Merge tag 'mmc-v4.20' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc

Pull MMC updates from Ulf Hansson: "MMC core: - Introduce a host helper function to share re-tuning progress MMC host: - sdhci: Add support for v4 host mode - sdhci-of-arasan: Add Support for AM654 MMC and PHY - sdhci-sprd: Add support for Spreadtrum's host controller - sdhci-tegra: Add support for HS400 enhanced strobe - sdhci-tegra: Enable UHS/HS200 modes for Tegra186/210 - sdhci-tegra: Add support for HS400 delay line calibration - sdhci-tegra: Add support for pad calibration - sdhci-of-dwcmshc: Address 128MB DMA boundary limitation - sdhci-of-esdhc: Add support for tuning erratum A008171 - sdhci-iproc: Add ACPI support - mediatek: Add support for MT8183 - mediatek: Improve the support for tuning - mediatek: Add bus clock control for MT2712 - jz4740: Add support for the JZ4725B - mmci: Add support for the stm32 sdmmc variant - mmci: Add support for an optional reset control - mmci: Add some new variant specific properties/callbacks - mmci: Re-structure DMA code to prepare for new variants - renesas_sdhi: Add support for r8a77470, r8a7744 and r8a774a1 - renesas_sdhi_internal_dmac: Whitelist r8a77970 and r8a774a1 - tmio/uniphier-sd: Add new UniPhier SD/eMMC controller driver - tmio/renesas_sdhi: Deal properly with SCC detection during re-tune - tmio/renesas_sdhi: Refactor/consolidate clock management - omap_hsmmc: Drop cover detection and some unused platform data - dw_mmc-exynos: Enable tuning for more speed modes - sunxi: Clarify the new timing mode and enable it for the A64 controller - various: Convert to slot GPIO descriptors" * tag 'mmc-v4.20' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc: (129 commits) mmc: mediatek: drop too much code of tuning method mmc: mediatek: add MT8183 MMC driver support mmc: mediatek: tune CMD/DATA together mmc: mediatek: fix cannot receive new request when msdc_cmd_is_ready fail mmc: mediatek: fill the actual clock for mmc debugfs mmc: dt-bindings: add support for MT8183 SoC mmc: uniphier-sd: avoid using broken DMA RX channel mmc: uniphier-sd: fix DMA disabling mmc: tmio: simplify the DMA mode test mmc: tmio: remove TMIO_MMC_HAVE_HIGH_REG flag mmc: tmio: move MFD variant reset to a platform hook mmc: renesas_sdhi: Add r8a77470 SDHI1 support dt-bindings: mmc: renesas_sdhi: Add r8a77470 support mmc: mmci: add stm32 sdmmc variant dt-bindings: mmci: add stm32 sdmmc variant mmc: mmci: add stm32 sdmmc registers mmc: mmci: add clock divider for stm32 sdmmc mmc: mmci: add optional reset property dt-bindings: mmci: add optional reset property mmc: mmci: add variant property to not read datacnt ...
author: Linus Torvalds <torvalds@linux-foundation.org> 2018-10-23 08:36:15 +0100
committer: Linus Torvalds <torvalds@linux-foundation.org> 2018-10-23 08:36:15 +0100
commit: 1650ac53066577a5e83fe3e9d992c9311597ff8c (patch)
tree: 58667f89d6b5ac5a40c386548c17646c561453ef /drivers/mmc
parent: Merge tag 'regulator-v5.0' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator (diff)
parent: mmc: mediatek: drop too much code of tuning method (diff)
download: linux-dev-1650ac53066577a5e83fe3e9d992c9311597ff8c.tar.xz
linux-dev-1650ac53066577a5e83fe3e9d992c9311597ff8c.zip
43 files changed, 3845 insertions, 912 deletions
diff --git a/drivers/mmc/core/Kconfig b/drivers/mmc/core/Kconfig
index 42e89060cd41..2f38a7ad07e0 100644
--- a/drivers/mmc/core/Kconfig
+++ b/drivers/mmc/core/Kconfig
@@ -14,7 +14,7 @@ config PWRSEQ_EMMC
 
 config PWRSEQ_SD8787
 	tristate "HW reset support for SD8787 BT + Wifi module"
-	depends on OF && (MWIFIEX || BT_MRVL_SDIO)
+	depends on OF && (MWIFIEX || BT_MRVL_SDIO || LIBERTAS_SDIO)
 	help
 	  This selects hardware reset support for the SD8787 BT + Wifi
 	  module. By default this option is set to n.
diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index 694d0828215d..1b58739d9744 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -34,6 +34,16 @@ config MMC_QCOM_DML
 
 	  if unsure, say N.
 
+config MMC_STM32_SDMMC
+	bool "STMicroelectronics STM32 SDMMC Controller"
+	depends on MMC_ARMMMCI
+	default y
+	help
+	  This selects the STMicroelectronics STM32 SDMMC host controller.
+	  If you have a STM32 sdmmc host with internal DMA say Y here.
+
+	  If unsure, say N.
+
 config MMC_PXA
 	tristate "Intel PXA25x/26x/27x Multimedia Card Interface support"
 	depends on ARCH_PXA
@@ -345,6 +355,7 @@ 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 MMC_SDHCI_PLTFM
+	depends on OF || ACPI
 	default ARCH_BCM_IPROC
 	select MMC_SDHCI_IO_ACCESSORS
 	help
@@ -592,6 +603,19 @@ config MMC_SDRICOH_CS
 	  To compile this driver as a module, choose M here: the
 	  module will be called sdricoh_cs.
 
+config MMC_SDHCI_SPRD
+	tristate "Spreadtrum SDIO host Controller"
+	depends on ARCH_SPRD
+	depends on MMC_SDHCI_PLTFM
+	select MMC_SDHCI_IO_ACCESSORS
+	help
+	  This selects the SDIO Host Controller in Spreadtrum
+	  SoCs, this driver supports R11(IP version: R11P0).
+
+	  If you have a controller with this interface, say Y or M here.
+
+	  If unsure, say N.
+
 config MMC_TMIO_CORE
 	tristate
 
@@ -622,14 +646,24 @@ config MMC_SDHI_SYS_DMAC
 
 config MMC_SDHI_INTERNAL_DMAC
 	tristate "DMA for SDHI SD/SDIO controllers using on-chip bus mastering"
-	depends on ARM64 || COMPILE_TEST
+	depends on ARM64 || ARCH_R8A77470 || COMPILE_TEST
 	depends on MMC_SDHI
-	default MMC_SDHI if ARM64
+	default MMC_SDHI if (ARM64 || ARCH_R8A77470)
 	help
 	  This provides DMA support for SDHI SD/SDIO controllers
 	  using on-chip bus mastering. This supports the controllers
 	  found in arm64 based SoCs.
 
+config MMC_UNIPHIER
+	tristate "UniPhier SD/eMMC Host Controller support"
+	depends on ARCH_UNIPHIER || COMPILE_TEST
+	depends on OF
+	select MMC_TMIO_CORE
+	help
+	  This provides support for the SD/eMMC controller found in
+	  UniPhier SoCs. The eMMC variant of this controller is used
+	  only for 32-bit SoCs.
+
 config MMC_CB710
 	tristate "ENE CB710 MMC/SD Interface support"
 	depends on PCI
@@ -772,7 +806,7 @@ config MMC_SH_MMCIF
 
 config MMC_JZ4740
 	tristate "Ingenic JZ47xx SD/Multimedia Card Interface support"
-	depends on MACH_JZ4740 || MACH_JZ4780
+	depends on MIPS
 	help
 	  This selects support for the SD/MMC controller on Ingenic
 	  JZ4740, JZ4750, JZ4770 and JZ4780 SoCs.
diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
index ce8398e6f2c0..720d37777098 100644
--- a/drivers/mmc/host/Makefile
+++ b/drivers/mmc/host/Makefile
@@ -6,6 +6,7 @@
 obj-$(CONFIG_MMC_ARMMMCI) += armmmci.o
 armmmci-y := mmci.o
 armmmci-$(CONFIG_MMC_QCOM_DML) += mmci_qcom_dml.o
+armmmci-$(CONFIG_MMC_STM32_SDMMC) += mmci_stm32_sdmmc.o
 obj-$(CONFIG_MMC_PXA)		+= pxamci.o
 obj-$(CONFIG_MMC_MXC)		+= mxcmmc.o
 obj-$(CONFIG_MMC_MXS)		+= mxs-mmc.o
@@ -42,6 +43,7 @@ obj-$(CONFIG_MMC_TMIO_CORE)	+= tmio_mmc_core.o
 obj-$(CONFIG_MMC_SDHI)		+= renesas_sdhi_core.o
 obj-$(CONFIG_MMC_SDHI_SYS_DMAC)		+= renesas_sdhi_sys_dmac.o
 obj-$(CONFIG_MMC_SDHI_INTERNAL_DMAC)	+= renesas_sdhi_internal_dmac.o
+obj-$(CONFIG_MMC_UNIPHIER)	+= uniphier-sd.o
 obj-$(CONFIG_MMC_CB710)		+= cb710-mmc.o
 obj-$(CONFIG_MMC_VIA_SDMMC)	+= via-sdmmc.o
 octeon-mmc-objs := cavium.o cavium-octeon.o
@@ -91,6 +93,7 @@ obj-$(CONFIG_MMC_SDHCI_ST)		+= sdhci-st.o
 obj-$(CONFIG_MMC_SDHCI_MICROCHIP_PIC32)	+= sdhci-pic32.o
 obj-$(CONFIG_MMC_SDHCI_BRCMSTB)		+= sdhci-brcmstb.o
 obj-$(CONFIG_MMC_SDHCI_OMAP)		+= sdhci-omap.o
+obj-$(CONFIG_MMC_SDHCI_SPRD)		+= sdhci-sprd.o
 obj-$(CONFIG_MMC_CQHCI)			+= cqhci.o
 
 ifeq ($(CONFIG_CB710_DEBUG),y)
diff --git a/drivers/mmc/host/dw_mmc-exynos.c b/drivers/mmc/host/dw_mmc-exynos.c
index ab47b018716a..d46c3439b508 100644
--- a/drivers/mmc/host/dw_mmc-exynos.c
+++ b/drivers/mmc/host/dw_mmc-exynos.c
@@ -253,6 +253,8 @@ static void dw_mci_exynos_config_hs400(struct dw_mci *host, u32 timing)
 	if (timing == MMC_TIMING_MMC_HS400) {
 		dqs |= DATA_STROBE_EN;
 		strobe = DQS_CTRL_RD_DELAY(strobe, priv->dqs_delay);
+	} else if (timing == MMC_TIMING_UHS_SDR104) {
+		dqs &= 0xffffff00;
 	} else {
 		dqs &= ~DATA_STROBE_EN;
 	}
@@ -312,6 +314,15 @@ static void dw_mci_exynos_set_ios(struct dw_mci *host, struct mmc_ios *ios)
 		if (ios->bus_width == MMC_BUS_WIDTH_8)
 			wanted <<= 1;
 		break;
+	case MMC_TIMING_UHS_SDR104:
+	case MMC_TIMING_UHS_SDR50:
+		clksel = (priv->sdr_timing & 0xfff8ffff) |
+			(priv->ciu_div << 16);
+		break;
+	case MMC_TIMING_UHS_DDR50:
+		clksel = (priv->ddr_timing & 0xfff8ffff) |
+			(priv->ciu_div << 16);
+		break;
 	default:
 		clksel = priv->sdr_timing;
 	}
diff --git a/drivers/mmc/host/dw_mmc-hi3798cv200.c b/drivers/mmc/host/dw_mmc-hi3798cv200.c
index f9b333ff259e..bc51cef47c47 100644
--- a/drivers/mmc/host/dw_mmc-hi3798cv200.c
+++ b/drivers/mmc/host/dw_mmc-hi3798cv200.c
@@ -23,6 +23,12 @@ struct hi3798cv200_priv {
 	struct clk *drive_clk;
 };
 
+static unsigned long dw_mci_hi3798cv200_caps[] = {
+	MMC_CAP_CMD23,
+	MMC_CAP_CMD23,
+	MMC_CAP_CMD23
+};
+
 static void dw_mci_hi3798cv200_set_ios(struct dw_mci *host, struct mmc_ios *ios)
 {
 	struct hi3798cv200_priv *priv = host->priv;
@@ -160,6 +166,8 @@ disable_sample_clk:
 }
 
 static const struct dw_mci_drv_data hi3798cv200_data = {
+	.caps = dw_mci_hi3798cv200_caps,
+	.num_caps = ARRAY_SIZE(dw_mci_hi3798cv200_caps),
 	.init = dw_mci_hi3798cv200_init,
 	.set_ios = dw_mci_hi3798cv200_set_ios,
 	.execute_tuning = dw_mci_hi3798cv200_execute_tuning,
diff --git a/drivers/mmc/host/jz4740_mmc.c b/drivers/mmc/host/jz4740_mmc.c
index 993386c9ea50..0c1efd5100b7 100644
--- a/drivers/mmc/host/jz4740_mmc.c
+++ b/drivers/mmc/host/jz4740_mmc.c
@@ -115,7 +115,7 @@
 
 enum jz4740_mmc_version {
 	JZ_MMC_JZ4740,
-	JZ_MMC_JZ4750,
+	JZ_MMC_JZ4725B,
 	JZ_MMC_JZ4780,
 };
 
@@ -176,7 +176,7 @@ struct jz4740_mmc_host {
 static void jz4740_mmc_write_irq_mask(struct jz4740_mmc_host *host,
 				      uint32_t val)
 {
-	if (host->version >= JZ_MMC_JZ4750)
+	if (host->version >= JZ_MMC_JZ4725B)
 		return writel(val, host->base + JZ_REG_MMC_IMASK);
 	else
 		return writew(val, host->base + JZ_REG_MMC_IMASK);
@@ -1012,6 +1012,7 @@ static void jz4740_mmc_free_gpios(struct platform_device *pdev)
 
 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,jz4780-mmc", .data = (void *) JZ_MMC_JZ4780 },
 	{},
 };
diff --git a/drivers/mmc/host/meson-mx-sdio.c b/drivers/mmc/host/meson-mx-sdio.c
index 2cfec33178c1..abe253c262a2 100644
--- a/drivers/mmc/host/meson-mx-sdio.c
+++ b/drivers/mmc/host/meson-mx-sdio.c
@@ -294,7 +294,7 @@ static void meson_mx_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	switch (ios->power_mode) {
 	case MMC_POWER_OFF:
 		vdd = 0;
-		/* fall-through: */
+		/* fall through */
 	case MMC_POWER_UP:
 		if (!IS_ERR(mmc->supply.vmmc)) {
 			host->error = mmc_regulator_set_ocr(mmc,
diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c
index 1841d250e9e2..82bab35fff41 100644
--- a/drivers/mmc/host/mmci.c
+++ b/drivers/mmc/host/mmci.c
@@ -28,8 +28,7 @@
 #include <linux/amba/bus.h>
 #include <linux/clk.h>
 #include <linux/scatterlist.h>
-#include <linux/gpio.h>
-#include <linux/of_gpio.h>
+#include <linux/of.h>
 #include <linux/regulator/consumer.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
@@ -37,6 +36,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/types.h>
 #include <linux/pinctrl/consumer.h>
+#include <linux/reset.h>
 
 #include <asm/div64.h>
 #include <asm/io.h>
@@ -46,41 +46,77 @@
 
 #define DRIVER_NAME "mmci-pl18x"
 
+#ifdef CONFIG_DMA_ENGINE
+void mmci_variant_init(struct mmci_host *host);
+#else
+static inline void mmci_variant_init(struct mmci_host *host) {}
+#endif
+
+#ifdef CONFIG_MMC_STM32_SDMMC
+void sdmmc_variant_init(struct mmci_host *host);
+#else
+static inline void sdmmc_variant_init(struct mmci_host *host) {}
+#endif
+
 static unsigned int fmax = 515633;
 
 static struct variant_data variant_arm = {
 	.fifosize		= 16 * 4,
 	.fifohalfsize		= 8 * 4,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 16,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.pwrreg_powerup		= MCI_PWR_UP,
 	.f_max			= 100000000,
 	.reversed_irq_handling	= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_ROD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_arm_extended_fifo = {
 	.fifosize		= 128 * 4,
 	.fifohalfsize		= 64 * 4,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 16,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.pwrreg_powerup		= MCI_PWR_UP,
 	.f_max			= 100000000,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_ROD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_arm_extended_fifo_hwfc = {
 	.fifosize		= 128 * 4,
 	.fifohalfsize		= 64 * 4,
 	.clkreg_enable		= MCI_ARM_HWFCEN,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 16,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.pwrreg_powerup		= MCI_PWR_UP,
 	.f_max			= 100000000,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_ROD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_u300 = {
@@ -88,7 +124,13 @@ static struct variant_data variant_u300 = {
 	.fifohalfsize		= 8 * 4,
 	.clkreg_enable		= MCI_ST_U300_HWFCEN,
 	.clkreg_8bit_bus_enable = MCI_ST_8BIT_BUS,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 16,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.datactrl_mask_sdio	= MCI_DPSM_ST_SDIOEN,
 	.st_sdio			= true,
 	.pwrreg_powerup		= MCI_PWR_ON,
@@ -97,8 +139,10 @@ static struct variant_data variant_u300 = {
 	.pwrreg_clkgate		= true,
 	.pwrreg_nopower		= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_OD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_nomadik = {
@@ -106,7 +150,13 @@ static struct variant_data variant_nomadik = {
 	.fifohalfsize		= 8 * 4,
 	.clkreg			= MCI_CLK_ENABLE,
 	.clkreg_8bit_bus_enable = MCI_ST_8BIT_BUS,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 24,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.datactrl_mask_sdio	= MCI_DPSM_ST_SDIOEN,
 	.st_sdio		= true,
 	.st_clkdiv		= true,
@@ -116,8 +166,10 @@ static struct variant_data variant_nomadik = {
 	.pwrreg_clkgate		= true,
 	.pwrreg_nopower		= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_OD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_ux500 = {
@@ -127,7 +179,13 @@ static struct variant_data variant_ux500 = {
 	.clkreg_enable		= MCI_ST_UX500_HWFCEN,
 	.clkreg_8bit_bus_enable = MCI_ST_8BIT_BUS,
 	.clkreg_neg_edge_enable	= MCI_ST_UX500_NEG_EDGE,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datalength_bits	= 24,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.datactrl_mask_sdio	= MCI_DPSM_ST_SDIOEN,
 	.st_sdio		= true,
 	.st_clkdiv		= true,
@@ -141,8 +199,10 @@ static struct variant_data variant_ux500 = {
 	.busy_detect_mask	= MCI_ST_BUSYENDMASK,
 	.pwrreg_nopower		= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_OD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_ux500v2 = {
@@ -152,8 +212,14 @@ static struct variant_data variant_ux500v2 = {
 	.clkreg_enable		= MCI_ST_UX500_HWFCEN,
 	.clkreg_8bit_bus_enable = MCI_ST_8BIT_BUS,
 	.clkreg_neg_edge_enable	= MCI_ST_UX500_NEG_EDGE,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.datactrl_mask_ddrmode	= MCI_DPSM_ST_DDRMODE,
 	.datalength_bits	= 24,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.datactrl_mask_sdio	= MCI_DPSM_ST_SDIOEN,
 	.st_sdio		= true,
 	.st_clkdiv		= true,
@@ -168,8 +234,10 @@ static struct variant_data variant_ux500v2 = {
 	.busy_detect_mask	= MCI_ST_BUSYENDMASK,
 	.pwrreg_nopower		= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_OD,
+	.init			= mmci_variant_init,
 };
 
 static struct variant_data variant_stm32 = {
@@ -179,7 +247,14 @@ static struct variant_data variant_stm32 = {
 	.clkreg_enable		= MCI_ST_UX500_HWFCEN,
 	.clkreg_8bit_bus_enable = MCI_ST_8BIT_BUS,
 	.clkreg_neg_edge_enable	= MCI_ST_UX500_NEG_EDGE,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.datalength_bits	= 24,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.datactrl_mask_sdio	= MCI_DPSM_ST_SDIOEN,
 	.st_sdio		= true,
 	.st_clkdiv		= true,
@@ -187,6 +262,26 @@ static struct variant_data variant_stm32 = {
 	.f_max			= 48000000,
 	.pwrreg_clkgate		= true,
 	.pwrreg_nopower		= true,
+	.init			= mmci_variant_init,
+};
+
+static struct variant_data variant_stm32_sdmmc = {
+	.fifosize		= 16 * 4,
+	.fifohalfsize		= 8 * 4,
+	.f_max			= 208000000,
+	.stm32_clkdiv		= true,
+	.cmdreg_cpsm_enable	= MCI_CPSM_STM32_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_STM32_LRSP_CRC,
+	.cmdreg_srsp_crc	= MCI_CPSM_STM32_SRSP_CRC,
+	.cmdreg_srsp		= MCI_CPSM_STM32_SRSP,
+	.data_cmd_enable	= MCI_CPSM_STM32_CMDTRANS,
+	.irq_pio_mask		= MCI_IRQ_PIO_STM32_MASK,
+	.datactrl_first		= true,
+	.datacnt_useless	= true,
+	.datalength_bits	= 25,
+	.datactrl_blocksz	= 14,
+	.stm32_idmabsize_mask	= GENMASK(12, 5),
+	.init			= sdmmc_variant_init,
 };
 
 static struct variant_data variant_qcom = {
@@ -197,15 +292,22 @@ static struct variant_data variant_qcom = {
 				  MCI_QCOM_CLK_SELECT_IN_FBCLK,
 	.clkreg_8bit_bus_enable = MCI_QCOM_CLK_WIDEBUS_8,
 	.datactrl_mask_ddrmode	= MCI_QCOM_CLK_SELECT_IN_DDR_MODE,
+	.cmdreg_cpsm_enable	= MCI_CPSM_ENABLE,
+	.cmdreg_lrsp_crc	= MCI_CPSM_RESPONSE | MCI_CPSM_LONGRSP,
+	.cmdreg_srsp_crc	= MCI_CPSM_RESPONSE,
+	.cmdreg_srsp		= MCI_CPSM_RESPONSE,
 	.data_cmd_enable	= MCI_CPSM_QCOM_DATCMD,
 	.blksz_datactrl4	= true,
 	.datalength_bits	= 24,
+	.datactrl_blocksz	= 11,
+	.datactrl_dpsm_enable	= MCI_DPSM_ENABLE,
 	.pwrreg_powerup		= MCI_PWR_UP,
 	.f_max			= 208000000,
 	.explicit_mclk_control	= true,
 	.qcom_fifo		= true,
 	.qcom_dml		= true,
 	.mmcimask1		= true,
+	.irq_pio_mask		= MCI_IRQ_PIO_MASK,
 	.start_err		= MCI_STARTBITERR,
 	.opendrain		= MCI_ROD,
 	.init			= qcom_variant_init,
@@ -226,24 +328,6 @@ static int mmci_card_busy(struct mmc_host *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)
 {
 	/*
@@ -262,7 +346,7 @@ static void mmci_reg_delay(struct mmci_host *host)
 /*
  * This must be called with host->lock held
  */
-static void mmci_write_clkreg(struct mmci_host *host, u32 clk)
+void mmci_write_clkreg(struct mmci_host *host, u32 clk)
 {
 	if (host->clk_reg != clk) {
 		host->clk_reg = clk;
@@ -273,7 +357,7 @@ static void mmci_write_clkreg(struct mmci_host *host, u32 clk)
 /*
  * This must be called with host->lock held
  */
-static void mmci_write_pwrreg(struct mmci_host *host, u32 pwr)
+void mmci_write_pwrreg(struct mmci_host *host, u32 pwr)
 {
 	if (host->pwr_reg != pwr) {
 		host->pwr_reg = pwr;
@@ -357,6 +441,135 @@ static void mmci_set_clkreg(struct mmci_host *host, unsigned int desired)
 	mmci_write_clkreg(host, clk);
 }
 
+void mmci_dma_release(struct mmci_host *host)
+{
+	if (host->ops && host->ops->dma_release)
+		host->ops->dma_release(host);
+
+	host->use_dma = false;
+}
+
+void mmci_dma_setup(struct mmci_host *host)
+{
+	if (!host->ops || !host->ops->dma_setup)
+		return;
+
+	if (host->ops->dma_setup(host))
+		return;
+
+	/* initialize pre request cookie */
+	host->next_cookie = 1;
+
+	host->use_dma = true;
+}
+
+/*
+ * 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;
+	}
+
+	if (host->ops && host->ops->validate_data)
+		return host->ops->validate_data(host, data);
+
+	return 0;
+}
+
+int mmci_prep_data(struct mmci_host *host, struct mmc_data *data, bool next)
+{
+	int err;
+
+	if (!host->ops || !host->ops->prep_data)
+		return 0;
+
+	err = host->ops->prep_data(host, data, next);
+
+	if (next && !err)
+		data->host_cookie = ++host->next_cookie < 0 ?
+			1 : host->next_cookie;
+
+	return err;
+}
+
+void mmci_unprep_data(struct mmci_host *host, struct mmc_data *data,
+		      int err)
+{
+	if (host->ops && host->ops->unprep_data)
+		host->ops->unprep_data(host, data, err);
+
+	data->host_cookie = 0;
+}
+
+void mmci_get_next_data(struct mmci_host *host, struct mmc_data *data)
+{
+	WARN_ON(data->host_cookie && data->host_cookie != host->next_cookie);
+
+	if (host->ops && host->ops->get_next_data)
+		host->ops->get_next_data(host, data);
+}
+
+int mmci_dma_start(struct mmci_host *host, unsigned int datactrl)
+{
+	struct mmc_data *data = host->data;
+	int ret;
+
+	if (!host->use_dma)
+		return -EINVAL;
+
+	ret = mmci_prep_data(host, data, false);
+	if (ret)
+		return ret;
+
+	if (!host->ops || !host->ops->dma_start)
+		return -EINVAL;
+
+	/* 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);
+
+	host->ops->dma_start(host, &datactrl);
+
+	/* 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;
+}
+
+void mmci_dma_finalize(struct mmci_host *host, struct mmc_data *data)
+{
+	if (!host->use_dma)
+		return;
+
+	if (host->ops && host->ops->dma_finalize)
+		host->ops->dma_finalize(host, data);
+}
+
+void mmci_dma_error(struct mmci_host *host)
+{
+	if (!host->use_dma)
+		return;
+
+	if (host->ops && host->ops->dma_error)
+		host->ops->dma_error(host);
+}
+
 static void
 mmci_request_end(struct mmci_host *host, struct mmc_request *mrq)
 {
@@ -378,7 +591,7 @@ static void mmci_set_mask1(struct mmci_host *host, unsigned int mask)
 	if (host->singleirq) {
 		unsigned int mask0 = readl(base + MMCIMASK0);
 
-		mask0 &= ~MCI_IRQ1MASK;
+		mask0 &= ~variant->irq_pio_mask;
 		mask0 |= mask;
 
 		writel(mask0, base + MMCIMASK0);
@@ -415,31 +628,50 @@ static void mmci_init_sg(struct mmci_host *host, struct mmc_data *data)
  * no custom DMA interfaces are supported.
  */
 #ifdef CONFIG_DMA_ENGINE
-static void mmci_dma_setup(struct mmci_host *host)
+struct mmci_dmae_next {
+	struct dma_async_tx_descriptor *desc;
+	struct dma_chan	*chan;
+};
+
+struct mmci_dmae_priv {
+	struct dma_chan	*cur;
+	struct dma_chan	*rx_channel;
+	struct dma_chan	*tx_channel;
+	struct dma_async_tx_descriptor	*desc_current;
+	struct mmci_dmae_next next_data;
+};
+
+int mmci_dmae_setup(struct mmci_host *host)
 {
 	const char *rxname, *txname;
+	struct mmci_dmae_priv *dmae;
 
-	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");
+	dmae = devm_kzalloc(mmc_dev(host->mmc), sizeof(*dmae), GFP_KERNEL);
+	if (!dmae)
+		return -ENOMEM;
 
-	/* initialize pre request cookie */
-	host->next_data.cookie = 1;
+	host->dma_priv = dmae;
+
+	dmae->rx_channel = dma_request_slave_channel(mmc_dev(host->mmc),
+						     "rx");
+	dmae->tx_channel = dma_request_slave_channel(mmc_dev(host->mmc),
+						     "tx");
 
 	/*
 	 * 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 (dmae->rx_channel && !dmae->tx_channel)
+		dmae->tx_channel = dmae->rx_channel;
 
-	if (host->dma_rx_channel)
-		rxname = dma_chan_name(host->dma_rx_channel);
+	if (dmae->rx_channel)
+		rxname = dma_chan_name(dmae->rx_channel);
 	else
 		rxname = "none";
 
-	if (host->dma_tx_channel)
-		txname = dma_chan_name(host->dma_tx_channel);
+	if (dmae->tx_channel)
+		txname = dma_chan_name(dmae->tx_channel);
 	else
 		txname = "none";
 
@@ -450,66 +682,84 @@ static void mmci_dma_setup(struct mmci_host *host)
 	 * 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;
+	if (dmae->tx_channel) {
+		struct device *dev = dmae->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;
+	if (dmae->rx_channel) {
+		struct device *dev = dmae->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;
 	}
 
-	if (host->ops && host->ops->dma_setup)
-		host->ops->dma_setup(host);
+	if (!dmae->tx_channel || !dmae->rx_channel) {
+		mmci_dmae_release(host);
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 /*
  * This is used in or so inline it
  * so it can be discarded.
  */
-static inline void mmci_dma_release(struct mmci_host *host)
+void mmci_dmae_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;
-}
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 
-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_in_progress = false;
-	host->dma_current = NULL;
-	host->dma_desc_current = NULL;
-	host->data->host_cookie = 0;
+	if (dmae->rx_channel)
+		dma_release_channel(dmae->rx_channel);
+	if (dmae->tx_channel)
+		dma_release_channel(dmae->tx_channel);
+	dmae->rx_channel = dmae->tx_channel = NULL;
 }
 
 static void mmci_dma_unmap(struct mmci_host *host, struct mmc_data *data)
 {
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 	struct dma_chan *chan;
 
 	if (data->flags & MMC_DATA_READ)
-		chan = host->dma_rx_channel;
+		chan = dmae->rx_channel;
 	else
-		chan = host->dma_tx_channel;
+		chan = dmae->tx_channel;
 
 	dma_unmap_sg(chan->device->dev, data->sg, data->sg_len,
 		     mmc_get_dma_dir(data));
 }
 
-static void mmci_dma_finalize(struct mmci_host *host, struct mmc_data *data)
+void mmci_dmae_error(struct mmci_host *host)
 {
+	struct mmci_dmae_priv *dmae = host->dma_priv;
+
+	if (!dma_inprogress(host))
+		return;
+
+	dev_err(mmc_dev(host->mmc), "error during DMA transfer!\n");
+	dmaengine_terminate_all(dmae->cur);
+	host->dma_in_progress = false;
+	dmae->cur = NULL;
+	dmae->desc_current = NULL;
+	host->data->host_cookie = 0;
+
+	mmci_dma_unmap(host, host->data);
+}
+
+void mmci_dmae_finalize(struct mmci_host *host, struct mmc_data *data)
+{
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 	u32 status;
 	int i;
 
+	if (!dma_inprogress(host))
+		return;
+
 	/* Wait up to 1ms for the DMA to complete */
 	for (i = 0; ; i++) {
 		status = readl(host->base + MMCISTATUS);
@@ -525,13 +775,12 @@ static void mmci_dma_finalize(struct mmci_host *host, struct mmc_data *data)
 	 * contiguous buffers.  On TX, we'll get a FIFO underrun error.
 	 */
 	if (status & MCI_RXDATAAVLBLMASK) {
-		mmci_dma_data_error(host);
+		mmci_dma_error(host);
 		if (!data->error)
 			data->error = -EIO;
-	}
-
-	if (!data->host_cookie)
+	} else if (!data->host_cookie) {
 		mmci_dma_unmap(host, data);
+	}
 
 	/*
 	 * Use of DMA with scatter-gather is impossible.
@@ -543,15 +792,16 @@ static void mmci_dma_finalize(struct mmci_host *host, struct mmc_data *data)
 	}
 
 	host->dma_in_progress = false;
-	host->dma_current = NULL;
-	host->dma_desc_current = NULL;
+	dmae->cur = NULL;
+	dmae->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,
+static int _mmci_dmae_prep_data(struct mmci_host *host, struct mmc_data *data,
 				struct dma_chan **dma_chan,
 				struct dma_async_tx_descriptor **dma_desc)
 {
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 	struct variant_data *variant = host->variant;
 	struct dma_slave_config conf = {
 		.src_addr = host->phybase + MMCIFIFO,
@@ -570,10 +820,10 @@ static int __mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
 
 	if (data->flags & MMC_DATA_READ) {
 		conf.direction = DMA_DEV_TO_MEM;
-		chan = host->dma_rx_channel;
+		chan = dmae->rx_channel;
 	} else {
 		conf.direction = DMA_MEM_TO_DEV;
-		chan = host->dma_tx_channel;
+		chan = dmae->tx_channel;
 	}
 
 	/* If there's no DMA channel, fall back to PIO */
@@ -610,160 +860,137 @@ static int __mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
 	return -ENOMEM;
 }
 
-static inline int mmci_dma_prep_data(struct mmci_host *host,
-				     struct mmc_data *data)
+int mmci_dmae_prep_data(struct mmci_host *host,
+			struct mmc_data *data,
+			bool next)
 {
+	struct mmci_dmae_priv *dmae = host->dma_priv;
+	struct mmci_dmae_next *nd = &dmae->next_data;
+
+	if (!host->use_dma)
+		return -EINVAL;
+
+	if (next)
+		return _mmci_dmae_prep_data(host, data, &nd->chan, &nd->desc);
 	/* Check if next job is already prepared. */
-	if (host->dma_current && host->dma_desc_current)
+	if (dmae->cur && dmae->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);
+	return _mmci_dmae_prep_data(host, data, &dmae->cur,
+				    &dmae->desc_current);
 }
 
-static int mmci_dma_start_data(struct mmci_host *host, unsigned int datactrl)
+int mmci_dmae_start(struct mmci_host *host, unsigned int *datactrl)
 {
-	int ret;
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 	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);
 	host->dma_in_progress = true;
-	dmaengine_submit(host->dma_desc_current);
-	dma_async_issue_pending(host->dma_current);
+	dmaengine_submit(dmae->desc_current);
+	dma_async_issue_pending(dmae->cur);
 
 	if (host->variant->qcom_dml)
 		dml_start_xfer(host, data);
 
-	datactrl |= MCI_DPSM_DMAENABLE;
+	*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)
+void mmci_dmae_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;
-}
+	struct mmci_dmae_priv *dmae = host->dma_priv;
+	struct mmci_dmae_next *next = &dmae->next_data;
 
-static void mmci_pre_request(struct mmc_host *mmc, struct mmc_request *mrq)
-{
-	struct mmci_host *host = mmc_priv(mmc);
-	struct mmc_data *data = mrq->data;
-	struct mmci_host_next *nd = &host->next_data;
-
-	if (!data)
+	if (!host->use_dma)
 		return;
 
-	BUG_ON(data->host_cookie);
+	WARN_ON(!data->host_cookie && (next->desc || next->chan));
 
-	if (mmci_validate_data(host, data))
-		return;
-
-	if (!mmci_dma_prep_next(host, data))
-		data->host_cookie = ++nd->cookie < 0 ? 1 : nd->cookie;
+	dmae->desc_current = next->desc;
+	dmae->cur = next->chan;
+	next->desc = NULL;
+	next->chan = NULL;
 }
 
-static void mmci_post_request(struct mmc_host *mmc, struct mmc_request *mrq,
-			      int err)
+void mmci_dmae_unprep_data(struct mmci_host *host,
+			   struct mmc_data *data, int err)
+
 {
-	struct mmci_host *host = mmc_priv(mmc);
-	struct mmc_data *data = mrq->data;
+	struct mmci_dmae_priv *dmae = host->dma_priv;
 
-	if (!data || !data->host_cookie)
+	if (!host->use_dma)
 		return;
 
 	mmci_dma_unmap(host, data);
 
 	if (err) {
-		struct mmci_host_next *next = &host->next_data;
+		struct mmci_dmae_next *next = &dmae->next_data;
 		struct dma_chan *chan;
 		if (data->flags & MMC_DATA_READ)
-			chan = host->dma_rx_channel;
+			chan = dmae->rx_channel;
 		else
-			chan = host->dma_tx_channel;
+			chan = dmae->tx_channel;
 		dmaengine_terminate_all(chan);
 
-		if (host->dma_desc_current == next->dma_desc)
-			host->dma_desc_current = NULL;
+		if (dmae->desc_current == next->desc)
+			dmae->desc_current = NULL;
 
-		if (host->dma_current == next->dma_chan) {
+		if (dmae->cur == next->chan) {
 			host->dma_in_progress = false;
-			host->dma_current = NULL;
+			dmae->cur = NULL;
 		}
 
-		next->dma_desc = NULL;
-		next->dma_chan = NULL;
-		data->host_cookie = 0;
+		next->desc = NULL;
+		next->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 struct mmci_host_ops mmci_variant_ops = {
+	.prep_data = mmci_dmae_prep_data,
+	.unprep_data = mmci_dmae_unprep_data,
+	.get_next_data = mmci_dmae_get_next_data,
+	.dma_setup = mmci_dmae_setup,
+	.dma_release = mmci_dmae_release,
+	.dma_start = mmci_dmae_start,
+	.dma_finalize = mmci_dmae_finalize,
+	.dma_error = mmci_dmae_error,
+};
 
-static inline void mmci_dma_release(struct mmci_host *host)
+void mmci_variant_init(struct mmci_host *host)
 {
+	host->ops = &mmci_variant_ops;
 }
+#endif
 
-static inline void mmci_dma_unmap(struct mmci_host *host, struct mmc_data *data)
+static void mmci_pre_request(struct mmc_host *mmc, struct mmc_request *mrq)
 {
-}
+	struct mmci_host *host = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
 
-static inline void mmci_dma_finalize(struct mmci_host *host,
-				     struct mmc_data *data)
-{
-}
+	if (!data)
+		return;
 
-static inline void mmci_dma_data_error(struct mmci_host *host)
-{
+	WARN_ON(data->host_cookie);
+
+	if (mmci_validate_data(host, data))
+		return;
+
+	mmci_prep_data(host, data, true);
 }
 
-static inline int mmci_dma_start_data(struct mmci_host *host, unsigned int datactrl)
+static void mmci_post_request(struct mmc_host *mmc, struct mmc_request *mrq,
+			      int err)
 {
-	return -ENOSYS;
-}
+	struct mmci_host *host = mmc_priv(mmc);
+	struct mmc_data *data = mrq->data;
 
-#define mmci_pre_request NULL
-#define mmci_post_request NULL
+	if (!data || !data->host_cookie)
+		return;
 
-#endif
+	mmci_unprep_data(host, data, err);
+}
 
 static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 {
@@ -793,11 +1020,11 @@ static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 	BUG_ON(1 << blksz_bits != data->blksz);
 
 	if (variant->blksz_datactrl16)
-		datactrl = MCI_DPSM_ENABLE | (data->blksz << 16);
+		datactrl = variant->datactrl_dpsm_enable | (data->blksz << 16);
 	else if (variant->blksz_datactrl4)
-		datactrl = MCI_DPSM_ENABLE | (data->blksz << 4);
+		datactrl = variant->datactrl_dpsm_enable | (data->blksz << 4);
 	else
-		datactrl = MCI_DPSM_ENABLE | blksz_bits << 4;
+		datactrl = variant->datactrl_dpsm_enable | blksz_bits << 4;
 
 	if (data->flags & MMC_DATA_READ)
 		datactrl |= MCI_DPSM_DIRECTION;
@@ -831,7 +1058,7 @@ static void mmci_start_data(struct mmci_host *host, struct mmc_data *data)
 	 * Attempt to use DMA operation mode, if this
 	 * should fail, fall back to PIO mode
 	 */
-	if (!mmci_dma_start_data(host, datactrl))
+	if (!mmci_dma_start(host, datactrl))
 		return;
 
 	/* IRQ mode, map the SG list for CPU reading/writing */
@@ -868,16 +1095,19 @@ mmci_start_command(struct mmci_host *host, struct mmc_command *cmd, u32 c)
 	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) {
+	if (readl(base + MMCICOMMAND) & host->variant->cmdreg_cpsm_enable) {
 		writel(0, base + MMCICOMMAND);
 		mmci_reg_delay(host);
 	}
 
-	c |= cmd->opcode | MCI_CPSM_ENABLE;
+	c |= cmd->opcode | host->variant->cmdreg_cpsm_enable;
 	if (cmd->flags & MMC_RSP_PRESENT) {
 		if (cmd->flags & MMC_RSP_136)
-			c |= MCI_CPSM_LONGRSP;
-		c |= MCI_CPSM_RESPONSE;
+			c |= host->variant->cmdreg_lrsp_crc;
+		else if (cmd->flags & MMC_RSP_CRC)
+			c |= host->variant->cmdreg_srsp_crc;
+		else
+			c |= host->variant->cmdreg_srsp;
 	}
 	if (/*interrupt*/0)
 		c |= MCI_CPSM_INTERRUPT;
@@ -895,21 +1125,22 @@ static void
 mmci_data_irq(struct mmci_host *host, struct mmc_data *data,
 	      unsigned int status)
 {
+	unsigned int status_err;
+
 	/* Make sure we have data to handle */
 	if (!data)
 		return;
 
 	/* First check for errors */
-	if (status & (MCI_DATACRCFAIL | MCI_DATATIMEOUT |
-		      host->variant->start_err |
-		      MCI_TXUNDERRUN | MCI_RXOVERRUN)) {
+	status_err = status & (host->variant->start_err |
+			       MCI_DATACRCFAIL | MCI_DATATIMEOUT |
+			       MCI_TXUNDERRUN | MCI_RXOVERRUN);
+
+	if (status_err) {
 		u32 remain, success;
 
 		/* Terminate the DMA transfer */
-		if (dma_inprogress(host)) {
-			mmci_dma_data_error(host);
-			mmci_dma_unmap(host, data);
-		}
+		mmci_dma_error(host);
 
 		/*
 		 * Calculate how far we are into the transfer.  Note that
@@ -918,22 +1149,26 @@ mmci_data_irq(struct mmci_host *host, struct mmc_data *data,
 		 * can be as much as a FIFO-worth of data ahead.  This
 		 * matters for FIFO overruns only.
 		 */
-		remain = readl(host->base + MMCIDATACNT);
-		success = data->blksz * data->blocks - remain;
+		if (!host->variant->datacnt_useless) {
+			remain = readl(host->base + MMCIDATACNT);
+			success = data->blksz * data->blocks - remain;
+		} else {
+			success = 0;
+		}
 
 		dev_dbg(mmc_dev(host->mmc), "MCI ERROR IRQ, status 0x%08x at 0x%08x\n",
-			status, success);
-		if (status & MCI_DATACRCFAIL) {
+			status_err, success);
+		if (status_err & MCI_DATACRCFAIL) {
 			/* Last block was not successful */
 			success -= 1;
 			data->error = -EILSEQ;
-		} else if (status & MCI_DATATIMEOUT) {
+		} else if (status_err & MCI_DATATIMEOUT) {
 			data->error = -ETIMEDOUT;
-		} else if (status & MCI_STARTBITERR) {
+		} else if (status_err & MCI_STARTBITERR) {
 			data->error = -ECOMM;
-		} else if (status & MCI_TXUNDERRUN) {
+		} else if (status_err & MCI_TXUNDERRUN) {
 			data->error = -EIO;
-		} else if (status & MCI_RXOVERRUN) {
+		} else if (status_err & MCI_RXOVERRUN) {
 			if (success > host->variant->fifosize)
 				success -= host->variant->fifosize;
 			else
@@ -947,8 +1182,8 @@ mmci_data_irq(struct mmci_host *host, struct mmc_data *data,
 		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_dma_finalize(host, data);
+
 		mmci_stop_data(host);
 
 		if (!data->error)
@@ -1055,16 +1290,15 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd,
 	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_dma_error(host);
+
 			mmci_stop_data(host);
 		}
 		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)) {
+	} else if (!host->variant->datactrl_first &&
+		   !(cmd->data->flags & MMC_DATA_READ)) {
 		mmci_start_data(host, cmd->data);
 	}
 }
@@ -1264,7 +1498,7 @@ static irqreturn_t mmci_irq(int irq, void *dev_id)
 			if (status & host->mask1_reg)
 				mmci_pio_irq(irq, dev_id);
 
-			status &= ~MCI_IRQ1MASK;
+			status &= ~host->variant->irq_pio_mask;
 		}
 
 		/*
@@ -1328,7 +1562,8 @@ static void mmci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	if (mrq->data)
 		mmci_get_next_data(host, mrq->data);
 
-	if (mrq->data && mrq->data->flags & MMC_DATA_READ)
+	if (mrq->data &&
+	    (host->variant->datactrl_first || mrq->data->flags & MMC_DATA_READ))
 		mmci_start_data(host, mrq->data);
 
 	if (mrq->sbc)
@@ -1438,8 +1673,16 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 	spin_lock_irqsave(&host->lock, flags);
 
-	mmci_set_clkreg(host, ios->clock);
-	mmci_write_pwrreg(host, pwr);
+	if (host->ops && host->ops->set_clkreg)
+		host->ops->set_clkreg(host, ios->clock);
+	else
+		mmci_set_clkreg(host, ios->clock);
+
+	if (host->ops && host->ops->set_pwrreg)
+		host->ops->set_pwrreg(host, pwr);
+	else
+		mmci_write_pwrreg(host, pwr);
+
 	mmci_reg_delay(host);
 
 	spin_unlock_irqrestore(&host->lock, flags);
@@ -1518,6 +1761,12 @@ static int mmci_of_parse(struct device_node *np, struct mmc_host *mmc)
 		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, "st,sig-dir", NULL))
+		host->pwr_reg_add |= MCI_STM32_DIRPOL;
+	if (of_get_property(np, "st,neg-edge", NULL))
+		host->clk_reg_add |= MCI_STM32_CLK_NEGEDGE;
+	if (of_get_property(np, "st,use-ckin", NULL))
+		host->clk_reg_add |= MCI_STM32_CLK_SELCKIN;
 
 	if (of_get_property(np, "mmc-cap-mmc-highspeed", NULL))
 		mmc->caps |= MMC_CAP_MMC_HIGHSPEED;
@@ -1644,6 +1893,8 @@ static int mmci_probe(struct amba_device *dev,
 	 */
 	if (variant->st_clkdiv)
 		mmc->f_min = DIV_ROUND_UP(host->mclk, 257);
+	else if (variant->stm32_clkdiv)
+		mmc->f_min = DIV_ROUND_UP(host->mclk, 2046);
 	else if (variant->explicit_mclk_control)
 		mmc->f_min = clk_round_rate(host->clk, 100000);
 	else
@@ -1665,6 +1916,12 @@ static int mmci_probe(struct amba_device *dev,
 
 	dev_dbg(mmc_dev(mmc), "clocking block at %u Hz\n", mmc->f_max);
 
+	host->rst = devm_reset_control_get_optional_exclusive(&dev->dev, NULL);
+	if (IS_ERR(host->rst)) {
+		ret = PTR_ERR(host->rst);
+		goto clk_disable;
+	}
+
 	/* Get regulators and the supported OCR mask */
 	ret = mmc_regulator_get_supply(mmc);
 	if (ret)
@@ -1675,13 +1932,6 @@ static int mmci_probe(struct amba_device *dev,
 	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;
 
@@ -1727,13 +1977,13 @@ static int mmci_probe(struct amba_device *dev,
 	/*
 	 * Block size can be up to 2048 bytes, but must be a power of two.
 	 */
-	mmc->max_blk_size = 1 << 11;
+	mmc->max_blk_size = 1 << variant->datactrl_blocksz;
 
 	/*
 	 * Limit the number of blocks transferred so that we don't overflow
 	 * the maximum request size.
 	 */
-	mmc->max_blk_count = mmc->max_req_size >> 11;
+	mmc->max_blk_count = mmc->max_req_size >> variant->datactrl_blocksz;
 
 	spin_lock_init(&host->lock);
 
@@ -1749,30 +1999,16 @@ static int mmci_probe(struct amba_device *dev,
 	 * - not using DT but using a descriptor table, or
 	 * - using a table of descriptors ALONGSIDE DT, or
 	 * look up these descriptors named "cd" and "wp" right here, fail
-	 * silently of these do not exist and proceed to try platform data
+	 * silently of these do not exist
 	 */
 	if (!np) {
 		ret = mmc_gpiod_request_cd(mmc, "cd", 0, false, 0, NULL);
-		if (ret < 0) {
-			if (ret == -EPROBE_DEFER)
-				goto clk_disable;
-			else if (gpio_is_valid(plat->gpio_cd)) {
-				ret = mmc_gpio_request_cd(mmc, plat->gpio_cd, 0);
-				if (ret)
-					goto clk_disable;
-			}
-		}
+		if (ret == -EPROBE_DEFER)
+			goto clk_disable;
 
 		ret = mmc_gpiod_request_ro(mmc, "wp", 0, false, 0, NULL);
-		if (ret < 0) {
-			if (ret == -EPROBE_DEFER)
-				goto clk_disable;
-			else if (gpio_is_valid(plat->gpio_wp)) {
-				ret = mmc_gpio_request_ro(mmc, plat->gpio_wp);
-				if (ret)
-					goto clk_disable;
-			}
-		}
+		if (ret == -EPROBE_DEFER)
+			goto clk_disable;
 	}
 
 	ret = devm_request_irq(&dev->dev, dev->irq[0], mmci_irq, IRQF_SHARED,
@@ -1789,7 +2025,7 @@ static int mmci_probe(struct amba_device *dev,
 			goto clk_disable;
 	}
 
-	writel(MCI_IRQENABLE, host->base + MMCIMASK0);
+	writel(MCI_IRQENABLE | variant->start_err, host->base + MMCIMASK0);
 
 	amba_set_drvdata(dev, mmc);
 
@@ -1876,7 +2112,8 @@ static void mmci_restore(struct mmci_host *host)
 		writel(host->datactrl_reg, host->base + MMCIDATACTRL);
 		writel(host->pwr_reg, host->base + MMCIPOWER);
 	}
-	writel(MCI_IRQENABLE, host->base + MMCIMASK0);
+	writel(MCI_IRQENABLE | host->variant->start_err,
+	       host->base + MMCIMASK0);
 	mmci_reg_delay(host);
 
 	spin_unlock_irqrestore(&host->lock, flags);
@@ -1971,6 +2208,11 @@ static const struct amba_id mmci_ids[] = {
 		.mask   = 0x00ffffff,
 		.data	= &variant_stm32,
 	},
+	{
+		.id     = 0x10153180,
+		.mask	= 0xf0ffffff,
+		.data	= &variant_stm32_sdmmc,
+	},
 	/* Qualcomm variants */
 	{
 		.id     = 0x00051180,
diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h
index 517591d219e9..550dd3914461 100644
--- a/drivers/mmc/host/mmci.h
+++ b/drivers/mmc/host/mmci.h
@@ -23,6 +23,14 @@
 #define MCI_ST_DATA31DIREN	(1 << 5)
 #define MCI_ST_FBCLKEN		(1 << 7)
 #define MCI_ST_DATA74DIREN	(1 << 8)
+/*
+ * The STM32 sdmmc does not have PWR_UP/OD/ROD
+ * and uses the power register for
+ */
+#define MCI_STM32_PWR_CYC	0x02
+#define MCI_STM32_VSWITCH	BIT(2)
+#define MCI_STM32_VSWITCHEN	BIT(3)
+#define MCI_STM32_DIRPOL	BIT(4)
 
 #define MMCICLOCK		0x004
 #define MCI_CLK_ENABLE		(1 << 8)
@@ -50,6 +58,19 @@
 #define MCI_QCOM_CLK_SELECT_IN_FBCLK	BIT(15)
 #define MCI_QCOM_CLK_SELECT_IN_DDR_MODE	(BIT(14) | BIT(15))
 
+/* Modified on STM32 sdmmc */
+#define MCI_STM32_CLK_CLKDIV_MSK	GENMASK(9, 0)
+#define MCI_STM32_CLK_WIDEBUS_4		BIT(14)
+#define MCI_STM32_CLK_WIDEBUS_8		BIT(15)
+#define MCI_STM32_CLK_NEGEDGE		BIT(16)
+#define MCI_STM32_CLK_HWFCEN		BIT(17)
+#define MCI_STM32_CLK_DDR		BIT(18)
+#define MCI_STM32_CLK_BUSSPEED		BIT(19)
+#define MCI_STM32_CLK_SEL_MSK		GENMASK(21, 20)
+#define MCI_STM32_CLK_SELCK		(0 << 20)
+#define MCI_STM32_CLK_SELCKIN		(1 << 20)
+#define MCI_STM32_CLK_SELFBCK		(2 << 20)
+
 #define MMCIARGUMENT		0x008
 
 /* The command register controls the Command Path State Machine (CPSM) */
@@ -72,6 +93,15 @@
 #define MCI_CPSM_QCOM_CCSDISABLE	BIT(15)
 #define MCI_CPSM_QCOM_AUTO_CMD19	BIT(16)
 #define MCI_CPSM_QCOM_AUTO_CMD21	BIT(21)
+/* Command register in STM32 sdmmc versions */
+#define MCI_CPSM_STM32_CMDTRANS		BIT(6)
+#define MCI_CPSM_STM32_CMDSTOP		BIT(7)
+#define MCI_CPSM_STM32_WAITRESP_MASK	GENMASK(9, 8)
+#define MCI_CPSM_STM32_NORSP		(0 << 8)
+#define MCI_CPSM_STM32_SRSP_CRC		(1 << 8)
+#define MCI_CPSM_STM32_SRSP		(2 << 8)
+#define MCI_CPSM_STM32_LRSP_CRC		(3 << 8)
+#define MCI_CPSM_STM32_ENABLE		BIT(12)
 
 #define MMCIRESPCMD		0x010
 #define MMCIRESPONSE0		0x014
@@ -130,6 +160,8 @@
 #define MCI_ST_SDIOIT		(1 << 22)
 #define MCI_ST_CEATAEND		(1 << 23)
 #define MCI_ST_CARDBUSY		(1 << 24)
+/* Extended status bits for the STM32 variants */
+#define MCI_STM32_BUSYD0	BIT(20)
 
 #define MMCICLEAR		0x038
 #define MCI_CMDCRCFAILCLR	(1 << 0)
@@ -175,21 +207,45 @@
 #define MCI_ST_SDIOITMASK	(1 << 22)
 #define MCI_ST_CEATAENDMASK	(1 << 23)
 #define MCI_ST_BUSYENDMASK	(1 << 24)
+/* Extended status bits for the STM32 variants */
+#define MCI_STM32_BUSYD0ENDMASK	BIT(21)
 
 #define MMCIMASK1		0x040
 #define MMCIFIFOCNT		0x048
 #define MMCIFIFO		0x080 /* to 0x0bc */
 
+/* STM32 sdmmc registers for IDMA (Internal DMA) */
+#define MMCI_STM32_IDMACTRLR	0x050
+#define MMCI_STM32_IDMAEN	BIT(0)
+#define MMCI_STM32_IDMALLIEN	BIT(1)
+
+#define MMCI_STM32_IDMABSIZER		0x054
+#define MMCI_STM32_IDMABNDT_SHIFT	5
+#define MMCI_STM32_IDMABNDT_MASK	GENMASK(12, 5)
+
+#define MMCI_STM32_IDMABASE0R	0x058
+
+#define MMCI_STM32_IDMALAR	0x64
+#define MMCI_STM32_IDMALA_MASK	GENMASK(13, 0)
+#define MMCI_STM32_ABR		BIT(29)
+#define MMCI_STM32_ULS		BIT(30)
+#define MMCI_STM32_ULA		BIT(31)
+
+#define MMCI_STM32_IDMABAR	0x68
+
 #define MCI_IRQENABLE	\
-	(MCI_CMDCRCFAILMASK|MCI_DATACRCFAILMASK|MCI_CMDTIMEOUTMASK|	\
-	MCI_DATATIMEOUTMASK|MCI_TXUNDERRUNMASK|MCI_RXOVERRUNMASK|	\
-	MCI_CMDRESPENDMASK|MCI_CMDSENTMASK|MCI_STARTBITERRMASK)
+	(MCI_CMDCRCFAILMASK | MCI_DATACRCFAILMASK | MCI_CMDTIMEOUTMASK | \
+	MCI_DATATIMEOUTMASK | MCI_TXUNDERRUNMASK | MCI_RXOVERRUNMASK |	\
+	MCI_CMDRESPENDMASK | MCI_CMDSENTMASK)
 
 /* These interrupts are directed to IRQ1 when two IRQ lines are available */
-#define MCI_IRQ1MASK \
+#define MCI_IRQ_PIO_MASK \
 	(MCI_RXFIFOHALFFULLMASK | MCI_RXDATAAVLBLMASK | \
 	 MCI_TXFIFOHALFEMPTYMASK)
 
+#define MCI_IRQ_PIO_STM32_MASK \
+	(MCI_RXFIFOHALFFULLMASK | MCI_TXFIFOHALFEMPTYMASK)
+
 #define NR_SG		128
 
 #define MMCI_PINCTRL_STATE_OPENDRAIN "opendrain"
@@ -204,6 +260,10 @@ struct mmci_host;
  * @clkreg_enable: enable value for MMCICLOCK register
  * @clkreg_8bit_bus_enable: enable value for 8 bit bus
  * @clkreg_neg_edge_enable: enable value for inverted data/cmd output
+ * @cmdreg_cpsm_enable: enable value for CPSM
+ * @cmdreg_lrsp_crc: enable value for long response with crc
+ * @cmdreg_srsp_crc: enable value for short response with crc
+ * @cmdreg_srsp: enable value for short response without crc
  * @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)
@@ -212,11 +272,17 @@ struct mmci_host;
  * @data_cmd_enable: enable value for data commands.
  * @st_sdio: enable ST specific SDIO logic
  * @st_clkdiv: true if using a ST-specific clock divider algorithm
+ * @stm32_clkdiv: true if using a STM32-specific clock divider algorithm
  * @datactrl_mask_ddrmode: ddr mode mask in datactrl register.
  * @blksz_datactrl16: true if Block size is at b16..b30 position in datactrl register
  * @blksz_datactrl4: true if Block size is at b4..b16 position in datactrl
  *		     register
  * @datactrl_mask_sdio: SDIO enable mask in datactrl register
+ * @datactrl_blksz: block size in power of two
+ * @datactrl_dpsm_enable: enable value for DPSM
+ * @datactrl_first: true if data must be setup before send command
+ * @datacnt_useless: true if you could not use datacnt register to read
+ *		     remaining data
  * @pwrreg_powerup: power up value for MMCIPOWER register
  * @f_max: maximum clk frequency supported by the controller.
  * @signal_direction: input/out direction of bus signals can be indicated
@@ -233,53 +299,75 @@ struct mmci_host;
  * @qcom_dml: enables qcom specific dma glue for dma transfers.
  * @reversed_irq_handling: handle data irq before cmd irq.
  * @mmcimask1: true if variant have a MMCIMASK1 register.
+ * @irq_pio_mask: bitmask used to manage interrupt pio transfert in mmcimask
+ *		  register
  * @start_err: bitmask identifying the STARTBITERR bit inside MMCISTATUS
  *	       register.
  * @opendrain: bitmask identifying the OPENDRAIN bit inside MMCIPOWER register
+ * @dma_lli: true if variant has dma link list feature.
+ * @stm32_idmabsize_mask: stm32 sdmmc idma buffer size.
  */
 struct variant_data {
 	unsigned int		clkreg;
 	unsigned int		clkreg_enable;
 	unsigned int		clkreg_8bit_bus_enable;
 	unsigned int		clkreg_neg_edge_enable;
+	unsigned int		cmdreg_cpsm_enable;
+	unsigned int		cmdreg_lrsp_crc;
+	unsigned int		cmdreg_srsp_crc;
+	unsigned int		cmdreg_srsp;
 	unsigned int		datalength_bits;
 	unsigned int		fifosize;
 	unsigned int		fifohalfsize;
 	unsigned int		data_cmd_enable;
 	unsigned int		datactrl_mask_ddrmode;
 	unsigned int		datactrl_mask_sdio;
-	bool			st_sdio;
-	bool			st_clkdiv;
-	bool			blksz_datactrl16;
-	bool			blksz_datactrl4;
+	unsigned int		datactrl_blocksz;
+	unsigned int		datactrl_dpsm_enable;
+	u8			datactrl_first:1;
+	u8			datacnt_useless:1;
+	u8			st_sdio:1;
+	u8			st_clkdiv:1;
+	u8			stm32_clkdiv:1;
+	u8			blksz_datactrl16:1;
+	u8			blksz_datactrl4:1;
 	u32			pwrreg_powerup;
 	u32			f_max;
-	bool			signal_direction;
-	bool			pwrreg_clkgate;
-	bool			busy_detect;
+	u8			signal_direction:1;
+	u8			pwrreg_clkgate:1;
+	u8			busy_detect:1;
 	u32			busy_dpsm_flag;
 	u32			busy_detect_flag;
 	u32			busy_detect_mask;
-	bool			pwrreg_nopower;
-	bool			explicit_mclk_control;
-	bool			qcom_fifo;
-	bool			qcom_dml;
-	bool			reversed_irq_handling;
-	bool			mmcimask1;
+	u8			pwrreg_nopower:1;
+	u8			explicit_mclk_control:1;
+	u8			qcom_fifo:1;
+	u8			qcom_dml:1;
+	u8			reversed_irq_handling:1;
+	u8			mmcimask1:1;
+	unsigned int		irq_pio_mask;
 	u32			start_err;
 	u32			opendrain;
+	u8			dma_lli:1;
+	u32			stm32_idmabsize_mask;
 	void (*init)(struct mmci_host *host);
 };
 
 /* mmci variant callbacks */
 struct mmci_host_ops {
-	void (*dma_setup)(struct mmci_host *host);
-};
-
-struct mmci_host_next {
-	struct dma_async_tx_descriptor	*dma_desc;
-	struct dma_chan			*dma_chan;
-	s32				cookie;
+	int (*validate_data)(struct mmci_host *host, struct mmc_data *data);
+	int (*prep_data)(struct mmci_host *host, struct mmc_data *data,
+			 bool next);
+	void (*unprep_data)(struct mmci_host *host, struct mmc_data *data,
+			    int err);
+	void (*get_next_data)(struct mmci_host *host, struct mmc_data *data);
+	int (*dma_setup)(struct mmci_host *host);
+	void (*dma_release)(struct mmci_host *host);
+	int (*dma_start)(struct mmci_host *host, unsigned int *datactrl);
+	void (*dma_finalize)(struct mmci_host *host, struct mmc_data *data);
+	void (*dma_error)(struct mmci_host *host);
+	void (*set_clkreg)(struct mmci_host *host, unsigned int desired);
+	void (*set_pwrreg)(struct mmci_host *host, unsigned int pwr);
 };
 
 struct mmci_host {
@@ -290,7 +378,9 @@ struct mmci_host {
 	struct mmc_data		*data;
 	struct mmc_host		*mmc;
 	struct clk		*clk;
-	bool			singleirq;
+	u8			singleirq:1;
+
+	struct reset_control	*rst;
 
 	spinlock_t		lock;
 
@@ -301,10 +391,11 @@ struct mmci_host {
 	u32			pwr_reg;
 	u32			pwr_reg_add;
 	u32			clk_reg;
+	u32			clk_reg_add;
 	u32			datactrl_reg;
 	u32			busy_status;
 	u32			mask1_reg;
-	bool			vqmmc_enabled;
+	u8			vqmmc_enabled:1;
 	struct mmci_platform_data *plat;
 	struct mmci_host_ops	*ops;
 	struct variant_data	*variant;
@@ -323,18 +414,25 @@ struct mmci_host {
 	unsigned int		size;
 	int (*get_rx_fifocnt)(struct mmci_host *h, u32 status, int remain);
 
-#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;
-	bool			dma_in_progress;
+	u8			use_dma:1;
+	u8			dma_in_progress:1;
+	void			*dma_priv;
 
-#define dma_inprogress(host)	((host)->dma_in_progress)
-#else
-#define dma_inprogress(host)	(0)
-#endif
+	s32			next_cookie;
 };
 
+#define dma_inprogress(host)	((host)->dma_in_progress)
+
+void mmci_write_clkreg(struct mmci_host *host, u32 clk);
+void mmci_write_pwrreg(struct mmci_host *host, u32 pwr);
+
+int mmci_dmae_prep_data(struct mmci_host *host, struct mmc_data *data,
+			bool next);
+void mmci_dmae_unprep_data(struct mmci_host *host, struct mmc_data *data,
+			   int err);
+void mmci_dmae_get_next_data(struct mmci_host *host, struct mmc_data *data);
+int mmci_dmae_setup(struct mmci_host *host);
+void mmci_dmae_release(struct mmci_host *host);
+int mmci_dmae_start(struct mmci_host *host, unsigned int *datactrl);
+void mmci_dmae_finalize(struct mmci_host *host, struct mmc_data *data);
+void mmci_dmae_error(struct mmci_host *host);
diff --git a/drivers/mmc/host/mmci_qcom_dml.c b/drivers/mmc/host/mmci_qcom_dml.c
index be3fab5db83f..25d0a75533ea 100644
--- a/drivers/mmc/host/mmci_qcom_dml.c
+++ b/drivers/mmc/host/mmci_qcom_dml.c
@@ -119,19 +119,23 @@ static int of_get_dml_pipe_index(struct device_node *np, const char *name)
 }
 
 /* Initialize the dml hardware connected to SD Card controller */
-static void qcom_dma_setup(struct mmci_host *host)
+static int qcom_dma_setup(struct mmci_host *host)
 {
 	u32 config;
 	void __iomem *base;
 	int consumer_id, producer_id;
 	struct device_node *np = host->mmc->parent->of_node;
 
+	if (mmci_dmae_setup(host))
+		return -EINVAL;
+
 	consumer_id = of_get_dml_pipe_index(np, "tx");
 	producer_id = of_get_dml_pipe_index(np, "rx");
 
 	if (producer_id < 0 || consumer_id < 0) {
 		host->variant->qcom_dml = false;
-		return;
+		mmci_dmae_release(host);
+		return -EINVAL;
 	}
 
 	base = host->base + DML_OFFSET;
@@ -175,10 +179,19 @@ static void qcom_dma_setup(struct mmci_host *host)
 
 	/* Make sure dml initialization is finished */
 	mb();
+
+	return 0;
 }
 
 static struct mmci_host_ops qcom_variant_ops = {
+	.prep_data = mmci_dmae_prep_data,
+	.unprep_data = mmci_dmae_unprep_data,
+	.get_next_data = mmci_dmae_get_next_data,
 	.dma_setup = qcom_dma_setup,
+	.dma_release = mmci_dmae_release,
+	.dma_start = mmci_dmae_start,
+	.dma_finalize = mmci_dmae_finalize,
+	.dma_error = mmci_dmae_error,
 };
 
 void qcom_variant_init(struct mmci_host *host)
diff --git a/drivers/mmc/host/mmci_stm32_sdmmc.c b/drivers/mmc/host/mmci_stm32_sdmmc.c
new file mode 100644
index 000000000000..cfbfc6f1048f
--- /dev/null
+++ b/drivers/mmc/host/mmci_stm32_sdmmc.c
@@ -0,0 +1,282 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) STMicroelectronics 2018 - All Rights Reserved
+ * Author: Ludovic.barre@st.com for STMicroelectronics.
+ */
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/card.h>
+#include <linux/reset.h>
+#include <linux/scatterlist.h>
+#include "mmci.h"
+
+#define SDMMC_LLI_BUF_LEN	PAGE_SIZE
+#define SDMMC_IDMA_BURST	BIT(MMCI_STM32_IDMABNDT_SHIFT)
+
+struct sdmmc_lli_desc {
+	u32 idmalar;
+	u32 idmabase;
+	u32 idmasize;
+};
+
+struct sdmmc_priv {
+	dma_addr_t sg_dma;
+	void *sg_cpu;
+};
+
+int sdmmc_idma_validate_data(struct mmci_host *host,
+			     struct mmc_data *data)
+{
+	struct scatterlist *sg;
+	int i;
+
+	/*
+	 * idma has constraints on idmabase & idmasize for each element
+	 * excepted the last element which has no constraint on idmasize
+	 */
+	for_each_sg(data->sg, sg, data->sg_len - 1, i) {
+		if (!IS_ALIGNED(sg_dma_address(data->sg), sizeof(u32)) ||
+		    !IS_ALIGNED(sg_dma_len(data->sg), SDMMC_IDMA_BURST)) {
+			dev_err(mmc_dev(host->mmc),
+				"unaligned scatterlist: ofst:%x length:%d\n",
+				data->sg->offset, data->sg->length);
+			return -EINVAL;
+		}
+	}
+
+	if (!IS_ALIGNED(sg_dma_address(data->sg), sizeof(u32))) {
+		dev_err(mmc_dev(host->mmc),
+			"unaligned last scatterlist: ofst:%x length:%d\n",
+			data->sg->offset, data->sg->length);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int _sdmmc_idma_prep_data(struct mmci_host *host,
+				 struct mmc_data *data)
+{
+	int n_elem;
+
+	n_elem = dma_map_sg(mmc_dev(host->mmc),
+			    data->sg,
+			    data->sg_len,
+			    mmc_get_dma_dir(data));
+
+	if (!n_elem) {
+		dev_err(mmc_dev(host->mmc), "dma_map_sg failed\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int sdmmc_idma_prep_data(struct mmci_host *host,
+				struct mmc_data *data, bool next)
+{
+	/* Check if job is already prepared. */
+	if (!next && data->host_cookie == host->next_cookie)
+		return 0;
+
+	return _sdmmc_idma_prep_data(host, data);
+}
+
+static void sdmmc_idma_unprep_data(struct mmci_host *host,
+				   struct mmc_data *data, int err)
+{
+	dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len,
+		     mmc_get_dma_dir(data));
+}
+
+static int sdmmc_idma_setup(struct mmci_host *host)
+{
+	struct sdmmc_priv *idma;
+
+	idma = devm_kzalloc(mmc_dev(host->mmc), sizeof(*idma), GFP_KERNEL);
+	if (!idma)
+		return -ENOMEM;
+
+	host->dma_priv = idma;
+
+	if (host->variant->dma_lli) {
+		idma->sg_cpu = dmam_alloc_coherent(mmc_dev(host->mmc),
+						   SDMMC_LLI_BUF_LEN,
+						   &idma->sg_dma, GFP_KERNEL);
+		if (!idma->sg_cpu) {
+			dev_err(mmc_dev(host->mmc),
+				"Failed to alloc IDMA descriptor\n");
+			return -ENOMEM;
+		}
+		host->mmc->max_segs = SDMMC_LLI_BUF_LEN /
+			sizeof(struct sdmmc_lli_desc);
+		host->mmc->max_seg_size = host->variant->stm32_idmabsize_mask;
+	} else {
+		host->mmc->max_segs = 1;
+		host->mmc->max_seg_size = host->mmc->max_req_size;
+	}
+
+	return 0;
+}
+
+static int sdmmc_idma_start(struct mmci_host *host, unsigned int *datactrl)
+
+{
+	struct sdmmc_priv *idma = host->dma_priv;
+	struct sdmmc_lli_desc *desc = (struct sdmmc_lli_desc *)idma->sg_cpu;
+	struct mmc_data *data = host->data;
+	struct scatterlist *sg;
+	int i;
+
+	if (!host->variant->dma_lli || data->sg_len == 1) {
+		writel_relaxed(sg_dma_address(data->sg),
+			       host->base + MMCI_STM32_IDMABASE0R);
+		writel_relaxed(MMCI_STM32_IDMAEN,
+			       host->base + MMCI_STM32_IDMACTRLR);
+		return 0;
+	}
+
+	for_each_sg(data->sg, sg, data->sg_len, i) {
+		desc[i].idmalar = (i + 1) * sizeof(struct sdmmc_lli_desc);
+		desc[i].idmalar |= MMCI_STM32_ULA | MMCI_STM32_ULS
+			| MMCI_STM32_ABR;
+		desc[i].idmabase = sg_dma_address(sg);
+		desc[i].idmasize = sg_dma_len(sg);
+	}
+
+	/* notice the end of link list */
+	desc[data->sg_len - 1].idmalar &= ~MMCI_STM32_ULA;
+
+	dma_wmb();
+	writel_relaxed(idma->sg_dma, host->base + MMCI_STM32_IDMABAR);
+	writel_relaxed(desc[0].idmalar, host->base + MMCI_STM32_IDMALAR);
+	writel_relaxed(desc[0].idmabase, host->base + MMCI_STM32_IDMABASE0R);
+	writel_relaxed(desc[0].idmasize, host->base + MMCI_STM32_IDMABSIZER);
+	writel_relaxed(MMCI_STM32_IDMAEN | MMCI_STM32_IDMALLIEN,
+		       host->base + MMCI_STM32_IDMACTRLR);
+
+	return 0;
+}
+
+static void sdmmc_idma_finalize(struct mmci_host *host, struct mmc_data *data)
+{
+	writel_relaxed(0, host->base + MMCI_STM32_IDMACTRLR);
+}
+
+static void mmci_sdmmc_set_clkreg(struct mmci_host *host, unsigned int desired)
+{
+	unsigned int clk = 0, ddr = 0;
+
+	if (host->mmc->ios.timing == MMC_TIMING_MMC_DDR52 ||
+	    host->mmc->ios.timing == MMC_TIMING_UHS_DDR50)
+		ddr = MCI_STM32_CLK_DDR;
+
+	/*
+	 * cclk = mclk / (2 * clkdiv)
+	 * clkdiv 0 => bypass
+	 * in ddr mode bypass is not possible
+	 */
+	if (desired) {
+		if (desired >= host->mclk && !ddr) {
+			host->cclk = host->mclk;
+		} else {
+			clk = DIV_ROUND_UP(host->mclk, 2 * desired);
+			if (clk > MCI_STM32_CLK_CLKDIV_MSK)
+				clk = MCI_STM32_CLK_CLKDIV_MSK;
+			host->cclk = host->mclk / (2 * clk);
+		}
+	} else {
+		/*
+		 * while power-on phase the clock can't be define to 0,
+		 * Only power-off and power-cyc deactivate the clock.
+		 * if desired clock is 0, set max divider
+		 */
+		clk = MCI_STM32_CLK_CLKDIV_MSK;
+		host->cclk = host->mclk / (2 * clk);
+	}
+
+	/* Set actual clock for debug */
+	if (host->mmc->ios.power_mode == MMC_POWER_ON)
+		host->mmc->actual_clock = host->cclk;
+	else
+		host->mmc->actual_clock = 0;
+
+	if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_4)
+		clk |= MCI_STM32_CLK_WIDEBUS_4;
+	if (host->mmc->ios.bus_width == MMC_BUS_WIDTH_8)
+		clk |= MCI_STM32_CLK_WIDEBUS_8;
+
+	clk |= MCI_STM32_CLK_HWFCEN;
+	clk |= host->clk_reg_add;
+	clk |= ddr;
+
+	/*
+	 * SDMMC_FBCK is selected when an external Delay Block is needed
+	 * with SDR104.
+	 */
+	if (host->mmc->ios.timing >= MMC_TIMING_UHS_SDR50) {
+		clk |= MCI_STM32_CLK_BUSSPEED;
+		if (host->mmc->ios.timing == MMC_TIMING_UHS_SDR104) {
+			clk &= ~MCI_STM32_CLK_SEL_MSK;
+			clk |= MCI_STM32_CLK_SELFBCK;
+		}
+	}
+
+	mmci_write_clkreg(host, clk);
+}
+
+static void mmci_sdmmc_set_pwrreg(struct mmci_host *host, unsigned int pwr)
+{
+	struct mmc_ios ios = host->mmc->ios;
+
+	pwr = host->pwr_reg_add;
+
+	if (ios.power_mode == MMC_POWER_OFF) {
+		/* Only a reset could power-off sdmmc */
+		reset_control_assert(host->rst);
+		udelay(2);
+		reset_control_deassert(host->rst);
+
+		/*
+		 * Set the SDMMC in Power-cycle state.
+		 * This will make that the SDMMC_D[7:0], SDMMC_CMD and SDMMC_CK
+		 * are driven low, to prevent the Card from being supplied
+		 * through the signal lines.
+		 */
+		mmci_write_pwrreg(host, MCI_STM32_PWR_CYC | pwr);
+	} else if (ios.power_mode == MMC_POWER_ON) {
+		/*
+		 * After power-off (reset): the irq mask defined in probe
+		 * functionis lost
+		 * ault irq mask (probe) must be activated
+		 */
+		writel(MCI_IRQENABLE | host->variant->start_err,
+		       host->base + MMCIMASK0);
+
+		/*
+		 * After a power-cycle state, we must set the SDMMC in
+		 * Power-off. The SDMMC_D[7:0], SDMMC_CMD and SDMMC_CK are
+		 * driven high. Then we can set the SDMMC to Power-on state
+		 */
+		mmci_write_pwrreg(host, MCI_PWR_OFF | pwr);
+		mdelay(1);
+		mmci_write_pwrreg(host, MCI_PWR_ON | pwr);
+	}
+}
+
+static struct mmci_host_ops sdmmc_variant_ops = {
+	.validate_data = sdmmc_idma_validate_data,
+	.prep_data = sdmmc_idma_prep_data,
+	.unprep_data = sdmmc_idma_unprep_data,
+	.dma_setup = sdmmc_idma_setup,
+	.dma_start = sdmmc_idma_start,
+	.dma_finalize = sdmmc_idma_finalize,
+	.set_clkreg = mmci_sdmmc_set_clkreg,
+	.set_pwrreg = mmci_sdmmc_set_pwrreg,
+};
+
+void sdmmc_variant_init(struct mmci_host *host)
+{
+	host->ops = &sdmmc_variant_ops;
+}
diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
index 04841386b65d..6334cc752d8b 100644
--- a/drivers/mmc/host/mtk-sd.c
+++ b/drivers/mmc/host/mtk-sd.c
@@ -87,6 +87,13 @@
 #define SDC_FIFO_CFG     0x228
 
 /*--------------------------------------------------------------------------*/
+/* Top Pad Register Offset                                                  */
+/*--------------------------------------------------------------------------*/
+#define EMMC_TOP_CONTROL	0x00
+#define EMMC_TOP_CMD		0x04
+#define EMMC50_PAD_DS_TUNE	0x0c
+
+/*--------------------------------------------------------------------------*/
 /* Register Mask                                                            */
 /*--------------------------------------------------------------------------*/
 
@@ -261,6 +268,23 @@
 #define SDC_FIFO_CFG_WRVALIDSEL   (0x1 << 24)  /* RW */
 #define SDC_FIFO_CFG_RDVALIDSEL   (0x1 << 25)  /* RW */
 
+/* EMMC_TOP_CONTROL mask */
+#define PAD_RXDLY_SEL           (0x1 << 0)      /* RW */
+#define DELAY_EN                (0x1 << 1)      /* RW */
+#define PAD_DAT_RD_RXDLY2       (0x1f << 2)     /* RW */
+#define PAD_DAT_RD_RXDLY        (0x1f << 7)     /* RW */
+#define PAD_DAT_RD_RXDLY2_SEL   (0x1 << 12)     /* RW */
+#define PAD_DAT_RD_RXDLY_SEL    (0x1 << 13)     /* RW */
+#define DATA_K_VALUE_SEL        (0x1 << 14)     /* RW */
+#define SDC_RX_ENH_EN           (0x1 << 15)     /* TW */
+
+/* EMMC_TOP_CMD mask */
+#define PAD_CMD_RXDLY2          (0x1f << 0)     /* RW */
+#define PAD_CMD_RXDLY           (0x1f << 5)     /* RW */
+#define PAD_CMD_RD_RXDLY2_SEL   (0x1 << 10)     /* RW */
+#define PAD_CMD_RD_RXDLY_SEL    (0x1 << 11)     /* RW */
+#define PAD_CMD_TX_DLY          (0x1f << 12)    /* RW */
+
 #define REQ_CMD_EIO  (0x1 << 0)
 #define REQ_CMD_TMO  (0x1 << 1)
 #define REQ_DAT_ERR  (0x1 << 2)
@@ -333,6 +357,9 @@ struct msdc_save_para {
 	u32 emmc50_cfg0;
 	u32 emmc50_cfg3;
 	u32 sdc_fifo_cfg;
+	u32 emmc_top_control;
+	u32 emmc_top_cmd;
+	u32 emmc50_pad_ds_tune;
 };
 
 struct mtk_mmc_compatible {
@@ -351,6 +378,8 @@ struct msdc_tune_para {
 	u32 iocon;
 	u32 pad_tune;
 	u32 pad_cmd_tune;
+	u32 emmc_top_control;
+	u32 emmc_top_cmd;
 };
 
 struct msdc_delay_phase {
@@ -372,6 +401,7 @@ struct msdc_host {
 	int error;
 
 	void __iomem *base;		/* host base address */
+	void __iomem *top_base;		/* host top register base address */
 
 	struct msdc_dma dma;	/* dma channel */
 	u64 dma_mask;
@@ -387,10 +417,10 @@ struct msdc_host {
 
 	struct clk *src_clk;	/* msdc source clock */
 	struct clk *h_clk;      /* msdc h_clk */
+	struct clk *bus_clk;	/* bus clock which used to access register */
 	struct clk *src_clk_cg; /* msdc source clock control gate */
 	u32 mclk;		/* mmc subsystem clock frequency */
 	u32 src_clk_freq;	/* source clock frequency */
-	u32 sclk;		/* SD/MS bus clock frequency */
 	unsigned char timing;
 	bool vqmmc_enabled;
 	u32 latch_ck;
@@ -429,6 +459,18 @@ static const struct mtk_mmc_compatible mt8173_compat = {
 	.support_64g = false,
 };
 
+static const struct mtk_mmc_compatible mt8183_compat = {
+	.clk_div_bits = 12,
+	.hs400_tune = false,
+	.pad_tune_reg = MSDC_PAD_TUNE0,
+	.async_fifo = true,
+	.data_tune = true,
+	.busy_check = true,
+	.stop_clk_fix = true,
+	.enhance_rx = true,
+	.support_64g = true,
+};
+
 static const struct mtk_mmc_compatible mt2701_compat = {
 	.clk_div_bits = 12,
 	.hs400_tune = false,
@@ -468,6 +510,7 @@ static const struct mtk_mmc_compatible mt7622_compat = {
 static const struct of_device_id msdc_of_ids[] = {
 	{ .compatible = "mediatek,mt8135-mmc", .data = &mt8135_compat},
 	{ .compatible = "mediatek,mt8173-mmc", .data = &mt8173_compat},
+	{ .compatible = "mediatek,mt8183-mmc", .data = &mt8183_compat},
 	{ .compatible = "mediatek,mt2701-mmc", .data = &mt2701_compat},
 	{ .compatible = "mediatek,mt2712-mmc", .data = &mt2712_compat},
 	{ .compatible = "mediatek,mt7622-mmc", .data = &mt7622_compat},
@@ -635,10 +678,10 @@ static void msdc_set_timeout(struct msdc_host *host, u32 ns, u32 clks)
 
 	host->timeout_ns = ns;
 	host->timeout_clks = clks;
-	if (host->sclk == 0) {
+	if (host->mmc->actual_clock == 0) {
 		timeout = 0;
 	} else {
-		clk_ns  = 1000000000UL / host->sclk;
+		clk_ns  = 1000000000UL / host->mmc->actual_clock;
 		timeout = (ns + clk_ns - 1) / clk_ns + clks;
 		/* in 1048576 sclk cycle unit */
 		timeout = (timeout + (0x1 << 20) - 1) >> 20;
@@ -660,12 +703,14 @@ static void msdc_gate_clock(struct msdc_host *host)
 {
 	clk_disable_unprepare(host->src_clk_cg);
 	clk_disable_unprepare(host->src_clk);
+	clk_disable_unprepare(host->bus_clk);
 	clk_disable_unprepare(host->h_clk);
 }
 
 static void msdc_ungate_clock(struct msdc_host *host)
 {
 	clk_prepare_enable(host->h_clk);
+	clk_prepare_enable(host->bus_clk);
 	clk_prepare_enable(host->src_clk);
 	clk_prepare_enable(host->src_clk_cg);
 	while (!(readl(host->base + MSDC_CFG) & MSDC_CFG_CKSTB))
@@ -683,6 +728,7 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	if (!hz) {
 		dev_dbg(host->dev, "set mclk to 0\n");
 		host->mclk = 0;
+		host->mmc->actual_clock = 0;
 		sdr_clr_bits(host->base + MSDC_CFG, MSDC_CFG_CKPDN);
 		return;
 	}
@@ -761,7 +807,7 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	while (!(readl(host->base + MSDC_CFG) & MSDC_CFG_CKSTB))
 		cpu_relax();
 	sdr_set_bits(host->base + MSDC_CFG, MSDC_CFG_CKPDN);
-	host->sclk = sclk;
+	host->mmc->actual_clock = sclk;
 	host->mclk = hz;
 	host->timing = timing;
 	/* need because clk changed. */
@@ -772,14 +818,30 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	 * mmc_select_hs400() will drop to 50Mhz and High speed mode,
 	 * tune result of hs200/200Mhz is not suitable for 50Mhz
 	 */
-	if (host->sclk <= 52000000) {
+	if (host->mmc->actual_clock <= 52000000) {
 		writel(host->def_tune_para.iocon, host->base + MSDC_IOCON);
-		writel(host->def_tune_para.pad_tune, host->base + tune_reg);
+		if (host->top_base) {
+			writel(host->def_tune_para.emmc_top_control,
+			       host->top_base + EMMC_TOP_CONTROL);
+			writel(host->def_tune_para.emmc_top_cmd,
+			       host->top_base + EMMC_TOP_CMD);
+		} else {
+			writel(host->def_tune_para.pad_tune,
+			       host->base + tune_reg);
+		}
 	} else {
 		writel(host->saved_tune_para.iocon, host->base + MSDC_IOCON);
-		writel(host->saved_tune_para.pad_tune, host->base + tune_reg);
 		writel(host->saved_tune_para.pad_cmd_tune,
 		       host->base + PAD_CMD_TUNE);
+		if (host->top_base) {
+			writel(host->saved_tune_para.emmc_top_control,
+			       host->top_base + EMMC_TOP_CONTROL);
+			writel(host->saved_tune_para.emmc_top_cmd,
+			       host->top_base + EMMC_TOP_CMD);
+		} else {
+			writel(host->saved_tune_para.pad_tune,
+			       host->base + tune_reg);
+		}
 	}
 
 	if (timing == MMC_TIMING_MMC_HS400 &&
@@ -787,7 +849,8 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 		sdr_set_field(host->base + PAD_CMD_TUNE,
 			      MSDC_PAD_TUNE_CMDRRDLY,
 			      host->hs400_cmd_int_delay);
-	dev_dbg(host->dev, "sclk: %d, timing: %d\n", host->sclk, timing);
+	dev_dbg(host->dev, "sclk: %d, timing: %d\n", host->mmc->actual_clock,
+		timing);
 }
 
 static inline u32 msdc_cmd_find_resp(struct msdc_host *host,
@@ -1055,6 +1118,7 @@ static void msdc_start_command(struct msdc_host *host,
 	WARN_ON(host->cmd);
 	host->cmd = cmd;
 
+	mod_delayed_work(system_wq, &host->req_timeout, DAT_TIMEOUT);
 	if (!msdc_cmd_is_ready(host, mrq, cmd))
 		return;
 
@@ -1066,7 +1130,6 @@ static void msdc_start_command(struct msdc_host *host,
 
 	cmd->error = 0;
 	rawcmd = msdc_cmd_prepare_raw_cmd(host, mrq, cmd);
-	mod_delayed_work(system_wq, &host->req_timeout, DAT_TIMEOUT);
 
 	sdr_set_bits(host->base + MSDC_INTEN, cmd_ints_mask);
 	writel(cmd->arg, host->base + SDC_ARG);
@@ -1351,7 +1414,12 @@ static void msdc_init_hw(struct msdc_host *host)
 	val = readl(host->base + MSDC_INT);
 	writel(val, host->base + MSDC_INT);
 
-	writel(0, host->base + tune_reg);
+	if (host->top_base) {
+		writel(0, host->top_base + EMMC_TOP_CONTROL);
+		writel(0, host->top_base + EMMC_TOP_CMD);
+	} else {
+		writel(0, host->base + tune_reg);
+	}
 	writel(0, host->base + MSDC_IOCON);
 	sdr_set_field(host->base + MSDC_IOCON, MSDC_IOCON_DDLSEL, 0);
 	writel(0x403c0046, host->base + MSDC_PATCH_BIT);
@@ -1375,8 +1443,12 @@ static void msdc_init_hw(struct msdc_host *host)
 		sdr_set_field(host->base + MSDC_PATCH_BIT2,
 			      MSDC_PB2_RESPWAIT, 3);
 		if (host->dev_comp->enhance_rx) {
-			sdr_set_bits(host->base + SDC_ADV_CFG0,
-				     SDC_RX_ENHANCE_EN);
+			if (host->top_base)
+				sdr_set_bits(host->top_base + EMMC_TOP_CONTROL,
+					     SDC_RX_ENH_EN);
+			else
+				sdr_set_bits(host->base + SDC_ADV_CFG0,
+					     SDC_RX_ENHANCE_EN);
 		} else {
 			sdr_set_field(host->base + MSDC_PATCH_BIT2,
 				      MSDC_PB2_RESPSTSENSEL, 2);
@@ -1394,11 +1466,26 @@ static void msdc_init_hw(struct msdc_host *host)
 		sdr_set_bits(host->base + MSDC_PATCH_BIT2,
 			     MSDC_PB2_SUPPORT_64G);
 	if (host->dev_comp->data_tune) {
-		sdr_set_bits(host->base + tune_reg,
-			     MSDC_PAD_TUNE_RD_SEL | MSDC_PAD_TUNE_CMD_SEL);
+		if (host->top_base) {
+			sdr_set_bits(host->top_base + EMMC_TOP_CONTROL,
+				     PAD_DAT_RD_RXDLY_SEL);
+			sdr_clr_bits(host->top_base + EMMC_TOP_CONTROL,
+				     DATA_K_VALUE_SEL);
+			sdr_set_bits(host->top_base + EMMC_TOP_CMD,
+				     PAD_CMD_RD_RXDLY_SEL);
+		} else {
+			sdr_set_bits(host->base + tune_reg,
+				     MSDC_PAD_TUNE_RD_SEL |
+				     MSDC_PAD_TUNE_CMD_SEL);
+		}
 	} else {
 		/* choose clock tune */
-		sdr_set_bits(host->base + tune_reg, MSDC_PAD_TUNE_RXDLYSEL);
+		if (host->top_base)
+			sdr_set_bits(host->top_base + EMMC_TOP_CONTROL,
+				     PAD_RXDLY_SEL);
+		else
+			sdr_set_bits(host->base + tune_reg,
+				     MSDC_PAD_TUNE_RXDLYSEL);
 	}
 
 	/* Configure to enable SDIO mode.
@@ -1413,9 +1500,20 @@ static void msdc_init_hw(struct msdc_host *host)
 	sdr_set_field(host->base + SDC_CFG, SDC_CFG_DTOC, 3);
 
 	host->def_tune_para.iocon = readl(host->base + MSDC_IOCON);
-	host->def_tune_para.pad_tune = readl(host->base + tune_reg);
 	host->saved_tune_para.iocon = readl(host->base + MSDC_IOCON);
-	host->saved_tune_para.pad_tune = readl(host->base + tune_reg);
+	if (host->top_base) {
+		host->def_tune_para.emmc_top_control =
+			readl(host->top_base + EMMC_TOP_CONTROL);
+		host->def_tune_para.emmc_top_cmd =
+			readl(host->top_base + EMMC_TOP_CMD);
+		host->saved_tune_para.emmc_top_control =
+			readl(host->top_base + EMMC_TOP_CONTROL);
+		host->saved_tune_para.emmc_top_cmd =
+			readl(host->top_base + EMMC_TOP_CMD);
+	} else {
+		host->def_tune_para.pad_tune = readl(host->base + tune_reg);
+		host->saved_tune_para.pad_tune = readl(host->base + tune_reg);
+	}
 	dev_dbg(host->dev, "init hardware done!");
 }
 
@@ -1563,6 +1661,30 @@ static struct msdc_delay_phase get_best_delay(struct msdc_host *host, u32 delay)
 	return delay_phase;
 }
 
+static inline void msdc_set_cmd_delay(struct msdc_host *host, u32 value)
+{
+	u32 tune_reg = host->dev_comp->pad_tune_reg;
+
+	if (host->top_base)
+		sdr_set_field(host->top_base + EMMC_TOP_CMD, PAD_CMD_RXDLY,
+			      value);
+	else
+		sdr_set_field(host->base + tune_reg, MSDC_PAD_TUNE_CMDRDLY,
+			      value);
+}
+
+static inline void msdc_set_data_delay(struct msdc_host *host, u32 value)
+{
+	u32 tune_reg = host->dev_comp->pad_tune_reg;
+
+	if (host->top_base)
+		sdr_set_field(host->top_base + EMMC_TOP_CONTROL,
+			      PAD_DAT_RD_RXDLY, value);
+	else
+		sdr_set_field(host->base + tune_reg, MSDC_PAD_TUNE_DATRRDLY,
+			      value);
+}
+
 static int msdc_tune_response(struct mmc_host *mmc, u32 opcode)
 {
 	struct msdc_host *host = mmc_priv(mmc);
@@ -1583,8 +1705,7 @@ static int msdc_tune_response(struct mmc_host *mmc, u32 opcode)
 
 	sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
 	for (i = 0 ; i < PAD_DELAY_MAX; i++) {
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_CMDRDLY, i);
+		msdc_set_cmd_delay(host, i);
 		/*
 		 * Using the same parameters, it may sometimes pass the test,
 		 * but sometimes it may fail. To make sure the parameters are
@@ -1608,8 +1729,7 @@ static int msdc_tune_response(struct mmc_host *mmc, u32 opcode)
 
 	sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
 	for (i = 0; i < PAD_DELAY_MAX; i++) {
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_CMDRDLY, i);
+		msdc_set_cmd_delay(host, i);
 		/*
 		 * Using the same parameters, it may sometimes pass the test,
 		 * but sometimes it may fail. To make sure the parameters are
@@ -1633,15 +1753,13 @@ skip_fall:
 		final_maxlen = final_fall_delay.maxlen;
 	if (final_maxlen == final_rise_delay.maxlen) {
 		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
-		sdr_set_field(host->base + tune_reg, MSDC_PAD_TUNE_CMDRDLY,
-			      final_rise_delay.final_phase);
 		final_delay = final_rise_delay.final_phase;
 	} else {
 		sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
-		sdr_set_field(host->base + tune_reg, MSDC_PAD_TUNE_CMDRDLY,
-			      final_fall_delay.final_phase);
 		final_delay = final_fall_delay.final_phase;
 	}
+	msdc_set_cmd_delay(host, final_delay);
+
 	if (host->dev_comp->async_fifo || host->hs200_cmd_int_delay)
 		goto skip_internal;
 
@@ -1716,7 +1834,6 @@ static int msdc_tune_data(struct mmc_host *mmc, u32 opcode)
 	u32 rise_delay = 0, fall_delay = 0;
 	struct msdc_delay_phase final_rise_delay, final_fall_delay = { 0,};
 	u8 final_delay, final_maxlen;
-	u32 tune_reg = host->dev_comp->pad_tune_reg;
 	int i, ret;
 
 	sdr_set_field(host->base + MSDC_PATCH_BIT, MSDC_INT_DAT_LATCH_CK_SEL,
@@ -1724,8 +1841,7 @@ static int msdc_tune_data(struct mmc_host *mmc, u32 opcode)
 	sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_DSPL);
 	sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_W_DSPL);
 	for (i = 0 ; i < PAD_DELAY_MAX; i++) {
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_DATRRDLY, i);
+		msdc_set_data_delay(host, i);
 		ret = mmc_send_tuning(mmc, opcode, NULL);
 		if (!ret)
 			rise_delay |= (1 << i);
@@ -1739,8 +1855,7 @@ static int msdc_tune_data(struct mmc_host *mmc, u32 opcode)
 	sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_DSPL);
 	sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_W_DSPL);
 	for (i = 0; i < PAD_DELAY_MAX; i++) {
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_DATRRDLY, i);
+		msdc_set_data_delay(host, i);
 		ret = mmc_send_tuning(mmc, opcode, NULL);
 		if (!ret)
 			fall_delay |= (1 << i);
@@ -1752,29 +1867,97 @@ skip_fall:
 	if (final_maxlen == final_rise_delay.maxlen) {
 		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_DSPL);
 		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_W_DSPL);
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_DATRRDLY,
-			      final_rise_delay.final_phase);
 		final_delay = final_rise_delay.final_phase;
 	} else {
 		sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_DSPL);
 		sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_W_DSPL);
-		sdr_set_field(host->base + tune_reg,
-			      MSDC_PAD_TUNE_DATRRDLY,
-			      final_fall_delay.final_phase);
 		final_delay = final_fall_delay.final_phase;
 	}
+	msdc_set_data_delay(host, final_delay);
 
 	dev_dbg(host->dev, "Final data pad delay: %x\n", final_delay);
 	return final_delay == 0xff ? -EIO : 0;
 }
 
+/*
+ * MSDC IP which supports data tune + async fifo can do CMD/DAT tune
+ * together, which can save the tuning time.
+ */
+static int msdc_tune_together(struct mmc_host *mmc, u32 opcode)
+{
+	struct msdc_host *host = mmc_priv(mmc);
+	u32 rise_delay = 0, fall_delay = 0;
+	struct msdc_delay_phase final_rise_delay, final_fall_delay = { 0,};
+	u8 final_delay, final_maxlen;
+	int i, ret;
+
+	sdr_set_field(host->base + MSDC_PATCH_BIT, MSDC_INT_DAT_LATCH_CK_SEL,
+		      host->latch_ck);
+
+	sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+	sdr_clr_bits(host->base + MSDC_IOCON,
+		     MSDC_IOCON_DSPL | MSDC_IOCON_W_DSPL);
+	for (i = 0 ; i < PAD_DELAY_MAX; i++) {
+		msdc_set_cmd_delay(host, i);
+		msdc_set_data_delay(host, i);
+		ret = mmc_send_tuning(mmc, opcode, NULL);
+		if (!ret)
+			rise_delay |= (1 << i);
+	}
+	final_rise_delay = get_best_delay(host, rise_delay);
+	/* if rising edge has enough margin, then do not scan falling edge */
+	if (final_rise_delay.maxlen >= 12 ||
+	    (final_rise_delay.start == 0 && final_rise_delay.maxlen >= 4))
+		goto skip_fall;
+
+	sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+	sdr_set_bits(host->base + MSDC_IOCON,
+		     MSDC_IOCON_DSPL | MSDC_IOCON_W_DSPL);
+	for (i = 0; i < PAD_DELAY_MAX; i++) {
+		msdc_set_cmd_delay(host, i);
+		msdc_set_data_delay(host, i);
+		ret = mmc_send_tuning(mmc, opcode, NULL);
+		if (!ret)
+			fall_delay |= (1 << i);
+	}
+	final_fall_delay = get_best_delay(host, fall_delay);
+
+skip_fall:
+	final_maxlen = max(final_rise_delay.maxlen, final_fall_delay.maxlen);
+	if (final_maxlen == final_rise_delay.maxlen) {
+		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+		sdr_clr_bits(host->base + MSDC_IOCON,
+			     MSDC_IOCON_DSPL | MSDC_IOCON_W_DSPL);
+		final_delay = final_rise_delay.final_phase;
+	} else {
+		sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+		sdr_set_bits(host->base + MSDC_IOCON,
+			     MSDC_IOCON_DSPL | MSDC_IOCON_W_DSPL);
+		final_delay = final_fall_delay.final_phase;
+	}
+
+	msdc_set_cmd_delay(host, final_delay);
+	msdc_set_data_delay(host, final_delay);
+
+	dev_dbg(host->dev, "Final pad delay: %x\n", final_delay);
+	return final_delay == 0xff ? -EIO : 0;
+}
+
 static int msdc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 {
 	struct msdc_host *host = mmc_priv(mmc);
 	int ret;
 	u32 tune_reg = host->dev_comp->pad_tune_reg;
 
+	if (host->dev_comp->data_tune && host->dev_comp->async_fifo) {
+		ret = msdc_tune_together(mmc, opcode);
+		if (host->hs400_mode) {
+			sdr_clr_bits(host->base + MSDC_IOCON,
+				     MSDC_IOCON_DSPL | MSDC_IOCON_W_DSPL);
+			msdc_set_data_delay(host, 0);
+		}
+		goto tune_done;
+	}
 	if (host->hs400_mode &&
 	    host->dev_comp->hs400_tune)
 		ret = hs400_tune_response(mmc, opcode);
@@ -1790,9 +1973,16 @@ static int msdc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 			dev_err(host->dev, "Tune data fail!\n");
 	}
 
+tune_done:
 	host->saved_tune_para.iocon = readl(host->base + MSDC_IOCON);
 	host->saved_tune_para.pad_tune = readl(host->base + tune_reg);
 	host->saved_tune_para.pad_cmd_tune = readl(host->base + PAD_CMD_TUNE);
+	if (host->top_base) {
+		host->saved_tune_para.emmc_top_control = readl(host->top_base +
+				EMMC_TOP_CONTROL);
+		host->saved_tune_para.emmc_top_cmd = readl(host->top_base +
+				EMMC_TOP_CMD);
+	}
 	return ret;
 }
 
@@ -1801,7 +1991,11 @@ static int msdc_prepare_hs400_tuning(struct mmc_host *mmc, struct mmc_ios *ios)
 	struct msdc_host *host = mmc_priv(mmc);
 	host->hs400_mode = true;
 
-	writel(host->hs400_ds_delay, host->base + PAD_DS_TUNE);
+	if (host->top_base)
+		writel(host->hs400_ds_delay,
+		       host->top_base + EMMC50_PAD_DS_TUNE);
+	else
+		writel(host->hs400_ds_delay, host->base + PAD_DS_TUNE);
 	/* hs400 mode must set it to 0 */
 	sdr_clr_bits(host->base + MSDC_PATCH_BIT2, MSDC_PATCH_BIT2_CFGCRCSTS);
 	/* to improve read performance, set outstanding to 2 */
@@ -1884,6 +2078,11 @@ static int msdc_drv_probe(struct platform_device *pdev)
 		goto host_free;
 	}
 
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	host->top_base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(host->top_base))
+		host->top_base = NULL;
+
 	ret = mmc_regulator_get_supply(mmc);
 	if (ret)
 		goto host_free;
@@ -1900,6 +2099,9 @@ static int msdc_drv_probe(struct platform_device *pdev)
 		goto host_free;
 	}
 
+	host->bus_clk = devm_clk_get(&pdev->dev, "bus_clk");
+	if (IS_ERR(host->bus_clk))
+		host->bus_clk = NULL;
 	/*source clock control gate is optional clock*/
 	host->src_clk_cg = devm_clk_get(&pdev->dev, "source_cg");
 	if (IS_ERR(host->src_clk_cg))
@@ -2049,7 +2251,6 @@ static void msdc_save_reg(struct msdc_host *host)
 	host->save_para.msdc_cfg = readl(host->base + MSDC_CFG);
 	host->save_para.iocon = readl(host->base + MSDC_IOCON);
 	host->save_para.sdc_cfg = readl(host->base + SDC_CFG);
-	host->save_para.pad_tune = readl(host->base + tune_reg);
 	host->save_para.patch_bit0 = readl(host->base + MSDC_PATCH_BIT);
 	host->save_para.patch_bit1 = readl(host->base + MSDC_PATCH_BIT1);
 	host->save_para.patch_bit2 = readl(host->base + MSDC_PATCH_BIT2);
@@ -2058,6 +2259,16 @@ static void msdc_save_reg(struct msdc_host *host)
 	host->save_para.emmc50_cfg0 = readl(host->base + EMMC50_CFG0);
 	host->save_para.emmc50_cfg3 = readl(host->base + EMMC50_CFG3);
 	host->save_para.sdc_fifo_cfg = readl(host->base + SDC_FIFO_CFG);
+	if (host->top_base) {
+		host->save_para.emmc_top_control =
+			readl(host->top_base + EMMC_TOP_CONTROL);
+		host->save_para.emmc_top_cmd =
+			readl(host->top_base + EMMC_TOP_CMD);
+		host->save_para.emmc50_pad_ds_tune =
+			readl(host->top_base + EMMC50_PAD_DS_TUNE);
+	} else {
+		host->save_para.pad_tune = readl(host->base + tune_reg);
+	}
 }
 
 static void msdc_restore_reg(struct msdc_host *host)
@@ -2067,7 +2278,6 @@ static void msdc_restore_reg(struct msdc_host *host)
 	writel(host->save_para.msdc_cfg, host->base + MSDC_CFG);
 	writel(host->save_para.iocon, host->base + MSDC_IOCON);
 	writel(host->save_para.sdc_cfg, host->base + SDC_CFG);
-	writel(host->save_para.pad_tune, host->base + tune_reg);
 	writel(host->save_para.patch_bit0, host->base + MSDC_PATCH_BIT);
 	writel(host->save_para.patch_bit1, host->base + MSDC_PATCH_BIT1);
 	writel(host->save_para.patch_bit2, host->base + MSDC_PATCH_BIT2);
@@ -2076,6 +2286,16 @@ static void msdc_restore_reg(struct msdc_host *host)
 	writel(host->save_para.emmc50_cfg0, host->base + EMMC50_CFG0);
 	writel(host->save_para.emmc50_cfg3, host->base + EMMC50_CFG3);
 	writel(host->save_para.sdc_fifo_cfg, host->base + SDC_FIFO_CFG);
+	if (host->top_base) {
+		writel(host->save_para.emmc_top_control,
+		       host->top_base + EMMC_TOP_CONTROL);
+		writel(host->save_para.emmc_top_cmd,
+		       host->top_base + EMMC_TOP_CMD);
+		writel(host->save_para.emmc50_pad_ds_tune,
+		       host->top_base + EMMC50_PAD_DS_TUNE);
+	} else {
+		writel(host->save_para.pad_tune, host->base + tune_reg);
+	}
 }
 
 static int msdc_runtime_suspend(struct device *dev)
diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c
index de4e6e5bf304..4d17032d15ee 100644
--- a/drivers/mmc/host/mxcmmc.c
+++ b/drivers/mmc/host/mxcmmc.c
@@ -728,7 +728,6 @@ static void mxcmci_cmd_done(struct mxcmci_host *host, unsigned int stat)
 static irqreturn_t mxcmci_irq(int irq, void *devid)
 {
 	struct mxcmci_host *host = devid;
-	unsigned long flags;
 	bool sdio_irq;
 	u32 stat;
 
@@ -740,9 +739,9 @@ static irqreturn_t mxcmci_irq(int irq, void *devid)
 
 	dev_dbg(mmc_dev(host->mmc), "%s: 0x%08x\n", __func__, stat);
 
-	spin_lock_irqsave(&host->lock, flags);
+	spin_lock(&host->lock);
 	sdio_irq = (stat & STATUS_SDIO_INT_ACTIVE) && host->use_sdio;
-	spin_unlock_irqrestore(&host->lock, flags);
+	spin_unlock(&host->lock);
 
 	if (mxcmci_use_dma(host) && (stat & (STATUS_WRITE_OP_DONE)))
 		mxcmci_writel(host, STATUS_WRITE_OP_DONE, MMC_REG_STATUS);
diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c
index 68760d4a5d3d..467d889a1638 100644
--- a/drivers/mmc/host/omap_hsmmc.c
+++ b/drivers/mmc/host/omap_hsmmc.c
@@ -30,7 +30,6 @@
 #include <linux/clk.h>
 #include <linux/of.h>
 #include <linux/of_irq.h>
-#include <linux/of_gpio.h>
 #include <linux/of_device.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/core.h>
@@ -38,7 +37,6 @@
 #include <linux/mmc/slot-gpio.h>
 #include <linux/io.h>
 #include <linux/irq.h>
-#include <linux/gpio.h>
 #include <linux/regulator/consumer.h>
 #include <linux/pinctrl/consumer.h>
 #include <linux/pm_runtime.h>
@@ -198,7 +196,6 @@ struct omap_hsmmc_host {
 	struct dma_chan		*rx_chan;
 	int			response_busy;
 	int			context_loss;
-	int			protect_card;
 	int			reqs_blocked;
 	int			req_in_progress;
 	unsigned long		clk_rate;
@@ -207,16 +204,6 @@ struct omap_hsmmc_host {
 #define HSMMC_SDIO_IRQ_ENABLED	(1 << 1)        /* SDIO irq enabled */
 	struct omap_hsmmc_next	next_data;
 	struct	omap_hsmmc_platform_data	*pdata;
-
-	/* return MMC cover switch state, can be NULL if not supported.
-	 *
-	 * possible return values:
-	 *   0 - closed
-	 *   1 - open
-	 */
-	int (*get_cover_state)(struct device *dev);
-
-	int (*card_detect)(struct device *dev);
 };
 
 struct omap_mmc_of_data {
@@ -226,20 +213,6 @@ struct omap_mmc_of_data {
 
 static void omap_hsmmc_start_dma_transfer(struct omap_hsmmc_host *host);
 
-static int omap_hsmmc_card_detect(struct device *dev)
-{
-	struct omap_hsmmc_host *host = dev_get_drvdata(dev);
-
-	return mmc_gpio_get_cd(host->mmc);
-}
-
-static int omap_hsmmc_get_cover_state(struct device *dev)
-{
-	struct omap_hsmmc_host *host = dev_get_drvdata(dev);
-
-	return mmc_gpio_get_cd(host->mmc);
-}
-
 static int omap_hsmmc_enable_supply(struct mmc_host *mmc)
 {
 	int ret;
@@ -484,38 +457,6 @@ static int omap_hsmmc_reg_get(struct omap_hsmmc_host *host)
 	return 0;
 }
 
-static irqreturn_t omap_hsmmc_cover_irq(int irq, void *dev_id);
-
-static int omap_hsmmc_gpio_init(struct mmc_host *mmc,
-				struct omap_hsmmc_host *host,
-				struct omap_hsmmc_platform_data *pdata)
-{
-	int ret;
-
-	if (gpio_is_valid(pdata->gpio_cod)) {
-		ret = mmc_gpio_request_cd(mmc, pdata->gpio_cod, 0);
-		if (ret)
-			return ret;
-
-		host->get_cover_state = omap_hsmmc_get_cover_state;
-		mmc_gpio_set_cd_isr(mmc, omap_hsmmc_cover_irq);
-	} else if (gpio_is_valid(pdata->gpio_cd)) {
-		ret = mmc_gpio_request_cd(mmc, pdata->gpio_cd, 0);
-		if (ret)
-			return ret;
-
-		host->card_detect = omap_hsmmc_card_detect;
-	}
-
-	if (gpio_is_valid(pdata->gpio_wp)) {
-		ret = mmc_gpio_request_ro(mmc, pdata->gpio_wp);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 /*
  * Start clock to the card
  */
@@ -781,9 +722,6 @@ static void send_init_stream(struct omap_hsmmc_host *host)
 	int reg = 0;
 	unsigned long timeout;
 
-	if (host->protect_card)
-		return;
-
 	disable_irq(host->irq);
 
 	OMAP_HSMMC_WRITE(host->base, IE, INT_EN_MASK);
@@ -804,29 +742,6 @@ static void send_init_stream(struct omap_hsmmc_host *host)
 	enable_irq(host->irq);
 }
 
-static inline
-int omap_hsmmc_cover_is_closed(struct omap_hsmmc_host *host)
-{
-	int r = 1;
-
-	if (host->get_cover_state)
-		r = host->get_cover_state(host->dev);
-	return r;
-}
-
-static ssize_t
-omap_hsmmc_show_cover_switch(struct device *dev, struct device_attribute *attr,
-			   char *buf)
-{
-	struct mmc_host *mmc = container_of(dev, struct mmc_host, class_dev);
-	struct omap_hsmmc_host *host = mmc_priv(mmc);
-
-	return sprintf(buf, "%s\n",
-			omap_hsmmc_cover_is_closed(host) ? "closed" : "open");
-}
-
-static DEVICE_ATTR(cover_switch, S_IRUGO, omap_hsmmc_show_cover_switch, NULL);
-
 static ssize_t
 omap_hsmmc_show_slot_name(struct device *dev, struct device_attribute *attr,
 			char *buf)
@@ -1247,44 +1162,6 @@ err:
 	return ret;
 }
 
-/* Protect the card while the cover is open */
-static void omap_hsmmc_protect_card(struct omap_hsmmc_host *host)
-{
-	if (!host->get_cover_state)
-		return;
-
-	host->reqs_blocked = 0;
-	if (host->get_cover_state(host->dev)) {
-		if (host->protect_card) {
-			dev_info(host->dev, "%s: cover is closed, "
-					 "card is now accessible\n",
-					 mmc_hostname(host->mmc));
-			host->protect_card = 0;
-		}
-	} else {
-		if (!host->protect_card) {
-			dev_info(host->dev, "%s: cover is open, "
-					 "card is now inaccessible\n",
-					 mmc_hostname(host->mmc));
-			host->protect_card = 1;
-		}
-	}
-}
-
-/*
- * irq handler when (cell-phone) cover is mounted/removed
- */
-static irqreturn_t omap_hsmmc_cover_irq(int irq, void *dev_id)
-{
-	struct omap_hsmmc_host *host = dev_id;
-
-	sysfs_notify(&host->mmc->class_dev.kobj, NULL, "cover_switch");
-
-	omap_hsmmc_protect_card(host);
-	mmc_detect_change(host->mmc, (HZ * 200) / 1000);
-	return IRQ_HANDLED;
-}
-
 static void omap_hsmmc_dma_callback(void *param)
 {
 	struct omap_hsmmc_host *host = param;
@@ -1555,24 +1432,7 @@ static void omap_hsmmc_request(struct mmc_host *mmc, struct mmc_request *req)
 
 	BUG_ON(host->req_in_progress);
 	BUG_ON(host->dma_ch != -1);
-	if (host->protect_card) {
-		if (host->reqs_blocked < 3) {
-			/*
-			 * Ensure the controller is left in a consistent
-			 * state by resetting the command and data state
-			 * machines.
-			 */
-			omap_hsmmc_reset_controller_fsm(host, SRD);
-			omap_hsmmc_reset_controller_fsm(host, SRC);
-			host->reqs_blocked += 1;
-		}
-		req->cmd->error = -EBADF;
-		if (req->data)
-			req->data->error = -EBADF;
-		req->cmd->retries = 0;
-		mmc_request_done(mmc, req);
-		return;
-	} else if (host->reqs_blocked)
+	if (host->reqs_blocked)
 		host->reqs_blocked = 0;
 	WARN_ON(host->mrq != NULL);
 	host->mrq = req;
@@ -1646,15 +1506,6 @@ static void omap_hsmmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	omap_hsmmc_set_bus_mode(host);
 }
 
-static int omap_hsmmc_get_cd(struct mmc_host *mmc)
-{
-	struct omap_hsmmc_host *host = mmc_priv(mmc);
-
-	if (!host->card_detect)
-		return -ENOSYS;
-	return host->card_detect(host->dev);
-}
-
 static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card)
 {
 	struct omap_hsmmc_host *host = mmc_priv(mmc);
@@ -1793,7 +1644,7 @@ static struct mmc_host_ops omap_hsmmc_ops = {
 	.pre_req = omap_hsmmc_pre_req,
 	.request = omap_hsmmc_request,
 	.set_ios = omap_hsmmc_set_ios,
-	.get_cd = omap_hsmmc_get_cd,
+	.get_cd = mmc_gpio_get_cd,
 	.get_ro = mmc_gpio_get_ro,
 	.init_card = omap_hsmmc_init_card,
 	.enable_sdio_irq = omap_hsmmc_enable_sdio_irq,
@@ -1920,10 +1771,6 @@ static struct omap_hsmmc_platform_data *of_get_hsmmc_pdata(struct device *dev)
 	if (of_find_property(np, "ti,dual-volt", NULL))
 		pdata->controller_flags |= OMAP_HSMMC_SUPPORTS_DUAL_VOLT;
 
-	pdata->gpio_cd = -EINVAL;
-	pdata->gpio_cod = -EINVAL;
-	pdata->gpio_wp = -EINVAL;
-
 	if (of_find_property(np, "ti,non-removable", NULL)) {
 		pdata->nonremovable = true;
 		pdata->no_regulator_off_init = true;
@@ -2008,10 +1855,6 @@ static int omap_hsmmc_probe(struct platform_device *pdev)
 	host->pbias_enabled = 0;
 	host->vqmmc_enabled = 0;
 
-	ret = omap_hsmmc_gpio_init(mmc, host, pdata);
-	if (ret)
-		goto err_gpio;
-
 	platform_set_drvdata(pdev, host);
 
 	if (pdev->dev.of_node)
@@ -2125,8 +1968,6 @@ static int omap_hsmmc_probe(struct platform_device *pdev)
 	if (!ret)
 		mmc->caps |= MMC_CAP_SDIO_IRQ;
 
-	omap_hsmmc_protect_card(host);
-
 	mmc_add_host(mmc);
 
 	if (mmc_pdata(host)->name != NULL) {
@@ -2134,12 +1975,6 @@ static int omap_hsmmc_probe(struct platform_device *pdev)
 		if (ret < 0)
 			goto err_slot_name;
 	}
-	if (host->get_cover_state) {
-		ret = device_create_file(&mmc->class_dev,
-					 &dev_attr_cover_switch);
-		if (ret < 0)
-			goto err_slot_name;
-	}
 
 	omap_hsmmc_debugfs(mmc);
 	pm_runtime_mark_last_busy(host->dev);
@@ -2161,7 +1996,6 @@ err_irq:
 	if (host->dbclk)
 		clk_disable_unprepare(host->dbclk);
 err1:
-err_gpio:
 	mmc_free_host(mmc);
 err:
 	return ret;
@@ -2231,7 +2065,6 @@ static int omap_hsmmc_resume(struct device *dev)
 	if (!(host->mmc->pm_flags & MMC_PM_KEEP_POWER))
 		omap_hsmmc_conf_bus_power(host);
 
-	omap_hsmmc_protect_card(host);
 	pm_runtime_mark_last_busy(host->dev);
 	pm_runtime_put_autosuspend(host->dev);
 	return 0;
diff --git a/drivers/mmc/host/renesas_sdhi.h b/drivers/mmc/host/renesas_sdhi.h
index f13f798d8506..da1e49c45bec 100644
--- a/drivers/mmc/host/renesas_sdhi.h
+++ b/drivers/mmc/host/renesas_sdhi.h
@@ -1,12 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Renesas Mobile SDHI
  *
  * Copyright (C) 2017 Horms Solutions Ltd., Simon Horman
  * Copyright (C) 2017 Renesas Electronics Corporation
- *
- * 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 RENESAS_SDHI_H
diff --git a/drivers/mmc/host/renesas_sdhi_core.c b/drivers/mmc/host/renesas_sdhi_core.c
index 777e32b0e410..d3ac43c3d0b6 100644
--- a/drivers/mmc/host/renesas_sdhi_core.c
+++ b/drivers/mmc/host/renesas_sdhi_core.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Renesas SDHI
  *
@@ -6,10 +7,6 @@
  * Copyright (C) 2016-17 Horms Solutions, Simon Horman
  * Copyright (C) 2009 Magnus Damm
  *
- * 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 "Compaq ASIC3 support":
  *
  * Copyright 2001 Compaq Computer Corporation.
@@ -155,6 +152,52 @@ static unsigned int renesas_sdhi_clk_update(struct tmio_mmc_host *host,
 	return ret == 0 ? best_freq : clk_get_rate(priv->clk);
 }
 
+static void renesas_sdhi_set_clock(struct tmio_mmc_host *host,
+				   unsigned int new_clock)
+{
+	u32 clk = 0, clock;
+
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, ~CLK_CTL_SCLKEN &
+		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
+
+	if (new_clock == 0)
+		goto out;
+
+	/*
+	 * Both HS400 and HS200/SD104 set 200MHz, but some devices need to
+	 * set 400MHz to distinguish the CPG settings in HS400.
+	 */
+	if (host->mmc->ios.timing == MMC_TIMING_MMC_HS400 &&
+	    host->pdata->flags & TMIO_MMC_HAVE_4TAP_HS400 &&
+	    new_clock == 200000000)
+		new_clock = 400000000;
+
+	clock = renesas_sdhi_clk_update(host, new_clock) / 512;
+
+	for (clk = 0x80000080; new_clock >= (clock << 1); clk >>= 1)
+		clock <<= 1;
+
+	/* 1/1 clock is option */
+	if ((host->pdata->flags & TMIO_MMC_CLK_ACTUAL) && ((clk >> 22) & 0x1)) {
+		if (!(host->mmc->ios.timing == MMC_TIMING_MMC_HS400))
+			clk |= 0xff;
+		else
+			clk &= ~0xff;
+	}
+
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, clk & CLK_CTL_DIV_MASK);
+	if (!(host->pdata->flags & TMIO_MMC_MIN_RCAR2))
+		usleep_range(10000, 11000);
+
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, CLK_CTL_SCLKEN |
+		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
+
+out:
+	/* HW engineers overrode docs: no sleep needed on R-Car2+ */
+	if (!(host->pdata->flags & TMIO_MMC_MIN_RCAR2))
+		usleep_range(10000, 11000);
+}
+
 static void renesas_sdhi_clk_disable(struct tmio_mmc_host *host)
 {
 	struct renesas_sdhi *priv = host_to_priv(host);
@@ -443,6 +486,19 @@ static int renesas_sdhi_select_tuning(struct tmio_mmc_host *host)
 static bool renesas_sdhi_check_scc_error(struct tmio_mmc_host *host)
 {
 	struct renesas_sdhi *priv = host_to_priv(host);
+	bool use_4tap = host->pdata->flags & TMIO_MMC_HAVE_4TAP_HS400;
+
+	/*
+	 * Skip checking SCC errors when running on 4 taps in HS400 mode as
+	 * any retuning would still result in the same 4 taps being used.
+	 */
+	if (!(host->mmc->ios.timing == MMC_TIMING_UHS_SDR104) &&
+	    !(host->mmc->ios.timing == MMC_TIMING_MMC_HS200) &&
+	    !(host->mmc->ios.timing == MMC_TIMING_MMC_HS400 && !use_4tap))
+		return false;
+
+	if (mmc_doing_retune(host->mmc))
+		return false;
 
 	/* Check SCC error */
 	if (sd_scc_read32(host, priv, SH_MOBILE_SDHI_SCC_RVSCNTL) &
@@ -620,8 +676,8 @@ int renesas_sdhi_probe(struct platform_device *pdev,
 
 	host->write16_hook	= renesas_sdhi_write16_hook;
 	host->clk_enable	= renesas_sdhi_clk_enable;
-	host->clk_update	= renesas_sdhi_clk_update;
 	host->clk_disable	= renesas_sdhi_clk_disable;
+	host->set_clock		= renesas_sdhi_set_clock;
 	host->multi_io_quirk	= renesas_sdhi_multi_io_quirk;
 	host->dma_ops		= dma_ops;
 
diff --git a/drivers/mmc/host/renesas_sdhi_internal_dmac.c b/drivers/mmc/host/renesas_sdhi_internal_dmac.c
index ca0b43973769..b6f54102bfdd 100644
--- a/drivers/mmc/host/renesas_sdhi_internal_dmac.c
+++ b/drivers/mmc/host/renesas_sdhi_internal_dmac.c
@@ -1,12 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * DMA support for Internal DMAC with SDHI SD/SDIO controller
  *
  * Copyright (C) 2016-17 Renesas Electronics Corporation
  * Copyright (C) 2016-17 Horms Solutions, Simon Horman
- *
- * 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/bitops.h>
@@ -35,8 +32,8 @@
 
 /* DM_CM_DTRAN_MODE */
 #define DTRAN_MODE_CH_NUM_CH0	0	/* "downstream" = for write commands */
-#define DTRAN_MODE_CH_NUM_CH1	BIT(16)	/* "uptream" = for read commands */
-#define DTRAN_MODE_BUS_WID_TH	(BIT(5) | BIT(4))
+#define DTRAN_MODE_CH_NUM_CH1	BIT(16)	/* "upstream" = for read commands */
+#define DTRAN_MODE_BUS_WIDTH	(BIT(5) | BIT(4))
 #define DTRAN_MODE_ADDR_MODE	BIT(0)	/* 1 = Increment address */
 
 /* DM_CM_DTRAN_CTRL */
@@ -116,6 +113,7 @@ static const struct renesas_sdhi_of_data of_rcar_gen3_compatible = {
 };
 
 static const struct of_device_id renesas_sdhi_internal_dmac_of_match[] = {
+	{ .compatible = "renesas,sdhi-mmc-r8a77470", .data = &of_rcar_gen3_compatible, },
 	{ .compatible = "renesas,sdhi-r8a7795", .data = &of_rcar_r8a7795_compatible, },
 	{ .compatible = "renesas,sdhi-r8a7796", .data = &of_rcar_r8a7795_compatible, },
 	{ .compatible = "renesas,rcar-gen3-sdhi", .data = &of_rcar_gen3_compatible, },
@@ -174,7 +172,7 @@ renesas_sdhi_internal_dmac_start_dma(struct tmio_mmc_host *host,
 				     struct mmc_data *data)
 {
 	struct scatterlist *sg = host->sg_ptr;
-	u32 dtran_mode = DTRAN_MODE_BUS_WID_TH | DTRAN_MODE_ADDR_MODE;
+	u32 dtran_mode = DTRAN_MODE_BUS_WIDTH | DTRAN_MODE_ADDR_MODE;
 
 	if (!dma_map_sg(&host->pdev->dev, sg, host->sg_len,
 			mmc_get_dma_dir(data)))
@@ -201,13 +199,14 @@ renesas_sdhi_internal_dmac_start_dma(struct tmio_mmc_host *host,
 	renesas_sdhi_internal_dmac_dm_write(host, DM_DTRAN_ADDR,
 					    sg_dma_address(sg));
 
+	host->dma_on = true;
+
 	return;
 
 force_pio_with_unmap:
 	dma_unmap_sg(&host->pdev->dev, sg, host->sg_len, mmc_get_dma_dir(data));
 
 force_pio:
-	host->force_pio = true;
 	renesas_sdhi_internal_dmac_enable_dma(host, false);
 }
 
@@ -291,16 +290,19 @@ static const struct tmio_mmc_dma_ops renesas_sdhi_internal_dmac_dma_ops = {
  * Whitelist of specific R-Car Gen3 SoC ES versions to use this DMAC
  * implementation as others may use a different implementation.
  */
-static const struct soc_device_attribute gen3_soc_whitelist[] = {
+static const struct soc_device_attribute soc_whitelist[] = {
 	/* specific ones */
 	{ .soc_id = "r8a7795", .revision = "ES1.*",
 	  .data = (void *)BIT(SDHI_INTERNAL_DMAC_ONE_RX_ONLY) },
 	{ .soc_id = "r8a7796", .revision = "ES1.0",
 	  .data = (void *)BIT(SDHI_INTERNAL_DMAC_ONE_RX_ONLY) },
 	/* generic ones */
+	{ .soc_id = "r8a774a1" },
+	{ .soc_id = "r8a77470" },
 	{ .soc_id = "r8a7795" },
 	{ .soc_id = "r8a7796" },
 	{ .soc_id = "r8a77965" },
+	{ .soc_id = "r8a77970" },
 	{ .soc_id = "r8a77980" },
 	{ .soc_id = "r8a77995" },
 	{ /* sentinel */ }
@@ -308,13 +310,21 @@ static const struct soc_device_attribute gen3_soc_whitelist[] = {
 
 static int renesas_sdhi_internal_dmac_probe(struct platform_device *pdev)
 {
-	const struct soc_device_attribute *soc = soc_device_match(gen3_soc_whitelist);
+	const struct soc_device_attribute *soc = soc_device_match(soc_whitelist);
+	struct device *dev = &pdev->dev;
 
 	if (!soc)
 		return -ENODEV;
 
 	global_flags |= (unsigned long)soc->data;
 
+	dev->dma_parms = devm_kzalloc(dev, sizeof(*dev->dma_parms), GFP_KERNEL);
+	if (!dev->dma_parms)
+		return -ENOMEM;
+
+	/* value is max of SD_SECCNT. Confirmed by HW engineers */
+	dma_set_max_seg_size(dev, 0xffffffff);
+
 	return renesas_sdhi_probe(pdev, &renesas_sdhi_internal_dmac_dma_ops);
 }
 
diff --git a/drivers/mmc/host/renesas_sdhi_sys_dmac.c b/drivers/mmc/host/renesas_sdhi_sys_dmac.c
index 5389c4821882..1a4016f635d3 100644
--- a/drivers/mmc/host/renesas_sdhi_sys_dmac.c
+++ b/drivers/mmc/host/renesas_sdhi_sys_dmac.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * DMA support use of SYS DMAC with SDHI SD/SDIO controller
  *
@@ -5,10 +6,6 @@
  * Copyright (C) 2016-17 Sang Engineering, Wolfram Sang
  * Copyright (C) 2017 Horms Solutions, Simon Horman
  * Copyright (C) 2010-2011 Guennadi Liakhovetski
- *
- * 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/device.h>
@@ -213,10 +210,8 @@ static void renesas_sdhi_sys_dmac_start_dma_rx(struct tmio_mmc_host *host)
 		goto pio;
 	}
 
-	if (sg->length < TMIO_MMC_MIN_DMA_LEN) {
-		host->force_pio = true;
+	if (sg->length < TMIO_MMC_MIN_DMA_LEN)
 		return;
-	}
 
 	/* The only sg element can be unaligned, use our bounce buffer then */
 	if (!aligned) {
@@ -240,6 +235,7 @@ static void renesas_sdhi_sys_dmac_start_dma_rx(struct tmio_mmc_host *host)
 			desc = NULL;
 			ret = cookie;
 		}
+		host->dma_on = true;
 	}
 pio:
 	if (!desc) {
@@ -286,10 +282,8 @@ static void renesas_sdhi_sys_dmac_start_dma_tx(struct tmio_mmc_host *host)
 		goto pio;
 	}
 
-	if (sg->length < TMIO_MMC_MIN_DMA_LEN) {
-		host->force_pio = true;
+	if (sg->length < TMIO_MMC_MIN_DMA_LEN)
 		return;
-	}
 
 	/* The only sg element can be unaligned, use our bounce buffer then */
 	if (!aligned) {
@@ -318,6 +312,7 @@ static void renesas_sdhi_sys_dmac_start_dma_tx(struct tmio_mmc_host *host)
 			desc = NULL;
 			ret = cookie;
 		}
+		host->dma_on = true;
 	}
 pio:
 	if (!desc) {
diff --git a/drivers/mmc/host/sdhci-acpi.c b/drivers/mmc/host/sdhci-acpi.c
index 32321bd596d8..82c9b9326e9e 100644
--- a/drivers/mmc/host/sdhci-acpi.c
+++ b/drivers/mmc/host/sdhci-acpi.c
@@ -76,6 +76,7 @@ struct sdhci_acpi_slot {
 	size_t		priv_size;
 	int (*probe_slot)(struct platform_device *, const char *, const char *);
 	int (*remove_slot)(struct platform_device *);
+	int (*free_slot)(struct platform_device *pdev);
 	int (*setup_host)(struct platform_device *pdev);
 };
 
@@ -470,10 +471,70 @@ static const struct sdhci_acpi_slot sdhci_acpi_slot_int_sd = {
 	.priv_size	= sizeof(struct intel_host),
 };
 
+#define VENDOR_SPECIFIC_PWRCTL_CLEAR_REG	0x1a8
+#define VENDOR_SPECIFIC_PWRCTL_CTL_REG		0x1ac
+static irqreturn_t sdhci_acpi_qcom_handler(int irq, void *ptr)
+{
+	struct sdhci_host *host = ptr;
+
+	sdhci_writel(host, 0x3, VENDOR_SPECIFIC_PWRCTL_CLEAR_REG);
+	sdhci_writel(host, 0x1, VENDOR_SPECIFIC_PWRCTL_CTL_REG);
+
+	return IRQ_HANDLED;
+}
+
+static int qcom_probe_slot(struct platform_device *pdev, const char *hid,
+			   const char *uid)
+{
+	struct sdhci_acpi_host *c = platform_get_drvdata(pdev);
+	struct sdhci_host *host = c->host;
+	int *irq = sdhci_acpi_priv(c);
+
+	*irq = -EINVAL;
+
+	if (strcmp(hid, "QCOM8051"))
+		return 0;
+
+	*irq = platform_get_irq(pdev, 1);
+	if (*irq < 0)
+		return 0;
+
+	return request_threaded_irq(*irq, NULL, sdhci_acpi_qcom_handler,
+				    IRQF_ONESHOT | IRQF_TRIGGER_HIGH,
+				    "sdhci_qcom", host);
+}
+
+static int qcom_free_slot(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct sdhci_acpi_host *c = platform_get_drvdata(pdev);
+	struct sdhci_host *host = c->host;
+	struct acpi_device *adev;
+	int *irq = sdhci_acpi_priv(c);
+	const char *hid;
+
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
+		return -ENODEV;
+
+	hid = acpi_device_hid(adev);
+	if (strcmp(hid, "QCOM8051"))
+		return 0;
+
+	if (*irq < 0)
+		return 0;
+
+	free_irq(*irq, host);
+	return 0;
+}
+
 static const struct sdhci_acpi_slot sdhci_acpi_slot_qcom_sd_3v = {
 	.quirks  = SDHCI_QUIRK_BROKEN_CARD_DETECTION,
 	.quirks2 = SDHCI_QUIRK2_NO_1_8_V,
 	.caps    = MMC_CAP_NONREMOVABLE,
+	.priv_size	= sizeof(int),
+	.probe_slot	= qcom_probe_slot,
+	.free_slot	= qcom_free_slot,
 };
 
 static const struct sdhci_acpi_slot sdhci_acpi_slot_qcom_sd = {
@@ -756,6 +817,9 @@ static int sdhci_acpi_probe(struct platform_device *pdev)
 err_cleanup:
 	sdhci_cleanup_host(c->host);
 err_free:
+	if (c->slot && c->slot->free_slot)
+		c->slot->free_slot(pdev);
+
 	sdhci_free_host(c->host);
 	return err;
 }
@@ -777,6 +841,10 @@ static int sdhci_acpi_remove(struct platform_device *pdev)
 
 	dead = (sdhci_readl(c->host, SDHCI_INT_STATUS) == ~0);
 	sdhci_remove_host(c->host, dead);
+
+	if (c->slot && c->slot->free_slot)
+		c->slot->free_slot(pdev);
+
 	sdhci_free_host(c->host);
 
 	return 0;
diff --git a/drivers/mmc/host/sdhci-esdhc.h b/drivers/mmc/host/sdhci-esdhc.h
index dfa58f8b8dfa..3f16d9c90ba2 100644
--- a/drivers/mmc/host/sdhci-esdhc.h
+++ b/drivers/mmc/host/sdhci-esdhc.h
@@ -60,6 +60,7 @@
 /* Tuning Block Control Register */
 #define ESDHC_TBCTL			0x120
 #define ESDHC_TB_EN			0x00000004
+#define ESDHC_TBPTR			0x128
 
 /* Control Register for DMA transfer */
 #define ESDHC_DMA_SYSCTL		0x40c
diff --git a/drivers/mmc/host/sdhci-iproc.c b/drivers/mmc/host/sdhci-iproc.c
index d0e83db42ae5..0db99057c44f 100644
--- a/drivers/mmc/host/sdhci-iproc.c
+++ b/drivers/mmc/host/sdhci-iproc.c
@@ -15,6 +15,7 @@
  * iProc SDHCI platform driver
  */
 
+#include <linux/acpi.h>
 #include <linux/delay.h>
 #include <linux/module.h>
 #include <linux/mmc/host.h>
@@ -162,9 +163,19 @@ static void sdhci_iproc_writeb(struct sdhci_host *host, u8 val, int reg)
 	sdhci_iproc_writel(host, newval, reg & ~3);
 }
 
+static unsigned int sdhci_iproc_get_max_clock(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+
+	if (pltfm_host->clk)
+		return sdhci_pltfm_clk_get_max_clock(host);
+	else
+		return pltfm_host->clock;
+}
+
 static const struct sdhci_ops sdhci_iproc_ops = {
 	.set_clock = sdhci_set_clock,
-	.get_max_clock = sdhci_pltfm_clk_get_max_clock,
+	.get_max_clock = sdhci_iproc_get_max_clock,
 	.set_bus_width = sdhci_set_bus_width,
 	.reset = sdhci_reset,
 	.set_uhs_signaling = sdhci_set_uhs_signaling,
@@ -178,7 +189,7 @@ static const struct sdhci_ops sdhci_iproc_32only_ops = {
 	.write_w = sdhci_iproc_writew,
 	.write_b = sdhci_iproc_writeb,
 	.set_clock = sdhci_set_clock,
-	.get_max_clock = sdhci_pltfm_clk_get_max_clock,
+	.get_max_clock = sdhci_iproc_get_max_clock,
 	.set_bus_width = sdhci_set_bus_width,
 	.reset = sdhci_reset,
 	.set_uhs_signaling = sdhci_set_uhs_signaling,
@@ -256,19 +267,25 @@ static const struct of_device_id sdhci_iproc_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, sdhci_iproc_of_match);
 
+static const struct acpi_device_id sdhci_iproc_acpi_ids[] = {
+	{ .id = "BRCM5871", .driver_data = (kernel_ulong_t)&iproc_cygnus_data },
+	{ .id = "BRCM5872", .driver_data = (kernel_ulong_t)&iproc_data },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(acpi, sdhci_iproc_acpi_ids);
+
 static int sdhci_iproc_probe(struct platform_device *pdev)
 {
-	const struct of_device_id *match;
-	const struct sdhci_iproc_data *iproc_data;
+	struct device *dev = &pdev->dev;
+	const struct sdhci_iproc_data *iproc_data = NULL;
 	struct sdhci_host *host;
 	struct sdhci_iproc_host *iproc_host;
 	struct sdhci_pltfm_host *pltfm_host;
 	int ret;
 
-	match = of_match_device(sdhci_iproc_of_match, &pdev->dev);
-	if (!match)
-		return -EINVAL;
-	iproc_data = match->data;
+	iproc_data = device_get_match_data(dev);
+	if (!iproc_data)
+		return -ENODEV;
 
 	host = sdhci_pltfm_init(pdev, iproc_data->pdata, sizeof(*iproc_host));
 	if (IS_ERR(host))
@@ -280,19 +297,21 @@ static int sdhci_iproc_probe(struct platform_device *pdev)
 	iproc_host->data = iproc_data;
 
 	mmc_of_parse(host->mmc);
-	sdhci_get_of_property(pdev);
+	sdhci_get_property(pdev);
 
 	host->mmc->caps |= iproc_host->data->mmc_caps;
 
-	pltfm_host->clk = devm_clk_get(&pdev->dev, NULL);
-	if (IS_ERR(pltfm_host->clk)) {
-		ret = PTR_ERR(pltfm_host->clk);
-		goto err;
-	}
-	ret = clk_prepare_enable(pltfm_host->clk);
-	if (ret) {
-		dev_err(&pdev->dev, "failed to enable host clk\n");
-		goto err;
+	if (dev->of_node) {
+		pltfm_host->clk = devm_clk_get(dev, NULL);
+		if (IS_ERR(pltfm_host->clk)) {
+			ret = PTR_ERR(pltfm_host->clk);
+			goto err;
+		}
+		ret = clk_prepare_enable(pltfm_host->clk);
+		if (ret) {
+			dev_err(dev, "failed to enable host clk\n");
+			goto err;
+		}
 	}
 
 	if (iproc_host->data->pdata->quirks & SDHCI_QUIRK_MISSING_CAPS) {
@@ -307,7 +326,8 @@ static int sdhci_iproc_probe(struct platform_device *pdev)
 	return 0;
 
 err_clk:
-	clk_disable_unprepare(pltfm_host->clk);
+	if (dev->of_node)
+		clk_disable_unprepare(pltfm_host->clk);
 err:
 	sdhci_pltfm_free(pdev);
 	return ret;
@@ -317,6 +337,7 @@ static struct platform_driver sdhci_iproc_driver = {
 	.driver = {
 		.name = "sdhci-iproc",
 		.of_match_table = sdhci_iproc_of_match,
+		.acpi_match_table = ACPI_PTR(sdhci_iproc_acpi_ids),
 		.pm = &sdhci_pltfm_pmops,
 	},
 	.probe = sdhci_iproc_probe,
diff --git a/drivers/mmc/host/sdhci-of-arasan.c b/drivers/mmc/host/sdhci-of-arasan.c
index a40bcc27f187..142c4b802f31 100644
--- a/drivers/mmc/host/sdhci-of-arasan.c
+++ b/drivers/mmc/host/sdhci-of-arasan.c
@@ -107,6 +107,11 @@ struct sdhci_arasan_data {
 #define SDHCI_ARASAN_QUIRK_CLOCK_UNSTABLE BIT(1)
 };
 
+struct sdhci_arasan_of_data {
+	const struct sdhci_arasan_soc_ctl_map *soc_ctl_map;
+	const struct sdhci_pltfm_data *pdata;
+};
+
 static const struct sdhci_arasan_soc_ctl_map rk3399_soc_ctl_map = {
 	.baseclkfreq = { .reg = 0xf000, .width = 8, .shift = 8 },
 	.clockmultiplier = { .reg = 0xf02c, .width = 8, .shift = 0},
@@ -226,6 +231,25 @@ static void sdhci_arasan_set_clock(struct sdhci_host *host, unsigned int clock)
 	}
 }
 
+static void sdhci_arasan_am654_set_clock(struct sdhci_host *host,
+					 unsigned int clock)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_arasan_data *sdhci_arasan = sdhci_pltfm_priv(pltfm_host);
+
+	if (sdhci_arasan->is_phy_on) {
+		phy_power_off(sdhci_arasan->phy);
+		sdhci_arasan->is_phy_on = false;
+	}
+
+	sdhci_set_clock(host, clock);
+
+	if (clock > PHY_CLK_TOO_SLOW_HZ) {
+		phy_power_on(sdhci_arasan->phy);
+		sdhci_arasan->is_phy_on = true;
+	}
+}
+
 static void sdhci_arasan_hs400_enhanced_strobe(struct mmc_host *mmc,
 					struct mmc_ios *ios)
 {
@@ -307,6 +331,33 @@ static const struct sdhci_pltfm_data sdhci_arasan_pdata = {
 			SDHCI_QUIRK2_STOP_WITH_TC,
 };
 
+static struct sdhci_arasan_of_data sdhci_arasan_data = {
+	.pdata = &sdhci_arasan_pdata,
+};
+
+static const struct sdhci_ops sdhci_arasan_am654_ops = {
+	.set_clock = sdhci_arasan_am654_set_clock,
+	.get_max_clock = sdhci_pltfm_clk_get_max_clock,
+	.get_timeout_clock = sdhci_pltfm_clk_get_max_clock,
+	.set_bus_width = sdhci_set_bus_width,
+	.reset = sdhci_arasan_reset,
+	.set_uhs_signaling = sdhci_set_uhs_signaling,
+};
+
+static const struct sdhci_pltfm_data sdhci_arasan_am654_pdata = {
+	.ops = &sdhci_arasan_am654_ops,
+	.quirks = SDHCI_QUIRK_CAP_CLOCK_BASE_BROKEN  |
+		  SDHCI_QUIRK_INVERTED_WRITE_PROTECT |
+		  SDHCI_QUIRK_MULTIBLOCK_READ_ACMD12,
+	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN |
+		   SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN |
+		   SDHCI_QUIRK2_CAPS_BIT63_FOR_HS400,
+};
+
+static const struct sdhci_arasan_of_data sdhci_arasan_am654_data = {
+	.pdata = &sdhci_arasan_am654_pdata,
+};
+
 static u32 sdhci_arasan_cqhci_irq(struct sdhci_host *host, u32 intmask)
 {
 	int cmd_error = 0;
@@ -363,6 +414,11 @@ static const struct sdhci_pltfm_data sdhci_arasan_cqe_pdata = {
 			SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
 };
 
+static struct sdhci_arasan_of_data sdhci_arasan_rk3399_data = {
+	.soc_ctl_map = &rk3399_soc_ctl_map,
+	.pdata = &sdhci_arasan_cqe_pdata,
+};
+
 #ifdef CONFIG_PM_SLEEP
 /**
  * sdhci_arasan_suspend - Suspend method for the driver
@@ -462,14 +518,25 @@ static const struct of_device_id sdhci_arasan_of_match[] = {
 	/* SoC-specific compatible strings w/ soc_ctl_map */
 	{
 		.compatible = "rockchip,rk3399-sdhci-5.1",
-		.data = &rk3399_soc_ctl_map,
+		.data = &sdhci_arasan_rk3399_data,
+	},
+	{
+		.compatible = "ti,am654-sdhci-5.1",
+		.data = &sdhci_arasan_am654_data,
 	},
-
 	/* Generic compatible below here */
-	{ .compatible = "arasan,sdhci-8.9a" },
-	{ .compatible = "arasan,sdhci-5.1" },
-	{ .compatible = "arasan,sdhci-4.9a" },
-
+	{
+		.compatible = "arasan,sdhci-8.9a",
+		.data = &sdhci_arasan_data,
+	},
+	{
+		.compatible = "arasan,sdhci-5.1",
+		.data = &sdhci_arasan_data,
+	},
+	{
+		.compatible = "arasan,sdhci-4.9a",
+		.data = &sdhci_arasan_data,
+	},
 	{ /* sentinel */ }
 };
 MODULE_DEVICE_TABLE(of, sdhci_arasan_of_match);
@@ -707,14 +774,11 @@ static int sdhci_arasan_probe(struct platform_device *pdev)
 	struct sdhci_pltfm_host *pltfm_host;
 	struct sdhci_arasan_data *sdhci_arasan;
 	struct device_node *np = pdev->dev.of_node;
-	const struct sdhci_pltfm_data *pdata;
-
-	if (of_device_is_compatible(pdev->dev.of_node, "arasan,sdhci-5.1"))
-		pdata = &sdhci_arasan_cqe_pdata;
-	else
-		pdata = &sdhci_arasan_pdata;
+	const struct sdhci_arasan_of_data *data;
 
-	host = sdhci_pltfm_init(pdev, pdata, sizeof(*sdhci_arasan));
+	match = of_match_node(sdhci_arasan_of_match, pdev->dev.of_node);
+	data = match->data;
+	host = sdhci_pltfm_init(pdev, data->pdata, sizeof(*sdhci_arasan));
 
 	if (IS_ERR(host))
 		return PTR_ERR(host);
@@ -723,8 +787,7 @@ static int sdhci_arasan_probe(struct platform_device *pdev)
 	sdhci_arasan = sdhci_pltfm_priv(pltfm_host);
 	sdhci_arasan->host = host;
 
-	match = of_match_node(sdhci_arasan_of_match, pdev->dev.of_node);
-	sdhci_arasan->soc_ctl_map = match->data;
+	sdhci_arasan->soc_ctl_map = data->soc_ctl_map;
 
 	node = of_parse_phandle(pdev->dev.of_node, "arasan,soc-ctl-syscon", 0);
 	if (node) {
@@ -788,7 +851,8 @@ static int sdhci_arasan_probe(struct platform_device *pdev)
 
 	ret = mmc_of_parse(host->mmc);
 	if (ret) {
-		dev_err(&pdev->dev, "parsing dt failed (%d)\n", ret);
+		if (ret != -EPROBE_DEFER)
+			dev_err(&pdev->dev, "parsing dt failed (%d)\n", ret);
 		goto unreg_clk;
 	}
 
diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
index 1b7cd144fb01..a5137845a1c7 100644
--- a/drivers/mmc/host/sdhci-of-dwcmshc.c
+++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
@@ -8,21 +8,51 @@
  */
 
 #include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/sizes.h>
 
 #include "sdhci-pltfm.h"
 
+#define BOUNDARY_OK(addr, len) \
+	((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
+
 struct dwcmshc_priv {
 	struct clk	*bus_clk;
 };
 
+/*
+ * If DMA addr spans 128MB boundary, we split the DMA transfer into two
+ * so that each DMA transfer doesn't exceed the boundary.
+ */
+static void dwcmshc_adma_write_desc(struct sdhci_host *host, void **desc,
+				    dma_addr_t addr, int len, unsigned int cmd)
+{
+	int tmplen, offset;
+
+	if (likely(!len || BOUNDARY_OK(addr, len))) {
+		sdhci_adma_write_desc(host, desc, addr, len, cmd);
+		return;
+	}
+
+	offset = addr & (SZ_128M - 1);
+	tmplen = SZ_128M - offset;
+	sdhci_adma_write_desc(host, desc, addr, tmplen, cmd);
+
+	addr += tmplen;
+	len -= tmplen;
+	sdhci_adma_write_desc(host, desc, addr, len, cmd);
+}
+
 static const struct sdhci_ops sdhci_dwcmshc_ops = {
 	.set_clock		= sdhci_set_clock,
 	.set_bus_width		= sdhci_set_bus_width,
 	.set_uhs_signaling	= sdhci_set_uhs_signaling,
 	.get_max_clock		= sdhci_pltfm_clk_get_max_clock,
 	.reset			= sdhci_reset,
+	.adma_write_desc	= dwcmshc_adma_write_desc,
 };
 
 static const struct sdhci_pltfm_data sdhci_dwcmshc_pdata = {
@@ -36,12 +66,21 @@ static int dwcmshc_probe(struct platform_device *pdev)
 	struct sdhci_host *host;
 	struct dwcmshc_priv *priv;
 	int err;
+	u32 extra;
 
 	host = sdhci_pltfm_init(pdev, &sdhci_dwcmshc_pdata,
 				sizeof(struct dwcmshc_priv));
 	if (IS_ERR(host))
 		return PTR_ERR(host);
 
+	/*
+	 * extra adma table cnt for cross 128M boundary handling.
+	 */
+	extra = DIV_ROUND_UP_ULL(dma_get_required_mask(&pdev->dev), SZ_128M);
+	if (extra > SDHCI_MAX_SEGS)
+		extra = SDHCI_MAX_SEGS;
+	host->adma_table_cnt += extra;
+
 	pltfm_host = sdhci_priv(host);
 	priv = sdhci_pltfm_priv(pltfm_host);
 
diff --git a/drivers/mmc/host/sdhci-of-esdhc.c b/drivers/mmc/host/sdhci-of-esdhc.c
index 9cb7554a463d..86fc9f022002 100644
--- a/drivers/mmc/host/sdhci-of-esdhc.c
+++ b/drivers/mmc/host/sdhci-of-esdhc.c
@@ -78,8 +78,10 @@ struct sdhci_esdhc {
 	u8 vendor_ver;
 	u8 spec_ver;
 	bool quirk_incorrect_hostver;
+	bool quirk_fixup_tuning;
 	unsigned int peripheral_clock;
 	const struct esdhc_clk_fixup *clk_fixup;
+	u32 div_ratio;
 };
 
 /**
@@ -580,6 +582,7 @@ static void esdhc_of_set_clock(struct sdhci_host *host, unsigned int clock)
 	dev_dbg(mmc_dev(host->mmc), "desired SD clock: %d, actual: %d\n",
 		clock, host->max_clk / pre_div / div);
 	host->mmc->actual_clock = host->max_clk / pre_div / div;
+	esdhc->div_ratio = pre_div * div;
 	pre_div >>= 1;
 	div--;
 
@@ -712,9 +715,24 @@ static int esdhc_signal_voltage_switch(struct mmc_host *mmc,
 	}
 }
 
+static struct soc_device_attribute soc_fixup_tuning[] = {
+	{ .family = "QorIQ T1040", .revision = "1.0", },
+	{ .family = "QorIQ T2080", .revision = "1.0", },
+	{ .family = "QorIQ T1023", .revision = "1.0", },
+	{ .family = "QorIQ LS1021A", .revision = "1.0", },
+	{ .family = "QorIQ LS1080A", .revision = "1.0", },
+	{ .family = "QorIQ LS2080A", .revision = "1.0", },
+	{ .family = "QorIQ LS1012A", .revision = "1.0", },
+	{ .family = "QorIQ LS1043A", .revision = "1.*", },
+	{ .family = "QorIQ LS1046A", .revision = "1.0", },
+	{ },
+};
+
 static int esdhc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 {
 	struct sdhci_host *host = mmc_priv(mmc);
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_esdhc *esdhc = sdhci_pltfm_priv(pltfm_host);
 	u32 val;
 
 	/* Use tuning block for tuning procedure */
@@ -728,7 +746,26 @@ static int esdhc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 	sdhci_writel(host, val, ESDHC_TBCTL);
 	esdhc_clock_enable(host, true);
 
-	return sdhci_execute_tuning(mmc, opcode);
+	sdhci_execute_tuning(mmc, opcode);
+	if (host->tuning_err == -EAGAIN && esdhc->quirk_fixup_tuning) {
+
+		/* program TBPTR[TB_WNDW_END_PTR] = 3*DIV_RATIO and
+		 * program TBPTR[TB_WNDW_START_PTR] = 5*DIV_RATIO
+		 */
+		val = sdhci_readl(host, ESDHC_TBPTR);
+		val = (val & ~((0x7f << 8) | 0x7f)) |
+		(3 * esdhc->div_ratio) | ((5 * esdhc->div_ratio) << 8);
+		sdhci_writel(host, val, ESDHC_TBPTR);
+
+		/* program the software tuning mode by setting
+		 * TBCTL[TB_MODE]=2'h3
+		 */
+		val = sdhci_readl(host, ESDHC_TBCTL);
+		val |= 0x3;
+		sdhci_writel(host, val, ESDHC_TBCTL);
+		sdhci_execute_tuning(mmc, opcode);
+	}
+	return 0;
 }
 
 #ifdef CONFIG_PM_SLEEP
@@ -903,6 +940,11 @@ static int sdhci_esdhc_probe(struct platform_device *pdev)
 
 	pltfm_host = sdhci_priv(host);
 	esdhc = sdhci_pltfm_priv(pltfm_host);
+	if (soc_device_match(soc_fixup_tuning))
+		esdhc->quirk_fixup_tuning = true;
+	else
+		esdhc->quirk_fixup_tuning = false;
+
 	if (esdhc->vendor_ver == VENDOR_V_22)
 		host->quirks2 |= SDHCI_QUIRK2_HOST_NO_CMD23;
 
diff --git a/drivers/mmc/host/sdhci-pci-o2micro.c b/drivers/mmc/host/sdhci-pci-o2micro.c
index 77e9bc4aaee9..cc3ffeffd7a2 100644
--- a/drivers/mmc/host/sdhci-pci-o2micro.c
+++ b/drivers/mmc/host/sdhci-pci-o2micro.c
@@ -490,6 +490,9 @@ int sdhci_pci_o2_probe(struct sdhci_pci_chip *chip)
 		pci_write_config_byte(chip->pdev, O2_SD_LOCK_WP, scratch);
 		break;
 	case PCI_DEVICE_ID_O2_SEABIRD0:
+		if (chip->pdev->revision == 0x01)
+			chip->quirks |= SDHCI_QUIRK_DELAY_AFTER_POWER;
+		/* fall through */
 	case PCI_DEVICE_ID_O2_SEABIRD1:
 		/* UnLock WP */
 		ret = pci_read_config_byte(chip->pdev,
diff --git a/drivers/mmc/host/sdhci-pltfm.c b/drivers/mmc/host/sdhci-pltfm.c
index 02bea6159d79..b231c9a3f888 100644
--- a/drivers/mmc/host/sdhci-pltfm.c
+++ b/drivers/mmc/host/sdhci-pltfm.c
@@ -30,6 +30,7 @@
 
 #include <linux/err.h>
 #include <linux/module.h>
+#include <linux/property.h>
 #include <linux/of.h>
 #ifdef CONFIG_PPC
 #include <asm/machdep.h>
@@ -51,11 +52,10 @@ static const struct sdhci_ops sdhci_pltfm_ops = {
 	.set_uhs_signaling = sdhci_set_uhs_signaling,
 };
 
-#ifdef CONFIG_OF
-static bool sdhci_of_wp_inverted(struct device_node *np)
+static bool sdhci_wp_inverted(struct device *dev)
 {
-	if (of_get_property(np, "sdhci,wp-inverted", NULL) ||
-	    of_get_property(np, "wp-inverted", NULL))
+	if (device_property_present(dev, "sdhci,wp-inverted") ||
+	    device_property_present(dev, "wp-inverted"))
 		return true;
 
 	/* Old device trees don't have the wp-inverted property. */
@@ -66,52 +66,64 @@ static bool sdhci_of_wp_inverted(struct device_node *np)
 #endif /* CONFIG_PPC */
 }
 
-void sdhci_get_of_property(struct platform_device *pdev)
+#ifdef CONFIG_OF
+static void sdhci_get_compatibility(struct platform_device *pdev)
 {
+	struct sdhci_host *host = platform_get_drvdata(pdev);
 	struct device_node *np = pdev->dev.of_node;
+
+	if (!np)
+		return;
+
+	if (of_device_is_compatible(np, "fsl,p2020-rev1-esdhc"))
+		host->quirks |= SDHCI_QUIRK_BROKEN_DMA;
+
+	if (of_device_is_compatible(np, "fsl,p2020-esdhc") ||
+	    of_device_is_compatible(np, "fsl,p1010-esdhc") ||
+	    of_device_is_compatible(np, "fsl,t4240-esdhc") ||
+	    of_device_is_compatible(np, "fsl,mpc8536-esdhc"))
+		host->quirks |= SDHCI_QUIRK_BROKEN_TIMEOUT_VAL;
+}
+#else
+void sdhci_get_compatibility(struct platform_device *pdev) {}
+#endif /* CONFIG_OF */
+
+void sdhci_get_property(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
 	struct sdhci_host *host = platform_get_drvdata(pdev);
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	u32 bus_width;
 
-	if (of_get_property(np, "sdhci,auto-cmd12", NULL))
+	if (device_property_present(dev, "sdhci,auto-cmd12"))
 		host->quirks |= SDHCI_QUIRK_MULTIBLOCK_READ_ACMD12;
 
-	if (of_get_property(np, "sdhci,1-bit-only", NULL) ||
-	    (of_property_read_u32(np, "bus-width", &bus_width) == 0 &&
+	if (device_property_present(dev, "sdhci,1-bit-only") ||
+	    (device_property_read_u32(dev, "bus-width", &bus_width) == 0 &&
 	    bus_width == 1))
 		host->quirks |= SDHCI_QUIRK_FORCE_1_BIT_DATA;
 
-	if (sdhci_of_wp_inverted(np))
+	if (sdhci_wp_inverted(dev))
 		host->quirks |= SDHCI_QUIRK_INVERTED_WRITE_PROTECT;
 
-	if (of_get_property(np, "broken-cd", NULL))
+	if (device_property_present(dev, "broken-cd"))
 		host->quirks |= SDHCI_QUIRK_BROKEN_CARD_DETECTION;
 
-	if (of_get_property(np, "no-1-8-v", NULL))
+	if (device_property_present(dev, "no-1-8-v"))
 		host->quirks2 |= SDHCI_QUIRK2_NO_1_8_V;
 
-	if (of_device_is_compatible(np, "fsl,p2020-rev1-esdhc"))
-		host->quirks |= SDHCI_QUIRK_BROKEN_DMA;
-
-	if (of_device_is_compatible(np, "fsl,p2020-esdhc") ||
-	    of_device_is_compatible(np, "fsl,p1010-esdhc") ||
-	    of_device_is_compatible(np, "fsl,t4240-esdhc") ||
-	    of_device_is_compatible(np, "fsl,mpc8536-esdhc"))
-		host->quirks |= SDHCI_QUIRK_BROKEN_TIMEOUT_VAL;
+	sdhci_get_compatibility(pdev);
 
-	of_property_read_u32(np, "clock-frequency", &pltfm_host->clock);
+	device_property_read_u32(dev, "clock-frequency", &pltfm_host->clock);
 
-	if (of_find_property(np, "keep-power-in-suspend", NULL))
+	if (device_property_present(dev, "keep-power-in-suspend"))
 		host->mmc->pm_caps |= MMC_PM_KEEP_POWER;
 
-	if (of_property_read_bool(np, "wakeup-source") ||
-	    of_property_read_bool(np, "enable-sdio-wakeup")) /* legacy */
+	if (device_property_read_bool(dev, "wakeup-source") ||
+	    device_property_read_bool(dev, "enable-sdio-wakeup")) /* legacy */
 		host->mmc->pm_caps |= MMC_PM_WAKE_SDIO_IRQ;
 }
-#else
-void sdhci_get_of_property(struct platform_device *pdev) {}
-#endif /* CONFIG_OF */
-EXPORT_SYMBOL_GPL(sdhci_get_of_property);
+EXPORT_SYMBOL_GPL(sdhci_get_property);
 
 struct sdhci_host *sdhci_pltfm_init(struct platform_device *pdev,
 				    const struct sdhci_pltfm_data *pdata,
@@ -184,7 +196,7 @@ int sdhci_pltfm_register(struct platform_device *pdev,
 	if (IS_ERR(host))
 		return PTR_ERR(host);
 
-	sdhci_get_of_property(pdev);
+	sdhci_get_property(pdev);
 
 	ret = sdhci_add_host(host);
 	if (ret)
diff --git a/drivers/mmc/host/sdhci-pltfm.h b/drivers/mmc/host/sdhci-pltfm.h
index 1e91fb1c020e..6109987fc3b5 100644
--- a/drivers/mmc/host/sdhci-pltfm.h
+++ b/drivers/mmc/host/sdhci-pltfm.h
@@ -90,7 +90,12 @@ static inline void sdhci_be32bs_writeb(struct sdhci_host *host, u8 val, int reg)
 }
 #endif /* CONFIG_MMC_SDHCI_BIG_ENDIAN_32BIT_BYTE_SWAPPER */
 
-extern void sdhci_get_of_property(struct platform_device *pdev);
+void sdhci_get_property(struct platform_device *pdev);
+
+static inline void sdhci_get_of_property(struct platform_device *pdev)
+{
+	return sdhci_get_property(pdev);
+}
 
 extern struct sdhci_host *sdhci_pltfm_init(struct platform_device *pdev,
 					  const struct sdhci_pltfm_data *pdata,
diff --git a/drivers/mmc/host/sdhci-pxav3.c b/drivers/mmc/host/sdhci-pxav3.c
index b8e96f392428..1783e29eae04 100644
--- a/drivers/mmc/host/sdhci-pxav3.c
+++ b/drivers/mmc/host/sdhci-pxav3.c
@@ -21,17 +21,14 @@
 #include <linux/platform_device.h>
 #include <linux/clk.h>
 #include <linux/io.h>
-#include <linux/gpio.h>
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
-#include <linux/mmc/slot-gpio.h>
 #include <linux/platform_data/pxa_sdhci.h>
 #include <linux/slab.h>
 #include <linux/delay.h>
 #include <linux/module.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
-#include <linux/of_gpio.h>
 #include <linux/pm.h>
 #include <linux/pm_runtime.h>
 #include <linux/mbus.h>
@@ -452,16 +449,6 @@ static int sdhci_pxav3_probe(struct platform_device *pdev)
 			host->mmc->caps2 |= pdata->host_caps2;
 		if (pdata->pm_caps)
 			host->mmc->pm_caps |= pdata->pm_caps;
-
-		if (gpio_is_valid(pdata->ext_cd_gpio)) {
-			ret = mmc_gpio_request_cd(host->mmc, pdata->ext_cd_gpio,
-						  0);
-			if (ret) {
-				dev_err(mmc_dev(host->mmc),
-					"failed to allocate card detect gpio\n");
-				goto err_cd_req;
-			}
-		}
 	}
 
 	pm_runtime_get_noresume(&pdev->dev);
@@ -486,7 +473,6 @@ err_add_host:
 	pm_runtime_disable(&pdev->dev);
 	pm_runtime_put_noidle(&pdev->dev);
 err_of_parse:
-err_cd_req:
 err_mbus_win:
 	clk_disable_unprepare(pxa->clk_io);
 	clk_disable_unprepare(pxa->clk_core);
diff --git a/drivers/mmc/host/sdhci-sirf.c b/drivers/mmc/host/sdhci-sirf.c
index 391d52b467ca..5eada6f87e60 100644
--- a/drivers/mmc/host/sdhci-sirf.c
+++ b/drivers/mmc/host/sdhci-sirf.c
@@ -11,7 +11,6 @@
 #include <linux/mmc/host.h>
 #include <linux/module.h>
 #include <linux/of.h>
-#include <linux/of_gpio.h>
 #include <linux/mmc/slot-gpio.h>
 #include "sdhci-pltfm.h"
 
@@ -19,10 +18,6 @@
 #define SDHCI_SIRF_8BITBUS BIT(3)
 #define SIRF_TUNING_COUNT 16384
 
-struct sdhci_sirf_priv {
-	int gpio_cd;
-};
-
 static void sdhci_sirf_set_bus_width(struct sdhci_host *host, int width)
 {
 	u8 ctrl;
@@ -170,9 +165,7 @@ static int sdhci_sirf_probe(struct platform_device *pdev)
 {
 	struct sdhci_host *host;
 	struct sdhci_pltfm_host *pltfm_host;
-	struct sdhci_sirf_priv *priv;
 	struct clk *clk;
-	int gpio_cd;
 	int ret;
 
 	clk = devm_clk_get(&pdev->dev, NULL);
@@ -181,19 +174,12 @@ static int sdhci_sirf_probe(struct platform_device *pdev)
 		return PTR_ERR(clk);
 	}
 
-	if (pdev->dev.of_node)
-		gpio_cd = of_get_named_gpio(pdev->dev.of_node, "cd-gpios", 0);
-	else
-		gpio_cd = -EINVAL;
-
-	host = sdhci_pltfm_init(pdev, &sdhci_sirf_pdata, sizeof(struct sdhci_sirf_priv));
+	host = sdhci_pltfm_init(pdev, &sdhci_sirf_pdata, 0);
 	if (IS_ERR(host))
 		return PTR_ERR(host);
 
 	pltfm_host = sdhci_priv(host);
 	pltfm_host->clk = clk;
-	priv = sdhci_pltfm_priv(pltfm_host);
-	priv->gpio_cd = gpio_cd;
 
 	sdhci_get_of_property(pdev);
 
@@ -209,15 +195,11 @@ static int sdhci_sirf_probe(struct platform_device *pdev)
 	 * We must request the IRQ after sdhci_add_host(), as the tasklet only
 	 * gets setup in sdhci_add_host() and we oops.
 	 */
-	if (gpio_is_valid(priv->gpio_cd)) {
-		ret = mmc_gpio_request_cd(host->mmc, priv->gpio_cd, 0);
-		if (ret) {
-			dev_err(&pdev->dev, "card detect irq request failed: %d\n",
-				ret);
-			goto err_request_cd;
-		}
+	ret = mmc_gpiod_request_cd(host->mmc, "cd", 0, false, 0, NULL);
+	if (ret == -EPROBE_DEFER)
+		goto err_request_cd;
+	if (!ret)
 		mmc_gpiod_request_cd_irq(host->mmc);
-	}
 
 	return 0;
 
diff --git a/drivers/mmc/host/sdhci-spear.c b/drivers/mmc/host/sdhci-spear.c
index 9247d51f2eed..916b5b09c3d1 100644
--- a/drivers/mmc/host/sdhci-spear.c
+++ b/drivers/mmc/host/sdhci-spear.c
@@ -15,13 +15,11 @@
 
 #include <linux/clk.h>
 #include <linux/delay.h>
-#include <linux/gpio.h>
 #include <linux/highmem.h>
 #include <linux/module.h>
 #include <linux/interrupt.h>
 #include <linux/irq.h>
 #include <linux/of.h>
-#include <linux/of_gpio.h>
 #include <linux/platform_device.h>
 #include <linux/pm.h>
 #include <linux/slab.h>
@@ -32,7 +30,6 @@
 
 struct spear_sdhci {
 	struct clk *clk;
-	int card_int_gpio;
 };
 
 /* sdhci ops */
@@ -43,18 +40,6 @@ static const struct sdhci_ops sdhci_pltfm_ops = {
 	.set_uhs_signaling = sdhci_set_uhs_signaling,
 };
 
-static void sdhci_probe_config_dt(struct device_node *np,
-				struct spear_sdhci *host)
-{
-	int cd_gpio;
-
-	cd_gpio = of_get_named_gpio(np, "cd-gpios", 0);
-	if (!gpio_is_valid(cd_gpio))
-		cd_gpio = -1;
-
-	host->card_int_gpio = cd_gpio;
-}
-
 static int sdhci_probe(struct platform_device *pdev)
 {
 	struct sdhci_host *host;
@@ -109,21 +94,13 @@ static int sdhci_probe(struct platform_device *pdev)
 		dev_dbg(&pdev->dev, "Error setting desired clk, clk=%lu\n",
 				clk_get_rate(sdhci->clk));
 
-	sdhci_probe_config_dt(pdev->dev.of_node, sdhci);
 	/*
-	 * It is optional to use GPIOs for sdhci card detection. If
-	 * sdhci->card_int_gpio < 0, then use original sdhci lines otherwise
-	 * GPIO lines. We use the built-in GPIO support for this.
+	 * It is optional to use GPIOs for sdhci card detection. If we
+	 * find a descriptor using slot GPIO, we use it.
 	 */
-	if (sdhci->card_int_gpio >= 0) {
-		ret = mmc_gpio_request_cd(host->mmc, sdhci->card_int_gpio, 0);
-		if (ret < 0) {
-			dev_dbg(&pdev->dev,
-				"failed to request card-detect gpio%d\n",
-				sdhci->card_int_gpio);
-			goto disable_clk;
-		}
-	}
+	ret = mmc_gpiod_request_cd(host->mmc, "cd", 0, false, 0, NULL);
+	if (ret == -EPROBE_DEFER)
+		goto disable_clk;
 
 	ret = sdhci_add_host(host);
 	if (ret)
diff --git a/drivers/mmc/host/sdhci-sprd.c b/drivers/mmc/host/sdhci-sprd.c
new file mode 100644
index 000000000000..9a822e2e9f0b
--- /dev/null
+++ b/drivers/mmc/host/sdhci-sprd.c
@@ -0,0 +1,498 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// Secure Digital Host Controller
+//
+// Copyright (C) 2018 Spreadtrum, Inc.
+// Author: Chunyan Zhang <chunyan.zhang@unisoc.com>
+
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/highmem.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/pm_runtime.h>
+#include <linux/regulator/consumer.h>
+#include <linux/slab.h>
+
+#include "sdhci-pltfm.h"
+
+/* SDHCI_ARGUMENT2 register high 16bit */
+#define SDHCI_SPRD_ARG2_STUFF		GENMASK(31, 16)
+
+#define SDHCI_SPRD_REG_32_DLL_DLY_OFFSET	0x208
+#define  SDHCIBSPRD_IT_WR_DLY_INV		BIT(5)
+#define  SDHCI_SPRD_BIT_CMD_DLY_INV		BIT(13)
+#define  SDHCI_SPRD_BIT_POSRD_DLY_INV		BIT(21)
+#define  SDHCI_SPRD_BIT_NEGRD_DLY_INV		BIT(29)
+
+#define SDHCI_SPRD_REG_32_BUSY_POSI		0x250
+#define  SDHCI_SPRD_BIT_OUTR_CLK_AUTO_EN	BIT(25)
+#define  SDHCI_SPRD_BIT_INNR_CLK_AUTO_EN	BIT(24)
+
+#define SDHCI_SPRD_REG_DEBOUNCE		0x28C
+#define  SDHCI_SPRD_BIT_DLL_BAK		BIT(0)
+#define  SDHCI_SPRD_BIT_DLL_VAL		BIT(1)
+
+#define  SDHCI_SPRD_INT_SIGNAL_MASK	0x1B7F410B
+
+/* SDHCI_HOST_CONTROL2 */
+#define  SDHCI_SPRD_CTRL_HS200		0x0005
+#define  SDHCI_SPRD_CTRL_HS400		0x0006
+
+/*
+ * According to the standard specification, BIT(3) of SDHCI_SOFTWARE_RESET is
+ * reserved, and only used on Spreadtrum's design, the hardware cannot work
+ * if this bit is cleared.
+ * 1 : normal work
+ * 0 : hardware reset
+ */
+#define  SDHCI_HW_RESET_CARD		BIT(3)
+
+#define SDHCI_SPRD_MAX_CUR		0xFFFFFF
+#define SDHCI_SPRD_CLK_MAX_DIV		1023
+
+#define SDHCI_SPRD_CLK_DEF_RATE		26000000
+
+struct sdhci_sprd_host {
+	u32 version;
+	struct clk *clk_sdio;
+	struct clk *clk_enable;
+	u32 base_rate;
+	int flags; /* backup of host attribute */
+};
+
+#define TO_SPRD_HOST(host) sdhci_pltfm_priv(sdhci_priv(host))
+
+static void sdhci_sprd_init_config(struct sdhci_host *host)
+{
+	u16 val;
+
+	/* set dll backup mode */
+	val = sdhci_readl(host, SDHCI_SPRD_REG_DEBOUNCE);
+	val |= SDHCI_SPRD_BIT_DLL_BAK | SDHCI_SPRD_BIT_DLL_VAL;
+	sdhci_writel(host, val, SDHCI_SPRD_REG_DEBOUNCE);
+}
+
+static inline u32 sdhci_sprd_readl(struct sdhci_host *host, int reg)
+{
+	if (unlikely(reg == SDHCI_MAX_CURRENT))
+		return SDHCI_SPRD_MAX_CUR;
+
+	return readl_relaxed(host->ioaddr + reg);
+}
+
+static inline void sdhci_sprd_writel(struct sdhci_host *host, u32 val, int reg)
+{
+	/* SDHCI_MAX_CURRENT is reserved on Spreadtrum's platform */
+	if (unlikely(reg == SDHCI_MAX_CURRENT))
+		return;
+
+	if (unlikely(reg == SDHCI_SIGNAL_ENABLE || reg == SDHCI_INT_ENABLE))
+		val = val & SDHCI_SPRD_INT_SIGNAL_MASK;
+
+	writel_relaxed(val, host->ioaddr + reg);
+}
+
+static inline void sdhci_sprd_writew(struct sdhci_host *host, u16 val, int reg)
+{
+	/* SDHCI_BLOCK_COUNT is Read Only on Spreadtrum's platform */
+	if (unlikely(reg == SDHCI_BLOCK_COUNT))
+		return;
+
+	writew_relaxed(val, host->ioaddr + reg);
+}
+
+static inline void sdhci_sprd_writeb(struct sdhci_host *host, u8 val, int reg)
+{
+	/*
+	 * Since BIT(3) of SDHCI_SOFTWARE_RESET is reserved according to the
+	 * standard specification, sdhci_reset() write this register directly
+	 * without checking other reserved bits, that will clear BIT(3) which
+	 * is defined as hardware reset on Spreadtrum's platform and clearing
+	 * it by mistake will lead the card not work. So here we need to work
+	 * around it.
+	 */
+	if (unlikely(reg == SDHCI_SOFTWARE_RESET)) {
+		if (readb_relaxed(host->ioaddr + reg) & SDHCI_HW_RESET_CARD)
+			val |= SDHCI_HW_RESET_CARD;
+	}
+
+	writeb_relaxed(val, host->ioaddr + reg);
+}
+
+static inline void sdhci_sprd_sd_clk_off(struct sdhci_host *host)
+{
+	u16 ctrl = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
+
+	ctrl &= ~SDHCI_CLOCK_CARD_EN;
+	sdhci_writew(host, ctrl, SDHCI_CLOCK_CONTROL);
+}
+
+static inline void
+sdhci_sprd_set_dll_invert(struct sdhci_host *host, u32 mask, bool en)
+{
+	u32 dll_dly_offset;
+
+	dll_dly_offset = sdhci_readl(host, SDHCI_SPRD_REG_32_DLL_DLY_OFFSET);
+	if (en)
+		dll_dly_offset |= mask;
+	else
+		dll_dly_offset &= ~mask;
+	sdhci_writel(host, dll_dly_offset, SDHCI_SPRD_REG_32_DLL_DLY_OFFSET);
+}
+
+static inline u32 sdhci_sprd_calc_div(u32 base_clk, u32 clk)
+{
+	u32 div;
+
+	/* select 2x clock source */
+	if (base_clk <= clk * 2)
+		return 0;
+
+	div = (u32) (base_clk / (clk * 2));
+
+	if ((base_clk / div) > (clk * 2))
+		div++;
+
+	if (div > SDHCI_SPRD_CLK_MAX_DIV)
+		div = SDHCI_SPRD_CLK_MAX_DIV;
+
+	if (div % 2)
+		div = (div + 1) / 2;
+	else
+		div = div / 2;
+
+	return div;
+}
+
+static inline void _sdhci_sprd_set_clock(struct sdhci_host *host,
+					unsigned int clk)
+{
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+	u32 div, val, mask;
+
+	div = sdhci_sprd_calc_div(sprd_host->base_rate, clk);
+
+	clk |= ((div & 0x300) >> 2) | ((div & 0xFF) << 8);
+	sdhci_enable_clk(host, clk);
+
+	/* enable auto gate sdhc_enable_auto_gate */
+	val = sdhci_readl(host, SDHCI_SPRD_REG_32_BUSY_POSI);
+	mask = SDHCI_SPRD_BIT_OUTR_CLK_AUTO_EN |
+	       SDHCI_SPRD_BIT_INNR_CLK_AUTO_EN;
+	if (mask != (val & mask)) {
+		val |= mask;
+		sdhci_writel(host, val, SDHCI_SPRD_REG_32_BUSY_POSI);
+	}
+}
+
+static void sdhci_sprd_set_clock(struct sdhci_host *host, unsigned int clock)
+{
+	bool en = false;
+
+	if (clock == 0) {
+		sdhci_writew(host, 0, SDHCI_CLOCK_CONTROL);
+	} else if (clock != host->clock) {
+		sdhci_sprd_sd_clk_off(host);
+		_sdhci_sprd_set_clock(host, clock);
+
+		if (clock <= 400000)
+			en = true;
+		sdhci_sprd_set_dll_invert(host, SDHCI_SPRD_BIT_CMD_DLY_INV |
+					  SDHCI_SPRD_BIT_POSRD_DLY_INV, en);
+	} else {
+		_sdhci_sprd_set_clock(host, clock);
+	}
+}
+
+static unsigned int sdhci_sprd_get_max_clock(struct sdhci_host *host)
+{
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+
+	return clk_round_rate(sprd_host->clk_sdio, ULONG_MAX);
+}
+
+static unsigned int sdhci_sprd_get_min_clock(struct sdhci_host *host)
+{
+	return 400000;
+}
+
+static void sdhci_sprd_set_uhs_signaling(struct sdhci_host *host,
+					 unsigned int timing)
+{
+	u16 ctrl_2;
+
+	if (timing == host->timing)
+		return;
+
+	ctrl_2 = sdhci_readw(host, SDHCI_HOST_CONTROL2);
+	/* Select Bus Speed Mode for host */
+	ctrl_2 &= ~SDHCI_CTRL_UHS_MASK;
+	switch (timing) {
+	case MMC_TIMING_UHS_SDR12:
+		ctrl_2 |= SDHCI_CTRL_UHS_SDR12;
+		break;
+	case MMC_TIMING_MMC_HS:
+	case MMC_TIMING_SD_HS:
+	case MMC_TIMING_UHS_SDR25:
+		ctrl_2 |= SDHCI_CTRL_UHS_SDR25;
+		break;
+	case MMC_TIMING_UHS_SDR50:
+		ctrl_2 |= SDHCI_CTRL_UHS_SDR50;
+		break;
+	case MMC_TIMING_UHS_SDR104:
+		ctrl_2 |= SDHCI_CTRL_UHS_SDR104;
+		break;
+	case MMC_TIMING_UHS_DDR50:
+	case MMC_TIMING_MMC_DDR52:
+		ctrl_2 |= SDHCI_CTRL_UHS_DDR50;
+		break;
+	case MMC_TIMING_MMC_HS200:
+		ctrl_2 |= SDHCI_SPRD_CTRL_HS200;
+		break;
+	case MMC_TIMING_MMC_HS400:
+		ctrl_2 |= SDHCI_SPRD_CTRL_HS400;
+		break;
+	default:
+		break;
+	}
+
+	sdhci_writew(host, ctrl_2, SDHCI_HOST_CONTROL2);
+}
+
+static void sdhci_sprd_hw_reset(struct sdhci_host *host)
+{
+	int val;
+
+	/*
+	 * Note: don't use sdhci_writeb() API here since it is redirected to
+	 * sdhci_sprd_writeb() in which we have a workaround for
+	 * SDHCI_SOFTWARE_RESET which would make bit SDHCI_HW_RESET_CARD can
+	 * not be cleared.
+	 */
+	val = readb_relaxed(host->ioaddr + SDHCI_SOFTWARE_RESET);
+	val &= ~SDHCI_HW_RESET_CARD;
+	writeb_relaxed(val, host->ioaddr + SDHCI_SOFTWARE_RESET);
+	/* wait for 10 us */
+	usleep_range(10, 20);
+
+	val |= SDHCI_HW_RESET_CARD;
+	writeb_relaxed(val, host->ioaddr + SDHCI_SOFTWARE_RESET);
+	usleep_range(300, 500);
+}
+
+static struct sdhci_ops sdhci_sprd_ops = {
+	.read_l = sdhci_sprd_readl,
+	.write_l = sdhci_sprd_writel,
+	.write_b = sdhci_sprd_writeb,
+	.set_clock = sdhci_sprd_set_clock,
+	.get_max_clock = sdhci_sprd_get_max_clock,
+	.get_min_clock = sdhci_sprd_get_min_clock,
+	.set_bus_width = sdhci_set_bus_width,
+	.reset = sdhci_reset,
+	.set_uhs_signaling = sdhci_sprd_set_uhs_signaling,
+	.hw_reset = sdhci_sprd_hw_reset,
+};
+
+static void sdhci_sprd_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct sdhci_host *host = mmc_priv(mmc);
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+
+	host->flags |= sprd_host->flags & SDHCI_AUTO_CMD23;
+
+	/*
+	 * From version 4.10 onward, ARGUMENT2 register is also as 32-bit
+	 * block count register which doesn't support stuff bits of
+	 * CMD23 argument on Spreadtrum's sd host controller.
+	 */
+	if (host->version >= SDHCI_SPEC_410 &&
+	    mrq->sbc && (mrq->sbc->arg & SDHCI_SPRD_ARG2_STUFF) &&
+	    (host->flags & SDHCI_AUTO_CMD23))
+		host->flags &= ~SDHCI_AUTO_CMD23;
+
+	sdhci_request(mmc, mrq);
+}
+
+static const struct sdhci_pltfm_data sdhci_sprd_pdata = {
+	.quirks = SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK,
+	.quirks2 = SDHCI_QUIRK2_BROKEN_HS200 |
+		   SDHCI_QUIRK2_USE_32BIT_BLK_CNT,
+	.ops = &sdhci_sprd_ops,
+};
+
+static int sdhci_sprd_probe(struct platform_device *pdev)
+{
+	struct sdhci_host *host;
+	struct sdhci_sprd_host *sprd_host;
+	struct clk *clk;
+	int ret = 0;
+
+	host = sdhci_pltfm_init(pdev, &sdhci_sprd_pdata, sizeof(*sprd_host));
+	if (IS_ERR(host))
+		return PTR_ERR(host);
+
+	host->dma_mask = DMA_BIT_MASK(64);
+	pdev->dev.dma_mask = &host->dma_mask;
+	host->mmc_host_ops.request = sdhci_sprd_request;
+
+	host->mmc->caps = MMC_CAP_SD_HIGHSPEED | MMC_CAP_MMC_HIGHSPEED |
+		MMC_CAP_ERASE | MMC_CAP_CMD23;
+	ret = mmc_of_parse(host->mmc);
+	if (ret)
+		goto pltfm_free;
+
+	sprd_host = TO_SPRD_HOST(host);
+
+	clk = devm_clk_get(&pdev->dev, "sdio");
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto pltfm_free;
+	}
+	sprd_host->clk_sdio = clk;
+	sprd_host->base_rate = clk_get_rate(sprd_host->clk_sdio);
+	if (!sprd_host->base_rate)
+		sprd_host->base_rate = SDHCI_SPRD_CLK_DEF_RATE;
+
+	clk = devm_clk_get(&pdev->dev, "enable");
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto pltfm_free;
+	}
+	sprd_host->clk_enable = clk;
+
+	ret = clk_prepare_enable(sprd_host->clk_sdio);
+	if (ret)
+		goto pltfm_free;
+
+	clk_prepare_enable(sprd_host->clk_enable);
+	if (ret)
+		goto clk_disable;
+
+	sdhci_sprd_init_config(host);
+	host->version = sdhci_readw(host, SDHCI_HOST_VERSION);
+	sprd_host->version = ((host->version & SDHCI_VENDOR_VER_MASK) >>
+			       SDHCI_VENDOR_VER_SHIFT);
+
+	pm_runtime_get_noresume(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, 50);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_suspend_ignore_children(&pdev->dev, 1);
+
+	sdhci_enable_v4_mode(host);
+
+	ret = sdhci_setup_host(host);
+	if (ret)
+		goto pm_runtime_disable;
+
+	sprd_host->flags = host->flags;
+
+	ret = __sdhci_add_host(host);
+	if (ret)
+		goto err_cleanup_host;
+
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+
+	return 0;
+
+err_cleanup_host:
+	sdhci_cleanup_host(host);
+
+pm_runtime_disable:
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	clk_disable_unprepare(sprd_host->clk_enable);
+
+clk_disable:
+	clk_disable_unprepare(sprd_host->clk_sdio);
+
+pltfm_free:
+	sdhci_pltfm_free(pdev);
+	return ret;
+}
+
+static int sdhci_sprd_remove(struct platform_device *pdev)
+{
+	struct sdhci_host *host = platform_get_drvdata(pdev);
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+	struct mmc_host *mmc = host->mmc;
+
+	mmc_remove_host(mmc);
+	clk_disable_unprepare(sprd_host->clk_sdio);
+	clk_disable_unprepare(sprd_host->clk_enable);
+
+	mmc_free_host(mmc);
+
+	return 0;
+}
+
+static const struct of_device_id sdhci_sprd_of_match[] = {
+	{ .compatible = "sprd,sdhci-r11", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, sdhci_sprd_of_match);
+
+#ifdef CONFIG_PM
+static int sdhci_sprd_runtime_suspend(struct device *dev)
+{
+	struct sdhci_host *host = dev_get_drvdata(dev);
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+
+	sdhci_runtime_suspend_host(host);
+
+	clk_disable_unprepare(sprd_host->clk_sdio);
+	clk_disable_unprepare(sprd_host->clk_enable);
+
+	return 0;
+}
+
+static int sdhci_sprd_runtime_resume(struct device *dev)
+{
+	struct sdhci_host *host = dev_get_drvdata(dev);
+	struct sdhci_sprd_host *sprd_host = TO_SPRD_HOST(host);
+	int ret;
+
+	ret = clk_prepare_enable(sprd_host->clk_enable);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(sprd_host->clk_sdio);
+	if (ret) {
+		clk_disable_unprepare(sprd_host->clk_enable);
+		return ret;
+	}
+
+	sdhci_runtime_resume_host(host);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops sdhci_sprd_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				pm_runtime_force_resume)
+	SET_RUNTIME_PM_OPS(sdhci_sprd_runtime_suspend,
+			   sdhci_sprd_runtime_resume, NULL)
+};
+
+static struct platform_driver sdhci_sprd_driver = {
+	.probe = sdhci_sprd_probe,
+	.remove = sdhci_sprd_remove,
+	.driver = {
+		.name = "sdhci_sprd_r11",
+		.of_match_table = of_match_ptr(sdhci_sprd_of_match),
+		.pm = &sdhci_sprd_pm_ops,
+	},
+};
+module_platform_driver(sdhci_sprd_driver);
+
+MODULE_DESCRIPTION("Spreadtrum sdio host controller r11 driver");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:sdhci-sprd-r11");
diff --git a/drivers/mmc/host/sdhci-tegra.c b/drivers/mmc/host/sdhci-tegra.c
index 908b23e6a03c..7b95d088fdef 100644
--- a/drivers/mmc/host/sdhci-tegra.c
+++ b/drivers/mmc/host/sdhci-tegra.c
@@ -16,17 +16,21 @@
 #include <linux/err.h>
 #include <linux/module.h>
 #include <linux/init.h>
+#include <linux/iopoll.h>
 #include <linux/platform_device.h>
 #include <linux/clk.h>
 #include <linux/io.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/regulator/consumer.h>
 #include <linux/reset.h>
 #include <linux/mmc/card.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
 #include <linux/mmc/slot-gpio.h>
 #include <linux/gpio/consumer.h>
+#include <linux/ktime.h>
 
 #include "sdhci-pltfm.h"
 
@@ -34,40 +38,96 @@
 #define SDHCI_TEGRA_VENDOR_CLOCK_CTRL			0x100
 #define SDHCI_CLOCK_CTRL_TAP_MASK			0x00ff0000
 #define SDHCI_CLOCK_CTRL_TAP_SHIFT			16
+#define SDHCI_CLOCK_CTRL_TRIM_MASK			0x1f000000
+#define SDHCI_CLOCK_CTRL_TRIM_SHIFT			24
 #define SDHCI_CLOCK_CTRL_SDR50_TUNING_OVERRIDE		BIT(5)
 #define SDHCI_CLOCK_CTRL_PADPIPE_CLKEN_OVERRIDE		BIT(3)
 #define SDHCI_CLOCK_CTRL_SPI_MODE_CLKEN_OVERRIDE	BIT(2)
 
-#define SDHCI_TEGRA_VENDOR_MISC_CTRL		0x120
-#define SDHCI_MISC_CTRL_ENABLE_SDR104		0x8
-#define SDHCI_MISC_CTRL_ENABLE_SDR50		0x10
-#define SDHCI_MISC_CTRL_ENABLE_SDHCI_SPEC_300	0x20
-#define SDHCI_MISC_CTRL_ENABLE_DDR50		0x200
+#define SDHCI_TEGRA_VENDOR_SYS_SW_CTRL			0x104
+#define SDHCI_TEGRA_SYS_SW_CTRL_ENHANCED_STROBE		BIT(31)
 
-#define SDHCI_TEGRA_AUTO_CAL_CONFIG		0x1e4
-#define SDHCI_AUTO_CAL_START			BIT(31)
-#define SDHCI_AUTO_CAL_ENABLE			BIT(29)
+#define SDHCI_TEGRA_VENDOR_CAP_OVERRIDES		0x10c
+#define SDHCI_TEGRA_CAP_OVERRIDES_DQS_TRIM_MASK		0x00003f00
+#define SDHCI_TEGRA_CAP_OVERRIDES_DQS_TRIM_SHIFT	8
 
-#define NVQUIRK_FORCE_SDHCI_SPEC_200	BIT(0)
-#define NVQUIRK_ENABLE_BLOCK_GAP_DET	BIT(1)
-#define NVQUIRK_ENABLE_SDHCI_SPEC_300	BIT(2)
-#define NVQUIRK_ENABLE_SDR50		BIT(3)
-#define NVQUIRK_ENABLE_SDR104		BIT(4)
-#define NVQUIRK_ENABLE_DDR50		BIT(5)
-#define NVQUIRK_HAS_PADCALIB		BIT(6)
+#define SDHCI_TEGRA_VENDOR_MISC_CTRL			0x120
+#define SDHCI_MISC_CTRL_ENABLE_SDR104			0x8
+#define SDHCI_MISC_CTRL_ENABLE_SDR50			0x10
+#define SDHCI_MISC_CTRL_ENABLE_SDHCI_SPEC_300		0x20
+#define SDHCI_MISC_CTRL_ENABLE_DDR50			0x200
+
+#define SDHCI_TEGRA_VENDOR_DLLCAL_CFG			0x1b0
+#define SDHCI_TEGRA_DLLCAL_CALIBRATE			BIT(31)
+
+#define SDHCI_TEGRA_VENDOR_DLLCAL_STA			0x1bc
+#define SDHCI_TEGRA_DLLCAL_STA_ACTIVE			BIT(31)
+
+#define SDHCI_VNDR_TUN_CTRL0_0				0x1c0
+#define SDHCI_VNDR_TUN_CTRL0_TUN_HW_TAP			0x20000
+
+#define SDHCI_TEGRA_AUTO_CAL_CONFIG			0x1e4
+#define SDHCI_AUTO_CAL_START				BIT(31)
+#define SDHCI_AUTO_CAL_ENABLE				BIT(29)
+#define SDHCI_AUTO_CAL_PDPU_OFFSET_MASK			0x0000ffff
+
+#define SDHCI_TEGRA_SDMEM_COMP_PADCTRL			0x1e0
+#define SDHCI_TEGRA_SDMEM_COMP_PADCTRL_VREF_SEL_MASK	0x0000000f
+#define SDHCI_TEGRA_SDMEM_COMP_PADCTRL_VREF_SEL_VAL	0x7
+#define SDHCI_TEGRA_SDMEM_COMP_PADCTRL_E_INPUT_E_PWRD	BIT(31)
+
+#define SDHCI_TEGRA_AUTO_CAL_STATUS			0x1ec
+#define SDHCI_TEGRA_AUTO_CAL_ACTIVE			BIT(31)
+
+#define NVQUIRK_FORCE_SDHCI_SPEC_200			BIT(0)
+#define NVQUIRK_ENABLE_BLOCK_GAP_DET			BIT(1)
+#define NVQUIRK_ENABLE_SDHCI_SPEC_300			BIT(2)
+#define NVQUIRK_ENABLE_SDR50				BIT(3)
+#define NVQUIRK_ENABLE_SDR104				BIT(4)
+#define NVQUIRK_ENABLE_DDR50				BIT(5)
+#define NVQUIRK_HAS_PADCALIB				BIT(6)
+#define NVQUIRK_NEEDS_PAD_CONTROL			BIT(7)
+#define NVQUIRK_DIS_CARD_CLK_CONFIG_TAP			BIT(8)
 
 struct sdhci_tegra_soc_data {
 	const struct sdhci_pltfm_data *pdata;
 	u32 nvquirks;
 };
 
+/* Magic pull up and pull down pad calibration offsets */
+struct sdhci_tegra_autocal_offsets {
+	u32 pull_up_3v3;
+	u32 pull_down_3v3;
+	u32 pull_up_3v3_timeout;
+	u32 pull_down_3v3_timeout;
+	u32 pull_up_1v8;
+	u32 pull_down_1v8;
+	u32 pull_up_1v8_timeout;
+	u32 pull_down_1v8_timeout;
+	u32 pull_up_sdr104;
+	u32 pull_down_sdr104;
+	u32 pull_up_hs400;
+	u32 pull_down_hs400;
+};
+
 struct sdhci_tegra {
 	const struct sdhci_tegra_soc_data *soc_data;
 	struct gpio_desc *power_gpio;
 	bool ddr_signaling;
 	bool pad_calib_required;
+	bool pad_control_available;
 
 	struct reset_control *rst;
+	struct pinctrl *pinctrl_sdmmc;
+	struct pinctrl_state *pinctrl_state_3v3;
+	struct pinctrl_state *pinctrl_state_1v8;
+
+	struct sdhci_tegra_autocal_offsets autocal_offsets;
+	ktime_t last_calib;
+
+	u32 default_tap;
+	u32 default_trim;
+	u32 dqs_trim;
 };
 
 static u16 tegra_sdhci_readw(struct sdhci_host *host, int reg)
@@ -133,23 +193,149 @@ static void tegra_sdhci_writel(struct sdhci_host *host, u32 val, int reg)
 	}
 }
 
+static bool tegra_sdhci_configure_card_clk(struct sdhci_host *host, bool enable)
+{
+	bool status;
+	u32 reg;
+
+	reg = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
+	status = !!(reg & SDHCI_CLOCK_CARD_EN);
+
+	if (status == enable)
+		return status;
+
+	if (enable)
+		reg |= SDHCI_CLOCK_CARD_EN;
+	else
+		reg &= ~SDHCI_CLOCK_CARD_EN;
+
+	sdhci_writew(host, reg, SDHCI_CLOCK_CONTROL);
+
+	return status;
+}
+
+static void tegra210_sdhci_writew(struct sdhci_host *host, u16 val, int reg)
+{
+	bool is_tuning_cmd = 0;
+	bool clk_enabled;
+	u8 cmd;
+
+	if (reg == SDHCI_COMMAND) {
+		cmd = SDHCI_GET_CMD(val);
+		is_tuning_cmd = cmd == MMC_SEND_TUNING_BLOCK ||
+				cmd == MMC_SEND_TUNING_BLOCK_HS200;
+	}
+
+	if (is_tuning_cmd)
+		clk_enabled = tegra_sdhci_configure_card_clk(host, 0);
+
+	writew(val, host->ioaddr + reg);
+
+	if (is_tuning_cmd) {
+		udelay(1);
+		tegra_sdhci_configure_card_clk(host, clk_enabled);
+	}
+}
+
 static unsigned int tegra_sdhci_get_ro(struct sdhci_host *host)
 {
 	return mmc_gpio_get_ro(host->mmc);
 }
 
+static bool tegra_sdhci_is_pad_and_regulator_valid(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	int has_1v8, has_3v3;
+
+	/*
+	 * The SoCs which have NVQUIRK_NEEDS_PAD_CONTROL require software pad
+	 * voltage configuration in order to perform voltage switching. This
+	 * means that valid pinctrl info is required on SDHCI instances capable
+	 * of performing voltage switching. Whether or not an SDHCI instance is
+	 * capable of voltage switching is determined based on the regulator.
+	 */
+
+	if (!(tegra_host->soc_data->nvquirks & NVQUIRK_NEEDS_PAD_CONTROL))
+		return true;
+
+	if (IS_ERR(host->mmc->supply.vqmmc))
+		return false;
+
+	has_1v8 = regulator_is_supported_voltage(host->mmc->supply.vqmmc,
+						 1700000, 1950000);
+
+	has_3v3 = regulator_is_supported_voltage(host->mmc->supply.vqmmc,
+						 2700000, 3600000);
+
+	if (has_1v8 == 1 && has_3v3 == 1)
+		return tegra_host->pad_control_available;
+
+	/* Fixed voltage, no pad control required. */
+	return true;
+}
+
+static void tegra_sdhci_set_tap(struct sdhci_host *host, unsigned int tap)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	const struct sdhci_tegra_soc_data *soc_data = tegra_host->soc_data;
+	bool card_clk_enabled = false;
+	u32 reg;
+
+	/*
+	 * Touching the tap values is a bit tricky on some SoC generations.
+	 * The quirk enables a workaround for a glitch that sometimes occurs if
+	 * the tap values are changed.
+	 */
+
+	if (soc_data->nvquirks & NVQUIRK_DIS_CARD_CLK_CONFIG_TAP)
+		card_clk_enabled = tegra_sdhci_configure_card_clk(host, false);
+
+	reg = sdhci_readl(host, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
+	reg &= ~SDHCI_CLOCK_CTRL_TAP_MASK;
+	reg |= tap << SDHCI_CLOCK_CTRL_TAP_SHIFT;
+	sdhci_writel(host, reg, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
+
+	if (soc_data->nvquirks & NVQUIRK_DIS_CARD_CLK_CONFIG_TAP &&
+	    card_clk_enabled) {
+		udelay(1);
+		sdhci_reset(host, SDHCI_RESET_CMD | SDHCI_RESET_DATA);
+		tegra_sdhci_configure_card_clk(host, card_clk_enabled);
+	}
+}
+
+static void tegra_sdhci_hs400_enhanced_strobe(struct mmc_host *mmc,
+					      struct mmc_ios *ios)
+{
+	struct sdhci_host *host = mmc_priv(mmc);
+	u32 val;
+
+	val = sdhci_readl(host, SDHCI_TEGRA_VENDOR_SYS_SW_CTRL);
+
+	if (ios->enhanced_strobe)
+		val |= SDHCI_TEGRA_SYS_SW_CTRL_ENHANCED_STROBE;
+	else
+		val &= ~SDHCI_TEGRA_SYS_SW_CTRL_ENHANCED_STROBE;
+
+	sdhci_writel(host, val, SDHCI_TEGRA_VENDOR_SYS_SW_CTRL);
+
+}
+
 static void tegra_sdhci_reset(struct sdhci_host *host, u8 mask)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
 	const struct sdhci_tegra_soc_data *soc_data = tegra_host->soc_data;
-	u32 misc_ctrl, clk_ctrl;
+	u32 misc_ctrl, clk_ctrl, pad_ctrl;
 
 	sdhci_reset(host, mask);
 
 	if (!(mask & SDHCI_RESET_ALL))
 		return;
 
+	tegra_sdhci_set_tap(host, tegra_host->default_tap);
+
 	misc_ctrl = sdhci_readl(host, SDHCI_TEGRA_VENDOR_MISC_CTRL);
 	clk_ctrl = sdhci_readl(host, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
 
@@ -158,15 +344,10 @@ static void tegra_sdhci_reset(struct sdhci_host *host, u8 mask)
 		       SDHCI_MISC_CTRL_ENABLE_DDR50 |
 		       SDHCI_MISC_CTRL_ENABLE_SDR104);
 
-	clk_ctrl &= ~SDHCI_CLOCK_CTRL_SPI_MODE_CLKEN_OVERRIDE;
+	clk_ctrl &= ~(SDHCI_CLOCK_CTRL_TRIM_MASK |
+		      SDHCI_CLOCK_CTRL_SPI_MODE_CLKEN_OVERRIDE);
 
-	/*
-	 * If the board does not define a regulator for the SDHCI
-	 * IO voltage, then don't advertise support for UHS modes
-	 * even if the device supports it because the IO voltage
-	 * cannot be configured.
-	 */
-	if (!IS_ERR(host->mmc->supply.vqmmc)) {
+	if (tegra_sdhci_is_pad_and_regulator_valid(host)) {
 		/* Erratum: Enable SDHCI spec v3.00 support */
 		if (soc_data->nvquirks & NVQUIRK_ENABLE_SDHCI_SPEC_300)
 			misc_ctrl |= SDHCI_MISC_CTRL_ENABLE_SDHCI_SPEC_300;
@@ -181,24 +362,237 @@ static void tegra_sdhci_reset(struct sdhci_host *host, u8 mask)
 			clk_ctrl |= SDHCI_CLOCK_CTRL_SDR50_TUNING_OVERRIDE;
 	}
 
+	clk_ctrl |= tegra_host->default_trim << SDHCI_CLOCK_CTRL_TRIM_SHIFT;
+
 	sdhci_writel(host, misc_ctrl, SDHCI_TEGRA_VENDOR_MISC_CTRL);
 	sdhci_writel(host, clk_ctrl, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
 
-	if (soc_data->nvquirks & NVQUIRK_HAS_PADCALIB)
+	if (soc_data->nvquirks & NVQUIRK_HAS_PADCALIB) {
+		pad_ctrl = sdhci_readl(host, SDHCI_TEGRA_SDMEM_COMP_PADCTRL);
+		pad_ctrl &= ~SDHCI_TEGRA_SDMEM_COMP_PADCTRL_VREF_SEL_MASK;
+		pad_ctrl |= SDHCI_TEGRA_SDMEM_COMP_PADCTRL_VREF_SEL_VAL;
+		sdhci_writel(host, pad_ctrl, SDHCI_TEGRA_SDMEM_COMP_PADCTRL);
+
 		tegra_host->pad_calib_required = true;
+	}
 
 	tegra_host->ddr_signaling = false;
 }
 
-static void tegra_sdhci_pad_autocalib(struct sdhci_host *host)
+static void tegra_sdhci_configure_cal_pad(struct sdhci_host *host, bool enable)
 {
 	u32 val;
 
-	mdelay(1);
+	/*
+	 * Enable or disable the additional I/O pad used by the drive strength
+	 * calibration process.
+	 */
+	val = sdhci_readl(host, SDHCI_TEGRA_SDMEM_COMP_PADCTRL);
+
+	if (enable)
+		val |= SDHCI_TEGRA_SDMEM_COMP_PADCTRL_E_INPUT_E_PWRD;
+	else
+		val &= ~SDHCI_TEGRA_SDMEM_COMP_PADCTRL_E_INPUT_E_PWRD;
+
+	sdhci_writel(host, val, SDHCI_TEGRA_SDMEM_COMP_PADCTRL);
+
+	if (enable)
+		usleep_range(1, 2);
+}
+
+static void tegra_sdhci_set_pad_autocal_offset(struct sdhci_host *host,
+					       u16 pdpu)
+{
+	u32 reg;
+
+	reg = sdhci_readl(host, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+	reg &= ~SDHCI_AUTO_CAL_PDPU_OFFSET_MASK;
+	reg |= pdpu;
+	sdhci_writel(host, reg, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+}
+
+static void tegra_sdhci_pad_autocalib(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	struct sdhci_tegra_autocal_offsets offsets =
+			tegra_host->autocal_offsets;
+	struct mmc_ios *ios = &host->mmc->ios;
+	bool card_clk_enabled;
+	u16 pdpu;
+	u32 reg;
+	int ret;
+
+	switch (ios->timing) {
+	case MMC_TIMING_UHS_SDR104:
+		pdpu = offsets.pull_down_sdr104 << 8 | offsets.pull_up_sdr104;
+		break;
+	case MMC_TIMING_MMC_HS400:
+		pdpu = offsets.pull_down_hs400 << 8 | offsets.pull_up_hs400;
+		break;
+	default:
+		if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_180)
+			pdpu = offsets.pull_down_1v8 << 8 | offsets.pull_up_1v8;
+		else
+			pdpu = offsets.pull_down_3v3 << 8 | offsets.pull_up_3v3;
+	}
+
+	tegra_sdhci_set_pad_autocal_offset(host, pdpu);
+
+	card_clk_enabled = tegra_sdhci_configure_card_clk(host, false);
+
+	tegra_sdhci_configure_cal_pad(host, true);
+
+	reg = sdhci_readl(host, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+	reg |= SDHCI_AUTO_CAL_ENABLE | SDHCI_AUTO_CAL_START;
+	sdhci_writel(host, reg, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+
+	usleep_range(1, 2);
+	/* 10 ms timeout */
+	ret = readl_poll_timeout(host->ioaddr + SDHCI_TEGRA_AUTO_CAL_STATUS,
+				 reg, !(reg & SDHCI_TEGRA_AUTO_CAL_ACTIVE),
+				 1000, 10000);
+
+	tegra_sdhci_configure_cal_pad(host, false);
+
+	tegra_sdhci_configure_card_clk(host, card_clk_enabled);
+
+	if (ret) {
+		dev_err(mmc_dev(host->mmc), "Pad autocal timed out\n");
+
+		if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_180)
+			pdpu = offsets.pull_down_1v8_timeout << 8 |
+			       offsets.pull_up_1v8_timeout;
+		else
+			pdpu = offsets.pull_down_3v3_timeout << 8 |
+			       offsets.pull_up_3v3_timeout;
+
+		/* Disable automatic calibration and use fixed offsets */
+		reg = sdhci_readl(host, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+		reg &= ~SDHCI_AUTO_CAL_ENABLE;
+		sdhci_writel(host, reg, SDHCI_TEGRA_AUTO_CAL_CONFIG);
 
-	val = sdhci_readl(host, SDHCI_TEGRA_AUTO_CAL_CONFIG);
-	val |= SDHCI_AUTO_CAL_ENABLE | SDHCI_AUTO_CAL_START;
-	sdhci_writel(host,val, SDHCI_TEGRA_AUTO_CAL_CONFIG);
+		tegra_sdhci_set_pad_autocal_offset(host, pdpu);
+	}
+}
+
+static void tegra_sdhci_parse_pad_autocal_dt(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	struct sdhci_tegra_autocal_offsets *autocal =
+			&tegra_host->autocal_offsets;
+	int err;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-3v3",
+			&autocal->pull_up_3v3);
+	if (err)
+		autocal->pull_up_3v3 = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-3v3",
+			&autocal->pull_down_3v3);
+	if (err)
+		autocal->pull_down_3v3 = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-1v8",
+			&autocal->pull_up_1v8);
+	if (err)
+		autocal->pull_up_1v8 = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-1v8",
+			&autocal->pull_down_1v8);
+	if (err)
+		autocal->pull_down_1v8 = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-3v3-timeout",
+			&autocal->pull_up_3v3);
+	if (err)
+		autocal->pull_up_3v3_timeout = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-3v3-timeout",
+			&autocal->pull_down_3v3);
+	if (err)
+		autocal->pull_down_3v3_timeout = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-1v8-timeout",
+			&autocal->pull_up_1v8);
+	if (err)
+		autocal->pull_up_1v8_timeout = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-1v8-timeout",
+			&autocal->pull_down_1v8);
+	if (err)
+		autocal->pull_down_1v8_timeout = 0;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-sdr104",
+			&autocal->pull_up_sdr104);
+	if (err)
+		autocal->pull_up_sdr104 = autocal->pull_up_1v8;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-sdr104",
+			&autocal->pull_down_sdr104);
+	if (err)
+		autocal->pull_down_sdr104 = autocal->pull_down_1v8;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-up-offset-hs400",
+			&autocal->pull_up_hs400);
+	if (err)
+		autocal->pull_up_hs400 = autocal->pull_up_1v8;
+
+	err = device_property_read_u32(host->mmc->parent,
+			"nvidia,pad-autocal-pull-down-offset-hs400",
+			&autocal->pull_down_hs400);
+	if (err)
+		autocal->pull_down_hs400 = autocal->pull_down_1v8;
+}
+
+static void tegra_sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq)
+{
+	struct sdhci_host *host = mmc_priv(mmc);
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	ktime_t since_calib = ktime_sub(ktime_get(), tegra_host->last_calib);
+
+	/* 100 ms calibration interval is specified in the TRM */
+	if (ktime_to_ms(since_calib) > 100) {
+		tegra_sdhci_pad_autocalib(host);
+		tegra_host->last_calib = ktime_get();
+	}
+
+	sdhci_request(mmc, mrq);
+}
+
+static void tegra_sdhci_parse_tap_and_trim(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	int err;
+
+	err = device_property_read_u32(host->mmc->parent, "nvidia,default-tap",
+				       &tegra_host->default_tap);
+	if (err)
+		tegra_host->default_tap = 0;
+
+	err = device_property_read_u32(host->mmc->parent, "nvidia,default-trim",
+				       &tegra_host->default_trim);
+	if (err)
+		tegra_host->default_trim = 0;
+
+	err = device_property_read_u32(host->mmc->parent, "nvidia,dqs-trim",
+				       &tegra_host->dqs_trim);
+	if (err)
+		tegra_host->dqs_trim = 0x11;
 }
 
 static void tegra_sdhci_set_clock(struct sdhci_host *host, unsigned int clock)
@@ -237,34 +631,82 @@ static void tegra_sdhci_set_clock(struct sdhci_host *host, unsigned int clock)
 	}
 }
 
-static void tegra_sdhci_set_uhs_signaling(struct sdhci_host *host,
-					  unsigned timing)
+static unsigned int tegra_sdhci_get_max_clock(struct sdhci_host *host)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
-	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
 
-	if (timing == MMC_TIMING_UHS_DDR50 ||
-	    timing == MMC_TIMING_MMC_DDR52)
-		tegra_host->ddr_signaling = true;
-
-	sdhci_set_uhs_signaling(host, timing);
+	return clk_round_rate(pltfm_host->clk, UINT_MAX);
 }
 
-static unsigned int tegra_sdhci_get_max_clock(struct sdhci_host *host)
+static void tegra_sdhci_set_dqs_trim(struct sdhci_host *host, u8 trim)
 {
-	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	u32 val;
 
-	return clk_round_rate(pltfm_host->clk, UINT_MAX);
+	val = sdhci_readl(host, SDHCI_TEGRA_VENDOR_CAP_OVERRIDES);
+	val &= ~SDHCI_TEGRA_CAP_OVERRIDES_DQS_TRIM_MASK;
+	val |= trim << SDHCI_TEGRA_CAP_OVERRIDES_DQS_TRIM_SHIFT;
+	sdhci_writel(host, val, SDHCI_TEGRA_VENDOR_CAP_OVERRIDES);
 }
 
-static void tegra_sdhci_set_tap(struct sdhci_host *host, unsigned int tap)
+static void tegra_sdhci_hs400_dll_cal(struct sdhci_host *host)
 {
 	u32 reg;
+	int err;
+
+	reg = sdhci_readl(host, SDHCI_TEGRA_VENDOR_DLLCAL_CFG);
+	reg |= SDHCI_TEGRA_DLLCAL_CALIBRATE;
+	sdhci_writel(host, reg, SDHCI_TEGRA_VENDOR_DLLCAL_CFG);
+
+	/* 1 ms sleep, 5 ms timeout */
+	err = readl_poll_timeout(host->ioaddr + SDHCI_TEGRA_VENDOR_DLLCAL_STA,
+				 reg, !(reg & SDHCI_TEGRA_DLLCAL_STA_ACTIVE),
+				 1000, 5000);
+	if (err)
+		dev_err(mmc_dev(host->mmc),
+			"HS400 delay line calibration timed out\n");
+}
 
-	reg = sdhci_readl(host, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
-	reg &= ~SDHCI_CLOCK_CTRL_TAP_MASK;
-	reg |= tap << SDHCI_CLOCK_CTRL_TAP_SHIFT;
-	sdhci_writel(host, reg, SDHCI_TEGRA_VENDOR_CLOCK_CTRL);
+static void tegra_sdhci_set_uhs_signaling(struct sdhci_host *host,
+					  unsigned timing)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	bool set_default_tap = false;
+	bool set_dqs_trim = false;
+	bool do_hs400_dll_cal = false;
+
+	switch (timing) {
+	case MMC_TIMING_UHS_SDR50:
+	case MMC_TIMING_UHS_SDR104:
+	case MMC_TIMING_MMC_HS200:
+		/* Don't set default tap on tunable modes. */
+		break;
+	case MMC_TIMING_MMC_HS400:
+		set_dqs_trim = true;
+		do_hs400_dll_cal = true;
+		break;
+	case MMC_TIMING_MMC_DDR52:
+	case MMC_TIMING_UHS_DDR50:
+		tegra_host->ddr_signaling = true;
+		set_default_tap = true;
+		break;
+	default:
+		set_default_tap = true;
+		break;
+	}
+
+	sdhci_set_uhs_signaling(host, timing);
+
+	tegra_sdhci_pad_autocalib(host);
+
+	if (set_default_tap)
+		tegra_sdhci_set_tap(host, tegra_host->default_tap);
+
+	if (set_dqs_trim)
+		tegra_sdhci_set_dqs_trim(host, tegra_host->dqs_trim);
+
+	if (do_hs400_dll_cal)
+		tegra_sdhci_hs400_dll_cal(host);
 }
 
 static int tegra_sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
@@ -301,6 +743,89 @@ static int tegra_sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
 	return mmc_send_tuning(host->mmc, opcode, NULL);
 }
 
+static int tegra_sdhci_set_padctrl(struct sdhci_host *host, int voltage)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	int ret;
+
+	if (!tegra_host->pad_control_available)
+		return 0;
+
+	if (voltage == MMC_SIGNAL_VOLTAGE_180) {
+		ret = pinctrl_select_state(tegra_host->pinctrl_sdmmc,
+					   tegra_host->pinctrl_state_1v8);
+		if (ret < 0)
+			dev_err(mmc_dev(host->mmc),
+				"setting 1.8V failed, ret: %d\n", ret);
+	} else {
+		ret = pinctrl_select_state(tegra_host->pinctrl_sdmmc,
+					   tegra_host->pinctrl_state_3v3);
+		if (ret < 0)
+			dev_err(mmc_dev(host->mmc),
+				"setting 3.3V failed, ret: %d\n", ret);
+	}
+
+	return ret;
+}
+
+static int sdhci_tegra_start_signal_voltage_switch(struct mmc_host *mmc,
+						   struct mmc_ios *ios)
+{
+	struct sdhci_host *host = mmc_priv(mmc);
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct sdhci_tegra *tegra_host = sdhci_pltfm_priv(pltfm_host);
+	int ret = 0;
+
+	if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
+		ret = tegra_sdhci_set_padctrl(host, ios->signal_voltage);
+		if (ret < 0)
+			return ret;
+		ret = sdhci_start_signal_voltage_switch(mmc, ios);
+	} else if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_180) {
+		ret = sdhci_start_signal_voltage_switch(mmc, ios);
+		if (ret < 0)
+			return ret;
+		ret = tegra_sdhci_set_padctrl(host, ios->signal_voltage);
+	}
+
+	if (tegra_host->pad_calib_required)
+		tegra_sdhci_pad_autocalib(host);
+
+	return ret;
+}
+
+static int tegra_sdhci_init_pinctrl_info(struct device *dev,
+					 struct sdhci_tegra *tegra_host)
+{
+	tegra_host->pinctrl_sdmmc = devm_pinctrl_get(dev);
+	if (IS_ERR(tegra_host->pinctrl_sdmmc)) {
+		dev_dbg(dev, "No pinctrl info, err: %ld\n",
+			PTR_ERR(tegra_host->pinctrl_sdmmc));
+		return -1;
+	}
+
+	tegra_host->pinctrl_state_3v3 =
+		pinctrl_lookup_state(tegra_host->pinctrl_sdmmc, "sdmmc-3v3");
+	if (IS_ERR(tegra_host->pinctrl_state_3v3)) {
+		dev_warn(dev, "Missing 3.3V pad state, err: %ld\n",
+			 PTR_ERR(tegra_host->pinctrl_state_3v3));
+		return -1;
+	}
+
+	tegra_host->pinctrl_state_1v8 =
+		pinctrl_lookup_state(tegra_host->pinctrl_sdmmc, "sdmmc-1v8");
+	if (IS_ERR(tegra_host->pinctrl_state_1v8)) {
+		dev_warn(dev, "Missing 1.8V pad state, err: %ld\n",
+			 PTR_ERR(tegra_host->pinctrl_state_1v8));
+		return -1;
+	}
+
+	tegra_host->pad_control_available = true;
+
+	return 0;
+}
+
 static void tegra_sdhci_voltage_switch(struct sdhci_host *host)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
@@ -421,6 +946,19 @@ static const struct sdhci_tegra_soc_data soc_data_tegra124 = {
 	.pdata = &sdhci_tegra124_pdata,
 };
 
+static const struct sdhci_ops tegra210_sdhci_ops = {
+	.get_ro     = tegra_sdhci_get_ro,
+	.read_w     = tegra_sdhci_readw,
+	.write_w    = tegra210_sdhci_writew,
+	.write_l    = tegra_sdhci_writel,
+	.set_clock  = tegra_sdhci_set_clock,
+	.set_bus_width = sdhci_set_bus_width,
+	.reset      = tegra_sdhci_reset,
+	.set_uhs_signaling = tegra_sdhci_set_uhs_signaling,
+	.voltage_switch = tegra_sdhci_voltage_switch,
+	.get_max_clock = tegra_sdhci_get_max_clock,
+};
+
 static const struct sdhci_pltfm_data sdhci_tegra210_pdata = {
 	.quirks = SDHCI_QUIRK_BROKEN_TIMEOUT_VAL |
 		  SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK |
@@ -429,11 +967,28 @@ static const struct sdhci_pltfm_data sdhci_tegra210_pdata = {
 		  SDHCI_QUIRK_BROKEN_ADMA_ZEROLEN_DESC |
 		  SDHCI_QUIRK_CAP_CLOCK_BASE_BROKEN,
 	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN,
-	.ops  = &tegra114_sdhci_ops,
+	.ops  = &tegra210_sdhci_ops,
 };
 
 static const struct sdhci_tegra_soc_data soc_data_tegra210 = {
 	.pdata = &sdhci_tegra210_pdata,
+	.nvquirks = NVQUIRK_NEEDS_PAD_CONTROL |
+		    NVQUIRK_HAS_PADCALIB |
+		    NVQUIRK_DIS_CARD_CLK_CONFIG_TAP |
+		    NVQUIRK_ENABLE_SDR50 |
+		    NVQUIRK_ENABLE_SDR104,
+};
+
+static const struct sdhci_ops tegra186_sdhci_ops = {
+	.get_ro     = tegra_sdhci_get_ro,
+	.read_w     = tegra_sdhci_readw,
+	.write_l    = tegra_sdhci_writel,
+	.set_clock  = tegra_sdhci_set_clock,
+	.set_bus_width = sdhci_set_bus_width,
+	.reset      = tegra_sdhci_reset,
+	.set_uhs_signaling = tegra_sdhci_set_uhs_signaling,
+	.voltage_switch = tegra_sdhci_voltage_switch,
+	.get_max_clock = tegra_sdhci_get_max_clock,
 };
 
 static const struct sdhci_pltfm_data sdhci_tegra186_pdata = {
@@ -452,11 +1007,16 @@ static const struct sdhci_pltfm_data sdhci_tegra186_pdata = {
 		    * But it is not supported as of now.
 		    */
 		   SDHCI_QUIRK2_BROKEN_64_BIT_DMA,
-	.ops  = &tegra114_sdhci_ops,
+	.ops  = &tegra186_sdhci_ops,
 };
 
 static const struct sdhci_tegra_soc_data soc_data_tegra186 = {
 	.pdata = &sdhci_tegra186_pdata,
+	.nvquirks = NVQUIRK_NEEDS_PAD_CONTROL |
+		    NVQUIRK_HAS_PADCALIB |
+		    NVQUIRK_DIS_CARD_CLK_CONFIG_TAP |
+		    NVQUIRK_ENABLE_SDR50 |
+		    NVQUIRK_ENABLE_SDR104,
 };
 
 static const struct of_device_id sdhci_tegra_dt_match[] = {
@@ -493,8 +1053,23 @@ static int sdhci_tegra_probe(struct platform_device *pdev)
 	tegra_host = sdhci_pltfm_priv(pltfm_host);
 	tegra_host->ddr_signaling = false;
 	tegra_host->pad_calib_required = false;
+	tegra_host->pad_control_available = false;
 	tegra_host->soc_data = soc_data;
 
+	if (soc_data->nvquirks & NVQUIRK_NEEDS_PAD_CONTROL) {
+		rc = tegra_sdhci_init_pinctrl_info(&pdev->dev, tegra_host);
+		if (rc == 0)
+			host->mmc_host_ops.start_signal_voltage_switch =
+				sdhci_tegra_start_signal_voltage_switch;
+	}
+
+	/* Hook to periodically rerun pad calibration */
+	if (soc_data->nvquirks & NVQUIRK_HAS_PADCALIB)
+		host->mmc_host_ops.request = tegra_sdhci_request;
+
+	host->mmc_host_ops.hs400_enhanced_strobe =
+			tegra_sdhci_hs400_enhanced_strobe;
+
 	rc = mmc_of_parse(host->mmc);
 	if (rc)
 		goto err_parse_dt;
@@ -502,6 +1077,10 @@ static int sdhci_tegra_probe(struct platform_device *pdev)
 	if (tegra_host->soc_data->nvquirks & NVQUIRK_ENABLE_DDR50)
 		host->mmc->caps |= MMC_CAP_1_8V_DDR;
 
+	tegra_sdhci_parse_pad_autocal_dt(host);
+
+	tegra_sdhci_parse_tap_and_trim(host);
+
 	tegra_host->power_gpio = devm_gpiod_get_optional(&pdev->dev, "power",
 							 GPIOD_OUT_HIGH);
 	if (IS_ERR(tegra_host->power_gpio)) {
diff --git a/drivers/mmc/host/sdhci-xenon-phy.c b/drivers/mmc/host/sdhci-xenon-phy.c
index c335052d0c02..5956e90380e8 100644
--- a/drivers/mmc/host/sdhci-xenon-phy.c
+++ b/drivers/mmc/host/sdhci-xenon-phy.c
@@ -660,8 +660,8 @@ static int get_dt_pad_ctrl_data(struct sdhci_host *host,
 		return 0;
 
 	if (of_address_to_resource(np, 1, &iomem)) {
-		dev_err(mmc_dev(host->mmc), "Unable to find SoC PAD ctrl register address for %s\n",
-			np->name);
+		dev_err(mmc_dev(host->mmc), "Unable to find SoC PAD ctrl register address for %pOFn\n",
+			np);
 		return -EINVAL;
 	}
 
diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index 1b3fbd9bd5c5..99bdae53fa2e 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -123,6 +123,29 @@ EXPORT_SYMBOL_GPL(sdhci_dumpregs);
  *                                                                           *
 \*****************************************************************************/
 
+static void sdhci_do_enable_v4_mode(struct sdhci_host *host)
+{
+	u16 ctrl2;
+
+	ctrl2 = sdhci_readb(host, SDHCI_HOST_CONTROL2);
+	if (ctrl2 & SDHCI_CTRL_V4_MODE)
+		return;
+
+	ctrl2 |= SDHCI_CTRL_V4_MODE;
+	sdhci_writeb(host, ctrl2, SDHCI_HOST_CONTROL);
+}
+
+/*
+ * This can be called before sdhci_add_host() by Vendor's host controller
+ * driver to enable v4 mode if supported.
+ */
+void sdhci_enable_v4_mode(struct sdhci_host *host)
+{
+	host->v4_mode = true;
+	sdhci_do_enable_v4_mode(host);
+}
+EXPORT_SYMBOL_GPL(sdhci_enable_v4_mode);
+
 static inline bool sdhci_data_line_cmd(struct mmc_command *cmd)
 {
 	return cmd->data || cmd->flags & MMC_RSP_BUSY;
@@ -243,6 +266,52 @@ static void sdhci_set_default_irqs(struct sdhci_host *host)
 	sdhci_writel(host, host->ier, SDHCI_SIGNAL_ENABLE);
 }
 
+static void sdhci_config_dma(struct sdhci_host *host)
+{
+	u8 ctrl;
+	u16 ctrl2;
+
+	if (host->version < SDHCI_SPEC_200)
+		return;
+
+	ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
+
+	/*
+	 * Always adjust the DMA selection as some controllers
+	 * (e.g. JMicron) can't do PIO properly when the selection
+	 * is ADMA.
+	 */
+	ctrl &= ~SDHCI_CTRL_DMA_MASK;
+	if (!(host->flags & SDHCI_REQ_USE_DMA))
+		goto out;
+
+	/* Note if DMA Select is zero then SDMA is selected */
+	if (host->flags & SDHCI_USE_ADMA)
+		ctrl |= SDHCI_CTRL_ADMA32;
+
+	if (host->flags & SDHCI_USE_64_BIT_DMA) {
+		/*
+		 * If v4 mode, all supported DMA can be 64-bit addressing if
+		 * controller supports 64-bit system address, otherwise only
+		 * ADMA can support 64-bit addressing.
+		 */
+		if (host->v4_mode) {
+			ctrl2 = sdhci_readw(host, SDHCI_HOST_CONTROL2);
+			ctrl2 |= SDHCI_CTRL_64BIT_ADDR;
+			sdhci_writew(host, ctrl2, SDHCI_HOST_CONTROL2);
+		} else if (host->flags & SDHCI_USE_ADMA) {
+			/*
+			 * Don't need to undo SDHCI_CTRL_ADMA32 in order to
+			 * set SDHCI_CTRL_ADMA64.
+			 */
+			ctrl |= SDHCI_CTRL_ADMA64;
+		}
+	}
+
+out:
+	sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
+}
+
 static void sdhci_init(struct sdhci_host *host, int soft)
 {
 	struct mmc_host *mmc = host->mmc;
@@ -252,6 +321,9 @@ static void sdhci_init(struct sdhci_host *host, int soft)
 	else
 		sdhci_do_reset(host, SDHCI_RESET_ALL);
 
+	if (host->v4_mode)
+		sdhci_do_enable_v4_mode(host);
+
 	sdhci_set_default_irqs(host);
 
 	host->cqe_on = false;
@@ -554,10 +626,10 @@ static void sdhci_kunmap_atomic(void *buffer, unsigned long *flags)
 	local_irq_restore(*flags);
 }
 
-static void sdhci_adma_write_desc(struct sdhci_host *host, void *desc,
-				  dma_addr_t addr, int len, unsigned cmd)
+void sdhci_adma_write_desc(struct sdhci_host *host, void **desc,
+			   dma_addr_t addr, int len, unsigned int cmd)
 {
-	struct sdhci_adma2_64_desc *dma_desc = desc;
+	struct sdhci_adma2_64_desc *dma_desc = *desc;
 
 	/* 32-bit and 64-bit descriptors have these members in same position */
 	dma_desc->cmd = cpu_to_le16(cmd);
@@ -566,6 +638,19 @@ static void sdhci_adma_write_desc(struct sdhci_host *host, void *desc,
 
 	if (host->flags & SDHCI_USE_64_BIT_DMA)
 		dma_desc->addr_hi = cpu_to_le32((u64)addr >> 32);
+
+	*desc += host->desc_sz;
+}
+EXPORT_SYMBOL_GPL(sdhci_adma_write_desc);
+
+static inline void __sdhci_adma_write_desc(struct sdhci_host *host,
+					   void **desc, dma_addr_t addr,
+					   int len, unsigned int cmd)
+{
+	if (host->ops->adma_write_desc)
+		host->ops->adma_write_desc(host, desc, addr, len, cmd);
+	else
+		sdhci_adma_write_desc(host, desc, addr, len, cmd);
 }
 
 static void sdhci_adma_mark_end(void *desc)
@@ -618,28 +703,24 @@ static void sdhci_adma_table_pre(struct sdhci_host *host,
 			}
 
 			/* tran, valid */
-			sdhci_adma_write_desc(host, desc, align_addr, offset,
-					      ADMA2_TRAN_VALID);
+			__sdhci_adma_write_desc(host, &desc, align_addr,
+						offset, ADMA2_TRAN_VALID);
 
 			BUG_ON(offset > 65536);
 
 			align += SDHCI_ADMA2_ALIGN;
 			align_addr += SDHCI_ADMA2_ALIGN;
 
-			desc += host->desc_sz;
-
 			addr += offset;
 			len -= offset;
 		}
 
 		BUG_ON(len > 65536);
 
-		if (len) {
-			/* tran, valid */
-			sdhci_adma_write_desc(host, desc, addr, len,
-					      ADMA2_TRAN_VALID);
-			desc += host->desc_sz;
-		}
+		/* tran, valid */
+		if (len)
+			__sdhci_adma_write_desc(host, &desc, addr, len,
+						ADMA2_TRAN_VALID);
 
 		/*
 		 * If this triggers then we have a calculation bug
@@ -656,7 +737,7 @@ static void sdhci_adma_table_pre(struct sdhci_host *host,
 		}
 	} else {
 		/* Add a terminating entry - nop, end, valid */
-		sdhci_adma_write_desc(host, desc, 0, 0, ADMA2_NOP_END_VALID);
+		__sdhci_adma_write_desc(host, &desc, 0, 0, ADMA2_NOP_END_VALID);
 	}
 }
 
@@ -701,7 +782,7 @@ static void sdhci_adma_table_post(struct sdhci_host *host,
 	}
 }
 
-static u32 sdhci_sdma_address(struct sdhci_host *host)
+static dma_addr_t sdhci_sdma_address(struct sdhci_host *host)
 {
 	if (host->bounce_buffer)
 		return host->bounce_addr;
@@ -709,6 +790,17 @@ static u32 sdhci_sdma_address(struct sdhci_host *host)
 		return sg_dma_address(host->data->sg);
 }
 
+static void sdhci_set_sdma_addr(struct sdhci_host *host, dma_addr_t addr)
+{
+	if (host->v4_mode) {
+		sdhci_writel(host, addr, SDHCI_ADMA_ADDRESS);
+		if (host->flags & SDHCI_USE_64_BIT_DMA)
+			sdhci_writel(host, (u64)addr >> 32, SDHCI_ADMA_ADDRESS_HI);
+	} else {
+		sdhci_writel(host, addr, SDHCI_DMA_ADDRESS);
+	}
+}
+
 static unsigned int sdhci_target_timeout(struct sdhci_host *host,
 					 struct mmc_command *cmd,
 					 struct mmc_data *data)
@@ -876,7 +968,6 @@ static void sdhci_set_timeout(struct sdhci_host *host, struct mmc_command *cmd)
 
 static void sdhci_prepare_data(struct sdhci_host *host, struct mmc_command *cmd)
 {
-	u8 ctrl;
 	struct mmc_data *data = cmd->data;
 
 	host->data_timeout = 0;
@@ -968,30 +1059,11 @@ static void sdhci_prepare_data(struct sdhci_host *host, struct mmc_command *cmd)
 					     SDHCI_ADMA_ADDRESS_HI);
 		} else {
 			WARN_ON(sg_cnt != 1);
-			sdhci_writel(host, sdhci_sdma_address(host),
-				     SDHCI_DMA_ADDRESS);
+			sdhci_set_sdma_addr(host, sdhci_sdma_address(host));
 		}
 	}
 
-	/*
-	 * Always adjust the DMA selection as some controllers
-	 * (e.g. JMicron) can't do PIO properly when the selection
-	 * is ADMA.
-	 */
-	if (host->version >= SDHCI_SPEC_200) {
-		ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
-		ctrl &= ~SDHCI_CTRL_DMA_MASK;
-		if ((host->flags & SDHCI_REQ_USE_DMA) &&
-			(host->flags & SDHCI_USE_ADMA)) {
-			if (host->flags & SDHCI_USE_64_BIT_DMA)
-				ctrl |= SDHCI_CTRL_ADMA64;
-			else
-				ctrl |= SDHCI_CTRL_ADMA32;
-		} else {
-			ctrl |= SDHCI_CTRL_SDMA;
-		}
-		sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
-	}
+	sdhci_config_dma(host);
 
 	if (!(host->flags & SDHCI_REQ_USE_DMA)) {
 		int flags;
@@ -1010,7 +1082,19 @@ static void sdhci_prepare_data(struct sdhci_host *host, struct mmc_command *cmd)
 	/* Set the DMA boundary value and block size */
 	sdhci_writew(host, SDHCI_MAKE_BLKSZ(host->sdma_boundary, data->blksz),
 		     SDHCI_BLOCK_SIZE);
-	sdhci_writew(host, data->blocks, SDHCI_BLOCK_COUNT);
+
+	/*
+	 * For Version 4.10 onwards, if v4 mode is enabled, 32-bit Block Count
+	 * can be supported, in that case 16-bit block count register must be 0.
+	 */
+	if (host->version >= SDHCI_SPEC_410 && host->v4_mode &&
+	    (host->quirks2 & SDHCI_QUIRK2_USE_32BIT_BLK_CNT)) {
+		if (sdhci_readw(host, SDHCI_BLOCK_COUNT))
+			sdhci_writew(host, 0, SDHCI_BLOCK_COUNT);
+		sdhci_writew(host, data->blocks, SDHCI_32BIT_BLK_CNT);
+	} else {
+		sdhci_writew(host, data->blocks, SDHCI_BLOCK_COUNT);
+	}
 }
 
 static inline bool sdhci_auto_cmd12(struct sdhci_host *host,
@@ -1020,6 +1104,43 @@ static inline bool sdhci_auto_cmd12(struct sdhci_host *host,
 	       !mrq->cap_cmd_during_tfr;
 }
 
+static inline void sdhci_auto_cmd_select(struct sdhci_host *host,
+					 struct mmc_command *cmd,
+					 u16 *mode)
+{
+	bool use_cmd12 = sdhci_auto_cmd12(host, cmd->mrq) &&
+			 (cmd->opcode != SD_IO_RW_EXTENDED);
+	bool use_cmd23 = cmd->mrq->sbc && (host->flags & SDHCI_AUTO_CMD23);
+	u16 ctrl2;
+
+	/*
+	 * In case of Version 4.10 or later, use of 'Auto CMD Auto
+	 * Select' is recommended rather than use of 'Auto CMD12
+	 * Enable' or 'Auto CMD23 Enable'.
+	 */
+	if (host->version >= SDHCI_SPEC_410 && (use_cmd12 || use_cmd23)) {
+		*mode |= SDHCI_TRNS_AUTO_SEL;
+
+		ctrl2 = sdhci_readw(host, SDHCI_HOST_CONTROL2);
+		if (use_cmd23)
+			ctrl2 |= SDHCI_CMD23_ENABLE;
+		else
+			ctrl2 &= ~SDHCI_CMD23_ENABLE;
+		sdhci_writew(host, ctrl2, SDHCI_HOST_CONTROL2);
+
+		return;
+	}
+
+	/*
+	 * If we are sending CMD23, CMD12 never gets sent
+	 * on successful completion (so no Auto-CMD12).
+	 */
+	if (use_cmd12)
+		*mode |= SDHCI_TRNS_AUTO_CMD12;
+	else if (use_cmd23)
+		*mode |= SDHCI_TRNS_AUTO_CMD23;
+}
+
 static void sdhci_set_transfer_mode(struct sdhci_host *host,
 	struct mmc_command *cmd)
 {
@@ -1048,17 +1169,9 @@ static void sdhci_set_transfer_mode(struct sdhci_host *host,
 
 	if (mmc_op_multi(cmd->opcode) || data->blocks > 1) {
 		mode = SDHCI_TRNS_BLK_CNT_EN | SDHCI_TRNS_MULTI;
-		/*
-		 * If we are sending CMD23, CMD12 never gets sent
-		 * on successful completion (so no Auto-CMD12).
-		 */
-		if (sdhci_auto_cmd12(host, cmd->mrq) &&
-		    (cmd->opcode != SD_IO_RW_EXTENDED))
-			mode |= SDHCI_TRNS_AUTO_CMD12;
-		else if (cmd->mrq->sbc && (host->flags & SDHCI_AUTO_CMD23)) {
-			mode |= SDHCI_TRNS_AUTO_CMD23;
+		sdhci_auto_cmd_select(host, cmd, &mode);
+		if (cmd->mrq->sbc && (host->flags & SDHCI_AUTO_CMD23))
 			sdhci_writel(host, cmd->mrq->sbc->arg, SDHCI_ARGUMENT2);
-		}
 	}
 
 	if (data->flags & MMC_DATA_READ)
@@ -1630,7 +1743,7 @@ EXPORT_SYMBOL_GPL(sdhci_set_power);
  *                                                                           *
 \*****************************************************************************/
 
-static void sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq)
+void sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 {
 	struct sdhci_host *host;
 	int present;
@@ -1669,6 +1782,7 @@ static void sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	mmiowb();
 	spin_unlock_irqrestore(&host->lock, flags);
 }
+EXPORT_SYMBOL_GPL(sdhci_request);
 
 void sdhci_set_bus_width(struct sdhci_host *host, int width)
 {
@@ -2219,7 +2333,7 @@ void sdhci_send_tuning(struct sdhci_host *host, u32 opcode)
 }
 EXPORT_SYMBOL_GPL(sdhci_send_tuning);
 
-static void __sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
+static int __sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
 {
 	int i;
 
@@ -2236,13 +2350,13 @@ static void __sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
 			pr_info("%s: Tuning timeout, falling back to fixed sampling clock\n",
 				mmc_hostname(host->mmc));
 			sdhci_abort_tuning(host, opcode);
-			return;
+			return -ETIMEDOUT;
 		}
 
 		ctrl = sdhci_readw(host, SDHCI_HOST_CONTROL2);
 		if (!(ctrl & SDHCI_CTRL_EXEC_TUNING)) {
 			if (ctrl & SDHCI_CTRL_TUNED_CLK)
-				return; /* Success! */
+				return 0; /* Success! */
 			break;
 		}
 
@@ -2254,6 +2368,7 @@ static void __sdhci_execute_tuning(struct sdhci_host *host, u32 opcode)
 	pr_info("%s: Tuning failed, falling back to fixed sampling clock\n",
 		mmc_hostname(host->mmc));
 	sdhci_reset_tuning(host);
+	return -EAGAIN;
 }
 
 int sdhci_execute_tuning(struct mmc_host *mmc, u32 opcode)
@@ -2315,7 +2430,7 @@ int sdhci_execute_tuning(struct mmc_host *mmc, u32 opcode)
 
 	sdhci_start_tuning(host);
 
-	__sdhci_execute_tuning(host, opcode);
+	host->tuning_err = __sdhci_execute_tuning(host, opcode);
 
 	sdhci_end_tuning(host);
 out:
@@ -2802,7 +2917,7 @@ static void sdhci_data_irq(struct sdhci_host *host, u32 intmask)
 		 * some controllers are faulty, don't trust them.
 		 */
 		if (intmask & SDHCI_INT_DMA_END) {
-			u32 dmastart, dmanow;
+			dma_addr_t dmastart, dmanow;
 
 			dmastart = sdhci_sdma_address(host);
 			dmanow = dmastart + host->data->bytes_xfered;
@@ -2810,12 +2925,12 @@ static void sdhci_data_irq(struct sdhci_host *host, u32 intmask)
 			 * Force update to the next DMA block boundary.
 			 */
 			dmanow = (dmanow &
-				~(SDHCI_DEFAULT_BOUNDARY_SIZE - 1)) +
+				~((dma_addr_t)SDHCI_DEFAULT_BOUNDARY_SIZE - 1)) +
 				SDHCI_DEFAULT_BOUNDARY_SIZE;
 			host->data->bytes_xfered = dmanow - dmastart;
-			DBG("DMA base 0x%08x, transferred 0x%06x bytes, next 0x%08x\n",
-			    dmastart, host->data->bytes_xfered, dmanow);
-			sdhci_writel(host, dmanow, SDHCI_DMA_ADDRESS);
+			DBG("DMA base %pad, transferred 0x%06x bytes, next %pad\n",
+			    &dmastart, host->data->bytes_xfered, &dmanow);
+			sdhci_set_sdma_addr(host, dmanow);
 		}
 
 		if (intmask & SDHCI_INT_DATA_END) {
@@ -3322,6 +3437,13 @@ struct sdhci_host *sdhci_alloc_host(struct device *dev,
 
 	host->sdma_boundary = SDHCI_DEFAULT_BOUNDARY_ARG;
 
+	/*
+	 * The DMA table descriptor count is calculated as the maximum
+	 * number of segments times 2, to allow for an alignment
+	 * descriptor for each segment, plus 1 for a nop end descriptor.
+	 */
+	host->adma_table_cnt = SDHCI_MAX_SEGS * 2 + 1;
+
 	return host;
 }
 
@@ -3376,6 +3498,9 @@ void __sdhci_read_caps(struct sdhci_host *host, u16 *ver, u32 *caps, u32 *caps1)
 
 	sdhci_do_reset(host, SDHCI_RESET_ALL);
 
+	if (host->v4_mode)
+		sdhci_do_enable_v4_mode(host);
+
 	of_property_read_u64(mmc_dev(host->mmc)->of_node,
 			     "sdhci-caps-mask", &dt_caps_mask);
 	of_property_read_u64(mmc_dev(host->mmc)->of_node,
@@ -3470,6 +3595,19 @@ static int sdhci_allocate_bounce_buffer(struct sdhci_host *host)
 	return 0;
 }
 
+static inline bool sdhci_can_64bit_dma(struct sdhci_host *host)
+{
+	/*
+	 * According to SD Host Controller spec v4.10, bit[27] added from
+	 * version 4.10 in Capabilities Register is used as 64-bit System
+	 * Address support for V4 mode.
+	 */
+	if (host->version >= SDHCI_SPEC_410 && host->v4_mode)
+		return host->caps & SDHCI_CAN_64BIT_V4;
+
+	return host->caps & SDHCI_CAN_64BIT;
+}
+
 int sdhci_setup_host(struct sdhci_host *host)
 {
 	struct mmc_host *mmc;
@@ -3506,7 +3644,7 @@ int sdhci_setup_host(struct sdhci_host *host)
 
 	override_timeout_clk = host->timeout_clk;
 
-	if (host->version > SDHCI_SPEC_300) {
+	if (host->version > SDHCI_SPEC_420) {
 		pr_err("%s: Unknown controller version (%d). You may experience problems.\n",
 		       mmc_hostname(mmc), host->version);
 	}
@@ -3541,7 +3679,7 @@ int sdhci_setup_host(struct sdhci_host *host)
 	 * SDHCI_QUIRK2_BROKEN_64_BIT_DMA must be left to the drivers to
 	 * implement.
 	 */
-	if (host->caps & SDHCI_CAN_64BIT)
+	if (sdhci_can_64bit_dma(host))
 		host->flags |= SDHCI_USE_64_BIT_DMA;
 
 	if (host->flags & (SDHCI_USE_SDMA | SDHCI_USE_ADMA)) {
@@ -3559,32 +3697,30 @@ int sdhci_setup_host(struct sdhci_host *host)
 		}
 	}
 
-	/* SDMA does not support 64-bit DMA */
-	if (host->flags & SDHCI_USE_64_BIT_DMA)
+	/* SDMA does not support 64-bit DMA if v4 mode not set */
+	if ((host->flags & SDHCI_USE_64_BIT_DMA) && !host->v4_mode)
 		host->flags &= ~SDHCI_USE_SDMA;
 
 	if (host->flags & SDHCI_USE_ADMA) {
 		dma_addr_t dma;
 		void *buf;
 
-		/*
-		 * The DMA descriptor table size is calculated as the maximum
-		 * number of segments times 2, to allow for an alignment
-		 * descriptor for each segment, plus 1 for a nop end descriptor,
-		 * all multipled by the descriptor size.
-		 */
 		if (host->flags & SDHCI_USE_64_BIT_DMA) {
-			host->adma_table_sz = (SDHCI_MAX_SEGS * 2 + 1) *
-					      SDHCI_ADMA2_64_DESC_SZ;
-			host->desc_sz = SDHCI_ADMA2_64_DESC_SZ;
+			host->adma_table_sz = host->adma_table_cnt *
+					      SDHCI_ADMA2_64_DESC_SZ(host);
+			host->desc_sz = SDHCI_ADMA2_64_DESC_SZ(host);
 		} else {
-			host->adma_table_sz = (SDHCI_MAX_SEGS * 2 + 1) *
+			host->adma_table_sz = host->adma_table_cnt *
 					      SDHCI_ADMA2_32_DESC_SZ;
 			host->desc_sz = SDHCI_ADMA2_32_DESC_SZ;
 		}
 
 		host->align_buffer_sz = SDHCI_MAX_SEGS * SDHCI_ADMA2_ALIGN;
-		buf = dma_alloc_coherent(mmc_dev(mmc), host->align_buffer_sz +
+		/*
+		 * Use zalloc to zero the reserved high 32-bits of 128-bit
+		 * descriptors so that they never need to be written.
+		 */
+		buf = dma_zalloc_coherent(mmc_dev(mmc), host->align_buffer_sz +
 					 host->adma_table_sz, &dma, GFP_KERNEL);
 		if (!buf) {
 			pr_warn("%s: Unable to allocate ADMA buffers - falling back to standard DMA\n",
@@ -3708,10 +3844,13 @@ int sdhci_setup_host(struct sdhci_host *host)
 	if (host->quirks & SDHCI_QUIRK_MULTIBLOCK_READ_ACMD12)
 		host->flags |= SDHCI_AUTO_CMD12;
 
-	/* Auto-CMD23 stuff only works in ADMA or PIO. */
+	/*
+	 * For v3 mode, Auto-CMD23 stuff only works in ADMA or PIO.
+	 * For v4 mode, SDMA may use Auto-CMD23 as well.
+	 */
 	if ((host->version >= SDHCI_SPEC_300) &&
 	    ((host->flags & SDHCI_USE_ADMA) ||
-	     !(host->flags & SDHCI_USE_SDMA)) &&
+	     !(host->flags & SDHCI_USE_SDMA) || host->v4_mode) &&
 	     !(host->quirks2 & SDHCI_QUIRK2_ACMD23_BROKEN)) {
 		host->flags |= SDHCI_AUTO_CMD23;
 		DBG("Auto-CMD23 available\n");
diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h
index f0bd36ce3817..b001cf4d3d7e 100644
--- a/drivers/mmc/host/sdhci.h
+++ b/drivers/mmc/host/sdhci.h
@@ -28,6 +28,7 @@
 
 #define SDHCI_DMA_ADDRESS	0x00
 #define SDHCI_ARGUMENT2		SDHCI_DMA_ADDRESS
+#define SDHCI_32BIT_BLK_CNT	SDHCI_DMA_ADDRESS
 
 #define SDHCI_BLOCK_SIZE	0x04
 #define  SDHCI_MAKE_BLKSZ(dma, blksz) (((dma & 0x7) << 12) | (blksz & 0xFFF))
@@ -41,6 +42,7 @@
 #define  SDHCI_TRNS_BLK_CNT_EN	0x02
 #define  SDHCI_TRNS_AUTO_CMD12	0x04
 #define  SDHCI_TRNS_AUTO_CMD23	0x08
+#define  SDHCI_TRNS_AUTO_SEL	0x0C
 #define  SDHCI_TRNS_READ	0x10
 #define  SDHCI_TRNS_MULTI	0x20
 
@@ -184,6 +186,9 @@
 #define   SDHCI_CTRL_DRV_TYPE_D		0x0030
 #define  SDHCI_CTRL_EXEC_TUNING		0x0040
 #define  SDHCI_CTRL_TUNED_CLK		0x0080
+#define  SDHCI_CMD23_ENABLE		0x0800
+#define  SDHCI_CTRL_V4_MODE		0x1000
+#define  SDHCI_CTRL_64BIT_ADDR		0x2000
 #define  SDHCI_CTRL_PRESET_VAL_ENABLE	0x8000
 
 #define SDHCI_CAPABILITIES	0x40
@@ -204,6 +209,7 @@
 #define  SDHCI_CAN_VDD_330	0x01000000
 #define  SDHCI_CAN_VDD_300	0x02000000
 #define  SDHCI_CAN_VDD_180	0x04000000
+#define  SDHCI_CAN_64BIT_V4	0x08000000
 #define  SDHCI_CAN_64BIT	0x10000000
 
 #define  SDHCI_SUPPORT_SDR50	0x00000001
@@ -270,6 +276,9 @@
 #define   SDHCI_SPEC_100	0
 #define   SDHCI_SPEC_200	1
 #define   SDHCI_SPEC_300	2
+#define   SDHCI_SPEC_400	3
+#define   SDHCI_SPEC_410	4
+#define   SDHCI_SPEC_420	5
 
 /*
  * End of controller registers.
@@ -305,8 +314,14 @@ struct sdhci_adma2_32_desc {
  */
 #define SDHCI_ADMA2_DESC_ALIGN	8
 
-/* ADMA2 64-bit DMA descriptor size */
-#define SDHCI_ADMA2_64_DESC_SZ	12
+/*
+ * ADMA2 64-bit DMA descriptor size
+ * According to SD Host Controller spec v4.10, there are two kinds of
+ * descriptors for 64-bit addressing mode: 96-bit Descriptor and 128-bit
+ * Descriptor, if Host Version 4 Enable is set in the Host Control 2
+ * register, 128-bit Descriptor will be selected.
+ */
+#define SDHCI_ADMA2_64_DESC_SZ(host)	((host)->v4_mode ? 16 : 12)
 
 /*
  * ADMA2 64-bit descriptor. Note 12-byte descriptor can't always be 8-byte
@@ -450,6 +465,13 @@ struct sdhci_host {
  * obtainable timeout.
  */
 #define SDHCI_QUIRK2_DISABLE_HW_TIMEOUT			(1<<17)
+/*
+ * 32-bit block count may not support eMMC where upper bits of CMD23 are used
+ * for other purposes.  Consequently we support 16-bit block count by default.
+ * Otherwise, SDHCI_QUIRK2_USE_32BIT_BLK_CNT can be selected to use 32-bit
+ * block count.
+ */
+#define SDHCI_QUIRK2_USE_32BIT_BLK_CNT			(1<<18)
 
 	int irq;		/* Device IRQ */
 	void __iomem *ioaddr;	/* Mapped address */
@@ -501,6 +523,7 @@ struct sdhci_host {
 	bool preset_enabled;	/* Preset is enabled */
 	bool pending_reset;	/* Cmd/data reset is pending */
 	bool irq_wake_enabled;	/* IRQ wakeup is enabled */
+	bool v4_mode;		/* Host Version 4 Enable */
 
 	struct mmc_request *mrqs_done[SDHCI_MAX_MRQS];	/* Requests done */
 	struct mmc_command *cmd;	/* Current command */
@@ -554,6 +577,7 @@ struct sdhci_host {
 
 	unsigned int		tuning_count;	/* Timer count for re-tuning */
 	unsigned int		tuning_mode;	/* Re-tuning mode supported by host */
+	unsigned int		tuning_err;	/* Error code for re-tuning */
 #define SDHCI_TUNING_MODE_1	0
 #define SDHCI_TUNING_MODE_2	1
 #define SDHCI_TUNING_MODE_3	2
@@ -563,6 +587,9 @@ struct sdhci_host {
 	/* Host SDMA buffer boundary. */
 	u32			sdma_boundary;
 
+	/* Host ADMA table count */
+	u32			adma_table_cnt;
+
 	u64			data_timeout;
 
 	unsigned long private[0] ____cacheline_aligned;
@@ -603,6 +630,8 @@ struct sdhci_ops {
 	void    (*adma_workaround)(struct sdhci_host *host, u32 intmask);
 	void    (*card_event)(struct sdhci_host *host);
 	void	(*voltage_switch)(struct sdhci_host *host);
+	void	(*adma_write_desc)(struct sdhci_host *host, void **desc,
+				   dma_addr_t addr, int len, unsigned int cmd);
 };
 
 #ifdef CONFIG_MMC_SDHCI_IO_ACCESSORS
@@ -725,6 +754,7 @@ void sdhci_set_power(struct sdhci_host *host, unsigned char mode,
 		     unsigned short vdd);
 void sdhci_set_power_noreg(struct sdhci_host *host, unsigned char mode,
 			   unsigned short vdd);
+void sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq);
 void sdhci_set_bus_width(struct sdhci_host *host, int width);
 void sdhci_reset(struct sdhci_host *host, u8 mask);
 void sdhci_set_uhs_signaling(struct sdhci_host *host, unsigned timing);
@@ -733,6 +763,8 @@ void sdhci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios);
 int sdhci_start_signal_voltage_switch(struct mmc_host *mmc,
 				      struct mmc_ios *ios);
 void sdhci_enable_sdio_irq(struct mmc_host *mmc, int enable);
+void sdhci_adma_write_desc(struct sdhci_host *host, void **desc,
+			   dma_addr_t addr, int len, unsigned int cmd);
 
 #ifdef CONFIG_PM
 int sdhci_suspend_host(struct sdhci_host *host);
@@ -747,6 +779,7 @@ bool sdhci_cqe_irq(struct sdhci_host *host, u32 intmask, int *cmd_error,
 		   int *data_error);
 
 void sdhci_dumpregs(struct sdhci_host *host);
+void sdhci_enable_v4_mode(struct sdhci_host *host);
 
 void sdhci_start_tuning(struct sdhci_host *host);
 void sdhci_end_tuning(struct sdhci_host *host);
diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c
index 4c2a1f8ddbf3..81bd9afb0980 100644
--- a/drivers/mmc/host/sh_mmcif.c
+++ b/drivers/mmc/host/sh_mmcif.c
@@ -1,12 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * MMCIF eMMC driver.
  *
  * Copyright (C) 2010 Renesas Solutions Corp.
  * Yusuke Goda <yusuke.goda.sx@renesas.com>
- *
- * 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.
  */
 
 /*
@@ -1573,6 +1570,6 @@ static struct platform_driver sh_mmcif_driver = {
 module_platform_driver(sh_mmcif_driver);
 
 MODULE_DESCRIPTION("SuperH on-chip MMC/eMMC interface driver");
-MODULE_LICENSE("GPL");
+MODULE_LICENSE("GPL v2");
 MODULE_ALIAS("platform:" DRIVER_NAME);
 MODULE_AUTHOR("Yusuke Goda <yusuke.goda.sx@renesas.com>");
diff --git a/drivers/mmc/host/sunxi-mmc.c b/drivers/mmc/host/sunxi-mmc.c
index 568349e1fbc2..279e326e397e 100644
--- a/drivers/mmc/host/sunxi-mmc.c
+++ b/drivers/mmc/host/sunxi-mmc.c
@@ -258,11 +258,16 @@ struct sunxi_mmc_cfg {
 	/* Does DATA0 needs to be masked while the clock is updated */
 	bool mask_data0;
 
-	/* hardware only supports new timing mode */
+	/*
+	 * hardware only supports new timing mode, either due to lack of
+	 * a mode switch in the clock controller, or the mmc controller
+	 * is permanently configured in the new timing mode, without the
+	 * NTSR mode switch.
+	 */
 	bool needs_new_timings;
 
-	/* hardware can switch between old and new timing modes */
-	bool has_timings_switch;
+	/* clock hardware can switch between old and new timing modes */
+	bool ccu_has_timings_switch;
 };
 
 struct sunxi_mmc_host {
@@ -787,7 +792,7 @@ static int sunxi_mmc_clk_set_rate(struct sunxi_mmc_host *host,
 		clock <<= 1;
 	}
 
-	if (host->use_new_timings && host->cfg->has_timings_switch) {
+	if (host->use_new_timings && host->cfg->ccu_has_timings_switch) {
 		ret = sunxi_ccu_set_mmc_timing_mode(host->clk_mmc, true);
 		if (ret) {
 			dev_err(mmc_dev(mmc),
@@ -822,6 +827,12 @@ static int sunxi_mmc_clk_set_rate(struct sunxi_mmc_host *host,
 	/* update card clock rate to account for internal divider */
 	rate /= div;
 
+	/*
+	 * Configure the controller to use the new timing mode if needed.
+	 * On controllers that only support the new timing mode, such as
+	 * the eMMC controller on the A64, this register does not exist,
+	 * and any writes to it are ignored.
+	 */
 	if (host->use_new_timings) {
 		/* Don't touch the delay bits */
 		rval = mmc_readl(host, REG_SD_NTSR);
@@ -1145,7 +1156,7 @@ static const struct sunxi_mmc_cfg sun8i_a83t_emmc_cfg = {
 	.idma_des_size_bits = 16,
 	.clk_delays = sunxi_mmc_clk_delays,
 	.can_calibrate = false,
-	.has_timings_switch = true,
+	.ccu_has_timings_switch = true,
 };
 
 static const struct sunxi_mmc_cfg sun9i_a80_cfg = {
@@ -1166,6 +1177,7 @@ static const struct sunxi_mmc_cfg sun50i_a64_emmc_cfg = {
 	.idma_des_size_bits = 13,
 	.clk_delays = NULL,
 	.can_calibrate = true,
+	.needs_new_timings = true,
 };
 
 static const struct of_device_id sunxi_mmc_of_match[] = {
@@ -1351,7 +1363,7 @@ static int sunxi_mmc_probe(struct platform_device *pdev)
 		goto error_free_host;
 	}
 
-	if (host->cfg->has_timings_switch) {
+	if (host->cfg->ccu_has_timings_switch) {
 		/*
 		 * Supports both old and new timing modes.
 		 * Try setting the clk to new timing mode.
diff --git a/drivers/mmc/host/tifm_sd.c b/drivers/mmc/host/tifm_sd.c
index a3d8380ab480..b6644ce296b2 100644
--- a/drivers/mmc/host/tifm_sd.c
+++ b/drivers/mmc/host/tifm_sd.c
@@ -336,7 +336,8 @@ static unsigned int tifm_sd_op_flags(struct mmc_command *cmd)
 		rc |= TIFM_MMCSD_RSP_R0;
 		break;
 	case MMC_RSP_R1B:
-		rc |= TIFM_MMCSD_RSP_BUSY; // deliberate fall-through
+		rc |= TIFM_MMCSD_RSP_BUSY;
+		/* fall-through */
 	case MMC_RSP_R1:
 		rc |= TIFM_MMCSD_RSP_R1;
 		break;
diff --git a/drivers/mmc/host/tmio_mmc.c b/drivers/mmc/host/tmio_mmc.c
index 43a2ea5cff24..93e83ad25976 100644
--- a/drivers/mmc/host/tmio_mmc.c
+++ b/drivers/mmc/host/tmio_mmc.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Driver for the MMC / SD / SDIO cell found in:
  *
@@ -7,12 +8,9 @@
  * Copyright (C) 2017 Horms Solutions, Simon Horman
  * Copyright (C) 2007 Ian Molton
  * Copyright (C) 2004 Ian Molton
- *
- * 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/delay.h>
 #include <linux/device.h>
 #include <linux/mfd/core.h>
 #include <linux/mfd/tmio.h>
@@ -23,6 +21,76 @@
 
 #include "tmio_mmc.h"
 
+/* Registers specific to this variant */
+#define CTL_SDIO_REGS		0x100
+#define CTL_CLK_AND_WAIT_CTL	0x138
+#define CTL_RESET_SDIO		0x1e0
+
+static void tmio_mmc_clk_start(struct tmio_mmc_host *host)
+{
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, CLK_CTL_SCLKEN |
+		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
+
+	usleep_range(10000, 11000);
+	sd_ctrl_write16(host, CTL_CLK_AND_WAIT_CTL, 0x0100);
+	usleep_range(10000, 11000);
+}
+
+static void tmio_mmc_clk_stop(struct tmio_mmc_host *host)
+{
+	sd_ctrl_write16(host, CTL_CLK_AND_WAIT_CTL, 0x0000);
+	usleep_range(10000, 11000);
+
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, ~CLK_CTL_SCLKEN &
+		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
+
+	usleep_range(10000, 11000);
+}
+
+static void tmio_mmc_set_clock(struct tmio_mmc_host *host,
+			       unsigned int new_clock)
+{
+	unsigned int divisor;
+	u32 clk = 0;
+	int clk_sel;
+
+	if (new_clock == 0) {
+		tmio_mmc_clk_stop(host);
+		return;
+	}
+
+	divisor = host->pdata->hclk / new_clock;
+
+	/* bit7 set: 1/512, ... bit0 set: 1/4, all bits clear: 1/2 */
+	clk_sel = (divisor <= 1);
+	clk = clk_sel ? 0 : (roundup_pow_of_two(divisor) >> 2);
+
+	host->pdata->set_clk_div(host->pdev, clk_sel);
+
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, ~CLK_CTL_SCLKEN &
+			sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
+	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, clk & CLK_CTL_DIV_MASK);
+	usleep_range(10000, 11000);
+
+	tmio_mmc_clk_start(host);
+}
+
+static void tmio_mmc_reset(struct tmio_mmc_host *host)
+{
+	/* FIXME - should we set stop clock reg here */
+	sd_ctrl_write16(host, CTL_RESET_SD, 0x0000);
+	sd_ctrl_write16(host, CTL_RESET_SDIO, 0x0000);
+	usleep_range(10000, 11000);
+	sd_ctrl_write16(host, CTL_RESET_SD, 0x0001);
+	sd_ctrl_write16(host, CTL_RESET_SDIO, 0x0001);
+	usleep_range(10000, 11000);
+
+	if (host->pdata->flags & TMIO_MMC_SDIO_IRQ) {
+		sd_ctrl_write16(host, CTL_SDIO_IRQ_MASK, host->sdio_irq_mask);
+		sd_ctrl_write16(host, CTL_TRANSACTION_CTL, 0x0001);
+	}
+}
+
 #ifdef CONFIG_PM_SLEEP
 static int tmio_mmc_suspend(struct device *dev)
 {
@@ -90,8 +158,6 @@ static int tmio_mmc_probe(struct platform_device *pdev)
 		goto cell_disable;
 	}
 
-	pdata->flags |= TMIO_MMC_HAVE_HIGH_REG;
-
 	host = tmio_mmc_host_alloc(pdev, pdata);
 	if (IS_ERR(host)) {
 		ret = PTR_ERR(host);
@@ -100,6 +166,8 @@ static int tmio_mmc_probe(struct platform_device *pdev)
 
 	/* SD control register space size is 0x200, 0x400 for bus_shift=1 */
 	host->bus_shift = resource_size(res) >> 10;
+	host->set_clock = tmio_mmc_set_clock;
+	host->reset = tmio_mmc_reset;
 
 	host->mmc->f_max = pdata->hclk;
 	host->mmc->f_min = pdata->hclk / 512;
diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h
index 5d141f79e175..1e317027bf53 100644
--- a/drivers/mmc/host/tmio_mmc.h
+++ b/drivers/mmc/host/tmio_mmc.h
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Driver for the MMC / SD / SDIO cell found in:
  *
@@ -8,11 +9,6 @@
  * Copyright (C) 2016-17 Horms Solutions, Simon Horman
  * Copyright (C) 2007 Ian Molton
  * Copyright (C) 2004 Ian Molton
- *
- * 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 TMIO_MMC_H
@@ -47,9 +43,6 @@
 #define CTL_RESET_SD 0xe0
 #define CTL_VERSION 0xe2
 #define CTL_SDIF_MODE 0xe6
-#define CTL_SDIO_REGS 0x100
-#define CTL_CLK_AND_WAIT_CTL 0x138
-#define CTL_RESET_SDIO 0x1e0
 
 /* Definitions for values the CTL_STOP_INTERNAL_ACTION register can take */
 #define TMIO_STOP_STP		BIT(0)
@@ -133,7 +126,6 @@ struct tmio_mmc_host {
 
 	/* Callbacks for clock / power control */
 	void (*set_pwr)(struct platform_device *host, int state);
-	void (*set_clk_div)(struct platform_device *host, int state);
 
 	/* pio related stuff */
 	struct scatterlist      *sg_ptr;
@@ -146,7 +138,7 @@ struct tmio_mmc_host {
 	struct tmio_mmc_data *pdata;
 
 	/* DMA support */
-	bool			force_pio;
+	bool			dma_on;
 	struct dma_chan		*chan_rx;
 	struct dma_chan		*chan_tx;
 	struct tasklet_struct	dma_issue;
@@ -170,14 +162,14 @@ struct tmio_mmc_host {
 
 	/* Mandatory callback */
 	int (*clk_enable)(struct tmio_mmc_host *host);
+	void (*set_clock)(struct tmio_mmc_host *host, unsigned int clock);
 
 	/* Optional callbacks */
-	unsigned int (*clk_update)(struct tmio_mmc_host *host,
-				   unsigned int new_clock);
 	void (*clk_disable)(struct tmio_mmc_host *host);
 	int (*multi_io_quirk)(struct mmc_card *card,
 			      unsigned int direction, int blk_size);
 	int (*write16_hook)(struct tmio_mmc_host *host, int addr);
+	void (*reset)(struct tmio_mmc_host *host);
 	void (*hw_reset)(struct tmio_mmc_host *host);
 	void (*prepare_tuning)(struct tmio_mmc_host *host, unsigned long tap);
 	bool (*check_scc_error)(struct tmio_mmc_host *host);
diff --git a/drivers/mmc/host/tmio_mmc_core.c b/drivers/mmc/host/tmio_mmc_core.c
index 261b4d62d2b1..8d64f6196f33 100644
--- a/drivers/mmc/host/tmio_mmc_core.c
+++ b/drivers/mmc/host/tmio_mmc_core.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Driver for the MMC / SD / SDIO IP found in:
  *
@@ -10,10 +11,6 @@
  * Copyright (C) 2007 Ian Molton
  * Copyright (C) 2004 Ian Molton
  *
- * 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 driver draws mainly on scattered spec sheets, Reverse engineering
  * of the toshiba e800  SD driver and some parts of the 2.4 ASIC3 driver (4 bit
  * support). (Further 4 bit support from a later datasheet).
@@ -160,100 +157,18 @@ static void tmio_mmc_enable_sdio_irq(struct mmc_host *mmc, int enable)
 	}
 }
 
-static void tmio_mmc_clk_start(struct tmio_mmc_host *host)
-{
-	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, CLK_CTL_SCLKEN |
-		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
-
-	/* HW engineers overrode docs: no sleep needed on R-Car2+ */
-	if (!(host->pdata->flags & TMIO_MMC_MIN_RCAR2))
-		usleep_range(10000, 11000);
-
-	if (host->pdata->flags & TMIO_MMC_HAVE_HIGH_REG) {
-		sd_ctrl_write16(host, CTL_CLK_AND_WAIT_CTL, 0x0100);
-		usleep_range(10000, 11000);
-	}
-}
-
-static void tmio_mmc_clk_stop(struct tmio_mmc_host *host)
-{
-	if (host->pdata->flags & TMIO_MMC_HAVE_HIGH_REG) {
-		sd_ctrl_write16(host, CTL_CLK_AND_WAIT_CTL, 0x0000);
-		usleep_range(10000, 11000);
-	}
-
-	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, ~CLK_CTL_SCLKEN &
-		sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
-
-	/* HW engineers overrode docs: no sleep needed on R-Car2+ */
-	if (!(host->pdata->flags & TMIO_MMC_MIN_RCAR2))
-		usleep_range(10000, 11000);
-}
-
-static void tmio_mmc_set_clock(struct tmio_mmc_host *host,
-			       unsigned int new_clock)
-{
-	u32 clk = 0, clock;
-
-	if (new_clock == 0) {
-		tmio_mmc_clk_stop(host);
-		return;
-	}
-	/*
-	 * Both HS400 and HS200/SD104 set 200MHz, but some devices need to
-	 * set 400MHz to distinguish the CPG settings in HS400.
-	 */
-	if (host->mmc->ios.timing == MMC_TIMING_MMC_HS400 &&
-	    host->pdata->flags & TMIO_MMC_HAVE_4TAP_HS400 &&
-	    new_clock == 200000000)
-		new_clock = 400000000;
-
-	if (host->clk_update)
-		clock = host->clk_update(host, new_clock) / 512;
-	else
-		clock = host->mmc->f_min;
-
-	for (clk = 0x80000080; new_clock >= (clock << 1); clk >>= 1)
-		clock <<= 1;
-
-	/* 1/1 clock is option */
-	if ((host->pdata->flags & TMIO_MMC_CLK_ACTUAL) &&
-	    ((clk >> 22) & 0x1)) {
-		if (!(host->mmc->ios.timing == MMC_TIMING_MMC_HS400))
-			clk |= 0xff;
-		else
-			clk &= ~0xff;
-	}
-
-	if (host->set_clk_div)
-		host->set_clk_div(host->pdev, (clk >> 22) & 1);
-
-	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, ~CLK_CTL_SCLKEN &
-			sd_ctrl_read16(host, CTL_SD_CARD_CLK_CTL));
-	sd_ctrl_write16(host, CTL_SD_CARD_CLK_CTL, clk & CLK_CTL_DIV_MASK);
-	if (!(host->pdata->flags & TMIO_MMC_MIN_RCAR2))
-		usleep_range(10000, 11000);
-
-	tmio_mmc_clk_start(host);
-}
-
 static void tmio_mmc_reset(struct tmio_mmc_host *host)
 {
 	/* FIXME - should we set stop clock reg here */
 	sd_ctrl_write16(host, CTL_RESET_SD, 0x0000);
-	if (host->pdata->flags & TMIO_MMC_HAVE_HIGH_REG)
-		sd_ctrl_write16(host, CTL_RESET_SDIO, 0x0000);
 	usleep_range(10000, 11000);
 	sd_ctrl_write16(host, CTL_RESET_SD, 0x0001);
-	if (host->pdata->flags & TMIO_MMC_HAVE_HIGH_REG)
-		sd_ctrl_write16(host, CTL_RESET_SDIO, 0x0001);
 	usleep_range(10000, 11000);
 
 	if (host->pdata->flags & TMIO_MMC_SDIO_IRQ) {
 		sd_ctrl_write16(host, CTL_SDIO_IRQ_MASK, host->sdio_irq_mask);
 		sd_ctrl_write16(host, CTL_TRANSACTION_CTL, 0x0001);
 	}
-
 }
 
 static void tmio_mmc_reset_work(struct work_struct *work)
@@ -294,7 +209,7 @@ static void tmio_mmc_reset_work(struct work_struct *work)
 
 	spin_unlock_irqrestore(&host->lock, flags);
 
-	tmio_mmc_reset(host);
+	host->reset(host);
 
 	/* Ready for new calls */
 	host->mrq = NULL;
@@ -446,7 +361,7 @@ static void tmio_mmc_pio_irq(struct tmio_mmc_host *host)
 	unsigned int count;
 	unsigned long flags;
 
-	if ((host->chan_tx || host->chan_rx) && !host->force_pio) {
+	if (host->dma_on) {
 		pr_err("PIO IRQ in DMA mode!\n");
 		return;
 	} else if (!data) {
@@ -518,7 +433,7 @@ void tmio_mmc_do_data_irq(struct tmio_mmc_host *host)
 	 */
 
 	if (data->flags & MMC_DATA_READ) {
-		if (host->chan_rx && !host->force_pio)
+		if (host->dma_on)
 			tmio_mmc_check_bounce_buffer(host);
 		dev_dbg(&host->pdev->dev, "Complete Rx request %p\n",
 			host->mrq);
@@ -555,7 +470,7 @@ static void tmio_mmc_data_irq(struct tmio_mmc_host *host, unsigned int stat)
 	if (stat & TMIO_STAT_CRCFAIL || stat & TMIO_STAT_STOPBIT_ERR ||
 	    stat & TMIO_STAT_TXUNDERRUN)
 		data->error = -EILSEQ;
-	if (host->chan_tx && (data->flags & MMC_DATA_WRITE) && !host->force_pio) {
+	if (host->dma_on && (data->flags & MMC_DATA_WRITE)) {
 		u32 status = sd_ctrl_read16_and_16_as_32(host, CTL_STATUS);
 		bool done = false;
 
@@ -579,7 +494,7 @@ static void tmio_mmc_data_irq(struct tmio_mmc_host *host, unsigned int stat)
 			tmio_mmc_disable_mmc_irqs(host, TMIO_STAT_DATAEND);
 			tmio_mmc_dataend_dma(host);
 		}
-	} else if (host->chan_rx && (data->flags & MMC_DATA_READ) && !host->force_pio) {
+	} else if (host->dma_on && (data->flags & MMC_DATA_READ)) {
 		tmio_mmc_disable_mmc_irqs(host, TMIO_STAT_DATAEND);
 		tmio_mmc_dataend_dma(host);
 	} else {
@@ -632,7 +547,7 @@ static void tmio_mmc_cmd_irq(struct tmio_mmc_host *host, unsigned int stat)
 	 */
 	if (host->data && (!cmd->error || cmd->error == -EILSEQ)) {
 		if (host->data->flags & MMC_DATA_READ) {
-			if (host->force_pio || !host->chan_rx) {
+			if (!host->dma_on) {
 				tmio_mmc_enable_mmc_irqs(host, TMIO_MASK_READOP);
 			} else {
 				tmio_mmc_disable_mmc_irqs(host,
@@ -640,7 +555,7 @@ static void tmio_mmc_cmd_irq(struct tmio_mmc_host *host, unsigned int stat)
 				tasklet_schedule(&host->dma_issue);
 			}
 		} else {
-			if (host->force_pio || !host->chan_tx) {
+			if (!host->dma_on) {
 				tmio_mmc_enable_mmc_irqs(host, TMIO_MASK_WRITEOP);
 			} else {
 				tmio_mmc_disable_mmc_irqs(host,
@@ -770,7 +685,7 @@ static int tmio_mmc_start_data(struct tmio_mmc_host *host,
 
 	tmio_mmc_init_sg(host, data);
 	host->data = data;
-	host->force_pio = false;
+	host->dma_on = false;
 
 	/* Set transfer length / blocksize */
 	sd_ctrl_write16(host, CTL_SD_XFER_LEN, data->blksz);
@@ -919,8 +834,8 @@ static void tmio_mmc_finish_request(struct tmio_mmc_host *host)
 	if (mrq->cmd->error || (mrq->data && mrq->data->error))
 		tmio_mmc_abort_dma(host);
 
-	if (host->check_scc_error)
-		host->check_scc_error(host);
+	if (host->check_scc_error && host->check_scc_error(host))
+		mrq->cmd->error = -EILSEQ;
 
 	/* If SET_BLOCK_COUNT, continue with main command */
 	if (host->mrq && !mrq->cmd->error) {
@@ -1043,15 +958,15 @@ static void tmio_mmc_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	switch (ios->power_mode) {
 	case MMC_POWER_OFF:
 		tmio_mmc_power_off(host);
-		tmio_mmc_clk_stop(host);
+		host->set_clock(host, 0);
 		break;
 	case MMC_POWER_UP:
 		tmio_mmc_power_on(host, ios->vdd);
-		tmio_mmc_set_clock(host, ios->clock);
+		host->set_clock(host, ios->clock);
 		tmio_mmc_set_bus_width(host, ios->bus_width);
 		break;
 	case MMC_POWER_ON:
-		tmio_mmc_set_clock(host, ios->clock);
+		host->set_clock(host, ios->clock);
 		tmio_mmc_set_bus_width(host, ios->bus_width);
 		break;
 	}
@@ -1237,7 +1152,7 @@ int tmio_mmc_host_probe(struct tmio_mmc_host *_host)
 	int ret;
 
 	/*
-	 * Check the sanity of mmc->f_min to prevent tmio_mmc_set_clock() from
+	 * Check the sanity of mmc->f_min to prevent host->set_clock() from
 	 * looping forever...
 	 */
 	if (mmc->f_min == 0)
@@ -1247,7 +1162,6 @@ int tmio_mmc_host_probe(struct tmio_mmc_host *_host)
 		_host->write16_hook = NULL;
 
 	_host->set_pwr = pdata->set_pwr;
-	_host->set_clk_div = pdata->set_clk_div;
 
 	ret = tmio_mmc_init_ocr(_host);
 	if (ret < 0)
@@ -1290,6 +1204,9 @@ int tmio_mmc_host_probe(struct tmio_mmc_host *_host)
 				  mmc->caps & MMC_CAP_NEEDS_POLL ||
 				  !mmc_card_is_removable(mmc));
 
+	if (!_host->reset)
+		_host->reset = tmio_mmc_reset;
+
 	/*
 	 * On Gen2+, eMMC with NONREMOVABLE currently fails because native
 	 * hotplug gets disabled. It seems RuntimePM related yet we need further
@@ -1310,8 +1227,8 @@ int tmio_mmc_host_probe(struct tmio_mmc_host *_host)
 	if (pdata->flags & TMIO_MMC_SDIO_IRQ)
 		_host->sdio_irq_mask = TMIO_SDIO_MASK_ALL;
 
-	tmio_mmc_clk_stop(_host);
-	tmio_mmc_reset(_host);
+	_host->set_clock(_host, 0);
+	_host->reset(_host);
 
 	_host->sdcard_irq_mask = sd_ctrl_read16_and_16_as_32(_host, CTL_IRQ_MASK);
 	tmio_mmc_disable_mmc_irqs(_host, TMIO_MASK_ALL);
@@ -1394,7 +1311,7 @@ int tmio_mmc_host_runtime_suspend(struct device *dev)
 	tmio_mmc_disable_mmc_irqs(host, TMIO_MASK_ALL);
 
 	if (host->clk_cache)
-		tmio_mmc_clk_stop(host);
+		host->set_clock(host, 0);
 
 	tmio_mmc_clk_disable(host);
 
@@ -1411,11 +1328,11 @@ int tmio_mmc_host_runtime_resume(struct device *dev)
 {
 	struct tmio_mmc_host *host = dev_get_drvdata(dev);
 
-	tmio_mmc_reset(host);
+	host->reset(host);
 	tmio_mmc_clk_enable(host);
 
 	if (host->clk_cache)
-		tmio_mmc_set_clock(host, host->clk_cache);
+		host->set_clock(host, host->clk_cache);
 
 	if (host->native_hotplug)
 		tmio_mmc_enable_mmc_irqs(host,
diff --git a/drivers/mmc/host/uniphier-sd.c b/drivers/mmc/host/uniphier-sd.c
new file mode 100644
index 000000000000..91a2be41edf6
--- /dev/null
+++ b/drivers/mmc/host/uniphier-sd.c
@@ -0,0 +1,698 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// Copyright (C) 2017-2018 Socionext Inc.
+//   Author: Masahiro Yamada <yamada.masahiro@socionext.com>
+
+#include <linux/bitfield.h>
+#include <linux/bitops.h>
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/mfd/tmio.h>
+#include <linux/mmc/host.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+
+#include "tmio_mmc.h"
+
+#define   UNIPHIER_SD_CLK_CTL_DIV1024		BIT(16)
+#define   UNIPHIER_SD_CLK_CTL_DIV1		BIT(10)
+#define   UNIPHIER_SD_CLKCTL_OFFEN		BIT(9)  // auto SDCLK stop
+#define UNIPHIER_SD_CC_EXT_MODE		0x1b0
+#define   UNIPHIER_SD_CC_EXT_MODE_DMA		BIT(1)
+#define UNIPHIER_SD_HOST_MODE		0x1c8
+#define UNIPHIER_SD_VOLT		0x1e4
+#define   UNIPHIER_SD_VOLT_MASK			GENMASK(1, 0)
+#define   UNIPHIER_SD_VOLT_OFF			0
+#define   UNIPHIER_SD_VOLT_330			1	// 3.3V signal
+#define   UNIPHIER_SD_VOLT_180			2	// 1.8V signal
+#define UNIPHIER_SD_DMA_MODE		0x410
+#define   UNIPHIER_SD_DMA_MODE_DIR_MASK		GENMASK(17, 16)
+#define   UNIPHIER_SD_DMA_MODE_DIR_TO_DEV	0
+#define   UNIPHIER_SD_DMA_MODE_DIR_FROM_DEV	1
+#define   UNIPHIER_SD_DMA_MODE_WIDTH_MASK	GENMASK(5, 4)
+#define   UNIPHIER_SD_DMA_MODE_WIDTH_8		0
+#define   UNIPHIER_SD_DMA_MODE_WIDTH_16		1
+#define   UNIPHIER_SD_DMA_MODE_WIDTH_32		2
+#define   UNIPHIER_SD_DMA_MODE_WIDTH_64		3
+#define   UNIPHIER_SD_DMA_MODE_ADDR_INC		BIT(0)	// 1: inc, 0: fixed
+#define UNIPHIER_SD_DMA_CTL		0x414
+#define   UNIPHIER_SD_DMA_CTL_START	BIT(0)	// start DMA (auto cleared)
+#define UNIPHIER_SD_DMA_RST		0x418
+#define   UNIPHIER_SD_DMA_RST_CH1	BIT(9)
+#define   UNIPHIER_SD_DMA_RST_CH0	BIT(8)
+#define UNIPHIER_SD_DMA_ADDR_L		0x440
+#define UNIPHIER_SD_DMA_ADDR_H		0x444
+
+/*
+ * IP is extended to support various features: built-in DMA engine,
+ * 1/1024 divisor, etc.
+ */
+#define UNIPHIER_SD_CAP_EXTENDED_IP		BIT(0)
+/* RX channel of the built-in DMA controller is broken (Pro5) */
+#define UNIPHIER_SD_CAP_BROKEN_DMA_RX		BIT(1)
+
+struct uniphier_sd_priv {
+	struct tmio_mmc_data tmio_data;
+	struct pinctrl *pinctrl;
+	struct pinctrl_state *pinstate_default;
+	struct pinctrl_state *pinstate_uhs;
+	struct clk *clk;
+	struct reset_control *rst;
+	struct reset_control *rst_br;
+	struct reset_control *rst_hw;
+	struct dma_chan *chan;
+	enum dma_data_direction dma_dir;
+	unsigned long clk_rate;
+	unsigned long caps;
+};
+
+static void *uniphier_sd_priv(struct tmio_mmc_host *host)
+{
+	return container_of(host->pdata, struct uniphier_sd_priv, tmio_data);
+}
+
+static void uniphier_sd_dma_endisable(struct tmio_mmc_host *host, int enable)
+{
+	sd_ctrl_write16(host, CTL_DMA_ENABLE, enable ? DMA_ENABLE_DMASDRW : 0);
+}
+
+/* external DMA engine */
+static void uniphier_sd_external_dma_issue(unsigned long arg)
+{
+	struct tmio_mmc_host *host = (void *)arg;
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	uniphier_sd_dma_endisable(host, 1);
+	dma_async_issue_pending(priv->chan);
+}
+
+static void uniphier_sd_external_dma_callback(void *param,
+					const struct dmaengine_result *result)
+{
+	struct tmio_mmc_host *host = param;
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	unsigned long flags;
+
+	dma_unmap_sg(mmc_dev(host->mmc), host->sg_ptr, host->sg_len,
+		     priv->dma_dir);
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	if (result->result == DMA_TRANS_NOERROR) {
+		/*
+		 * When the external DMA engine is enabled, strangely enough,
+		 * the DATAEND flag can be asserted even if the DMA engine has
+		 * not been kicked yet.  Enable the TMIO_STAT_DATAEND irq only
+		 * after we make sure the DMA engine finishes the transfer,
+		 * hence, in this callback.
+		 */
+		tmio_mmc_enable_mmc_irqs(host, TMIO_STAT_DATAEND);
+	} else {
+		host->data->error = -ETIMEDOUT;
+		tmio_mmc_do_data_irq(host);
+	}
+
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void uniphier_sd_external_dma_start(struct tmio_mmc_host *host,
+					   struct mmc_data *data)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	enum dma_transfer_direction dma_tx_dir;
+	struct dma_async_tx_descriptor *desc;
+	dma_cookie_t cookie;
+	int sg_len;
+
+	if (!priv->chan)
+		goto force_pio;
+
+	if (data->flags & MMC_DATA_READ) {
+		priv->dma_dir = DMA_FROM_DEVICE;
+		dma_tx_dir = DMA_DEV_TO_MEM;
+	} else {
+		priv->dma_dir = DMA_TO_DEVICE;
+		dma_tx_dir = DMA_MEM_TO_DEV;
+	}
+
+	sg_len = dma_map_sg(mmc_dev(host->mmc), host->sg_ptr, host->sg_len,
+			    priv->dma_dir);
+	if (sg_len == 0)
+		goto force_pio;
+
+	desc = dmaengine_prep_slave_sg(priv->chan, host->sg_ptr, sg_len,
+				       dma_tx_dir, DMA_CTRL_ACK);
+	if (!desc)
+		goto unmap_sg;
+
+	desc->callback_result = uniphier_sd_external_dma_callback;
+	desc->callback_param = host;
+
+	cookie = dmaengine_submit(desc);
+	if (cookie < 0)
+		goto unmap_sg;
+
+	host->dma_on = true;
+
+	return;
+
+unmap_sg:
+	dma_unmap_sg(mmc_dev(host->mmc), host->sg_ptr, host->sg_len,
+		     priv->dma_dir);
+force_pio:
+	uniphier_sd_dma_endisable(host, 0);
+}
+
+static void uniphier_sd_external_dma_enable(struct tmio_mmc_host *host,
+					    bool enable)
+{
+}
+
+static void uniphier_sd_external_dma_request(struct tmio_mmc_host *host,
+					     struct tmio_mmc_data *pdata)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	struct dma_chan *chan;
+
+	chan = dma_request_chan(mmc_dev(host->mmc), "rx-tx");
+	if (IS_ERR(chan)) {
+		dev_warn(mmc_dev(host->mmc),
+			 "failed to request DMA channel. falling back to PIO\n");
+		return;	/* just use PIO even for -EPROBE_DEFER */
+	}
+
+	/* this driver uses a single channel for both RX an TX */
+	priv->chan = chan;
+	host->chan_rx = chan;
+	host->chan_tx = chan;
+
+	tasklet_init(&host->dma_issue, uniphier_sd_external_dma_issue,
+		     (unsigned long)host);
+}
+
+static void uniphier_sd_external_dma_release(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	if (priv->chan)
+		dma_release_channel(priv->chan);
+}
+
+static void uniphier_sd_external_dma_abort(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	uniphier_sd_dma_endisable(host, 0);
+
+	if (priv->chan)
+		dmaengine_terminate_sync(priv->chan);
+}
+
+static void uniphier_sd_external_dma_dataend(struct tmio_mmc_host *host)
+{
+	uniphier_sd_dma_endisable(host, 0);
+
+	tmio_mmc_do_data_irq(host);
+}
+
+static const struct tmio_mmc_dma_ops uniphier_sd_external_dma_ops = {
+	.start = uniphier_sd_external_dma_start,
+	.enable = uniphier_sd_external_dma_enable,
+	.request = uniphier_sd_external_dma_request,
+	.release = uniphier_sd_external_dma_release,
+	.abort = uniphier_sd_external_dma_abort,
+	.dataend = uniphier_sd_external_dma_dataend,
+};
+
+static void uniphier_sd_internal_dma_issue(unsigned long arg)
+{
+	struct tmio_mmc_host *host = (void *)arg;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+	tmio_mmc_enable_mmc_irqs(host, TMIO_STAT_DATAEND);
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	uniphier_sd_dma_endisable(host, 1);
+	writel(UNIPHIER_SD_DMA_CTL_START, host->ctl + UNIPHIER_SD_DMA_CTL);
+}
+
+static void uniphier_sd_internal_dma_start(struct tmio_mmc_host *host,
+					   struct mmc_data *data)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	struct scatterlist *sg = host->sg_ptr;
+	dma_addr_t dma_addr;
+	unsigned int dma_mode_dir;
+	u32 dma_mode;
+	int sg_len;
+
+	if ((data->flags & MMC_DATA_READ) && !host->chan_rx)
+		goto force_pio;
+
+	if (WARN_ON(host->sg_len != 1))
+		goto force_pio;
+
+	if (!IS_ALIGNED(sg->offset, 8))
+		goto force_pio;
+
+	if (data->flags & MMC_DATA_READ) {
+		priv->dma_dir = DMA_FROM_DEVICE;
+		dma_mode_dir = UNIPHIER_SD_DMA_MODE_DIR_FROM_DEV;
+	} else {
+		priv->dma_dir = DMA_TO_DEVICE;
+		dma_mode_dir = UNIPHIER_SD_DMA_MODE_DIR_TO_DEV;
+	}
+
+	sg_len = dma_map_sg(mmc_dev(host->mmc), sg, 1, priv->dma_dir);
+	if (sg_len == 0)
+		goto force_pio;
+
+	dma_mode = FIELD_PREP(UNIPHIER_SD_DMA_MODE_DIR_MASK, dma_mode_dir);
+	dma_mode |= FIELD_PREP(UNIPHIER_SD_DMA_MODE_WIDTH_MASK,
+			       UNIPHIER_SD_DMA_MODE_WIDTH_64);
+	dma_mode |= UNIPHIER_SD_DMA_MODE_ADDR_INC;
+
+	writel(dma_mode, host->ctl + UNIPHIER_SD_DMA_MODE);
+
+	dma_addr = sg_dma_address(data->sg);
+	writel(lower_32_bits(dma_addr), host->ctl + UNIPHIER_SD_DMA_ADDR_L);
+	writel(upper_32_bits(dma_addr), host->ctl + UNIPHIER_SD_DMA_ADDR_H);
+
+	host->dma_on = true;
+
+	return;
+force_pio:
+	uniphier_sd_dma_endisable(host, 0);
+}
+
+static void uniphier_sd_internal_dma_enable(struct tmio_mmc_host *host,
+					    bool enable)
+{
+}
+
+static void uniphier_sd_internal_dma_request(struct tmio_mmc_host *host,
+					     struct tmio_mmc_data *pdata)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	/*
+	 * Due to a hardware bug, Pro5 cannot use DMA for RX.
+	 * We can still use DMA for TX, but PIO for RX.
+	 */
+	if (!(priv->caps & UNIPHIER_SD_CAP_BROKEN_DMA_RX))
+		host->chan_rx = (void *)0xdeadbeaf;
+
+	host->chan_tx = (void *)0xdeadbeaf;
+
+	tasklet_init(&host->dma_issue, uniphier_sd_internal_dma_issue,
+		     (unsigned long)host);
+}
+
+static void uniphier_sd_internal_dma_release(struct tmio_mmc_host *host)
+{
+	/* Each value is set to zero to assume "disabling" each DMA */
+	host->chan_rx = NULL;
+	host->chan_tx = NULL;
+}
+
+static void uniphier_sd_internal_dma_abort(struct tmio_mmc_host *host)
+{
+	u32 tmp;
+
+	uniphier_sd_dma_endisable(host, 0);
+
+	tmp = readl(host->ctl + UNIPHIER_SD_DMA_RST);
+	tmp &= ~(UNIPHIER_SD_DMA_RST_CH1 | UNIPHIER_SD_DMA_RST_CH0);
+	writel(tmp, host->ctl + UNIPHIER_SD_DMA_RST);
+
+	tmp |= UNIPHIER_SD_DMA_RST_CH1 | UNIPHIER_SD_DMA_RST_CH0;
+	writel(tmp, host->ctl + UNIPHIER_SD_DMA_RST);
+}
+
+static void uniphier_sd_internal_dma_dataend(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	uniphier_sd_dma_endisable(host, 0);
+	dma_unmap_sg(mmc_dev(host->mmc), host->sg_ptr, 1, priv->dma_dir);
+
+	tmio_mmc_do_data_irq(host);
+}
+
+static const struct tmio_mmc_dma_ops uniphier_sd_internal_dma_ops = {
+	.start = uniphier_sd_internal_dma_start,
+	.enable = uniphier_sd_internal_dma_enable,
+	.request = uniphier_sd_internal_dma_request,
+	.release = uniphier_sd_internal_dma_release,
+	.abort = uniphier_sd_internal_dma_abort,
+	.dataend = uniphier_sd_internal_dma_dataend,
+};
+
+static int uniphier_sd_clk_enable(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	struct mmc_host *mmc = host->mmc;
+	int ret;
+
+	ret = clk_prepare_enable(priv->clk);
+	if (ret)
+		return ret;
+
+	ret = clk_set_rate(priv->clk, ULONG_MAX);
+	if (ret)
+		goto disable_clk;
+
+	priv->clk_rate = clk_get_rate(priv->clk);
+
+	/* If max-frequency property is set, use it. */
+	if (!mmc->f_max)
+		mmc->f_max = priv->clk_rate;
+
+	/*
+	 * 1/512 is the finest divisor in the original IP.  Newer versions
+	 * also supports 1/1024 divisor. (UniPhier-specific extension)
+	 */
+	if (priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP)
+		mmc->f_min = priv->clk_rate / 1024;
+	else
+		mmc->f_min = priv->clk_rate / 512;
+
+	ret = reset_control_deassert(priv->rst);
+	if (ret)
+		goto disable_clk;
+
+	ret = reset_control_deassert(priv->rst_br);
+	if (ret)
+		goto assert_rst;
+
+	return 0;
+
+assert_rst:
+	reset_control_assert(priv->rst);
+disable_clk:
+	clk_disable_unprepare(priv->clk);
+
+	return ret;
+}
+
+static void uniphier_sd_clk_disable(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	reset_control_assert(priv->rst_br);
+	reset_control_assert(priv->rst);
+	clk_disable_unprepare(priv->clk);
+}
+
+static void uniphier_sd_hw_reset(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+
+	reset_control_assert(priv->rst_hw);
+	/* For eMMC, minimum is 1us but give it 9us for good measure */
+	udelay(9);
+	reset_control_deassert(priv->rst_hw);
+	/* For eMMC, minimum is 200us but give it 300us for good measure */
+	usleep_range(300, 1000);
+}
+
+static void uniphier_sd_set_clock(struct tmio_mmc_host *host,
+				  unsigned int clock)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	unsigned long divisor;
+	u32 tmp;
+
+	tmp = readl(host->ctl + (CTL_SD_CARD_CLK_CTL << 1));
+
+	/* stop the clock before changing its rate to avoid a glitch signal */
+	tmp &= ~CLK_CTL_SCLKEN;
+	writel(tmp, host->ctl + (CTL_SD_CARD_CLK_CTL << 1));
+
+	if (clock == 0)
+		return;
+
+	tmp &= ~UNIPHIER_SD_CLK_CTL_DIV1024;
+	tmp &= ~UNIPHIER_SD_CLK_CTL_DIV1;
+	tmp &= ~CLK_CTL_DIV_MASK;
+
+	divisor = priv->clk_rate / clock;
+
+	/*
+	 * In the original IP, bit[7:0] represents the divisor.
+	 * bit7 set: 1/512, ... bit0 set:1/4, all bits clear: 1/2
+	 *
+	 * The IP does not define a way to achieve 1/1.  For UniPhier variants,
+	 * bit10 is used for 1/1.  Newer versions of UniPhier variants use
+	 * bit16 for 1/1024.
+	 */
+	if (divisor <= 1)
+		tmp |= UNIPHIER_SD_CLK_CTL_DIV1;
+	else if (priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP && divisor > 512)
+		tmp |= UNIPHIER_SD_CLK_CTL_DIV1024;
+	else
+		tmp |= roundup_pow_of_two(divisor) >> 2;
+
+	writel(tmp, host->ctl + (CTL_SD_CARD_CLK_CTL << 1));
+
+	tmp |= CLK_CTL_SCLKEN;
+	writel(tmp, host->ctl + (CTL_SD_CARD_CLK_CTL << 1));
+}
+
+static void uniphier_sd_host_init(struct tmio_mmc_host *host)
+{
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	u32 val;
+
+	/*
+	 * Connected to 32bit AXI.
+	 * This register holds settings for SoC-specific internal bus
+	 * connection.  What is worse, the register spec was changed,
+	 * breaking the backward compatibility.  Write an appropriate
+	 * value depending on a flag associated with a compatible string.
+	 */
+	if (priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP)
+		val = 0x00000101;
+	else
+		val = 0x00000000;
+
+	writel(val, host->ctl + UNIPHIER_SD_HOST_MODE);
+
+	val = 0;
+	/*
+	 * If supported, the controller can automatically
+	 * enable/disable the clock line to the card.
+	 */
+	if (priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP)
+		val |= UNIPHIER_SD_CLKCTL_OFFEN;
+
+	writel(val, host->ctl + (CTL_SD_CARD_CLK_CTL << 1));
+}
+
+static int uniphier_sd_start_signal_voltage_switch(struct mmc_host *mmc,
+						   struct mmc_ios *ios)
+{
+	struct tmio_mmc_host *host = mmc_priv(mmc);
+	struct uniphier_sd_priv *priv = uniphier_sd_priv(host);
+	struct pinctrl_state *pinstate;
+	u32 val, tmp;
+
+	switch (ios->signal_voltage) {
+	case MMC_SIGNAL_VOLTAGE_330:
+		val = UNIPHIER_SD_VOLT_330;
+		pinstate = priv->pinstate_default;
+		break;
+	case MMC_SIGNAL_VOLTAGE_180:
+		val = UNIPHIER_SD_VOLT_180;
+		pinstate = priv->pinstate_uhs;
+		break;
+	default:
+		return -ENOTSUPP;
+	}
+
+	tmp = readl(host->ctl + UNIPHIER_SD_VOLT);
+	tmp &= ~UNIPHIER_SD_VOLT_MASK;
+	tmp |= FIELD_PREP(UNIPHIER_SD_VOLT_MASK, val);
+	writel(tmp, host->ctl + UNIPHIER_SD_VOLT);
+
+	pinctrl_select_state(priv->pinctrl, pinstate);
+
+	return 0;
+}
+
+static int uniphier_sd_uhs_init(struct tmio_mmc_host *host,
+				struct uniphier_sd_priv *priv)
+{
+	priv->pinctrl = devm_pinctrl_get(mmc_dev(host->mmc));
+	if (IS_ERR(priv->pinctrl))
+		return PTR_ERR(priv->pinctrl);
+
+	priv->pinstate_default = pinctrl_lookup_state(priv->pinctrl,
+						      PINCTRL_STATE_DEFAULT);
+	if (IS_ERR(priv->pinstate_default))
+		return PTR_ERR(priv->pinstate_default);
+
+	priv->pinstate_uhs = pinctrl_lookup_state(priv->pinctrl, "uhs");
+	if (IS_ERR(priv->pinstate_uhs))
+		return PTR_ERR(priv->pinstate_uhs);
+
+	host->ops.start_signal_voltage_switch =
+					uniphier_sd_start_signal_voltage_switch;
+
+	return 0;
+}
+
+static int uniphier_sd_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct uniphier_sd_priv *priv;
+	struct tmio_mmc_data *tmio_data;
+	struct tmio_mmc_host *host;
+	int irq, ret;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "failed to get IRQ number");
+		return irq;
+	}
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->caps = (unsigned long)of_device_get_match_data(dev);
+
+	priv->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(priv->clk)) {
+		dev_err(dev, "failed to get clock\n");
+		return PTR_ERR(priv->clk);
+	}
+
+	priv->rst = devm_reset_control_get_shared(dev, "host");
+	if (IS_ERR(priv->rst)) {
+		dev_err(dev, "failed to get host reset\n");
+		return PTR_ERR(priv->rst);
+	}
+
+	/* old version has one more reset */
+	if (!(priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP)) {
+		priv->rst_br = devm_reset_control_get_shared(dev, "bridge");
+		if (IS_ERR(priv->rst_br)) {
+			dev_err(dev, "failed to get bridge reset\n");
+			return PTR_ERR(priv->rst_br);
+		}
+	}
+
+	tmio_data = &priv->tmio_data;
+	tmio_data->flags |= TMIO_MMC_32BIT_DATA_PORT;
+
+	host = tmio_mmc_host_alloc(pdev, tmio_data);
+	if (IS_ERR(host))
+		return PTR_ERR(host);
+
+	if (host->mmc->caps & MMC_CAP_HW_RESET) {
+		priv->rst_hw = devm_reset_control_get_exclusive(dev, "hw");
+		if (IS_ERR(priv->rst_hw)) {
+			dev_err(dev, "failed to get hw reset\n");
+			ret = PTR_ERR(priv->rst_hw);
+			goto free_host;
+		}
+		host->hw_reset = uniphier_sd_hw_reset;
+	}
+
+	if (host->mmc->caps & MMC_CAP_UHS) {
+		ret = uniphier_sd_uhs_init(host, priv);
+		if (ret) {
+			dev_warn(dev,
+				 "failed to setup UHS (error %d).  Disabling UHS.",
+				 ret);
+			host->mmc->caps &= ~MMC_CAP_UHS;
+		}
+	}
+
+	ret = devm_request_irq(dev, irq, tmio_mmc_irq, IRQF_SHARED,
+			       dev_name(dev), host);
+	if (ret)
+		goto free_host;
+
+	if (priv->caps & UNIPHIER_SD_CAP_EXTENDED_IP)
+		host->dma_ops = &uniphier_sd_internal_dma_ops;
+	else
+		host->dma_ops = &uniphier_sd_external_dma_ops;
+
+	host->bus_shift = 1;
+	host->clk_enable = uniphier_sd_clk_enable;
+	host->clk_disable = uniphier_sd_clk_disable;
+	host->set_clock = uniphier_sd_set_clock;
+
+	ret = uniphier_sd_clk_enable(host);
+	if (ret)
+		goto free_host;
+
+	uniphier_sd_host_init(host);
+
+	tmio_data->ocr_mask = MMC_VDD_32_33 | MMC_VDD_33_34;
+	if (host->mmc->caps & MMC_CAP_UHS)
+		tmio_data->ocr_mask |= MMC_VDD_165_195;
+
+	tmio_data->max_segs = 1;
+	tmio_data->max_blk_count = U16_MAX;
+
+	ret = tmio_mmc_host_probe(host);
+	if (ret)
+		goto free_host;
+
+	return 0;
+
+free_host:
+	tmio_mmc_host_free(host);
+
+	return ret;
+}
+
+static int uniphier_sd_remove(struct platform_device *pdev)
+{
+	struct tmio_mmc_host *host = platform_get_drvdata(pdev);
+
+	tmio_mmc_host_remove(host);
+	uniphier_sd_clk_disable(host);
+
+	return 0;
+}
+
+static const struct of_device_id uniphier_sd_match[] = {
+	{
+		.compatible = "socionext,uniphier-sd-v2.91",
+	},
+	{
+		.compatible = "socionext,uniphier-sd-v3.1",
+		.data = (void *)(UNIPHIER_SD_CAP_EXTENDED_IP |
+				 UNIPHIER_SD_CAP_BROKEN_DMA_RX),
+	},
+	{
+		.compatible = "socionext,uniphier-sd-v3.1.1",
+		.data = (void *)UNIPHIER_SD_CAP_EXTENDED_IP,
+	},
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, uniphier_sd_match);
+
+static struct platform_driver uniphier_sd_driver = {
+	.probe = uniphier_sd_probe,
+	.remove = uniphier_sd_remove,
+	.driver = {
+		.name = "uniphier-sd",
+		.of_match_table = uniphier_sd_match,
+	},
+};
+module_platform_driver(uniphier_sd_driver);
+
+MODULE_AUTHOR("Masahiro Yamada <yamada.masahiro@socionext.com>");
+MODULE_DESCRIPTION("UniPhier SD/eMMC host controller driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/mmc/host/usdhi6rol0.c b/drivers/mmc/host/usdhi6rol0.c
index cdfeb15b6f05..cd8b1b9d4d8a 100644
--- a/drivers/mmc/host/usdhi6rol0.c
+++ b/drivers/mmc/host/usdhi6rol0.c
@@ -1,10 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2013-2014 Renesas Electronics Europe Ltd.
  * Author: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of version 2 of the GNU General Public License as
- * published by the Free Software Foundation.
  */
 
 #include <linux/clk.h>
author	Linus Torvalds <torvalds@linux-foundation.org>	2018-10-23 08:36:15 +0100
committer	Linus Torvalds <torvalds@linux-foundation.org>	2018-10-23 08:36:15 +0100
commit	1650ac53066577a5e83fe3e9d992c9311597ff8c (patch)
tree	58667f89d6b5ac5a40c386548c17646c561453ef /drivers/mmc
parent	Merge tag 'regulator-v5.0' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator (diff)
parent	mmc: mediatek: drop too much code of tuning method (diff)
download	linux-dev-1650ac53066577a5e83fe3e9d992c9311597ff8c.tar.xz linux-dev-1650ac53066577a5e83fe3e9d992c9311597ff8c.zip