diff --git a/Documentation/devicetree/bindings/timer/img,pistachio-gptimer.txt b/Documentation/devicetree/bindings/timer/img,pistachio-gptimer.txt
new file mode 100644
index 0000000..7afce80
--- /dev/null
+++ b/Documentation/devicetree/bindings/timer/img,pistachio-gptimer.txt
@@ -0,0 +1,28 @@
+* Pistachio general-purpose timer based clocksource
+
+Required properties:
+ - compatible: "img,pistachio-gptimer".
+ - reg: Address range of the timer registers.
+ - interrupts: An interrupt for each of the four timers
+ - clocks: Should contain a clock specifier for each entry in clock-names
+ - clock-names: Should contain the following entries:
+                "sys", interface clock
+                "slow", slow counter clock
+                "fast", fast counter clock
+ - img,cr-periph: Must contain a phandle to the peripheral control
+		  syscon node.
+
+Example:
+	timer: timer@18102000 {
+		compatible = "img,pistachio-gptimer";
+		reg = <0x18102000 0x100>;
+		interrupts = <GIC_SHARED 60 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SHARED 61 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SHARED 62 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SHARED 63 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&clk_periph PERIPH_CLK_COUNTER_FAST>,
+		         <&clk_periph PERIPH_CLK_COUNTER_SLOW>,
+			 <&cr_periph SYS_CLK_TIMER>;
+		clock-names = "fast", "slow", "sys";
+		img,cr-periph = <&cr_periph>;
+	};
diff --git a/Documentation/features/debug/uprobes/arch-support.txt b/Documentation/features/debug/uprobes/arch-support.txt
index 4efe36c..d605c3f 100644
--- a/Documentation/features/debug/uprobes/arch-support.txt
+++ b/Documentation/features/debug/uprobes/arch-support.txt
@@ -22,7 +22,7 @@
     |        m68k: | TODO |
     |       metag: | TODO |
     |  microblaze: | TODO |
-    |        mips: | TODO |
+    |        mips: |  ok  |
     |     mn10300: | TODO |
     |       nios2: | TODO |
     |    openrisc: | TODO |
diff --git a/Documentation/sysrq.txt b/Documentation/sysrq.txt
index 0e307c9..267f393 100644
--- a/Documentation/sysrq.txt
+++ b/Documentation/sysrq.txt
@@ -119,6 +119,7 @@ On all -  write a character to /proc/sysrq-trigger.  e.g.:
 
 'x'	- Used by xmon interface on ppc/powerpc platforms.
           Show global PMU Registers on sparc64.
+          Dump all TLB entries on MIPS.
 
 'y'	- Show global CPU Registers [SPARC-64 specific]
 
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 4ab9a79..752acca 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -1,8 +1,10 @@
 config MIPS
 	bool
 	default y
+	select ARCH_SUPPORTS_UPROBES
 	select ARCH_MIGHT_HAVE_PC_PARPORT
 	select ARCH_MIGHT_HAVE_PC_SERIO
+	select ARCH_USE_CMPXCHG_LOCKREF if 64BIT
 	select HAVE_CONTEXT_TRACKING
 	select HAVE_GENERIC_DMA_COHERENT
 	select HAVE_IDE
@@ -13,7 +15,6 @@ config MIPS
 	select HAVE_ARCH_SECCOMP_FILTER
 	select HAVE_ARCH_TRACEHOOK
 	select HAVE_BPF_JIT if !CPU_MICROMIPS
-	select ARCH_HAVE_CUSTOM_GPIO_H
 	select HAVE_FUNCTION_TRACER
 	select HAVE_DYNAMIC_FTRACE
 	select HAVE_FTRACE_MCOUNT_RECORD
@@ -409,6 +410,7 @@ config MIPS_MALTA
 	select CEVT_R4K
 	select CSRC_R4K
 	select CLKSRC_MIPS_GIC
+	select COMMON_CLK
 	select DMA_MAYBE_COHERENT
 	select GENERIC_ISA_DMA
 	select HAVE_PCSPKR_PLATFORM
@@ -459,6 +461,7 @@ config MIPS_SEAD3
 	select CEVT_R4K
 	select CSRC_R4K
 	select CLKSRC_MIPS_GIC
+	select COMMON_CLK
 	select CPU_MIPSR2_IRQ_VI
 	select CPU_MIPSR2_IRQ_EI
 	select DMA_NONCOHERENT
@@ -899,6 +902,7 @@ config NLM_XLP_BOARD
 	select SYS_SUPPORTS_32BIT_KERNEL
 	select SYS_SUPPORTS_64BIT_KERNEL
 	select ARCH_PHYS_ADDR_T_64BIT
+	select ARCH_REQUIRE_GPIOLIB
 	select SYS_SUPPORTS_BIG_ENDIAN
 	select SYS_SUPPORTS_LITTLE_ENDIAN
 	select SYS_SUPPORTS_HIGHMEM
@@ -948,6 +952,7 @@ source "arch/mips/jazz/Kconfig"
 source "arch/mips/jz4740/Kconfig"
 source "arch/mips/lantiq/Kconfig"
 source "arch/mips/lasat/Kconfig"
+source "arch/mips/pistachio/Kconfig"
 source "arch/mips/pmcs-msp71xx/Kconfig"
 source "arch/mips/ralink/Kconfig"
 source "arch/mips/sgi-ip27/Kconfig"
@@ -1041,6 +1046,9 @@ config FW_CFE
 config ARCH_DMA_ADDR_T_64BIT
 	def_bool (HIGHMEM && ARCH_PHYS_ADDR_T_64BIT) || 64BIT
 
+config ARCH_SUPPORTS_UPROBES
+	bool
+
 config DMA_MAYBE_COHERENT
 	select DMA_NONCOHERENT
 	bool
@@ -1364,7 +1372,7 @@ config CPU_MIPS32_R2
 	  otherwise CPU_MIPS32_R1 is a safe bet for any MIPS32 system.
 
 config CPU_MIPS32_R6
-	bool "MIPS32 Release 6 (EXPERIMENTAL)"
+	bool "MIPS32 Release 6"
 	depends on SYS_HAS_CPU_MIPS32_R6
 	select CPU_HAS_PREFETCH
 	select CPU_SUPPORTS_32BIT_KERNEL
@@ -1415,7 +1423,7 @@ config CPU_MIPS64_R2
 	  otherwise CPU_MIPS64_R1 is a safe bet for any MIPS64 system.
 
 config CPU_MIPS64_R6
-	bool "MIPS64 Release 6 (EXPERIMENTAL)"
+	bool "MIPS64 Release 6"
 	depends on SYS_HAS_CPU_MIPS64_R6
 	select CPU_HAS_PREFETCH
 	select CPU_SUPPORTS_32BIT_KERNEL
@@ -1965,6 +1973,7 @@ config 32BIT
 	select TRAD_SIGNALS
 	help
 	  Select this option if you want to build a 32-bit kernel.
+
 config 64BIT
 	bool "64-bit kernel"
 	depends on CPU_SUPPORTS_64BIT_KERNEL && SYS_SUPPORTS_64BIT_KERNEL
@@ -2110,7 +2119,7 @@ config CPU_R4K_CACHE_TLB
 
 config MIPS_MT_SMP
 	bool "MIPS MT SMP support (1 TC on each available VPE)"
-	depends on SYS_SUPPORTS_MULTITHREADING
+	depends on SYS_SUPPORTS_MULTITHREADING && !CPU_MIPSR6
 	select CPU_MIPSR2_IRQ_VI
 	select CPU_MIPSR2_IRQ_EI
 	select SYNC_R4K
@@ -2211,7 +2220,7 @@ config MIPS_VPE_APSP_API_MT
 
 config MIPS_CMP
 	bool "MIPS CMP framework support (DEPRECATED)"
-	depends on SYS_SUPPORTS_MIPS_CMP
+	depends on SYS_SUPPORTS_MIPS_CMP && !CPU_MIPSR6
 	select MIPS_GIC_IPI
 	select SMP
 	select SYNC_R4K
@@ -2228,7 +2237,7 @@ config MIPS_CMP
 
 config MIPS_CPS
 	bool "MIPS Coherent Processing System support"
-	depends on SYS_SUPPORTS_MIPS_CPS
+	depends on SYS_SUPPORTS_MIPS_CPS && !CPU_MIPSR6
 	select MIPS_CM
 	select MIPS_CPC
 	select MIPS_CPS_PM if HOTPLUG_CPU
@@ -2303,7 +2312,7 @@ config CPU_MICROMIPS
 endchoice
 
 config CPU_HAS_MSA
-	bool "Support for the MIPS SIMD Architecture (EXPERIMENTAL)"
+	bool "Support for the MIPS SIMD Architecture"
 	depends on CPU_SUPPORTS_MSA
 	depends on 64BIT || MIPS_O32_FP64_SUPPORT
 	help
@@ -2643,7 +2652,7 @@ config SECCOMP
 	  If unsure, say Y. Only embedded should say N here.
 
 config MIPS_O32_FP64_SUPPORT
-	bool "Support for O32 binaries using 64-bit FP (EXPERIMENTAL)"
+	bool "Support for O32 binaries using 64-bit FP"
 	depends on 32BIT || MIPS32_O32
 	help
 	  When this is enabled, the kernel will support use of 64-bit floating
diff --git a/arch/mips/Kconfig.debug b/arch/mips/Kconfig.debug
index 3a2b775..e250524 100644
--- a/arch/mips/Kconfig.debug
+++ b/arch/mips/Kconfig.debug
@@ -87,15 +87,6 @@ config SB1XXX_CORELIS
 	  Select compile flags that produce code that can be processed by the
 	  Corelis mksym utility and UDB Emulator.
 
-config RUNTIME_DEBUG
-	bool "Enable run-time debugging"
-	depends on DEBUG_KERNEL
-	help
-	  If you say Y here, some debugging macros will do run-time checking.
-	  If you say N here, those macros will mostly turn to no-ops.  See
-	  arch/mips/include/asm/debug.h for debugging macros.
-	  If unsure, say N.
-
 config DEBUG_ZBOOT
 	bool "Enable compressed kernel support debugging"
 	depends on DEBUG_KERNEL && SYS_SUPPORTS_ZBOOT
diff --git a/arch/mips/alchemy/Kconfig b/arch/mips/alchemy/Kconfig
index b962898..7fa2488 100644
--- a/arch/mips/alchemy/Kconfig
+++ b/arch/mips/alchemy/Kconfig
@@ -6,13 +6,6 @@ config ALCHEMY_GPIOINT_AU1000
 config ALCHEMY_GPIOINT_AU1300
 	bool
 
-# select this in your board config if you don't want to use the gpio
-# namespace as documented in the manuals.  In this case however you need
-# to create the necessary gpio_* functions in your board code/headers!
-# see arch/mips/include/asm/mach-au1x00/gpio.h   for more information.
-config ALCHEMY_GPIO_INDIRECT
-	def_bool n
-
 choice
 	prompt "Machine type"
 	depends on MIPS_ALCHEMY
diff --git a/arch/mips/alchemy/board-gpr.c b/arch/mips/alchemy/board-gpr.c
index acf9a2a..79efe4c 100644
--- a/arch/mips/alchemy/board-gpr.c
+++ b/arch/mips/alchemy/board-gpr.c
@@ -34,6 +34,7 @@
 #include <asm/idle.h>
 #include <asm/reboot.h>
 #include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/gpio-au1000.h>
 #include <prom.h>
 
 const char *get_system_type(void)
diff --git a/arch/mips/alchemy/board-mtx1.c b/arch/mips/alchemy/board-mtx1.c
index 1e3b102..85bb756 100644
--- a/arch/mips/alchemy/board-mtx1.c
+++ b/arch/mips/alchemy/board-mtx1.c
@@ -32,6 +32,7 @@
 #include <asm/bootinfo.h>
 #include <asm/reboot.h>
 #include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/gpio-au1000.h>
 #include <asm/mach-au1x00/au1xxx_eth.h>
 #include <prom.h>
 
diff --git a/arch/mips/alchemy/common/Makefile b/arch/mips/alchemy/common/Makefile
index f64744f..23800b8 100644
--- a/arch/mips/alchemy/common/Makefile
+++ b/arch/mips/alchemy/common/Makefile
@@ -5,10 +5,5 @@
 # Makefile for the Alchemy Au1xx0 CPUs, generic files.
 #
 
-obj-y += prom.o time.o clock.o platform.o power.o \
+obj-y += prom.o time.o clock.o platform.o power.o gpiolib.o \
 	 setup.o sleeper.o dma.o dbdma.o vss.o irq.o usb.o
-
-# optional gpiolib support
-ifeq ($(CONFIG_ALCHEMY_GPIO_INDIRECT),)
- obj-$(CONFIG_GPIOLIB) += gpiolib.o
-endif
diff --git a/arch/mips/alchemy/common/irq.c b/arch/mips/alchemy/common/irq.c
index 6cb60ab..4c496c5 100644
--- a/arch/mips/alchemy/common/irq.c
+++ b/arch/mips/alchemy/common/irq.c
@@ -491,7 +491,7 @@ static int au1x_ic_settype(struct irq_data *d, unsigned int flow_type)
 	default:
 		ret = -EINVAL;
 	}
-	__irq_set_chip_handler_name_locked(d->irq, chip, handler, name);
+	irq_set_chip_handler_name_locked(d, chip, handler, name);
 
 	wmb();
 
@@ -703,7 +703,7 @@ static int au1300_gpic_settype(struct irq_data *d, unsigned int type)
 		return -EINVAL;
 	}
 
-	__irq_set_chip_handler_name_locked(d->irq, &au1300_gpic, hdl, name);
+	irq_set_chip_handler_name_locked(d, &au1300_gpic, hdl, name);
 
 	au1300_gpic_chgcfg(d->irq - ALCHEMY_GPIC_INT_BASE, GPIC_CFG_IC_MASK, s);
 
diff --git a/arch/mips/alchemy/common/time.c b/arch/mips/alchemy/common/time.c
index 50e17e1..f99d3ec 100644
--- a/arch/mips/alchemy/common/time.c
+++ b/arch/mips/alchemy/common/time.c
@@ -69,11 +69,6 @@ static int au1x_rtcmatch2_set_next_event(unsigned long delta,
 	return 0;
 }
 
-static void au1x_rtcmatch2_set_mode(enum clock_event_mode mode,
-				    struct clock_event_device *cd)
-{
-}
-
 static irqreturn_t au1x_rtcmatch2_irq(int irq, void *dev_id)
 {
 	struct clock_event_device *cd = dev_id;
@@ -86,7 +81,6 @@ static struct clock_event_device au1x_rtcmatch2_clockdev = {
 	.features	= CLOCK_EVT_FEAT_ONESHOT,
 	.rating		= 1500,
 	.set_next_event = au1x_rtcmatch2_set_next_event,
-	.set_mode	= au1x_rtcmatch2_set_mode,
 	.cpumask	= cpu_all_mask,
 };
 
diff --git a/arch/mips/alchemy/devboards/bcsr.c b/arch/mips/alchemy/devboards/bcsr.c
index c98c9ea..324ad72 100644
--- a/arch/mips/alchemy/devboards/bcsr.c
+++ b/arch/mips/alchemy/devboards/bcsr.c
@@ -8,6 +8,7 @@
  */
 
 #include <linux/interrupt.h>
+#include <linux/irqchip/chained_irq.h>
 #include <linux/module.h>
 #include <linux/spinlock.h>
 #include <linux/irq.h>
@@ -88,10 +89,11 @@ EXPORT_SYMBOL_GPL(bcsr_mod);
 static void bcsr_csc_handler(unsigned int irq, struct irq_desc *d)
 {
 	unsigned short bisr = __raw_readw(bcsr_virt + BCSR_REG_INTSTAT);
+	struct irq_chip *chip = irq_desc_get_chip(d);
 
-	disable_irq_nosync(irq);
+	chained_irq_enter(chip, d);
 	generic_handle_irq(bcsr_csc_base + __ffs(bisr));
-	enable_irq(irq);
+	chained_irq_exit(chip, d);
 }
 
 static void bcsr_irq_mask(struct irq_data *d)
diff --git a/arch/mips/alchemy/devboards/db1000.c b/arch/mips/alchemy/devboards/db1000.c
index 001102e..bdeed9d 100644
--- a/arch/mips/alchemy/devboards/db1000.c
+++ b/arch/mips/alchemy/devboards/db1000.c
@@ -33,6 +33,7 @@
 #include <linux/spi/spi_gpio.h>
 #include <linux/spi/ads7846.h>
 #include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/gpio-au1000.h>
 #include <asm/mach-au1x00/au1000_dma.h>
 #include <asm/mach-au1x00/au1100_mmc.h>
 #include <asm/mach-db1x00/bcsr.h>
diff --git a/arch/mips/alchemy/devboards/db1300.c b/arch/mips/alchemy/devboards/db1300.c
index 1c64fdb..b580770 100644
--- a/arch/mips/alchemy/devboards/db1300.c
+++ b/arch/mips/alchemy/devboards/db1300.c
@@ -24,6 +24,7 @@
 #include <linux/wm97xx.h>
 
 #include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/gpio-au1300.h>
 #include <asm/mach-au1x00/au1100_mmc.h>
 #include <asm/mach-au1x00/au1200fb.h>
 #include <asm/mach-au1x00/au1xxx_dbdma.h>
diff --git a/arch/mips/alchemy/devboards/db1550.c b/arch/mips/alchemy/devboards/db1550.c
index 0fd5177..5740bcf 100644
--- a/arch/mips/alchemy/devboards/db1550.c
+++ b/arch/mips/alchemy/devboards/db1550.c
@@ -20,6 +20,7 @@
 #include <linux/spi/flash.h>
 #include <asm/bootinfo.h>
 #include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/gpio-au1000.h>
 #include <asm/mach-au1x00/au1xxx_eth.h>
 #include <asm/mach-au1x00/au1xxx_dbdma.h>
 #include <asm/mach-au1x00/au1xxx_psc.h>
diff --git a/arch/mips/alchemy/devboards/pm.c b/arch/mips/alchemy/devboards/pm.c
index bfeb8f3..93024dc 100644
--- a/arch/mips/alchemy/devboards/pm.c
+++ b/arch/mips/alchemy/devboards/pm.c
@@ -9,7 +9,7 @@
 #include <linux/suspend.h>
 #include <linux/sysfs.h>
 #include <asm/mach-au1x00/au1000.h>
-#include <asm/mach-au1x00/gpio.h>
+#include <asm/mach-au1x00/gpio-au1000.h>
 #include <asm/mach-db1x00/bcsr.h>
 
 /*
diff --git a/arch/mips/ar7/gpio.c b/arch/mips/ar7/gpio.c
index d8dbd8f..f493045 100644
--- a/arch/mips/ar7/gpio.c
+++ b/arch/mips/ar7/gpio.c
@@ -21,7 +21,10 @@
 #include <linux/module.h>
 #include <linux/gpio.h>
 
-#include <asm/mach-ar7/gpio.h>
+#include <asm/mach-ar7/ar7.h>
+
+#define AR7_GPIO_MAX 32
+#define TITAN_GPIO_MAX 51
 
 struct ar7_gpio_chip {
 	void __iomem		*regs;
diff --git a/arch/mips/ar7/platform.c b/arch/mips/ar7/platform.c
index 298b977..58fca9a 100644
--- a/arch/mips/ar7/platform.c
+++ b/arch/mips/ar7/platform.c
@@ -39,7 +39,6 @@
 
 #include <asm/addrspace.h>
 #include <asm/mach-ar7/ar7.h>
-#include <asm/mach-ar7/gpio.h>
 #include <asm/mach-ar7/prom.h>
 
 /*****************************************************************************
diff --git a/arch/mips/ar7/setup.c b/arch/mips/ar7/setup.c
index 820b7a3..7bb9a67 100644
--- a/arch/mips/ar7/setup.c
+++ b/arch/mips/ar7/setup.c
@@ -23,7 +23,6 @@
 #include <asm/reboot.h>
 #include <asm/mach-ar7/ar7.h>
 #include <asm/mach-ar7/prom.h>
-#include <asm/mach-ar7/gpio.h>
 
 static void ar7_machine_restart(char *command)
 {
diff --git a/arch/mips/ath79/Makefile b/arch/mips/ath79/Makefile
index 5c9ff69..fcc382c 100644
--- a/arch/mips/ath79/Makefile
+++ b/arch/mips/ath79/Makefile
@@ -8,7 +8,7 @@
 # under the terms of the GNU General Public License version 2 as published
 # by the Free Software Foundation.
 
-obj-y	:= prom.o setup.o irq.o common.o clock.o gpio.o
+obj-y	:= prom.o setup.o irq.o common.o clock.o
 
 obj-$(CONFIG_EARLY_PRINTK)		+= early_printk.o
 obj-$(CONFIG_PCI)			+= pci.o
diff --git a/arch/mips/ath79/common.h b/arch/mips/ath79/common.h
index e5ea712..ca7cc19 100644
--- a/arch/mips/ath79/common.h
+++ b/arch/mips/ath79/common.h
@@ -25,9 +25,6 @@ unsigned long ath79_get_sys_clk_rate(const char *id);
 void ath79_ddr_ctrl_init(void);
 void ath79_ddr_wb_flush(unsigned int reg);
 
-void ath79_gpio_function_enable(u32 mask);
-void ath79_gpio_function_disable(u32 mask);
-void ath79_gpio_function_setup(u32 set, u32 clear);
 void ath79_gpio_init(void);
 
 #endif /* __ATH79_COMMON_H */
diff --git a/arch/mips/ath79/irq.c b/arch/mips/ath79/irq.c
index 2021be2..807132b 100644
--- a/arch/mips/ath79/irq.c
+++ b/arch/mips/ath79/irq.c
@@ -123,8 +123,6 @@ static void ar934x_ip2_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 {
 	u32 status;
 
-	disable_irq_nosync(irq);
-
 	status = ath79_reset_rr(AR934X_RESET_REG_PCIE_WMAC_INT_STATUS);
 
 	if (status & AR934X_PCIE_WMAC_INT_PCIE_ALL) {
@@ -136,8 +134,6 @@ static void ar934x_ip2_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 	} else {
 		spurious_interrupt();
 	}
-
-	enable_irq(irq);
 }
 
 static void ar934x_ip2_irq_init(void)
@@ -156,14 +152,12 @@ static void qca955x_ip2_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 {
 	u32 status;
 
-	disable_irq_nosync(irq);
-
 	status = ath79_reset_rr(QCA955X_RESET_REG_EXT_INT_STATUS);
 	status &= QCA955X_EXT_INT_PCIE_RC1_ALL | QCA955X_EXT_INT_WMAC_ALL;
 
 	if (status == 0) {
 		spurious_interrupt();
-		goto enable;
+		return;
 	}
 
 	if (status & QCA955X_EXT_INT_PCIE_RC1_ALL) {
@@ -175,17 +169,12 @@ static void qca955x_ip2_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 		/* TODO: flush DDR? */
 		generic_handle_irq(ATH79_IP2_IRQ(1));
 	}
-
-enable:
-	enable_irq(irq);
 }
 
 static void qca955x_ip3_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 {
 	u32 status;
 
-	disable_irq_nosync(irq);
-
 	status = ath79_reset_rr(QCA955X_RESET_REG_EXT_INT_STATUS);
 	status &= QCA955X_EXT_INT_PCIE_RC2_ALL |
 		  QCA955X_EXT_INT_USB1 |
@@ -193,7 +182,7 @@ static void qca955x_ip3_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 
 	if (status == 0) {
 		spurious_interrupt();
-		goto enable;
+		return;
 	}
 
 	if (status & QCA955X_EXT_INT_USB1) {
@@ -210,9 +199,6 @@ static void qca955x_ip3_irq_dispatch(unsigned int irq, struct irq_desc *desc)
 		/* TODO: flush DDR? */
 		generic_handle_irq(ATH79_IP3_IRQ(2));
 	}
-
-enable:
-	enable_irq(irq);
 }
 
 static void qca955x_irq_init(void)
diff --git a/arch/mips/bcm47xx/buttons.c b/arch/mips/bcm47xx/buttons.c
index 08a4abf..52caa75 100644
--- a/arch/mips/bcm47xx/buttons.c
+++ b/arch/mips/bcm47xx/buttons.c
@@ -396,10 +396,9 @@ static int __init bcm47xx_buttons_copy(const struct gpio_keys_button *buttons,
 {
 	size_t size = nbuttons * sizeof(*buttons);
 
-	bcm47xx_button_pdata.buttons = kmalloc(size, GFP_KERNEL);
+	bcm47xx_button_pdata.buttons = kmemdup(buttons, size, GFP_KERNEL);
 	if (!bcm47xx_button_pdata.buttons)
 		return -ENOMEM;
-	memcpy(bcm47xx_button_pdata.buttons, buttons, size);
 	bcm47xx_button_pdata.nbuttons = nbuttons;
 
 	return 0;
diff --git a/arch/mips/bcm63xx/irq.c b/arch/mips/bcm63xx/irq.c
index e3e808a..1a47ec2 100644
--- a/arch/mips/bcm63xx/irq.c
+++ b/arch/mips/bcm63xx/irq.c
@@ -60,7 +60,7 @@ static inline int enable_irq_for_cpu(int cpu, struct irq_data *d,
 	if (m)
 		enable &= cpumask_test_cpu(cpu, m);
 	else if (irqd_affinity_was_set(d))
-		enable &= cpumask_test_cpu(cpu, d->affinity);
+		enable &= cpumask_test_cpu(cpu, irq_data_get_affinity_mask(d));
 #endif
 	return enable;
 }
@@ -365,9 +365,9 @@ static int bcm63xx_external_irq_set_type(struct irq_data *d,
 
 	irqd_set_trigger_type(d, flow_type);
 	if (flow_type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH))
-		__irq_set_handler_locked(d->irq, handle_level_irq);
+		irq_set_handler_locked(d, handle_level_irq);
 	else
-		__irq_set_handler_locked(d->irq, handle_edge_irq);
+		irq_set_handler_locked(d, handle_edge_irq);
 
 	return IRQ_SET_MASK_OK_NOCOPY;
 }
diff --git a/arch/mips/boot/compressed/Makefile b/arch/mips/boot/compressed/Makefile
index dc91bde..d5bdee1 100644
--- a/arch/mips/boot/compressed/Makefile
+++ b/arch/mips/boot/compressed/Makefile
@@ -78,7 +78,7 @@ endif
 
 vmlinuzobjs-y += $(obj)/piggy.o
 
-quiet_cmd_zld = LD	$@
+quiet_cmd_zld = LD      $@
       cmd_zld = $(LD) $(LDFLAGS) -Ttext $(VMLINUZ_LOAD_ADDRESS) -T $< $(vmlinuzobjs-y) -o $@
 quiet_cmd_strip = STRIP	  $@
       cmd_strip = $(STRIP) -s $@
diff --git a/arch/mips/boot/dts/netlogic/xlp_evp.dts b/arch/mips/boot/dts/netlogic/xlp_evp.dts
index 89ad048..ec16ec2 100644
--- a/arch/mips/boot/dts/netlogic/xlp_evp.dts
+++ b/arch/mips/boot/dts/netlogic/xlp_evp.dts
@@ -110,6 +110,18 @@
 				read-only;
 			};
 		};
+
+		gpio: xlp_gpio@34100 {
+			compatible = "netlogic,xlp832-gpio";
+			reg = <0 0x34100 0x1000>;
+			#gpio-cells = <2>;
+			gpio-controller;
+
+			#interrupt-cells = <2>;
+			interrupt-parent = <&pic>;
+			interrupts = <39>;
+			interrupt-controller;
+		};
 	};
 
 	chosen {
diff --git a/arch/mips/boot/dts/netlogic/xlp_fvp.dts b/arch/mips/boot/dts/netlogic/xlp_fvp.dts
index 63e62b7..4bcebe6 100644
--- a/arch/mips/boot/dts/netlogic/xlp_fvp.dts
+++ b/arch/mips/boot/dts/netlogic/xlp_fvp.dts
@@ -110,6 +110,18 @@
 				read-only;
 			};
 		};
+
+		gpio: xlp_gpio@34100 {
+			compatible = "netlogic,xlp208-gpio";
+			reg = <0 0x34100 0x1000>;
+			#gpio-cells = <2>;
+			gpio-controller;
+
+			#interrupt-cells = <2>;
+			interrupt-parent = <&pic>;
+			interrupts = <39>;
+			interrupt-controller;
+		};
 	};
 
 	chosen {
diff --git a/arch/mips/boot/dts/netlogic/xlp_gvp.dts b/arch/mips/boot/dts/netlogic/xlp_gvp.dts
index bb4ecd1..b3ccb82 100644
--- a/arch/mips/boot/dts/netlogic/xlp_gvp.dts
+++ b/arch/mips/boot/dts/netlogic/xlp_gvp.dts
@@ -69,6 +69,17 @@
 			};
 		};
 
+		gpio: xlp_gpio@114100 {
+			compatible = "netlogic,xlp980-gpio";
+			reg = <0 0x114100 0x1000>;
+			#gpio-cells = <2>;
+			gpio-controller;
+
+			#interrupt-cells = <2>;
+			interrupt-parent = <&pic>;
+			interrupts = <39>;
+			interrupt-controller;
+		};
 	};
 
 	chosen {
diff --git a/arch/mips/boot/dts/netlogic/xlp_rvp.dts b/arch/mips/boot/dts/netlogic/xlp_rvp.dts
index 7188aed..3783639 100644
--- a/arch/mips/boot/dts/netlogic/xlp_rvp.dts
+++ b/arch/mips/boot/dts/netlogic/xlp_rvp.dts
@@ -69,6 +69,17 @@
 			};
 		};
 
+		gpio: xlp_gpio@114100 {
+			compatible = "netlogic,xlp532-gpio";
+			reg = <0 0x114100 0x1000>;
+			#gpio-cells = <2>;
+			gpio-controller;
+
+			#interrupt-cells = <2>;
+			interrupt-parent = <&pic>;
+			interrupts = <39>;
+			interrupt-controller;
+		};
 	};
 
 	chosen {
diff --git a/arch/mips/boot/dts/netlogic/xlp_svp.dts b/arch/mips/boot/dts/netlogic/xlp_svp.dts
index 1ebd00e..44d6640 100644
--- a/arch/mips/boot/dts/netlogic/xlp_svp.dts
+++ b/arch/mips/boot/dts/netlogic/xlp_svp.dts
@@ -110,6 +110,18 @@
 				read-only;
 			};
 		};
+
+		gpio: xlp_gpio@34100 {
+			compatible = "netlogic,xlp316-gpio";
+			reg = <0 0x34100 0x1000>;
+			#gpio-cells = <2>;
+			gpio-controller;
+
+			#interrupt-cells = <2>;
+			interrupt-parent = <&pic>;
+			interrupts = <39>;
+			interrupt-controller;
+		};
 	};
 
 	chosen {
diff --git a/arch/mips/cavium-octeon/executive/cvmx-helper-board.c b/arch/mips/cavium-octeon/executive/cvmx-helper-board.c
index 9eb0fee..36e30d6 100644
--- a/arch/mips/cavium-octeon/executive/cvmx-helper-board.c
+++ b/arch/mips/cavium-octeon/executive/cvmx-helper-board.c
@@ -195,6 +195,12 @@ int cvmx_helper_board_get_mii_address(int ipd_port)
 			return 8;
 		else
 			return -1;
+	case CVMX_BOARD_TYPE_KONTRON_S1901:
+		if (ipd_port == CVMX_HELPER_BOARD_MGMT_IPD_PORT)
+			return 1;
+		else
+			return -1;
+
 	}
 
 	/* Some unknown board. Somebody forgot to update this function... */
diff --git a/arch/mips/cavium-octeon/executive/cvmx-helper-util.c b/arch/mips/cavium-octeon/executive/cvmx-helper-util.c
index 453d7f6..b45b297 100644
--- a/arch/mips/cavium-octeon/executive/cvmx-helper-util.c
+++ b/arch/mips/cavium-octeon/executive/cvmx-helper-util.c
@@ -95,9 +95,9 @@ int cvmx_helper_dump_packet(cvmx_wqe_t *work)
 	uint8_t *data_address;
 	uint8_t *end_of_data;
 
-	cvmx_dprintf("Packet Length:   %u\n", work->len);
-	cvmx_dprintf("	  Input Port:  %u\n", work->ipprt);
-	cvmx_dprintf("	  QoS:	       %u\n", work->qos);
+	cvmx_dprintf("Packet Length:   %u\n", work->word1.len);
+	cvmx_dprintf("	  Input Port:  %u\n", cvmx_wqe_get_port(work));
+	cvmx_dprintf("	  QoS:	       %u\n", cvmx_wqe_get_qos(work));
 	cvmx_dprintf("	  Buffers:     %u\n", work->word2.s.bufs);
 
 	if (work->word2.s.bufs == 0) {
@@ -127,7 +127,7 @@ int cvmx_helper_dump_packet(cvmx_wqe_t *work)
 		}
 	} else
 		buffer_ptr = work->packet_ptr;
-	remaining_bytes = work->len;
+	remaining_bytes = work->word1.len;
 
 	while (remaining_bytes) {
 		start_of_buffer =
@@ -382,6 +382,10 @@ int cvmx_helper_get_ipd_port(int interface, int port)
 		return port + 32;
 	case 3:
 		return port + 36;
+	case 4:
+		return port + 40;
+	case 5:
+		return port + 44;
 	}
 	return -1;
 }
@@ -404,6 +408,10 @@ int cvmx_helper_get_interface_num(int ipd_port)
 		return 2;
 	else if (ipd_port < 40)
 		return 3;
+	else if (ipd_port < 44)
+		return 4;
+	else if (ipd_port < 48)
+		return 5;
 	else
 		cvmx_dprintf("cvmx_helper_get_interface_num: Illegal IPD "
 			     "port number\n");
@@ -428,6 +436,10 @@ int cvmx_helper_get_interface_index_num(int ipd_port)
 		return ipd_port & 3;
 	else if (ipd_port < 40)
 		return ipd_port & 3;
+	else if (ipd_port < 44)
+		return ipd_port & 3;
+	else if (ipd_port < 48)
+		return ipd_port & 3;
 	else
 		cvmx_dprintf("cvmx_helper_get_interface_index_num: "
 			     "Illegal IPD port number\n");
diff --git a/arch/mips/cavium-octeon/executive/cvmx-helper-xaui.c b/arch/mips/cavium-octeon/executive/cvmx-helper-xaui.c
index 7653b7e..a56ee59 100644
--- a/arch/mips/cavium-octeon/executive/cvmx-helper-xaui.c
+++ b/arch/mips/cavium-octeon/executive/cvmx-helper-xaui.c
@@ -124,6 +124,13 @@ int __cvmx_helper_xaui_enable(int interface)
 	union cvmx_gmxx_tx_int_en gmx_tx_int_en;
 	union cvmx_pcsxx_int_en_reg pcsx_int_en_reg;
 
+	/* Setup PKND */
+	if (octeon_has_feature(OCTEON_FEATURE_PKND)) {
+		gmx_cfg.u64 = cvmx_read_csr(CVMX_GMXX_PRTX_CFG(0, interface));
+		gmx_cfg.s.pknd = cvmx_helper_get_ipd_port(interface, 0);
+		cvmx_write_csr(CVMX_GMXX_PRTX_CFG(0, interface), gmx_cfg.u64);
+	}
+
 	/* (1) Interface has already been enabled. */
 
 	/* (2) Disable GMX. */
@@ -151,7 +158,12 @@ int __cvmx_helper_xaui_enable(int interface)
 	/* (4)c Aply reset sequence */
 	xauiCtl.u64 = cvmx_read_csr(CVMX_PCSXX_CONTROL1_REG(interface));
 	xauiCtl.s.lo_pwr = 0;
-	xauiCtl.s.reset = 1;
+
+	/* Issuing a reset here seems to hang some CN68XX chips. */
+	if (!OCTEON_IS_MODEL(OCTEON_CN68XX_PASS1_X) &&
+	    !OCTEON_IS_MODEL(OCTEON_CN68XX_PASS2_X))
+		xauiCtl.s.reset = 1;
+
 	cvmx_write_csr(CVMX_PCSXX_CONTROL1_REG(interface), xauiCtl.u64);
 
 	/* Wait for PCS to come out of reset */
diff --git a/arch/mips/cavium-octeon/executive/cvmx-helper.c b/arch/mips/cavium-octeon/executive/cvmx-helper.c
index 7e5cf7a..376701f 100644
--- a/arch/mips/cavium-octeon/executive/cvmx-helper.c
+++ b/arch/mips/cavium-octeon/executive/cvmx-helper.c
@@ -83,6 +83,8 @@ static cvmx_helper_link_info_t
  */
 int cvmx_helper_get_number_of_interfaces(void)
 {
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX))
+		return 9;
 	if (OCTEON_IS_MODEL(OCTEON_CN56XX) || OCTEON_IS_MODEL(OCTEON_CN52XX))
 		return 4;
 	else
@@ -656,6 +658,21 @@ static int __cvmx_helper_global_setup_pko(void)
 	fau_to.s.tout_val = 0xfff;
 	fau_to.s.tout_enb = 0;
 	cvmx_write_csr(CVMX_IOB_FAU_TIMEOUT, fau_to.u64);
+
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+		union cvmx_pko_reg_min_pkt min_pkt;
+
+		min_pkt.u64 = 0;
+		min_pkt.s.size1 = 59;
+		min_pkt.s.size2 = 59;
+		min_pkt.s.size3 = 59;
+		min_pkt.s.size4 = 59;
+		min_pkt.s.size5 = 59;
+		min_pkt.s.size6 = 59;
+		min_pkt.s.size7 = 59;
+		cvmx_write_csr(CVMX_PKO_REG_MIN_PKT, min_pkt.u64);
+	}
+
 	return 0;
 }
 
diff --git a/arch/mips/cavium-octeon/executive/cvmx-pko.c b/arch/mips/cavium-octeon/executive/cvmx-pko.c
index 008b881..87be167 100644
--- a/arch/mips/cavium-octeon/executive/cvmx-pko.c
+++ b/arch/mips/cavium-octeon/executive/cvmx-pko.c
@@ -39,6 +39,143 @@
  * Internal state of packet output
  */
 
+static int __cvmx_pko_int(int interface, int index)
+{
+	switch (interface) {
+	case 0:
+		return index;
+	case 1:
+		return 4;
+	case 2:
+		return index + 0x08;
+	case 3:
+		return index + 0x0c;
+	case 4:
+		return index + 0x10;
+	case 5:
+		return 0x1c;
+	case 6:
+		return 0x1d;
+	case 7:
+		return 0x1e;
+	case 8:
+		return 0x1f;
+	default:
+		return -1;
+	}
+}
+
+static void __cvmx_pko_iport_config(int pko_port)
+{
+	int queue;
+	const int num_queues = 1;
+	const int base_queue = pko_port;
+	const int static_priority_end = 1;
+	const int static_priority_base = 1;
+
+	for (queue = 0; queue < num_queues; queue++) {
+		union cvmx_pko_mem_iqueue_ptrs config;
+		cvmx_cmd_queue_result_t cmd_res;
+		uint64_t *buf_ptr;
+
+		config.u64		= 0;
+		config.s.index		= queue;
+		config.s.qid		= base_queue + queue;
+		config.s.ipid		= pko_port;
+		config.s.tail		= (queue == (num_queues - 1));
+		config.s.s_tail		= (queue == static_priority_end);
+		config.s.static_p	= (static_priority_base >= 0);
+		config.s.static_q	= (queue <= static_priority_end);
+		config.s.qos_mask	= 0xff;
+
+		cmd_res = cvmx_cmd_queue_initialize(
+				CVMX_CMD_QUEUE_PKO(base_queue + queue),
+				CVMX_PKO_MAX_QUEUE_DEPTH,
+				CVMX_FPA_OUTPUT_BUFFER_POOL,
+				(CVMX_FPA_OUTPUT_BUFFER_POOL_SIZE -
+				 CVMX_PKO_COMMAND_BUFFER_SIZE_ADJUST * 8));
+
+		WARN(cmd_res,
+		     "%s: cmd_res=%d pko_port=%d base_queue=%d num_queues=%d queue=%d\n",
+			__func__, (int)cmd_res, pko_port, base_queue,
+			num_queues, queue);
+
+		buf_ptr = (uint64_t *)cvmx_cmd_queue_buffer(
+				CVMX_CMD_QUEUE_PKO(base_queue + queue));
+		config.s.buf_ptr = cvmx_ptr_to_phys(buf_ptr) >> 7;
+		CVMX_SYNCWS;
+		cvmx_write_csr(CVMX_PKO_MEM_IQUEUE_PTRS, config.u64);
+	}
+}
+
+static void __cvmx_pko_queue_alloc_o68(void)
+{
+	int port;
+
+	for (port = 0; port < 48; port++)
+		__cvmx_pko_iport_config(port);
+}
+
+static void __cvmx_pko_port_map_o68(void)
+{
+	int port;
+	int interface, index;
+	cvmx_helper_interface_mode_t mode;
+	union cvmx_pko_mem_iport_ptrs config;
+
+	/*
+	 * Initialize every iport with the invalid eid.
+	 */
+	config.u64 = 0;
+	config.s.eid = 31; /* Invalid */
+	for (port = 0; port < 128; port++) {
+		config.s.ipid = port;
+		cvmx_write_csr(CVMX_PKO_MEM_IPORT_PTRS, config.u64);
+	}
+
+	/*
+	 * Set up PKO_MEM_IPORT_PTRS
+	 */
+	for (port = 0; port < 48; port++) {
+		interface = cvmx_helper_get_interface_num(port);
+		index = cvmx_helper_get_interface_index_num(port);
+		mode = cvmx_helper_interface_get_mode(interface);
+		if (mode == CVMX_HELPER_INTERFACE_MODE_DISABLED)
+			continue;
+
+		config.s.ipid = port;
+		config.s.qos_mask = 0xff;
+		config.s.crc = 1;
+		config.s.min_pkt = 1;
+		config.s.intr = __cvmx_pko_int(interface, index);
+		config.s.eid = config.s.intr;
+		config.s.pipe = (mode == CVMX_HELPER_INTERFACE_MODE_LOOP) ?
+			index : port;
+		cvmx_write_csr(CVMX_PKO_MEM_IPORT_PTRS, config.u64);
+	}
+}
+
+static void __cvmx_pko_chip_init(void)
+{
+	int i;
+
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+		__cvmx_pko_port_map_o68();
+		__cvmx_pko_queue_alloc_o68();
+		return;
+	}
+
+	/*
+	 * Initialize queues
+	 */
+	for (i = 0; i < CVMX_PKO_MAX_OUTPUT_QUEUES; i++) {
+		const uint64_t priority = 8;
+
+		cvmx_pko_config_port(CVMX_PKO_MEM_QUEUE_PTRS_ILLEGAL_PID, i, 1,
+				     &priority);
+	}
+}
+
 /**
  * Call before any other calls to initialize the packet
  * output system.  This does chip global config, and should only be
@@ -47,8 +184,6 @@
 
 void cvmx_pko_initialize_global(void)
 {
-	int i;
-	uint64_t priority = 8;
 	union cvmx_pko_reg_cmd_buf config;
 
 	/*
@@ -62,9 +197,10 @@ void cvmx_pko_initialize_global(void)
 
 	cvmx_write_csr(CVMX_PKO_REG_CMD_BUF, config.u64);
 
-	for (i = 0; i < CVMX_PKO_MAX_OUTPUT_QUEUES; i++)
-		cvmx_pko_config_port(CVMX_PKO_MEM_QUEUE_PTRS_ILLEGAL_PID, i, 1,
-				     &priority);
+	/*
+	 * Chip-specific setup.
+	 */
+	__cvmx_pko_chip_init();
 
 	/*
 	 * If we aren't using all of the queues optimize PKO's
@@ -212,6 +348,9 @@ cvmx_pko_status_t cvmx_pko_config_port(uint64_t port, uint64_t base_queue,
 	int static_priority_base = -1;
 	int static_priority_end = -1;
 
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX))
+		return CVMX_PKO_SUCCESS;
+
 	if ((port >= CVMX_PKO_NUM_OUTPUT_PORTS)
 	    && (port != CVMX_PKO_MEM_QUEUE_PTRS_ILLEGAL_PID)) {
 		cvmx_dprintf("ERROR: cvmx_pko_config_port: Invalid port %llu\n",
diff --git a/arch/mips/cavium-octeon/octeon-irq.c b/arch/mips/cavium-octeon/octeon-irq.c
index d8124a3..f26c3c6 100644
--- a/arch/mips/cavium-octeon/octeon-irq.c
+++ b/arch/mips/cavium-octeon/octeon-irq.c
@@ -225,13 +225,14 @@ static int next_cpu_for_irq(struct irq_data *data)
 
 #ifdef CONFIG_SMP
 	int cpu;
-	int weight = cpumask_weight(data->affinity);
+	struct cpumask *mask = irq_data_get_affinity_mask(data);
+	int weight = cpumask_weight(mask);
 	struct octeon_ciu_chip_data *cd = irq_data_get_irq_chip_data(data);
 
 	if (weight > 1) {
 		cpu = cd->current_cpu;
 		for (;;) {
-			cpu = cpumask_next(cpu, data->affinity);
+			cpu = cpumask_next(cpu, mask);
 			if (cpu >= nr_cpu_ids) {
 				cpu = -1;
 				continue;
@@ -240,7 +241,7 @@ static int next_cpu_for_irq(struct irq_data *data)
 			}
 		}
 	} else if (weight == 1) {
-		cpu = cpumask_first(data->affinity);
+		cpu = cpumask_first(mask);
 	} else {
 		cpu = smp_processor_id();
 	}
@@ -662,6 +663,11 @@ static int octeon_irq_ciu_gpio_set_type(struct irq_data *data, unsigned int t)
 	irqd_set_trigger_type(data, t);
 	octeon_irq_gpio_setup(data);
 
+	if (irqd_get_trigger_type(data) & IRQ_TYPE_EDGE_BOTH)
+		irq_set_handler_locked(data, handle_edge_irq);
+	else
+		irq_set_handler_locked(data, handle_level_irq);
+
 	return IRQ_SET_MASK_OK;
 }
 
@@ -696,32 +702,23 @@ static void octeon_irq_ciu_gpio_ack(struct irq_data *data)
 	cvmx_write_csr(CVMX_GPIO_INT_CLR, mask);
 }
 
-static void octeon_irq_handle_trigger(unsigned int irq, struct irq_desc *desc)
-{
-	struct irq_data *data = irq_desc_get_irq_data(desc);
-
-	if (irqd_get_trigger_type(data) & IRQ_TYPE_EDGE_BOTH)
-		handle_edge_irq(irq, desc);
-	else
-		handle_level_irq(irq, desc);
-}
-
 #ifdef CONFIG_SMP
 
 static void octeon_irq_cpu_offline_ciu(struct irq_data *data)
 {
 	int cpu = smp_processor_id();
 	cpumask_t new_affinity;
+	struct cpumask *mask = irq_data_get_affinity_mask(data);
 
-	if (!cpumask_test_cpu(cpu, data->affinity))
+	if (!cpumask_test_cpu(cpu, mask))
 		return;
 
-	if (cpumask_weight(data->affinity) > 1) {
+	if (cpumask_weight(mask) > 1) {
 		/*
 		 * It has multi CPU affinity, just remove this CPU
 		 * from the affinity set.
 		 */
-		cpumask_copy(&new_affinity, data->affinity);
+		cpumask_copy(&new_affinity, mask);
 		cpumask_clear_cpu(cpu, &new_affinity);
 	} else {
 		/* Otherwise, put it on lowest numbered online CPU. */
@@ -1227,8 +1224,13 @@ static int octeon_irq_gpio_map(struct irq_domain *d,
 		octeon_irq_ciu_to_irq[line][bit] != 0)
 		return -EINVAL;
 
+	/*
+	 * Default to handle_level_irq. If the DT contains a different
+	 * trigger type, it will call the irq_set_type callback and
+	 * the handler gets updated.
+	 */
 	r = octeon_irq_set_ciu_mapping(virq, line, bit, hw,
-		octeon_irq_gpio_chip, octeon_irq_handle_trigger);
+				       octeon_irq_gpio_chip, handle_level_irq);
 	return r;
 }
 
diff --git a/arch/mips/include/asm/Kbuild b/arch/mips/include/asm/Kbuild
index 1f85460..40ec4ca 100644
--- a/arch/mips/include/asm/Kbuild
+++ b/arch/mips/include/asm/Kbuild
@@ -16,6 +16,5 @@ generic-y += sections.h
 generic-y += segment.h
 generic-y += serial.h
 generic-y += trace_clock.h
-generic-y += ucontext.h
 generic-y += user.h
 generic-y += xor.h
diff --git a/arch/mips/include/asm/abi.h b/arch/mips/include/asm/abi.h
index 7186bb5..37f8407 100644
--- a/arch/mips/include/asm/abi.h
+++ b/arch/mips/include/asm/abi.h
@@ -20,6 +20,10 @@ struct mips_abi {
 				     struct pt_regs *regs, sigset_t *set);
 	const unsigned long	rt_signal_return_offset;
 	const unsigned long	restart;
+
+	unsigned	off_sc_fpregs;
+	unsigned	off_sc_fpc_csr;
+	unsigned	off_sc_used_math;
 };
 
 #endif /* _ASM_ABI_H */
diff --git a/arch/mips/include/asm/asmmacro.h b/arch/mips/include/asm/asmmacro.h
index 76317a7..867f924 100644
--- a/arch/mips/include/asm/asmmacro.h
+++ b/arch/mips/include/asm/asmmacro.h
@@ -232,6 +232,30 @@
 	.set	pop
 	.endm
 
+	.macro	ld_b	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	ld.b	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
+	.macro	ld_h	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	ld.h	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
+	.macro	ld_w	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	ld.w	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
 	.macro	ld_d	wd, off, base
 	.set	push
 	.set	mips32r2
@@ -241,6 +265,30 @@
 	.set	pop
 	.endm
 
+	.macro	st_b	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	st.b	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
+	.macro	st_h	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	st.h	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
+	.macro	st_w	wd, off, base
+	.set	push
+	.set	mips32r2
+	.set	msa
+	st.w	$w\wd, \off(\base)
+	.set	pop
+	.endm
+
 	.macro	st_d	wd, off, base
 	.set	push
 	.set	mips32r2
@@ -290,7 +338,13 @@
 #ifdef CONFIG_CPU_MICROMIPS
 #define CFC_MSA_INSN		0x587e0056
 #define CTC_MSA_INSN		0x583e0816
+#define LDB_MSA_INSN		0x58000807
+#define LDH_MSA_INSN		0x58000817
+#define LDW_MSA_INSN		0x58000827
 #define LDD_MSA_INSN		0x58000837
+#define STB_MSA_INSN		0x5800080f
+#define STH_MSA_INSN		0x5800081f
+#define STW_MSA_INSN		0x5800082f
 #define STD_MSA_INSN		0x5800083f
 #define COPY_UW_MSA_INSN	0x58f00056
 #define COPY_UD_MSA_INSN	0x58f80056
@@ -299,7 +353,13 @@
 #else
 #define CFC_MSA_INSN		0x787e0059
 #define CTC_MSA_INSN		0x783e0819
+#define LDB_MSA_INSN		0x78000820
+#define LDH_MSA_INSN		0x78000821
+#define LDW_MSA_INSN		0x78000822
 #define LDD_MSA_INSN		0x78000823
+#define STB_MSA_INSN		0x78000824
+#define STH_MSA_INSN		0x78000825
+#define STW_MSA_INSN		0x78000826
 #define STD_MSA_INSN		0x78000827
 #define COPY_UW_MSA_INSN	0x78f00059
 #define COPY_UD_MSA_INSN	0x78f80059
@@ -329,6 +389,33 @@
 	.set	pop
 	.endm
 
+	.macro	ld_b	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	LDB_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
+	.macro	ld_h	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	LDH_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
+	.macro	ld_w	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	LDW_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
 	.macro	ld_d	wd, off, base
 	.set	push
 	.set	noat
@@ -338,6 +425,33 @@
 	.set	pop
 	.endm
 
+	.macro	st_b	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	STB_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
+	.macro	st_h	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	STH_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
+	.macro	st_w	wd, off, base
+	.set	push
+	.set	noat
+	SET_HARDFLOAT
+	addu	$1, \base, \off
+	.word	STW_MSA_INSN | (\wd << 6)
+	.set	pop
+	.endm
+
 	.macro	st_d	wd, off, base
 	.set	push
 	.set	noat
diff --git a/arch/mips/include/asm/cdmm.h b/arch/mips/include/asm/cdmm.h
index 16e22ce..bece206 100644
--- a/arch/mips/include/asm/cdmm.h
+++ b/arch/mips/include/asm/cdmm.h
@@ -53,7 +53,7 @@ struct mips_cdmm_driver {
  * mips_cdmm_phys_base() - Choose a physical base address for CDMM region.
  *
  * Picking a suitable physical address at which to map the CDMM region is
- * platform specific, so this weak function can be defined by platform code to
+ * platform specific, so this function can be defined by platform code to
  * pick a suitable value if none is configured by the bootloader.
  *
  * This address must be 32kB aligned, and the region occupies a maximum of 32kB
@@ -61,7 +61,7 @@ struct mips_cdmm_driver {
  *
  * Returns:	Physical base address for CDMM region, or 0 on failure.
  */
-phys_addr_t __weak mips_cdmm_phys_base(void);
+phys_addr_t mips_cdmm_phys_base(void);
 
 extern struct bus_type mips_cdmm_bustype;
 void __iomem *mips_cdmm_early_probe(unsigned int dev_type);
diff --git a/arch/mips/include/asm/cevt-r4k.h b/arch/mips/include/asm/cevt-r4k.h
index f0edf6f..2e13a03 100644
--- a/arch/mips/include/asm/cevt-r4k.h
+++ b/arch/mips/include/asm/cevt-r4k.h
@@ -21,7 +21,6 @@ DECLARE_PER_CPU(struct clock_event_device, mips_clockevent_device);
 
 void mips_event_handler(struct clock_event_device *dev);
 int c0_compare_int_usable(void);
-void mips_set_clock_mode(enum clock_event_mode, struct clock_event_device *);
 irqreturn_t c0_compare_interrupt(int, void *);
 
 extern struct irqaction c0_compare_irqaction;
diff --git a/arch/mips/include/asm/cpu-features.h b/arch/mips/include/asm/cpu-features.h
index f25de77..9801ac9 100644
--- a/arch/mips/include/asm/cpu-features.h
+++ b/arch/mips/include/asm/cpu-features.h
@@ -411,4 +411,8 @@
 # define cpu_has_cdmm		(cpu_data[0].options & MIPS_CPU_CDMM)
 #endif
 
+#ifndef cpu_has_small_pages
+# define cpu_has_small_pages	(cpu_data[0].options & MIPS_CPU_SP)
+#endif
+
 #endif /* __ASM_CPU_FEATURES_H */
diff --git a/arch/mips/include/asm/cpu-type.h b/arch/mips/include/asm/cpu-type.h
index d41e8e2..abee2bf 100644
--- a/arch/mips/include/asm/cpu-type.h
+++ b/arch/mips/include/asm/cpu-type.h
@@ -77,6 +77,10 @@ static inline int __pure __get_cpu_type(const int cpu_type)
 	 */
 #endif
 
+#ifdef CONFIG_SYS_HAS_CPU_MIPS64_R6
+	case CPU_I6400:
+#endif
+
 #ifdef CONFIG_SYS_HAS_CPU_R3000
 	case CPU_R2000:
 	case CPU_R3000:
diff --git a/arch/mips/include/asm/cpu.h b/arch/mips/include/asm/cpu.h
index e46e406..cd89e98 100644
--- a/arch/mips/include/asm/cpu.h
+++ b/arch/mips/include/asm/cpu.h
@@ -120,6 +120,7 @@
 #define PRID_IMP_PROAPTIV_MP	0xa300
 #define PRID_IMP_M5150		0xa700
 #define PRID_IMP_P5600		0xa800
+#define PRID_IMP_I6400		0xa900
 
 /*
  * These are the PRID's for when 23:16 == PRID_COMP_SIBYTE
@@ -307,6 +308,7 @@ enum cpu_type_enum {
 	CPU_ALCHEMY, CPU_PR4450, CPU_BMIPS32, CPU_BMIPS3300, CPU_BMIPS4350,
 	CPU_BMIPS4380, CPU_BMIPS5000, CPU_JZRISC, CPU_LOONGSON1, CPU_M14KC,
 	CPU_M14KEC, CPU_INTERAPTIV, CPU_P5600, CPU_PROAPTIV, CPU_1074K, CPU_M5150,
+	CPU_I6400,
 
 	/*
 	 * MIPS64 class processors
@@ -382,6 +384,7 @@ enum cpu_type_enum {
 #define MIPS_CPU_XPA		0x2000000000ull /* CPU supports Extended Physical Addressing */
 #define MIPS_CPU_CDMM		0x4000000000ull	/* CPU has Common Device Memory Map */
 #define MIPS_CPU_BP_GHIST	0x8000000000ull /* R12K+ Branch Prediction Global History */
+#define MIPS_CPU_SP		0x10000000000ull /* Small (1KB) page support */
 
 /*
  * CPU ASE encodings
diff --git a/arch/mips/include/asm/debug.h b/arch/mips/include/asm/debug.h
deleted file mode 100644
index 1fd5a2b..0000000
--- a/arch/mips/include/asm/debug.h
+++ /dev/null
@@ -1,48 +0,0 @@
-/*
- * Debug macros for run-time debugging.
- * Turned on/off with CONFIG_RUNTIME_DEBUG option.
- *
- * Copyright (C) 2001 MontaVista Software Inc.
- * Author: Jun Sun, jsun@mvista.com or jsun@junsun.net
- *
- * 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 _ASM_DEBUG_H
-#define _ASM_DEBUG_H
-
-
-/*
- * run-time macros for catching spurious errors.  Eable CONFIG_RUNTIME_DEBUG in
- * kernel hacking config menu to use them.
- *
- * Use them as run-time debugging aid.  NEVER USE THEM AS ERROR HANDLING CODE!!!
- */
-
-#ifdef CONFIG_RUNTIME_DEBUG
-
-#include <linux/kernel.h>
-
-#define db_assert(x)  if (!(x)) { \
-	panic("assertion failed at %s:%d: %s", __FILE__, __LINE__, #x); }
-#define db_warn(x)  if (!(x)) { \
-	printk(KERN_WARNING "warning at %s:%d: %s", __FILE__, __LINE__, #x); }
-#define db_verify(x, y) db_assert(x y)
-#define db_verify_warn(x, y) db_warn(x y)
-#define db_run(x)  do { x; } while (0)
-
-#else
-
-#define db_assert(x)
-#define db_warn(x)
-#define db_verify(x, y) x
-#define db_verify_warn(x, y) x
-#define db_run(x)
-
-#endif
-
-#endif /* _ASM_DEBUG_H */
diff --git a/arch/mips/include/asm/elf.h b/arch/mips/include/asm/elf.h
index f19e890..53b2693 100644
--- a/arch/mips/include/asm/elf.h
+++ b/arch/mips/include/asm/elf.h
@@ -382,7 +382,9 @@ do {									\
    instruction set this cpu supports.  This could be done in userspace,
    but it's not easy, and we've already done it here.  */
 
-#define ELF_HWCAP	(0)
+#define ELF_HWCAP	(elf_hwcap)
+extern unsigned int elf_hwcap;
+#include <asm/hwcap.h>
 
 /*
  * This yields a string that ld.so will use to load implementation
diff --git a/arch/mips/include/asm/fpu.h b/arch/mips/include/asm/fpu.h
index 1b06251..9cbf383 100644
--- a/arch/mips/include/asm/fpu.h
+++ b/arch/mips/include/asm/fpu.h
@@ -164,25 +164,30 @@ static inline int own_fpu(int restore)
 	return ret;
 }
 
-static inline void lose_fpu(int save)
+static inline void lose_fpu_inatomic(int save, struct task_struct *tsk)
 {
-	preempt_disable();
 	if (is_msa_enabled()) {
 		if (save) {
-			save_msa(current);
-			current->thread.fpu.fcr31 =
+			save_msa(tsk);
+			tsk->thread.fpu.fcr31 =
 					read_32bit_cp1_register(CP1_STATUS);
 		}
 		disable_msa();
-		clear_thread_flag(TIF_USEDMSA);
+		clear_tsk_thread_flag(tsk, TIF_USEDMSA);
 		__disable_fpu();
 	} else if (is_fpu_owner()) {
 		if (save)
-			_save_fp(current);
+			_save_fp(tsk);
 		__disable_fpu();
 	}
-	KSTK_STATUS(current) &= ~ST0_CU1;
-	clear_thread_flag(TIF_USEDFPU);
+	KSTK_STATUS(tsk) &= ~ST0_CU1;
+	clear_tsk_thread_flag(tsk, TIF_USEDFPU);
+}
+
+static inline void lose_fpu(int save)
+{
+	preempt_disable();
+	lose_fpu_inatomic(save, current);
 	preempt_enable();
 }
 
diff --git a/arch/mips/include/asm/gpio.h b/arch/mips/include/asm/gpio.h
deleted file mode 100644
index 06e46fa..0000000
--- a/arch/mips/include/asm/gpio.h
+++ /dev/null
@@ -1,6 +0,0 @@
-#ifndef __ASM_MIPS_GPIO_H
-#define __ASM_MIPS_GPIO_H
-
-#include <gpio.h>
-
-#endif /* __ASM_MIPS_GPIO_H */
diff --git a/arch/mips/include/asm/irq.h b/arch/mips/include/asm/irq.h
index f0db99f..15e0fec 100644
--- a/arch/mips/include/asm/irq.h
+++ b/arch/mips/include/asm/irq.h
@@ -49,7 +49,7 @@ extern int cp0_compare_irq_shift;
 extern int cp0_perfcount_irq;
 extern int cp0_fdc_irq;
 
-extern int __weak get_c0_fdc_int(void);
+extern int get_c0_fdc_int(void);
 
 void arch_trigger_all_cpu_backtrace(bool);
 #define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace
diff --git a/arch/mips/include/asm/kdebug.h b/arch/mips/include/asm/kdebug.h
index cba22ab..8e3d08e 100644
--- a/arch/mips/include/asm/kdebug.h
+++ b/arch/mips/include/asm/kdebug.h
@@ -11,7 +11,9 @@ enum die_val {
 	DIE_PAGE_FAULT,
 	DIE_BREAK,
 	DIE_SSTEPBP,
-	DIE_MSAFP
+	DIE_MSAFP,
+	DIE_UPROBE,
+	DIE_UPROBE_XOL,
 };
 
 #endif /* _ASM_MIPS_KDEBUG_H */
diff --git a/arch/mips/include/asm/linkage.h b/arch/mips/include/asm/linkage.h
index 2767dda..99651b0 100644
--- a/arch/mips/include/asm/linkage.h
+++ b/arch/mips/include/asm/linkage.h
@@ -5,7 +5,6 @@
 #include <asm/asm.h>
 #endif
 
-#define __weak __attribute__((weak))
 #define cond_syscall(x) asm(".weak\t" #x "\n" #x "\t=\tsys_ni_syscall")
 #define SYSCALL_ALIAS(alias, name)					\
 	asm ( #alias " = " #name "\n\t.globl " #alias)
diff --git a/arch/mips/include/asm/maar.h b/arch/mips/include/asm/maar.h
index 6c62b0f..b02891f 100644
--- a/arch/mips/include/asm/maar.h
+++ b/arch/mips/include/asm/maar.h
@@ -26,7 +26,7 @@
  *
  * Return:	The number of MAAR pairs configured.
  */
-unsigned __weak platform_maar_init(unsigned num_pairs);
+unsigned platform_maar_init(unsigned num_pairs);
 
 /**
  * write_maar_pair() - write to a pair of MAARs
diff --git a/arch/mips/include/asm/mach-ar7/ar7.h b/arch/mips/include/asm/mach-ar7/ar7.h
index a47ea0c..468cbd6 100644
--- a/arch/mips/include/asm/mach-ar7/ar7.h
+++ b/arch/mips/include/asm/mach-ar7/ar7.h
@@ -203,4 +203,8 @@ static inline void ar7_device_off(u32 bit)
 int __init ar7_gpio_init(void);
 void __init ar7_init_clocks(void);
 
+/* Board specific GPIO functions */
+int ar7_gpio_enable(unsigned gpio);
+int ar7_gpio_disable(unsigned gpio);
+
 #endif /* __AR7_H__ */
diff --git a/arch/mips/include/asm/mach-ar7/gpio.h b/arch/mips/include/asm/mach-ar7/gpio.h
deleted file mode 100644
index c177cd1..0000000
--- a/arch/mips/include/asm/mach-ar7/gpio.h
+++ /dev/null
@@ -1,41 +0,0 @@
-/*
- * Copyright (C) 2007-2009 Florian Fainelli <florian@openwrt.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- */
-
-#ifndef __AR7_GPIO_H__
-#define __AR7_GPIO_H__
-
-#include <asm/mach-ar7/ar7.h>
-
-#define AR7_GPIO_MAX 32
-#define TITAN_GPIO_MAX	51
-#define NR_BUILTIN_GPIO TITAN_GPIO_MAX
-
-#define gpio_to_irq(gpio)	-1
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-
-#define gpio_cansleep __gpio_cansleep
-
-/* Board specific GPIO functions */
-int ar7_gpio_enable(unsigned gpio);
-int ar7_gpio_disable(unsigned gpio);
-
-#include <asm-generic/gpio.h>
-
-#endif
diff --git a/arch/mips/include/asm/mach-ath25/gpio.h b/arch/mips/include/asm/mach-ath25/gpio.h
deleted file mode 100644
index 713564b..0000000
--- a/arch/mips/include/asm/mach-ath25/gpio.h
+++ /dev/null
@@ -1,16 +0,0 @@
-#ifndef __ASM_MACH_ATH25_GPIO_H
-#define __ASM_MACH_ATH25_GPIO_H
-
-#include <asm-generic/gpio.h>
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-#define gpio_cansleep __gpio_cansleep
-#define gpio_to_irq __gpio_to_irq
-
-static inline int irq_to_gpio(unsigned irq)
-{
-	return -EINVAL;
-}
-
-#endif	/* __ASM_MACH_ATH25_GPIO_H */
diff --git a/arch/mips/include/asm/mach-ath79/gpio.h b/arch/mips/include/asm/mach-ath79/gpio.h
deleted file mode 100644
index 60dcb62..0000000
--- a/arch/mips/include/asm/mach-ath79/gpio.h
+++ /dev/null
@@ -1,26 +0,0 @@
-/*
- *  Atheros AR71XX/AR724X/AR913X GPIO API definitions
- *
- *  Copyright (C) 2008-2010 Gabor Juhos <juhosg@openwrt.org>
- *  Copyright (C) 2008 Imre Kaloz <kaloz@openwrt.org>
- *
- *  This program is free software; you can redistribute it and/or modify it
- *  under the terms of the GNU General Public License version 2 as published
- *  by the Free Software Foundation.
- *
- */
-
-#ifndef __ASM_MACH_ATH79_GPIO_H
-#define __ASM_MACH_ATH79_GPIO_H
-
-#define ARCH_NR_GPIOS	64
-#include <asm-generic/gpio.h>
-
-int gpio_to_irq(unsigned gpio);
-int irq_to_gpio(unsigned irq);
-int gpio_get_value(unsigned gpio);
-void gpio_set_value(unsigned gpio, int value);
-
-#define gpio_cansleep	__gpio_cansleep
-
-#endif /* __ASM_MACH_ATH79_GPIO_H */
diff --git a/arch/mips/include/asm/mach-au1x00/gpio-au1000.h b/arch/mips/include/asm/mach-au1x00/gpio-au1000.h
index 9785e4e..adde1fa 100644
--- a/arch/mips/include/asm/mach-au1x00/gpio-au1000.h
+++ b/arch/mips/include/asm/mach-au1x00/gpio-au1000.h
@@ -266,6 +266,17 @@ static inline int alchemy_gpio1_to_irq(int gpio)
 	return -ENXIO;
 }
 
+/* On Au1000, Au1500 and Au1100 GPIOs won't work as inputs before
+ * SYS_PININPUTEN is written to at least once.  On Au1550/Au1200/Au1300 this
+ * register enables use of GPIOs as wake source.
+ */
+static inline void alchemy_gpio1_input_enable(void)
+{
+	void __iomem *base = (void __iomem *)KSEG1ADDR(AU1000_SYS_PHYS_ADDR);
+	__raw_writel(0, base + 0x110);		/* the write op is key */
+	wmb();
+}
+
 /*
  * GPIO2 block macros for common linux GPIO functions. The 'gpio'
  * parameter must be in range of ALCHEMY_GPIO2_BASE..ALCHEMY_GPIO2_MAX.
@@ -518,141 +529,4 @@ static inline int alchemy_irq_to_gpio(int irq)
 	return -ENXIO;
 }
 
-/**********************************************************************/
-
-/* Linux gpio framework integration.
- *
- * 4 use cases of Au1000-Au1200 GPIOS:
- *(1) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=y:
- *	Board must register gpiochips.
- *(2) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=n:
- *	2 (1 for Au1000) gpio_chips are registered.
- *
- *(3) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=y:
- *	the boards' gpio.h must provide the linux gpio wrapper functions,
- *
- *(4) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=n:
- *	inlinable gpio functions are provided which enable access to the
- *	Au1000 gpios only by using the numbers straight out of the data-
- *	sheets.
-
- * Cases 1 and 3 are intended for boards which want to provide their own
- * GPIO namespace and -operations (i.e. for example you have 8 GPIOs
- * which are in part provided by spare Au1000 GPIO pins and in part by
- * an external FPGA but you still want them to be accssible in linux
- * as gpio0-7. The board can of course use the alchemy_gpioX_* functions
- * as required).
- */
-
-#ifndef CONFIG_GPIOLIB
-
-#ifdef CONFIG_ALCHEMY_GPIOINT_AU1000
-
-#ifndef CONFIG_ALCHEMY_GPIO_INDIRECT	/* case (4) */
-
-static inline int gpio_direction_input(int gpio)
-{
-	return alchemy_gpio_direction_input(gpio);
-}
-
-static inline int gpio_direction_output(int gpio, int v)
-{
-	return alchemy_gpio_direction_output(gpio, v);
-}
-
-static inline int gpio_get_value(int gpio)
-{
-	return alchemy_gpio_get_value(gpio);
-}
-
-static inline void gpio_set_value(int gpio, int v)
-{
-	alchemy_gpio_set_value(gpio, v);
-}
-
-static inline int gpio_get_value_cansleep(unsigned gpio)
-{
-	return gpio_get_value(gpio);
-}
-
-static inline void gpio_set_value_cansleep(unsigned gpio, int value)
-{
-	gpio_set_value(gpio, value);
-}
-
-static inline int gpio_is_valid(int gpio)
-{
-	return alchemy_gpio_is_valid(gpio);
-}
-
-static inline int gpio_cansleep(int gpio)
-{
-	return alchemy_gpio_cansleep(gpio);
-}
-
-static inline int gpio_to_irq(int gpio)
-{
-	return alchemy_gpio_to_irq(gpio);
-}
-
-static inline int irq_to_gpio(int irq)
-{
-	return alchemy_irq_to_gpio(irq);
-}
-
-static inline int gpio_request(unsigned gpio, const char *label)
-{
-	return 0;
-}
-
-static inline int gpio_request_one(unsigned gpio,
-					unsigned long flags, const char *label)
-{
-	return 0;
-}
-
-static inline int gpio_request_array(struct gpio *array, size_t num)
-{
-	return 0;
-}
-
-static inline void gpio_free(unsigned gpio)
-{
-}
-
-static inline void gpio_free_array(struct gpio *array, size_t num)
-{
-}
-
-static inline int gpio_set_debounce(unsigned gpio, unsigned debounce)
-{
-	return -ENOSYS;
-}
-
-static inline int gpio_export(unsigned gpio, bool direction_may_change)
-{
-	return -ENOSYS;
-}
-
-static inline int gpio_export_link(struct device *dev, const char *name,
-				   unsigned gpio)
-{
-	return -ENOSYS;
-}
-
-static inline int gpio_sysfs_set_active_low(unsigned gpio, int value)
-{
-	return -ENOSYS;
-}
-
-static inline void gpio_unexport(unsigned gpio)
-{
-}
-
-#endif	/* !CONFIG_ALCHEMY_GPIO_INDIRECT */
-
-#endif	/* CONFIG_ALCHEMY_GPIOINT_AU1000 */
-
-#endif	/* !CONFIG_GPIOLIB */
-
 #endif /* _ALCHEMY_GPIO_AU1000_H_ */
diff --git a/arch/mips/include/asm/mach-au1x00/gpio.h b/arch/mips/include/asm/mach-au1x00/gpio.h
deleted file mode 100644
index 22e7ff1..0000000
--- a/arch/mips/include/asm/mach-au1x00/gpio.h
+++ /dev/null
@@ -1,86 +0,0 @@
-/*
- * Alchemy GPIO support.
- *
- * With CONFIG_GPIOLIB=y different types of on-chip GPIO can be supported within
- *  the same kernel image.
- * With CONFIG_GPIOLIB=n, your board must select ALCHEMY_GPIOINT_AU1XXX for the
- *  appropriate CPU type (AU1000 currently).
- */
-
-#ifndef _ALCHEMY_GPIO_H_
-#define _ALCHEMY_GPIO_H_
-
-#include <asm/mach-au1x00/au1000.h>
-#include <asm/mach-au1x00/gpio-au1000.h>
-#include <asm/mach-au1x00/gpio-au1300.h>
-
-/* On Au1000, Au1500 and Au1100 GPIOs won't work as inputs before
- * SYS_PININPUTEN is written to at least once.  On Au1550/Au1200/Au1300 this
- * register enables use of GPIOs as wake source.
- */
-static inline void alchemy_gpio1_input_enable(void)
-{
-	void __iomem *base = (void __iomem *)KSEG1ADDR(AU1000_SYS_PHYS_ADDR);
-	__raw_writel(0, base + 0x110);		/* the write op is key */
-	wmb();
-}
-
-
-/* Linux gpio framework integration.
-*
-* 4 use cases of Alchemy GPIOS:
-*(1) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=y:
-*	Board must register gpiochips.
-*(2) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=n:
-*	A gpiochip for the 75 GPIOs is registered.
-*
-*(3) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=y:
-*	the boards' gpio.h must provide	the linux gpio wrapper functions,
-*
-*(4) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=n:
-*	inlinable gpio functions are provided which enable access to the
-*	Au1300 gpios only by using the numbers straight out of the data-
-*	sheets.
-
-* Cases 1 and 3 are intended for boards which want to provide their own
-* GPIO namespace and -operations (i.e. for example you have 8 GPIOs
-* which are in part provided by spare Au1300 GPIO pins and in part by
-* an external FPGA but you still want them to be accssible in linux
-* as gpio0-7. The board can of course use the alchemy_gpioX_* functions
-* as required).
-*/
-
-#ifdef CONFIG_GPIOLIB
-
-/* wraps the cpu-dependent irq_to_gpio functions */
-/* FIXME: gpiolib needs an irq_to_gpio hook */
-static inline int __au_irq_to_gpio(unsigned int irq)
-{
-	switch (alchemy_get_cputype()) {
-	case ALCHEMY_CPU_AU1000...ALCHEMY_CPU_AU1200:
-		return alchemy_irq_to_gpio(irq);
-	case ALCHEMY_CPU_AU1300:
-		return au1300_irq_to_gpio(irq);
-	}
-	return -EINVAL;
-}
-
-
-/* using gpiolib to provide up to 2 gpio_chips for on-chip gpios */
-#ifndef CONFIG_ALCHEMY_GPIO_INDIRECT	/* case (2) */
-
-/* get everything through gpiolib */
-#define gpio_to_irq	__gpio_to_irq
-#define gpio_get_value	__gpio_get_value
-#define gpio_set_value	__gpio_set_value
-#define gpio_cansleep	__gpio_cansleep
-#define irq_to_gpio	__au_irq_to_gpio
-
-#include <asm-generic/gpio.h>
-
-#endif	/* !CONFIG_ALCHEMY_GPIO_INDIRECT */
-
-
-#endif	/* CONFIG_GPIOLIB */
-
-#endif	/* _ALCHEMY_GPIO_H_ */
diff --git a/arch/mips/include/asm/mach-bcm47xx/gpio.h b/arch/mips/include/asm/mach-bcm47xx/gpio.h
deleted file mode 100644
index 90daefa..0000000
--- a/arch/mips/include/asm/mach-bcm47xx/gpio.h
+++ /dev/null
@@ -1,17 +0,0 @@
-#ifndef __ASM_MIPS_MACH_BCM47XX_GPIO_H
-#define __ASM_MIPS_MACH_BCM47XX_GPIO_H
-
-#include <asm-generic/gpio.h>
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-
-#define gpio_cansleep __gpio_cansleep
-#define gpio_to_irq __gpio_to_irq
-
-static inline int irq_to_gpio(unsigned int irq)
-{
-	return -EINVAL;
-}
-
-#endif
diff --git a/arch/mips/include/asm/mach-bcm63xx/gpio.h b/arch/mips/include/asm/mach-bcm63xx/gpio.h
deleted file mode 100644
index 1eb534d..0000000
--- a/arch/mips/include/asm/mach-bcm63xx/gpio.h
+++ /dev/null
@@ -1,15 +0,0 @@
-#ifndef __ASM_MIPS_MACH_BCM63XX_GPIO_H
-#define __ASM_MIPS_MACH_BCM63XX_GPIO_H
-
-#include <bcm63xx_gpio.h>
-
-#define gpio_to_irq(gpio)	-1
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-
-#define gpio_cansleep __gpio_cansleep
-
-#include <asm-generic/gpio.h>
-
-#endif /* __ASM_MIPS_MACH_BCM63XX_GPIO_H */
diff --git a/arch/mips/include/asm/mach-cavium-octeon/gpio.h b/arch/mips/include/asm/mach-cavium-octeon/gpio.h
deleted file mode 100644
index 34e9f7a..0000000
--- a/arch/mips/include/asm/mach-cavium-octeon/gpio.h
+++ /dev/null
@@ -1,21 +0,0 @@
-#ifndef __ASM_MACH_CAVIUM_OCTEON_GPIO_H
-#define __ASM_MACH_CAVIUM_OCTEON_GPIO_H
-
-#ifdef CONFIG_GPIOLIB
-#define gpio_get_value	__gpio_get_value
-#define gpio_set_value	__gpio_set_value
-#define gpio_cansleep	__gpio_cansleep
-#else
-int gpio_request(unsigned gpio, const char *label);
-void gpio_free(unsigned gpio);
-int gpio_direction_input(unsigned gpio);
-int gpio_direction_output(unsigned gpio, int value);
-int gpio_get_value(unsigned gpio);
-void gpio_set_value(unsigned gpio, int value);
-#endif
-
-#include <asm-generic/gpio.h>
-
-#define gpio_to_irq	__gpio_to_irq
-
-#endif /* __ASM_MACH_GENERIC_GPIO_H */
diff --git a/arch/mips/include/asm/mach-generic/gpio.h b/arch/mips/include/asm/mach-generic/gpio.h
deleted file mode 100644
index b4e7020..0000000
--- a/arch/mips/include/asm/mach-generic/gpio.h
+++ /dev/null
@@ -1,21 +0,0 @@
-#ifndef __ASM_MACH_GENERIC_GPIO_H
-#define __ASM_MACH_GENERIC_GPIO_H
-
-#ifdef CONFIG_GPIOLIB
-#define gpio_get_value	__gpio_get_value
-#define gpio_set_value	__gpio_set_value
-#define gpio_cansleep	__gpio_cansleep
-#else
-int gpio_request(unsigned gpio, const char *label);
-void gpio_free(unsigned gpio);
-int gpio_direction_input(unsigned gpio);
-int gpio_direction_output(unsigned gpio, int value);
-int gpio_get_value(unsigned gpio);
-void gpio_set_value(unsigned gpio, int value);
-#endif
-int gpio_to_irq(unsigned gpio);
-int irq_to_gpio(unsigned irq);
-
-#include <asm-generic/gpio.h>		/* cansleep wrappers */
-
-#endif /* __ASM_MACH_GENERIC_GPIO_H */
diff --git a/arch/mips/include/asm/mach-jz4740/gpio.h b/arch/mips/include/asm/mach-jz4740/gpio.h
index eaacba7..bf8c3e1 100644
--- a/arch/mips/include/asm/mach-jz4740/gpio.h
+++ b/arch/mips/include/asm/mach-jz4740/gpio.h
@@ -73,8 +73,6 @@ int jz_gpio_port_direction_output(int port, uint32_t mask);
 void jz_gpio_port_set_value(int port, uint32_t value, uint32_t mask);
 uint32_t jz_gpio_port_get_value(int port, uint32_t mask);
 
-#include <asm/mach-generic/gpio.h>
-
 #define JZ_GPIO_PORTA(x) ((x) + 32 * 0)
 #define JZ_GPIO_PORTB(x) ((x) + 32 * 1)
 #define JZ_GPIO_PORTC(x) ((x) + 32 * 2)
diff --git a/arch/mips/include/asm/mach-lantiq/gpio.h b/arch/mips/include/asm/mach-lantiq/gpio.h
deleted file mode 100644
index 9ba1cae..0000000
--- a/arch/mips/include/asm/mach-lantiq/gpio.h
+++ /dev/null
@@ -1,13 +0,0 @@
-#ifndef __ASM_MIPS_MACH_LANTIQ_GPIO_H
-#define __ASM_MIPS_MACH_LANTIQ_GPIO_H
-
-#define gpio_to_irq __gpio_to_irq
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-
-#define gpio_cansleep __gpio_cansleep
-
-#include <asm-generic/gpio.h>
-
-#endif
diff --git a/arch/mips/include/asm/mach-loongson64/gpio.h b/arch/mips/include/asm/mach-loongson64/gpio.h
deleted file mode 100644
index b3b2169..0000000
--- a/arch/mips/include/asm/mach-loongson64/gpio.h
+++ /dev/null
@@ -1,36 +0,0 @@
-/*
- * Loongson GPIO Support
- *
- * Copyright (c) 2008  Richard Liu, STMicroelectronics <richard.liu@st.com>
- * Copyright (c) 2008-2010  Arnaud Patard <apatard@mandriva.com>
- * Copyright (c) 2014  Huacai Chen <chenhc@lemote.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 __LOONGSON_GPIO_H
-#define __LOONGSON_GPIO_H
-
-#include <asm-generic/gpio.h>
-
-#define gpio_get_value __gpio_get_value
-#define gpio_set_value __gpio_set_value
-#define gpio_cansleep __gpio_cansleep
-
-/* The chip can do interrupt
- * but it has not been tested and doc not clear
- */
-static inline int gpio_to_irq(int gpio)
-{
-	return -EINVAL;
-}
-
-static inline int irq_to_gpio(int gpio)
-{
-	return -EINVAL;
-}
-
-#endif	/* __LOONGSON_GPIO_H */
diff --git a/arch/mips/include/asm/mach-pistachio/gpio.h b/arch/mips/include/asm/mach-pistachio/gpio.h
deleted file mode 100644
index 6c1649c..0000000
--- a/arch/mips/include/asm/mach-pistachio/gpio.h
+++ /dev/null
@@ -1,21 +0,0 @@
-/*
- * Pistachio IRQ setup
- *
- * Copyright (C) 2014 Google, Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- */
-
-#ifndef __ASM_MACH_PISTACHIO_GPIO_H
-#define __ASM_MACH_PISTACHIO_GPIO_H
-
-#include <asm-generic/gpio.h>
-
-#define gpio_get_value	__gpio_get_value
-#define gpio_set_value	__gpio_set_value
-#define gpio_cansleep	__gpio_cansleep
-#define gpio_to_irq	__gpio_to_irq
-
-#endif /* __ASM_MACH_PISTACHIO_GPIO_H */
diff --git a/arch/mips/include/asm/mach-rc32434/gpio.h b/arch/mips/include/asm/mach-rc32434/gpio.h
index 4dee0a3..db21121 100644
--- a/arch/mips/include/asm/mach-rc32434/gpio.h
+++ b/arch/mips/include/asm/mach-rc32434/gpio.h
@@ -13,18 +13,6 @@
 #ifndef _RC32434_GPIO_H_
 #define _RC32434_GPIO_H_
 
-#include <linux/types.h>
-#include <asm-generic/gpio.h>
-
-#define NR_BUILTIN_GPIO		32
-
-#define gpio_get_value	__gpio_get_value
-#define gpio_set_value	__gpio_set_value
-#define gpio_cansleep	__gpio_cansleep
-
-#define gpio_to_irq(gpio)	(8 + 4 * 32 + gpio)
-#define irq_to_gpio(irq)	(irq - (8 + 4 * 32))
-
 struct rb532_gpio_reg {
 	u32   gpiofunc;	  /* GPIO Function Register
 			   * gpiofunc[x]==0 bit = gpio
diff --git a/arch/mips/include/asm/mips-cm.h b/arch/mips/include/asm/mips-cm.h
index edc7ee9..d75b75e 100644
--- a/arch/mips/include/asm/mips-cm.h
+++ b/arch/mips/include/asm/mips-cm.h
@@ -33,6 +33,29 @@ extern void __iomem *mips_cm_l2sync_base;
  */
 extern phys_addr_t __mips_cm_phys_base(void);
 
+/*
+ * mips_cm_is64 - determine CM register width
+ *
+ * The CM register width is processor and CM specific. A 64-bit processor
+ * usually has a 64-bit CM and a 32-bit one has a 32-bit CM but a 64-bit
+ * processor could come with a 32-bit CM. Moreover, accesses on 64-bit CMs
+ * can be done either using regular 64-bit load/store instructions, or 32-bit
+ * load/store instruction on 32-bit register pairs. We opt for using 64-bit
+ * accesses on 64-bit CMs and kernels and 32-bit in any other case.
+ *
+ * It's set to 0 for 32-bit accesses and 1 for 64-bit accesses.
+ */
+extern int mips_cm_is64;
+
+/**
+ * mips_cm_error_report - Report CM cache errors
+ */
+#ifdef CONFIG_MIPS_CM
+extern void mips_cm_error_report(void);
+#else
+static inline void mips_cm_error_report(void) {}
+#endif
+
 /**
  * mips_cm_probe - probe for a Coherence Manager
  *
@@ -90,20 +113,46 @@ static inline bool mips_cm_has_l2sync(void)
 
 /* Macros to ease the creation of register access functions */
 #define BUILD_CM_R_(name, off)					\
-static inline u32 __iomem *addr_gcr_##name(void)		\
+static inline unsigned long __iomem *addr_gcr_##name(void)	\
 {								\
-	return (u32 __iomem *)(mips_cm_base + (off));		\
+	return (unsigned long __iomem *)(mips_cm_base + (off));	\
 }								\
 								\
-static inline u32 read_gcr_##name(void)				\
+static inline u32 read32_gcr_##name(void)			\
 {								\
 	return __raw_readl(addr_gcr_##name());			\
+}								\
+								\
+static inline u64 read64_gcr_##name(void)			\
+{								\
+	return __raw_readq(addr_gcr_##name());			\
+}								\
+								\
+static inline unsigned long read_gcr_##name(void)		\
+{								\
+	if (mips_cm_is64)					\
+		return read64_gcr_##name();			\
+	else							\
+		return read32_gcr_##name();			\
 }
 
 #define BUILD_CM__W(name, off)					\
-static inline void write_gcr_##name(u32 value)			\
+static inline void write32_gcr_##name(u32 value)		\
 {								\
 	__raw_writel(value, addr_gcr_##name());			\
+}								\
+								\
+static inline void write64_gcr_##name(u64 value)		\
+{								\
+	__raw_writeq(value, addr_gcr_##name());			\
+}								\
+								\
+static inline void write_gcr_##name(unsigned long value)	\
+{								\
+	if (mips_cm_is64)					\
+		write64_gcr_##name(value);			\
+	else							\
+		write32_gcr_##name(value);			\
 }
 
 #define BUILD_CM_RW(name, off)					\
@@ -144,6 +193,7 @@ BUILD_CM_RW(reg3_base,		MIPS_CM_GCB_OFS + 0xc0)
 BUILD_CM_RW(reg3_mask,		MIPS_CM_GCB_OFS + 0xc8)
 BUILD_CM_R_(gic_status,		MIPS_CM_GCB_OFS + 0xd0)
 BUILD_CM_R_(cpc_status,		MIPS_CM_GCB_OFS + 0xf0)
+BUILD_CM_RW(l2_config,		MIPS_CM_GCB_OFS + 0x130)
 
 /* Core Local & Core Other register accessor functions */
 BUILD_CM_Cx_RW(reset_release,	0x00)
@@ -189,6 +239,13 @@ BUILD_CM_Cx_R_(tcid_8_priority,	0x80)
 #define CM_GCR_REV_MINOR_SHF			0
 #define CM_GCR_REV_MINOR_MSK			(_ULCAST_(0xff) << 0)
 
+#define CM_ENCODE_REV(major, minor) \
+		(((major) << CM_GCR_REV_MAJOR_SHF) | \
+		 ((minor) << CM_GCR_REV_MINOR_SHF))
+
+#define CM_REV_CM2				CM_ENCODE_REV(6, 0)
+#define CM_REV_CM3				CM_ENCODE_REV(8, 0)
+
 /* GCR_ERROR_CAUSE register fields */
 #define CM_GCR_ERROR_CAUSE_ERRTYPE_SHF		27
 #define CM_GCR_ERROR_CAUSE_ERRTYPE_MSK		(_ULCAST_(0x1f) << 27)
@@ -249,6 +306,16 @@ BUILD_CM_Cx_R_(tcid_8_priority,	0x80)
 #define CM_GCR_CPC_STATUS_EX_SHF		0
 #define CM_GCR_CPC_STATUS_EX_MSK		(_ULCAST_(0x1) << 0)
 
+/* GCR_L2_CONFIG register fields */
+#define CM_GCR_L2_CONFIG_BYPASS_SHF		20
+#define CM_GCR_L2_CONFIG_BYPASS_MSK		(_ULCAST_(0x1) << 20)
+#define CM_GCR_L2_CONFIG_SET_SIZE_SHF		12
+#define CM_GCR_L2_CONFIG_SET_SIZE_MSK		(_ULCAST_(0xf) << 12)
+#define CM_GCR_L2_CONFIG_LINE_SIZE_SHF		8
+#define CM_GCR_L2_CONFIG_LINE_SIZE_MSK		(_ULCAST_(0xf) << 8)
+#define CM_GCR_L2_CONFIG_ASSOC_SHF		0
+#define CM_GCR_L2_CONFIG_ASSOC_MSK		(_ULCAST_(0xff) << 0)
+
 /* GCR_Cx_COHERENCE register fields */
 #define CM_GCR_Cx_COHERENCE_COHDOMAINEN_SHF	0
 #define CM_GCR_Cx_COHERENCE_COHDOMAINEN_MSK	(_ULCAST_(0xff) << 0)
@@ -324,4 +391,18 @@ static inline int mips_cm_l2sync(void)
 	return 0;
 }
 
+/**
+ * mips_cm_revision() - return CM revision
+ *
+ * Return: The revision of the CM, from GCR_REV, or 0 if no CM is present. The
+ * return value should be checked against the CM_REV_* macros.
+ */
+static inline int mips_cm_revision(void)
+{
+	if (!mips_cm_present())
+		return 0;
+
+	return read_gcr_rev();
+}
+
 #endif /* __MIPS_ASM_MIPS_CM_H__ */
diff --git a/arch/mips/include/asm/mips-cpc.h b/arch/mips/include/asm/mips-cpc.h
index 1cebe8c..f386f32 100644
--- a/arch/mips/include/asm/mips-cpc.h
+++ b/arch/mips/include/asm/mips-cpc.h
@@ -28,16 +28,6 @@ extern void __iomem *mips_cpc_base;
 extern phys_addr_t mips_cpc_default_phys_base(void);
 
 /**
- * mips_cpc_phys_base - retrieve the physical base address of the CPC
- *
- * This function returns the physical base address of the Cluster Power
- * Controller memory mapped registers, or 0 if no Cluster Power Controller
- * is present. It may be overriden by individual platforms which determine
- * this address in a different way.
- */
-extern phys_addr_t __weak mips_cpc_phys_base(void);
-
-/**
  * mips_cpc_probe - probe for a Cluster Power Controller
  *
  * Attempt to detect the presence of a Cluster Power Controller. Returns 0 if
diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h
index c5b0956..d3cd8ea 100644
--- a/arch/mips/include/asm/mipsregs.h
+++ b/arch/mips/include/asm/mipsregs.h
@@ -112,6 +112,30 @@
 #define CP0_TX39_CACHE	$7
 
 
+/* Generic EntryLo bit definitions */
+#define ENTRYLO_G		(_ULCAST_(1) << 0)
+#define ENTRYLO_V		(_ULCAST_(1) << 1)
+#define ENTRYLO_D		(_ULCAST_(1) << 2)
+#define ENTRYLO_C_SHIFT		3
+#define ENTRYLO_C		(_ULCAST_(7) << ENTRYLO_C_SHIFT)
+
+/* R3000 EntryLo bit definitions */
+#define R3K_ENTRYLO_G		(_ULCAST_(1) << 8)
+#define R3K_ENTRYLO_V		(_ULCAST_(1) << 9)
+#define R3K_ENTRYLO_D		(_ULCAST_(1) << 10)
+#define R3K_ENTRYLO_N		(_ULCAST_(1) << 11)
+
+/* MIPS32/64 EntryLo bit definitions */
+#ifdef CONFIG_64BIT
+/* as read by dmfc0 */
+#define MIPS_ENTRYLO_XI		(_ULCAST_(1) << 62)
+#define MIPS_ENTRYLO_RI		(_ULCAST_(1) << 63)
+#else
+/* as read by mfc0 */
+#define MIPS_ENTRYLO_XI		(_ULCAST_(1) << 30)
+#define MIPS_ENTRYLO_RI		(_ULCAST_(1) << 31)
+#endif
+
 /*
  * Values for PageMask register
  */
@@ -203,6 +227,9 @@
 #define PG_ESP		(_ULCAST_(1) <<	 28)
 #define PG_IEC		(_ULCAST_(1) <<  27)
 
+/* MIPS32/64 EntryHI bit definitions */
+#define MIPS_ENTRYHI_EHINV	(_ULCAST_(1) << 10)
+
 /*
  * R4x00 interrupt enable / cause bits
  */
@@ -579,6 +606,8 @@
 
 #define MIPS_CONF7_IAR		(_ULCAST_(1) << 10)
 #define MIPS_CONF7_AR		(_ULCAST_(1) << 16)
+/* FTLB probability bits for R6 */
+#define MIPS_CONF7_FTLBP_SHIFT	(18)
 
 /* MAAR bit definitions */
 #define MIPS_MAAR_ADDR		((BIT_ULL(BITS_PER_LONG - 12) - 1) << 12)
@@ -586,31 +615,6 @@
 #define MIPS_MAAR_S		(_ULCAST_(1) << 1)
 #define MIPS_MAAR_V		(_ULCAST_(1) << 0)
 
-/*  EntryHI bit definition */
-#define MIPS_ENTRYHI_EHINV	(_ULCAST_(1) << 10)
-
-/* R3000 EntryLo bit definitions */
-#define R3K_ENTRYLO_G		(_ULCAST_(1) << 8)
-#define R3K_ENTRYLO_V		(_ULCAST_(1) << 9)
-#define R3K_ENTRYLO_D		(_ULCAST_(1) << 10)
-#define R3K_ENTRYLO_N		(_ULCAST_(1) << 11)
-
-/* R4000 compatible EntryLo bit definitions */
-#define MIPS_ENTRYLO_G		(_ULCAST_(1) << 0)
-#define MIPS_ENTRYLO_V		(_ULCAST_(1) << 1)
-#define MIPS_ENTRYLO_D		(_ULCAST_(1) << 2)
-#define MIPS_ENTRYLO_C_SHIFT	3
-#define MIPS_ENTRYLO_C		(_ULCAST_(7) << MIPS_ENTRYLO_C_SHIFT)
-#ifdef CONFIG_64BIT
-/* as read by dmfc0 */
-#define MIPS_ENTRYLO_XI		(_ULCAST_(1) << 62)
-#define MIPS_ENTRYLO_RI		(_ULCAST_(1) << 63)
-#else
-/* as read by mfc0 */
-#define MIPS_ENTRYLO_XI		(_ULCAST_(1) << 30)
-#define MIPS_ENTRYLO_RI		(_ULCAST_(1) << 31)
-#endif
-
 /* CMGCRBase bit definitions */
 #define MIPS_CMGCRB_BASE	11
 #define MIPS_CMGCRF_BASE	(~_ULCAST_((1 << MIPS_CMGCRB_BASE) - 1))
@@ -932,7 +936,7 @@ do {								\
  */
 
 #define __read_32bit_c0_register(source, sel)				\
-({ int __res;								\
+({ unsigned int __res;							\
 	if (sel == 0)							\
 		__asm__ __volatile__(					\
 			"mfc0\t%0, " #source "\n\t"			\
@@ -1014,7 +1018,7 @@ do {									\
  * On RM7000/RM9000 these are uses to access cop0 set 1 registers
  */
 #define __read_32bit_c0_ctrl_register(source)				\
-({ int __res;								\
+({ unsigned int __res;							\
 	__asm__ __volatile__(						\
 		"cfc0\t%0, " #source "\n\t"				\
 		: "=r" (__res));					\
@@ -1471,7 +1475,7 @@ do {									\
  */
 #define _read_32bit_cp1_register(source, gas_hardfloat)			\
 ({									\
-	int __res;							\
+	unsigned int __res;						\
 									\
 	__asm__ __volatile__(						\
 	"	.set	push					\n"	\
diff --git a/arch/mips/include/asm/msa.h b/arch/mips/include/asm/msa.h
index af5638b..bbb85fe 100644
--- a/arch/mips/include/asm/msa.h
+++ b/arch/mips/include/asm/msa.h
@@ -14,10 +14,90 @@
 
 #ifndef __ASSEMBLY__
 
+#include <asm/inst.h>
+
 extern void _save_msa(struct task_struct *);
 extern void _restore_msa(struct task_struct *);
 extern void _init_msa_upper(void);
 
+extern void read_msa_wr_b(unsigned idx, union fpureg *to);
+extern void read_msa_wr_h(unsigned idx, union fpureg *to);
+extern void read_msa_wr_w(unsigned idx, union fpureg *to);
+extern void read_msa_wr_d(unsigned idx, union fpureg *to);
+
+/**
+ * read_msa_wr() - Read a single MSA vector register
+ * @idx:	The index of the vector register to read
+ * @to:		The FPU register union to store the registers value in
+ * @fmt:	The format of the data in the vector register
+ *
+ * Read the value of MSA vector register idx into the FPU register
+ * union to, using the format fmt.
+ */
+static inline void read_msa_wr(unsigned idx, union fpureg *to,
+			       enum msa_2b_fmt fmt)
+{
+	switch (fmt) {
+	case msa_fmt_b:
+		read_msa_wr_b(idx, to);
+		break;
+
+	case msa_fmt_h:
+		read_msa_wr_h(idx, to);
+		break;
+
+	case msa_fmt_w:
+		read_msa_wr_w(idx, to);
+		break;
+
+	case msa_fmt_d:
+		read_msa_wr_d(idx, to);
+		break;
+
+	default:
+		BUG();
+	}
+}
+
+extern void write_msa_wr_b(unsigned idx, union fpureg *from);
+extern void write_msa_wr_h(unsigned idx, union fpureg *from);
+extern void write_msa_wr_w(unsigned idx, union fpureg *from);
+extern void write_msa_wr_d(unsigned idx, union fpureg *from);
+
+/**
+ * write_msa_wr() - Write a single MSA vector register
+ * @idx:	The index of the vector register to write
+ * @from:	The FPU register union to take the registers value from
+ * @fmt:	The format of the data in the vector register
+ *
+ * Write the value from the FPU register union from into MSA vector
+ * register idx, using the format fmt.
+ */
+static inline void write_msa_wr(unsigned idx, union fpureg *from,
+				enum msa_2b_fmt fmt)
+{
+	switch (fmt) {
+	case msa_fmt_b:
+		write_msa_wr_b(idx, from);
+		break;
+
+	case msa_fmt_h:
+		write_msa_wr_h(idx, from);
+		break;
+
+	case msa_fmt_w:
+		write_msa_wr_w(idx, from);
+		break;
+
+	case msa_fmt_d:
+		write_msa_wr_d(idx, from);
+		break;
+
+	default:
+		BUG();
+	}
+}
+
 static inline void enable_msa(void)
 {
 	if (cpu_has_msa) {
diff --git a/arch/mips/include/asm/octeon/cvmx-bootinfo.h b/arch/mips/include/asm/octeon/cvmx-bootinfo.h
index c373d95..d92cf59 100644
--- a/arch/mips/include/asm/octeon/cvmx-bootinfo.h
+++ b/arch/mips/include/asm/octeon/cvmx-bootinfo.h
@@ -284,6 +284,7 @@ enum cvmx_board_types_enum {
 	CVMX_BOARD_TYPE_CUST_PRIVATE_MIN = 20001,
 	CVMX_BOARD_TYPE_UBNT_E100 = 20002,
 	CVMX_BOARD_TYPE_CUST_DSR1000N = 20006,
+	CVMX_BOARD_TYPE_KONTRON_S1901 = 21901,
 	CVMX_BOARD_TYPE_CUST_PRIVATE_MAX = 30000,
 
 	/* The remaining range is reserved for future use. */
@@ -384,6 +385,7 @@ static inline const char *cvmx_board_type_to_string(enum
 		ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_CUST_PRIVATE_MIN)
 		ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_UBNT_E100)
 		ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_CUST_DSR1000N)
+		ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_KONTRON_S1901)
 		ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_CUST_PRIVATE_MAX)
 	}
 	return "Unsupported Board";
diff --git a/arch/mips/include/asm/octeon/cvmx-pip.h b/arch/mips/include/asm/octeon/cvmx-pip.h
index df69bfd..c210154 100644
--- a/arch/mips/include/asm/octeon/cvmx-pip.h
+++ b/arch/mips/include/asm/octeon/cvmx-pip.h
@@ -37,7 +37,7 @@
 #include <asm/octeon/cvmx-fpa.h>
 #include <asm/octeon/cvmx-pip-defs.h>
 
-#define CVMX_PIP_NUM_INPUT_PORTS		40
+#define CVMX_PIP_NUM_INPUT_PORTS		48
 #define CVMX_PIP_NUM_WATCHERS			4
 
 /*
diff --git a/arch/mips/include/asm/octeon/cvmx-pko.h b/arch/mips/include/asm/octeon/cvmx-pko.h
index 3da59bb..5f47f76 100644
--- a/arch/mips/include/asm/octeon/cvmx-pko.h
+++ b/arch/mips/include/asm/octeon/cvmx-pko.h
@@ -542,6 +542,9 @@ static inline int cvmx_pko_get_base_queue_per_core(int port, int core)
  */
 static inline int cvmx_pko_get_base_queue(int port)
 {
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX))
+		return port;
+
 	return cvmx_pko_get_base_queue_per_core(port, 0);
 }
 
diff --git a/arch/mips/include/asm/octeon/cvmx-pow-defs.h b/arch/mips/include/asm/octeon/cvmx-pow-defs.h
index 9020ef4..6a3db4b 100644
--- a/arch/mips/include/asm/octeon/cvmx-pow-defs.h
+++ b/arch/mips/include/asm/octeon/cvmx-pow-defs.h
@@ -52,6 +52,12 @@
 #define CVMX_POW_WQ_INT_THRX(offset) (CVMX_ADD_IO_SEG(0x0001670000000080ull) + ((offset) & 15) * 8)
 #define CVMX_POW_WS_PCX(offset) (CVMX_ADD_IO_SEG(0x0001670000000280ull) + ((offset) & 15) * 8)
 
+#define CVMX_SSO_WQ_INT (CVMX_ADD_IO_SEG(0x0001670000001000ull))
+#define CVMX_SSO_WQ_IQ_DIS (CVMX_ADD_IO_SEG(0x0001670000001010ull))
+#define CVMX_SSO_WQ_INT_PC (CVMX_ADD_IO_SEG(0x0001670000001020ull))
+#define CVMX_SSO_PPX_GRP_MSK(offset) (CVMX_ADD_IO_SEG(0x0001670000006000ull) + ((offset) & 31) * 8)
+#define CVMX_SSO_WQ_INT_THRX(offset) (CVMX_ADD_IO_SEG(0x0001670000007000ull) + ((offset) & 63) * 8)
+
 union cvmx_pow_bist_stat {
 	uint64_t u64;
 	struct cvmx_pow_bist_stat_s {
@@ -1286,4 +1292,27 @@ union cvmx_pow_ws_pcx {
 	struct cvmx_pow_ws_pcx_s cnf71xx;
 };
 
+union cvmx_sso_wq_int_thrx {
+	uint64_t u64;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_33_63:31;
+		uint64_t tc_en:1;
+		uint64_t tc_thr:4;
+		uint64_t reserved_26_27:2;
+		uint64_t ds_thr:12;
+		uint64_t reserved_12_13:2;
+		uint64_t iq_thr:12;
+#else
+		uint64_t iq_thr:12;
+		uint64_t reserved_12_13:2;
+		uint64_t ds_thr:12;
+		uint64_t reserved_26_27:2;
+		uint64_t tc_thr:4;
+		uint64_t tc_en:1;
+		uint64_t reserved_33_63:31;
+#endif
+	} s;
+};
+
 #endif
diff --git a/arch/mips/include/asm/octeon/cvmx-pow.h b/arch/mips/include/asm/octeon/cvmx-pow.h
index d5565d7..5153156 100644
--- a/arch/mips/include/asm/octeon/cvmx-pow.h
+++ b/arch/mips/include/asm/octeon/cvmx-pow.h
@@ -1810,10 +1810,11 @@ static inline void cvmx_pow_work_submit(cvmx_wqe_t *wqp, uint32_t tag,
 	cvmx_addr_t ptr;
 	cvmx_pow_tag_req_t tag_req;
 
-	wqp->qos = qos;
-	wqp->tag = tag;
-	wqp->tag_type = tag_type;
-	wqp->grp = grp;
+	wqp->word1.tag = tag;
+	wqp->word1.tag_type = tag_type;
+
+	cvmx_wqe_set_qos(wqp, qos);
+	cvmx_wqe_set_grp(wqp, grp);
 
 	tag_req.u64 = 0;
 	tag_req.s.op = CVMX_POW_TAG_OP_ADDWQ;
diff --git a/arch/mips/include/asm/octeon/cvmx-wqe.h b/arch/mips/include/asm/octeon/cvmx-wqe.h
index 2d6d0c7..0d697aa 100644
--- a/arch/mips/include/asm/octeon/cvmx-wqe.h
+++ b/arch/mips/include/asm/octeon/cvmx-wqe.h
@@ -193,6 +193,53 @@ typedef union {
 	        uint64_t bufs:8;
 #endif
 	} s;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t bufs:8;
+		uint64_t ip_offset:8;
+		uint64_t vlan_valid:1;
+		uint64_t vlan_stacked:1;
+		uint64_t unassigned:1;
+		uint64_t vlan_cfi:1;
+		uint64_t vlan_id:12;
+		uint64_t port:12;		/* MAC/PIP port number. */
+		uint64_t dec_ipcomp:1;
+		uint64_t tcp_or_udp:1;
+		uint64_t dec_ipsec:1;
+		uint64_t is_v6:1;
+		uint64_t software:1;
+		uint64_t L4_error:1;
+		uint64_t is_frag:1;
+		uint64_t IP_exc:1;
+		uint64_t is_bcast:1;
+		uint64_t is_mcast:1;
+		uint64_t not_IP:1;
+		uint64_t rcv_error:1;
+		uint64_t err_code:8;
+#else
+		uint64_t err_code:8;
+		uint64_t rcv_error:1;
+		uint64_t not_IP:1;
+		uint64_t is_mcast:1;
+		uint64_t is_bcast:1;
+		uint64_t IP_exc:1;
+		uint64_t is_frag:1;
+		uint64_t L4_error:1;
+		uint64_t software:1;
+		uint64_t is_v6:1;
+		uint64_t dec_ipsec:1;
+		uint64_t tcp_or_udp:1;
+		uint64_t dec_ipcomp:1;
+		uint64_t port:12;
+		uint64_t vlan_id:12;
+		uint64_t vlan_cfi:1;
+		uint64_t unassigned:1;
+		uint64_t vlan_stacked:1;
+		uint64_t vlan_valid:1;
+		uint64_t ip_offset:8;
+		uint64_t bufs:8;
+#endif
+	} s_cn68xx;
 
 	/* use this to get at the 16 vlan bits */
 	struct {
@@ -355,6 +402,146 @@ typedef union {
 
 } cvmx_pip_wqe_word2;
 
+union cvmx_pip_wqe_word0 {
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		/**
+		 * raw chksum result generated by the HW
+		 */
+		uint16_t hw_chksum;
+		/**
+		 * Field unused by hardware - available for software
+		 */
+		uint8_t unused;
+		/**
+		 * Next pointer used by hardware for list maintenance.
+		 * May be written/read by HW before the work queue
+		 * entry is scheduled to a PP (Only 36 bits used in
+		 * Octeon 1)
+		 */
+		uint64_t next_ptr:40;
+#else
+		uint64_t next_ptr:40;
+		uint8_t unused;
+		uint16_t hw_chksum;
+#endif
+	} cn38xx;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t l4ptr:8;       /* 56..63 */
+		uint64_t unused0:8;     /* 48..55 */
+		uint64_t l3ptr:8;       /* 40..47 */
+		uint64_t l2ptr:8;       /* 32..39 */
+		uint64_t unused1:18;    /* 14..31 */
+		uint64_t bpid:6;        /* 8..13 */
+		uint64_t unused2:2;     /* 6..7 */
+		uint64_t pknd:6;        /* 0..5 */
+#else
+		uint64_t pknd:6;        /* 0..5 */
+		uint64_t unused2:2;     /* 6..7 */
+		uint64_t bpid:6;        /* 8..13 */
+		uint64_t unused1:18;    /* 14..31 */
+		uint64_t l2ptr:8;       /* 32..39 */
+		uint64_t l3ptr:8;       /* 40..47 */
+		uint64_t unused0:8;     /* 48..55 */
+		uint64_t l4ptr:8;       /* 56..63 */
+#endif
+	} cn68xx;
+};
+
+union cvmx_wqe_word0 {
+	uint64_t u64;
+	union cvmx_pip_wqe_word0 pip;
+};
+
+union cvmx_wqe_word1 {
+	uint64_t u64;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t len:16;
+		uint64_t varies:14;
+		/**
+		 * the type of the tag (ORDERED, ATOMIC, NULL)
+		 */
+		uint64_t tag_type:2;
+		uint64_t tag:32;
+#else
+		uint64_t tag:32;
+		uint64_t tag_type:2;
+		uint64_t varies:14;
+		uint64_t len:16;
+#endif
+	};
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t len:16;
+		uint64_t zero_0:1;
+		/**
+		 * HW sets this to what it thought the priority of
+		 * the input packet was
+		 */
+		uint64_t qos:3;
+
+		uint64_t zero_1:1;
+		/**
+		 * the group that the work queue entry will be scheduled to
+		 */
+		uint64_t grp:6;
+		uint64_t zero_2:3;
+		uint64_t tag_type:2;
+		uint64_t tag:32;
+#else
+		uint64_t tag:32;
+		uint64_t tag_type:2;
+		uint64_t zero_2:3;
+		uint64_t grp:6;
+		uint64_t zero_1:1;
+		uint64_t qos:3;
+		uint64_t zero_0:1;
+		uint64_t len:16;
+#endif
+	} cn68xx;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		/**
+		 * HW sets to the total number of bytes in the packet
+		 */
+		uint64_t len:16;
+		/**
+		 * HW sets this to input physical port
+		 */
+		uint64_t ipprt:6;
+
+		/**
+		 * HW sets this to what it thought the priority of
+		 * the input packet was
+		 */
+		uint64_t qos:3;
+
+		/**
+		 * the group that the work queue entry will be scheduled to
+		 */
+		uint64_t grp:4;
+		/**
+		 * the type of the tag (ORDERED, ATOMIC, NULL)
+		 */
+		uint64_t tag_type:3;
+		/**
+		 * the synchronization/ordering tag
+		 */
+		uint64_t tag:32;
+#else
+		uint64_t tag:32;
+		uint64_t tag_type:2;
+		uint64_t zero_2:1;
+		uint64_t grp:4;
+		uint64_t qos:3;
+		uint64_t ipprt:6;
+		uint64_t len:16;
+#endif
+	} cn38xx;
+};
+
 /**
  * Work queue entry format
  *
@@ -366,70 +553,13 @@ typedef struct {
      * WORD 0
      *	HW WRITE: the following 64 bits are filled by HW when a packet arrives
      */
-
-#ifdef __BIG_ENDIAN_BITFIELD
-    /**
-     * raw chksum result generated by the HW
-     */
-	uint16_t hw_chksum;
-    /**
-     * Field unused by hardware - available for software
-     */
-	uint8_t unused;
-    /**
-     * Next pointer used by hardware for list maintenance.
-     * May be written/read by HW before the work queue
-     *		 entry is scheduled to a PP
-     * (Only 36 bits used in Octeon 1)
-     */
-	uint64_t next_ptr:40;
-#else
-	uint64_t next_ptr:40;
-	uint8_t unused;
-	uint16_t hw_chksum;
-#endif
+	union cvmx_wqe_word0 word0;
 
     /*****************************************************************
      * WORD 1
      *	HW WRITE: the following 64 bits are filled by HW when a packet arrives
      */
-
-#ifdef __BIG_ENDIAN_BITFIELD
-    /**
-     * HW sets to the total number of bytes in the packet
-     */
-	uint64_t len:16;
-    /**
-     * HW sets this to input physical port
-     */
-	uint64_t ipprt:6;
-
-    /**
-     * HW sets this to what it thought the priority of the input packet was
-     */
-	uint64_t qos:3;
-
-    /**
-     * the group that the work queue entry will be scheduled to
-     */
-	uint64_t grp:4;
-    /**
-     * the type of the tag (ORDERED, ATOMIC, NULL)
-     */
-	uint64_t tag_type:3;
-    /**
-     * the synchronization/ordering tag
-     */
-	uint64_t tag:32;
-#else
-	uint64_t tag:32;
-	uint64_t tag_type:2;
-	uint64_t zero_2:1;
-	uint64_t grp:4;
-	uint64_t qos:3;
-	uint64_t ipprt:6;
-	uint64_t len:16;
-#endif
+	union cvmx_wqe_word1 word1;
 
     /**
      * WORD 2 HW WRITE: the following 64-bits are filled in by
@@ -465,4 +595,64 @@ typedef struct {
 
 } CVMX_CACHE_LINE_ALIGNED cvmx_wqe_t;
 
+static inline int cvmx_wqe_get_port(cvmx_wqe_t *work)
+{
+	int port;
+
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		port = work->word2.s_cn68xx.port;
+	else
+		port = work->word1.cn38xx.ipprt;
+
+	return port;
+}
+
+static inline void cvmx_wqe_set_port(cvmx_wqe_t *work, int port)
+{
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		work->word2.s_cn68xx.port = port;
+	else
+		work->word1.cn38xx.ipprt = port;
+}
+
+static inline int cvmx_wqe_get_grp(cvmx_wqe_t *work)
+{
+	int grp;
+
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		grp = work->word1.cn68xx.grp;
+	else
+		grp = work->word1.cn38xx.grp;
+
+	return grp;
+}
+
+static inline void cvmx_wqe_set_grp(cvmx_wqe_t *work, int grp)
+{
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		work->word1.cn68xx.grp = grp;
+	else
+		work->word1.cn38xx.grp = grp;
+}
+
+static inline int cvmx_wqe_get_qos(cvmx_wqe_t *work)
+{
+	int qos;
+
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		qos = work->word1.cn68xx.qos;
+	else
+		qos = work->word1.cn38xx.qos;
+
+	return qos;
+}
+
+static inline void cvmx_wqe_set_qos(cvmx_wqe_t *work, int qos)
+{
+	if (octeon_has_feature(OCTEON_FEATURE_CN68XX_WQE))
+		work->word1.cn68xx.qos = qos;
+	else
+		work->word1.cn38xx.qos = qos;
+}
+
 #endif /* __CVMX_WQE_H__ */
diff --git a/arch/mips/include/asm/pgtable-bits.h b/arch/mips/include/asm/pgtable-bits.h
index c28a849..ff7ad91 100644
--- a/arch/mips/include/asm/pgtable-bits.h
+++ b/arch/mips/include/asm/pgtable-bits.h
@@ -133,20 +133,13 @@
 #define _PAGE_HUGE		(1 << _PAGE_HUGE_SHIFT)
 #define _PAGE_SPLITTING_SHIFT	(_PAGE_HUGE_SHIFT + 1)
 #define _PAGE_SPLITTING		(1 << _PAGE_SPLITTING_SHIFT)
-
-/* Only R2 or newer cores have the XI bit */
-#if defined(CONFIG_CPU_MIPSR2) || defined(CONFIG_CPU_MIPSR6)
-#define _PAGE_NO_EXEC_SHIFT	(_PAGE_SPLITTING_SHIFT + 1)
-#else
-#define _PAGE_GLOBAL_SHIFT	(_PAGE_SPLITTING_SHIFT + 1)
-#define _PAGE_GLOBAL		(1 << _PAGE_GLOBAL_SHIFT)
-#endif	/* CONFIG_CPU_MIPSR2 || CONFIG_CPU_MIPSR6 */
-
 #endif	/* CONFIG_64BIT && CONFIG_MIPS_HUGE_TLB_SUPPORT */
 
 #if defined(CONFIG_CPU_MIPSR2) || defined(CONFIG_CPU_MIPSR6)
 /* XI - page cannot be executed */
-#ifndef _PAGE_NO_EXEC_SHIFT
+#ifdef _PAGE_SPLITTING_SHIFT
+#define _PAGE_NO_EXEC_SHIFT	(_PAGE_SPLITTING_SHIFT + 1)
+#else
 #define _PAGE_NO_EXEC_SHIFT	(_PAGE_MODIFIED_SHIFT + 1)
 #endif
 #define _PAGE_NO_EXEC		(cpu_has_rixi ? (1 << _PAGE_NO_EXEC_SHIFT) : 0)
@@ -156,14 +149,16 @@
 #define _PAGE_READ		(cpu_has_rixi ? 0 : (1 << _PAGE_READ_SHIFT))
 #define _PAGE_NO_READ_SHIFT	_PAGE_READ_SHIFT
 #define _PAGE_NO_READ		(cpu_has_rixi ? (1 << _PAGE_READ_SHIFT) : 0)
+#endif	/* defined(CONFIG_CPU_MIPSR2) || defined(CONFIG_CPU_MIPSR6) */
 
+#if defined(_PAGE_NO_READ_SHIFT)
 #define _PAGE_GLOBAL_SHIFT	(_PAGE_NO_READ_SHIFT + 1)
-#define _PAGE_GLOBAL		(1 << _PAGE_GLOBAL_SHIFT)
-
-#else	/* !CONFIG_CPU_MIPSR2 && !CONFIG_CPU_MIPSR6 */
+#elif defined(_PAGE_SPLITTING_SHIFT)
+#define _PAGE_GLOBAL_SHIFT	(_PAGE_SPLITTING_SHIFT + 1)
+#else
 #define _PAGE_GLOBAL_SHIFT	(_PAGE_MODIFIED_SHIFT + 1)
+#endif
 #define _PAGE_GLOBAL		(1 << _PAGE_GLOBAL_SHIFT)
-#endif	/* CONFIG_CPU_MIPSR2 || CONFIG_CPU_MIPSR6 */
 
 #define _PAGE_VALID_SHIFT	(_PAGE_GLOBAL_SHIFT + 1)
 #define _PAGE_VALID		(1 << _PAGE_VALID_SHIFT)
@@ -249,7 +244,7 @@ static inline uint64_t pte_to_entrylo(unsigned long pte_val)
 #define _CACHE_CACHABLE_NONCOHERENT (3<<_CACHE_SHIFT)  /* LOONGSON       */
 #define _CACHE_CACHABLE_COHERENT    (3<<_CACHE_SHIFT)  /* LOONGSON-3     */
 
-#elif defined(CONFIG_MACH_JZ4740)
+#elif defined(CONFIG_MACH_INGENIC)
 
 /* Ingenic uses the WA bit to achieve write-combine memory writes */
 #define _CACHE_UNCACHED_ACCELERATED (1<<_CACHE_SHIFT)
diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index ae85694..8957f15 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -393,6 +393,8 @@ static inline pgprot_t pgprot_noncached(pgprot_t _prot)
 	return __pgprot(prot);
 }
 
+#define pgprot_writecombine pgprot_writecombine
+
 static inline pgprot_t pgprot_writecombine(pgprot_t _prot)
 {
 	unsigned long prot = pgprot_val(_prot);
diff --git a/arch/mips/include/asm/processor.h b/arch/mips/include/asm/processor.h
index 9b3b48e..59ee6dc 100644
--- a/arch/mips/include/asm/processor.h
+++ b/arch/mips/include/asm/processor.h
@@ -275,6 +275,7 @@ struct thread_struct {
 	unsigned long cp0_badvaddr;	/* Last user fault */
 	unsigned long cp0_baduaddr;	/* Last kernel fault accessing USEG */
 	unsigned long error_code;
+	unsigned long trap_nr;
 #ifdef CONFIG_CPU_CAVIUM_OCTEON
 	struct octeon_cop2_state cp2 __attribute__ ((__aligned__(128)));
 	struct octeon_cvmseg_state cvmseg __attribute__ ((__aligned__(128)));
@@ -341,6 +342,7 @@ struct thread_struct {
 	.cp0_badvaddr		= 0,				\
 	.cp0_baduaddr		= 0,				\
 	.error_code		= 0,				\
+	.trap_nr		= 0,				\
 	/*							\
 	 * Platform specific cop2 registers(null if no COP2)	\
 	 */							\
diff --git a/arch/mips/include/asm/ptrace.h b/arch/mips/include/asm/ptrace.h
index ffc3203..f6fc6aa 100644
--- a/arch/mips/include/asm/ptrace.h
+++ b/arch/mips/include/asm/ptrace.h
@@ -14,11 +14,16 @@
 #include <linux/linkage.h>
 #include <linux/types.h>
 #include <asm/isadep.h>
+#include <asm/page.h>
+#include <asm/thread_info.h>
 #include <uapi/asm/ptrace.h>
 
 /*
  * This struct defines the way the registers are stored on the stack during a
  * system call/exception. As usual the registers k0/k1 aren't being saved.
+ *
+ * If you add a register here, also add it to regoffset_table[] in
+ * arch/mips/kernel/ptrace.c.
  */
 struct pt_regs {
 #ifdef CONFIG_32BIT
@@ -43,8 +48,83 @@ struct pt_regs {
 	unsigned long long mpl[6];        /* MTM{0-5} */
 	unsigned long long mtp[6];        /* MTP{0-5} */
 #endif
+	unsigned long __last[0];
 } __aligned(8);
 
+static inline unsigned long kernel_stack_pointer(struct pt_regs *regs)
+{
+	return regs->regs[31];
+}
+
+/*
+ * Don't use asm-generic/ptrace.h it defines FP accessors that don't make
+ * sense on MIPS.  We rather want an error if they get invoked.
+ */
+
+static inline void instruction_pointer_set(struct pt_regs *regs,
+                                           unsigned long val)
+{
+	regs->cp0_epc = val;
+}
+
+/* Query offset/name of register from its name/offset */
+extern int regs_query_register_offset(const char *name);
+#define MAX_REG_OFFSET (offsetof(struct pt_regs, __last))
+
+/**
+ * regs_get_register() - get register value from its offset
+ * @regs:       pt_regs from which register value is gotten.
+ * @offset:     offset number of the register.
+ *
+ * regs_get_register returns the value of a register. The @offset is the
+ * offset of the register in struct pt_regs address which specified by @regs.
+ * If @offset is bigger than MAX_REG_OFFSET, this returns 0.
+ */
+static inline unsigned long regs_get_register(struct pt_regs *regs,
+                                              unsigned int offset)
+{
+	if (unlikely(offset > MAX_REG_OFFSET))
+		return 0;
+
+	return *(unsigned long *)((unsigned long)regs + offset);
+}
+
+/**
+ * regs_within_kernel_stack() - check the address in the stack
+ * @regs:       pt_regs which contains kernel stack pointer.
+ * @addr:       address which is checked.
+ *
+ * regs_within_kernel_stack() checks @addr is within the kernel stack page(s).
+ * If @addr is within the kernel stack, it returns true. If not, returns false.
+ */
+static inline int regs_within_kernel_stack(struct pt_regs *regs,
+                                           unsigned long addr)
+{
+	return ((addr & ~(THREAD_SIZE - 1))  ==
+		(kernel_stack_pointer(regs) & ~(THREAD_SIZE - 1)));
+}
+
+/**
+ * regs_get_kernel_stack_nth() - get Nth entry of the stack
+ * @regs:       pt_regs which contains kernel stack pointer.
+ * @n:          stack entry number.
+ *
+ * regs_get_kernel_stack_nth() returns @n th entry of the kernel stack which
+ * is specified by @regs. If the @n th entry is NOT in the kernel stack,
+ * this returns 0.
+ */
+static inline unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs,
+                                                      unsigned int n)
+{
+	unsigned long *addr = (unsigned long *)kernel_stack_pointer(regs);
+
+	addr += n;
+	if (regs_within_kernel_stack(regs, (unsigned long)addr))
+		return *addr;
+	else
+		return 0;
+}
+
 struct task_struct;
 
 extern int ptrace_getregs(struct task_struct *child,
diff --git a/arch/mips/include/asm/signal.h b/arch/mips/include/asm/signal.h
index 8efe5a9..003e273 100644
--- a/arch/mips/include/asm/signal.h
+++ b/arch/mips/include/asm/signal.h
@@ -23,4 +23,7 @@
 
 #define __ARCH_HAS_IRIX_SIGACTION
 
+extern int protected_save_fp_context(void __user *sc);
+extern int protected_restore_fp_context(void __user *sc);
+
 #endif /* _ASM_SIGNAL_H */
diff --git a/arch/mips/include/asm/spinlock.h b/arch/mips/include/asm/spinlock.h
index 9de4ba4..40196be 100644
--- a/arch/mips/include/asm/spinlock.h
+++ b/arch/mips/include/asm/spinlock.h
@@ -42,6 +42,11 @@ static inline int arch_spin_is_locked(arch_spinlock_t *lock)
 	return ((counters >> 16) ^ counters) & 0xffff;
 }
 
+static inline int arch_spin_value_unlocked(arch_spinlock_t lock)
+{
+	return lock.h.serving_now == lock.h.ticket;
+}
+
 #define arch_spin_lock_flags(lock, flags) arch_spin_lock(lock)
 #define arch_spin_unlock_wait(x) \
 	while (arch_spin_is_locked(x)) { cpu_relax(); }
diff --git a/arch/mips/include/asm/switch_to.h b/arch/mips/include/asm/switch_to.h
index 9733cd0..28b5d84 100644
--- a/arch/mips/include/asm/switch_to.h
+++ b/arch/mips/include/asm/switch_to.h
@@ -16,29 +16,21 @@
 #include <asm/watch.h>
 #include <asm/dsp.h>
 #include <asm/cop2.h>
-#include <asm/msa.h>
+#include <asm/fpu.h>
 
 struct task_struct;
 
-enum {
-	FP_SAVE_NONE	= 0,
-	FP_SAVE_VECTOR	= -1,
-	FP_SAVE_SCALAR	= 1,
-};
-
 /**
  * resume - resume execution of a task
  * @prev:	The task previously executed.
  * @next:	The task to begin executing.
  * @next_ti:	task_thread_info(next).
- * @fp_save:	Which, if any, FP context to save for prev.
  *
  * This function is used whilst scheduling to save the context of prev & load
  * the context of next. Returns prev.
  */
 extern asmlinkage struct task_struct *resume(struct task_struct *prev,
-		struct task_struct *next, struct thread_info *next_ti,
-		s32 fp_save);
+		struct task_struct *next, struct thread_info *next_ti);
 
 extern unsigned int ll_bit;
 extern struct task_struct *ll_task;
@@ -91,8 +83,8 @@ do {	if (cpu_has_rw_llb) {						\
  */
 #define switch_to(prev, next, last)					\
 do {									\
-	s32 __fpsave = FP_SAVE_NONE;					\
 	__mips_mt_fpaff_switch_to(prev);				\
+	lose_fpu_inatomic(1, prev);					\
 	if (cpu_has_dsp) {						\
 		__save_dsp(prev);					\
 		__restore_dsp(next);					\
@@ -111,15 +103,10 @@ do {									\
 		clear_c0_status(ST0_CU2);				\
 	}								\
 	__clear_software_ll_bit();					\
-	if (test_and_clear_tsk_thread_flag(prev, TIF_USEDFPU))		\
-		__fpsave = FP_SAVE_SCALAR;				\
-	if (test_and_clear_tsk_thread_flag(prev, TIF_USEDMSA))		\
-		__fpsave = FP_SAVE_VECTOR;				\
 	if (cpu_has_userlocal)						\
 		write_c0_userlocal(task_thread_info(next)->tp_value);	\
 	__restore_watch();						\
-	disable_msa();							\
-	(last) = resume(prev, next, task_thread_info(next), __fpsave);	\
+	(last) = resume(prev, next, task_thread_info(next));		\
 } while (0)
 
 #endif /* _ASM_SWITCH_TO_H */
diff --git a/arch/mips/include/asm/thread_info.h b/arch/mips/include/asm/thread_info.h
index 9c0014e..e309d8f 100644
--- a/arch/mips/include/asm/thread_info.h
+++ b/arch/mips/include/asm/thread_info.h
@@ -99,6 +99,7 @@ static inline struct thread_info *current_thread_info(void)
 #define TIF_SYSCALL_AUDIT	3	/* syscall auditing active */
 #define TIF_SECCOMP		4	/* secure computing */
 #define TIF_NOTIFY_RESUME	5	/* callback before returning to user */
+#define TIF_UPROBE		6	/* breakpointed or singlestepping */
 #define TIF_RESTORE_SIGMASK	9	/* restore signal mask in do_signal() */
 #define TIF_USEDFPU		16	/* FPU was used by this task this quantum (SMP) */
 #define TIF_MEMDIE		18	/* is terminating due to OOM killer */
@@ -122,6 +123,7 @@ static inline struct thread_info *current_thread_info(void)
 #define _TIF_SYSCALL_AUDIT	(1<<TIF_SYSCALL_AUDIT)
 #define _TIF_SECCOMP		(1<<TIF_SECCOMP)
 #define _TIF_NOTIFY_RESUME	(1<<TIF_NOTIFY_RESUME)
+#define _TIF_UPROBE		(1<<TIF_UPROBE)
 #define _TIF_USEDFPU		(1<<TIF_USEDFPU)
 #define _TIF_NOHZ		(1<<TIF_NOHZ)
 #define _TIF_FIXADE		(1<<TIF_FIXADE)
@@ -146,7 +148,8 @@ static inline struct thread_info *current_thread_info(void)
 
 /* work to do on interrupt/exception return */
 #define _TIF_WORK_MASK		\
-	(_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_NOTIFY_RESUME)
+	(_TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_NOTIFY_RESUME |	\
+	 _TIF_UPROBE)
 /* work to do on any return to u-space */
 #define _TIF_ALLWORK_MASK	(_TIF_NOHZ | _TIF_WORK_MASK |		\
 				 _TIF_WORK_SYSCALL_EXIT |		\
diff --git a/arch/mips/include/asm/time.h b/arch/mips/include/asm/time.h
index 8ab2874..17d4cd2 100644
--- a/arch/mips/include/asm/time.h
+++ b/arch/mips/include/asm/time.h
@@ -51,7 +51,7 @@ extern int __weak get_c0_perfcount_int(void);
 /*
  * Initialize the calling CPU's compare interrupt as clockevent device
  */
-extern unsigned int __weak get_c0_compare_int(void);
+extern unsigned int get_c0_compare_int(void);
 extern int r4k_clockevent_init(void);
 
 static inline int mips_clockevent_init(void)
diff --git a/arch/mips/include/asm/tlbdebug.h b/arch/mips/include/asm/tlbdebug.h
index bb8f5c2..3a25a87 100644
--- a/arch/mips/include/asm/tlbdebug.h
+++ b/arch/mips/include/asm/tlbdebug.h
@@ -11,6 +11,7 @@
 /*
  * TLB debugging functions:
  */
+extern void dump_tlb_regs(void);
 extern void dump_tlb_all(void);
 
 #endif /* __ASM_TLBDEBUG_H */
diff --git a/arch/mips/include/asm/uprobes.h b/arch/mips/include/asm/uprobes.h
new file mode 100644
index 0000000..34c325c
--- /dev/null
+++ b/arch/mips/include/asm/uprobes.h
@@ -0,0 +1,58 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+#ifndef __ASM_UPROBES_H
+#define __ASM_UPROBES_H
+
+#include <linux/notifier.h>
+#include <linux/types.h>
+
+#include <asm/break.h>
+#include <asm/inst.h>
+
+/*
+ * We want this to be defined as union mips_instruction but that makes the
+ * generic code blow up.
+ */
+typedef u32 uprobe_opcode_t;
+
+/*
+ * Classic MIPS (note this implementation doesn't consider microMIPS yet)
+ * instructions are always 4 bytes but in order to deal with branches and
+ * their delay slots, we treat instructions as having 8 bytes maximum.
+ */
+#define MAX_UINSN_BYTES			8
+#define UPROBE_XOL_SLOT_BYTES		128	/* Max. cache line size */
+
+#define UPROBE_BRK_UPROBE		0x000d000d	/* break 13 */
+#define UPROBE_BRK_UPROBE_XOL		0x000e000d	/* break 14 */
+
+#define UPROBE_SWBP_INSN		UPROBE_BRK_UPROBE
+#define UPROBE_SWBP_INSN_SIZE		4
+
+struct arch_uprobe {
+	unsigned long	resume_epc;
+	u32	insn[2];
+	u32	ixol[2];
+	union	mips_instruction orig_inst[MAX_UINSN_BYTES / 4];
+};
+
+struct arch_uprobe_task {
+	unsigned long saved_trap_nr;
+};
+
+extern int arch_uprobe_analyze_insn(struct arch_uprobe *aup,
+	struct mm_struct *mm, unsigned long addr);
+extern int arch_uprobe_pre_xol(struct arch_uprobe *aup, struct pt_regs *regs);
+extern int arch_uprobe_post_xol(struct arch_uprobe *aup, struct pt_regs *regs);
+extern bool arch_uprobe_xol_was_trapped(struct task_struct *tsk);
+extern int arch_uprobe_exception_notify(struct notifier_block *self,
+	unsigned long val, void *data);
+extern void arch_uprobe_abort_xol(struct arch_uprobe *aup,
+	struct pt_regs *regs);
+extern unsigned long arch_uretprobe_hijack_return_addr(
+	unsigned long trampoline_vaddr, struct pt_regs *regs);
+
+#endif /* __ASM_UPROBES_H */
diff --git a/arch/mips/include/asm/vpe.h b/arch/mips/include/asm/vpe.h
index 7849f39..80e70db 100644
--- a/arch/mips/include/asm/vpe.h
+++ b/arch/mips/include/asm/vpe.h
@@ -122,7 +122,7 @@ void release_vpe(struct vpe *v);
 void *alloc_progmem(unsigned long len);
 void release_progmem(void *ptr);
 
-int __weak vpe_run(struct vpe *v);
+int vpe_run(struct vpe *v);
 void cleanup_tc(struct tc *tc);
 
 int __init vpe_module_init(void);
diff --git a/arch/mips/include/uapi/asm/break.h b/arch/mips/include/uapi/asm/break.h
index 002c39e..9c4265c 100644
--- a/arch/mips/include/uapi/asm/break.h
+++ b/arch/mips/include/uapi/asm/break.h
@@ -21,6 +21,8 @@
 #define BRK_DIVZERO	7	/* Divide by zero check */
 #define BRK_RANGE	8	/* Range error check */
 #define BRK_BUG		12	/* Used by BUG() */
+#define BRK_UPROBE	13	/* See <asm/uprobes.h> */
+#define BRK_UPROBE_XOL	14	/* See <asm/uprobes.h> */
 #define BRK_MEMU	514	/* Used by FPU emulator */
 #define BRK_KPROBE_BP	515	/* Kprobe break */
 #define BRK_KPROBE_SSTEPBP 516	/* Kprobe single step software implementation */
diff --git a/arch/mips/include/uapi/asm/hwcap.h b/arch/mips/include/uapi/asm/hwcap.h
new file mode 100644
index 0000000..c7484a7
--- /dev/null
+++ b/arch/mips/include/uapi/asm/hwcap.h
@@ -0,0 +1,8 @@
+#ifndef _UAPI_ASM_HWCAP_H
+#define _UAPI_ASM_HWCAP_H
+
+/* HWCAP flags */
+#define HWCAP_MIPS_R6		(1 << 0)
+#define HWCAP_MIPS_MSA		(1 << 1)
+
+#endif /* _UAPI_ASM_HWCAP_H */
diff --git a/arch/mips/include/uapi/asm/inst.h b/arch/mips/include/uapi/asm/inst.h
index fc0cf5a..9b44d5a 100644
--- a/arch/mips/include/uapi/asm/inst.h
+++ b/arch/mips/include/uapi/asm/inst.h
@@ -26,7 +26,7 @@ enum major_op {
 	cop0_op, cop1_op, cop2_op, cop1x_op,
 	beql_op, bnel_op, blezl_op, bgtzl_op,
 	daddi_op, cbcond1_op = daddi_op, daddiu_op, ldl_op, ldr_op,
-	spec2_op, jalx_op, mdmx_op, spec3_op,
+	spec2_op, jalx_op, mdmx_op, msa_op = mdmx_op, spec3_op,
 	lb_op, lh_op, lwl_op, lw_op,
 	lbu_op, lhu_op, lwr_op, lwu_op,
 	sb_op, sh_op, swl_op, sw_op,
@@ -167,8 +167,13 @@ enum cop1_sdw_func {
 	fround_op    =	0x0c, ftrunc_op	   =  0x0d,
 	fceil_op     =	0x0e, ffloor_op	   =  0x0f,
 	fmovc_op     =	0x11, fmovz_op	   =  0x12,
-	fmovn_op     =	0x13, frecip_op	   =  0x15,
-	frsqrt_op    =	0x16, fcvts_op	   =  0x20,
+	fmovn_op     =	0x13, fseleqz_op   =  0x14,
+	frecip_op    =  0x15, frsqrt_op    =  0x16,
+	fselnez_op   =  0x17, fmaddf_op    =  0x18,
+	fmsubf_op    =  0x19, frint_op     =  0x1a,
+	fclass_op    =  0x1b, fmin_op      =  0x1c,
+	fmina_op     =  0x1d, fmax_op      =  0x1e,
+	fmaxa_op     =  0x1f, fcvts_op     =  0x20,
 	fcvtd_op     =	0x21, fcvte_op	   =  0x22,
 	fcvtw_op     =	0x24, fcvtl_op	   =  0x25,
 	fcmp_op	     =	0x30
@@ -221,6 +226,24 @@ enum bshfl_func {
 };
 
 /*
+ * func field for MSA MI10 format.
+ */
+enum msa_mi10_func {
+	msa_ld_op = 8,
+	msa_st_op = 9,
+};
+
+/*
+ * MSA 2 bit format fields.
+ */
+enum msa_2b_fmt {
+	msa_fmt_b = 0,
+	msa_fmt_h = 1,
+	msa_fmt_w = 2,
+	msa_fmt_d = 3,
+};
+
+/*
  * (microMIPS) Major opcodes.
  */
 enum mm_major_op {
@@ -611,6 +634,16 @@ struct v_format {				/* MDMX vector format */
 	;)))))))
 };
 
+struct msa_mi10_format {		/* MSA MI10 */
+	__BITFIELD_FIELD(unsigned int opcode : 6,
+	__BITFIELD_FIELD(signed int s10 : 10,
+	__BITFIELD_FIELD(unsigned int rs : 5,
+	__BITFIELD_FIELD(unsigned int wd : 5,
+	__BITFIELD_FIELD(unsigned int func : 4,
+	__BITFIELD_FIELD(unsigned int df : 2,
+	;))))))
+};
+
 struct spec3_format {   /* SPEC3 */
 	__BITFIELD_FIELD(unsigned int opcode:6,
 	__BITFIELD_FIELD(unsigned int rs:5,
@@ -888,6 +921,7 @@ union mips_instruction {
 	struct p_format p_format;
 	struct f_format f_format;
 	struct ma_format ma_format;
+	struct msa_mi10_format msa_mi10_format;
 	struct b_format b_format;
 	struct ps_format ps_format;
 	struct v_format v_format;
diff --git a/arch/mips/include/uapi/asm/sigcontext.h b/arch/mips/include/uapi/asm/sigcontext.h
index 9081d88..5cbd9ae 100644
--- a/arch/mips/include/uapi/asm/sigcontext.h
+++ b/arch/mips/include/uapi/asm/sigcontext.h
@@ -12,6 +12,18 @@
 #include <linux/types.h>
 #include <asm/sgidefs.h>
 
+/* scalar FP context was used */
+#define USED_FP			(1 << 0)
+
+/* the value of Status.FR when context was saved */
+#define USED_FR1		(1 << 1)
+
+/* FR=1, but with odd singles in bits 63:32 of preceding even double */
+#define USED_HYBRID_FPRS	(1 << 2)
+
+/* extended context was used, see struct extcontext for details */
+#define USED_EXTCONTEXT		(1 << 3)
+
 #if _MIPS_SIM == _MIPS_SIM_ABI32
 
 /*
diff --git a/arch/mips/include/uapi/asm/swab.h b/arch/mips/include/uapi/asm/swab.h
index 8f2d184..c4ddc4f 100644
--- a/arch/mips/include/uapi/asm/swab.h
+++ b/arch/mips/include/uapi/asm/swab.h
@@ -16,11 +16,13 @@
 #if (defined(__mips_isa_rev) && (__mips_isa_rev >= 2)) ||		\
     defined(_MIPS_ARCH_LOONGSON3A)
 
-static inline __attribute_const__ __u16 __arch_swab16(__u16 x)
+static inline __attribute__((nomips16)) __attribute_const__
+		__u16 __arch_swab16(__u16 x)
 {
 	__asm__(
 	"	.set	push			\n"
 	"	.set	arch=mips32r2		\n"
+	"	.set	nomips16		\n"
 	"	wsbh	%0, %1			\n"
 	"	.set	pop			\n"
 	: "=r" (x)
@@ -30,11 +32,13 @@ static inline __attribute_const__ __u16 __arch_swab16(__u16 x)
 }
 #define __arch_swab16 __arch_swab16
 
-static inline __attribute_const__ __u32 __arch_swab32(__u32 x)
+static inline __attribute__((nomips16)) __attribute_const__
+		__u32 __arch_swab32(__u32 x)
 {
 	__asm__(
 	"	.set	push			\n"
 	"	.set	arch=mips32r2		\n"
+	"	.set	nomips16		\n"
 	"	wsbh	%0, %1			\n"
 	"	rotr	%0, %0, 16		\n"
 	"	.set	pop			\n"
@@ -50,11 +54,13 @@ static inline __attribute_const__ __u32 __arch_swab32(__u32 x)
  * 64-bit kernel on r2 CPUs.
  */
 #ifdef __mips64
-static inline __attribute_const__ __u64 __arch_swab64(__u64 x)
+static inline __attribute__((nomips16)) __attribute_const__
+		__u64 __arch_swab64(__u64 x)
 {
 	__asm__(
 	"	.set	push			\n"
 	"	.set	arch=mips64r2		\n"
+	"	.set	nomips16		\n"
 	"	dsbh	%0, %1			\n"
 	"	dshd	%0, %0			\n"
 	"	.set	pop			\n"
diff --git a/arch/mips/include/uapi/asm/ucontext.h b/arch/mips/include/uapi/asm/ucontext.h
new file mode 100644
index 0000000..2320144
--- /dev/null
+++ b/arch/mips/include/uapi/asm/ucontext.h
@@ -0,0 +1,65 @@
+#ifndef __MIPS_UAPI_ASM_UCONTEXT_H
+#define __MIPS_UAPI_ASM_UCONTEXT_H
+
+/**
+ * struct extcontext - extended context header structure
+ * @magic:	magic value identifying the type of extended context
+ * @size:	the size in bytes of the enclosing structure
+ *
+ * Extended context structures provide context which does not fit within struct
+ * sigcontext. They are placed sequentially in memory at the end of struct
+ * ucontext and struct sigframe, with each extended context structure beginning
+ * with a header defined by this struct. The type of context represented is
+ * indicated by the magic field. Userland may check each extended context
+ * structure against magic values that it recognises. The size field allows any
+ * unrecognised context to be skipped, allowing for future expansion. The end
+ * of the extended context data is indicated by the magic value
+ * END_EXTCONTEXT_MAGIC.
+ */
+struct extcontext {
+	unsigned int		magic;
+	unsigned int		size;
+};
+
+/**
+ * struct msa_extcontext - MSA extended context structure
+ * @ext:	the extended context header, with magic == MSA_EXTCONTEXT_MAGIC
+ * @wr:		the most significant 64 bits of each MSA vector register
+ * @csr:	the value of the MSA control & status register
+ *
+ * If MSA context is live for a task at the time a signal is delivered to it,
+ * this structure will hold the MSA context of the task as it was prior to the
+ * signal delivery.
+ */
+struct msa_extcontext {
+	struct extcontext	ext;
+#define MSA_EXTCONTEXT_MAGIC	0x784d5341	/* xMSA */
+
+	unsigned long long	wr[32];
+	unsigned int		csr;
+};
+
+#define END_EXTCONTEXT_MAGIC	0x78454e44	/* xEND */
+
+/**
+ * struct ucontext - user context structure
+ * @uc_flags:
+ * @uc_link:
+ * @uc_stack:
+ * @uc_mcontext:	holds basic processor state
+ * @uc_sigmask:
+ * @uc_extcontext:	holds extended processor state
+ */
+struct ucontext {
+	/* Historic fields matching asm-generic */
+	unsigned long		uc_flags;
+	struct ucontext		*uc_link;
+	stack_t			uc_stack;
+	struct sigcontext	uc_mcontext;
+	sigset_t		uc_sigmask;
+
+	/* Extended context structures may follow ucontext */
+	unsigned long long	uc_extcontext[0];
+};
+
+#endif /* __MIPS_UAPI_ASM_UCONTEXT_H */
diff --git a/arch/mips/jazz/irq.c b/arch/mips/jazz/irq.c
index e1ea4f6..5d6828b 100644
--- a/arch/mips/jazz/irq.c
+++ b/arch/mips/jazz/irq.c
@@ -110,18 +110,11 @@ asmlinkage void plat_irq_dispatch(void)
 	}
 }
 
-static void r4030_set_mode(enum clock_event_mode mode,
-			   struct clock_event_device *evt)
-{
-	/* Nothing to do ...  */
-}
-
 struct clock_event_device r4030_clockevent = {
 	.name		= "r4030",
 	.features	= CLOCK_EVT_FEAT_PERIODIC,
 	.rating		= 300,
 	.irq		= JAZZ_TIMER_IRQ,
-	.set_mode	= r4030_set_mode,
 };
 
 static irqreturn_t r4030_timer_interrupt(int irq, void *dev_id)
diff --git a/arch/mips/jz4740/gpio.c b/arch/mips/jz4740/gpio.c
index 54c80d4..6cd69fd 100644
--- a/arch/mips/jz4740/gpio.c
+++ b/arch/mips/jz4740/gpio.c
@@ -231,6 +231,13 @@ static int jz_gpio_direction_input(struct gpio_chip *chip, unsigned gpio)
 	return 0;
 }
 
+static int jz_gpio_to_irq(struct gpio_chip *chip, unsigned gpio)
+{
+	struct jz_gpio_chip *jz_gpio = gpio_chip_to_jz_gpio_chip(chip);
+
+	return jz_gpio->irq_base + gpio;
+}
+
 int jz_gpio_port_direction_input(int port, uint32_t mask)
 {
 	writel(mask, GPIO_TO_REG(port, JZ_REG_GPIO_DIRECTION_CLEAR));
@@ -262,18 +269,6 @@ uint32_t jz_gpio_port_get_value(int port, uint32_t mask)
 }
 EXPORT_SYMBOL(jz_gpio_port_get_value);
 
-int gpio_to_irq(unsigned gpio)
-{
-	return JZ4740_IRQ_GPIO(0) + gpio;
-}
-EXPORT_SYMBOL_GPL(gpio_to_irq);
-
-int irq_to_gpio(unsigned irq)
-{
-	return irq - JZ4740_IRQ_GPIO(0);
-}
-EXPORT_SYMBOL_GPL(irq_to_gpio);
-
 #define IRQ_TO_BIT(irq) BIT(irq_to_gpio(irq) & 0x1f)
 
 static void jz_gpio_check_trigger_both(struct jz_gpio_chip *chip, unsigned int irq)
@@ -403,6 +398,7 @@ static int jz_gpio_irq_set_wake(struct irq_data *data, unsigned int on)
 		.get = jz_gpio_get_value, \
 		.direction_output = jz_gpio_direction_output, \
 		.direction_input = jz_gpio_direction_input, \
+		.to_irq = jz_gpio_to_irq, \
 		.base = JZ4740_GPIO_BASE_ ## _bank, \
 		.ngpio = JZ4740_GPIO_NUM_ ## _bank, \
 	}, \
@@ -423,8 +419,8 @@ static void jz4740_gpio_chip_init(struct jz_gpio_chip *chip, unsigned int id)
 	chip->base = ioremap(JZ4740_GPIO_BASE_ADDR + (id * 0x100), 0x100);
 
 	chip->irq = JZ4740_IRQ_INTC_GPIO(id);
-	irq_set_handler_data(chip->irq, chip);
-	irq_set_chained_handler(chip->irq, jz_gpio_irq_demux_handler);
+	irq_set_chained_handler_and_data(chip->irq,
+					 jz_gpio_irq_demux_handler, chip);
 
 	gc = irq_alloc_generic_chip(chip->gpio_chip.label, 1, chip->irq_base,
 		chip->base, handle_level_irq);
diff --git a/arch/mips/jz4740/time.c b/arch/mips/jz4740/time.c
index 7ab47fe..1f7ca2c 100644
--- a/arch/mips/jz4740/time.c
+++ b/arch/mips/jz4740/time.c
@@ -58,7 +58,7 @@ static irqreturn_t jz4740_clockevent_irq(int irq, void *devid)
 
 	jz4740_timer_ack_full(TIMER_CLOCKEVENT);
 
-	if (cd->mode != CLOCK_EVT_MODE_PERIODIC)
+	if (!clockevent_state_periodic(cd))
 		jz4740_timer_disable(TIMER_CLOCKEVENT);
 
 	cd->event_handler(cd);
@@ -66,24 +66,29 @@ static irqreturn_t jz4740_clockevent_irq(int irq, void *devid)
 	return IRQ_HANDLED;
 }
 
-static void jz4740_clockevent_set_mode(enum clock_event_mode mode,
-	struct clock_event_device *cd)
+static int jz4740_clockevent_set_periodic(struct clock_event_device *evt)
 {
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		jz4740_timer_set_count(TIMER_CLOCKEVENT, 0);
-		jz4740_timer_set_period(TIMER_CLOCKEVENT, jz4740_jiffies_per_tick);
-	case CLOCK_EVT_MODE_RESUME:
-		jz4740_timer_irq_full_enable(TIMER_CLOCKEVENT);
-		jz4740_timer_enable(TIMER_CLOCKEVENT);
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		jz4740_timer_disable(TIMER_CLOCKEVENT);
-		break;
-	default:
-		break;
-	}
+	jz4740_timer_set_count(TIMER_CLOCKEVENT, 0);
+	jz4740_timer_set_period(TIMER_CLOCKEVENT, jz4740_jiffies_per_tick);
+	jz4740_timer_irq_full_enable(TIMER_CLOCKEVENT);
+	jz4740_timer_enable(TIMER_CLOCKEVENT);
+
+	return 0;
+}
+
+static int jz4740_clockevent_resume(struct clock_event_device *evt)
+{
+	jz4740_timer_irq_full_enable(TIMER_CLOCKEVENT);
+	jz4740_timer_enable(TIMER_CLOCKEVENT);
+
+	return 0;
+}
+
+static int jz4740_clockevent_shutdown(struct clock_event_device *evt)
+{
+	jz4740_timer_disable(TIMER_CLOCKEVENT);
+
+	return 0;
 }
 
 static int jz4740_clockevent_set_next(unsigned long evt,
@@ -100,7 +105,10 @@ static struct clock_event_device jz4740_clockevent = {
 	.name = "jz4740-timer",
 	.features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
 	.set_next_event = jz4740_clockevent_set_next,
-	.set_mode = jz4740_clockevent_set_mode,
+	.set_state_shutdown = jz4740_clockevent_shutdown,
+	.set_state_periodic = jz4740_clockevent_set_periodic,
+	.set_state_oneshot = jz4740_clockevent_shutdown,
+	.tick_resume = jz4740_clockevent_resume,
 	.rating = 200,
 #ifdef CONFIG_MACH_JZ4740
 	.irq = JZ4740_IRQ_TCU0,
diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile
index 3156c8d..d982be1 100644
--- a/arch/mips/kernel/Makefile
+++ b/arch/mips/kernel/Makefile
@@ -99,6 +99,7 @@ obj-$(CONFIG_PERF_EVENTS)	+= perf_event.o
 obj-$(CONFIG_HW_PERF_EVENTS)	+= perf_event_mipsxx.o
 
 obj-$(CONFIG_JUMP_LABEL)	+= jump_label.o
+obj-$(CONFIG_UPROBES)		+= uprobes.o
 
 obj-$(CONFIG_MIPS_CM)		+= mips-cm.o
 obj-$(CONFIG_MIPS_CPC)		+= mips-cpc.o
diff --git a/arch/mips/kernel/asm-offsets.c b/arch/mips/kernel/asm-offsets.c
index 072fab1..154e203 100644
--- a/arch/mips/kernel/asm-offsets.c
+++ b/arch/mips/kernel/asm-offsets.c
@@ -128,6 +128,7 @@ void output_thread_defines(void)
 	       thread.cp0_baduaddr);
 	OFFSET(THREAD_ECODE, task_struct, \
 	       thread.error_code);
+	OFFSET(THREAD_TRAPNO, task_struct, thread.trap_nr);
 	BLANK();
 }
 
@@ -245,17 +246,6 @@ void output_sc_defines(void)
 }
 #endif
 
-#ifdef CONFIG_MIPS32_COMPAT
-void output_sc32_defines(void)
-{
-	COMMENT("Linux 32-bit sigcontext offsets.");
-	OFFSET(SC32_FPREGS, sigcontext32, sc_fpregs);
-	OFFSET(SC32_FPC_CSR, sigcontext32, sc_fpc_csr);
-	OFFSET(SC32_FPC_EIR, sigcontext32, sc_fpc_eir);
-	BLANK();
-}
-#endif
-
 void output_signal_defined(void)
 {
 	COMMENT("Linux signal numbers.");
diff --git a/arch/mips/kernel/cevt-bcm1480.c b/arch/mips/kernel/cevt-bcm1480.c
index 7976457..940ac00 100644
--- a/arch/mips/kernel/cevt-bcm1480.c
+++ b/arch/mips/kernel/cevt-bcm1480.c
@@ -40,8 +40,8 @@
  * The general purpose timer ticks at 1MHz independent if
  * the rest of the system
  */
-static void sibyte_set_mode(enum clock_event_mode mode,
-			   struct clock_event_device *evt)
+
+static int sibyte_set_periodic(struct clock_event_device *evt)
 {
 	unsigned int cpu = smp_processor_id();
 	void __iomem *cfg, *init;
@@ -49,24 +49,22 @@ static void sibyte_set_mode(enum clock_event_mode mode,
 	cfg = IOADDR(A_SCD_TIMER_REGISTER(cpu, R_SCD_TIMER_CFG));
 	init = IOADDR(A_SCD_TIMER_REGISTER(cpu, R_SCD_TIMER_INIT));
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		__raw_writeq(0, cfg);
-		__raw_writeq((V_SCD_TIMER_FREQ / HZ) - 1, init);
-		__raw_writeq(M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS,
-			     cfg);
-		break;
-
-	case CLOCK_EVT_MODE_ONESHOT:
-		/* Stop the timer until we actually program a shot */
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		__raw_writeq(0, cfg);
-		break;
-
-	case CLOCK_EVT_MODE_UNUSED:	/* shuddup gcc */
-	case CLOCK_EVT_MODE_RESUME:
-		;
-	}
+	__raw_writeq(0, cfg);
+	__raw_writeq((V_SCD_TIMER_FREQ / HZ) - 1, init);
+	__raw_writeq(M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS, cfg);
+	return 0;
+}
+
+static int sibyte_shutdown(struct clock_event_device *evt)
+{
+	unsigned int cpu = smp_processor_id();
+	void __iomem *cfg;
+
+	cfg = IOADDR(A_SCD_TIMER_REGISTER(cpu, R_SCD_TIMER_CFG));
+
+	/* Stop the timer until we actually program a shot */
+	__raw_writeq(0, cfg);
+	return 0;
 }
 
 static int sibyte_next_event(unsigned long delta, struct clock_event_device *cd)
@@ -91,7 +89,7 @@ static irqreturn_t sibyte_counter_handler(int irq, void *dev_id)
 	void __iomem *cfg;
 	unsigned long tmode;
 
-	if (cd->mode == CLOCK_EVT_MODE_PERIODIC)
+	if (clockevent_state_periodic(cd))
 		tmode = M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS;
 	else
 		tmode = 0;
@@ -130,7 +128,9 @@ void sb1480_clockevent_init(void)
 	cd->irq			= irq;
 	cd->cpumask		= cpumask_of(cpu);
 	cd->set_next_event	= sibyte_next_event;
-	cd->set_mode		= sibyte_set_mode;
+	cd->set_state_shutdown	= sibyte_shutdown;
+	cd->set_state_periodic	= sibyte_set_periodic;
+	cd->set_state_oneshot	= sibyte_shutdown;
 	clockevents_register_device(cd);
 
 	bcm1480_mask_irq(cpu, irq);
diff --git a/arch/mips/kernel/cevt-ds1287.c b/arch/mips/kernel/cevt-ds1287.c
index ff1f01b..77a5ddf 100644
--- a/arch/mips/kernel/cevt-ds1287.c
+++ b/arch/mips/kernel/cevt-ds1287.c
@@ -59,27 +59,32 @@ static int ds1287_set_next_event(unsigned long delta,
 	return -EINVAL;
 }
 
-static void ds1287_set_mode(enum clock_event_mode mode,
-			    struct clock_event_device *evt)
+static int ds1287_shutdown(struct clock_event_device *evt)
 {
 	u8 val;
 
 	spin_lock(&rtc_lock);
 
 	val = CMOS_READ(RTC_REG_B);
+	val &= ~RTC_PIE;
+	CMOS_WRITE(val, RTC_REG_B);
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		val |= RTC_PIE;
-		break;
-	default:
-		val &= ~RTC_PIE;
-		break;
-	}
+	spin_unlock(&rtc_lock);
+	return 0;
+}
 
+static int ds1287_set_periodic(struct clock_event_device *evt)
+{
+	u8 val;
+
+	spin_lock(&rtc_lock);
+
+	val = CMOS_READ(RTC_REG_B);
+	val |= RTC_PIE;
 	CMOS_WRITE(val, RTC_REG_B);
 
 	spin_unlock(&rtc_lock);
+	return 0;
 }
 
 static void ds1287_event_handler(struct clock_event_device *dev)
@@ -87,11 +92,13 @@ static void ds1287_event_handler(struct clock_event_device *dev)
 }
 
 static struct clock_event_device ds1287_clockevent = {
-	.name		= "ds1287",
-	.features	= CLOCK_EVT_FEAT_PERIODIC,
-	.set_next_event = ds1287_set_next_event,
-	.set_mode	= ds1287_set_mode,
-	.event_handler	= ds1287_event_handler,
+	.name			= "ds1287",
+	.features		= CLOCK_EVT_FEAT_PERIODIC,
+	.set_next_event		= ds1287_set_next_event,
+	.set_state_shutdown	= ds1287_shutdown,
+	.set_state_periodic	= ds1287_set_periodic,
+	.tick_resume		= ds1287_shutdown,
+	.event_handler		= ds1287_event_handler,
 };
 
 static irqreturn_t ds1287_interrupt(int irq, void *dev_id)
diff --git a/arch/mips/kernel/cevt-gt641xx.c b/arch/mips/kernel/cevt-gt641xx.c
index f069460..6604005 100644
--- a/arch/mips/kernel/cevt-gt641xx.c
+++ b/arch/mips/kernel/cevt-gt641xx.c
@@ -64,8 +64,7 @@ static int gt641xx_timer0_set_next_event(unsigned long delta,
 	return 0;
 }
 
-static void gt641xx_timer0_set_mode(enum clock_event_mode mode,
-				    struct clock_event_device *evt)
+static int gt641xx_timer0_shutdown(struct clock_event_device *evt)
 {
 	u32 ctrl;
 
@@ -73,21 +72,39 @@ static void gt641xx_timer0_set_mode(enum clock_event_mode mode,
 
 	ctrl = GT_READ(GT_TC_CONTROL_OFS);
 	ctrl &= ~(GT_TC_CONTROL_ENTC0_MSK | GT_TC_CONTROL_SELTC0_MSK);
+	GT_WRITE(GT_TC_CONTROL_OFS, ctrl);
+
+	raw_spin_unlock(&gt641xx_timer_lock);
+	return 0;
+}
+
+static int gt641xx_timer0_set_oneshot(struct clock_event_device *evt)
+{
+	u32 ctrl;
+
+	raw_spin_lock(&gt641xx_timer_lock);
+
+	ctrl = GT_READ(GT_TC_CONTROL_OFS);
+	ctrl &= ~GT_TC_CONTROL_SELTC0_MSK;
+	ctrl |= GT_TC_CONTROL_ENTC0_MSK;
+	GT_WRITE(GT_TC_CONTROL_OFS, ctrl);
+
+	raw_spin_unlock(&gt641xx_timer_lock);
+	return 0;
+}
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		ctrl |= GT_TC_CONTROL_ENTC0_MSK | GT_TC_CONTROL_SELTC0_MSK;
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-		ctrl |= GT_TC_CONTROL_ENTC0_MSK;
-		break;
-	default:
-		break;
-	}
+static int gt641xx_timer0_set_periodic(struct clock_event_device *evt)
+{
+	u32 ctrl;
+
+	raw_spin_lock(&gt641xx_timer_lock);
 
+	ctrl = GT_READ(GT_TC_CONTROL_OFS);
+	ctrl |= GT_TC_CONTROL_ENTC0_MSK | GT_TC_CONTROL_SELTC0_MSK;
 	GT_WRITE(GT_TC_CONTROL_OFS, ctrl);
 
 	raw_spin_unlock(&gt641xx_timer_lock);
+	return 0;
 }
 
 static void gt641xx_timer0_event_handler(struct clock_event_device *dev)
@@ -95,12 +112,16 @@ static void gt641xx_timer0_event_handler(struct clock_event_device *dev)
 }
 
 static struct clock_event_device gt641xx_timer0_clockevent = {
-	.name		= "gt641xx-timer0",
-	.features	= CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
-	.irq		= GT641XX_TIMER0_IRQ,
-	.set_next_event = gt641xx_timer0_set_next_event,
-	.set_mode	= gt641xx_timer0_set_mode,
-	.event_handler	= gt641xx_timer0_event_handler,
+	.name			= "gt641xx-timer0",
+	.features		= CLOCK_EVT_FEAT_PERIODIC |
+				  CLOCK_EVT_FEAT_ONESHOT,
+	.irq			= GT641XX_TIMER0_IRQ,
+	.set_next_event		= gt641xx_timer0_set_next_event,
+	.set_state_shutdown	= gt641xx_timer0_shutdown,
+	.set_state_periodic	= gt641xx_timer0_set_periodic,
+	.set_state_oneshot	= gt641xx_timer0_set_oneshot,
+	.tick_resume		= gt641xx_timer0_shutdown,
+	.event_handler		= gt641xx_timer0_event_handler,
 };
 
 static irqreturn_t gt641xx_timer0_interrupt(int irq, void *dev_id)
diff --git a/arch/mips/kernel/cevt-r4k.c b/arch/mips/kernel/cevt-r4k.c
index d70c4d8..8dfe6a6 100644
--- a/arch/mips/kernel/cevt-r4k.c
+++ b/arch/mips/kernel/cevt-r4k.c
@@ -28,12 +28,6 @@ static int mips_next_event(unsigned long delta,
 	return res;
 }
 
-void mips_set_clock_mode(enum clock_event_mode mode,
-				struct clock_event_device *evt)
-{
-	/* Nothing to do ...  */
-}
-
 DEFINE_PER_CPU(struct clock_event_device, mips_clockevent_device);
 int cp0_timer_irq_installed;
 
@@ -174,6 +168,11 @@ int c0_compare_int_usable(void)
 	return 1;
 }
 
+unsigned int __weak get_c0_compare_int(void)
+{
+	return MIPS_CPU_IRQ_BASE + cp0_compare_irq;
+}
+
 int r4k_clockevent_init(void)
 {
 	unsigned int cpu = smp_processor_id();
@@ -189,11 +188,9 @@ int r4k_clockevent_init(void)
 	/*
 	 * With vectored interrupts things are getting platform specific.
 	 * get_c0_compare_int is a hook to allow a platform to return the
-	 * interrupt number of it's liking.
+	 * interrupt number of its liking.
 	 */
-	irq = MIPS_CPU_IRQ_BASE + cp0_compare_irq;
-	if (get_c0_compare_int)
-		irq = get_c0_compare_int();
+	irq = get_c0_compare_int();
 
 	cd = &per_cpu(mips_clockevent_device, cpu);
 
@@ -212,7 +209,6 @@ int r4k_clockevent_init(void)
 	cd->irq			= irq;
 	cd->cpumask		= cpumask_of(cpu);
 	cd->set_next_event	= mips_next_event;
-	cd->set_mode		= mips_set_clock_mode;
 	cd->event_handler	= mips_event_handler;
 
 	clockevents_register_device(cd);
diff --git a/arch/mips/kernel/cevt-sb1250.c b/arch/mips/kernel/cevt-sb1250.c
index 5ea6d6b..3d860ef 100644
--- a/arch/mips/kernel/cevt-sb1250.c
+++ b/arch/mips/kernel/cevt-sb1250.c
@@ -38,8 +38,20 @@
  * The general purpose timer ticks at 1MHz independent if
  * the rest of the system
  */
-static void sibyte_set_mode(enum clock_event_mode mode,
-			   struct clock_event_device *evt)
+
+static int sibyte_shutdown(struct clock_event_device *evt)
+{
+	void __iomem *cfg;
+
+	cfg = IOADDR(A_SCD_TIMER_REGISTER(smp_processor_id(), R_SCD_TIMER_CFG));
+
+	/* Stop the timer until we actually program a shot */
+	__raw_writeq(0, cfg);
+
+	return 0;
+}
+
+static int sibyte_set_periodic(struct clock_event_device *evt)
 {
 	unsigned int cpu = smp_processor_id();
 	void __iomem *cfg, *init;
@@ -47,24 +59,11 @@ static void sibyte_set_mode(enum clock_event_mode mode,
 	cfg = IOADDR(A_SCD_TIMER_REGISTER(cpu, R_SCD_TIMER_CFG));
 	init = IOADDR(A_SCD_TIMER_REGISTER(cpu, R_SCD_TIMER_INIT));
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		__raw_writeq(0, cfg);
-		__raw_writeq((V_SCD_TIMER_FREQ / HZ) - 1, init);
-		__raw_writeq(M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS,
-			     cfg);
-		break;
-
-	case CLOCK_EVT_MODE_ONESHOT:
-		/* Stop the timer until we actually program a shot */
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		__raw_writeq(0, cfg);
-		break;
-
-	case CLOCK_EVT_MODE_UNUSED:	/* shuddup gcc */
-	case CLOCK_EVT_MODE_RESUME:
-		;
-	}
+	__raw_writeq(0, cfg);
+	__raw_writeq((V_SCD_TIMER_FREQ / HZ) - 1, init);
+	__raw_writeq(M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS, cfg);
+
+	return 0;
 }
 
 static int sibyte_next_event(unsigned long delta, struct clock_event_device *cd)
@@ -89,7 +88,7 @@ static irqreturn_t sibyte_counter_handler(int irq, void *dev_id)
 	void __iomem *cfg;
 	unsigned long tmode;
 
-	if (cd->mode == CLOCK_EVT_MODE_PERIODIC)
+	if (clockevent_state_periodic(cd))
 		tmode = M_SCD_TIMER_ENABLE | M_SCD_TIMER_MODE_CONTINUOUS;
 	else
 		tmode = 0;
@@ -129,7 +128,9 @@ void sb1250_clockevent_init(void)
 	cd->irq			= irq;
 	cd->cpumask		= cpumask_of(cpu);
 	cd->set_next_event	= sibyte_next_event;
-	cd->set_mode		= sibyte_set_mode;
+	cd->set_state_shutdown	= sibyte_shutdown;
+	cd->set_state_periodic	= sibyte_set_periodic;
+	cd->set_state_oneshot	= sibyte_shutdown;
 	clockevents_register_device(cd);
 
 	sb1250_mask_irq(cpu, irq);
diff --git a/arch/mips/kernel/cevt-txx9.c b/arch/mips/kernel/cevt-txx9.c
index 7239324..537eefd 100644
--- a/arch/mips/kernel/cevt-txx9.c
+++ b/arch/mips/kernel/cevt-txx9.c
@@ -85,36 +85,54 @@ static void txx9tmr_stop_and_clear(struct txx9_tmr_reg __iomem *tmrptr)
 	__raw_writel(0, &tmrptr->tisr);
 }
 
-static void txx9tmr_set_mode(enum clock_event_mode mode,
-			     struct clock_event_device *evt)
+static int txx9tmr_set_state_periodic(struct clock_event_device *evt)
 {
 	struct txx9_clock_event_device *txx9_cd =
 		container_of(evt, struct txx9_clock_event_device, cd);
 	struct txx9_tmr_reg __iomem *tmrptr = txx9_cd->tmrptr;
 
 	txx9tmr_stop_and_clear(tmrptr);
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		__raw_writel(TXx9_TMITMR_TIIE | TXx9_TMITMR_TZCE,
-			     &tmrptr->itmr);
-		/* start timer */
-		__raw_writel(((u64)(NSEC_PER_SEC / HZ) * evt->mult) >>
-			     evt->shift,
-			     &tmrptr->cpra);
-		__raw_writel(TCR_BASE | TXx9_TMTCR_TCE, &tmrptr->tcr);
-		break;
-	case CLOCK_EVT_MODE_SHUTDOWN:
-	case CLOCK_EVT_MODE_UNUSED:
-		__raw_writel(0, &tmrptr->itmr);
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-		__raw_writel(TXx9_TMITMR_TIIE, &tmrptr->itmr);
-		break;
-	case CLOCK_EVT_MODE_RESUME:
-		__raw_writel(TIMER_CCD, &tmrptr->ccdr);
-		__raw_writel(0, &tmrptr->itmr);
-		break;
-	}
+
+	__raw_writel(TXx9_TMITMR_TIIE | TXx9_TMITMR_TZCE, &tmrptr->itmr);
+	/* start timer */
+	__raw_writel(((u64)(NSEC_PER_SEC / HZ) * evt->mult) >> evt->shift,
+		     &tmrptr->cpra);
+	__raw_writel(TCR_BASE | TXx9_TMTCR_TCE, &tmrptr->tcr);
+	return 0;
+}
+
+static int txx9tmr_set_state_oneshot(struct clock_event_device *evt)
+{
+	struct txx9_clock_event_device *txx9_cd =
+		container_of(evt, struct txx9_clock_event_device, cd);
+	struct txx9_tmr_reg __iomem *tmrptr = txx9_cd->tmrptr;
+
+	txx9tmr_stop_and_clear(tmrptr);
+	__raw_writel(TXx9_TMITMR_TIIE, &tmrptr->itmr);
+	return 0;
+}
+
+static int txx9tmr_set_state_shutdown(struct clock_event_device *evt)
+{
+	struct txx9_clock_event_device *txx9_cd =
+		container_of(evt, struct txx9_clock_event_device, cd);
+	struct txx9_tmr_reg __iomem *tmrptr = txx9_cd->tmrptr;
+
+	txx9tmr_stop_and_clear(tmrptr);
+	__raw_writel(0, &tmrptr->itmr);
+	return 0;
+}
+
+static int txx9tmr_tick_resume(struct clock_event_device *evt)
+{
+	struct txx9_clock_event_device *txx9_cd =
+		container_of(evt, struct txx9_clock_event_device, cd);
+	struct txx9_tmr_reg __iomem *tmrptr = txx9_cd->tmrptr;
+
+	txx9tmr_stop_and_clear(tmrptr);
+	__raw_writel(TIMER_CCD, &tmrptr->ccdr);
+	__raw_writel(0, &tmrptr->itmr);
+	return 0;
 }
 
 static int txx9tmr_set_next_event(unsigned long delta,
@@ -133,12 +151,15 @@ static int txx9tmr_set_next_event(unsigned long delta,
 
 static struct txx9_clock_event_device txx9_clock_event_device = {
 	.cd = {
-		.name		= "TXx9",
-		.features	= CLOCK_EVT_FEAT_PERIODIC |
-				  CLOCK_EVT_FEAT_ONESHOT,
-		.rating		= 200,
-		.set_mode	= txx9tmr_set_mode,
-		.set_next_event = txx9tmr_set_next_event,
+		.name			= "TXx9",
+		.features		= CLOCK_EVT_FEAT_PERIODIC |
+					  CLOCK_EVT_FEAT_ONESHOT,
+		.rating			= 200,
+		.set_state_shutdown	= txx9tmr_set_state_shutdown,
+		.set_state_periodic	= txx9tmr_set_state_periodic,
+		.set_state_oneshot	= txx9tmr_set_state_oneshot,
+		.tick_resume		= txx9tmr_tick_resume,
+		.set_next_event		= txx9tmr_set_next_event,
 	},
 };
 
diff --git a/arch/mips/kernel/cps-vec.S b/arch/mips/kernel/cps-vec.S
index 1b6ca63..9f71c06 100644
--- a/arch/mips/kernel/cps-vec.S
+++ b/arch/mips/kernel/cps-vec.S
@@ -152,7 +152,7 @@ dcache_done:
 
 	/* Enter the coherent domain */
 	li	t0, 0xff
-	PTR_S	t0, GCR_CL_COHERENCE_OFS(v1)
+	sw	t0, GCR_CL_COHERENCE_OFS(v1)
 	ehb
 
 	/* Jump to kseg0 */
@@ -302,7 +302,7 @@ LEAF(mips_cps_boot_vpes)
 	PTR_L	t0, 0(t0)
 
 	/* Calculate a pointer to this cores struct core_boot_config */
-	PTR_L	t0, GCR_CL_ID_OFS(t0)
+	lw	t0, GCR_CL_ID_OFS(t0)
 	li	t1, COREBOOTCFG_SIZE
 	mul	t0, t0, t1
 	PTR_LA	t1, mips_cps_core_bootcfg
diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c
index dbe0792..571a8e6 100644
--- a/arch/mips/kernel/cpu-probe.c
+++ b/arch/mips/kernel/cpu-probe.c
@@ -32,6 +32,9 @@
 #include <asm/spram.h>
 #include <asm/uaccess.h>
 
+/* Hardware capabilities */
+unsigned int elf_hwcap __read_mostly;
+
 /*
  * Get the FPU Implementation/Revision.
  */
@@ -188,7 +191,7 @@ __setup("nohtw", htw_disable);
 static int mips_ftlb_disabled;
 static int mips_has_ftlb_configured;
 
-static void set_ftlb_enable(struct cpuinfo_mips *c, int enable);
+static int set_ftlb_enable(struct cpuinfo_mips *c, int enable);
 
 static int __init ftlb_disable(char *s)
 {
@@ -202,7 +205,10 @@ static int __init ftlb_disable(char *s)
 		return 1;
 
 	/* Disable it in the boot cpu */
-	set_ftlb_enable(&cpu_data[0], 0);
+	if (set_ftlb_enable(&cpu_data[0], 0)) {
+		pr_warn("Can't turn FTLB off\n");
+		return 1;
+	}
 
 	back_to_back_c0_hazard();
 
@@ -364,30 +370,41 @@ static unsigned int calculate_ftlb_probability(struct cpuinfo_mips *c)
 		return 3;
 }
 
-static void set_ftlb_enable(struct cpuinfo_mips *c, int enable)
+static int set_ftlb_enable(struct cpuinfo_mips *c, int enable)
 {
-	unsigned int config6;
+	unsigned int config;
 
 	/* It's implementation dependent how the FTLB can be enabled */
 	switch (c->cputype) {
 	case CPU_PROAPTIV:
 	case CPU_P5600:
 		/* proAptiv & related cores use Config6 to enable the FTLB */
-		config6 = read_c0_config6();
+		config = read_c0_config6();
 		/* Clear the old probability value */
-		config6 &= ~(3 << MIPS_CONF6_FTLBP_SHIFT);
+		config &= ~(3 << MIPS_CONF6_FTLBP_SHIFT);
 		if (enable)
 			/* Enable FTLB */
-			write_c0_config6(config6 |
+			write_c0_config6(config |
 					 (calculate_ftlb_probability(c)
 					  << MIPS_CONF6_FTLBP_SHIFT)
 					 | MIPS_CONF6_FTLBEN);
 		else
 			/* Disable FTLB */
-			write_c0_config6(config6 &  ~MIPS_CONF6_FTLBEN);
-		back_to_back_c0_hazard();
+			write_c0_config6(config &  ~MIPS_CONF6_FTLBEN);
 		break;
+	case CPU_I6400:
+		/* I6400 & related cores use Config7 to configure FTLB */
+		config = read_c0_config7();
+		/* Clear the old probability value */
+		config &= ~(3 << MIPS_CONF7_FTLBP_SHIFT);
+		write_c0_config7(config | (calculate_ftlb_probability(c)
+					   << MIPS_CONF7_FTLBP_SHIFT));
+		break;
+	default:
+		return 1;
 	}
+
+	return 0;
 }
 
 static inline unsigned int decode_config0(struct cpuinfo_mips *c)
@@ -524,6 +541,8 @@ static inline unsigned int decode_config3(struct cpuinfo_mips *c)
 	}
 	if (config3 & MIPS_CONF3_CDMM)
 		c->options |= MIPS_CPU_CDMM;
+	if (config3 & MIPS_CONF3_SP)
+		c->options |= MIPS_CPU_SP;
 
 	return config3 & MIPS_CONF_M;
 }
@@ -540,7 +559,16 @@ static inline unsigned int decode_config4(struct cpuinfo_mips *c)
 	if (cpu_has_tlb) {
 		if (((config4 & MIPS_CONF4_IE) >> 29) == 2)
 			c->options |= MIPS_CPU_TLBINV;
-		mmuextdef = config4 & MIPS_CONF4_MMUEXTDEF;
+		/*
+		 * This is a bit ugly. R6 has dropped that field from
+		 * config4 and the only valid configuration is VTLB+FTLB so
+		 * set a good value for mmuextdef for that case.
+		 */
+		if (cpu_has_mips_r6)
+			mmuextdef = MIPS_CONF4_MMUEXTDEF_VTLBSIZEEXT;
+		else
+			mmuextdef = config4 & MIPS_CONF4_MMUEXTDEF;
+
 		switch (mmuextdef) {
 		case MIPS_CONF4_MMUEXTDEF_MMUSIZEEXT:
 			c->tlbsize += (config4 & MIPS_CONF4_MMUSIZEEXT) * 0x40;
@@ -1121,6 +1149,10 @@ static inline void cpu_probe_mips(struct cpuinfo_mips *c, unsigned int cpu)
 		c->cputype = CPU_P5600;
 		__cpu_name[cpu] = "MIPS P5600";
 		break;
+	case PRID_IMP_I6400:
+		c->cputype = CPU_I6400;
+		__cpu_name[cpu] = "MIPS I6400";
+		break;
 	case PRID_IMP_M5150:
 		c->cputype = CPU_M5150;
 		__cpu_name[cpu] = "MIPS M5150";
@@ -1492,10 +1524,14 @@ void cpu_probe(void)
 	else
 		c->srsets = 1;
 
+	if (cpu_has_mips_r6)
+		elf_hwcap |= HWCAP_MIPS_R6;
+
 	if (cpu_has_msa) {
 		c->msa_id = cpu_get_msa_id();
 		WARN(c->msa_id & MSA_IR_WRPF,
 		     "Vector register partitioning unimplemented!");
+		elf_hwcap |= HWCAP_MIPS_MSA;
 	}
 
 	cpu_probe_vmbits(c);
diff --git a/arch/mips/kernel/idle.c b/arch/mips/kernel/idle.c
index e4f62b7..ab1478d 100644
--- a/arch/mips/kernel/idle.c
+++ b/arch/mips/kernel/idle.c
@@ -196,6 +196,7 @@ void __init check_wait(void)
 	case CPU_INTERAPTIV:
 	case CPU_M5150:
 	case CPU_QEMU_GENERIC:
+	case CPU_I6400:
 		cpu_wait = r4k_wait;
 		if (read_c0_config7() & MIPS_CONF7_WII)
 			cpu_wait = r4k_wait_irqoff;
diff --git a/arch/mips/kernel/mips-cm.c b/arch/mips/kernel/mips-cm.c
index 85bbe9b..b8ceee5 100644
--- a/arch/mips/kernel/mips-cm.c
+++ b/arch/mips/kernel/mips-cm.c
@@ -15,11 +15,131 @@
 
 void __iomem *mips_cm_base;
 void __iomem *mips_cm_l2sync_base;
+int mips_cm_is64;
+
+static char *cm2_tr[8] = {
+	"mem",	"gcr",	"gic",	"mmio",
+	"0x04", "cpc", "0x06", "0x07"
+};
+
+/* CM3 Tag ECC transation type */
+static char *cm3_tr[16] = {
+	[0x0] = "ReqNoData",
+	[0x1] = "0x1",
+	[0x2] = "ReqWData",
+	[0x3] = "0x3",
+	[0x4] = "IReqNoResp",
+	[0x5] = "IReqWResp",
+	[0x6] = "IReqNoRespDat",
+	[0x7] = "IReqWRespDat",
+	[0x8] = "RespNoData",
+	[0x9] = "RespDataFol",
+	[0xa] = "RespWData",
+	[0xb] = "RespDataOnly",
+	[0xc] = "IRespNoData",
+	[0xd] = "IRespDataFol",
+	[0xe] = "IRespWData",
+	[0xf] = "IRespDataOnly"
+};
+
+static char *cm2_cmd[32] = {
+	[0x00] = "0x00",
+	[0x01] = "Legacy Write",
+	[0x02] = "Legacy Read",
+	[0x03] = "0x03",
+	[0x04] = "0x04",
+	[0x05] = "0x05",
+	[0x06] = "0x06",
+	[0x07] = "0x07",
+	[0x08] = "Coherent Read Own",
+	[0x09] = "Coherent Read Share",
+	[0x0a] = "Coherent Read Discard",
+	[0x0b] = "Coherent Ready Share Always",
+	[0x0c] = "Coherent Upgrade",
+	[0x0d] = "Coherent Writeback",
+	[0x0e] = "0x0e",
+	[0x0f] = "0x0f",
+	[0x10] = "Coherent Copyback",
+	[0x11] = "Coherent Copyback Invalidate",
+	[0x12] = "Coherent Invalidate",
+	[0x13] = "Coherent Write Invalidate",
+	[0x14] = "Coherent Completion Sync",
+	[0x15] = "0x15",
+	[0x16] = "0x16",
+	[0x17] = "0x17",
+	[0x18] = "0x18",
+	[0x19] = "0x19",
+	[0x1a] = "0x1a",
+	[0x1b] = "0x1b",
+	[0x1c] = "0x1c",
+	[0x1d] = "0x1d",
+	[0x1e] = "0x1e",
+	[0x1f] = "0x1f"
+};
+
+/* CM3 Tag ECC command type */
+static char *cm3_cmd[16] = {
+	[0x0] = "Legacy Read",
+	[0x1] = "Legacy Write",
+	[0x2] = "Coherent Read Own",
+	[0x3] = "Coherent Read Share",
+	[0x4] = "Coherent Read Discard",
+	[0x5] = "Coherent Evicted",
+	[0x6] = "Coherent Upgrade",
+	[0x7] = "Coherent Upgrade for Store Conditional",
+	[0x8] = "Coherent Writeback",
+	[0x9] = "Coherent Write Invalidate",
+	[0xa] = "0xa",
+	[0xb] = "0xb",
+	[0xc] = "0xc",
+	[0xd] = "0xd",
+	[0xe] = "0xe",
+	[0xf] = "0xf"
+};
+
+/* CM3 Tag ECC command group */
+static char *cm3_cmd_group[8] = {
+	[0x0] = "Normal",
+	[0x1] = "Registers",
+	[0x2] = "TLB",
+	[0x3] = "0x3",
+	[0x4] = "L1I",
+	[0x5] = "L1D",
+	[0x6] = "L3",
+	[0x7] = "L2"
+};
+
+static char *cm2_core[8] = {
+	"Invalid/OK",	"Invalid/Data",
+	"Shared/OK",	"Shared/Data",
+	"Modified/OK",	"Modified/Data",
+	"Exclusive/OK", "Exclusive/Data"
+};
+
+static char *cm2_causes[32] = {
+	"None", "GC_WR_ERR", "GC_RD_ERR", "COH_WR_ERR",
+	"COH_RD_ERR", "MMIO_WR_ERR", "MMIO_RD_ERR", "0x07",
+	"0x08", "0x09", "0x0a", "0x0b",
+	"0x0c", "0x0d", "0x0e", "0x0f",
+	"0x10", "0x11", "0x12", "0x13",
+	"0x14", "0x15", "0x16", "INTVN_WR_ERR",
+	"INTVN_RD_ERR", "0x19", "0x1a", "0x1b",
+	"0x1c", "0x1d", "0x1e", "0x1f"
+};
+
+static char *cm3_causes[32] = {
+	"0x0", "MP_CORRECTABLE_ECC_ERR", "MP_REQUEST_DECODE_ERR",
+	"MP_UNCORRECTABLE_ECC_ERR", "MP_PARITY_ERR", "MP_COHERENCE_ERR",
+	"CMBIU_REQUEST_DECODE_ERR", "CMBIU_PARITY_ERR", "CMBIU_AXI_RESP_ERR",
+	"0x9", "RBI_BUS_ERR", "0xb", "0xc", "0xd", "0xe", "0xf", "0x10",
+	"0x11", "0x12", "0x13", "0x14", "0x15", "0x16", "0x17", "0x18",
+	"0x19", "0x1a", "0x1b", "0x1c", "0x1d", "0x1e", "0x1f"
+};
 
 phys_addr_t __mips_cm_phys_base(void)
 {
 	u32 config3 = read_c0_config3();
-	u32 cmgcr;
+	unsigned long cmgcr;
 
 	/* Check the CMGCRBase register is implemented */
 	if (!(config3 & MIPS_CONF3_CMGCR))
@@ -81,6 +201,13 @@ int mips_cm_probe(void)
 	phys_addr_t addr;
 	u32 base_reg;
 
+	/*
+	 * No need to probe again if we have already been
+	 * here before.
+	 */
+	if (mips_cm_base)
+		return 0;
+
 	addr = mips_cm_phys_base();
 	BUG_ON((addr & CM_GCR_BASE_GCRBASE_MSK) != addr);
 	if (!addr)
@@ -117,5 +244,133 @@ int mips_cm_probe(void)
 	/* probe for an L2-only sync region */
 	mips_cm_probe_l2sync();
 
+	/* determine register width for this CM */
+	mips_cm_is64 = config_enabled(CONFIG_64BIT) && (mips_cm_revision() >= CM_REV_CM3);
+
 	return 0;
 }
+
+void mips_cm_error_report(void)
+{
+	unsigned long revision = mips_cm_revision();
+	/*
+	 * CM3 has a 64-bit Error cause register with 0:57 containing the error
+	 * info and 63:58 the error type. For old CMs, everything is contained
+	 * in a single 32-bit register (0:26 and 31:27 respectively). Even
+	 * though the cm_error is u64, we will simply ignore the upper word
+	 * for CM2.
+	 */
+	u64 cm_error = read_gcr_error_cause();
+	int cm_error_cause_sft = CM_GCR_ERROR_CAUSE_ERRTYPE_SHF +
+				 ((revision >= CM_REV_CM3) ? 31 : 0);
+	unsigned long cm_addr = read_gcr_error_addr();
+	unsigned long cm_other = read_gcr_error_mult();
+	int ocause, cause;
+	char buf[256];
+
+	if (!mips_cm_present())
+		return;
+
+	cause = cm_error >> cm_error_cause_sft;
+
+	if (!cause)
+		/* All good */
+		return;
+
+	ocause = cm_other >> CM_GCR_ERROR_MULT_ERR2ND_SHF;
+	if (revision < CM_REV_CM3) { /* CM2 */
+		if (cause < 16) {
+			unsigned long cca_bits = (cm_error >> 15) & 7;
+			unsigned long tr_bits = (cm_error >> 12) & 7;
+			unsigned long cmd_bits = (cm_error >> 7) & 0x1f;
+			unsigned long stag_bits = (cm_error >> 3) & 15;
+			unsigned long sport_bits = (cm_error >> 0) & 7;
+
+			snprintf(buf, sizeof(buf),
+				 "CCA=%lu TR=%s MCmd=%s STag=%lu "
+				 "SPort=%lu\n", cca_bits, cm2_tr[tr_bits],
+				 cm2_cmd[cmd_bits], stag_bits, sport_bits);
+		} else {
+			/* glob state & sresp together */
+			unsigned long c3_bits = (cm_error >> 18) & 7;
+			unsigned long c2_bits = (cm_error >> 15) & 7;
+			unsigned long c1_bits = (cm_error >> 12) & 7;
+			unsigned long c0_bits = (cm_error >> 9) & 7;
+			unsigned long sc_bit = (cm_error >> 8) & 1;
+			unsigned long cmd_bits = (cm_error >> 3) & 0x1f;
+			unsigned long sport_bits = (cm_error >> 0) & 7;
+
+			snprintf(buf, sizeof(buf),
+				 "C3=%s C2=%s C1=%s C0=%s SC=%s "
+				 "MCmd=%s SPort=%lu\n",
+				 cm2_core[c3_bits], cm2_core[c2_bits],
+				 cm2_core[c1_bits], cm2_core[c0_bits],
+				 sc_bit ? "True" : "False",
+				 cm2_cmd[cmd_bits], sport_bits);
+		}
+			pr_err("CM_ERROR=%08llx %s <%s>\n", cm_error,
+			       cm2_causes[cause], buf);
+		pr_err("CM_ADDR =%08lx\n", cm_addr);
+		pr_err("CM_OTHER=%08lx %s\n", cm_other, cm2_causes[ocause]);
+	} else { /* CM3 */
+	/* Used by cause == {1,2,3} */
+		unsigned long core_id_bits = (cm_error >> 22) & 0xf;
+		unsigned long vp_id_bits = (cm_error >> 18) & 0xf;
+		unsigned long cmd_bits = (cm_error >> 14) & 0xf;
+		unsigned long cmd_group_bits = (cm_error >> 11) & 0xf;
+		unsigned long cm3_cca_bits = (cm_error >> 8) & 7;
+		unsigned long mcp_bits = (cm_error >> 5) & 0xf;
+		unsigned long cm3_tr_bits = (cm_error >> 1) & 0xf;
+		unsigned long sched_bit = cm_error & 0x1;
+
+		if (cause == 1 || cause == 3) { /* Tag ECC */
+			unsigned long tag_ecc = (cm_error >> 57) & 0x1;
+			unsigned long tag_way_bits = (cm_error >> 29) & 0xffff;
+			unsigned long dword_bits = (cm_error >> 49) & 0xff;
+			unsigned long data_way_bits = (cm_error >> 45) & 0xf;
+			unsigned long data_sets_bits = (cm_error >> 29) & 0xfff;
+			unsigned long bank_bit = (cm_error >> 28) & 0x1;
+			snprintf(buf, sizeof(buf),
+				 "%s ECC Error: Way=%lu (DWORD=%lu, Sets=%lu)"
+				 "Bank=%lu CoreID=%lu VPID=%lu Command=%s"
+				 "Command Group=%s CCA=%lu MCP=%d"
+				 "Transaction type=%s Scheduler=%lu\n",
+				 tag_ecc ? "TAG" : "DATA",
+				 tag_ecc ? (unsigned long)ffs(tag_way_bits) - 1 :
+				 data_way_bits, bank_bit, dword_bits,
+				 data_sets_bits,
+				 core_id_bits, vp_id_bits,
+				 cm3_cmd[cmd_bits],
+				 cm3_cmd_group[cmd_group_bits],
+				 cm3_cca_bits, 1 << mcp_bits,
+				 cm3_tr[cm3_tr_bits], sched_bit);
+		} else if (cause == 2) {
+			unsigned long data_error_type = (cm_error >> 41) & 0xfff;
+			unsigned long data_decode_cmd = (cm_error >> 37) & 0xf;
+			unsigned long data_decode_group = (cm_error >> 34) & 0x7;
+			unsigned long data_decode_destination_id = (cm_error >> 28) & 0x3f;
+
+			snprintf(buf, sizeof(buf),
+				 "Decode Request Error: Type=%lu, Command=%lu"
+				 "Command Group=%lu Destination ID=%lu"
+				 "CoreID=%lu VPID=%lu Command=%s"
+				 "Command Group=%s CCA=%lu MCP=%d"
+				 "Transaction type=%s Scheduler=%lu\n",
+				 data_error_type, data_decode_cmd,
+				 data_decode_group, data_decode_destination_id,
+				 core_id_bits, vp_id_bits,
+				 cm3_cmd[cmd_bits],
+				 cm3_cmd_group[cmd_group_bits],
+				 cm3_cca_bits, 1 << mcp_bits,
+				 cm3_tr[cm3_tr_bits], sched_bit);
+		}
+
+		pr_err("CM_ERROR=%llx %s <%s>\n", cm_error,
+		       cm3_causes[cause], buf);
+		pr_err("CM_ADDR =%lx\n", cm_addr);
+		pr_err("CM_OTHER=%lx %s\n", cm_other, cm3_causes[ocause]);
+	}
+
+	/* reprime cause register */
+	write_gcr_error_cause(0);
+}
diff --git a/arch/mips/kernel/mips-cpc.c b/arch/mips/kernel/mips-cpc.c
index 1196450..8af4d62 100644
--- a/arch/mips/kernel/mips-cpc.c
+++ b/arch/mips/kernel/mips-cpc.c
@@ -21,9 +21,16 @@ static DEFINE_PER_CPU_ALIGNED(spinlock_t, cpc_core_lock);
 
 static DEFINE_PER_CPU_ALIGNED(unsigned long, cpc_core_lock_flags);
 
-phys_addr_t __weak mips_cpc_phys_base(void)
+/**
+ * mips_cpc_phys_base - retrieve the physical base address of the CPC
+ *
+ * This function returns the physical base address of the Cluster Power
+ * Controller memory mapped registers, or 0 if no Cluster Power Controller
+ * is present.
+ */
+static phys_addr_t mips_cpc_phys_base(void)
 {
-	u32 cpc_base;
+	unsigned long cpc_base;
 
 	if (!mips_cm_present())
 		return 0;
diff --git a/arch/mips/kernel/perf_event_mipsxx.c b/arch/mips/kernel/perf_event_mipsxx.c
index cc1b6fa..d7b8dd4 100644
--- a/arch/mips/kernel/perf_event_mipsxx.c
+++ b/arch/mips/kernel/perf_event_mipsxx.c
@@ -1556,6 +1556,7 @@ static const struct mips_perf_event *mipsxx_pmu_map_raw_event(u64 config)
 #endif
 		break;
 	case CPU_P5600:
+	case CPU_I6400:
 		/* 8-bit event numbers */
 		raw_id = config & 0x1ff;
 		base_id = raw_id & 0xff;
@@ -1717,6 +1718,11 @@ init_hw_perf_events(void)
 		mipspmu.general_event_map = &mipsxxcore_event_map2;
 		mipspmu.cache_event_map = &mipsxxcore_cache_map2;
 		break;
+	case CPU_I6400:
+		mipspmu.name = "mips/I6400";
+		mipspmu.general_event_map = &mipsxxcore_event_map2;
+		mipspmu.cache_event_map = &mipsxxcore_cache_map2;
+		break;
 	case CPU_1004K:
 		mipspmu.name = "mips/1004K";
 		mipspmu.general_event_map = &mipsxxcore_event_map;
diff --git a/arch/mips/kernel/pm-cps.c b/arch/mips/kernel/pm-cps.c
index 0614717..f63a289 100644
--- a/arch/mips/kernel/pm-cps.c
+++ b/arch/mips/kernel/pm-cps.c
@@ -267,6 +267,7 @@ static int __init cps_gen_flush_fsb(u32 **pp, struct uasm_label **pl,
 
 	/* CPUs which do not require the workaround */
 	case CPU_P5600:
+	case CPU_I6400:
 		return 0;
 
 	default:
@@ -671,6 +672,7 @@ static int __init cps_pm_init(void)
 	case CPU_PROAPTIV:
 	case CPU_M5150:
 	case CPU_P5600:
+	case CPU_I6400:
 		stype_intervention = 0x2;
 		stype_memory = 0x3;
 		stype_ordering = 0x10;
diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c
index e933a30..4f0ac78 100644
--- a/arch/mips/kernel/ptrace.c
+++ b/arch/mips/kernel/ptrace.c
@@ -25,6 +25,7 @@
 #include <linux/regset.h>
 #include <linux/smp.h>
 #include <linux/security.h>
+#include <linux/stddef.h>
 #include <linux/tracehook.h>
 #include <linux/audit.h>
 #include <linux/seccomp.h>
@@ -490,6 +491,93 @@ enum mips_regset {
 	REGSET_FPR,
 };
 
+struct pt_regs_offset {
+	const char *name;
+	int offset;
+};
+
+#define REG_OFFSET_NAME(reg, r) {					\
+	.name = #reg,							\
+	.offset = offsetof(struct pt_regs, r)				\
+}
+
+#define REG_OFFSET_END {						\
+	.name = NULL,							\
+	.offset = 0							\
+}
+
+static const struct pt_regs_offset regoffset_table[] = {
+	REG_OFFSET_NAME(r0, regs[0]),
+	REG_OFFSET_NAME(r1, regs[1]),
+	REG_OFFSET_NAME(r2, regs[2]),
+	REG_OFFSET_NAME(r3, regs[3]),
+	REG_OFFSET_NAME(r4, regs[4]),
+	REG_OFFSET_NAME(r5, regs[5]),
+	REG_OFFSET_NAME(r6, regs[6]),
+	REG_OFFSET_NAME(r7, regs[7]),
+	REG_OFFSET_NAME(r8, regs[8]),
+	REG_OFFSET_NAME(r9, regs[9]),
+	REG_OFFSET_NAME(r10, regs[10]),
+	REG_OFFSET_NAME(r11, regs[11]),
+	REG_OFFSET_NAME(r12, regs[12]),
+	REG_OFFSET_NAME(r13, regs[13]),
+	REG_OFFSET_NAME(r14, regs[14]),
+	REG_OFFSET_NAME(r15, regs[15]),
+	REG_OFFSET_NAME(r16, regs[16]),
+	REG_OFFSET_NAME(r17, regs[17]),
+	REG_OFFSET_NAME(r18, regs[18]),
+	REG_OFFSET_NAME(r19, regs[19]),
+	REG_OFFSET_NAME(r20, regs[20]),
+	REG_OFFSET_NAME(r21, regs[21]),
+	REG_OFFSET_NAME(r22, regs[22]),
+	REG_OFFSET_NAME(r23, regs[23]),
+	REG_OFFSET_NAME(r24, regs[24]),
+	REG_OFFSET_NAME(r25, regs[25]),
+	REG_OFFSET_NAME(r26, regs[26]),
+	REG_OFFSET_NAME(r27, regs[27]),
+	REG_OFFSET_NAME(r28, regs[28]),
+	REG_OFFSET_NAME(r29, regs[29]),
+	REG_OFFSET_NAME(r30, regs[30]),
+	REG_OFFSET_NAME(r31, regs[31]),
+	REG_OFFSET_NAME(c0_status, cp0_status),
+	REG_OFFSET_NAME(hi, hi),
+	REG_OFFSET_NAME(lo, lo),
+#ifdef CONFIG_CPU_HAS_SMARTMIPS
+	REG_OFFSET_NAME(acx, acx),
+#endif
+	REG_OFFSET_NAME(c0_badvaddr, cp0_badvaddr),
+	REG_OFFSET_NAME(c0_cause, cp0_cause),
+	REG_OFFSET_NAME(c0_epc, cp0_epc),
+#ifdef CONFIG_MIPS_MT_SMTC
+	REG_OFFSET_NAME(c0_tcstatus, cp0_tcstatus),
+#endif
+#ifdef CONFIG_CPU_CAVIUM_OCTEON
+	REG_OFFSET_NAME(mpl0, mpl[0]),
+	REG_OFFSET_NAME(mpl1, mpl[1]),
+	REG_OFFSET_NAME(mpl2, mpl[2]),
+	REG_OFFSET_NAME(mtp0, mtp[0]),
+	REG_OFFSET_NAME(mtp1, mtp[1]),
+	REG_OFFSET_NAME(mtp2, mtp[2]),
+#endif
+	REG_OFFSET_END,
+};
+
+/**
+ * regs_query_register_offset() - query register offset from its name
+ * @name:       the name of a register
+ *
+ * regs_query_register_offset() returns the offset of a register in struct
+ * pt_regs from its name. If the name is invalid, this returns -EINVAL;
+ */
+int regs_query_register_offset(const char *name)
+{
+        const struct pt_regs_offset *roff;
+        for (roff = regoffset_table; roff->name != NULL; roff++)
+                if (!strcmp(roff->name, name))
+                        return roff->offset;
+        return -EINVAL;
+}
+
 #if defined(CONFIG_32BIT) || defined(CONFIG_MIPS32_O32)
 
 static const struct user_regset mips_regsets[] = {
diff --git a/arch/mips/kernel/r4k_fpu.S b/arch/mips/kernel/r4k_fpu.S
index 1d88af2..f09546e 100644
--- a/arch/mips/kernel/r4k_fpu.S
+++ b/arch/mips/kernel/r4k_fpu.S
@@ -13,6 +13,7 @@
  * Copyright (C) 1999, 2001 Silicon Graphics, Inc.
  */
 #include <asm/asm.h>
+#include <asm/asmmacro.h>
 #include <asm/errno.h>
 #include <asm/fpregdef.h>
 #include <asm/mipsregs.h>
@@ -35,6 +36,14 @@
 
 	.set	noreorder
 
+/**
+ * _save_fp_context() - save FP context from the FPU
+ * @a0 - pointer to fpregs field of sigcontext
+ * @a1 - pointer to fpc_csr field of sigcontext
+ *
+ * Save FP context, including the 32 FP data registers and the FP
+ * control & status register, from the FPU to signal context.
+ */
 LEAF(_save_fp_context)
 	.set	push
 	SET_HARDFLOAT
@@ -54,117 +63,60 @@ LEAF(_save_fp_context)
 	 nop
 #endif
 	/* Store the 16 odd double precision registers */
-	EX	sdc1 $f1, SC_FPREGS+8(a0)
-	EX	sdc1 $f3, SC_FPREGS+24(a0)
-	EX	sdc1 $f5, SC_FPREGS+40(a0)
-	EX	sdc1 $f7, SC_FPREGS+56(a0)
-	EX	sdc1 $f9, SC_FPREGS+72(a0)
-	EX	sdc1 $f11, SC_FPREGS+88(a0)
-	EX	sdc1 $f13, SC_FPREGS+104(a0)
-	EX	sdc1 $f15, SC_FPREGS+120(a0)
-	EX	sdc1 $f17, SC_FPREGS+136(a0)
-	EX	sdc1 $f19, SC_FPREGS+152(a0)
-	EX	sdc1 $f21, SC_FPREGS+168(a0)
-	EX	sdc1 $f23, SC_FPREGS+184(a0)
-	EX	sdc1 $f25, SC_FPREGS+200(a0)
-	EX	sdc1 $f27, SC_FPREGS+216(a0)
-	EX	sdc1 $f29, SC_FPREGS+232(a0)
-	EX	sdc1 $f31, SC_FPREGS+248(a0)
+	EX	sdc1 $f1, 8(a0)
+	EX	sdc1 $f3, 24(a0)
+	EX	sdc1 $f5, 40(a0)
+	EX	sdc1 $f7, 56(a0)
+	EX	sdc1 $f9, 72(a0)
+	EX	sdc1 $f11, 88(a0)
+	EX	sdc1 $f13, 104(a0)
+	EX	sdc1 $f15, 120(a0)
+	EX	sdc1 $f17, 136(a0)
+	EX	sdc1 $f19, 152(a0)
+	EX	sdc1 $f21, 168(a0)
+	EX	sdc1 $f23, 184(a0)
+	EX	sdc1 $f25, 200(a0)
+	EX	sdc1 $f27, 216(a0)
+	EX	sdc1 $f29, 232(a0)
+	EX	sdc1 $f31, 248(a0)
 1:	.set	pop
 #endif
 
 	.set push
 	SET_HARDFLOAT
 	/* Store the 16 even double precision registers */
-	EX	sdc1 $f0, SC_FPREGS+0(a0)
-	EX	sdc1 $f2, SC_FPREGS+16(a0)
-	EX	sdc1 $f4, SC_FPREGS+32(a0)
-	EX	sdc1 $f6, SC_FPREGS+48(a0)
-	EX	sdc1 $f8, SC_FPREGS+64(a0)
-	EX	sdc1 $f10, SC_FPREGS+80(a0)
-	EX	sdc1 $f12, SC_FPREGS+96(a0)
-	EX	sdc1 $f14, SC_FPREGS+112(a0)
-	EX	sdc1 $f16, SC_FPREGS+128(a0)
-	EX	sdc1 $f18, SC_FPREGS+144(a0)
-	EX	sdc1 $f20, SC_FPREGS+160(a0)
-	EX	sdc1 $f22, SC_FPREGS+176(a0)
-	EX	sdc1 $f24, SC_FPREGS+192(a0)
-	EX	sdc1 $f26, SC_FPREGS+208(a0)
-	EX	sdc1 $f28, SC_FPREGS+224(a0)
-	EX	sdc1 $f30, SC_FPREGS+240(a0)
-	EX	sw t1, SC_FPC_CSR(a0)
+	EX	sdc1 $f0, 0(a0)
+	EX	sdc1 $f2, 16(a0)
+	EX	sdc1 $f4, 32(a0)
+	EX	sdc1 $f6, 48(a0)
+	EX	sdc1 $f8, 64(a0)
+	EX	sdc1 $f10, 80(a0)
+	EX	sdc1 $f12, 96(a0)
+	EX	sdc1 $f14, 112(a0)
+	EX	sdc1 $f16, 128(a0)
+	EX	sdc1 $f18, 144(a0)
+	EX	sdc1 $f20, 160(a0)
+	EX	sdc1 $f22, 176(a0)
+	EX	sdc1 $f24, 192(a0)
+	EX	sdc1 $f26, 208(a0)
+	EX	sdc1 $f28, 224(a0)
+	EX	sdc1 $f30, 240(a0)
+	EX	sw t1, 0(a1)
 	jr	ra
 	 li	v0, 0					# success
 	.set pop
 	END(_save_fp_context)
 
-#ifdef CONFIG_MIPS32_COMPAT
-	/* Save 32-bit process floating point context */
-LEAF(_save_fp_context32)
-	.set push
-	.set MIPS_ISA_ARCH_LEVEL_RAW
-	SET_HARDFLOAT
-	cfc1	t1, fcr31
-
-#ifndef CONFIG_CPU_MIPS64_R6
-	mfc0	t0, CP0_STATUS
-	sll	t0, t0, 5
-	bgez	t0, 1f			# skip storing odd if FR=0
-	 nop
-#endif
-
-	/* Store the 16 odd double precision registers */
-	EX      sdc1 $f1, SC32_FPREGS+8(a0)
-	EX      sdc1 $f3, SC32_FPREGS+24(a0)
-	EX      sdc1 $f5, SC32_FPREGS+40(a0)
-	EX      sdc1 $f7, SC32_FPREGS+56(a0)
-	EX      sdc1 $f9, SC32_FPREGS+72(a0)
-	EX      sdc1 $f11, SC32_FPREGS+88(a0)
-	EX      sdc1 $f13, SC32_FPREGS+104(a0)
-	EX      sdc1 $f15, SC32_FPREGS+120(a0)
-	EX      sdc1 $f17, SC32_FPREGS+136(a0)
-	EX      sdc1 $f19, SC32_FPREGS+152(a0)
-	EX      sdc1 $f21, SC32_FPREGS+168(a0)
-	EX      sdc1 $f23, SC32_FPREGS+184(a0)
-	EX      sdc1 $f25, SC32_FPREGS+200(a0)
-	EX      sdc1 $f27, SC32_FPREGS+216(a0)
-	EX      sdc1 $f29, SC32_FPREGS+232(a0)
-	EX      sdc1 $f31, SC32_FPREGS+248(a0)
-
-	/* Store the 16 even double precision registers */
-1:	EX	sdc1 $f0, SC32_FPREGS+0(a0)
-	EX	sdc1 $f2, SC32_FPREGS+16(a0)
-	EX	sdc1 $f4, SC32_FPREGS+32(a0)
-	EX	sdc1 $f6, SC32_FPREGS+48(a0)
-	EX	sdc1 $f8, SC32_FPREGS+64(a0)
-	EX	sdc1 $f10, SC32_FPREGS+80(a0)
-	EX	sdc1 $f12, SC32_FPREGS+96(a0)
-	EX	sdc1 $f14, SC32_FPREGS+112(a0)
-	EX	sdc1 $f16, SC32_FPREGS+128(a0)
-	EX	sdc1 $f18, SC32_FPREGS+144(a0)
-	EX	sdc1 $f20, SC32_FPREGS+160(a0)
-	EX	sdc1 $f22, SC32_FPREGS+176(a0)
-	EX	sdc1 $f24, SC32_FPREGS+192(a0)
-	EX	sdc1 $f26, SC32_FPREGS+208(a0)
-	EX	sdc1 $f28, SC32_FPREGS+224(a0)
-	EX	sdc1 $f30, SC32_FPREGS+240(a0)
-	EX	sw t1, SC32_FPC_CSR(a0)
-	cfc1	t0, $0				# implementation/version
-	EX	sw t0, SC32_FPC_EIR(a0)
-	.set pop
-
-	jr	ra
-	 li	v0, 0					# success
-	END(_save_fp_context32)
-#endif
-
-/*
- * Restore FPU state:
- *  - fp gp registers
- *  - cp1 status/control register
+/**
+ * _restore_fp_context() - restore FP context to the FPU
+ * @a0 - pointer to fpregs field of sigcontext
+ * @a1 - pointer to fpc_csr field of sigcontext
+ *
+ * Restore FP context, including the 32 FP data registers and the FP
+ * control & status register, from signal context to the FPU.
  */
 LEAF(_restore_fp_context)
-	EX	lw t1, SC_FPC_CSR(a0)
+	EX	lw t1, 0(a1)
 
 #if defined(CONFIG_64BIT) || defined(CONFIG_CPU_MIPS32_R2)  || \
 		defined(CONFIG_CPU_MIPS32_R6)
@@ -178,101 +130,231 @@ LEAF(_restore_fp_context)
 	bgez	t0, 1f			# skip loading odd if FR=0
 	 nop
 #endif
-	EX	ldc1 $f1, SC_FPREGS+8(a0)
-	EX	ldc1 $f3, SC_FPREGS+24(a0)
-	EX	ldc1 $f5, SC_FPREGS+40(a0)
-	EX	ldc1 $f7, SC_FPREGS+56(a0)
-	EX	ldc1 $f9, SC_FPREGS+72(a0)
-	EX	ldc1 $f11, SC_FPREGS+88(a0)
-	EX	ldc1 $f13, SC_FPREGS+104(a0)
-	EX	ldc1 $f15, SC_FPREGS+120(a0)
-	EX	ldc1 $f17, SC_FPREGS+136(a0)
-	EX	ldc1 $f19, SC_FPREGS+152(a0)
-	EX	ldc1 $f21, SC_FPREGS+168(a0)
-	EX	ldc1 $f23, SC_FPREGS+184(a0)
-	EX	ldc1 $f25, SC_FPREGS+200(a0)
-	EX	ldc1 $f27, SC_FPREGS+216(a0)
-	EX	ldc1 $f29, SC_FPREGS+232(a0)
-	EX	ldc1 $f31, SC_FPREGS+248(a0)
+	EX	ldc1 $f1, 8(a0)
+	EX	ldc1 $f3, 24(a0)
+	EX	ldc1 $f5, 40(a0)
+	EX	ldc1 $f7, 56(a0)
+	EX	ldc1 $f9, 72(a0)
+	EX	ldc1 $f11, 88(a0)
+	EX	ldc1 $f13, 104(a0)
+	EX	ldc1 $f15, 120(a0)
+	EX	ldc1 $f17, 136(a0)
+	EX	ldc1 $f19, 152(a0)
+	EX	ldc1 $f21, 168(a0)
+	EX	ldc1 $f23, 184(a0)
+	EX	ldc1 $f25, 200(a0)
+	EX	ldc1 $f27, 216(a0)
+	EX	ldc1 $f29, 232(a0)
+	EX	ldc1 $f31, 248(a0)
 1:	.set pop
 #endif
 	.set push
 	SET_HARDFLOAT
-	EX	ldc1 $f0, SC_FPREGS+0(a0)
-	EX	ldc1 $f2, SC_FPREGS+16(a0)
-	EX	ldc1 $f4, SC_FPREGS+32(a0)
-	EX	ldc1 $f6, SC_FPREGS+48(a0)
-	EX	ldc1 $f8, SC_FPREGS+64(a0)
-	EX	ldc1 $f10, SC_FPREGS+80(a0)
-	EX	ldc1 $f12, SC_FPREGS+96(a0)
-	EX	ldc1 $f14, SC_FPREGS+112(a0)
-	EX	ldc1 $f16, SC_FPREGS+128(a0)
-	EX	ldc1 $f18, SC_FPREGS+144(a0)
-	EX	ldc1 $f20, SC_FPREGS+160(a0)
-	EX	ldc1 $f22, SC_FPREGS+176(a0)
-	EX	ldc1 $f24, SC_FPREGS+192(a0)
-	EX	ldc1 $f26, SC_FPREGS+208(a0)
-	EX	ldc1 $f28, SC_FPREGS+224(a0)
-	EX	ldc1 $f30, SC_FPREGS+240(a0)
+	EX	ldc1 $f0, 0(a0)
+	EX	ldc1 $f2, 16(a0)
+	EX	ldc1 $f4, 32(a0)
+	EX	ldc1 $f6, 48(a0)
+	EX	ldc1 $f8, 64(a0)
+	EX	ldc1 $f10, 80(a0)
+	EX	ldc1 $f12, 96(a0)
+	EX	ldc1 $f14, 112(a0)
+	EX	ldc1 $f16, 128(a0)
+	EX	ldc1 $f18, 144(a0)
+	EX	ldc1 $f20, 160(a0)
+	EX	ldc1 $f22, 176(a0)
+	EX	ldc1 $f24, 192(a0)
+	EX	ldc1 $f26, 208(a0)
+	EX	ldc1 $f28, 224(a0)
+	EX	ldc1 $f30, 240(a0)
 	ctc1	t1, fcr31
 	.set pop
 	jr	ra
 	 li	v0, 0					# success
 	END(_restore_fp_context)
 
-#ifdef CONFIG_MIPS32_COMPAT
-LEAF(_restore_fp_context32)
-	/* Restore an o32 sigcontext.  */
-	.set push
-	SET_HARDFLOAT
-	EX	lw t1, SC32_FPC_CSR(a0)
+#ifdef CONFIG_CPU_HAS_MSA
 
-#ifndef CONFIG_CPU_MIPS64_R6
-	mfc0	t0, CP0_STATUS
-	sll	t0, t0, 5
-	bgez	t0, 1f			# skip loading odd if FR=0
+	.macro	op_one_wr	op, idx, base
+	.align	4
+\idx:	\op	\idx, 0, \base
+	jr	ra
 	 nop
-#endif
+	.endm
 
-	EX      ldc1 $f1, SC32_FPREGS+8(a0)
-	EX      ldc1 $f3, SC32_FPREGS+24(a0)
-	EX      ldc1 $f5, SC32_FPREGS+40(a0)
-	EX      ldc1 $f7, SC32_FPREGS+56(a0)
-	EX      ldc1 $f9, SC32_FPREGS+72(a0)
-	EX      ldc1 $f11, SC32_FPREGS+88(a0)
-	EX      ldc1 $f13, SC32_FPREGS+104(a0)
-	EX      ldc1 $f15, SC32_FPREGS+120(a0)
-	EX      ldc1 $f17, SC32_FPREGS+136(a0)
-	EX      ldc1 $f19, SC32_FPREGS+152(a0)
-	EX      ldc1 $f21, SC32_FPREGS+168(a0)
-	EX      ldc1 $f23, SC32_FPREGS+184(a0)
-	EX      ldc1 $f25, SC32_FPREGS+200(a0)
-	EX      ldc1 $f27, SC32_FPREGS+216(a0)
-	EX      ldc1 $f29, SC32_FPREGS+232(a0)
-	EX      ldc1 $f31, SC32_FPREGS+248(a0)
+	.macro	op_msa_wr	name, op
+LEAF(\name)
+	.set		push
+	.set		noreorder
+	sll		t0, a0, 4
+	PTR_LA		t1, 0f
+	PTR_ADDU	t0, t0, t1
+	jr		t0
+	  nop
+	op_one_wr	\op, 0, a1
+	op_one_wr	\op, 1, a1
+	op_one_wr	\op, 2, a1
+	op_one_wr	\op, 3, a1
+	op_one_wr	\op, 4, a1
+	op_one_wr	\op, 5, a1
+	op_one_wr	\op, 6, a1
+	op_one_wr	\op, 7, a1
+	op_one_wr	\op, 8, a1
+	op_one_wr	\op, 9, a1
+	op_one_wr	\op, 10, a1
+	op_one_wr	\op, 11, a1
+	op_one_wr	\op, 12, a1
+	op_one_wr	\op, 13, a1
+	op_one_wr	\op, 14, a1
+	op_one_wr	\op, 15, a1
+	op_one_wr	\op, 16, a1
+	op_one_wr	\op, 17, a1
+	op_one_wr	\op, 18, a1
+	op_one_wr	\op, 19, a1
+	op_one_wr	\op, 20, a1
+	op_one_wr	\op, 21, a1
+	op_one_wr	\op, 22, a1
+	op_one_wr	\op, 23, a1
+	op_one_wr	\op, 24, a1
+	op_one_wr	\op, 25, a1
+	op_one_wr	\op, 26, a1
+	op_one_wr	\op, 27, a1
+	op_one_wr	\op, 28, a1
+	op_one_wr	\op, 29, a1
+	op_one_wr	\op, 30, a1
+	op_one_wr	\op, 31, a1
+	.set		pop
+	END(\name)
+	.endm
 
-1:	EX	ldc1 $f0, SC32_FPREGS+0(a0)
-	EX	ldc1 $f2, SC32_FPREGS+16(a0)
-	EX	ldc1 $f4, SC32_FPREGS+32(a0)
-	EX	ldc1 $f6, SC32_FPREGS+48(a0)
-	EX	ldc1 $f8, SC32_FPREGS+64(a0)
-	EX	ldc1 $f10, SC32_FPREGS+80(a0)
-	EX	ldc1 $f12, SC32_FPREGS+96(a0)
-	EX	ldc1 $f14, SC32_FPREGS+112(a0)
-	EX	ldc1 $f16, SC32_FPREGS+128(a0)
-	EX	ldc1 $f18, SC32_FPREGS+144(a0)
-	EX	ldc1 $f20, SC32_FPREGS+160(a0)
-	EX	ldc1 $f22, SC32_FPREGS+176(a0)
-	EX	ldc1 $f24, SC32_FPREGS+192(a0)
-	EX	ldc1 $f26, SC32_FPREGS+208(a0)
-	EX	ldc1 $f28, SC32_FPREGS+224(a0)
-	EX	ldc1 $f30, SC32_FPREGS+240(a0)
-	ctc1	t1, fcr31
+	op_msa_wr	read_msa_wr_b, st_b
+	op_msa_wr	read_msa_wr_h, st_h
+	op_msa_wr	read_msa_wr_w, st_w
+	op_msa_wr	read_msa_wr_d, st_d
+
+	op_msa_wr	write_msa_wr_b, ld_b
+	op_msa_wr	write_msa_wr_h, ld_h
+	op_msa_wr	write_msa_wr_w, ld_w
+	op_msa_wr	write_msa_wr_d, ld_d
+
+#endif /* CONFIG_CPU_HAS_MSA */
+
+#ifdef CONFIG_CPU_HAS_MSA
+
+	.macro	save_msa_upper	wr, off, base
+	.set	push
+	.set	noat
+#ifdef CONFIG_64BIT
+	copy_u_d \wr, 1
+	EX sd	$1, \off(\base)
+#elif defined(CONFIG_CPU_LITTLE_ENDIAN)
+	copy_u_w \wr, 2
+	EX sw	$1, \off(\base)
+	copy_u_w \wr, 3
+	EX sw	$1, (\off+4)(\base)
+#else /* CONFIG_CPU_BIG_ENDIAN */
+	copy_u_w \wr, 2
+	EX sw	$1, (\off+4)(\base)
+	copy_u_w \wr, 3
+	EX sw	$1, \off(\base)
+#endif
+	.set	pop
+	.endm
+
+LEAF(_save_msa_all_upper)
+	save_msa_upper	0, 0x00, a0
+	save_msa_upper	1, 0x08, a0
+	save_msa_upper	2, 0x10, a0
+	save_msa_upper	3, 0x18, a0
+	save_msa_upper	4, 0x20, a0
+	save_msa_upper	5, 0x28, a0
+	save_msa_upper	6, 0x30, a0
+	save_msa_upper	7, 0x38, a0
+	save_msa_upper	8, 0x40, a0
+	save_msa_upper	9, 0x48, a0
+	save_msa_upper	10, 0x50, a0
+	save_msa_upper	11, 0x58, a0
+	save_msa_upper	12, 0x60, a0
+	save_msa_upper	13, 0x68, a0
+	save_msa_upper	14, 0x70, a0
+	save_msa_upper	15, 0x78, a0
+	save_msa_upper	16, 0x80, a0
+	save_msa_upper	17, 0x88, a0
+	save_msa_upper	18, 0x90, a0
+	save_msa_upper	19, 0x98, a0
+	save_msa_upper	20, 0xa0, a0
+	save_msa_upper	21, 0xa8, a0
+	save_msa_upper	22, 0xb0, a0
+	save_msa_upper	23, 0xb8, a0
+	save_msa_upper	24, 0xc0, a0
+	save_msa_upper	25, 0xc8, a0
+	save_msa_upper	26, 0xd0, a0
+	save_msa_upper	27, 0xd8, a0
+	save_msa_upper	28, 0xe0, a0
+	save_msa_upper	29, 0xe8, a0
+	save_msa_upper	30, 0xf0, a0
+	save_msa_upper	31, 0xf8, a0
 	jr	ra
-	 li	v0, 0					# success
-	.set pop
-	END(_restore_fp_context32)
+	 li	v0, 0
+	END(_save_msa_all_upper)
+
+	.macro	restore_msa_upper	wr, off, base
+	.set	push
+	.set	noat
+#ifdef CONFIG_64BIT
+	EX ld	$1, \off(\base)
+	insert_d \wr, 1
+#elif defined(CONFIG_CPU_LITTLE_ENDIAN)
+	EX lw	$1, \off(\base)
+	insert_w \wr, 2
+	EX lw	$1, (\off+4)(\base)
+	insert_w \wr, 3
+#else /* CONFIG_CPU_BIG_ENDIAN */
+	EX lw	$1, (\off+4)(\base)
+	insert_w \wr, 2
+	EX lw	$1, \off(\base)
+	insert_w \wr, 3
 #endif
+	.set	pop
+	.endm
+
+LEAF(_restore_msa_all_upper)
+	restore_msa_upper	0, 0x00, a0
+	restore_msa_upper	1, 0x08, a0
+	restore_msa_upper	2, 0x10, a0
+	restore_msa_upper	3, 0x18, a0
+	restore_msa_upper	4, 0x20, a0
+	restore_msa_upper	5, 0x28, a0
+	restore_msa_upper	6, 0x30, a0
+	restore_msa_upper	7, 0x38, a0
+	restore_msa_upper	8, 0x40, a0
+	restore_msa_upper	9, 0x48, a0
+	restore_msa_upper	10, 0x50, a0
+	restore_msa_upper	11, 0x58, a0
+	restore_msa_upper	12, 0x60, a0
+	restore_msa_upper	13, 0x68, a0
+	restore_msa_upper	14, 0x70, a0
+	restore_msa_upper	15, 0x78, a0
+	restore_msa_upper	16, 0x80, a0
+	restore_msa_upper	17, 0x88, a0
+	restore_msa_upper	18, 0x90, a0
+	restore_msa_upper	19, 0x98, a0
+	restore_msa_upper	20, 0xa0, a0
+	restore_msa_upper	21, 0xa8, a0
+	restore_msa_upper	22, 0xb0, a0
+	restore_msa_upper	23, 0xb8, a0
+	restore_msa_upper	24, 0xc0, a0
+	restore_msa_upper	25, 0xc8, a0
+	restore_msa_upper	26, 0xd0, a0
+	restore_msa_upper	27, 0xd8, a0
+	restore_msa_upper	28, 0xe0, a0
+	restore_msa_upper	29, 0xe8, a0
+	restore_msa_upper	30, 0xf0, a0
+	restore_msa_upper	31, 0xf8, a0
+	jr	ra
+	 li	v0, 0
+	END(_restore_msa_all_upper)
+
+#endif /* CONFIG_CPU_HAS_MSA */
 
 	.set	reorder
 
diff --git a/arch/mips/kernel/r4k_switch.S b/arch/mips/kernel/r4k_switch.S
index 04cbbde..92cd051 100644
--- a/arch/mips/kernel/r4k_switch.S
+++ b/arch/mips/kernel/r4k_switch.S
@@ -34,7 +34,7 @@
 #ifndef USE_ALTERNATE_RESUME_IMPL
 /*
  * task_struct *resume(task_struct *prev, task_struct *next,
- *		       struct thread_info *next_ti, s32 fp_save)
+ *		       struct thread_info *next_ti)
  */
 	.align	5
 	LEAF(resume)
@@ -43,45 +43,6 @@
 	cpu_save_nonscratch a0
 	LONG_S	ra, THREAD_REG31(a0)
 
-	/*
-	 * Check whether we need to save any FP context. FP context is saved
-	 * iff the process has used the context with the scalar FPU or the MSA
-	 * ASE in the current time slice, as indicated by _TIF_USEDFPU and
-	 * _TIF_USEDMSA respectively. switch_to will have set fp_save
-	 * accordingly to an FP_SAVE_ enum value.
-	 */
-	beqz	a3, 2f
-
-	/*
-	 * We do. Clear the saved CU1 bit for prev, such that next time it is
-	 * scheduled it will start in userland with the FPU disabled. If the
-	 * task uses the FPU then it will be enabled again via the do_cpu trap.
-	 * This allows us to lazily restore the FP context.
-	 */
-	PTR_L	t3, TASK_THREAD_INFO(a0)
-	LONG_L	t0, ST_OFF(t3)
-	li	t1, ~ST0_CU1
-	and	t0, t0, t1
-	LONG_S	t0, ST_OFF(t3)
-
-	/* Check whether we're saving scalar or vector context. */
-	bgtz	a3, 1f
-
-	/* Save 128b MSA vector context + scalar FP control & status. */
-	.set push
-	SET_HARDFLOAT
-	cfc1	t1, fcr31
-	msa_save_all	a0
-	.set pop	/* SET_HARDFLOAT */
-
-	sw	t1, THREAD_FCR31(a0)
-	b	2f
-
-1:	/* Save 32b/64b scalar FP context. */
-	fpu_save_double a0 t0 t1		# c0_status passed in t0
-						# clobbers t1
-2:
-
 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
 	PTR_LA	t8, __stack_chk_guard
 	LONG_L	t9, TASK_STACK_CANARY(a1)
diff --git a/arch/mips/kernel/signal-common.h b/arch/mips/kernel/signal-common.h
index 0b85f82..f50d484 100644
--- a/arch/mips/kernel/signal-common.h
+++ b/arch/mips/kernel/signal-common.h
@@ -31,4 +31,13 @@ extern int fpcsr_pending(unsigned int __user *fpcsr);
 #define lock_fpu_owner()	({ preempt_disable(); pagefault_disable(); })
 #define unlock_fpu_owner()	({ pagefault_enable(); preempt_enable(); })
 
+/* Assembly functions to move context to/from the FPU */
+extern asmlinkage int
+_save_fp_context(void __user *fpregs, void __user *csr);
+extern asmlinkage int
+_restore_fp_context(void __user *fpregs, void __user *csr);
+
+extern asmlinkage int _save_msa_all_upper(void __user *buf);
+extern asmlinkage int _restore_msa_all_upper(void __user *buf);
+
 #endif	/* __SIGNAL_COMMON_H */
diff --git a/arch/mips/kernel/signal.c b/arch/mips/kernel/signal.c
index 6a28c79..2fec67b 100644
--- a/arch/mips/kernel/signal.c
+++ b/arch/mips/kernel/signal.c
@@ -21,6 +21,7 @@
 #include <linux/wait.h>
 #include <linux/ptrace.h>
 #include <linux/unistd.h>
+#include <linux/uprobes.h>
 #include <linux/compiler.h>
 #include <linux/syscalls.h>
 #include <linux/uaccess.h>
@@ -38,20 +39,21 @@
 #include <asm/vdso.h>
 #include <asm/dsp.h>
 #include <asm/inst.h>
+#include <asm/msa.h>
 
 #include "signal-common.h"
 
-static int (*save_fp_context)(struct sigcontext __user *sc);
-static int (*restore_fp_context)(struct sigcontext __user *sc);
-
-extern asmlinkage int _save_fp_context(struct sigcontext __user *sc);
-extern asmlinkage int _restore_fp_context(struct sigcontext __user *sc);
+static int (*save_fp_context)(void __user *sc);
+static int (*restore_fp_context)(void __user *sc);
 
 struct sigframe {
 	u32 sf_ass[4];		/* argument save space for o32 */
 	u32 sf_pad[2];		/* Was: signal trampoline */
+
+	/* Matches struct ucontext from its uc_mcontext field onwards */
 	struct sigcontext sf_sc;
 	sigset_t sf_mask;
+	unsigned long long sf_extcontext[0];
 };
 
 struct rt_sigframe {
@@ -65,43 +67,255 @@ struct rt_sigframe {
  * Thread saved context copy to/from a signal context presumed to be on the
  * user stack, and therefore accessed with appropriate macros from uaccess.h.
  */
-static int copy_fp_to_sigcontext(struct sigcontext __user *sc)
+static int copy_fp_to_sigcontext(void __user *sc)
 {
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
 	int i;
 	int err = 0;
+	int inc = test_thread_flag(TIF_32BIT_FPREGS) ? 2 : 1;
 
-	for (i = 0; i < NUM_FPU_REGS; i++) {
+	for (i = 0; i < NUM_FPU_REGS; i += inc) {
 		err |=
 		    __put_user(get_fpr64(&current->thread.fpu.fpr[i], 0),
-			       &sc->sc_fpregs[i]);
+			       &fpregs[i]);
 	}
-	err |= __put_user(current->thread.fpu.fcr31, &sc->sc_fpc_csr);
+	err |= __put_user(current->thread.fpu.fcr31, csr);
 
 	return err;
 }
 
-static int copy_fp_from_sigcontext(struct sigcontext __user *sc)
+static int copy_fp_from_sigcontext(void __user *sc)
 {
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
 	int i;
 	int err = 0;
+	int inc = test_thread_flag(TIF_32BIT_FPREGS) ? 2 : 1;
 	u64 fpr_val;
 
-	for (i = 0; i < NUM_FPU_REGS; i++) {
-		err |= __get_user(fpr_val, &sc->sc_fpregs[i]);
+	for (i = 0; i < NUM_FPU_REGS; i += inc) {
+		err |= __get_user(fpr_val, &fpregs[i]);
 		set_fpr64(&current->thread.fpu.fpr[i], 0, fpr_val);
 	}
-	err |= __get_user(current->thread.fpu.fcr31, &sc->sc_fpc_csr);
+	err |= __get_user(current->thread.fpu.fcr31, csr);
 
 	return err;
 }
 
 /*
+ * Wrappers for the assembly _{save,restore}_fp_context functions.
+ */
+static int save_hw_fp_context(void __user *sc)
+{
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
+
+	return _save_fp_context(fpregs, csr);
+}
+
+static int restore_hw_fp_context(void __user *sc)
+{
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
+
+	return _restore_fp_context(fpregs, csr);
+}
+
+/*
+ * Extended context handling.
+ */
+
+static inline void __user *sc_to_extcontext(void __user *sc)
+{
+	struct ucontext __user *uc;
+
+	/*
+	 * We can just pretend the sigcontext is always embedded in a struct
+	 * ucontext here, because the offset from sigcontext to extended
+	 * context is the same in the struct sigframe case.
+	 */
+	uc = container_of(sc, struct ucontext, uc_mcontext);
+	return &uc->uc_extcontext;
+}
+
+static int save_msa_extcontext(void __user *buf)
+{
+	struct msa_extcontext __user *msa = buf;
+	uint64_t val;
+	int i, err;
+
+	if (!thread_msa_context_live())
+		return 0;
+
+	/*
+	 * Ensure that we can't lose the live MSA context between checking
+	 * for it & writing it to memory.
+	 */
+	preempt_disable();
+
+	if (is_msa_enabled()) {
+		/*
+		 * There are no EVA versions of the vector register load/store
+		 * instructions, so MSA context has to be saved to kernel memory
+		 * and then copied to user memory. The save to kernel memory
+		 * should already have been done when handling scalar FP
+		 * context.
+		 */
+		BUG_ON(config_enabled(CONFIG_EVA));
+
+		err = __put_user(read_msa_csr(), &msa->csr);
+		err |= _save_msa_all_upper(&msa->wr);
+
+		preempt_enable();
+	} else {
+		preempt_enable();
+
+		err = __put_user(current->thread.fpu.msacsr, &msa->csr);
+
+		for (i = 0; i < NUM_FPU_REGS; i++) {
+			val = get_fpr64(&current->thread.fpu.fpr[i], 1);
+			err |= __put_user(val, &msa->wr[i]);
+		}
+	}
+
+	err |= __put_user(MSA_EXTCONTEXT_MAGIC, &msa->ext.magic);
+	err |= __put_user(sizeof(*msa), &msa->ext.size);
+
+	return err ? -EFAULT : sizeof(*msa);
+}
+
+static int restore_msa_extcontext(void __user *buf, unsigned int size)
+{
+	struct msa_extcontext __user *msa = buf;
+	unsigned long long val;
+	unsigned int csr;
+	int i, err;
+
+	if (size != sizeof(*msa))
+		return -EINVAL;
+
+	err = get_user(csr, &msa->csr);
+	if (err)
+		return err;
+
+	preempt_disable();
+
+	if (is_msa_enabled()) {
+		/*
+		 * There are no EVA versions of the vector register load/store
+		 * instructions, so MSA context has to be copied to kernel
+		 * memory and later loaded to registers. The same is true of
+		 * scalar FP context, so FPU & MSA should have already been
+		 * disabled whilst handling scalar FP context.
+		 */
+		BUG_ON(config_enabled(CONFIG_EVA));
+
+		write_msa_csr(csr);
+		err |= _restore_msa_all_upper(&msa->wr);
+		preempt_enable();
+	} else {
+		preempt_enable();
+
+		current->thread.fpu.msacsr = csr;
+
+		for (i = 0; i < NUM_FPU_REGS; i++) {
+			err |= __get_user(val, &msa->wr[i]);
+			set_fpr64(&current->thread.fpu.fpr[i], 1, val);
+		}
+	}
+
+	return err;
+}
+
+static int save_extcontext(void __user *buf)
+{
+	int sz;
+
+	sz = save_msa_extcontext(buf);
+	if (sz < 0)
+		return sz;
+	buf += sz;
+
+	/* If no context was saved then trivially return */
+	if (!sz)
+		return 0;
+
+	/* Write the end marker */
+	if (__put_user(END_EXTCONTEXT_MAGIC, (u32 *)buf))
+		return -EFAULT;
+
+	sz += sizeof(((struct extcontext *)NULL)->magic);
+	return sz;
+}
+
+static int restore_extcontext(void __user *buf)
+{
+	struct extcontext ext;
+	int err;
+
+	while (1) {
+		err = __get_user(ext.magic, (unsigned int *)buf);
+		if (err)
+			return err;
+
+		if (ext.magic == END_EXTCONTEXT_MAGIC)
+			return 0;
+
+		err = __get_user(ext.size, (unsigned int *)(buf
+			+ offsetof(struct extcontext, size)));
+		if (err)
+			return err;
+
+		switch (ext.magic) {
+		case MSA_EXTCONTEXT_MAGIC:
+			err = restore_msa_extcontext(buf, ext.size);
+			break;
+
+		default:
+			err = -EINVAL;
+			break;
+		}
+
+		if (err)
+			return err;
+
+		buf += ext.size;
+	}
+}
+
+/*
  * Helper routines
  */
-static int protected_save_fp_context(struct sigcontext __user *sc)
+int protected_save_fp_context(void __user *sc)
 {
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
+	uint32_t __user *used_math = sc + abi->off_sc_used_math;
+	unsigned int used, ext_sz;
 	int err;
-#ifndef CONFIG_EVA
+
+	used = used_math() ? USED_FP : 0;
+	if (!used)
+		goto fp_done;
+
+	if (!test_thread_flag(TIF_32BIT_FPREGS))
+		used |= USED_FR1;
+	if (test_thread_flag(TIF_HYBRID_FPREGS))
+		used |= USED_HYBRID_FPRS;
+
+	/*
+	 * EVA does not have userland equivalents of ldc1 or sdc1, so
+	 * save to the kernel FP context & copy that to userland below.
+	 */
+	if (config_enabled(CONFIG_EVA))
+		lose_fpu(1);
+
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
@@ -114,27 +328,57 @@ static int protected_save_fp_context(struct sigcontext __user *sc)
 		if (likely(!err))
 			break;
 		/* touch the sigcontext and try again */
-		err = __put_user(0, &sc->sc_fpregs[0]) |
-			__put_user(0, &sc->sc_fpregs[31]) |
-			__put_user(0, &sc->sc_fpc_csr);
+		err = __put_user(0, &fpregs[0]) |
+			__put_user(0, &fpregs[31]) |
+			__put_user(0, csr);
 		if (err)
-			break;	/* really bad sigcontext */
+			return err;	/* really bad sigcontext */
 	}
-#else
-	/*
-	 * EVA does not have FPU EVA instructions so saving fpu context directly
-	 * does not work.
-	 */
-	lose_fpu(1);
-	err = save_fp_context(sc); /* this might fail */
-#endif
-	return err;
+
+fp_done:
+	ext_sz = err = save_extcontext(sc_to_extcontext(sc));
+	if (err < 0)
+		return err;
+	used |= ext_sz ? USED_EXTCONTEXT : 0;
+
+	return __put_user(used, used_math);
 }
 
-static int protected_restore_fp_context(struct sigcontext __user *sc)
+int protected_restore_fp_context(void __user *sc)
 {
-	int err, tmp __maybe_unused;
-#ifndef CONFIG_EVA
+	struct mips_abi *abi = current->thread.abi;
+	uint64_t __user *fpregs = sc + abi->off_sc_fpregs;
+	uint32_t __user *csr = sc + abi->off_sc_fpc_csr;
+	uint32_t __user *used_math = sc + abi->off_sc_used_math;
+	unsigned int used;
+	int err, sig = 0, tmp __maybe_unused;
+
+	err = __get_user(used, used_math);
+	conditional_used_math(used & USED_FP);
+
+	/*
+	 * The signal handler may have used FPU; give it up if the program
+	 * doesn't want it following sigreturn.
+	 */
+	if (err || !(used & USED_FP))
+		lose_fpu(0);
+	if (err)
+		return err;
+	if (!(used & USED_FP))
+		goto fp_done;
+
+	err = sig = fpcsr_pending(csr);
+	if (err < 0)
+		return err;
+
+	/*
+	 * EVA does not have userland equivalents of ldc1 or sdc1, so we
+	 * disable the FPU here such that the code below simply copies to
+	 * the kernel FP context.
+	 */
+	if (config_enabled(CONFIG_EVA))
+		lose_fpu(0);
+
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
@@ -147,28 +391,24 @@ static int protected_restore_fp_context(struct sigcontext __user *sc)
 		if (likely(!err))
 			break;
 		/* touch the sigcontext and try again */
-		err = __get_user(tmp, &sc->sc_fpregs[0]) |
-			__get_user(tmp, &sc->sc_fpregs[31]) |
-			__get_user(tmp, &sc->sc_fpc_csr);
+		err = __get_user(tmp, &fpregs[0]) |
+			__get_user(tmp, &fpregs[31]) |
+			__get_user(tmp, csr);
 		if (err)
 			break;	/* really bad sigcontext */
 	}
-#else
-	/*
-	 * EVA does not have FPU EVA instructions so restoring fpu context
-	 * directly does not work.
-	 */
-	lose_fpu(0);
-	err = restore_fp_context(sc); /* this might fail */
-#endif
-	return err;
+
+fp_done:
+	if (used & USED_EXTCONTEXT)
+		err |= restore_extcontext(sc_to_extcontext(sc));
+
+	return err ?: sig;
 }
 
 int setup_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 {
 	int err = 0;
 	int i;
-	unsigned int used_math;
 
 	err |= __put_user(regs->cp0_epc, &sc->sc_pc);
 
@@ -191,19 +431,38 @@ int setup_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 		err |= __put_user(rddsp(DSP_MASK), &sc->sc_dsp);
 	}
 
-	used_math = !!used_math();
-	err |= __put_user(used_math, &sc->sc_used_math);
 
-	if (used_math) {
-		/*
-		 * Save FPU state to signal context. Signal handler
-		 * will "inherit" current FPU state.
-		 */
-		err |= protected_save_fp_context(sc);
-	}
+	/*
+	 * Save FPU state to signal context. Signal handler
+	 * will "inherit" current FPU state.
+	 */
+	err |= protected_save_fp_context(sc);
+
 	return err;
 }
 
+static size_t extcontext_max_size(void)
+{
+	size_t sz = 0;
+
+	/*
+	 * The assumption here is that between this point & the point at which
+	 * the extended context is saved the size of the context should only
+	 * ever be able to shrink (if the task is preempted), but never grow.
+	 * That is, what this function returns is an upper bound on the size of
+	 * the extended context for the current task at the current time.
+	 */
+
+	if (thread_msa_context_live())
+		sz += sizeof(struct msa_extcontext);
+
+	/* If any context is saved then we'll append the end marker */
+	if (sz)
+		sz += sizeof(((struct extcontext *)NULL)->magic);
+
+	return sz;
+}
+
 int fpcsr_pending(unsigned int __user *fpcsr)
 {
 	int err, sig = 0;
@@ -223,21 +482,8 @@ int fpcsr_pending(unsigned int __user *fpcsr)
 	return err ?: sig;
 }
 
-static int
-check_and_restore_fp_context(struct sigcontext __user *sc)
-{
-	int err, sig;
-
-	err = sig = fpcsr_pending(&sc->sc_fpc_csr);
-	if (err > 0)
-		err = 0;
-	err |= protected_restore_fp_context(sc);
-	return err ?: sig;
-}
-
 int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 {
-	unsigned int used_math;
 	unsigned long treg;
 	int err = 0;
 	int i;
@@ -265,19 +511,7 @@ int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 	for (i = 1; i < 32; i++)
 		err |= __get_user(regs->regs[i], &sc->sc_regs[i]);
 
-	err |= __get_user(used_math, &sc->sc_used_math);
-	conditional_used_math(used_math);
-
-	if (used_math) {
-		/* restore fpu context if we have used it before */
-		if (!err)
-			err = check_and_restore_fp_context(sc);
-	} else {
-		/* signal handler may have used FPU.  Give it up. */
-		lose_fpu(0);
-	}
-
-	return err;
+	return err ?: protected_restore_fp_context(sc);
 }
 
 void __user *get_sigframe(struct ksignal *ksig, struct pt_regs *regs,
@@ -285,6 +519,9 @@ void __user *get_sigframe(struct ksignal *ksig, struct pt_regs *regs,
 {
 	unsigned long sp;
 
+	/* Leave space for potential extended context */
+	frame_size += extcontext_max_size();
+
 	/* Default to using normal stack */
 	sp = regs->regs[29];
 
@@ -520,7 +757,11 @@ struct mips_abi mips_abi = {
 	.setup_rt_frame = setup_rt_frame,
 	.rt_signal_return_offset =
 		offsetof(struct mips_vdso, rt_signal_trampoline),
-	.restart	= __NR_restart_syscall
+	.restart	= __NR_restart_syscall,
+
+	.off_sc_fpregs = offsetof(struct sigcontext, sc_fpregs),
+	.off_sc_fpc_csr = offsetof(struct sigcontext, sc_fpc_csr),
+	.off_sc_used_math = offsetof(struct sigcontext, sc_used_math),
 };
 
 static void handle_signal(struct ksignal *ksig, struct pt_regs *regs)
@@ -616,6 +857,9 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, void *unused,
 
 	user_exit();
 
+	if (thread_info_flags & _TIF_UPROBE)
+		uprobe_notify_resume(regs);
+
 	/* deal with pending signal delivery */
 	if (thread_info_flags & _TIF_SIGPENDING)
 		do_signal(regs);
@@ -629,43 +873,46 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, void *unused,
 }
 
 #ifdef CONFIG_SMP
-#ifndef CONFIG_EVA
-static int smp_save_fp_context(struct sigcontext __user *sc)
+static int smp_save_fp_context(void __user *sc)
 {
 	return raw_cpu_has_fpu
-	       ? _save_fp_context(sc)
+	       ? save_hw_fp_context(sc)
 	       : copy_fp_to_sigcontext(sc);
 }
 
-static int smp_restore_fp_context(struct sigcontext __user *sc)
+static int smp_restore_fp_context(void __user *sc)
 {
 	return raw_cpu_has_fpu
-	       ? _restore_fp_context(sc)
+	       ? restore_hw_fp_context(sc)
 	       : copy_fp_from_sigcontext(sc);
 }
-#endif /* CONFIG_EVA */
 #endif
 
 static int signal_setup(void)
 {
-#ifndef CONFIG_EVA
+	/*
+	 * The offset from sigcontext to extended context should be the same
+	 * regardless of the type of signal, such that userland can always know
+	 * where to look if it wishes to find the extended context structures.
+	 */
+	BUILD_BUG_ON((offsetof(struct sigframe, sf_extcontext) -
+		      offsetof(struct sigframe, sf_sc)) !=
+		     (offsetof(struct rt_sigframe, rs_uc.uc_extcontext) -
+		      offsetof(struct rt_sigframe, rs_uc.uc_mcontext)));
+
 #ifdef CONFIG_SMP
 	/* For now just do the cpu_has_fpu check when the functions are invoked */
 	save_fp_context = smp_save_fp_context;
 	restore_fp_context = smp_restore_fp_context;
 #else
 	if (cpu_has_fpu) {
-		save_fp_context = _save_fp_context;
-		restore_fp_context = _restore_fp_context;
+		save_fp_context = save_hw_fp_context;
+		restore_fp_context = restore_hw_fp_context;
 	} else {
 		save_fp_context = copy_fp_to_sigcontext;
 		restore_fp_context = copy_fp_from_sigcontext;
 	}
 #endif /* CONFIG_SMP */
-#else
-	save_fp_context = copy_fp_to_sigcontext;
-	restore_fp_context = copy_fp_from_sigcontext;
-#endif
 
 	return 0;
 }
diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c
index 5d7f263..f7e8952 100644
--- a/arch/mips/kernel/signal32.c
+++ b/arch/mips/kernel/signal32.c
@@ -36,12 +36,6 @@
 
 #include "signal-common.h"
 
-static int (*save_fp_context32)(struct sigcontext32 __user *sc);
-static int (*restore_fp_context32)(struct sigcontext32 __user *sc);
-
-extern asmlinkage int _save_fp_context32(struct sigcontext32 __user *sc);
-extern asmlinkage int _restore_fp_context32(struct sigcontext32 __user *sc);
-
 /*
  * Including <asm/unistd.h> would give use the 64-bit syscall numbers ...
  */
@@ -74,99 +68,11 @@ struct rt_sigframe32 {
 	struct ucontext32 rs_uc;
 };
 
-/*
- * Thread saved context copy to/from a signal context presumed to be on the
- * user stack, and therefore accessed with appropriate macros from uaccess.h.
- */
-static int copy_fp_to_sigcontext32(struct sigcontext32 __user *sc)
-{
-	int i;
-	int err = 0;
-	int inc = test_thread_flag(TIF_32BIT_FPREGS) ? 2 : 1;
-
-	for (i = 0; i < NUM_FPU_REGS; i += inc) {
-		err |=
-		    __put_user(get_fpr64(&current->thread.fpu.fpr[i], 0),
-			       &sc->sc_fpregs[i]);
-	}
-	err |= __put_user(current->thread.fpu.fcr31, &sc->sc_fpc_csr);
-
-	return err;
-}
-
-static int copy_fp_from_sigcontext32(struct sigcontext32 __user *sc)
-{
-	int i;
-	int err = 0;
-	int inc = test_thread_flag(TIF_32BIT_FPREGS) ? 2 : 1;
-	u64 fpr_val;
-
-	for (i = 0; i < NUM_FPU_REGS; i += inc) {
-		err |= __get_user(fpr_val, &sc->sc_fpregs[i]);
-		set_fpr64(&current->thread.fpu.fpr[i], 0, fpr_val);
-	}
-	err |= __get_user(current->thread.fpu.fcr31, &sc->sc_fpc_csr);
-
-	return err;
-}
-
-/*
- * sigcontext handlers
- */
-static int protected_save_fp_context32(struct sigcontext32 __user *sc)
-{
-	int err;
-	while (1) {
-		lock_fpu_owner();
-		if (is_fpu_owner()) {
-			err = save_fp_context32(sc);
-			unlock_fpu_owner();
-		} else {
-			unlock_fpu_owner();
-			err = copy_fp_to_sigcontext32(sc);
-		}
-		if (likely(!err))
-			break;
-		/* touch the sigcontext and try again */
-		err = __put_user(0, &sc->sc_fpregs[0]) |
-			__put_user(0, &sc->sc_fpregs[31]) |
-			__put_user(0, &sc->sc_fpc_csr);
-		if (err)
-			break;	/* really bad sigcontext */
-	}
-	return err;
-}
-
-static int protected_restore_fp_context32(struct sigcontext32 __user *sc)
-{
-	int err, tmp __maybe_unused;
-	while (1) {
-		lock_fpu_owner();
-		if (is_fpu_owner()) {
-			err = restore_fp_context32(sc);
-			unlock_fpu_owner();
-		} else {
-			unlock_fpu_owner();
-			err = copy_fp_from_sigcontext32(sc);
-		}
-		if (likely(!err))
-			break;
-		/* touch the sigcontext and try again */
-		err = __get_user(tmp, &sc->sc_fpregs[0]) |
-			__get_user(tmp, &sc->sc_fpregs[31]) |
-			__get_user(tmp, &sc->sc_fpc_csr);
-		if (err)
-			break;	/* really bad sigcontext */
-	}
-	return err;
-}
-
 static int setup_sigcontext32(struct pt_regs *regs,
 			      struct sigcontext32 __user *sc)
 {
 	int err = 0;
 	int i;
-	u32 used_math;
 
 	err |= __put_user(regs->cp0_epc, &sc->sc_pc);
 
@@ -186,35 +92,18 @@ static int setup_sigcontext32(struct pt_regs *regs,
 		err |= __put_user(mflo3(), &sc->sc_lo3);
 	}
 
-	used_math = !!used_math();
-	err |= __put_user(used_math, &sc->sc_used_math);
+	/*
+	 * Save FPU state to signal context.  Signal handler
+	 * will "inherit" current FPU state.
+	 */
+	err |= protected_save_fp_context(sc);
 
-	if (used_math) {
-		/*
-		 * Save FPU state to signal context.  Signal handler
-		 * will "inherit" current FPU state.
-		 */
-		err |= protected_save_fp_context32(sc);
-	}
 	return err;
 }
 
-static int
-check_and_restore_fp_context32(struct sigcontext32 __user *sc)
-{
-	int err, sig;
-
-	err = sig = fpcsr_pending(&sc->sc_fpc_csr);
-	if (err > 0)
-		err = 0;
-	err |= protected_restore_fp_context32(sc);
-	return err ?: sig;
-}
-
 static int restore_sigcontext32(struct pt_regs *regs,
 				struct sigcontext32 __user *sc)
 {
-	u32 used_math;
 	int err = 0;
 	s32 treg;
 	int i;
@@ -238,70 +127,7 @@ static int restore_sigcontext32(struct pt_regs *regs,
 	for (i = 1; i < 32; i++)
 		err |= __get_user(regs->regs[i], &sc->sc_regs[i]);
 
-	err |= __get_user(used_math, &sc->sc_used_math);
-	conditional_used_math(used_math);
-
-	if (used_math) {
-		/* restore fpu context if we have used it before */
-		if (!err)
-			err = check_and_restore_fp_context32(sc);
-	} else {
-		/* signal handler may have used FPU.  Give it up. */
-		lose_fpu(0);
-	}
-
-	return err;
-}
-
-/*
- *
- */
-extern void __put_sigset_unknown_nsig(void);
-extern void __get_sigset_unknown_nsig(void);
-
-static inline int put_sigset(const sigset_t *kbuf, compat_sigset_t __user *ubuf)
-{
-	int err = 0;
-
-	if (!access_ok(VERIFY_WRITE, ubuf, sizeof(*ubuf)))
-		return -EFAULT;
-
-	switch (_NSIG_WORDS) {
-	default:
-		__put_sigset_unknown_nsig();
-	case 2:
-		err |= __put_user(kbuf->sig[1] >> 32, &ubuf->sig[3]);
-		err |= __put_user(kbuf->sig[1] & 0xffffffff, &ubuf->sig[2]);
-	case 1:
-		err |= __put_user(kbuf->sig[0] >> 32, &ubuf->sig[1]);
-		err |= __put_user(kbuf->sig[0] & 0xffffffff, &ubuf->sig[0]);
-	}
-
-	return err;
-}
-
-static inline int get_sigset(sigset_t *kbuf, const compat_sigset_t __user *ubuf)
-{
-	int err = 0;
-	unsigned long sig[4];
-
-	if (!access_ok(VERIFY_READ, ubuf, sizeof(*ubuf)))
-		return -EFAULT;
-
-	switch (_NSIG_WORDS) {
-	default:
-		__get_sigset_unknown_nsig();
-	case 2:
-		err |= __get_user(sig[3], &ubuf->sig[3]);
-		err |= __get_user(sig[2], &ubuf->sig[2]);
-		kbuf->sig[1] = sig[2] | (sig[3] << 32);
-	case 1:
-		err |= __get_user(sig[1], &ubuf->sig[1]);
-		err |= __get_user(sig[0], &ubuf->sig[0]);
-		kbuf->sig[0] = sig[0] | (sig[1] << 32);
-	}
-
-	return err;
+	return err ?: protected_restore_fp_context(sc);
 }
 
 /*
@@ -585,20 +411,9 @@ struct mips_abi mips_abi_32 = {
 	.setup_rt_frame = setup_rt_frame_32,
 	.rt_signal_return_offset =
 		offsetof(struct mips_vdso, o32_rt_signal_trampoline),
-	.restart	= __NR_O32_restart_syscall
-};
-
-static int signal32_init(void)
-{
-	if (cpu_has_fpu) {
-		save_fp_context32 = _save_fp_context32;
-		restore_fp_context32 = _restore_fp_context32;
-	} else {
-		save_fp_context32 = copy_fp_to_sigcontext32;
-		restore_fp_context32 = copy_fp_from_sigcontext32;
-	}
+	.restart	= __NR_O32_restart_syscall,
 
-	return 0;
-}
-
-arch_initcall(signal32_init);
+	.off_sc_fpregs = offsetof(struct sigcontext32, sc_fpregs),
+	.off_sc_fpc_csr = offsetof(struct sigcontext32, sc_fpc_csr),
+	.off_sc_used_math = offsetof(struct sigcontext32, sc_used_math),
+};
diff --git a/arch/mips/kernel/signal_n32.c b/arch/mips/kernel/signal_n32.c
index f1d4751..0d017fd 100644
--- a/arch/mips/kernel/signal_n32.c
+++ b/arch/mips/kernel/signal_n32.c
@@ -153,5 +153,9 @@ struct mips_abi mips_abi_n32 = {
 	.setup_rt_frame = setup_rt_frame_n32,
 	.rt_signal_return_offset =
 		offsetof(struct mips_vdso, n32_rt_signal_trampoline),
-	.restart	= __NR_N32_restart_syscall
+	.restart	= __NR_N32_restart_syscall,
+
+	.off_sc_fpregs = offsetof(struct sigcontext, sc_fpregs),
+	.off_sc_fpc_csr = offsetof(struct sigcontext, sc_fpc_csr),
+	.off_sc_used_math = offsetof(struct sigcontext, sc_used_math),
 };
diff --git a/arch/mips/kernel/spram.c b/arch/mips/kernel/spram.c
index d1168d7..8489c88 100644
--- a/arch/mips/kernel/spram.c
+++ b/arch/mips/kernel/spram.c
@@ -209,6 +209,7 @@ void spram_config(void)
 	case CPU_PROAPTIV:
 	case CPU_P5600:
 	case CPU_QEMU_GENERIC:
+	case CPU_I6400:
 		config0 = read_c0_config();
 		/* FIXME: addresses are Malta specific */
 		if (config0 & (1<<24)) {
diff --git a/arch/mips/kernel/sysrq.c b/arch/mips/kernel/sysrq.c
index 5b539f5..5f05539 100644
--- a/arch/mips/kernel/sysrq.c
+++ b/arch/mips/kernel/sysrq.c
@@ -21,24 +21,12 @@ static DEFINE_SPINLOCK(show_lock);
 
 static void sysrq_tlbdump_single(void *dummy)
 {
-	const int field = 2 * sizeof(unsigned long);
 	unsigned long flags;
 
 	spin_lock_irqsave(&show_lock, flags);
 
 	pr_info("CPU%d:\n", smp_processor_id());
-	pr_info("Index	: %0x\n", read_c0_index());
-	pr_info("Pagemask: %0x\n", read_c0_pagemask());
-	pr_info("EntryHi : %0*lx\n", field, read_c0_entryhi());
-	pr_info("EntryLo0: %0*lx\n", field, read_c0_entrylo0());
-	pr_info("EntryLo1: %0*lx\n", field, read_c0_entrylo1());
-	pr_info("Wired   : %0x\n", read_c0_wired());
-	pr_info("Pagegrain: %0x\n", read_c0_pagegrain());
-	if (cpu_has_htw) {
-		pr_info("PWField : %0*lx\n", field, read_c0_pwfield());
-		pr_info("PWSize  : %0*lx\n", field, read_c0_pwsize());
-		pr_info("PWCtl   : %0x\n", read_c0_pwctl());
-	}
+	dump_tlb_regs();
 	pr_info("\n");
 	dump_tlb_all();
 	pr_info("\n");
diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c
index 8ea28e6..fdb392b 100644
--- a/arch/mips/kernel/traps.c
+++ b/arch/mips/kernel/traps.c
@@ -370,11 +370,6 @@ void show_registers(struct pt_regs *regs)
 	set_fs(old_fs);
 }
 
-static int regs_to_trapnr(struct pt_regs *regs)
-{
-	return (regs->cp0_cause >> 2) & 0x1f;
-}
-
 static DEFINE_RAW_SPINLOCK(die_lock);
 
 void __noreturn die(const char *str, struct pt_regs *regs)
@@ -384,7 +379,7 @@ void __noreturn die(const char *str, struct pt_regs *regs)
 
 	oops_enter();
 
-	if (notify_die(DIE_OOPS, str, regs, 0, regs_to_trapnr(regs),
+	if (notify_die(DIE_OOPS, str, regs, 0, current->thread.trap_nr,
 		       SIGSEGV) == NOTIFY_STOP)
 		sig = 0;
 
@@ -470,7 +465,7 @@ asmlinkage void do_be(struct pt_regs *regs)
 	printk(KERN_ALERT "%s bus error, epc == %0*lx, ra == %0*lx\n",
 	       data ? "Data" : "Instruction",
 	       field, regs->cp0_epc, field, regs->regs[31]);
-	if (notify_die(DIE_OOPS, "bus error", regs, 0, regs_to_trapnr(regs),
+	if (notify_die(DIE_OOPS, "bus error", regs, 0, current->thread.trap_nr,
 		       SIGBUS) == NOTIFY_STOP)
 		goto out;
 
@@ -826,7 +821,7 @@ asmlinkage void do_fpe(struct pt_regs *regs, unsigned long fcr31)
 	int sig;
 
 	prev_state = exception_enter();
-	if (notify_die(DIE_FP, "FP exception", regs, 0, regs_to_trapnr(regs),
+	if (notify_die(DIE_FP, "FP exception", regs, 0, current->thread.trap_nr,
 		       SIGFPE) == NOTIFY_STOP)
 		goto out;
 
@@ -882,11 +877,12 @@ void do_trap_or_bp(struct pt_regs *regs, unsigned int code,
 	char b[40];
 
 #ifdef CONFIG_KGDB_LOW_LEVEL_TRAP
-	if (kgdb_ll_trap(DIE_TRAP, str, regs, code, regs_to_trapnr(regs), SIGTRAP) == NOTIFY_STOP)
+	if (kgdb_ll_trap(DIE_TRAP, str, regs, code, current->thread.trap_nr,
+			 SIGTRAP) == NOTIFY_STOP)
 		return;
 #endif /* CONFIG_KGDB_LOW_LEVEL_TRAP */
 
-	if (notify_die(DIE_TRAP, str, regs, code, regs_to_trapnr(regs),
+	if (notify_die(DIE_TRAP, str, regs, code, current->thread.trap_nr,
 		       SIGTRAP) == NOTIFY_STOP)
 		return;
 
@@ -948,6 +944,7 @@ asmlinkage void do_bp(struct pt_regs *regs)
 		set_fs(KERNEL_DS);
 
 	prev_state = exception_enter();
+	current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 	if (get_isa16_mode(regs->cp0_epc)) {
 		u16 instr[2];
 
@@ -987,15 +984,27 @@ asmlinkage void do_bp(struct pt_regs *regs)
 	 * pertain to them.
 	 */
 	switch (bcode) {
+	case BRK_UPROBE:
+		if (notify_die(DIE_UPROBE, "uprobe", regs, bcode,
+			       current->thread.trap_nr, SIGTRAP) == NOTIFY_STOP)
+			goto out;
+		else
+			break;
+	case BRK_UPROBE_XOL:
+		if (notify_die(DIE_UPROBE_XOL, "uprobe_xol", regs, bcode,
+			       current->thread.trap_nr, SIGTRAP) == NOTIFY_STOP)
+			goto out;
+		else
+			break;
 	case BRK_KPROBE_BP:
 		if (notify_die(DIE_BREAK, "debug", regs, bcode,
-			       regs_to_trapnr(regs), SIGTRAP) == NOTIFY_STOP)
+			       current->thread.trap_nr, SIGTRAP) == NOTIFY_STOP)
 			goto out;
 		else
 			break;
 	case BRK_KPROBE_SSTEPBP:
 		if (notify_die(DIE_SSTEPBP, "single_step", regs, bcode,
-			       regs_to_trapnr(regs), SIGTRAP) == NOTIFY_STOP)
+			       current->thread.trap_nr, SIGTRAP) == NOTIFY_STOP)
 			goto out;
 		else
 			break;
@@ -1028,6 +1037,7 @@ asmlinkage void do_tr(struct pt_regs *regs)
 		set_fs(get_ds());
 
 	prev_state = exception_enter();
+	current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 	if (get_isa16_mode(regs->cp0_epc)) {
 		if (__get_user(instr[0], (u16 __user *)(epc + 0)) ||
 		    __get_user(instr[1], (u16 __user *)(epc + 2)))
@@ -1094,8 +1104,9 @@ asmlinkage void do_ri(struct pt_regs *regs)
 no_r2_instr:
 
 	prev_state = exception_enter();
+	current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 
-	if (notify_die(DIE_RI, "RI Fault", regs, 0, regs_to_trapnr(regs),
+	if (notify_die(DIE_RI, "RI Fault", regs, 0, current->thread.trap_nr,
 		       SIGILL) == NOTIFY_STOP)
 		goto out;
 
@@ -1444,8 +1455,9 @@ asmlinkage void do_msa_fpe(struct pt_regs *regs, unsigned int msacsr)
 	enum ctx_state prev_state;
 
 	prev_state = exception_enter();
+	current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 	if (notify_die(DIE_MSAFP, "MSA FP exception", regs, 0,
-		       regs_to_trapnr(regs), SIGFPE) == NOTIFY_STOP)
+		       current->thread.trap_nr, SIGFPE) == NOTIFY_STOP)
 		goto out;
 
 	/* Clear MSACSR.Cause before enabling interrupts */
@@ -1523,7 +1535,6 @@ asmlinkage void do_watch(struct pt_regs *regs)
 
 asmlinkage void do_mcheck(struct pt_regs *regs)
 {
-	const int field = 2 * sizeof(unsigned long);
 	int multi_match = regs->cp0_status & ST0_TS;
 	enum ctx_state prev_state;
 	mm_segment_t old_fs = get_fs();
@@ -1532,19 +1543,8 @@ asmlinkage void do_mcheck(struct pt_regs *regs)
 	show_regs(regs);
 
 	if (multi_match) {
-		pr_err("Index	: %0x\n", read_c0_index());
-		pr_err("Pagemask: %0x\n", read_c0_pagemask());
-		pr_err("EntryHi : %0*lx\n", field, read_c0_entryhi());
-		pr_err("EntryLo0: %0*lx\n", field, read_c0_entrylo0());
-		pr_err("EntryLo1: %0*lx\n", field, read_c0_entrylo1());
-		pr_err("Wired   : %0x\n", read_c0_wired());
-		pr_err("Pagegrain: %0x\n", read_c0_pagegrain());
-		if (cpu_has_htw) {
-			pr_err("PWField : %0*lx\n", field, read_c0_pwfield());
-			pr_err("PWSize  : %0*lx\n", field, read_c0_pwsize());
-			pr_err("PWCtl   : %0x\n", read_c0_pwctl());
-		}
-		pr_err("\n");
+		dump_tlb_regs();
+		pr_info("\n");
 		dump_tlb_all();
 	}
 
@@ -1651,6 +1651,7 @@ static inline void parity_protection_init(void)
 	case CPU_PROAPTIV:
 	case CPU_P5600:
 	case CPU_QEMU_GENERIC:
+	case CPU_I6400:
 		{
 #define ERRCTL_PE	0x80000000
 #define ERRCTL_L2P	0x00800000
diff --git a/arch/mips/kernel/unaligned.c b/arch/mips/kernel/unaligned.c
index eb3efd1..990354d 100644
--- a/arch/mips/kernel/unaligned.c
+++ b/arch/mips/kernel/unaligned.c
@@ -891,6 +891,9 @@ static void emulate_load_store_insn(struct pt_regs *regs,
 #ifdef	CONFIG_EVA
 	mm_segment_t seg;
 #endif
+	union fpureg *fpr;
+	enum msa_2b_fmt df;
+	unsigned int wd;
 	origpc = (unsigned long)pc;
 	orig31 = regs->regs[31];
 
@@ -1202,6 +1205,75 @@ static void emulate_load_store_insn(struct pt_regs *regs,
 			break;
 		return;
 
+	case msa_op:
+		if (!cpu_has_msa)
+			goto sigill;
+
+		/*
+		 * If we've reached this point then userland should have taken
+		 * the MSA disabled exception & initialised vector context at
+		 * some point in the past.
+		 */
+		BUG_ON(!thread_msa_context_live());
+
+		df = insn.msa_mi10_format.df;
+		wd = insn.msa_mi10_format.wd;
+		fpr = &current->thread.fpu.fpr[wd];
+
+		switch (insn.msa_mi10_format.func) {
+		case msa_ld_op:
+			if (!access_ok(VERIFY_READ, addr, sizeof(*fpr)))
+				goto sigbus;
+
+			/*
+			 * Disable preemption to avoid a race between copying
+			 * state from userland, migrating to another CPU and
+			 * updating the hardware vector register below.
+			 */
+			preempt_disable();
+
+			res = __copy_from_user_inatomic(fpr, addr,
+							sizeof(*fpr));
+			if (res)
+				goto fault;
+
+			/*
+			 * Update the hardware register if it is in use by the
+			 * task in this quantum, in order to avoid having to
+			 * save & restore the whole vector context.
+			 */
+			if (test_thread_flag(TIF_USEDMSA))
+				write_msa_wr(wd, fpr, df);
+
+			preempt_enable();
+			break;
+
+		case msa_st_op:
+			if (!access_ok(VERIFY_WRITE, addr, sizeof(*fpr)))
+				goto sigbus;
+
+			/*
+			 * Update from the hardware register if it is in use by
+			 * the task in this quantum, in order to avoid having to
+			 * save & restore the whole vector context.
+			 */
+			preempt_disable();
+			if (test_thread_flag(TIF_USEDMSA))
+				read_msa_wr(wd, fpr, df);
+			preempt_enable();
+
+			res = __copy_to_user_inatomic(addr, fpr, sizeof(*fpr));
+			if (res)
+				goto fault;
+			break;
+
+		default:
+			goto sigbus;
+		}
+
+		compute_return_epc(regs);
+		break;
+
 #ifndef CONFIG_CPU_MIPSR6
 	/*
 	 * COP2 is available to implementor for application specific use.
@@ -2240,5 +2312,5 @@ static int __init debugfs_unaligned(void)
 		return -ENOMEM;
 	return 0;
 }
-__initcall(debugfs_unaligned);
+arch_initcall(debugfs_unaligned);
 #endif
diff --git a/arch/mips/kernel/uprobes.c b/arch/mips/kernel/uprobes.c
new file mode 100644
index 0000000..8452d93
--- /dev/null
+++ b/arch/mips/kernel/uprobes.c
@@ -0,0 +1,341 @@
+#include <linux/highmem.h>
+#include <linux/kdebug.h>
+#include <linux/types.h>
+#include <linux/notifier.h>
+#include <linux/sched.h>
+#include <linux/uprobes.h>
+
+#include <asm/branch.h>
+#include <asm/cpu-features.h>
+#include <asm/ptrace.h>
+#include <asm/inst.h>
+
+static inline int insn_has_delay_slot(const union mips_instruction insn)
+{
+	switch (insn.i_format.opcode) {
+	/*
+	 * jr and jalr are in r_format format.
+	 */
+	case spec_op:
+		switch (insn.r_format.func) {
+		case jalr_op:
+		case jr_op:
+			return 1;
+		}
+		break;
+
+	/*
+	 * This group contains:
+	 * bltz_op, bgez_op, bltzl_op, bgezl_op,
+	 * bltzal_op, bgezal_op, bltzall_op, bgezall_op.
+	 */
+	case bcond_op:
+		switch (insn.i_format.rt) {
+		case bltz_op:
+		case bltzl_op:
+		case bgez_op:
+		case bgezl_op:
+		case bltzal_op:
+		case bltzall_op:
+		case bgezal_op:
+		case bgezall_op:
+		case bposge32_op:
+			return 1;
+		}
+		break;
+
+	/*
+	 * These are unconditional and in j_format.
+	 */
+	case jal_op:
+	case j_op:
+	case beq_op:
+	case beql_op:
+	case bne_op:
+	case bnel_op:
+	case blez_op: /* not really i_format */
+	case blezl_op:
+	case bgtz_op:
+	case bgtzl_op:
+		return 1;
+
+	/*
+	 * And now the FPA/cp1 branch instructions.
+	 */
+	case cop1_op:
+#ifdef CONFIG_CPU_CAVIUM_OCTEON
+	case lwc2_op: /* This is bbit0 on Octeon */
+	case ldc2_op: /* This is bbit032 on Octeon */
+	case swc2_op: /* This is bbit1 on Octeon */
+	case sdc2_op: /* This is bbit132 on Octeon */
+#endif
+		return 1;
+	}
+
+	return 0;
+}
+
+/**
+ * arch_uprobe_analyze_insn - instruction analysis including validity and fixups.
+ * @mm: the probed address space.
+ * @arch_uprobe: the probepoint information.
+ * @addr: virtual address at which to install the probepoint
+ * Return 0 on success or a -ve number on error.
+ */
+int arch_uprobe_analyze_insn(struct arch_uprobe *aup,
+	struct mm_struct *mm, unsigned long addr)
+{
+	union mips_instruction inst;
+
+	/*
+	 * For the time being this also blocks attempts to use uprobes with
+	 * MIPS16 and microMIPS.
+	 */
+	if (addr & 0x03)
+		return -EINVAL;
+
+	inst.word = aup->insn[0];
+	aup->ixol[0] = aup->insn[insn_has_delay_slot(inst)];
+	aup->ixol[1] = UPROBE_BRK_UPROBE_XOL;		/* NOP  */
+
+	return 0;
+}
+
+/**
+ * is_trap_insn - check if the instruction is a trap variant
+ * @insn: instruction to be checked.
+ * Returns true if @insn is a trap variant.
+ *
+ * This definition overrides the weak definition in kernel/events/uprobes.c.
+ * and is needed for the case where an architecture has multiple trap
+ * instructions (like PowerPC or MIPS).  We treat BREAK just like the more
+ * modern conditional trap instructions.
+ */
+bool is_trap_insn(uprobe_opcode_t *insn)
+{
+	union mips_instruction inst;
+
+	inst.word = *insn;
+
+	switch (inst.i_format.opcode) {
+	case spec_op:
+		switch (inst.r_format.func) {
+		case break_op:
+		case teq_op:
+		case tge_op:
+		case tgeu_op:
+		case tlt_op:
+		case tltu_op:
+		case tne_op:
+			return 1;
+		}
+		break;
+
+	case bcond_op:	/* Yes, really ...  */
+		switch (inst.u_format.rt) {
+		case teqi_op:
+		case tgei_op:
+		case tgeiu_op:
+		case tlti_op:
+		case tltiu_op:
+		case tnei_op:
+			return 1;
+		}
+		break;
+	}
+
+	return 0;
+}
+
+#define UPROBE_TRAP_NR	ULONG_MAX
+
+/*
+ * arch_uprobe_pre_xol - prepare to execute out of line.
+ * @auprobe: the probepoint information.
+ * @regs: reflects the saved user state of current task.
+ */
+int arch_uprobe_pre_xol(struct arch_uprobe *aup, struct pt_regs *regs)
+{
+	struct uprobe_task *utask = current->utask;
+	union mips_instruction insn;
+
+	/*
+	 * Now find the EPC where to resume after the breakpoint has been
+	 * dealt with.  This may require emulation of a branch.
+	 */
+	aup->resume_epc = regs->cp0_epc + 4;
+	if (insn_has_delay_slot((union mips_instruction) aup->insn[0])) {
+		unsigned long epc;
+
+		epc = regs->cp0_epc;
+		__compute_return_epc_for_insn(regs, insn);
+		aup->resume_epc = regs->cp0_epc;
+	}
+
+	utask->autask.saved_trap_nr = current->thread.trap_nr;
+	current->thread.trap_nr = UPROBE_TRAP_NR;
+	regs->cp0_epc = current->utask->xol_vaddr;
+
+	return 0;
+}
+
+int arch_uprobe_post_xol(struct arch_uprobe *aup, struct pt_regs *regs)
+{
+	struct uprobe_task *utask = current->utask;
+
+	current->thread.trap_nr = utask->autask.saved_trap_nr;
+	regs->cp0_epc = aup->resume_epc;
+
+	return 0;
+}
+
+/*
+ * If xol insn itself traps and generates a signal(Say,
+ * SIGILL/SIGSEGV/etc), then detect the case where a singlestepped
+ * instruction jumps back to its own address. It is assumed that anything
+ * like do_page_fault/do_trap/etc sets thread.trap_nr != -1.
+ *
+ * arch_uprobe_pre_xol/arch_uprobe_post_xol save/restore thread.trap_nr,
+ * arch_uprobe_xol_was_trapped() simply checks that ->trap_nr is not equal to
+ * UPROBE_TRAP_NR == -1 set by arch_uprobe_pre_xol().
+ */
+bool arch_uprobe_xol_was_trapped(struct task_struct *tsk)
+{
+	if (tsk->thread.trap_nr != UPROBE_TRAP_NR)
+		return true;
+
+	return false;
+}
+
+int arch_uprobe_exception_notify(struct notifier_block *self,
+	unsigned long val, void *data)
+{
+	struct die_args *args = data;
+	struct pt_regs *regs = args->regs;
+
+	/* regs == NULL is a kernel bug */
+	if (WARN_ON(!regs))
+		return NOTIFY_DONE;
+
+	/* We are only interested in userspace traps */
+	if (!user_mode(regs))
+		return NOTIFY_DONE;
+
+	switch (val) {
+	case DIE_BREAK:
+		if (uprobe_pre_sstep_notifier(regs))
+			return NOTIFY_STOP;
+		break;
+	case DIE_UPROBE_XOL:
+		if (uprobe_post_sstep_notifier(regs))
+			return NOTIFY_STOP;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+/*
+ * This function gets called when XOL instruction either gets trapped or
+ * the thread has a fatal signal. Reset the instruction pointer to its
+ * probed address for the potential restart or for post mortem analysis.
+ */
+void arch_uprobe_abort_xol(struct arch_uprobe *aup,
+	struct pt_regs *regs)
+{
+	struct uprobe_task *utask = current->utask;
+
+	instruction_pointer_set(regs, utask->vaddr);
+}
+
+unsigned long arch_uretprobe_hijack_return_addr(
+	unsigned long trampoline_vaddr, struct pt_regs *regs)
+{
+	unsigned long ra;
+
+	ra = regs->regs[31];
+
+	/* Replace the return address with the trampoline address */
+	regs->regs[31] = ra;
+
+	return ra;
+}
+
+/**
+ * set_swbp - store breakpoint at a given address.
+ * @auprobe: arch specific probepoint information.
+ * @mm: the probed process address space.
+ * @vaddr: the virtual address to insert the opcode.
+ *
+ * For mm @mm, store the breakpoint instruction at @vaddr.
+ * Return 0 (success) or a negative errno.
+ *
+ * This version overrides the weak version in kernel/events/uprobes.c.
+ * It is required to handle MIPS16 and microMIPS.
+ */
+int __weak set_swbp(struct arch_uprobe *auprobe, struct mm_struct *mm,
+	unsigned long vaddr)
+{
+	return uprobe_write_opcode(mm, vaddr, UPROBE_SWBP_INSN);
+}
+
+/**
+ * set_orig_insn - Restore the original instruction.
+ * @mm: the probed process address space.
+ * @auprobe: arch specific probepoint information.
+ * @vaddr: the virtual address to insert the opcode.
+ *
+ * For mm @mm, restore the original opcode (opcode) at @vaddr.
+ * Return 0 (success) or a negative errno.
+ *
+ * This overrides the weak version in kernel/events/uprobes.c.
+ */
+int set_orig_insn(struct arch_uprobe *auprobe, struct mm_struct *mm,
+		 unsigned long vaddr)
+{
+	return uprobe_write_opcode(mm, vaddr,
+			*(uprobe_opcode_t *)&auprobe->orig_inst[0].word);
+}
+
+void __weak arch_uprobe_copy_ixol(struct page *page, unsigned long vaddr,
+				  void *src, unsigned long len)
+{
+	void *kaddr;
+
+	/* Initialize the slot */
+	kaddr = kmap_atomic(page);
+	memcpy(kaddr + (vaddr & ~PAGE_MASK), src, len);
+	kunmap_atomic(kaddr);
+
+	/*
+	 * The MIPS version of flush_icache_range will operate safely on
+	 * user space addresses and more importantly, it doesn't require a
+	 * VMA argument.
+	 */
+	flush_icache_range(vaddr, vaddr + len);
+}
+
+/**
+ * uprobe_get_swbp_addr - compute address of swbp given post-swbp regs
+ * @regs: Reflects the saved state of the task after it has hit a breakpoint
+ * instruction.
+ * Return the address of the breakpoint instruction.
+ *
+ * This overrides the weak version in kernel/events/uprobes.c.
+ */
+unsigned long uprobe_get_swbp_addr(struct pt_regs *regs)
+{
+	return instruction_pointer(regs);
+}
+
+/*
+ * See if the instruction can be emulated.
+ * Returns true if instruction was emulated, false otherwise.
+ *
+ * For now we always emulate so this function just returns 0.
+ */
+bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	return 0;
+}
diff --git a/arch/mips/kernel/vpe.c b/arch/mips/kernel/vpe.c
index 11da314..9067b65 100644
--- a/arch/mips/kernel/vpe.c
+++ b/arch/mips/kernel/vpe.c
@@ -817,6 +817,7 @@ static int vpe_open(struct inode *inode, struct file *filp)
 
 static int vpe_release(struct inode *inode, struct file *filp)
 {
+#if defined(CONFIG_MIPS_VPE_LOADER_MT) || defined(CONFIG_MIPS_VPE_LOADER_CMP)
 	struct vpe *v;
 	Elf_Ehdr *hdr;
 	int ret = 0;
@@ -827,7 +828,7 @@ static int vpe_release(struct inode *inode, struct file *filp)
 
 	hdr = (Elf_Ehdr *) v->pbuffer;
 	if (memcmp(hdr->e_ident, ELFMAG, SELFMAG) == 0) {
-		if ((vpe_elfload(v) >= 0) && vpe_run) {
+		if (vpe_elfload(v) >= 0) {
 			vpe_run(v);
 		} else {
 			pr_warn("VPE loader: ELF load failed.\n");
@@ -850,6 +851,10 @@ static int vpe_release(struct inode *inode, struct file *filp)
 	v->plen = 0;
 
 	return ret;
+#else
+	pr_warn("VPE loader: ELF load failed.\n");
+	return -ENOEXEC;
+#endif
 }
 
 static ssize_t vpe_write(struct file *file, const char __user *buffer,
diff --git a/arch/mips/lasat/sysctl.c b/arch/mips/lasat/sysctl.c
index a57959e..c710d96 100644
--- a/arch/mips/lasat/sysctl.c
+++ b/arch/mips/lasat/sysctl.c
@@ -270,4 +270,4 @@ static int __init lasat_register_sysctl(void)
 	return 0;
 }
 
-__initcall(lasat_register_sysctl);
+arch_initcall(lasat_register_sysctl);
diff --git a/arch/mips/lib/dump_tlb.c b/arch/mips/lib/dump_tlb.c
index 167f356..92a3731 100644
--- a/arch/mips/lib/dump_tlb.c
+++ b/arch/mips/lib/dump_tlb.c
@@ -13,6 +13,33 @@
 #include <asm/pgtable.h>
 #include <asm/tlbdebug.h>
 
+void dump_tlb_regs(void)
+{
+	const int field = 2 * sizeof(unsigned long);
+
+	pr_info("Index    : %0x\n", read_c0_index());
+	pr_info("PageMask : %0x\n", read_c0_pagemask());
+	pr_info("EntryHi  : %0*lx\n", field, read_c0_entryhi());
+	pr_info("EntryLo0 : %0*lx\n", field, read_c0_entrylo0());
+	pr_info("EntryLo1 : %0*lx\n", field, read_c0_entrylo1());
+	pr_info("Wired    : %0x\n", read_c0_wired());
+	switch (current_cpu_type()) {
+	case CPU_R10000:
+	case CPU_R12000:
+	case CPU_R14000:
+	case CPU_R16000:
+		pr_info("FrameMask: %0x\n", read_c0_framemask());
+		break;
+	}
+	if (cpu_has_small_pages || cpu_has_rixi || cpu_has_xpa)
+		pr_info("PageGrain: %0x\n", read_c0_pagegrain());
+	if (cpu_has_htw) {
+		pr_info("PWField  : %0*lx\n", field, read_c0_pwfield());
+		pr_info("PWSize   : %0*lx\n", field, read_c0_pwsize());
+		pr_info("PWCtl    : %0x\n", read_c0_pwctl());
+	}
+}
+
 static inline const char *msk2str(unsigned int mask)
 {
 	switch (mask) {
@@ -87,7 +114,7 @@ static void dump_tlb(int first, int last)
 		 * leave only a single G bit set after a machine check exception
 		 * due to duplicate TLB entry.
 		 */
-		if (!((entrylo0 | entrylo1) & MIPS_ENTRYLO_G) &&
+		if (!((entrylo0 | entrylo1) & ENTRYLO_G) &&
 		    (entryhi & 0xff) != asid)
 			continue;
 
@@ -96,8 +123,8 @@ static void dump_tlb(int first, int last)
 		 */
 		printk("Index: %2d pgmask=%s ", i, msk2str(pagemask));
 
-		c0 = (entrylo0 & MIPS_ENTRYLO_C) >> MIPS_ENTRYLO_C_SHIFT;
-		c1 = (entrylo1 & MIPS_ENTRYLO_C) >> MIPS_ENTRYLO_C_SHIFT;
+		c0 = (entrylo0 & ENTRYLO_C) >> ENTRYLO_C_SHIFT;
+		c1 = (entrylo1 & ENTRYLO_C) >> ENTRYLO_C_SHIFT;
 
 		printk("va=%0*lx asid=%02lx\n",
 		       vwidth, (entryhi & ~0x1fffUL),
@@ -114,9 +141,9 @@ static void dump_tlb(int first, int last)
 			       (entrylo0 & MIPS_ENTRYLO_XI) ? 1 : 0);
 		printk("pa=%0*llx c=%d d=%d v=%d g=%d] [",
 		       pwidth, pa, c0,
-		       (entrylo0 & MIPS_ENTRYLO_D) ? 1 : 0,
-		       (entrylo0 & MIPS_ENTRYLO_V) ? 1 : 0,
-		       (entrylo0 & MIPS_ENTRYLO_G) ? 1 : 0);
+		       (entrylo0 & ENTRYLO_D) ? 1 : 0,
+		       (entrylo0 & ENTRYLO_V) ? 1 : 0,
+		       (entrylo0 & ENTRYLO_G) ? 1 : 0);
 		/* RI/XI are in awkward places, so mask them off separately */
 		pa = entrylo1 & ~(MIPS_ENTRYLO_RI | MIPS_ENTRYLO_XI);
 		if (xpa)
@@ -128,9 +155,9 @@ static void dump_tlb(int first, int last)
 			       (entrylo1 & MIPS_ENTRYLO_XI) ? 1 : 0);
 		printk("pa=%0*llx c=%d d=%d v=%d g=%d]\n",
 		       pwidth, pa, c1,
-		       (entrylo1 & MIPS_ENTRYLO_D) ? 1 : 0,
-		       (entrylo1 & MIPS_ENTRYLO_V) ? 1 : 0,
-		       (entrylo1 & MIPS_ENTRYLO_G) ? 1 : 0);
+		       (entrylo1 & ENTRYLO_D) ? 1 : 0,
+		       (entrylo1 & ENTRYLO_V) ? 1 : 0,
+		       (entrylo1 & ENTRYLO_G) ? 1 : 0);
 	}
 	printk("\n");
 
diff --git a/arch/mips/lib/r3k_dump_tlb.c b/arch/mips/lib/r3k_dump_tlb.c
index 8e0d3cf..cfcbb52 100644
--- a/arch/mips/lib/r3k_dump_tlb.c
+++ b/arch/mips/lib/r3k_dump_tlb.c
@@ -14,6 +14,17 @@
 #include <asm/pgtable.h>
 #include <asm/tlbdebug.h>
 
+extern int r3k_have_wired_reg;
+
+void dump_tlb_regs(void)
+{
+	pr_info("Index    : %0x\n", read_c0_index());
+	pr_info("EntryHi  : %0lx\n", read_c0_entryhi());
+	pr_info("EntryLo  : %0lx\n", read_c0_entrylo0());
+	if (r3k_have_wired_reg)
+		pr_info("Wired    : %0x\n", read_c0_wired());
+}
+
 static void dump_tlb(int first, int last)
 {
 	int	i;
diff --git a/arch/mips/loongson32/common/time.c b/arch/mips/loongson32/common/time.c
index df0f850..0996b02 100644
--- a/arch/mips/loongson32/common/time.c
+++ b/arch/mips/loongson32/common/time.c
@@ -126,26 +126,34 @@ static irqreturn_t ls1x_clockevent_isr(int irq, void *devid)
 	return IRQ_HANDLED;
 }
 
-static void ls1x_clockevent_set_mode(enum clock_event_mode mode,
-				     struct clock_event_device *cd)
+static int ls1x_clockevent_set_state_periodic(struct clock_event_device *cd)
 {
 	raw_spin_lock(&ls1x_timer_lock);
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		ls1x_pwmtimer_set_period(ls1x_jiffies_per_tick);
-		ls1x_pwmtimer_restart();
-	case CLOCK_EVT_MODE_RESUME:
-		__raw_writel(INT_EN | CNT_EN, timer_base + PWM_CTRL);
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		__raw_writel(__raw_readl(timer_base + PWM_CTRL) & ~CNT_EN,
-			     timer_base + PWM_CTRL);
-		break;
-	default:
-		break;
-	}
+	ls1x_pwmtimer_set_period(ls1x_jiffies_per_tick);
+	ls1x_pwmtimer_restart();
+	__raw_writel(INT_EN | CNT_EN, timer_base + PWM_CTRL);
 	raw_spin_unlock(&ls1x_timer_lock);
+
+	return 0;
+}
+
+static int ls1x_clockevent_tick_resume(struct clock_event_device *cd)
+{
+	raw_spin_lock(&ls1x_timer_lock);
+	__raw_writel(INT_EN | CNT_EN, timer_base + PWM_CTRL);
+	raw_spin_unlock(&ls1x_timer_lock);
+
+	return 0;
+}
+
+static int ls1x_clockevent_set_state_shutdown(struct clock_event_device *cd)
+{
+	raw_spin_lock(&ls1x_timer_lock);
+	__raw_writel(__raw_readl(timer_base + PWM_CTRL) & ~CNT_EN,
+		     timer_base + PWM_CTRL);
+	raw_spin_unlock(&ls1x_timer_lock);
+
+	return 0;
 }
 
 static int ls1x_clockevent_set_next(unsigned long evt,
@@ -160,12 +168,15 @@ static int ls1x_clockevent_set_next(unsigned long evt,
 }
 
 static struct clock_event_device ls1x_clockevent = {
-	.name		= "ls1x-pwmtimer",
-	.features	= CLOCK_EVT_FEAT_PERIODIC,
-	.rating		= 300,
-	.irq		= LS1X_TIMER_IRQ,
-	.set_next_event	= ls1x_clockevent_set_next,
-	.set_mode	= ls1x_clockevent_set_mode,
+	.name			= "ls1x-pwmtimer",
+	.features		= CLOCK_EVT_FEAT_PERIODIC,
+	.rating			= 300,
+	.irq			= LS1X_TIMER_IRQ,
+	.set_next_event		= ls1x_clockevent_set_next,
+	.set_state_shutdown	= ls1x_clockevent_set_state_shutdown,
+	.set_state_periodic	= ls1x_clockevent_set_state_periodic,
+	.set_state_oneshot	= ls1x_clockevent_set_state_shutdown,
+	.tick_resume		= ls1x_clockevent_tick_resume,
 };
 
 static struct irqaction ls1x_pwmtimer_irqaction = {
diff --git a/arch/mips/loongson64/common/cs5536/cs5536_mfgpt.c b/arch/mips/loongson64/common/cs5536/cs5536_mfgpt.c
index 8750370..da77d41 100644
--- a/arch/mips/loongson64/common/cs5536/cs5536_mfgpt.c
+++ b/arch/mips/loongson64/common/cs5536/cs5536_mfgpt.c
@@ -51,40 +51,36 @@ void enable_mfgpt0_counter(void)
 }
 EXPORT_SYMBOL(enable_mfgpt0_counter);
 
-static void init_mfgpt_timer(enum clock_event_mode mode,
-			     struct clock_event_device *evt)
+static int mfgpt_timer_set_periodic(struct clock_event_device *evt)
 {
 	raw_spin_lock(&mfgpt_lock);
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		outw(COMPARE, MFGPT0_CMP2);	/* set comparator2 */
-		outw(0, MFGPT0_CNT);	/* set counter to 0 */
-		enable_mfgpt0_counter();
-		break;
-
-	case CLOCK_EVT_MODE_SHUTDOWN:
-	case CLOCK_EVT_MODE_UNUSED:
-		if (evt->mode == CLOCK_EVT_MODE_PERIODIC ||
-		    evt->mode == CLOCK_EVT_MODE_ONESHOT)
-			disable_mfgpt0_counter();
-		break;
-
-	case CLOCK_EVT_MODE_ONESHOT:
-		/* The oneshot mode have very high deviation, Not use it! */
-		break;
-
-	case CLOCK_EVT_MODE_RESUME:
-		/* Nothing to do here */
-		break;
-	}
+	outw(COMPARE, MFGPT0_CMP2);	/* set comparator2 */
+	outw(0, MFGPT0_CNT);		/* set counter to 0 */
+	enable_mfgpt0_counter();
+
 	raw_spin_unlock(&mfgpt_lock);
+	return 0;
+}
+
+static int mfgpt_timer_shutdown(struct clock_event_device *evt)
+{
+	if (clockevent_state_periodic(evt) || clockevent_state_oneshot(evt)) {
+		raw_spin_lock(&mfgpt_lock);
+		disable_mfgpt0_counter();
+		raw_spin_unlock(&mfgpt_lock);
+	}
+
+	return 0;
 }
 
 static struct clock_event_device mfgpt_clockevent = {
 	.name = "mfgpt",
 	.features = CLOCK_EVT_FEAT_PERIODIC,
-	.set_mode = init_mfgpt_timer,
+
+	/* The oneshot mode have very high deviation, don't use it! */
+	.set_state_shutdown = mfgpt_timer_shutdown,
+	.set_state_periodic = mfgpt_timer_set_periodic,
 	.irq = CS5536_MFGPT_INTR,
 };
 
diff --git a/arch/mips/loongson64/loongson-3/hpet.c b/arch/mips/loongson64/loongson-3/hpet.c
index 5c21cd3..bf9f1a7 100644
--- a/arch/mips/loongson64/loongson-3/hpet.c
+++ b/arch/mips/loongson64/loongson-3/hpet.c
@@ -78,55 +78,77 @@ static void hpet_enable_legacy_int(void)
 	/* Do nothing on Loongson-3 */
 }
 
-static void hpet_set_mode(enum clock_event_mode mode,
-				struct clock_event_device *evt)
+static int hpet_set_state_periodic(struct clock_event_device *evt)
 {
-	int cfg = 0;
+	int cfg;
 
 	spin_lock(&hpet_lock);
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		pr_info("set clock event to periodic mode!\n");
-		/* stop counter */
-		hpet_stop_counter();
-
-		/* enables the timer0 to generate a periodic interrupt */
-		cfg = hpet_read(HPET_T0_CFG);
-		cfg &= ~HPET_TN_LEVEL;
-		cfg |= HPET_TN_ENABLE | HPET_TN_PERIODIC |
-				HPET_TN_SETVAL | HPET_TN_32BIT;
-		hpet_write(HPET_T0_CFG, cfg);
-
-		/* set the comparator */
-		hpet_write(HPET_T0_CMP, HPET_COMPARE_VAL);
-		udelay(1);
-		hpet_write(HPET_T0_CMP, HPET_COMPARE_VAL);
-
-		/* start counter */
-		hpet_start_counter();
-		break;
-	case CLOCK_EVT_MODE_SHUTDOWN:
-	case CLOCK_EVT_MODE_UNUSED:
-		cfg = hpet_read(HPET_T0_CFG);
-		cfg &= ~HPET_TN_ENABLE;
-		hpet_write(HPET_T0_CFG, cfg);
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-		pr_info("set clock event to one shot mode!\n");
-		cfg = hpet_read(HPET_T0_CFG);
-		/* set timer0 type
-		 * 1 : periodic interrupt
-		 * 0 : non-periodic(oneshot) interrupt
-		 */
-		cfg &= ~HPET_TN_PERIODIC;
-		cfg |= HPET_TN_ENABLE | HPET_TN_32BIT;
-		hpet_write(HPET_T0_CFG, cfg);
-		break;
-	case CLOCK_EVT_MODE_RESUME:
-		hpet_enable_legacy_int();
-		break;
-	}
+
+	pr_info("set clock event to periodic mode!\n");
+	/* stop counter */
+	hpet_stop_counter();
+
+	/* enables the timer0 to generate a periodic interrupt */
+	cfg = hpet_read(HPET_T0_CFG);
+	cfg &= ~HPET_TN_LEVEL;
+	cfg |= HPET_TN_ENABLE | HPET_TN_PERIODIC | HPET_TN_SETVAL |
+		HPET_TN_32BIT;
+	hpet_write(HPET_T0_CFG, cfg);
+
+	/* set the comparator */
+	hpet_write(HPET_T0_CMP, HPET_COMPARE_VAL);
+	udelay(1);
+	hpet_write(HPET_T0_CMP, HPET_COMPARE_VAL);
+
+	/* start counter */
+	hpet_start_counter();
+
+	spin_unlock(&hpet_lock);
+	return 0;
+}
+
+static int hpet_set_state_shutdown(struct clock_event_device *evt)
+{
+	int cfg;
+
+	spin_lock(&hpet_lock);
+
+	cfg = hpet_read(HPET_T0_CFG);
+	cfg &= ~HPET_TN_ENABLE;
+	hpet_write(HPET_T0_CFG, cfg);
+
 	spin_unlock(&hpet_lock);
+	return 0;
+}
+
+static int hpet_set_state_oneshot(struct clock_event_device *evt)
+{
+	int cfg;
+
+	spin_lock(&hpet_lock);
+
+	pr_info("set clock event to one shot mode!\n");
+	cfg = hpet_read(HPET_T0_CFG);
+	/*
+	 * set timer0 type
+	 * 1 : periodic interrupt
+	 * 0 : non-periodic(oneshot) interrupt
+	 */
+	cfg &= ~HPET_TN_PERIODIC;
+	cfg |= HPET_TN_ENABLE | HPET_TN_32BIT;
+	hpet_write(HPET_T0_CFG, cfg);
+
+	spin_unlock(&hpet_lock);
+	return 0;
+}
+
+static int hpet_tick_resume(struct clock_event_device *evt)
+{
+	spin_lock(&hpet_lock);
+	hpet_enable_legacy_int();
+	spin_unlock(&hpet_lock);
+
+	return 0;
 }
 
 static int hpet_next_event(unsigned long delta,
@@ -206,7 +228,10 @@ void __init setup_hpet_timer(void)
 	cd->name = "hpet";
 	cd->rating = 320;
 	cd->features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT;
-	cd->set_mode = hpet_set_mode;
+	cd->set_state_shutdown = hpet_set_state_shutdown;
+	cd->set_state_periodic = hpet_set_state_periodic;
+	cd->set_state_oneshot = hpet_set_state_oneshot;
+	cd->tick_resume = hpet_tick_resume;
 	cd->set_next_event = hpet_next_event;
 	cd->irq = HPET_T0_IRQ;
 	cd->cpumask = cpumask_of(cpu);
diff --git a/arch/mips/math-emu/Makefile b/arch/mips/math-emu/Makefile
index 2e5f962..a19641d 100644
--- a/arch/mips/math-emu/Makefile
+++ b/arch/mips/math-emu/Makefile
@@ -4,9 +4,9 @@
 
 obj-y	+= cp1emu.o ieee754dp.o ieee754sp.o ieee754.o \
 	   dp_div.o dp_mul.o dp_sub.o dp_add.o dp_fsp.o dp_cmp.o dp_simple.o \
-	   dp_tint.o dp_fint.o \
+	   dp_tint.o dp_fint.o dp_maddf.o dp_msubf.o dp_2008class.o dp_fmin.o dp_fmax.o \
 	   sp_div.o sp_mul.o sp_sub.o sp_add.o sp_fdp.o sp_cmp.o sp_simple.o \
-	   sp_tint.o sp_fint.o \
+	   sp_tint.o sp_fint.o sp_maddf.o sp_msubf.o sp_2008class.o sp_fmin.o sp_fmax.o \
 	   dsemul.o
 
 lib-y	+= ieee754d.o \
diff --git a/arch/mips/math-emu/cp1emu.c b/arch/mips/math-emu/cp1emu.c
index 712f17a..32f0e19 100644
--- a/arch/mips/math-emu/cp1emu.c
+++ b/arch/mips/math-emu/cp1emu.c
@@ -1137,7 +1137,7 @@ emul:
 			break;
 
 		case mfhc_op:
-			if (!cpu_has_mips_r2)
+			if (!cpu_has_mips_r2_r6)
 				goto sigill;
 
 			/* copregister rd -> gpr[rt] */
@@ -1148,7 +1148,7 @@ emul:
 			break;
 
 		case mthc_op:
-			if (!cpu_has_mips_r2)
+			if (!cpu_has_mips_r2_r6)
 				goto sigill;
 
 			/* copregister rd <- gpr[rt] */
@@ -1181,6 +1181,24 @@ emul:
 			}
 			break;
 
+		case bc1eqz_op:
+		case bc1nez_op:
+			if (!cpu_has_mips_r6 || delay_slot(xcp))
+				return SIGILL;
+
+			cond = likely = 0;
+			switch (MIPSInst_RS(ir)) {
+			case bc1eqz_op:
+				if (get_fpr32(&current->thread.fpu.fpr[MIPSInst_RT(ir)], 0) & 0x1)
+				    cond = 1;
+				break;
+			case bc1nez_op:
+				if (!(get_fpr32(&current->thread.fpu.fpr[MIPSInst_RT(ir)], 0) & 0x1))
+				    cond = 1;
+				break;
+			}
+			goto branch_common;
+
 		case bc_op:
 			if (delay_slot(xcp))
 				return SIGILL;
@@ -1207,7 +1225,7 @@ emul:
 			case bct_op:
 				break;
 			}
-
+branch_common:
 			set_delay_slot(xcp);
 			if (cond) {
 				/*
@@ -1376,6 +1394,14 @@ static const unsigned char cmptab[8] = {
 	IEEE754_CLT | IEEE754_CEQ | IEEE754_CUN,	/* cmp_ule (sig) cmp_ngt */
 };
 
+static const unsigned char negative_cmptab[8] = {
+	0, /* Reserved */
+	IEEE754_CLT | IEEE754_CGT | IEEE754_CEQ,
+	IEEE754_CLT | IEEE754_CGT | IEEE754_CUN,
+	IEEE754_CLT | IEEE754_CGT,
+	/* Reserved */
+};
+
 
 /*
  * Additional MIPS4 instructions
@@ -1717,6 +1743,126 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx,
 			SPFROMREG(rv.s, MIPSInst_FS(ir));
 			break;
 
+		case fseleqz_op:
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(rv.s, MIPSInst_FT(ir));
+			if (rv.w & 0x1)
+				rv.w = 0;
+			else
+				SPFROMREG(rv.s, MIPSInst_FS(ir));
+			break;
+
+		case fselnez_op:
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(rv.s, MIPSInst_FT(ir));
+			if (rv.w & 0x1)
+				SPFROMREG(rv.s, MIPSInst_FS(ir));
+			else
+				rv.w = 0;
+			break;
+
+		case fmaddf_op: {
+			union ieee754sp ft, fs, fd;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			SPFROMREG(fd, MIPSInst_FD(ir));
+			rv.s = ieee754sp_maddf(fd, fs, ft);
+			break;
+		}
+
+		case fmsubf_op: {
+			union ieee754sp ft, fs, fd;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			SPFROMREG(fd, MIPSInst_FD(ir));
+			rv.s = ieee754sp_msubf(fd, fs, ft);
+			break;
+		}
+
+		case frint_op: {
+			union ieee754sp fs;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.l = ieee754sp_tlong(fs);
+			rv.s = ieee754sp_flong(rv.l);
+			goto copcsr;
+		}
+
+		case fclass_op: {
+			union ieee754sp fs;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.w = ieee754sp_2008class(fs);
+			rfmt = w_fmt;
+			break;
+		}
+
+		case fmin_op: {
+			union ieee754sp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.s = ieee754sp_fmin(fs, ft);
+			break;
+		}
+
+		case fmina_op: {
+			union ieee754sp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.s = ieee754sp_fmina(fs, ft);
+			break;
+		}
+
+		case fmax_op: {
+			union ieee754sp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.s = ieee754sp_fmax(fs, ft);
+			break;
+		}
+
+		case fmaxa_op: {
+			union ieee754sp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			SPFROMREG(ft, MIPSInst_FT(ir));
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			rv.s = ieee754sp_fmaxa(fs, ft);
+			break;
+		}
+
 		case fabs_op:
 			handler.u = ieee754sp_abs;
 			goto scopuop;
@@ -1820,7 +1966,7 @@ copcsr:
 			goto copcsr;
 
 		default:
-			if (MIPSInst_FUNC(ir) >= fcmp_op) {
+			if (!NO_R6EMU && MIPSInst_FUNC(ir) >= fcmp_op) {
 				unsigned cmpop = MIPSInst_FUNC(ir) - fcmp_op;
 				union ieee754sp fs, ft;
 
@@ -1914,6 +2060,127 @@ copcsr:
 				return 0;
 			DPFROMREG(rv.d, MIPSInst_FS(ir));
 			break;
+
+		case fseleqz_op:
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(rv.d, MIPSInst_FT(ir));
+			if (rv.l & 0x1)
+				rv.l = 0;
+			else
+				DPFROMREG(rv.d, MIPSInst_FS(ir));
+			break;
+
+		case fselnez_op:
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(rv.d, MIPSInst_FT(ir));
+			if (rv.l & 0x1)
+				DPFROMREG(rv.d, MIPSInst_FS(ir));
+			else
+				rv.l = 0;
+			break;
+
+		case fmaddf_op: {
+			union ieee754dp ft, fs, fd;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			DPFROMREG(fd, MIPSInst_FD(ir));
+			rv.d = ieee754dp_maddf(fd, fs, ft);
+			break;
+		}
+
+		case fmsubf_op: {
+			union ieee754dp ft, fs, fd;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			DPFROMREG(fd, MIPSInst_FD(ir));
+			rv.d = ieee754dp_msubf(fd, fs, ft);
+			break;
+		}
+
+		case frint_op: {
+			union ieee754dp fs;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.l = ieee754dp_tlong(fs);
+			rv.d = ieee754dp_flong(rv.l);
+			goto copcsr;
+		}
+
+		case fclass_op: {
+			union ieee754dp fs;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.w = ieee754dp_2008class(fs);
+			rfmt = w_fmt;
+			break;
+		}
+
+		case fmin_op: {
+			union ieee754dp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.d = ieee754dp_fmin(fs, ft);
+			break;
+		}
+
+		case fmina_op: {
+			union ieee754dp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.d = ieee754dp_fmina(fs, ft);
+			break;
+		}
+
+		case fmax_op: {
+			union ieee754dp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.d = ieee754dp_fmax(fs, ft);
+			break;
+		}
+
+		case fmaxa_op: {
+			union ieee754dp fs, ft;
+
+			if (!cpu_has_mips_r6)
+				return SIGILL;
+
+			DPFROMREG(ft, MIPSInst_FT(ir));
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			rv.d = ieee754dp_fmaxa(fs, ft);
+			break;
+		}
+
 		case fabs_op:
 			handler.u = ieee754dp_abs;
 			goto dcopuop;
@@ -1997,7 +2264,7 @@ dcopuop:
 			goto copcsr;
 
 		default:
-			if (MIPSInst_FUNC(ir) >= fcmp_op) {
+			if (!NO_R6EMU && MIPSInst_FUNC(ir) >= fcmp_op) {
 				unsigned cmpop = MIPSInst_FUNC(ir) - fcmp_op;
 				union ieee754dp fs, ft;
 
@@ -2021,8 +2288,11 @@ dcopuop:
 			break;
 		}
 		break;
+	}
+
+	case w_fmt: {
+		union ieee754dp fs;
 
-	case w_fmt:
 		switch (MIPSInst_FUNC(ir)) {
 		case fcvts_op:
 			/* convert word to single precision real */
@@ -2036,10 +2306,65 @@ dcopuop:
 			rv.d = ieee754dp_fint(fs.bits);
 			rfmt = d_fmt;
 			goto copcsr;
-		default:
-			return SIGILL;
+		default: {
+			/* Emulating the new CMP.condn.fmt R6 instruction */
+#define CMPOP_MASK	0x7
+#define SIGN_BIT	(0x1 << 3)
+#define PREDICATE_BIT	(0x1 << 4)
+
+			int cmpop = MIPSInst_FUNC(ir) & CMPOP_MASK;
+			int sig = MIPSInst_FUNC(ir) & SIGN_BIT;
+			union ieee754sp fs, ft;
+
+			/* This is an R6 only instruction */
+			if (!cpu_has_mips_r6 ||
+			    (MIPSInst_FUNC(ir) & 0x20))
+				return SIGILL;
+
+			/* fmt is w_fmt for single precision so fix it */
+			rfmt = s_fmt;
+			/* default to false */
+			rv.w = 0;
+
+			/* CMP.condn.S */
+			SPFROMREG(fs, MIPSInst_FS(ir));
+			SPFROMREG(ft, MIPSInst_FT(ir));
+
+			/* positive predicates */
+			if (!(MIPSInst_FUNC(ir) & PREDICATE_BIT)) {
+				if (ieee754sp_cmp(fs, ft, cmptab[cmpop],
+						  sig))
+				    rv.w = -1; /* true, all 1s */
+				if ((sig) &&
+				    ieee754_cxtest(IEEE754_INVALID_OPERATION))
+					rcsr = FPU_CSR_INV_X | FPU_CSR_INV_S;
+				else
+					goto copcsr;
+			} else {
+				/* negative predicates */
+				switch (cmpop) {
+				case 1:
+				case 2:
+				case 3:
+					if (ieee754sp_cmp(fs, ft,
+							  negative_cmptab[cmpop],
+							  sig))
+						rv.w = -1; /* true, all 1s */
+					if (sig &&
+					    ieee754_cxtest(IEEE754_INVALID_OPERATION))
+						rcsr = FPU_CSR_INV_X | FPU_CSR_INV_S;
+					else
+						goto copcsr;
+					break;
+				default:
+					/* Reserved R6 ops */
+					pr_err("Reserved MIPS R6 CMP.condn.S operation\n");
+					return SIGILL;
+				}
+			}
+			break;
+			}
 		}
-		break;
 	}
 
 	case l_fmt:
@@ -2060,11 +2385,60 @@ dcopuop:
 			rv.d = ieee754dp_flong(bits);
 			rfmt = d_fmt;
 			goto copcsr;
-		default:
-			return SIGILL;
-		}
-		break;
+		default: {
+			/* Emulating the new CMP.condn.fmt R6 instruction */
+			int cmpop = MIPSInst_FUNC(ir) & CMPOP_MASK;
+			int sig = MIPSInst_FUNC(ir) & SIGN_BIT;
+			union ieee754dp fs, ft;
+
+			if (!cpu_has_mips_r6 ||
+			    (MIPSInst_FUNC(ir) & 0x20))
+				return SIGILL;
+
+			/* fmt is l_fmt for double precision so fix it */
+			rfmt = d_fmt;
+			/* default to false */
+			rv.l = 0;
 
+			/* CMP.condn.D */
+			DPFROMREG(fs, MIPSInst_FS(ir));
+			DPFROMREG(ft, MIPSInst_FT(ir));
+
+			/* positive predicates */
+			if (!(MIPSInst_FUNC(ir) & PREDICATE_BIT)) {
+				if (ieee754dp_cmp(fs, ft,
+						  cmptab[cmpop], sig))
+				    rv.l = -1LL; /* true, all 1s */
+				if (sig &&
+				    ieee754_cxtest(IEEE754_INVALID_OPERATION))
+					rcsr = FPU_CSR_INV_X | FPU_CSR_INV_S;
+				else
+					goto copcsr;
+			} else {
+				/* negative predicates */
+				switch (cmpop) {
+				case 1:
+				case 2:
+				case 3:
+					if (ieee754dp_cmp(fs, ft,
+							  negative_cmptab[cmpop],
+							  sig))
+						rv.l = -1LL; /* true, all 1s */
+					if (sig &&
+					    ieee754_cxtest(IEEE754_INVALID_OPERATION))
+						rcsr = FPU_CSR_INV_X | FPU_CSR_INV_S;
+					else
+						goto copcsr;
+					break;
+				default:
+					/* Reserved R6 ops */
+					pr_err("Reserved MIPS R6 CMP.condn.D operation\n");
+					return SIGILL;
+				}
+			}
+			break;
+			}
+		}
 	default:
 		return SIGILL;
 	}
diff --git a/arch/mips/math-emu/dp_2008class.c b/arch/mips/math-emu/dp_2008class.c
new file mode 100644
index 0000000..9dc39fc
--- /dev/null
+++ b/arch/mips/math-emu/dp_2008class.c
@@ -0,0 +1,55 @@
+/*
+ * IEEE754 floating point arithmetic
+ * double precision: CLASS.f
+ * FPR[fd] = class(FPR[fs])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754dp.h"
+
+int ieee754dp_2008class(union ieee754dp x)
+{
+	COMPXDP;
+
+	EXPLODEXDP;
+
+	/*
+	 * 10 bit mask as follows:
+	 *
+	 * bit0 = SNAN
+	 * bit1 = QNAN
+	 * bit2 = -INF
+	 * bit3 = -NORM
+	 * bit4 = -DNORM
+	 * bit5 = -ZERO
+	 * bit6 = INF
+	 * bit7 = NORM
+	 * bit8 = DNORM
+	 * bit9 = ZERO
+	 */
+
+	switch(xc) {
+	case IEEE754_CLASS_SNAN:
+		return 0x01;
+	case IEEE754_CLASS_QNAN:
+		return 0x02;
+	case IEEE754_CLASS_INF:
+		return 0x04 << (xs ? 0 : 4);
+	case IEEE754_CLASS_NORM:
+		return 0x08 << (xs ? 0 : 4);
+	case IEEE754_CLASS_DNORM:
+		return 0x10 << (xs ? 0 : 4);
+	case IEEE754_CLASS_ZERO:
+		return 0x20 << (xs ? 0 : 4);
+	default:
+		pr_err("Unknown class: %d\n", xc);
+		return 0;
+	}
+}
diff --git a/arch/mips/math-emu/dp_fmax.c b/arch/mips/math-emu/dp_fmax.c
new file mode 100644
index 0000000..fd71b8d
--- /dev/null
+++ b/arch/mips/math-emu/dp_fmax.c
@@ -0,0 +1,213 @@
+/*
+ * IEEE754 floating point arithmetic
+ * double precision: MIN{,A}.f
+ * MIN : Scalar Floating-Point Minimum
+ * MINA: Scalar Floating-Point argument with Minimum Absolute Value
+ *
+ * MIN.D : FPR[fd] = minNum(FPR[fs],FPR[ft])
+ * MINA.D: FPR[fd] = maxNumMag(FPR[fs],FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754dp.h"
+
+union ieee754dp ieee754dp_fmax(union ieee754dp x, union ieee754dp y)
+{
+	COMPXDP;
+	COMPYDP;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+
+	FLUSHXDP;
+	FLUSHYDP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return xs ? y : x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return ys ? x : y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754dp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		DPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	/* Compare signs */
+	if (xs > ys)
+		return y;
+	else if (xs < ys)
+		return x;
+
+	/* Compare exponent */
+	if (xe > ye)
+		return x;
+	else if (xe < ye)
+		return y;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return y;
+	return x;
+}
+
+union ieee754dp ieee754dp_fmaxa(union ieee754dp x, union ieee754dp y)
+{
+	COMPXDP;
+	COMPYDP;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+
+	FLUSHXDP;
+	FLUSHYDP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754dp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		DPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	/* Compare exponent */
+	if (xe > ye)
+		return x;
+	else if (xe < ye)
+		return y;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return y;
+	return x;
+}
diff --git a/arch/mips/math-emu/dp_fmin.c b/arch/mips/math-emu/dp_fmin.c
new file mode 100644
index 0000000..c1072b0
--- /dev/null
+++ b/arch/mips/math-emu/dp_fmin.c
@@ -0,0 +1,213 @@
+/*
+ * IEEE754 floating point arithmetic
+ * double precision: MIN{,A}.f
+ * MIN : Scalar Floating-Point Minimum
+ * MINA: Scalar Floating-Point argument with Minimum Absolute Value
+ *
+ * MIN.D : FPR[fd] = minNum(FPR[fs],FPR[ft])
+ * MINA.D: FPR[fd] = maxNumMag(FPR[fs],FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754dp.h"
+
+union ieee754dp ieee754dp_fmin(union ieee754dp x, union ieee754dp y)
+{
+	COMPXDP;
+	COMPYDP;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+
+	FLUSHXDP;
+	FLUSHYDP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return xs ? x : y;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return ys ? y : x;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754dp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		DPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	/* Compare signs */
+	if (xs > ys)
+		return x;
+	else if (xs < ys)
+		return y;
+
+	/* Compare exponent */
+	if (xe > ye)
+		return y;
+	else if (xe < ye)
+		return x;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return x;
+	return y;
+}
+
+union ieee754dp ieee754dp_fmina(union ieee754dp x, union ieee754dp y)
+{
+	COMPXDP;
+	COMPYDP;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+
+	FLUSHXDP;
+	FLUSHYDP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754dp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		DPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	/* Compare exponent */
+	if (xe > ye)
+		return y;
+	else if (xe < ye)
+		return x;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return x;
+	return y;
+}
diff --git a/arch/mips/math-emu/dp_maddf.c b/arch/mips/math-emu/dp_maddf.c
new file mode 100644
index 0000000..119eda9
--- /dev/null
+++ b/arch/mips/math-emu/dp_maddf.c
@@ -0,0 +1,265 @@
+/*
+ * IEEE754 floating point arithmetic
+ * double precision: MADDF.f (Fused Multiply Add)
+ * MADDF.fmt: FPR[fd] = FPR[fd] + (FPR[fs] x FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754dp.h"
+
+union ieee754dp ieee754dp_maddf(union ieee754dp z, union ieee754dp x,
+				union ieee754dp y)
+{
+	int re;
+	int rs;
+	u64 rm;
+	unsigned lxm;
+	unsigned hxm;
+	unsigned lym;
+	unsigned hym;
+	u64 lrm;
+	u64 hrm;
+	u64 t;
+	u64 at;
+	int s;
+
+	COMPXDP;
+	COMPYDP;
+
+	u64 zm; int ze; int zs __maybe_unused; int zc;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+	EXPLODEDP(z, zc, zs, ze, zm)
+
+	FLUSHXDP;
+	FLUSHYDP;
+	FLUSHDP(z, zc, zs, ze, zm);
+
+	ieee754_clearcx();
+
+	switch (zc) {
+	case IEEE754_CLASS_SNAN:
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754dp_nanxcpt(z);
+	case IEEE754_CLASS_DNORM:
+		DPDNORMx(zm, ze);
+	/* QNAN is handled separately below */
+	}
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return x;
+
+
+	/*
+	 * Infinity handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754dp_indef();
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		return ieee754dp_inf(xs ^ ys);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		/* Multiplication is 0 so just return z */
+		return z;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		DPDNORMX;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		/* fall through to real computations */
+	}
+
+	/* Finally get to do some computation */
+
+	/*
+	 * Do the multiplication bit first
+	 *
+	 * rm = xm * ym, re = xe + ye basically
+	 *
+	 * At this point xm and ym should have been normalized.
+	 */
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	re = xe + ye;
+	rs = xs ^ ys;
+
+	/* shunt to top of word */
+	xm <<= 64 - (DP_FBITS + 1);
+	ym <<= 64 - (DP_FBITS + 1);
+
+	/*
+	 * Multiply 32 bits xm, ym to give high 32 bits rm with stickness.
+	 */
+
+	/* 32 * 32 => 64 */
+#define DPXMULT(x, y)	((u64)(x) * (u64)y)
+
+	lxm = xm;
+	hxm = xm >> 32;
+	lym = ym;
+	hym = ym >> 32;
+
+	lrm = DPXMULT(lxm, lym);
+	hrm = DPXMULT(hxm, hym);
+
+	t = DPXMULT(lxm, hym);
+
+	at = lrm + (t << 32);
+	hrm += at < lrm;
+	lrm = at;
+
+	hrm = hrm + (t >> 32);
+
+	t = DPXMULT(hxm, lym);
+
+	at = lrm + (t << 32);
+	hrm += at < lrm;
+	lrm = at;
+
+	hrm = hrm + (t >> 32);
+
+	rm = hrm | (lrm != 0);
+
+	/*
+	 * Sticky shift down to normal rounding precision.
+	 */
+	if ((s64) rm < 0) {
+		rm = (rm >> (64 - (DP_FBITS + 1 + 3))) |
+		     ((rm << (DP_FBITS + 1 + 3)) != 0);
+			re++;
+	} else {
+		rm = (rm >> (64 - (DP_FBITS + 1 + 3 + 1))) |
+		     ((rm << (DP_FBITS + 1 + 3 + 1)) != 0);
+	}
+	assert(rm & (DP_HIDDEN_BIT << 3));
+
+	/* And now the addition */
+	assert(zm & DP_HIDDEN_BIT);
+
+	/*
+	 * Provide guard,round and stick bit space.
+	 */
+	zm <<= 3;
+
+	if (ze > re) {
+		/*
+		 * Have to shift y fraction right to align.
+		 */
+		s = ze - re;
+		rm = XDPSRS(rm, s);
+		re += s;
+	} else if (re > ze) {
+		/*
+		 * Have to shift x fraction right to align.
+		 */
+		s = re - ze;
+		zm = XDPSRS(zm, s);
+		ze += s;
+	}
+	assert(ze == re);
+	assert(ze <= DP_EMAX);
+
+	if (zs == rs) {
+		/*
+		 * Generate 28 bit result of adding two 27 bit numbers
+		 * leaving result in xm, xs and xe.
+		 */
+		zm = zm + rm;
+
+		if (zm >> (DP_FBITS + 1 + 3)) { /* carry out */
+			zm = XDPSRS1(zm);
+			ze++;
+		}
+	} else {
+		if (zm >= rm) {
+			zm = zm - rm;
+		} else {
+			zm = rm - zm;
+			zs = rs;
+		}
+		if (zm == 0)
+			return ieee754dp_zero(ieee754_csr.rm == FPU_CSR_RD);
+
+		/*
+		 * Normalize to rounding precision.
+		 */
+		while ((zm >> (DP_FBITS + 3)) == 0) {
+			zm <<= 1;
+			ze--;
+		}
+	}
+
+	return ieee754dp_format(zs, ze, zm);
+}
diff --git a/arch/mips/math-emu/dp_msubf.c b/arch/mips/math-emu/dp_msubf.c
new file mode 100644
index 0000000..1224126
--- /dev/null
+++ b/arch/mips/math-emu/dp_msubf.c
@@ -0,0 +1,269 @@
+/*
+ * IEEE754 floating point arithmetic
+ * double precision: MSUB.f (Fused Multiply Subtract)
+ * MSUBF.fmt: FPR[fd] = FPR[fd] - (FPR[fs] x FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754dp.h"
+
+union ieee754dp ieee754dp_msubf(union ieee754dp z, union ieee754dp x,
+				union ieee754dp y)
+{
+	int re;
+	int rs;
+	u64 rm;
+	unsigned lxm;
+	unsigned hxm;
+	unsigned lym;
+	unsigned hym;
+	u64 lrm;
+	u64 hrm;
+	u64 t;
+	u64 at;
+	int s;
+
+	COMPXDP;
+	COMPYDP;
+
+	u64 zm; int ze; int zs __maybe_unused; int zc;
+
+	EXPLODEXDP;
+	EXPLODEYDP;
+	EXPLODEDP(z, zc, zs, ze, zm)
+
+	FLUSHXDP;
+	FLUSHYDP;
+	FLUSHDP(z, zc, zs, ze, zm);
+
+	ieee754_clearcx();
+
+	switch (zc) {
+	case IEEE754_CLASS_SNAN:
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754dp_nanxcpt(z);
+	case IEEE754_CLASS_DNORM:
+		DPDNORMx(zm, ze);
+	/* QNAN is handled separately below */
+	}
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754dp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754dp_nanxcpt(x);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return x;
+
+
+	/*
+	 * Infinity handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754dp_indef();
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		return ieee754dp_inf(xs ^ ys);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		/* Multiplication is 0 so just return z */
+		return z;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		DPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		DPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		DPDNORMX;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754dp_inf(zs);
+		/* fall through to real computations */
+	}
+
+	/* Finally get to do some computation */
+
+	/*
+	 * Do the multiplication bit first
+	 *
+	 * rm = xm * ym, re = xe + ye basically
+	 *
+	 * At this point xm and ym should have been normalized.
+	 */
+	assert(xm & DP_HIDDEN_BIT);
+	assert(ym & DP_HIDDEN_BIT);
+
+	re = xe + ye;
+	rs = xs ^ ys;
+
+	/* shunt to top of word */
+	xm <<= 64 - (DP_FBITS + 1);
+	ym <<= 64 - (DP_FBITS + 1);
+
+	/*
+	 * Multiply 32 bits xm, ym to give high 32 bits rm with stickness.
+	 */
+
+	/* 32 * 32 => 64 */
+#define DPXMULT(x, y)	((u64)(x) * (u64)y)
+
+	lxm = xm;
+	hxm = xm >> 32;
+	lym = ym;
+	hym = ym >> 32;
+
+	lrm = DPXMULT(lxm, lym);
+	hrm = DPXMULT(hxm, hym);
+
+	t = DPXMULT(lxm, hym);
+
+	at = lrm + (t << 32);
+	hrm += at < lrm;
+	lrm = at;
+
+	hrm = hrm + (t >> 32);
+
+	t = DPXMULT(hxm, lym);
+
+	at = lrm + (t << 32);
+	hrm += at < lrm;
+	lrm = at;
+
+	hrm = hrm + (t >> 32);
+
+	rm = hrm | (lrm != 0);
+
+	/*
+	 * Sticky shift down to normal rounding precision.
+	 */
+	if ((s64) rm < 0) {
+		rm = (rm >> (64 - (DP_FBITS + 1 + 3))) |
+		     ((rm << (DP_FBITS + 1 + 3)) != 0);
+			re++;
+	} else {
+		rm = (rm >> (64 - (DP_FBITS + 1 + 3 + 1))) |
+		     ((rm << (DP_FBITS + 1 + 3 + 1)) != 0);
+	}
+	assert(rm & (DP_HIDDEN_BIT << 3));
+
+	/* And now the subtraction */
+
+	/* flip sign of r and handle as add */
+	rs ^= 1;
+
+	assert(zm & DP_HIDDEN_BIT);
+
+	/*
+	 * Provide guard,round and stick bit space.
+	 */
+	zm <<= 3;
+
+	if (ze > re) {
+		/*
+		 * Have to shift y fraction right to align.
+		 */
+		s = ze - re;
+		rm = XDPSRS(rm, s);
+		re += s;
+	} else if (re > ze) {
+		/*
+		 * Have to shift x fraction right to align.
+		 */
+		s = re - ze;
+		zm = XDPSRS(zm, s);
+		ze += s;
+	}
+	assert(ze == re);
+	assert(ze <= DP_EMAX);
+
+	if (zs == rs) {
+		/*
+		 * Generate 28 bit result of adding two 27 bit numbers
+		 * leaving result in xm, xs and xe.
+		 */
+		zm = zm + rm;
+
+		if (zm >> (DP_FBITS + 1 + 3)) { /* carry out */
+			zm = XDPSRS1(zm);
+			ze++;
+		}
+	} else {
+		if (zm >= rm) {
+			zm = zm - rm;
+		} else {
+			zm = rm - zm;
+			zs = rs;
+		}
+		if (zm == 0)
+			return ieee754dp_zero(ieee754_csr.rm == FPU_CSR_RD);
+
+		/*
+		 * Normalize to rounding precision.
+		 */
+		while ((zm >> (DP_FBITS + 3)) == 0) {
+			zm <<= 1;
+			ze--;
+		}
+	}
+
+	return ieee754dp_format(zs, ze, zm);
+}
diff --git a/arch/mips/math-emu/dsemul.c b/arch/mips/math-emu/dsemul.c
index e0b5cc2..cbb36c1 100644
--- a/arch/mips/math-emu/dsemul.c
+++ b/arch/mips/math-emu/dsemul.c
@@ -33,7 +33,6 @@ struct emuframe {
 
 int mips_dsemul(struct pt_regs *regs, mips_instruction ir, unsigned long cpc)
 {
-	extern asmlinkage void handle_dsemulret(void);
 	struct emuframe __user *fr;
 	int err;
 
diff --git a/arch/mips/math-emu/ieee754.h b/arch/mips/math-emu/ieee754.h
index a5ca108..df94720 100644
--- a/arch/mips/math-emu/ieee754.h
+++ b/arch/mips/math-emu/ieee754.h
@@ -75,6 +75,16 @@ int ieee754sp_cmp(union ieee754sp x, union ieee754sp y, int cop, int sig);
 
 union ieee754sp ieee754sp_sqrt(union ieee754sp x);
 
+union ieee754sp ieee754sp_maddf(union ieee754sp z, union ieee754sp x,
+				union ieee754sp y);
+union ieee754sp ieee754sp_msubf(union ieee754sp z, union ieee754sp x,
+				union ieee754sp y);
+int ieee754sp_2008class(union ieee754sp x);
+union ieee754sp ieee754sp_fmin(union ieee754sp x, union ieee754sp y);
+union ieee754sp ieee754sp_fmina(union ieee754sp x, union ieee754sp y);
+union ieee754sp ieee754sp_fmax(union ieee754sp x, union ieee754sp y);
+union ieee754sp ieee754sp_fmaxa(union ieee754sp x, union ieee754sp y);
+
 /*
  * double precision (often aka double)
 */
@@ -99,6 +109,15 @@ int ieee754dp_cmp(union ieee754dp x, union ieee754dp y, int cop, int sig);
 
 union ieee754dp ieee754dp_sqrt(union ieee754dp x);
 
+union ieee754dp ieee754dp_maddf(union ieee754dp z, union ieee754dp x,
+				union ieee754dp y);
+union ieee754dp ieee754dp_msubf(union ieee754dp z, union ieee754dp x,
+				union ieee754dp y);
+int ieee754dp_2008class(union ieee754dp x);
+union ieee754dp ieee754dp_fmin(union ieee754dp x, union ieee754dp y);
+union ieee754dp ieee754dp_fmina(union ieee754dp x, union ieee754dp y);
+union ieee754dp ieee754dp_fmax(union ieee754dp x, union ieee754dp y);
+union ieee754dp ieee754dp_fmaxa(union ieee754dp x, union ieee754dp y);
 
 
 /* 5 types of floating point number
diff --git a/arch/mips/math-emu/ieee754int.h b/arch/mips/math-emu/ieee754int.h
index 05389d5..6383e2c 100644
--- a/arch/mips/math-emu/ieee754int.h
+++ b/arch/mips/math-emu/ieee754int.h
@@ -65,8 +65,8 @@ static inline int ieee754_class_nan(int xc)
 			vc = IEEE754_CLASS_INF;				\
 		else if (vm & SP_MBIT(SP_FBITS-1))			\
 			vc = IEEE754_CLASS_SNAN;			\
-	else								\
-		vc = IEEE754_CLASS_QNAN;				\
+		else							\
+			vc = IEEE754_CLASS_QNAN;			\
 	} else if (ve == SP_EMIN-1+SP_EBIAS) {				\
 		if (vm) {						\
 			ve = SP_EMIN;					\
@@ -105,8 +105,8 @@ static inline int ieee754_class_nan(int xc)
 		if (vm) {						\
 			ve = DP_EMIN;					\
 			vc = IEEE754_CLASS_DNORM;			\
-	} else								\
-		vc = IEEE754_CLASS_ZERO;				\
+		} else							\
+			vc = IEEE754_CLASS_ZERO;			\
 	} else {							\
 		ve -= DP_EBIAS;						\
 		vm |= DP_HIDDEN_BIT;					\
diff --git a/arch/mips/math-emu/me-debugfs.c b/arch/mips/math-emu/me-debugfs.c
index f308e0f..506a67a 100644
--- a/arch/mips/math-emu/me-debugfs.c
+++ b/arch/mips/math-emu/me-debugfs.c
@@ -65,4 +65,4 @@ do {									\
 
 	return 0;
 }
-__initcall(debugfs_fpuemu);
+arch_initcall(debugfs_fpuemu);
diff --git a/arch/mips/math-emu/sp_2008class.c b/arch/mips/math-emu/sp_2008class.c
new file mode 100644
index 0000000..ff62606
--- /dev/null
+++ b/arch/mips/math-emu/sp_2008class.c
@@ -0,0 +1,55 @@
+/*
+ * IEEE754 floating point arithmetic
+ * single precision: CLASS.f
+ * FPR[fd] = class(FPR[fs])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754sp.h"
+
+int ieee754sp_2008class(union ieee754sp x)
+{
+	COMPXSP;
+
+	EXPLODEXSP;
+
+	/*
+	 * 10 bit mask as follows:
+	 *
+	 * bit0 = SNAN
+	 * bit1 = QNAN
+	 * bit2 = -INF
+	 * bit3 = -NORM
+	 * bit4 = -DNORM
+	 * bit5 = -ZERO
+	 * bit6 = INF
+	 * bit7 = NORM
+	 * bit8 = DNORM
+	 * bit9 = ZERO
+	 */
+
+	switch(xc) {
+	case IEEE754_CLASS_SNAN:
+		return 0x01;
+	case IEEE754_CLASS_QNAN:
+		return 0x02;
+	case IEEE754_CLASS_INF:
+		return 0x04 << (xs ? 0 : 4);
+	case IEEE754_CLASS_NORM:
+		return 0x08 << (xs ? 0 : 4);
+	case IEEE754_CLASS_DNORM:
+		return 0x10 << (xs ? 0 : 4);
+	case IEEE754_CLASS_ZERO:
+		return 0x20 << (xs ? 0 : 4);
+	default:
+		pr_err("Unknown class: %d\n", xc);
+		return 0;
+	}
+}
diff --git a/arch/mips/math-emu/sp_fmax.c b/arch/mips/math-emu/sp_fmax.c
new file mode 100644
index 0000000..4d00084
--- /dev/null
+++ b/arch/mips/math-emu/sp_fmax.c
@@ -0,0 +1,213 @@
+/*
+ * IEEE754 floating point arithmetic
+ * single precision: MAX{,A}.f
+ * MAX : Scalar Floating-Point Maximum
+ * MAXA: Scalar Floating-Point argument with Maximum Absolute Value
+ *
+ * MAX.S : FPR[fd] = maxNum(FPR[fs],FPR[ft])
+ * MAXA.S: FPR[fd] = maxNumMag(FPR[fs],FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754sp.h"
+
+union ieee754sp ieee754sp_fmax(union ieee754sp x, union ieee754sp y)
+{
+	COMPXSP;
+	COMPYSP;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+
+	FLUSHXSP;
+	FLUSHYSP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return xs ? y : x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return ys ? x : y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754sp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		SPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	/* Compare signs */
+	if (xs > ys)
+		return y;
+	else if (xs < ys)
+		return x;
+
+	/* Compare exponent */
+	if (xe > ye)
+		return x;
+	else if (xe < ye)
+		return y;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return y;
+	return x;
+}
+
+union ieee754sp ieee754sp_fmaxa(union ieee754sp x, union ieee754sp y)
+{
+	COMPXSP;
+	COMPYSP;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+
+	FLUSHXSP;
+	FLUSHYSP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754sp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		SPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	/* Compare exponent */
+	if (xe > ye)
+		return x;
+	else if (xe < ye)
+		return y;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return y;
+	return x;
+}
diff --git a/arch/mips/math-emu/sp_fmin.c b/arch/mips/math-emu/sp_fmin.c
new file mode 100644
index 0000000..4eb1bb9
--- /dev/null
+++ b/arch/mips/math-emu/sp_fmin.c
@@ -0,0 +1,213 @@
+/*
+ * IEEE754 floating point arithmetic
+ * single precision: MIN{,A}.f
+ * MIN : Scalar Floating-Point Minimum
+ * MINA: Scalar Floating-Point argument with Minimum Absolute Value
+ *
+ * MIN.S : FPR[fd] = minNum(FPR[fs],FPR[ft])
+ * MINA.S: FPR[fd] = maxNumMag(FPR[fs],FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754sp.h"
+
+union ieee754sp ieee754sp_fmin(union ieee754sp x, union ieee754sp y)
+{
+	COMPXSP;
+	COMPYSP;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+
+	FLUSHXSP;
+	FLUSHYSP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return xs ? x : y;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return ys ? y : x;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754sp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		SPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	/* Compare signs */
+	if (xs > ys)
+		return x;
+	else if (xs < ys)
+		return y;
+
+	/* Compare exponent */
+	if (xe > ye)
+		return y;
+	else if (xe < ye)
+		return x;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return x;
+	return y;
+}
+
+union ieee754sp ieee754sp_fmina(union ieee754sp x, union ieee754sp y)
+{
+	COMPXSP;
+	COMPYSP;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+
+	FLUSHXSP;
+	FLUSHYSP;
+
+	ieee754_clearcx();
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	/* numbers are preferred to NaNs */
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return y;
+
+	/*
+	 * Infinity and zero handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		return x;
+
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+		if (xs == ys)
+			return x;
+		return ieee754sp_zero(1);
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		SPDNORMX;
+	}
+
+	/* Finally get to do some computation */
+
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	/* Compare exponent */
+	if (xe > ye)
+		return y;
+	else if (xe < ye)
+		return x;
+
+	/* Compare mantissa */
+	if (xm <= ym)
+		return x;
+	return y;
+}
diff --git a/arch/mips/math-emu/sp_maddf.c b/arch/mips/math-emu/sp_maddf.c
new file mode 100644
index 0000000..dd1dd83
--- /dev/null
+++ b/arch/mips/math-emu/sp_maddf.c
@@ -0,0 +1,255 @@
+/*
+ * IEEE754 floating point arithmetic
+ * single precision: MADDF.f (Fused Multiply Add)
+ * MADDF.fmt: FPR[fd] = FPR[fd] + (FPR[fs] x FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754sp.h"
+
+union ieee754sp ieee754sp_maddf(union ieee754sp z, union ieee754sp x,
+				union ieee754sp y)
+{
+	int re;
+	int rs;
+	unsigned rm;
+	unsigned short lxm;
+	unsigned short hxm;
+	unsigned short lym;
+	unsigned short hym;
+	unsigned lrm;
+	unsigned hrm;
+	unsigned t;
+	unsigned at;
+	int s;
+
+	COMPXSP;
+	COMPYSP;
+	u32 zm; int ze; int zs __maybe_unused; int zc;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+	EXPLODESP(z, zc, zs, ze, zm)
+
+	FLUSHXSP;
+	FLUSHYSP;
+	FLUSHSP(z, zc, zs, ze, zm);
+
+	ieee754_clearcx();
+
+	switch (zc) {
+	case IEEE754_CLASS_SNAN:
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754sp_nanxcpt(z);
+	case IEEE754_CLASS_DNORM:
+		SPDNORMx(zm, ze);
+	/* QNAN is handled separately below */
+	}
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return x;
+
+	/*
+	 * Infinity handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754sp_indef();
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		return ieee754sp_inf(xs ^ ys);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		/* Multiplication is 0 so just return z */
+		return z;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		SPDNORMX;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		/* fall through to real computations */
+	}
+
+	/* Finally get to do some computation */
+
+	/*
+	 * Do the multiplication bit first
+	 *
+	 * rm = xm * ym, re = xe + ye basically
+	 *
+	 * At this point xm and ym should have been normalized.
+	 */
+
+	/* rm = xm * ym, re = xe+ye basically */
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	re = xe + ye;
+	rs = xs ^ ys;
+
+	/* shunt to top of word */
+	xm <<= 32 - (SP_FBITS + 1);
+	ym <<= 32 - (SP_FBITS + 1);
+
+	/*
+	 * Multiply 32 bits xm, ym to give high 32 bits rm with stickness.
+	 */
+	lxm = xm & 0xffff;
+	hxm = xm >> 16;
+	lym = ym & 0xffff;
+	hym = ym >> 16;
+
+	lrm = lxm * lym;	/* 16 * 16 => 32 */
+	hrm = hxm * hym;	/* 16 * 16 => 32 */
+
+	t = lxm * hym; /* 16 * 16 => 32 */
+	at = lrm + (t << 16);
+	hrm += at < lrm;
+	lrm = at;
+	hrm = hrm + (t >> 16);
+
+	t = hxm * lym; /* 16 * 16 => 32 */
+	at = lrm + (t << 16);
+	hrm += at < lrm;
+	lrm = at;
+	hrm = hrm + (t >> 16);
+
+	rm = hrm | (lrm != 0);
+
+	/*
+	 * Sticky shift down to normal rounding precision.
+	 */
+	if ((int) rm < 0) {
+		rm = (rm >> (32 - (SP_FBITS + 1 + 3))) |
+		    ((rm << (SP_FBITS + 1 + 3)) != 0);
+		re++;
+	} else {
+		rm = (rm >> (32 - (SP_FBITS + 1 + 3 + 1))) |
+		     ((rm << (SP_FBITS + 1 + 3 + 1)) != 0);
+	}
+	assert(rm & (SP_HIDDEN_BIT << 3));
+
+	/* And now the addition */
+
+	assert(zm & SP_HIDDEN_BIT);
+
+	/*
+	 * Provide guard,round and stick bit space.
+	 */
+	zm <<= 3;
+
+	if (ze > re) {
+		/*
+		 * Have to shift y fraction right to align.
+		 */
+		s = ze - re;
+		SPXSRSYn(s);
+	} else if (re > ze) {
+		/*
+		 * Have to shift x fraction right to align.
+		 */
+		s = re - ze;
+		SPXSRSYn(s);
+	}
+	assert(ze == re);
+	assert(ze <= SP_EMAX);
+
+	if (zs == rs) {
+		/*
+		 * Generate 28 bit result of adding two 27 bit numbers
+		 * leaving result in zm, zs and ze.
+		 */
+		zm = zm + rm;
+
+		if (zm >> (SP_FBITS + 1 + 3)) { /* carry out */
+			SPXSRSX1();
+		}
+	} else {
+		if (zm >= rm) {
+			zm = zm - rm;
+		} else {
+			zm = rm - zm;
+			zs = rs;
+		}
+		if (zm == 0)
+			return ieee754sp_zero(ieee754_csr.rm == FPU_CSR_RD);
+
+		/*
+		 * Normalize in extended single precision
+		 */
+		while ((zm >> (SP_MBITS + 3)) == 0) {
+			zm <<= 1;
+			ze--;
+		}
+
+	}
+	return ieee754sp_format(zs, ze, zm);
+}
diff --git a/arch/mips/math-emu/sp_msubf.c b/arch/mips/math-emu/sp_msubf.c
new file mode 100644
index 0000000..81c38b9
--- /dev/null
+++ b/arch/mips/math-emu/sp_msubf.c
@@ -0,0 +1,258 @@
+/*
+ * IEEE754 floating point arithmetic
+ * single precision: MSUB.f (Fused Multiply Subtract)
+ * MSUBF.fmt: FPR[fd] = FPR[fd] - (FPR[fs] x FPR[ft])
+ *
+ * MIPS floating point support
+ * Copyright (C) 2015 Imagination Technologies, Ltd.
+ * Author: Markos Chandras <markos.chandras@imgtec.com>
+ *
+ *  This program is free software; you can distribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; version 2 of the License.
+ */
+
+#include "ieee754sp.h"
+
+union ieee754sp ieee754sp_msubf(union ieee754sp z, union ieee754sp x,
+				union ieee754sp y)
+{
+	int re;
+	int rs;
+	unsigned rm;
+	unsigned short lxm;
+	unsigned short hxm;
+	unsigned short lym;
+	unsigned short hym;
+	unsigned lrm;
+	unsigned hrm;
+	unsigned t;
+	unsigned at;
+	int s;
+
+	COMPXSP;
+	COMPYSP;
+	u32 zm; int ze; int zs __maybe_unused; int zc;
+
+	EXPLODEXSP;
+	EXPLODEYSP;
+	EXPLODESP(z, zc, zs, ze, zm)
+
+	FLUSHXSP;
+	FLUSHYSP;
+	FLUSHSP(z, zc, zs, ze, zm);
+
+	ieee754_clearcx();
+
+	switch (zc) {
+	case IEEE754_CLASS_SNAN:
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754sp_nanxcpt(z);
+	case IEEE754_CLASS_DNORM:
+		SPDNORMx(zm, ze);
+	/* QNAN is handled separately below */
+	}
+
+	switch (CLPAIR(xc, yc)) {
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_SNAN):
+		return ieee754sp_nanxcpt(y);
+
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_SNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF):
+		return ieee754sp_nanxcpt(x);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_QNAN):
+		return y;
+
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_QNAN):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_QNAN, IEEE754_CLASS_INF):
+		return x;
+
+	/*
+	 * Infinity handling
+	 */
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		ieee754_setcx(IEEE754_INVALID_OPERATION);
+		return ieee754sp_indef();
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		return ieee754sp_inf(xs ^ ys);
+
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM):
+	case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM):
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO):
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO):
+		if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		/* Multiplication is 0 so just return z */
+		return z;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM):
+		SPDNORMX;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		SPDNORMY;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		SPDNORMX;
+		break;
+
+	case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_NORM):
+		if (zc == IEEE754_CLASS_QNAN)
+			return z;
+		else if (zc == IEEE754_CLASS_INF)
+			return ieee754sp_inf(zs);
+		/* fall through to real compuation */
+	}
+
+	/* Finally get to do some computation */
+
+	/*
+	 * Do the multiplication bit first
+	 *
+	 * rm = xm * ym, re = xe + ye basically
+	 *
+	 * At this point xm and ym should have been normalized.
+	 */
+
+	/* rm = xm * ym, re = xe+ye basically */
+	assert(xm & SP_HIDDEN_BIT);
+	assert(ym & SP_HIDDEN_BIT);
+
+	re = xe + ye;
+	rs = xs ^ ys;
+
+	/* shunt to top of word */
+	xm <<= 32 - (SP_FBITS + 1);
+	ym <<= 32 - (SP_FBITS + 1);
+
+	/*
+	 * Multiply 32 bits xm, ym to give high 32 bits rm with stickness.
+	 */
+	lxm = xm & 0xffff;
+	hxm = xm >> 16;
+	lym = ym & 0xffff;
+	hym = ym >> 16;
+
+	lrm = lxm * lym;	/* 16 * 16 => 32 */
+	hrm = hxm * hym;	/* 16 * 16 => 32 */
+
+	t = lxm * hym; /* 16 * 16 => 32 */
+	at = lrm + (t << 16);
+	hrm += at < lrm;
+	lrm = at;
+	hrm = hrm + (t >> 16);
+
+	t = hxm * lym; /* 16 * 16 => 32 */
+	at = lrm + (t << 16);
+	hrm += at < lrm;
+	lrm = at;
+	hrm = hrm + (t >> 16);
+
+	rm = hrm | (lrm != 0);
+
+	/*
+	 * Sticky shift down to normal rounding precision.
+	 */
+	if ((int) rm < 0) {
+		rm = (rm >> (32 - (SP_FBITS + 1 + 3))) |
+		    ((rm << (SP_FBITS + 1 + 3)) != 0);
+		re++;
+	} else {
+		rm = (rm >> (32 - (SP_FBITS + 1 + 3 + 1))) |
+		     ((rm << (SP_FBITS + 1 + 3 + 1)) != 0);
+	}
+	assert(rm & (SP_HIDDEN_BIT << 3));
+
+	/* And now the subtraction */
+
+	/* Flip sign of r and handle as add */
+	rs ^= 1;
+
+	assert(zm & SP_HIDDEN_BIT);
+
+	/*
+	 * Provide guard,round and stick bit space.
+	 */
+	zm <<= 3;
+
+	if (ze > re) {
+		/*
+		 * Have to shift y fraction right to align.
+		 */
+		s = ze - re;
+		SPXSRSYn(s);
+	} else if (re > ze) {
+		/*
+		 * Have to shift x fraction right to align.
+		 */
+		s = re - ze;
+		SPXSRSYn(s);
+	}
+	assert(ze == re);
+	assert(ze <= SP_EMAX);
+
+	if (zs == rs) {
+		/*
+		 * Generate 28 bit result of adding two 27 bit numbers
+		 * leaving result in zm, zs and ze.
+		 */
+		zm = zm + rm;
+
+		if (zm >> (SP_FBITS + 1 + 3)) { /* carry out */
+			SPXSRSX1(); /* shift preserving sticky */
+		}
+	} else {
+		if (zm >= rm) {
+			zm = zm - rm;
+		} else {
+			zm = rm - zm;
+			zs = rs;
+		}
+		if (zm == 0)
+			return ieee754sp_zero(ieee754_csr.rm == FPU_CSR_RD);
+
+		/*
+		 * Normalize in extended single precision
+		 */
+		while ((zm >> (SP_MBITS + 3)) == 0) {
+			zm <<= 1;
+			ze--;
+		}
+
+	}
+	return ieee754sp_format(zs, ze, zm);
+}
diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
index fbea443..5d3a25e 100644
--- a/arch/mips/mm/c-r4k.c
+++ b/arch/mips/mm/c-r4k.c
@@ -1276,6 +1276,7 @@ static void probe_pcache(void)
 	case CPU_PROAPTIV:
 	case CPU_M5150:
 	case CPU_QEMU_GENERIC:
+	case CPU_I6400:
 		if (!(read_c0_config7() & MIPS_CONF7_IAR) &&
 		    (c->icache.waysize > PAGE_SIZE))
 			c->icache.flags |= MIPS_CACHE_ALIASES;
diff --git a/arch/mips/mm/dma-default.c b/arch/mips/mm/dma-default.c
index eeaf024..8f23cf0 100644
--- a/arch/mips/mm/dma-default.c
+++ b/arch/mips/mm/dma-default.c
@@ -194,6 +194,40 @@ static void mips_dma_free_coherent(struct device *dev, size_t size, void *vaddr,
 		__free_pages(page, get_order(size));
 }
 
+static int mips_dma_mmap(struct device *dev, struct vm_area_struct *vma,
+	void *cpu_addr, dma_addr_t dma_addr, size_t size,
+	struct dma_attrs *attrs)
+{
+	unsigned long user_count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
+	unsigned long count = PAGE_ALIGN(size) >> PAGE_SHIFT;
+	unsigned long addr = (unsigned long)cpu_addr;
+	unsigned long off = vma->vm_pgoff;
+	unsigned long pfn;
+	int ret = -ENXIO;
+
+	if (!plat_device_is_coherent(dev) && !hw_coherentio)
+		addr = CAC_ADDR(addr);
+
+	pfn = page_to_pfn(virt_to_page((void *)addr));
+
+	if (dma_get_attr(DMA_ATTR_WRITE_COMBINE, attrs))
+		vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
+	else
+		vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+
+	if (dma_mmap_from_coherent(dev, vma, cpu_addr, size, &ret))
+		return ret;
+
+	if (off < count && user_count <= (count - off)) {
+		ret = remap_pfn_range(vma, vma->vm_start,
+				      pfn + off,
+				      user_count << PAGE_SHIFT,
+				      vma->vm_page_prot);
+	}
+
+	return ret;
+}
+
 static inline void __dma_sync_virtual(void *addr, size_t size,
 	enum dma_data_direction direction)
 {
@@ -380,6 +414,7 @@ EXPORT_SYMBOL(dma_cache_sync);
 static struct dma_map_ops mips_default_dma_map_ops = {
 	.alloc = mips_dma_alloc_coherent,
 	.free = mips_dma_free_coherent,
+	.mmap = mips_dma_mmap,
 	.map_page = mips_dma_map_page,
 	.unmap_page = mips_dma_unmap_page,
 	.map_sg = mips_dma_map_sg,
diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c
index 852a41c..4b88fa0 100644
--- a/arch/mips/mm/fault.c
+++ b/arch/mips/mm/fault.c
@@ -57,12 +57,10 @@ static void __kprobes __do_page_fault(struct pt_regs *regs, unsigned long write,
 
 #ifdef CONFIG_KPROBES
 	/*
-	 * This is to notify the fault handler of the kprobes.	The
-	 * exception code is redundant as it is also carried in REGS,
-	 * but we pass it anyhow.
+	 * This is to notify the fault handler of the kprobes.
 	 */
 	if (notify_die(DIE_PAGE_FAULT, "page fault", regs, -1,
-		       (regs->cp0_cause >> 2) & 0x1f, SIGSEGV) == NOTIFY_STOP)
+		       current->thread.trap_nr, SIGSEGV) == NOTIFY_STOP)
 		return;
 #endif
 
@@ -224,6 +222,7 @@ bad_area_nosemaphore:
 			print_vma_addr(" ", regs->regs[31]);
 			pr_info("\n");
 		}
+		current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 		info.si_signo = SIGSEGV;
 		info.si_errno = 0;
 		/* info.si_code has been set above */
@@ -282,6 +281,7 @@ do_sigbus:
 		       field, (unsigned long) regs->cp0_epc,
 		       field, (unsigned long) regs->regs[31]);
 #endif
+	current->thread.trap_nr = (regs->cp0_cause >> 2) & 0x1f;
 	tsk->thread.cp0_badvaddr = address;
 	info.si_signo = SIGBUS;
 	info.si_errno = 0;
diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c
index 198a314..66d0f49 100644
--- a/arch/mips/mm/init.c
+++ b/arch/mips/mm/init.c
@@ -37,6 +37,7 @@
 #include <asm/cpu.h>
 #include <asm/dma.h>
 #include <asm/kmap_types.h>
+#include <asm/maar.h>
 #include <asm/mmu_context.h>
 #include <asm/sections.h>
 #include <asm/pgtable.h>
@@ -333,9 +334,40 @@ static inline void mem_init_free_highmem(void)
 #endif
 }
 
-unsigned __weak platform_maar_init(unsigned num_maars)
+unsigned __weak platform_maar_init(unsigned num_pairs)
 {
-	return 0;
+	struct maar_config cfg[BOOT_MEM_MAP_MAX];
+	unsigned i, num_configured, num_cfg = 0;
+	phys_addr_t skip;
+
+	for (i = 0; i < boot_mem_map.nr_map; i++) {
+		switch (boot_mem_map.map[i].type) {
+		case BOOT_MEM_RAM:
+		case BOOT_MEM_INIT_RAM:
+			break;
+		default:
+			continue;
+		}
+
+		skip = 0x10000 - (boot_mem_map.map[i].addr & 0xffff);
+
+		cfg[num_cfg].lower = boot_mem_map.map[i].addr;
+		cfg[num_cfg].lower += skip;
+
+		cfg[num_cfg].upper = cfg[num_cfg].lower;
+		cfg[num_cfg].upper += boot_mem_map.map[i].size - 1;
+		cfg[num_cfg].upper -= skip;
+
+		cfg[num_cfg].attrs = MIPS_MAAR_S;
+		num_cfg++;
+	}
+
+	num_configured = maar_config(cfg, num_cfg, num_pairs);
+	if (num_configured < num_cfg)
+		pr_warn("Not enough MAAR pairs (%u) for all bootmem regions (%u)\n",
+			num_pairs, num_cfg);
+
+	return num_configured;
 }
 
 static void maar_init(void)
diff --git a/arch/mips/mm/sc-mips.c b/arch/mips/mm/sc-mips.c
index 4ceafd1..53ea839 100644
--- a/arch/mips/mm/sc-mips.c
+++ b/arch/mips/mm/sc-mips.c
@@ -14,6 +14,7 @@
 #include <asm/pgtable.h>
 #include <asm/mmu_context.h>
 #include <asm/r4kcache.h>
+#include <asm/mips-cm.h>
 
 /*
  * MIPS32/MIPS64 L2 cache handling
@@ -94,6 +95,38 @@ static inline int mips_sc_is_activated(struct cpuinfo_mips *c)
 	return 1;
 }
 
+static int __init mips_sc_probe_cm3(void)
+{
+	struct cpuinfo_mips *c = &current_cpu_data;
+	unsigned long cfg = read_gcr_l2_config();
+	unsigned long sets, line_sz, assoc;
+
+	if (cfg & CM_GCR_L2_CONFIG_BYPASS_MSK)
+		return 0;
+
+	sets = cfg & CM_GCR_L2_CONFIG_SET_SIZE_MSK;
+	sets >>= CM_GCR_L2_CONFIG_SET_SIZE_SHF;
+	c->scache.sets = 64 << sets;
+
+	line_sz = cfg & CM_GCR_L2_CONFIG_LINE_SIZE_MSK;
+	line_sz >>= CM_GCR_L2_CONFIG_LINE_SIZE_SHF;
+	c->scache.linesz = 2 << line_sz;
+
+	assoc = cfg & CM_GCR_L2_CONFIG_ASSOC_MSK;
+	assoc >>= CM_GCR_L2_CONFIG_ASSOC_SHF;
+	c->scache.ways = assoc + 1;
+	c->scache.waysize = c->scache.sets * c->scache.linesz;
+	c->scache.waybit = __ffs(c->scache.waysize);
+
+	c->scache.flags &= ~MIPS_CACHE_NOT_PRESENT;
+
+	return 1;
+}
+
+void __weak platform_early_l2_init(void)
+{
+}
+
 static inline int __init mips_sc_probe(void)
 {
 	struct cpuinfo_mips *c = &current_cpu_data;
@@ -103,6 +136,15 @@ static inline int __init mips_sc_probe(void)
 	/* Mark as not present until probe completed */
 	c->scache.flags |= MIPS_CACHE_NOT_PRESENT;
 
+	/*
+	 * Do we need some platform specific probing before
+	 * we configure L2?
+	 */
+	platform_early_l2_init();
+
+	if (mips_cm_revision() >= CM_REV_CM3)
+		return mips_sc_probe_cm3();
+
 	/* Ignore anything but MIPSxx processors */
 	if (!(c->isa_level & (MIPS_CPU_ISA_M32R1 | MIPS_CPU_ISA_M32R2 |
 			      MIPS_CPU_ISA_M32R6 | MIPS_CPU_ISA_M64R1 |
diff --git a/arch/mips/mm/tlb-r3k.c b/arch/mips/mm/tlb-r3k.c
index 2b75b8f..b4f366f 100644
--- a/arch/mips/mm/tlb-r3k.c
+++ b/arch/mips/mm/tlb-r3k.c
@@ -36,7 +36,7 @@ extern void build_tlb_refill_handler(void);
 		"nop\n\t"		\
 		".set	pop\n\t")
 
-static int r3k_have_wired_reg;			/* Should be in cpu_data? */
+int r3k_have_wired_reg;			/* Should be in cpu_data? */
 
 /* TLB operations. */
 static void local_flush_tlb_from(int entry)
diff --git a/arch/mips/mti-malta/malta-init.c b/arch/mips/mti-malta/malta-init.c
index cec3e18..53c2478 100644
--- a/arch/mips/mti-malta/malta-init.c
+++ b/arch/mips/mti-malta/malta-init.c
@@ -303,3 +303,10 @@ mips_pci_controller:
 	if (!register_vsmp_smp_ops())
 		return;
 }
+
+void platform_early_l2_init(void)
+{
+	/* L2 configuration lives in the CM3 */
+	if (mips_cm_revision() >= CM_REV_CM3)
+		mips_cm_probe();
+}
diff --git a/arch/mips/mti-malta/malta-int.c b/arch/mips/mti-malta/malta-int.c
index fa8f591..c6a6c7a 100644
--- a/arch/mips/mti-malta/malta-int.c
+++ b/arch/mips/mti-malta/malta-int.c
@@ -382,122 +382,12 @@ void malta_be_init(void)
 	/* Could change CM error mask register. */
 }
 
-
-static char *tr[8] = {
-	"mem",	"gcr",	"gic",	"mmio",
-	"0x04", "0x05", "0x06", "0x07"
-};
-
-static char *mcmd[32] = {
-	[0x00] = "0x00",
-	[0x01] = "Legacy Write",
-	[0x02] = "Legacy Read",
-	[0x03] = "0x03",
-	[0x04] = "0x04",
-	[0x05] = "0x05",
-	[0x06] = "0x06",
-	[0x07] = "0x07",
-	[0x08] = "Coherent Read Own",
-	[0x09] = "Coherent Read Share",
-	[0x0a] = "Coherent Read Discard",
-	[0x0b] = "Coherent Ready Share Always",
-	[0x0c] = "Coherent Upgrade",
-	[0x0d] = "Coherent Writeback",
-	[0x0e] = "0x0e",
-	[0x0f] = "0x0f",
-	[0x10] = "Coherent Copyback",
-	[0x11] = "Coherent Copyback Invalidate",
-	[0x12] = "Coherent Invalidate",
-	[0x13] = "Coherent Write Invalidate",
-	[0x14] = "Coherent Completion Sync",
-	[0x15] = "0x15",
-	[0x16] = "0x16",
-	[0x17] = "0x17",
-	[0x18] = "0x18",
-	[0x19] = "0x19",
-	[0x1a] = "0x1a",
-	[0x1b] = "0x1b",
-	[0x1c] = "0x1c",
-	[0x1d] = "0x1d",
-	[0x1e] = "0x1e",
-	[0x1f] = "0x1f"
-};
-
-static char *core[8] = {
-	"Invalid/OK",	"Invalid/Data",
-	"Shared/OK",	"Shared/Data",
-	"Modified/OK",	"Modified/Data",
-	"Exclusive/OK", "Exclusive/Data"
-};
-
-static char *causes[32] = {
-	"None", "GC_WR_ERR", "GC_RD_ERR", "COH_WR_ERR",
-	"COH_RD_ERR", "MMIO_WR_ERR", "MMIO_RD_ERR", "0x07",
-	"0x08", "0x09", "0x0a", "0x0b",
-	"0x0c", "0x0d", "0x0e", "0x0f",
-	"0x10", "0x11", "0x12", "0x13",
-	"0x14", "0x15", "0x16", "INTVN_WR_ERR",
-	"INTVN_RD_ERR", "0x19", "0x1a", "0x1b",
-	"0x1c", "0x1d", "0x1e", "0x1f"
-};
-
 int malta_be_handler(struct pt_regs *regs, int is_fixup)
 {
 	/* This duplicates the handling in do_be which seems wrong */
 	int retval = is_fixup ? MIPS_BE_FIXUP : MIPS_BE_FATAL;
 
-	if (mips_cm_present()) {
-		unsigned long cm_error = read_gcr_error_cause();
-		unsigned long cm_addr = read_gcr_error_addr();
-		unsigned long cm_other = read_gcr_error_mult();
-		unsigned long cause, ocause;
-		char buf[256];
-
-		cause = cm_error & CM_GCR_ERROR_CAUSE_ERRTYPE_MSK;
-		if (cause != 0) {
-			cause >>= CM_GCR_ERROR_CAUSE_ERRTYPE_SHF;
-			if (cause < 16) {
-				unsigned long cca_bits = (cm_error >> 15) & 7;
-				unsigned long tr_bits = (cm_error >> 12) & 7;
-				unsigned long cmd_bits = (cm_error >> 7) & 0x1f;
-				unsigned long stag_bits = (cm_error >> 3) & 15;
-				unsigned long sport_bits = (cm_error >> 0) & 7;
-
-				snprintf(buf, sizeof(buf),
-					 "CCA=%lu TR=%s MCmd=%s STag=%lu "
-					 "SPort=%lu\n",
-					 cca_bits, tr[tr_bits], mcmd[cmd_bits],
-					 stag_bits, sport_bits);
-			} else {
-				/* glob state & sresp together */
-				unsigned long c3_bits = (cm_error >> 18) & 7;
-				unsigned long c2_bits = (cm_error >> 15) & 7;
-				unsigned long c1_bits = (cm_error >> 12) & 7;
-				unsigned long c0_bits = (cm_error >> 9) & 7;
-				unsigned long sc_bit = (cm_error >> 8) & 1;
-				unsigned long cmd_bits = (cm_error >> 3) & 0x1f;
-				unsigned long sport_bits = (cm_error >> 0) & 7;
-				snprintf(buf, sizeof(buf),
-					 "C3=%s C2=%s C1=%s C0=%s SC=%s "
-					 "MCmd=%s SPort=%lu\n",
-					 core[c3_bits], core[c2_bits],
-					 core[c1_bits], core[c0_bits],
-					 sc_bit ? "True" : "False",
-					 mcmd[cmd_bits], sport_bits);
-			}
-
-			ocause = (cm_other & CM_GCR_ERROR_MULT_ERR2ND_MSK) >>
-				 CM_GCR_ERROR_MULT_ERR2ND_SHF;
-
-			pr_err("CM_ERROR=%08lx %s <%s>\n", cm_error,
-			       causes[cause], buf);
-			pr_err("CM_ADDR =%08lx\n", cm_addr);
-			pr_err("CM_OTHER=%08lx %s\n", cm_other, causes[ocause]);
-
-			/* reprime cause register */
-			write_gcr_error_cause(0);
-		}
-	}
+	mips_cm_error_report();
 
 	return retval;
 }
diff --git a/arch/mips/mti-malta/malta-memory.c b/arch/mips/mti-malta/malta-memory.c
index b769657..dadeb83 100644
--- a/arch/mips/mti-malta/malta-memory.c
+++ b/arch/mips/mti-malta/malta-memory.c
@@ -179,31 +179,6 @@ void __init prom_free_prom_memory(void)
 	}
 }
 
-unsigned platform_maar_init(unsigned num_pairs)
-{
-	phys_addr_t mem_end = (physical_memsize & ~0xffffull) - 1;
-	struct maar_config cfg[] = {
-		/* DRAM preceding I/O */
-		{ 0x00000000, 0x0fffffff, MIPS_MAAR_S },
-
-		/* DRAM following I/O */
-		{ 0x20000000, mem_end, MIPS_MAAR_S },
-
-		/* DRAM alias in upper half of physical */
-		{ 0x80000000, 0x80000000 + mem_end, MIPS_MAAR_S },
-	};
-	unsigned i, num_cfg = ARRAY_SIZE(cfg);
-
-	/* If DRAM fits before I/O, drop the region following it */
-	if (physical_memsize <= 0x10000000) {
-		num_cfg--;
-		for (i = 1; i < num_cfg; i++)
-			cfg[i] = cfg[i + 1];
-	}
-
-	return maar_config(cfg, num_cfg, num_pairs);
-}
-
 phys_addr_t mips_cdmm_phys_base(void)
 {
 	/* This address is "typically unused" */
diff --git a/arch/mips/netlogic/common/irq.c b/arch/mips/netlogic/common/irq.c
index 5f5d18b..3660dc6 100644
--- a/arch/mips/netlogic/common/irq.c
+++ b/arch/mips/netlogic/common/irq.c
@@ -87,7 +87,7 @@ struct nlm_pic_irq {
 static void xlp_pic_enable(struct irq_data *d)
 {
 	unsigned long flags;
-	struct nlm_pic_irq *pd = irq_data_get_irq_handler_data(d);
+	struct nlm_pic_irq *pd = irq_data_get_irq_chip_data(d);
 
 	BUG_ON(!pd);
 	spin_lock_irqsave(&pd->node->piclock, flags);
@@ -97,7 +97,7 @@ static void xlp_pic_enable(struct irq_data *d)
 
 static void xlp_pic_disable(struct irq_data *d)
 {
-	struct nlm_pic_irq *pd = irq_data_get_irq_handler_data(d);
+	struct nlm_pic_irq *pd = irq_data_get_irq_chip_data(d);
 	unsigned long flags;
 
 	BUG_ON(!pd);
@@ -108,7 +108,7 @@ static void xlp_pic_disable(struct irq_data *d)
 
 static void xlp_pic_mask_ack(struct irq_data *d)
 {
-	struct nlm_pic_irq *pd = irq_data_get_irq_handler_data(d);
+	struct nlm_pic_irq *pd = irq_data_get_irq_chip_data(d);
 
 	clear_c0_eimr(pd->picirq);
 	ack_c0_eirr(pd->picirq);
@@ -116,7 +116,7 @@ static void xlp_pic_mask_ack(struct irq_data *d)
 
 static void xlp_pic_unmask(struct irq_data *d)
 {
-	struct nlm_pic_irq *pd = irq_data_get_irq_handler_data(d);
+	struct nlm_pic_irq *pd = irq_data_get_irq_chip_data(d);
 
 	BUG_ON(!pd);
 
@@ -193,7 +193,7 @@ void nlm_setup_pic_irq(int node, int picirq, int irq, int irt)
 	pic_data->picirq = picirq;
 	pic_data->node = nlm_get_node(node);
 	irq_set_chip_and_handler(xirq, &xlp_pic, handle_level_irq);
-	irq_set_handler_data(xirq, pic_data);
+	irq_set_chip_data(xirq, pic_data);
 }
 
 void nlm_set_pic_extra_ack(int node, int irq, void (*xack)(struct irq_data *))
@@ -202,7 +202,7 @@ void nlm_set_pic_extra_ack(int node, int irq, void (*xack)(struct irq_data *))
 	int xirq;
 
 	xirq = nlm_irq_to_xirq(node, irq);
-	pic_data = irq_get_handler_data(xirq);
+	pic_data = irq_get_chip_data(xirq);
 	if (WARN_ON(!pic_data))
 		return;
 	pic_data->extra_ack = xack;
diff --git a/arch/mips/netlogic/common/smp.c b/arch/mips/netlogic/common/smp.c
index f5fff22..0136b4f 100644
--- a/arch/mips/netlogic/common/smp.c
+++ b/arch/mips/netlogic/common/smp.c
@@ -82,8 +82,9 @@ void nlm_send_ipi_mask(const struct cpumask *mask, unsigned int action)
 }
 
 /* IRQ_IPI_SMP_FUNCTION Handler */
-void nlm_smp_function_ipi_handler(unsigned int irq, struct irq_desc *desc)
+void nlm_smp_function_ipi_handler(unsigned int __irq, struct irq_desc *desc)
 {
+	unsigned int irq = irq_desc_get_irq(desc);
 	clear_c0_eimr(irq);
 	ack_c0_eirr(irq);
 	generic_smp_call_function_interrupt();
@@ -91,8 +92,9 @@ void nlm_smp_function_ipi_handler(unsigned int irq, struct irq_desc *desc)
 }
 
 /* IRQ_IPI_SMP_RESCHEDULE  handler */
-void nlm_smp_resched_ipi_handler(unsigned int irq, struct irq_desc *desc)
+void nlm_smp_resched_ipi_handler(unsigned int __irq, struct irq_desc *desc)
 {
+	unsigned int irq = irq_desc_get_irq(desc);
 	clear_c0_eimr(irq);
 	ack_c0_eirr(irq);
 	scheduler_ipi();
diff --git a/arch/mips/netlogic/xlp/ahci-init-xlp2.c b/arch/mips/netlogic/xlp/ahci-init-xlp2.c
index 7b066a4..c11b9c7 100644
--- a/arch/mips/netlogic/xlp/ahci-init-xlp2.c
+++ b/arch/mips/netlogic/xlp/ahci-init-xlp2.c
@@ -152,7 +152,7 @@ static const u8 sata_phy_config1[]  = {
 	0xC9, 0xC9, 0x07, 0x07, 0x18, 0x18, 0x01, 0x01, 0x22, 0x00
 };
 
-/* SATA PHY config for register block 2 0x0x8065 .. 0x0x80A4 */
+/* SATA PHY config for register block 2 0x8065 .. 0x80A4 */
 static const u8 sata_phy_config2[]  = {
 	0xAA, 0x00, 0x4C, 0xC9, 0xC9, 0x07, 0x07, 0x18,
 	0x18, 0x05, 0x0C, 0x10, 0x00, 0x10, 0x00, 0xFF,
diff --git a/arch/mips/netlogic/xlp/nlm_hal.c b/arch/mips/netlogic/xlp/nlm_hal.c
index a8f4144..80ec929 100644
--- a/arch/mips/netlogic/xlp/nlm_hal.c
+++ b/arch/mips/netlogic/xlp/nlm_hal.c
@@ -91,6 +91,8 @@ static int xlp9xx_irq_to_irt(int irq)
 		return 134;
 	case PIC_SATA_IRQ:
 		return 143;
+	case PIC_NAND_IRQ:
+		return 151;
 	case PIC_SPI_IRQ:
 		return 152;
 	case PIC_MMC_IRQ:
diff --git a/arch/mips/oprofile/common.c b/arch/mips/oprofile/common.c
index 81f5895..3c9ec3d 100644
--- a/arch/mips/oprofile/common.c
+++ b/arch/mips/oprofile/common.c
@@ -91,6 +91,7 @@ int __init oprofile_arch_init(struct oprofile_operations *ops)
 	case CPU_INTERAPTIV:
 	case CPU_PROAPTIV:
 	case CPU_P5600:
+	case CPU_I6400:
 	case CPU_M5150:
 	case CPU_LOONGSON1:
 	case CPU_SB1:
diff --git a/arch/mips/oprofile/op_model_mipsxx.c b/arch/mips/oprofile/op_model_mipsxx.c
index 6a6e2cc..8f988a6 100644
--- a/arch/mips/oprofile/op_model_mipsxx.c
+++ b/arch/mips/oprofile/op_model_mipsxx.c
@@ -392,6 +392,10 @@ static int __init mipsxx_init(void)
 		op_model_mipsxx_ops.cpu_type = "mips/P5600";
 		break;
 
+	case CPU_I6400:
+		op_model_mipsxx_ops.cpu_type = "mips/I6400";
+		break;
+
 	case CPU_M5150:
 		op_model_mipsxx_ops.cpu_type = "mips/M5150";
 		break;
diff --git a/arch/mips/pci/msi-xlp.c b/arch/mips/pci/msi-xlp.c
index 3407495..bb14335 100644
--- a/arch/mips/pci/msi-xlp.c
+++ b/arch/mips/pci/msi-xlp.c
@@ -131,7 +131,7 @@ struct xlp_msi_data {
  */
 static void xlp_msi_enable(struct irq_data *d)
 {
-	struct xlp_msi_data *md = irq_data_get_irq_handler_data(d);
+	struct xlp_msi_data *md = irq_data_get_irq_chip_data(d);
 	unsigned long flags;
 	int vec;
 
@@ -148,7 +148,7 @@ static void xlp_msi_enable(struct irq_data *d)
 
 static void xlp_msi_disable(struct irq_data *d)
 {
-	struct xlp_msi_data *md = irq_data_get_irq_handler_data(d);
+	struct xlp_msi_data *md = irq_data_get_irq_chip_data(d);
 	unsigned long flags;
 	int vec;
 
@@ -165,7 +165,7 @@ static void xlp_msi_disable(struct irq_data *d)
 
 static void xlp_msi_mask_ack(struct irq_data *d)
 {
-	struct xlp_msi_data *md = irq_data_get_irq_handler_data(d);
+	struct xlp_msi_data *md = irq_data_get_irq_chip_data(d);
 	int link, vec;
 
 	link = nlm_irq_msilink(d->irq);
@@ -211,7 +211,7 @@ static void xlp_msix_mask_ack(struct irq_data *d)
 	msixvec = nlm_irq_msixvec(d->irq);
 	link = nlm_irq_msixlink(msixvec);
 	pci_msi_mask_irq(d);
-	md = irq_data_get_irq_handler_data(d);
+	md = irq_data_get_irq_chip_data(d);
 
 	/* Ack MSI on bridge */
 	if (cpu_is_xlp9xx()) {
@@ -302,7 +302,7 @@ static int xlp_setup_msi(uint64_t lnkbase, int node, int link,
 	/* Get MSI data for the link */
 	lirq = PIC_PCIE_LINK_MSI_IRQ(link);
 	xirq = nlm_irq_to_xirq(node, nlm_link_msiirq(link, 0));
-	md = irq_get_handler_data(xirq);
+	md = irq_get_chip_data(xirq);
 	msiaddr = MSI_LINK_ADDR(node, link);
 
 	spin_lock_irqsave(&md->msi_lock, flags);
@@ -409,7 +409,7 @@ static int xlp_setup_msix(uint64_t lnkbase, int node, int link,
 	/* Get MSI data for the link */
 	lirq = PIC_PCIE_MSIX_IRQ(link);
 	xirq = nlm_irq_to_xirq(node, nlm_link_msixirq(link, 0));
-	md = irq_get_handler_data(xirq);
+	md = irq_get_chip_data(xirq);
 	msixaddr = MSIX_LINK_ADDR(node, link);
 
 	spin_lock_irqsave(&md->msi_lock, flags);
@@ -485,7 +485,7 @@ void __init xlp_init_node_msi_irqs(int node, int link)
 	irq = nlm_irq_to_xirq(node, nlm_link_msiirq(link, 0));
 	for (i = irq; i < irq + XLP_MSIVEC_PER_LINK; i++) {
 		irq_set_chip_and_handler(i, &xlp_msi_chip, handle_level_irq);
-		irq_set_handler_data(i, md);
+		irq_set_chip_data(i, md);
 	}
 
 	for (i = 0; i < XLP_MSIXVEC_PER_LINK ; i++) {
@@ -508,7 +508,7 @@ void __init xlp_init_node_msi_irqs(int node, int link)
 		/* Initialize MSI-X extended irq space for the link  */
 		irq = nlm_irq_to_xirq(node, nlm_link_msixirq(link, i));
 		irq_set_chip_and_handler(irq, &xlp_msix_chip, handle_level_irq);
-		irq_set_handler_data(irq, md);
+		irq_set_chip_data(irq, md);
 	}
 }
 
@@ -520,7 +520,7 @@ void nlm_dispatch_msi(int node, int lirq)
 
 	link = lirq - PIC_PCIE_LINK_MSI_IRQ_BASE;
 	irqbase = nlm_irq_to_xirq(node, nlm_link_msiirq(link, 0));
-	md = irq_get_handler_data(irqbase);
+	md = irq_get_chip_data(irqbase);
 	if (cpu_is_xlp9xx())
 		status = nlm_read_reg(md->lnkbase, PCIE_9XX_MSI_STATUS) &
 						md->msi_enabled_mask;
@@ -550,7 +550,7 @@ void nlm_dispatch_msix(int node, int lirq)
 
 	link = lirq - PIC_PCIE_MSIX_IRQ_BASE;
 	irqbase = nlm_irq_to_xirq(node, nlm_link_msixirq(link, 0));
-	md = irq_get_handler_data(irqbase);
+	md = irq_get_chip_data(irqbase);
 	if (cpu_is_xlp9xx())
 		status = nlm_read_reg(md->lnkbase, PCIE_9XX_MSIX_STATUSX(link));
 	else
diff --git a/arch/mips/pci/ops-emma2rh.c b/arch/mips/pci/ops-emma2rh.c
index 710aef5..2dc97c4 100644
--- a/arch/mips/pci/ops-emma2rh.c
+++ b/arch/mips/pci/ops-emma2rh.c
@@ -25,7 +25,6 @@
 #include <linux/types.h>
 
 #include <asm/addrspace.h>
-#include <asm/debug.h>
 
 #include <asm/emma/emma2rh.h>
 
@@ -40,10 +39,9 @@
 static int check_args(struct pci_bus *bus, u32 devfn, u32 * bus_num)
 {
 	/* check if the bus is top-level */
-	if (bus->parent != NULL) {
+	if (bus->parent != NULL)
 		*bus_num = bus->number;
-		db_assert(bus_num != NULL);
-	} else
+	else
 		*bus_num = 0;
 
 	if (*bus_num == 0) {
diff --git a/arch/mips/pci/pci-ar71xx.c b/arch/mips/pci/pci-ar71xx.c
index 283157f..ad35a5e 100644
--- a/arch/mips/pci/pci-ar71xx.c
+++ b/arch/mips/pci/pci-ar71xx.c
@@ -312,8 +312,8 @@ static void ar71xx_pci_irq_init(struct ar71xx_pci_controller *apc)
 		irq_set_chip_data(i, apc);
 	}
 
-	irq_set_handler_data(apc->irq, apc);
-	irq_set_chained_handler(apc->irq, ar71xx_pci_irq_handler);
+	irq_set_chained_handler_and_data(apc->irq, ar71xx_pci_irq_handler,
+					 apc);
 }
 
 static void ar71xx_pci_reset(void)
diff --git a/arch/mips/pci/pci-ar724x.c b/arch/mips/pci/pci-ar724x.c
index 0af362b..907d11d 100644
--- a/arch/mips/pci/pci-ar724x.c
+++ b/arch/mips/pci/pci-ar724x.c
@@ -321,8 +321,8 @@ static void ar724x_pci_irq_init(struct ar724x_pci_controller *apc,
 		irq_set_chip_data(i, apc);
 	}
 
-	irq_set_handler_data(apc->irq, apc);
-	irq_set_chained_handler(apc->irq, ar724x_pci_irq_handler);
+	irq_set_chained_handler_and_data(apc->irq, ar724x_pci_irq_handler,
+					 apc);
 }
 
 static int ar724x_pci_probe(struct platform_device *pdev)
diff --git a/arch/mips/pci/pci-lantiq.c b/arch/mips/pci/pci-lantiq.c
index c5347d9..6a15dbd 100644
--- a/arch/mips/pci/pci-lantiq.c
+++ b/arch/mips/pci/pci-lantiq.c
@@ -20,7 +20,6 @@
 #include <linux/of_irq.h>
 #include <linux/of_pci.h>
 
-#include <asm/gpio.h>
 #include <asm/addrspace.h>
 
 #include <lantiq_soc.h>
diff --git a/arch/mips/pci/pci-rt3883.c b/arch/mips/pci/pci-rt3883.c
index 80fafe6..53c8efa 100644
--- a/arch/mips/pci/pci-rt3883.c
+++ b/arch/mips/pci/pci-rt3883.c
@@ -129,7 +129,7 @@ static void rt3883_pci_write_cfg32(struct rt3883_pci_controller *rpc,
 	rt3883_pci_w32(rpc, val, RT3883_PCI_REG_CFGDATA);
 }
 
-static void rt3883_pci_irq_handler(unsigned int irq, struct irq_desc *desc)
+static void rt3883_pci_irq_handler(unsigned int __irq, struct irq_desc *desc)
 {
 	struct rt3883_pci_controller *rpc;
 	u32 pending;
@@ -145,7 +145,7 @@ static void rt3883_pci_irq_handler(unsigned int irq, struct irq_desc *desc)
 	}
 
 	while (pending) {
-		unsigned bit = __ffs(pending);
+		unsigned irq, bit = __ffs(pending);
 
 		irq = irq_find_mapping(rpc->irq_domain, bit);
 		generic_handle_irq(irq);
@@ -225,8 +225,7 @@ static int rt3883_pci_irq_init(struct device *dev,
 		return -ENODEV;
 	}
 
-	irq_set_handler_data(irq, rpc);
-	irq_set_chained_handler(irq, rt3883_pci_irq_handler);
+	irq_set_chained_handler_and_data(irq, rt3883_pci_irq_handler, rpc);
 
 	return 0;
 }
diff --git a/arch/mips/pistachio/Kconfig b/arch/mips/pistachio/Kconfig
new file mode 100644
index 0000000..97731ea
--- /dev/null
+++ b/arch/mips/pistachio/Kconfig
@@ -0,0 +1,13 @@
+config PISTACHIO_GPTIMER_CLKSRC
+	bool "Enable General Purpose Timer based clocksource"
+	depends on MACH_PISTACHIO
+	select CLKSRC_PISTACHIO
+	select MIPS_EXTERNAL_TIMER
+	help
+	  This option enables a clocksource driver based on a Pistachio
+	  SoC General Purpose external timer.
+
+	  If you want to enable the CPUFreq, you need to enable
+	  this option.
+
+	  If you don't want to enable CPUFreq, you can leave this disabled.
diff --git a/arch/mips/pmcs-msp71xx/msp_irq_cic.c b/arch/mips/pmcs-msp71xx/msp_irq_cic.c
index 1207ec4..8b9cf64 100644
--- a/arch/mips/pmcs-msp71xx/msp_irq_cic.c
+++ b/arch/mips/pmcs-msp71xx/msp_irq_cic.c
@@ -88,7 +88,8 @@ static void unmask_cic_irq(struct irq_data *d)
 	* Make sure we have IRQ affinity.  It may have changed while
 	* we were processing the IRQ.
 	*/
-	if (!cpumask_test_cpu(smp_processor_id(), d->affinity))
+	if (!cpumask_test_cpu(smp_processor_id(),
+			      irq_data_get_affinity_mask(d)))
 		return;
 #endif
 
diff --git a/arch/mips/ralink/cevt-rt3352.c b/arch/mips/ralink/cevt-rt3352.c
index 24bf057..a8e70a9 100644
--- a/arch/mips/ralink/cevt-rt3352.c
+++ b/arch/mips/ralink/cevt-rt3352.c
@@ -36,8 +36,8 @@ struct systick_device {
 	int freq_scale;
 };
 
-static void systick_set_clock_mode(enum clock_event_mode mode,
-				struct clock_event_device *evt);
+static int systick_set_oneshot(struct clock_event_device *evt);
+static int systick_shutdown(struct clock_event_device *evt);
 
 static int systick_next_event(unsigned long delta,
 				struct clock_event_device *evt)
@@ -73,11 +73,12 @@ static struct systick_device systick = {
 		 * cevt-r4k uses 300, make sure systick
 		 * gets used if available
 		 */
-		.rating		= 310,
-		.features	= CLOCK_EVT_FEAT_ONESHOT,
-		.set_next_event	= systick_next_event,
-		.set_mode	= systick_set_clock_mode,
-		.event_handler	= systick_event_handler,
+		.rating			= 310,
+		.features		= CLOCK_EVT_FEAT_ONESHOT,
+		.set_next_event		= systick_next_event,
+		.set_state_shutdown	= systick_shutdown,
+		.set_state_oneshot	= systick_set_oneshot,
+		.event_handler		= systick_event_handler,
 	},
 };
 
@@ -87,33 +88,33 @@ static struct irqaction systick_irqaction = {
 	.dev_id = &systick.dev,
 };
 
-static void systick_set_clock_mode(enum clock_event_mode mode,
-				struct clock_event_device *evt)
+static int systick_shutdown(struct clock_event_device *evt)
 {
 	struct systick_device *sdev;
 
 	sdev = container_of(evt, struct systick_device, dev);
 
-	switch (mode) {
-	case CLOCK_EVT_MODE_ONESHOT:
-		if (!sdev->irq_requested)
-			setup_irq(systick.dev.irq, &systick_irqaction);
-		sdev->irq_requested = 1;
-		iowrite32(CFG_EXT_STK_EN | CFG_CNT_EN,
-				systick.membase + SYSTICK_CONFIG);
-		break;
-
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		if (sdev->irq_requested)
-			free_irq(systick.dev.irq, &systick_irqaction);
-		sdev->irq_requested = 0;
-		iowrite32(0, systick.membase + SYSTICK_CONFIG);
-		break;
-
-	default:
-		pr_err("%s: Unhandeled mips clock_mode\n", systick.dev.name);
-		break;
-	}
+	if (sdev->irq_requested)
+		free_irq(systick.dev.irq, &systick_irqaction);
+	sdev->irq_requested = 0;
+	iowrite32(0, systick.membase + SYSTICK_CONFIG);
+
+	return 0;
+}
+
+static int systick_set_oneshot(struct clock_event_device *evt)
+{
+	struct systick_device *sdev;
+
+	sdev = container_of(evt, struct systick_device, dev);
+
+	if (!sdev->irq_requested)
+		setup_irq(systick.dev.irq, &systick_irqaction);
+	sdev->irq_requested = 1;
+	iowrite32(CFG_EXT_STK_EN | CFG_CNT_EN,
+		  systick.membase + SYSTICK_CONFIG);
+
+	return 0;
 }
 
 static void __init ralink_systick_init(struct device_node *np)
diff --git a/arch/mips/rb532/devices.c b/arch/mips/rb532/devices.c
index e31e8cd..9bd7a2d 100644
--- a/arch/mips/rb532/devices.c
+++ b/arch/mips/rb532/devices.c
@@ -23,6 +23,7 @@
 #include <linux/mtd/nand.h>
 #include <linux/mtd/mtd.h>
 #include <linux/mtd/partitions.h>
+#include <linux/gpio.h>
 #include <linux/gpio_keys.h>
 #include <linux/input.h>
 #include <linux/serial_8250.h>
diff --git a/arch/mips/rb532/gpio.c b/arch/mips/rb532/gpio.c
index 5aa3df8..650d5d3 100644
--- a/arch/mips/rb532/gpio.c
+++ b/arch/mips/rb532/gpio.c
@@ -140,6 +140,11 @@ static int rb532_gpio_direction_output(struct gpio_chip *chip,
 	return 0;
 }
 
+static int rb532_gpio_to_irq(struct gpio_chip *chip, unsigned gpio)
+{
+	return 8 + 4 * 32 + gpio;
+}
+
 static struct rb532_gpio_chip rb532_gpio_chip[] = {
 	[0] = {
 		.chip = {
@@ -148,6 +153,7 @@ static struct rb532_gpio_chip rb532_gpio_chip[] = {
 			.direction_output	= rb532_gpio_direction_output,
 			.get			= rb532_gpio_get,
 			.set			= rb532_gpio_set,
+			.to_irq			= rb532_gpio_to_irq,
 			.base			= 0,
 			.ngpio			= 32,
 		},
diff --git a/arch/mips/sgi-ip27/ip27-timer.c b/arch/mips/sgi-ip27/ip27-timer.c
index a6d10f6..42d6cb9 100644
--- a/arch/mips/sgi-ip27/ip27-timer.c
+++ b/arch/mips/sgi-ip27/ip27-timer.c
@@ -64,12 +64,6 @@ static int rt_next_event(unsigned long delta, struct clock_event_device *evt)
 	return LOCAL_HUB_L(PI_RT_COUNT) >= cnt ? -ETIME : 0;
 }
 
-static void rt_set_mode(enum clock_event_mode mode,
-		struct clock_event_device *evt)
-{
-	/* Nothing to do ...  */
-}
-
 unsigned int rt_timer_irq;
 
 static DEFINE_PER_CPU(struct clock_event_device, hub_rt_clockevent);
@@ -124,7 +118,6 @@ void hub_rt_clock_event_init(void)
 	cd->irq			= irq;
 	cd->cpumask		= cpumask_of(cpu);
 	cd->set_next_event	= rt_next_event;
-	cd->set_mode		= rt_set_mode;
 	clockevents_register_device(cd);
 }
 
diff --git a/arch/mips/sibyte/common/bus_watcher.c b/arch/mips/sibyte/common/bus_watcher.c
index 41a1d22..a4e5599 100644
--- a/arch/mips/sibyte/common/bus_watcher.c
+++ b/arch/mips/sibyte/common/bus_watcher.c
@@ -250,4 +250,4 @@ int __init sibyte_bus_watcher(void)
 	return 0;
 }
 
-__initcall(sibyte_bus_watcher);
+device_initcall(sibyte_bus_watcher);
diff --git a/arch/mips/sni/time.c b/arch/mips/sni/time.c
index cf8ec56..fb4b352 100644
--- a/arch/mips/sni/time.c
+++ b/arch/mips/sni/time.c
@@ -14,44 +14,33 @@
 #define SNI_COUNTER2_DIV	64
 #define SNI_COUNTER0_DIV	((SNI_CLOCK_TICK_RATE / SNI_COUNTER2_DIV) / HZ)
 
-static void a20r_set_mode(enum clock_event_mode mode,
-			  struct clock_event_device *evt)
+static int a20r_set_periodic(struct clock_event_device *evt)
 {
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0x34;
-		wmb();
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE +  0) = SNI_COUNTER0_DIV;
-		wmb();
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE +  0) = SNI_COUNTER0_DIV >> 8;
-		wmb();
-
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0xb4;
-		wmb();
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE +  8) = SNI_COUNTER2_DIV;
-		wmb();
-		*(volatile u8 *)(A20R_PT_CLOCK_BASE +  8) = SNI_COUNTER2_DIV >> 8;
-		wmb();
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0x34;
+	wmb();
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 0) = SNI_COUNTER0_DIV;
+	wmb();
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 0) = SNI_COUNTER0_DIV >> 8;
+	wmb();
 
-		break;
-	case CLOCK_EVT_MODE_ONESHOT:
-	case CLOCK_EVT_MODE_UNUSED:
-	case CLOCK_EVT_MODE_SHUTDOWN:
-		break;
-	case CLOCK_EVT_MODE_RESUME:
-		break;
-	}
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0xb4;
+	wmb();
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 8) = SNI_COUNTER2_DIV;
+	wmb();
+	*(volatile u8 *)(A20R_PT_CLOCK_BASE + 8) = SNI_COUNTER2_DIV >> 8;
+	wmb();
+	return 0;
 }
 
 static struct clock_event_device a20r_clockevent_device = {
-	.name		= "a20r-timer",
-	.features	= CLOCK_EVT_FEAT_PERIODIC,
+	.name			= "a20r-timer",
+	.features		= CLOCK_EVT_FEAT_PERIODIC,
 
 	/* .mult, .shift, .max_delta_ns and .min_delta_ns left uninitialized */
 
-	.rating		= 300,
-	.irq		= SNI_A20R_IRQ_TIMER,
-	.set_mode	= a20r_set_mode,
+	.rating			= 300,
+	.irq			= SNI_A20R_IRQ_TIMER,
+	.set_state_periodic	= a20r_set_periodic,
 };
 
 static irqreturn_t a20r_interrupt(int irq, void *dev_id)
diff --git a/arch/mips/txx9/generic/setup.c b/arch/mips/txx9/generic/setup.c
index 2791b86..9d9962a 100644
--- a/arch/mips/txx9/generic/setup.c
+++ b/arch/mips/txx9/generic/setup.c
@@ -117,22 +117,6 @@ void clk_put(struct clk *clk)
 }
 EXPORT_SYMBOL(clk_put);
 
-/* GPIO support */
-
-#ifdef CONFIG_GPIOLIB
-int gpio_to_irq(unsigned gpio)
-{
-	return -EINVAL;
-}
-EXPORT_SYMBOL(gpio_to_irq);
-
-int irq_to_gpio(unsigned irq)
-{
-	return -EINVAL;
-}
-EXPORT_SYMBOL(irq_to_gpio);
-#endif
-
 #define BOARD_VEC(board)	extern struct txx9_board_vec board;
 #include <asm/txx9/boards.h>
 #undef BOARD_VEC
diff --git a/drivers/ata/pata_rb532_cf.c b/drivers/ata/pata_rb532_cf.c
index 6d08446..12fe0f3 100644
--- a/drivers/ata/pata_rb532_cf.c
+++ b/drivers/ata/pata_rb532_cf.c
@@ -27,12 +27,11 @@
 #include <linux/io.h>
 #include <linux/interrupt.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 
 #include <linux/libata.h>
 #include <scsi/scsi_host.h>
 
-#include <asm/gpio.h>
-
 #define DRV_NAME	"pata-rb532-cf"
 #define DRV_VERSION	"0.1.0"
 #define DRV_DESC	"PATA driver for RouterBOARD 532 Compact Flash"
diff --git a/drivers/bus/mips_cdmm.c b/drivers/bus/mips_cdmm.c
index ab3bde1..1c543ef 100644
--- a/drivers/bus/mips_cdmm.c
+++ b/drivers/bus/mips_cdmm.c
@@ -332,6 +332,18 @@ static phys_addr_t mips_cdmm_cur_base(void)
 }
 
 /**
+ * mips_cdmm_phys_base() - Choose a physical base address for CDMM region.
+ *
+ * Picking a suitable physical address at which to map the CDMM region is
+ * platform specific, so this weak function can be overridden by platform
+ * code to pick a suitable value if none is configured by the bootloader.
+ */
+phys_addr_t __weak mips_cdmm_phys_base(void)
+{
+	return 0;
+}
+
+/**
  * mips_cdmm_setup() - Ensure the CDMM bus is initialised and usable.
  * @bus:	Pointer to bus information for current CPU.
  *		IS_ERR(bus) is checked, so no need for caller to check.
@@ -368,7 +380,7 @@ static int mips_cdmm_setup(struct mips_cdmm_bus *bus)
 	if (!bus->phys)
 		bus->phys = mips_cdmm_cur_base();
 	/* Otherwise, ask platform code for suggestions */
-	if (!bus->phys && mips_cdmm_phys_base)
+	if (!bus->phys)
 		bus->phys = mips_cdmm_phys_base();
 	/* Otherwise, copy what other CPUs have done */
 	if (!bus->phys)
diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig
index c03f04d..552c9b1 100644
--- a/drivers/clocksource/Kconfig
+++ b/drivers/clocksource/Kconfig
@@ -111,6 +111,10 @@ config CLKSRC_LPC32XX
 	select CLKSRC_MMIO
 	select CLKSRC_OF
 
+config CLKSRC_PISTACHIO
+	bool
+	select CLKSRC_OF
+
 config CLKSRC_STM32
 	bool "Clocksource for STM32 SoCs" if !ARCH_STM32
 	depends on OF && ARM && (ARCH_STM32 || COMPILE_TEST)
diff --git a/drivers/clocksource/Makefile b/drivers/clocksource/Makefile
index f228354..066337e 100644
--- a/drivers/clocksource/Makefile
+++ b/drivers/clocksource/Makefile
@@ -44,6 +44,7 @@ obj-$(CONFIG_FSL_FTM_TIMER)	+= fsl_ftm_timer.o
 obj-$(CONFIG_VF_PIT_TIMER)	+= vf_pit_timer.o
 obj-$(CONFIG_CLKSRC_QCOM)	+= qcom-timer.o
 obj-$(CONFIG_MTK_TIMER)		+= mtk_timer.o
+obj-$(CONFIG_CLKSRC_PISTACHIO)	+= time-pistachio.o
 
 obj-$(CONFIG_ARM_ARCH_TIMER)		+= arm_arch_timer.o
 obj-$(CONFIG_ARM_GLOBAL_TIMER)		+= arm_global_timer.o
diff --git a/drivers/clocksource/mips-gic-timer.c b/drivers/clocksource/mips-gic-timer.c
index c3810b6..02a1945 100644
--- a/drivers/clocksource/mips-gic-timer.c
+++ b/drivers/clocksource/mips-gic-timer.c
@@ -72,6 +72,13 @@ static void gic_clockevent_cpu_exit(struct clock_event_device *cd)
 	disable_percpu_irq(gic_timer_irq);
 }
 
+static void gic_update_frequency(void *data)
+{
+	unsigned long rate = (unsigned long)data;
+
+	clockevents_update_freq(this_cpu_ptr(&gic_clockevent_device), rate);
+}
+
 static int gic_cpu_notifier(struct notifier_block *nb, unsigned long action,
 				void *data)
 {
@@ -87,18 +94,40 @@ static int gic_cpu_notifier(struct notifier_block *nb, unsigned long action,
 	return NOTIFY_OK;
 }
 
+static int gic_clk_notifier(struct notifier_block *nb, unsigned long action,
+			    void *data)
+{
+	struct clk_notifier_data *cnd = data;
+
+	if (action == POST_RATE_CHANGE)
+		on_each_cpu(gic_update_frequency, (void *)cnd->new_rate, 1);
+
+	return NOTIFY_OK;
+}
+
+
 static struct notifier_block gic_cpu_nb = {
 	.notifier_call = gic_cpu_notifier,
 };
 
+static struct notifier_block gic_clk_nb = {
+	.notifier_call = gic_clk_notifier,
+};
+
 static int gic_clockevent_init(void)
 {
+	int ret;
+
 	if (!cpu_has_counter || !gic_frequency)
 		return -ENXIO;
 
-	setup_percpu_irq(gic_timer_irq, &gic_compare_irqaction);
+	ret = setup_percpu_irq(gic_timer_irq, &gic_compare_irqaction);
+	if (ret < 0)
+		return ret;
 
-	register_cpu_notifier(&gic_cpu_nb);
+	ret = register_cpu_notifier(&gic_cpu_nb);
+	if (ret < 0)
+		pr_warn("GIC: Unable to register CPU notifier\n");
 
 	gic_clockevent_cpu_init(this_cpu_ptr(&gic_clockevent_device));
 
@@ -118,18 +147,17 @@ static struct clocksource gic_clocksource = {
 
 static void __init __gic_clocksource_init(void)
 {
+	int ret;
+
 	/* Set clocksource mask. */
 	gic_clocksource.mask = CLOCKSOURCE_MASK(gic_get_count_width());
 
 	/* Calculate a somewhat reasonable rating value. */
 	gic_clocksource.rating = 200 + gic_frequency / 10000000;
 
-	clocksource_register_hz(&gic_clocksource, gic_frequency);
-
-	gic_clockevent_init();
-
-	/* And finally start the counter */
-	gic_start_count();
+	ret = clocksource_register_hz(&gic_clocksource, gic_frequency);
+	if (ret < 0)
+		pr_warn("GIC: Unable to register clocksource\n");
 }
 
 void __init gic_clocksource_init(unsigned int frequency)
@@ -139,11 +167,16 @@ void __init gic_clocksource_init(unsigned int frequency)
 		GIC_LOCAL_TO_HWIRQ(GIC_LOCAL_INT_COMPARE);
 
 	__gic_clocksource_init();
+	gic_clockevent_init();
+
+	/* And finally start the counter */
+	gic_start_count();
 }
 
 static void __init gic_clocksource_of_init(struct device_node *node)
 {
 	struct clk *clk;
+	int ret;
 
 	if (WARN_ON(!gic_present || !node->parent ||
 		    !of_device_is_compatible(node->parent, "mti,gic")))
@@ -151,8 +184,13 @@ static void __init gic_clocksource_of_init(struct device_node *node)
 
 	clk = of_clk_get(node, 0);
 	if (!IS_ERR(clk)) {
+		if (clk_prepare_enable(clk) < 0) {
+			pr_err("GIC failed to enable clock\n");
+			clk_put(clk);
+			return;
+		}
+
 		gic_frequency = clk_get_rate(clk);
-		clk_put(clk);
 	} else if (of_property_read_u32(node, "clock-frequency",
 					&gic_frequency)) {
 		pr_err("GIC frequency not specified.\n");
@@ -165,6 +203,15 @@ static void __init gic_clocksource_of_init(struct device_node *node)
 	}
 
 	__gic_clocksource_init();
+
+	ret = gic_clockevent_init();
+	if (!ret && !IS_ERR(clk)) {
+		if (clk_notifier_register(clk, &gic_clk_nb) < 0)
+			pr_warn("GIC: Unable to register clock notifier\n");
+	}
+
+	/* And finally start the counter */
+	gic_start_count();
 }
 CLOCKSOURCE_OF_DECLARE(mips_gic_timer, "mti,gic-timer",
 		       gic_clocksource_of_init);
diff --git a/drivers/clocksource/time-pistachio.c b/drivers/clocksource/time-pistachio.c
new file mode 100644
index 0000000..18d4266
--- /dev/null
+++ b/drivers/clocksource/time-pistachio.c
@@ -0,0 +1,217 @@
+/*
+ * Pistachio clocksource based on general-purpose timers
+ *
+ * Copyright (C) 2015 Imagination Technologies
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+
+#define pr_fmt(fmt) "%s: " fmt, __func__
+
+#include <linux/clk.h>
+#include <linux/clocksource.h>
+#include <linux/clockchips.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/spinlock.h>
+#include <linux/mfd/syscon.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/sched_clock.h>
+#include <linux/time.h>
+
+/* Top level reg */
+#define CR_TIMER_CTRL_CFG		0x00
+#define TIMER_ME_GLOBAL			BIT(0)
+#define CR_TIMER_REV			0x10
+
+/* Timer specific registers */
+#define TIMER_CFG			0x20
+#define TIMER_ME_LOCAL			BIT(0)
+#define TIMER_RELOAD_VALUE		0x24
+#define TIMER_CURRENT_VALUE		0x28
+#define TIMER_CURRENT_OVERFLOW_VALUE	0x2C
+#define TIMER_IRQ_STATUS		0x30
+#define TIMER_IRQ_CLEAR			0x34
+#define TIMER_IRQ_MASK			0x38
+
+#define PERIP_TIMER_CONTROL		0x90
+
+/* Timer specific configuration Values */
+#define RELOAD_VALUE			0xffffffff
+
+struct pistachio_clocksource {
+	void __iomem *base;
+	raw_spinlock_t lock;
+	struct clocksource cs;
+};
+
+static struct pistachio_clocksource pcs_gpt;
+
+#define to_pistachio_clocksource(cs)	\
+	container_of(cs, struct pistachio_clocksource, cs)
+
+static inline u32 gpt_readl(void __iomem *base, u32 offset, u32 gpt_id)
+{
+	return readl(base + 0x20 * gpt_id + offset);
+}
+
+static inline void gpt_writel(void __iomem *base, u32 value, u32 offset,
+		u32 gpt_id)
+{
+	writel(value, base + 0x20 * gpt_id + offset);
+}
+
+static cycle_t pistachio_clocksource_read_cycles(struct clocksource *cs)
+{
+	struct pistachio_clocksource *pcs = to_pistachio_clocksource(cs);
+	u32 counter, overflw;
+	unsigned long flags;
+
+	/*
+	 * The counter value is only refreshed after the overflow value is read.
+	 * And they must be read in strict order, hence raw spin lock added.
+	 */
+
+	raw_spin_lock_irqsave(&pcs->lock, flags);
+	overflw = gpt_readl(pcs->base, TIMER_CURRENT_OVERFLOW_VALUE, 0);
+	counter = gpt_readl(pcs->base, TIMER_CURRENT_VALUE, 0);
+	raw_spin_unlock_irqrestore(&pcs->lock, flags);
+
+	return ~(cycle_t)counter;
+}
+
+static u64 notrace pistachio_read_sched_clock(void)
+{
+	return pistachio_clocksource_read_cycles(&pcs_gpt.cs);
+}
+
+static void pistachio_clksrc_set_mode(struct clocksource *cs, int timeridx,
+			int enable)
+{
+	struct pistachio_clocksource *pcs = to_pistachio_clocksource(cs);
+	u32 val;
+
+	val = gpt_readl(pcs->base, TIMER_CFG, timeridx);
+	if (enable)
+		val |= TIMER_ME_LOCAL;
+	else
+		val &= ~TIMER_ME_LOCAL;
+
+	gpt_writel(pcs->base, val, TIMER_CFG, timeridx);
+}
+
+static void pistachio_clksrc_enable(struct clocksource *cs, int timeridx)
+{
+	struct pistachio_clocksource *pcs = to_pistachio_clocksource(cs);
+
+	/* Disable GPT local before loading reload value */
+	pistachio_clksrc_set_mode(cs, timeridx, false);
+	gpt_writel(pcs->base, RELOAD_VALUE, TIMER_RELOAD_VALUE, timeridx);
+	pistachio_clksrc_set_mode(cs, timeridx, true);
+}
+
+static void pistachio_clksrc_disable(struct clocksource *cs, int timeridx)
+{
+	/* Disable GPT local */
+	pistachio_clksrc_set_mode(cs, timeridx, false);
+}
+
+static int pistachio_clocksource_enable(struct clocksource *cs)
+{
+	pistachio_clksrc_enable(cs, 0);
+	return 0;
+}
+
+static void pistachio_clocksource_disable(struct clocksource *cs)
+{
+	pistachio_clksrc_disable(cs, 0);
+}
+
+/* Desirable clock source for pistachio platform */
+static struct pistachio_clocksource pcs_gpt = {
+	.cs =	{
+		.name		= "gptimer",
+		.rating		= 300,
+		.enable		= pistachio_clocksource_enable,
+		.disable	= pistachio_clocksource_disable,
+		.read		= pistachio_clocksource_read_cycles,
+		.mask		= CLOCKSOURCE_MASK(32),
+		.flags		= CLOCK_SOURCE_IS_CONTINUOUS |
+				  CLOCK_SOURCE_SUSPEND_NONSTOP,
+		},
+};
+
+static void __init pistachio_clksrc_of_init(struct device_node *node)
+{
+	struct clk *sys_clk, *fast_clk;
+	struct regmap *periph_regs;
+	unsigned long rate;
+	int ret;
+
+	pcs_gpt.base = of_iomap(node, 0);
+	if (!pcs_gpt.base) {
+		pr_err("cannot iomap\n");
+		return;
+	}
+
+	periph_regs = syscon_regmap_lookup_by_phandle(node, "img,cr-periph");
+	if (IS_ERR(periph_regs)) {
+		pr_err("cannot get peripheral regmap (%lu)\n",
+		       PTR_ERR(periph_regs));
+		return;
+	}
+
+	/* Switch to using the fast counter clock */
+	ret = regmap_update_bits(periph_regs, PERIP_TIMER_CONTROL,
+				 0xf, 0x0);
+	if (ret)
+		return;
+
+	sys_clk = of_clk_get_by_name(node, "sys");
+	if (IS_ERR(sys_clk)) {
+		pr_err("clock get failed (%lu)\n", PTR_ERR(sys_clk));
+		return;
+	}
+
+	fast_clk = of_clk_get_by_name(node, "fast");
+	if (IS_ERR(fast_clk)) {
+		pr_err("clock get failed (%lu)\n", PTR_ERR(fast_clk));
+		return;
+	}
+
+	ret = clk_prepare_enable(sys_clk);
+	if (ret < 0) {
+		pr_err("failed to enable clock (%d)\n", ret);
+		return;
+	}
+
+	ret = clk_prepare_enable(fast_clk);
+	if (ret < 0) {
+		pr_err("failed to enable clock (%d)\n", ret);
+		clk_disable_unprepare(sys_clk);
+		return;
+	}
+
+	rate = clk_get_rate(fast_clk);
+
+	/* Disable irq's for clocksource usage */
+	gpt_writel(&pcs_gpt.base, 0, TIMER_IRQ_MASK, 0);
+	gpt_writel(&pcs_gpt.base, 0, TIMER_IRQ_MASK, 1);
+	gpt_writel(&pcs_gpt.base, 0, TIMER_IRQ_MASK, 2);
+	gpt_writel(&pcs_gpt.base, 0, TIMER_IRQ_MASK, 3);
+
+	/* Enable timer block */
+	writel(TIMER_ME_GLOBAL, pcs_gpt.base);
+
+	raw_spin_lock_init(&pcs_gpt.lock);
+	sched_clock_register(pistachio_read_sched_clock, 32, rate);
+	clocksource_register_hz(&pcs_gpt.cs, rate);
+}
+CLOCKSOURCE_OF_DECLARE(pistachio_gptimer, "img,pistachio-gptimer",
+		       pistachio_clksrc_of_init);
diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile
index f82cd67..2b64f61 100644
--- a/drivers/gpio/Makefile
+++ b/drivers/gpio/Makefile
@@ -20,6 +20,7 @@ obj-$(CONFIG_GPIO_ADP5588)	+= gpio-adp5588.o
 obj-$(CONFIG_GPIO_ALTERA)  	+= gpio-altera.o
 obj-$(CONFIG_GPIO_AMD8111)	+= gpio-amd8111.o
 obj-$(CONFIG_GPIO_ARIZONA)	+= gpio-arizona.o
+obj-$(CONFIG_ATH79)		+= gpio-ath79.o
 obj-$(CONFIG_GPIO_BCM_KONA)	+= gpio-bcm-kona.o
 obj-$(CONFIG_GPIO_BRCMSTB)	+= gpio-brcmstb.o
 obj-$(CONFIG_GPIO_BT8XX)	+= gpio-bt8xx.o
diff --git a/arch/mips/ath79/gpio.c b/drivers/gpio/gpio-ath79.c
similarity index 80%
rename from arch/mips/ath79/gpio.c
rename to drivers/gpio/gpio-ath79.c
index f59ccb2..03b9953 100644
--- a/arch/mips/ath79/gpio.c
+++ b/drivers/gpio/gpio-ath79.c
@@ -24,8 +24,6 @@
 #include <linux/of_device.h>
 
 #include <asm/mach-ath79/ar71xx_regs.h>
-#include <asm/mach-ath79/ath79.h>
-#include "common.h"
 
 static void __iomem *ath79_gpio_base;
 static u32 ath79_gpio_count;
@@ -139,47 +137,6 @@ static struct gpio_chip ath79_gpio_chip = {
 	.base			= 0,
 };
 
-static void __iomem *ath79_gpio_get_function_reg(void)
-{
-	u32 reg = 0;
-
-	if (soc_is_ar71xx() ||
-	    soc_is_ar724x() ||
-	    soc_is_ar913x() ||
-	    soc_is_ar933x())
-		reg = AR71XX_GPIO_REG_FUNC;
-	else if (soc_is_ar934x())
-		reg = AR934X_GPIO_REG_FUNC;
-	else
-		BUG();
-
-	return ath79_gpio_base + reg;
-}
-
-void ath79_gpio_function_setup(u32 set, u32 clear)
-{
-	void __iomem *reg = ath79_gpio_get_function_reg();
-	unsigned long flags;
-
-	spin_lock_irqsave(&ath79_gpio_lock, flags);
-
-	__raw_writel((__raw_readl(reg) & ~clear) | set, reg);
-	/* flush write */
-	__raw_readl(reg);
-
-	spin_unlock_irqrestore(&ath79_gpio_lock, flags);
-}
-
-void ath79_gpio_function_enable(u32 mask)
-{
-	ath79_gpio_function_setup(mask, 0);
-}
-
-void ath79_gpio_function_disable(u32 mask)
-{
-	ath79_gpio_function_setup(0, mask);
-}
-
 static const struct of_device_id ath79_gpio_of_match[] = {
 	{ .compatible = "qca,ar7100-gpio" },
 	{ .compatible = "qca,ar9340-gpio" },
@@ -245,35 +202,3 @@ static struct platform_driver ath79_gpio_driver = {
 };
 
 module_platform_driver(ath79_gpio_driver);
-
-int gpio_get_value(unsigned gpio)
-{
-	if (gpio < ath79_gpio_count)
-		return __ath79_gpio_get_value(gpio);
-
-	return __gpio_get_value(gpio);
-}
-EXPORT_SYMBOL(gpio_get_value);
-
-void gpio_set_value(unsigned gpio, int value)
-{
-	if (gpio < ath79_gpio_count)
-		__ath79_gpio_set_value(gpio, value);
-	else
-		__gpio_set_value(gpio, value);
-}
-EXPORT_SYMBOL(gpio_set_value);
-
-int gpio_to_irq(unsigned gpio)
-{
-	/* FIXME */
-	return -EINVAL;
-}
-EXPORT_SYMBOL(gpio_to_irq);
-
-int irq_to_gpio(unsigned irq)
-{
-	/* FIXME */
-	return -EINVAL;
-}
-EXPORT_SYMBOL(irq_to_gpio);
diff --git a/drivers/input/misc/rb532_button.c b/drivers/input/misc/rb532_button.c
index e956e81..62c5814 100644
--- a/drivers/input/misc/rb532_button.c
+++ b/drivers/input/misc/rb532_button.c
@@ -7,6 +7,7 @@
 #include <linux/input-polldev.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/gpio.h>
 
 #include <asm/mach-rc32434/gpio.h>
 #include <asm/mach-rc32434/rb.h>
diff --git a/drivers/irqchip/irq-mips-gic.c b/drivers/irqchip/irq-mips-gic.c
index dae5914..1764bcf 100644
--- a/drivers/irqchip/irq-mips-gic.c
+++ b/drivers/irqchip/irq-mips-gic.c
@@ -41,20 +41,46 @@ static struct irq_chip gic_level_irq_controller, gic_edge_irq_controller;
 
 static void __gic_irq_dispatch(void);
 
-static inline unsigned int gic_read(unsigned int reg)
+static inline u32 gic_read32(unsigned int reg)
 {
 	return __raw_readl(gic_base + reg);
 }
 
-static inline void gic_write(unsigned int reg, unsigned int val)
+static inline u64 gic_read64(unsigned int reg)
 {
-	__raw_writel(val, gic_base + reg);
+	return __raw_readq(gic_base + reg);
 }
 
-static inline void gic_update_bits(unsigned int reg, unsigned int mask,
-				   unsigned int val)
+static inline unsigned long gic_read(unsigned int reg)
 {
-	unsigned int regval;
+	if (!mips_cm_is64)
+		return gic_read32(reg);
+	else
+		return gic_read64(reg);
+}
+
+static inline void gic_write32(unsigned int reg, u32 val)
+{
+	return __raw_writel(val, gic_base + reg);
+}
+
+static inline void gic_write64(unsigned int reg, u64 val)
+{
+	return __raw_writeq(val, gic_base + reg);
+}
+
+static inline void gic_write(unsigned int reg, unsigned long val)
+{
+	if (!mips_cm_is64)
+		return gic_write32(reg, (u32)val);
+	else
+		return gic_write64(reg, (u64)val);
+}
+
+static inline void gic_update_bits(unsigned int reg, unsigned long mask,
+				   unsigned long val)
+{
+	unsigned long regval;
 
 	regval = gic_read(reg);
 	regval &= ~mask;
@@ -65,40 +91,40 @@ static inline void gic_update_bits(unsigned int reg, unsigned int mask,
 static inline void gic_reset_mask(unsigned int intr)
 {
 	gic_write(GIC_REG(SHARED, GIC_SH_RMASK) + GIC_INTR_OFS(intr),
-		  1 << GIC_INTR_BIT(intr));
+		  1ul << GIC_INTR_BIT(intr));
 }
 
 static inline void gic_set_mask(unsigned int intr)
 {
 	gic_write(GIC_REG(SHARED, GIC_SH_SMASK) + GIC_INTR_OFS(intr),
-		  1 << GIC_INTR_BIT(intr));
+		  1ul << GIC_INTR_BIT(intr));
 }
 
 static inline void gic_set_polarity(unsigned int intr, unsigned int pol)
 {
 	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_POLARITY) +
-			GIC_INTR_OFS(intr), 1 << GIC_INTR_BIT(intr),
-			pol << GIC_INTR_BIT(intr));
+			GIC_INTR_OFS(intr), 1ul << GIC_INTR_BIT(intr),
+			(unsigned long)pol << GIC_INTR_BIT(intr));
 }
 
 static inline void gic_set_trigger(unsigned int intr, unsigned int trig)
 {
 	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_TRIGGER) +
-			GIC_INTR_OFS(intr), 1 << GIC_INTR_BIT(intr),
-			trig << GIC_INTR_BIT(intr));
+			GIC_INTR_OFS(intr), 1ul << GIC_INTR_BIT(intr),
+			(unsigned long)trig << GIC_INTR_BIT(intr));
 }
 
 static inline void gic_set_dual_edge(unsigned int intr, unsigned int dual)
 {
 	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_DUAL) + GIC_INTR_OFS(intr),
-			1 << GIC_INTR_BIT(intr),
-			dual << GIC_INTR_BIT(intr));
+			1ul << GIC_INTR_BIT(intr),
+			(unsigned long)dual << GIC_INTR_BIT(intr));
 }
 
 static inline void gic_map_to_pin(unsigned int intr, unsigned int pin)
 {
-	gic_write(GIC_REG(SHARED, GIC_SH_INTR_MAP_TO_PIN_BASE) +
-		  GIC_SH_MAP_TO_PIN(intr), GIC_MAP_TO_PIN_MSK | pin);
+	gic_write32(GIC_REG(SHARED, GIC_SH_INTR_MAP_TO_PIN_BASE) +
+		    GIC_SH_MAP_TO_PIN(intr), GIC_MAP_TO_PIN_MSK | pin);
 }
 
 static inline void gic_map_to_vpe(unsigned int intr, unsigned int vpe)
@@ -113,10 +139,13 @@ cycle_t gic_read_count(void)
 {
 	unsigned int hi, hi2, lo;
 
+	if (mips_cm_is64)
+		return (cycle_t)gic_read(GIC_REG(SHARED, GIC_SH_COUNTER));
+
 	do {
-		hi = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
-		lo = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_31_00));
-		hi2 = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
+		hi = gic_read32(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
+		lo = gic_read32(GIC_REG(SHARED, GIC_SH_COUNTER_31_00));
+		hi2 = gic_read32(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
 	} while (hi2 != hi);
 
 	return (((cycle_t) hi) << 32) + lo;
@@ -135,10 +164,14 @@ unsigned int gic_get_count_width(void)
 
 void gic_write_compare(cycle_t cnt)
 {
-	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI),
-				(int)(cnt >> 32));
-	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO),
-				(int)(cnt & 0xffffffff));
+	if (mips_cm_is64) {
+		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE), cnt);
+	} else {
+		gic_write32(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI),
+					(int)(cnt >> 32));
+		gic_write32(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO),
+					(int)(cnt & 0xffffffff));
+	}
 }
 
 void gic_write_cpu_compare(cycle_t cnt, int cpu)
@@ -148,10 +181,15 @@ void gic_write_cpu_compare(cycle_t cnt, int cpu)
 	local_irq_save(flags);
 
 	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), cpu);
-	gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_HI),
-				(int)(cnt >> 32));
-	gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_LO),
-				(int)(cnt & 0xffffffff));
+
+	if (mips_cm_is64) {
+		gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE), cnt);
+	} else {
+		gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_HI),
+					(int)(cnt >> 32));
+		gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_LO),
+					(int)(cnt & 0xffffffff));
+	}
 
 	local_irq_restore(flags);
 }
@@ -160,8 +198,11 @@ cycle_t gic_read_compare(void)
 {
 	unsigned int hi, lo;
 
-	hi = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI));
-	lo = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO));
+	if (mips_cm_is64)
+		return (cycle_t)gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE));
+
+	hi = gic_read32(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI));
+	lo = gic_read32(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO));
 
 	return (((cycle_t) hi) << 32) + lo;
 }
@@ -196,7 +237,7 @@ static bool gic_local_irq_is_routable(int intr)
 	if (cpu_has_veic)
 		return true;
 
-	vpe_ctl = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_CTL));
+	vpe_ctl = gic_read32(GIC_REG(VPE_LOCAL, GIC_VPE_CTL));
 	switch (intr) {
 	case GIC_LOCAL_INT_TIMER:
 		return vpe_ctl & GIC_VPE_CTL_TIMER_RTBL_MSK;
@@ -262,7 +303,7 @@ int gic_get_c0_fdc_int(void)
 
 static void gic_handle_shared_int(bool chained)
 {
-	unsigned int i, intr, virq;
+	unsigned int i, intr, virq, gic_reg_step = mips_cm_is64 ? 8 : 4;
 	unsigned long *pcpu_mask;
 	unsigned long pending_reg, intrmask_reg;
 	DECLARE_BITMAP(pending, GIC_MAX_INTRS);
@@ -277,8 +318,8 @@ static void gic_handle_shared_int(bool chained)
 	for (i = 0; i < BITS_TO_LONGS(gic_shared_intrs); i++) {
 		pending[i] = gic_read(pending_reg);
 		intrmask[i] = gic_read(intrmask_reg);
-		pending_reg += 0x4;
-		intrmask_reg += 0x4;
+		pending_reg += gic_reg_step;
+		intrmask_reg += gic_reg_step;
 	}
 
 	bitmap_and(pending, pending, intrmask, gic_shared_intrs);
@@ -425,8 +466,8 @@ static void gic_handle_local_int(bool chained)
 	unsigned long pending, masked;
 	unsigned int intr, virq;
 
-	pending = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_PEND));
-	masked = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_MASK));
+	pending = gic_read32(GIC_REG(VPE_LOCAL, GIC_VPE_PEND));
+	masked = gic_read32(GIC_REG(VPE_LOCAL, GIC_VPE_MASK));
 
 	bitmap_and(&pending, &pending, &masked, GIC_NUM_LOCAL_INTRS);
 
@@ -449,14 +490,14 @@ static void gic_mask_local_irq(struct irq_data *d)
 {
 	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
 
-	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_RMASK), 1 << intr);
+	gic_write32(GIC_REG(VPE_LOCAL, GIC_VPE_RMASK), 1 << intr);
 }
 
 static void gic_unmask_local_irq(struct irq_data *d)
 {
 	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
 
-	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_SMASK), 1 << intr);
+	gic_write32(GIC_REG(VPE_LOCAL, GIC_VPE_SMASK), 1 << intr);
 }
 
 static struct irq_chip gic_local_irq_controller = {
@@ -474,7 +515,7 @@ static void gic_mask_local_irq_all_vpes(struct irq_data *d)
 	spin_lock_irqsave(&gic_lock, flags);
 	for (i = 0; i < gic_vpes; i++) {
 		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
-		gic_write(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << intr);
+		gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << intr);
 	}
 	spin_unlock_irqrestore(&gic_lock, flags);
 }
@@ -488,7 +529,7 @@ static void gic_unmask_local_irq_all_vpes(struct irq_data *d)
 	spin_lock_irqsave(&gic_lock, flags);
 	for (i = 0; i < gic_vpes; i++) {
 		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
-		gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SMASK), 1 << intr);
+		gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_SMASK), 1 << intr);
 	}
 	spin_unlock_irqrestore(&gic_lock, flags);
 }
@@ -608,7 +649,7 @@ static void __init gic_basic_init(void)
 		for (j = 0; j < GIC_NUM_LOCAL_INTRS; j++) {
 			if (!gic_local_irq_is_routable(j))
 				continue;
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << j);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << j);
 		}
 	}
 }
@@ -653,27 +694,32 @@ static int gic_local_irq_domain_map(struct irq_domain *d, unsigned int virq,
 
 		switch (intr) {
 		case GIC_LOCAL_INT_WD:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_WD_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_WD_MAP), val);
 			break;
 		case GIC_LOCAL_INT_COMPARE:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_MAP),
+				    val);
 			break;
 		case GIC_LOCAL_INT_TIMER:
 			/* CONFIG_MIPS_CMP workaround (see __gic_init) */
 			val = GIC_MAP_TO_PIN_MSK | timer_cpu_pin;
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_TIMER_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_TIMER_MAP),
+				    val);
 			break;
 		case GIC_LOCAL_INT_PERFCTR:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_PERFCTR_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_PERFCTR_MAP),
+				    val);
 			break;
 		case GIC_LOCAL_INT_SWINT0:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SWINT0_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_SWINT0_MAP),
+				    val);
 			break;
 		case GIC_LOCAL_INT_SWINT1:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SWINT1_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_SWINT1_MAP),
+				    val);
 			break;
 		case GIC_LOCAL_INT_FDC:
-			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_FDC_MAP), val);
+			gic_write32(GIC_REG(VPE_OTHER, GIC_VPE_FDC_MAP), val);
 			break;
 		default:
 			pr_err("Invalid local IRQ %d\n", intr);
@@ -778,7 +824,7 @@ static void __init __gic_init(unsigned long gic_base_addr,
 		 */
 		if (IS_ENABLED(CONFIG_MIPS_CMP) &&
 		    gic_local_irq_is_routable(GIC_LOCAL_INT_TIMER)) {
-			timer_cpu_pin = gic_read(GIC_REG(VPE_LOCAL,
+			timer_cpu_pin = gic_read32(GIC_REG(VPE_LOCAL,
 							 GIC_VPE_TIMER_MAP)) &
 					GIC_MAP_MSK;
 			irq_set_chained_handler(MIPS_CPU_IRQ_BASE +
diff --git a/drivers/net/ethernet/ti/cpmac.c b/drivers/net/ethernet/ti/cpmac.c
index dd94300..cba3d9f 100644
--- a/drivers/net/ethernet/ti/cpmac.c
+++ b/drivers/net/ethernet/ti/cpmac.c
@@ -41,6 +41,8 @@
 #include <linux/gpio.h>
 #include <linux/atomic.h>
 
+#include <asm/mach-ar7/ar7.h>
+
 MODULE_AUTHOR("Eugene Konev <ejka@imfi.kspu.ru>");
 MODULE_DESCRIPTION("TI AR7 ethernet driver (CPMAC)");
 MODULE_LICENSE("GPL");
diff --git a/drivers/staging/octeon/ethernet-rx.c b/drivers/staging/octeon/ethernet-rx.c
index 22853d3..d1a33a9 100644
--- a/drivers/staging/octeon/ethernet-rx.c
+++ b/drivers/staging/octeon/ethernet-rx.c
@@ -70,7 +70,14 @@ static irqreturn_t cvm_oct_do_interrupt(int cpl, void *dev_id)
  */
 static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 {
-	if ((work->word2.snoip.err_code == 10) && (work->len <= 64)) {
+	int port;
+
+	if (octeon_has_feature(OCTEON_FEATURE_PKND))
+		port = work->word0.pip.cn68xx.pknd;
+	else
+		port = work->word1.cn38xx.ipprt;
+
+	if ((work->word2.snoip.err_code == 10) && (work->word1.len <= 64)) {
 		/*
 		 * Ignore length errors on min size packets. Some
 		 * equipment incorrectly pads packets to 64+4FCS
@@ -87,8 +94,8 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 		 * packet to determine if we can remove a non spec
 		 * preamble and generate a correct packet.
 		 */
-		int interface = cvmx_helper_get_interface_num(work->ipprt);
-		int index = cvmx_helper_get_interface_index_num(work->ipprt);
+		int interface = cvmx_helper_get_interface_num(port);
+		int index = cvmx_helper_get_interface_index_num(port);
 		union cvmx_gmxx_rxx_frm_ctl gmxx_rxx_frm_ctl;
 
 		gmxx_rxx_frm_ctl.u64 =
@@ -99,7 +106,7 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 			    cvmx_phys_to_ptr(work->packet_ptr.s.addr);
 			int i = 0;
 
-			while (i < work->len - 1) {
+			while (i < work->word1.len - 1) {
 				if (*ptr != 0x55)
 					break;
 				ptr++;
@@ -109,18 +116,18 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 			if (*ptr == 0xd5) {
 				/*
 				  printk_ratelimited("Port %d received 0xd5 preamble\n",
-					  work->ipprt);
+					  port);
 				 */
 				work->packet_ptr.s.addr += i + 1;
-				work->len -= i + 5;
+				work->word1.len -= i + 5;
 			} else if ((*ptr & 0xf) == 0xd) {
 				/*
 				  printk_ratelimited("Port %d received 0x?d preamble\n",
-					  work->ipprt);
+					  port);
 				 */
 				work->packet_ptr.s.addr += i;
-				work->len -= i + 4;
-				for (i = 0; i < work->len; i++) {
+				work->word1.len -= i + 4;
+				for (i = 0; i < work->word1.len; i++) {
 					*ptr =
 					    ((*ptr & 0xf0) >> 4) |
 					    ((*(ptr + 1) & 0xf) << 4);
@@ -128,7 +135,7 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 				}
 			} else {
 				printk_ratelimited("Port %d unknown preamble, packet dropped\n",
-						   work->ipprt);
+						   port);
 				/*
 				   cvmx_helper_dump_packet(work);
 				 */
@@ -138,7 +145,7 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t *work)
 		}
 	} else {
 		printk_ratelimited("Port %d receive error code %d, packet dropped\n",
-				   work->ipprt, work->word2.snoip.err_code);
+				   port, work->word2.snoip.err_code);
 		cvm_oct_free_work(work);
 		return 1;
 	}
@@ -172,9 +179,16 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 	}
 
 	/* Only allow work for our group (and preserve priorities) */
-	old_group_mask = cvmx_read_csr(CVMX_POW_PP_GRP_MSKX(coreid));
-	cvmx_write_csr(CVMX_POW_PP_GRP_MSKX(coreid),
-		       (old_group_mask & ~0xFFFFull) | 1 << pow_receive_group);
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+		old_group_mask = cvmx_read_csr(CVMX_SSO_PPX_GRP_MSK(coreid));
+		cvmx_write_csr(CVMX_SSO_PPX_GRP_MSK(coreid),
+				1ull << pow_receive_group);
+		cvmx_read_csr(CVMX_SSO_PPX_GRP_MSK(coreid)); /* Flush */
+	} else {
+		old_group_mask = cvmx_read_csr(CVMX_POW_PP_GRP_MSKX(coreid));
+		cvmx_write_csr(CVMX_POW_PP_GRP_MSKX(coreid),
+			(old_group_mask & ~0xFFFFull) | 1 << pow_receive_group);
+	}
 
 	if (USE_ASYNC_IOBDMA) {
 		cvmx_pow_work_request_async(CVMX_SCR_SCRATCH, CVMX_POW_NO_WAIT);
@@ -186,6 +200,7 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 		struct sk_buff **pskb = NULL;
 		int skb_in_hw;
 		cvmx_wqe_t *work;
+		int port;
 
 		if (USE_ASYNC_IOBDMA && did_work_request)
 			work = cvmx_pow_work_response_async(CVMX_SCR_SCRATCH);
@@ -195,12 +210,19 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 		prefetch(work);
 		did_work_request = 0;
 		if (work == NULL) {
-			union cvmx_pow_wq_int wq_int;
+			if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+				cvmx_write_csr(CVMX_SSO_WQ_IQ_DIS,
+					       1ull << pow_receive_group);
+				cvmx_write_csr(CVMX_SSO_WQ_INT,
+					       1ull << pow_receive_group);
+			} else {
+				union cvmx_pow_wq_int wq_int;
 
-			wq_int.u64 = 0;
-			wq_int.s.iq_dis = 1 << pow_receive_group;
-			wq_int.s.wq_int = 1 << pow_receive_group;
-			cvmx_write_csr(CVMX_POW_WQ_INT, wq_int.u64);
+				wq_int.u64 = 0;
+				wq_int.s.iq_dis = 1 << pow_receive_group;
+				wq_int.s.wq_int = 1 << pow_receive_group;
+				cvmx_write_csr(CVMX_POW_WQ_INT, wq_int.u64);
+			}
 			break;
 		}
 		pskb = (struct sk_buff **)(cvm_oct_get_buffer_ptr(work->packet_ptr) -
@@ -220,7 +242,13 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 			prefetch(&skb->head);
 			prefetch(&skb->len);
 		}
-		prefetch(cvm_oct_device[work->ipprt]);
+
+		if (octeon_has_feature(OCTEON_FEATURE_PKND))
+			port = work->word0.pip.cn68xx.pknd;
+		else
+			port = work->word1.cn38xx.ipprt;
+
+		prefetch(cvm_oct_device[port]);
 
 		/* Immediately throw away all packets with receive errors */
 		if (unlikely(work->word2.snoip.rcv_error)) {
@@ -237,7 +265,7 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 			skb->data = skb->head + work->packet_ptr.s.addr -
 				cvmx_ptr_to_phys(skb->head);
 			prefetch(skb->data);
-			skb->len = work->len;
+			skb->len = work->word1.len;
 			skb_set_tail_pointer(skb, skb->len);
 			packet_not_copied = 1;
 		} else {
@@ -245,7 +273,7 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 			 * We have to copy the packet. First allocate
 			 * an skbuff for it.
 			 */
-			skb = dev_alloc_skb(work->len);
+			skb = dev_alloc_skb(work->word1.len);
 			if (!skb) {
 				cvm_oct_free_work(work);
 				continue;
@@ -268,13 +296,14 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 					else
 						ptr += 6;
 				}
-				memcpy(skb_put(skb, work->len), ptr, work->len);
+				memcpy(skb_put(skb, work->word1.len), ptr,
+				       work->word1.len);
 				/* No packet buffers to free */
 			} else {
 				int segments = work->word2.s.bufs;
 				union cvmx_buf_ptr segment_ptr =
 				    work->packet_ptr;
-				int len = work->len;
+				int len = work->word1.len;
 
 				while (segments--) {
 					union cvmx_buf_ptr next_ptr =
@@ -310,10 +339,9 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 			}
 			packet_not_copied = 0;
 		}
-
-		if (likely((work->ipprt < TOTAL_NUMBER_OF_PORTS) &&
-			   cvm_oct_device[work->ipprt])) {
-			struct net_device *dev = cvm_oct_device[work->ipprt];
+		if (likely((port < TOTAL_NUMBER_OF_PORTS) &&
+			   cvm_oct_device[port])) {
+			struct net_device *dev = cvm_oct_device[port];
 			struct octeon_ethernet *priv = netdev_priv(dev);
 
 			/*
@@ -333,7 +361,7 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 					skb->ip_summed = CHECKSUM_UNNECESSARY;
 
 				/* Increment RX stats for virtual ports */
-				if (work->ipprt >= CVMX_PIP_NUM_INPUT_PORTS) {
+				if (port >= CVMX_PIP_NUM_INPUT_PORTS) {
 #ifdef CONFIG_64BIT
 					atomic64_add(1,
 						     (atomic64_t *)&priv->stats.rx_packets);
@@ -368,7 +396,7 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 			 * doesn't exist.
 			 */
 			printk_ratelimited("Port %d not controlled by Linux, packet dropped\n",
-				   work->ipprt);
+				   port);
 			dev_kfree_skb_irq(skb);
 		}
 		/*
@@ -390,7 +418,13 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, int budget)
 		}
 	}
 	/* Restore the original POW group mask */
-	cvmx_write_csr(CVMX_POW_PP_GRP_MSKX(coreid), old_group_mask);
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+		cvmx_write_csr(CVMX_SSO_PPX_GRP_MSK(coreid), old_group_mask);
+		cvmx_read_csr(CVMX_SSO_PPX_GRP_MSK(coreid)); /* Flush */
+	} else {
+		cvmx_write_csr(CVMX_POW_PP_GRP_MSKX(coreid), old_group_mask);
+	}
+
 	if (USE_ASYNC_IOBDMA) {
 		/* Restore the scratch area */
 		cvmx_scratch_write64(CVMX_SCR_SCRATCH, old_scratch);
@@ -422,8 +456,6 @@ void cvm_oct_rx_initialize(void)
 {
 	int i;
 	struct net_device *dev_for_napi = NULL;
-	union cvmx_pow_wq_int_thrx int_thr;
-	union cvmx_pow_wq_int_pc int_pc;
 
 	for (i = 0; i < TOTAL_NUMBER_OF_PORTS; i++) {
 		if (cvm_oct_device[i]) {
@@ -449,15 +481,34 @@ void cvm_oct_rx_initialize(void)
 
 	disable_irq_nosync(OCTEON_IRQ_WORKQ0 + pow_receive_group);
 
-	int_thr.u64 = 0;
-	int_thr.s.tc_en = 1;
-	int_thr.s.tc_thr = 1;
 	/* Enable POW interrupt when our port has at least one packet */
-	cvmx_write_csr(CVMX_POW_WQ_INT_THRX(pow_receive_group), int_thr.u64);
-
-	int_pc.u64 = 0;
-	int_pc.s.pc_thr = 5;
-	cvmx_write_csr(CVMX_POW_WQ_INT_PC, int_pc.u64);
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
+		union cvmx_sso_wq_int_thrx int_thr;
+		union cvmx_pow_wq_int_pc int_pc;
+
+		int_thr.u64 = 0;
+		int_thr.s.tc_en = 1;
+		int_thr.s.tc_thr = 1;
+		cvmx_write_csr(CVMX_SSO_WQ_INT_THRX(pow_receive_group),
+			       int_thr.u64);
+
+		int_pc.u64 = 0;
+		int_pc.s.pc_thr = 5;
+		cvmx_write_csr(CVMX_SSO_WQ_INT_PC, int_pc.u64);
+	} else {
+		union cvmx_pow_wq_int_thrx int_thr;
+		union cvmx_pow_wq_int_pc int_pc;
+
+		int_thr.u64 = 0;
+		int_thr.s.tc_en = 1;
+		int_thr.s.tc_thr = 1;
+		cvmx_write_csr(CVMX_POW_WQ_INT_THRX(pow_receive_group),
+			       int_thr.u64);
+
+		int_pc.u64 = 0;
+		int_pc.s.pc_thr = 5;
+		cvmx_write_csr(CVMX_POW_WQ_INT_PC, int_pc.u64);
+	}
 
 	/* Schedule NAPI now. This will indirectly enable the interrupt. */
 	napi_schedule(&cvm_oct_napi);
diff --git a/drivers/staging/octeon/ethernet-tx.c b/drivers/staging/octeon/ethernet-tx.c
index e2df041..9e2116f 100644
--- a/drivers/staging/octeon/ethernet-tx.c
+++ b/drivers/staging/octeon/ethernet-tx.c
@@ -591,13 +591,14 @@ int cvm_oct_xmit_pow(struct sk_buff *skb, struct net_device *dev)
 	 * Fill in some of the work queue fields. We may need to add
 	 * more if the software at the other end needs them.
 	 */
-	work->hw_chksum = skb->csum;
-	work->len = skb->len;
-	work->ipprt = priv->port;
-	work->qos = priv->port & 0x7;
-	work->grp = pow_send_group;
-	work->tag_type = CVMX_HELPER_INPUT_TAG_TYPE;
-	work->tag = pow_send_group;	/* FIXME */
+	if (!OCTEON_IS_MODEL(OCTEON_CN68XX))
+		work->word0.pip.cn38xx.hw_chksum = skb->csum;
+	work->word1.len = skb->len;
+	cvmx_wqe_set_port(work, priv->port);
+	cvmx_wqe_set_qos(work, priv->port & 0x7);
+	cvmx_wqe_set_grp(work, pow_send_group);
+	work->word1.tag_type = CVMX_HELPER_INPUT_TAG_TYPE;
+	work->word1.tag = pow_send_group;	/* FIXME */
 	/* Default to zero. Sets of zero later are commented out */
 	work->word2.u64 = 0;
 	work->word2.s.bufs = 1;
@@ -677,8 +678,8 @@ int cvm_oct_xmit_pow(struct sk_buff *skb, struct net_device *dev)
 	}
 
 	/* Submit the packet to the POW */
-	cvmx_pow_work_submit(work, work->tag, work->tag_type, work->qos,
-			     work->grp);
+	cvmx_pow_work_submit(work, work->word1.tag, work->word1.tag_type,
+			     cvmx_wqe_get_qos(work), cvmx_wqe_get_grp(work));
 	priv->stats.tx_packets++;
 	priv->stats.tx_bytes += skb->len;
 	dev_consume_skb_any(skb);
diff --git a/drivers/staging/octeon/ethernet-util.h b/drivers/staging/octeon/ethernet-util.h
index 1ba789a..45f024b 100644
--- a/drivers/staging/octeon/ethernet-util.h
+++ b/drivers/staging/octeon/ethernet-util.h
@@ -8,6 +8,10 @@
  * published by the Free Software Foundation.
  */
 
+#include <asm/octeon/cvmx-pip.h>
+#include <asm/octeon/cvmx-helper.h>
+#include <asm/octeon/cvmx-helper-util.h>
+
 /**
  * cvm_oct_get_buffer_ptr - convert packet data address to pointer
  * @packet_ptr: Packet data hardware address
@@ -28,14 +32,12 @@ static inline void *cvm_oct_get_buffer_ptr(union cvmx_buf_ptr packet_ptr)
  */
 static inline int INTERFACE(int ipd_port)
 {
-	if (ipd_port < 32)	/* Interface 0 or 1 for RGMII,GMII,SPI, etc */
-		return ipd_port >> 4;
-	else if (ipd_port < 36)	/* Interface 2 for NPI */
-		return 2;
-	else if (ipd_port < 40)	/* Interface 3 for loopback */
-		return 3;
-	else if (ipd_port == 40)	/* Non existent interface for POW0 */
-		return 4;
+	int interface = cvmx_helper_get_interface_num(ipd_port);
+
+	if (interface >= 0)
+		return interface;
+	else if (ipd_port == CVMX_PIP_NUM_INPUT_PORTS)
+		return 10;
 	panic("Illegal ipd_port %d passed to INTERFACE\n", ipd_port);
 }
 
@@ -47,7 +49,5 @@ static inline int INTERFACE(int ipd_port)
  */
 static inline int INDEX(int ipd_port)
 {
-	if (ipd_port < 32)
-		return ipd_port & 15;
-	return ipd_port & 3;
+	return cvmx_helper_get_interface_index_num(ipd_port);
 }
diff --git a/drivers/staging/octeon/ethernet.c b/drivers/staging/octeon/ethernet.c
index 0718b35..7274fda 100644
--- a/drivers/staging/octeon/ethernet.c
+++ b/drivers/staging/octeon/ethernet.c
@@ -152,7 +152,7 @@ static void cvm_oct_configure_common_hw(void)
 			     num_packet_buffers);
 	if (CVMX_FPA_OUTPUT_BUFFER_POOL != CVMX_FPA_PACKET_POOL)
 		cvm_oct_mem_fill_fpa(CVMX_FPA_OUTPUT_BUFFER_POOL,
-				     CVMX_FPA_OUTPUT_BUFFER_POOL_SIZE, 128);
+				     CVMX_FPA_OUTPUT_BUFFER_POOL_SIZE, 1024);
 
 #ifdef __LITTLE_ENDIAN
 	{
@@ -860,7 +860,10 @@ static int cvm_oct_remove(struct platform_device *pdev)
 	int port;
 
 	/* Disable POW interrupt */
-	cvmx_write_csr(CVMX_POW_WQ_INT_THRX(pow_receive_group), 0);
+	if (OCTEON_IS_MODEL(OCTEON_CN68XX))
+		cvmx_write_csr(CVMX_SSO_WQ_INT_THRX(pow_receive_group), 0);
+	else
+		cvmx_write_csr(CVMX_POW_WQ_INT_THRX(pow_receive_group), 0);
 
 	cvmx_ipd_disable();
 
diff --git a/drivers/tty/mips_ejtag_fdc.c b/drivers/tty/mips_ejtag_fdc.c
index 358323c..a8c8cfd 100644
--- a/drivers/tty/mips_ejtag_fdc.c
+++ b/drivers/tty/mips_ejtag_fdc.c
@@ -879,6 +879,11 @@ static const struct tty_operations mips_ejtag_fdc_tty_ops = {
 	.chars_in_buffer	= mips_ejtag_fdc_tty_chars_in_buffer,
 };
 
+int __weak get_c0_fdc_int(void)
+{
+	return -1;
+}
+
 static int mips_ejtag_fdc_tty_probe(struct mips_cdmm_device *dev)
 {
 	int ret, nport;
@@ -967,9 +972,7 @@ static int mips_ejtag_fdc_tty_probe(struct mips_cdmm_device *dev)
 	wake_up_process(priv->thread);
 
 	/* Look for an FDC IRQ */
-	priv->irq = -1;
-	if (get_c0_fdc_int)
-		priv->irq = get_c0_fdc_int();
+	priv->irq = get_c0_fdc_int();
 
 	/* Try requesting the IRQ */
 	if (priv->irq >= 0) {
diff --git a/include/linux/irqchip/mips-gic.h b/include/linux/irqchip/mips-gic.h
index 9b1ad37..4e68616 100644
--- a/include/linux/irqchip/mips-gic.h
+++ b/include/linux/irqchip/mips-gic.h
@@ -41,12 +41,20 @@
 
 /* Shared Global Counter */
 #define GIC_SH_COUNTER_31_00_OFS	0x0010
+/* 64-bit counter register for CM3 */
+#define GIC_SH_COUNTER_OFS		GIC_SH_COUNTER_31_00_OFS
 #define GIC_SH_COUNTER_63_32_OFS	0x0014
 #define GIC_SH_REVISIONID_OFS		0x0020
 
 /* Convert an interrupt number to a byte offset/bit for multi-word registers */
-#define GIC_INTR_OFS(intr)		(((intr) / 32) * 4)
-#define GIC_INTR_BIT(intr)		((intr) % 32)
+#define GIC_INTR_OFS(intr) ({				\
+	unsigned bits = mips_cm_is64 ? 64 : 32;		\
+	unsigned reg_idx = (intr) / bits;		\
+	unsigned reg_width = bits / 8;			\
+							\
+	reg_idx * reg_width;				\
+})
+#define GIC_INTR_BIT(intr)		((intr) % (mips_cm_is64 ? 64 : 32))
 
 /* Polarity : Reset Value is always 0 */
 #define GIC_SH_SET_POLARITY_OFS		0x0100
@@ -98,6 +106,8 @@
 #define GIC_VPE_WD_COUNT0_OFS		0x0094
 #define GIC_VPE_WD_INITIAL0_OFS		0x0098
 #define GIC_VPE_COMPARE_LO_OFS		0x00a0
+/* 64-bit Compare register on CM3 */
+#define GIC_VPE_COMPARE_OFS		GIC_VPE_COMPARE_LO_OFS
 #define GIC_VPE_COMPARE_HI_OFS		0x00a4
 
 #define GIC_VPE_EIC_SHADOW_SET_BASE_OFS	0x0100
