		0x010
+#define SDMMC_TMOUT		0x014
+#define SDMMC_CTYPE		0x018
+#define SDMMC_BLKSIZ		0x01c
+#define SDMMC_BYTCNT		0x020
+#define SDMMC_INTMASK		0x024
+#define SDMMC_CMDARG		0x028
+#define SDMMC_CMD		0x02c
+#define SDMMC_RESP0		0x030
+#define SDMMC_RESP1		0x034
+#define SDMMC_RESP2		0x038
+#define SDMMC_RESP3		0x03c
+#define SDMMC_MINTSTS		0x040
+#define SDMMC_RINTSTS		0x044
+#define SDMMC_STATUS		0x048
+#define SDMMC_FIFOTH		0x04c
+#define SDMMC_CDETECT		0x050
+#define SDMMC_WRTPRT		0x054
+#define SDMMC_GPIO		0x058
+#define SDMMC_TCBCNT		0x05c
+#define SDMMC_TBBCNT		0x060
+#define SDMMC_DEBNCE		0x064
+#define SDMMC_USRID		0x068
+#define SDMMC_VERID		0x06c
+#define SDMMC_HCON		0x070
+#define SDMMC_UHS_REG		0x074
+#define SDMMC_BMOD		0x080
+#define SDMMC_PLDMND		0x084
+#define SDMMC_DBADDR		0x088
+#define SDMMC_IDSTS		0x08c
+#define SDMMC_IDINTEN		0x090
+#define SDMMC_DSCADDR		0x094
+#define SDMMC_BUFADDR		0x098
+#define SDMMC_CDTHRCTL		0x100
+#define SDMMC_DATA(x)		(x)
+
+/*
+ * Data offset is difference according to Version
+ * Lower than 2.40a : data register offest is 0x100
+ */
+#define DATA_OFFSET		0x100
+#define DATA_240A_OFFSET	0x200
+
+/* shift bit field */
+#define _SBF(f, v)		((v) << (f))
+
+/* Control register defines */
+#define SDMMC_CTRL_USE_IDMAC		BIT(25)
+#define SDMMC_CTRL_CEATA_INT_EN		BIT(11)
+#define SDMMC_CTRL_SEND_AS_CCSD		BIT(10)
+#define SDMMC_CTRL_SEND_CCSD		BIT(9)
+#define SDMMC_CTRL_ABRT_READ_DATA	BIT(8)
+#define SDMMC_CTRL_SEND_IRQ_RESP	BIT(7)
+#define SDMMC_CTRL_READ_WAIT		BIT(6)
+#define SDMMC_CTRL_DMA_ENABLE		BIT(5)
+#define SDMMC_CTRL_INT_ENABLE		BIT(4)
+#define SDMMC_CTRL_DMA_RESET		BIT(2)
+#define SDMMC_CTRL_FIFO_RESET		BIT(1)
+#define SDMMC_CTRL_RESET		BIT(0)
+/* Clock Enable register defines */
+#define SDMMC_CLKEN_LOW_PWR		BIT(16)
+#define SDMMC_CLKEN_ENABLE		BIT(0)
+/* time-out register defines */
+#define SDMMC_TMOUT_DATA(n)		_SBF(8, (n))
+#define SDMMC_TMOUT_DATA_MSK		0xFFFFFF00
+#define SDMMC_TMOUT_RESP(n)		((n) & 0xFF)
+#define SDMMC_TMOUT_RESP_MSK		0xFF
+/* card-type register defines */
+#define SDMMC_CTYPE_8BIT		BIT(16)
+#define SDMMC_CTYPE_4BIT		BIT(0)
+#define SDMMC_CTYPE_1BIT		0
+/* Interrupt status & mask register defines */
+#define SDMMC_INT_SDIO(n)		BIT(16 + (n))
+#define SDMMC_INT_EBE			BIT(15)
+#define SDMMC_INT_ACD			BIT(14)
+#define SDMMC_INT_SBE			BIT(13)
+#define SDMMC_INT_HLE			BIT(12)
+#define SDMMC_INT_FRUN			BIT(11)
+#define SDMMC_INT_HTO			BIT(10)
+#define SDMMC_INT_DRTO			BIT(9)
+#define SDMMC_INT_RTO			BIT(8)
+#define SDMMC_INT_DCRC			BIT(7)
+#define SDMMC_INT_RCRC			BIT(6)
+#define SDMMC_INT_RXDR			BIT(5)
+#define SDMMC_INT_TXDR			BIT(4)
+#define SDMMC_INT_DATA_OVER		BIT(3)
+#define SDMMC_INT_CMD_DONE		BIT(2)
+#define SDMMC_INT_RESP_ERR		BIT(1)
+#define SDMMC_INT_CD			BIT(0)
+#define SDMMC_INT_ERROR			0xbfc2
+/* Command register defines */
+#define SDMMC_CMD_START			BIT(31)
+#define SDMMC_CMD_USE_HOLD_REG	BIT(29)
+#define SDMMC_CMD_CCS_EXP		BIT(23)
+#define SDMMC_CMD_CEATA_RD		BIT(22)
+#define SDMMC_CMD_UPD_CLK		BIT(21)
+#define SDMMC_CMD_INIT			BIT(15)
+#define SDMMC_CMD_STOP			BIT(14)
+#define SDMMC_CMD_PRV_DAT_WAIT		BIT(13)
+#define SDMMC_CMD_SEND_STOP		BIT(12)
+#define SDMMC_CMD_STRM_MODE		BIT(11)
+#define SDMMC_CMD_DAT_WR		BIT(10)
+#define SDMMC_CMD_DAT_EXP		BIT(9)
+#define SDMMC_CMD_RESP_CRC		BIT(8)
+#define SDMMC_CMD_RESP_LONG		BIT(7)
+#define SDMMC_CMD_RESP_EXP		BIT(6)
+#define SDMMC_CMD_INDX(n)		((n) & 0x1F)
+/* Status register defines */
+#define SDMMC_GET_FCNT(x)		(((x)>>17) & 0x1FFF)
+/* FIFOTH register defines */
+#define SDMMC_SET_FIFOTH(m, r, t)	(((m) & 0x7) << 28 | \
+					 ((r) & 0xFFF) << 16 | \
+					 ((t) & 0xFFF))
+/* Internal DMAC interrupt defines */
+#define SDMMC_IDMAC_INT_AI		BIT(9)
+#define SDMMC_IDMAC_INT_NI		BIT(8)
+#define SDMMC_IDMAC_INT_CES		BIT(5)
+#define SDMMC_IDMAC_INT_DU		BIT(4)
+#define SDMMC_IDMAC_INT_FBE		BIT(2)
+#define SDMMC_IDMAC_INT_RI		BIT(1)
+#define SDMMC_IDMAC_INT_TI		BIT(0)
+/* Internal DMAC bus mode bits */
+#define SDMMC_IDMAC_ENABLE		BIT(7)
+#define SDMMC_IDMAC_FB			BIT(1)
+#define SDMMC_IDMAC_SWRESET		BIT(0)
+/* Version ID register define */
+#define SDMMC_GET_VERID(x)		((x) & 0xFFFF)
+/* Card read threshold */
+#define SDMMC_SET_RD_THLD(v, x)		(((v) & 0x1FFF) << 16 | (x))
+
+/* Register access macros */
+#define mci_readl(dev, reg)			\
+	__raw_readl((dev)->regs + SDMMC_##reg)
+#define mci_writel(dev, reg, value)			\
+	__raw_writel((value), (dev)->regs + SDMMC_##reg)
+
+/* 16-bit FIFO access macros */
+#define mci_readw(dev, reg)			\
+	__raw_readw((dev)->regs + SDMMC_##reg)
+#define mci_writew(dev, reg, value)			\
+	__raw_writew((value), (dev)->regs + SDMMC_##reg)
+
+/* 64-bit FIFO access macros */
+#ifdef readq
+#define mci_readq(dev, reg)			\
+	__raw_readq((dev)->regs + SDMMC_##reg)
+#define mci_writeq(dev, reg, value)			\
+	__raw_writeq((value), (dev)->regs + SDMMC_##reg)
+#else
+/*
+ * Dummy readq implementation for architectures that don't define it.
+ *
+ * We would assume that none of these architectures would configure
+ * the IP block with a 64bit FIFO width, so this code will never be
+ * executed on those machines. Defining these macros here keeps the
+ * rest of the code free from ifdefs.
+ */
+#define mci_readq(dev, reg)			\
+	(*(volatile u64 __force *)((dev)->regs + SDMMC_##reg))
+#define mci_writeq(dev, reg, value)			\
+	(*(volatile u64 __force *)((dev)->regs + SDMMC_##reg) = (value))
+#endif
+
+extern int dw_mci_probe(struct dw_mci *host);
+extern void dw_mci_remove(struct dw_mci *host);
+#ifdef CONFIG_PM_SLEEP
+extern int dw_mci_suspend(struct dw_mci *host);
+extern int dw_mci_resume(struct dw_mci *host);
+#endif
+
+/**
+ * struct dw_mci_slot - MMC slot state
+ * @mmc: The mmc_host representing this slot.
+ * @host: The MMC controller this slot is using.
+ * @quirks: Slot-level quirks (DW_MCI_SLOT_QUIRK_XXX)
+ * @ctype: Card type for this slot.
+ * @mrq: mmc_request currently being processed or waiting to be
+ *	processed, or NULL when the slot is idle.
+ * @queue_node: List node for placing this node in the @queue list of
+ *	&struct dw_mci.
+ * @clock: Clock rate configured by set_ios(). Protected by host->lock.
+ * @__clk_old: The last updated clock with reflecting clock divider.
+ *	Keeping track of this helps us to avoid spamming the console
+ *	with CONFIG_MMC_CLKGATE.
+ * @flags: Random state bits associated with the slot.
+ * @id: Number of this slot.
+ * @last_detect_state: Most recently observed card detect state.
+ */
+struct dw_mci_slot {
+	struct mmc_host		*mmc;
+	struct dw_mci		*host;
+
+	int			quirks;
+
+	u32			ctype;
+
+	struct mmc_request	*mrq;
+	struct list_head	queue_node;
+
+	unsigned int		clock;
+	unsigned int		__clk_old;
+
+	unsigned long		flags;
+#define DW_MMC_CARD_PRESENT	0
+#define DW_MMC_CARD_NEED_INIT	1
+	int			id;
+	int			last_detect_state;
+};
+
+struct dw_mci_tuning_data {
+	const u8 *blk_pattern;
+	unsigned int blksz;
+};
+
+/**
+ * dw_mci driver data - dw-mshc implementation specific driver data.
+ * @caps: mmc subsystem specified capabilities of the controller(s).
+ * @init: early implementation specific initialization.
+ * @setup_clock: implementation specific clock configuration.
+ * @prepare_command: handle CMD register extensions.
+ * @set_ios: handle bus specific extensions.
+ * @parse_dt: parse implementation specific device tree properties.
+ * @execute_tuning: implementation specific tuning procedure.
+ *
+ * Provide controller implementation specific extensions. The usage of this
+ * data structure is fully optional and usage of each member in this structure
+ * is optional as well.
+ */
+struct dw_mci_drv_data {
+	unsigned long	*caps;
+	int		(*init)(struct dw_mci *host);
+	int		(*setup_clock)(struct dw_mci *host);
+	void		(*prepare_command)(struct dw_mci *host, u32 *cmdr);
+	void		(*set_ios)(struct dw_mci *host, struct mmc_ios *ios);
+	int		(*parse_dt)(struct dw_mci *host);
+	int		(*execute_tuning)(struct dw_mci_slot *slot, u32 opcode,
+					struct dw_mci_tuning_data *tuning_data);
+};
+#endif /* _DW_MMC_H_ */
diff --git a/drivers/mmc/host/imxmmc.c b/drivers/mmc/host/imxmmc.c
deleted file mode 100644
index e0be21a4a69..00000000000
--- a/drivers/mmc/host/imxmmc.c
+++ /dev/null
@@ -1,1169 +0,0 @@
-/*
- *  linux/drivers/mmc/host/imxmmc.c - Motorola i.MX MMCI driver
- *
- *  Copyright (C) 2004 Sascha Hauer, Pengutronix <sascha@saschahauer.de>
- *  Copyright (C) 2006 Pavel Pisa, PiKRON <ppisa@pikron.com>
- *
- *  derived from pxamci.c by Russell King
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- */
-
-#include <linux/module.h>
-#include <linux/init.h>
-#include <linux/ioport.h>
-#include <linux/platform_device.h>
-#include <linux/interrupt.h>
-#include <linux/blkdev.h>
-#include <linux/dma-mapping.h>
-#include <linux/mmc/host.h>
-#include <linux/mmc/card.h>
-#include <linux/delay.h>
-#include <linux/clk.h>
-#include <linux/io.h>
-
-#include <asm/dma.h>
-#include <asm/irq.h>
-#include <asm/sizes.h>
-#include <mach/mmc.h>
-#include <mach/imx-dma.h>
-
-#include "imxmmc.h"
-
-#define DRIVER_NAME "imx-mmc"
-
-#define IMXMCI_INT_MASK_DEFAULT (INT_MASK_BUF_READY | INT_MASK_DATA_TRAN | \
-				 INT_MASK_WRITE_OP_DONE | INT_MASK_END_CMD_RES | \
-				 INT_MASK_AUTO_CARD_DETECT | INT_MASK_DAT0_EN | INT_MASK_SDIO)
-
-struct imxmci_host {
-	struct mmc_host		*mmc;
-	spinlock_t		lock;
-	struct resource		*res;
-	void __iomem		*base;
-	int			irq;
-	imx_dmach_t		dma;
-	volatile unsigned int	imask;
-	unsigned int		power_mode;
-	unsigned int		present;
-	struct imxmmc_platform_data *pdata;
-
-	struct mmc_request	*req;
-	struct mmc_command	*cmd;
-	struct mmc_data		*data;
-
-	struct timer_list	timer;
-	struct tasklet_struct	tasklet;
-	unsigned int		status_reg;
-	unsigned long		pending_events;
-	/* Next two fields are there for CPU driven transfers to overcome SDHC deficiencies */
-	u16			*data_ptr;
-	unsigned int		data_cnt;
-	atomic_t		stuck_timeout;
-
-	unsigned int		dma_nents;
-	unsigned int		dma_size;
-	unsigned int		dma_dir;
-	int			dma_allocated;
-
-	unsigned char		actual_bus_width;
-
-	int			prev_cmd_code;
-
-	struct clk		*clk;
-};
-
-#define IMXMCI_PEND_IRQ_b	0
-#define IMXMCI_PEND_DMA_END_b	1
-#define IMXMCI_PEND_DMA_ERR_b	2
-#define IMXMCI_PEND_WAIT_RESP_b	3
-#define IMXMCI_PEND_DMA_DATA_b	4
-#define IMXMCI_PEND_CPU_DATA_b	5
-#define IMXMCI_PEND_CARD_XCHG_b	6
-#define IMXMCI_PEND_SET_INIT_b	7
-#define IMXMCI_PEND_STARTED_b	8
-
-#define IMXMCI_PEND_IRQ_m	(1 << IMXMCI_PEND_IRQ_b)
-#define IMXMCI_PEND_DMA_END_m	(1 << IMXMCI_PEND_DMA_END_b)
-#define IMXMCI_PEND_DMA_ERR_m	(1 << IMXMCI_PEND_DMA_ERR_b)
-#define IMXMCI_PEND_WAIT_RESP_m	(1 << IMXMCI_PEND_WAIT_RESP_b)
-#define IMXMCI_PEND_DMA_DATA_m	(1 << IMXMCI_PEND_DMA_DATA_b)
-#define IMXMCI_PEND_CPU_DATA_m	(1 << IMXMCI_PEND_CPU_DATA_b)
-#define IMXMCI_PEND_CARD_XCHG_m	(1 << IMXMCI_PEND_CARD_XCHG_b)
-#define IMXMCI_PEND_SET_INIT_m	(1 << IMXMCI_PEND_SET_INIT_b)
-#define IMXMCI_PEND_STARTED_m	(1 << IMXMCI_PEND_STARTED_b)
-
-static void imxmci_stop_clock(struct imxmci_host *host)
-{
-	int i = 0;
-	u16 reg;
-
-	reg = readw(host->base + MMC_REG_STR_STP_CLK);
-	writew(reg & ~STR_STP_CLK_START_CLK, host->base + MMC_REG_STR_STP_CLK);
-	while (i < 0x1000) {
-		if (!(i & 0x7f)) {
-			reg = readw(host->base + MMC_REG_STR_STP_CLK);
-			writew(reg | STR_STP_CLK_STOP_CLK,
-					host->base + MMC_REG_STR_STP_CLK);
-		}
-
-		reg = readw(host->base + MMC_REG_STATUS);
-		if (!(reg & STATUS_CARD_BUS_CLK_RUN)) {
-			/* Check twice before cut */
-			reg = readw(host->base + MMC_REG_STATUS);
-			if (!(reg & STATUS_CARD_BUS_CLK_RUN))
-				return;
-		}
-
-		i++;
-	}
-	dev_dbg(mmc_dev(host->mmc), "imxmci_stop_clock blocked, no luck\n");
-}
-
-static int imxmci_start_clock(struct imxmci_host *host)
-{
-	unsigned int trials = 0;
-	unsigned int delay_limit = 128;
-	unsigned long flags;
-	u16 reg;
-
-	reg = readw(host->base + MMC_REG_STR_STP_CLK);
-	writew(reg & ~STR_STP_CLK_STOP_CLK, host->base + MMC_REG_STR_STP_CLK);
-
-	clear_bit(IMXMCI_PEND_STARTED_b, &host->pending_events);
-
-	/*
-	 * Command start of the clock, this usually succeeds in less
-	 * then 6 delay loops, but during card detection (low clockrate)
-	 * it takes up to 5000 delay loops and sometimes fails for the first time
-	 */
-	reg = readw(host->base + MMC_REG_STR_STP_CLK);
-	writew(reg | STR_STP_CLK_START_CLK, host->base + MMC_REG_STR_STP_CLK);
-
-	do {
-		unsigned int delay = delay_limit;
-
-		while (delay--) {
-			reg = readw(host->base + MMC_REG_STATUS);
-			if (reg & STATUS_CARD_BUS_CLK_RUN)
-				/* Check twice before cut */
-				reg = readw(host->base + MMC_REG_STATUS);
-				if (reg & STATUS_CARD_BUS_CLK_RUN)
-					return 0;
-
-			if (test_bit(IMXMCI_PEND_STARTED_b, &host->pending_events))
-				return 0;
-		}
-
-		local_irq_save(flags);
-		/*
-		 * Ensure, that request is not doubled under all possible circumstances.
-		 * It is possible, that cock running state is missed, because some other
-		 * IRQ or schedule delays this function execution and the clocks has
-		 * been already stopped by other means (response processing, SDHC HW)
-		 */
-		if (!test_bit(IMXMCI_PEND_STARTED_b, &host->pending_events)) {
-			reg = readw(host->base + MMC_REG_STR_STP_CLK);
-			writew(reg | STR_STP_CLK_START_CLK,
-					host->base + MMC_REG_STR_STP_CLK);
-		}
-		local_irq_restore(flags);
-
-	} while (++trials < 256);
-
-	dev_err(mmc_dev(host->mmc), "imxmci_start_clock blocked, no luck\n");
-
-	return -1;
-}
-
-static void imxmci_softreset(struct imxmci_host *host)
-{
-	int i;
-
-	/* reset sequence */
-	writew(0x08, host->base + MMC_REG_STR_STP_CLK);
-	writew(0x0D, host->base + MMC_REG_STR_STP_CLK);
-
-	for (i = 0; i < 8; i++)
-		writew(0x05, host->base + MMC_REG_STR_STP_CLK);
-
-	writew(0xff, host->base + MMC_REG_RES_TO);
-	writew(512, host->base + MMC_REG_BLK_LEN);
-	writew(1, host->base + MMC_REG_NOB);
-}
-
-static int imxmci_busy_wait_for_status(struct imxmci_host *host,
-				       unsigned int *pstat, unsigned int stat_mask,
-				       int timeout, const char *where)
-{
-	int loops = 0;
-
-	while (!(*pstat & stat_mask)) {
-		loops += 2;
-		if (loops >= timeout) {
-			dev_dbg(mmc_dev(host->mmc), "busy wait timeout in %s, STATUS = 0x%x (0x%x)\n",
-				where, *pstat, stat_mask);
-			return -1;
-		}
-		udelay(2);
-		*pstat |= readw(host->base + MMC_REG_STATUS);
-	}
-	if (!loops)
-		return 0;
-
-	/* The busy-wait is expected there for clock <8MHz due to SDHC hardware flaws */
-	if (!(stat_mask & STATUS_END_CMD_RESP) || (host->mmc->ios.clock >= 8000000))
-		dev_info(mmc_dev(host->mmc), "busy wait for %d usec in %s, STATUS = 0x%x (0x%x)\n",
-			 loops, where, *pstat, stat_mask);
-	return loops;
-}
-
-static void imxmci_setup_data(struct imxmci_host *host, struct mmc_data *data)
-{
-	unsigned int nob = data->blocks;
-	unsigned int blksz = data->blksz;
-	unsigned int datasz = nob * blksz;
-	int i;
-
-	if (data->flags & MMC_DATA_STREAM)
-		nob = 0xffff;
-
-	host->data = data;
-	data->bytes_xfered = 0;
-
-	writew(nob, host->base + MMC_REG_NOB);
-	writew(blksz, host->base + MMC_REG_BLK_LEN);
-
-	/*
-	 * DMA cannot be used for small block sizes, we have to use CPU driven transfers otherwise.
-	 * We are in big troubles for non-512 byte transfers according to note in the paragraph
-	 * 20.6.7 of User Manual anyway, but we need to be able to transfer SCR at least.
-	 * The situation is even more complex in reality. The SDHC in not able to handle wll
-	 * partial FIFO fills and reads. The length has to be rounded up to burst size multiple.
-	 * This is required for SCR read at least.
-	 */
-	if (datasz < 512) {
-		host->dma_size = datasz;
-		if (data->flags & MMC_DATA_READ) {
-			host->dma_dir = DMA_FROM_DEVICE;
-
-			/* Hack to enable read SCR */
-			writew(1, host->base + MMC_REG_NOB);
-			writew(512, host->base + MMC_REG_BLK_LEN);
-		} else {
-			host->dma_dir = DMA_TO_DEVICE;
-		}
-
-		/* Convert back to virtual address */
-		host->data_ptr = (u16 *)sg_virt(data->sg);
-		host->data_cnt = 0;
-
-		clear_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events);
-		set_bit(IMXMCI_PEND_CPU_DATA_b, &host->pending_events);
-
-		return;
-	}
-
-	if (data->flags & MMC_DATA_READ) {
-		host->dma_dir = DMA_FROM_DEVICE;
-		host->dma_nents = dma_map_sg(mmc_dev(host->mmc), data->sg,
-					     data->sg_len,  host->dma_dir);
-
-		imx_dma_setup_sg(host->dma, data->sg, data->sg_len, datasz,
-				 host->res->start + MMC_REG_BUFFER_ACCESS,
-				 DMA_MODE_READ);
-
-		/*imx_dma_setup_mem2dev_ccr(host->dma, DMA_MODE_READ, IMX_DMA_WIDTH_16, CCR_REN);*/
-		CCR(host->dma) = CCR_DMOD_LINEAR | CCR_DSIZ_32 | CCR_SMOD_FIFO | CCR_SSIZ_16 | CCR_REN;
-	} else {
-		host->dma_dir = DMA_TO_DEVICE;
-
-		host->dma_nents = dma_map_sg(mmc_dev(host->mmc), data->sg,
-					     data->sg_len,  host->dma_dir);
-
-		imx_dma_setup_sg(host->dma, data->sg, data->sg_len, datasz,
-				 host->res->start + MMC_REG_BUFFER_ACCESS,
-				 DMA_MODE_WRITE);
-
-		/*imx_dma_setup_mem2dev_ccr(host->dma, DMA_MODE_WRITE, IMX_DMA_WIDTH_16, CCR_REN);*/
-		CCR(host->dma) = CCR_SMOD_LINEAR | CCR_SSIZ_32 | CCR_DMOD_FIFO | CCR_DSIZ_16 | CCR_REN;
-	}
-
-#if 1	/* This code is there only for consistency checking and can be disabled in future */
-	host->dma_size = 0;
-	for (i = 0; i < host->dma_nents; i++)
-		host->dma_size += data->sg[i].length;
-
-	if (datasz > host->dma_size) {
-		dev_err(mmc_dev(host->mmc), "imxmci_setup_data datasz 0x%x > 0x%x dm_size\n",
-			datasz, host->dma_size);
-	}
-#endif
-
-	host->dma_size = datasz;
-
-	wmb();
-
-	set_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events);
-	clear_bit(IMXMCI_PEND_CPU_DATA_b, &host->pending_events);
-
-	/* start DMA engine for read, write is delayed after initial response */
-	if (host->dma_dir == DMA_FROM_DEVICE)
-		imx_dma_enable(host->dma);
-}
-
-static void imxmci_start_cmd(struct imxmci_host *host, struct mmc_command *cmd, unsigned int cmdat)
-{
-	unsigned long flags;
-	u32 imask;
-
-	WARN_ON(host->cmd != NULL);
-	host->cmd = cmd;
-
-	/* Ensure, that clock are stopped else command programming and start fails */
-	imxmci_stop_clock(host);
-
-	if (cmd->flags & MMC_RSP_BUSY)
-		cmdat |= CMD_DAT_CONT_BUSY;
-
-	switch (mmc_resp_type(cmd)) {
-	case MMC_RSP_R1: /* short CRC, OPCODE */
-	case MMC_RSP_R1B:/* short CRC, OPCODE, BUSY */
-		cmdat |= CMD_DAT_CONT_RESPONSE_FORMAT_R1;
-		break;
-	case MMC_RSP_R2: /* long 136 bit + CRC */
-		cmdat |= CMD_DAT_CONT_RESPONSE_FORMAT_R2;
-		break;
-	case MMC_RSP_R3: /* short */
-		cmdat |= CMD_DAT_CONT_RESPONSE_FORMAT_R3;
-		break;
-	default:
-		break;
-	}
-
-	if (test_and_clear_bit(IMXMCI_PEND_SET_INIT_b, &host->pending_events))
-		cmdat |= CMD_DAT_CONT_INIT; /* This command needs init */
-
-	if (host->actual_bus_width == MMC_BUS_WIDTH_4)
-		cmdat |= CMD_DAT_CONT_BUS_WIDTH_4;
-
-	writew(cmd->opcode, host->base + MMC_REG_CMD);
-	writew(cmd->arg >> 16, host->base + MMC_REG_ARGH);
-	writew(cmd->arg & 0xffff, host->base + MMC_REG_ARGL);
-	writew(cmdat, host->base + MMC_REG_CMD_DAT_CONT);
-
-	atomic_set(&host->stuck_timeout, 0);
-	set_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events);
-
-
-	imask = IMXMCI_INT_MASK_DEFAULT;
-	imask &= ~INT_MASK_END_CMD_RES;
-	if (cmdat & CMD_DAT_CONT_DATA_ENABLE) {
-		/* imask &= ~INT_MASK_BUF_READY; */
-		imask &= ~INT_MASK_DATA_TRAN;
-		if (cmdat & CMD_DAT_CONT_WRITE)
-			imask &= ~INT_MASK_WRITE_OP_DONE;
-		if (test_bit(IMXMCI_PEND_CPU_DATA_b, &host->pending_events))
-			imask &= ~INT_MASK_BUF_READY;
-	}
-
-	spin_lock_irqsave(&host->lock, flags);
-	host->imask = imask;
-	writew(host->imask, host->base + MMC_REG_INT_MASK);
-	spin_unlock_irqrestore(&host->lock, flags);
-
-	dev_dbg(mmc_dev(host->mmc), "CMD%02d (0x%02x) mask set to 0x%04x\n",
-		cmd->opcode, cmd->opcode, imask);
-
-	imxmci_start_clock(host);
-}
-
-static void imxmci_finish_request(struct imxmci_host *host, struct mmc_request *req)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&host->lock, flags);
-
-	host->pending_events &= ~(IMXMCI_PEND_WAIT_RESP_m | IMXMCI_PEND_DMA_END_m |
-				  IMXMCI_PEND_DMA_DATA_m | IMXMCI_PEND_CPU_DATA_m);
-
-	host->imask = IMXMCI_INT_MASK_DEFAULT;
-	writew(host->imask, host->base + MMC_REG_INT_MASK);
-
-	spin_unlock_irqrestore(&host->lock, flags);
-
-	if (req && req->cmd)
-		host->prev_cmd_code = req->cmd->opcode;
-
-	host->req = NULL;
-	host->cmd = NULL;
-	host->data = NULL;
-	mmc_request_done(host->mmc, req);
-}
-
-static int imxmci_finish_data(struct imxmci_host *host, unsigned int stat)
-{
-	struct mmc_data *data = host->data;
-	int data_error;
-
-	if (test_and_clear_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events)) {
-		imx_dma_disable(host->dma);
-		dma_unmap_sg(mmc_dev(host->mmc), data->sg, host->dma_nents,
-			     host->dma_dir);
-	}
-
-	if (stat & STATUS_ERR_MASK) {
-		dev_dbg(mmc_dev(host->mmc), "request failed. status: 0x%08x\n", stat);
-		if (stat & (STATUS_CRC_READ_ERR | STATUS_CRC_WRITE_ERR))
-			data->error = -EILSEQ;
-		else if (stat & STATUS_TIME_OUT_READ)
-			data->error = -ETIMEDOUT;
-		else
-			data->error = -EIO;
-	} else {
-		data->bytes_xfered = host->dma_size;
-	}
-
-	data_error = data->error;
-
-	host->data = NULL;
-
-	return data_error;
-}
-
-static int imxmci_cmd_done(struct imxmci_host *host, unsigned int stat)
-{
-	struct mmc_command *cmd = host->cmd;
-	int i;
-	u32 a, b, c;
-	struct mmc_data *data = host->data;
-
-	if (!cmd)
-		return 0;
-
-	host->cmd = NULL;
-
-	if (stat & STATUS_TIME_OUT_RESP) {
-		dev_dbg(mmc_dev(host->mmc), "CMD TIMEOUT\n");
-		cmd->error = -ETIMEDOUT;
-	} else if (stat & STATUS_RESP_CRC_ERR && cmd->flags & MMC_RSP_CRC) {
-		dev_dbg(mmc_dev(host->mmc), "cmd crc error\n");
-		cmd->error = -EILSEQ;
-	}
-
-	if (cmd->flags & MMC_RSP_PRESENT) {
-		if (cmd->flags & MMC_RSP_136) {
-			for (i = 0; i < 4; i++) {
-				a = readw(host->base + MMC_REG_RES_FIFO);
-				b = readw(host->base + MMC_REG_RES_FIFO);
-				cmd->resp[i] = a << 16 | b;
-			}
-		} else {
-			a = readw(host->base + MMC_REG_RES_FIFO);
-			b = readw(host->base + MMC_REG_RES_FIFO);
-			c = readw(host->base + MMC_REG_RES_FIFO);
-			cmd->resp[0] = a << 24 | b << 8 | c >> 8;
-		}
-	}
-
-	dev_dbg(mmc_dev(host->mmc), "RESP 0x%08x, 0x%08x, 0x%08x, 0x%08x, error %d\n",
-		cmd->resp[0], cmd->resp[1], cmd->resp[2], cmd->resp[3], cmd->error);
-
-	if (data && !cmd->error && !(stat & STATUS_ERR_MASK)) {
-		if (host->req->data->flags & MMC_DATA_WRITE) {
-
-			/* Wait for FIFO to be empty before starting DMA write */
-
-			stat = readw(host->base + MMC_REG_STATUS);
-			if (imxmci_busy_wait_for_status(host, &stat,
-							STATUS_APPL_BUFF_FE,
-							40, "imxmci_cmd_done DMA WR") < 0) {
-				cmd->error = -EIO;
-				imxmci_finish_data(host, stat);
-				if (host->req)
-					imxmci_finish_request(host, host->req);
-				dev_warn(mmc_dev(host->mmc), "STATUS = 0x%04x\n",
-					 stat);
-				return 0;
-			}
-
-			if (test_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events))
-				imx_dma_enable(host->dma);
-		}
-	} else {
-		struct mmc_request *req;
-		imxmci_stop_clock(host);
-		req = host->req;
-
-		if (data)
-			imxmci_finish_data(host, stat);
-
-		if (req)
-			imxmci_finish_request(host, req);
-		else
-			dev_warn(mmc_dev(host->mmc), "imxmci_cmd_done: no request to finish\n");
-	}
-
-	return 1;
-}
-
-static int imxmci_data_done(struct imxmci_host *host, unsigned int stat)
-{
-	struct mmc_data *data = host->data;
-	int data_error;
-
-	if (!data)
-		return 0;
-
-	data_error = imxmci_finish_data(host, stat);
-
-	if (host->req->stop) {
-		imxmci_stop_clock(host);
-		imxmci_start_cmd(host, host->req->stop, 0);
-	} else {
-		struct mmc_request *req;
-		req = host->req;
-		if (req)
-			imxmci_finish_request(host, req);
-		else
-			dev_warn(mmc_dev(host->mmc), "imxmci_data_done: no request to finish\n");
-	}
-
-	return 1;
-}
-
-static int imxmci_cpu_driven_data(struct imxmci_host *host, unsigned int *pstat)
-{
-	int i;
-	int burst_len;
-	int trans_done = 0;
-	unsigned int stat = *pstat;
-
-	if (host->actual_bus_width != MMC_BUS_WIDTH_4)
-		burst_len = 16;
-	else
-		burst_len = 64;
-
-	/* This is unfortunately required */
-	dev_dbg(mmc_dev(host->mmc), "imxmci_cpu_driven_data running STATUS = 0x%x\n",
-		stat);
-
-	udelay(20);	/* required for clocks < 8MHz*/
-
-	if (host->dma_dir == DMA_FROM_DEVICE) {
-		imxmci_busy_wait_for_status(host, &stat,
-					    STATUS_APPL_BUFF_FF | STATUS_DATA_TRANS_DONE |
-					    STATUS_TIME_OUT_READ,
-					    50, "imxmci_cpu_driven_data read");
-
-		while ((stat & (STATUS_APPL_BUFF_FF | STATUS_DATA_TRANS_DONE)) &&
-		       !(stat & STATUS_TIME_OUT_READ) &&
-		       (host->data_cnt < 512)) {
-
-			udelay(20);	/* required for clocks < 8MHz*/
-
-			for (i = burst_len; i >= 2 ; i -= 2) {
-				u16 data;
-				data = readw(host->base + MMC_REG_BUFFER_ACCESS);
-				udelay(10);	/* required for clocks < 8MHz*/
-				if (host->data_cnt+2 <= host->dma_size) {
-					*(host->data_ptr++) = data;
-				} else {
-					if (host->data_cnt < host->dma_size)
-						*(u8 *)(host->data_ptr) = data;
-				}
-				host->data_cnt += 2;
-			}
-
-			stat = readw(host->base + MMC_REG_STATUS);
-
-			dev_dbg(mmc_dev(host->mmc), "imxmci_cpu_driven_data read %d burst %d STATUS = 0x%x\n",
-				host->data_cnt, burst_len, stat);
-		}
-
-		if ((stat & STATUS_DATA_TRANS_DONE) && (host->data_cnt >= 512))
-			trans_done = 1;
-
-		if (host->dma_size & 0x1ff)
-			stat &= ~STATUS_CRC_READ_ERR;
-
-		if (stat & STATUS_TIME_OUT_READ) {
-			dev_dbg(mmc_dev(host->mmc), "imxmci_cpu_driven_data read timeout STATUS = 0x%x\n",
-				stat);
-			trans_done = -1;
-		}
-
-	} else {
-		imxmci_busy_wait_for_status(host, &stat,
-					    STATUS_APPL_BUFF_FE,
-					    20, "imxmci_cpu_driven_data write");
-
-		while ((stat & STATUS_APPL_BUFF_FE) &&
-		       (host->data_cnt < host->dma_size)) {
-			if (burst_len >= host->dma_size - host->data_cnt) {
-				burst_len = host->dma_size - host->data_cnt;
-				host->data_cnt = host->dma_size;
-				trans_done = 1;
-			} else {
-				host->data_cnt += burst_len;
-			}
-
-			for (i = burst_len; i > 0 ; i -= 2)
-				writew(*(host->data_ptr++), host->base + MMC_REG_BUFFER_ACCESS);
-
-			stat = readw(host->base + MMC_REG_STATUS);
-
-			dev_dbg(mmc_dev(host->mmc), "imxmci_cpu_driven_data write burst %d STATUS = 0x%x\n",
-				burst_len, stat);
-		}
-	}
-
-	*pstat = stat;
-
-	return trans_done;
-}
-
-static void imxmci_dma_irq(int dma, void *devid)
-{
-	struct imxmci_host *host = devid;
-	u32 stat = readw(host->base + MMC_REG_STATUS);
-
-	atomic_set(&host->stuck_timeout, 0);
-	host->status_reg = stat;
-	set_bit(IMXMCI_PEND_DMA_END_b, &host->pending_events);
-	tasklet_schedule(&host->tasklet);
-}
-
-static irqreturn_t imxmci_irq(int irq, void *devid)
-{
-	struct imxmci_host *host = devid;
-	u32 stat = readw(host->base + MMC_REG_STATUS);
-	int handled = 1;
-
-	writew(host->imask | INT_MASK_SDIO | INT_MASK_AUTO_CARD_DETECT,
-			host->base + MMC_REG_INT_MASK);
-
-	atomic_set(&host->stuck_timeout, 0);
-	host->status_reg = stat;
-	set_bit(IMXMCI_PEND_IRQ_b, &host->pending_events);
-	set_bit(IMXMCI_PEND_STARTED_b, &host->pending_events);
-	tasklet_schedule(&host->tasklet);
-
-	return IRQ_RETVAL(handled);;
-}
-
-static void imxmci_tasklet_fnc(unsigned long data)
-{
-	struct imxmci_host *host = (struct imxmci_host *)data;
-	u32 stat;
-	unsigned int data_dir_mask = 0;	/* STATUS_WR_CRC_ERROR_CODE_MASK */
-	int timeout = 0;
-
-	if (atomic_read(&host->stuck_timeout) > 4) {
-		char *what;
-		timeout = 1;
-		stat = readw(host->base + MMC_REG_STATUS);
-		host->status_reg = stat;
-		if (test_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events))
-			if (test_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events))
-				what = "RESP+DMA";
-			else
-				what = "RESP";
-		else
-			if (test_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events))
-				if (test_bit(IMXMCI_PEND_DMA_END_b, &host->pending_events))
-					what = "DATA";
-				else
-					what = "DMA";
-			else
-				what = "???";
-
-		dev_err(mmc_dev(host->mmc),
-			"%s TIMEOUT, hardware stucked STATUS = 0x%04x IMASK = 0x%04x\n",
-			what, stat,
-			readw(host->base + MMC_REG_INT_MASK));
-		dev_err(mmc_dev(host->mmc),
-			"CMD_DAT_CONT = 0x%04x, MMC_BLK_LEN = 0x%04x, MMC_NOB = 0x%04x, DMA_CCR = 0x%08x\n",
-			readw(host->base + MMC_REG_CMD_DAT_CONT),
-			readw(host->base + MMC_REG_BLK_LEN),
-			readw(host->base + MMC_REG_NOB),
-			CCR(host->dma));
-		dev_err(mmc_dev(host->mmc), "CMD%d, prevCMD%d, bus %d-bit, dma_size = 0x%x\n",
-			host->cmd ? host->cmd->opcode : 0,
-			host->prev_cmd_code,
-			1 << host->actual_bus_width, host->dma_size);
-	}
-
-	if (!host->present || timeout)
-		host->status_reg = STATUS_TIME_OUT_RESP | STATUS_TIME_OUT_READ |
-			STATUS_CRC_READ_ERR | STATUS_CRC_WRITE_ERR;
-
-	if (test_bit(IMXMCI_PEND_IRQ_b, &host->pending_events) || timeout) {
-		clear_bit(IMXMCI_PEND_IRQ_b, &host->pending_events);
-
-		stat = readw(host->base + MMC_REG_STATUS);
-		/*
-		 * This is not required in theory, but there is chance to miss some flag
-		 * which clears automatically by mask write, FreeScale original code keeps
-		 * stat from IRQ time so do I
-		 */
-		stat |= host->status_reg;
-
-		if (test_bit(IMXMCI_PEND_CPU_DATA_b, &host->pending_events))
-			stat &= ~STATUS_CRC_READ_ERR;
-
-		if (test_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events)) {
-			imxmci_busy_wait_for_status(host, &stat,
-						    STATUS_END_CMD_RESP | STATUS_ERR_MASK,
-						    20, "imxmci_tasklet_fnc resp (ERRATUM #4)");
-		}
-
-		if (stat & (STATUS_END_CMD_RESP | STATUS_ERR_MASK)) {
-			if (test_and_clear_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events))
-				imxmci_cmd_done(host, stat);
-			if (host->data && (stat & STATUS_ERR_MASK))
-				imxmci_data_done(host, stat);
-		}
-
-		if (test_bit(IMXMCI_PEND_CPU_DATA_b, &host->pending_events)) {
-			stat |= readw(host->base + MMC_REG_STATUS);
-			if (imxmci_cpu_driven_data(host, &stat)) {
-				if (test_and_clear_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events))
-					imxmci_cmd_done(host, stat);
-				atomic_clear_mask(IMXMCI_PEND_IRQ_m|IMXMCI_PEND_CPU_DATA_m,
-						  &host->pending_events);
-				imxmci_data_done(host, stat);
-			}
-		}
-	}
-
-	if (test_bit(IMXMCI_PEND_DMA_END_b, &host->pending_events) &&
-	    !test_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events)) {
-
-		stat = readw(host->base + MMC_REG_STATUS);
-		/* Same as above */
-		stat |= host->status_reg;
-
-		if (host->dma_dir == DMA_TO_DEVICE)
-			data_dir_mask = STATUS_WRITE_OP_DONE;
-		else
-			data_dir_mask = STATUS_DATA_TRANS_DONE;
-
-		if (stat & data_dir_mask) {
-			clear_bit(IMXMCI_PEND_DMA_END_b, &host->pending_events);
-			imxmci_data_done(host, stat);
-		}
-	}
-
-	if (test_and_clear_bit(IMXMCI_PEND_CARD_XCHG_b, &host->pending_events)) {
-
-		if (host->cmd)
-			imxmci_cmd_done(host, STATUS_TIME_OUT_RESP);
-
-		if (host->data)
-			imxmci_data_done(host, STATUS_TIME_OUT_READ |
-					 STATUS_CRC_READ_ERR | STATUS_CRC_WRITE_ERR);
-
-		if (host->req)
-			imxmci_finish_request(host, host->req);
-
-		mmc_detect_change(host->mmc, msecs_to_jiffies(100));
-
-	}
-}
-
-static void imxmci_request(struct mmc_host *mmc, struct mmc_request *req)
-{
-	struct imxmci_host *host = mmc_priv(mmc);
-	unsigned int cmdat;
-
-	WARN_ON(host->req != NULL);
-
-	host->req = req;
-
-	cmdat = 0;
-
-	if (req->data) {
-		imxmci_setup_data(host, req->data);
-
-		cmdat |= CMD_DAT_CONT_DATA_ENABLE;
-
-		if (req->data->flags & MMC_DATA_WRITE)
-			cmdat |= CMD_DAT_CONT_WRITE;
-
-		if (req->data->flags & MMC_DATA_STREAM)
-			cmdat |= CMD_DAT_CONT_STREAM_BLOCK;
-	}
-
-	imxmci_start_cmd(host, req->cmd, cmdat);
-}
-
-#define CLK_RATE 19200000
-
-static void imxmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
-{
-	struct imxmci_host *host = mmc_priv(mmc);
-	int prescaler;
-
-	if (ios->bus_width == MMC_BUS_WIDTH_4) {
-		host->actual_bus_width = MMC_BUS_WIDTH_4;
-		imx_gpio_mode(PB11_PF_SD_DAT3);
-		BLR(host->dma) = 0;	/* burst 64 byte read/write */
-	} else {
-		host->actual_bus_width = MMC_BUS_WIDTH_1;
-		imx_gpio_mode(GPIO_PORTB | GPIO_IN | GPIO_PUEN | 11);
-		BLR(host->dma) = 16;	/* burst 16 byte read/write */
-	}
-
-	if (host->power_mode != ios->power_mode) {
-		switch (ios->power_mode) {
-		case MMC_POWER_OFF:
-			break;
-		case MMC_POWER_UP:
-			set_bit(IMXMCI_PEND_SET_INIT_b, &host->pending_events);
-			break;
-		case MMC_POWER_ON:
-			break;
-		}
-		host->power_mode = ios->power_mode;
-	}
-
-	if (ios->clock) {
-		unsigned int clk;
-		u16 reg;
-
-		/* The prescaler is 5 for PERCLK2 equal to 96MHz
-		 * then 96MHz / 5 = 19.2 MHz
-		 */
-		clk = clk_get_rate(host->clk);
-		prescaler = (clk + (CLK_RATE * 7) / 8) / CLK_RATE;
-		switch (prescaler) {
-		case 0:
-		case 1:	prescaler = 0;
-			break;
-		case 2:	prescaler = 1;
-			break;
-		case 3:	prescaler = 2;
-			break;
-		case 4:	prescaler = 4;
-			break;
-		default:
-		case 5:	prescaler = 5;
-			break;
-		}
-
-		dev_dbg(mmc_dev(host->mmc), "PERCLK2 %d MHz -> prescaler %d\n",
-			clk, prescaler);
-
-		for (clk = 0; clk < 8; clk++) {
-			int x;
-			x = CLK_RATE / (1 << clk);
-			if (x <= ios->clock)
-				break;
-		}
-
-		/* enable controller */
-		reg = readw(host->base + MMC_REG_STR_STP_CLK);
-		writew(reg | STR_STP_CLK_ENABLE,
-				host->base + MMC_REG_STR_STP_CLK);
-
-		imxmci_stop_clock(host);
-		writew((prescaler << 3) | clk, host->base + MMC_REG_CLK_RATE);
-		/*
-		 * Under my understanding, clock should not be started there, because it would
-		 * initiate SDHC sequencer and send last or random command into card
-		 */
-		/* imxmci_start_clock(host); */
-
-		dev_dbg(mmc_dev(host->mmc),
-			"MMC_CLK_RATE: 0x%08x\n",
-			readw(host->base + MMC_REG_CLK_RATE));
-	} else {
-		imxmci_stop_clock(host);
-	}
-}
-
-static int imxmci_get_ro(struct mmc_host *mmc)
-{
-	struct imxmci_host *host = mmc_priv(mmc);
-
-	if (host->pdata && host->pdata->get_ro)
-		return !!host->pdata->get_ro(mmc_dev(mmc));
-	/*
-	 * Board doesn't support read only detection; let the mmc core
-	 * decide what to do.
-	 */
-	return -ENOSYS;
-}
-
-
-static const struct mmc_host_ops imxmci_ops = {
-	.request	= imxmci_request,
-	.set_ios	= imxmci_set_ios,
-	.get_ro		= imxmci_get_ro,
-};
-
-static void imxmci_check_status(unsigned long data)
-{
-	struct imxmci_host *host = (struct imxmci_host *)data;
-
-	if (host->pdata && host->pdata->card_present &&
-	    host->pdata->card_present(mmc_dev(host->mmc)) != host->present) {
-		host->present ^= 1;
-		dev_info(mmc_dev(host->mmc), "card %s\n",
-		      host->present ? "inserted" : "removed");
-
-		set_bit(IMXMCI_PEND_CARD_XCHG_b, &host->pending_events);
-		tasklet_schedule(&host->tasklet);
-	}
-
-	if (test_bit(IMXMCI_PEND_WAIT_RESP_b, &host->pending_events) ||
-	    test_bit(IMXMCI_PEND_DMA_DATA_b, &host->pending_events)) {
-		atomic_inc(&host->stuck_timeout);
-		if (atomic_read(&host->stuck_timeout) > 4)
-			tasklet_schedule(&host->tasklet);
-	} else {
-		atomic_set(&host->stuck_timeout, 0);
-
-	}
-
-	mod_timer(&host->timer, jiffies + (HZ>>1));
-}
-
-static int __init imxmci_probe(struct platform_device *pdev)
-{
-	struct mmc_host *mmc;
-	struct imxmci_host *host = NULL;
-	struct resource *r;
-	int ret = 0, irq;
-	u16 rev_no;
-
-	printk(KERN_INFO "i.MX mmc driver\n");
-
-	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	irq = platform_get_irq(pdev, 0);
-	if (!r || irq < 0)
-		return -ENXIO;
-
-	r = request_mem_region(r->start, resource_size(r), pdev->name);
-	if (!r)
-		return -EBUSY;
-
-	mmc = mmc_alloc_host(sizeof(struct imxmci_host), &pdev->dev);
-	if (!mmc) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	mmc->ops = &imxmci_ops;
-	mmc->f_min = 150000;
-	mmc->f_max = CLK_RATE/2;
-	mmc->ocr_avail = MMC_VDD_32_33;
-	mmc->caps = MMC_CAP_4_BIT_DATA;
-
-	/* MMC core transfer sizes tunable parameters */
-	mmc->max_hw_segs = 64;
-	mmc->max_phys_segs = 64;
-	mmc->max_seg_size = 64*512;	/* default PAGE_CACHE_SIZE */
-	mmc->max_req_size = 64*512;	/* default PAGE_CACHE_SIZE */
-	mmc->max_blk_size = 2048;
-	mmc->max_blk_count = 65535;
-
-	host = mmc_priv(mmc);
-	host->base = ioremap(r->start, resource_size(r));
-	if (!host->base) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	host->mmc = mmc;
-	host->dma_allocated = 0;
-	host->pdata = pdev->dev.platform_data;
-	if (!host->pdata)
-		dev_warn(&pdev->dev, "No platform data provided!\n");
-
-	spin_lock_init(&host->lock);
-	host->res = r;
-	host->irq = irq;
-
-	host->clk = clk_get(&pdev->dev, "perclk2");
-	if (IS_ERR(host->clk)) {
-		ret = PTR_ERR(host->clk);
-		goto out;
-	}
-	clk_enable(host->clk);
-
-	imx_gpio_mode(PB8_PF_SD_DAT0);
-	imx_gpio_mode(PB9_PF_SD_DAT1);
-	imx_gpio_mode(PB10_PF_SD_DAT2);
-	/* Configured as GPIO with pull-up to ensure right MCC card mode */
-	/* Switched to PB11_PF_SD_DAT3 if 4 bit bus is configured */
-	imx_gpio_mode(GPIO_PORTB | GPIO_IN | GPIO_PUEN | 11);
-	/* imx_gpio_mode(PB11_PF_SD_DAT3); */
-	imx_gpio_mode(PB12_PF_SD_CLK);
-	imx_gpio_mode(PB13_PF_SD_CMD);
-
-	imxmci_softreset(host);
-
-	rev_no = readw(host->base + MMC_REG_REV_NO);
-	if (rev_no != 0x390) {
-		dev_err(mmc_dev(host->mmc), "wrong rev.no. 0x%08x. aborting.\n",
-			readw(host->base + MMC_REG_REV_NO));
-		goto out;
-	}
-
-	/* recommended in data sheet */
-	writew(0x2db4, host->base + MMC_REG_READ_TO);
-
-	host->imask = IMXMCI_INT_MASK_DEFAULT;
-	writew(host->imask, host->base + MMC_REG_INT_MASK);
-
-	host->dma = imx_dma_request_by_prio(DRIVER_NAME, DMA_PRIO_LOW);
-	if(host->dma < 0) {
-		dev_err(mmc_dev(host->mmc), "imx_dma_request_by_prio failed\n");
-		ret = -EBUSY;
-		goto out;
-	}
-	host->dma_allocated = 1;
-	imx_dma_setup_handlers(host->dma, imxmci_dma_irq, NULL, host);
-	RSSR(host->dma) = DMA_REQ_SDHC;
-
-	tasklet_init(&host->tasklet, imxmci_tasklet_fnc, (unsigned long)host);
-	host->status_reg=0;
-	host->pending_events=0;
-
-	ret = request_irq(host->irq, imxmci_irq, 0, DRIVER_NAME, host);
-	if (ret)
-		goto out;
-
-	if (host->pdata && host->pdata->card_present)
-		host->present = host->pdata->card_present(mmc_dev(mmc));
-	else	/* if there is no way to detect assume that card is present */
-		host->present = 1;
-
-	init_timer(&host->timer);
-	host->timer.data = (unsigned long)host;
-	host->timer.function = imxmci_check_status;
-	add_timer(&host->timer);
-	mod_timer(&host->timer, jiffies + (HZ >> 1));
-
-	platform_set_drvdata(pdev, mmc);
-
-	mmc_add_host(mmc);
-
-	return 0;
-
-out:
-	if (host) {
-		if (host->dma_allocated) {
-			imx_dma_free(host->dma);
-			host->dma_allocated = 0;
-		}
-		if (host->clk) {
-			clk_disable(host->clk);
-			clk_put(host->clk);
-		}
-		if (host->base)
-			iounmap(host->base);
-	}
-	if (mmc)
-		mmc_free_host(mmc);
-	release_mem_region(r->start, resource_size(r));
-	return ret;
-}
-
-static int __exit imxmci_remove(struct platform_device *pdev)
-{
-	struct mmc_host *mmc = platform_get_drvdata(pdev);
-
-	platform_set_drvdata(pdev, NULL);
-
-	if (mmc) {
-		struct imxmci_host *host = mmc_priv(mmc);
-
-		tasklet_disable(&host->tasklet);
-
-		del_timer_sync(&host->timer);
-		mmc_remove_host(mmc);
-
-		free_irq(host->irq, host);
-		iounmap(host->base);
-		if (host->dma_allocated) {
-			imx_dma_free(host->dma);
-			host->dma_allocated = 0;
-		}
-
-		tasklet_kill(&host->tasklet);
-
-		clk_disable(host->clk);
-		clk_put(host->clk);
-
-		release_mem_region(host->res->start, resource_size(host->res));
-
-		mmc_free_host(mmc);
-	}
-	return 0;
-}
-
-#ifdef CONFIG_PM
-static int imxmci_suspend(struct platform_device *dev, pm_message_t state)
-{
-	struct mmc_host *mmc = platform_get_drvdata(dev);
-	int ret = 0;
-
-	if (mmc)
-		ret = mmc_suspend_host(mmc, state);
-
-	return ret;
-}
-
-static int imxmci_resume(struct platform_device *dev)
-{
-	struct mmc_host *mmc = platform_get_drvdata(dev);
-	struct imxmci_host *host;
-	int ret = 0;
-
-	if (mmc) {
-		host = mmc_priv(mmc);
-		if (host)
-			set_bit(IMXMCI_PEND_SET_INIT_b, &host->pending_events);
-		ret = mmc_resume_host(mmc);
-	}
-
-	return ret;
-}
-#else
-#define imxmci_suspend  NULL
-#define imxmci_resume   NULL
-#endif /* CONFIG_PM */
-
-static struct platform_driver imxmci_driver = {
-	.remove		= __exit_p(imxmci_remove),
-	.suspend	= imxmci_suspend,
-	.resume		= imxmci_resume,
-	.driver		= {
-		.name		= DRIVER_NAME,
-		.owner		= THIS_MODULE,
-	}
-};
-
-static int __init imxmci_init(void)
-{
-	return platform_driver_probe(&imxmci_driver, imxmci_probe);
-}
-
-static void __exit imxmci_exit(void)
-{
-	platform_driver_unregister(&imxmci_driver);
-}
-
-module_init(imxmci_init);
-module_exit(imxmci_exit);
-
-MODULE_DESCRIPTION("i.MX Multimedia Card Interface Driver");
-MODULE_AUTHOR("Sascha Hauer, Pengutronix");
-MODULE_LICENSE("GPL");
-MODULE_ALIAS("platform:imx-mmc");
diff --git a/drivers/mmc/host/imxmmc.h b/drivers/mmc/host/imxmmc.h
deleted file mode 100644
index 09d5d4ee3a7..00000000000
--- a/drivers/mmc/host/imxmmc.h
+++ /dev/null
@@ -1,64 +0,0 @@
-#define MMC_REG_STR_STP_CLK		0x00
-#define MMC_REG_STATUS			0x04
-#define MMC_REG_CLK_RATE		0x08
-#define MMC_REG_CMD_DAT_CONT		0x0C
-#define MMC_REG_RES_TO			0x10
-#define MMC_REG_READ_TO			0x14
-#define MMC_REG_BLK_LEN			0x18
-#define MMC_REG_NOB			0x1C
-#define MMC_REG_REV_NO			0x20
-#define MMC_REG_INT_MASK		0x24
-#define MMC_REG_CMD			0x28
-#define MMC_REG_ARGH			0x2C
-#define MMC_REG_ARGL			0x30
-#define MMC_REG_RES_FIFO		0x34
-#define MMC_REG_BUFFER_ACCESS		0x38
-
-#define STR_STP_CLK_IPG_CLK_GATE_DIS    (1<<15)
-#define STR_STP_CLK_IPG_PERCLK_GATE_DIS (1<<14)
-#define STR_STP_CLK_ENDIAN              (1<<5)
-#define STR_STP_CLK_RESET               (1<<3)
-#define STR_STP_CLK_ENABLE              (1<<2)
-#define STR_STP_CLK_START_CLK           (1<<1)
-#define STR_STP_CLK_STOP_CLK            (1<<0)
-#define STATUS_CARD_PRESENCE            (1<<15)
-#define STATUS_SDIO_INT_ACTIVE          (1<<14)
-#define STATUS_END_CMD_RESP             (1<<13)
-#define STATUS_WRITE_OP_DONE            (1<<12)
-#define STATUS_DATA_TRANS_DONE          (1<<11)
-#define STATUS_WR_CRC_ERROR_CODE_MASK   (3<<10)
-#define STATUS_CARD_BUS_CLK_RUN         (1<<8)
-#define STATUS_APPL_BUFF_FF             (1<<7)
-#define STATUS_APPL_BUFF_FE             (1<<6)
-#define STATUS_RESP_CRC_ERR             (1<<5)
-#define STATUS_CRC_READ_ERR             (1<<3)
-#define STATUS_CRC_WRITE_ERR            (1<<2)
-#define STATUS_TIME_OUT_RESP            (1<<1)
-#define STATUS_TIME_OUT_READ            (1<<0)
-#define STATUS_ERR_MASK                 0x2f
-#define CLK_RATE_PRESCALER(x)           ((x) & 0x7)
-#define CLK_RATE_CLK_RATE(x)            (((x) & 0x7) << 3)
-#define CMD_DAT_CONT_CMD_RESP_LONG_OFF  (1<<12)
-#define CMD_DAT_CONT_STOP_READWAIT      (1<<11)
-#define CMD_DAT_CONT_START_READWAIT     (1<<10)
-#define CMD_DAT_CONT_BUS_WIDTH_1        (0<<8)
-#define CMD_DAT_CONT_BUS_WIDTH_4        (2<<8)
-#define CMD_DAT_CONT_INIT               (1<<7)
-#define CMD_DAT_CONT_BUSY               (1<<6)
-#define CMD_DAT_CONT_STREAM_BLOCK       (1<<5)
-#define CMD_DAT_CONT_WRITE              (1<<4)
-#define CMD_DAT_CONT_DATA_ENABLE        (1<<3)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R1 (1)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R2 (2)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R3 (3)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R4 (4)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R5 (5)
-#define CMD_DAT_CONT_RESPONSE_FORMAT_R6 (6)
-#define INT_MASK_AUTO_CARD_DETECT       (1<<6)
-#define INT_MASK_DAT0_EN                (1<<5)
-#define INT_MASK_SDIO                   (1<<4)
-#define INT_MASK_BUF_READY              (1<<3)
-#define INT_MASK_END_CMD_RES            (1<<2)
-#define INT_MASK_WRITE_OP_DONE          (1<<1)
-#define INT_MASK_DATA_TRAN              (1<<0)
-#define INT_ALL                         (0x7f)
diff --git a/drivers/mmc/host/jz4740_mmc.c b/drivers/mmc/host/jz4740_mmc.c
new file mode 100644
index 00000000000..537d6c7a5ae
--- /dev/null
+++ b/drivers/mmc/host/jz4740_mmc.c
@@ -0,0 +1,921 @@
+/*
+ *  Copyright (C) 2009-2010, Lars-Peter Clausen <lars@metafoo.de>
+ *  JZ4740 SD/MMC controller driver
+ *
+ *  This program is free software; you can redistribute  it and/or modify it
+ *  under  the terms of  the GNU General  Public License as published by the
+ *  Free Software Foundation;  either version 2 of the  License, or (at your
+ *  option) any later version.
+ *
+ *  You should have received a copy of the  GNU General Public License along
+ *  with this program; if not, write  to the Free Software Foundation, Inc.,
+ *  675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ */
+
+#include <linux/mmc/host.h>
+#include <linux/mmc/slot-gpio.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/scatterlist.h>
+#include <linux/clk.h>
+
+#include <linux/bitops.h>
+#include <linux/gpio.h>
+#include <asm/mach-jz4740/gpio.h>
+#include <asm/cacheflush.h>
+#include <linux/dma-mapping.h>
+
+#include <asm/mach-jz4740/jz4740_mmc.h>
+
+#define JZ_REG_MMC_STRPCL	0x00
+#define JZ_REG_MMC_STATUS	0x04
+#define JZ_REG_MMC_CLKRT	0x08
+#define JZ_REG_MMC_CMDAT	0x0C
+#define JZ_REG_MMC_RESTO	0x10
+#define JZ_REG_MMC_RDTO		0x14
+#define JZ_REG_MMC_BLKLEN	0x18
+#define JZ_REG_MMC_NOB		0x1C
+#define JZ_REG_MMC_SNOB		0x20
+#define JZ_REG_MMC_IMASK	0x24
+#define JZ_REG_MMC_IREG		0x28
+#define JZ_REG_MMC_CMD		0x2C
+#define JZ_REG_MMC_ARG		0x30
+#define JZ_REG_MMC_RESP_FIFO	0x34
+#define JZ_REG_MMC_RXFIFO	0x38
+#define JZ_REG_MMC_TXFIFO	0x3C
+
+#define JZ_MMC_STRPCL_EXIT_MULTIPLE BIT(7)
+#define JZ_MMC_STRPCL_EXIT_TRANSFER BIT(6)
+#define JZ_MMC_STRPCL_START_READWAIT BIT(5)
+#define JZ_MMC_STRPCL_STOP_READWAIT BIT(4)
+#define JZ_MMC_STRPCL_RESET BIT(3)
+#define JZ_MMC_STRPCL_START_OP BIT(2)
+#define JZ_MMC_STRPCL_CLOCK_CONTROL (BIT(1) | BIT(0))
+#define JZ_MMC_STRPCL_CLOCK_STOP BIT(0)
+#define JZ_MMC_STRPCL_CLOCK_START BIT(1)
+
+
+#define JZ_MMC_STATUS_IS_RESETTING BIT(15)
+#define JZ_MMC_STATUS_SDIO_INT_ACTIVE BIT(14)
+#define JZ_MMC_STATUS_PRG_DONE BIT(13)
+#define JZ_MMC_STATUS_DATA_TRAN_DONE BIT(12)
+#define JZ_MMC_STATUS_END_CMD_RES BIT(11)
+#define JZ_MMC_STATUS_DATA_FIFO_AFULL BIT(10)
+#define JZ_MMC_STATUS_IS_READWAIT BIT(9)
+#define JZ_MMC_STATUS_CLK_EN BIT(8)
+#define JZ_MMC_STATUS_DATA_FIFO_FULL BIT(7)
+#define JZ_MMC_STATUS_DATA_FIFO_EMPTY BIT(6)
+#define JZ_MMC_STATUS_CRC_RES_ERR BIT(5)
+#define JZ_MMC_STATUS_CRC_READ_ERROR BIT(4)
+#define JZ_MMC_STATUS_TIMEOUT_WRITE BIT(3)
+#define JZ_MMC_STATUS_CRC_WRITE_ERROR BIT(2)
+#define JZ_MMC_STATUS_TIMEOUT_RES BIT(1)
+#define JZ_MMC_STATUS_TIMEOUT_READ BIT(0)
+
+#define JZ_MMC_STATUS_READ_ERROR_MASK (BIT(4) | BIT(0))
+#define JZ_MMC_STATUS_WRITE_ERROR_MASK (BIT(3) | BIT(2))
+
+
+#define JZ_MMC_CMDAT_IO_ABORT BIT(11)
+#define JZ_MMC_CMDAT_BUS_WIDTH_4BIT BIT(10)
+#define JZ_MMC_CMDAT_DMA_EN BIT(8)
+#define JZ_MMC_CMDAT_INIT BIT(7)
+#define JZ_MMC_CMDAT_BUSY BIT(6)
+#define JZ_MMC_CMDAT_STREAM BIT(5)
+#define JZ_MMC_CMDAT_WRITE BIT(4)
+#define JZ_MMC_CMDAT_DATA_EN BIT(3)
+#define JZ_MMC_CMDAT_RESPONSE_FORMAT (BIT(2) | BIT(1) | BIT(0))
+#define JZ_MMC_CMDAT_RSP_R1 1
+#define JZ_MMC_CMDAT_RSP_R2 2
+#define JZ_MMC_CMDAT_RSP_R3 3
+
+#define JZ_MMC_IRQ_SDIO BIT(7)
+#define JZ_MMC_IRQ_TXFIFO_WR_REQ BIT(6)
+#define JZ_MMC_IRQ_RXFIFO_RD_REQ BIT(5)
+#define JZ_MMC_IRQ_END_CMD_RES BIT(2)
+#define JZ_MMC_IRQ_PRG_DONE BIT(1)
+#define JZ_MMC_IRQ_DATA_TRAN_DONE BIT(0)
+
+
+#define JZ_MMC_CLK_RATE 24000000
+
+enum jz4740_mmc_state {
+	JZ4740_MMC_STATE_READ_RESPONSE,
+	JZ4740_MMC_STATE_TRANSFER_DATA,
+	JZ4740_MMC_STATE_SEND_STOP,
+	JZ4740_MMC_STATE_DONE,
+};
+
+struct jz4740_mmc_host {
+	struct mmc_host *mmc;
+	struct platform_device *pdev;
+	struct jz4740_mmc_platform_data *pdata;
+	struct clk *clk;
+
+	int irq;
+	int card_detect_irq;
+
+	void __iomem *base;
+	struct mmc_request *req;
+	struct mmc_command *cmd;
+
+	unsigned long waiting;
+
+	uint32_t cmdat;
+
+	uint16_t irq_mask;
+
+	spinlock_t lock;
+
+	struct timer_list timeout_timer;
+	struct sg_mapping_iter miter;
+	enum jz4740_mmc_state state;
+};
+
+static void jz4740_mmc_set_irq_enabled(struct jz4740_mmc_host *host,
+	unsigned int irq, bool enabled)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+	if (enabled)
+		host->irq_mask &= ~irq;
+	else
+		host->irq_mask |= irq;
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	writew(host->irq_mask, host->base + JZ_REG_MMC_IMASK);
+}
+
+static void jz4740_mmc_clock_enable(struct jz4740_mmc_host *host,
+	bool start_transfer)
+{
+	uint16_t val = JZ_MMC_STRPCL_CLOCK_START;
+
+	if (start_transfer)
+		val |= JZ_MMC_STRPCL_START_OP;
+
+	writew(val, host->base + JZ_REG_MMC_STRPCL);
+}
+
+static void jz4740_mmc_clock_disable(struct jz4740_mmc_host *host)
+{
+	uint32_t status;
+	unsigned int timeout = 1000;
+
+	writew(JZ_MMC_STRPCL_CLOCK_STOP, host->base + JZ_REG_MMC_STRPCL);
+	do {
+		status = readl(host->base + JZ_REG_MMC_STATUS);
+	} while (status & JZ_MMC_STATUS_CLK_EN && --timeout);
+}
+
+static void jz4740_mmc_reset(struct jz4740_mmc_host *host)
+{
+	uint32_t status;
+	unsigned int timeout = 1000;
+
+	writew(JZ_MMC_STRPCL_RESET, host->base + JZ_REG_MMC_STRPCL);
+	udelay(10);
+	do {
+		status = readl(host->base + JZ_REG_MMC_STATUS);
+	} while (status & JZ_MMC_STATUS_IS_RESETTING && --timeout);
+}
+
+static void jz4740_mmc_request_done(struct jz4740_mmc_host *host)
+{
+	struct mmc_request *req;
+
+	req = host->req;
+	host->req = NULL;
+
+	mmc_request_done(host->mmc, req);
+}
+
+static unsigned int jz4740_mmc_poll_irq(struct jz4740_mmc_host *host,
+	unsigned int irq)
+{
+	unsigned int timeout = 0x800;
+	uint16_t status;
+
+	do {
+		status = readw(host->base + JZ_REG_MMC_IREG);
+	} while (!(status & irq) && --timeout);
+
+	if (timeout == 0) {
+		set_bit(0, &host->waiting);
+		mod_timer(&host->timeout_timer, jiffies + 5*HZ);
+		jz4740_mmc_set_irq_enabled(host, irq, true);
+		return true;
+	}
+
+	return false;
+}
+
+static void jz4740_mmc_transfer_check_state(struct jz4740_mmc_host *host,
+	struct mmc_data *data)
+{
+	int status;
+
+	status = readl(host->base + JZ_REG_MMC_STATUS);
+	if (status & JZ_MMC_STATUS_WRITE_ERROR_MASK) {
+		if (status & (JZ_MMC_STATUS_TIMEOUT_WRITE)) {
+			host->req->cmd->error = -ETIMEDOUT;
+			data->error = -ETIMEDOUT;
+		} else {
+			host->req->cmd->error = -EIO;
+			data->error = -EIO;
+		}
+	} else if (status & JZ_MMC_STATUS_READ_ERROR_MASK) {
+		if (status & (JZ_MMC_STATUS_TIMEOUT_READ)) {
+			host->req->cmd->error = -ETIMEDOUT;
+			data->error = -ETIMEDOUT;
+		} else {
+			host->req->cmd->error = -EIO;
+			data->error = -EIO;
+		}
+	}
+}
+
+static bool jz4740_mmc_write_data(struct jz4740_mmc_host *host,
+	struct mmc_data *data)
+{
+	struct sg_mapping_iter *miter = &host->miter;
+	void __iomem *fifo_addr = host->base + JZ_REG_MMC_TXFIFO;
+	uint32_t *buf;
+	bool timeout;
+	size_t i, j;
+
+	while (sg_miter_next(miter)) {
+		buf = miter->addr;
+		i = miter->length / 4;
+		j = i / 8;
+		i = i & 0x7;
+		while (j) {
+			timeout = jz4740_mmc_poll_irq(host, JZ_MMC_IRQ_TXFIFO_WR_REQ);
+			if (unlikely(timeout))
+				goto poll_timeout;
+
+			writel(buf[0], fifo_addr);
+			writel(buf[1], fifo_addr);
+			writel(buf[2], fifo_addr);
+			writel(buf[3], fifo_addr);
+			writel(buf[4], fifo_addr);
+			writel(buf[5], fifo_addr);
+			writel(buf[6], fifo_addr);
+			writel(buf[7], fifo_addr);
+			buf += 8;
+			--j;
+		}
+		if (unlikely(i)) {
+			timeout = jz4740_mmc_poll_irq(host, JZ_MMC_IRQ_TXFIFO_WR_REQ);
+			if (unlikely(timeout))
+				goto poll_timeout;
+
+			while (i) {
+				writel(*buf, fifo_addr);
+				++buf;
+				--i;
+			}
+		}
+		data->bytes_xfered += miter->length;
+	}
+	sg_miter_stop(miter);
+
+	return false;
+
+poll_timeout:
+	miter->consumed = (void *)buf - miter->addr;
+	data->bytes_xfered += miter->consumed;
+	sg_miter_stop(miter);
+
+	return true;
+}
+
+static bool jz4740_mmc_read_data(struct jz4740_mmc_host *host,
+				struct mmc_data *data)
+{
+	struct sg_mapping_iter *miter = &host->miter;
+	void __iomem *fifo_addr = host->base + JZ_REG_MMC_RXFIFO;
+	uint32_t *buf;
+	uint32_t d;
+	uint16_t status;
+	size_t i, j;
+	unsigned int timeout;
+
+	while (sg_miter_next(miter)) {
+		buf = miter->addr;
+		i = miter->length;
+		j = i / 32;
+		i = i & 0x1f;
+		while (j) {
+			timeout = jz4740_mmc_poll_irq(host, JZ_MMC_IRQ_RXFIFO_RD_REQ);
+			if (unlikely(timeout))
+				goto poll_timeout;
+
+			buf[0] = readl(fifo_addr);
+			buf[1] = readl(fifo_addr);
+			buf[2] = readl(fifo_addr);
+			buf[3] = readl(fifo_addr);
+			buf[4] = readl(fifo_addr);
+			buf[5] = readl(fifo_addr);
+			buf[6] = readl(fifo_addr);
+			buf[7] = readl(fifo_addr);
+
+			buf += 8;
+			--j;
+		}
+
+		if (unlikely(i)) {
+			timeout = jz4740_mmc_poll_irq(host, JZ_MMC_IRQ_RXFIFO_RD_REQ);
+			if (unlikely(timeout))
+				goto poll_timeout;
+
+			while (i >= 4) {
+				*buf++ = readl(fifo_addr);
+				i -= 4;
+			}
+			if (unlikely(i > 0)) {
+				d = readl(fifo_addr);
+				memcpy(buf, &d, i);
+			}
+		}
+		data->bytes_xfered += miter->length;
+
+		/* This can go away once MIPS implements
+		 * flush_kernel_dcache_page */
+		flush_dcache_page(miter->page);
+	}
+	sg_miter_stop(miter);
+
+	/* For whatever reason there is sometime one word more in the fifo then
+	 * requested */
+	timeout = 1000;
+	status = readl(host->base + JZ_REG_MMC_STATUS);
+	while (!(status & JZ_MMC_STATUS_DATA_FIFO_EMPTY) && --timeout) {
+		d = readl(fifo_addr);
+		status = readl(host->base + JZ_REG_MMC_STATUS);
+	}
+
+	return false;
+
+poll_timeout:
+	miter->consumed = (void *)buf - miter->addr;
+	data->bytes_xfered += miter->consumed;
+	sg_miter_stop(miter);
+
+	return true;
+}
+
+static void jz4740_mmc_timeout(unsigned long data)
+{
+	struct jz4740_mmc_host *host = (struct jz4740_mmc_host *)data;
+
+	if (!test_and_clear_bit(0, &host->waiting))
+		return;
+
+	jz4740_mmc_set_irq_enabled(host, JZ_MMC_IRQ_END_CMD_RES, false);
+
+	host->req->cmd->error = -ETIMEDOUT;
+	jz4740_mmc_request_done(host);
+}
+
+static void jz4740_mmc_read_response(struct jz4740_mmc_host *host,
+	struct mmc_command *cmd)
+{
+	int i;
+	uint16_t tmp;
+	void __iomem *fifo_addr = host->base + JZ_REG_MMC_RESP_FIFO;
+
+	if (cmd->flags & MMC_RSP_136) {
+		tmp = readw(fifo_addr);
+		for (i = 0; i < 4; ++i) {
+			cmd->resp[i] = tmp << 24;
+			tmp = readw(fifo_addr);
+			cmd->resp[i] |= tmp << 8;
+			tmp = readw(fifo_addr);
+			cmd->resp[i] |= tmp >> 8;
+		}
+	} else {
+		cmd->resp[0] = readw(fifo_addr) << 24;
+		cmd->resp[0] |= readw(fifo_addr) << 8;
+		cmd->resp[0] |= readw(fifo_addr) & 0xff;
+	}
+}
+
+static void jz4740_mmc_send_command(struct jz4740_mmc_host *host,
+	struct mmc_command *cmd)
+{
+	uint32_t cmdat = host->cmdat;
+
+	host->cmdat &= ~JZ_MMC_CMDAT_INIT;
+	jz4740_mmc_clock_disable(host);
+
+	host->cmd = cmd;
+
+	if (cmd->flags & MMC_RSP_BUSY)
+		cmdat |= JZ_MMC_CMDAT_BUSY;
+
+	switch (mmc_resp_type(cmd)) {
+	case MMC_RSP_R1B:
+	case MMC_RSP_R1:
+		cmdat |= JZ_MMC_CMDAT_RSP_R1;
+		break;
+	case MMC_RSP_R2:
+		cmdat |= JZ_MMC_CMDAT_RSP_R2;
+		break;
+	case MMC_RSP_R3:
+		cmdat |= JZ_MMC_CMDAT_RSP_R3;
+		break;
+	default:
+		break;
+	}
+
+	if (cmd->data) {
+		cmdat |= JZ_MMC_CMDAT_DATA_EN;
+		if (cmd->data->flags & MMC_DATA_WRITE)
+			cmdat |= JZ_MMC_CMDAT_WRITE;
+		if (cmd->data->flags & MMC_DATA_STREAM)
+			cmdat |= JZ_MMC_CMDAT_STREAM;
+
+		writew(cmd->data->blksz, host->base + JZ_REG_MMC_BLKLEN);
+		writew(cmd->data->blocks, host->base + JZ_REG_MMC_NOB);
+	}
+
+	writeb(cmd->opcode, host->base + JZ_REG_MMC_CMD);
+	writel(cmd->arg, host->base + JZ_REG_MMC_ARG);
+	writel(cmdat, host->base + JZ_REG_MMC_CMDAT);
+
+	jz4740_mmc_clock_enable(host, 1);
+}
+
+static void jz_mmc_prepare_data_transfer(struct jz4740_mmc_host *host)
+{
+	struct mmc_command *cmd = host->req->cmd;
+	struct mmc_data *data = cmd->data;
+	int direction;
+
+	if (data->flags & MMC_DATA_READ)
+		direction = SG_MITER_TO_SG;
+	else
+		direction = SG_MITER_FROM_SG;
+
+	sg_miter_start(&host->miter, data->sg, data->sg_len, direction);
+}
+
+
+static irqreturn_t jz_mmc_irq_worker(int irq, void *devid)
+{
+	struct jz4740_mmc_host *host = (struct jz4740_mmc_host *)devid;
+	struct mmc_command *cmd = host->req->cmd;
+	struct mmc_request *req = host->req;
+	bool timeout = false;
+
+	if (cmd->error)
+		host->state = JZ4740_MMC_STATE_DONE;
+
+	switch (host->state) {
+	case JZ4740_MMC_STATE_READ_RESPONSE:
+		if (cmd->flags & MMC_RSP_PRESENT)
+			jz4740_mmc_read_response(host, cmd);
+
+		if (!cmd->data)
+			break;
+
+		jz_mmc_prepare_data_transfer(host);
+
+	case JZ4740_MMC_STATE_TRANSFER_DATA:
+		if (cmd->data->flags & MMC_DATA_READ)
+			timeout = jz4740_mmc_read_data(host, cmd->data);
+		else
+			timeout = jz4740_mmc_write_data(host, cmd->data);
+
+		if (unlikely(timeout)) {
+			host->state = JZ4740_MMC_STATE_TRANSFER_DATA;
+			break;
+		}
+
+		jz4740_mmc_transfer_check_state(host, cmd->data);
+
+		timeout = jz4740_mmc_poll_irq(host, JZ_MMC_IRQ_DATA_TRAN_DONE);
+		if (unlikely(timeout)) {
+			host->state = JZ4740_MMC_STATE_SEND_STOP;
+			break;
+		}
+		writew(JZ_MMC_IRQ_DATA_TRAN_DONE, host->base + JZ_REG_MMC_IREG);
+
+	case JZ4740_MMC_STATE_SEND_STOP:
+		if (!req->stop)
+			break;
+
+		jz4740_mmc_send_command(host, req->stop);
+
+		if (mmc_resp_type(req->stop) & MMC_RSP_BUSY) {
+			timeout = jz4740_mmc_poll_irq(host,
+						      JZ_MMC_IRQ_PRG_DONE);
+			if (timeout) {
+				host->state = JZ4740_MMC_STATE_DONE;
+				break;
+			}
+		}
+	case JZ4740_MMC_STATE_DONE:
+		break;
+	}
+
+	if (!timeout)
+		jz4740_mmc_request_done(host);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t jz_mmc_irq(int irq, void *devid)
+{
+	struct jz4740_mmc_host *host = devid;
+	struct mmc_command *cmd = host->cmd;
+	uint16_t irq_reg, status, tmp;
+
+	irq_reg = readw(host->base + JZ_REG_MMC_IREG);
+
+	tmp = irq_reg;
+	irq_reg &= ~host->irq_mask;
+
+	tmp &= ~(JZ_MMC_IRQ_TXFIFO_WR_REQ | JZ_MMC_IRQ_RXFIFO_RD_REQ |
+		JZ_MMC_IRQ_PRG_DONE | JZ_MMC_IRQ_DATA_TRAN_DONE);
+
+	if (tmp != irq_reg)
+		writew(tmp & ~irq_reg, host->base + JZ_REG_MMC_IREG);
+
+	if (irq_reg & JZ_MMC_IRQ_SDIO) {
+		writew(JZ_MMC_IRQ_SDIO, host->base + JZ_REG_MMC_IREG);
+		mmc_signal_sdio_irq(host->mmc);
+		irq_reg &= ~JZ_MMC_IRQ_SDIO;
+	}
+
+	if (host->req && cmd && irq_reg) {
+		if (test_and_clear_bit(0, &host->waiting)) {
+			del_timer(&host->timeout_timer);
+
+			status = readl(host->base + JZ_REG_MMC_STATUS);
+
+			if (status & JZ_MMC_STATUS_TIMEOUT_RES) {
+					cmd->error = -ETIMEDOUT;
+			} else if (status & JZ_MMC_STATUS_CRC_RES_ERR) {
+					cmd->error = -EIO;
+			} else if (status & (JZ_MMC_STATUS_CRC_READ_ERROR |
+				    JZ_MMC_STATUS_CRC_WRITE_ERROR)) {
+					if (cmd->data)
+							cmd->data->error = -EIO;
+					cmd->error = -EIO;
+			}
+
+			jz4740_mmc_set_irq_enabled(host, irq_reg, false);
+			writew(irq_reg, host->base + JZ_REG_MMC_IREG);
+
+			return IRQ_WAKE_THREAD;
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int jz4740_mmc_set_clock_rate(struct jz4740_mmc_host *host, int rate)
+{
+	int div = 0;
+	int real_rate;
+
+	jz4740_mmc_clock_disable(host);
+	clk_set_rate(host->clk, JZ_MMC_CLK_RATE);
+
+	real_rate = clk_get_rate(host->clk);
+
+	while (real_rate > rate && div < 7) {
+		++div;
+		real_rate >>= 1;
+	}
+
+	writew(div, host->base + JZ_REG_MMC_CLKRT);
+	return real_rate;
+}
+
+static void jz4740_mmc_request(struct mmc_host *mmc, struct mmc_request *req)
+{
+	struct jz4740_mmc_host *host = mmc_priv(mmc);
+
+	host->req = req;
+
+	writew(0xffff, host->base + JZ_REG_MMC_IREG);
+
+	writew(JZ_MMC_IRQ_END_CMD_RES, host->base + JZ_REG_MMC_IREG);
+	jz4740_mmc_set_irq_enabled(host, JZ_MMC_IRQ_END_CMD_RES, true);
+
+	host->state = JZ4740_MMC_STATE_READ_RESPONSE;
+	set_bit(0, &host->waiting);
+	mod_timer(&host->timeout_timer, jiffies + 5*HZ);
+	jz4740_mmc_send_command(host, req->cmd);
+}
+
+static void jz4740_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct jz4740_mmc_host *host = mmc_priv(mmc);
+	if (ios->clock)
+		jz4740_mmc_set_clock_rate(host, ios->clock);
+
+	switch (ios->power_mode) {
+	case MMC_POWER_UP:
+		jz4740_mmc_reset(host);
+		if (gpio_is_valid(host->pdata->gpio_power))
+			gpio_set_value(host->pdata->gpio_power,
+					!host->pdata->power_active_low);
+		host->cmdat |= JZ_MMC_CMDAT_INIT;
+		clk_prepare_enable(host->clk);
+		break;
+	case MMC_POWER_ON:
+		break;
+	default:
+		if (gpio_is_valid(host->pdata->gpio_power))
+			gpio_set_value(host->pdata->gpio_power,
+					host->pdata->power_active_low);
+		clk_disable_unprepare(host->clk);
+		break;
+	}
+
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_1:
+		host->cmdat &= ~JZ_MMC_CMDAT_BUS_WIDTH_4BIT;
+		break;
+	case MMC_BUS_WIDTH_4:
+		host->cmdat |= JZ_MMC_CMDAT_BUS_WIDTH_4BIT;
+		break;
+	default:
+		break;
+	}
+}
+
+static void jz4740_mmc_enable_sdio_irq(struct mmc_host *mmc, int enable)
+{
+	struct jz4740_mmc_host *host = mmc_priv(mmc);
+	jz4740_mmc_set_irq_enabled(host, JZ_MMC_IRQ_SDIO, enable);
+}
+
+static const struct mmc_host_ops jz4740_mmc_ops = {
+	.request	= jz4740_mmc_request,
+	.set_ios	= jz4740_mmc_set_ios,
+	.get_ro		= mmc_gpio_get_ro,
+	.get_cd		= mmc_gpio_get_cd,
+	.enable_sdio_irq = jz4740_mmc_enable_sdio_irq,
+};
+
+static const struct jz_gpio_bulk_request jz4740_mmc_pins[] = {
+	JZ_GPIO_BULK_PIN(MSC_CMD),
+	JZ_GPIO_BULK_PIN(MSC_CLK),
+	JZ_GPIO_BULK_PIN(MSC_DATA0),
+	JZ_GPIO_BULK_PIN(MSC_DATA1),
+	JZ_GPIO_BULK_PIN(MSC_DATA2),
+	JZ_GPIO_BULK_PIN(MSC_DATA3),
+};
+
+static int jz4740_mmc_request_gpio(struct device *dev, int gpio,
+	const char *name, bool output, int value)
+{
+	int ret;
+
+	if (!gpio_is_valid(gpio))
+		return 0;
+
+	ret = gpio_request(gpio, name);
+	if (ret) {
+		dev_err(dev, "Failed to request %s gpio: %d\n", name, ret);
+		return ret;
+	}
+
+	if (output)
+		gpio_direction_output(gpio, value);
+	else
+		gpio_direction_input(gpio);
+
+	return 0;
+}
+
+static int jz4740_mmc_request_gpios(struct mmc_host *mmc,
+	struct platform_device *pdev)
+{
+	struct jz4740_mmc_platform_data *pdata = pdev->dev.platform_data;
+	int ret = 0;
+
+	if (!pdata)
+		return 0;
+
+	if (!pdata->card_detect_active_low)
+		mmc->caps2 |= MMC_CAP2_CD_ACTIVE_HIGH;
+	if (!pdata->read_only_active_low)
+		mmc->caps2 |= MMC_CAP2_RO_ACTIVE_HIGH;
+
+	if (gpio_is_valid(pdata->gpio_card_detect)) {
+		ret = mmc_gpio_request_cd(mmc, pdata->gpio_card_detect, 0);
+		if (ret)
+			return ret;
+	}
+
+	if (gpio_is_valid(pdata->gpio_read_only)) {
+		ret = mmc_gpio_request_ro(mmc, pdata->gpio_read_only);
+		if (ret)
+			return ret;
+	}
+
+	return jz4740_mmc_request_gpio(&pdev->dev, pdata->gpio_power,
+			"MMC read only", true, pdata->power_active_low);
+}
+
+static void jz4740_mmc_free_gpios(struct platform_device *pdev)
+{
+	struct jz4740_mmc_platform_data *pdata = pdev->dev.platform_data;
+
+	if (!pdata)
+		return;
+
+	if (gpio_is_valid(pdata->gpio_power))
+		gpio_free(pdata->gpio_power);
+}
+
+static inline size_t jz4740_mmc_num_pins(struct jz4740_mmc_host *host)
+{
+	size_t num_pins = ARRAY_SIZE(jz4740_mmc_pins);
+	if (host->pdata && host->pdata->data_1bit)
+		num_pins -= 3;
+
+	return num_pins;
+}
+
+static int jz4740_mmc_probe(struct platform_device* pdev)
+{
+	int ret;
+	struct mmc_host *mmc;
+	struct jz4740_mmc_host *host;
+	struct jz4740_mmc_platform_data *pdata;
+	struct resource *res;
+
+	pdata = pdev->dev.platform_data;
+
+	mmc = mmc_alloc_host(sizeof(struct jz4740_mmc_host), &pdev->dev);
+	if (!mmc) {
+		dev_err(&pdev->dev, "Failed to alloc mmc host structure\n");
+		return -ENOMEM;
+	}
+
+	host = mmc_priv(mmc);
+	host->pdata = pdata;
+
+	host->irq = platform_get_irq(pdev, 0);
+	if (host->irq < 0) {
+		ret = host->irq;
+		dev_err(&pdev->dev, "Failed to get platform irq: %d\n", ret);
+		goto err_free_host;
+	}
+
+	host->clk = devm_clk_get(&pdev->dev, "mmc");
+	if (IS_ERR(host->clk)) {
+		ret = PTR_ERR(host->clk);
+		dev_err(&pdev->dev, "Failed to get mmc clock\n");
+		goto err_free_host;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	host->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(host->base)) {
+		ret = PTR_ERR(host->base);
+		goto err_free_host;
+	}
+
+	ret = jz_gpio_bulk_request(jz4740_mmc_pins, jz4740_mmc_num_pins(host));
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to request mmc pins: %d\n", ret);
+		goto err_free_host;
+	}
+
+	ret = jz4740_mmc_request_gpios(mmc, pdev);
+	if (ret)
+		goto err_gpio_bulk_free;
+
+	mmc->ops = &jz4740_mmc_ops;
+	mmc->f_min = JZ_MMC_CLK_RATE / 128;
+	mmc->f_max = JZ_MMC_CLK_RATE;
+	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+	mmc->caps = (pdata && pdata->data_1bit) ? 0 : MMC_CAP_4_BIT_DATA;
+	mmc->caps |= MMC_CAP_SDIO_IRQ;
+
+	mmc->max_blk_size = (1 << 10) - 1;
+	mmc->max_blk_count = (1 << 15) - 1;
+	mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count;
+
+	mmc->max_segs = 128;
+	mmc->max_seg_size = mmc->max_req_size;
+
+	host->mmc = mmc;
+	host->pdev = pdev;
+	spin_lock_init(&host->lock);
+	host->irq_mask = 0xffff;
+
+	ret = request_threaded_irq(host->irq, jz_mmc_irq, jz_mmc_irq_worker, 0,
+			dev_name(&pdev->dev), host);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to request irq: %d\n", ret);
+		goto err_free_gpios;
+	}
+
+	jz4740_mmc_reset(host);
+	jz4740_mmc_clock_disable(host);
+	setup_timer(&host->timeout_timer, jz4740_mmc_timeout,
+			(unsigned long)host);
+	/* It is not important when it times out, it just needs to timeout. */
+	set_timer_slack(&host->timeout_timer, HZ);
+
+	platform_set_drvdata(pdev, host);
+	ret = mmc_add_host(mmc);
+
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to add mmc host: %d\n", ret);
+		goto err_free_irq;
+	}
+	dev_info(&pdev->dev, "JZ SD/MMC card driver registered\n");
+
+	return 0;
+
+err_free_irq:
+	free_irq(host->irq, host);
+err_free_gpios:
+	jz4740_mmc_free_gpios(pdev);
+err_gpio_bulk_free:
+	jz_gpio_bulk_free(jz4740_mmc_pins, jz4740_mmc_num_pins(host));
+err_free_host:
+	mmc_free_host(mmc);
+
+	return ret;
+}
+
+static int jz4740_mmc_remove(struct platform_device *pdev)
+{
+	struct jz4740_mmc_host *host = platform_get_drvdata(pdev);
+
+	del_timer_sync(&host->timeout_timer);
+	jz4740_mmc_set_irq_enabled(host, 0xff, false);
+	jz4740_mmc_reset(host);
+
+	mmc_remove_host(host->mmc);
+
+	free_irq(host->irq, host);
+
+	jz4740_mmc_free_gpios(pdev);
+	jz_gpio_bulk_free(jz4740_mmc_pins, jz4740_mmc_num_pins(host));
+
+	mmc_free_host(host->mmc);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+
+static int jz4740_mmc_suspend(struct device *dev)
+{
+	struct jz4740_mmc_host *host = dev_get_drvdata(dev);
+
+	jz_gpio_bulk_suspend(jz4740_mmc_pins, jz4740_mmc_num_pins(host));
+
+	return 0;
+}
+
+static int jz4740_mmc_resume(struct device *dev)
+{
+	struct jz4740_mmc_host *host = dev_get_drvdata(dev);
+
+	jz_gpio_bulk_resume(jz4740_mmc_pins, jz4740_mmc_num_pins(host));
+
+	return 0;
+}
+
+static SIMPLE_DEV_PM_OPS(jz4740_mmc_pm_ops, jz4740_mmc_suspend,
+	jz4740_mmc_resume);
+#define JZ4740_MMC_PM_OPS (&jz4740_mmc_pm_ops)
+#else
+#define JZ4740_MMC_PM_OPS NULL
+#endif
+
+static struct platform_driver jz4740_mmc_driver = {
+	.probe = jz4740_mmc_probe,
+	.remove = jz4740_mmc_remove,
+	.driver = {
+		.name = "jz4740-mmc",
+		.owner = THIS_MODULE,
+		.pm = JZ4740_MMC_PM_OPS,
+	},
+};
+
+module_platform_driver(jz4740_mmc_driver);
+
+MODULE_DESCRIPTION("JZ4740 SD/MMC controller driver");
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Lars-Peter Clausen <lars@metafoo.de>");
diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c
index f48349d18c9..cc8d4a6099c 100644
--- a/drivers/mmc/host/mmc_spi.c
+++ b/drivers/mmc/host/mmc_spi.c
@@ -26,6 +26,8 @@
  */
 #include <linux/sched.h>
 #include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/module.h>
 #include <linux/bio.h>
 #include <linux/dma-mapping.h>
 #include <linux/crc7.h>
@@ -34,6 +36,7 @@
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>		/* for R1_SPI_* bit values */
+#include <linux/mmc/slot-gpio.h>
 
 #include <linux/spi/spi.h>
 #include <linux/spi/mmc_spi.h>
@@ -97,6 +100,14 @@
  */
 #define r1b_timeout		(HZ * 3)
 
+/* One of the critical speed parameters is the amount of data which may
+ * be transferred in one command. If this value is too low, the SD card
+ * controller has to do multiple partial block writes (argggh!). With
+ * today (2008) SD cards there is little speed gain if we transfer more
+ * than 64 KBytes at a time. So use this value until there is any indication
+ * that we should do more here.
+ */
+#define MMC_SPI_BLOCKSATONCE	128
 
 /****************************************************************************/
 
@@ -173,7 +184,7 @@ mmc_spi_readbytes(struct mmc_spi_host *host, unsigned len)
 				host->data_dma, sizeof(*host->data),
 				DMA_FROM_DEVICE);
 
-	status = spi_sync(host->spi, &host->readback);
+	status = spi_sync_locked(host->spi, &host->readback);
 
 	if (host->dma_dev)
 		dma_sync_single_for_cpu(host->dma_dev,
@@ -327,15 +338,16 @@ checkstatus:
 
 	/* Status byte: the entire seven-bit R1 response.  */
 	if (cmd->resp[0] != 0) {
-		if ((R1_SPI_PARAMETER | R1_SPI_ADDRESS
-				      | R1_SPI_ILLEGAL_COMMAND)
+		if ((R1_SPI_PARAMETER | R1_SPI_ADDRESS)
 				& cmd->resp[0])
-			value = -EINVAL;
+			value = -EFAULT; /* Bad address */
+		else if (R1_SPI_ILLEGAL_COMMAND & cmd->resp[0])
+			value = -ENOSYS; /* Function not implemented */
 		else if (R1_SPI_COM_CRC & cmd->resp[0])
-			value = -EILSEQ;
+			value = -EILSEQ; /* Illegal byte sequence */
 		else if ((R1_SPI_ERASE_SEQ | R1_SPI_ERASE_RESET)
 				& cmd->resp[0])
-			value = -EIO;
+			value = -EIO;    /* I/O error */
 		/* else R1_SPI_IDLE, "it's resetting" */
 	}
 
@@ -436,7 +448,6 @@ mmc_spi_command_send(struct mmc_spi_host *host,
 {
 	struct scratch		*data = host->data;
 	u8			*cp = data->status;
-	u32			arg = cmd->arg;
 	int			status;
 	struct spi_transfer	*t;
 
@@ -453,14 +464,12 @@ mmc_spi_command_send(struct mmc_spi_host *host,
 	 * We init the whole buffer to all-ones, which is what we need
 	 * to write while we're reading (later) response data.
 	 */
-	memset(cp++, 0xff, sizeof(data->status));
+	memset(cp, 0xff, sizeof(data->status));
 
-	*cp++ = 0x40 | cmd->opcode;
-	*cp++ = (u8)(arg >> 24);
-	*cp++ = (u8)(arg >> 16);
-	*cp++ = (u8)(arg >> 8);
-	*cp++ = (u8)arg;
-	*cp++ = (crc7(0, &data->status[1], 5) << 1) | 0x01;
+	cp[1] = 0x40 | cmd->opcode;
+	put_unaligned_be32(cmd->arg, cp+2);
+	cp[6] = crc7_be(0, cp+1, 5) | 0x01;
+	cp += 7;
 
 	/* Then, read up to 13 bytes (while writing all-ones):
 	 *  - N(CR) (== 1..8) bytes of all-ones
@@ -531,7 +540,7 @@ mmc_spi_command_send(struct mmc_spi_host *host,
 				host->data_dma, sizeof(*host->data),
 				DMA_BIDIRECTIONAL);
 	}
-	status = spi_sync(host->spi, &host->m);
+	status = spi_sync_locked(host->spi, &host->m);
 
 	if (host->dma_dev)
 		dma_sync_single_for_cpu(host->dma_dev,
@@ -675,7 +684,7 @@ mmc_spi_writeblock(struct mmc_spi_host *host, struct spi_transfer *t,
 				host->data_dma, sizeof(*scratch),
 				DMA_BIDIRECTIONAL);
 
-	status = spi_sync(spi, &host->m);
+	status = spi_sync_locked(spi, &host->m);
 
 	if (status != 0) {
 		dev_dbg(&spi->dev, "write error (%d)\n", status);
@@ -699,10 +708,7 @@ mmc_spi_writeblock(struct mmc_spi_host *host, struct spi_transfer *t,
 	 * so we have to cope with this situation and check the response
 	 * bit-by-bit. Arggh!!!
 	 */
-	pattern  = scratch->status[0] << 24;
-	pattern |= scratch->status[1] << 16;
-	pattern |= scratch->status[2] << 8;
-	pattern |= scratch->status[3];
+	pattern = get_unaligned_be32(scratch->status);
 
 	/* First 3 bit of pattern are undefined */
 	pattern |= 0xE0000000;
@@ -812,7 +818,7 @@ mmc_spi_readblock(struct mmc_spi_host *host, struct spi_transfer *t,
 				DMA_FROM_DEVICE);
 	}
 
-	status = spi_sync(spi, &host->m);
+	status = spi_sync_locked(spi, &host->m);
 
 	if (host->dma_dev) {
 		dma_sync_single_for_cpu(host->dma_dev,
@@ -1008,7 +1014,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 					host->data_dma, sizeof(*scratch),
 					DMA_BIDIRECTIONAL);
 
-		tmp = spi_sync(spi, &host->m);
+		tmp = spi_sync_locked(spi, &host->m);
 
 		if (host->dma_dev)
 			dma_sync_single_for_cpu(host->dma_dev,
@@ -1045,6 +1051,8 @@ static void mmc_spi_request(struct mmc_host *mmc, struct mmc_request *mrq)
 {
 	struct mmc_spi_host	*host = mmc_priv(mmc);
 	int			status = -EINVAL;
+	int			crc_retry = 5;
+	struct mmc_command	stop;
 
 #ifdef DEBUG
 	/* MMC core and layered drivers *MUST* issue SPI-aware commands */
@@ -1074,16 +1082,41 @@ static void mmc_spi_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	}
 #endif
 
+	/* request exclusive bus access */
+	spi_bus_lock(host->spi->master);
+
+crc_recover:
 	/* issue command; then optionally data and stop */
 	status = mmc_spi_command_send(host, mrq, mrq->cmd, mrq->data != NULL);
 	if (status == 0 && mrq->data) {
 		mmc_spi_data_do(host, mrq->cmd, mrq->data, mrq->data->blksz);
+
+		/*
+		 * The SPI bus is not always reliable for large data transfers.
+		 * If an occasional crc error is reported by the SD device with
+		 * data read/write over SPI, it may be recovered by repeating
+		 * the last SD command again. The retry count is set to 5 to
+		 * ensure the driver passes stress tests.
+		 */
+		if (mrq->data->error == -EILSEQ && crc_retry) {
+			stop.opcode = MMC_STOP_TRANSMISSION;
+			stop.arg = 0;
+			stop.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
+			status = mmc_spi_command_send(host, mrq, &stop, 0);
+			crc_retry--;
+			mrq->data->error = 0;
+			goto crc_recover;
+		}
+
 		if (mrq->stop)
 			status = mmc_spi_command_send(host, mrq, mrq->stop, 0);
 		else
 			mmc_cs_off(host);
 	}
 
+	/* release the bus */
+	spi_bus_unlock(host->spi->master);
+
 	mmc_request_done(host->mmc, mrq);
 }
 
@@ -1234,33 +1267,11 @@ static void mmc_spi_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	}
 }
 
-static int mmc_spi_get_ro(struct mmc_host *mmc)
-{
-	struct mmc_spi_host *host = mmc_priv(mmc);
-
-	if (host->pdata && host->pdata->get_ro)
-		return !!host->pdata->get_ro(mmc->parent);
-	/*
-	 * Board doesn't support read only detection; let the mmc core
-	 * decide what to do.
-	 */
-	return -ENOSYS;
-}
-
-static int mmc_spi_get_cd(struct mmc_host *mmc)
-{
-	struct mmc_spi_host *host = mmc_priv(mmc);
-
-	if (host->pdata && host->pdata->get_cd)
-		return !!host->pdata->get_cd(mmc->parent);
-	return -ENOSYS;
-}
-
 static const struct mmc_host_ops mmc_spi_ops = {
 	.request	= mmc_spi_request,
 	.set_ios	= mmc_spi_set_ios,
-	.get_ro		= mmc_spi_get_ro,
-	.get_cd		= mmc_spi_get_cd,
+	.get_ro		= mmc_gpio_get_ro,
+	.get_cd		= mmc_gpio_get_cd,
 };
 
 
@@ -1280,29 +1291,19 @@ mmc_spi_detect_irq(int irq, void *mmc)
 	return IRQ_HANDLED;
 }
 
-struct count_children {
-	unsigned	n;
-	struct bus_type	*bus;
-};
-
-static int maybe_count_child(struct device *dev, void *c)
-{
-	struct count_children *ccp = c;
-
-	if (dev->bus == ccp->bus) {
-		if (ccp->n)
-			return -EBUSY;
-		ccp->n++;
-	}
-	return 0;
-}
-
 static int mmc_spi_probe(struct spi_device *spi)
 {
 	void			*ones;
 	struct mmc_host		*mmc;
 	struct mmc_spi_host	*host;
 	int			status;
+	bool			has_ro = false;
+
+	/* We rely on full duplex transfers, mostly to reduce
+	 * per-transfer overheads (by making fewer transfers).
+	 */
+	if (spi->master->flags & SPI_MASTER_HALF_DUPLEX)
+		return -EINVAL;
 
 	/* MMC and SD specs only seem to care that sampling is on the
 	 * rising edge ... meaning SPI modes 0 or 3.  So either SPI mode
@@ -1322,32 +1323,6 @@ static int mmc_spi_probe(struct spi_device *spi)
 		return status;
 	}
 
-	/* We can use the bus safely iff nobody else will interfere with us.
-	 * Most commands consist of one SPI message to issue a command, then
-	 * several more to collect its response, then possibly more for data
-	 * transfer.  Clocking access to other devices during that period will
-	 * corrupt the command execution.
-	 *
-	 * Until we have software primitives which guarantee non-interference,
-	 * we'll aim for a hardware-level guarantee.
-	 *
-	 * REVISIT we can't guarantee another device won't be added later...
-	 */
-	if (spi->master->num_chipselect > 1) {
-		struct count_children cc;
-
-		cc.n = 0;
-		cc.bus = spi->dev.bus;
-		status = device_for_each_child(spi->dev.parent, &cc,
-				maybe_count_child);
-		if (status < 0) {
-			dev_err(&spi->dev, "can't share SPI bus\n");
-			return status;
-		}
-
-		dev_warn(&spi->dev, "ASSUMING SPI bus stays unshared!\n");
-	}
-
 	/* We need a supply of ones to transmit.  This is the only time
 	 * the CPU touches these, so cache coherency isn't a concern.
 	 *
@@ -1366,6 +1341,9 @@ static int mmc_spi_probe(struct spi_device *spi)
 
 	mmc->ops = &mmc_spi_ops;
 	mmc->max_blk_size = MMC_SPI_BLOCKSIZE;
+	mmc->max_segs = MMC_SPI_BLOCKSATONCE;
+	mmc->max_req_size = MMC_SPI_BLOCKSATONCE * MMC_SPI_BLOCKSIZE;
+	mmc->max_blk_count = MMC_SPI_BLOCKSATONCE;
 
 	mmc->caps = MMC_CAP_SPI;
 
@@ -1444,18 +1422,33 @@ static int mmc_spi_probe(struct spi_device *spi)
 	}
 
 	/* pass platform capabilities, if any */
-	if (host->pdata)
+	if (host->pdata) {
 		mmc->caps |= host->pdata->caps;
+		mmc->caps2 |= host->pdata->caps2;
+	}
 
 	status = mmc_add_host(mmc);
 	if (status != 0)
 		goto fail_add_host;
 
+	if (host->pdata && host->pdata->flags & MMC_SPI_USE_CD_GPIO) {
+		status = mmc_gpio_request_cd(mmc, host->pdata->cd_gpio,
+					     host->pdata->cd_debounce);
+		if (status != 0)
+			goto fail_add_host;
+	}
+
+	if (host->pdata && host->pdata->flags & MMC_SPI_USE_RO_GPIO) {
+		has_ro = true;
+		status = mmc_gpio_request_ro(mmc, host->pdata->ro_gpio);
+		if (status != 0)
+			goto fail_add_host;
+	}
+
 	dev_info(&spi->dev, "SD/MMC host %s%s%s%s%s\n",
 			dev_name(&mmc->class_dev),
 			host->dma_dev ? "" : ", no DMA",
-			(host->pdata && host->pdata->get_ro)
-				? "" : ", no WP",
+			has_ro ? "" : ", no WP",
 			(host->pdata && host->pdata->setpower)
 				? "" : ", no poweroff",
 			(mmc->caps & MMC_CAP_NEEDS_POLL)
@@ -1481,7 +1474,7 @@ nomem:
 }
 
 
-static int __devexit mmc_spi_remove(struct spi_device *spi)
+static int mmc_spi_remove(struct spi_device *spi)
 {
 	struct mmc_host		*mmc = dev_get_drvdata(&spi->dev);
 	struct mmc_spi_host	*host;
@@ -1513,33 +1506,25 @@ static int __devexit mmc_spi_remove(struct spi_device *spi)
 	return 0;
 }
 
+static struct of_device_id mmc_spi_of_match_table[] = {
+	{ .compatible = "mmc-spi-slot", },
+	{},
+};
 
 static struct spi_driver mmc_spi_driver = {
 	.driver = {
 		.name =		"mmc_spi",
-		.bus =		&spi_bus_type,
 		.owner =	THIS_MODULE,
+		.of_match_table = mmc_spi_of_match_table,
 	},
 	.probe =	mmc_spi_probe,
-	.remove =	__devexit_p(mmc_spi_remove),
+	.remove =	mmc_spi_remove,
 };
 
-
-static int __init mmc_spi_init(void)
-{
-	return spi_register_driver(&mmc_spi_driver);
-}
-module_init(mmc_spi_init);
-
-
-static void __exit mmc_spi_exit(void)
-{
-	spi_unregister_driver(&mmc_spi_driver);
-}
-module_exit(mmc_spi_exit);
-
+module_spi_driver(mmc_spi_driver);
 
 MODULE_AUTHOR("Mike Lavender, David Brownell, "
 		"Hans-Peter Nilsson, Jan Nikitenko");
 MODULE_DESCRIPTION("SPI SD/MMC host driver");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("spi:mmc_spi");
diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c
index a663429b3d5..7ad463e9741 100644
--- a/drivers/mmc/host/mmci.c
+++ b/drivers/mmc/host/mmci.c
@@ -2,6 +2,7 @@
  *  linux/drivers/mmc/host/mmci.c - ARM PrimeCell MMCI PL180/1 driver
  *
  *  Copyright (C) 2003 Deep Blue Solutions, Ltd, All Rights Reserved.
+ *  Copyright (C) 2010 ST-Ericsson SA
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -12,31 +13,301 @@
 #include <linux/init.h>
 #include <linux/ioport.h>
 #include <linux/device.h>
+#include <linux/io.h>
 #include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
 #include <linux/delay.h>
 #include <linux/err.h>
 #include <linux/highmem.h>
 #include <linux/log2.h>
+#include <linux/mmc/pm.h>
 #include <linux/mmc/host.h>
+#include <linux/mmc/card.h>
+#include <linux/mmc/slot-gpio.h>
 #include <linux/amba/bus.h>
 #include <linux/clk.h>
 #include <linux/scatterlist.h>
+#include <linux/gpio.h>
+#include <linux/of_gpio.h>
+#include <linux/regulator/consumer.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/amba/mmci.h>
+#include <linux/pm_runtime.h>
+#include <linux/types.h>
+#include <linux/pinctrl/consumer.h>
 
-#include <asm/cacheflush.h>
 #include <asm/div64.h>
 #include <asm/io.h>
 #include <asm/sizes.h>
-#include <asm/mach/mmc.h>
 
 #include "mmci.h"
 
 #define DRIVER_NAME "mmci-pl18x"
 
-#define DBG(host,fmt,args...)	\
-	pr_debug("%s: %s: " fmt, mmc_hostname(host->mmc), __func__ , args)
-
 static unsigned int fmax = 515633;
 
+/**
+ * struct variant_data - MMCI variant-specific quirks
+ * @clkreg: default value for MCICLOCK register
+ * @clkreg_enable: enable value for MMCICLOCK register
+ * @datalength_bits: number of bits in the MMCIDATALENGTH register
+ * @fifosize: number of bytes that can be written when MMCI_TXFIFOEMPTY
+ *	      is asserted (likewise for RX)
+ * @fifohalfsize: number of bytes that can be written when MCI_TXFIFOHALFEMPTY
+ *		  is asserted (likewise for RX)
+ * @sdio: variant supports SDIO
+ * @st_clkdiv: true if using a ST-specific clock divider algorithm
+ * @blksz_datactrl16: true if Block size is at b16..b30 position in datactrl register
+ * @pwrreg_powerup: power up value for MMCIPOWER register
+ * @signal_direction: input/out direction of bus signals can be indicated
+ * @pwrreg_clkgate: MMCIPOWER register must be used to gate the clock
+ * @busy_detect: true if busy detection on dat0 is supported
+ * @pwrreg_nopower: bits in MMCIPOWER don't controls ext. power supply
+ */
+struct variant_data {
+	unsigned int		clkreg;
+	unsigned int		clkreg_enable;
+	unsigned int		datalength_bits;
+	unsigned int		fifosize;
+	unsigned int		fifohalfsize;
+	bool			sdio;
+	bool			st_clkdiv;
+	bool			blksz_datactrl16;
+	u32			pwrreg_powerup;
+	bool			signal_direction;
+	bool			pwrreg_clkgate;
+	bool			busy_detect;
+	bool			pwrreg_nopower;
+};
+
+static struct variant_data variant_arm = {
+	.fifosize		= 16 * 4,
+	.fifohalfsize		= 8 * 4,
+	.datalength_bits	= 16,
+	.pwrreg_powerup		= MCI_PWR_UP,
+};
+
+static struct variant_data variant_arm_extended_fifo = {
+	.fifosize		= 128 * 4,
+	.fifohalfsize		= 64 * 4,
+	.datalength_bits	= 16,
+	.pwrreg_powerup		= MCI_PWR_UP,
+};
+
+static struct variant_data variant_arm_extended_fifo_hwfc = {
+	.fifosize		= 128 * 4,
+	.fifohalfsize		= 64 * 4,
+	.clkreg_enable		= MCI_ARM_HWFCEN,
+	.datalength_bits	= 16,
+	.pwrreg_powerup		= MCI_PWR_UP,
+};
+
+static struct variant_data variant_u300 = {
+	.fifosize		= 16 * 4,
+	.fifohalfsize		= 8 * 4,
+	.clkreg_enable		= MCI_ST_U300_HWFCEN,
+	.datalength_bits	= 16,
+	.sdio			= true,
+	.pwrreg_powerup		= MCI_PWR_ON,
+	.signal_direction	= true,
+	.pwrreg_clkgate		= true,
+	.pwrreg_nopower		= true,
+};
+
+static struct variant_data variant_nomadik = {
+	.fifosize		= 16 * 4,
+	.fifohalfsize		= 8 * 4,
+	.clkreg			= MCI_CLK_ENABLE,
+	.datalength_bits	= 24,
+	.sdio			= true,
+	.st_clkdiv		= true,
+	.pwrreg_powerup		= MCI_PWR_ON,
+	.signal_direction	= true,
+	.pwrreg_clkgate		= true,
+	.pwrreg_nopower		= true,
+};
+
+static struct variant_data variant_ux500 = {
+	.fifosize		= 30 * 4,
+	.fifohalfsize		= 8 * 4,
+	.clkreg			= MCI_CLK_ENABLE,
+	.clkreg_enable		= MCI_ST_UX500_HWFCEN,
+	.datalength_bits	= 24,
+	.sdio			= true,
+	.st_clkdiv		= true,
+	.pwrreg_powerup		= MCI_PWR_ON,
+	.signal_direction	= true,
+	.pwrreg_clkgate		= true,
+	.busy_detect		= true,
+	.pwrreg_nopower		= true,
+};
+
+static struct variant_data variant_ux500v2 = {
+	.fifosize		= 30 * 4,
+	.fifohalfsize		= 8 * 4,
+	.clkreg			= MCI_CLK_ENABLE,
+	.clkreg_enable		= MCI_ST_UX500_HWFCEN,
+	.datalength_bits	= 24,
+	.sdio			= true,
+	.st_clkdiv		= true,
+	.blksz_datactrl16	= true,
+	.pwrreg_powerup		= MCI_PWR_ON,
+	.signal_direction	= true,
+	.pwrreg_clkgate		= true,
+	.busy_detect		= true,
+	.pwrreg_nopower		= true,
+};
+
+static int mmci_card_busy(struct mmc_host *mmc)
+{
+	struct mmci_host *host = mmc_priv(mmc);
+	unsigned long flags;
+	int busy = 0;
+
+	pm_runtime_get_sync(mmc_dev(mmc));
+
+	spin_lock_irqsave(&host->lock, flags);
+	if (readl(host->base + MMCISTATUS) & MCI_ST_CARDBUSY)
+		busy = 1;
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	pm_runtime_mark_last_busy(mmc_dev(mmc));
+	pm_runtime_put_autosuspend(mmc_dev(mmc));
+
+	return busy;
+}
+
+/*
+ * Validate mmc prerequisites
+ */
+static int mmci_validate_data(struct mmci_host *host,
+			      struct mmc_data *data)
+{
+	if (!data)
+		return 0;
+
+	if (!is_power_of_2(data->blksz)) {
+		dev_err(mmc_dev(host->mmc),
+			"unsupported block size (%d bytes)\n", data->blksz);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void mmci_reg_delay(struct mmci_host *host)
+{
+	/*
+	 * According to the spec, at least three feedback clock cycles
+	 * of max 52 MHz must pass between two writes to the MMCICLOCK reg.
+	 * Three MCLK clock cycles must pass between two MMCIPOWER reg writes.
+	 * Worst delay time during card init is at 100 kHz => 30 us.
+	 * Worst delay time when up and running is at 25 MHz => 120 ns.
+	 */
+	if (host->cclk < 25000000)
+		udelay(30);
+	else
+		ndelay(120);
+}
+
+/*
+ * This must be called with host->lock held
+ */
+static void mmci_write_clkreg(struct mmci_host *host, u32 clk)
+{
+	if (host->clk_reg != clk) {
+		host->clk_reg = clk;
+		writel(clk, host->base + MMCICLOCK);
+	}
+}
+
+/*
+ * This must be called with host->lock held
+ */
+static void mmci_write_pwrreg(struct mmci_host *host, u32 pwr)
+{
+	if (host->pwr_reg != pwr) {
+		host->pwr_reg = pwr;
+		writel(pwr, host->base + MMCIPOWER);
+	}
+}
+
+/*
+ * This must be called with host->lock held
+ */
+static void mmci_write_datactrlreg(struct mmci_host *host, u32 datactrl)
+{
+	/* Keep ST Micro busy mode if enabled */
+	datactrl |= host->datactrl_reg & MCI_ST_DPSM_BUSYMODE;
+
+	if (host->datactrl_reg != datactrl) {
+		host->datactrl_reg = datactrl;
+		writel(datactrl, host->base + MMCIDATACTRL);
+	}
+}
+
+/*
+ * This must be called with host->lock held
+ */
+static void mmci_set_clkreg(struct mmci_host *host, unsigned int desired)
+{
+	struct variant_data *variant = host->variant;
+	u32 clk = variant->clkreg;
+
+	/* Make sure cclk reflects the current calculated clock */
+	host->cclk = 0;
+
+	if (desired) {
+		if (desired >= host->mclk) {
+			clk = MCI_CLK_BYPASS;
+			if (variant->st_clkdiv)
+				clk |= MCI_ST_UX500_NEG_EDGE;
+			host->cclk = host->mclk;
+		} else if (variant->st_clkdiv) {
+			/*
+			 * DB8500 TRM says f = mclk / (clkdiv + 2)
+			 * => clkdiv = (mclk / f) - 2
+			 * Round the divider up so we don't exceed the max
+			 * frequency
+			 */
+			clk = DIV_ROUND_UP(host->mclk, desired) - 2;
+			if (clk >= 256)
+				clk = 255;
+			host->cclk = host->mclk / (clk + 2);
+		} else {
+			/*
+			 * PL180 TRM says f = mclk / (2 * (clkdiv + 1))
+			 * => clkdiv = mclk / (2 * f) - 1
+			 */
+			clk = host->mclk / (2 * desired) - 1;
+			if (clk >= 256)
+				clk = 255;
+			host->cclk = host->mclk / (2 * (clk + 1));
+		}
+
+		clk |= variant->clkreg_enable;
+		clk |= MCI_CLK_ENABLE;
+		/* This hasn't proven to be worthwhile */
+		/* clk |= MCI_CLK_PWRSAVE; */
+	}
+
+	/* Set actual clock for debug */
+	host->mmc->actual_clock = host->cclk;
+
+	if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_4)
+		clk |= MCI_4BIT_BUS;
+	if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_8)
+		clk |= MCI_ST_8BIT_BUS;
+
+	if (host->mmc->ios.timing == MMC_TIMING_UHS_DDR50 ||
+	    host->mmc->ios.timing == MMC_TIMING_MMC_DDR52)
+		clk |= MCI_ST_UX500_NEG_EDGE;
+
+	mmci_write_clkreg(host, clk);
+}
+
 static void
 mmci_request_end(struct mmci_host *host, struct mmc_request *mrq)
 {
@@ -47,40 +318,405 @@ mmci_request_end(struct mmci_host *host, struct mmc_request *mrq)
 	host->mrq = NULL;
 	host->cmd = NULL;
 
-	if (mrq->data)
-		mrq->data->bytes_xfered = host->data_xfered;
-
-	/*
-	 * Need to drop the host lock here; mmc_request_done may call
-	 * back into the driver...
-	 */
-	spin_unlock(&host->lock);
 	mmc_request_done(host->mmc, mrq);
-	spin_lock(&host->lock);
+
+	pm_runtime_mark_last_busy(mmc_dev(host->mmc));
+	pm_runtime_put_autosuspend(mmc_dev(host->mmc));
+}
+
+static void mmci_set_mask1(struct mmci_host *host, unsigned int mask)
+{
+	void __iomem *base = host->base;
+
+	if (host->singleirq) {
+		unsigned int mask0 = readl(base + MMCIMASK0);
+
+		mask0 &= ~MCI_IRQ1MASK;
+		mask0 |= mask;
+
+		writel(mask0, base + MMCIMASK0);
+	}
+
+	writel(mask, base + MMCIMASK1);
 }
 
 static void mmci_stop_data(struct mmci_host *host)
 {
-	writel(0, host->base + MMCIDATACTRL);
-	writel(0, host->base + MMCIMASK1);
+	mmci_write_datactrlreg(host, 0);
+	mmci_set_mask1(host, 0);
 	host->data = NULL;
 }
 
+static void mmci_init_sg(struct mmci_host *host, struct mmc_data *data)
+{
+	unsigned int flags = SG_MITER_ATOMIC;
+
+	if (data->flags & MMC_DATA_READ)
+		flags |= SG_MITER_TO_SG;
+	else
+		flags |= SG_MITER_FROM_SG;
+
+	sg_miter_start(&host->sg_miter, data->sg, data->sg_len, flags);
+}
+
+/*
+ * All the DMA operation mode stuff goes inside this ifdef.
+ * This assumes that you have a generic DMA device interface,
+ * no custom DMA interfaces are supported.
+ */
+#ifdef CONFIG_DMA_ENGINE
+static void mmci_dma_setup(struct mmci_host *host)
+{
+	const char *rxname, *txname;
+	dma_cap_mask_t mask;
+
+	host->dma_rx_channel = dma_request_slave_channel(mmc_dev(host->mmc), "rx");
+	host->dma_tx_channel = dma_request_slave_channel(mmc_dev(host->mmc), "tx");
+
+	/* initialize pre request cookie */
+	host->next_data.cookie = 1;
+
+	/* Try to acquire a generic DMA engine slave channel */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	/*
+	 * If only an RX channel is specified, the driver will
+	 * attempt to use it bidirectionally, however if it is
+	 * is specified but cannot be located, DMA will be disabled.
+	 */
+	if (host->dma_rx_channel && !host->dma_tx_channel)
+		host->dma_tx_channel = host->dma_rx_channel;
+
+	if (host->dma_rx_channel)
+		rxname = dma_chan_name(host->dma_rx_channel);
+	else
+		rxname = "none";
+
+	if (host->dma_tx_channel)
+		txname = dma_chan_name(host->dma_tx_channel);
+	else
+		txname = "none";
+
+	dev_info(mmc_dev(host->mmc), "DMA channels RX %s, TX %s\n",
+		 rxname, txname);
+
+	/*
+	 * Limit the maximum segment size in any SG entry according to
+	 * the parameters of the DMA engine device.
+	 */
+	if (host->dma_tx_channel) {
+		struct device *dev = host->dma_tx_channel->device->dev;
+		unsigned int max_seg_size = dma_get_max_seg_size(dev);
+
+		if (max_seg_size < host->mmc->max_seg_size)
+			host->mmc->max_seg_size = max_seg_size;
+	}
+	if (host->dma_rx_channel) {
+		struct device *dev = host->dma_rx_channel->device->dev;
+		unsigned int max_seg_size = dma_get_max_seg_size(dev);
+
+		if (max_seg_size < host->mmc->max_seg_size)
+			host->mmc->max_seg_size = max_seg_size;
+	}
+}
+
+/*
+ * This is used in or so inline it
+ * so it can be discarded.
+ */
+static inline void mmci_dma_release(struct mmci_host *host)
+{
+	if (host->dma_rx_channel)
+		dma_release_channel(host->dma_rx_channel);
+	if (host->dma_tx_channel)
+		dma_release_channel(host->dma_tx_channel);
+	host->dma_rx_channel = host->dma_tx_channel = NULL;
+}
+
+static void mmci_dma_data_error(struct mmci_host *host)
+{
+	dev_err(mmc_dev(host->mmc), "error during DMA transfer!\n");
+	dmaengine_terminate_all(host->dma_current);
+	host->dma_current = NULL;
+	host->dma_desc_current = NULL;
+	host->data->host_cookie = 0;
+}
+
+static void mmci_dma_unmap(struct mmci_host *host, struct mmc_data *data)
+{
+	struct dma_chan *chan;
+	enum dma_data_direction dir;
+
+	if (data->flags & MMC_DATA_READ) {
+		dir = DMA_FROM_DEVICE;
+		chan = host->dma_rx_channel;
+	} else {
+		dir = DMA_TO_DEVICE;
+		chan = host->dma_tx_channel;
+	}
+
+	dma_unmap_sg(chan->device->dev, data->sg, data->sg_len, dir);
+}
+
+static void mmci_dma_finalize(struct mmci_host *host, struct mmc_data *data)
+{
+	u32 status;
+	int i;
+
+	/* Wait up to 1ms for the DMA to complete */
+	for (i = 0; ; i++) {
+		status = readl(host->base + MMCISTATUS);
+		if (!(status & MCI_RXDATAAVLBLMASK) || i >= 100)
+			break;
+		udelay(10);
+	}
+
+	/*
+	 * Check to see whether we still have some data left in the FIFO -
+	 * this catches DMA controllers which are unable to monitor the
+	 * DMALBREQ and DMALSREQ signals while allowing us to DMA to non-
+	 * contiguous buffers.  On TX, we'll get a FIFO underrun error.
+	 */
+	if (status & MCI_RXDATAAVLBLMASK) {
+		mmci_dma_data_error(host);
+		if (!data->error)
+			data->error = -EIO;
+	}
+
+	if (!data->host_cookie)
+		mmci_dma_unmap(host, data);
+
+	/*
+	 * Use of DMA with scatter-gather is impossible.
+	 * Give up with DMA and switch back to PIO mode.
+	 */
+	if (status & MCI_RXDATAAVLBLMASK) {
+		dev_err(mmc_dev(host->mmc), "buggy DMA detected. Taking evasive action.\n");
+		mmci_dma_release(host);
+	}
+
+	host->dma_current = NULL;
+	host->dma_desc_current = NULL;
+}
+
+/* prepares DMA channel and DMA descriptor, returns non-zero on failure */
+static int __mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
+				struct dma_chan **dma_chan,
+				struct dma_async_tx_descriptor **dma_desc)
+{
+	struct variant_data *variant = host->variant;
+	struct dma_slave_config conf = {
+		.src_addr = host->phybase + MMCIFIFO,
+		.dst_addr = host->phybase + MMCIFIFO,
+		.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
+		.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
+		.src_maxburst = variant->fifohalfsize >> 2, /* # of words */
+		.dst_maxburst = variant->fifohalfsize >> 2, /* # of words */
+		.device_fc = false,
+	};
+	struct dma_chan *chan;
+	struct dma_device *device;
+	struct dma_async_tx_descriptor *desc;
+	enum dma_data_direction buffer_dirn;
+	int nr_sg;
+
+	if (data->flags & MMC_DATA_READ) {
+		conf.direction = DMA_DEV_TO_MEM;
+		buffer_dirn = DMA_FROM_DEVICE;
+		chan = host->dma_rx_channel;
+	} else {
+		conf.direction = DMA_MEM_TO_DEV;
+		buffer_dirn = DMA_TO_DEVICE;
+		chan = host->dma_tx_channel;
+	}
+
+	/* If there's no DMA channel, fall back to PIO */
+	if (!chan)
+		return -EINVAL;
+
+	/* If less than or equal to the fifo size, don't bother with DMA */
+	if (data->blksz * data->blocks <= variant->fifosize)
+		return -EINVAL;
+
+	device = chan->device;
+	nr_sg = dma_map_sg(device->dev, data->sg, data->sg_len, buffer_dirn);
+	if (nr_sg == 0)
+		return -EINVAL;
+
+	dmaengine_slave_config(chan, &conf);
+	desc = dmaengine_prep_slave_sg(chan, data->sg, nr_sg,
+					    conf.direction, DMA_CTRL_ACK);
+	if (!desc)
+		goto unmap_exit;
+
+	*dma_chan = chan;
+	*dma_desc = desc;
+
+	return 0;
+
+ unmap_exit:
+	dma_unmap_sg(device->dev, data->sg, data->sg_len, buffer_dirn);
+	return -ENOMEM;
+}
+
+static inline int mmci_dma_prep_data(struct mmci_host *host,
+				     struct mmc_data *data)
+{
+	/* Check if next job is already prepared. */
+	if (host->dma_current && host->dma_desc_current)
+		return 0;
+
+	/* No job were prepared thus do it now. */
+	return __mmci_dma_prep_data(host, data, &host->dma_current,
+				    &host->dma_desc_current);
+}
+
+static inline int mmci_dma_prep_next(struct mmci_host *host,
+				     struct mmc_data *data)
+{
+	struct mmci_host_next *nd = &host->next_data;
+	return __mmci_dma_prep_data(host, data, &nd->dma_chan, &nd->dma_desc);
+}
+
+static int mmci_dma_start_data(struct mmci_host *host, unsigned int datactrl)
+{
+	int ret;
+	struct mmc_data *data = host->data;
+
+	ret = mmci_dma_prep_data(host, host->data);
+	if (ret)
+		return ret;
+
+	/* Okay, go for it. */
+	dev_vdbg(mmc_dev(host->mmc),
+		 "Submit MMCI DMA job, sglen %d blksz %04x blks %04x flags %08x\n",
+		 data->sg_len, data->blksz, data->blocks, data->flags);
+	dmaengine_submit(host->dma_desc_current);
+	dma_async_issue_pending(host->dma_current);
+
+	datactrl |= MCI_DPSM_DMAENABLE;
+
+	/* Trigger the DMA transfer */
+	mmci_write_datactrlreg(host, datactrl);
+
+	/*
+	 * Let the MMCI say when the data is ended and it's time
+	 * to fire next DMA request. When that happens, MMCI will
+	 * call mmci_data_end()
+	 */
+	writel(readl(host->base + MMCIMASK0) | MCI_DATAENDMASK,
+	       host->base + MMCIMASK0);
+	return 0;
+}
+
+static void mmci_get_next_data(struct mmci_host *host, struct mmc_data *data)
+{
+	struct mmci_host_next *next = &host->next_data;
+
+	WARN_ON(data->host_cookie && data->host_cookie != next->cookie);
+	WARN_ON(!data->host_cookie && (next->dma_desc || next->dma_chan));
+
+	host->dma_desc_current = next->dma_desc;
+	host->dma_current = next->dma_chan;
+	next->dma_desc = NULL;
+	next->dma_chan = NULL;
+}
+
+static void mmci_pre_request(struct mmc_host *mmc, struct mmc_request *mrq,
+			     bool is_first_req)
+{
+	struct mmci_host *host = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
+	struct mmci_host_next *nd = &host->next_data;
+
+	if (!data)
+		return;
+
+	BUG_ON(data->host_cookie);
+
+	if (mmci_validate_data(host, data))
+		return;
+
+	if (!mmci_dma_prep_next(host, data))
+		data->host_cookie = ++nd->cookie < 0 ? 1 : nd->cookie;
+}
+
+static void mmci_post_request(struct mmc_host *mmc, struct mmc_request *mrq,
+			      int err)
+{
+	struct mmci_host *host = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
+
+	if (!data || !data->host_cookie)
+		return;
+
+	mmci_dma_unmap(host, data);
+
+	if (err) {
+		struct mmci_host_next *next = &host->next_data;
+		struct dma_chan *chan;
+		if (data->flags & MMC_DATA_READ)
+			chan = host->dma_rx_channel;
+		else
+			chan = host->dma_tx_channel;
+		dmaengine_terminate_all(chan);
+
+		next->dma_desc = NULL;
+		next->dma_chan = NULL;
+	}
+}
+
+#else
+/* Blank functions if the DMA engine is not available */
+static void mmci_get_next_data(struct mmci_host *host, struct mmc_data *data)
+{
+}
+static inline void mmci_dma_setup(struct mmci_host *host)
+{
+}
+
+static inline void mmci_dma_release(struct mmci_host *host)
+{
+}
+
+static inline void mmci_dma_unmap(struct mmci_host *host, struct mmc_data *data)
+{
+}
+
+static inline void mmci_dma_finalize(struct mmci_host *host,
+				     struct mmc_data *data)
+{
+}
+
+static inline void mmci_dma_data_error(struct mmci_host *host)
+{
+}
+
+static inline int mmci_dma_start_data(struct mmci_host *host, unsigned int datactrl)
+{
+	return -ENOSYS;
+}
+
+#define mmci_pre_request NULL
+#define mmci_post_request NULL
+
+#endif
+
 static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 {
+	struct variant_data *variant = host->variant;
 	unsigned int datactrl, timeout, irqmask;
 	unsigned long long clks;
 	void __iomem *base;
 	int blksz_bits;
 
-	DBG(host, "blksz %04x blks %04x flags %08x\n",
-	    data->blksz, data->blocks, data->flags);
+	dev_dbg(mmc_dev(host->mmc), "blksz %04x blks %04x flags %08x\n",
+		data->blksz, data->blocks, data->flags);
 
 	host->data = data;
-	host->size = data->blksz;
-	host->data_xfered = 0;
-
-	mmci_init_sg(host, data);
+	host->size = data->blksz * data->blocks;
+	data->bytes_xfered = 0;
 
 	clks = (unsigned long long)data->timeout_ns * host->cclk;
 	do_div(clks, 1000000000UL);
@@ -94,16 +730,64 @@ static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 	blksz_bits = ffs(data->blksz) - 1;
 	BUG_ON(1 << blksz_bits != data->blksz);
 
-	datactrl = MCI_DPSM_ENABLE | blksz_bits << 4;
-	if (data->flags & MMC_DATA_READ) {
+	if (variant->blksz_datactrl16)
+		datactrl = MCI_DPSM_ENABLE | (data->blksz << 16);
+	else
+		datactrl = MCI_DPSM_ENABLE | blksz_bits << 4;
+
+	if (data->flags & MMC_DATA_READ)
 		datactrl |= MCI_DPSM_DIRECTION;
+
+	/* The ST Micro variants has a special bit to enable SDIO */
+	if (variant->sdio && host->mmc->card)
+		if (mmc_card_sdio(host->mmc->card)) {
+			/*
+			 * The ST Micro variants has a special bit
+			 * to enable SDIO.
+			 */
+			u32 clk;
+
+			datactrl |= MCI_ST_DPSM_SDIOEN;
+
+			/*
+			 * The ST Micro variant for SDIO small write transfers
+			 * needs to have clock H/W flow control disabled,
+			 * otherwise the transfer will not start. The threshold
+			 * depends on the rate of MCLK.
+			 */
+			if (data->flags & MMC_DATA_WRITE &&
+			    (host->size < 8 ||
+			     (host->size <= 8 && host->mclk > 50000000)))
+				clk = host->clk_reg & ~variant->clkreg_enable;
+			else
+				clk = host->clk_reg | variant->clkreg_enable;
+
+			mmci_write_clkreg(host, clk);
+		}
+
+	if (host->mmc->ios.timing == MMC_TIMING_UHS_DDR50 ||
+	    host->mmc->ios.timing == MMC_TIMING_MMC_DDR52)
+		datactrl |= MCI_ST_DPSM_DDRMODE;
+
+	/*
+	 * Attempt to use DMA operation mode, if this
+	 * should fail, fall back to PIO mode
+	 */
+	if (!mmci_dma_start_data(host, datactrl))
+		return;
+
+	/* IRQ mode, map the SG list for CPU reading/writing */
+	mmci_init_sg(host, data);
+
+	if (data->flags & MMC_DATA_READ) {
 		irqmask = MCI_RXFIFOHALFFULLMASK;
 
 		/*
-		 * If we have less than a FIFOSIZE of bytes to transfer,
-		 * trigger a PIO interrupt as soon as any data is available.
+		 * If we have less than the fifo 'half-full' threshold to
+		 * transfer, trigger a PIO interrupt as soon as any data
+		 * is available.
 		 */
-		if (host->size < MCI_FIFOSIZE)
+		if (host->size < variant->fifohalfsize)
 			irqmask |= MCI_RXDATAAVLBLMASK;
 	} else {
 		/*
@@ -113,9 +797,9 @@ static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 		irqmask = MCI_TXFIFOHALFEMPTYMASK;
 	}
 
-	writel(datactrl, base + MMCIDATACTRL);
+	mmci_write_datactrlreg(host, datactrl);
 	writel(readl(base + MMCIMASK0) & ~MCI_DATAENDMASK, base + MMCIMASK0);
-	writel(irqmask, base + MMCIMASK1);
+	mmci_set_mask1(host, irqmask);
 }
 
 static void
@@ -123,7 +807,7 @@ mmci_start_command(struct mmci_host *host, struct mmc_command *cmd, u32 c)
 {
 	void __iomem *base = host->base;
 
-	DBG(host, "op %02x arg %08x flags %08x\n",
+	dev_dbg(mmc_dev(host->mmc), "op %02x arg %08x flags %08x\n",
 	    cmd->opcode, cmd->arg, cmd->flags);
 
 	if (readl(base + MMCICOMMAND) & MCI_CPSM_ENABLE) {
@@ -150,29 +834,62 @@ static void
 mmci_data_irq(struct mmci_host *host, struct mmc_data *data,
 	      unsigned int status)
 {
-	if (status & MCI_DATABLOCKEND) {
-		host->data_xfered += data->blksz;
-	}
-	if (status & (MCI_DATACRCFAIL|MCI_DATATIMEOUT|MCI_TXUNDERRUN|MCI_RXOVERRUN)) {
-		if (status & MCI_DATACRCFAIL)
-			data->error = -EILSEQ;
-		else if (status & MCI_DATATIMEOUT)
-			data->error = -ETIMEDOUT;
-		else if (status & (MCI_TXUNDERRUN|MCI_RXOVERRUN))
-			data->error = -EIO;
-		status |= MCI_DATAEND;
+	/* First check for errors */
+	if (status & (MCI_DATACRCFAIL|MCI_DATATIMEOUT|MCI_STARTBITERR|
+		      MCI_TXUNDERRUN|MCI_RXOVERRUN)) {
+		u32 remain, success;
+
+		/* Terminate the DMA transfer */
+		if (dma_inprogress(host)) {
+			mmci_dma_data_error(host);
+			mmci_dma_unmap(host, data);
+		}
 
 		/*
-		 * We hit an error condition.  Ensure that any data
-		 * partially written to a page is properly coherent.
+		 * Calculate how far we are into the transfer.  Note that
+		 * the data counter gives the number of bytes transferred
+		 * on the MMC bus, not on the host side.  On reads, this
+		 * can be as much as a FIFO-worth of data ahead.  This
+		 * matters for FIFO overruns only.
 		 */
-		if (host->sg_len && data->flags & MMC_DATA_READ)
-			flush_dcache_page(sg_page(host->sg_ptr));
+		remain = readl(host->base + MMCIDATACNT);
+		success = data->blksz * data->blocks - remain;
+
+		dev_dbg(mmc_dev(host->mmc), "MCI ERROR IRQ, status 0x%08x at 0x%08x\n",
+			status, success);
+		if (status & MCI_DATACRCFAIL) {
+			/* Last block was not successful */
+			success -= 1;
+			data->error = -EILSEQ;
+		} else if (status & MCI_DATATIMEOUT) {
+			data->error = -ETIMEDOUT;
+		} else if (status & MCI_STARTBITERR) {
+			data->error = -ECOMM;
+		} else if (status & MCI_TXUNDERRUN) {
+			data->error = -EIO;
+		} else if (status & MCI_RXOVERRUN) {
+			if (success > host->variant->fifosize)
+				success -= host->variant->fifosize;
+			else
+				success = 0;
+			data->error = -EIO;
+		}
+		data->bytes_xfered = round_down(success, data->blksz);
 	}
-	if (status & MCI_DATAEND) {
+
+	if (status & MCI_DATABLOCKEND)
+		dev_err(mmc_dev(host->mmc), "stray MCI_DATABLOCKEND interrupt\n");
+
+	if (status & MCI_DATAEND || data->error) {
+		if (dma_inprogress(host))
+			mmci_dma_finalize(host, data);
 		mmci_stop_data(host);
 
-		if (!data->stop) {
+		if (!data->error)
+			/* The error clause is handled above, success! */
+			data->bytes_xfered = data->blksz * data->blocks;
+
+		if (!data->stop || host->mrq->sbc) {
 			mmci_request_end(host, data->mrq);
 		} else {
 			mmci_start_command(host, data->stop, 0);
@@ -185,24 +902,56 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd,
 	     unsigned int status)
 {
 	void __iomem *base = host->base;
+	bool sbc = (cmd == host->mrq->sbc);
+	bool busy_resp = host->variant->busy_detect &&
+			(cmd->flags & MMC_RSP_BUSY);
 
-	host->cmd = NULL;
+	/* Check if we need to wait for busy completion. */
+	if (host->busy_status && (status & MCI_ST_CARDBUSY))
+		return;
+
+	/* Enable busy completion if needed and supported. */
+	if (!host->busy_status && busy_resp &&
+		!(status & (MCI_CMDCRCFAIL|MCI_CMDTIMEOUT)) &&
+		(readl(base + MMCISTATUS) & MCI_ST_CARDBUSY)) {
+		writel(readl(base + MMCIMASK0) | MCI_ST_BUSYEND,
+			base + MMCIMASK0);
+		host->busy_status = status & (MCI_CMDSENT|MCI_CMDRESPEND);
+		return;
+	}
+
+	/* At busy completion, mask the IRQ and complete the request. */
+	if (host->busy_status) {
+		writel(readl(base + MMCIMASK0) & ~MCI_ST_BUSYEND,
+			base + MMCIMASK0);
+		host->busy_status = 0;
+	}
 
-	cmd->resp[0] = readl(base + MMCIRESPONSE0);
-	cmd->resp[1] = readl(base + MMCIRESPONSE1);
-	cmd->resp[2] = readl(base + MMCIRESPONSE2);
-	cmd->resp[3] = readl(base + MMCIRESPONSE3);
+	host->cmd = NULL;
 
 	if (status & MCI_CMDTIMEOUT) {
 		cmd->error = -ETIMEDOUT;
 	} else if (status & MCI_CMDCRCFAIL && cmd->flags & MMC_RSP_CRC) {
 		cmd->error = -EILSEQ;
+	} else {
+		cmd->resp[0] = readl(base + MMCIRESPONSE0);
+		cmd->resp[1] = readl(base + MMCIRESPONSE1);
+		cmd->resp[2] = readl(base + MMCIRESPONSE2);
+		cmd->resp[3] = readl(base + MMCIRESPONSE3);
 	}
 
-	if (!cmd->data || cmd->error) {
-		if (host->data)
+	if ((!sbc && !cmd->data) || cmd->error) {
+		if (host->data) {
+			/* Terminate the DMA transfer */
+			if (dma_inprogress(host)) {
+				mmci_dma_data_error(host);
+				mmci_dma_unmap(host, host->data);
+			}
 			mmci_stop_data(host);
-		mmci_request_end(host, cmd->mrq);
+		}
+		mmci_request_end(host, host->mrq);
+	} else if (sbc) {
+		mmci_start_command(host, host->mrq->cmd, 0);
 	} else if (!(cmd->data->flags & MMC_DATA_READ)) {
 		mmci_start_data(host, cmd->data);
 	}
@@ -224,7 +973,24 @@ static int mmci_pio_read(struct mmci_host *host, char *buffer, unsigned int rema
 		if (count <= 0)
 			break;
 
-		readsl(base + MMCIFIFO, ptr, count >> 2);
+		/*
+		 * SDIO especially may want to send something that is
+		 * not divisible by 4 (as opposed to card sectors
+		 * etc). Therefore make sure to always read the last bytes
+		 * while only doing full 32-bit reads towards the FIFO.
+		 */
+		if (unlikely(count & 0x3)) {
+			if (count < 4) {
+				unsigned char buf[4];
+				ioread32_rep(base + MMCIFIFO, buf, 1);
+				memcpy(ptr, buf, count);
+			} else {
+				ioread32_rep(base + MMCIFIFO, ptr, count >> 2);
+				count &= ~0x3;
+			}
+		} else {
+			ioread32_rep(base + MMCIFIFO, ptr, count >> 2);
+		}
 
 		ptr += count;
 		remain -= count;
@@ -241,16 +1007,26 @@ static int mmci_pio_read(struct mmci_host *host, char *buffer, unsigned int rema
 
 static int mmci_pio_write(struct mmci_host *host, char *buffer, unsigned int remain, u32 status)
 {
+	struct variant_data *variant = host->variant;
 	void __iomem *base = host->base;
 	char *ptr = buffer;
 
 	do {
 		unsigned int count, maxcnt;
 
-		maxcnt = status & MCI_TXFIFOEMPTY ? MCI_FIFOSIZE : MCI_FIFOHALFSIZE;
+		maxcnt = status & MCI_TXFIFOEMPTY ?
+			 variant->fifosize : variant->fifohalfsize;
 		count = min(remain, maxcnt);
 
-		writesl(base + MMCIFIFO, ptr, count >> 2);
+		/*
+		 * SDIO especially may want to send something that is
+		 * not divisible by 4 (as opposed to card sectors
+		 * etc), and the FIFO only accept full 32-bit writes.
+		 * So compensate by adding +3 on the count, a single
+		 * byte become a 32bit write, 7 bytes will be two
+		 * 32bit writes etc.
+		 */
+		iowrite32_rep(base + MMCIFIFO, ptr, (count + 3) >> 2);
 
 		ptr += count;
 		remain -= count;
@@ -270,15 +1046,19 @@ static int mmci_pio_write(struct mmci_host *host, char *buffer, unsigned int rem
 static irqreturn_t mmci_pio_irq(int irq, void *dev_id)
 {
 	struct mmci_host *host = dev_id;
+	struct sg_mapping_iter *sg_miter = &host->sg_miter;
+	struct variant_data *variant = host->variant;
 	void __iomem *base = host->base;
+	unsigned long flags;
 	u32 status;
 
 	status = readl(base + MMCISTATUS);
 
-	DBG(host, "irq1 %08x\n", status);
+	dev_dbg(mmc_dev(host->mmc), "irq1 (pio) %08x\n", status);
+
+	local_irq_save(flags);
 
 	do {
-		unsigned long flags;
 		unsigned int remain, len;
 		char *buffer;
 
@@ -292,11 +1072,11 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id)
 		if (!(status & (MCI_TXFIFOHALFEMPTY|MCI_RXDATAAVLBL)))
 			break;
 
-		/*
-		 * Map the current scatter buffer.
-		 */
-		buffer = mmci_kmap_atomic(host, &flags) + host->sg_off;
-		remain = host->sg_ptr->length - host->sg_off;
+		if (!sg_miter_next(sg_miter))
+			break;
+
+		buffer = sg_miter->addr;
+		remain = sg_miter->length;
 
 		len = 0;
 		if (status & MCI_RXACTIVE)
@@ -304,37 +1084,27 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id)
 		if (status & MCI_TXACTIVE)
 			len = mmci_pio_write(host, buffer, remain, status);
 
-		/*
-		 * Unmap the buffer.
-		 */
-		mmci_kunmap_atomic(host, buffer, &flags);
+		sg_miter->consumed = len;
 
-		host->sg_off += len;
 		host->size -= len;
 		remain -= len;
 
 		if (remain)
 			break;
 
-		/*
-		 * If we were reading, and we have completed this
-		 * page, ensure that the data cache is coherent.
-		 */
-		if (status & MCI_RXACTIVE)
-			flush_dcache_page(sg_page(host->sg_ptr));
-
-		if (!mmci_next_sg(host))
-			break;
-
 		status = readl(base + MMCISTATUS);
 	} while (1);
 
+	sg_miter_stop(sg_miter);
+
+	local_irq_restore(flags);
+
 	/*
-	 * If we're nearing the end of the read, switch to
-	 * "any data available" mode.
+	 * If we have less than the fifo 'half-full' threshold to transfer,
+	 * trigger a PIO interrupt as soon as any data is available.
 	 */
-	if (status & MCI_RXACTIVE && host->size < MCI_FIFOSIZE)
-		writel(MCI_RXDATAAVLBLMASK, base + MMCIMASK1);
+	if (status & MCI_RXACTIVE && host->size < variant->fifohalfsize)
+		mmci_set_mask1(host, MCI_RXDATAAVLBLMASK);
 
 	/*
 	 * If we run out of data, disable the data IRQs; this
@@ -343,7 +1113,7 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id)
 	 * stops us racing with our data end IRQ.
 	 */
 	if (host->size == 0) {
-		writel(0, base + MMCIMASK1);
+		mmci_set_mask1(host, 0);
 		writel(readl(base + MMCIMASK0) | MCI_DATAENDMASK, base + MMCIMASK0);
 	}
 
@@ -366,19 +1136,38 @@ static irqreturn_t mmci_irq(int irq, void *dev_id)
 		struct mmc_data *data;
 
 		status = readl(host->base + MMCISTATUS);
+
+		if (host->singleirq) {
+			if (status & readl(host->base + MMCIMASK1))
+				mmci_pio_irq(irq, dev_id);
+
+			status &= ~MCI_IRQ1MASK;
+		}
+
+		/*
+		 * We intentionally clear the MCI_ST_CARDBUSY IRQ here (if it's
+		 * enabled) since the HW seems to be triggering the IRQ on both
+		 * edges while monitoring DAT0 for busy completion.
+		 */
 		status &= readl(host->base + MMCIMASK0);
 		writel(status, host->base + MMCICLEAR);
 
-		DBG(host, "irq0 %08x\n", status);
+		dev_dbg(mmc_dev(host->mmc), "irq0 (data+cmd) %08x\n", status);
+
+		cmd = host->cmd;
+		if ((status|host->busy_status) & (MCI_CMDCRCFAIL|MCI_CMDTIMEOUT|
+			MCI_CMDSENT|MCI_CMDRESPEND) && cmd)
+			mmci_cmd_irq(host, cmd, status);
 
 		data = host->data;
-		if (status & (MCI_DATACRCFAIL|MCI_DATATIMEOUT|MCI_TXUNDERRUN|
-			      MCI_RXOVERRUN|MCI_DATAEND|MCI_DATABLOCKEND) && data)
+		if (status & (MCI_DATACRCFAIL|MCI_DATATIMEOUT|MCI_STARTBITERR|
+			      MCI_TXUNDERRUN|MCI_RXOVERRUN|MCI_DATAEND|
+			      MCI_DATABLOCKEND) && data)
 			mmci_data_irq(host, data, status);
 
-		cmd = host->cmd;
-		if (status & (MCI_CMDCRCFAIL|MCI_CMDTIMEOUT|MCI_CMDSENT|MCI_CMDRESPEND) && cmd)
-			mmci_cmd_irq(host, cmd, status);
+		/* Don't poll for busy completion in irq context. */
+		if (host->busy_status)
+			status &= ~MCI_ST_CARDBUSY;
 
 		ret = 1;
 	} while (status);
@@ -395,22 +1184,28 @@ static void mmci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 
 	WARN_ON(host->mrq != NULL);
 
-	if (mrq->data && !is_power_of_2(mrq->data->blksz)) {
-		printk(KERN_ERR "%s: Unsupported block size (%d bytes)\n",
-			mmc_hostname(mmc), mrq->data->blksz);
-		mrq->cmd->error = -EINVAL;
+	mrq->cmd->error = mmci_validate_data(host, mrq->data);
+	if (mrq->cmd->error) {
 		mmc_request_done(mmc, mrq);
 		return;
 	}
 
+	pm_runtime_get_sync(mmc_dev(mmc));
+
 	spin_lock_irqsave(&host->lock, flags);
 
 	host->mrq = mrq;
 
+	if (mrq->data)
+		mmci_get_next_data(host, mrq->data);
+
 	if (mrq->data && mrq->data->flags & MMC_DATA_READ)
 		mmci_start_data(host, mrq->data);
 
-	mmci_start_command(host, mrq->cmd, 0);
+	if (mrq->sbc)
+		mmci_start_command(host, mrq->sbc, 0);
+	else
+		mmci_start_command(host, mrq->cmd, 0);
 
 	spin_unlock_irqrestore(&host->lock, flags);
 }
@@ -418,42 +1213,72 @@ static void mmci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 {
 	struct mmci_host *host = mmc_priv(mmc);
-	u32 clk = 0, pwr = 0;
+	struct variant_data *variant = host->variant;
+	u32 pwr = 0;
+	unsigned long flags;
+	int ret;
 
-	if (ios->clock) {
-		if (ios->clock >= host->mclk) {
-			clk = MCI_CLK_BYPASS;
-			host->cclk = host->mclk;
-		} else {
-			clk = host->mclk / (2 * ios->clock) - 1;
-			if (clk >= 256)
-				clk = 255;
-			host->cclk = host->mclk / (2 * (clk + 1));
-		}
-		if (host->hw_designer == 0x80)
-			clk |= MCI_FCEN; /* Bug fix in ST IP block */
-		clk |= MCI_CLK_ENABLE;
-	}
+	pm_runtime_get_sync(mmc_dev(mmc));
 
-	if (host->plat->translate_vdd)
-		pwr |= host->plat->translate_vdd(mmc_dev(mmc), ios->vdd);
+	if (host->plat->ios_handler &&
+		host->plat->ios_handler(mmc_dev(mmc), ios))
+			dev_err(mmc_dev(mmc), "platform ios_handler failed\n");
 
 	switch (ios->power_mode) {
 	case MMC_POWER_OFF:
+		if (!IS_ERR(mmc->supply.vmmc))
+			mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, 0);
+
+		if (!IS_ERR(mmc->supply.vqmmc) && host->vqmmc_enabled) {
+			regulator_disable(mmc->supply.vqmmc);
+			host->vqmmc_enabled = false;
+		}
+
 		break;
 	case MMC_POWER_UP:
-		/* The ST version does not have this, fall through to POWER_ON */
-		if (host->hw_designer != 0x80) {
-			pwr |= MCI_PWR_UP;
-			break;
-		}
+		if (!IS_ERR(mmc->supply.vmmc))
+			mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, ios->vdd);
+
+		/*
+		 * The ST Micro variant doesn't have the PL180s MCI_PWR_UP
+		 * and instead uses MCI_PWR_ON so apply whatever value is
+		 * configured in the variant data.
+		 */
+		pwr |= variant->pwrreg_powerup;
+
+		break;
 	case MMC_POWER_ON:
+		if (!IS_ERR(mmc->supply.vqmmc) && !host->vqmmc_enabled) {
+			ret = regulator_enable(mmc->supply.vqmmc);
+			if (ret < 0)
+				dev_err(mmc_dev(mmc),
+					"failed to enable vqmmc regulator\n");
+			else
+				host->vqmmc_enabled = true;
+		}
+
 		pwr |= MCI_PWR_ON;
 		break;
 	}
 
+	if (variant->signal_direction && ios->power_mode != MMC_POWER_OFF) {
+		/*
+		 * The ST Micro variant has some additional bits
+		 * indicating signal direction for the signals in
+		 * the SD/MMC bus and feedback-clock usage.
+		 */
+		pwr |= host->pwr_reg_add;
+
+		if (ios->bus_width == MMC_BUS_WIDTH_4)
+			pwr &= ~MCI_ST_DATA74DIREN;
+		else if (ios->bus_width == MMC_BUS_WIDTH_1)
+			pwr &= (~MCI_ST_DATA74DIREN &
+				~MCI_ST_DATA31DIREN &
+				~MCI_ST_DATA2DIREN);
+	}
+
 	if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN) {
-		if (host->hw_designer != 0x80)
+		if (host->hw_designer != AMBA_VENDOR_ST)
 			pwr |= MCI_ROD;
 		else {
 			/*
@@ -464,74 +1289,162 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		}
 	}
 
-	writel(clk, host->base + MMCICLOCK);
+	/*
+	 * If clock = 0 and the variant requires the MMCIPOWER to be used for
+	 * gating the clock, the MCI_PWR_ON bit is cleared.
+	 */
+	if (!ios->clock && variant->pwrreg_clkgate)
+		pwr &= ~MCI_PWR_ON;
 
-	if (host->pwr != pwr) {
-		host->pwr = pwr;
-		writel(pwr, host->base + MMCIPOWER);
+	spin_lock_irqsave(&host->lock, flags);
+
+	mmci_set_clkreg(host, ios->clock);
+	mmci_write_pwrreg(host, pwr);
+	mmci_reg_delay(host);
+
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	pm_runtime_mark_last_busy(mmc_dev(mmc));
+	pm_runtime_put_autosuspend(mmc_dev(mmc));
+}
+
+static int mmci_get_cd(struct mmc_host *mmc)
+{
+	struct mmci_host *host = mmc_priv(mmc);
+	struct mmci_platform_data *plat = host->plat;
+	unsigned int status = mmc_gpio_get_cd(mmc);
+
+	if (status == -ENOSYS) {
+		if (!plat->status)
+			return 1; /* Assume always present */
+
+		status = plat->status(mmc_dev(host->mmc));
 	}
+	return status;
 }
 
-static const struct mmc_host_ops mmci_ops = {
+static int mmci_sig_volt_switch(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	int ret = 0;
+
+	if (!IS_ERR(mmc->supply.vqmmc)) {
+
+		pm_runtime_get_sync(mmc_dev(mmc));
+
+		switch (ios->signal_voltage) {
+		case MMC_SIGNAL_VOLTAGE_330:
+			ret = regulator_set_voltage(mmc->supply.vqmmc,
+						2700000, 3600000);
+			break;
+		case MMC_SIGNAL_VOLTAGE_180:
+			ret = regulator_set_voltage(mmc->supply.vqmmc,
+						1700000, 1950000);
+			break;
+		case MMC_SIGNAL_VOLTAGE_120:
+			ret = regulator_set_voltage(mmc->supply.vqmmc,
+						1100000, 1300000);
+			break;
+		}
+
+		if (ret)
+			dev_warn(mmc_dev(mmc), "Voltage switch failed\n");
+
+		pm_runtime_mark_last_busy(mmc_dev(mmc));
+		pm_runtime_put_autosuspend(mmc_dev(mmc));
+	}
+
+	return ret;
+}
+
+static struct mmc_host_ops mmci_ops = {
 	.request	= mmci_request,
+	.pre_req	= mmci_pre_request,
+	.post_req	= mmci_post_request,
 	.set_ios	= mmci_set_ios,
+	.get_ro		= mmc_gpio_get_ro,
+	.get_cd		= mmci_get_cd,
+	.start_signal_voltage_switch = mmci_sig_volt_switch,
 };
 
-static void mmci_check_status(unsigned long data)
+static int mmci_of_parse(struct device_node *np, struct mmc_host *mmc)
 {
-	struct mmci_host *host = (struct mmci_host *)data;
-	unsigned int status;
+	struct mmci_host *host = mmc_priv(mmc);
+	int ret = mmc_of_parse(mmc);
 
-	status = host->plat->status(mmc_dev(host->mmc));
-	if (status ^ host->oldstat)
-		mmc_detect_change(host->mmc, 0);
+	if (ret)
+		return ret;
+
+	if (of_get_property(np, "st,sig-dir-dat0", NULL))
+		host->pwr_reg_add |= MCI_ST_DATA0DIREN;
+	if (of_get_property(np, "st,sig-dir-dat2", NULL))
+		host->pwr_reg_add |= MCI_ST_DATA2DIREN;
+	if (of_get_property(np, "st,sig-dir-dat31", NULL))
+		host->pwr_reg_add |= MCI_ST_DATA31DIREN;
+	if (of_get_property(np, "st,sig-dir-dat74", NULL))
+		host->pwr_reg_add |= MCI_ST_DATA74DIREN;
+	if (of_get_property(np, "st,sig-dir-cmd", NULL))
+		host->pwr_reg_add |= MCI_ST_CMDDIREN;
+	if (of_get_property(np, "st,sig-pin-fbclk", NULL))
+		host->pwr_reg_add |= MCI_ST_FBCLKEN;
+
+	if (of_get_property(np, "mmc-cap-mmc-highspeed", NULL))
+		mmc->caps |= MMC_CAP_MMC_HIGHSPEED;
+	if (of_get_property(np, "mmc-cap-sd-highspeed", NULL))
+		mmc->caps |= MMC_CAP_SD_HIGHSPEED;
 
-	host->oldstat = status;
-	mod_timer(&host->timer, jiffies + HZ);
+	return 0;
 }
 
-static int __devinit mmci_probe(struct amba_device *dev, void *id)
+static int mmci_probe(struct amba_device *dev,
+	const struct amba_id *id)
 {
-	struct mmc_platform_data *plat = dev->dev.platform_data;
+	struct mmci_platform_data *plat = dev->dev.platform_data;
+	struct device_node *np = dev->dev.of_node;
+	struct variant_data *variant = id->data;
 	struct mmci_host *host;
 	struct mmc_host *mmc;
 	int ret;
 
-	/* must have platform data */
-	if (!plat) {
-		ret = -EINVAL;
-		goto out;
+	/* Must have platform data or Device Tree. */
+	if (!plat && !np) {
+		dev_err(&dev->dev, "No plat data or DT found\n");
+		return -EINVAL;
 	}
 
-	ret = amba_request_regions(dev, DRIVER_NAME);
-	if (ret)
-		goto out;
+	if (!plat) {
+		plat = devm_kzalloc(&dev->dev, sizeof(*plat), GFP_KERNEL);
+		if (!plat)
+			return -ENOMEM;
+	}
 
 	mmc = mmc_alloc_host(sizeof(struct mmci_host), &dev->dev);
-	if (!mmc) {
-		ret = -ENOMEM;
-		goto rel_regions;
-	}
+	if (!mmc)
+		return -ENOMEM;
+
+	ret = mmci_of_parse(np, mmc);
+	if (ret)
+		goto host_free;
 
 	host = mmc_priv(mmc);
-	/* Bits 12 thru 19 is the designer */
-	host->hw_designer = (dev->periphid >> 12) & 0xff;
-	/* Bits 20 thru 23 is the revison */
-	host->hw_revision = (dev->periphid >> 20) & 0xf;
-	DBG(host, "designer ID = 0x%02x\n", host->hw_designer);
-	DBG(host, "revision = 0x%01x\n", host->hw_revision);
-	host->clk = clk_get(&dev->dev, NULL);
+	host->mmc = mmc;
+
+	host->hw_designer = amba_manf(dev);
+	host->hw_revision = amba_rev(dev);
+	dev_dbg(mmc_dev(mmc), "designer ID = 0x%02x\n", host->hw_designer);
+	dev_dbg(mmc_dev(mmc), "revision = 0x%01x\n", host->hw_revision);
+
+	host->clk = devm_clk_get(&dev->dev, NULL);
 	if (IS_ERR(host->clk)) {
 		ret = PTR_ERR(host->clk);
-		host->clk = NULL;
 		goto host_free;
 	}
 
-	ret = clk_enable(host->clk);
+	ret = clk_prepare_enable(host->clk);
 	if (ret)
-		goto clk_free;
+		goto host_free;
 
 	host->plat = plat;
+	host->variant = variant;
 	host->mclk = clk_get_rate(host->clk);
 	/*
 	 * According to the spec, mclk is max 100 MHz,
@@ -543,31 +1456,78 @@ static int __devinit mmci_probe(struct amba_device *dev, void *id)
 		if (ret < 0)
 			goto clk_disable;
 		host->mclk = clk_get_rate(host->clk);
-		DBG(host, "eventual mclk rate: %u Hz\n", host->mclk);
+		dev_dbg(mmc_dev(mmc), "eventual mclk rate: %u Hz\n",
+			host->mclk);
 	}
-	host->mmc = mmc;
-	host->base = ioremap(dev->res.start, SZ_4K);
-	if (!host->base) {
-		ret = -ENOMEM;
+
+	host->phybase = dev->res.start;
+	host->base = devm_ioremap_resource(&dev->dev, &dev->res);
+	if (IS_ERR(host->base)) {
+		ret = PTR_ERR(host->base);
 		goto clk_disable;
 	}
 
+	/*
+	 * The ARM and ST versions of the block have slightly different
+	 * clock divider equations which means that the minimum divider
+	 * differs too.
+	 */
+	if (variant->st_clkdiv)
+		mmc->f_min = DIV_ROUND_UP(host->mclk, 257);
+	else
+		mmc->f_min = DIV_ROUND_UP(host->mclk, 512);
+	/*
+	 * If no maximum operating frequency is supplied, fall back to use
+	 * the module parameter, which has a (low) default value in case it
+	 * is not specified. Either value must not exceed the clock rate into
+	 * the block, of course.
+	 */
+	if (mmc->f_max)
+		mmc->f_max = min(host->mclk, mmc->f_max);
+	else
+		mmc->f_max = min(host->mclk, fmax);
+	dev_dbg(mmc_dev(mmc), "clocking block at %u Hz\n", mmc->f_max);
+
+	/* Get regulators and the supported OCR mask */
+	mmc_regulator_get_supply(mmc);
+	if (!mmc->ocr_avail)
+		mmc->ocr_avail = plat->ocr_mask;
+	else if (plat->ocr_mask)
+		dev_warn(mmc_dev(mmc), "Platform OCR mask is ignored\n");
+
+	/* DT takes precedence over platform data. */
+	if (!np) {
+		if (!plat->cd_invert)
+			mmc->caps2 |= MMC_CAP2_CD_ACTIVE_HIGH;
+		mmc->caps2 |= MMC_CAP2_RO_ACTIVE_HIGH;
+	}
+
+	/* We support these capabilities. */
+	mmc->caps |= MMC_CAP_CMD23;
+
+	if (variant->busy_detect) {
+		mmci_ops.card_busy = mmci_card_busy;
+		mmci_write_datactrlreg(host, MCI_ST_DPSM_BUSYMODE);
+		mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
+		mmc->max_busy_timeout = 0;
+	}
+
 	mmc->ops = &mmci_ops;
-	mmc->f_min = (host->mclk + 511) / 512;
-	mmc->f_max = min(host->mclk, fmax);
-	mmc->ocr_avail = plat->ocr_mask;
+
+	/* We support these PM capabilities. */
+	mmc->pm_caps |= MMC_PM_KEEP_POWER;
 
 	/*
 	 * We can do SGIO
 	 */
-	mmc->max_hw_segs = 16;
-	mmc->max_phys_segs = NR_SG;
+	mmc->max_segs = NR_SG;
 
 	/*
-	 * Since we only have a 16-bit data length register, we must
-	 * ensure that we don't exceed 2^16-1 bytes in a single request.
+	 * Since only a certain number of bits are valid in the data length
+	 * register, we must ensure that we don't exceed 2^num-1 bytes in a
+	 * single request.
 	 */
-	mmc->max_req_size = 65535;
+	mmc->max_req_size = (1 << variant->datalength_bits) - 1;
 
 	/*
 	 * Set the maximum segment size.  Since we aren't doing DMA
@@ -578,12 +1538,13 @@ static int __devinit mmci_probe(struct amba_device *dev, void *id)
 	/*
 	 * Block size can be up to 2048 bytes, but must be a power of two.
 	 */
-	mmc->max_blk_size = 2048;
+	mmc->max_blk_size = 1 << 11;
 
 	/*
-	 * No limit on the number of blocks transferred.
+	 * Limit the number of blocks transferred so that we don't overflow
+	 * the maximum request size.
 	 */
-	mmc->max_blk_count = mmc->max_req_size;
+	mmc->max_blk_count = mmc->max_req_size >> 11;
 
 	spin_lock_init(&host->lock);
 
@@ -591,58 +1552,70 @@ static int __devinit mmci_probe(struct amba_device *dev, void *id)
 	writel(0, host->base + MMCIMASK1);
 	writel(0xfff, host->base + MMCICLEAR);
 
-	ret = request_irq(dev->irq[0], mmci_irq, IRQF_SHARED, DRIVER_NAME " (cmd)", host);
-	if (ret)
-		goto unmap;
+	/* If DT, cd/wp gpios must be supplied through it. */
+	if (!np && gpio_is_valid(plat->gpio_cd)) {
+		ret = mmc_gpio_request_cd(mmc, plat->gpio_cd, 0);
+		if (ret)
+			goto clk_disable;
+	}
+	if (!np && gpio_is_valid(plat->gpio_wp)) {
+		ret = mmc_gpio_request_ro(mmc, plat->gpio_wp);
+		if (ret)
+			goto clk_disable;
+	}
 
-	ret = request_irq(dev->irq[1], mmci_pio_irq, IRQF_SHARED, DRIVER_NAME " (pio)", host);
+	ret = devm_request_irq(&dev->dev, dev->irq[0], mmci_irq, IRQF_SHARED,
+			DRIVER_NAME " (cmd)", host);
 	if (ret)
-		goto irq0_free;
+		goto clk_disable;
+
+	if (!dev->irq[1])
+		host->singleirq = true;
+	else {
+		ret = devm_request_irq(&dev->dev, dev->irq[1], mmci_pio_irq,
+				IRQF_SHARED, DRIVER_NAME " (pio)", host);
+		if (ret)
+			goto clk_disable;
+	}
 
 	writel(MCI_IRQENABLE, host->base + MMCIMASK0);
 
 	amba_set_drvdata(dev, mmc);
 
-	mmc_add_host(mmc);
+	dev_info(&dev->dev, "%s: PL%03x manf %x rev%u at 0x%08llx irq %d,%d (pio)\n",
+		 mmc_hostname(mmc), amba_part(dev), amba_manf(dev),
+		 amba_rev(dev), (unsigned long long)dev->res.start,
+		 dev->irq[0], dev->irq[1]);
 
-	printk(KERN_INFO "%s: MMCI rev %x cfg %02x at 0x%016llx irq %d,%d\n",
-		mmc_hostname(mmc), amba_rev(dev), amba_config(dev),
-		(unsigned long long)dev->res.start, dev->irq[0], dev->irq[1]);
+	mmci_dma_setup(host);
 
-	init_timer(&host->timer);
-	host->timer.data = (unsigned long)host;
-	host->timer.function = mmci_check_status;
-	host->timer.expires = jiffies + HZ;
-	add_timer(&host->timer);
+	pm_runtime_set_autosuspend_delay(&dev->dev, 50);
+	pm_runtime_use_autosuspend(&dev->dev);
+	pm_runtime_put(&dev->dev);
+
+	mmc_add_host(mmc);
 
 	return 0;
 
- irq0_free:
-	free_irq(dev->irq[0], host);
- unmap:
-	iounmap(host->base);
  clk_disable:
-	clk_disable(host->clk);
- clk_free:
-	clk_put(host->clk);
+	clk_disable_unprepare(host->clk);
  host_free:
 	mmc_free_host(mmc);
- rel_regions:
-	amba_release_regions(dev);
- out:
 	return ret;
 }
 
-static int __devexit mmci_remove(struct amba_device *dev)
+static int mmci_remove(struct amba_device *dev)
 {
 	struct mmc_host *mmc = amba_get_drvdata(dev);
 
-	amba_set_drvdata(dev, NULL);
-
 	if (mmc) {
 		struct mmci_host *host = mmc_priv(mmc);
 
-		del_timer_sync(&host->timer);
+		/*
+		 * Undo pm_runtime_put() in probe.  We use the _sync
+		 * version here so that we can access the primecell.
+		 */
+		pm_runtime_get_sync(&dev->dev);
 
 		mmc_remove_host(mmc);
 
@@ -652,102 +1625,150 @@ static int __devexit mmci_remove(struct amba_device *dev)
 		writel(0, host->base + MMCICOMMAND);
 		writel(0, host->base + MMCIDATACTRL);
 
-		free_irq(dev->irq[0], host);
-		free_irq(dev->irq[1], host);
-
-		iounmap(host->base);
-		clk_disable(host->clk);
-		clk_put(host->clk);
-
+		mmci_dma_release(host);
+		clk_disable_unprepare(host->clk);
 		mmc_free_host(mmc);
-
-		amba_release_regions(dev);
 	}
 
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int mmci_suspend(struct amba_device *dev, pm_message_t state)
+static void mmci_save(struct mmci_host *host)
 {
-	struct mmc_host *mmc = amba_get_drvdata(dev);
-	int ret = 0;
+	unsigned long flags;
 
-	if (mmc) {
-		struct mmci_host *host = mmc_priv(mmc);
+	spin_lock_irqsave(&host->lock, flags);
 
-		ret = mmc_suspend_host(mmc, state);
-		if (ret == 0)
-			writel(0, host->base + MMCIMASK0);
+	writel(0, host->base + MMCIMASK0);
+	if (host->variant->pwrreg_nopower) {
+		writel(0, host->base + MMCIDATACTRL);
+		writel(0, host->base + MMCIPOWER);
+		writel(0, host->base + MMCICLOCK);
 	}
+	mmci_reg_delay(host);
 
-	return ret;
+	spin_unlock_irqrestore(&host->lock, flags);
 }
 
-static int mmci_resume(struct amba_device *dev)
+static void mmci_restore(struct mmci_host *host)
 {
-	struct mmc_host *mmc = amba_get_drvdata(dev);
-	int ret = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	if (host->variant->pwrreg_nopower) {
+		writel(host->clk_reg, host->base + MMCICLOCK);
+		writel(host->datactrl_reg, host->base + MMCIDATACTRL);
+		writel(host->pwr_reg, host->base + MMCIPOWER);
+	}
+	writel(MCI_IRQENABLE, host->base + MMCIMASK0);
+	mmci_reg_delay(host);
+
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static int mmci_runtime_suspend(struct device *dev)
+{
+	struct amba_device *adev = to_amba_device(dev);
+	struct mmc_host *mmc = amba_get_drvdata(adev);
 
 	if (mmc) {
 		struct mmci_host *host = mmc_priv(mmc);
+		pinctrl_pm_select_sleep_state(dev);
+		mmci_save(host);
+		clk_disable_unprepare(host->clk);
+	}
+
+	return 0;
+}
 
-		writel(MCI_IRQENABLE, host->base + MMCIMASK0);
+static int mmci_runtime_resume(struct device *dev)
+{
+	struct amba_device *adev = to_amba_device(dev);
+	struct mmc_host *mmc = amba_get_drvdata(adev);
 
-		ret = mmc_resume_host(mmc);
+	if (mmc) {
+		struct mmci_host *host = mmc_priv(mmc);
+		clk_prepare_enable(host->clk);
+		mmci_restore(host);
+		pinctrl_pm_select_default_state(dev);
 	}
 
-	return ret;
+	return 0;
 }
-#else
-#define mmci_suspend	NULL
-#define mmci_resume	NULL
 #endif
 
+static const struct dev_pm_ops mmci_dev_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				pm_runtime_force_resume)
+	SET_PM_RUNTIME_PM_OPS(mmci_runtime_suspend, mmci_runtime_resume, NULL)
+};
+
 static struct amba_id mmci_ids[] = {
 	{
 		.id	= 0x00041180,
-		.mask	= 0x000fffff,
+		.mask	= 0xff0fffff,
+		.data	= &variant_arm,
+	},
+	{
+		.id	= 0x01041180,
+		.mask	= 0xff0fffff,
+		.data	= &variant_arm_extended_fifo,
+	},
+	{
+		.id	= 0x02041180,
+		.mask	= 0xff0fffff,
+		.data	= &variant_arm_extended_fifo_hwfc,
 	},
 	{
 		.id	= 0x00041181,
 		.mask	= 0x000fffff,
+		.data	= &variant_arm,
 	},
 	/* ST Micro variants */
 	{
 		.id     = 0x00180180,
 		.mask   = 0x00ffffff,
+		.data	= &variant_u300,
+	},
+	{
+		.id     = 0x10180180,
+		.mask   = 0xf0ffffff,
+		.data	= &variant_nomadik,
 	},
 	{
 		.id     = 0x00280180,
 		.mask   = 0x00ffffff,
+		.data	= &variant_u300,
+	},
+	{
+		.id     = 0x00480180,
+		.mask   = 0xf0ffffff,
+		.data	= &variant_ux500,
+	},
+	{
+		.id     = 0x10480180,
+		.mask   = 0xf0ffffff,
+		.data	= &variant_ux500v2,
 	},
 	{ 0, 0 },
 };
 
+MODULE_DEVICE_TABLE(amba, mmci_ids);
+
 static struct amba_driver mmci_driver = {
 	.drv		= {
 		.name	= DRIVER_NAME,
+		.pm	= &mmci_dev_pm_ops,
 	},
 	.probe		= mmci_probe,
-	.remove		= __devexit_p(mmci_remove),
-	.suspend	= mmci_suspend,
-	.resume		= mmci_resume,
+	.remove		= mmci_remove,
 	.id_table	= mmci_ids,
 };
 
-static int __init mmci_init(void)
-{
-	return amba_driver_register(&mmci_driver);
-}
-
-static void __exit mmci_exit(void)
-{
-	amba_driver_unregister(&mmci_driver);
-}
+module_amba_driver(mmci_driver);
 
-module_init(mmci_init);
-module_exit(mmci_exit);
 module_param(fmax, uint, 0444);
 
 MODULE_DESCRIPTION("ARM PrimeCell PL180/181 Multimedia Card Interface driver");
diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h
index 0441bac1c0e..347d942d740 100644
--- a/drivers/mmc/host/mmci.h
+++ b/drivers/mmc/host/mmci.h
@@ -11,23 +11,35 @@
 #define MCI_PWR_OFF		0x00
 #define MCI_PWR_UP		0x02
 #define MCI_PWR_ON		0x03
-#define MCI_DATA2DIREN		(1 << 2)
-#define MCI_CMDDIREN		(1 << 3)
-#define MCI_DATA0DIREN		(1 << 4)
-#define MCI_DATA31DIREN		(1 << 5)
 #define MCI_OD			(1 << 6)
 #define MCI_ROD			(1 << 7)
-/* The ST Micro version does not have ROD */
-#define MCI_FBCLKEN		(1 << 7)
-#define MCI_DATA74DIREN		(1 << 8)
+/*
+ * The ST Micro version does not have ROD and reuse the voltage registers for
+ * direction settings.
+ */
+#define MCI_ST_DATA2DIREN	(1 << 2)
+#define MCI_ST_CMDDIREN		(1 << 3)
+#define MCI_ST_DATA0DIREN	(1 << 4)
+#define MCI_ST_DATA31DIREN	(1 << 5)
+#define MCI_ST_FBCLKEN		(1 << 7)
+#define MCI_ST_DATA74DIREN	(1 << 8)
 
 #define MMCICLOCK		0x004
 #define MCI_CLK_ENABLE		(1 << 8)
 #define MCI_CLK_PWRSAVE		(1 << 9)
 #define MCI_CLK_BYPASS		(1 << 10)
-#define MCI_WIDE_BUS		(1 << 11)
-/* HW flow control on the ST Micro version */
-#define MCI_FCEN		(1 << 13)
+#define MCI_4BIT_BUS		(1 << 11)
+/*
+ * 8bit wide buses, hardware flow contronl, negative edges and clock inversion
+ * supported in ST Micro U300 and Ux500 versions
+ */
+#define MCI_ST_8BIT_BUS		(1 << 12)
+#define MCI_ST_U300_HWFCEN	(1 << 13)
+#define MCI_ST_UX500_NEG_EDGE	(1 << 13)
+#define MCI_ST_UX500_HWFCEN	(1 << 14)
+#define MCI_ST_UX500_CLK_INV	(1 << 15)
+/* Modified PL180 on Versatile Express platform */
+#define MCI_ARM_HWFCEN		(1 << 12)
 
 #define MMCIARGUMENT		0x008
 #define MMCICOMMAND		0x00c
@@ -36,10 +48,11 @@
 #define MCI_CPSM_INTERRUPT	(1 << 8)
 #define MCI_CPSM_PENDING	(1 << 9)
 #define MCI_CPSM_ENABLE		(1 << 10)
-#define MCI_SDIO_SUSP		(1 << 11)
-#define MCI_ENCMD_COMPL		(1 << 12)
-#define MCI_NIEN		(1 << 13)
-#define MCI_CE_ATACMD		(1 << 14)
+/* Argument flag extenstions in the ST Micro versions */
+#define MCI_ST_SDIO_SUSP	(1 << 11)
+#define MCI_ST_ENCMD_COMPL	(1 << 12)
+#define MCI_ST_NIEN		(1 << 13)
+#define MCI_ST_CE_ATACMD	(1 << 14)
 
 #define MMCIRESPCMD		0x010
 #define MMCIRESPONSE0		0x014
@@ -54,10 +67,16 @@
 #define MCI_DPSM_MODE		(1 << 2)
 #define MCI_DPSM_DMAENABLE	(1 << 3)
 #define MCI_DPSM_BLOCKSIZE	(1 << 4)
-#define MCI_DPSM_RWSTART	(1 << 8)
-#define MCI_DPSM_RWSTOP		(1 << 9)
-#define MCI_DPSM_RWMOD		(1 << 10)
-#define MCI_DPSM_SDIOEN		(1 << 11)
+/* Control register extensions in the ST Micro U300 and Ux500 versions */
+#define MCI_ST_DPSM_RWSTART	(1 << 8)
+#define MCI_ST_DPSM_RWSTOP	(1 << 9)
+#define MCI_ST_DPSM_RWMOD	(1 << 10)
+#define MCI_ST_DPSM_SDIOEN	(1 << 11)
+/* Control register extensions in the ST Micro Ux500 versions */
+#define MCI_ST_DPSM_DMAREQCTL	(1 << 12)
+#define MCI_ST_DPSM_DBOOTMODEEN	(1 << 13)
+#define MCI_ST_DPSM_BUSYMODE	(1 << 14)
+#define MCI_ST_DPSM_DDRMODE	(1 << 15)
 
 #define MMCIDATACNT		0x030
 #define MMCISTATUS		0x034
@@ -70,6 +89,7 @@
 #define MCI_CMDRESPEND		(1 << 6)
 #define MCI_CMDSENT		(1 << 7)
 #define MCI_DATAEND		(1 << 8)
+#define MCI_STARTBITERR		(1 << 9)
 #define MCI_DATABLOCKEND	(1 << 10)
 #define MCI_CMDACTIVE		(1 << 11)
 #define MCI_TXACTIVE		(1 << 12)
@@ -82,8 +102,10 @@
 #define MCI_RXFIFOEMPTY		(1 << 19)
 #define MCI_TXDATAAVLBL		(1 << 20)
 #define MCI_RXDATAAVLBL		(1 << 21)
-#define MCI_SDIOIT		(1 << 22)
-#define MCI_CEATAEND		(1 << 23)
+/* Extended status bits for the ST Micro variants */
+#define MCI_ST_SDIOIT		(1 << 22)
+#define MCI_ST_CEATAEND		(1 << 23)
+#define MCI_ST_CARDBUSY		(1 << 24)
 
 #define MMCICLEAR		0x038
 #define MCI_CMDCRCFAILCLR	(1 << 0)
@@ -95,9 +117,12 @@
 #define MCI_CMDRESPENDCLR	(1 << 6)
 #define MCI_CMDSENTCLR		(1 << 7)
 #define MCI_DATAENDCLR		(1 << 8)
+#define MCI_STARTBITERRCLR	(1 << 9)
 #define MCI_DATABLOCKENDCLR	(1 << 10)
-#define MCI_SDIOITC		(1 << 22)
-#define MCI_CEATAENDC		(1 << 23)
+/* Extended status bits for the ST Micro variants */
+#define MCI_ST_SDIOITC		(1 << 22)
+#define MCI_ST_CEATAENDC	(1 << 23)
+#define MCI_ST_BUSYENDC		(1 << 24)
 
 #define MMCIMASK0		0x03c
 #define MCI_CMDCRCFAILMASK	(1 << 0)
@@ -109,6 +134,7 @@
 #define MCI_CMDRESPENDMASK	(1 << 6)
 #define MCI_CMDSENTMASK		(1 << 7)
 #define MCI_DATAENDMASK		(1 << 8)
+#define MCI_STARTBITERRMASK	(1 << 9)
 #define MCI_DATABLOCKENDMASK	(1 << 10)
 #define MCI_CMDACTIVEMASK	(1 << 11)
 #define MCI_TXACTIVEMASK	(1 << 12)
@@ -121,8 +147,10 @@
 #define MCI_RXFIFOEMPTYMASK	(1 << 19)
 #define MCI_TXDATAAVLBLMASK	(1 << 20)
 #define MCI_RXDATAAVLBLMASK	(1 << 21)
-#define MCI_SDIOITMASK		(1 << 22)
-#define MCI_CEATAENDMASK	(1 << 23)
+/* Extended status bits for the ST Micro variants */
+#define MCI_ST_SDIOITMASK	(1 << 22)
+#define MCI_ST_CEATAENDMASK	(1 << 23)
+#define MCI_ST_BUSYEND		(1 << 24)
 
 #define MMCIMASK1		0x040
 #define MMCIFIFOCNT		0x048
@@ -131,35 +159,47 @@
 #define MCI_IRQENABLE	\
 	(MCI_CMDCRCFAILMASK|MCI_DATACRCFAILMASK|MCI_CMDTIMEOUTMASK|	\
 	MCI_DATATIMEOUTMASK|MCI_TXUNDERRUNMASK|MCI_RXOVERRUNMASK|	\
-	MCI_CMDRESPENDMASK|MCI_CMDSENTMASK|MCI_DATABLOCKENDMASK)
+	MCI_CMDRESPENDMASK|MCI_CMDSENTMASK|MCI_STARTBITERRMASK)
 
-/*
- * The size of the FIFO in bytes.
- */
-#define MCI_FIFOSIZE	(16*4)
-	
-#define MCI_FIFOHALFSIZE (MCI_FIFOSIZE / 2)
+/* These interrupts are directed to IRQ1 when two IRQ lines are available */
+#define MCI_IRQ1MASK \
+	(MCI_RXFIFOHALFFULLMASK | MCI_RXDATAAVLBLMASK | \
+	 MCI_TXFIFOHALFEMPTYMASK)
 
-#define NR_SG		16
+#define NR_SG		128
 
 struct clk;
+struct variant_data;
+struct dma_chan;
+
+struct mmci_host_next {
+	struct dma_async_tx_descriptor	*dma_desc;
+	struct dma_chan			*dma_chan;
+	s32				cookie;
+};
 
 struct mmci_host {
+	phys_addr_t		phybase;
 	void __iomem		*base;
 	struct mmc_request	*mrq;
 	struct mmc_command	*cmd;
 	struct mmc_data		*data;
 	struct mmc_host		*mmc;
 	struct clk		*clk;
-
-	unsigned int		data_xfered;
+	bool			singleirq;
 
 	spinlock_t		lock;
 
 	unsigned int		mclk;
 	unsigned int		cclk;
-	u32			pwr;
-	struct mmc_platform_data *plat;
+	u32			pwr_reg;
+	u32			pwr_reg_add;
+	u32			clk_reg;
+	u32			datactrl_reg;
+	u32			busy_status;
+	bool			vqmmc_enabled;
+	struct mmci_platform_data *plat;
+	struct variant_data	*variant;
 
 	u8			hw_designer;
 	u8			hw_revision:4;
@@ -167,41 +207,21 @@ struct mmci_host {
 	struct timer_list	timer;
 	unsigned int		oldstat;
 
-	unsigned int		sg_len;
-
 	/* pio stuff */
-	struct scatterlist	*sg_ptr;
-	unsigned int		sg_off;
+	struct sg_mapping_iter	sg_miter;
 	unsigned int		size;
+
+#ifdef CONFIG_DMA_ENGINE
+	/* DMA stuff */
+	struct dma_chan		*dma_current;
+	struct dma_chan		*dma_rx_channel;
+	struct dma_chan		*dma_tx_channel;
+	struct dma_async_tx_descriptor	*dma_desc_current;
+	struct mmci_host_next	next_data;
+
+#define dma_inprogress(host)	((host)->dma_current)
+#else
+#define dma_inprogress(host)	(0)
+#endif
 };
 
-static inline void mmci_init_sg(struct mmci_host *host, struct mmc_data *data)
-{
-	/*
-	 * Ideally, we want the higher levels to pass us a scatter list.
-	 */
-	host->sg_len = data->sg_len;
-	host->sg_ptr = data->sg;
-	host->sg_off = 0;
-}
-
-static inline int mmci_next_sg(struct mmci_host *host)
-{
-	host->sg_ptr++;
-	host->sg_off = 0;
-	return --host->sg_len;
-}
-
-static inline char *mmci_kmap_atomic(struct mmci_host *host, unsigned long *flags)
-{
-	struct scatterlist *sg = host->sg_ptr;
-
-	local_irq_save(*flags);
-	return kmap_atomic(sg_page(sg), KM_BIO_SRC_IRQ) + sg->offset;
-}
-
-static inline void mmci_kunmap_atomic(struct mmci_host *host, void *buffer, unsigned long *flags)
-{
-	kunmap_atomic(buffer, KM_BIO_SRC_IRQ);
-	local_irq_restore(*flags);
-}
diff --git a/drivers/mmc/host/moxart-mmc.c b/drivers/mmc/host/moxart-mmc.c
new file mode 100644
index 00000000000..74924a04026
--- /dev/null
+++ b/drivers/mmc/host/moxart-mmc.c
@@ -0,0 +1,730 @@
+/*
+ * MOXA ART MMC host driver.
+ *
+ * Copyright (C) 2014 Jonas Jensen
+ *
+ * Jonas Jensen <jonas.jensen@gmail.com>
+ *
+ * Based on code from
+ * Moxa Technologies Co., Ltd. <www.moxa.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2.  This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/version.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/blkdev.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/sd.h>
+#include <linux/sched.h>
+#include <linux/io.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/clk.h>
+#include <linux/bitops.h>
+#include <linux/of_dma.h>
+#include <linux/spinlock.h>
+
+#define REG_COMMAND		0
+#define REG_ARGUMENT		4
+#define REG_RESPONSE0		8
+#define REG_RESPONSE1		12
+#define REG_RESPONSE2		16
+#define REG_RESPONSE3		20
+#define REG_RESPONSE_COMMAND	24
+#define REG_DATA_CONTROL	28
+#define REG_DATA_TIMER		32
+#define REG_DATA_LENGTH		36
+#define REG_STATUS		40
+#define REG_CLEAR		44
+#define REG_INTERRUPT_MASK	48
+#define REG_POWER_CONTROL	52
+#define REG_CLOCK_CONTROL	56
+#define REG_BUS_WIDTH		60
+#define REG_DATA_WINDOW		64
+#define REG_FEATURE		68
+#define REG_REVISION		72
+
+/* REG_COMMAND */
+#define CMD_SDC_RESET		BIT(10)
+#define CMD_EN			BIT(9)
+#define CMD_APP_CMD		BIT(8)
+#define CMD_LONG_RSP		BIT(7)
+#define CMD_NEED_RSP		BIT(6)
+#define CMD_IDX_MASK		0x3f
+
+/* REG_RESPONSE_COMMAND */
+#define RSP_CMD_APP		BIT(6)
+#define RSP_CMD_IDX_MASK	0x3f
+
+/* REG_DATA_CONTROL */
+#define DCR_DATA_FIFO_RESET     BIT(8)
+#define DCR_DATA_THRES          BIT(7)
+#define DCR_DATA_EN		BIT(6)
+#define DCR_DMA_EN		BIT(5)
+#define DCR_DATA_WRITE		BIT(4)
+#define DCR_BLK_SIZE		0x0f
+
+/* REG_DATA_LENGTH */
+#define DATA_LEN_MASK		0xffffff
+
+/* REG_STATUS */
+#define WRITE_PROT		BIT(12)
+#define CARD_DETECT		BIT(11)
+/* 1-10 below can be sent to either registers, interrupt or clear. */
+#define CARD_CHANGE		BIT(10)
+#define FIFO_ORUN		BIT(9)
+#define FIFO_URUN		BIT(8)
+#define DATA_END		BIT(7)
+#define CMD_SENT		BIT(6)
+#define DATA_CRC_OK		BIT(5)
+#define RSP_CRC_OK		BIT(4)
+#define DATA_TIMEOUT		BIT(3)
+#define RSP_TIMEOUT		BIT(2)
+#define DATA_CRC_FAIL		BIT(1)
+#define RSP_CRC_FAIL		BIT(0)
+
+#define MASK_RSP		(RSP_TIMEOUT | RSP_CRC_FAIL | \
+				 RSP_CRC_OK  | CARD_DETECT  | CMD_SENT)
+
+#define MASK_DATA		(DATA_CRC_OK   | DATA_END | \
+				 DATA_CRC_FAIL | DATA_TIMEOUT)
+
+#define MASK_INTR_PIO		(FIFO_URUN | FIFO_ORUN | CARD_CHANGE)
+
+/* REG_POWER_CONTROL */
+#define SD_POWER_ON		BIT(4)
+#define SD_POWER_MASK		0x0f
+
+/* REG_CLOCK_CONTROL */
+#define CLK_HISPD		BIT(9)
+#define CLK_OFF			BIT(8)
+#define CLK_SD			BIT(7)
+#define CLK_DIV_MASK		0x7f
+
+/* REG_BUS_WIDTH */
+#define BUS_WIDTH_8		BIT(2)
+#define BUS_WIDTH_4		BIT(1)
+#define BUS_WIDTH_1		BIT(0)
+
+#define MMC_VDD_360		23
+#define MIN_POWER		(MMC_VDD_360 - SD_POWER_MASK)
+#define MAX_RETRIES		500000
+
+struct moxart_host {
+	spinlock_t			lock;
+
+	void __iomem			*base;
+
+	phys_addr_t			reg_phys;
+
+	struct dma_chan			*dma_chan_tx;
+	struct dma_chan                 *dma_chan_rx;
+	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;
+	u32				data_len;
+	u32				fifo_width;
+	u32				timeout;
+	u32				rate;
+
+	long				sysclk;
+
+	bool				have_dma;
+	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)
+{
+	int ret = -ETIMEDOUT;
+	u32 i;
+
+	for (i = 0; i < MAX_RETRIES; i++) {
+		*status = readl(host->base + REG_STATUS);
+		if (!(*status & mask)) {
+			udelay(5);
+			continue;
+		}
+		writel(*status & mask, host->base + REG_CLEAR);
+		ret = 0;
+		break;
+	}
+
+	if (ret)
+		dev_err(mmc_dev(host->mmc), "timed out waiting for status\n");
+
+	return ret;
+}
+
+
+static void moxart_send_command(struct moxart_host *host,
+	struct mmc_command *cmd)
+{
+	u32 status, cmdctrl;
+
+	writel(RSP_TIMEOUT  | RSP_CRC_OK |
+	       RSP_CRC_FAIL | CMD_SENT, host->base + REG_CLEAR);
+	writel(cmd->arg, host->base + REG_ARGUMENT);
+
+	cmdctrl = cmd->opcode & CMD_IDX_MASK;
+	if (cmdctrl == SD_APP_SET_BUS_WIDTH    || cmdctrl == SD_APP_OP_COND   ||
+	    cmdctrl == SD_APP_SEND_SCR         || cmdctrl == SD_APP_SD_STATUS ||
+	    cmdctrl == SD_APP_SEND_NUM_WR_BLKS)
+		cmdctrl |= CMD_APP_CMD;
+
+	if (cmd->flags & MMC_RSP_PRESENT)
+		cmdctrl |= CMD_NEED_RSP;
+
+	if (cmd->flags & MMC_RSP_136)
+		cmdctrl |= CMD_LONG_RSP;
+
+	writel(cmdctrl | CMD_EN, host->base + REG_COMMAND);
+
+	if (moxart_wait_for_status(host, MASK_RSP, &status) == -ETIMEDOUT)
+		cmd->error = -ETIMEDOUT;
+
+	if (status & RSP_TIMEOUT) {
+		cmd->error = -ETIMEDOUT;
+		return;
+	}
+	if (status & RSP_CRC_FAIL) {
+		cmd->error = -EIO;
+		return;
+	}
+	if (status & RSP_CRC_OK) {
+		if (cmd->flags & MMC_RSP_136) {
+			cmd->resp[3] = readl(host->base + REG_RESPONSE0);
+			cmd->resp[2] = readl(host->base + REG_RESPONSE1);
+			cmd->resp[1] = readl(host->base + REG_RESPONSE2);
+			cmd->resp[0] = readl(host->base + REG_RESPONSE3);
+		} else {
+			cmd->resp[0] = readl(host->base + REG_RESPONSE0);
+		}
+	}
+}
+
+static void moxart_dma_complete(void *param)
+{
+	struct moxart_host *host = param;
+
+	complete(&host->dma_complete);
+}
+
+static void moxart_transfer_dma(struct mmc_data *data, struct moxart_host *host)
+{
+	u32 len, dir_data, dir_slave;
+	unsigned long dma_time;
+	struct dma_async_tx_descriptor *desc = NULL;
+	struct dma_chan *dma_chan;
+
+	if (host->data_len == data->bytes_xfered)
+		return;
+
+	if (data->flags & MMC_DATA_WRITE) {
+		dma_chan = host->dma_chan_tx;
+		dir_data = DMA_TO_DEVICE;
+		dir_slave = DMA_MEM_TO_DEV;
+	} else {
+		dma_chan = host->dma_chan_rx;
+		dir_data = DMA_FROM_DEVICE;
+		dir_slave = DMA_DEV_TO_MEM;
+	}
+
+	len = dma_map_sg(dma_chan->device->dev, data->sg,
+			 data->sg_len, dir_data);
+
+	if (len > 0) {
+		desc = dmaengine_prep_slave_sg(dma_chan, data->sg,
+					       len, dir_slave,
+					       DMA_PREP_INTERRUPT |
+					       DMA_CTRL_ACK);
+	} else {
+		dev_err(mmc_dev(host->mmc), "dma_map_sg returned zero length\n");
+	}
+
+	if (desc) {
+		host->tx_desc = desc;
+		desc->callback = moxart_dma_complete;
+		desc->callback_param = host;
+		dmaengine_submit(desc);
+		dma_async_issue_pending(dma_chan);
+	}
+
+	data->bytes_xfered += host->data_remain;
+
+	dma_time = wait_for_completion_interruptible_timeout(
+		   &host->dma_complete, host->timeout);
+
+	dma_unmap_sg(dma_chan->device->dev,
+		     data->sg, data->sg_len,
+		     dir_data);
+}
+
+
+static void moxart_transfer_pio(struct moxart_host *host)
+{
+	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;
+
+	if (data->flags & MMC_DATA_WRITE) {
+		while (remain > 0) {
+			if (moxart_wait_for_status(host, FIFO_URUN, &status)
+			     == -ETIMEDOUT) {
+				data->error = -ETIMEDOUT;
+				complete(&host->pio_complete);
+				return;
+			}
+			for (len = 0; len < remain && len < host->fifo_width;) {
+				iowrite32(*sgp, host->base + REG_DATA_WINDOW);
+				sgp++;
+				len += 4;
+			}
+			remain -= len;
+		}
+
+	} else {
+		while (remain > 0) {
+			if (moxart_wait_for_status(host, FIFO_ORUN, &status)
+			    == -ETIMEDOUT) {
+				data->error = -ETIMEDOUT;
+				complete(&host->pio_complete);
+				return;
+			}
+			for (len = 0; len < remain && len < host->fifo_width;) {
+				/* SCR data must be read in big endian. */
+				if (data->mrq->cmd->opcode == SD_APP_SEND_SCR)
+					*sgp = ioread32be(host->base +
+							  REG_DATA_WINDOW);
+				else
+					*sgp = ioread32(host->base +
+							REG_DATA_WINDOW);
+				sgp++;
+				len += 4;
+			}
+			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
+		complete(&host->pio_complete);
+}
+
+static void moxart_prepare_data(struct moxart_host *host)
+{
+	struct mmc_data *data = host->mrq->cmd->data;
+	u32 datactrl;
+	int blksz_bits;
+
+	if (!data)
+		return;
+
+	host->data_len = data->blocks * data->blksz;
+	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)
+		datactrl |= DCR_DATA_WRITE;
+
+	if ((host->data_len > host->fifo_width) && host->have_dma)
+		datactrl |= DCR_DMA_EN;
+
+	writel(DCR_DATA_FIFO_RESET, host->base + REG_DATA_CONTROL);
+	writel(MASK_DATA | FIFO_URUN | FIFO_ORUN, host->base + REG_CLEAR);
+	writel(host->rate, host->base + REG_DATA_TIMER);
+	writel(host->data_len, host->base + REG_DATA_LENGTH);
+	writel(datactrl, host->base + REG_DATA_CONTROL);
+}
+
+static void moxart_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct moxart_host *host = mmc_priv(mmc);
+	unsigned long pio_time, flags;
+	u32 status;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	init_completion(&host->dma_complete);
+	init_completion(&host->pio_complete);
+
+	host->mrq = mrq;
+
+	if (readl(host->base + REG_STATUS) & CARD_DETECT) {
+		mrq->cmd->error = -ETIMEDOUT;
+		goto request_done;
+	}
+
+	moxart_prepare_data(host);
+	moxart_send_command(host, host->mrq->cmd);
+
+	if (mrq->cmd->data) {
+		if ((host->data_len > host->fifo_width) && host->have_dma) {
+
+			writel(CARD_CHANGE, host->base + REG_INTERRUPT_MASK);
+
+			spin_unlock_irqrestore(&host->lock, flags);
+
+			moxart_transfer_dma(mrq->cmd->data, host);
+
+			spin_lock_irqsave(&host->lock, flags);
+		} else {
+
+			writel(MASK_INTR_PIO, host->base + REG_INTERRUPT_MASK);
+
+			spin_unlock_irqrestore(&host->lock, flags);
+
+			/* PIO transfers start from interrupt. */
+			pio_time = wait_for_completion_interruptible_timeout(
+				   &host->pio_complete, host->timeout);
+
+			spin_lock_irqsave(&host->lock, flags);
+		}
+
+		if (host->is_removed) {
+			dev_err(mmc_dev(host->mmc), "card removed\n");
+			mrq->cmd->error = -ETIMEDOUT;
+			goto request_done;
+		}
+
+		if (moxart_wait_for_status(host, MASK_DATA, &status)
+		    == -ETIMEDOUT) {
+			mrq->cmd->data->error = -ETIMEDOUT;
+			goto request_done;
+		}
+
+		if (status & DATA_CRC_FAIL)
+			mrq->cmd->data->error = -ETIMEDOUT;
+
+		if (mrq->cmd->data->stop)
+			moxart_send_command(host, mrq->cmd->data->stop);
+	}
+
+request_done:
+	spin_unlock_irqrestore(&host->lock, flags);
+	mmc_request_done(host->mmc, mrq);
+}
+
+static irqreturn_t moxart_irq(int irq, void *devid)
+{
+	struct moxart_host *host = (struct moxart_host *)devid;
+	u32 status;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	status = readl(host->base + REG_STATUS);
+	if (status & CARD_CHANGE) {
+		host->is_removed = status & CARD_DETECT;
+		if (host->is_removed && host->have_dma) {
+			dmaengine_terminate_all(host->dma_chan_tx);
+			dmaengine_terminate_all(host->dma_chan_rx);
+		}
+		host->mrq = NULL;
+		writel(MASK_INTR_PIO, host->base + REG_CLEAR);
+		writel(CARD_CHANGE, host->base + REG_INTERRUPT_MASK);
+		mmc_detect_change(host->mmc, 0);
+	}
+	if (status & (FIFO_ORUN | FIFO_URUN) && host->mrq)
+		moxart_transfer_pio(host);
+
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	return IRQ_HANDLED;
+}
+
+static void moxart_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct moxart_host *host = mmc_priv(mmc);
+	unsigned long flags;
+	u8 power, div;
+	u32 ctrl;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	if (ios->clock) {
+		for (div = 0; div < CLK_DIV_MASK; ++div) {
+			if (ios->clock >= host->sysclk / (2 * (div + 1)))
+				break;
+		}
+		ctrl = CLK_SD | div;
+		host->rate = host->sysclk / (2 * (div + 1));
+		if (host->rate > host->sysclk)
+			ctrl |= CLK_HISPD;
+		writel(ctrl, host->base + REG_CLOCK_CONTROL);
+	}
+
+	if (ios->power_mode == MMC_POWER_OFF) {
+		writel(readl(host->base + REG_POWER_CONTROL) & ~SD_POWER_ON,
+		       host->base + REG_POWER_CONTROL);
+	} else {
+		if (ios->vdd < MIN_POWER)
+			power = 0;
+		else
+			power = ios->vdd - MIN_POWER;
+
+		writel(SD_POWER_ON | (u32) power,
+		       host->base + REG_POWER_CONTROL);
+	}
+
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_4:
+		writel(BUS_WIDTH_4, host->base + REG_BUS_WIDTH);
+		break;
+	case MMC_BUS_WIDTH_8:
+		writel(BUS_WIDTH_8, host->base + REG_BUS_WIDTH);
+		break;
+	default:
+		writel(BUS_WIDTH_1, host->base + REG_BUS_WIDTH);
+		break;
+	}
+
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+
+static int moxart_get_ro(struct mmc_host *mmc)
+{
+	struct moxart_host *host = mmc_priv(mmc);
+
+	return !!(readl(host->base + REG_STATUS) & WRITE_PROT);
+}
+
+static struct mmc_host_ops moxart_ops = {
+	.request = moxart_request,
+	.set_ios = moxart_set_ios,
+	.get_ro = moxart_get_ro,
+};
+
+static int moxart_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *node = dev->of_node;
+	struct resource res_mmc;
+	struct mmc_host *mmc;
+	struct moxart_host *host = NULL;
+	struct dma_slave_config cfg;
+	struct clk *clk;
+	void __iomem *reg_mmc;
+	dma_cap_mask_t mask;
+	int irq, ret;
+	u32 i;
+
+	mmc = mmc_alloc_host(sizeof(struct moxart_host), dev);
+	if (!mmc) {
+		dev_err(dev, "mmc_alloc_host failed\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = of_address_to_resource(node, 0, &res_mmc);
+	if (ret) {
+		dev_err(dev, "of_address_to_resource failed\n");
+		goto out;
+	}
+
+	irq = irq_of_parse_and_map(node, 0);
+	if (irq <= 0) {
+		dev_err(dev, "irq_of_parse_and_map failed\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	clk = of_clk_get(node, 0);
+	if (IS_ERR(clk)) {
+		dev_err(dev, "of_clk_get failed\n");
+		ret = PTR_ERR(clk);
+		goto out;
+	}
+
+	reg_mmc = devm_ioremap_resource(dev, &res_mmc);
+	if (IS_ERR(reg_mmc)) {
+		ret = PTR_ERR(reg_mmc);
+		goto out;
+	}
+
+	mmc_of_parse(mmc);
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	host = mmc_priv(mmc);
+	host->mmc = mmc;
+	host->base = reg_mmc;
+	host->reg_phys = res_mmc.start;
+	host->timeout = msecs_to_jiffies(1000);
+	host->sysclk = clk_get_rate(clk);
+	host->fifo_width = readl(host->base + REG_FEATURE) << 2;
+	host->dma_chan_tx = of_dma_request_slave_channel(node, "tx");
+	host->dma_chan_rx = of_dma_request_slave_channel(node, "rx");
+
+	spin_lock_init(&host->lock);
+
+	mmc->ops = &moxart_ops;
+	mmc->f_max = DIV_ROUND_CLOSEST(host->sysclk, 2);
+	mmc->f_min = DIV_ROUND_CLOSEST(host->sysclk, CLK_DIV_MASK * 2);
+	mmc->ocr_avail = 0xffff00;	/* Support 2.0v - 3.6v power. */
+
+	if (IS_ERR(host->dma_chan_tx) || IS_ERR(host->dma_chan_rx)) {
+		dev_dbg(dev, "PIO mode transfer enabled\n");
+		host->have_dma = false;
+	} else {
+		dev_dbg(dev, "DMA channels found (%p,%p)\n",
+			 host->dma_chan_tx, host->dma_chan_rx);
+		host->have_dma = true;
+
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+
+		cfg.direction = DMA_MEM_TO_DEV;
+		cfg.src_addr = 0;
+		cfg.dst_addr = host->reg_phys + REG_DATA_WINDOW;
+		dmaengine_slave_config(host->dma_chan_tx, &cfg);
+
+		cfg.direction = DMA_DEV_TO_MEM;
+		cfg.src_addr = host->reg_phys + REG_DATA_WINDOW;
+		cfg.dst_addr = 0;
+		dmaengine_slave_config(host->dma_chan_rx, &cfg);
+	}
+
+	switch ((readl(host->base + REG_BUS_WIDTH) >> 3) & 3) {
+	case 1:
+		mmc->caps |= MMC_CAP_4_BIT_DATA;
+		break;
+	case 2:
+		mmc->caps |= MMC_CAP_4_BIT_DATA | MMC_CAP_8_BIT_DATA;
+		break;
+	default:
+		break;
+	}
+
+	writel(0, host->base + REG_INTERRUPT_MASK);
+
+	writel(CMD_SDC_RESET, host->base + REG_COMMAND);
+	for (i = 0; i < MAX_RETRIES; i++) {
+		if (!(readl(host->base + REG_COMMAND) & CMD_SDC_RESET))
+			break;
+		udelay(5);
+	}
+
+	ret = devm_request_irq(dev, irq, moxart_irq, 0, "moxart-mmc", host);
+	if (ret)
+		goto out;
+
+	dev_set_drvdata(dev, mmc);
+	mmc_add_host(mmc);
+
+	dev_dbg(dev, "IRQ=%d, FIFO is %d bytes\n", irq, host->fifo_width);
+
+	return 0;
+
+out:
+	if (mmc)
+		mmc_free_host(mmc);
+	return ret;
+}
+
+static int moxart_remove(struct platform_device *pdev)
+{
+	struct mmc_host *mmc = dev_get_drvdata(&pdev->dev);
+	struct moxart_host *host = mmc_priv(mmc);
+
+	dev_set_drvdata(&pdev->dev, NULL);
+
+	if (mmc) {
+		if (!IS_ERR(host->dma_chan_tx))
+			dma_release_channel(host->dma_chan_tx);
+		if (!IS_ERR(host->dma_chan_rx))
+			dma_release_channel(host->dma_chan_rx);
+		mmc_remove_host(mmc);
+		mmc_free_host(mmc);
+
+		writel(0, host->base + REG_INTERRUPT_MASK);
+		writel(0, host->base + REG_POWER_CONTROL);
+		writel(readl(host->base + REG_CLOCK_CONTROL) | CLK_OFF,
+		       host->base + REG_CLOCK_CONTROL);
+	}
+
+	kfree(host);
+
+	return 0;
+}
+
+static const struct of_device_id moxart_mmc_match[] = {
+	{ .compatible = "moxa,moxart-mmc" },
+	{ .compatible = "faraday,ftsdc010" },
+	{ }
+};
+
+static struct platform_driver moxart_mmc_driver = {
+	.probe      = moxart_probe,
+	.remove     = moxart_remove,
+	.driver     = {
+		.name		= "mmc-moxart",
+		.owner		= THIS_MODULE,
+		.of_match_table	= moxart_mmc_match,
+	},
+};
+module_platform_driver(moxart_mmc_driver);
+
+MODULE_ALIAS("platform:mmc-moxart");
+MODULE_DESCRIPTION("MOXA ART MMC driver");
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Jonas Jensen <jonas.jensen@gmail.com>");
diff --git a/drivers/mmc/host/msm_sdcc.c b/drivers/mmc/host/msm_sdcc.c
new file mode 100644
index 00000000000..9405ecdaf6c
--- /dev/null
+++ b/drivers/mmc/host/msm_sdcc.c
@@ -0,0 +1,1474 @@
+/*
+ *  linux/drivers/mmc/host/msm_sdcc.c - Qualcomm MSM 7X00A SDCC Driver
+ *
+ *  Copyright (C) 2007 Google Inc,
+ *  Copyright (C) 2003 Deep Blue Solutions, Ltd, All Rights Reserved.
+ *  Copyright (C) 2009, Code Aurora Forum. All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * Based on mmci.c
+ *
+ * Author: San Mehat (san@android.com)
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/ioport.h>
+#include <linux/device.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/highmem.h>
+#include <linux/log2.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/card.h>
+#include <linux/mmc/sdio.h>
+#include <linux/clk.h>
+#include <linux/scatterlist.h>
+#include <linux/platform_device.h>
+#include <linux/dma-mapping.h>
+#include <linux/debugfs.h>
+#include <linux/io.h>
+#include <linux/memory.h>
+#include <linux/gfp.h>
+#include <linux/gpio.h>
+
+#include <asm/cacheflush.h>
+#include <asm/div64.h>
+#include <asm/sizes.h>
+
+#include <linux/platform_data/mmc-msm_sdcc.h>
+#include <mach/dma.h>
+#include <mach/clk.h>
+
+#include "msm_sdcc.h"
+
+#define DRIVER_NAME "msm-sdcc"
+
+#define BUSCLK_PWRSAVE 1
+#define BUSCLK_TIMEOUT (HZ)
+static unsigned int msmsdcc_fmin = 144000;
+static unsigned int msmsdcc_fmax = 50000000;
+static unsigned int msmsdcc_4bit = 1;
+static unsigned int msmsdcc_pwrsave = 1;
+static unsigned int msmsdcc_piopoll = 1;
+static unsigned int msmsdcc_sdioirq;
+
+#define PIO_SPINMAX 30
+#define CMD_SPINMAX 20
+
+
+static inline void
+msmsdcc_disable_clocks(struct msmsdcc_host *host, int deferr)
+{
+	WARN_ON(!host->clks_on);
+
+	BUG_ON(host->curr.mrq);
+
+	if (deferr) {
+		mod_timer(&host->busclk_timer, jiffies + BUSCLK_TIMEOUT);
+	} else {
+		del_timer_sync(&host->busclk_timer);
+		/* Need to check clks_on again in case the busclk
+		 * timer fired
+		 */
+		if (host->clks_on) {
+			clk_disable(host->clk);
+			clk_disable(host->pclk);
+			host->clks_on = 0;
+		}
+	}
+}
+
+static inline int
+msmsdcc_enable_clocks(struct msmsdcc_host *host)
+{
+	int rc;
+
+	del_timer_sync(&host->busclk_timer);
+
+	if (!host->clks_on) {
+		rc = clk_enable(host->pclk);
+		if (rc)
+			return rc;
+		rc = clk_enable(host->clk);
+		if (rc) {
+			clk_disable(host->pclk);
+			return rc;
+		}
+		udelay(1 + ((3 * USEC_PER_SEC) /
+		       (host->clk_rate ? host->clk_rate : msmsdcc_fmin)));
+		host->clks_on = 1;
+	}
+	return 0;
+}
+
+static inline unsigned int
+msmsdcc_readl(struct msmsdcc_host *host, unsigned int reg)
+{
+	return readl(host->base + reg);
+}
+
+static inline void
+msmsdcc_writel(struct msmsdcc_host *host, u32 data, unsigned int reg)
+{
+	writel(data, host->base + reg);
+	/* 3 clk delay required! */
+	udelay(1 + ((3 * USEC_PER_SEC) /
+	       (host->clk_rate ? host->clk_rate : msmsdcc_fmin)));
+}
+
+static void
+msmsdcc_start_command(struct msmsdcc_host *host, struct mmc_command *cmd,
+		      u32 c);
+
+static void msmsdcc_reset_and_restore(struct msmsdcc_host *host)
+{
+	u32	mci_clk = 0;
+	u32	mci_mask0 = 0;
+	int	ret = 0;
+
+	/* Save the controller state */
+	mci_clk = readl(host->base + MMCICLOCK);
+	mci_mask0 = readl(host->base + MMCIMASK0);
+
+	/* Reset the controller */
+	ret = clk_reset(host->clk, CLK_RESET_ASSERT);
+	if (ret)
+		pr_err("%s: Clock assert failed at %u Hz with err %d\n",
+				mmc_hostname(host->mmc), host->clk_rate, ret);
+
+	ret = clk_reset(host->clk, CLK_RESET_DEASSERT);
+	if (ret)
+		pr_err("%s: Clock deassert failed at %u Hz with err %d\n",
+				mmc_hostname(host->mmc), host->clk_rate, ret);
+
+	pr_info("%s: Controller has been re-initialiazed\n",
+			mmc_hostname(host->mmc));
+
+	/* Restore the contoller state */
+	writel(host->pwr, host->base + MMCIPOWER);
+	writel(mci_clk, host->base + MMCICLOCK);
+	writel(mci_mask0, host->base + MMCIMASK0);
+	ret = clk_set_rate(host->clk, host->clk_rate);
+	if (ret)
+		pr_err("%s: Failed to set clk rate %u Hz (%d)\n",
+				mmc_hostname(host->mmc), host->clk_rate, ret);
+}
+
+static void
+msmsdcc_request_end(struct msmsdcc_host *host, struct mmc_request *mrq)
+{
+	BUG_ON(host->curr.data);
+
+	host->curr.mrq = NULL;
+	host->curr.cmd = NULL;
+
+	if (mrq->data)
+		mrq->data->bytes_xfered = host->curr.data_xfered;
+	if (mrq->cmd->error == -ETIMEDOUT)
+		mdelay(5);
+
+#if BUSCLK_PWRSAVE
+	msmsdcc_disable_clocks(host, 1);
+#endif
+	/*
+	 * Need to drop the host lock here; mmc_request_done may call
+	 * back into the driver...
+	 */
+	spin_unlock(&host->lock);
+	mmc_request_done(host->mmc, mrq);
+	spin_lock(&host->lock);
+}
+
+static void
+msmsdcc_stop_data(struct msmsdcc_host *host)
+{
+	host->curr.data = NULL;
+	host->curr.got_dataend = 0;
+}
+
+uint32_t msmsdcc_fifo_addr(struct msmsdcc_host *host)
+{
+	return host->memres->start + MMCIFIFO;
+}
+
+static inline void
+msmsdcc_start_command_exec(struct msmsdcc_host *host, u32 arg, u32 c) {
+       msmsdcc_writel(host, arg, MMCIARGUMENT);
+       msmsdcc_writel(host, c, MMCICOMMAND);
+}
+
+static void
+msmsdcc_dma_exec_func(struct msm_dmov_cmd *cmd)
+{
+	struct msmsdcc_host *host = (struct msmsdcc_host *)cmd->data;
+
+	msmsdcc_writel(host, host->cmd_timeout, MMCIDATATIMER);
+	msmsdcc_writel(host, (unsigned int)host->curr.xfer_size,
+		       MMCIDATALENGTH);
+	msmsdcc_writel(host, (msmsdcc_readl(host, MMCIMASK0) &
+			(~MCI_IRQ_PIO)) | host->cmd_pio_irqmask, MMCIMASK0);
+	msmsdcc_writel(host, host->cmd_datactrl, MMCIDATACTRL);
+
+	if (host->cmd_cmd) {
+		msmsdcc_start_command_exec(host,
+					   (u32) host->cmd_cmd->arg,
+					   (u32) host->cmd_c);
+	}
+	host->dma.active = 1;
+}
+
+static void
+msmsdcc_dma_complete_tlet(unsigned long data)
+{
+	struct msmsdcc_host *host = (struct msmsdcc_host *)data;
+	unsigned long		flags;
+	struct mmc_request	*mrq;
+	struct msm_dmov_errdata err;
+
+	spin_lock_irqsave(&host->lock, flags);
+	host->dma.active = 0;
+
+	err = host->dma.err;
+	mrq = host->curr.mrq;
+	BUG_ON(!mrq);
+	WARN_ON(!mrq->data);
+
+	if (!(host->dma.result & DMOV_RSLT_VALID)) {
+		pr_err("msmsdcc: Invalid DataMover result\n");
+		goto out;
+	}
+
+	if (host->dma.result & DMOV_RSLT_DONE) {
+		host->curr.data_xfered = host->curr.xfer_size;
+	} else {
+		/* Error or flush  */
+		if (host->dma.result & DMOV_RSLT_ERROR)
+			pr_err("%s: DMA error (0x%.8x)\n",
+			       mmc_hostname(host->mmc), host->dma.result);
+		if (host->dma.result & DMOV_RSLT_FLUSH)
+			pr_err("%s: DMA channel flushed (0x%.8x)\n",
+			       mmc_hostname(host->mmc), host->dma.result);
+
+		pr_err("Flush data: %.8x %.8x %.8x %.8x %.8x %.8x\n",
+		       err.flush[0], err.flush[1], err.flush[2],
+		       err.flush[3], err.flush[4], err.flush[5]);
+
+		msmsdcc_reset_and_restore(host);
+		if (!mrq->data->error)
+			mrq->data->error = -EIO;
+	}
+	dma_unmap_sg(mmc_dev(host->mmc), host->dma.sg, host->dma.num_ents,
+		     host->dma.dir);
+
+	host->dma.sg = NULL;
+	host->dma.busy = 0;
+
+	if (host->curr.got_dataend || mrq->data->error) {
+
+		/*
+		 * If we've already gotten our DATAEND / DATABLKEND
+		 * for this request, then complete it through here.
+		 */
+		msmsdcc_stop_data(host);
+
+		if (!mrq->data->error)
+			host->curr.data_xfered = host->curr.xfer_size;
+		if (!mrq->data->stop || mrq->cmd->error) {
+			host->curr.mrq = NULL;
+			host->curr.cmd = NULL;
+			mrq->data->bytes_xfered = host->curr.data_xfered;
+
+			spin_unlock_irqrestore(&host->lock, flags);
+#if BUSCLK_PWRSAVE
+			msmsdcc_disable_clocks(host, 1);
+#endif
+			mmc_request_done(host->mmc, mrq);
+			return;
+		} else
+			msmsdcc_start_command(host, mrq->data->stop, 0);
+	}
+
+out:
+	spin_unlock_irqrestore(&host->lock, flags);
+	return;
+}
+
+static void
+msmsdcc_dma_complete_func(struct msm_dmov_cmd *cmd,
+			  unsigned int result,
+			  struct msm_dmov_errdata *err)
+{
+	struct msmsdcc_dma_data	*dma_data =
+		container_of(cmd, struct msmsdcc_dma_data, hdr);
+	struct msmsdcc_host *host = dma_data->host;
+
+	dma_data->result = result;
+	if (err)
+		memcpy(&dma_data->err, err, sizeof(struct msm_dmov_errdata));
+
+	tasklet_schedule(&host->dma_tlet);
+}
+
+static int validate_dma(struct msmsdcc_host *host, struct mmc_data *data)
+{
+	if (host->dma.channel == -1)
+		return -ENOENT;
+
+	if ((data->blksz * data->blocks) < MCI_FIFOSIZE)
+		return -EINVAL;
+	if ((data->blksz * data->blocks) % MCI_FIFOSIZE)
+		return -EINVAL;
+	return 0;
+}
+
+static int msmsdcc_config_dma(struct msmsdcc_host *host, struct mmc_data *data)
+{
+	struct msmsdcc_nc_dmadata *nc;
+	dmov_box *box;
+	uint32_t rows;
+	uint32_t crci;
+	unsigned int n;
+	int i, rc;
+	struct scatterlist *sg = data->sg;
+
+	rc = validate_dma(host, data);
+	if (rc)
+		return rc;
+
+	host->dma.sg = data->sg;
+	host->dma.num_ents = data->sg_len;
+
+       BUG_ON(host->dma.num_ents > NR_SG); /* Prevent memory corruption */
+
+	nc = host->dma.nc;
+
+	switch (host->pdev_id) {
+	case 1:
+		crci = MSMSDCC_CRCI_SDC1;
+		break;
+	case 2:
+		crci = MSMSDCC_CRCI_SDC2;
+		break;
+	case 3:
+		crci = MSMSDCC_CRCI_SDC3;
+		break;
+	case 4:
+		crci = MSMSDCC_CRCI_SDC4;
+		break;
+	default:
+		host->dma.sg = NULL;
+		host->dma.num_ents = 0;
+		return -ENOENT;
+	}
+
+	if (data->flags & MMC_DATA_READ)
+		host->dma.dir = DMA_FROM_DEVICE;
+	else
+		host->dma.dir = DMA_TO_DEVICE;
+
+	host->curr.user_pages = 0;
+
+	box = &nc->cmd[0];
+
+	/* location of command block must be 64 bit aligned */
+	BUG_ON(host->dma.cmd_busaddr & 0x07);
+
+	nc->cmdptr = (host->dma.cmd_busaddr >> 3) | CMD_PTR_LP;
+	host->dma.hdr.cmdptr = DMOV_CMD_PTR_LIST |
+			       DMOV_CMD_ADDR(host->dma.cmdptr_busaddr);
+	host->dma.hdr.complete_func = msmsdcc_dma_complete_func;
+
+	n = dma_map_sg(mmc_dev(host->mmc), host->dma.sg,
+			host->dma.num_ents, host->dma.dir);
+	if (n == 0) {
+		pr_err("%s: Unable to map in all sg elements\n",
+			mmc_hostname(host->mmc));
+		host->dma.sg = NULL;
+		host->dma.num_ents = 0;
+		return -ENOMEM;
+	}
+
+	for_each_sg(host->dma.sg, sg, n, i) {
+
+		box->cmd = CMD_MODE_BOX;
+
+		if (i == n - 1)
+			box->cmd |= CMD_LC;
+		rows = (sg_dma_len(sg) % MCI_FIFOSIZE) ?
+			(sg_dma_len(sg) / MCI_FIFOSIZE) + 1 :
+			(sg_dma_len(sg) / MCI_FIFOSIZE) ;
+
+		if (data->flags & MMC_DATA_READ) {
+			box->src_row_addr = msmsdcc_fifo_addr(host);
+			box->dst_row_addr = sg_dma_address(sg);
+
+			box->src_dst_len = (MCI_FIFOSIZE << 16) |
+					   (MCI_FIFOSIZE);
+			box->row_offset = MCI_FIFOSIZE;
+
+			box->num_rows = rows * ((1 << 16) + 1);
+			box->cmd |= CMD_SRC_CRCI(crci);
+		} else {
+			box->src_row_addr = sg_dma_address(sg);
+			box->dst_row_addr = msmsdcc_fifo_addr(host);
+
+			box->src_dst_len = (MCI_FIFOSIZE << 16) |
+					   (MCI_FIFOSIZE);
+			box->row_offset = (MCI_FIFOSIZE << 16);
+
+			box->num_rows = rows * ((1 << 16) + 1);
+			box->cmd |= CMD_DST_CRCI(crci);
+		}
+		box++;
+	}
+
+	return 0;
+}
+
+static int
+snoop_cccr_abort(struct mmc_command *cmd)
+{
+	if ((cmd->opcode == 52) &&
+	    (cmd->arg & 0x80000000) &&
+	    (((cmd->arg >> 9) & 0x1ffff) == SDIO_CCCR_ABORT))
+		return 1;
+	return 0;
+}
+
+static void
+msmsdcc_start_command_deferred(struct msmsdcc_host *host,
+				struct mmc_command *cmd, u32 *c)
+{
+	*c |= (cmd->opcode | MCI_CPSM_ENABLE);
+
+	if (cmd->flags & MMC_RSP_PRESENT) {
+		if (cmd->flags & MMC_RSP_136)
+			*c |= MCI_CPSM_LONGRSP;
+		*c |= MCI_CPSM_RESPONSE;
+	}
+
+	if (/*interrupt*/0)
+		*c |= MCI_CPSM_INTERRUPT;
+
+	if ((((cmd->opcode == 17) || (cmd->opcode == 18))  ||
+	     ((cmd->opcode == 24) || (cmd->opcode == 25))) ||
+	      (cmd->opcode == 53))
+		*c |= MCI_CSPM_DATCMD;
+
+	if (host->prog_scan && (cmd->opcode == 12)) {
+		*c |= MCI_CPSM_PROGENA;
+		host->prog_enable = true;
+	}
+
+	if (cmd == cmd->mrq->stop)
+		*c |= MCI_CSPM_MCIABORT;
+
+	if (snoop_cccr_abort(cmd))
+		*c |= MCI_CSPM_MCIABORT;
+
+	if (host->curr.cmd != NULL) {
+		pr_err("%s: Overlapping command requests\n",
+			mmc_hostname(host->mmc));
+	}
+	host->curr.cmd = cmd;
+}
+
+static void
+msmsdcc_start_data(struct msmsdcc_host *host, struct mmc_data *data,
+			struct mmc_command *cmd, u32 c)
+{
+	unsigned int datactrl, timeout;
+	unsigned long long clks;
+	unsigned int pio_irqmask = 0;
+
+	host->curr.data = data;
+	host->curr.xfer_size = data->blksz * data->blocks;
+	host->curr.xfer_remain = host->curr.xfer_size;
+	host->curr.data_xfered = 0;
+	host->curr.got_dataend = 0;
+
+	memset(&host->pio, 0, sizeof(host->pio));
+
+	datactrl = MCI_DPSM_ENABLE | (data->blksz << 4);
+
+	if (!msmsdcc_config_dma(host, data))
+		datactrl |= MCI_DPSM_DMAENABLE;
+	else {
+		host->pio.sg = data->sg;
+		host->pio.sg_len = data->sg_len;
+		host->pio.sg_off = 0;
+
+		if (data->flags & MMC_DATA_READ) {
+			pio_irqmask = MCI_RXFIFOHALFFULLMASK;
+			if (host->curr.xfer_remain < MCI_FIFOSIZE)
+				pio_irqmask |= MCI_RXDATAAVLBLMASK;
+		} else
+			pio_irqmask = MCI_TXFIFOHALFEMPTYMASK;
+	}
+
+	if (data->flags & MMC_DATA_READ)
+		datactrl |= MCI_DPSM_DIRECTION;
+
+	clks = (unsigned long long)data->timeout_ns * host->clk_rate;
+	do_div(clks, NSEC_PER_SEC);
+	timeout = data->timeout_clks + (unsigned int)clks*2 ;
+
+	if (datactrl & MCI_DPSM_DMAENABLE) {
+		/* Save parameters for the exec function */
+		host->cmd_timeout = timeout;
+		host->cmd_pio_irqmask = pio_irqmask;
+		host->cmd_datactrl = datactrl;
+		host->cmd_cmd = cmd;
+
+		host->dma.hdr.execute_func = msmsdcc_dma_exec_func;
+		host->dma.hdr.data = (void *)host;
+		host->dma.busy = 1;
+
+		if (cmd) {
+			msmsdcc_start_command_deferred(host, cmd, &c);
+			host->cmd_c = c;
+		}
+		msm_dmov_enqueue_cmd(host->dma.channel, &host->dma.hdr);
+		if (data->flags & MMC_DATA_WRITE)
+			host->prog_scan = true;
+	} else {
+		msmsdcc_writel(host, timeout, MMCIDATATIMER);
+
+		msmsdcc_writel(host, host->curr.xfer_size, MMCIDATALENGTH);
+
+		msmsdcc_writel(host, (msmsdcc_readl(host, MMCIMASK0) &
+				(~MCI_IRQ_PIO)) | pio_irqmask, MMCIMASK0);
+
+		msmsdcc_writel(host, datactrl, MMCIDATACTRL);
+
+		if (cmd) {
+			/* Daisy-chain the command if requested */
+			msmsdcc_start_command(host, cmd, c);
+		}
+	}
+}
+
+static void
+msmsdcc_start_command(struct msmsdcc_host *host, struct mmc_command *cmd, u32 c)
+{
+	if (cmd == cmd->mrq->stop)
+		c |= MCI_CSPM_MCIABORT;
+
+	host->stats.cmds++;
+
+	msmsdcc_start_command_deferred(host, cmd, &c);
+	msmsdcc_start_command_exec(host, cmd->arg, c);
+}
+
+static void
+msmsdcc_data_err(struct msmsdcc_host *host, struct mmc_data *data,
+		 unsigned int status)
+{
+	if (status & MCI_DATACRCFAIL) {
+		pr_err("%s: Data CRC error\n", mmc_hostname(host->mmc));
+		pr_err("%s: opcode 0x%.8x\n", __func__,
+		       data->mrq->cmd->opcode);
+		pr_err("%s: blksz %d, blocks %d\n", __func__,
+		       data->blksz, data->blocks);
+		data->error = -EILSEQ;
+	} else if (status & MCI_DATATIMEOUT) {
+		pr_err("%s: Data timeout\n", mmc_hostname(host->mmc));
+		data->error = -ETIMEDOUT;
+	} else if (status & MCI_RXOVERRUN) {
+		pr_err("%s: RX overrun\n", mmc_hostname(host->mmc));
+		data->error = -EIO;
+	} else if (status & MCI_TXUNDERRUN) {
+		pr_err("%s: TX underrun\n", mmc_hostname(host->mmc));
+		data->error = -EIO;
+	} else {
+		pr_err("%s: Unknown error (0x%.8x)\n",
+		       mmc_hostname(host->mmc), status);
+		data->error = -EIO;
+	}
+}
+
+
+static int
+msmsdcc_pio_read(struct msmsdcc_host *host, char *buffer, unsigned int remain)
+{
+	uint32_t	*ptr = (uint32_t *) buffer;
+	int		count = 0;
+
+	if (remain % 4)
+		remain = ((remain >> 2) + 1) << 2;
+
+	while (msmsdcc_readl(host, MMCISTATUS) & MCI_RXDATAAVLBL) {
+		*ptr = msmsdcc_readl(host, MMCIFIFO + (count % MCI_FIFOSIZE));
+		ptr++;
+		count += sizeof(uint32_t);
+
+		remain -=  sizeof(uint32_t);
+		if (remain == 0)
+			break;
+	}
+	return count;
+}
+
+static int
+msmsdcc_pio_write(struct msmsdcc_host *host, char *buffer,
+		  unsigned int remain, u32 status)
+{
+	void __iomem *base = host->base;
+	char *ptr = buffer;
+
+	do {
+		unsigned int count, maxcnt, sz;
+
+		maxcnt = status & MCI_TXFIFOEMPTY ? MCI_FIFOSIZE :
+						    MCI_FIFOHALFSIZE;
+		count = min(remain, maxcnt);
+
+		sz = count % 4 ? (count >> 2) + 1 : (count >> 2);
+		writesl(base + MMCIFIFO, ptr, sz);
+		ptr += count;
+		remain -= count;
+
+		if (remain == 0)
+			break;
+
+		status = msmsdcc_readl(host, MMCISTATUS);
+	} while (status & MCI_TXFIFOHALFEMPTY);
+
+	return ptr - buffer;
+}
+
+static int
+msmsdcc_spin_on_status(struct msmsdcc_host *host, uint32_t mask, int maxspin)
+{
+	while (maxspin) {
+		if ((msmsdcc_readl(host, MMCISTATUS) & mask))
+			return 0;
+		udelay(1);
+		--maxspin;
+	}
+	return -ETIMEDOUT;
+}
+
+static irqreturn_t
+msmsdcc_pio_irq(int irq, void *dev_id)
+{
+	struct msmsdcc_host	*host = dev_id;
+	uint32_t		status;
+	u32 mci_mask0;
+
+	status = msmsdcc_readl(host, MMCISTATUS);
+	mci_mask0 = msmsdcc_readl(host, MMCIMASK0);
+
+	if (((mci_mask0 & status) & MCI_IRQ_PIO) == 0)
+		return IRQ_NONE;
+
+	do {
+		unsigned long flags;
+		unsigned int remain, len;
+		char *buffer;
+
+		if (!(status & (MCI_TXFIFOHALFEMPTY | MCI_RXDATAAVLBL))) {
+			if (host->curr.xfer_remain == 0 || !msmsdcc_piopoll)
+				break;
+
+			if (msmsdcc_spin_on_status(host,
+						   (MCI_TXFIFOHALFEMPTY |
+						   MCI_RXDATAAVLBL),
+						   PIO_SPINMAX)) {
+				break;
+			}
+		}
+
+		/* Map the current scatter buffer */
+		local_irq_save(flags);
+		buffer = kmap_atomic(sg_page(host->pio.sg))
+				     + host->pio.sg->offset;
+		buffer += host->pio.sg_off;
+		remain = host->pio.sg->length - host->pio.sg_off;
+		len = 0;
+		if (status & MCI_RXACTIVE)
+			len = msmsdcc_pio_read(host, buffer, remain);
+		if (status & MCI_TXACTIVE)
+			len = msmsdcc_pio_write(host, buffer, remain, status);
+
+		/* Unmap the buffer */
+		kunmap_atomic(buffer);
+		local_irq_restore(flags);
+
+		host->pio.sg_off += len;
+		host->curr.xfer_remain -= len;
+		host->curr.data_xfered += len;
+		remain -= len;
+
+		if (remain == 0) {
+			/* This sg page is full - do some housekeeping */
+			if (status & MCI_RXACTIVE && host->curr.user_pages)
+				flush_dcache_page(sg_page(host->pio.sg));
+
+			if (!--host->pio.sg_len) {
+				memset(&host->pio, 0, sizeof(host->pio));
+				break;
+			}
+
+			/* Advance to next sg */
+			host->pio.sg++;
+			host->pio.sg_off = 0;
+		}
+
+		status = msmsdcc_readl(host, MMCISTATUS);
+	} while (1);
+
+	if (status & MCI_RXACTIVE && host->curr.xfer_remain < MCI_FIFOSIZE)
+		msmsdcc_writel(host, (mci_mask0 & (~MCI_IRQ_PIO)) |
+					MCI_RXDATAAVLBLMASK, MMCIMASK0);
+
+	if (!host->curr.xfer_remain)
+		msmsdcc_writel(host, (mci_mask0 & (~MCI_IRQ_PIO)) | 0,
+					MMCIMASK0);
+
+	return IRQ_HANDLED;
+}
+
+static void msmsdcc_do_cmdirq(struct msmsdcc_host *host, uint32_t status)
+{
+	struct mmc_command *cmd = host->curr.cmd;
+
+	host->curr.cmd = NULL;
+	cmd->resp[0] = msmsdcc_readl(host, MMCIRESPONSE0);
+	cmd->resp[1] = msmsdcc_readl(host, MMCIRESPONSE1);
+	cmd->resp[2] = msmsdcc_readl(host, MMCIRESPONSE2);
+	cmd->resp[3] = msmsdcc_readl(host, MMCIRESPONSE3);
+
+	if (status & MCI_CMDTIMEOUT) {
+		cmd->error = -ETIMEDOUT;
+	} else if (status & MCI_CMDCRCFAIL &&
+		   cmd->flags & MMC_RSP_CRC) {
+		pr_err("%s: Command CRC error\n", mmc_hostname(host->mmc));
+		cmd->error = -EILSEQ;
+	}
+
+	if (!cmd->data || cmd->error) {
+		if (host->curr.data && host->dma.sg)
+			msm_dmov_stop_cmd(host->dma.channel,
+					  &host->dma.hdr, 0);
+		else if (host->curr.data) { /* Non DMA */
+			msmsdcc_reset_and_restore(host);
+			msmsdcc_stop_data(host);
+			msmsdcc_request_end(host, cmd->mrq);
+		} else { /* host->data == NULL */
+			if (!cmd->error && host->prog_enable) {
+				if (status & MCI_PROGDONE) {
+					host->prog_scan = false;
+					host->prog_enable = false;
+					msmsdcc_request_end(host, cmd->mrq);
+				} else {
+					host->curr.cmd = cmd;
+				}
+			} else {
+				if (host->prog_enable) {
+					host->prog_scan = false;
+					host->prog_enable = false;
+				}
+				msmsdcc_request_end(host, cmd->mrq);
+			}
+		}
+	} else if (cmd->data)
+		if (!(cmd->data->flags & MMC_DATA_READ))
+			msmsdcc_start_data(host, cmd->data,
+						NULL, 0);
+}
+
+static void
+msmsdcc_handle_irq_data(struct msmsdcc_host *host, u32 status,
+			void __iomem *base)
+{
+	struct mmc_data *data = host->curr.data;
+
+	if (status & (MCI_CMDSENT | MCI_CMDRESPEND | MCI_CMDCRCFAIL |
+			MCI_CMDTIMEOUT | MCI_PROGDONE) && host->curr.cmd) {
+		msmsdcc_do_cmdirq(host, status);
+	}
+
+	if (!data)
+		return;
+
+	/* Check for data errors */
+	if (status & (MCI_DATACRCFAIL | MCI_DATATIMEOUT |
+		      MCI_TXUNDERRUN | MCI_RXOVERRUN)) {
+		msmsdcc_data_err(host, data, status);
+		host->curr.data_xfered = 0;
+		if (host->dma.sg)
+			msm_dmov_stop_cmd(host->dma.channel,
+					  &host->dma.hdr, 0);
+		else {
+			msmsdcc_reset_and_restore(host);
+			if (host->curr.data)
+				msmsdcc_stop_data(host);
+			if (!data->stop)
+				msmsdcc_request_end(host, data->mrq);
+			else
+				msmsdcc_start_command(host, data->stop, 0);
+		}
+	}
+
+	/* Check for data done */
+	if (!host->curr.got_dataend && (status & MCI_DATAEND))
+		host->curr.got_dataend = 1;
+
+	/*
+	 * If DMA is still in progress, we complete via the completion handler
+	 */
+	if (host->curr.got_dataend && !host->dma.busy) {
+		/*
+		 * There appears to be an issue in the controller where
+		 * if you request a small block transfer (< fifo size),
+		 * you may get your DATAEND/DATABLKEND irq without the
+		 * PIO data irq.
+		 *
+		 * Check to see if there is still data to be read,
+		 * and simulate a PIO irq.
+		 */
+		if (readl(base + MMCISTATUS) & MCI_RXDATAAVLBL)
+			msmsdcc_pio_irq(1, host);
+
+		msmsdcc_stop_data(host);
+		if (!data->error)
+			host->curr.data_xfered = host->curr.xfer_size;
+
+		if (!data->stop)
+			msmsdcc_request_end(host, data->mrq);
+		else
+			msmsdcc_start_command(host, data->stop, 0);
+	}
+}
+
+static irqreturn_t
+msmsdcc_irq(int irq, void *dev_id)
+{
+	struct msmsdcc_host	*host = dev_id;
+	void __iomem		*base = host->base;
+	u32			status;
+	int			ret = 0;
+	int			cardint = 0;
+
+	spin_lock(&host->lock);
+
+	do {
+		status = msmsdcc_readl(host, MMCISTATUS);
+		status &= msmsdcc_readl(host, MMCIMASK0);
+		if ((status & (~MCI_IRQ_PIO)) == 0)
+			break;
+		msmsdcc_writel(host, status, MMCICLEAR);
+
+		if (status & MCI_SDIOINTR)
+			status &= ~MCI_SDIOINTR;
+
+		if (!status)
+			break;
+
+		msmsdcc_handle_irq_data(host, status, base);
+
+		if (status & MCI_SDIOINTOPER) {
+			cardint = 1;
+			status &= ~MCI_SDIOINTOPER;
+		}
+		ret = 1;
+	} while (status);
+
+	spin_unlock(&host->lock);
+
+	/*
+	 * We have to delay handling the card interrupt as it calls
+	 * back into the driver.
+	 */
+	if (cardint)
+		mmc_signal_sdio_irq(host->mmc);
+
+	return IRQ_RETVAL(ret);
+}
+
+static void
+msmsdcc_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct msmsdcc_host *host = mmc_priv(mmc);
+	unsigned long flags;
+
+	WARN_ON(host->curr.mrq != NULL);
+	WARN_ON(host->pwr == 0);
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	host->stats.reqs++;
+
+	if (host->eject) {
+		if (mrq->data && !(mrq->data->flags & MMC_DATA_READ)) {
+			mrq->cmd->error = 0;
+			mrq->data->bytes_xfered = mrq->data->blksz *
+						  mrq->data->blocks;
+		} else
+			mrq->cmd->error = -ENOMEDIUM;
+
+		spin_unlock_irqrestore(&host->lock, flags);
+		mmc_request_done(mmc, mrq);
+		return;
+	}
+
+	msmsdcc_enable_clocks(host);
+
+	host->curr.mrq = mrq;
+
+	if (mrq->data && mrq->data->flags & MMC_DATA_READ)
+		/* Queue/read data, daisy-chain command when data starts */
+		msmsdcc_start_data(host, mrq->data, mrq->cmd, 0);
+	else
+		msmsdcc_start_command(host, mrq->cmd, 0);
+
+	if (host->cmdpoll && !msmsdcc_spin_on_status(host,
+				MCI_CMDRESPEND|MCI_CMDCRCFAIL|MCI_CMDTIMEOUT,
+				CMD_SPINMAX)) {
+		uint32_t status = msmsdcc_readl(host, MMCISTATUS);
+		msmsdcc_do_cmdirq(host, status);
+		msmsdcc_writel(host,
+			       MCI_CMDRESPEND | MCI_CMDCRCFAIL | MCI_CMDTIMEOUT,
+			       MMCICLEAR);
+		host->stats.cmdpoll_hits++;
+	} else {
+		host->stats.cmdpoll_misses++;
+	}
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void msmsdcc_setup_gpio(struct msmsdcc_host *host, bool enable)
+{
+	struct msm_mmc_gpio_data *curr;
+	int i, rc = 0;
+
+	if (!host->plat->gpio_data || host->gpio_config_status == enable)
+		return;
+
+	curr = host->plat->gpio_data;
+	for (i = 0; i < curr->size; i++) {
+		if (enable) {
+			rc = gpio_request(curr->gpio[i].no,
+						curr->gpio[i].name);
+			if (rc) {
+				pr_err("%s: gpio_request(%d, %s) failed %d\n",
+					mmc_hostname(host->mmc),
+					curr->gpio[i].no,
+					curr->gpio[i].name, rc);
+				goto free_gpios;
+			}
+		} else {
+			gpio_free(curr->gpio[i].no);
+		}
+	}
+	host->gpio_config_status = enable;
+	return;
+
+free_gpios:
+	for (; i >= 0; i--)
+		gpio_free(curr->gpio[i].no);
+}
+
+static void
+msmsdcc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct msmsdcc_host *host = mmc_priv(mmc);
+	u32 clk = 0, pwr = 0;
+	int rc;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	msmsdcc_enable_clocks(host);
+
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	if (ios->clock) {
+		if (ios->clock != host->clk_rate) {
+			rc = clk_set_rate(host->clk, ios->clock);
+			if (rc < 0)
+				pr_err("%s: Error setting clock rate (%d)\n",
+				       mmc_hostname(host->mmc), rc);
+			else
+				host->clk_rate = ios->clock;
+		}
+		clk |= MCI_CLK_ENABLE;
+	}
+
+	if (ios->bus_width == MMC_BUS_WIDTH_4)
+		clk |= (2 << 10); /* Set WIDEBUS */
+
+	if (ios->clock > 400000 && msmsdcc_pwrsave)
+		clk |= (1 << 9); /* PWRSAVE */
+
+	clk |= (1 << 12); /* FLOW_ENA */
+	clk |= (1 << 15); /* feedback clock */
+
+	if (host->plat->translate_vdd)
+		pwr |= host->plat->translate_vdd(mmc_dev(mmc), ios->vdd);
+
+	switch (ios->power_mode) {
+	case MMC_POWER_OFF:
+		msmsdcc_setup_gpio(host, false);
+		break;
+	case MMC_POWER_UP:
+		pwr |= MCI_PWR_UP;
+		msmsdcc_setup_gpio(host, true);
+		break;
+	case MMC_POWER_ON:
+		pwr |= MCI_PWR_ON;
+		break;
+	}
+
+	if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN)
+		pwr |= MCI_OD;
+
+	msmsdcc_writel(host, clk, MMCICLOCK);
+
+	if (host->pwr != pwr) {
+		host->pwr = pwr;
+		msmsdcc_writel(host, pwr, MMCIPOWER);
+	}
+#if BUSCLK_PWRSAVE
+	spin_lock_irqsave(&host->lock, flags);
+	msmsdcc_disable_clocks(host, 1);
+	spin_unlock_irqrestore(&host->lock, flags);
+#endif
+}
+
+static void msmsdcc_enable_sdio_irq(struct mmc_host *mmc, int enable)
+{
+	struct msmsdcc_host *host = mmc_priv(mmc);
+	unsigned long flags;
+	u32 status;
+
+	spin_lock_irqsave(&host->lock, flags);
+	if (msmsdcc_sdioirq == 1) {
+		status = msmsdcc_readl(host, MMCIMASK0);
+		if (enable)
+			status |= MCI_SDIOINTOPERMASK;
+		else
+			status &= ~MCI_SDIOINTOPERMASK;
+		host->saved_irq0mask = status;
+		msmsdcc_writel(host, status, MMCIMASK0);
+	}
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void msmsdcc_init_card(struct mmc_host *mmc, struct mmc_card *card)
+{
+	struct msmsdcc_host *host = mmc_priv(mmc);
+
+	if (host->plat->init_card)
+		host->plat->init_card(card);
+}
+
+static const struct mmc_host_ops msmsdcc_ops = {
+	.request	= msmsdcc_request,
+	.set_ios	= msmsdcc_set_ios,
+	.enable_sdio_irq = msmsdcc_enable_sdio_irq,
+	.init_card	= msmsdcc_init_card,
+};
+
+static void
+msmsdcc_check_status(unsigned long data)
+{
+	struct msmsdcc_host *host = (struct msmsdcc_host *)data;
+	unsigned int status;
+
+	if (!host->plat->status) {
+		mmc_detect_change(host->mmc, 0);
+		goto out;
+	}
+
+	status = host->plat->status(mmc_dev(host->mmc));
+	host->eject = !status;
+	if (status ^ host->oldstat) {
+		pr_info("%s: Slot status change detected (%d -> %d)\n",
+			mmc_hostname(host->mmc), host->oldstat, status);
+		if (status)
+			mmc_detect_change(host->mmc, (5 * HZ) / 2);
+		else
+			mmc_detect_change(host->mmc, 0);
+	}
+
+	host->oldstat = status;
+
+out:
+	if (host->timer.function)
+		mod_timer(&host->timer, jiffies + HZ);
+}
+
+static irqreturn_t
+msmsdcc_platform_status_irq(int irq, void *dev_id)
+{
+	struct msmsdcc_host *host = dev_id;
+
+	pr_debug("%s: %d\n", __func__, irq);
+	msmsdcc_check_status((unsigned long) host);
+	return IRQ_HANDLED;
+}
+
+static void
+msmsdcc_status_notify_cb(int card_present, void *dev_id)
+{
+	struct msmsdcc_host *host = dev_id;
+
+	pr_debug("%s: card_present %d\n", mmc_hostname(host->mmc),
+	       card_present);
+	msmsdcc_check_status((unsigned long) host);
+}
+
+static void
+msmsdcc_busclk_expired(unsigned long _data)
+{
+	struct msmsdcc_host	*host = (struct msmsdcc_host *) _data;
+
+	if (host->clks_on)
+		msmsdcc_disable_clocks(host, 0);
+}
+
+static int
+msmsdcc_init_dma(struct msmsdcc_host *host)
+{
+	memset(&host->dma, 0, sizeof(struct msmsdcc_dma_data));
+	host->dma.host = host;
+	host->dma.channel = -1;
+
+	if (!host->dmares)
+		return -ENODEV;
+
+	host->dma.nc = dma_alloc_coherent(NULL,
+					  sizeof(struct msmsdcc_nc_dmadata),
+					  &host->dma.nc_busaddr,
+					  GFP_KERNEL);
+	if (host->dma.nc == NULL) {
+		pr_err("Unable to allocate DMA buffer\n");
+		return -ENOMEM;
+	}
+	memset(host->dma.nc, 0x00, sizeof(struct msmsdcc_nc_dmadata));
+	host->dma.cmd_busaddr = host->dma.nc_busaddr;
+	host->dma.cmdptr_busaddr = host->dma.nc_busaddr +
+				offsetof(struct msmsdcc_nc_dmadata, cmdptr);
+	host->dma.channel = host->dmares->start;
+
+	return 0;
+}
+
+static int
+msmsdcc_probe(struct platform_device *pdev)
+{
+	struct msm_mmc_platform_data *plat = pdev->dev.platform_data;
+	struct msmsdcc_host *host;
+	struct mmc_host *mmc;
+	struct resource *cmd_irqres = NULL;
+	struct resource *stat_irqres = NULL;
+	struct resource *memres = NULL;
+	struct resource *dmares = NULL;
+	int ret;
+
+	/* must have platform data */
+	if (!plat) {
+		pr_err("%s: Platform data not available\n", __func__);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (pdev->id < 1 || pdev->id > 4)
+		return -EINVAL;
+
+	if (pdev->resource == NULL || pdev->num_resources < 2) {
+		pr_err("%s: Invalid resource\n", __func__);
+		return -ENXIO;
+	}
+
+	memres = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dmares = platform_get_resource(pdev, IORESOURCE_DMA, 0);
+	cmd_irqres = platform_get_resource_byname(pdev, IORESOURCE_IRQ,
+						  "cmd_irq");
+	stat_irqres = platform_get_resource_byname(pdev, IORESOURCE_IRQ,
+						   "status_irq");
+
+	if (!cmd_irqres || !memres) {
+		pr_err("%s: Invalid resource\n", __func__);
+		return -ENXIO;
+	}
+
+	/*
+	 * Setup our host structure
+	 */
+
+	mmc = mmc_alloc_host(sizeof(struct msmsdcc_host), &pdev->dev);
+	if (!mmc) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	host = mmc_priv(mmc);
+	host->pdev_id = pdev->id;
+	host->plat = plat;
+	host->mmc = mmc;
+	host->curr.cmd = NULL;
+	init_timer(&host->busclk_timer);
+	host->busclk_timer.data = (unsigned long) host;
+	host->busclk_timer.function = msmsdcc_busclk_expired;
+
+
+	host->cmdpoll = 1;
+
+	host->base = ioremap(memres->start, PAGE_SIZE);
+	if (!host->base) {
+		ret = -ENOMEM;
+		goto host_free;
+	}
+
+	host->cmd_irqres = cmd_irqres;
+	host->memres = memres;
+	host->dmares = dmares;
+	spin_lock_init(&host->lock);
+
+	tasklet_init(&host->dma_tlet, msmsdcc_dma_complete_tlet,
+			(unsigned long)host);
+
+	/*
+	 * Setup DMA
+	 */
+	if (host->dmares) {
+		ret = msmsdcc_init_dma(host);
+		if (ret)
+			goto ioremap_free;
+	} else {
+		host->dma.channel = -1;
+	}
+
+	/* Get our clocks */
+	host->pclk = clk_get(&pdev->dev, "sdc_pclk");
+	if (IS_ERR(host->pclk)) {
+		ret = PTR_ERR(host->pclk);
+		goto dma_free;
+	}
+
+	host->clk = clk_get(&pdev->dev, "sdc_clk");
+	if (IS_ERR(host->clk)) {
+		ret = PTR_ERR(host->clk);
+		goto pclk_put;
+	}
+
+	ret = clk_set_rate(host->clk, msmsdcc_fmin);
+	if (ret) {
+		pr_err("%s: Clock rate set failed (%d)\n", __func__, ret);
+		goto clk_put;
+	}
+
+	ret = clk_prepare(host->pclk);
+	if (ret)
+		goto clk_put;
+
+	ret = clk_prepare(host->clk);
+	if (ret)
+		goto clk_unprepare_p;
+
+	/* Enable clocks */
+	ret = msmsdcc_enable_clocks(host);
+	if (ret)
+		goto clk_unprepare;
+
+	host->pclk_rate = clk_get_rate(host->pclk);
+	host->clk_rate = clk_get_rate(host->clk);
+
+	/*
+	 * Setup MMC host structure
+	 */
+	mmc->ops = &msmsdcc_ops;
+	mmc->f_min = msmsdcc_fmin;
+	mmc->f_max = msmsdcc_fmax;
+	mmc->ocr_avail = plat->ocr_mask;
+
+	if (msmsdcc_4bit)
+		mmc->caps |= MMC_CAP_4_BIT_DATA;
+	if (msmsdcc_sdioirq)
+		mmc->caps |= MMC_CAP_SDIO_IRQ;
+	mmc->caps |= MMC_CAP_MMC_HIGHSPEED | MMC_CAP_SD_HIGHSPEED;
+
+	mmc->max_segs = NR_SG;
+	mmc->max_blk_size = 4096;	/* MCI_DATA_CTL BLOCKSIZE up to 4096 */
+	mmc->max_blk_count = 65536;
+
+	mmc->max_req_size = 33554432;	/* MCI_DATA_LENGTH is 25 bits */
+	mmc->max_seg_size = mmc->max_req_size;
+
+	msmsdcc_writel(host, 0, MMCIMASK0);
+	msmsdcc_writel(host, 0x5e007ff, MMCICLEAR);
+
+	msmsdcc_writel(host, MCI_IRQENABLE, MMCIMASK0);
+	host->saved_irq0mask = MCI_IRQENABLE;
+
+	/*
+	 * Setup card detect change
+	 */
+
+	memset(&host->timer, 0, sizeof(host->timer));
+
+	if (stat_irqres && !(stat_irqres->flags & IORESOURCE_DISABLED)) {
+		unsigned long irqflags = IRQF_SHARED |
+			(stat_irqres->flags & IRQF_TRIGGER_MASK);
+
+		host->stat_irq = stat_irqres->start;
+		ret = request_irq(host->stat_irq,
+				  msmsdcc_platform_status_irq,
+				  irqflags,
+				  DRIVER_NAME " (slot)",
+				  host);
+		if (ret) {
+			pr_err("%s: Unable to get slot IRQ %d (%d)\n",
+			       mmc_hostname(mmc), host->stat_irq, ret);
+			goto clk_disable;
+		}
+	} else if (plat->register_status_notify) {
+		plat->register_status_notify(msmsdcc_status_notify_cb, host);
+	} else if (!plat->status)
+		pr_err("%s: No card detect facilities available\n",
+		       mmc_hostname(mmc));
+	else {
+		init_timer(&host->timer);
+		host->timer.data = (unsigned long)host;
+		host->timer.function = msmsdcc_check_status;
+		host->timer.expires = jiffies + HZ;
+		add_timer(&host->timer);
+	}
+
+	if (plat->status) {
+		host->oldstat = host->plat->status(mmc_dev(host->mmc));
+		host->eject = !host->oldstat;
+	}
+
+	ret = request_irq(cmd_irqres->start, msmsdcc_irq, IRQF_SHARED,
+			  DRIVER_NAME " (cmd)", host);
+	if (ret)
+		goto stat_irq_free;
+
+	ret = request_irq(cmd_irqres->start, msmsdcc_pio_irq, IRQF_SHARED,
+			  DRIVER_NAME " (pio)", host);
+	if (ret)
+		goto cmd_irq_free;
+
+	mmc_set_drvdata(pdev, mmc);
+	mmc_add_host(mmc);
+
+	pr_info("%s: Qualcomm MSM SDCC at 0x%016llx irq %d,%d dma %d\n",
+		mmc_hostname(mmc), (unsigned long long)memres->start,
+		(unsigned int) cmd_irqres->start,
+		(unsigned int) host->stat_irq, host->dma.channel);
+	pr_info("%s: 4 bit data mode %s\n", mmc_hostname(mmc),
+		(mmc->caps & MMC_CAP_4_BIT_DATA ? "enabled" : "disabled"));
+	pr_info("%s: MMC clock %u -> %u Hz, PCLK %u Hz\n",
+		mmc_hostname(mmc), msmsdcc_fmin, msmsdcc_fmax, host->pclk_rate);
+	pr_info("%s: Slot eject status = %d\n", mmc_hostname(mmc), host->eject);
+	pr_info("%s: Power save feature enable = %d\n",
+		mmc_hostname(mmc), msmsdcc_pwrsave);
+
+	if (host->dma.channel != -1) {
+		pr_info("%s: DM non-cached buffer at %p, dma_addr 0x%.8x\n",
+			mmc_hostname(mmc), host->dma.nc, host->dma.nc_busaddr);
+		pr_info("%s: DM cmd busaddr 0x%.8x, cmdptr busaddr 0x%.8x\n",
+			mmc_hostname(mmc), host->dma.cmd_busaddr,
+			host->dma.cmdptr_busaddr);
+	} else
+		pr_info("%s: PIO transfer enabled\n", mmc_hostname(mmc));
+	if (host->timer.function)
+		pr_info("%s: Polling status mode enabled\n", mmc_hostname(mmc));
+
+	return 0;
+ cmd_irq_free:
+	free_irq(cmd_irqres->start, host);
+ stat_irq_free:
+	if (host->stat_irq)
+		free_irq(host->stat_irq, host);
+ clk_disable:
+	msmsdcc_disable_clocks(host, 0);
+ clk_unprepare:
+	clk_unprepare(host->clk);
+ clk_unprepare_p:
+	clk_unprepare(host->pclk);
+ clk_put:
+	clk_put(host->clk);
+ pclk_put:
+	clk_put(host->pclk);
+dma_free:
+	if (host->dmares)
+		dma_free_coherent(NULL, sizeof(struct msmsdcc_nc_dmadata),
+					host->dma.nc, host->dma.nc_busaddr);
+ioremap_free:
+	tasklet_kill(&host->dma_tlet);
+	iounmap(host->base);
+ host_free:
+	mmc_free_host(mmc);
+ out:
+	return ret;
+}
+
+#ifdef CONFIG_PM
+static int
+msmsdcc_suspend(struct platform_device *dev, pm_message_t state)
+{
+	struct mmc_host *mmc = mmc_get_drvdata(dev);
+
+	if (mmc) {
+		struct msmsdcc_host *host = mmc_priv(mmc);
+
+		if (host->stat_irq)
+			disable_irq(host->stat_irq);
+
+		msmsdcc_writel(host, 0, MMCIMASK0);
+		if (host->clks_on)
+			msmsdcc_disable_clocks(host, 0);
+	}
+	return 0;
+}
+
+static int
+msmsdcc_resume(struct platform_device *dev)
+{
+	struct mmc_host *mmc = mmc_get_drvdata(dev);
+
+	if (mmc) {
+		struct msmsdcc_host *host = mmc_priv(mmc);
+
+		msmsdcc_enable_clocks(host);
+
+		msmsdcc_writel(host, host->saved_irq0mask, MMCIMASK0);
+
+		if (host->stat_irq)
+			enable_irq(host->stat_irq);
+#if BUSCLK_PWRSAVE
+		msmsdcc_disable_clocks(host, 1);
+#endif
+	}
+	return 0;
+}
+#else
+#define msmsdcc_suspend	0
+#define msmsdcc_resume 0
+#endif
+
+static struct platform_driver msmsdcc_driver = {
+	.probe		= msmsdcc_probe,
+	.suspend	= msmsdcc_suspend,
+	.resume		= msmsdcc_resume,
+	.driver		= {
+		.name	= "msm_sdcc",
+	},
+};
+
+module_platform_driver(msmsdcc_driver);
+
+MODULE_DESCRIPTION("Qualcomm MSM 7X00A Multimedia Card Interface driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/mmc/host/msm_sdcc.h b/drivers/mmc/host/msm_sdcc.h
new file mode 100644
index 00000000000..402028d16b8
--- /dev/null
+++ b/drivers/mmc/host/msm_sdcc.h
@@ -0,0 +1,256 @@
+/*
+ *  linux/drivers/mmc/host/msmsdcc.h - QCT MSM7K SDC Controller
+ *
+ *  Copyright (C) 2008 Google, All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * - Based on mmci.h
+ */
+
+#ifndef _MSM_SDCC_H
+#define _MSM_SDCC_H
+
+#define MSMSDCC_CRCI_SDC1	6
+#define MSMSDCC_CRCI_SDC2	7
+#define MSMSDCC_CRCI_SDC3	12
+#define MSMSDCC_CRCI_SDC4	13
+
+#define MMCIPOWER		0x000
+#define MCI_PWR_OFF		0x00
+#define MCI_PWR_UP		0x02
+#define MCI_PWR_ON		0x03
+#define MCI_OD			(1 << 6)
+
+#define MMCICLOCK		0x004
+#define MCI_CLK_ENABLE		(1 << 8)
+#define MCI_CLK_PWRSAVE		(1 << 9)
+#define MCI_CLK_WIDEBUS		(1 << 10)
+#define MCI_CLK_FLOWENA		(1 << 12)
+#define MCI_CLK_INVERTOUT	(1 << 13)
+#define MCI_CLK_SELECTIN	(1 << 14)
+
+#define MMCIARGUMENT		0x008
+#define MMCICOMMAND		0x00c
+#define MCI_CPSM_RESPONSE	(1 << 6)
+#define MCI_CPSM_LONGRSP	(1 << 7)
+#define MCI_CPSM_INTERRUPT	(1 << 8)
+#define MCI_CPSM_PENDING	(1 << 9)
+#define MCI_CPSM_ENABLE		(1 << 10)
+#define MCI_CPSM_PROGENA	(1 << 11)
+#define MCI_CSPM_DATCMD		(1 << 12)
+#define MCI_CSPM_MCIABORT	(1 << 13)
+#define MCI_CSPM_CCSENABLE	(1 << 14)
+#define MCI_CSPM_CCSDISABLE	(1 << 15)
+
+
+#define MMCIRESPCMD		0x010
+#define MMCIRESPONSE0		0x014
+#define MMCIRESPONSE1		0x018
+#define MMCIRESPONSE2		0x01c
+#define MMCIRESPONSE3		0x020
+#define MMCIDATATIMER		0x024
+#define MMCIDATALENGTH		0x028
+
+#define MMCIDATACTRL		0x02c
+#define MCI_DPSM_ENABLE		(1 << 0)
+#define MCI_DPSM_DIRECTION	(1 << 1)
+#define MCI_DPSM_MODE		(1 << 2)
+#define MCI_DPSM_DMAENABLE	(1 << 3)
+
+#define MMCIDATACNT		0x030
+#define MMCISTATUS		0x034
+#define MCI_CMDCRCFAIL		(1 << 0)
+#define MCI_DATACRCFAIL		(1 << 1)
+#define MCI_CMDTIMEOUT		(1 << 2)
+#define MCI_DATATIMEOUT		(1 << 3)
+#define MCI_TXUNDERRUN		(1 << 4)
+#define MCI_RXOVERRUN		(1 << 5)
+#define MCI_CMDRESPEND		(1 << 6)
+#define MCI_CMDSENT		(1 << 7)
+#define MCI_DATAEND		(1 << 8)
+#define MCI_DATABLOCKEND	(1 << 10)
+#define MCI_CMDACTIVE		(1 << 11)
+#define MCI_TXACTIVE		(1 << 12)
+#define MCI_RXACTIVE		(1 << 13)
+#define MCI_TXFIFOHALFEMPTY	(1 << 14)
+#define MCI_RXFIFOHALFFULL	(1 << 15)
+#define MCI_TXFIFOFULL		(1 << 16)
+#define MCI_RXFIFOFULL		(1 << 17)
+#define MCI_TXFIFOEMPTY		(1 << 18)
+#define MCI_RXFIFOEMPTY		(1 << 19)
+#define MCI_TXDATAAVLBL		(1 << 20)
+#define MCI_RXDATAAVLBL		(1 << 21)
+#define MCI_SDIOINTR		(1 << 22)
+#define MCI_PROGDONE		(1 << 23)
+#define MCI_ATACMDCOMPL		(1 << 24)
+#define MCI_SDIOINTOPER		(1 << 25)
+#define MCI_CCSTIMEOUT		(1 << 26)
+
+#define MMCICLEAR		0x038
+#define MCI_CMDCRCFAILCLR	(1 << 0)
+#define MCI_DATACRCFAILCLR	(1 << 1)
+#define MCI_CMDTIMEOUTCLR	(1 << 2)
+#define MCI_DATATIMEOUTCLR	(1 << 3)
+#define MCI_TXUNDERRUNCLR	(1 << 4)
+#define MCI_RXOVERRUNCLR	(1 << 5)
+#define MCI_CMDRESPENDCLR	(1 << 6)
+#define MCI_CMDSENTCLR		(1 << 7)
+#define MCI_DATAENDCLR		(1 << 8)
+#define MCI_DATABLOCKENDCLR	(1 << 10)
+
+#define MMCIMASK0		0x03c
+#define MCI_CMDCRCFAILMASK	(1 << 0)
+#define MCI_DATACRCFAILMASK	(1 << 1)
+#define MCI_CMDTIMEOUTMASK	(1 << 2)
+#define MCI_DATATIMEOUTMASK	(1 << 3)
+#define MCI_TXUNDERRUNMASK	(1 << 4)
+#define MCI_RXOVERRUNMASK	(1 << 5)
+#define MCI_CMDRESPENDMASK	(1 << 6)
+#define MCI_CMDSENTMASK		(1 << 7)
+#define MCI_DATAENDMASK		(1 << 8)
+#define MCI_DATABLOCKENDMASK	(1 << 10)
+#define MCI_CMDACTIVEMASK	(1 << 11)
+#define MCI_TXACTIVEMASK	(1 << 12)
+#define MCI_RXACTIVEMASK	(1 << 13)
+#define MCI_TXFIFOHALFEMPTYMASK	(1 << 14)
+#define MCI_RXFIFOHALFFULLMASK	(1 << 15)
+#define MCI_TXFIFOFULLMASK	(1 << 16)
+#define MCI_RXFIFOFULLMASK	(1 << 17)
+#define MCI_TXFIFOEMPTYMASK	(1 << 18)
+#define MCI_RXFIFOEMPTYMASK	(1 << 19)
+#define MCI_TXDATAAVLBLMASK	(1 << 20)
+#define MCI_RXDATAAVLBLMASK	(1 << 21)
+#define MCI_SDIOINTMASK		(1 << 22)
+#define MCI_PROGDONEMASK	(1 << 23)
+#define MCI_ATACMDCOMPLMASK	(1 << 24)
+#define MCI_SDIOINTOPERMASK	(1 << 25)
+#define MCI_CCSTIMEOUTMASK	(1 << 26)
+
+#define MMCIMASK1		0x040
+#define MMCIFIFOCNT		0x044
+#define MCICCSTIMER		0x058
+
+#define MMCIFIFO		0x080 /* to 0x0bc */
+
+#define MCI_IRQENABLE	\
+	(MCI_CMDCRCFAILMASK|MCI_DATACRCFAILMASK|MCI_CMDTIMEOUTMASK|	\
+	MCI_DATATIMEOUTMASK|MCI_TXUNDERRUNMASK|MCI_RXOVERRUNMASK|	\
+	MCI_CMDRESPENDMASK|MCI_CMDSENTMASK|MCI_DATAENDMASK|MCI_PROGDONEMASK)
+
+#define MCI_IRQ_PIO \
+	(MCI_RXDATAAVLBLMASK | MCI_TXDATAAVLBLMASK | MCI_RXFIFOEMPTYMASK | \
+	 MCI_TXFIFOEMPTYMASK | MCI_RXFIFOFULLMASK | MCI_TXFIFOFULLMASK | \
+	 MCI_RXFIFOHALFFULLMASK | MCI_TXFIFOHALFEMPTYMASK | \
+	 MCI_RXACTIVEMASK | MCI_TXACTIVEMASK)
+/*
+ * The size of the FIFO in bytes.
+ */
+#define MCI_FIFOSIZE	(16*4)
+
+#define MCI_FIFOHALFSIZE (MCI_FIFOSIZE / 2)
+
+#define NR_SG		32
+
+struct clk;
+
+struct msmsdcc_nc_dmadata {
+	dmov_box	cmd[NR_SG];
+	uint32_t	cmdptr;
+};
+
+struct msmsdcc_dma_data {
+	struct msmsdcc_nc_dmadata	*nc;
+	dma_addr_t			nc_busaddr;
+	dma_addr_t			cmd_busaddr;
+	dma_addr_t			cmdptr_busaddr;
+
+	struct msm_dmov_cmd		hdr;
+	enum dma_data_direction		dir;
+
+	struct scatterlist		*sg;
+	int				num_ents;
+
+	int				channel;
+	struct msmsdcc_host		*host;
+	int				busy; /* Set if DM is busy */
+	int				active;
+	unsigned int			result;
+	struct msm_dmov_errdata		err;
+};
+
+struct msmsdcc_pio_data {
+	struct scatterlist	*sg;
+	unsigned int		sg_len;
+	unsigned int		sg_off;
+};
+
+struct msmsdcc_curr_req {
+	struct mmc_request	*mrq;
+	struct mmc_command	*cmd;
+	struct mmc_data		*data;
+	unsigned int		xfer_size;	/* Total data size */
+	unsigned int		xfer_remain;	/* Bytes remaining to send */
+	unsigned int		data_xfered;	/* Bytes acked by BLKEND irq */
+	int			got_dataend;
+	int			user_pages;
+};
+
+struct msmsdcc_stats {
+	unsigned int reqs;
+	unsigned int cmds;
+	unsigned int cmdpoll_hits;
+	unsigned int cmdpoll_misses;
+};
+
+struct msmsdcc_host {
+	struct resource		*cmd_irqres;
+	struct resource		*memres;
+	struct resource		*dmares;
+	void __iomem		*base;
+	int			pdev_id;
+	unsigned int		stat_irq;
+
+	struct msmsdcc_curr_req	curr;
+
+	struct mmc_host		*mmc;
+	struct clk		*clk;		/* main MMC bus clock */
+	struct clk		*pclk;		/* SDCC peripheral bus clock */
+	unsigned int		clks_on;	/* set if clocks are enabled */
+	struct timer_list	busclk_timer;
+
+	unsigned int		eject;		/* eject state */
+
+	spinlock_t		lock;
+
+	unsigned int		clk_rate;	/* Current clock rate */
+	unsigned int		pclk_rate;
+
+	u32			pwr;
+	u32			saved_irq0mask;	/* MMCIMASK0 reg value */
+	struct msm_mmc_platform_data *plat;
+
+	struct timer_list	timer;
+	unsigned int		oldstat;
+
+	struct msmsdcc_dma_data	dma;
+	struct msmsdcc_pio_data	pio;
+	int			cmdpoll;
+	struct msmsdcc_stats	stats;
+
+	struct tasklet_struct	dma_tlet;
+	/* Command parameters */
+	unsigned int		cmd_timeout;
+	unsigned int		cmd_pio_irqmask;
+	unsigned int		cmd_datactrl;
+	struct mmc_command	*cmd_cmd;
+	u32			cmd_c;
+	bool			gpio_config_status;
+
+	bool prog_scan;
+	bool prog_enable;
+};
+
+#endif
diff --git a/drivers/mmc/host/mvsdio.c b/drivers/mmc/host/mvsdio.c