Merge tag 'drivers-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc

Pull ARM SoC driver changes from Arnd Bergmann:
 "These changes are mostly for ARM specific device drivers that either
  don't have an upstream maintainer, or that had the maintainer ask us
  to pick up the changes to avoid conflicts.

  A large chunk of this are clock drivers (bcm281xx, exynos, versatile,
  shmobile), aside from that, reset controllers for STi as well as a
  large rework of the Marvell Orion/EBU watchdog driver are notable"

* tag 'drivers-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (99 commits)
  Revert "dts: socfpga: Add DTS entry for adding the stmmac glue layer for stmmac."
  Revert "net: stmmac: Add SOCFPGA glue driver"
  ARM: shmobile: r8a7791: Fix SCIFA3-5 clocks
  ARM: STi: Add reset controller support to mach-sti Kconfig
  drivers: reset: stih416: add softreset controller
  drivers: reset: stih415: add softreset controller
  drivers: reset: Reset controller driver for STiH416
  drivers: reset: Reset controller driver for STiH415
  drivers: reset: STi SoC system configuration reset controller support
  dts: socfpga: Add sysmgr node so the gmac can use to reference
  dts: socfpga: Add support for SD/MMC on the SOCFPGA platform
  reset: Add optional resets and stubs
  ARM: shmobile: r7s72100: fix bus clock calculation
  Power: Reset: Generalize qnap-poweroff to work on Synology devices.
  dts: socfpga: Update clock entry to support multiple parents
  ARM: socfpga: Update socfpga_defconfig
  dts: socfpga: Add DTS entry for adding the stmmac glue layer for stmmac.
  net: stmmac: Add SOCFPGA glue driver
  watchdog: orion_wdt: Use %pa to print 'phys_addr_t'
  drivers: cci: Export CCI PMU revision
  ...

54 files changed, 4370 insertions, 349 deletions

diff --git a/drivers/amba/tegra-ahb.c b/drivers/amba/tegra-ahb.c
index 1f44e56..558a239 100644
--- a/drivers/amba/tegra-ahb.c
+++ b/drivers/amba/tegra-ahb.c
@@ -256,8 +256,6 @@ static int tegra_ahb_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!res)
-		return -ENODEV;
 	ahb->regs = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(ahb->regs))
 		return PTR_ERR(ahb->regs);
diff --git a/drivers/bus/arm-cci.c b/drivers/bus/arm-cci.c
index 962fd35..5a86da9 100644
--- a/drivers/bus/arm-cci.c
+++ b/drivers/bus/arm-cci.c
@@ -31,7 +31,6 @@
 
 #define DRIVER_NAME		"CCI-400"
 #define DRIVER_NAME_PMU		DRIVER_NAME " PMU"
-#define PMU_NAME		"CCI_400"
 
 #define CCI_PORT_CTRL		0x0
 #define CCI_CTRL_STATUS		0xc
@@ -88,8 +87,7 @@ static unsigned long cci_ctrl_phys;
 
 #define CCI_REV_R0		0
 #define CCI_REV_R1		1
-#define CCI_REV_R0_P4		4
-#define CCI_REV_R1_P2		6
+#define CCI_REV_R1_PX		5
 
 #define CCI_PMU_EVT_SEL		0x000
 #define CCI_PMU_CNTR		0x004
@@ -163,6 +161,15 @@ static struct pmu_port_event_ranges port_event_range[] = {
 	},
 };
 
+/*
+ * Export different PMU names for the different revisions so userspace knows
+ * because the event ids are different
+ */
+static char *const pmu_names[] = {
+	[CCI_REV_R0] = "CCI_400",
+	[CCI_REV_R1] = "CCI_400_r1",
+};
+
 struct cci_pmu_drv_data {
 	void __iomem *base;
 	struct arm_pmu *cci_pmu;
@@ -193,21 +200,16 @@ static int probe_cci_revision(void)
 	rev = readl_relaxed(cci_ctrl_base + CCI_PID2) & CCI_PID2_REV_MASK;
 	rev >>= CCI_PID2_REV_SHIFT;
 
-	if (rev <= CCI_REV_R0_P4)
+	if (rev < CCI_REV_R1_PX)
 		return CCI_REV_R0;
-	else if (rev <= CCI_REV_R1_P2)
+	else
 		return CCI_REV_R1;
-
-	return -ENOENT;
 }
 
 static struct pmu_port_event_ranges *port_range_by_rev(void)
 {
 	int rev = probe_cci_revision();
 
-	if (rev < 0)
-		return NULL;
-
 	return &port_event_range[rev];
 }
 
@@ -526,7 +528,7 @@ static void pmu_write_counter(struct perf_event *event, u32 value)
 static int cci_pmu_init(struct arm_pmu *cci_pmu, struct platform_device *pdev)
 {
 	*cci_pmu = (struct arm_pmu){
-		.name             = PMU_NAME,
+		.name		  = pmu_names[probe_cci_revision()],
 		.max_period       = (1LLU << 32) - 1,
 		.get_hw_events    = pmu_get_hw_events,
 		.get_event_idx    = pmu_get_event_idx,
diff --git a/drivers/bus/mvebu-mbus.c b/drivers/bus/mvebu-mbus.c
index 2ac754e..293e2e0 100644
--- a/drivers/bus/mvebu-mbus.c
+++ b/drivers/bus/mvebu-mbus.c
@@ -890,13 +890,12 @@ int __init mvebu_mbus_dt_init(void)
 	const __be32 *prop;
 	int ret;
 
-	np = of_find_matching_node(NULL, of_mvebu_mbus_ids);
+	np = of_find_matching_node_and_match(NULL, of_mvebu_mbus_ids, &of_id);
 	if (!np) {
 		pr_err("could not find a matching SoC family\n");
 		return -ENODEV;
 	}
 
-	of_id = of_match_node(of_mvebu_mbus_ids, np);
 	mbus_state.soc = of_id->data;
 
 	prop = of_get_property(np, "controller", NULL);
diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig
index 2f2b084..244759b 100644
--- a/drivers/char/hw_random/Kconfig
+++ b/drivers/char/hw_random/Kconfig
@@ -342,11 +342,11 @@ config HW_RANDOM_TPM
 	  If unsure, say Y.
 
 config HW_RANDOM_MSM
-	tristate "Qualcomm MSM Random Number Generator support"
-	depends on HW_RANDOM && ARCH_MSM
+	tristate "Qualcomm SoCs Random Number Generator support"
+	depends on HW_RANDOM && ARCH_QCOM
 	---help---
 	  This driver provides kernel-side support for the Random Number
-	  Generator hardware found on Qualcomm MSM SoCs.
+	  Generator hardware found on Qualcomm SoCs.
 
 	  To compile this driver as a module, choose M here. the
 	  module will be called msm-rng.
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index 7641965..f9f6056 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -111,4 +111,5 @@ source "drivers/clk/qcom/Kconfig"
 
 endmenu
 
+source "drivers/clk/bcm/Kconfig"
 source "drivers/clk/mvebu/Kconfig"
diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
index a367a98..88af4a3 100644
--- a/drivers/clk/Makefile
+++ b/drivers/clk/Makefile
@@ -29,6 +29,7 @@ obj-$(CONFIG_ARCH_VT8500)		+= clk-vt8500.o
 obj-$(CONFIG_COMMON_CLK_WM831X)		+= clk-wm831x.o
 obj-$(CONFIG_COMMON_CLK_XGENE)		+= clk-xgene.o
 obj-$(CONFIG_COMMON_CLK_AT91)		+= at91/
+obj-$(CONFIG_ARCH_BCM_MOBILE)		+= bcm/
 obj-$(CONFIG_ARCH_HI3xxx)		+= hisilicon/
 obj-$(CONFIG_COMMON_CLK_KEYSTONE)	+= keystone/
 ifeq ($(CONFIG_COMMON_CLK), y)
diff --git a/drivers/clk/bcm/Kconfig b/drivers/clk/bcm/Kconfig
new file mode 100644
index 0000000..a7262fb
--- /dev/null
+++ b/drivers/clk/bcm/Kconfig
@@ -0,0 +1,9 @@
+config CLK_BCM_KONA
+	bool "Broadcom Kona CCU clock support"
+	depends on ARCH_BCM_MOBILE
+	depends on COMMON_CLK
+	default y
+	help
+	  Enable common clock framework support for Broadcom SoCs
+	  using "Kona" style clock control units, including those
+	  in the BCM281xx family.
diff --git a/drivers/clk/bcm/Makefile b/drivers/clk/bcm/Makefile
new file mode 100644
index 0000000..cf93359
--- /dev/null
+++ b/drivers/clk/bcm/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_CLK_BCM_KONA)	+= clk-kona.o
+obj-$(CONFIG_CLK_BCM_KONA)	+= clk-kona-setup.o
+obj-$(CONFIG_CLK_BCM_KONA)	+= clk-bcm281xx.o
diff --git a/drivers/clk/bcm/clk-bcm281xx.c b/drivers/clk/bcm/clk-bcm281xx.c
new file mode 100644
index 0000000..3c66de6
--- /dev/null
+++ b/drivers/clk/bcm/clk-bcm281xx.c
@@ -0,0 +1,416 @@
+/*
+ * Copyright (C) 2013 Broadcom Corporation
+ * Copyright 2013 Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include "clk-kona.h"
+#include "dt-bindings/clock/bcm281xx.h"
+
+/* bcm11351 CCU device tree "compatible" strings */
+#define BCM11351_DT_ROOT_CCU_COMPAT	"brcm,bcm11351-root-ccu"
+#define BCM11351_DT_AON_CCU_COMPAT	"brcm,bcm11351-aon-ccu"
+#define BCM11351_DT_HUB_CCU_COMPAT	"brcm,bcm11351-hub-ccu"
+#define BCM11351_DT_MASTER_CCU_COMPAT	"brcm,bcm11351-master-ccu"
+#define BCM11351_DT_SLAVE_CCU_COMPAT	"brcm,bcm11351-slave-ccu"
+
+/* Root CCU clocks */
+
+static struct peri_clk_data frac_1m_data = {
+	.gate		= HW_SW_GATE(0x214, 16, 0, 1),
+	.trig		= TRIGGER(0x0e04, 0),
+	.div		= FRAC_DIVIDER(0x0e00, 0, 22, 16),
+	.clocks		= CLOCKS("ref_crystal"),
+};
+
+/* AON CCU clocks */
+
+static struct peri_clk_data hub_timer_data = {
+	.gate		= HW_SW_GATE(0x0414, 16, 0, 1),
+	.clocks		= CLOCKS("bbl_32k",
+				 "frac_1m",
+				 "dft_19_5m"),
+	.sel		= SELECTOR(0x0a10, 0, 2),
+	.trig		= TRIGGER(0x0a40, 4),
+};
+
+static struct peri_clk_data pmu_bsc_data = {
+	.gate		= HW_SW_GATE(0x0418, 16, 0, 1),
+	.clocks		= CLOCKS("ref_crystal",
+				 "pmu_bsc_var",
+				 "bbl_32k"),
+	.sel		= SELECTOR(0x0a04, 0, 2),
+	.div		= DIVIDER(0x0a04, 3, 4),
+	.trig		= TRIGGER(0x0a40, 0),
+};
+
+static struct peri_clk_data pmu_bsc_var_data = {
+	.clocks		= CLOCKS("var_312m",
+				 "ref_312m"),
+	.sel		= SELECTOR(0x0a00, 0, 2),
+	.div		= DIVIDER(0x0a00, 4, 5),
+	.trig		= TRIGGER(0x0a40, 2),
+};
+
+/* Hub CCU clocks */
+
+static struct peri_clk_data tmon_1m_data = {
+	.gate		= HW_SW_GATE(0x04a4, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "frac_1m"),
+	.sel		= SELECTOR(0x0e74, 0, 2),
+	.trig		= TRIGGER(0x0e84, 1),
+};
+
+/* Master CCU clocks */
+
+static struct peri_clk_data sdio1_data = {
+	.gate		= HW_SW_GATE(0x0358, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_52m",
+				 "ref_52m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a28, 0, 3),
+	.div		= DIVIDER(0x0a28, 4, 14),
+	.trig		= TRIGGER(0x0afc, 9),
+};
+
+static struct peri_clk_data sdio2_data = {
+	.gate		= HW_SW_GATE(0x035c, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_52m",
+				 "ref_52m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a2c, 0, 3),
+	.div		= DIVIDER(0x0a2c, 4, 14),
+	.trig		= TRIGGER(0x0afc, 10),
+};
+
+static struct peri_clk_data sdio3_data = {
+	.gate		= HW_SW_GATE(0x0364, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_52m",
+				 "ref_52m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a34, 0, 3),
+	.div		= DIVIDER(0x0a34, 4, 14),
+	.trig		= TRIGGER(0x0afc, 12),
+};
+
+static struct peri_clk_data sdio4_data = {
+	.gate		= HW_SW_GATE(0x0360, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_52m",
+				 "ref_52m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a30, 0, 3),
+	.div		= DIVIDER(0x0a30, 4, 14),
+	.trig		= TRIGGER(0x0afc, 11),
+};
+
+static struct peri_clk_data usb_ic_data = {
+	.gate		= HW_SW_GATE(0x0354, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_96m",
+				 "ref_96m"),
+	.div		= FIXED_DIVIDER(2),
+	.sel		= SELECTOR(0x0a24, 0, 2),
+	.trig		= TRIGGER(0x0afc, 7),
+};
+
+/* also called usbh_48m */
+static struct peri_clk_data hsic2_48m_data = {
+	.gate		= HW_SW_GATE(0x0370, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a38, 0, 2),
+	.div		= FIXED_DIVIDER(2),
+	.trig		= TRIGGER(0x0afc, 5),
+};
+
+/* also called usbh_12m */
+static struct peri_clk_data hsic2_12m_data = {
+	.gate		= HW_SW_GATE(0x0370, 20, 4, 5),
+	.div		= DIVIDER(0x0a38, 12, 2),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_96m",
+				 "ref_96m"),
+	.pre_div	= FIXED_DIVIDER(2),
+	.sel		= SELECTOR(0x0a38, 0, 2),
+	.trig		= TRIGGER(0x0afc, 5),
+};
+
+/* Slave CCU clocks */
+
+static struct peri_clk_data uartb_data = {
+	.gate		= HW_SW_GATE(0x0400, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_156m",
+				 "ref_156m"),
+	.sel		= SELECTOR(0x0a10, 0, 2),
+	.div		= FRAC_DIVIDER(0x0a10, 4, 12, 8),
+	.trig		= TRIGGER(0x0afc, 2),
+};
+
+static struct peri_clk_data uartb2_data = {
+	.gate		= HW_SW_GATE(0x0404, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_156m",
+				 "ref_156m"),
+	.sel		= SELECTOR(0x0a14, 0, 2),
+	.div		= FRAC_DIVIDER(0x0a14, 4, 12, 8),
+	.trig		= TRIGGER(0x0afc, 3),
+};
+
+static struct peri_clk_data uartb3_data = {
+	.gate		= HW_SW_GATE(0x0408, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_156m",
+				 "ref_156m"),
+	.sel		= SELECTOR(0x0a18, 0, 2),
+	.div		= FRAC_DIVIDER(0x0a18, 4, 12, 8),
+	.trig		= TRIGGER(0x0afc, 4),
+};
+
+static struct peri_clk_data uartb4_data = {
+	.gate		= HW_SW_GATE(0x0408, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_156m",
+				 "ref_156m"),
+	.sel		= SELECTOR(0x0a1c, 0, 2),
+	.div		= FRAC_DIVIDER(0x0a1c, 4, 12, 8),
+	.trig		= TRIGGER(0x0afc, 5),
+};
+
+static struct peri_clk_data ssp0_data = {
+	.gate		= HW_SW_GATE(0x0410, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_104m",
+				 "ref_104m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a20, 0, 3),
+	.div		= DIVIDER(0x0a20, 4, 14),
+	.trig		= TRIGGER(0x0afc, 6),
+};
+
+static struct peri_clk_data ssp2_data = {
+	.gate		= HW_SW_GATE(0x0418, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_104m",
+				 "ref_104m",
+				 "var_96m",
+				 "ref_96m"),
+	.sel		= SELECTOR(0x0a28, 0, 3),
+	.div		= DIVIDER(0x0a28, 4, 14),
+	.trig		= TRIGGER(0x0afc, 8),
+};
+
+static struct peri_clk_data bsc1_data = {
+	.gate		= HW_SW_GATE(0x0458, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_104m",
+				 "ref_104m",
+				 "var_13m",
+				 "ref_13m"),
+	.sel		= SELECTOR(0x0a64, 0, 3),
+	.trig		= TRIGGER(0x0afc, 23),
+};
+
+static struct peri_clk_data bsc2_data = {
+	.gate		= HW_SW_GATE(0x045c, 18, 2, 3),
+	.clocks	= CLOCKS("ref_crystal",
+				 "var_104m",
+				 "ref_104m",
+				 "var_13m",
+				 "ref_13m"),
+	.sel		= SELECTOR(0x0a68, 0, 3),
+	.trig		= TRIGGER(0x0afc, 24),
+};
+
+static struct peri_clk_data bsc3_data = {
+	.gate		= HW_SW_GATE(0x0484, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_104m",
+				 "ref_104m",
+				 "var_13m",
+				 "ref_13m"),
+	.sel		= SELECTOR(0x0a84, 0, 3),
+	.trig		= TRIGGER(0x0b00, 2),
+};
+
+static struct peri_clk_data pwm_data = {
+	.gate		= HW_SW_GATE(0x0468, 18, 2, 3),
+	.clocks		= CLOCKS("ref_crystal",
+				 "var_104m"),
+	.sel		= SELECTOR(0x0a70, 0, 2),
+	.div		= DIVIDER(0x0a70, 4, 3),
+	.trig		= TRIGGER(0x0afc, 15),
+};
+
+/*
+ * CCU setup routines
+ *
+ * These are called from kona_dt_ccu_setup() to initialize the array
+ * of clocks provided by the CCU.  Once allocated, the entries in
+ * the array are initialized by calling kona_clk_setup() with the
+ * initialization data for each clock.  They return 0 if successful
+ * or an error code otherwise.
+ */
+static int __init bcm281xx_root_ccu_clks_setup(struct ccu_data *ccu)
+{
+	struct clk **clks;
+	size_t count = BCM281XX_ROOT_CCU_CLOCK_COUNT;
+
+	clks = kzalloc(count * sizeof(*clks), GFP_KERNEL);
+	if (!clks) {
+		pr_err("%s: failed to allocate root clocks\n", __func__);
+		return -ENOMEM;
+	}
+	ccu->data.clks = clks;
+	ccu->data.clk_num = count;
+
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_ROOT_CCU_FRAC_1M, frac_1m);
+
+	return 0;
+}
+
+static int __init bcm281xx_aon_ccu_clks_setup(struct ccu_data *ccu)
+{
+	struct clk **clks;
+	size_t count = BCM281XX_AON_CCU_CLOCK_COUNT;
+
+	clks = kzalloc(count * sizeof(*clks), GFP_KERNEL);
+	if (!clks) {
+		pr_err("%s: failed to allocate aon clocks\n", __func__);
+		return -ENOMEM;
+	}
+	ccu->data.clks = clks;
+	ccu->data.clk_num = count;
+
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_AON_CCU_HUB_TIMER, hub_timer);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_AON_CCU_PMU_BSC, pmu_bsc);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_AON_CCU_PMU_BSC_VAR, pmu_bsc_var);
+
+	return 0;
+}
+
+static int __init bcm281xx_hub_ccu_clks_setup(struct ccu_data *ccu)
+{
+	struct clk **clks;
+	size_t count = BCM281XX_HUB_CCU_CLOCK_COUNT;
+
+	clks = kzalloc(count * sizeof(*clks), GFP_KERNEL);
+	if (!clks) {
+		pr_err("%s: failed to allocate hub clocks\n", __func__);
+		return -ENOMEM;
+	}
+	ccu->data.clks = clks;
+	ccu->data.clk_num = count;
+
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_HUB_CCU_TMON_1M, tmon_1m);
+
+	return 0;
+}
+
+static int __init bcm281xx_master_ccu_clks_setup(struct ccu_data *ccu)
+{
+	struct clk **clks;
+	size_t count = BCM281XX_MASTER_CCU_CLOCK_COUNT;
+
+	clks = kzalloc(count * sizeof(*clks), GFP_KERNEL);
+	if (!clks) {
+		pr_err("%s: failed to allocate master clocks\n", __func__);
+		return -ENOMEM;
+	}
+	ccu->data.clks = clks;
+	ccu->data.clk_num = count;
+
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_SDIO1, sdio1);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_SDIO2, sdio2);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_SDIO3, sdio3);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_SDIO4, sdio4);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_USB_IC, usb_ic);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_HSIC2_48M, hsic2_48m);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_MASTER_CCU_HSIC2_12M, hsic2_12m);
+
+	return 0;
+}
+
+static int __init bcm281xx_slave_ccu_clks_setup(struct ccu_data *ccu)
+{
+	struct clk **clks;
+	size_t count = BCM281XX_SLAVE_CCU_CLOCK_COUNT;
+
+	clks = kzalloc(count * sizeof(*clks), GFP_KERNEL);
+	if (!clks) {
+		pr_err("%s: failed to allocate slave clocks\n", __func__);
+		return -ENOMEM;
+	}
+	ccu->data.clks = clks;
+	ccu->data.clk_num = count;
+
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_UARTB, uartb);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_UARTB2, uartb2);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_UARTB3, uartb3);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_UARTB4, uartb4);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_SSP0, ssp0);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_SSP2, ssp2);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_BSC1, bsc1);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_BSC2, bsc2);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_BSC3, bsc3);
+	PERI_CLK_SETUP(clks, ccu, BCM281XX_SLAVE_CCU_PWM, pwm);
+
+	return 0;
+}
+
+/* Device tree match table callback functions */
+
+static void __init kona_dt_root_ccu_setup(struct device_node *node)
+{
+	kona_dt_ccu_setup(node, bcm281xx_root_ccu_clks_setup);
+}
+
+static void __init kona_dt_aon_ccu_setup(struct device_node *node)
+{
+	kona_dt_ccu_setup(node, bcm281xx_aon_ccu_clks_setup);
+}
+
+static void __init kona_dt_hub_ccu_setup(struct device_node *node)
+{
+	kona_dt_ccu_setup(node, bcm281xx_hub_ccu_clks_setup);
+}
+
+static void __init kona_dt_master_ccu_setup(struct device_node *node)
+{
+	kona_dt_ccu_setup(node, bcm281xx_master_ccu_clks_setup);
+}
+
+static void __init kona_dt_slave_ccu_setup(struct device_node *node)
+{
+	kona_dt_ccu_setup(node, bcm281xx_slave_ccu_clks_setup);
+}
+
+CLK_OF_DECLARE(bcm11351_root_ccu, BCM11351_DT_ROOT_CCU_COMPAT,
+			kona_dt_root_ccu_setup);
+CLK_OF_DECLARE(bcm11351_aon_ccu, BCM11351_DT_AON_CCU_COMPAT,
+			kona_dt_aon_ccu_setup);
+CLK_OF_DECLARE(bcm11351_hub_ccu, BCM11351_DT_HUB_CCU_COMPAT,
+			kona_dt_hub_ccu_setup);
+CLK_OF_DECLARE(bcm11351_master_ccu, BCM11351_DT_MASTER_CCU_COMPAT,
+			kona_dt_master_ccu_setup);
+CLK_OF_DECLARE(bcm11351_slave_ccu, BCM11351_DT_SLAVE_CCU_COMPAT,
+			kona_dt_slave_ccu_setup);
diff --git a/drivers/clk/bcm/clk-kona-setup.c b/drivers/clk/bcm/clk-kona-setup.c
new file mode 100644
index 0000000..c7607fe
--- /dev/null
+++ b/drivers/clk/bcm/clk-kona-setup.c
@@ -0,0 +1,769 @@
+/*
+ * Copyright (C) 2013 Broadcom Corporation
+ * Copyright 2013 Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/io.h>
+#include <linux/of_address.h>
+
+#include "clk-kona.h"
+
+/* These are used when a selector or trigger is found to be unneeded */
+#define selector_clear_exists(sel)	((sel)->width = 0)
+#define trigger_clear_exists(trig)	FLAG_CLEAR(trig, TRIG, EXISTS)
+
+LIST_HEAD(ccu_list);	/* The list of set up CCUs */
+
+/* Validity checking */
+
+static bool clk_requires_trigger(struct kona_clk *bcm_clk)
+{
+	struct peri_clk_data *peri = bcm_clk->peri;
+	struct bcm_clk_sel *sel;
+	struct bcm_clk_div *div;
+
+	if (bcm_clk->type != bcm_clk_peri)
+		return false;
+
+	sel = &peri->sel;
+	if (sel->parent_count && selector_exists(sel))
+		return true;
+
+	div = &peri->div;
+	if (!divider_exists(div))
+		return false;
+
+	/* Fixed dividers don't need triggers */
+	if (!divider_is_fixed(div))
+		return true;
+
+	div = &peri->pre_div;
+
+	return divider_exists(div) && !divider_is_fixed(div);
+}
+
+static bool peri_clk_data_offsets_valid(struct kona_clk *bcm_clk)
+{
+	struct peri_clk_data *peri;
+	struct bcm_clk_gate *gate;
+	struct bcm_clk_div *div;
+	struct bcm_clk_sel *sel;
+	struct bcm_clk_trig *trig;
+	const char *name;
+	u32 range;
+	u32 limit;
+
+	BUG_ON(bcm_clk->type != bcm_clk_peri);
+	peri = bcm_clk->peri;
+	name = bcm_clk->name;
+	range = bcm_clk->ccu->range;
+
+	limit = range - sizeof(u32);
+	limit = round_down(limit, sizeof(u32));
+
+	gate = &peri->gate;
+	if (gate_exists(gate)) {
+		if (gate->offset > limit) {
+			pr_err("%s: bad gate offset for %s (%u > %u)\n",
+				__func__, name, gate->offset, limit);
+			return false;
+		}
+	}
+
+	div = &peri->div;
+	if (divider_exists(div)) {
+		if (div->offset > limit) {
+			pr_err("%s: bad divider offset for %s (%u > %u)\n",
+				__func__, name, div->offset, limit);
+			return false;
+		}
+	}
+
+	div = &peri->pre_div;
+	if (divider_exists(div)) {
+		if (div->offset > limit) {
+			pr_err("%s: bad pre-divider offset for %s "
+					"(%u > %u)\n",
+				__func__, name, div->offset, limit);
+			return false;
+		}
+	}
+
+	sel = &peri->sel;
+	if (selector_exists(sel)) {
+		if (sel->offset > limit) {
+			pr_err("%s: bad selector offset for %s (%u > %u)\n",
+				__func__, name, sel->offset, limit);
+			return false;
+		}
+	}
+
+	trig = &peri->trig;
+	if (trigger_exists(trig)) {
+		if (trig->offset > limit) {
+			pr_err("%s: bad trigger offset for %s (%u > %u)\n",
+				__func__, name, trig->offset, limit);
+			return false;
+		}
+	}
+
+	trig = &peri->pre_trig;
+	if (trigger_exists(trig)) {
+		if (trig->offset > limit) {
+			pr_err("%s: bad pre-trigger offset for %s (%u > %u)\n",
+				__func__, name, trig->offset, limit);
+			return false;
+		}
+	}
+
+	return true;
+}
+
+/* A bit position must be less than the number of bits in a 32-bit register. */
+static bool bit_posn_valid(u32 bit_posn, const char *field_name,
+			const char *clock_name)
+{
+	u32 limit = BITS_PER_BYTE * sizeof(u32) - 1;
+
+	if (bit_posn > limit) {
+		pr_err("%s: bad %s bit for %s (%u > %u)\n", __func__,
+			field_name, clock_name, bit_posn, limit);
+		return false;
+	}
+	return true;
+}
+
+/*
+ * A bitfield must be at least 1 bit wide.  Both the low-order and
+ * high-order bits must lie within a 32-bit register.  We require
+ * fields to be less than 32 bits wide, mainly because we use
+ * shifting to produce field masks, and shifting a full word width
+ * is not well-defined by the C standard.
+ */
+static bool bitfield_valid(u32 shift, u32 width, const char *field_name,
+			const char *clock_name)
+{
+	u32 limit = BITS_PER_BYTE * sizeof(u32);
+
+	if (!width) {
+		pr_err("%s: bad %s field width 0 for %s\n", __func__,
+			field_name, clock_name);
+		return false;
+	}
+	if (shift + width > limit) {
+		pr_err("%s: bad %s for %s (%u + %u > %u)\n", __func__,
+			field_name, clock_name, shift, width, limit);
+		return false;
+	}
+	return true;
+}
+
+/*
+ * All gates, if defined, have a status bit, and for hardware-only
+ * gates, that's it.  Gates that can be software controlled also
+ * have an enable bit.  And a gate that can be hardware or software
+ * controlled will have a hardware/software select bit.
+ */
+static bool gate_valid(struct bcm_clk_gate *gate, const char *field_name,
+			const char *clock_name)
+{
+	if (!bit_posn_valid(gate->status_bit, "gate status", clock_name))
+		return false;
+
+	if (gate_is_sw_controllable(gate)) {
+		if (!bit_posn_valid(gate->en_bit, "gate enable", clock_name))
+			return false;
+
+		if (gate_is_hw_controllable(gate)) {
+			if (!bit_posn_valid(gate->hw_sw_sel_bit,
+						"gate hw/sw select",
+						clock_name))
+				return false;
+		}
+	} else {
+		BUG_ON(!gate_is_hw_controllable(gate));
+	}
+
+	return true;
+}
+
+/*
+ * A selector bitfield must be valid.  Its parent_sel array must
+ * also be reasonable for the field.
+ */
+static bool sel_valid(struct bcm_clk_sel *sel, const char *field_name,
+			const char *clock_name)
+{
+	if (!bitfield_valid(sel->shift, sel->width, field_name, clock_name))
+		return false;
+
+	if (sel->parent_count) {
+		u32 max_sel;
+		u32 limit;
+
+		/*
+		 * Make sure the selector field can hold all the
+		 * selector values we expect to be able to use.  A
+		 * clock only needs to have a selector defined if it
+		 * has more than one parent.  And in that case the
+		 * highest selector value will be in the last entry
+		 * in the array.
+		 */
+		max_sel = sel->parent_sel[sel->parent_count - 1];
+		limit = (1 << sel->width) - 1;
+		if (max_sel > limit) {
+			pr_err("%s: bad selector for %s "
+					"(%u needs > %u bits)\n",
+				__func__, clock_name, max_sel,
+				sel->width);
+			return false;
+		}
+	} else {
+		pr_warn("%s: ignoring selector for %s (no parents)\n",
+			__func__, clock_name);
+		selector_clear_exists(sel);
+		kfree(sel->parent_sel);
+		sel->parent_sel = NULL;
+	}
+
+	return true;
+}
+
+/*
+ * A fixed divider just needs to be non-zero.  A variable divider
+ * has to have a valid divider bitfield, and if it has a fraction,
+ * the width of the fraction must not be no more than the width of
+ * the divider as a whole.
+ */
+static bool div_valid(struct bcm_clk_div *div, const char *field_name,
+			const char *clock_name)
+{
+	if (divider_is_fixed(div)) {
+		/* Any fixed divider value but 0 is OK */
+		if (div->fixed == 0) {
+			pr_err("%s: bad %s fixed value 0 for %s\n", __func__,
+				field_name, clock_name);
+			return false;
+		}
+		return true;
+	}
+	if (!bitfield_valid(div->shift, div->width, field_name, clock_name))
+		return false;
+
+	if (divider_has_fraction(div))
+		if (div->frac_width > div->width) {
+			pr_warn("%s: bad %s fraction width for %s (%u > %u)\n",
+				__func__, field_name, clock_name,
+				div->frac_width, div->width);
+			return false;
+		}
+
+	return true;
+}
+
+/*
+ * If a clock has two dividers, the combined number of fractional
+ * bits must be representable in a 32-bit unsigned value.  This
+ * is because we scale up a dividend using both dividers before
+ * dividing to improve accuracy, and we need to avoid overflow.
+ */
+static bool kona_dividers_valid(struct kona_clk *bcm_clk)
+{
+	struct peri_clk_data *peri = bcm_clk->peri;
+	struct bcm_clk_div *div;
+	struct bcm_clk_div *pre_div;
+	u32 limit;
+
+	BUG_ON(bcm_clk->type != bcm_clk_peri);
+
+	if (!divider_exists(&peri->div) || !divider_exists(&peri->pre_div))
+		return true;
+
+	div = &peri->div;
+	pre_div = &peri->pre_div;
+	if (divider_is_fixed(div) || divider_is_fixed(pre_div))
+		return true;
+
+	limit = BITS_PER_BYTE * sizeof(u32);
+
+	return div->frac_width + pre_div->frac_width <= limit;
+}
+
+
+/* A trigger just needs to represent a valid bit position */
+static bool trig_valid(struct bcm_clk_trig *trig, const char *field_name,
+			const char *clock_name)
+{
+	return bit_posn_valid(trig->bit, field_name, clock_name);
+}
+
+/* Determine whether the set of peripheral clock registers are valid. */
+static bool
+peri_clk_data_valid(struct kona_clk *bcm_clk)
+{
+	struct peri_clk_data *peri;
+	struct bcm_clk_gate *gate;
+	struct bcm_clk_sel *sel;
+	struct bcm_clk_div *div;
+	struct bcm_clk_div *pre_div;
+	struct bcm_clk_trig *trig;
+	const char *name;
+
+	BUG_ON(bcm_clk->type != bcm_clk_peri);
+
+	/*
+	 * First validate register offsets.  This is the only place
+	 * where we need something from the ccu, so we do these
+	 * together.
+	 */
+	if (!peri_clk_data_offsets_valid(bcm_clk))
+		return false;
+
+	peri = bcm_clk->peri;
+	name = bcm_clk->name;
+	gate = &peri->gate;
+	if (gate_exists(gate) && !gate_valid(gate, "gate", name))
+		return false;
+
+	sel = &peri->sel;
+	if (selector_exists(sel)) {
+		if (!sel_valid(sel, "selector", name))
+			return false;
+
+	} else if (sel->parent_count > 1) {
+		pr_err("%s: multiple parents but no selector for %s\n",
+			__func__, name);
+
+		return false;
+	}
+
+	div = &peri->div;
+	pre_div = &peri->pre_div;
+	if (divider_exists(div)) {
+		if (!div_valid(div, "divider", name))
+			return false;
+
+		if (divider_exists(pre_div))
+			if (!div_valid(pre_div, "pre-divider", name))
+				return false;
+	} else if (divider_exists(pre_div)) {
+		pr_err("%s: pre-divider but no divider for %s\n", __func__,
+			name);
+		return false;
+	}
+
+	trig = &peri->trig;
+	if (trigger_exists(trig)) {
+		if (!trig_valid(trig, "trigger", name))
+			return false;
+
+		if (trigger_exists(&peri->pre_trig)) {
+			if (!trig_valid(trig, "pre-trigger", name)) {
+				return false;
+			}
+		}
+		if (!clk_requires_trigger(bcm_clk)) {
+			pr_warn("%s: ignoring trigger for %s (not needed)\n",
+				__func__, name);
+			trigger_clear_exists(trig);
+		}
+	} else if (trigger_exists(&peri->pre_trig)) {
+		pr_err("%s: pre-trigger but no trigger for %s\n", __func__,
+			name);
+		return false;
+	} else if (clk_requires_trigger(bcm_clk)) {
+		pr_err("%s: required trigger missing for %s\n", __func__,
+			name);
+		return false;
+	}
+
+	return kona_dividers_valid(bcm_clk);
+}
+
+static bool kona_clk_valid(struct kona_clk *bcm_clk)
+{
+	switch (bcm_clk->type) {
+	case bcm_clk_peri:
+		if (!peri_clk_data_valid(bcm_clk))
+			return false;
+		break;
+	default:
+		pr_err("%s: unrecognized clock type (%d)\n", __func__,
+			(int)bcm_clk->type);
+		return false;
+	}
+	return true;
+}
+
+/*
+ * Scan an array of parent clock names to determine whether there
+ * are any entries containing BAD_CLK_NAME.  Such entries are
+ * placeholders for non-supported clocks.  Keep track of the
+ * position of each clock name in the original array.
+ *
+ * Allocates an array of pointers to to hold the names of all
+ * non-null entries in the original array, and returns a pointer to
+ * that array in *names.  This will be used for registering the
+ * clock with the common clock code.  On successful return,
+ * *count indicates how many entries are in that names array.
+ *
+ * If there is more than one entry in the resulting names array,
+ * another array is allocated to record the parent selector value
+ * for each (defined) parent clock.  This is the value that
+ * represents this parent clock in the clock's source selector
+ * register.  The position of the clock in the original parent array
+ * defines that selector value.  The number of entries in this array
+ * is the same as the number of entries in the parent names array.
+ *
+ * The array of selector values is returned.  If the clock has no
+ * parents, no selector is required and a null pointer is returned.
+ *
+ * Returns a null pointer if the clock names array supplied was
+ * null.  (This is not an error.)
+ *
+ * Returns a pointer-coded error if an error occurs.
+ */
+static u32 *parent_process(const char *clocks[],
+			u32 *count, const char ***names)
+{
+	static const char **parent_names;
+	static u32 *parent_sel;
+	const char **clock;
+	u32 parent_count;
+	u32 bad_count = 0;
+	u32 orig_count;
+	u32 i;
+	u32 j;
+
+	*count = 0;	/* In case of early return */
+	*names = NULL;
+	if (!clocks)
+		return NULL;
+
+	/*
+	 * Count the number of names in the null-terminated array,
+	 * and find out how many of those are actually clock names.
+	 */
+	for (clock = clocks; *clock; clock++)
+		if (*clock == BAD_CLK_NAME)
+			bad_count++;
+	orig_count = (u32)(clock - clocks);
+	parent_count = orig_count - bad_count;
+
+	/* If all clocks are unsupported, we treat it as no clock */
+	if (!parent_count)
+		return NULL;
+
+	/* Avoid exceeding our parent clock limit */
+	if (parent_count > PARENT_COUNT_MAX) {
+		pr_err("%s: too many parents (%u > %u)\n", __func__,
+			parent_count, PARENT_COUNT_MAX);
+		return ERR_PTR(-EINVAL);
+	}
+
+	/*
+	 * There is one parent name for each defined parent clock.
+	 * We also maintain an array containing the selector value
+	 * for each defined clock.  If there's only one clock, the
+	 * selector is not required, but we allocate space for the
+	 * array anyway to keep things simple.
+	 */
+	parent_names = kmalloc(parent_count * sizeof(parent_names), GFP_KERNEL);
+	if (!parent_names) {
+		pr_err("%s: error allocating %u parent names\n", __func__,
+				parent_count);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/* There is at least one parent, so allocate a selector array */
+
+	parent_sel = kmalloc(parent_count * sizeof(*parent_sel), GFP_KERNEL);
+	if (!parent_sel) {
+		pr_err("%s: error allocating %u parent selectors\n", __func__,
+				parent_count);
+		kfree(parent_names);
+
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/* Now fill in the parent names and selector arrays */
+	for (i = 0, j = 0; i < orig_count; i++) {
+		if (clocks[i] != BAD_CLK_NAME) {
+			parent_names[j] = clocks[i];
+			parent_sel[j] = i;
+			j++;
+		}
+	}
+	*names = parent_names;
+	*count = parent_count;
+
+	return parent_sel;
+}
+
+static int
+clk_sel_setup(const char **clocks, struct bcm_clk_sel *sel,
+		struct clk_init_data *init_data)
+{
+	const char **parent_names = NULL;
+	u32 parent_count = 0;
+	u32 *parent_sel;
+
+	/*
+	 * If a peripheral clock has multiple parents, the value
+	 * used by the hardware to select that parent is represented
+	 * by the parent clock's position in the "clocks" list.  Some
+	 * values don't have defined or supported clocks; these will
+	 * have BAD_CLK_NAME entries in the parents[] array.  The
+	 * list is terminated by a NULL entry.
+	 *
+	 * We need to supply (only) the names of defined parent
+	 * clocks when registering a clock though, so we use an
+	 * array of parent selector values to map between the
+	 * indexes the common clock code uses and the selector
+	 * values we need.
+	 */
+	parent_sel = parent_process(clocks, &parent_count, &parent_names);
+	if (IS_ERR(parent_sel)) {
+		int ret = PTR_ERR(parent_sel);
+
+		pr_err("%s: error processing parent clocks for %s (%d)\n",
+			__func__, init_data->name, ret);
+
+		return ret;
+	}
+
+	init_data->parent_names = parent_names;
+	init_data->num_parents = parent_count;
+
+	sel->parent_count = parent_count;
+	sel->parent_sel = parent_sel;
+
+	return 0;
+}
+
+static void clk_sel_teardown(struct bcm_clk_sel *sel,
+		struct clk_init_data *init_data)
+{
+	kfree(sel->parent_sel);
+	sel->parent_sel = NULL;
+	sel->parent_count = 0;
+
+	init_data->num_parents = 0;
+	kfree(init_data->parent_names);
+	init_data->parent_names = NULL;
+}
+
+static void peri_clk_teardown(struct peri_clk_data *data,
+				struct clk_init_data *init_data)
+{
+	clk_sel_teardown(&data->sel, init_data);
+	init_data->ops = NULL;
+}
+
+/*
+ * Caller is responsible for freeing the parent_names[] and
+ * parent_sel[] arrays in the peripheral clock's "data" structure
+ * that can be assigned if the clock has one or more parent clocks
+ * associated with it.
+ */
+static int peri_clk_setup(struct ccu_data *ccu, struct peri_clk_data *data,
+			struct clk_init_data *init_data)
+{
+	init_data->ops = &kona_peri_clk_ops;
+	init_data->flags = CLK_IGNORE_UNUSED;
+
+	return clk_sel_setup(data->clocks, &data->sel, init_data);
+}
+
+static void bcm_clk_teardown(struct kona_clk *bcm_clk)
+{
+	switch (bcm_clk->type) {
+	case bcm_clk_peri:
+		peri_clk_teardown(bcm_clk->data, &bcm_clk->init_data);
+		break;
+	default:
+		break;
+	}
+	bcm_clk->data = NULL;
+	bcm_clk->type = bcm_clk_none;
+}
+
+static void kona_clk_teardown(struct clk *clk)
+{
+	struct clk_hw *hw;
+	struct kona_clk *bcm_clk;
+
+	if (!clk)
+		return;
+
+	hw = __clk_get_hw(clk);
+	if (!hw) {
+		pr_err("%s: clk %p has null hw pointer\n", __func__, clk);
+		return;
+	}
+	clk_unregister(clk);
+
+	bcm_clk = to_kona_clk(hw);
+	bcm_clk_teardown(bcm_clk);
+}
+
+struct clk *kona_clk_setup(struct ccu_data *ccu, const char *name,
+			enum bcm_clk_type type, void *data)
+{
+	struct kona_clk *bcm_clk;
+	struct clk_init_data *init_data;
+	struct clk *clk = NULL;
+
+	bcm_clk = kzalloc(sizeof(*bcm_clk), GFP_KERNEL);
+	if (!bcm_clk) {
+		pr_err("%s: failed to allocate bcm_clk for %s\n", __func__,
+			name);
+		return NULL;
+	}
+	bcm_clk->ccu = ccu;
+	bcm_clk->name = name;
+
+	init_data = &bcm_clk->init_data;
+	init_data->name = name;
+	switch (type) {
+	case bcm_clk_peri:
+		if (peri_clk_setup(ccu, data, init_data))
+			goto out_free;
+		break;
+	default:
+		data = NULL;
+		break;
+	}
+	bcm_clk->type = type;
+	bcm_clk->data = data;
+
+	/* Make sure everything makes sense before we set it up */
+	if (!kona_clk_valid(bcm_clk)) {
+		pr_err("%s: clock data invalid for %s\n", __func__, name);
+		goto out_teardown;
+	}
+
+	bcm_clk->hw.init = init_data;
+	clk = clk_register(NULL, &bcm_clk->hw);
+	if (IS_ERR(clk)) {
+		pr_err("%s: error registering clock %s (%ld)\n", __func__,
+				name, PTR_ERR(clk));
+		goto out_teardown;
+	}
+	BUG_ON(!clk);
+
+	return clk;
+out_teardown:
+	bcm_clk_teardown(bcm_clk);
+out_free:
+	kfree(bcm_clk);
+
+	return NULL;
+}
+
+static void ccu_clks_teardown(struct ccu_data *ccu)
+{
+	u32 i;
+
+	for (i = 0; i < ccu->data.clk_num; i++)
+		kona_clk_teardown(ccu->data.clks[i]);
+	kfree(ccu->data.clks);
+}
+
+static void kona_ccu_teardown(struct ccu_data *ccu)
+{
+	if (!ccu)
+		return;
+
+	if (!ccu->base)
+		goto done;
+
+	of_clk_del_provider(ccu->node);	/* safe if never added */
+	ccu_clks_teardown(ccu);
+	list_del(&ccu->links);
+	of_node_put(ccu->node);
+	iounmap(ccu->base);
+done:
+	kfree(ccu->name);
+	kfree(ccu);
+}
+
+/*
+ * Set up a CCU.  Call the provided ccu_clks_setup callback to
+ * initialize the array of clocks provided by the CCU.
+ */
+void __init kona_dt_ccu_setup(struct device_node *node,
+			int (*ccu_clks_setup)(struct ccu_data *))
+{
+	struct ccu_data *ccu;
+	struct resource res = { 0 };
+	resource_size_t range;
+	int ret;
+
+	ccu = kzalloc(sizeof(*ccu), GFP_KERNEL);
+	if (ccu)
+		ccu->name = kstrdup(node->name, GFP_KERNEL);
+	if (!ccu || !ccu->name) {
+		pr_err("%s: unable to allocate CCU struct for %s\n",
+			__func__, node->name);
+		kfree(ccu);
+
+		return;
+	}
+
+	ret = of_address_to_resource(node, 0, &res);
+	if (ret) {
+		pr_err("%s: no valid CCU registers found for %s\n", __func__,
+			node->name);
+		goto out_err;
+	}
+
+	range = resource_size(&res);
+	if (range > (resource_size_t)U32_MAX) {
+		pr_err("%s: address range too large for %s\n", __func__,
+			node->name);
+		goto out_err;
+	}
+
+	ccu->range = (u32)range;
+	ccu->base = ioremap(res.start, ccu->range);
+	if (!ccu->base) {
+		pr_err("%s: unable to map CCU registers for %s\n", __func__,
+			node->name);
+		goto out_err;
+	}
+
+	spin_lock_init(&ccu->lock);
+	INIT_LIST_HEAD(&ccu->links);
+	ccu->node = of_node_get(node);
+
+	list_add_tail(&ccu->links, &ccu_list);
+
+	/* Set up clocks array (in ccu->data) */
+	if (ccu_clks_setup(ccu))
+		goto out_err;
+
+	ret = of_clk_add_provider(node, of_clk_src_onecell_get, &ccu->data);
+	if (ret) {
+		pr_err("%s: error adding ccu %s as provider (%d)\n", __func__,
+				node->name, ret);
+		goto out_err;
+	}
+
+	if (!kona_ccu_init(ccu))
+		pr_err("Broadcom %s initialization had errors\n", node->name);
+
+	return;
+out_err:
+	kona_ccu_teardown(ccu);
+	pr_err("Broadcom %s setup aborted\n", node->name);
+}
diff --git a/drivers/clk/bcm/clk-kona.c b/drivers/clk/bcm/clk-kona.c
new file mode 100644
index 0000000..e3d339e
--- /dev/null
+++ b/drivers/clk/bcm/clk-kona.c
@@ -0,0 +1,1033 @@
+/*
+ * Copyright (C) 2013 Broadcom Corporation
+ * Copyright 2013 Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include "clk-kona.h"
+
+#include <linux/delay.h>
+
+#define CCU_ACCESS_PASSWORD      0xA5A500
+#define CLK_GATE_DELAY_LOOP      2000
+
+/* Bitfield operations */
+
+/* Produces a mask of set bits covering a range of a 32-bit value */
+static inline u32 bitfield_mask(u32 shift, u32 width)
+{
+	return ((1 << width) - 1) << shift;
+}
+
+/* Extract the value of a bitfield found within a given register value */
+static inline u32 bitfield_extract(u32 reg_val, u32 shift, u32 width)
+{
+	return (reg_val & bitfield_mask(shift, width)) >> shift;
+}
+
+/* Replace the value of a bitfield found within a given register value */
+static inline u32 bitfield_replace(u32 reg_val, u32 shift, u32 width, u32 val)
+{
+	u32 mask = bitfield_mask(shift, width);
+
+	return (reg_val & ~mask) | (val << shift);
+}
+
+/* Divider and scaling helpers */
+
+/*
+ * Implement DIV_ROUND_CLOSEST() for 64-bit dividend and both values
+ * unsigned.  Note that unlike do_div(), the remainder is discarded
+ * and the return value is the quotient (not the remainder).
+ */
+u64 do_div_round_closest(u64 dividend, unsigned long divisor)
+{
+	u64 result;
+
+	result = dividend + ((u64)divisor >> 1);
+	(void)do_div(result, divisor);
+
+	return result;
+}
+
+/* Convert a divider into the scaled divisor value it represents. */
+static inline u64 scaled_div_value(struct bcm_clk_div *div, u32 reg_div)
+{
+	return (u64)reg_div + ((u64)1 << div->frac_width);
+}
+
+/*
+ * Build a scaled divider value as close as possible to the
+ * given whole part (div_value) and fractional part (expressed
+ * in billionths).
+ */
+u64 scaled_div_build(struct bcm_clk_div *div, u32 div_value, u32 billionths)
+{
+	u64 combined;
+
+	BUG_ON(!div_value);
+	BUG_ON(billionths >= BILLION);
+
+	combined = (u64)div_value * BILLION + billionths;
+	combined <<= div->frac_width;
+
+	return do_div_round_closest(combined, BILLION);
+}
+
+/* The scaled minimum divisor representable by a divider */
+static inline u64
+scaled_div_min(struct bcm_clk_div *div)
+{
+	if (divider_is_fixed(div))
+		return (u64)div->fixed;
+
+	return scaled_div_value(div, 0);
+}
+
+/* The scaled maximum divisor representable by a divider */
+u64 scaled_div_max(struct bcm_clk_div *div)
+{
+	u32 reg_div;
+
+	if (divider_is_fixed(div))
+		return (u64)div->fixed;
+
+	reg_div = ((u32)1 << div->width) - 1;
+
+	return scaled_div_value(div, reg_div);
+}
+
+/*
+ * Convert a scaled divisor into its divider representation as
+ * stored in a divider register field.
+ */
+static inline u32
+divider(struct bcm_clk_div *div, u64 scaled_div)
+{
+	BUG_ON(scaled_div < scaled_div_min(div));
+	BUG_ON(scaled_div > scaled_div_max(div));
+
+	return (u32)(scaled_div - ((u64)1 << div->frac_width));
+}
+
+/* Return a rate scaled for use when dividing by a scaled divisor. */
+static inline u64
+scale_rate(struct bcm_clk_div *div, u32 rate)
+{
+	if (divider_is_fixed(div))
+		return (u64)rate;
+
+	return (u64)rate << div->frac_width;
+}
+
+/* CCU access */
+
+/* Read a 32-bit register value from a CCU's address space. */
+static inline u32 __ccu_read(struct ccu_data *ccu, u32 reg_offset)
+{
+	return readl(ccu->base + reg_offset);
+}
+
+/* Write a 32-bit register value into a CCU's address space. */
+static inline void
+__ccu_write(struct ccu_data *ccu, u32 reg_offset, u32 reg_val)
+{
+	writel(reg_val, ccu->base + reg_offset);
+}
+
+static inline unsigned long ccu_lock(struct ccu_data *ccu)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ccu->lock, flags);
+
+	return flags;
+}
+static inline void ccu_unlock(struct ccu_data *ccu, unsigned long flags)
+{
+	spin_unlock_irqrestore(&ccu->lock, flags);
+}
+
+/*
+ * Enable/disable write access to CCU protected registers.  The
+ * WR_ACCESS register for all CCUs is at offset 0.
+ */
+static inline void __ccu_write_enable(struct ccu_data *ccu)
+{
+	if (ccu->write_enabled) {
+		pr_err("%s: access already enabled for %s\n", __func__,
+			ccu->name);
+		return;
+	}
+	ccu->write_enabled = true;
+	__ccu_write(ccu, 0, CCU_ACCESS_PASSWORD | 1);
+}
+
+static inline void __ccu_write_disable(struct ccu_data *ccu)
+{
+	if (!ccu->write_enabled) {
+		pr_err("%s: access wasn't enabled for %s\n", __func__,
+			ccu->name);
+		return;
+	}
+
+	__ccu_write(ccu, 0, CCU_ACCESS_PASSWORD);
+	ccu->write_enabled = false;
+}
+
+/*
+ * Poll a register in a CCU's address space, returning when the
+ * specified bit in that register's value is set (or clear).  Delay
+ * a microsecond after each read of the register.  Returns true if
+ * successful, or false if we gave up trying.
+ *
+ * Caller must ensure the CCU lock is held.
+ */
+static inline bool
+__ccu_wait_bit(struct ccu_data *ccu, u32 reg_offset, u32 bit, bool want)
+{
+	unsigned int tries;
+	u32 bit_mask = 1 << bit;
+
+	for (tries = 0; tries < CLK_GATE_DELAY_LOOP; tries++) {
+		u32 val;
+		bool bit_val;
+
+		val = __ccu_read(ccu, reg_offset);
+		bit_val = (val & bit_mask) != 0;
+		if (bit_val == want)
+			return true;
+		udelay(1);
+	}
+	return false;
+}
+
+/* Gate operations */
+
+/* Determine whether a clock is gated.  CCU lock must be held.  */
+static bool
+__is_clk_gate_enabled(struct ccu_data *ccu, struct bcm_clk_gate *gate)
+{
+	u32 bit_mask;
+	u32 reg_val;
+
+	/* If there is no gate we can assume it's enabled. */
+	if (!gate_exists(gate))
+		return true;
+
+	bit_mask = 1 << gate->status_bit;
+	reg_val = __ccu_read(ccu, gate->offset);
+
+	return (reg_val & bit_mask) != 0;
+}
+
+/* Determine whether a clock is gated. */
+static bool
+is_clk_gate_enabled(struct ccu_data *ccu, struct bcm_clk_gate *gate)
+{
+	long flags;
+	bool ret;
+
+	/* Avoid taking the lock if we can */
+	if (!gate_exists(gate))
+		return true;
+
+	flags = ccu_lock(ccu);
+	ret = __is_clk_gate_enabled(ccu, gate);
+	ccu_unlock(ccu, flags);
+
+	return ret;
+}
+
+/*
+ * Commit our desired gate state to the hardware.
+ * Returns true if successful, false otherwise.
+ */
+static bool
+__gate_commit(struct ccu_data *ccu, struct bcm_clk_gate *gate)
+{
+	u32 reg_val;
+	u32 mask;
+	bool enabled = false;
+
+	BUG_ON(!gate_exists(gate));
+	if (!gate_is_sw_controllable(gate))
+		return true;		/* Nothing we can change */
+
+	reg_val = __ccu_read(ccu, gate->offset);
+
+	/* For a hardware/software gate, set which is in control */
+	if (gate_is_hw_controllable(gate)) {
+		mask = (u32)1 << gate->hw_sw_sel_bit;
+		if (gate_is_sw_managed(gate))
+			reg_val |= mask;
+		else
+			reg_val &= ~mask;
+	}
+
+	/*
+	 * If software is in control, enable or disable the gate.
+	 * If hardware is, clear the enabled bit for good measure.
+	 * If a software controlled gate can't be disabled, we're
+	 * required to write a 0 into the enable bit (but the gate
+	 * will be enabled).
+	 */
+	mask = (u32)1 << gate->en_bit;
+	if (gate_is_sw_managed(gate) && (enabled = gate_is_enabled(gate)) &&
+			!gate_is_no_disable(gate))
+		reg_val |= mask;
+	else
+		reg_val &= ~mask;
+
+	__ccu_write(ccu, gate->offset, reg_val);
+
+	/* For a hardware controlled gate, we're done */
+	if (!gate_is_sw_managed(gate))
+		return true;
+
+	/* Otherwise wait for the gate to be in desired state */
+	return __ccu_wait_bit(ccu, gate->offset, gate->status_bit, enabled);
+}
+
+/*
+ * Initialize a gate.  Our desired state (hardware/software select,
+ * and if software, its enable state) is committed to hardware
+ * without the usual checks to see if it's already set up that way.
+ * Returns true if successful, false otherwise.
+ */
+static bool gate_init(struct ccu_data *ccu, struct bcm_clk_gate *gate)
+{
+	if (!gate_exists(gate))
+		return true;
+	return __gate_commit(ccu, gate);
+}
+
+/*
+ * Set a gate to enabled or disabled state.  Does nothing if the
+ * gate is not currently under software control, or if it is already
+ * in the requested state.  Returns true if successful, false
+ * otherwise.  CCU lock must be held.
+ */
+static bool
+__clk_gate(struct ccu_data *ccu, struct bcm_clk_gate *gate, bool enable)
+{
+	bool ret;
+
+	if (!gate_exists(gate) || !gate_is_sw_managed(gate))
+		return true;	/* Nothing to do */
+
+	if (!enable && gate_is_no_disable(gate)) {
+		pr_warn("%s: invalid gate disable request (ignoring)\n",
+			__func__);
+		return true;
+	}
+
+	if (enable == gate_is_enabled(gate))
+		return true;	/* No change */
+
+	gate_flip_enabled(gate);
+	ret = __gate_commit(ccu, gate);
+	if (!ret)
+		gate_flip_enabled(gate);	/* Revert the change */
+
+	return ret;
+}
+
+/* Enable or disable a gate.  Returns 0 if successful, -EIO otherwise */
+static int clk_gate(struct ccu_data *ccu, const char *name,
+			struct bcm_clk_gate *gate, bool enable)
+{
+	unsigned long flags;
+	bool success;
+
+	/*
+	 * Avoid taking the lock if we can.  We quietly ignore
+	 * requests to change state that don't make sense.
+	 */
+	if (!gate_exists(gate) || !gate_is_sw_managed(gate))
+		return 0;
+	if (!enable && gate_is_no_disable(gate))
+		return 0;
+
+	flags = ccu_lock(ccu);
+	__ccu_write_enable(ccu);
+
+	success = __clk_gate(ccu, gate, enable);
+
+	__ccu_write_disable(ccu);
+	ccu_unlock(ccu, flags);
+
+	if (success)
+		return 0;
+
+	pr_err("%s: failed to %s gate for %s\n", __func__,
+		enable ? "enable" : "disable", name);
+
+	return -EIO;
+}
+
+/* Trigger operations */
+
+/*
+ * Caller must ensure CCU lock is held and access is enabled.
+ * Returns true if successful, false otherwise.
+ */
+static bool __clk_trigger(struct ccu_data *ccu, struct bcm_clk_trig *trig)
+{
+	/* Trigger the clock and wait for it to finish */
+	__ccu_write(ccu, trig->offset, 1 << trig->bit);
+
+	return __ccu_wait_bit(ccu, trig->offset, trig->bit, false);
+}
+
+/* Divider operations */
+
+/* Read a divider value and return the scaled divisor it represents. */
+static u64 divider_read_scaled(struct ccu_data *ccu, struct bcm_clk_div *div)
+{
+	unsigned long flags;
+	u32 reg_val;
+	u32 reg_div;
+
+	if (divider_is_fixed(div))
+		return (u64)div->fixed;
+
+	flags = ccu_lock(ccu);
+	reg_val = __ccu_read(ccu, div->offset);
+	ccu_unlock(ccu, flags);
+
+	/* Extract the full divider field from the register value */
+	reg_div = bitfield_extract(reg_val, div->shift, div->width);
+
+	/* Return the scaled divisor value it represents */
+	return scaled_div_value(div, reg_div);
+}
+
+/*
+ * Convert a divider's scaled divisor value into its recorded form
+ * and commit it into the hardware divider register.
+ *
+ * Returns 0 on success.  Returns -EINVAL for invalid arguments.
+ * Returns -ENXIO if gating failed, and -EIO if a trigger failed.
+ */
+static int __div_commit(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_div *div, struct bcm_clk_trig *trig)
+{
+	bool enabled;
+	u32 reg_div;
+	u32 reg_val;
+	int ret = 0;
+
+	BUG_ON(divider_is_fixed(div));
+
+	/*
+	 * If we're just initializing the divider, and no initial
+	 * state was defined in the device tree, we just find out
+	 * what its current value is rather than updating it.
+	 */
+	if (div->scaled_div == BAD_SCALED_DIV_VALUE) {
+		reg_val = __ccu_read(ccu, div->offset);
+		reg_div = bitfield_extract(reg_val, div->shift, div->width);
+		div->scaled_div = scaled_div_value(div, reg_div);
+
+		return 0;
+	}
+
+	/* Convert the scaled divisor to the value we need to record */
+	reg_div = divider(div, div->scaled_div);
+
+	/* Clock needs to be enabled before changing the rate */
+	enabled = __is_clk_gate_enabled(ccu, gate);
+	if (!enabled && !__clk_gate(ccu, gate, true)) {
+		ret = -ENXIO;
+		goto out;
+	}
+
+	/* Replace the divider value and record the result */
+	reg_val = __ccu_read(ccu, div->offset);
+	reg_val = bitfield_replace(reg_val, div->shift, div->width, reg_div);
+	__ccu_write(ccu, div->offset, reg_val);
+
+	/* If the trigger fails we still want to disable the gate */
+	if (!__clk_trigger(ccu, trig))
+		ret = -EIO;
+
+	/* Disable the clock again if it was disabled to begin with */
+	if (!enabled && !__clk_gate(ccu, gate, false))
+		ret = ret ? ret : -ENXIO;	/* return first error */
+out:
+	return ret;
+}
+
+/*
+ * Initialize a divider by committing our desired state to hardware
+ * without the usual checks to see if it's already set up that way.
+ * Returns true if successful, false otherwise.
+ */
+static bool div_init(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_div *div, struct bcm_clk_trig *trig)
+{
+	if (!divider_exists(div) || divider_is_fixed(div))
+		return true;
+	return !__div_commit(ccu, gate, div, trig);
+}
+
+static int divider_write(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_div *div, struct bcm_clk_trig *trig,
+			u64 scaled_div)
+{
+	unsigned long flags;
+	u64 previous;
+	int ret;
+
+	BUG_ON(divider_is_fixed(div));
+
+	previous = div->scaled_div;
+	if (previous == scaled_div)
+		return 0;	/* No change */
+
+	div->scaled_div = scaled_div;
+
+	flags = ccu_lock(ccu);
+	__ccu_write_enable(ccu);
+
+	ret = __div_commit(ccu, gate, div, trig);
+
+	__ccu_write_disable(ccu);
+	ccu_unlock(ccu, flags);
+
+	if (ret)
+		div->scaled_div = previous;		/* Revert the change */
+
+	return ret;
+
+}
+
+/* Common clock rate helpers */
+
+/*
+ * Implement the common clock framework recalc_rate method, taking
+ * into account a divider and an optional pre-divider.  The
+ * pre-divider register pointer may be NULL.
+ */
+static unsigned long clk_recalc_rate(struct ccu_data *ccu,
+			struct bcm_clk_div *div, struct bcm_clk_div *pre_div,
+			unsigned long parent_rate)
+{
+	u64 scaled_parent_rate;
+	u64 scaled_div;
+	u64 result;
+
+	if (!divider_exists(div))
+		return parent_rate;
+
+	if (parent_rate > (unsigned long)LONG_MAX)
+		return 0;	/* actually this would be a caller bug */
+
+	/*
+	 * If there is a pre-divider, divide the scaled parent rate
+	 * by the pre-divider value first.  In this case--to improve
+	 * accuracy--scale the parent rate by *both* the pre-divider
+	 * value and the divider before actually computing the
+	 * result of the pre-divider.
+	 *
+	 * If there's only one divider, just scale the parent rate.
+	 */
+	if (pre_div && divider_exists(pre_div)) {
+		u64 scaled_rate;
+
+		scaled_rate = scale_rate(pre_div, parent_rate);
+		scaled_rate = scale_rate(div, scaled_rate);
+		scaled_div = divider_read_scaled(ccu, pre_div);
+		scaled_parent_rate = do_div_round_closest(scaled_rate,
+							scaled_div);
+	} else  {
+		scaled_parent_rate = scale_rate(div, parent_rate);
+	}
+
+	/*
+	 * Get the scaled divisor value, and divide the scaled
+	 * parent rate by that to determine this clock's resulting
+	 * rate.
+	 */
+	scaled_div = divider_read_scaled(ccu, div);
+	result = do_div_round_closest(scaled_parent_rate, scaled_div);
+
+	return (unsigned long)result;
+}
+
+/*
+ * Compute the output rate produced when a given parent rate is fed
+ * into two dividers.  The pre-divider can be NULL, and even if it's
+ * non-null it may be nonexistent.  It's also OK for the divider to
+ * be nonexistent, and in that case the pre-divider is also ignored.
+ *
+ * If scaled_div is non-null, it is used to return the scaled divisor
+ * value used by the (downstream) divider to produce that rate.
+ */
+static long round_rate(struct ccu_data *ccu, struct bcm_clk_div *div,
+				struct bcm_clk_div *pre_div,
+				unsigned long rate, unsigned long parent_rate,
+				u64 *scaled_div)
+{
+	u64 scaled_parent_rate;
+	u64 min_scaled_div;
+	u64 max_scaled_div;
+	u64 best_scaled_div;
+	u64 result;
+
+	BUG_ON(!divider_exists(div));
+	BUG_ON(!rate);
+	BUG_ON(parent_rate > (u64)LONG_MAX);
+
+	/*
+	 * If there is a pre-divider, divide the scaled parent rate
+	 * by the pre-divider value first.  In this case--to improve
+	 * accuracy--scale the parent rate by *both* the pre-divider
+	 * value and the divider before actually computing the
+	 * result of the pre-divider.
+	 *
+	 * If there's only one divider, just scale the parent rate.
+	 *
+	 * For simplicity we treat the pre-divider as fixed (for now).
+	 */
+	if (divider_exists(pre_div)) {
+		u64 scaled_rate;
+		u64 scaled_pre_div;
+
+		scaled_rate = scale_rate(pre_div, parent_rate);
+		scaled_rate = scale_rate(div, scaled_rate);
+		scaled_pre_div = divider_read_scaled(ccu, pre_div);
+		scaled_parent_rate = do_div_round_closest(scaled_rate,
+							scaled_pre_div);
+	} else {
+		scaled_parent_rate = scale_rate(div, parent_rate);
+	}
+
+	/*
+	 * Compute the best possible divider and ensure it is in
+	 * range.  A fixed divider can't be changed, so just report
+	 * the best we can do.
+	 */
+	if (!divider_is_fixed(div)) {
+		best_scaled_div = do_div_round_closest(scaled_parent_rate,
+							rate);
+		min_scaled_div = scaled_div_min(div);
+		max_scaled_div = scaled_div_max(div);
+		if (best_scaled_div > max_scaled_div)
+			best_scaled_div = max_scaled_div;
+		else if (best_scaled_div < min_scaled_div)
+			best_scaled_div = min_scaled_div;
+	} else {
+		best_scaled_div = divider_read_scaled(ccu, div);
+	}
+
+	/* OK, figure out the resulting rate */
+	result = do_div_round_closest(scaled_parent_rate, best_scaled_div);
+
+	if (scaled_div)
+		*scaled_div = best_scaled_div;
+
+	return (long)result;
+}
+
+/* Common clock parent helpers */
+
+/*
+ * For a given parent selector (register field) value, find the
+ * index into a selector's parent_sel array that contains it.
+ * Returns the index, or BAD_CLK_INDEX if it's not found.
+ */
+static u8 parent_index(struct bcm_clk_sel *sel, u8 parent_sel)
+{
+	u8 i;
+
+	BUG_ON(sel->parent_count > (u32)U8_MAX);
+	for (i = 0; i < sel->parent_count; i++)
+		if (sel->parent_sel[i] == parent_sel)
+			return i;
+	return BAD_CLK_INDEX;
+}
+
+/*
+ * Fetch the current value of the selector, and translate that into
+ * its corresponding index in the parent array we registered with
+ * the clock framework.
+ *
+ * Returns parent array index that corresponds with the value found,
+ * or BAD_CLK_INDEX if the found value is out of range.
+ */
+static u8 selector_read_index(struct ccu_data *ccu, struct bcm_clk_sel *sel)
+{
+	unsigned long flags;
+	u32 reg_val;
+	u32 parent_sel;
+	u8 index;
+
+	/* If there's no selector, there's only one parent */
+	if (!selector_exists(sel))
+		return 0;
+
+	/* Get the value in the selector register */
+	flags = ccu_lock(ccu);
+	reg_val = __ccu_read(ccu, sel->offset);
+	ccu_unlock(ccu, flags);
+
+	parent_sel = bitfield_extract(reg_val, sel->shift, sel->width);
+
+	/* Look up that selector's parent array index and return it */
+	index = parent_index(sel, parent_sel);
+	if (index == BAD_CLK_INDEX)
+		pr_err("%s: out-of-range parent selector %u (%s 0x%04x)\n",
+			__func__, parent_sel, ccu->name, sel->offset);
+
+	return index;
+}
+
+/*
+ * Commit our desired selector value to the hardware.
+ *
+ * Returns 0 on success.  Returns -EINVAL for invalid arguments.
+ * Returns -ENXIO if gating failed, and -EIO if a trigger failed.
+ */
+static int
+__sel_commit(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_sel *sel, struct bcm_clk_trig *trig)
+{
+	u32 parent_sel;
+	u32 reg_val;
+	bool enabled;
+	int ret = 0;
+
+	BUG_ON(!selector_exists(sel));
+
+	/*
+	 * If we're just initializing the selector, and no initial
+	 * state was defined in the device tree, we just find out
+	 * what its current value is rather than updating it.
+	 */
+	if (sel->clk_index == BAD_CLK_INDEX) {
+		u8 index;
+
+		reg_val = __ccu_read(ccu, sel->offset);
+		parent_sel = bitfield_extract(reg_val, sel->shift, sel->width);
+		index = parent_index(sel, parent_sel);
+		if (index == BAD_CLK_INDEX)
+			return -EINVAL;
+		sel->clk_index = index;
+
+		return 0;
+	}
+
+	BUG_ON((u32)sel->clk_index >= sel->parent_count);
+	parent_sel = sel->parent_sel[sel->clk_index];
+
+	/* Clock needs to be enabled before changing the parent */
+	enabled = __is_clk_gate_enabled(ccu, gate);
+	if (!enabled && !__clk_gate(ccu, gate, true))
+		return -ENXIO;
+
+	/* Replace the selector value and record the result */
+	reg_val = __ccu_read(ccu, sel->offset);
+	reg_val = bitfield_replace(reg_val, sel->shift, sel->width, parent_sel);
+	__ccu_write(ccu, sel->offset, reg_val);
+
+	/* If the trigger fails we still want to disable the gate */
+	if (!__clk_trigger(ccu, trig))
+		ret = -EIO;
+
+	/* Disable the clock again if it was disabled to begin with */
+	if (!enabled && !__clk_gate(ccu, gate, false))
+		ret = ret ? ret : -ENXIO;	/* return first error */
+
+	return ret;
+}
+
+/*
+ * Initialize a selector by committing our desired state to hardware
+ * without the usual checks to see if it's already set up that way.
+ * Returns true if successful, false otherwise.
+ */
+static bool sel_init(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_sel *sel, struct bcm_clk_trig *trig)
+{
+	if (!selector_exists(sel))
+		return true;
+	return !__sel_commit(ccu, gate, sel, trig);
+}
+
+/*
+ * Write a new value into a selector register to switch to a
+ * different parent clock.  Returns 0 on success, or an error code
+ * (from __sel_commit()) otherwise.
+ */
+static int selector_write(struct ccu_data *ccu, struct bcm_clk_gate *gate,
+			struct bcm_clk_sel *sel, struct bcm_clk_trig *trig,
+			u8 index)
+{
+	unsigned long flags;
+	u8 previous;
+	int ret;
+
+	previous = sel->clk_index;
+	if (previous == index)
+		return 0;	/* No change */
+
+	sel->clk_index = index;
+
+	flags = ccu_lock(ccu);
+	__ccu_write_enable(ccu);
+
+	ret = __sel_commit(ccu, gate, sel, trig);
+
+	__ccu_write_disable(ccu);
+	ccu_unlock(ccu, flags);
+
+	if (ret)
+		sel->clk_index = previous;	/* Revert the change */
+
+	return ret;
+}
+
+/* Clock operations */
+
+static int kona_peri_clk_enable(struct clk_hw *hw)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct bcm_clk_gate *gate = &bcm_clk->peri->gate;
+
+	return clk_gate(bcm_clk->ccu, bcm_clk->name, gate, true);
+}
+
+static void kona_peri_clk_disable(struct clk_hw *hw)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct bcm_clk_gate *gate = &bcm_clk->peri->gate;
+
+	(void)clk_gate(bcm_clk->ccu, bcm_clk->name, gate, false);
+}
+
+static int kona_peri_clk_is_enabled(struct clk_hw *hw)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct bcm_clk_gate *gate = &bcm_clk->peri->gate;
+
+	return is_clk_gate_enabled(bcm_clk->ccu, gate) ? 1 : 0;
+}
+
+static unsigned long kona_peri_clk_recalc_rate(struct clk_hw *hw,
+			unsigned long parent_rate)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct peri_clk_data *data = bcm_clk->peri;
+
+	return clk_recalc_rate(bcm_clk->ccu, &data->div, &data->pre_div,
+				parent_rate);
+}
+
+static long kona_peri_clk_round_rate(struct clk_hw *hw, unsigned long rate,
+			unsigned long *parent_rate)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct bcm_clk_div *div = &bcm_clk->peri->div;
+
+	if (!divider_exists(div))
+		return __clk_get_rate(hw->clk);
+
+	/* Quietly avoid a zero rate */
+	return round_rate(bcm_clk->ccu, div, &bcm_clk->peri->pre_div,
+				rate ? rate : 1, *parent_rate, NULL);
+}
+
+static int kona_peri_clk_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct peri_clk_data *data = bcm_clk->peri;
+	struct bcm_clk_sel *sel = &data->sel;
+	struct bcm_clk_trig *trig;
+	int ret;
+
+	BUG_ON(index >= sel->parent_count);
+
+	/* If there's only one parent we don't require a selector */
+	if (!selector_exists(sel))
+		return 0;
+
+	/*
+	 * The regular trigger is used by default, but if there's a
+	 * pre-trigger we want to use that instead.
+	 */
+	trig = trigger_exists(&data->pre_trig) ? &data->pre_trig
+					       : &data->trig;
+
+	ret = selector_write(bcm_clk->ccu, &data->gate, sel, trig, index);
+	if (ret == -ENXIO) {
+		pr_err("%s: gating failure for %s\n", __func__, bcm_clk->name);
+		ret = -EIO;	/* Don't proliferate weird errors */
+	} else if (ret == -EIO) {
+		pr_err("%s: %strigger failed for %s\n", __func__,
+			trig == &data->pre_trig ? "pre-" : "",
+			bcm_clk->name);
+	}
+
+	return ret;
+}
+
+static u8 kona_peri_clk_get_parent(struct clk_hw *hw)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct peri_clk_data *data = bcm_clk->peri;
+	u8 index;
+
+	index = selector_read_index(bcm_clk->ccu, &data->sel);
+
+	/* Not all callers would handle an out-of-range value gracefully */
+	return index == BAD_CLK_INDEX ? 0 : index;
+}
+
+static int kona_peri_clk_set_rate(struct clk_hw *hw, unsigned long rate,
+			unsigned long parent_rate)
+{
+	struct kona_clk *bcm_clk = to_kona_clk(hw);
+	struct peri_clk_data *data = bcm_clk->peri;
+	struct bcm_clk_div *div = &data->div;
+	u64 scaled_div = 0;
+	int ret;
+
+	if (parent_rate > (unsigned long)LONG_MAX)
+		return -EINVAL;
+
+	if (rate == __clk_get_rate(hw->clk))
+		return 0;
+
+	if (!divider_exists(div))
+		return rate == parent_rate ? 0 : -EINVAL;
+
+	/*
+	 * A fixed divider can't be changed.  (Nor can a fixed
+	 * pre-divider be, but for now we never actually try to
+	 * change that.)  Tolerate a request for a no-op change.
+	 */
+	if (divider_is_fixed(&data->div))
+		return rate == parent_rate ? 0 : -EINVAL;
+
+	/*
+	 * Get the scaled divisor value needed to achieve a clock
+	 * rate as close as possible to what was requested, given
+	 * the parent clock rate supplied.
+	 */
+	(void)round_rate(bcm_clk->ccu, div, &data->pre_div,
+				rate ? rate : 1, parent_rate, &scaled_div);
+
+	/*
+	 * We aren't updating any pre-divider at this point, so
+	 * we'll use the regular trigger.
+	 */
+	ret = divider_write(bcm_clk->ccu, &data->gate, &data->div,
+				&data->trig, scaled_div);
+	if (ret == -ENXIO) {
+		pr_err("%s: gating failure for %s\n", __func__, bcm_clk->name);
+		ret = -EIO;	/* Don't proliferate weird errors */
+	} else if (ret == -EIO) {
+		pr_err("%s: trigger failed for %s\n", __func__, bcm_clk->name);
+	}
+
+	return ret;
+}
+
+struct clk_ops kona_peri_clk_ops = {
+	.enable = kona_peri_clk_enable,
+	.disable = kona_peri_clk_disable,
+	.is_enabled = kona_peri_clk_is_enabled,
+	.recalc_rate = kona_peri_clk_recalc_rate,
+	.round_rate = kona_peri_clk_round_rate,
+	.set_parent = kona_peri_clk_set_parent,
+	.get_parent = kona_peri_clk_get_parent,
+	.set_rate = kona_peri_clk_set_rate,
+};
+
+/* Put a peripheral clock into its initial state */
+static bool __peri_clk_init(struct kona_clk *bcm_clk)
+{
+	struct ccu_data *ccu = bcm_clk->ccu;
+	struct peri_clk_data *peri = bcm_clk->peri;
+	const char *name = bcm_clk->name;
+	struct bcm_clk_trig *trig;
+
+	BUG_ON(bcm_clk->type != bcm_clk_peri);
+
+	if (!gate_init(ccu, &peri->gate)) {
+		pr_err("%s: error initializing gate for %s\n", __func__, name);
+		return false;
+	}
+	if (!div_init(ccu, &peri->gate, &peri->div, &peri->trig)) {
+		pr_err("%s: error initializing divider for %s\n", __func__,
+			name);
+		return false;
+	}
+
+	/*
+	 * For the pre-divider and selector, the pre-trigger is used
+	 * if it's present, otherwise we just use the regular trigger.
+	 */
+	trig = trigger_exists(&peri->pre_trig) ? &peri->pre_trig
+					       : &peri->trig;
+
+	if (!div_init(ccu, &peri->gate, &peri->pre_div, trig)) {
+		pr_err("%s: error initializing pre-divider for %s\n", __func__,
+			name);
+		return false;
+	}
+
+	if (!sel_init(ccu, &peri->gate, &peri->sel, trig)) {
+		pr_err("%s: error initializing selector for %s\n", __func__,
+			name);
+		return false;
+	}
+
+	return true;
+}
+
+static bool __kona_clk_init(struct kona_clk *bcm_clk)
+{
+	switch (bcm_clk->type) {
+	case bcm_clk_peri:
+		return __peri_clk_init(bcm_clk);
+	default:
+		BUG();
+	}
+	return -EINVAL;
+}
+
+/* Set a CCU and all its clocks into their desired initial state */
+bool __init kona_ccu_init(struct ccu_data *ccu)
+{
+	unsigned long flags;
+	unsigned int which;
+	struct clk **clks = ccu->data.clks;
+	bool success = true;
+
+	flags = ccu_lock(ccu);
+	__ccu_write_enable(ccu);
+
+	for (which = 0; which < ccu->data.clk_num; which++) {
+		struct kona_clk *bcm_clk;
+
+		if (!clks[which])
+			continue;
+		bcm_clk = to_kona_clk(__clk_get_hw(clks[which]));
+		success &= __kona_clk_init(bcm_clk);
+	}
+
+	__ccu_write_disable(ccu);
+	ccu_unlock(ccu, flags);
+	return success;
+}
diff --git a/drivers/clk/bcm/clk-kona.h b/drivers/clk/bcm/clk-kona.h
new file mode 100644
index 0000000..5e139ad
--- /dev/null
+++ b/drivers/clk/bcm/clk-kona.h
@@ -0,0 +1,410 @@
+/*
+ * Copyright (C) 2013 Broadcom Corporation
+ * Copyright 2013 Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _CLK_KONA_H
+#define _CLK_KONA_H
+
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/device.h>
+#include <linux/of.h>
+#include <linux/clk-provider.h>
+
+#define	BILLION		1000000000
+
+/* The common clock framework uses u8 to represent a parent index */
+#define PARENT_COUNT_MAX	((u32)U8_MAX)
+
+#define BAD_CLK_INDEX		U8_MAX	/* Can't ever be valid */
+#define BAD_CLK_NAME		((const char *)-1)
+
+#define BAD_SCALED_DIV_VALUE	U64_MAX
+
+/*
+ * Utility macros for object flag management.  If possible, flags
+ * should be defined such that 0 is the desired default value.
+ */
+#define FLAG(type, flag)		BCM_CLK_ ## type ## _FLAGS_ ## flag
+#define FLAG_SET(obj, type, flag)	((obj)->flags |= FLAG(type, flag))
+#define FLAG_CLEAR(obj, type, flag)	((obj)->flags &= ~(FLAG(type, flag)))
+#define FLAG_FLIP(obj, type, flag)	((obj)->flags ^= FLAG(type, flag))
+#define FLAG_TEST(obj, type, flag)	(!!((obj)->flags & FLAG(type, flag)))
+
+/* Clock field state tests */
+
+#define gate_exists(gate)		FLAG_TEST(gate, GATE, EXISTS)
+#define gate_is_enabled(gate)		FLAG_TEST(gate, GATE, ENABLED)
+#define gate_is_hw_controllable(gate)	FLAG_TEST(gate, GATE, HW)
+#define gate_is_sw_controllable(gate)	FLAG_TEST(gate, GATE, SW)
+#define gate_is_sw_managed(gate)	FLAG_TEST(gate, GATE, SW_MANAGED)
+#define gate_is_no_disable(gate)	FLAG_TEST(gate, GATE, NO_DISABLE)
+
+#define gate_flip_enabled(gate)		FLAG_FLIP(gate, GATE, ENABLED)
+
+#define divider_exists(div)		FLAG_TEST(div, DIV, EXISTS)
+#define divider_is_fixed(div)		FLAG_TEST(div, DIV, FIXED)
+#define divider_has_fraction(div)	(!divider_is_fixed(div) && \
+						(div)->frac_width > 0)
+
+#define selector_exists(sel)		((sel)->width != 0)
+#define trigger_exists(trig)		FLAG_TEST(trig, TRIG, EXISTS)
+
+/* Clock type, used to tell common block what it's part of */
+enum bcm_clk_type {
+	bcm_clk_none,		/* undefined clock type */
+	bcm_clk_bus,
+	bcm_clk_core,
+	bcm_clk_peri
+};
+
+/*
+ * Each CCU defines a mapped area of memory containing registers
+ * used to manage clocks implemented by the CCU.  Access to memory
+ * within the CCU's space is serialized by a spinlock.  Before any
+ * (other) address can be written, a special access "password" value
+ * must be written to its WR_ACCESS register (located at the base
+ * address of the range).  We keep track of the name of each CCU as
+ * it is set up, and maintain them in a list.
+ */
+struct ccu_data {
+	void __iomem *base;	/* base of mapped address space */
+	spinlock_t lock;	/* serialization lock */
+	bool write_enabled;	/* write access is currently enabled */
+	struct list_head links;	/* for ccu_list */
+	struct device_node *node;
+	struct clk_onecell_data data;
+	const char *name;
+	u32 range;		/* byte range of address space */
+};
+
+/*
+ * Gating control and status is managed by a 32-bit gate register.
+ *
+ * There are several types of gating available:
+ * - (no gate)
+ *     A clock with no gate is assumed to be always enabled.
+ * - hardware-only gating (auto-gating)
+ *     Enabling or disabling clocks with this type of gate is
+ *     managed automatically by the hardware.  Such clocks can be
+ *     considered by the software to be enabled.  The current status
+ *     of auto-gated clocks can be read from the gate status bit.
+ * - software-only gating
+ *     Auto-gating is not available for this type of clock.
+ *     Instead, software manages whether it's enabled by setting or
+ *     clearing the enable bit.  The current gate status of a gate
+ *     under software control can be read from the gate status bit.
+ *     To ensure a change to the gating status is complete, the
+ *     status bit can be polled to verify that the gate has entered
+ *     the desired state.
+ * - selectable hardware or software gating
+ *     Gating for this type of clock can be configured to be either
+ *     under software or hardware control.  Which type is in use is
+ *     determined by the hw_sw_sel bit of the gate register.
+ */
+struct bcm_clk_gate {
+	u32 offset;		/* gate register offset */
+	u32 status_bit;		/* 0: gate is disabled; 0: gatge is enabled */
+	u32 en_bit;		/* 0: disable; 1: enable */
+	u32 hw_sw_sel_bit;	/* 0: hardware gating; 1: software gating */
+	u32 flags;		/* BCM_CLK_GATE_FLAGS_* below */
+};
+
+/*
+ * Gate flags:
+ *   HW         means this gate can be auto-gated
+ *   SW         means the state of this gate can be software controlled
+ *   NO_DISABLE means this gate is (only) enabled if under software control
+ *   SW_MANAGED means the status of this gate is under software control
+ *   ENABLED    means this software-managed gate is *supposed* to be enabled
+ */
+#define BCM_CLK_GATE_FLAGS_EXISTS	((u32)1 << 0)	/* Gate is valid */
+#define BCM_CLK_GATE_FLAGS_HW		((u32)1 << 1)	/* Can auto-gate */
+#define BCM_CLK_GATE_FLAGS_SW		((u32)1 << 2)	/* Software control */
+#define BCM_CLK_GATE_FLAGS_NO_DISABLE	((u32)1 << 3)	/* HW or enabled */
+#define BCM_CLK_GATE_FLAGS_SW_MANAGED	((u32)1 << 4)	/* SW now in control */
+#define BCM_CLK_GATE_FLAGS_ENABLED	((u32)1 << 5)	/* If SW_MANAGED */
+
+/*
+ * Gate initialization macros.
+ *
+ * Any gate initially under software control will be enabled.
+ */
+
+/* A hardware/software gate initially under software control */
+#define HW_SW_GATE(_offset, _status_bit, _en_bit, _hw_sw_sel_bit)	\
+	{								\
+		.offset = (_offset),					\
+		.status_bit = (_status_bit),				\
+		.en_bit = (_en_bit),					\
+		.hw_sw_sel_bit = (_hw_sw_sel_bit),			\
+		.flags = FLAG(GATE, HW)|FLAG(GATE, SW)|			\
+			FLAG(GATE, SW_MANAGED)|FLAG(GATE, ENABLED)|	\
+			FLAG(GATE, EXISTS),				\
+	}
+
+/* A hardware/software gate initially under hardware control */
+#define HW_SW_GATE_AUTO(_offset, _status_bit, _en_bit, _hw_sw_sel_bit)	\
+	{								\
+		.offset = (_offset),					\
+		.status_bit = (_status_bit),				\
+		.en_bit = (_en_bit),					\
+		.hw_sw_sel_bit = (_hw_sw_sel_bit),			\
+		.flags = FLAG(GATE, HW)|FLAG(GATE, SW)|			\
+			FLAG(GATE, EXISTS),				\
+	}
+
+/* A hardware-or-enabled gate (enabled if not under hardware control) */
+#define HW_ENABLE_GATE(_offset, _status_bit, _en_bit, _hw_sw_sel_bit)	\
+	{								\
+		.offset = (_offset),					\
+		.status_bit = (_status_bit),				\
+		.en_bit = (_en_bit),					\
+		.hw_sw_sel_bit = (_hw_sw_sel_bit),			\
+		.flags = FLAG(GATE, HW)|FLAG(GATE, SW)|			\
+			FLAG(GATE, NO_DISABLE)|FLAG(GATE, EXISTS),	\
+	}
+
+/* A software-only gate */
+#define SW_ONLY_GATE(_offset, _status_bit, _en_bit)			\
+	{								\
+		.offset = (_offset),					\
+		.status_bit = (_status_bit),				\
+		.en_bit = (_en_bit),					\
+		.flags = FLAG(GATE, SW)|FLAG(GATE, SW_MANAGED)|		\
+			FLAG(GATE, ENABLED)|FLAG(GATE, EXISTS),		\
+	}
+
+/* A hardware-only gate */
+#define HW_ONLY_GATE(_offset, _status_bit)				\
+	{								\
+		.offset = (_offset),					\
+		.status_bit = (_status_bit),				\
+		.flags = FLAG(GATE, HW)|FLAG(GATE, EXISTS),		\
+	}
+
+/*
+ * Each clock can have zero, one, or two dividers which change the
+ * output rate of the clock.  Each divider can be either fixed or
+ * variable.  If there are two dividers, they are the "pre-divider"
+ * and the "regular" or "downstream" divider.  If there is only one,
+ * there is no pre-divider.
+ *
+ * A fixed divider is any non-zero (positive) value, and it
+ * indicates how the input rate is affected by the divider.
+ *
+ * The value of a variable divider is maintained in a sub-field of a
+ * 32-bit divider register.  The position of the field in the
+ * register is defined by its offset and width.  The value recorded
+ * in this field is always 1 less than the value it represents.
+ *
+ * In addition, a variable divider can indicate that some subset
+ * of its bits represent a "fractional" part of the divider.  Such
+ * bits comprise the low-order portion of the divider field, and can
+ * be viewed as representing the portion of the divider that lies to
+ * the right of the decimal point.  Most variable dividers have zero
+ * fractional bits.  Variable dividers with non-zero fraction width
+ * still record a value 1 less than the value they represent; the
+ * added 1 does *not* affect the low-order bit in this case, it
+ * affects the bits above the fractional part only.  (Often in this
+ * code a divider field value is distinguished from the value it
+ * represents by referring to the latter as a "divisor".)
+ *
+ * In order to avoid dealing with fractions, divider arithmetic is
+ * performed using "scaled" values.  A scaled value is one that's
+ * been left-shifted by the fractional width of a divider.  Dividing
+ * a scaled value by a scaled divisor produces the desired quotient
+ * without loss of precision and without any other special handling
+ * for fractions.
+ *
+ * The recorded value of a variable divider can be modified.  To
+ * modify either divider (or both), a clock must be enabled (i.e.,
+ * using its gate).  In addition, a trigger register (described
+ * below) must be used to commit the change, and polled to verify
+ * the change is complete.
+ */
+struct bcm_clk_div {
+	union {
+		struct {	/* variable divider */
+			u32 offset;	/* divider register offset */
+			u32 shift;	/* field shift */
+			u32 width;	/* field width */
+			u32 frac_width;	/* field fraction width */
+
+			u64 scaled_div;	/* scaled divider value */
+		};
+		u32 fixed;	/* non-zero fixed divider value */
+	};
+	u32 flags;		/* BCM_CLK_DIV_FLAGS_* below */
+};
+
+/*
+ * Divider flags:
+ *   EXISTS means this divider exists
+ *   FIXED means it is a fixed-rate divider
+ */
+#define BCM_CLK_DIV_FLAGS_EXISTS	((u32)1 << 0)	/* Divider is valid */
+#define BCM_CLK_DIV_FLAGS_FIXED		((u32)1 << 1)	/* Fixed-value */
+
+/* Divider initialization macros */
+
+/* A fixed (non-zero) divider */
+#define FIXED_DIVIDER(_value)						\
+	{								\
+		.fixed = (_value),					\
+		.flags = FLAG(DIV, EXISTS)|FLAG(DIV, FIXED),		\
+	}
+
+/* A divider with an integral divisor */
+#define DIVIDER(_offset, _shift, _width)				\
+	{								\
+		.offset = (_offset),					\
+		.shift = (_shift),					\
+		.width = (_width),					\
+		.scaled_div = BAD_SCALED_DIV_VALUE,			\
+		.flags = FLAG(DIV, EXISTS),				\
+	}
+
+/* A divider whose divisor has an integer and fractional part */
+#define FRAC_DIVIDER(_offset, _shift, _width, _frac_width)		\
+	{								\
+		.offset = (_offset),					\
+		.shift = (_shift),					\
+		.width = (_width),					\
+		.frac_width = (_frac_width),				\
+		.scaled_div = BAD_SCALED_DIV_VALUE,			\
+		.flags = FLAG(DIV, EXISTS),				\
+	}
+
+/*
+ * Clocks may have multiple "parent" clocks.  If there is more than
+ * one, a selector must be specified to define which of the parent
+ * clocks is currently in use.  The selected clock is indicated in a
+ * sub-field of a 32-bit selector register.  The range of
+ * representable selector values typically exceeds the number of
+ * available parent clocks.  Occasionally the reset value of a
+ * selector field is explicitly set to a (specific) value that does
+ * not correspond to a defined input clock.
+ *
+ * We register all known parent clocks with the common clock code
+ * using a packed array (i.e., no empty slots) of (parent) clock
+ * names, and refer to them later using indexes into that array.
+ * We maintain an array of selector values indexed by common clock
+ * index values in order to map between these common clock indexes
+ * and the selector values used by the hardware.
+ *
+ * Like dividers, a selector can be modified, but to do so a clock
+ * must be enabled, and a trigger must be used to commit the change.
+ */
+struct bcm_clk_sel {
+	u32 offset;		/* selector register offset */
+	u32 shift;		/* field shift */
+	u32 width;		/* field width */
+
+	u32 parent_count;	/* number of entries in parent_sel[] */
+	u32 *parent_sel;	/* array of parent selector values */
+	u8 clk_index;		/* current selected index in parent_sel[] */
+};
+
+/* Selector initialization macro */
+#define SELECTOR(_offset, _shift, _width)				\
+	{								\
+		.offset = (_offset),					\
+		.shift = (_shift),					\
+		.width = (_width),					\
+		.clk_index = BAD_CLK_INDEX,				\
+	}
+
+/*
+ * Making changes to a variable divider or a selector for a clock
+ * requires the use of a trigger.  A trigger is defined by a single
+ * bit within a register.  To signal a change, a 1 is written into
+ * that bit.  To determine when the change has been completed, that
+ * trigger bit is polled; the read value will be 1 while the change
+ * is in progress, and 0 when it is complete.
+ *
+ * Occasionally a clock will have more than one trigger.  In this
+ * case, the "pre-trigger" will be used when changing a clock's
+ * selector and/or its pre-divider.
+ */
+struct bcm_clk_trig {
+	u32 offset;		/* trigger register offset */
+	u32 bit;		/* trigger bit */
+	u32 flags;		/* BCM_CLK_TRIG_FLAGS_* below */
+};
+
+/*
+ * Trigger flags:
+ *   EXISTS means this trigger exists
+ */
+#define BCM_CLK_TRIG_FLAGS_EXISTS	((u32)1 << 0)	/* Trigger is valid */
+
+/* Trigger initialization macro */
+#define TRIGGER(_offset, _bit)						\
+	{								\
+		.offset = (_offset),					\
+		.bit = (_bit),						\
+		.flags = FLAG(TRIG, EXISTS),				\
+	}
+
+struct peri_clk_data {
+	struct bcm_clk_gate gate;
+	struct bcm_clk_trig pre_trig;
+	struct bcm_clk_div pre_div;
+	struct bcm_clk_trig trig;
+	struct bcm_clk_div div;
+	struct bcm_clk_sel sel;
+	const char *clocks[];	/* must be last; use CLOCKS() to declare */
+};
+#define CLOCKS(...)	{ __VA_ARGS__, NULL, }
+#define NO_CLOCKS	{ NULL, }	/* Must use of no parent clocks */
+
+struct kona_clk {
+	struct clk_hw hw;
+	struct clk_init_data init_data;
+	const char *name;	/* name of this clock */
+	struct ccu_data *ccu;	/* ccu this clock is associated with */
+	enum bcm_clk_type type;
+	union {
+		void *data;
+		struct peri_clk_data *peri;
+	};
+};
+#define to_kona_clk(_hw) \
+	container_of(_hw, struct kona_clk, hw)
+
+/* Exported globals */
+
+extern struct clk_ops kona_peri_clk_ops;
+
+/* Help functions */
+
+#define PERI_CLK_SETUP(clks, ccu, id, name) \
+	clks[id] = kona_clk_setup(ccu, #name, bcm_clk_peri, &name ## _data)
+
+/* Externally visible functions */
+
+extern u64 do_div_round_closest(u64 dividend, unsigned long divisor);
+extern u64 scaled_div_max(struct bcm_clk_div *div);
+extern u64 scaled_div_build(struct bcm_clk_div *div, u32 div_value,
+				u32 billionths);
+
+extern struct clk *kona_clk_setup(struct ccu_data *ccu, const char *name,
+			enum bcm_clk_type type, void *data);
+extern void __init kona_dt_ccu_setup(struct device_node *node,
+			int (*ccu_clks_setup)(struct ccu_data *));
+extern bool __init kona_ccu_init(struct ccu_data *ccu);
+
+#endif /* _CLK_KONA_H */
diff --git a/drivers/clk/samsung/clk-exynos4.c b/drivers/clk/samsung/clk-exynos4.c
index 010f071..b4f9672 100644
--- a/drivers/clk/samsung/clk-exynos4.c
+++ b/drivers/clk/samsung/clk-exynos4.c
@@ -16,6 +16,7 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 
 #include "clk.h"
 
@@ -130,6 +131,17 @@ enum exynos4_plls {
 	nr_plls			/* number of PLLs */
 };
 
+static void __iomem *reg_base;
+static enum exynos4_soc exynos4_soc;
+
+/*
+ * Support for CMU save/restore across system suspends
+ */
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos4_save_common;
+static struct samsung_clk_reg_dump *exynos4_save_soc;
+static struct samsung_clk_reg_dump *exynos4_save_pll;
+
 /*
  * list of controller registers to be saved and restored during a
  * suspend/resume cycle.
@@ -154,6 +166,17 @@ static unsigned long exynos4x12_clk_save[] __initdata = {
 	E4X12_MPLL_CON0,
 };
 
+static unsigned long exynos4_clk_pll_regs[] __initdata = {
+	EPLL_LOCK,
+	VPLL_LOCK,
+	EPLL_CON0,
+	EPLL_CON1,
+	EPLL_CON2,
+	VPLL_CON0,
+	VPLL_CON1,
+	VPLL_CON2,
+};
+
 static unsigned long exynos4_clk_regs[] __initdata = {
 	SRC_LEFTBUS,
 	DIV_LEFTBUS,
@@ -161,12 +184,6 @@ static unsigned long exynos4_clk_regs[] __initdata = {
 	SRC_RIGHTBUS,
 	DIV_RIGHTBUS,
 	GATE_IP_RIGHTBUS,
-	EPLL_CON0,
-	EPLL_CON1,
-	EPLL_CON2,
-	VPLL_CON0,
-	VPLL_CON1,
-	VPLL_CON2,
 	SRC_TOP0,
 	SRC_TOP1,
 	SRC_CAM,
@@ -227,6 +244,124 @@ static unsigned long exynos4_clk_regs[] __initdata = {
 	GATE_IP_CPU,
 };
 
+static const struct samsung_clk_reg_dump src_mask_suspend[] = {
+	{ .offset = SRC_MASK_TOP,		.value = 0x00000001, },
+	{ .offset = SRC_MASK_CAM,		.value = 0x11111111, },
+	{ .offset = SRC_MASK_TV,		.value = 0x00000111, },
+	{ .offset = SRC_MASK_LCD0,		.value = 0x00001111, },
+	{ .offset = SRC_MASK_MAUDIO,		.value = 0x00000001, },
+	{ .offset = SRC_MASK_FSYS,		.value = 0x01011111, },
+	{ .offset = SRC_MASK_PERIL0,		.value = 0x01111111, },
+	{ .offset = SRC_MASK_PERIL1,		.value = 0x01110111, },
+	{ .offset = SRC_MASK_DMC,		.value = 0x00010000, },
+};
+
+static const struct samsung_clk_reg_dump src_mask_suspend_e4210[] = {
+	{ .offset = E4210_SRC_MASK_LCD1,	.value = 0x00001111, },
+};
+
+#define PLL_ENABLED	(1 << 31)
+#define PLL_LOCKED	(1 << 29)
+
+static void exynos4_clk_wait_for_pll(u32 reg)
+{
+	u32 pll_con;
+
+	pll_con = readl(reg_base + reg);
+	if (!(pll_con & PLL_ENABLED))
+		return;
+
+	while (!(pll_con & PLL_LOCKED)) {
+		cpu_relax();
+		pll_con = readl(reg_base + reg);
+	}
+}
+
+static int exynos4_clk_suspend(void)
+{
+	samsung_clk_save(reg_base, exynos4_save_common,
+				ARRAY_SIZE(exynos4_clk_regs));
+	samsung_clk_save(reg_base, exynos4_save_pll,
+				ARRAY_SIZE(exynos4_clk_pll_regs));
+
+	if (exynos4_soc == EXYNOS4210) {
+		samsung_clk_save(reg_base, exynos4_save_soc,
+					ARRAY_SIZE(exynos4210_clk_save));
+		samsung_clk_restore(reg_base, src_mask_suspend_e4210,
+					ARRAY_SIZE(src_mask_suspend_e4210));
+	} else {
+		samsung_clk_save(reg_base, exynos4_save_soc,
+					ARRAY_SIZE(exynos4x12_clk_save));
+	}
+
+	samsung_clk_restore(reg_base, src_mask_suspend,
+					ARRAY_SIZE(src_mask_suspend));
+
+	return 0;
+}
+
+static void exynos4_clk_resume(void)
+{
+	samsung_clk_restore(reg_base, exynos4_save_pll,
+				ARRAY_SIZE(exynos4_clk_pll_regs));
+
+	exynos4_clk_wait_for_pll(EPLL_CON0);
+	exynos4_clk_wait_for_pll(VPLL_CON0);
+
+	samsung_clk_restore(reg_base, exynos4_save_common,
+				ARRAY_SIZE(exynos4_clk_regs));
+
+	if (exynos4_soc == EXYNOS4210)
+		samsung_clk_restore(reg_base, exynos4_save_soc,
+					ARRAY_SIZE(exynos4210_clk_save));
+	else
+		samsung_clk_restore(reg_base, exynos4_save_soc,
+					ARRAY_SIZE(exynos4x12_clk_save));
+}
+
+static struct syscore_ops exynos4_clk_syscore_ops = {
+	.suspend = exynos4_clk_suspend,
+	.resume = exynos4_clk_resume,
+};
+
+static void exynos4_clk_sleep_init(void)
+{
+	exynos4_save_common = samsung_clk_alloc_reg_dump(exynos4_clk_regs,
+					ARRAY_SIZE(exynos4_clk_regs));
+	if (!exynos4_save_common)
+		goto err_warn;
+
+	if (exynos4_soc == EXYNOS4210)
+		exynos4_save_soc = samsung_clk_alloc_reg_dump(
+					exynos4210_clk_save,
+					ARRAY_SIZE(exynos4210_clk_save));
+	else
+		exynos4_save_soc = samsung_clk_alloc_reg_dump(
+					exynos4x12_clk_save,
+					ARRAY_SIZE(exynos4x12_clk_save));
+	if (!exynos4_save_soc)
+		goto err_common;
+
+	exynos4_save_pll = samsung_clk_alloc_reg_dump(exynos4_clk_pll_regs,
+					ARRAY_SIZE(exynos4_clk_pll_regs));
+	if (!exynos4_save_pll)
+		goto err_soc;
+
+	register_syscore_ops(&exynos4_clk_syscore_ops);
+	return;
+
+err_soc:
+	kfree(exynos4_save_soc);
+err_common:
+	kfree(exynos4_save_common);
+err_warn:
+	pr_warn("%s: failed to allocate sleep save data, no sleep support!\n",
+		__func__);
+}
+#else
+static void exynos4_clk_sleep_init(void) {}
+#endif
+
 /* list of all parent clock list */
 PNAME(mout_apll_p)	= { "fin_pll", "fout_apll", };
 PNAME(mout_mpll_p)	= { "fin_pll", "fout_mpll", };
@@ -908,12 +1043,13 @@ static unsigned long exynos4_get_xom(void)
 	return xom;
 }
 
-static void __init exynos4_clk_register_finpll(unsigned long xom)
+static void __init exynos4_clk_register_finpll(void)
 {
 	struct samsung_fixed_rate_clock fclk;
 	struct clk *clk;
 	unsigned long finpll_f = 24000000;
 	char *parent_name;
+	unsigned int xom = exynos4_get_xom();
 
 	parent_name = xom & 1 ? "xusbxti" : "xxti";
 	clk = clk_get(NULL, parent_name);
@@ -1038,27 +1174,21 @@ static struct samsung_pll_clock exynos4x12_plls[nr_plls] __initdata = {
 
 /* register exynos4 clocks */
 static void __init exynos4_clk_init(struct device_node *np,
-				    enum exynos4_soc exynos4_soc,
-				    void __iomem *reg_base, unsigned long xom)
+				    enum exynos4_soc soc)
 {
+	exynos4_soc = soc;
+
 	reg_base = of_iomap(np, 0);
 	if (!reg_base)
 		panic("%s: failed to map registers\n", __func__);
 
-	if (exynos4_soc == EXYNOS4210)
-		samsung_clk_init(np, reg_base, CLK_NR_CLKS,
-			exynos4_clk_regs, ARRAY_SIZE(exynos4_clk_regs),
-			exynos4210_clk_save, ARRAY_SIZE(exynos4210_clk_save));
-	else
-		samsung_clk_init(np, reg_base, CLK_NR_CLKS,
-			exynos4_clk_regs, ARRAY_SIZE(exynos4_clk_regs),
-			exynos4x12_clk_save, ARRAY_SIZE(exynos4x12_clk_save));
+	samsung_clk_init(np, reg_base, CLK_NR_CLKS);
 
 	samsung_clk_of_register_fixed_ext(exynos4_fixed_rate_ext_clks,
 			ARRAY_SIZE(exynos4_fixed_rate_ext_clks),
 			ext_clk_match);
 
-	exynos4_clk_register_finpll(xom);
+	exynos4_clk_register_finpll();
 
 	if (exynos4_soc == EXYNOS4210) {
 		samsung_clk_register_mux(exynos4210_mux_early,
@@ -1125,6 +1255,8 @@ static void __init exynos4_clk_init(struct device_node *np,
 	samsung_clk_register_alias(exynos4_aliases,
 			ARRAY_SIZE(exynos4_aliases));
 
+	exynos4_clk_sleep_init();
+
 	pr_info("%s clocks: sclk_apll = %ld, sclk_mpll = %ld\n"
 		"\tsclk_epll = %ld, sclk_vpll = %ld, arm_clk = %ld\n",
 		exynos4_soc == EXYNOS4210 ? "Exynos4210" : "Exynos4x12",
@@ -1136,12 +1268,12 @@ static void __init exynos4_clk_init(struct device_node *np,
 
 static void __init exynos4210_clk_init(struct device_node *np)
 {
-	exynos4_clk_init(np, EXYNOS4210, NULL, exynos4_get_xom());
+	exynos4_clk_init(np, EXYNOS4210);
 }
 CLK_OF_DECLARE(exynos4210_clk, "samsung,exynos4210-clock", exynos4210_clk_init);
 
 static void __init exynos4412_clk_init(struct device_node *np)
 {
-	exynos4_clk_init(np, EXYNOS4X12, NULL, exynos4_get_xom());
+	exynos4_clk_init(np, EXYNOS4X12);
 }
 CLK_OF_DECLARE(exynos4412_clk, "samsung,exynos4412-clock", exynos4412_clk_init);
diff --git a/drivers/clk/samsung/clk-exynos5250.c b/drivers/clk/samsung/clk-exynos5250.c
index ff4beeb..e7ee442 100644
--- a/drivers/clk/samsung/clk-exynos5250.c
+++ b/drivers/clk/samsung/clk-exynos5250.c
@@ -16,6 +16,7 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 
 #include "clk.h"
 
@@ -85,6 +86,11 @@ enum exynos5250_plls {
 	nr_plls			/* number of PLLs */
 };
 
+static void __iomem *reg_base;
+
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos5250_save;
+
 /*
  * list of controller registers to be saved and restored during a
  * suspend/resume cycle.
@@ -137,6 +143,41 @@ static unsigned long exynos5250_clk_regs[] __initdata = {
 	GATE_IP_ACP,
 };
 
+static int exynos5250_clk_suspend(void)
+{
+	samsung_clk_save(reg_base, exynos5250_save,
+				ARRAY_SIZE(exynos5250_clk_regs));
+
+	return 0;
+}
+
+static void exynos5250_clk_resume(void)
+{
+	samsung_clk_restore(reg_base, exynos5250_save,
+				ARRAY_SIZE(exynos5250_clk_regs));
+}
+
+static struct syscore_ops exynos5250_clk_syscore_ops = {
+	.suspend = exynos5250_clk_suspend,
+	.resume = exynos5250_clk_resume,
+};
+
+static void exynos5250_clk_sleep_init(void)
+{
+	exynos5250_save = samsung_clk_alloc_reg_dump(exynos5250_clk_regs,
+					ARRAY_SIZE(exynos5250_clk_regs));
+	if (!exynos5250_save) {
+		pr_warn("%s: failed to allocate sleep save data, no sleep support!\n",
+			__func__);
+		return;
+	}
+
+	register_syscore_ops(&exynos5250_clk_syscore_ops);
+}
+#else
+static void exynos5250_clk_sleep_init(void) {}
+#endif
+
 /* list of all parent clock list */
 PNAME(mout_apll_p)	= { "fin_pll", "fout_apll", };
 PNAME(mout_cpu_p)	= { "mout_apll", "mout_mpll", };
@@ -645,8 +686,6 @@ static struct of_device_id ext_clk_match[] __initdata = {
 /* register exynox5250 clocks */
 static void __init exynos5250_clk_init(struct device_node *np)
 {
-	void __iomem *reg_base;
-
 	if (np) {
 		reg_base = of_iomap(np, 0);
 		if (!reg_base)
@@ -655,9 +694,7 @@ static void __init exynos5250_clk_init(struct device_node *np)
 		panic("%s: unable to determine soc\n", __func__);
 	}
 
-	samsung_clk_init(np, reg_base, CLK_NR_CLKS,
-			exynos5250_clk_regs, ARRAY_SIZE(exynos5250_clk_regs),
-			NULL, 0);
+	samsung_clk_init(np, reg_base, CLK_NR_CLKS);
 	samsung_clk_of_register_fixed_ext(exynos5250_fixed_rate_ext_clks,
 			ARRAY_SIZE(exynos5250_fixed_rate_ext_clks),
 			ext_clk_match);
@@ -685,6 +722,8 @@ static void __init exynos5250_clk_init(struct device_node *np)
 	samsung_clk_register_gate(exynos5250_gate_clks,
 			ARRAY_SIZE(exynos5250_gate_clks));
 
+	exynos5250_clk_sleep_init();
+
 	pr_info("Exynos5250: clock setup completed, armclk=%ld\n",
 			_get_rate("div_arm2"));
 }
diff --git a/drivers/clk/samsung/clk-exynos5420.c b/drivers/clk/samsung/clk-exynos5420.c
index ab4f2f7d..60b2681 100644
--- a/drivers/clk/samsung/clk-exynos5420.c
+++ b/drivers/clk/samsung/clk-exynos5420.c
@@ -16,6 +16,7 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 
 #include "clk.h"
 
@@ -108,6 +109,11 @@ enum exynos5420_plls {
 	nr_plls			/* number of PLLs */
 };
 
+static void __iomem *reg_base;
+
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos5420_save;
+
 /*
  * list of controller registers to be saved and restored during a
  * suspend/resume cycle.
@@ -174,6 +180,41 @@ static unsigned long exynos5420_clk_regs[] __initdata = {
 	DIV_KFC0,
 };
 
+static int exynos5420_clk_suspend(void)
+{
+	samsung_clk_save(reg_base, exynos5420_save,
+				ARRAY_SIZE(exynos5420_clk_regs));
+
+	return 0;
+}
+
+static void exynos5420_clk_resume(void)
+{
+	samsung_clk_restore(reg_base, exynos5420_save,
+				ARRAY_SIZE(exynos5420_clk_regs));
+}
+
+static struct syscore_ops exynos5420_clk_syscore_ops = {
+	.suspend = exynos5420_clk_suspend,
+	.resume = exynos5420_clk_resume,
+};
+
+static void exynos5420_clk_sleep_init(void)
+{
+	exynos5420_save = samsung_clk_alloc_reg_dump(exynos5420_clk_regs,
+					ARRAY_SIZE(exynos5420_clk_regs));
+	if (!exynos5420_save) {
+		pr_warn("%s: failed to allocate sleep save data, no sleep support!\n",
+			__func__);
+		return;
+	}
+
+	register_syscore_ops(&exynos5420_clk_syscore_ops);
+}
+#else
+static void exynos5420_clk_sleep_init(void) {}
+#endif
+
 /* list of all parent clocks */
 PNAME(mspll_cpu_p)	= { "sclk_cpll", "sclk_dpll",
 				"sclk_mpll", "sclk_spll" };
@@ -737,8 +778,6 @@ static struct of_device_id ext_clk_match[] __initdata = {
 /* register exynos5420 clocks */
 static void __init exynos5420_clk_init(struct device_node *np)
 {
-	void __iomem *reg_base;
-
 	if (np) {
 		reg_base = of_iomap(np, 0);
 		if (!reg_base)
@@ -747,9 +786,7 @@ static void __init exynos5420_clk_init(struct device_node *np)
 		panic("%s: unable to determine soc\n", __func__);
 	}
 
-	samsung_clk_init(np, reg_base, CLK_NR_CLKS,
-			exynos5420_clk_regs, ARRAY_SIZE(exynos5420_clk_regs),
-			NULL, 0);
+	samsung_clk_init(np, reg_base, CLK_NR_CLKS);
 	samsung_clk_of_register_fixed_ext(exynos5420_fixed_rate_ext_clks,
 			ARRAY_SIZE(exynos5420_fixed_rate_ext_clks),
 			ext_clk_match);
@@ -765,5 +802,7 @@ static void __init exynos5420_clk_init(struct device_node *np)
 			ARRAY_SIZE(exynos5420_div_clks));
 	samsung_clk_register_gate(exynos5420_gate_clks,
 			ARRAY_SIZE(exynos5420_gate_clks));
+
+	exynos5420_clk_sleep_init();
 }
 CLK_OF_DECLARE(exynos5420_clk, "samsung,exynos5420-clock", exynos5420_clk_init);
diff --git a/drivers/clk/samsung/clk-exynos5440.c b/drivers/clk/samsung/clk-exynos5440.c
index cbc15b5..2bfad5a 100644
--- a/drivers/clk/samsung/clk-exynos5440.c
+++ b/drivers/clk/samsung/clk-exynos5440.c
@@ -101,7 +101,7 @@ static void __init exynos5440_clk_init(struct device_node *np)
 		return;
 	}
 
-	samsung_clk_init(np, reg_base, CLK_NR_CLKS, NULL, 0, NULL, 0);
+	samsung_clk_init(np, reg_base, CLK_NR_CLKS);
 	samsung_clk_of_register_fixed_ext(exynos5440_fixed_rate_ext_clks,
 		ARRAY_SIZE(exynos5440_fixed_rate_ext_clks), ext_clk_match);
 
diff --git a/drivers/clk/samsung/clk-s3c64xx.c b/drivers/clk/samsung/clk-s3c64xx.c
index 8e27aee..8bda658 100644
--- a/drivers/clk/samsung/clk-s3c64xx.c
+++ b/drivers/clk/samsung/clk-s3c64xx.c
@@ -13,6 +13,7 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 
 #include <dt-bindings/clock/samsung,s3c64xx-clock.h>
 
@@ -61,6 +62,13 @@ enum s3c64xx_plls {
 	apll, mpll, epll,
 };
 
+static void __iomem *reg_base;
+static bool is_s3c6400;
+
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *s3c64xx_save_common;
+static struct samsung_clk_reg_dump *s3c64xx_save_soc;
+
 /*
  * List of controller registers to be saved and restored during
  * a suspend/resume cycle.
@@ -87,6 +95,60 @@ static unsigned long s3c6410_clk_regs[] __initdata = {
 	MEM0_GATE,
 };
 
+static int s3c64xx_clk_suspend(void)
+{
+	samsung_clk_save(reg_base, s3c64xx_save_common,
+				ARRAY_SIZE(s3c64xx_clk_regs));
+
+	if (!is_s3c6400)
+		samsung_clk_save(reg_base, s3c64xx_save_soc,
+					ARRAY_SIZE(s3c6410_clk_regs));
+
+	return 0;
+}
+
+static void s3c64xx_clk_resume(void)
+{
+	samsung_clk_restore(reg_base, s3c64xx_save_common,
+				ARRAY_SIZE(s3c64xx_clk_regs));
+
+	if (!is_s3c6400)
+		samsung_clk_restore(reg_base, s3c64xx_save_soc,
+					ARRAY_SIZE(s3c6410_clk_regs));
+}
+
+static struct syscore_ops s3c64xx_clk_syscore_ops = {
+	.suspend = s3c64xx_clk_suspend,
+	.resume = s3c64xx_clk_resume,
+};
+
+static void s3c64xx_clk_sleep_init(void)
+{
+	s3c64xx_save_common = samsung_clk_alloc_reg_dump(s3c64xx_clk_regs,
+						ARRAY_SIZE(s3c64xx_clk_regs));
+	if (!s3c64xx_save_common)
+		goto err_warn;
+
+	if (!is_s3c6400) {
+		s3c64xx_save_soc = samsung_clk_alloc_reg_dump(s3c6410_clk_regs,
+						ARRAY_SIZE(s3c6410_clk_regs));
+		if (!s3c64xx_save_soc)
+			goto err_soc;
+	}
+
+	register_syscore_ops(&s3c64xx_clk_syscore_ops);
+	return;
+
+err_soc:
+	kfree(s3c64xx_save_common);
+err_warn:
+	pr_warn("%s: failed to allocate sleep save data, no sleep support!\n",
+		__func__);
+}
+#else
+static void s3c64xx_clk_sleep_init(void) {}
+#endif
+
 /* List of parent clocks common for all S3C64xx SoCs. */
 PNAME(spi_mmc_p)	= { "mout_epll", "dout_mpll", "fin_pll", "clk27m" };
 PNAME(uart_p)		= { "mout_epll", "dout_mpll" };
@@ -391,11 +453,11 @@ static void __init s3c64xx_clk_register_fixed_ext(unsigned long fin_pll_f,
 
 /* Register s3c64xx clocks. */
 void __init s3c64xx_clk_init(struct device_node *np, unsigned long xtal_f,
-			     unsigned long xusbxti_f, bool is_s3c6400,
-			     void __iomem *reg_base)
+			     unsigned long xusbxti_f, bool s3c6400,
+			     void __iomem *base)
 {
-	unsigned long *soc_regs = NULL;
-	unsigned long nr_soc_regs = 0;
+	reg_base = base;
+	is_s3c6400 = s3c6400;
 
 	if (np) {
 		reg_base = of_iomap(np, 0);
@@ -403,13 +465,7 @@ void __init s3c64xx_clk_init(struct device_node *np, unsigned long xtal_f,
 			panic("%s: failed to map registers\n", __func__);
 	}
 
-	if (!is_s3c6400) {
-		soc_regs = s3c6410_clk_regs;
-		nr_soc_regs = ARRAY_SIZE(s3c6410_clk_regs);
-	}
-
-	samsung_clk_init(np, reg_base, NR_CLKS, s3c64xx_clk_regs,
-			ARRAY_SIZE(s3c64xx_clk_regs), soc_regs, nr_soc_regs);
+	samsung_clk_init(np, reg_base, NR_CLKS);
 
 	/* Register external clocks. */
 	if (!np)
@@ -452,6 +508,7 @@ void __init s3c64xx_clk_init(struct device_node *np, unsigned long xtal_f,
 
 	samsung_clk_register_alias(s3c64xx_clock_aliases,
 					ARRAY_SIZE(s3c64xx_clock_aliases));
+	s3c64xx_clk_sleep_init();
 
 	pr_info("%s clocks: apll = %lu, mpll = %lu\n"
 		"\tepll = %lu, arm_clk = %lu\n",
diff --git a/drivers/clk/samsung/clk.c b/drivers/clk/samsung/clk.c
index f503f32..91bec3e 100644
--- a/drivers/clk/samsung/clk.c
+++ b/drivers/clk/samsung/clk.c
@@ -21,64 +21,45 @@ static void __iomem *reg_base;
 static struct clk_onecell_data clk_data;
 #endif
 
-#ifdef CONFIG_PM_SLEEP
-static struct samsung_clk_reg_dump *reg_dump;
-static unsigned long nr_reg_dump;
-
-static int samsung_clk_suspend(void)
+void samsung_clk_save(void __iomem *base,
+				    struct samsung_clk_reg_dump *rd,
+				    unsigned int num_regs)
 {
-	struct samsung_clk_reg_dump *rd = reg_dump;
-	unsigned long i;
-
-	for (i = 0; i < nr_reg_dump; i++, rd++)
-		rd->value = __raw_readl(reg_base + rd->offset);
+	for (; num_regs > 0; --num_regs, ++rd)
+		rd->value = readl(base + rd->offset);
+}
 
-	return 0;
+void samsung_clk_restore(void __iomem *base,
+				      const struct samsung_clk_reg_dump *rd,
+				      unsigned int num_regs)
+{
+	for (; num_regs > 0; --num_regs, ++rd)
+		writel(rd->value, base + rd->offset);
 }
 
-static void samsung_clk_resume(void)
+struct samsung_clk_reg_dump *samsung_clk_alloc_reg_dump(
+						const unsigned long *rdump,
+						unsigned long nr_rdump)
 {
-	struct samsung_clk_reg_dump *rd = reg_dump;
-	unsigned long i;
+	struct samsung_clk_reg_dump *rd;
+	unsigned int i;
 
-	for (i = 0; i < nr_reg_dump; i++, rd++)
-		__raw_writel(rd->value, reg_base + rd->offset);
-}
+	rd = kcalloc(nr_rdump, sizeof(*rd), GFP_KERNEL);
+	if (!rd)
+		return NULL;
+
+	for (i = 0; i < nr_rdump; ++i)
+		rd[i].offset = rdump[i];
 
-static struct syscore_ops samsung_clk_syscore_ops = {
-	.suspend	= samsung_clk_suspend,
-	.resume		= samsung_clk_resume,
-};
-#endif /* CONFIG_PM_SLEEP */
+	return rd;
+}
 
 /* setup the essentials required to support clock lookup using ccf */
 void __init samsung_clk_init(struct device_node *np, void __iomem *base,
-		unsigned long nr_clks, unsigned long *rdump,
-		unsigned long nr_rdump, unsigned long *soc_rdump,
-		unsigned long nr_soc_rdump)
+			     unsigned long nr_clks)
 {
 	reg_base = base;
 
-#ifdef CONFIG_PM_SLEEP
-	if (rdump && nr_rdump) {
-		unsigned int idx;
-		reg_dump = kzalloc(sizeof(struct samsung_clk_reg_dump)
-				* (nr_rdump + nr_soc_rdump), GFP_KERNEL);
-		if (!reg_dump) {
-			pr_err("%s: memory alloc for register dump failed\n",
-					__func__);
-			return;
-		}
-
-		for (idx = 0; idx < nr_rdump; idx++)
-			reg_dump[idx].offset = rdump[idx];
-		for (idx = 0; idx < nr_soc_rdump; idx++)
-			reg_dump[nr_rdump + idx].offset = soc_rdump[idx];
-		nr_reg_dump = nr_rdump + nr_soc_rdump;
-		register_syscore_ops(&samsung_clk_syscore_ops);
-	}
-#endif
-
 	clk_table = kzalloc(sizeof(struct clk *) * nr_clks, GFP_KERNEL);
 	if (!clk_table)
 		panic("could not allocate clock lookup table\n");
diff --git a/drivers/clk/samsung/clk.h b/drivers/clk/samsung/clk.h
index 31b4174..c7141ba 100644
--- a/drivers/clk/samsung/clk.h
+++ b/drivers/clk/samsung/clk.h
@@ -313,9 +313,7 @@ struct samsung_pll_clock {
 		_lock, _con, _rtable, _alias)
 
 extern void __init samsung_clk_init(struct device_node *np, void __iomem *base,
-		unsigned long nr_clks, unsigned long *rdump,
-		unsigned long nr_rdump, unsigned long *soc_rdump,
-		unsigned long nr_soc_rdump);
+				    unsigned long nr_clks);
 extern void __init samsung_clk_of_register_fixed_ext(
 		struct samsung_fixed_rate_clock *fixed_rate_clk,
 		unsigned int nr_fixed_rate_clk,
@@ -340,4 +338,14 @@ extern void __init samsung_clk_register_pll(struct samsung_pll_clock *pll_list,
 
 extern unsigned long _get_rate(const char *clk_name);
 
+extern void samsung_clk_save(void __iomem *base,
+			     struct samsung_clk_reg_dump *rd,
+			     unsigned int num_regs);
+extern void samsung_clk_restore(void __iomem *base,
+				const struct samsung_clk_reg_dump *rd,
+				unsigned int num_regs);
+extern struct samsung_clk_reg_dump *samsung_clk_alloc_reg_dump(
+						const unsigned long *rdump,
+						unsigned long nr_rdump);
+
 #endif /* __SAMSUNG_CLK_H */
diff --git a/drivers/clk/versatile/clk-icst.c b/drivers/clk/versatile/clk-icst.c
index 8cbfcf8..a820b0cf 100644
--- a/drivers/clk/versatile/clk-icst.c
+++ b/drivers/clk/versatile/clk-icst.c
@@ -33,7 +33,7 @@ struct clk_icst {
 	struct clk_hw hw;
 	void __iomem *vcoreg;
 	void __iomem *lockreg;
-	const struct icst_params *params;
+	struct icst_params *params;
 	unsigned long rate;
 };
 
@@ -84,6 +84,8 @@ static unsigned long icst_recalc_rate(struct clk_hw *hw,
 	struct clk_icst *icst = to_icst(hw);
 	struct icst_vco vco;
 
+	if (parent_rate)
+		icst->params->ref = parent_rate;
 	vco = vco_get(icst->vcoreg);
 	icst->rate = icst_hz(icst->params, vco);
 	return icst->rate;
@@ -105,6 +107,8 @@ static int icst_set_rate(struct clk_hw *hw, unsigned long rate,
 	struct clk_icst *icst = to_icst(hw);
 	struct icst_vco vco;
 
+	if (parent_rate)
+		icst->params->ref = parent_rate;
 	vco = icst_hz_to_vco(icst->params, rate);
 	icst->rate = icst_hz(icst->params, vco);
 	vco_set(icst->lockreg, icst->vcoreg, vco);
@@ -120,24 +124,33 @@ static const struct clk_ops icst_ops = {
 struct clk *icst_clk_register(struct device *dev,
 			const struct clk_icst_desc *desc,
 			const char *name,
+			const char *parent_name,
 			void __iomem *base)
 {
 	struct clk *clk;
 	struct clk_icst *icst;
 	struct clk_init_data init;
+	struct icst_params *pclone;
 
 	icst = kzalloc(sizeof(struct clk_icst), GFP_KERNEL);
 	if (!icst) {
 		pr_err("could not allocate ICST clock!\n");
 		return ERR_PTR(-ENOMEM);
 	}
+
+	pclone = kmemdup(desc->params, sizeof(*pclone), GFP_KERNEL);
+	if (!pclone) {
+		pr_err("could not clone ICST params\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
 	init.name = name;
 	init.ops = &icst_ops;
 	init.flags = CLK_IS_ROOT;
-	init.parent_names = NULL;
-	init.num_parents = 0;
+	init.parent_names = (parent_name ? &parent_name : NULL);
+	init.num_parents = (parent_name ? 1 : 0);
 	icst->hw.init = &init;
-	icst->params = desc->params;
+	icst->params = pclone;
 	icst->vcoreg = base + desc->vco_offset;
 	icst->lockreg = base + desc->lock_offset;
 
diff --git a/drivers/clk/versatile/clk-icst.h b/drivers/clk/versatile/clk-icst.h
index be99dd0..04e6f0a 100644
--- a/drivers/clk/versatile/clk-icst.h
+++ b/drivers/clk/versatile/clk-icst.h
@@ -16,4 +16,5 @@ struct clk_icst_desc {
 struct clk *icst_clk_register(struct device *dev,
 			      const struct clk_icst_desc *desc,
 			      const char *name,
+			      const char *parent_name,
 			      void __iomem *base);
diff --git a/drivers/clk/versatile/clk-impd1.c b/drivers/clk/versatile/clk-impd1.c
index 844f8d7..6d8b8e1 100644
--- a/drivers/clk/versatile/clk-impd1.c
+++ b/drivers/clk/versatile/clk-impd1.c
@@ -93,13 +93,15 @@ void integrator_impd1_clk_init(void __iomem *base, unsigned int id)
 	imc = &impd1_clks[id];
 
 	imc->vco1name = kasprintf(GFP_KERNEL, "lm%x-vco1", id);
-	clk = icst_clk_register(NULL, &impd1_icst1_desc, imc->vco1name, base);
+	clk = icst_clk_register(NULL, &impd1_icst1_desc, imc->vco1name, NULL,
+				base);
 	imc->vco1clk = clk;
 	imc->clks[0] = clkdev_alloc(clk, NULL, "lm%x:01000", id);
 
 	/* VCO2 is also called "CLK2" */
 	imc->vco2name = kasprintf(GFP_KERNEL, "lm%x-vco2", id);
-	clk = icst_clk_register(NULL, &impd1_icst2_desc, imc->vco2name, base);
+	clk = icst_clk_register(NULL, &impd1_icst2_desc, imc->vco2name, NULL,
+				base);
 	imc->vco2clk = clk;
 
 	/* MMCI uses CLK2 right off */
diff --git a/drivers/clk/versatile/clk-integrator.c b/drivers/clk/versatile/clk-integrator.c
index bda8967..734c4b8 100644
--- a/drivers/clk/versatile/clk-integrator.c
+++ b/drivers/clk/versatile/clk-integrator.c
@@ -10,21 +10,17 @@
 #include <linux/clk.h>
 #include <linux/clkdev.h>
 #include <linux/err.h>
-#include <linux/platform_data/clk-integrator.h>
-
-#include <mach/hardware.h>
-#include <mach/platform.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
 
 #include "clk-icst.h"
 
-/*
- * Implementation of the ARM Integrator/AP and Integrator/CP clock tree.
- * Inspired by portions of:
- * plat-versatile/clock.c and plat-versatile/include/plat/clock.h
- */
+#define INTEGRATOR_HDR_LOCK_OFFSET	0x14
 
-static const struct icst_params cp_auxvco_params = {
-	.ref		= 24000000,
+/* Base offset for the core module */
+static void __iomem *cm_base;
+
+static const struct icst_params cp_auxosc_params = {
 	.vco_max	= ICST525_VCO_MAX_5V,
 	.vco_min	= ICST525_VCO_MIN,
 	.vd_min 	= 8,
@@ -35,50 +31,39 @@ static const struct icst_params cp_auxvco_params = {
 	.idx2s		= icst525_idx2s,
 };
 
-static const struct clk_icst_desc __initdata cp_icst_desc = {
-	.params = &cp_auxvco_params,
+static const struct clk_icst_desc __initdata cm_auxosc_desc = {
+	.params = &cp_auxosc_params,
 	.vco_offset = 0x1c,
 	.lock_offset = INTEGRATOR_HDR_LOCK_OFFSET,
 };
 
-/*
- * integrator_clk_init() - set up the integrator clock tree
- * @is_cp: pass true if it's the Integrator/CP else AP is assumed
- */
-void __init integrator_clk_init(bool is_cp)
+static void __init of_integrator_cm_osc_setup(struct device_node *np)
 {
-	struct clk *clk;
-
-	/* APB clock dummy */
-	clk = clk_register_fixed_rate(NULL, "apb_pclk", NULL, CLK_IS_ROOT, 0);
-	clk_register_clkdev(clk, "apb_pclk", NULL);
-
-	/* UART reference clock */
-	clk = clk_register_fixed_rate(NULL, "uartclk", NULL, CLK_IS_ROOT,
-				14745600);
-	clk_register_clkdev(clk, NULL, "uart0");
-	clk_register_clkdev(clk, NULL, "uart1");
-	if (is_cp)
-		clk_register_clkdev(clk, NULL, "mmci");
-
-	/* 24 MHz clock */
-	clk = clk_register_fixed_rate(NULL, "clk24mhz", NULL, CLK_IS_ROOT,
-				24000000);
-	clk_register_clkdev(clk, NULL, "kmi0");
-	clk_register_clkdev(clk, NULL, "kmi1");
-	if (!is_cp)
-		clk_register_clkdev(clk, NULL, "ap_timer");
+	struct clk *clk = ERR_PTR(-EINVAL);
+	const char *clk_name = np->name;
+	const struct clk_icst_desc *desc = &cm_auxosc_desc;
+	const char *parent_name;
 
-	if (!is_cp)
-		return;
+	if (!cm_base) {
+		/* Remap the core module base if not done yet */
+		struct device_node *parent;
 
-	/* 1 MHz clock */
-	clk = clk_register_fixed_rate(NULL, "clk1mhz", NULL, CLK_IS_ROOT,
-				1000000);
-	clk_register_clkdev(clk, NULL, "sp804");
+		parent = of_get_parent(np);
+		if (!np) {
+			pr_err("no parent on core module clock\n");
+			return;
+		}
+		cm_base = of_iomap(parent, 0);
+		if (!cm_base) {
+			pr_err("could not remap core module base\n");
+			return;
+		}
+	}
 
-	/* ICST VCO clock used on the Integrator/CP CLCD */
-	clk = icst_clk_register(NULL, &cp_icst_desc, "icst",
-				__io_address(INTEGRATOR_HDR_BASE));
-	clk_register_clkdev(clk, NULL, "clcd");
+	parent_name = of_clk_get_parent_name(np, 0);
+	clk = icst_clk_register(NULL, desc, clk_name, parent_name, cm_base);
+	if (!IS_ERR(clk))
+		of_clk_add_provider(np, of_clk_src_simple_get, clk);
 }
+CLK_OF_DECLARE(integrator_cm_auxosc_clk,
+	"arm,integrator-cm-auxosc", of_integrator_cm_osc_setup);
diff --git a/drivers/clk/versatile/clk-realview.c b/drivers/clk/versatile/clk-realview.c
index 747e7b3..c8b5231 100644
--- a/drivers/clk/versatile/clk-realview.c
+++ b/drivers/clk/versatile/clk-realview.c
@@ -85,10 +85,10 @@ void __init realview_clk_init(void __iomem *sysbase, bool is_pb1176)
 	/* ICST VCO clock */
 	if (is_pb1176)
 		clk = icst_clk_register(NULL, &realview_osc0_desc,
-					"osc0", sysbase);
+					"osc0", NULL, sysbase);
 	else
 		clk = icst_clk_register(NULL, &realview_osc4_desc,
-					"osc4", sysbase);
+					"osc4", NULL, sysbase);
 
 	clk_register_clkdev(clk, NULL, "dev:clcd");
 	clk_register_clkdev(clk, NULL, "issp:clcd");
diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm
index 9fb6270..1e2b9db 100644
--- a/drivers/cpufreq/Kconfig.arm
+++ b/drivers/cpufreq/Kconfig.arm
@@ -122,7 +122,7 @@ config ARM_INTEGRATOR
 	  If in doubt, say Y.
 
 config ARM_KIRKWOOD_CPUFREQ
-	def_bool ARCH_KIRKWOOD && OF
+	def_bool MACH_KIRKWOOD
 	help
 	  This adds the CPUFreq driver for Marvell Kirkwood
 	  SoCs.
diff --git a/drivers/cpuidle/Kconfig.arm b/drivers/cpuidle/Kconfig.arm
index d988948..97ccc31 100644
--- a/drivers/cpuidle/Kconfig.arm
+++ b/drivers/cpuidle/Kconfig.arm
@@ -22,7 +22,7 @@ config ARM_HIGHBANK_CPUIDLE
 
 config ARM_KIRKWOOD_CPUIDLE
 	bool "CPU Idle Driver for Marvell Kirkwood SoCs"
-	depends on ARCH_KIRKWOOD
+	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
 	help
 	  This adds the CPU Idle driver for Marvell Kirkwood SoCs.
 
diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig
index 92d8e9a..a86c49a 100644
--- a/drivers/gpio/Kconfig
+++ b/drivers/gpio/Kconfig
@@ -210,7 +210,7 @@ config GPIO_MSM_V1
 
 config GPIO_MSM_V2
 	tristate "Qualcomm MSM GPIO v2"
-	depends on GPIOLIB && OF && ARCH_MSM
+	depends on GPIOLIB && OF && ARCH_QCOM
 	help
 	  Say yes here to support the GPIO interface on ARM v7 based
 	  Qualcomm MSM chips.  Most of the pins on the MSM can be
diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
index c69d1e0..b698497 100644
--- a/drivers/gpu/drm/msm/Kconfig
+++ b/drivers/gpu/drm/msm/Kconfig
@@ -3,7 +3,7 @@ config DRM_MSM
 	tristate "MSM DRM"
 	depends on DRM
 	depends on MSM_IOMMU
-	depends on (ARCH_MSM && ARCH_MSM8960) || (ARM && COMPILE_TEST)
+	depends on ARCH_MSM8960 || (ARM && COMPILE_TEST)
 	select DRM_KMS_HELPER
 	select SHMEM
 	select TMPFS
diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig
index ec42d2d..d770f74 100644
--- a/drivers/irqchip/Kconfig
+++ b/drivers/irqchip/Kconfig
@@ -77,3 +77,11 @@ config VERSATILE_FPGA_IRQ_NR
 config XTENSA_MX
 	bool
 	select IRQ_DOMAIN
+
+config IRQ_CROSSBAR
+	bool
+	help
+	  Support for a CROSSBAR ip that preceeds the main interrupt controller.
+	  The primary irqchip invokes the crossbar's callback which inturn allocates
+	  a free irq and configures the IP. Thus the peripheral interrupts are
+	  routed to one of the free irqchip interrupt lines.
diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
index 6cee9ef..f180f8d 100644
--- a/drivers/irqchip/Makefile
+++ b/drivers/irqchip/Makefile
@@ -28,3 +28,4 @@ obj-$(CONFIG_ARCH_VT8500)		+= irq-vt8500.o
 obj-$(CONFIG_TB10X_IRQC)		+= irq-tb10x.o
 obj-$(CONFIG_XTENSA)			+= irq-xtensa-pic.o
 obj-$(CONFIG_XTENSA_MX)			+= irq-xtensa-mx.o
+obj-$(CONFIG_IRQ_CROSSBAR)		+= irq-crossbar.o
diff --git a/drivers/irqchip/irq-crossbar.c b/drivers/irqchip/irq-crossbar.c
new file mode 100644
index 0000000..fc817d2
--- /dev/null
+++ b/drivers/irqchip/irq-crossbar.c
@@ -0,0 +1,208 @@
+/*
+ *  drivers/irqchip/irq-crossbar.c
+ *
+ *  Copyright (C) 2013 Texas Instruments Incorporated - http://www.ti.com
+ *  Author: Sricharan R <r.sricharan@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/slab.h>
+#include <linux/irqchip/arm-gic.h>
+
+#define IRQ_FREE	-1
+#define GIC_IRQ_START	32
+
+/*
+ * @int_max: maximum number of supported interrupts
+ * @irq_map: array of interrupts to crossbar number mapping
+ * @crossbar_base: crossbar base address
+ * @register_offsets: offsets for each irq number
+ */
+struct crossbar_device {
+	uint int_max;
+	uint *irq_map;
+	void __iomem *crossbar_base;
+	int *register_offsets;
+	void (*write) (int, int);
+};
+
+static struct crossbar_device *cb;
+
+static inline void crossbar_writel(int irq_no, int cb_no)
+{
+	writel(cb_no, cb->crossbar_base + cb->register_offsets[irq_no]);
+}
+
+static inline void crossbar_writew(int irq_no, int cb_no)
+{
+	writew(cb_no, cb->crossbar_base + cb->register_offsets[irq_no]);
+}
+
+static inline void crossbar_writeb(int irq_no, int cb_no)
+{
+	writeb(cb_no, cb->crossbar_base + cb->register_offsets[irq_no]);
+}
+
+static inline int allocate_free_irq(int cb_no)
+{
+	int i;
+
+	for (i = 0; i < cb->int_max; i++) {
+		if (cb->irq_map[i] == IRQ_FREE) {
+			cb->irq_map[i] = cb_no;
+			return i;
+		}
+	}
+
+	return -ENODEV;
+}
+
+static int crossbar_domain_map(struct irq_domain *d, unsigned int irq,
+			       irq_hw_number_t hw)
+{
+	cb->write(hw - GIC_IRQ_START, cb->irq_map[hw - GIC_IRQ_START]);
+	return 0;
+}
+
+static void crossbar_domain_unmap(struct irq_domain *d, unsigned int irq)
+{
+	irq_hw_number_t hw = irq_get_irq_data(irq)->hwirq;
+
+	if (hw > GIC_IRQ_START)
+		cb->irq_map[hw - GIC_IRQ_START] = IRQ_FREE;
+}
+
+static int crossbar_domain_xlate(struct irq_domain *d,
+				 struct device_node *controller,
+				 const u32 *intspec, unsigned int intsize,
+				 unsigned long *out_hwirq,
+				 unsigned int *out_type)
+{
+	unsigned long ret;
+
+	ret = allocate_free_irq(intspec[1]);
+
+	if (IS_ERR_VALUE(ret))
+		return ret;
+
+	*out_hwirq = ret + GIC_IRQ_START;
+	return 0;
+}
+
+const struct irq_domain_ops routable_irq_domain_ops = {
+	.map = crossbar_domain_map,
+	.unmap = crossbar_domain_unmap,
+	.xlate = crossbar_domain_xlate
+};
+
+static int __init crossbar_of_init(struct device_node *node)
+{
+	int i, size, max, reserved = 0, entry;
+	const __be32 *irqsr;
+
+	cb = kzalloc(sizeof(struct cb_device *), GFP_KERNEL);
+
+	if (!cb)
+		return -ENOMEM;
+
+	cb->crossbar_base = of_iomap(node, 0);
+	if (!cb->crossbar_base)
+		goto err1;
+
+	of_property_read_u32(node, "ti,max-irqs", &max);
+	cb->irq_map = kzalloc(max * sizeof(int), GFP_KERNEL);
+	if (!cb->irq_map)
+		goto err2;
+
+	cb->int_max = max;
+
+	for (i = 0; i < max; i++)
+		cb->irq_map[i] = IRQ_FREE;
+
+	/* Get and mark reserved irqs */
+	irqsr = of_get_property(node, "ti,irqs-reserved", &size);
+	if (irqsr) {
+		size /= sizeof(__be32);
+
+		for (i = 0; i < size; i++) {
+			of_property_read_u32_index(node,
+						   "ti,irqs-reserved",
+						   i, &entry);
+			if (entry > max) {
+				pr_err("Invalid reserved entry\n");
+				goto err3;
+			}
+			cb->irq_map[entry] = 0;
+		}
+	}
+
+	cb->register_offsets = kzalloc(max * sizeof(int), GFP_KERNEL);
+	if (!cb->register_offsets)
+		goto err3;
+
+	of_property_read_u32(node, "ti,reg-size", &size);
+
+	switch (size) {
+	case 1:
+		cb->write = crossbar_writeb;
+		break;
+	case 2:
+		cb->write = crossbar_writew;
+		break;
+	case 4:
+		cb->write = crossbar_writel;
+		break;
+	default:
+		pr_err("Invalid reg-size property\n");
+		goto err4;
+		break;
+	}
+
+	/*
+	 * Register offsets are not linear because of the
+	 * reserved irqs. so find and store the offsets once.
+	 */
+	for (i = 0; i < max; i++) {
+		if (!cb->irq_map[i])
+			continue;
+
+		cb->register_offsets[i] = reserved;
+		reserved += size;
+	}
+
+	register_routable_domain_ops(&routable_irq_domain_ops);
+	return 0;
+
+err4:
+	kfree(cb->register_offsets);
+err3:
+	kfree(cb->irq_map);
+err2:
+	iounmap(cb->crossbar_base);
+err1:
+	kfree(cb);
+	return -ENOMEM;
+}
+
+static const struct of_device_id crossbar_match[] __initconst = {
+	{ .compatible = "ti,irq-crossbar" },
+	{}
+};
+
+int __init irqcrossbar_init(void)
+{
+	struct device_node *np;
+	np = of_find_matching_node(NULL, crossbar_match);
+	if (!np)
+		return -ENODEV;
+
+	crossbar_of_init(np);
+	return 0;
+}
diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
index 63922b9..4300b66 100644
--- a/drivers/irqchip/irq-gic.c
+++ b/drivers/irqchip/irq-gic.c
@@ -824,16 +824,25 @@ static int gic_irq_domain_map(struct irq_domain *d, unsigned int irq,
 		irq_set_chip_and_handler(irq, &gic_chip,
 					 handle_fasteoi_irq);
 		set_irq_flags(irq, IRQF_VALID | IRQF_PROBE);
+
+		gic_routable_irq_domain_ops->map(d, irq, hw);
 	}
 	irq_set_chip_data(irq, d->host_data);
 	return 0;
 }
 
+static void gic_irq_domain_unmap(struct irq_domain *d, unsigned int irq)
+{
+	gic_routable_irq_domain_ops->unmap(d, irq);
+}
+
 static int gic_irq_domain_xlate(struct irq_domain *d,
 				struct device_node *controller,
 				const u32 *intspec, unsigned int intsize,
 				unsigned long *out_hwirq, unsigned int *out_type)
 {
+	unsigned long ret = 0;
+
 	if (d->of_node != controller)
 		return -EINVAL;
 	if (intsize < 3)
@@ -843,11 +852,20 @@ static int gic_irq_domain_xlate(struct irq_domain *d,
 	*out_hwirq = intspec[1] + 16;
 
 	/* For SPIs, we need to add 16 more to get the GIC irq ID number */
-	if (!intspec[0])
-		*out_hwirq += 16;
+	if (!intspec[0]) {
+		ret = gic_routable_irq_domain_ops->xlate(d, controller,
+							 intspec,
+							 intsize,
+							 out_hwirq,
+							 out_type);
+
+		if (IS_ERR_VALUE(ret))
+			return ret;
+	}
 
 	*out_type = intspec[2] & IRQ_TYPE_SENSE_MASK;
-	return 0;
+
+	return ret;
 }
 
 #ifdef CONFIG_SMP
@@ -871,9 +889,41 @@ static struct notifier_block gic_cpu_notifier = {
 
 static const struct irq_domain_ops gic_irq_domain_ops = {
 	.map = gic_irq_domain_map,
+	.unmap = gic_irq_domain_unmap,
 	.xlate = gic_irq_domain_xlate,
 };
 
+/* Default functions for routable irq domain */
+static int gic_routable_irq_domain_map(struct irq_domain *d, unsigned int irq,
+			      irq_hw_number_t hw)
+{
+	return 0;
+}
+
+static void gic_routable_irq_domain_unmap(struct irq_domain *d,
+					  unsigned int irq)
+{
+}
+
+static int gic_routable_irq_domain_xlate(struct irq_domain *d,
+				struct device_node *controller,
+				const u32 *intspec, unsigned int intsize,
+				unsigned long *out_hwirq,
+				unsigned int *out_type)
+{
+	*out_hwirq += 16;
+	return 0;
+}
+
+const struct irq_domain_ops gic_default_routable_irq_domain_ops = {
+	.map = gic_routable_irq_domain_map,
+	.unmap = gic_routable_irq_domain_unmap,
+	.xlate = gic_routable_irq_domain_xlate,
+};
+
+const struct irq_domain_ops *gic_routable_irq_domain_ops =
+					&gic_default_routable_irq_domain_ops;
+
 void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 			   void __iomem *dist_base, void __iomem *cpu_base,
 			   u32 percpu_offset, struct device_node *node)
@@ -881,6 +931,7 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 	irq_hw_number_t hwirq_base;
 	struct gic_chip_data *gic;
 	int gic_irqs, irq_base, i;
+	int nr_routable_irqs;
 
 	BUG_ON(gic_nr >= MAX_GIC_NR);
 
@@ -946,14 +997,25 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 	gic->gic_irqs = gic_irqs;
 
 	gic_irqs -= hwirq_base; /* calculate # of irqs to allocate */
-	irq_base = irq_alloc_descs(irq_start, 16, gic_irqs, numa_node_id());
-	if (IS_ERR_VALUE(irq_base)) {
-		WARN(1, "Cannot allocate irq_descs @ IRQ%d, assuming pre-allocated\n",
-		     irq_start);
-		irq_base = irq_start;
+
+	if (of_property_read_u32(node, "arm,routable-irqs",
+				 &nr_routable_irqs)) {
+		irq_base = irq_alloc_descs(irq_start, 16, gic_irqs,
+					   numa_node_id());
+		if (IS_ERR_VALUE(irq_base)) {
+			WARN(1, "Cannot allocate irq_descs @ IRQ%d, assuming pre-allocated\n",
+			     irq_start);
+			irq_base = irq_start;
+		}
+
+		gic->domain = irq_domain_add_legacy(node, gic_irqs, irq_base,
+					hwirq_base, &gic_irq_domain_ops, gic);
+	} else {
+		gic->domain = irq_domain_add_linear(node, nr_routable_irqs,
+						    &gic_irq_domain_ops,
+						    gic);
 	}
-	gic->domain = irq_domain_add_legacy(node, gic_irqs, irq_base,
-				    hwirq_base, &gic_irq_domain_ops, gic);
+
 	if (WARN_ON(!gic->domain))
 		return;
 
diff --git a/drivers/irqchip/irq-vic.c b/drivers/irqchip/irq-vic.c
index 473f09a..37dab0b 100644
--- a/drivers/irqchip/irq-vic.c
+++ b/drivers/irqchip/irq-vic.c
@@ -57,6 +57,7 @@
 
 /**
  * struct vic_device - VIC PM device
+ * @parent_irq: The parent IRQ number of the VIC if cascaded, or 0.
  * @irq: The IRQ number for the base of the VIC.
  * @base: The register base for the VIC.
  * @valid_sources: A bitmask of valid interrupts
@@ -224,6 +225,17 @@ static int handle_one_vic(struct vic_device *vic, struct pt_regs *regs)
 	return handled;
 }
 
+static void vic_handle_irq_cascaded(unsigned int irq, struct irq_desc *desc)
+{
+	u32 stat, hwirq;
+	struct vic_device *vic = irq_desc_get_handler_data(desc);
+
+	while ((stat = readl_relaxed(vic->base + VIC_IRQ_STATUS))) {
+		hwirq = ffs(stat) - 1;
+		generic_handle_irq(irq_find_mapping(vic->domain, hwirq));
+	}
+}
+
 /*
  * Keep iterating over all registered VIC's until there are no pending
  * interrupts.
@@ -246,6 +258,7 @@ static struct irq_domain_ops vic_irqdomain_ops = {
 /**
  * vic_register() - Register a VIC.
  * @base: The base address of the VIC.
+ * @parent_irq: The parent IRQ if cascaded, else 0.
  * @irq: The base IRQ for the VIC.
  * @valid_sources: bitmask of valid interrupts
  * @resume_sources: bitmask of interrupts allowed for resume sources.
@@ -257,7 +270,8 @@ static struct irq_domain_ops vic_irqdomain_ops = {
  *
  * This also configures the IRQ domain for the VIC.
  */
-static void __init vic_register(void __iomem *base, unsigned int irq,
+static void __init vic_register(void __iomem *base, unsigned int parent_irq,
+				unsigned int irq,
 				u32 valid_sources, u32 resume_sources,
 				struct device_node *node)
 {
@@ -273,15 +287,25 @@ static void __init vic_register(void __iomem *base, unsigned int irq,
 	v->base = base;
 	v->valid_sources = valid_sources;
 	v->resume_sources = resume_sources;
-	v->irq = irq;
 	set_handle_irq(vic_handle_irq);
 	vic_id++;
+
+	if (parent_irq) {
+		irq_set_handler_data(parent_irq, v);
+		irq_set_chained_handler(parent_irq, vic_handle_irq_cascaded);
+	}
+
 	v->domain = irq_domain_add_simple(node, fls(valid_sources), irq,
 					  &vic_irqdomain_ops, v);
 	/* create an IRQ mapping for each valid IRQ */
 	for (i = 0; i < fls(valid_sources); i++)
 		if (valid_sources & (1 << i))
 			irq_create_mapping(v->domain, i);
+	/* If no base IRQ was passed, figure out our allocated base */
+	if (irq)
+		v->irq = irq;
+	else
+		v->irq = irq_find_mapping(v->domain, 0);
 }
 
 static void vic_ack_irq(struct irq_data *d)
@@ -409,10 +433,10 @@ static void __init vic_init_st(void __iomem *base, unsigned int irq_start,
 		writel(32, base + VIC_PL190_DEF_VECT_ADDR);
 	}
 
-	vic_register(base, irq_start, vic_sources, 0, node);
+	vic_register(base, 0, irq_start, vic_sources, 0, node);
 }
 
-void __init __vic_init(void __iomem *base, int irq_start,
+void __init __vic_init(void __iomem *base, int parent_irq, int irq_start,
 			      u32 vic_sources, u32 resume_sources,
 			      struct device_node *node)
 {
@@ -449,7 +473,7 @@ void __init __vic_init(void __iomem *base, int irq_start,
 
 	vic_init2(base);
 
-	vic_register(base, irq_start, vic_sources, resume_sources, node);
+	vic_register(base, parent_irq, irq_start, vic_sources, resume_sources, node);
 }
 
 /**
@@ -462,8 +486,30 @@ void __init __vic_init(void __iomem *base, int irq_start,
 void __init vic_init(void __iomem *base, unsigned int irq_start,
 		     u32 vic_sources, u32 resume_sources)
 {
-	__vic_init(base, irq_start, vic_sources, resume_sources, NULL);
+	__vic_init(base, 0, irq_start, vic_sources, resume_sources, NULL);
+}
+
+/**
+ * vic_init_cascaded() - initialise a cascaded vectored interrupt controller
+ * @base: iomem base address
+ * @parent_irq: the parent IRQ we're cascaded off
+ * @irq_start: starting interrupt number, must be muliple of 32
+ * @vic_sources: bitmask of interrupt sources to allow
+ * @resume_sources: bitmask of interrupt sources to allow for resume
+ *
+ * This returns the base for the new interrupts or negative on error.
+ */
+int __init vic_init_cascaded(void __iomem *base, unsigned int parent_irq,
+			      u32 vic_sources, u32 resume_sources)
+{
+	struct vic_device *v;
+
+	v = &vic_devices[vic_id];
+	__vic_init(base, parent_irq, 0, vic_sources, resume_sources, NULL);
+	/* Return out acquired base */
+	return v->irq;
 }
+EXPORT_SYMBOL_GPL(vic_init_cascaded);
 
 #ifdef CONFIG_OF
 int __init vic_of_init(struct device_node *node, struct device_node *parent)
@@ -485,7 +531,7 @@ int __init vic_of_init(struct device_node *node, struct device_node *parent)
 	/*
 	 * Passing 0 as first IRQ makes the simple domain allocate descriptors
 	 */
-	__vic_init(regs, 0, interrupt_mask, wakeup_mask, node);
+	__vic_init(regs, 0, 0, interrupt_mask, wakeup_mask, node);
 
 	return 0;
 }
diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig
index 72156c1..44c358e 100644
--- a/drivers/leds/Kconfig
+++ b/drivers/leds/Kconfig
@@ -421,7 +421,7 @@ config LEDS_MC13783
 config LEDS_NS2
 	tristate "LED support for Network Space v2 GPIO LEDs"
 	depends on LEDS_CLASS
-	depends on ARCH_KIRKWOOD
+	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
 	default y
 	help
 	  This option enable support for the dual-GPIO LED found on the
@@ -431,7 +431,7 @@ config LEDS_NS2
 config LEDS_NETXBIG
 	tristate "LED support for Big Network series LEDs"
 	depends on LEDS_CLASS
-	depends on ARCH_KIRKWOOD
+	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
 	default y
 	help
 	  This option enable support for LEDs found on the LaCie 2Big
diff --git a/drivers/mtd/nand/davinci_nand.c b/drivers/mtd/nand/davinci_nand.c
index a4989ec..8eb6a36 100644
--- a/drivers/mtd/nand/davinci_nand.c
+++ b/drivers/mtd/nand/davinci_nand.c
@@ -746,28 +746,6 @@ static int nand_davinci_probe(struct platform_device *pdev)
 		goto err_clk_enable;
 	}
 
-	/*
-	 * Setup Async configuration register in case we did not boot from
-	 * NAND and so bootloader did not bother to set it up.
-	 */
-	val = davinci_nand_readl(info, A1CR_OFFSET + info->core_chipsel * 4);
-
-	/* Extended Wait is not valid and Select Strobe mode is not used */
-	val &= ~(ACR_ASIZE_MASK | ACR_EW_MASK | ACR_SS_MASK);
-	if (info->chip.options & NAND_BUSWIDTH_16)
-		val |= 0x1;
-
-	davinci_nand_writel(info, A1CR_OFFSET + info->core_chipsel * 4, val);
-
-	ret = 0;
-	if (info->timing)
-		ret = davinci_aemif_setup_timing(info->timing, info->base,
-							info->core_chipsel);
-	if (ret < 0) {
-		dev_dbg(&pdev->dev, "NAND timing values setup fail\n");
-		goto err;
-	}
-
 	spin_lock_irq(&davinci_nand_lock);
 
 	/* put CSxNAND into NAND mode */
diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig
index 8d3c49c..3bb05f1 100644
--- a/drivers/phy/Kconfig
+++ b/drivers/phy/Kconfig
@@ -27,7 +27,7 @@ config PHY_EXYNOS_MIPI_VIDEO
 
 config PHY_MVEBU_SATA
 	def_bool y
-	depends on ARCH_KIRKWOOD || ARCH_DOVE || MACH_DOVE
+	depends on ARCH_KIRKWOOD || ARCH_DOVE || MACH_DOVE || MACH_KIRKWOOD
 	depends on OF
 	select GENERIC_PHY
 
diff --git a/drivers/power/reset/Kconfig b/drivers/power/reset/Kconfig
index 6d452a7..fa0e4e0 100644
--- a/drivers/power/reset/Kconfig
+++ b/drivers/power/reset/Kconfig
@@ -22,7 +22,7 @@ config POWER_RESET_GPIO
 
 config POWER_RESET_MSM
 	bool "Qualcomm MSM power-off driver"
-	depends on POWER_RESET && ARCH_MSM
+	depends on POWER_RESET && ARCH_QCOM
 	help
 	  Power off and restart support for Qualcomm boards.
 
diff --git a/drivers/power/reset/qnap-poweroff.c b/drivers/power/reset/qnap-poweroff.c
index 37f56f7..a75db7f 100644
--- a/drivers/power/reset/qnap-poweroff.c
+++ b/drivers/power/reset/qnap-poweroff.c
@@ -1,5 +1,5 @@
 /*
- * QNAP Turbo NAS Board power off
+ * QNAP Turbo NAS Board power off. Can also be used on Synology devices.
  *
  * Copyright (C) 2012 Andrew Lunn <andrew@lunn.ch>
  *
@@ -25,17 +25,43 @@
 
 #define UART1_REG(x)	(base + ((UART_##x) << 2))
 
+struct power_off_cfg {
+	u32 baud;
+	char cmd;
+};
+
+static const struct power_off_cfg qnap_power_off_cfg = {
+	.baud = 19200,
+	.cmd = 'A',
+};
+
+static const struct power_off_cfg synology_power_off_cfg = {
+	.baud = 9600,
+	.cmd = '1',
+};
+
+static const struct of_device_id qnap_power_off_of_match_table[] = {
+	{ .compatible = "qnap,power-off",
+	  .data = &qnap_power_off_cfg,
+	},
+	{ .compatible = "synology,power-off",
+	  .data = &synology_power_off_cfg,
+	},
+	{}
+};
+MODULE_DEVICE_TABLE(of, qnap_power_off_of_match_table);
+
 static void __iomem *base;
 static unsigned long tclk;
+static const struct power_off_cfg *cfg;
 
 static void qnap_power_off(void)
 {
-	/* 19200 baud divisor */
-	const unsigned divisor = ((tclk + (8 * 19200)) / (16 * 19200));
+	const unsigned divisor = ((tclk + (8 * cfg->baud)) / (16 * cfg->baud));
 
 	pr_err("%s: triggering power-off...\n", __func__);
 
-	/* hijack UART1 and reset into sane state (19200,8n1) */
+	/* hijack UART1 and reset into sane state */
 	writel(0x83, UART1_REG(LCR));
 	writel(divisor & 0xff, UART1_REG(DLL));
 	writel((divisor >> 8) & 0xff, UART1_REG(DLM));
@@ -44,16 +70,21 @@ static void qnap_power_off(void)
 	writel(0x00, UART1_REG(FCR));
 	writel(0x00, UART1_REG(MCR));
 
-	/* send the power-off command 'A' to PIC */
-	writel('A', UART1_REG(TX));
+	/* send the power-off command to PIC */
+	writel(cfg->cmd, UART1_REG(TX));
 }
 
 static int qnap_power_off_probe(struct platform_device *pdev)
 {
+	struct device_node *np = pdev->dev.of_node;
 	struct resource *res;
 	struct clk *clk;
 	char symname[KSYM_NAME_LEN];
 
+	const struct of_device_id *match =
+		of_match_node(qnap_power_off_of_match_table, np);
+	cfg = match->data;
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(&pdev->dev, "Missing resource");
@@ -94,12 +125,6 @@ static int qnap_power_off_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct of_device_id qnap_power_off_of_match_table[] = {
-	{ .compatible = "qnap,power-off", },
-	{}
-};
-MODULE_DEVICE_TABLE(of, qnap_power_off_of_match_table);
-
 static struct platform_driver qnap_power_off_driver = {
 	.probe	= qnap_power_off_probe,
 	.remove	= qnap_power_off_remove,
diff --git a/drivers/reset/Kconfig b/drivers/reset/Kconfig
index c9d04f7..0615f50 100644
--- a/drivers/reset/Kconfig
+++ b/drivers/reset/Kconfig
@@ -11,3 +11,5 @@ menuconfig RESET_CONTROLLER
 	  via GPIOs or SoC-internal reset controller modules.
 
 	  If unsure, say no.
+
+source "drivers/reset/sti/Kconfig"
diff --git a/drivers/reset/Makefile b/drivers/reset/Makefile
index cc29832..4f60caf 100644
--- a/drivers/reset/Makefile
+++ b/drivers/reset/Makefile
@@ -1,2 +1,3 @@
 obj-$(CONFIG_RESET_CONTROLLER) += core.o
 obj-$(CONFIG_ARCH_SUNXI) += reset-sunxi.o
+obj-$(CONFIG_ARCH_STI) += sti/
diff --git a/drivers/reset/core.c b/drivers/reset/core.c
index d1b6089..baeaf82 100644
--- a/drivers/reset/core.c
+++ b/drivers/reset/core.c
@@ -43,7 +43,7 @@ struct reset_control {
  * This simple translation function should be used for reset controllers
  * with 1:1 mapping, where reset lines can be indexed by number without gaps.
  */
-int of_reset_simple_xlate(struct reset_controller_dev *rcdev,
+static int of_reset_simple_xlate(struct reset_controller_dev *rcdev,
 			  const struct of_phandle_args *reset_spec)
 {
 	if (WARN_ON(reset_spec->args_count != rcdev->of_reset_n_cells))
@@ -54,7 +54,6 @@ int of_reset_simple_xlate(struct reset_controller_dev *rcdev,
 
 	return reset_spec->args[0];
 }
-EXPORT_SYMBOL_GPL(of_reset_simple_xlate);
 
 /**
  * reset_controller_register - register a reset controller device
@@ -127,15 +126,16 @@ int reset_control_deassert(struct reset_control *rstc)
 EXPORT_SYMBOL_GPL(reset_control_deassert);
 
 /**
- * reset_control_get - Lookup and obtain a reference to a reset controller.
- * @dev: device to be reset by the controller
+ * of_reset_control_get - Lookup and obtain a reference to a reset controller.
+ * @node: device to be reset by the controller
  * @id: reset line name
  *
  * Returns a struct reset_control or IS_ERR() condition containing errno.
  *
  * Use of id names is optional.
  */
-struct reset_control *reset_control_get(struct device *dev, const char *id)
+struct reset_control *of_reset_control_get(struct device_node *node,
+					   const char *id)
 {
 	struct reset_control *rstc = ERR_PTR(-EPROBE_DEFER);
 	struct reset_controller_dev *r, *rcdev;
@@ -144,13 +144,10 @@ struct reset_control *reset_control_get(struct device *dev, const char *id)
 	int rstc_id;
 	int ret;
 
-	if (!dev)
-		return ERR_PTR(-EINVAL);
-
 	if (id)
-		index = of_property_match_string(dev->of_node,
+		index = of_property_match_string(node,
 						 "reset-names", id);
-	ret = of_parse_phandle_with_args(dev->of_node, "resets", "#reset-cells",
+	ret = of_parse_phandle_with_args(node, "resets", "#reset-cells",
 					 index, &args);
 	if (ret)
 		return ERR_PTR(ret);
@@ -167,7 +164,7 @@ struct reset_control *reset_control_get(struct device *dev, const char *id)
 
 	if (!rcdev) {
 		mutex_unlock(&reset_controller_list_mutex);
-		return ERR_PTR(-ENODEV);
+		return ERR_PTR(-EPROBE_DEFER);
 	}
 
 	rstc_id = rcdev->of_xlate(rcdev, &args);
@@ -185,12 +182,35 @@ struct reset_control *reset_control_get(struct device *dev, const char *id)
 		return ERR_PTR(-ENOMEM);
 	}
 
-	rstc->dev = dev;
 	rstc->rcdev = rcdev;
 	rstc->id = rstc_id;
 
 	return rstc;
 }
+EXPORT_SYMBOL_GPL(of_reset_control_get);
+
+/**
+ * reset_control_get - Lookup and obtain a reference to a reset controller.
+ * @dev: device to be reset by the controller
+ * @id: reset line name
+ *
+ * Returns a struct reset_control or IS_ERR() condition containing errno.
+ *
+ * Use of id names is optional.
+ */
+struct reset_control *reset_control_get(struct device *dev, const char *id)
+{
+	struct reset_control *rstc;
+
+	if (!dev)
+		return ERR_PTR(-EINVAL);
+
+	rstc = of_reset_control_get(dev->of_node, id);
+	if (!IS_ERR(rstc))
+		rstc->dev = dev;
+
+	return rstc;
+}
 EXPORT_SYMBOL_GPL(reset_control_get);
 
 /**
@@ -243,33 +263,6 @@ struct reset_control *devm_reset_control_get(struct device *dev, const char *id)
 }
 EXPORT_SYMBOL_GPL(devm_reset_control_get);
 
-static int devm_reset_control_match(struct device *dev, void *res, void *data)
-{
-	struct reset_control **rstc = res;
-	if (WARN_ON(!rstc || !*rstc))
-		return 0;
-	return *rstc == data;
-}
-
-/**
- * devm_reset_control_put - resource managed reset_control_put()
- * @rstc: reset controller to free
- *
- * Deallocate a reset control allocated withd devm_reset_control_get().
- * This function will not need to be called normally, as devres will take
- * care of freeing the resource.
- */
-void devm_reset_control_put(struct reset_control *rstc)
-{
-	int ret;
-
-	ret = devres_release(rstc->dev, devm_reset_control_release,
-			     devm_reset_control_match, rstc);
-	if (ret)
-		WARN_ON(ret);
-}
-EXPORT_SYMBOL_GPL(devm_reset_control_put);
-
 /**
  * device_reset - find reset controller associated with the device
  *                and perform reset
diff --git a/drivers/reset/sti/Kconfig b/drivers/reset/sti/Kconfig
new file mode 100644
index 0000000..88d2d03
--- /dev/null
+++ b/drivers/reset/sti/Kconfig
@@ -0,0 +1,15 @@
+if ARCH_STI
+
+config STI_RESET_SYSCFG
+	bool
+	select RESET_CONTROLLER
+
+config STIH415_RESET
+	bool
+	select STI_RESET_SYSCFG
+
+config STIH416_RESET
+	bool
+	select STI_RESET_SYSCFG
+
+endif
diff --git a/drivers/reset/sti/Makefile b/drivers/reset/sti/Makefile
new file mode 100644
index 0000000..be1c976
--- /dev/null
+++ b/drivers/reset/sti/Makefile
@@ -0,0 +1,4 @@
+obj-$(CONFIG_STI_RESET_SYSCFG) += reset-syscfg.o
+
+obj-$(CONFIG_STIH415_RESET) += reset-stih415.o
+obj-$(CONFIG_STIH416_RESET) += reset-stih416.o
diff --git a/drivers/reset/sti/reset-stih415.c b/drivers/reset/sti/reset-stih415.c
new file mode 100644
index 0000000..e6f6c41
--- /dev/null
+++ b/drivers/reset/sti/reset-stih415.c
@@ -0,0 +1,112 @@
+/*
+ * Copyright (C) 2013 STMicroelectronics (R&D) Limited
+ * Author: Stephen Gallimore <stephen.gallimore@st.com>
+ * Author: Srinivas Kandagatla <srinivas.kandagatla@st.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, or
+ * (at your option) any later version.
+ */
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+
+#include <dt-bindings/reset-controller/stih415-resets.h>
+
+#include "reset-syscfg.h"
+
+/*
+ * STiH415 Peripheral powerdown definitions.
+ */
+static const char stih415_front[] = "st,stih415-front-syscfg";
+static const char stih415_rear[] = "st,stih415-rear-syscfg";
+static const char stih415_sbc[] = "st,stih415-sbc-syscfg";
+static const char stih415_lpm[] = "st,stih415-lpm-syscfg";
+
+#define STIH415_PDN_FRONT(_bit) \
+	_SYSCFG_RST_CH(stih415_front, SYSCFG_114, _bit, SYSSTAT_187, _bit)
+
+#define STIH415_PDN_REAR(_cntl, _stat) \
+	_SYSCFG_RST_CH(stih415_rear, SYSCFG_336, _cntl, SYSSTAT_384, _stat)
+
+#define STIH415_SRST_REAR(_reg, _bit) \
+	_SYSCFG_RST_CH_NO_ACK(stih415_rear, _reg, _bit)
+
+#define STIH415_SRST_SBC(_reg, _bit) \
+	_SYSCFG_RST_CH_NO_ACK(stih415_sbc, _reg, _bit)
+
+#define STIH415_SRST_FRONT(_reg, _bit) \
+	_SYSCFG_RST_CH_NO_ACK(stih415_front, _reg, _bit)
+
+#define STIH415_SRST_LPM(_reg, _bit) \
+	_SYSCFG_RST_CH_NO_ACK(stih415_lpm, _reg, _bit)
+
+#define SYSCFG_114	0x38 /* Powerdown request EMI/NAND/Keyscan */
+#define SYSSTAT_187	0x15c /* Powerdown status EMI/NAND/Keyscan */
+
+#define SYSCFG_336	0x90 /* Powerdown request USB/SATA/PCIe */
+#define SYSSTAT_384	0x150 /* Powerdown status USB/SATA/PCIe */
+
+#define SYSCFG_376	0x130 /* Reset generator 0 control 0 */
+#define SYSCFG_166	0x108 /* Softreset Ethernet 0 */
+#define SYSCFG_31	0x7c /* Softreset Ethernet 1 */
+#define LPM_SYSCFG_1	0x4 /* Softreset IRB */
+
+static const struct syscfg_reset_channel_data stih415_powerdowns[] = {
+	[STIH415_EMISS_POWERDOWN]	= STIH415_PDN_FRONT(0),
+	[STIH415_NAND_POWERDOWN]	= STIH415_PDN_FRONT(1),
+	[STIH415_KEYSCAN_POWERDOWN]	= STIH415_PDN_FRONT(2),
+	[STIH415_USB0_POWERDOWN]	= STIH415_PDN_REAR(0, 0),
+	[STIH415_USB1_POWERDOWN]	= STIH415_PDN_REAR(1, 1),
+	[STIH415_USB2_POWERDOWN]	= STIH415_PDN_REAR(2, 2),
+	[STIH415_SATA0_POWERDOWN]	= STIH415_PDN_REAR(3, 3),
+	[STIH415_SATA1_POWERDOWN]	= STIH415_PDN_REAR(4, 4),
+	[STIH415_PCIE_POWERDOWN]	= STIH415_PDN_REAR(5, 8),
+};
+
+static const struct syscfg_reset_channel_data stih415_softresets[] = {
+	[STIH415_ETH0_SOFTRESET] = STIH415_SRST_FRONT(SYSCFG_166, 0),
+	[STIH415_ETH1_SOFTRESET] = STIH415_SRST_SBC(SYSCFG_31, 0),
+	[STIH415_IRB_SOFTRESET]	 = STIH415_SRST_LPM(LPM_SYSCFG_1, 6),
+	[STIH415_USB0_SOFTRESET] = STIH415_SRST_REAR(SYSCFG_376, 9),
+	[STIH415_USB1_SOFTRESET] = STIH415_SRST_REAR(SYSCFG_376, 10),
+	[STIH415_USB2_SOFTRESET] = STIH415_SRST_REAR(SYSCFG_376, 11),
+};
+
+static struct syscfg_reset_controller_data stih415_powerdown_controller = {
+	.wait_for_ack = true,
+	.nr_channels = ARRAY_SIZE(stih415_powerdowns),
+	.channels = stih415_powerdowns,
+};
+
+static struct syscfg_reset_controller_data stih415_softreset_controller = {
+	.wait_for_ack = false,
+	.active_low = true,
+	.nr_channels = ARRAY_SIZE(stih415_softresets),
+	.channels = stih415_softresets,
+};
+
+static struct of_device_id stih415_reset_match[] = {
+	{ .compatible = "st,stih415-powerdown",
+	  .data = &stih415_powerdown_controller, },
+	{ .compatible = "st,stih415-softreset",
+	  .data = &stih415_softreset_controller, },
+	{},
+};
+
+static struct platform_driver stih415_reset_driver = {
+	.probe = syscfg_reset_probe,
+	.driver = {
+		.name = "reset-stih415",
+		.owner = THIS_MODULE,
+		.of_match_table = stih415_reset_match,
+	},
+};
+
+static int __init stih415_reset_init(void)
+{
+	return platform_driver_register(&stih415_reset_driver);
+}
+arch_initcall(stih415_reset_init);
diff --git a/drivers/reset/sti/reset-stih416.c b/drivers/reset/sti/reset-stih416.c
new file mode 100644
index 0000000..fe3bf02
--- /dev/null
+++ b/drivers/reset/sti/reset-stih416.c
@@ -0,0 +1,143 @@
+/*
+ * Copyright (C) 2013 STMicroelectronics (R&D) Limited
+ * Author: Stephen Gallimore <stephen.gallimore@st.com>
+ * Author: Srinivas Kandagatla <srinivas.kandagatla@st.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, or
+ * (at your option) any later version.
+ */
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+
+#include <dt-bindings/reset-controller/stih416-resets.h>
+
+#include "reset-syscfg.h"
+
+/*
+ * STiH416 Peripheral powerdown definitions.
+ */
+static const char stih416_front[] = "st,stih416-front-syscfg";
+static const char stih416_rear[] = "st,stih416-rear-syscfg";
+static const char stih416_sbc[] = "st,stih416-sbc-syscfg";
+static const char stih416_lpm[] = "st,stih416-lpm-syscfg";
+static const char stih416_cpu[] = "st,stih416-cpu-syscfg";
+
+#define STIH416_PDN_FRONT(_bit) \
+	_SYSCFG_RST_CH(stih416_front, SYSCFG_1500, _bit, SYSSTAT_1578, _bit)
+
+#define STIH416_PDN_REAR(_cntl, _stat) \
+	_SYSCFG_RST_CH(stih416_rear, SYSCFG_2525, _cntl, SYSSTAT_2583, _stat)
+
+#define SYSCFG_1500	0x7d0 /* Powerdown request EMI/NAND/Keyscan */
+#define SYSSTAT_1578	0x908 /* Powerdown status EMI/NAND/Keyscan */
+
+#define SYSCFG_2525	0x834 /* Powerdown request USB/SATA/PCIe */
+#define SYSSTAT_2583	0x91c /* Powerdown status USB/SATA/PCIe */
+
+#define SYSCFG_2552	0x8A0 /* Reset Generator control 0 */
+#define SYSCFG_1539	0x86c /* Softreset Ethernet 0 */
+#define SYSCFG_510	0x7f8 /* Softreset Ethernet 1 */
+#define LPM_SYSCFG_1	0x4 /* Softreset IRB */
+#define SYSCFG_2553	0x8a4 /* Softreset SATA0/1, PCIE0/1 */
+#define SYSCFG_7563	0x8cc /* MPE softresets 0 */
+#define SYSCFG_7564	0x8d0 /* MPE softresets 1 */
+
+#define STIH416_SRST_CPU(_reg, _bit) \
+	 _SYSCFG_RST_CH_NO_ACK(stih416_cpu, _reg, _bit)
+
+#define STIH416_SRST_FRONT(_reg, _bit) \
+	 _SYSCFG_RST_CH_NO_ACK(stih416_front, _reg, _bit)
+
+#define STIH416_SRST_REAR(_reg, _bit) \
+	 _SYSCFG_RST_CH_NO_ACK(stih416_rear, _reg, _bit)
+
+#define STIH416_SRST_LPM(_reg, _bit) \
+	 _SYSCFG_RST_CH_NO_ACK(stih416_lpm, _reg, _bit)
+
+#define STIH416_SRST_SBC(_reg, _bit) \
+	 _SYSCFG_RST_CH_NO_ACK(stih416_sbc, _reg, _bit)
+
+static const struct syscfg_reset_channel_data stih416_powerdowns[] = {
+	[STIH416_EMISS_POWERDOWN]	= STIH416_PDN_FRONT(0),
+	[STIH416_NAND_POWERDOWN]	= STIH416_PDN_FRONT(1),
+	[STIH416_KEYSCAN_POWERDOWN]	= STIH416_PDN_FRONT(2),
+	[STIH416_USB0_POWERDOWN]	= STIH416_PDN_REAR(0, 0),
+	[STIH416_USB1_POWERDOWN]	= STIH416_PDN_REAR(1, 1),
+	[STIH416_USB2_POWERDOWN]	= STIH416_PDN_REAR(2, 2),
+	[STIH416_USB3_POWERDOWN]	= STIH416_PDN_REAR(6, 5),
+	[STIH416_SATA0_POWERDOWN]	= STIH416_PDN_REAR(3, 3),
+	[STIH416_SATA1_POWERDOWN]	= STIH416_PDN_REAR(4, 4),
+	[STIH416_PCIE0_POWERDOWN]	= STIH416_PDN_REAR(7, 9),
+	[STIH416_PCIE1_POWERDOWN]	= STIH416_PDN_REAR(5, 8),
+};
+
+static const struct syscfg_reset_channel_data stih416_softresets[] = {
+	[STIH416_ETH0_SOFTRESET] = STIH416_SRST_FRONT(SYSCFG_1539, 0),
+	[STIH416_ETH1_SOFTRESET] = STIH416_SRST_SBC(SYSCFG_510, 0),
+	[STIH416_IRB_SOFTRESET]	 = STIH416_SRST_LPM(LPM_SYSCFG_1, 6),
+	[STIH416_USB0_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 9),
+	[STIH416_USB1_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 10),
+	[STIH416_USB2_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 11),
+	[STIH416_USB3_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 28),
+	[STIH416_SATA0_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 7),
+	[STIH416_SATA1_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 3),
+	[STIH416_PCIE0_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 15),
+	[STIH416_PCIE1_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 2),
+	[STIH416_AUD_DAC_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 14),
+	[STIH416_HDTVOUT_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 5),
+	[STIH416_VTAC_M_RX_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 25),
+	[STIH416_VTAC_A_RX_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2552, 26),
+	[STIH416_SYNC_HD_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 5),
+	[STIH416_SYNC_SD_SOFTRESET] = STIH416_SRST_REAR(SYSCFG_2553, 6),
+	[STIH416_BLITTER_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 10),
+	[STIH416_GPU_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 11),
+	[STIH416_VTAC_M_TX_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 18),
+	[STIH416_VTAC_A_TX_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 19),
+	[STIH416_VTG_AUX_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 21),
+	[STIH416_JPEG_DEC_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7563, 23),
+	[STIH416_HVA_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7564, 2),
+	[STIH416_COMPO_M_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7564, 3),
+	[STIH416_COMPO_A_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7564, 4),
+	[STIH416_VP8_DEC_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7564, 10),
+	[STIH416_VTG_MAIN_SOFTRESET] = STIH416_SRST_CPU(SYSCFG_7564, 16),
+};
+
+static struct syscfg_reset_controller_data stih416_powerdown_controller = {
+	.wait_for_ack	= true,
+	.nr_channels	= ARRAY_SIZE(stih416_powerdowns),
+	.channels	= stih416_powerdowns,
+};
+
+static struct syscfg_reset_controller_data stih416_softreset_controller = {
+	.wait_for_ack = false,
+	.active_low = true,
+	.nr_channels = ARRAY_SIZE(stih416_softresets),
+	.channels = stih416_softresets,
+};
+
+static struct of_device_id stih416_reset_match[] = {
+	{ .compatible = "st,stih416-powerdown",
+	  .data = &stih416_powerdown_controller, },
+	{ .compatible = "st,stih416-softreset",
+	  .data = &stih416_softreset_controller, },
+	{},
+};
+
+static struct platform_driver stih416_reset_driver = {
+	.probe = syscfg_reset_probe,
+	.driver = {
+		.name = "reset-stih416",
+		.owner = THIS_MODULE,
+		.of_match_table = stih416_reset_match,
+	},
+};
+
+static int __init stih416_reset_init(void)
+{
+	return platform_driver_register(&stih416_reset_driver);
+}
+arch_initcall(stih416_reset_init);
diff --git a/drivers/reset/sti/reset-syscfg.c b/drivers/reset/sti/reset-syscfg.c
new file mode 100644
index 0000000..a145cc0
--- /dev/null
+++ b/drivers/reset/sti/reset-syscfg.c
@@ -0,0 +1,186 @@
+/*
+ * Copyright (C) 2013 STMicroelectronics Limited
+ * Author: Stephen Gallimore <stephen.gallimore@st.com>
+ *
+ * Inspired by mach-imx/src.c
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/module.h>
+#include <linux/err.h>
+#include <linux/types.h>
+#include <linux/of_device.h>
+#include <linux/regmap.h>
+#include <linux/mfd/syscon.h>
+
+#include "reset-syscfg.h"
+
+/**
+ * Reset channel regmap configuration
+ *
+ * @reset: regmap field for the channel's reset bit.
+ * @ack: regmap field for the channel's ack bit (optional).
+ */
+struct syscfg_reset_channel {
+	struct regmap_field *reset;
+	struct regmap_field *ack;
+};
+
+/**
+ * A reset controller which groups together a set of related reset bits, which
+ * may be located in different system configuration registers.
+ *
+ * @rst: base reset controller structure.
+ * @active_low: are the resets in this controller active low, i.e. clearing
+ *              the reset bit puts the hardware into reset.
+ * @channels: An array of reset channels for this controller.
+ */
+struct syscfg_reset_controller {
+	struct reset_controller_dev rst;
+	bool active_low;
+	struct syscfg_reset_channel *channels;
+};
+
+#define to_syscfg_reset_controller(_rst) \
+	container_of(_rst, struct syscfg_reset_controller, rst)
+
+static int syscfg_reset_program_hw(struct reset_controller_dev *rcdev,
+				   unsigned long idx, int assert)
+{
+	struct syscfg_reset_controller *rst = to_syscfg_reset_controller(rcdev);
+	const struct syscfg_reset_channel *ch;
+	u32 ctrl_val = rst->active_low ? !assert : !!assert;
+	int err;
+
+	if (idx >= rcdev->nr_resets)
+		return -EINVAL;
+
+	ch = &rst->channels[idx];
+
+	err = regmap_field_write(ch->reset, ctrl_val);
+	if (err)
+		return err;
+
+	if (ch->ack) {
+		unsigned long timeout = jiffies + msecs_to_jiffies(1000);
+		u32 ack_val;
+
+		while (true) {
+			err = regmap_field_read(ch->ack, &ack_val);
+			if (err)
+				return err;
+
+			if (ack_val == ctrl_val)
+				break;
+
+			if (time_after(jiffies, timeout))
+				return -ETIME;
+
+			cpu_relax();
+		}
+	}
+
+	return 0;
+}
+
+static int syscfg_reset_assert(struct reset_controller_dev *rcdev,
+			       unsigned long idx)
+{
+	return syscfg_reset_program_hw(rcdev, idx, true);
+}
+
+static int syscfg_reset_deassert(struct reset_controller_dev *rcdev,
+				 unsigned long idx)
+{
+	return syscfg_reset_program_hw(rcdev, idx, false);
+}
+
+static int syscfg_reset_dev(struct reset_controller_dev *rcdev,
+			    unsigned long idx)
+{
+	int err = syscfg_reset_assert(rcdev, idx);
+	if (err)
+		return err;
+
+	return syscfg_reset_deassert(rcdev, idx);
+}
+
+static struct reset_control_ops syscfg_reset_ops = {
+	.reset    = syscfg_reset_dev,
+	.assert   = syscfg_reset_assert,
+	.deassert = syscfg_reset_deassert,
+};
+
+static int syscfg_reset_controller_register(struct device *dev,
+				const struct syscfg_reset_controller_data *data)
+{
+	struct syscfg_reset_controller *rc;
+	size_t size;
+	int i, err;
+
+	rc = devm_kzalloc(dev, sizeof(*rc), GFP_KERNEL);
+	if (!rc)
+		return -ENOMEM;
+
+	size = sizeof(struct syscfg_reset_channel) * data->nr_channels;
+
+	rc->channels = devm_kzalloc(dev, size, GFP_KERNEL);
+	if (!rc->channels)
+		return -ENOMEM;
+
+	rc->rst.ops = &syscfg_reset_ops,
+	rc->rst.of_node = dev->of_node;
+	rc->rst.nr_resets = data->nr_channels;
+	rc->active_low = data->active_low;
+
+	for (i = 0; i < data->nr_channels; i++) {
+		struct regmap *map;
+		struct regmap_field *f;
+		const char *compatible = data->channels[i].compatible;
+
+		map = syscon_regmap_lookup_by_compatible(compatible);
+		if (IS_ERR(map))
+			return PTR_ERR(map);
+
+		f = devm_regmap_field_alloc(dev, map, data->channels[i].reset);
+		if (IS_ERR(f))
+			return PTR_ERR(f);
+
+		rc->channels[i].reset = f;
+
+		if (!data->wait_for_ack)
+			continue;
+
+		f = devm_regmap_field_alloc(dev, map, data->channels[i].ack);
+		if (IS_ERR(f))
+			return PTR_ERR(f);
+
+		rc->channels[i].ack = f;
+	}
+
+	err = reset_controller_register(&rc->rst);
+	if (!err)
+		dev_info(dev, "registered\n");
+
+	return err;
+}
+
+int syscfg_reset_probe(struct platform_device *pdev)
+{
+	struct device *dev = pdev ? &pdev->dev : NULL;
+	const struct of_device_id *match;
+
+	if (!dev || !dev->driver)
+		return -ENODEV;
+
+	match = of_match_device(dev->driver->of_match_table, dev);
+	if (!match || !match->data)
+		return -EINVAL;
+
+	return syscfg_reset_controller_register(dev, match->data);
+}
diff --git a/drivers/reset/sti/reset-syscfg.h b/drivers/reset/sti/reset-syscfg.h
new file mode 100644
index 0000000..2cc2283
--- /dev/null
+++ b/drivers/reset/sti/reset-syscfg.h
@@ -0,0 +1,69 @@
+/*
+ * Copyright (C) 2013 STMicroelectronics (R&D) Limited
+ * Author: Stephen Gallimore <stephen.gallimore@st.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, or
+ * (at your option) any later version.
+ */
+#ifndef __STI_RESET_SYSCFG_H
+#define __STI_RESET_SYSCFG_H
+
+#include <linux/device.h>
+#include <linux/regmap.h>
+#include <linux/reset-controller.h>
+
+/**
+ * Reset channel description for a system configuration register based
+ * reset controller.
+ *
+ * @compatible: Compatible string of the syscon regmap containing this
+ *              channel's control and ack (status) bits.
+ * @reset: Regmap field description of the channel's reset bit.
+ * @ack: Regmap field description of the channel's acknowledge bit.
+ */
+struct syscfg_reset_channel_data {
+	const char *compatible;
+	struct reg_field reset;
+	struct reg_field ack;
+};
+
+#define _SYSCFG_RST_CH(_c, _rr, _rb, _ar, _ab)		\
+	{ .compatible	= _c,				\
+	  .reset	= REG_FIELD(_rr, _rb, _rb),	\
+	  .ack		= REG_FIELD(_ar, _ab, _ab), }
+
+#define _SYSCFG_RST_CH_NO_ACK(_c, _rr, _rb)		\
+	{ .compatible	= _c,			\
+	  .reset	= REG_FIELD(_rr, _rb, _rb), }
+
+/**
+ * Description of a system configuration register based reset controller.
+ *
+ * @wait_for_ack: The controller will wait for reset assert and de-assert to
+ *                be "ack'd" in a channel's ack field.
+ * @active_low: Are the resets in this controller active low, i.e. clearing
+ *              the reset bit puts the hardware into reset.
+ * @nr_channels: The number of reset channels in this controller.
+ * @channels: An array of reset channel descriptions.
+ */
+struct syscfg_reset_controller_data {
+	bool wait_for_ack;
+	bool active_low;
+	int nr_channels;
+	const struct syscfg_reset_channel_data *channels;
+};
+
+/**
+ * syscfg_reset_probe(): platform device probe function used by syscfg
+ *                       reset controller drivers. This registers a reset
+ *                       controller configured by the OF match data for
+ *                       the compatible device which should be of type
+ *                       "struct syscfg_reset_controller_data".
+ *
+ * @pdev: platform device
+ */
+int syscfg_reset_probe(struct platform_device *pdev);
+
+#endif /* __STI_RESET_SYSCFG_H */
diff --git a/drivers/rtc/rtc-isl12057.c b/drivers/rtc/rtc-isl12057.c
index 7e5ead9..41bd76a 100644
--- a/drivers/rtc/rtc-isl12057.c
+++ b/drivers/rtc/rtc-isl12057.c
@@ -274,10 +274,7 @@ static int isl12057_probe(struct i2c_client *client,
 	dev_set_drvdata(dev, data);
 
 	rtc = devm_rtc_device_register(dev, DRV_NAME, &rtc_ops, THIS_MODULE);
-	if (IS_ERR(rtc))
-		return PTR_ERR(rtc);
-
-	return 0;
+	return PTR_ERR_OR_ZERO(rtc);
 }
 
 #ifdef CONFIG_OF
diff --git a/drivers/rtc/rtc-mv.c b/drivers/rtc/rtc-mv.c
index d536c59..d15a999 100644
--- a/drivers/rtc/rtc-mv.c
+++ b/drivers/rtc/rtc-mv.c
@@ -222,6 +222,7 @@ static int __init mv_rtc_probe(struct platform_device *pdev)
 	struct resource *res;
 	struct rtc_plat_data *pdata;
 	u32 rtc_time;
+	u32 rtc_date;
 	int ret = 0;
 
 	pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
@@ -257,6 +258,17 @@ static int __init mv_rtc_probe(struct platform_device *pdev)
 		}
 	}
 
+	/*
+	 * A date after January 19th, 2038 does not fit on 32 bits and
+	 * will confuse the kernel and userspace. Reset to a sane date
+	 * (January 1st, 2013) if we're after 2038.
+	 */
+	rtc_date = readl(pdata->ioaddr + RTC_DATE_REG_OFFS);
+	if (bcd2bin((rtc_date >> RTC_YEAR_OFFS) & 0xff) >= 38) {
+		dev_info(&pdev->dev, "invalid RTC date, resetting to January 1st, 2013\n");
+		writel(0x130101, pdata->ioaddr + RTC_DATE_REG_OFFS);
+	}
+
 	pdata->irq = platform_get_irq(pdev, 0);
 
 	platform_set_drvdata(pdev, pdata);
diff --git a/drivers/sh/clk/cpg.c b/drivers/sh/clk/cpg.c
index 1ebe67c..7442bc1 100644
--- a/drivers/sh/clk/cpg.c
+++ b/drivers/sh/clk/cpg.c
@@ -36,9 +36,47 @@ static void sh_clk_write(int value, struct clk *clk)
 		iowrite32(value, clk->mapped_reg);
 }
 
+static unsigned int r8(const void __iomem *addr)
+{
+	return ioread8(addr);
+}
+
+static unsigned int r16(const void __iomem *addr)
+{
+	return ioread16(addr);
+}
+
+static unsigned int r32(const void __iomem *addr)
+{
+	return ioread32(addr);
+}
+
 static int sh_clk_mstp_enable(struct clk *clk)
 {
 	sh_clk_write(sh_clk_read(clk) & ~(1 << clk->enable_bit), clk);
+	if (clk->status_reg) {
+		unsigned int (*read)(const void __iomem *addr);
+		int i;
+		void __iomem *mapped_status = (phys_addr_t)clk->status_reg -
+			(phys_addr_t)clk->enable_reg + clk->mapped_reg;
+
+		if (clk->flags & CLK_ENABLE_REG_8BIT)
+			read = r8;
+		else if (clk->flags & CLK_ENABLE_REG_16BIT)
+			read = r16;
+		else
+			read = r32;
+
+		for (i = 1000;
+		     (read(mapped_status) & (1 << clk->enable_bit)) && i;
+		     i--)
+			cpu_relax();
+		if (!i) {
+			pr_err("cpg: failed to enable %p[%d]\n",
+			       clk->enable_reg, clk->enable_bit);
+			return -ETIMEDOUT;
+		}
+	}
 	return 0;
 }
 
diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig
index 5f88d76..2d51912 100644
--- a/drivers/thermal/Kconfig
+++ b/drivers/thermal/Kconfig
@@ -143,7 +143,7 @@ config RCAR_THERMAL
 
 config KIRKWOOD_THERMAL
 	tristate "Temperature sensor on Marvell Kirkwood SoCs"
-	depends on ARCH_KIRKWOOD
+	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
 	depends on OF
 	help
 	  Support for the Kirkwood thermal sensor driver into the Linux thermal
diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index 2577d67..2e6d8dd 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig
@@ -1024,7 +1024,7 @@ config SERIAL_SGI_IOC3
 
 config SERIAL_MSM
 	bool "MSM on-chip serial port support"
-	depends on ARCH_MSM
+	depends on ARCH_MSM || ARCH_QCOM
 	select SERIAL_CORE
 
 config SERIAL_MSM_CONSOLE
diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index 0c6048d..74ec8fc 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig
@@ -301,7 +301,7 @@ config DAVINCI_WATCHDOG
 
 config ORION_WATCHDOG
 	tristate "Orion watchdog"
-	depends on ARCH_ORION5X || ARCH_KIRKWOOD || ARCH_DOVE || MACH_DOVE
+	depends on ARCH_ORION5X || ARCH_KIRKWOOD || ARCH_DOVE || MACH_DOVE || ARCH_MVEBU
 	select WATCHDOG_CORE
 	help
 	  Say Y here if to include support for the watchdog timer
diff --git a/drivers/watchdog/orion_wdt.c b/drivers/watchdog/orion_wdt.c
index 4981634..9b3c41d 100644
--- a/drivers/watchdog/orion_wdt.c
+++ b/drivers/watchdog/orion_wdt.c
@@ -18,101 +18,204 @@
 #include <linux/kernel.h>
 #include <linux/platform_device.h>
 #include <linux/watchdog.h>
+#include <linux/interrupt.h>
 #include <linux/io.h>
-#include <linux/spinlock.h>
 #include <linux/clk.h>
 #include <linux/err.h>
 #include <linux/of.h>
-#include <mach/bridge-regs.h>
+#include <linux/of_device.h>
+
+/* RSTOUT mask register physical address for Orion5x, Kirkwood and Dove */
+#define ORION_RSTOUT_MASK_OFFSET	0x20108
+
+/* Internal registers can be configured at any 1 MiB aligned address */
+#define INTERNAL_REGS_MASK		~(SZ_1M - 1)
 
 /*
  * Watchdog timer block registers.
  */
 #define TIMER_CTRL		0x0000
-#define WDT_EN			0x0010
-#define WDT_VAL			0x0024
+#define TIMER_A370_STATUS	0x04
 
 #define WDT_MAX_CYCLE_COUNT	0xffffffff
-#define WDT_IN_USE		0
-#define WDT_OK_TO_CLOSE		1
 
-#define WDT_RESET_OUT_EN	BIT(1)
-#define WDT_INT_REQ		BIT(3)
+#define WDT_A370_RATIO_MASK(v)	((v) << 16)
+#define WDT_A370_RATIO_SHIFT	5
+#define WDT_A370_RATIO		(1 << WDT_A370_RATIO_SHIFT)
+
+#define WDT_AXP_FIXED_ENABLE_BIT BIT(10)
+#define WDT_A370_EXPIRED	BIT(31)
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
 static int heartbeat = -1;		/* module parameter (seconds) */
-static unsigned int wdt_max_duration;	/* (seconds) */
-static struct clk *clk;
-static unsigned int wdt_tclk;
-static void __iomem *wdt_reg;
-static DEFINE_SPINLOCK(wdt_lock);
 
-static int orion_wdt_ping(struct watchdog_device *wdt_dev)
+struct orion_watchdog;
+
+struct orion_watchdog_data {
+	int wdt_counter_offset;
+	int wdt_enable_bit;
+	int rstout_enable_bit;
+	int (*clock_init)(struct platform_device *,
+			  struct orion_watchdog *);
+	int (*start)(struct watchdog_device *);
+};
+
+struct orion_watchdog {
+	struct watchdog_device wdt;
+	void __iomem *reg;
+	void __iomem *rstout;
+	unsigned long clk_rate;
+	struct clk *clk;
+	const struct orion_watchdog_data *data;
+};
+
+static int orion_wdt_clock_init(struct platform_device *pdev,
+				struct orion_watchdog *dev)
 {
-	spin_lock(&wdt_lock);
+	int ret;
 
-	/* Reload watchdog duration */
-	writel(wdt_tclk * wdt_dev->timeout, wdt_reg + WDT_VAL);
+	dev->clk = clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dev->clk))
+		return PTR_ERR(dev->clk);
+	ret = clk_prepare_enable(dev->clk);
+	if (ret) {
+		clk_put(dev->clk);
+		return ret;
+	}
 
-	spin_unlock(&wdt_lock);
+	dev->clk_rate = clk_get_rate(dev->clk);
 	return 0;
 }
 
-static int orion_wdt_start(struct watchdog_device *wdt_dev)
+static int armada370_wdt_clock_init(struct platform_device *pdev,
+				    struct orion_watchdog *dev)
 {
-	u32 reg;
+	int ret;
 
-	spin_lock(&wdt_lock);
+	dev->clk = clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dev->clk))
+		return PTR_ERR(dev->clk);
+	ret = clk_prepare_enable(dev->clk);
+	if (ret) {
+		clk_put(dev->clk);
+		return ret;
+	}
+
+	/* Setup watchdog input clock */
+	atomic_io_modify(dev->reg + TIMER_CTRL,
+			WDT_A370_RATIO_MASK(WDT_A370_RATIO_SHIFT),
+			WDT_A370_RATIO_MASK(WDT_A370_RATIO_SHIFT));
+
+	dev->clk_rate = clk_get_rate(dev->clk) / WDT_A370_RATIO;
+	return 0;
+}
+
+static int armadaxp_wdt_clock_init(struct platform_device *pdev,
+				   struct orion_watchdog *dev)
+{
+	int ret;
+
+	dev->clk = of_clk_get_by_name(pdev->dev.of_node, "fixed");
+	if (IS_ERR(dev->clk))
+		return PTR_ERR(dev->clk);
+	ret = clk_prepare_enable(dev->clk);
+	if (ret) {
+		clk_put(dev->clk);
+		return ret;
+	}
+
+	/* Enable the fixed watchdog clock input */
+	atomic_io_modify(dev->reg + TIMER_CTRL,
+			 WDT_AXP_FIXED_ENABLE_BIT,
+			 WDT_AXP_FIXED_ENABLE_BIT);
+
+	dev->clk_rate = clk_get_rate(dev->clk);
+	return 0;
+}
+
+static int orion_wdt_ping(struct watchdog_device *wdt_dev)
+{
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
+	/* Reload watchdog duration */
+	writel(dev->clk_rate * wdt_dev->timeout,
+	       dev->reg + dev->data->wdt_counter_offset);
+	return 0;
+}
+
+static int armada370_start(struct watchdog_device *wdt_dev)
+{
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
 
 	/* Set watchdog duration */
-	writel(wdt_tclk * wdt_dev->timeout, wdt_reg + WDT_VAL);
+	writel(dev->clk_rate * wdt_dev->timeout,
+	       dev->reg + dev->data->wdt_counter_offset);
 
-	/* Clear watchdog timer interrupt */
-	writel(~WDT_INT_REQ, BRIDGE_CAUSE);
+	/* Clear the watchdog expiration bit */
+	atomic_io_modify(dev->reg + TIMER_A370_STATUS, WDT_A370_EXPIRED, 0);
 
 	/* Enable watchdog timer */
-	reg = readl(wdt_reg + TIMER_CTRL);
-	reg |= WDT_EN;
-	writel(reg, wdt_reg + TIMER_CTRL);
+	atomic_io_modify(dev->reg + TIMER_CTRL, dev->data->wdt_enable_bit,
+						dev->data->wdt_enable_bit);
+
+	atomic_io_modify(dev->rstout, dev->data->rstout_enable_bit,
+				      dev->data->rstout_enable_bit);
+	return 0;
+}
+
+static int orion_start(struct watchdog_device *wdt_dev)
+{
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
+
+	/* Set watchdog duration */
+	writel(dev->clk_rate * wdt_dev->timeout,
+	       dev->reg + dev->data->wdt_counter_offset);
+
+	/* Enable watchdog timer */
+	atomic_io_modify(dev->reg + TIMER_CTRL, dev->data->wdt_enable_bit,
+						dev->data->wdt_enable_bit);
 
 	/* Enable reset on watchdog */
-	reg = readl(RSTOUTn_MASK);
-	reg |= WDT_RESET_OUT_EN;
-	writel(reg, RSTOUTn_MASK);
+	atomic_io_modify(dev->rstout, dev->data->rstout_enable_bit,
+				      dev->data->rstout_enable_bit);
 
-	spin_unlock(&wdt_lock);
 	return 0;
 }
 
-static int orion_wdt_stop(struct watchdog_device *wdt_dev)
+static int orion_wdt_start(struct watchdog_device *wdt_dev)
 {
-	u32 reg;
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
 
-	spin_lock(&wdt_lock);
+	/* There are some per-SoC quirks to handle */
+	return dev->data->start(wdt_dev);
+}
+
+static int orion_wdt_stop(struct watchdog_device *wdt_dev)
+{
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
 
 	/* Disable reset on watchdog */
-	reg = readl(RSTOUTn_MASK);
-	reg &= ~WDT_RESET_OUT_EN;
-	writel(reg, RSTOUTn_MASK);
+	atomic_io_modify(dev->rstout, dev->data->rstout_enable_bit, 0);
 
 	/* Disable watchdog timer */
-	reg = readl(wdt_reg + TIMER_CTRL);
-	reg &= ~WDT_EN;
-	writel(reg, wdt_reg + TIMER_CTRL);
+	atomic_io_modify(dev->reg + TIMER_CTRL, dev->data->wdt_enable_bit, 0);
 
-	spin_unlock(&wdt_lock);
 	return 0;
 }
 
-static unsigned int orion_wdt_get_timeleft(struct watchdog_device *wdt_dev)
+static int orion_wdt_enabled(struct orion_watchdog *dev)
 {
-	unsigned int time_left;
+	bool enabled, running;
+
+	enabled = readl(dev->rstout) & dev->data->rstout_enable_bit;
+	running = readl(dev->reg + TIMER_CTRL) & dev->data->wdt_enable_bit;
 
-	spin_lock(&wdt_lock);
-	time_left = readl(wdt_reg + WDT_VAL) / wdt_tclk;
-	spin_unlock(&wdt_lock);
+	return enabled && running;
+}
 
-	return time_left;
+static unsigned int orion_wdt_get_timeleft(struct watchdog_device *wdt_dev)
+{
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
+	return readl(dev->reg + dev->data->wdt_counter_offset) / dev->clk_rate;
 }
 
 static int orion_wdt_set_timeout(struct watchdog_device *wdt_dev,
@@ -136,68 +239,188 @@ static const struct watchdog_ops orion_wdt_ops = {
 	.get_timeleft = orion_wdt_get_timeleft,
 };
 
-static struct watchdog_device orion_wdt = {
-	.info = &orion_wdt_info,
-	.ops = &orion_wdt_ops,
-	.min_timeout = 1,
+static irqreturn_t orion_wdt_irq(int irq, void *devid)
+{
+	panic("Watchdog Timeout");
+	return IRQ_HANDLED;
+}
+
+/*
+ * The original devicetree binding for this driver specified only
+ * one memory resource, so in order to keep DT backwards compatibility
+ * we try to fallback to a hardcoded register address, if the resource
+ * is missing from the devicetree.
+ */
+static void __iomem *orion_wdt_ioremap_rstout(struct platform_device *pdev,
+					      phys_addr_t internal_regs)
+{
+	struct resource *res;
+	phys_addr_t rstout;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (res)
+		return devm_ioremap(&pdev->dev, res->start,
+				    resource_size(res));
+
+	/* This workaround works only for "orion-wdt", DT-enabled */
+	if (!of_device_is_compatible(pdev->dev.of_node, "marvell,orion-wdt"))
+		return NULL;
+
+	rstout = internal_regs + ORION_RSTOUT_MASK_OFFSET;
+
+	WARN(1, FW_BUG "falling back to harcoded RSTOUT reg %pa\n", &rstout);
+	return devm_ioremap(&pdev->dev, rstout, 0x4);
+}
+
+static const struct orion_watchdog_data orion_data = {
+	.rstout_enable_bit = BIT(1),
+	.wdt_enable_bit = BIT(4),
+	.wdt_counter_offset = 0x24,
+	.clock_init = orion_wdt_clock_init,
+	.start = orion_start,
+};
+
+static const struct orion_watchdog_data armada370_data = {
+	.rstout_enable_bit = BIT(8),
+	.wdt_enable_bit = BIT(8),
+	.wdt_counter_offset = 0x34,
+	.clock_init = armada370_wdt_clock_init,
+	.start = armada370_start,
 };
 
+static const struct orion_watchdog_data armadaxp_data = {
+	.rstout_enable_bit = BIT(8),
+	.wdt_enable_bit = BIT(8),
+	.wdt_counter_offset = 0x34,
+	.clock_init = armadaxp_wdt_clock_init,
+	.start = armada370_start,
+};
+
+static const struct of_device_id orion_wdt_of_match_table[] = {
+	{
+		.compatible = "marvell,orion-wdt",
+		.data = &orion_data,
+	},
+	{
+		.compatible = "marvell,armada-370-wdt",
+		.data = &armada370_data,
+	},
+	{
+		.compatible = "marvell,armada-xp-wdt",
+		.data = &armadaxp_data,
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, orion_wdt_of_match_table);
+
 static int orion_wdt_probe(struct platform_device *pdev)
 {
+	struct orion_watchdog *dev;
+	const struct of_device_id *match;
+	unsigned int wdt_max_duration;	/* (seconds) */
 	struct resource *res;
-	int ret;
+	int ret, irq;
 
-	clk = devm_clk_get(&pdev->dev, NULL);
-	if (IS_ERR(clk)) {
-		dev_err(&pdev->dev, "Orion Watchdog missing clock\n");
-		return -ENODEV;
-	}
-	clk_prepare_enable(clk);
-	wdt_tclk = clk_get_rate(clk);
+	dev = devm_kzalloc(&pdev->dev, sizeof(struct orion_watchdog),
+			   GFP_KERNEL);
+	if (!dev)
+		return -ENOMEM;
+
+	match = of_match_device(orion_wdt_of_match_table, &pdev->dev);
+	if (!match)
+		/* Default legacy match */
+		match = &orion_wdt_of_match_table[0];
+
+	dev->wdt.info = &orion_wdt_info;
+	dev->wdt.ops = &orion_wdt_ops;
+	dev->wdt.min_timeout = 1;
+	dev->data = match->data;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res)
 		return -ENODEV;
-	wdt_reg = devm_ioremap(&pdev->dev, res->start, resource_size(res));
-	if (!wdt_reg)
-		return -ENOMEM;
 
-	wdt_max_duration = WDT_MAX_CYCLE_COUNT / wdt_tclk;
+	dev->reg = devm_ioremap(&pdev->dev, res->start,
+			       resource_size(res));
+	if (!dev->reg)
+		return -ENOMEM;
 
-	orion_wdt.timeout = wdt_max_duration;
-	orion_wdt.max_timeout = wdt_max_duration;
-	watchdog_init_timeout(&orion_wdt, heartbeat, &pdev->dev);
+	dev->rstout = orion_wdt_ioremap_rstout(pdev, res->start &
+						     INTERNAL_REGS_MASK);
+	if (!dev->rstout)
+		return -ENODEV;
 
-	watchdog_set_nowayout(&orion_wdt, nowayout);
-	ret = watchdog_register_device(&orion_wdt);
+	ret = dev->data->clock_init(pdev, dev);
 	if (ret) {
-		clk_disable_unprepare(clk);
+		dev_err(&pdev->dev, "cannot initialize clock\n");
 		return ret;
 	}
 
+	wdt_max_duration = WDT_MAX_CYCLE_COUNT / dev->clk_rate;
+
+	dev->wdt.timeout = wdt_max_duration;
+	dev->wdt.max_timeout = wdt_max_duration;
+	watchdog_init_timeout(&dev->wdt, heartbeat, &pdev->dev);
+
+	platform_set_drvdata(pdev, &dev->wdt);
+	watchdog_set_drvdata(&dev->wdt, dev);
+
+	/*
+	 * Let's make sure the watchdog is fully stopped, unless it's
+	 * explicitly enabled. This may be the case if the module was
+	 * removed and re-insterted, or if the bootloader explicitly
+	 * set a running watchdog before booting the kernel.
+	 */
+	if (!orion_wdt_enabled(dev))
+		orion_wdt_stop(&dev->wdt);
+
+	/* Request the IRQ only after the watchdog is disabled */
+	irq = platform_get_irq(pdev, 0);
+	if (irq > 0) {
+		/*
+		 * Not all supported platforms specify an interrupt for the
+		 * watchdog, so let's make it optional.
+		 */
+		ret = devm_request_irq(&pdev->dev, irq, orion_wdt_irq, 0,
+				       pdev->name, dev);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to request IRQ\n");
+			goto disable_clk;
+		}
+	}
+
+	watchdog_set_nowayout(&dev->wdt, nowayout);
+	ret = watchdog_register_device(&dev->wdt);
+	if (ret)
+		goto disable_clk;
+
 	pr_info("Initial timeout %d sec%s\n",
-		orion_wdt.timeout, nowayout ? ", nowayout" : "");
+		dev->wdt.timeout, nowayout ? ", nowayout" : "");
 	return 0;
+
+disable_clk:
+	clk_disable_unprepare(dev->clk);
+	clk_put(dev->clk);
+	return ret;
 }
 
 static int orion_wdt_remove(struct platform_device *pdev)
 {
-	watchdog_unregister_device(&orion_wdt);
-	clk_disable_unprepare(clk);
+	struct watchdog_device *wdt_dev = platform_get_drvdata(pdev);
+	struct orion_watchdog *dev = watchdog_get_drvdata(wdt_dev);
+
+	watchdog_unregister_device(wdt_dev);
+	clk_disable_unprepare(dev->clk);
+	clk_put(dev->clk);
 	return 0;
 }
 
 static void orion_wdt_shutdown(struct platform_device *pdev)
 {
-	orion_wdt_stop(&orion_wdt);
+	struct watchdog_device *wdt_dev = platform_get_drvdata(pdev);
+	orion_wdt_stop(wdt_dev);
 }
 
-static const struct of_device_id orion_wdt_of_match_table[] = {
-	{ .compatible = "marvell,orion-wdt", },
-	{},
-};
-MODULE_DEVICE_TABLE(of, orion_wdt_of_match_table);
-
 static struct platform_driver orion_wdt_driver = {
 	.probe		= orion_wdt_probe,
 	.remove		= orion_wdt_remove,