Merge tag 'mmc-v5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc

Pull MMC and MEMSTICK updates from Ulf Hansson:
 "MMC core:
   - Add support for Cache Ctrl for SD cards
   - Add support for Power Off Notification for SD cards
   - Add support for read/write of the SD function extension registers
   - Allow broken eMMC HS400 mode to be disabled via DT
   - Allow UHS-I voltage switch for SDSC cards if supported
   - Disable command queueing in the ioctl path
   - Enable eMMC sleep commands to use HW busy polling to minimize delay
   - Extend re-use of the common polling loop to standardize behaviour
   - Take into account MMC_CAP_NEED_RSP_BUSY for eMMC HPI commands

  MMC host:
   - jz4740: Add support for the JZ4775 variant
   - sdhci-acpi: Disable write protect detection on Toshiba Encore 2 WT8-B
   - sdhci-esdhc-imx: Advertise HS400 support through MMC caps
   - sdhci-esdhc-imx: Enable support for system wakeup for SDIO
   - sdhci-iproc: Add support for the legacy sdhci controller on the BCM7211
   - vub3000: Fix control-request direction

  MEMSTICK:
   - A couple of fixes/cleanups"

* tag 'mmc-v5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc: (54 commits)
  mmc: sdhci-iproc: Add support for the legacy sdhci controller on the BCM7211
  dt-bindings: mmc: sdhci-iproc: Add brcm,bcm7211a0-sdhci
  mmc: JZ4740: Add support for JZ4775
  dt-bindings: mmc: JZ4740: Add bindings for JZ4775
  mmc: sdhci-esdhc-imx: Enable support for system wakeup for SDIO
  mmc: Improve function name when aborting a tuning cmd
  mmc: sdhci-of-aspeed: Turn down a phase correction warning
  mmc: debugfs: add description for module parameter
  mmc: via-sdmmc: add a check against NULL pointer dereference
  mmc: sdhci-sprd: use sdhci_sprd_writew
  mmc: sdhci-esdhc-imx: remove unused is_imx6q_usdhc
  mmc: core: Allow UHS-I voltage switch for SDSC cards if supported
  mmc: mmc_spi: Imply container_of() to be no-op
  mmc: mmc_spi: Drop duplicate 'mmc_spi' in the debug messages
  mmc: dw_mmc-pltfm: Remove unused <linux/clk.h>
  mmc: sdhci-of-aspeed: Configure the SDHCIs as specified by the devicetree.
  mmc: core: Add a missing SPDX license header
  mmc: vub3000: fix control-request direction
  mmc: sdhci-omap: Use pm_runtime_resume_and_get() to replace open coding
  mmc: sdhci_am654: Use pm_runtime_resume_and_get() to replace open coding
  ...

35 files changed, 813 insertions, 261 deletions

diff --git a/drivers/memstick/core/ms_block.c b/drivers/memstick/core/ms_block.c
index 8004dd64d09a..d971acd98236 100644
--- a/drivers/memstick/core/ms_block.c
+++ b/drivers/memstick/core/ms_block.c
@@ -129,7 +129,7 @@ static int msb_sg_compare_to_buffer(struct scatterlist *sg,
  * Each zone consists of 512 eraseblocks, out of which in first
  * zone 494 are used and 496 are for all following zones.
  * Therefore zone #0 hosts blocks 0-493, zone #1 blocks 494-988, etc...
-*/
+ */
 static int msb_get_zone_from_lba(int lba)
 {
 	if (lba < 494)
@@ -348,8 +348,9 @@ again:
 	switch (msb->state) {
 	case MSB_RP_SEND_BLOCK_ADDRESS:
 		/* msb_write_regs sometimes "fails" because it needs to update
-			the reg window, and thus it returns request for that.
-			Then we stay in this state and retry */
+		 * the reg window, and thus it returns request for that.
+		 * Then we stay in this state and retry
+		 */
 		if (!msb_write_regs(msb,
 			offsetof(struct ms_register, param),
 			sizeof(struct ms_param_register),
@@ -368,7 +369,8 @@ again:
 	case MSB_RP_SEND_INT_REQ:
 		msb->state = MSB_RP_RECEIVE_INT_REQ_RESULT;
 		/* If dont actually need to send the int read request (only in
-			serial mode), then just fall through */
+		 * serial mode), then just fall through
+		 */
 		if (msb_read_int_reg(msb, -1))
 			return 0;
 		fallthrough;
@@ -702,7 +704,8 @@ static int h_msb_parallel_switch(struct memstick_dev *card,
 
 	case MSB_PS_SWICH_HOST:
 		 /* Set parallel interface on our side + send a dummy request
-			to see if card responds */
+		  * to see if card responds
+		  */
 		host->set_param(host, MEMSTICK_INTERFACE, MEMSTICK_PAR4);
 		memstick_init_req(mrq, MS_TPC_GET_INT, NULL, 1);
 		msb->state = MSB_PS_CONFIRM;
@@ -821,6 +824,7 @@ static int msb_mark_page_bad(struct msb_data *msb, int pba, int page)
 static int msb_erase_block(struct msb_data *msb, u16 pba)
 {
 	int error, try;
+
 	if (msb->read_only)
 		return -EROFS;
 
@@ -997,6 +1001,7 @@ static int msb_write_block(struct msb_data *msb,
 			u16 pba, u32 lba, struct scatterlist *sg, int offset)
 {
 	int error, current_try = 1;
+
 	BUG_ON(sg->length < msb->page_size);
 
 	if (msb->read_only)
@@ -1045,11 +1050,12 @@ static int msb_write_block(struct msb_data *msb,
 		error = msb_run_state_machine(msb, h_msb_write_block);
 
 		/* Sector we just wrote to is assumed erased since its pba
-			was erased. If it wasn't erased, write will succeed
-			and will just clear the bits that were set in the block
-			thus test that what we have written,
-			matches what we expect.
-			We do trust the blocks that we erased */
+		 * was erased. If it wasn't erased, write will succeed
+		 * and will just clear the bits that were set in the block
+		 * thus test that what we have written,
+		 * matches what we expect.
+		 * We do trust the blocks that we erased
+		 */
 		if (!error && (verify_writes ||
 				!test_bit(pba, msb->erased_blocks_bitmap)))
 			error = msb_verify_block(msb, pba, sg, offset);
@@ -1493,6 +1499,7 @@ static int msb_ftl_scan(struct msb_data *msb)
 static void msb_cache_flush_timer(struct timer_list *t)
 {
 	struct msb_data *msb = from_timer(msb, t, cache_flush_timer);
+
 	msb->need_flush_cache = true;
 	queue_work(msb->io_queue, &msb->io_work);
 }
@@ -1673,7 +1680,8 @@ static int msb_cache_read(struct msb_data *msb, int lba,
  * This table content isn't that importaint,
  * One could put here different values, providing that they still
  * cover whole disk.
- * 64 MB entry is what windows reports for my 64M memstick */
+ * 64 MB entry is what windows reports for my 64M memstick
+ */
 
 static const struct chs_entry chs_table[] = {
 /*        size sectors cylynders  heads */
@@ -1706,8 +1714,9 @@ static int msb_init_card(struct memstick_dev *card)
 		return error;
 
 	/* Due to a bug in Jmicron driver written by Alex Dubov,
-	 its serial mode barely works,
-	 so we switch to parallel mode right away */
+	 * its serial mode barely works,
+	 * so we switch to parallel mode right away
+	 */
 	if (host->caps & MEMSTICK_CAP_PAR4)
 		msb_switch_to_parallel(msb);
 
@@ -2033,6 +2042,7 @@ static blk_status_t msb_queue_rq(struct blk_mq_hw_ctx *hctx,
 static int msb_check_card(struct memstick_dev *card)
 {
 	struct msb_data *msb = memstick_get_drvdata(card);
+
 	return (msb->card_dead == 0);
 }
 
@@ -2333,6 +2343,7 @@ static struct memstick_driver msb_driver = {
 static int __init msb_init(void)
 {
 	int rc = memstick_register_driver(&msb_driver);
+
 	if (rc)
 		pr_err("failed to register memstick driver (error %d)\n", rc);
 
diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index 102dbb8080da..29271ad4728a 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -799,9 +799,9 @@ static int rtsx_usb_ms_drv_probe(struct platform_device *pdev)
 
 	return 0;
 err_out:
-	memstick_free_host(msh);
 	pm_runtime_disable(ms_dev(host));
 	pm_runtime_put_noidle(ms_dev(host));
+	memstick_free_host(msh);
 	return err;
 }
 
@@ -828,9 +828,6 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	}
 	mutex_unlock(&host->host_mutex);
 
-	memstick_remove_host(msh);
-	memstick_free_host(msh);
-
 	/* Balance possible unbalanced usage count
 	 * e.g. unconditional module removal
 	 */
@@ -838,10 +835,11 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 		pm_runtime_put(ms_dev(host));
 
 	pm_runtime_disable(ms_dev(host));
-	platform_set_drvdata(pdev, NULL);
-
+	memstick_remove_host(msh);
 	dev_dbg(ms_dev(host),
 		": Realtek USB Memstick controller has been removed\n");
+	memstick_free_host(msh);
+	platform_set_drvdata(pdev, NULL);
 
 	return 0;
 }
diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c
index 689eb9afeeed..88f4c215caa6 100644
--- a/drivers/mmc/core/block.c
+++ b/drivers/mmc/core/block.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Block driver for media (i.e., flash cards)
  *
@@ -1004,6 +1005,12 @@ static void mmc_blk_issue_drv_op(struct mmc_queue *mq, struct request *req)
 
 	switch (mq_rq->drv_op) {
 	case MMC_DRV_OP_IOCTL:
+		if (card->ext_csd.cmdq_en) {
+			ret = mmc_cmdq_disable(card);
+			if (ret)
+				break;
+		}
+		fallthrough;
 	case MMC_DRV_OP_IOCTL_RPMB:
 		idata = mq_rq->drv_op_data;
 		for (i = 0, ret = 0; i < mq_rq->ioc_count; i++) {
@@ -1014,6 +1021,8 @@ static void mmc_blk_issue_drv_op(struct mmc_queue *mq, struct request *req)
 		/* Always switch back to main area after RPMB access */
 		if (rpmb_ioctl)
 			mmc_blk_part_switch(card, 0);
+		else if (card->reenable_cmdq && !card->ext_csd.cmdq_en)
+			mmc_cmdq_enable(card);
 		break;
 	case MMC_DRV_OP_BOOT_WP:
 		ret = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_BOOT_WP,
@@ -1159,7 +1168,7 @@ static void mmc_blk_issue_flush(struct mmc_queue *mq, struct request *req)
 	struct mmc_card *card = md->queue.card;
 	int ret = 0;
 
-	ret = mmc_flush_cache(card);
+	ret = mmc_flush_cache(card->host);
 	blk_mq_end_request(req, ret ? BLK_STS_IOERR : BLK_STS_OK);
 }
 
diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
index f194940c5974..b039dcff17f8 100644
--- a/drivers/mmc/core/core.c
+++ b/drivers/mmc/core/core.c
@@ -1582,7 +1582,7 @@ static int mmc_do_erase(struct mmc_card *card, unsigned int from,
 {
 	struct mmc_command cmd = {};
 	unsigned int qty = 0, busy_timeout = 0;
-	bool use_r1b_resp = false;
+	bool use_r1b_resp;
 	int err;
 
 	mmc_retune_hold(card->host);
@@ -1650,23 +1650,7 @@ static int mmc_do_erase(struct mmc_card *card, unsigned int from,
 	cmd.opcode = MMC_ERASE;
 	cmd.arg = arg;
 	busy_timeout = mmc_erase_timeout(card, arg, qty);
-	/*
-	 * If the host controller supports busy signalling and the timeout for
-	 * the erase operation does not exceed the max_busy_timeout, we should
-	 * use R1B response. Or we need to prevent the host from doing hw busy
-	 * detection, which is done by converting to a R1 response instead.
-	 * Note, some hosts requires R1B, which also means they are on their own
-	 * when it comes to deal with the busy timeout.
-	 */
-	if (!(card->host->caps & MMC_CAP_NEED_RSP_BUSY) &&
-	    card->host->max_busy_timeout &&
-	    busy_timeout > card->host->max_busy_timeout) {
-		cmd.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_AC;
-	} else {
-		cmd.flags = MMC_RSP_SPI_R1B | MMC_RSP_R1B | MMC_CMD_AC;
-		cmd.busy_timeout = busy_timeout;
-		use_r1b_resp = true;
-	}
+	use_r1b_resp = mmc_prepare_busy_cmd(card->host, &cmd, busy_timeout);
 
 	err = mmc_wait_for_cmd(card->host, &cmd, 0);
 	if (err) {
@@ -1687,7 +1671,7 @@ static int mmc_do_erase(struct mmc_card *card, unsigned int from,
 		goto out;
 
 	/* Let's poll to find out when the erase operation completes. */
-	err = mmc_poll_for_busy(card, busy_timeout, MMC_BUSY_ERASE);
+	err = mmc_poll_for_busy(card, busy_timeout, false, MMC_BUSY_ERASE);
 
 out:
 	mmc_retune_release(card->host);
diff --git a/drivers/mmc/core/core.h b/drivers/mmc/core/core.h
index db3c9c68875d..0c4de2030b3f 100644
--- a/drivers/mmc/core/core.h
+++ b/drivers/mmc/core/core.h
@@ -30,6 +30,7 @@ struct mmc_bus_ops {
 	int (*hw_reset)(struct mmc_host *);
 	int (*sw_reset)(struct mmc_host *);
 	bool (*cache_enabled)(struct mmc_host *);
+	int (*flush_cache)(struct mmc_host *);
 };
 
 void mmc_attach_bus(struct mmc_host *host, const struct mmc_bus_ops *ops);
@@ -172,4 +173,12 @@ static inline bool mmc_cache_enabled(struct mmc_host *host)
 	return false;
 }
 
+static inline int mmc_flush_cache(struct mmc_host *host)
+{
+	if (host->bus_ops->flush_cache)
+		return host->bus_ops->flush_cache(host);
+
+	return 0;
+}
+
 #endif
diff --git a/drivers/mmc/core/debugfs.c b/drivers/mmc/core/debugfs.c
index 9ec84c86c46a..3fdbc801e64a 100644
--- a/drivers/mmc/core/debugfs.c
+++ b/drivers/mmc/core/debugfs.c
@@ -26,6 +26,7 @@
 static DECLARE_FAULT_ATTR(fail_default_attr);
 static char *fail_request;
 module_param(fail_request, charp, 0);
+MODULE_PARM_DESC(fail_request, "default fault injection attributes");
 
 #endif /* CONFIG_FAIL_MMC_REQUEST */
 
diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
index 0b0577990ddc..eda4a1892c33 100644
--- a/drivers/mmc/core/host.c
+++ b/drivers/mmc/core/host.c
@@ -388,6 +388,9 @@ int mmc_of_parse(struct mmc_host *host)
 		host->caps2 |= MMC_CAP2_NO_SD;
 	if (device_property_read_bool(dev, "no-mmc"))
 		host->caps2 |= MMC_CAP2_NO_MMC;
+	if (device_property_read_bool(dev, "no-mmc-hs400"))
+		host->caps2 &= ~(MMC_CAP2_HS400_1_8V | MMC_CAP2_HS400_1_2V |
+				 MMC_CAP2_HS400_ES);
 
 	/* Must be after "non-removable" check */
 	if (device_property_read_u32(dev, "fixed-emmc-driver-type", &drv_type) == 0) {
diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c
index 8674c3e0c02c..838726b68ff3 100644
--- a/drivers/mmc/core/mmc.c
+++ b/drivers/mmc/core/mmc.c
@@ -28,6 +28,7 @@
 
 #define DEFAULT_CMD6_TIMEOUT_MS	500
 #define MIN_CACHE_EN_TIMEOUT_MS 1600
+#define CACHE_FLUSH_TIMEOUT_MS 30000 /* 30s */
 
 static const unsigned int tran_exp[] = {
 	10000,		100000,		1000000,	10000000,
@@ -1905,11 +1906,20 @@ static int mmc_can_sleep(struct mmc_card *card)
 	return card->ext_csd.rev >= 3;
 }
 
+static int mmc_sleep_busy_cb(void *cb_data, bool *busy)
+{
+	struct mmc_host *host = cb_data;
+
+	*busy = host->ops->card_busy(host);
+	return 0;
+}
+
 static int mmc_sleep(struct mmc_host *host)
 {
 	struct mmc_command cmd = {};
 	struct mmc_card *card = host->card;
 	unsigned int timeout_ms = DIV_ROUND_UP(card->ext_csd.sa_timeout, 10000);
+	bool use_r1b_resp;
 	int err;
 
 	/* Re-tuning can't be done once the card is deselected */
@@ -1922,35 +1932,27 @@ static int mmc_sleep(struct mmc_host *host)
 	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. Note, some hosts requires R1B, which also
-	 * means they are on their own when it comes to deal with the busy
-	 * timeout.
-	 */
-	if (!(host->caps & MMC_CAP_NEED_RSP_BUSY) && 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;
-	}
+	use_r1b_resp = mmc_prepare_busy_cmd(host, &cmd, timeout_ms);
 
 	err = mmc_wait_for_cmd(host, &cmd, 0);
 	if (err)
 		goto out_release;
 
 	/*
-	 * 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 the host does not wait while the card signals busy, then we can
+	 * try to poll, but only if the host supports HW polling, as the
+	 * SEND_STATUS cmd is not allowed. If we can't poll, then we simply need
+	 * to wait the sleep/awake timeout.
 	 */
-	if (!cmd.busy_timeout || !(host->caps & MMC_CAP_WAIT_WHILE_BUSY))
+	if (host->caps & MMC_CAP_WAIT_WHILE_BUSY && use_r1b_resp)
+		goto out_release;
+
+	if (!host->ops->card_busy) {
 		mmc_delay(timeout_ms);
+		goto out_release;
+	}
+
+	err = __mmc_poll_for_busy(card, timeout_ms, &mmc_sleep_busy_cb, host);
 
 out_release:
 	mmc_retune_release(host);
@@ -2035,6 +2037,25 @@ static bool _mmc_cache_enabled(struct mmc_host *host)
 	       host->card->ext_csd.cache_ctrl & 1;
 }
 
+/*
+ * Flush the internal cache of the eMMC to non-volatile storage.
+ */
+static int _mmc_flush_cache(struct mmc_host *host)
+{
+	int err = 0;
+
+	if (_mmc_cache_enabled(host)) {
+		err = mmc_switch(host->card, EXT_CSD_CMD_SET_NORMAL,
+				 EXT_CSD_FLUSH_CACHE, 1,
+				 CACHE_FLUSH_TIMEOUT_MS);
+		if (err)
+			pr_err("%s: cache flush error %d\n",
+			       mmc_hostname(host), err);
+	}
+
+	return err;
+}
+
 static int _mmc_suspend(struct mmc_host *host, bool is_suspend)
 {
 	int err = 0;
@@ -2046,7 +2067,7 @@ static int _mmc_suspend(struct mmc_host *host, bool is_suspend)
 	if (mmc_card_suspended(host->card))
 		goto out;
 
-	err = mmc_flush_cache(host->card);
+	err = _mmc_flush_cache(host);
 	if (err)
 		goto out;
 
@@ -2187,7 +2208,7 @@ static int _mmc_hw_reset(struct mmc_host *host)
 	 * In the case of recovery, we can't expect flushing the cache to work
 	 * always, but we have a go and ignore errors.
 	 */
-	mmc_flush_cache(host->card);
+	_mmc_flush_cache(host);
 
 	if ((host->caps & MMC_CAP_HW_RESET) && host->ops->hw_reset &&
 	     mmc_can_reset(card)) {
@@ -2215,6 +2236,7 @@ static const struct mmc_bus_ops mmc_ops = {
 	.shutdown = mmc_shutdown,
 	.hw_reset = _mmc_hw_reset,
 	.cache_enabled = _mmc_cache_enabled,
+	.flush_cache = _mmc_flush_cache,
 };
 
 /*
diff --git a/drivers/mmc/core/mmc_ops.c b/drivers/mmc/core/mmc_ops.c
index 5756781fef37..973756ed4016 100644
--- a/drivers/mmc/core/mmc_ops.c
+++ b/drivers/mmc/core/mmc_ops.c
@@ -20,7 +20,6 @@
 #include "mmc_ops.h"
 
 #define MMC_BKOPS_TIMEOUT_MS		(120 * 1000) /* 120s */
-#define MMC_CACHE_FLUSH_TIMEOUT_MS	(30 * 1000) /* 30s */
 #define MMC_SANITIZE_TIMEOUT_MS		(240 * 1000) /* 240s */
 
 static const u8 tuning_blk_pattern_4bit[] = {
@@ -53,6 +52,12 @@ static const u8 tuning_blk_pattern_8bit[] = {
 	0xff, 0x77, 0x77, 0xff, 0x77, 0xbb, 0xdd, 0xee,
 };
 
+struct mmc_busy_data {
+	struct mmc_card *card;
+	bool retry_crc_err;
+	enum mmc_busy_cmd busy_cmd;
+};
+
 int __mmc_send_status(struct mmc_card *card, u32 *status, unsigned int retries)
 {
 	int err;
@@ -246,9 +251,8 @@ mmc_send_cxd_native(struct mmc_host *host, u32 arg, u32 *cxd, int opcode)
  * 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)
+int mmc_send_adtc_data(struct mmc_card *card, struct mmc_host *host, u32 opcode,
+		       u32 args, void *buf, unsigned len)
 {
 	struct mmc_request mrq = {};
 	struct mmc_command cmd = {};
@@ -259,7 +263,7 @@ mmc_send_cxd_data(struct mmc_card *card, struct mmc_host *host,
 	mrq.data = &data;
 
 	cmd.opcode = opcode;
-	cmd.arg = 0;
+	cmd.arg = args;
 
 	/* 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
@@ -305,7 +309,7 @@ static int mmc_spi_send_cxd(struct mmc_host *host, u32 *cxd, u32 opcode)
 	if (!cxd_tmp)
 		return -ENOMEM;
 
-	ret = mmc_send_cxd_data(NULL, host, opcode, cxd_tmp, 16);
+	ret = mmc_send_adtc_data(NULL, host, opcode, 0, cxd_tmp, 16);
 	if (ret)
 		goto err;
 
@@ -353,7 +357,7 @@ int mmc_get_ext_csd(struct mmc_card *card, u8 **new_ext_csd)
 	if (!ext_csd)
 		return -ENOMEM;
 
-	err = mmc_send_cxd_data(card, card->host, MMC_SEND_EXT_CSD, ext_csd,
+	err = mmc_send_adtc_data(card, card->host, MMC_SEND_EXT_CSD, 0, ext_csd,
 				512);
 	if (err)
 		kfree(ext_csd);
@@ -424,10 +428,10 @@ int mmc_switch_status(struct mmc_card *card, bool crc_err_fatal)
 	return mmc_switch_status_error(card->host, status);
 }
 
-static int mmc_busy_status(struct mmc_card *card, bool retry_crc_err,
-			   enum mmc_busy_cmd busy_cmd, bool *busy)
+static int mmc_busy_cb(void *cb_data, bool *busy)
 {
-	struct mmc_host *host = card->host;
+	struct mmc_busy_data *data = cb_data;
+	struct mmc_host *host = data->card->host;
 	u32 status = 0;
 	int err;
 
@@ -436,22 +440,23 @@ static int mmc_busy_status(struct mmc_card *card, bool retry_crc_err,
 		return 0;
 	}
 
-	err = mmc_send_status(card, &status);
-	if (retry_crc_err && err == -EILSEQ) {
+	err = mmc_send_status(data->card, &status);
+	if (data->retry_crc_err && err == -EILSEQ) {
 		*busy = true;
 		return 0;
 	}
 	if (err)
 		return err;
 
-	switch (busy_cmd) {
+	switch (data->busy_cmd) {
 	case MMC_BUSY_CMD6:
-		err = mmc_switch_status_error(card->host, status);
+		err = mmc_switch_status_error(host, status);
 		break;
 	case MMC_BUSY_ERASE:
 		err = R1_STATUS(status) ? -EIO : 0;
 		break;
 	case MMC_BUSY_HPI:
+	case MMC_BUSY_EXTR_SINGLE:
 		break;
 	default:
 		err = -EINVAL;
@@ -464,9 +469,9 @@ static int mmc_busy_status(struct mmc_card *card, bool retry_crc_err,
 	return 0;
 }
 
-static int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
-			       bool send_status, bool retry_crc_err,
-			       enum mmc_busy_cmd busy_cmd)
+int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
+			int (*busy_cb)(void *cb_data, bool *busy),
+			void *cb_data)
 {
 	struct mmc_host *host = card->host;
 	int err;
@@ -475,16 +480,6 @@ static int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
 	bool expired = false;
 	bool busy = false;
 
-	/*
-	 * In cases when not allowed to poll by using CMD13 or because we aren't
-	 * capable of polling by using ->card_busy(), then rely on waiting the
-	 * stated timeout to be sufficient.
-	 */
-	if (!send_status && !host->ops->card_busy) {
-		mmc_delay(timeout_ms);
-		return 0;
-	}
-
 	timeout = jiffies + msecs_to_jiffies(timeout_ms) + 1;
 	do {
 		/*
@@ -493,7 +488,7 @@ static int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
 		 */
 		expired = time_after(jiffies, timeout);
 
-		err = mmc_busy_status(card, retry_crc_err, busy_cmd, &busy);
+		err = (*busy_cb)(cb_data, &busy);
 		if (err)
 			return err;
 
@@ -516,9 +511,36 @@ static int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
 }
 
 int mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
-		      enum mmc_busy_cmd busy_cmd)
+		      bool retry_crc_err, enum mmc_busy_cmd busy_cmd)
+{
+	struct mmc_busy_data cb_data;
+
+	cb_data.card = card;
+	cb_data.retry_crc_err = retry_crc_err;
+	cb_data.busy_cmd = busy_cmd;
+
+	return __mmc_poll_for_busy(card, timeout_ms, &mmc_busy_cb, &cb_data);
+}
+
+bool mmc_prepare_busy_cmd(struct mmc_host *host, struct mmc_command *cmd,
+			  unsigned int timeout_ms)
 {
-	return __mmc_poll_for_busy(card, timeout_ms, true, false, busy_cmd);
+	/*
+	 * If the max_busy_timeout of the host is specified, make sure it's
+	 * enough to fit the used timeout_ms. In case it's not, let's instruct
+	 * the host to avoid HW busy detection, by converting to a R1 response
+	 * instead of a R1B. Note, some hosts requires R1B, which also means
+	 * they are on their own when it comes to deal with the busy timeout.
+	 */
+	if (!(host->caps & MMC_CAP_NEED_RSP_BUSY) && host->max_busy_timeout &&
+	    (timeout_ms > host->max_busy_timeout)) {
+		cmd->flags = MMC_CMD_AC | MMC_RSP_SPI_R1 | MMC_RSP_R1;
+		return false;
+	}
+
+	cmd->flags = MMC_CMD_AC | MMC_RSP_SPI_R1B | MMC_RSP_R1B;
+	cmd->busy_timeout = timeout_ms;
+	return true;
 }
 
 /**
@@ -543,7 +565,7 @@ int __mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
 	struct mmc_host *host = card->host;
 	int err;
 	struct mmc_command cmd = {};
-	bool use_r1b_resp = true;
+	bool use_r1b_resp;
 	unsigned char old_timing = host->ios.timing;
 
 	mmc_retune_hold(host);
@@ -554,29 +576,12 @@ int __mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
 		timeout_ms = card->ext_csd.generic_cmd6_time;
 	}
 
-	/*
-	 * If the max_busy_timeout of the host is specified, make sure it's
-	 * enough to fit the used timeout_ms. In case it's not, let's instruct
-	 * the host to avoid HW busy detection, by converting to a R1 response
-	 * instead of a R1B. Note, some hosts requires R1B, which also means
-	 * they are on their own when it comes to deal with the busy timeout.
-	 */
-	if (!(host->caps & MMC_CAP_NEED_RSP_BUSY) && 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_CMD_AC;
-	if (use_r1b_resp) {
-		cmd.flags |= MMC_RSP_SPI_R1B | MMC_RSP_R1B;
-		cmd.busy_timeout = timeout_ms;
-	} else {
-		cmd.flags |= MMC_RSP_SPI_R1 | MMC_RSP_R1;
-	}
+	use_r1b_resp = mmc_prepare_busy_cmd(host, &cmd, timeout_ms);
 
 	err = mmc_wait_for_cmd(host, &cmd, retries);
 	if (err)
@@ -587,9 +592,18 @@ int __mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
 		mmc_host_is_spi(host))
 		goto out_tim;
 
+	/*
+	 * If the host doesn't support HW polling via the ->card_busy() ops and
+	 * when it's not allowed to poll by using CMD13, then we need to rely on
+	 * waiting the stated timeout to be sufficient.
+	 */
+	if (!send_status && !host->ops->card_busy) {
+		mmc_delay(timeout_ms);
+		goto out_tim;
+	}
+
 	/* Let's try to poll to find out when the command is completed. */
-	err = __mmc_poll_for_busy(card, timeout_ms, send_status, retry_crc_err,
-				  MMC_BUSY_CMD6);
+	err = mmc_poll_for_busy(card, timeout_ms, retry_crc_err, MMC_BUSY_CMD6);
 	if (err)
 		goto out;
 
@@ -686,7 +700,7 @@ out:
 }
 EXPORT_SYMBOL_GPL(mmc_send_tuning);
 
-int mmc_abort_tuning(struct mmc_host *host, u32 opcode)
+int mmc_send_abort_tuning(struct mmc_host *host, u32 opcode)
 {
 	struct mmc_command cmd = {};
 
@@ -709,7 +723,7 @@ int mmc_abort_tuning(struct mmc_host *host, u32 opcode)
 
 	return mmc_wait_for_cmd(host, &cmd, 0);
 }
-EXPORT_SYMBOL_GPL(mmc_abort_tuning);
+EXPORT_SYMBOL_GPL(mmc_send_abort_tuning);
 
 static int
 mmc_send_bus_test(struct mmc_card *card, struct mmc_host *host, u8 opcode,
@@ -813,28 +827,17 @@ static int mmc_send_hpi_cmd(struct mmc_card *card)
 {
 	unsigned int busy_timeout_ms = card->ext_csd.out_of_int_time;
 	struct mmc_host *host = card->host;
-	bool use_r1b_resp = true;
+	bool use_r1b_resp = false;
 	struct mmc_command cmd = {};
 	int err;
 
 	cmd.opcode = card->ext_csd.hpi_cmd;
 	cmd.arg = card->rca << 16 | 1;
+	cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
 
-	/*
-	 * Make sure the host's max_busy_timeout fit the needed timeout for HPI.
-	 * In case it doesn't, let's instruct the host to avoid HW busy
-	 * detection, by using a R1 response instead of R1B.
-	 */
-	if (host->max_busy_timeout && busy_timeout_ms > host->max_busy_timeout)
-		use_r1b_resp = false;
-
-	if (cmd.opcode == MMC_STOP_TRANSMISSION && use_r1b_resp) {
-		cmd.flags = MMC_RSP_R1B | MMC_CMD_AC;
-		cmd.busy_timeout = busy_timeout_ms;
-	} else {
-		cmd.flags = MMC_RSP_R1 | MMC_CMD_AC;
-		use_r1b_resp = false;
-	}
+	if (cmd.opcode == MMC_STOP_TRANSMISSION)
+		use_r1b_resp = mmc_prepare_busy_cmd(host, &cmd,
+						    busy_timeout_ms);
 
 	err = mmc_wait_for_cmd(host, &cmd, 0);
 	if (err) {
@@ -848,7 +851,7 @@ static int mmc_send_hpi_cmd(struct mmc_card *card)
 		return 0;
 
 	/* Let's poll to find out when the HPI request completes. */
-	return mmc_poll_for_busy(card, busy_timeout_ms, MMC_BUSY_HPI);
+	return mmc_poll_for_busy(card, busy_timeout_ms, false, MMC_BUSY_HPI);
 }
 
 /**
@@ -961,26 +964,6 @@ void mmc_run_bkops(struct mmc_card *card)
 }
 EXPORT_SYMBOL(mmc_run_bkops);
 
-/*
- * Flush the cache to the non-volatile storage.
- */
-int mmc_flush_cache(struct mmc_card *card)
-{
-	int err = 0;
-
-	if (mmc_cache_enabled(card->host)) {
-		err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL,
-				 EXT_CSD_FLUSH_CACHE, 1,
-				 MMC_CACHE_FLUSH_TIMEOUT_MS);
-		if (err)
-			pr_err("%s: cache flush error %d\n",
-					mmc_hostname(card->host), err);
-	}
-
-	return err;
-}
-EXPORT_SYMBOL(mmc_flush_cache);
-
 static int mmc_cmdq_switch(struct mmc_card *card, bool enable)
 {
 	u8 val = enable ? EXT_CSD_CMDQ_MODE_ENABLED : 0;
diff --git a/drivers/mmc/core/mmc_ops.h b/drivers/mmc/core/mmc_ops.h
index 7bc1cfb0654c..41ab4f573a31 100644
--- a/drivers/mmc/core/mmc_ops.h
+++ b/drivers/mmc/core/mmc_ops.h
@@ -14,10 +14,12 @@ enum mmc_busy_cmd {
 	MMC_BUSY_CMD6,
 	MMC_BUSY_ERASE,
 	MMC_BUSY_HPI,
+	MMC_BUSY_EXTR_SINGLE,
 };
 
 struct mmc_host;
 struct mmc_card;
+struct mmc_command;
 
 int mmc_select_card(struct mmc_card *card);
 int mmc_deselect_cards(struct mmc_host *host);
@@ -25,6 +27,8 @@ int mmc_set_dsr(struct mmc_host *host);
 int mmc_go_idle(struct mmc_host *host);
 int mmc_send_op_cond(struct mmc_host *host, u32 ocr, u32 *rocr);
 int mmc_set_relative_addr(struct mmc_card *card);
+int mmc_send_adtc_data(struct mmc_card *card, struct mmc_host *host, u32 opcode,
+		       u32 args, void *buf, unsigned len);
 int mmc_send_csd(struct mmc_card *card, u32 *csd);
 int __mmc_send_status(struct mmc_card *card, u32 *status, unsigned int retries);
 int mmc_send_status(struct mmc_card *card, u32 *status);
@@ -35,15 +39,19 @@ int mmc_bus_test(struct mmc_card *card, u8 bus_width);
 int mmc_can_ext_csd(struct mmc_card *card);
 int mmc_get_ext_csd(struct mmc_card *card, u8 **new_ext_csd);
 int mmc_switch_status(struct mmc_card *card, bool crc_err_fatal);
+bool mmc_prepare_busy_cmd(struct mmc_host *host, struct mmc_command *cmd,
+			  unsigned int timeout_ms);
+int __mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
+			int (*busy_cb)(void *cb_data, bool *busy),
+			void *cb_data);
 int mmc_poll_for_busy(struct mmc_card *card, unsigned int timeout_ms,
-		      enum mmc_busy_cmd busy_cmd);
+		      bool retry_crc_err, enum mmc_busy_cmd busy_cmd);
 int __mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
 		unsigned int timeout_ms, unsigned char timing,
 		bool send_status, bool retry_crc_err, unsigned int retries);
 int mmc_switch(struct mmc_card *card, u8 set, u8 index, u8 value,
 		unsigned int timeout_ms);
 void mmc_run_bkops(struct mmc_card *card);
-int mmc_flush_cache(struct mmc_card *card);
 int mmc_cmdq_enable(struct mmc_card *card);
 int mmc_cmdq_disable(struct mmc_card *card);
 int mmc_sanitize(struct mmc_card *card, unsigned int timeout_ms);
diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
index 2c48d6504101..4646b7a03db6 100644
--- a/drivers/mmc/core/sd.c
+++ b/drivers/mmc/core/sd.c
@@ -66,6 +66,14 @@ static const unsigned int sd_au_size[] = {
 		__res & __mask;						\
 	})
 
+#define SD_POWEROFF_NOTIFY_TIMEOUT_MS 2000
+#define SD_WRITE_EXTR_SINGLE_TIMEOUT_MS 1000
+
+struct sd_busy_data {
+	struct mmc_card *card;
+	u8 *reg_buf;
+};
+
 /*
  * Given the decoded CSD structure, decode the raw CID to our CID structure.
  */
@@ -222,7 +230,9 @@ static int mmc_decode_scr(struct mmc_card *card)
 	else
 		card->erased_byte = 0x0;
 
-	if (scr->sda_spec3)
+	if (scr->sda_spec4)
+		scr->cmds = UNSTUFF_BITS(resp, 32, 4);
+	else if (scr->sda_spec3)
 		scr->cmds = UNSTUFF_BITS(resp, 32, 2);
 
 	/* SD Spec says: any SD Card shall set at least bits 0 and 2 */
@@ -847,11 +857,13 @@ try_again:
 		return err;
 
 	/*
-	 * In case CCS and S18A in the response is set, start Signal Voltage
-	 * Switch procedure. SPI mode doesn't support CMD11.
+	 * In case the S18A bit is set in the response, let's start the signal
+	 * voltage switch procedure. SPI mode doesn't support CMD11.
+	 * Note that, according to the spec, the S18A bit is not valid unless
+	 * the CCS bit is set as well. We deliberately deviate from the spec in
+	 * regards to this, which allows UHS-I to be supported for SDSC cards.
 	 */
-	if (!mmc_host_is_spi(host) && rocr &&
-	   ((*rocr & 0x41000000) == 0x41000000)) {
+	if (!mmc_host_is_spi(host) && rocr && (*rocr & 0x01000000)) {
 		err = mmc_set_uhs_voltage(host, pocr);
 		if (err == -EAGAIN) {
 			retries--;
@@ -994,6 +1006,380 @@ static bool mmc_sd_card_using_v18(struct mmc_card *card)
 	       (SD_MODE_UHS_SDR50 | SD_MODE_UHS_SDR104 | SD_MODE_UHS_DDR50);
 }
 
+static int sd_write_ext_reg(struct mmc_card *card, u8 fno, u8 page, u16 offset,
+			    u8 reg_data)
+{
+	struct mmc_host *host = card->host;
+	struct mmc_request mrq = {};
+	struct mmc_command cmd = {};
+	struct mmc_data data = {};
+	struct scatterlist sg;
+	u8 *reg_buf;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	mrq.cmd = &cmd;
+	mrq.data = &data;
+
+	/*
+	 * Arguments of CMD49:
+	 * [31:31] MIO (0 = memory).
+	 * [30:27] FNO (function number).
+	 * [26:26] MW - mask write mode (0 = disable).
+	 * [25:18] page number.
+	 * [17:9] offset address.
+	 * [8:0] length (0 = 1 byte).
+	 */
+	cmd.arg = fno << 27 | page << 18 | offset << 9;
+
+	/* The first byte in the buffer is the data to be written. */
+	reg_buf[0] = reg_data;
+
+	data.flags = MMC_DATA_WRITE;
+	data.blksz = 512;
+	data.blocks = 1;
+	data.sg = &sg;
+	data.sg_len = 1;
+	sg_init_one(&sg, reg_buf, 512);
+
+	cmd.opcode = SD_WRITE_EXTR_SINGLE;
+	cmd.flags = MMC_RSP_R1 | MMC_CMD_ADTC;
+
+	mmc_set_data_timeout(&data, card);
+	mmc_wait_for_req(host, &mrq);
+
+	kfree(reg_buf);
+
+	/*
+	 * Note that, the SD card is allowed to signal busy on DAT0 up to 1s
+	 * after the CMD49. Although, let's leave this to be managed by the
+	 * caller.
+	 */
+
+	if (cmd.error)
+		return cmd.error;
+	if (data.error)
+		return data.error;
+
+	return 0;
+}
+
+static int sd_read_ext_reg(struct mmc_card *card, u8 fno, u8 page,
+			   u16 offset, u16 len, u8 *reg_buf)
+{
+	u32 cmd_args;
+
+	/*
+	 * Command arguments of CMD48:
+	 * [31:31] MIO (0 = memory).
+	 * [30:27] FNO (function number).
+	 * [26:26] reserved (0).
+	 * [25:18] page number.
+	 * [17:9] offset address.
+	 * [8:0] length (0 = 1 byte, 1ff = 512 bytes).
+	 */
+	cmd_args = fno << 27 | page << 18 | offset << 9 | (len -1);
+
+	return mmc_send_adtc_data(card, card->host, SD_READ_EXTR_SINGLE,
+				  cmd_args, reg_buf, 512);
+}
+
+static int sd_parse_ext_reg_power(struct mmc_card *card, u8 fno, u8 page,
+				  u16 offset)
+{
+	int err;
+	u8 *reg_buf;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	/* Read the extension register for power management function. */
+	err = sd_read_ext_reg(card, fno, page, offset, 512, reg_buf);
+	if (err) {
+		pr_warn("%s: error %d reading PM func of ext reg\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	/* PM revision consists of 4 bits. */
+	card->ext_power.rev = reg_buf[0] & 0xf;
+
+	/* Power Off Notification support at bit 4. */
+	if (reg_buf[1] & BIT(4))
+		card->ext_power.feature_support |= SD_EXT_POWER_OFF_NOTIFY;
+
+	/* Power Sustenance support at bit 5. */
+	if (reg_buf[1] & BIT(5))
+		card->ext_power.feature_support |= SD_EXT_POWER_SUSTENANCE;
+
+	/* Power Down Mode support at bit 6. */
+	if (reg_buf[1] & BIT(6))
+		card->ext_power.feature_support |= SD_EXT_POWER_DOWN_MODE;
+
+	card->ext_power.fno = fno;
+	card->ext_power.page = page;
+	card->ext_power.offset = offset;
+
+out:
+	kfree(reg_buf);
+	return err;
+}
+
+static int sd_parse_ext_reg_perf(struct mmc_card *card, u8 fno, u8 page,
+				 u16 offset)
+{
+	int err;
+	u8 *reg_buf;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	err = sd_read_ext_reg(card, fno, page, offset, 512, reg_buf);
+	if (err) {
+		pr_warn("%s: error %d reading PERF func of ext reg\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	/* PERF revision. */
+	card->ext_perf.rev = reg_buf[0];
+
+	/* FX_EVENT support at bit 0. */
+	if (reg_buf[1] & BIT(0))
+		card->ext_perf.feature_support |= SD_EXT_PERF_FX_EVENT;
+
+	/* Card initiated self-maintenance support at bit 0. */
+	if (reg_buf[2] & BIT(0))
+		card->ext_perf.feature_support |= SD_EXT_PERF_CARD_MAINT;
+
+	/* Host initiated self-maintenance support at bit 1. */
+	if (reg_buf[2] & BIT(1))
+		card->ext_perf.feature_support |= SD_EXT_PERF_HOST_MAINT;
+
+	/* Cache support at bit 0. */
+	if (reg_buf[4] & BIT(0))
+		card->ext_perf.feature_support |= SD_EXT_PERF_CACHE;
+
+	/* Command queue support indicated via queue depth bits (0 to 4). */
+	if (reg_buf[6] & 0x1f)
+		card->ext_perf.feature_support |= SD_EXT_PERF_CMD_QUEUE;
+
+	card->ext_perf.fno = fno;
+	card->ext_perf.page = page;
+	card->ext_perf.offset = offset;
+
+out:
+	kfree(reg_buf);
+	return err;
+}
+
+static int sd_parse_ext_reg(struct mmc_card *card, u8 *gen_info_buf,
+			    u16 *next_ext_addr)
+{
+	u8 num_regs, fno, page;
+	u16 sfc, offset, ext = *next_ext_addr;
+	u32 reg_addr;
+
+	/*
+	 * Parse only one register set per extension, as that is sufficient to
+	 * support the standard functions. This means another 48 bytes in the
+	 * buffer must be available.
+	 */
+	if (ext + 48 > 512)
+		return -EFAULT;
+
+	/* Standard Function Code */
+	memcpy(&sfc, &gen_info_buf[ext], 2);
+
+	/* Address to the next extension. */
+	memcpy(next_ext_addr, &gen_info_buf[ext + 40], 2);
+
+	/* Number of registers for this extension. */
+	num_regs = gen_info_buf[ext + 42];
+
+	/* We support only one register per extension. */
+	if (num_regs != 1)
+		return 0;
+
+	/* Extension register address. */
+	memcpy(&reg_addr, &gen_info_buf[ext + 44], 4);
+
+	/* 9 bits (0 to 8) contains the offset address. */
+	offset = reg_addr & 0x1ff;
+
+	/* 8 bits (9 to 16) contains the page number. */
+	page = reg_addr >> 9 & 0xff ;
+
+	/* 4 bits (18 to 21) contains the function number. */
+	fno = reg_addr >> 18 & 0xf;
+
+	/* Standard Function Code for power management. */
+	if (sfc == 0x1)
+		return sd_parse_ext_reg_power(card, fno, page, offset);
+
+	/* Standard Function Code for performance enhancement. */
+	if (sfc == 0x2)
+		return sd_parse_ext_reg_perf(card, fno, page, offset);
+
+	return 0;
+}
+
+static int sd_read_ext_regs(struct mmc_card *card)
+{
+	int err, i;
+	u8 num_ext, *gen_info_buf;
+	u16 rev, len, next_ext_addr;
+
+	if (mmc_host_is_spi(card->host))
+		return 0;
+
+	if (!(card->scr.cmds & SD_SCR_CMD48_SUPPORT))
+		return 0;
+
+	gen_info_buf = kzalloc(512, GFP_KERNEL);
+	if (!gen_info_buf)
+		return -ENOMEM;
+
+	/*
+	 * Read 512 bytes of general info, which is found at function number 0,
+	 * at page 0 and with no offset.
+	 */
+	err = sd_read_ext_reg(card, 0, 0, 0, 512, gen_info_buf);
+	if (err) {
+		pr_warn("%s: error %d reading general info of SD ext reg\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	/* General info structure revision. */
+	memcpy(&rev, &gen_info_buf[0], 2);
+
+	/* Length of general info in bytes. */
+	memcpy(&len, &gen_info_buf[2], 2);
+
+	/* Number of extensions to be find. */
+	num_ext = gen_info_buf[4];
+
+	/* We support revision 0, but limit it to 512 bytes for simplicity. */
+	if (rev != 0 || len > 512) {
+		pr_warn("%s: non-supported SD ext reg layout\n",
+			mmc_hostname(card->host));
+		goto out;
+	}
+
+	/*
+	 * Parse the extension registers. The first extension should start
+	 * immediately after the general info header (16 bytes).
+	 */
+	next_ext_addr = 16;
+	for (i = 0; i < num_ext; i++) {
+		err = sd_parse_ext_reg(card, gen_info_buf, &next_ext_addr);
+		if (err) {
+			pr_warn("%s: error %d parsing SD ext reg\n",
+				mmc_hostname(card->host), err);
+			goto out;
+		}
+	}
+
+out:
+	kfree(gen_info_buf);
+	return err;
+}
+
+static bool sd_cache_enabled(struct mmc_host *host)
+{
+	return host->card->ext_perf.feature_enabled & SD_EXT_PERF_CACHE;
+}
+
+static int sd_flush_cache(struct mmc_host *host)
+{
+	struct mmc_card *card = host->card;
+	u8 *reg_buf, fno, page;
+	u16 offset;
+	int err;
+
+	if (!sd_cache_enabled(host))
+		return 0;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	/*
+	 * Set Flush Cache at bit 0 in the performance enhancement register at
+	 * 261 bytes offset.
+	 */
+	fno = card->ext_perf.fno;
+	page = card->ext_perf.page;
+	offset = card->ext_perf.offset + 261;
+
+	err = sd_write_ext_reg(card, fno, page, offset, BIT(0));
+	if (err) {
+		pr_warn("%s: error %d writing Cache Flush bit\n",
+			mmc_hostname(host), err);
+		goto out;
+	}
+
+	err = mmc_poll_for_busy(card, SD_WRITE_EXTR_SINGLE_TIMEOUT_MS, false,
+				MMC_BUSY_EXTR_SINGLE);
+	if (err)
+		goto out;
+
+	/*
+	 * Read the Flush Cache bit. The card shall reset it, to confirm that
+	 * it's has completed the flushing of the cache.
+	 */
+	err = sd_read_ext_reg(card, fno, page, offset, 1, reg_buf);
+	if (err) {
+		pr_warn("%s: error %d reading Cache Flush bit\n",
+			mmc_hostname(host), err);
+		goto out;
+	}
+
+	if (reg_buf[0] & BIT(0))
+		err = -ETIMEDOUT;
+out:
+	kfree(reg_buf);
+	return err;
+}
+
+static int sd_enable_cache(struct mmc_card *card)
+{
+	u8 *reg_buf;
+	int err;
+
+	card->ext_perf.feature_enabled &= ~SD_EXT_PERF_CACHE;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	/*
+	 * Set Cache Enable at bit 0 in the performance enhancement register at
+	 * 260 bytes offset.
+	 */
+	err = sd_write_ext_reg(card, card->ext_perf.fno, card->ext_perf.page,
+			       card->ext_perf.offset + 260, BIT(0));
+	if (err) {
+		pr_warn("%s: error %d writing Cache Enable bit\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	err = mmc_poll_for_busy(card, SD_WRITE_EXTR_SINGLE_TIMEOUT_MS, false,
+				MMC_BUSY_EXTR_SINGLE);
+	if (!err)
+		card->ext_perf.feature_enabled |= SD_EXT_PERF_CACHE;
+
+out:
+	kfree(reg_buf);
+	return err;
+}
+
 /*
  * Handle the detection and initialisation of a card.
  *
@@ -1142,6 +1528,20 @@ retry:
 		}
 	}
 
+	if (!oldcard) {
+		/* Read/parse the extension registers. */
+		err = sd_read_ext_regs(card);
+		if (err)
+			goto free_card;
+	}
+
+	/* Enable internal SD cache if supported. */
+	if (card->ext_perf.feature_support & SD_EXT_PERF_CACHE) {
+		err = sd_enable_cache(card);
+		if (err)
+			goto free_card;
+	}
+
 	if (host->cqe_ops && !host->cqe_enabled) {
 		err = host->cqe_ops->cqe_enable(host, card);
 		if (!err) {
@@ -1213,21 +1613,84 @@ static void mmc_sd_detect(struct mmc_host *host)
 	}
 }
 
+static int sd_can_poweroff_notify(struct mmc_card *card)
+{
+	return card->ext_power.feature_support & SD_EXT_POWER_OFF_NOTIFY;
+}
+
+static int sd_busy_poweroff_notify_cb(void *cb_data, bool *busy)
+{
+	struct sd_busy_data *data = cb_data;
+	struct mmc_card *card = data->card;
+	int err;
+
+	/*
+	 * Read the status register for the power management function. It's at
+	 * one byte offset and is one byte long. The Power Off Notification
+	 * Ready is bit 0.
+	 */
+	err = sd_read_ext_reg(card, card->ext_power.fno, card->ext_power.page,
+			      card->ext_power.offset + 1, 1, data->reg_buf);
+	if (err) {
+		pr_warn("%s: error %d reading status reg of PM func\n",
+			mmc_hostname(card->host), err);
+		return err;
+	}
+
+	*busy = !(data->reg_buf[0] & BIT(0));
+	return 0;
+}
+
+static int sd_poweroff_notify(struct mmc_card *card)
+{
+	struct sd_busy_data cb_data;
+	u8 *reg_buf;
+	int err;
+
+	reg_buf = kzalloc(512, GFP_KERNEL);
+	if (!reg_buf)
+		return -ENOMEM;
+
+	/*
+	 * Set the Power Off Notification bit in the power management settings
+	 * register at 2 bytes offset.
+	 */
+	err = sd_write_ext_reg(card, card->ext_power.fno, card->ext_power.page,
+			       card->ext_power.offset + 2, BIT(0));
+	if (err) {
+		pr_warn("%s: error %d writing Power Off Notify bit\n",
+			mmc_hostname(card->host), err);
+		goto out;
+	}
+
+	cb_data.card = card;
+	cb_data.reg_buf = reg_buf;
+	err = __mmc_poll_for_busy(card, SD_POWEROFF_NOTIFY_TIMEOUT_MS,
+				  &sd_busy_poweroff_notify_cb, &cb_data);
+
+out:
+	kfree(reg_buf);
+	return err;
+}
+
 static int _mmc_sd_suspend(struct mmc_host *host)
 {
+	struct mmc_card *card = host->card;
 	int err = 0;
 
 	mmc_claim_host(host);
 
-	if (mmc_card_suspended(host->card))
+	if (mmc_card_suspended(card))
 		goto out;
 
-	if (!mmc_host_is_spi(host))
+	if (sd_can_poweroff_notify(card))
+		err = sd_poweroff_notify(card);
+	else if (!mmc_host_is_spi(host))
 		err = mmc_deselect_cards(host);
 
 	if (!err) {
 		mmc_power_off(host);
-		mmc_card_set_suspended(host->card);
+		mmc_card_set_suspended(card);
 	}
 
 out:
@@ -1331,6 +1794,8 @@ static const struct mmc_bus_ops mmc_sd_ops = {
 	.alive = mmc_sd_alive,
 	.shutdown = mmc_sd_suspend,
 	.hw_reset = mmc_sd_hw_reset,
+	.cache_enabled = sd_cache_enabled,
+	.flush_cache = sd_flush_cache,
 };
 
 /*
diff --git a/drivers/mmc/core/sd_ops.c b/drivers/mmc/core/sd_ops.c
index d61ff811218c..ef8d1dce5af1 100644
--- a/drivers/mmc/core/sd_ops.c
+++ b/drivers/mmc/core/sd_ops.c
@@ -17,6 +17,7 @@
 
 #include "core.h"
 #include "sd_ops.h"
+#include "mmc_ops.h"
 
 int mmc_app_cmd(struct mmc_host *host, struct mmc_card *card)
 {
@@ -309,43 +310,18 @@ int mmc_app_send_scr(struct mmc_card *card)
 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 scatterlist sg;
+	u32 cmd_args;
 
 	/* NOTE: caller guarantees resp is heap-allocated */
 
 	mode = !!mode;
 	value &= 0xF;
+	cmd_args = mode << 31 | 0x00FFFFFF;
+	cmd_args &= ~(0xF << (group * 4));
+	cmd_args |= value << (group * 4);
 
-	mrq.cmd = &cmd;
-	mrq.data = &data;
-
-	cmd.opcode = SD_SWITCH;
-	cmd.arg = mode << 31 | 0x00FFFFFF;
-	cmd.arg &= ~(0xF << (group * 4));
-	cmd.arg |= value << (group * 4);
-	cmd.flags = MMC_RSP_SPI_R1 | 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, resp, 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;
+	return mmc_send_adtc_data(card, card->host, SD_SWITCH, cmd_args, resp,
+				  64);
 }
 
 int mmc_app_sd_status(struct mmc_card *card, void *ssr)
diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c
index 3eb94ac2712e..68edf7a615be 100644
--- a/drivers/mmc/core/sdio.c
+++ b/drivers/mmc/core/sdio.c
@@ -937,11 +937,9 @@ static void mmc_sdio_detect(struct mmc_host *host)
 
 	/* 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);
+		err = pm_runtime_resume_and_get(&host->card->dev);
+		if (err < 0)
 			goto out;
-		}
 	}
 
 	mmc_claim_host(host);
diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index a4d4c757eea0..561184fa7eb9 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -412,7 +412,7 @@ config MMC_SDHCI_MILBEAUT
 
 config MMC_SDHCI_IPROC
 	tristate "SDHCI support for the BCM2835 & iProc SD/MMC Controller"
-	depends on ARCH_BCM2835 || ARCH_BCM_IPROC || COMPILE_TEST
+	depends on ARCH_BCM2835 || ARCH_BCM_IPROC || ARCH_BRCMSTB || COMPILE_TEST
 	depends on MMC_SDHCI_PLTFM
 	depends on OF || ACPI
 	default ARCH_BCM_IPROC
diff --git a/drivers/mmc/host/cqhci-core.c b/drivers/mmc/host/cqhci-core.c
index 93b0432bb601..38559a956330 100644
--- a/drivers/mmc/host/cqhci-core.c
+++ b/drivers/mmc/host/cqhci-core.c
@@ -45,17 +45,23 @@ static inline u8 *get_link_desc(struct cqhci_host *cq_host, u8 tag)
 	return desc + cq_host->task_desc_len;
 }
 
+static inline size_t get_trans_desc_offset(struct cqhci_host *cq_host, u8 tag)
+{
+	return cq_host->trans_desc_len * cq_host->mmc->max_segs * tag;
+}
+
 static inline dma_addr_t get_trans_desc_dma(struct cqhci_host *cq_host, u8 tag)
 {
-	return cq_host->trans_desc_dma_base +
-		(cq_host->mmc->max_segs * tag *
-		 cq_host->trans_desc_len);
+	size_t offset = get_trans_desc_offset(cq_host, tag);
+
+	return cq_host->trans_desc_dma_base + offset;
 }
 
 static inline u8 *get_trans_desc(struct cqhci_host *cq_host, u8 tag)
 {
-	return cq_host->trans_desc_base +
-		(cq_host->trans_desc_len * cq_host->mmc->max_segs * tag);
+	size_t offset = get_trans_desc_offset(cq_host, tag);
+
+	return cq_host->trans_desc_base + offset;
 }
 
 static void setup_trans_desc(struct cqhci_host *cq_host, u8 tag)
@@ -146,7 +152,7 @@ static void cqhci_dumpregs(struct cqhci_host *cq_host)
 }
 
 /*
- * The allocated descriptor table for task, link & transfer descritors
+ * The allocated descriptor table for task, link & transfer descriptors
  * looks like:
  * |----------|
  * |task desc |  |->|----------|
@@ -194,8 +200,7 @@ static int cqhci_host_alloc_tdl(struct cqhci_host *cq_host)
 
 	cq_host->desc_size = cq_host->slot_sz * cq_host->num_slots;
 
-	cq_host->data_size = cq_host->trans_desc_len * cq_host->mmc->max_segs *
-		cq_host->mmc->cqe_qdepth;
+	cq_host->data_size = get_trans_desc_offset(cq_host, cq_host->mmc->cqe_qdepth);
 
 	pr_debug("%s: cqhci: desc_size: %zu data_sz: %zu slot-sz: %d\n",
 		 mmc_hostname(cq_host->mmc), cq_host->desc_size, cq_host->data_size,
diff --git a/drivers/mmc/host/dw_mmc-pltfm.c b/drivers/mmc/host/dw_mmc-pltfm.c
index 73731cd3ba23..9901208be797 100644
--- a/drivers/mmc/host/dw_mmc-pltfm.c
+++ b/drivers/mmc/host/dw_mmc-pltfm.c
@@ -17,7 +17,6 @@
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
 #include <linux/of.h>
-#include <linux/clk.h>
 
 #include "dw_mmc.h"
 #include "dw_mmc-pltfm.h"
diff --git a/drivers/mmc/host/jz4740_mmc.c b/drivers/mmc/host/jz4740_mmc.c
index b3c636edbb46..0db17bcc9c16 100644
--- a/drivers/mmc/host/jz4740_mmc.c
+++ b/drivers/mmc/host/jz4740_mmc.c
@@ -674,7 +674,7 @@ static void jz4740_mmc_send_command(struct jz4740_mmc_host *host,
 			cmdat |= JZ_MMC_CMDAT_WRITE;
 		if (host->use_dma) {
 			/*
-			 * The 4780's MMC controller has integrated DMA ability
+			 * The JZ4780's MMC controller has integrated DMA ability
 			 * in addition to being able to use the external DMA
 			 * controller. It moves DMA control bits to a separate
 			 * register. The DMA_SEL bit chooses the external
@@ -866,7 +866,7 @@ static int jz4740_mmc_set_clock_rate(struct jz4740_mmc_host *host, int rate)
 	writew(div, host->base + JZ_REG_MMC_CLKRT);
 
 	if (real_rate > 25000000) {
-		if (host->version >= JZ_MMC_X1000) {
+		if (host->version >= JZ_MMC_JZ4780) {
 			writel(JZ_MMC_LPM_DRV_RISING_QTR_PHASE_DLY |
 				   JZ_MMC_LPM_SMP_RISING_QTR_OR_HALF_PHASE_DLY |
 				   JZ_MMC_LPM_LOW_POWER_MODE_EN,
@@ -959,6 +959,7 @@ static const struct of_device_id jz4740_mmc_of_match[] = {
 	{ .compatible = "ingenic,jz4740-mmc", .data = (void *) JZ_MMC_JZ4740 },
 	{ .compatible = "ingenic,jz4725b-mmc", .data = (void *)JZ_MMC_JZ4725B },
 	{ .compatible = "ingenic,jz4760-mmc", .data = (void *) JZ_MMC_JZ4760 },
+	{ .compatible = "ingenic,jz4775-mmc", .data = (void *) JZ_MMC_JZ4780 },
 	{ .compatible = "ingenic,jz4780-mmc", .data = (void *) JZ_MMC_JZ4780 },
 	{ .compatible = "ingenic,x1000-mmc", .data = (void *) JZ_MMC_X1000 },
 	{},
@@ -1013,7 +1014,6 @@ static int jz4740_mmc_probe(struct platform_device* pdev)
 	host->base = devm_ioremap_resource(&pdev->dev, host->mem_res);
 	if (IS_ERR(host->base)) {
 		ret = PTR_ERR(host->base);
-		dev_err(&pdev->dev, "Failed to ioremap base memory\n");
 		goto err_free_host;
 	}
 
diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c
index 9776a03a10f5..65c65bb5737f 100644
--- a/drivers/mmc/host/mmc_spi.c
+++ b/drivers/mmc/host/mmc_spi.c
@@ -504,7 +504,7 @@ mmc_spi_command_send(struct mmc_spi_host *host,
 		/* else:  R1 (most commands) */
 	}
 
-	dev_dbg(&host->spi->dev, "  mmc_spi: CMD%d, resp %s\n",
+	dev_dbg(&host->spi->dev, "  CMD%d, resp %s\n",
 		cmd->opcode, maptype(cmd));
 
 	/* send command, leaving chipselect active */
@@ -928,8 +928,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 		while (length) {
 			t->len = min(length, blk_size);
 
-			dev_dbg(&host->spi->dev,
-				"    mmc_spi: %s block, %d bytes\n",
+			dev_dbg(&host->spi->dev, "    %s block, %d bytes\n",
 				(direction == DMA_TO_DEVICE) ? "write" : "read",
 				t->len);
 
@@ -974,7 +973,7 @@ mmc_spi_data_do(struct mmc_spi_host *host, struct mmc_command *cmd,
 		int		tmp;
 		const unsigned	statlen = sizeof(scratch->status);
 
-		dev_dbg(&spi->dev, "    mmc_spi: STOP_TRAN\n");
+		dev_dbg(&spi->dev, "    STOP_TRAN\n");
 
 		/* Tweak the per-block message we set up earlier by morphing
 		 * it to hold single buffer with the token followed by some
@@ -1175,7 +1174,7 @@ static void mmc_spi_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 		canpower = host->pdata && host->pdata->setpower;
 
-		dev_dbg(&host->spi->dev, "mmc_spi: power %s (%d)%s\n",
+		dev_dbg(&host->spi->dev, "power %s (%d)%s\n",
 				mmc_powerstring(ios->power_mode),
 				ios->vdd,
 				canpower ? ", can switch" : "");
@@ -1248,8 +1247,7 @@ static void mmc_spi_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 		host->spi->max_speed_hz = ios->clock;
 		status = spi_setup(host->spi);
-		dev_dbg(&host->spi->dev,
-			"mmc_spi:  clock to %d Hz, %d\n",
+		dev_dbg(&host->spi->dev, "  clock to %d Hz, %d\n",
 			host->spi->max_speed_hz, status);
 	}
 }
diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
index 898ed1b023df..4dfc246c5f95 100644
--- a/drivers/mmc/host/mtk-sd.c
+++ b/drivers/mmc/host/mtk-sd.c
@@ -724,10 +724,8 @@ static inline void msdc_dma_setup(struct msdc_host *host, struct msdc_dma *dma,
 	writel(lower_32_bits(dma->gpd_addr), host->base + MSDC_DMA_SA);
 }
 
-static void msdc_prepare_data(struct msdc_host *host, struct mmc_request *mrq)
+static void msdc_prepare_data(struct msdc_host *host, struct mmc_data *data)
 {
-	struct mmc_data *data = mrq->data;
-
 	if (!(data->host_cookie & MSDC_PREPARE_FLAG)) {
 		data->host_cookie |= MSDC_PREPARE_FLAG;
 		data->sg_count = dma_map_sg(host->dev, data->sg, data->sg_len,
@@ -735,10 +733,8 @@ static void msdc_prepare_data(struct msdc_host *host, struct mmc_request *mrq)
 	}
 }
 
-static void msdc_unprepare_data(struct msdc_host *host, struct mmc_request *mrq)
+static void msdc_unprepare_data(struct msdc_host *host, struct mmc_data *data)
 {
-	struct mmc_data *data = mrq->data;
-
 	if (data->host_cookie & MSDC_ASYNC_FLAG)
 		return;
 
@@ -1140,7 +1136,7 @@ static void msdc_request_done(struct msdc_host *host, struct mmc_request *mrq)
 
 	msdc_track_cmd_data(host, mrq->cmd, mrq->data);
 	if (mrq->data)
-		msdc_unprepare_data(host, mrq);
+		msdc_unprepare_data(host, mrq->data);
 	if (host->error)
 		msdc_reset_hw(host);
 	mmc_request_done(mmc_from_priv(host), mrq);
@@ -1311,7 +1307,7 @@ static void msdc_ops_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	host->mrq = mrq;
 
 	if (mrq->data)
-		msdc_prepare_data(host, mrq);
+		msdc_prepare_data(host, mrq->data);
 
 	/* if SBC is required, we have HW option and SW option.
 	 * if HW option is enabled, and SBC does not have "special" flags,
@@ -1332,7 +1328,7 @@ static void msdc_pre_req(struct mmc_host *mmc, struct mmc_request *mrq)
 	if (!data)
 		return;
 
-	msdc_prepare_data(host, mrq);
+	msdc_prepare_data(host, data);
 	data->host_cookie |= MSDC_ASYNC_FLAG;
 }
 
@@ -1340,19 +1336,18 @@ static void msdc_post_req(struct mmc_host *mmc, struct mmc_request *mrq,
 		int err)
 {
 	struct msdc_host *host = mmc_priv(mmc);
-	struct mmc_data *data;
+	struct mmc_data *data = mrq->data;
 
-	data = mrq->data;
 	if (!data)
 		return;
+
 	if (data->host_cookie) {
 		data->host_cookie &= ~MSDC_ASYNC_FLAG;
-		msdc_unprepare_data(host, mrq);
+		msdc_unprepare_data(host, data);
 	}
 }
 
-static void msdc_data_xfer_next(struct msdc_host *host,
-				struct mmc_request *mrq, struct mmc_data *data)
+static void msdc_data_xfer_next(struct msdc_host *host, struct mmc_request *mrq)
 {
 	if (mmc_op_multi(mrq->cmd->opcode) && mrq->stop && !mrq->stop->error &&
 	    !mrq->sbc)
@@ -1411,7 +1406,7 @@ static bool msdc_data_xfer_done(struct msdc_host *host, u32 events,
 				(int)data->error, data->bytes_xfered);
 		}
 
-		msdc_data_xfer_next(host, mrq, data);
+		msdc_data_xfer_next(host, mrq);
 		done = true;
 	}
 	return done;
diff --git a/drivers/mmc/host/of_mmc_spi.c b/drivers/mmc/host/of_mmc_spi.c
index 9d480a05f655..3629550528b6 100644
--- a/drivers/mmc/host/of_mmc_spi.c
+++ b/drivers/mmc/host/of_mmc_spi.c
@@ -22,8 +22,8 @@
 MODULE_LICENSE("GPL");
 
 struct of_mmc_spi {
-	int detect_irq;
 	struct mmc_spi_platform_data pdata;
+	int detect_irq;
 };
 
 static struct of_mmc_spi *to_of_mmc_spi(struct device *dev)
diff --git a/drivers/mmc/host/renesas_sdhi_core.c b/drivers/mmc/host/renesas_sdhi_core.c
index baab4c2e1b53..e49ca0f7fe9a 100644
--- a/drivers/mmc/host/renesas_sdhi_core.c
+++ b/drivers/mmc/host/renesas_sdhi_core.c
@@ -704,7 +704,7 @@ static int renesas_sdhi_execute_tuning(struct mmc_host *mmc, u32 opcode)
 			set_bit(i, priv->smpcmp);
 
 		if (cmd_error)
-			mmc_abort_tuning(mmc, opcode);
+			mmc_send_abort_tuning(mmc, opcode);
 	}
 
 	ret = renesas_sdhi_select_tuning(host);
diff --git a/drivers/mmc/host/s3cmci.c b/drivers/mmc/host/s3cmci.c
index 0ca6f6d30b75..8d5929a32d34 100644
--- a/drivers/mmc/host/s3cmci.c
+++ b/drivers/mmc/host/s3cmci.c
@@ -1578,17 +1578,12 @@ static int s3cmci_probe(struct platform_device *pdev)
 		goto probe_iounmap;
 	}
 
-	if (request_irq(host->irq, s3cmci_irq, 0, DRIVER_NAME, host)) {
+	if (request_irq(host->irq, s3cmci_irq, IRQF_NO_AUTOEN, DRIVER_NAME, host)) {
 		dev_err(&pdev->dev, "failed to request mci interrupt.\n");
 		ret = -ENOENT;
 		goto probe_iounmap;
 	}
 
-	/* We get spurious interrupts even when we have set the IMSK
-	 * register to ignore everything, so use disable_irq() to make
-	 * ensure we don't lock the system with un-serviceable requests. */
-
-	disable_irq(host->irq);
 	host->irq_state = false;
 
 	/* Depending on the dma state, get a DMA channel to use. */
diff --git a/drivers/mmc/host/sdhci-acpi.c b/drivers/mmc/host/sdhci-acpi.c
index c3fbf8c825c4..8fe65f172a61 100644
--- a/drivers/mmc/host/sdhci-acpi.c
+++ b/drivers/mmc/host/sdhci-acpi.c
@@ -822,6 +822,17 @@ static const struct dmi_system_id sdhci_acpi_quirks[] = {
 		},
 		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
 	},
+	{
+		/*
+		 * The Toshiba WT8-B's microSD slot always reports the card being
+		 * write-protected.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "TOSHIBA"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "TOSHIBA ENCORE 2 WT8-B"),
+		},
+		.driver_data = (void *)DMI_QUIRK_SD_NO_WRITE_PROTECT,
+	},
 	{} /* Terminating entry */
 };
 
diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
index b991cf0e60c5..72c0bf0c1887 100644
--- a/drivers/mmc/host/sdhci-esdhc-imx.c
+++ b/drivers/mmc/host/sdhci-esdhc-imx.c
@@ -324,11 +324,6 @@ static inline int is_imx53_esdhc(struct pltfm_imx_data *data)
 	return data->socdata == &esdhc_imx53_data;
 }
 
-static inline int is_imx6q_usdhc(struct pltfm_imx_data *data)
-{
-	return data->socdata == &usdhc_imx6q_data;
-}
-
 static inline int esdhc_is_usdhc(struct pltfm_imx_data *data)
 {
 	return !!(data->socdata->flags & ESDHC_FLAG_USDHC);
@@ -427,9 +422,6 @@ static u32 esdhc_readl_le(struct sdhci_host *host, int reg)
 					| FIELD_PREP(SDHCI_RETUNING_MODE_MASK,
 						     SDHCI_TUNING_MODE_3);
 
-			if (imx_data->socdata->flags & ESDHC_FLAG_HS400)
-				val |= SDHCI_SUPPORT_HS400;
-
 			/*
 			 * Do not advertise faster UHS modes if there are no
 			 * pinctrl states for 100MHz/200MHz.
@@ -1591,7 +1583,7 @@ static int sdhci_esdhc_imx_probe(struct platform_device *pdev)
 		host->quirks |= SDHCI_QUIRK_BROKEN_ADMA;
 
 	if (imx_data->socdata->flags & ESDHC_FLAG_HS400)
-		host->quirks2 |= SDHCI_QUIRK2_CAPS_BIT63_FOR_HS400;
+		host->mmc->caps2 |= MMC_CAP2_HS400;
 
 	if (imx_data->socdata->flags & ESDHC_FLAG_BROKEN_AUTO_CMD23)
 		host->quirks2 |= SDHCI_QUIRK2_ACMD23_BROKEN;
@@ -1628,6 +1620,14 @@ static int sdhci_esdhc_imx_probe(struct platform_device *pdev)
 	if (err)
 		goto disable_ahb_clk;
 
+	/*
+	 * Setup the wakeup capability here, let user to decide
+	 * whether need to enable this wakeup through sysfs interface.
+	 */
+	if ((host->mmc->pm_caps & MMC_PM_KEEP_POWER) &&
+			(host->mmc->pm_caps & MMC_PM_WAKE_SDIO_IRQ))
+		device_set_wakeup_capable(&pdev->dev, true);
+
 	pm_runtime_set_active(&pdev->dev);
 	pm_runtime_set_autosuspend_delay(&pdev->dev, 50);
 	pm_runtime_use_autosuspend(&pdev->dev);
diff --git a/drivers/mmc/host/sdhci-iproc.c b/drivers/mmc/host/sdhci-iproc.c
index ddeaf8e1f72f..cce390fe9cf3 100644
--- a/drivers/mmc/host/sdhci-iproc.c
+++ b/drivers/mmc/host/sdhci-iproc.c
@@ -286,11 +286,35 @@ static const struct sdhci_iproc_data bcm2711_data = {
 	.mmc_caps = MMC_CAP_3_3V_DDR,
 };
 
+static const struct sdhci_pltfm_data sdhci_bcm7211a0_pltfm_data = {
+	.quirks = SDHCI_QUIRK_MISSING_CAPS |
+		SDHCI_QUIRK_BROKEN_TIMEOUT_VAL |
+		SDHCI_QUIRK_BROKEN_DMA |
+		SDHCI_QUIRK_BROKEN_ADMA,
+	.ops = &sdhci_iproc_ops,
+};
+
+#define BCM7211A0_BASE_CLK_MHZ 100
+static const struct sdhci_iproc_data bcm7211a0_data = {
+	.pdata = &sdhci_bcm7211a0_pltfm_data,
+	.caps = ((BCM7211A0_BASE_CLK_MHZ / 2) << SDHCI_TIMEOUT_CLK_SHIFT) |
+		(BCM7211A0_BASE_CLK_MHZ << SDHCI_CLOCK_BASE_SHIFT) |
+		((0x2 << SDHCI_MAX_BLOCK_SHIFT)
+			& SDHCI_MAX_BLOCK_MASK) |
+		SDHCI_CAN_VDD_330 |
+		SDHCI_CAN_VDD_180 |
+		SDHCI_CAN_DO_SUSPEND |
+		SDHCI_CAN_DO_HISPD,
+	.caps1 = SDHCI_DRIVER_TYPE_C |
+		 SDHCI_DRIVER_TYPE_D,
+};
+
 static const struct of_device_id sdhci_iproc_of_match[] = {
 	{ .compatible = "brcm,bcm2835-sdhci", .data = &bcm2835_data },
 	{ .compatible = "brcm,bcm2711-emmc2", .data = &bcm2711_data },
 	{ .compatible = "brcm,sdhci-iproc-cygnus", .data = &iproc_cygnus_data},
 	{ .compatible = "brcm,sdhci-iproc", .data = &iproc_data },
+	{ .compatible = "brcm,bcm7211a0-sdhci", .data = &bcm7211a0_data },
 	{ }
 };
 MODULE_DEVICE_TABLE(of, sdhci_iproc_of_match);
@@ -384,6 +408,11 @@ err:
 	return ret;
 }
 
+static void sdhci_iproc_shutdown(struct platform_device *pdev)
+{
+	sdhci_pltfm_suspend(&pdev->dev);
+}
+
 static struct platform_driver sdhci_iproc_driver = {
 	.driver = {
 		.name = "sdhci-iproc",
@@ -394,6 +423,7 @@ static struct platform_driver sdhci_iproc_driver = {
 	},
 	.probe = sdhci_iproc_probe,
 	.remove = sdhci_pltfm_unregister,
+	.shutdown = sdhci_iproc_shutdown,
 };
 module_platform_driver(sdhci_iproc_driver);
 
diff --git a/drivers/mmc/host/sdhci-of-aspeed.c b/drivers/mmc/host/sdhci-of-aspeed.c
index d001c51074a0..6e4e132903a6 100644
--- a/drivers/mmc/host/sdhci-of-aspeed.c
+++ b/drivers/mmc/host/sdhci-of-aspeed.c
@@ -31,6 +31,11 @@
 #define   ASPEED_SDC_S0_PHASE_OUT_EN	GENMASK(1, 0)
 #define   ASPEED_SDC_PHASE_MAX		31
 
+/* SDIO{10,20} */
+#define ASPEED_SDC_CAP1_1_8V           (0 * 32 + 26)
+/* SDIO{14,24} */
+#define ASPEED_SDC_CAP2_SDR104         (1 * 32 + 1)
+
 struct aspeed_sdc {
 	struct clk *clk;
 	struct resource *res;
@@ -72,6 +77,37 @@ struct aspeed_sdhci {
 	const struct aspeed_sdhci_phase_desc *phase_desc;
 };
 
+/*
+ * The function sets the mirror register for updating
+ * capbilities of the current slot.
+ *
+ *   slot | capability  | caps_reg | mirror_reg
+ *   -----|-------------|----------|------------
+ *     0  | CAP1_1_8V   | SDIO140  |   SDIO10
+ *     0  | CAP2_SDR104 | SDIO144  |   SDIO14
+ *     1  | CAP1_1_8V   | SDIO240  |   SDIO20
+ *     1  | CAP2_SDR104 | SDIO244  |   SDIO24
+ */
+static void aspeed_sdc_set_slot_capability(struct sdhci_host *host, struct aspeed_sdc *sdc,
+					   int capability, bool enable, u8 slot)
+{
+	u32 mirror_reg_offset;
+	u32 cap_val;
+	u8 cap_reg;
+
+	if (slot > 1)
+		return;
+
+	cap_reg = capability / 32;
+	cap_val = sdhci_readl(host, 0x40 + (cap_reg * 4));
+	if (enable)
+		cap_val |= BIT(capability % 32);
+	else
+		cap_val &= ~BIT(capability % 32);
+	mirror_reg_offset = ((slot + 1) * 0x10) + (cap_reg * 4);
+	writel(cap_val, sdc->regs + mirror_reg_offset);
+}
+
 static void aspeed_sdc_configure_8bit_mode(struct aspeed_sdc *sdc,
 					   struct aspeed_sdhci *sdhci,
 					   bool bus8)
@@ -150,7 +186,7 @@ static int aspeed_sdhci_phase_to_tap(struct device *dev, unsigned long rate_hz,
 
 	tap = div_u64(phase_period_ps, prop_delay_ps);
 	if (tap > ASPEED_SDHCI_NR_TAPS) {
-		dev_warn(dev,
+		dev_dbg(dev,
 			 "Requested out of range phase tap %d for %d degrees of phase compensation at %luHz, clamping to tap %d\n",
 			 tap, phase_deg, rate_hz, ASPEED_SDHCI_NR_TAPS);
 		tap = ASPEED_SDHCI_NR_TAPS;
@@ -328,6 +364,7 @@ static inline int aspeed_sdhci_calculate_slot(struct aspeed_sdhci *dev,
 static int aspeed_sdhci_probe(struct platform_device *pdev)
 {
 	const struct aspeed_sdhci_pdata *aspeed_pdata;
+	struct device_node *np = pdev->dev.of_node;
 	struct sdhci_pltfm_host *pltfm_host;
 	struct aspeed_sdhci *dev;
 	struct sdhci_host *host;
@@ -372,6 +409,17 @@ static int aspeed_sdhci_probe(struct platform_device *pdev)
 
 	sdhci_get_of_property(pdev);
 
+	if (of_property_read_bool(np, "mmc-hs200-1_8v") ||
+	    of_property_read_bool(np, "sd-uhs-sdr104")) {
+		aspeed_sdc_set_slot_capability(host, dev->parent, ASPEED_SDC_CAP1_1_8V,
+					       true, slot);
+	}
+
+	if (of_property_read_bool(np, "sd-uhs-sdr104")) {
+		aspeed_sdc_set_slot_capability(host, dev->parent, ASPEED_SDC_CAP2_SDR104,
+					       true, slot);
+	}
+
 	pltfm_host->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(pltfm_host->clk))
 		return PTR_ERR(pltfm_host->clk);
diff --git a/drivers/mmc/host/sdhci-omap.c b/drivers/mmc/host/sdhci-omap.c
index 7893fd3599b6..8f4d1f003f65 100644
--- a/drivers/mmc/host/sdhci-omap.c
+++ b/drivers/mmc/host/sdhci-omap.c
@@ -1173,10 +1173,9 @@ static int sdhci_omap_probe(struct platform_device *pdev)
 	 * as part of pm_runtime_get_sync.
 	 */
 	pm_runtime_enable(dev);
-	ret = pm_runtime_get_sync(dev);
-	if (ret < 0) {
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret) {
 		dev_err(dev, "pm_runtime_get_sync failed\n");
-		pm_runtime_put_noidle(dev);
 		goto err_rpm_disable;
 	}
 
diff --git a/drivers/mmc/host/sdhci-pci-gli.c b/drivers/mmc/host/sdhci-pci-gli.c
index 061618aa247f..4fd99c1e82ba 100644
--- a/drivers/mmc/host/sdhci-pci-gli.c
+++ b/drivers/mmc/host/sdhci-pci-gli.c
@@ -94,7 +94,7 @@
 
 #define PCIE_GLI_9763E_CFG2      0x8A4
 #define   GLI_9763E_CFG2_L1DLY     GENMASK(28, 19)
-#define   GLI_9763E_CFG2_L1DLY_MID 0x50
+#define   GLI_9763E_CFG2_L1DLY_MID 0x54
 
 #define PCIE_GLI_9763E_MMC_CTRL  0x960
 #define   GLI_9763E_HS400_SLOW     BIT(3)
@@ -847,7 +847,7 @@ static void gli_set_gl9763e(struct sdhci_pci_slot *slot)
 
 	pci_read_config_dword(pdev, PCIE_GLI_9763E_CFG2, &value);
 	value &= ~GLI_9763E_CFG2_L1DLY;
-	/* set ASPM L1 entry delay to 20us */
+	/* set ASPM L1 entry delay to 21us */
 	value |= FIELD_PREP(GLI_9763E_CFG2_L1DLY, GLI_9763E_CFG2_L1DLY_MID);
 	pci_write_config_dword(pdev, PCIE_GLI_9763E_CFG2, value);
 
diff --git a/drivers/mmc/host/sdhci-sprd.c b/drivers/mmc/host/sdhci-sprd.c
index 5dc36efff47f..11e375579cfb 100644
--- a/drivers/mmc/host/sdhci-sprd.c
+++ b/drivers/mmc/host/sdhci-sprd.c
@@ -393,6 +393,7 @@ static void sdhci_sprd_request_done(struct sdhci_host *host,
 static struct sdhci_ops sdhci_sprd_ops = {
 	.read_l = sdhci_sprd_readl,
 	.write_l = sdhci_sprd_writel,
+	.write_w = sdhci_sprd_writew,
 	.write_b = sdhci_sprd_writeb,
 	.set_clock = sdhci_sprd_set_clock,
 	.get_max_clock = sdhci_sprd_get_max_clock,
diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index bf238ade1602..6aaf5c3ce34c 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -2680,7 +2680,7 @@ void sdhci_abort_tuning(struct sdhci_host *host, u32 opcode)
 
 	sdhci_end_tuning(host);
 
-	mmc_abort_tuning(host->mmc, opcode);
+	mmc_send_abort_tuning(host->mmc, opcode);
 }
 EXPORT_SYMBOL_GPL(sdhci_abort_tuning);
 
diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h
index 0770c036e2ff..c35ed4be75b7 100644
--- a/drivers/mmc/host/sdhci.h
+++ b/drivers/mmc/host/sdhci.h
@@ -201,8 +201,10 @@
 
 #define SDHCI_CAPABILITIES	0x40
 #define  SDHCI_TIMEOUT_CLK_MASK		GENMASK(5, 0)
+#define  SDHCI_TIMEOUT_CLK_SHIFT 0
 #define  SDHCI_TIMEOUT_CLK_UNIT	0x00000080
 #define  SDHCI_CLOCK_BASE_MASK		GENMASK(13, 8)
+#define  SDHCI_CLOCK_BASE_SHIFT	8
 #define  SDHCI_CLOCK_V3_BASE_MASK	GENMASK(15, 8)
 #define  SDHCI_MAX_BLOCK_MASK	0x00030000
 #define  SDHCI_MAX_BLOCK_SHIFT  16
diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c
index 1fad6e442688..f654afbe8e83 100644
--- a/drivers/mmc/host/sdhci_am654.c
+++ b/drivers/mmc/host/sdhci_am654.c
@@ -809,11 +809,9 @@ static int sdhci_am654_probe(struct platform_device *pdev)
 
 	/* Clocks are enabled using pm_runtime */
 	pm_runtime_enable(dev);
-	ret = pm_runtime_get_sync(dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(dev);
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret)
 		goto pm_runtime_disable;
-	}
 
 	base = devm_platform_ioremap_resource(pdev, 1);
 	if (IS_ERR(base)) {
diff --git a/drivers/mmc/host/usdhi6rol0.c b/drivers/mmc/host/usdhi6rol0.c
index 615f3d008af1..b9b79b1089a0 100644
--- a/drivers/mmc/host/usdhi6rol0.c
+++ b/drivers/mmc/host/usdhi6rol0.c
@@ -1801,6 +1801,7 @@ static int usdhi6_probe(struct platform_device *pdev)
 
 	version = usdhi6_read(host, USDHI6_VERSION);
 	if ((version & 0xfff) != 0xa0d) {
+		ret = -EPERM;
 		dev_err(dev, "Version not recognized %x\n", version);
 		goto e_clk_off;
 	}
diff --git a/drivers/mmc/host/via-sdmmc.c b/drivers/mmc/host/via-sdmmc.c
index a1d098560099..c32df5530b94 100644
--- a/drivers/mmc/host/via-sdmmc.c
+++ b/drivers/mmc/host/via-sdmmc.c
@@ -857,6 +857,9 @@ static void via_sdc_data_isr(struct via_crdr_mmc_host *host, u16 intmask)
 {
 	BUG_ON(intmask == 0);
 
+	if (!host->data)
+		return;
+
 	if (intmask & VIA_CRDR_SDSTS_DT)
 		host->data->error = -ETIMEDOUT;
 	else if (intmask & (VIA_CRDR_SDSTS_RC | VIA_CRDR_SDSTS_WC))
diff --git a/drivers/mmc/host/vub300.c b/drivers/mmc/host/vub300.c
index 739cf63ef6e2..4950d10d3a19 100644
--- a/drivers/mmc/host/vub300.c
+++ b/drivers/mmc/host/vub300.c
@@ -2279,7 +2279,7 @@ static int vub300_probe(struct usb_interface *interface,
 	if (retval < 0)
 		goto error5;
 	retval =
-		usb_control_msg(vub300->udev, usb_rcvctrlpipe(vub300->udev, 0),
+		usb_control_msg(vub300->udev, usb_sndctrlpipe(vub300->udev, 0),
 				SET_ROM_WAIT_STATES,
 				USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE,
 				firmware_rom_wait_states, 0x0000, NULL, 0, HZ);