Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/torvalds/linux-2.6 into common/serial-rework

Conflicts:
	arch/sh/kernel/cpu/sh2/setup-sh7619.c
	arch/sh/kernel/cpu/sh2a/setup-mxg.c
	arch/sh/kernel/cpu/sh2a/setup-sh7201.c
	arch/sh/kernel/cpu/sh2a/setup-sh7203.c
	arch/sh/kernel/cpu/sh2a/setup-sh7206.c
	arch/sh/kernel/cpu/sh3/setup-sh7705.c
	arch/sh/kernel/cpu/sh3/setup-sh770x.c
	arch/sh/kernel/cpu/sh3/setup-sh7710.c
	arch/sh/kernel/cpu/sh3/setup-sh7720.c
	arch/sh/kernel/cpu/sh4/setup-sh4-202.c
	arch/sh/kernel/cpu/sh4/setup-sh7750.c
	arch/sh/kernel/cpu/sh4/setup-sh7760.c
	arch/sh/kernel/cpu/sh4a/setup-sh7343.c
	arch/sh/kernel/cpu/sh4a/setup-sh7366.c
	arch/sh/kernel/cpu/sh4a/setup-sh7722.c
	arch/sh/kernel/cpu/sh4a/setup-sh7723.c
	arch/sh/kernel/cpu/sh4a/setup-sh7724.c
	arch/sh/kernel/cpu/sh4a/setup-sh7763.c
	arch/sh/kernel/cpu/sh4a/setup-sh7770.c
	arch/sh/kernel/cpu/sh4a/setup-sh7780.c
	arch/sh/kernel/cpu/sh4a/setup-sh7785.c
	arch/sh/kernel/cpu/sh4a/setup-sh7786.c
	arch/sh/kernel/cpu/sh4a/setup-shx3.c
	arch/sh/kernel/cpu/sh5/setup-sh5.c
	drivers/serial/sh-sci.c
	drivers/serial/sh-sci.h
	include/linux/serial_sci.h

10 files changed, 462 insertions, 258 deletions

diff --git a/arch/sh/kernel/cpu/sh4/Makefile b/arch/sh/kernel/cpu/sh4/Makefile
index 203b18347b83..3a1dbc709831 100644
--- a/arch/sh/kernel/cpu/sh4/Makefile
+++ b/arch/sh/kernel/cpu/sh4/Makefile
@@ -9,6 +9,11 @@ obj-$(CONFIG_HIBERNATION)		+= $(addprefix ../sh3/, swsusp.o)
 obj-$(CONFIG_SH_FPU)			+= fpu.o softfloat.o
 obj-$(CONFIG_SH_STORE_QUEUES)		+= sq.o
 
+# Perf events
+perf-$(CONFIG_CPU_SUBTYPE_SH7750)	:= perf_event.o
+perf-$(CONFIG_CPU_SUBTYPE_SH7750S)	:= perf_event.o
+perf-$(CONFIG_CPU_SUBTYPE_SH7091)	:= perf_event.o
+
 # CPU subtype setup
 obj-$(CONFIG_CPU_SUBTYPE_SH7750)	+= setup-sh7750.o
 obj-$(CONFIG_CPU_SUBTYPE_SH7750R)	+= setup-sh7750.o
@@ -27,4 +32,5 @@ endif
 # Additional clocks by subtype
 clock-$(CONFIG_CPU_SUBTYPE_SH4_202)	+= clock-sh4-202.o
 
-obj-y	+= $(clock-y)
+obj-y					+= $(clock-y)
+obj-$(CONFIG_PERF_EVENTS)		+= $(perf-y)
diff --git a/arch/sh/kernel/cpu/sh4/clock-sh4-202.c b/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
index 21421e34e7d5..3f6f8e98635c 100644
--- a/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
+++ b/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
@@ -12,9 +12,10 @@
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/err.h>
+#include <linux/io.h>
+#include <linux/clkdev.h>
 #include <asm/clock.h>
 #include <asm/freq.h>
-#include <asm/io.h>
 
 #define CPG2_FRQCR3	0xfe0a0018
 
@@ -23,7 +24,7 @@ static int frqcr3_values[]   = { 0, 1, 2, 3, 4, 5, 6  };
 
 static unsigned long emi_clk_recalc(struct clk *clk)
 {
-	int idx = ctrl_inl(CPG2_FRQCR3) & 0x0007;
+	int idx = __raw_readl(CPG2_FRQCR3) & 0x0007;
 	return clk->parent->rate / frqcr3_divisors[idx];
 }
 
@@ -45,14 +46,13 @@ static struct clk_ops sh4202_emi_clk_ops = {
 };
 
 static struct clk sh4202_emi_clk = {
-	.name		= "emi_clk",
 	.flags		= CLK_ENABLE_ON_INIT,
 	.ops		= &sh4202_emi_clk_ops,
 };
 
 static unsigned long femi_clk_recalc(struct clk *clk)
 {
-	int idx = (ctrl_inl(CPG2_FRQCR3) >> 3) & 0x0007;
+	int idx = (__raw_readl(CPG2_FRQCR3) >> 3) & 0x0007;
 	return clk->parent->rate / frqcr3_divisors[idx];
 }
 
@@ -61,7 +61,6 @@ static struct clk_ops sh4202_femi_clk_ops = {
 };
 
 static struct clk sh4202_femi_clk = {
-	.name		= "femi_clk",
 	.flags		= CLK_ENABLE_ON_INIT,
 	.ops		= &sh4202_femi_clk_ops,
 };
@@ -82,8 +81,7 @@ static void shoc_clk_init(struct clk *clk)
 	for (i = 0; i < ARRAY_SIZE(frqcr3_divisors); i++) {
 		int divisor = frqcr3_divisors[i];
 
-		if (clk->ops->set_rate(clk, clk->parent->rate /
-						divisor, 0) == 0)
+		if (clk->ops->set_rate(clk, clk->parent->rate / divisor) == 0)
 			break;
 	}
 
@@ -92,7 +90,7 @@ static void shoc_clk_init(struct clk *clk)
 
 static unsigned long shoc_clk_recalc(struct clk *clk)
 {
-	int idx = (ctrl_inl(CPG2_FRQCR3) >> 6) & 0x0007;
+	int idx = (__raw_readl(CPG2_FRQCR3) >> 6) & 0x0007;
 	return clk->parent->rate / frqcr3_divisors[idx];
 }
 
@@ -111,7 +109,7 @@ static int shoc_clk_verify_rate(struct clk *clk, unsigned long rate)
 	return 0;
 }
 
-static int shoc_clk_set_rate(struct clk *clk, unsigned long rate, int algo_id)
+static int shoc_clk_set_rate(struct clk *clk, unsigned long rate)
 {
 	unsigned long frqcr3;
 	unsigned int tmp;
@@ -122,10 +120,10 @@ static int shoc_clk_set_rate(struct clk *clk, unsigned long rate, int algo_id)
 
 	tmp = frqcr3_lookup(clk, rate);
 
-	frqcr3 = ctrl_inl(CPG2_FRQCR3);
+	frqcr3 = __raw_readl(CPG2_FRQCR3);
 	frqcr3 &= ~(0x0007 << 6);
 	frqcr3 |= tmp << 6;
-	ctrl_outl(frqcr3, CPG2_FRQCR3);
+	__raw_writel(frqcr3, CPG2_FRQCR3);
 
 	clk->rate = clk->parent->rate / frqcr3_divisors[tmp];
 
@@ -139,7 +137,6 @@ static struct clk_ops sh4202_shoc_clk_ops = {
 };
 
 static struct clk sh4202_shoc_clk = {
-	.name		= "shoc_clk",
 	.flags		= CLK_ENABLE_ON_INIT,
 	.ops		= &sh4202_shoc_clk_ops,
 };
@@ -150,6 +147,15 @@ static struct clk *sh4202_onchip_clocks[] = {
 	&sh4202_shoc_clk,
 };
 
+#define CLKDEV_CON_ID(_id, _clk) { .con_id = _id, .clk = _clk }
+
+static struct clk_lookup lookups[] = {
+	/* main clocks */
+	CLKDEV_CON_ID("emi_clk", &sh4202_emi_clk),
+	CLKDEV_CON_ID("femi_clk", &sh4202_femi_clk),
+	CLKDEV_CON_ID("shoc_clk", &sh4202_shoc_clk),
+};
+
 int __init arch_clk_init(void)
 {
 	struct clk *clk;
@@ -167,5 +173,7 @@ int __init arch_clk_init(void)
 
 	clk_put(clk);
 
+	clkdev_add_table(lookups, ARRAY_SIZE(lookups));
+
 	return ret;
 }
diff --git a/arch/sh/kernel/cpu/sh4/clock-sh4.c b/arch/sh/kernel/cpu/sh4/clock-sh4.c
index 73294d9cd049..5add75c1f539 100644
--- a/arch/sh/kernel/cpu/sh4/clock-sh4.c
+++ b/arch/sh/kernel/cpu/sh4/clock-sh4.c
@@ -28,7 +28,7 @@ static int pfc_divisors[] = { 2, 3, 4, 6, 8, 2, 2, 2 };
 
 static void master_clk_init(struct clk *clk)
 {
-	clk->rate *= pfc_divisors[ctrl_inw(FRQCR) & 0x0007];
+	clk->rate *= pfc_divisors[__raw_readw(FRQCR) & 0x0007];
 }
 
 static struct clk_ops sh4_master_clk_ops = {
@@ -37,7 +37,7 @@ static struct clk_ops sh4_master_clk_ops = {
 
 static unsigned long module_clk_recalc(struct clk *clk)
 {
-	int idx = (ctrl_inw(FRQCR) & 0x0007);
+	int idx = (__raw_readw(FRQCR) & 0x0007);
 	return clk->parent->rate / pfc_divisors[idx];
 }
 
@@ -47,7 +47,7 @@ static struct clk_ops sh4_module_clk_ops = {
 
 static unsigned long bus_clk_recalc(struct clk *clk)
 {
-	int idx = (ctrl_inw(FRQCR) >> 3) & 0x0007;
+	int idx = (__raw_readw(FRQCR) >> 3) & 0x0007;
 	return clk->parent->rate / bfc_divisors[idx];
 }
 
@@ -57,7 +57,7 @@ static struct clk_ops sh4_bus_clk_ops = {
 
 static unsigned long cpu_clk_recalc(struct clk *clk)
 {
-	int idx = (ctrl_inw(FRQCR) >> 6) & 0x0007;
+	int idx = (__raw_readw(FRQCR) >> 6) & 0x0007;
 	return clk->parent->rate / ifc_divisors[idx];
 }
 
diff --git a/arch/sh/kernel/cpu/sh4/fpu.c b/arch/sh/kernel/cpu/sh4/fpu.c
index e3ea5411da6d..447482d7f65e 100644
--- a/arch/sh/kernel/cpu/sh4/fpu.c
+++ b/arch/sh/kernel/cpu/sh4/fpu.c
@@ -41,13 +41,11 @@ static unsigned int fpu_exception_flags;
 
 /*
  * Save FPU registers onto task structure.
- * Assume called with FPU enabled (SR.FD=0).
  */
-void save_fpu(struct task_struct *tsk, struct pt_regs *regs)
+void save_fpu(struct task_struct *tsk)
 {
 	unsigned long dummy;
 
-	clear_tsk_thread_flag(tsk, TIF_USEDFPU);
 	enable_fpu();
 	asm volatile ("sts.l	fpul, @-%0\n\t"
 		      "sts.l	fpscr, @-%0\n\t"
@@ -87,15 +85,14 @@ void save_fpu(struct task_struct *tsk, struct pt_regs *regs)
 		      "fmov.s	fr1, @-%0\n\t"
 		      "fmov.s	fr0, @-%0\n\t"
 		      "lds	%3, fpscr\n\t":"=r" (dummy)
-		      :"0"((char *)(&tsk->thread.fpu.hard.status)),
+		      :"0"((char *)(&tsk->thread.xstate->hardfpu.status)),
 		      "r"(FPSCR_RCHG), "r"(FPSCR_INIT)
 		      :"memory");
 
 	disable_fpu();
-	release_fpu(regs);
 }
 
-static void restore_fpu(struct task_struct *tsk)
+void restore_fpu(struct task_struct *tsk)
 {
 	unsigned long dummy;
 
@@ -138,62 +135,11 @@ static void restore_fpu(struct task_struct *tsk)
 		      "lds.l	@%0+, fpscr\n\t"
 		      "lds.l	@%0+, fpul\n\t"
 		      :"=r" (dummy)
-		      :"0"(&tsk->thread.fpu), "r"(FPSCR_RCHG)
+		      :"0" (tsk->thread.xstate), "r" (FPSCR_RCHG)
 		      :"memory");
 	disable_fpu();
 }
 
-/*
- * Load the FPU with signalling NANS.  This bit pattern we're using
- * has the property that no matter wether considered as single or as
- * double precision represents signaling NANS.
- */
-
-static void fpu_init(void)
-{
-	enable_fpu();
-	asm volatile (	"lds	%0, fpul\n\t"
-			"lds	%1, fpscr\n\t"
-			"fsts	fpul, fr0\n\t"
-			"fsts	fpul, fr1\n\t"
-			"fsts	fpul, fr2\n\t"
-			"fsts	fpul, fr3\n\t"
-			"fsts	fpul, fr4\n\t"
-			"fsts	fpul, fr5\n\t"
-			"fsts	fpul, fr6\n\t"
-			"fsts	fpul, fr7\n\t"
-			"fsts	fpul, fr8\n\t"
-			"fsts	fpul, fr9\n\t"
-			"fsts	fpul, fr10\n\t"
-			"fsts	fpul, fr11\n\t"
-			"fsts	fpul, fr12\n\t"
-			"fsts	fpul, fr13\n\t"
-			"fsts	fpul, fr14\n\t"
-			"fsts	fpul, fr15\n\t"
-			"frchg\n\t"
-			"fsts	fpul, fr0\n\t"
-			"fsts	fpul, fr1\n\t"
-			"fsts	fpul, fr2\n\t"
-			"fsts	fpul, fr3\n\t"
-			"fsts	fpul, fr4\n\t"
-			"fsts	fpul, fr5\n\t"
-			"fsts	fpul, fr6\n\t"
-			"fsts	fpul, fr7\n\t"
-			"fsts	fpul, fr8\n\t"
-			"fsts	fpul, fr9\n\t"
-			"fsts	fpul, fr10\n\t"
-			"fsts	fpul, fr11\n\t"
-			"fsts	fpul, fr12\n\t"
-			"fsts	fpul, fr13\n\t"
-			"fsts	fpul, fr14\n\t"
-			"fsts	fpul, fr15\n\t"
-			"frchg\n\t"
-			"lds	%2, fpscr\n\t"
-			:	/* no output */
-			:"r" (0), "r"(FPSCR_RCHG), "r"(FPSCR_INIT));
-	disable_fpu();
-}
-
 /**
  *      denormal_to_double - Given denormalized float number,
  *                           store double float
@@ -285,10 +231,9 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 		/* fcnvsd */
 		struct task_struct *tsk = current;
 
-		save_fpu(tsk, regs);
-		if ((tsk->thread.fpu.hard.fpscr & FPSCR_CAUSE_ERROR))
+		if ((tsk->thread.xstate->hardfpu.fpscr & FPSCR_CAUSE_ERROR))
 			/* FPU error */
-			denormal_to_double(&tsk->thread.fpu.hard,
+			denormal_to_double(&tsk->thread.xstate->hardfpu,
 					   (finsn >> 8) & 0xf);
 		else
 			return 0;
@@ -304,9 +249,9 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 		n = (finsn >> 8) & 0xf;
 		m = (finsn >> 4) & 0xf;
-		hx = tsk->thread.fpu.hard.fp_regs[n];
-		hy = tsk->thread.fpu.hard.fp_regs[m];
-		fpscr = tsk->thread.fpu.hard.fpscr;
+		hx = tsk->thread.xstate->hardfpu.fp_regs[n];
+		hy = tsk->thread.xstate->hardfpu.fp_regs[m];
+		fpscr = tsk->thread.xstate->hardfpu.fpscr;
 		prec = fpscr & FPSCR_DBL_PRECISION;
 
 		if ((fpscr & FPSCR_CAUSE_ERROR)
@@ -316,18 +261,18 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 			/* FPU error because of denormal (doubles) */
 			llx = ((long long)hx << 32)
-			    | tsk->thread.fpu.hard.fp_regs[n + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[n + 1];
 			lly = ((long long)hy << 32)
-			    | tsk->thread.fpu.hard.fp_regs[m + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[m + 1];
 			llx = float64_mul(llx, lly);
-			tsk->thread.fpu.hard.fp_regs[n] = llx >> 32;
-			tsk->thread.fpu.hard.fp_regs[n + 1] = llx & 0xffffffff;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = llx >> 32;
+			tsk->thread.xstate->hardfpu.fp_regs[n + 1] = llx & 0xffffffff;
 		} else if ((fpscr & FPSCR_CAUSE_ERROR)
 			   && (!prec && ((hx & 0x7fffffff) < 0x00800000
 					 || (hy & 0x7fffffff) < 0x00800000))) {
 			/* FPU error because of denormal (floats) */
 			hx = float32_mul(hx, hy);
-			tsk->thread.fpu.hard.fp_regs[n] = hx;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = hx;
 		} else
 			return 0;
 
@@ -342,9 +287,9 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 		n = (finsn >> 8) & 0xf;
 		m = (finsn >> 4) & 0xf;
-		hx = tsk->thread.fpu.hard.fp_regs[n];
-		hy = tsk->thread.fpu.hard.fp_regs[m];
-		fpscr = tsk->thread.fpu.hard.fpscr;
+		hx = tsk->thread.xstate->hardfpu.fp_regs[n];
+		hy = tsk->thread.xstate->hardfpu.fp_regs[m];
+		fpscr = tsk->thread.xstate->hardfpu.fpscr;
 		prec = fpscr & FPSCR_DBL_PRECISION;
 
 		if ((fpscr & FPSCR_CAUSE_ERROR)
@@ -354,15 +299,15 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 			/* FPU error because of denormal (doubles) */
 			llx = ((long long)hx << 32)
-			    | tsk->thread.fpu.hard.fp_regs[n + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[n + 1];
 			lly = ((long long)hy << 32)
-			    | tsk->thread.fpu.hard.fp_regs[m + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[m + 1];
 			if ((finsn & 0xf00f) == 0xf000)
 				llx = float64_add(llx, lly);
 			else
 				llx = float64_sub(llx, lly);
-			tsk->thread.fpu.hard.fp_regs[n] = llx >> 32;
-			tsk->thread.fpu.hard.fp_regs[n + 1] = llx & 0xffffffff;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = llx >> 32;
+			tsk->thread.xstate->hardfpu.fp_regs[n + 1] = llx & 0xffffffff;
 		} else if ((fpscr & FPSCR_CAUSE_ERROR)
 			   && (!prec && ((hx & 0x7fffffff) < 0x00800000
 					 || (hy & 0x7fffffff) < 0x00800000))) {
@@ -371,7 +316,7 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 				hx = float32_add(hx, hy);
 			else
 				hx = float32_sub(hx, hy);
-			tsk->thread.fpu.hard.fp_regs[n] = hx;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = hx;
 		} else
 			return 0;
 
@@ -386,9 +331,9 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 		n = (finsn >> 8) & 0xf;
 		m = (finsn >> 4) & 0xf;
-		hx = tsk->thread.fpu.hard.fp_regs[n];
-		hy = tsk->thread.fpu.hard.fp_regs[m];
-		fpscr = tsk->thread.fpu.hard.fpscr;
+		hx = tsk->thread.xstate->hardfpu.fp_regs[n];
+		hy = tsk->thread.xstate->hardfpu.fp_regs[m];
+		fpscr = tsk->thread.xstate->hardfpu.fpscr;
 		prec = fpscr & FPSCR_DBL_PRECISION;
 
 		if ((fpscr & FPSCR_CAUSE_ERROR)
@@ -398,20 +343,20 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 
 			/* FPU error because of denormal (doubles) */
 			llx = ((long long)hx << 32)
-			    | tsk->thread.fpu.hard.fp_regs[n + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[n + 1];
 			lly = ((long long)hy << 32)
-			    | tsk->thread.fpu.hard.fp_regs[m + 1];
+			    | tsk->thread.xstate->hardfpu.fp_regs[m + 1];
 
 			llx = float64_div(llx, lly);
 
-			tsk->thread.fpu.hard.fp_regs[n] = llx >> 32;
-			tsk->thread.fpu.hard.fp_regs[n + 1] = llx & 0xffffffff;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = llx >> 32;
+			tsk->thread.xstate->hardfpu.fp_regs[n + 1] = llx & 0xffffffff;
 		} else if ((fpscr & FPSCR_CAUSE_ERROR)
 			   && (!prec && ((hx & 0x7fffffff) < 0x00800000
 					 || (hy & 0x7fffffff) < 0x00800000))) {
 			/* FPU error because of denormal (floats) */
 			hx = float32_div(hx, hy);
-			tsk->thread.fpu.hard.fp_regs[n] = hx;
+			tsk->thread.xstate->hardfpu.fp_regs[n] = hx;
 		} else
 			return 0;
 
@@ -424,17 +369,17 @@ static int ieee_fpe_handler(struct pt_regs *regs)
 		unsigned int hx;
 
 		m = (finsn >> 8) & 0x7;
-		hx = tsk->thread.fpu.hard.fp_regs[m];
+		hx = tsk->thread.xstate->hardfpu.fp_regs[m];
 
-		if ((tsk->thread.fpu.hard.fpscr & FPSCR_CAUSE_ERROR)
+		if ((tsk->thread.xstate->hardfpu.fpscr & FPSCR_CAUSE_ERROR)
 			&& ((hx & 0x7fffffff) < 0x00100000)) {
 			/* subnormal double to float conversion */
 			long long llx;
 
-			llx = ((long long)tsk->thread.fpu.hard.fp_regs[m] << 32)
-			    | tsk->thread.fpu.hard.fp_regs[m + 1];
+			llx = ((long long)tsk->thread.xstate->hardfpu.fp_regs[m] << 32)
+			    | tsk->thread.xstate->hardfpu.fp_regs[m + 1];
 
-			tsk->thread.fpu.hard.fpul = float64_to_float32(llx);
+			tsk->thread.xstate->hardfpu.fpul = float64_to_float32(llx);
 		} else
 			return 0;
 
@@ -453,7 +398,7 @@ void float_raise(unsigned int flags)
 int float_rounding_mode(void)
 {
 	struct task_struct *tsk = current;
-	int roundingMode = FPSCR_ROUNDING_MODE(tsk->thread.fpu.hard.fpscr);
+	int roundingMode = FPSCR_ROUNDING_MODE(tsk->thread.xstate->hardfpu.fpscr);
 	return roundingMode;
 }
 
@@ -462,19 +407,19 @@ BUILD_TRAP_HANDLER(fpu_error)
 	struct task_struct *tsk = current;
 	TRAP_HANDLER_DECL;
 
-	save_fpu(tsk, regs);
+	__unlazy_fpu(tsk, regs);
 	fpu_exception_flags = 0;
 	if (ieee_fpe_handler(regs)) {
-		tsk->thread.fpu.hard.fpscr &=
+		tsk->thread.xstate->hardfpu.fpscr &=
 		    ~(FPSCR_CAUSE_MASK | FPSCR_FLAG_MASK);
-		tsk->thread.fpu.hard.fpscr |= fpu_exception_flags;
+		tsk->thread.xstate->hardfpu.fpscr |= fpu_exception_flags;
 		/* Set the FPSCR flag as well as cause bits - simply
 		 * replicate the cause */
-		tsk->thread.fpu.hard.fpscr |= (fpu_exception_flags >> 10);
+		tsk->thread.xstate->hardfpu.fpscr |= (fpu_exception_flags >> 10);
 		grab_fpu(regs);
 		restore_fpu(tsk);
-		set_tsk_thread_flag(tsk, TIF_USEDFPU);
-		if ((((tsk->thread.fpu.hard.fpscr & FPSCR_ENABLE_MASK) >> 7) &
+		task_thread_info(tsk)->status |= TS_USEDFPU;
+		if ((((tsk->thread.xstate->hardfpu.fpscr & FPSCR_ENABLE_MASK) >> 7) &
 		     (fpu_exception_flags >> 2)) == 0) {
 			return;
 		}
@@ -482,25 +427,3 @@ BUILD_TRAP_HANDLER(fpu_error)
 
 	force_sig(SIGFPE, tsk);
 }
-
-BUILD_TRAP_HANDLER(fpu_state_restore)
-{
-	struct task_struct *tsk = current;
-	TRAP_HANDLER_DECL;
-
-	grab_fpu(regs);
-	if (!user_mode(regs)) {
-		printk(KERN_ERR "BUG: FPU is used in kernel mode.\n");
-		return;
-	}
-
-	if (used_math()) {
-		/* Using the FPU again.  */
-		restore_fpu(tsk);
-	} else {
-		/* First time FPU user.  */
-		fpu_init();
-		set_used_math();
-	}
-	set_tsk_thread_flag(tsk, TIF_USEDFPU);
-}
diff --git a/arch/sh/kernel/cpu/sh4/perf_event.c b/arch/sh/kernel/cpu/sh4/perf_event.c
new file mode 100644
index 000000000000..748955df018d
--- /dev/null
+++ b/arch/sh/kernel/cpu/sh4/perf_event.c
@@ -0,0 +1,253 @@
+/*
+ * Performance events support for SH7750-style performance counters
+ *
+ *  Copyright (C) 2009  Paul Mundt
+ *
+ * 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.
+ */
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/perf_event.h>
+#include <asm/processor.h>
+
+#define PM_CR_BASE	0xff000084	/* 16-bit */
+#define PM_CTR_BASE	0xff100004	/* 32-bit */
+
+#define PMCR(n)		(PM_CR_BASE + ((n) * 0x04))
+#define PMCTRH(n)	(PM_CTR_BASE + 0x00 + ((n) * 0x08))
+#define PMCTRL(n)	(PM_CTR_BASE + 0x04 + ((n) * 0x08))
+
+#define PMCR_PMM_MASK	0x0000003f
+
+#define PMCR_CLKF	0x00000100
+#define PMCR_PMCLR	0x00002000
+#define PMCR_PMST	0x00004000
+#define PMCR_PMEN	0x00008000
+
+static struct sh_pmu sh7750_pmu;
+
+/*
+ * There are a number of events supported by each counter (33 in total).
+ * Since we have 2 counters, each counter will take the event code as it
+ * corresponds to the PMCR PMM setting. Each counter can be configured
+ * independently.
+ *
+ *	Event Code	Description
+ *	----------	-----------
+ *
+ *	0x01		Operand read access
+ *	0x02		Operand write access
+ *	0x03		UTLB miss
+ *	0x04		Operand cache read miss
+ *	0x05		Operand cache write miss
+ *	0x06		Instruction fetch (w/ cache)
+ *	0x07		Instruction TLB miss
+ *	0x08		Instruction cache miss
+ *	0x09		All operand accesses
+ *	0x0a		All instruction accesses
+ *	0x0b		OC RAM operand access
+ *	0x0d		On-chip I/O space access
+ *	0x0e		Operand access (r/w)
+ *	0x0f		Operand cache miss (r/w)
+ *	0x10		Branch instruction
+ *	0x11		Branch taken
+ *	0x12		BSR/BSRF/JSR
+ *	0x13		Instruction execution
+ *	0x14		Instruction execution in parallel
+ *	0x15		FPU Instruction execution
+ *	0x16		Interrupt
+ *	0x17		NMI
+ *	0x18		trapa instruction execution
+ *	0x19		UBCA match
+ *	0x1a		UBCB match
+ *	0x21		Instruction cache fill
+ *	0x22		Operand cache fill
+ *	0x23		Elapsed time
+ *	0x24		Pipeline freeze by I-cache miss
+ *	0x25		Pipeline freeze by D-cache miss
+ *	0x27		Pipeline freeze by branch instruction
+ *	0x28		Pipeline freeze by CPU register
+ *	0x29		Pipeline freeze by FPU
+ */
+
+static const int sh7750_general_events[] = {
+	[PERF_COUNT_HW_CPU_CYCLES]		= 0x0023,
+	[PERF_COUNT_HW_INSTRUCTIONS]		= 0x000a,
+	[PERF_COUNT_HW_CACHE_REFERENCES]	= 0x0006,	/* I-cache */
+	[PERF_COUNT_HW_CACHE_MISSES]		= 0x0008,	/* I-cache */
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= 0x0010,
+	[PERF_COUNT_HW_BRANCH_MISSES]		= -1,
+	[PERF_COUNT_HW_BUS_CYCLES]		= -1,
+};
+
+#define C(x)	PERF_COUNT_HW_CACHE_##x
+
+static const int sh7750_cache_events
+			[PERF_COUNT_HW_CACHE_MAX]
+			[PERF_COUNT_HW_CACHE_OP_MAX]
+			[PERF_COUNT_HW_CACHE_RESULT_MAX] =
+{
+	[ C(L1D) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = 0x0001,
+			[ C(RESULT_MISS)   ] = 0x0004,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = 0x0002,
+			[ C(RESULT_MISS)   ] = 0x0005,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+	},
+
+	[ C(L1I) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = 0x0006,
+			[ C(RESULT_MISS)   ] = 0x0008,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+	},
+
+	[ C(LL) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+	},
+
+	[ C(DTLB) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0x0003,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0,
+		},
+	},
+
+	[ C(ITLB) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = 0,
+			[ C(RESULT_MISS)   ] = 0x0007,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+	},
+
+	[ C(BPU) ] = {
+		[ C(OP_READ) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+		[ C(OP_WRITE) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+		[ C(OP_PREFETCH) ] = {
+			[ C(RESULT_ACCESS) ] = -1,
+			[ C(RESULT_MISS)   ] = -1,
+		},
+	},
+};
+
+static int sh7750_event_map(int event)
+{
+	return sh7750_general_events[event];
+}
+
+static u64 sh7750_pmu_read(int idx)
+{
+	return (u64)((u64)(__raw_readl(PMCTRH(idx)) & 0xffff) << 32) |
+			   __raw_readl(PMCTRL(idx));
+}
+
+static void sh7750_pmu_disable(struct hw_perf_event *hwc, int idx)
+{
+	unsigned int tmp;
+
+	tmp = __raw_readw(PMCR(idx));
+	tmp &= ~(PMCR_PMM_MASK | PMCR_PMEN);
+	__raw_writew(tmp, PMCR(idx));
+}
+
+static void sh7750_pmu_enable(struct hw_perf_event *hwc, int idx)
+{
+	__raw_writew(__raw_readw(PMCR(idx)) | PMCR_PMCLR, PMCR(idx));
+	__raw_writew(hwc->config | PMCR_PMEN | PMCR_PMST, PMCR(idx));
+}
+
+static void sh7750_pmu_disable_all(void)
+{
+	int i;
+
+	for (i = 0; i < sh7750_pmu.num_events; i++)
+		__raw_writew(__raw_readw(PMCR(i)) & ~PMCR_PMEN, PMCR(i));
+}
+
+static void sh7750_pmu_enable_all(void)
+{
+	int i;
+
+	for (i = 0; i < sh7750_pmu.num_events; i++)
+		__raw_writew(__raw_readw(PMCR(i)) | PMCR_PMEN, PMCR(i));
+}
+
+static struct sh_pmu sh7750_pmu = {
+	.name		= "sh7750",
+	.num_events	= 2,
+	.event_map	= sh7750_event_map,
+	.max_events	= ARRAY_SIZE(sh7750_general_events),
+	.raw_event_mask	= PMCR_PMM_MASK,
+	.cache_events	= &sh7750_cache_events,
+	.read		= sh7750_pmu_read,
+	.disable	= sh7750_pmu_disable,
+	.enable		= sh7750_pmu_enable,
+	.disable_all	= sh7750_pmu_disable_all,
+	.enable_all	= sh7750_pmu_enable_all,
+};
+
+static int __init sh7750_pmu_init(void)
+{
+	/*
+	 * Make sure this CPU actually has perf counters.
+	 */
+	if (!(boot_cpu_data.flags & CPU_HAS_PERF_COUNTER)) {
+		pr_notice("HW perf events unsupported, software events only.\n");
+		return -ENODEV;
+	}
+
+	return register_sh_pmu(&sh7750_pmu);
+}
+early_initcall(sh7750_pmu_init);
diff --git a/arch/sh/kernel/cpu/sh4/probe.c b/arch/sh/kernel/cpu/sh4/probe.c
index 6c78d0a9c857..b93458f33b74 100644
--- a/arch/sh/kernel/cpu/sh4/probe.c
+++ b/arch/sh/kernel/cpu/sh4/probe.c
@@ -15,7 +15,7 @@
 #include <asm/processor.h>
 #include <asm/cache.h>
 
-int __init detect_cpu_and_cache_system(void)
+void __cpuinit cpu_probe(void)
 {
 	unsigned long pvr, prr, cvr;
 	unsigned long size;
@@ -28,9 +28,9 @@ int __init detect_cpu_and_cache_system(void)
 		[9] = (1 << 16)
 	};
 
-	pvr = (ctrl_inl(CCN_PVR) >> 8) & 0xffffff;
-	prr = (ctrl_inl(CCN_PRR) >> 4) & 0xff;
-	cvr = (ctrl_inl(CCN_CVR));
+	pvr = (__raw_readl(CCN_PVR) >> 8) & 0xffffff;
+	prr = (__raw_readl(CCN_PRR) >> 4) & 0xff;
+	cvr = (__raw_readl(CCN_CVR));
 
 	/*
 	 * Setup some sane SH-4 defaults for the icache
@@ -57,8 +57,12 @@ int __init detect_cpu_and_cache_system(void)
 	 * Setup some generic flags we can probe on SH-4A parts
 	 */
 	if (((pvr >> 16) & 0xff) == 0x10) {
-		if ((cvr & 0x10000000) == 0)
+		boot_cpu_data.family = CPU_FAMILY_SH4A;
+
+		if ((cvr & 0x10000000) == 0) {
 			boot_cpu_data.flags |= CPU_HAS_DSP;
+			boot_cpu_data.family = CPU_FAMILY_SH4AL_DSP;
+		}
 
 		boot_cpu_data.flags |= CPU_HAS_LLSC | CPU_HAS_PERF_COUNTER;
 		boot_cpu_data.cut_major = pvr & 0x7f;
@@ -67,10 +71,11 @@ int __init detect_cpu_and_cache_system(void)
 		boot_cpu_data.dcache.ways = 4;
 	} else {
 		/* And some SH-4 defaults.. */
-		boot_cpu_data.flags |= CPU_HAS_PTEA;
+		boot_cpu_data.flags |= CPU_HAS_PTEA | CPU_HAS_FPU;
+		boot_cpu_data.family = CPU_FAMILY_SH4;
 	}
 
-	/* FPU detection works for everyone */
+	/* FPU detection works for almost everyone */
 	if ((cvr & 0x20000000))
 		boot_cpu_data.flags |= CPU_HAS_FPU;
 
@@ -119,6 +124,7 @@ int __init detect_cpu_and_cache_system(void)
 		boot_cpu_data.type = CPU_SH7785;
 		break;
 	case 0x4004:
+	case 0x4005:
 		boot_cpu_data.type = CPU_SH7786;
 		boot_cpu_data.flags |= CPU_HAS_PTEAEX | CPU_HAS_L2_CACHE;
 		break;
@@ -139,8 +145,15 @@ int __init detect_cpu_and_cache_system(void)
 		}
 		break;
 	case 0x300b:
-		boot_cpu_data.type = CPU_SH7724;
-		boot_cpu_data.flags |= CPU_HAS_L2_CACHE;
+		switch (prr) {
+		case 0x20:
+			boot_cpu_data.type = CPU_SH7724;
+			boot_cpu_data.flags |= CPU_HAS_L2_CACHE;
+			break;
+		case 0x10:
+			boot_cpu_data.type = CPU_SH7757;
+			break;
+		}
 		break;
 	case 0x4000:	/* 1st cut */
 	case 0x4001:	/* 2nd cut */
@@ -148,6 +161,7 @@ int __init detect_cpu_and_cache_system(void)
 		break;
 	case 0x700:
 		boot_cpu_data.type = CPU_SH4_501;
+		boot_cpu_data.flags &= ~CPU_HAS_FPU;
 		boot_cpu_data.icache.ways = 2;
 		boot_cpu_data.dcache.ways = 2;
 		break;
@@ -173,9 +187,6 @@ int __init detect_cpu_and_cache_system(void)
 		boot_cpu_data.dcache.ways = 2;
 
 		break;
-	default:
-		boot_cpu_data.type = CPU_SH_NONE;
-		break;
 	}
 
 	/*
@@ -218,7 +229,7 @@ int __init detect_cpu_and_cache_system(void)
 			 * Size calculation is much more sensible
 			 * than it is for the L1.
 			 *
-			 * Sizes are 128KB, 258KB, 512KB, and 1MB.
+			 * Sizes are 128KB, 256KB, 512KB, and 1MB.
 			 */
 			size = (cvr & 0xf) << 17;
 
@@ -240,6 +251,4 @@ int __init detect_cpu_and_cache_system(void)
 				 boot_cpu_data.scache.linesz);
 		}
 	}
-
-	return 0;
 }
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
index ec2104b49ef7..5b2833159b7d 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
@@ -15,38 +15,31 @@
 #include <linux/sh_timer.h>
 #include <linux/io.h>
 
-static struct plat_sci_port sci_platform_data[] = {
-	{
-		.mapbase	= 0xffe80000,
-		.flags		= UPF_BOOT_AUTOCONF,
-		.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
-		.scbrr_algo_id	= SCBRR_ALGO_2,
-		.type		= PORT_SCIF,
-		.irqs		= { 40, 41, 43, 42 },
-	}, {
-		.flags = 0,
-	}
+static struct plat_sci_port scif0_platform_data = {
+	.mapbase	= 0xffe80000,
+	.flags		= UPF_BOOT_AUTOCONF,
+	.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
+	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.type		= PORT_SCIF,
+	.irqs		= { 40, 41, 43, 42 },
 };
 
-static struct platform_device sci_device = {
+static struct platform_device scif0_device = {
 	.name		= "sh-sci",
-	.id		= -1,
+	.id		= 0,
 	.dev		= {
-		.platform_data	= sci_platform_data,
+		.platform_data	= &scif0_platform_data,
 	},
 };
 
 static struct sh_timer_config tmu0_platform_data = {
-	.name = "TMU0",
 	.channel_offset = 0x04,
 	.timer_bit = 0,
-	.clk = "peripheral_clk",
 	.clockevent_rating = 200,
 };
 
 static struct resource tmu0_resources[] = {
 	[0] = {
-		.name	= "TMU0",
 		.start	= 0xffd80008,
 		.end	= 0xffd80013,
 		.flags	= IORESOURCE_MEM,
@@ -68,16 +61,13 @@ static struct platform_device tmu0_device = {
 };
 
 static struct sh_timer_config tmu1_platform_data = {
-	.name = "TMU1",
 	.channel_offset = 0x10,
 	.timer_bit = 1,
-	.clk = "peripheral_clk",
 	.clocksource_rating = 200,
 };
 
 static struct resource tmu1_resources[] = {
 	[0] = {
-		.name	= "TMU1",
 		.start	= 0xffd80014,
 		.end	= 0xffd8001f,
 		.flags	= IORESOURCE_MEM,
@@ -99,15 +89,12 @@ static struct platform_device tmu1_device = {
 };
 
 static struct sh_timer_config tmu2_platform_data = {
-	.name = "TMU2",
 	.channel_offset = 0x1c,
 	.timer_bit = 2,
-	.clk = "peripheral_clk",
 };
 
 static struct resource tmu2_resources[] = {
 	[0] = {
-		.name	= "TMU2",
 		.start	= 0xffd80020,
 		.end	= 0xffd8002f,
 		.flags	= IORESOURCE_MEM,
@@ -129,7 +116,7 @@ static struct platform_device tmu2_device = {
 };
 
 static struct platform_device *sh4202_devices[] __initdata = {
-	&sci_device,
+	&scif0_device,
 	&tmu0_device,
 	&tmu1_device,
 	&tmu2_device,
@@ -140,9 +127,10 @@ static int __init sh4202_devices_setup(void)
 	return platform_add_devices(sh4202_devices,
 				    ARRAY_SIZE(sh4202_devices));
 }
-__initcall(sh4202_devices_setup);
+arch_initcall(sh4202_devices_setup);
 
 static struct platform_device *sh4202_early_devices[] __initdata = {
+	&scif0_device,
 	&tmu0_device,
 	&tmu1_device,
 	&tmu2_device,
@@ -203,7 +191,7 @@ void __init plat_irq_setup_pins(int mode)
 {
 	switch (mode) {
 	case IRQ_MODE_IRQ: /* individual interrupt mode for IRL3-0 */
-		ctrl_outw(ctrl_inw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
+		__raw_writew(__raw_readw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
 		register_intc_controller(&intc_desc_irlm);
 		break;
 	default:
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7750.c b/arch/sh/kernel/cpu/sh4/setup-sh7750.c
index 51a945e0d72c..c2b0aaaedcae 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh7750.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh7750.c
@@ -39,16 +39,17 @@ static struct platform_device rtc_device = {
 static struct plat_sci_port sci_platform_data = {
 	.mapbase	= 0xffe00000,
 	.flags		= UPF_BOOT_AUTOCONF,
-	.type		= PORT_SCI,
 	.scscr		= SCSCR_TE | SCSCR_RE,
 	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.type		= PORT_SCI,
 	.irqs		= { 23, 23, 23, 0 },
 };
 
 static struct platform_device sci_device = {
 	.name		= "sh-sci",
+	.id		= 0,
 	.dev		= {
-		.platform_data	= sci_platform_data,
+		.platform_data	= &sci_platform_data,
 	},
 };
 
@@ -63,22 +64,20 @@ static struct plat_sci_port scif_platform_data = {
 
 static struct platform_device scif_device = {
 	.name		= "sh-sci",
+	.id		= 1,
 	.dev		= {
-		.platform_data	= scif_platform_data,
+		.platform_data	= &scif_platform_data,
 	},
 };
 
 static struct sh_timer_config tmu0_platform_data = {
-	.name = "TMU0",
 	.channel_offset = 0x04,
 	.timer_bit = 0,
-	.clk = "peripheral_clk",
 	.clockevent_rating = 200,
 };
 
 static struct resource tmu0_resources[] = {
 	[0] = {
-		.name	= "TMU0",
 		.start	= 0xffd80008,
 		.end	= 0xffd80013,
 		.flags	= IORESOURCE_MEM,
@@ -100,16 +99,13 @@ static struct platform_device tmu0_device = {
 };
 
 static struct sh_timer_config tmu1_platform_data = {
-	.name = "TMU1",
 	.channel_offset = 0x10,
 	.timer_bit = 1,
-	.clk = "peripheral_clk",
 	.clocksource_rating = 200,
 };
 
 static struct resource tmu1_resources[] = {
 	[0] = {
-		.name	= "TMU1",
 		.start	= 0xffd80014,
 		.end	= 0xffd8001f,
 		.flags	= IORESOURCE_MEM,
@@ -131,15 +127,12 @@ static struct platform_device tmu1_device = {
 };
 
 static struct sh_timer_config tmu2_platform_data = {
-	.name = "TMU2",
 	.channel_offset = 0x1c,
 	.timer_bit = 2,
-	.clk = "peripheral_clk",
 };
 
 static struct resource tmu2_resources[] = {
 	[0] = {
-		.name	= "TMU2",
 		.start	= 0xffd80020,
 		.end	= 0xffd8002f,
 		.flags	= IORESOURCE_MEM,
@@ -166,15 +159,12 @@ static struct platform_device tmu2_device = {
 	defined(CONFIG_CPU_SUBTYPE_SH7751R)
 
 static struct sh_timer_config tmu3_platform_data = {
-	.name = "TMU3",
 	.channel_offset = 0x04,
 	.timer_bit = 0,
-	.clk = "peripheral_clk",
 };
 
 static struct resource tmu3_resources[] = {
 	[0] = {
-		.name	= "TMU3",
 		.start	= 0xfe100008,
 		.end	= 0xfe100013,
 		.flags	= IORESOURCE_MEM,
@@ -196,15 +186,12 @@ static struct platform_device tmu3_device = {
 };
 
 static struct sh_timer_config tmu4_platform_data = {
-	.name = "TMU4",
 	.channel_offset = 0x10,
 	.timer_bit = 1,
-	.clk = "peripheral_clk",
 };
 
 static struct resource tmu4_resources[] = {
 	[0] = {
-		.name	= "TMU4",
 		.start	= 0xfe100014,
 		.end	= 0xfe10001f,
 		.flags	= IORESOURCE_MEM,
@@ -243,7 +230,6 @@ static struct platform_device *sh7750_devices[] __initdata = {
 static int __init sh7750_devices_setup(void)
 {
 	if (mach_is_rts7751r2d()) {
-		scif_platform_data.scscr |= SCSCR_CKE1;
 		platform_register_device(&scif_device);
 	} else {
 		platform_register_device(&sci_device);
@@ -253,7 +239,7 @@ static int __init sh7750_devices_setup(void)
 	return platform_add_devices(sh7750_devices,
 				    ARRAY_SIZE(sh7750_devices));
 }
-__initcall(sh7750_devices_setup);
+arch_initcall(sh7750_devices_setup);
 
 static struct platform_device *sh7750_early_devices[] __initdata = {
 	&tmu0_device,
@@ -269,6 +255,14 @@ static struct platform_device *sh7750_early_devices[] __initdata = {
 
 void __init plat_early_device_setup(void)
 {
+	if (mach_is_rts7751r2d()) {
+		scif_platform_data.scscr |= SCSCR_CKE1;
+		early_platform_add_devices(&scif_device, 1);
+	} else {
+		early_platform_add_devices(&sci_device, 1);
+		early_platform_add_devices(&scif_device, 1);
+	}
+
 	early_platform_add_devices(sh7750_early_devices,
 				   ARRAY_SIZE(sh7750_early_devices));
 }
@@ -449,7 +443,7 @@ void __init plat_irq_setup_pins(int mode)
 
 	switch (mode) {
 	case IRQ_MODE_IRQ: /* individual interrupt mode for IRL3-0 */
-		ctrl_outw(ctrl_inw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
+		__raw_writew(__raw_readw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
 		register_intc_controller(&intc_desc_irlm);
 		break;
 	default:
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7760.c b/arch/sh/kernel/cpu/sh4/setup-sh7760.c
index cee660fe1d90..78bbf232e391 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh7760.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh7760.c
@@ -126,59 +126,82 @@ static struct intc_vect vectors_irq[] __initdata = {
 static DECLARE_INTC_DESC(intc_desc_irq, "sh7760-irq", vectors_irq, groups,
 			 mask_registers, prio_registers, NULL);
 
-static struct plat_sci_port sci_platform_data[] = {
-	{
-		.mapbase	= 0xfe600000,
-		.flags		= UPF_BOOT_AUTOCONF,
-		.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
-		.scbrr_algo_id	= SCBRR_ALGO_2,
-		.type		= PORT_SCIF,
-		.irqs		= { 52, 53, 55, 54 },
-	}, {
-		.mapbase	= 0xfe610000,
-		.flags		= UPF_BOOT_AUTOCONF,
-		.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
-		.scbrr_algo_id	= SCBRR_ALGO_2,
-		.type		= PORT_SCIF,
-		.irqs		= { 72, 73, 75, 74 },
-	}, {
-		.mapbase	= 0xfe620000,
-		.flags		= UPF_BOOT_AUTOCONF,
-		.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
-		.scbrr_algo_id	= SCBRR_ALGO_2,
-		.type		= PORT_SCIF,
-		.irqs		= { 76, 77, 79, 78 },
-	}, {
-		.mapbase	= 0xfe480000,
-		.flags		= UPF_BOOT_AUTOCONF,
-		.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
-		.scbrr_algo_id	= SCBRR_ALGO_2,
-		.type		= PORT_SCI,
-		.irqs		= { 80, 81, 82, 0 },
-	}, {
-		.flags = 0,
-	}
+static struct plat_sci_port scif0_platform_data = {
+	.mapbase	= 0xfe600000,
+	.flags		= UPF_BOOT_AUTOCONF,
+	.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
+	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.type		= PORT_SCIF,
+	.irqs		= { 52, 53, 55, 54 },
+};
+
+static struct platform_device scif0_device = {
+	.name		= "sh-sci",
+	.id		= 0,
+	.dev		= {
+		.platform_data	= &scif0_platform_data,
+	},
+};
+
+static struct plat_sci_port scif1_platform_data = {
+	.mapbase	= 0xfe610000,
+	.flags		= UPF_BOOT_AUTOCONF,
+	.type		= PORT_SCIF,
+	.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
+	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.irqs		= { 72, 73, 75, 74 },
+};
+
+static struct platform_device scif1_device = {
+	.name		= "sh-sci",
+	.id		= 1,
+	.dev		= {
+		.platform_data	= &scif1_platform_data,
+	},
+};
+
+static struct plat_sci_port scif2_platform_data = {
+	.mapbase	= 0xfe620000,
+	.flags		= UPF_BOOT_AUTOCONF,
+	.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
+	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.type		= PORT_SCIF,
+	.irqs		= { 76, 77, 79, 78 },
+};
+
+static struct platform_device scif2_device = {
+	.name		= "sh-sci",
+	.id		= 2,
+	.dev		= {
+		.platform_data	= &scif2_platform_data,
+	},
+};
+
+static struct plat_sci_port scif3_platform_data = {
+	.mapbase	= 0xfe480000,
+	.flags		= UPF_BOOT_AUTOCONF,
+	.scscr		= SCSCR_RE | SCSCR_TE | SCSCR_REIE,
+	.scbrr_algo_id	= SCBRR_ALGO_2,
+	.type		= PORT_SCI,
+	.irqs		= { 80, 81, 82, 0 },
 };
 
-static struct platform_device sci_device = {
+static struct platform_device scif3_device = {
 	.name		= "sh-sci",
-	.id		= -1,
+	.id		= 3,
 	.dev		= {
-		.platform_data	= sci_platform_data,
+		.platform_data	= &scif3_platform_data,
 	},
 };
 
 static struct sh_timer_config tmu0_platform_data = {
-	.name = "TMU0",
 	.channel_offset = 0x04,
 	.timer_bit = 0,
-	.clk = "peripheral_clk",
 	.clockevent_rating = 200,
 };
 
 static struct resource tmu0_resources[] = {
 	[0] = {
-		.name	= "TMU0",
 		.start	= 0xffd80008,
 		.end	= 0xffd80013,
 		.flags	= IORESOURCE_MEM,
@@ -200,16 +223,13 @@ static struct platform_device tmu0_device = {
 };
 
 static struct sh_timer_config tmu1_platform_data = {
-	.name = "TMU1",
 	.channel_offset = 0x10,
 	.timer_bit = 1,
-	.clk = "peripheral_clk",
 	.clocksource_rating = 200,
 };
 
 static struct resource tmu1_resources[] = {
 	[0] = {
-		.name	= "TMU1",
 		.start	= 0xffd80014,
 		.end	= 0xffd8001f,
 		.flags	= IORESOURCE_MEM,
@@ -231,15 +251,12 @@ static struct platform_device tmu1_device = {
 };
 
 static struct sh_timer_config tmu2_platform_data = {
-	.name = "TMU2",
 	.channel_offset = 0x1c,
 	.timer_bit = 2,
-	.clk = "peripheral_clk",
 };
 
 static struct resource tmu2_resources[] = {
 	[0] = {
-		.name	= "TMU2",
 		.start	= 0xffd80020,
 		.end	= 0xffd8002f,
 		.flags	= IORESOURCE_MEM,
@@ -262,7 +279,10 @@ static struct platform_device tmu2_device = {
 
 
 static struct platform_device *sh7760_devices[] __initdata = {
-	&sci_device,
+	&scif0_device,
+	&scif1_device,
+	&scif2_device,
+	&scif3_device,
 	&tmu0_device,
 	&tmu1_device,
 	&tmu2_device,
@@ -273,9 +293,13 @@ static int __init sh7760_devices_setup(void)
 	return platform_add_devices(sh7760_devices,
 				    ARRAY_SIZE(sh7760_devices));
 }
-__initcall(sh7760_devices_setup);
+arch_initcall(sh7760_devices_setup);
 
 static struct platform_device *sh7760_early_devices[] __initdata = {
+	&scif0_device,
+	&scif1_device,
+	&scif2_device,
+	&scif3_device,
 	&tmu0_device,
 	&tmu1_device,
 	&tmu2_device,
@@ -294,7 +318,7 @@ void __init plat_irq_setup_pins(int mode)
 {
 	switch (mode) {
 	case IRQ_MODE_IRQ:
-		ctrl_outw(ctrl_inw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
+		__raw_writew(__raw_readw(INTC_ICR) | INTC_ICR_IRLM, INTC_ICR);
 		register_intc_controller(&intc_desc_irq);
 		break;
 	default:
diff --git a/arch/sh/kernel/cpu/sh4/sq.c b/arch/sh/kernel/cpu/sh4/sq.c
index 8a8a993f55ea..14726eef1ce0 100644
--- a/arch/sh/kernel/cpu/sh4/sq.c
+++ b/arch/sh/kernel/cpu/sh4/sq.c
@@ -43,9 +43,9 @@ static unsigned long *sq_bitmap;
 
 #define store_queue_barrier()			\
 do {						\
-	(void)ctrl_inl(P4SEG_STORE_QUE);	\
-	ctrl_outl(0, P4SEG_STORE_QUE + 0);	\
-	ctrl_outl(0, P4SEG_STORE_QUE + 8);	\
+	(void)__raw_readl(P4SEG_STORE_QUE);	\
+	__raw_writel(0, P4SEG_STORE_QUE + 0);	\
+	__raw_writel(0, P4SEG_STORE_QUE + 8);	\
 } while (0);
 
 /**
@@ -100,7 +100,7 @@ static inline void sq_mapping_list_del(struct sq_mapping *map)
 	spin_unlock_irq(&sq_mapping_lock);
 }
 
-static int __sq_remap(struct sq_mapping *map, unsigned long flags)
+static int __sq_remap(struct sq_mapping *map, pgprot_t prot)
 {
 #if defined(CONFIG_MMU)
 	struct vm_struct *vma;
@@ -113,7 +113,7 @@ static int __sq_remap(struct sq_mapping *map, unsigned long flags)
 
 	if (ioremap_page_range((unsigned long)vma->addr,
 			       (unsigned long)vma->addr + map->size,
-			       vma->phys_addr, __pgprot(flags))) {
+			       vma->phys_addr, prot)) {
 		vunmap(vma->addr);
 		return -EAGAIN;
 	}
@@ -123,8 +123,8 @@ static int __sq_remap(struct sq_mapping *map, unsigned long flags)
 	 * straightforward, as we can just load up each queue's QACR with
 	 * the physical address appropriately masked.
 	 */
-	ctrl_outl(((map->addr >> 26) << 2) & 0x1c, SQ_QACR0);
-	ctrl_outl(((map->addr >> 26) << 2) & 0x1c, SQ_QACR1);
+	__raw_writel(((map->addr >> 26) << 2) & 0x1c, SQ_QACR0);
+	__raw_writel(((map->addr >> 26) << 2) & 0x1c, SQ_QACR1);
 #endif
 
 	return 0;
@@ -135,14 +135,14 @@ static int __sq_remap(struct sq_mapping *map, unsigned long flags)
  * @phys: Physical address of mapping.
  * @size: Length of mapping.
  * @name: User invoking mapping.
- * @flags: Protection flags.
+ * @prot: Protection bits.
  *
  * Remaps the physical address @phys through the next available store queue
  * address of @size length. @name is logged at boot time as well as through
  * the sysfs interface.
  */
 unsigned long sq_remap(unsigned long phys, unsigned int size,
-		       const char *name, unsigned long flags)
+		       const char *name, pgprot_t prot)
 {
 	struct sq_mapping *map;
 	unsigned long end;
@@ -177,7 +177,7 @@ unsigned long sq_remap(unsigned long phys, unsigned int size,
 
 	map->sq_addr = P4SEG_STORE_QUE + (page << PAGE_SHIFT);
 
-	ret = __sq_remap(map, pgprot_val(PAGE_KERNEL_NOCACHE) | flags);
+	ret = __sq_remap(map, prot);
 	if (unlikely(ret != 0))
 		goto out;
 
@@ -309,8 +309,7 @@ static ssize_t mapping_store(const char *buf, size_t count)
 		return -EIO;
 
 	if (likely(len)) {
-		int ret = sq_remap(base, len, "Userspace",
-				   pgprot_val(PAGE_SHARED));
+		int ret = sq_remap(base, len, "Userspace", PAGE_SHARED);
 		if (ret < 0)
 			return ret;
 	} else
@@ -327,7 +326,7 @@ static struct attribute *sq_sysfs_attrs[] = {
 	NULL,
 };
 
-static struct sysfs_ops sq_sysfs_ops = {
+static const struct sysfs_ops sq_sysfs_ops = {
 	.show	= sq_sysfs_show,
 	.store	= sq_sysfs_store,
 };