Merge branch 'master' into sh/cachetlb

Conflicts:
	arch/sh/kernel/Makefile_64

24 files changed, 390 insertions, 211 deletions

diff --git a/arch/sh/boards/mach-se/7724/setup.c b/arch/sh/boards/mach-se/7724/setup.c
index 36a4ca3..9162081 100644
--- a/arch/sh/boards/mach-se/7724/setup.c
+++ b/arch/sh/boards/mach-se/7724/setup.c
@@ -246,7 +246,7 @@ static struct platform_device ceu1_device = {
 	},
 };
 
-/* KEYSC */
+/* KEYSC in SoC (Needs SW33-2 set to ON) */
 static struct sh_keysc_info keysc_info = {
 	.mode = SH_KEYSC_MODE_1,
 	.scan_timing = 10,
@@ -263,12 +263,13 @@ static struct sh_keysc_info keysc_info = {
 
 static struct resource keysc_resources[] = {
 	[0] = {
-		.start  = 0x1a204000,
-		.end    = 0x1a20400f,
+		.name	= "KEYSC",
+		.start  = 0x044b0000,
+		.end    = 0x044b000f,
 		.flags  = IORESOURCE_MEM,
 	},
 	[1] = {
-		.start  = IRQ0_KEY,
+		.start  = 79,
 		.flags  = IORESOURCE_IRQ,
 	},
 };
diff --git a/arch/sh/drivers/heartbeat.c b/arch/sh/drivers/heartbeat.c
index 938817e..a9339a6 100644
--- a/arch/sh/drivers/heartbeat.c
+++ b/arch/sh/drivers/heartbeat.c
@@ -40,14 +40,19 @@ static inline void heartbeat_toggle_bit(struct heartbeat_data *hd,
 	if (inverted)
 		new = ~new;
 
+	new &= hd->mask;
+
 	switch (hd->regsize) {
 	case 32:
+		new |= ioread32(hd->base) & ~hd->mask;
 		iowrite32(new, hd->base);
 		break;
 	case 16:
+		new |= ioread16(hd->base) & ~hd->mask;
 		iowrite16(new, hd->base);
 		break;
 	default:
+		new |= ioread8(hd->base) & ~hd->mask;
 		iowrite8(new, hd->base);
 		break;
 	}
@@ -72,6 +77,7 @@ static int heartbeat_drv_probe(struct platform_device *pdev)
 {
 	struct resource *res;
 	struct heartbeat_data *hd;
+	int i;
 
 	if (unlikely(pdev->num_resources != 1)) {
 		dev_err(&pdev->dev, "invalid number of resources\n");
@@ -107,6 +113,10 @@ static int heartbeat_drv_probe(struct platform_device *pdev)
 		hd->nr_bits = ARRAY_SIZE(default_bit_pos);
 	}
 
+	hd->mask = 0;
+	for (i = 0; i < hd->nr_bits; i++)
+		hd->mask |= (1 << hd->bit_pos[i]);
+
 	if (!hd->regsize)
 		hd->regsize = 8;	/* default access size */
 
diff --git a/arch/sh/include/asm/dwarf.h b/arch/sh/include/asm/dwarf.h
index 60b1807..d3d3837 100644
--- a/arch/sh/include/asm/dwarf.h
+++ b/arch/sh/include/asm/dwarf.h
@@ -340,6 +340,10 @@ struct dwarf_stack {
 #define DW_CFA_lo_user		0x1c
 #define DW_CFA_hi_user		0x3f
 
+/* GNU extension opcodes  */
+#define DW_CFA_GNU_args_size	0x2e
+#define DW_CFA_GNU_negative_offset_extended 0x2f
+
 /*
  * Some call frame instructions encode their operands in the opcode. We
  * need some helper functions to extract both the opcode and operands
diff --git a/arch/sh/include/asm/heartbeat.h b/arch/sh/include/asm/heartbeat.h
index 724a43e..caaafe5 100644
--- a/arch/sh/include/asm/heartbeat.h
+++ b/arch/sh/include/asm/heartbeat.h
@@ -11,6 +11,7 @@ struct heartbeat_data {
 	unsigned int nr_bits;
 	struct timer_list timer;
 	unsigned int regsize;
+	unsigned int mask;
 	unsigned long flags;
 };
 
diff --git a/arch/sh/include/mach-kfr2r09/mach/partner-jet-setup.txt b/arch/sh/include/mach-kfr2r09/mach/partner-jet-setup.txt
index 9c85088..25801d4 100644
--- a/arch/sh/include/mach-kfr2r09/mach/partner-jet-setup.txt
+++ b/arch/sh/include/mach-kfr2r09/mach/partner-jet-setup.txt
@@ -24,6 +24,7 @@ LIST "setup clocks"
 ED 0xa4150004, 0x00000050
 ED 0xa4150000, 0x91053508
 WAIT 1
+ED 0xa4150050, 0x00000340
 ED 0xa4150024, 0x00005000
 
 LIST "setup pins"
diff --git a/arch/sh/kernel/Makefile b/arch/sh/kernel/Makefile
index 349d833..f37cf02 100644
--- a/arch/sh/kernel/Makefile
+++ b/arch/sh/kernel/Makefile
@@ -1,5 +1,40 @@
-ifeq ($(CONFIG_SUPERH32),y)
-include ${srctree}/arch/sh/kernel/Makefile_32
-else
-include ${srctree}/arch/sh/kernel/Makefile_64
+#
+# Makefile for the Linux/SuperH kernel.
+#
+
+extra-y	:= head_$(BITS).o init_task.o vmlinux.lds
+
+ifdef CONFIG_FUNCTION_TRACER
+# Do not profile debug and lowlevel utilities
+CFLAGS_REMOVE_ftrace.o = -pg
 endif
+
+obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
+	   machvec.o process_$(BITS).o ptrace_$(BITS).o setup.o 	\
+	   signal_$(BITS).o sys_sh.o sys_sh$(BITS).o syscalls_$(BITS).o	\
+	   time.o topology.o traps.o traps_$(BITS).o unwinder.o
+
+obj-y				+= cpu/
+obj-$(CONFIG_VSYSCALL)		+= vsyscall/
+obj-$(CONFIG_SMP)		+= smp.o
+obj-$(CONFIG_SH_STANDARD_BIOS)	+= sh_bios.o
+obj-$(CONFIG_KGDB)		+= kgdb.o
+obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
+obj-$(CONFIG_MODULES)		+= sh_ksyms_$(BITS).o module.o
+obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
+obj-$(CONFIG_KEXEC)		+= machine_kexec.o relocate_kernel.o
+obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
+obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
+obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
+obj-$(CONFIG_KPROBES)		+= kprobes.o
+obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
+obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
+obj-$(CONFIG_FTRACE_SYSCALLS)	+= ftrace.o
+obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
+obj-$(CONFIG_DUMP_CODE)		+= disassemble.o
+obj-$(CONFIG_HIBERNATION)	+= swsusp.o
+obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
+
+obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
+
+EXTRA_CFLAGS += -Werror
diff --git a/arch/sh/kernel/Makefile_32 b/arch/sh/kernel/Makefile_32
deleted file mode 100644
index f2245eb..0000000
--- a/arch/sh/kernel/Makefile_32
+++ /dev/null
@@ -1,40 +0,0 @@
-#
-# Makefile for the Linux/SuperH kernel.
-#
-
-extra-y	:= head_32.o init_task.o vmlinux.lds
-
-ifdef CONFIG_FUNCTION_TRACER
-# Do not profile debug and lowlevel utilities
-CFLAGS_REMOVE_ftrace.o = -pg
-endif
-
-obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
-	   machvec.o process_32.o ptrace_32.o setup.o signal_32.o	\
-	   sys_sh.o sys_sh32.o syscalls_32.o time.o topology.o		\
-	   traps.o traps_32.o unwinder.o
-
-obj-y				+= cpu/
-obj-$(CONFIG_VSYSCALL)		+= vsyscall/
-obj-$(CONFIG_SMP)		+= smp.o
-obj-$(CONFIG_SH_STANDARD_BIOS)	+= sh_bios.o
-obj-$(CONFIG_KGDB)		+= kgdb.o
-obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
-obj-$(CONFIG_MODULES)		+= sh_ksyms_32.o module.o
-obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
-obj-$(CONFIG_KEXEC)		+= machine_kexec.o relocate_kernel.o
-obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
-obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
-obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
-obj-$(CONFIG_KPROBES)		+= kprobes.o
-obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
-obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
-obj-$(CONFIG_FTRACE_SYSCALLS)	+= ftrace.o
-obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
-obj-$(CONFIG_DUMP_CODE)		+= disassemble.o
-obj-$(CONFIG_HIBERNATION)	+= swsusp.o
-obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
-
-obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
-
-EXTRA_CFLAGS += -Werror
diff --git a/arch/sh/kernel/Makefile_64 b/arch/sh/kernel/Makefile_64
deleted file mode 100644
index cdfec1e..0000000
--- a/arch/sh/kernel/Makefile_64
+++ /dev/null
@@ -1,21 +0,0 @@
-extra-y	:= head_64.o init_task.o vmlinux.lds
-
-obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
-	   machvec.o process_64.o ptrace_64.o setup.o signal_64.o	\
-	   sys_sh.o sys_sh64.o syscalls_64.o time.o topology.o 		\
-	   traps.o traps_64.o unwinder.o
-
-obj-y				+= cpu/
-obj-$(CONFIG_SMP)		+= smp.o
-obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
-obj-$(CONFIG_MODULES)		+= sh_ksyms_64.o module.o
-obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
-obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
-obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
-obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
-obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
-obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
-
-obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
-
-EXTRA_CFLAGS += -Werror
diff --git a/arch/sh/kernel/cpu/init.c b/arch/sh/kernel/cpu/init.c
index c832fa4..e932ebe 100644
--- a/arch/sh/kernel/cpu/init.c
+++ b/arch/sh/kernel/cpu/init.c
@@ -3,7 +3,7 @@
  *
  * CPU init code
  *
- * Copyright (C) 2002 - 2007  Paul Mundt
+ * Copyright (C) 2002 - 2009  Paul Mundt
  * Copyright (C) 2003  Richard Curnow
  *
  * This file is subject to the terms and conditions of the GNU General Public
@@ -62,6 +62,37 @@ static void __init speculative_execution_init(void)
 #define speculative_execution_init()	do { } while (0)
 #endif
 
+#ifdef CONFIG_CPU_SH4A
+#define EXPMASK			0xff2f0004
+#define EXPMASK_RTEDS		(1 << 0)
+#define EXPMASK_BRDSSLP		(1 << 1)
+#define EXPMASK_MMCAW		(1 << 4)
+
+static void __init expmask_init(void)
+{
+	unsigned long expmask = __raw_readl(EXPMASK);
+
+	/*
+	 * Future proofing.
+	 *
+	 * Disable support for slottable sleep instruction
+	 * and non-nop instructions in the rte delay slot.
+	 */
+	expmask &= ~(EXPMASK_RTEDS | EXPMASK_BRDSSLP);
+
+	/*
+	 * Enable associative writes to the memory-mapped cache array
+	 * until the cache flush ops have been rewritten.
+	 */
+	expmask |= EXPMASK_MMCAW;
+
+	__raw_writel(expmask, EXPMASK);
+	ctrl_barrier();
+}
+#else
+#define expmask_init()	do { } while (0)
+#endif
+
 /* 2nd-level cache init */
 void __uses_jump_to_uncached __attribute__ ((weak)) l2_cache_init(void)
 {
@@ -319,4 +350,5 @@ asmlinkage void __init sh_cpu_init(void)
 #endif
 
 	speculative_execution_init();
+	expmask_init();
 }
diff --git a/arch/sh/kernel/cpu/sh2/entry.S b/arch/sh/kernel/cpu/sh2/entry.S
index becc54c45..c8a4331 100644
--- a/arch/sh/kernel/cpu/sh2/entry.S
+++ b/arch/sh/kernel/cpu/sh2/entry.S
@@ -227,8 +227,9 @@ ENTRY(sh_bios_handler)
 	mov.l	@r15+, r14
 	add	#8,r15
 	lds.l	@r15+, pr
+	mov.l	@r15+,r15
 	rte
-	 mov.l	@r15+,r15
+	 nop
 	.align	2
 1:	.long	gdb_vbr_vector
 #endif /* CONFIG_SH_STANDARD_BIOS */
diff --git a/arch/sh/kernel/cpu/sh2a/entry.S b/arch/sh/kernel/cpu/sh2a/entry.S
index ab3903e..222742d 100644
--- a/arch/sh/kernel/cpu/sh2a/entry.S
+++ b/arch/sh/kernel/cpu/sh2a/entry.S
@@ -176,8 +176,9 @@ ENTRY(sh_bios_handler)
 	movml.l	@r15+,r14
 	add	#8,r15
 	lds.l	@r15+, pr
+	mov.l	@r15+,r15
 	rte
-	 mov.l	@r15+,r15
+	 nop
 	.align	2
 1:	.long	gdb_vbr_vector
 #endif /* CONFIG_SH_STANDARD_BIOS */
diff --git a/arch/sh/kernel/cpu/sh3/entry.S b/arch/sh/kernel/cpu/sh3/entry.S
index 854921c..f94f25e 100644
--- a/arch/sh/kernel/cpu/sh3/entry.S
+++ b/arch/sh/kernel/cpu/sh3/entry.S
@@ -516,6 +516,14 @@ ENTRY(handle_interrupt)
 	bsr	save_regs	! needs original pr value in k3
 	 mov	#-1, k2		! default vector kept in k2
 
+	stc	sr, r0	! get status register
+	shlr2	r0
+	and	#0x3c, r0
+	cmp/eq	#0x3c, r0
+	bf	9f
+	TRACE_IRQS_OFF
+9:
+
 	! Setup return address and jump to do_IRQ
 	mov.l	4f, r9		! fetch return address
 	lds	r9, pr		! put return address in pr
diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c
index 34611d9..627588d 100644
--- a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c
+++ b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c
@@ -127,7 +127,7 @@ struct clk *main_clks[] = {
 	&div3_clk,
 };
 
-static int divisors[] = { 2, 0, 4, 6, 8, 12, 16, 0, 24, 32, 36, 48, 0, 72 };
+static int divisors[] = { 2, 3, 4, 6, 8, 12, 16, 0, 24, 32, 36, 48, 0, 72 };
 
 static struct clk_div_mult_table div4_table = {
 	.divisors = divisors,
diff --git a/arch/sh/kernel/cpu/shmobile/cpuidle.c b/arch/sh/kernel/cpu/shmobile/cpuidle.c
index 4afdd97..1c504bd 100644
--- a/arch/sh/kernel/cpu/shmobile/cpuidle.c
+++ b/arch/sh/kernel/cpu/shmobile/cpuidle.c
@@ -21,6 +21,7 @@
 static unsigned long cpuidle_mode[] = {
 	SUSP_SH_SLEEP, /* regular sleep mode */
 	SUSP_SH_SLEEP | SUSP_SH_SF, /* sleep mode + self refresh */
+	SUSP_SH_STANDBY | SUSP_SH_SF, /* software standby mode + self refresh */
 };
 
 static int cpuidle_sleep_enter(struct cpuidle_device *dev,
@@ -96,6 +97,16 @@ void sh_mobile_setup_cpuidle(void)
 	state->flags |= CPUIDLE_FLAG_TIME_VALID;
 	state->enter = cpuidle_sleep_enter;
 
+	state = &dev->states[i++];
+	snprintf(state->name, CPUIDLE_NAME_LEN, "C2");
+	strncpy(state->desc, "SuperH Mobile Standby Mode [SF]", CPUIDLE_DESC_LEN);
+	state->exit_latency = 2300;
+	state->target_residency = 1 * 2;
+	state->power_usage = 1;
+	state->flags = 0;
+	state->flags |= CPUIDLE_FLAG_TIME_VALID;
+	state->enter = cpuidle_sleep_enter;
+
 	dev->state_count = i;
 
 	cpuidle_register_device(dev);
diff --git a/arch/sh/kernel/cpu/shmobile/pm.c b/arch/sh/kernel/cpu/shmobile/pm.c
index de078d2..ee3c2aa 100644
--- a/arch/sh/kernel/cpu/shmobile/pm.c
+++ b/arch/sh/kernel/cpu/shmobile/pm.c
@@ -41,23 +41,11 @@ extern const unsigned int sh_mobile_standby_size;
 
 void sh_mobile_call_standby(unsigned long mode)
 {
-	extern void *vbr_base;
 	void *onchip_mem = (void *)ILRAM_BASE;
-	void (*standby_onchip_mem)(unsigned long) = onchip_mem;
-
-	/* Note: Wake up from sleep may generate exceptions!
-	 * Setup VBR to point to on-chip ram if self-refresh is
-	 * going to be used.
-	 */
-	if (mode & SUSP_SH_SF)
-		asm volatile("ldc %0, vbr" : : "r" (onchip_mem) : "memory");
+	void (*standby_onchip_mem)(unsigned long, unsigned long) = onchip_mem;
 
 	/* Let assembly snippet in on-chip memory handle the rest */
-	standby_onchip_mem(mode);
-
-	/* Put VBR back in System RAM again */
-	if (mode & SUSP_SH_SF)
-		asm volatile("ldc %0, vbr" : : "r" (&vbr_base) : "memory");
+	standby_onchip_mem(mode, ILRAM_BASE);
 }
 
 static int sh_pm_enter(suspend_state_t state)
diff --git a/arch/sh/kernel/cpu/shmobile/sleep.S b/arch/sh/kernel/cpu/shmobile/sleep.S
index 5d888ef..a439e6c 100644
--- a/arch/sh/kernel/cpu/shmobile/sleep.S
+++ b/arch/sh/kernel/cpu/shmobile/sleep.S
@@ -16,18 +16,73 @@
 #include <asm/asm-offsets.h>
 #include <asm/suspend.h>
 
+/*
+ * Kernel mode register usage, see entry.S:
+ *	k0	scratch
+ *	k1	scratch
+ *	k4	scratch
+ */
+#define k0	r0
+#define k1	r1
+#define k4	r4
+
 /* manage self-refresh and enter standby mode.
  * this code will be copied to on-chip memory and executed from there.
  */
 
 	.balign 	4096,0,4096
 ENTRY(sh_mobile_standby)
+
+	/* save original vbr */
+	stc	vbr, r1
+	mova	saved_vbr, r0
+	mov.l	r1, @r0
+
+	/* point vbr to our on-chip memory page */
+	ldc	r5, vbr
+
+	/* save return address */
+	mova	saved_spc, r0
+	sts	pr, r5
+	mov.l	r5, @r0
+
+	/* save sr */
+	mova	saved_sr, r0
+	stc	sr, r5
+	mov.l	r5, @r0
+
+	/* save mode flags */
+	mova	saved_mode, r0
+	mov.l	r4, @r0
+
+	/* put mode flags in r0 */
 	mov	r4, r0
 
 	tst	#SUSP_SH_SF, r0
 	bt	skip_set_sf
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+	/* DBSC: put memory in self-refresh mode */
+	mov.l	dben_reg, r4
+	mov.l	dben_data0, r1
+	mov.l	r1, @r4
 
-	/* SDRAM: disable power down and put in self-refresh mode */
+	mov.l	dbrfpdn0_reg, r4
+	mov.l	dbrfpdn0_data0, r1
+	mov.l	r1, @r4
+
+	mov.l	dbcmdcnt_reg, r4
+	mov.l	dbcmdcnt_data0, r1
+	mov.l	r1, @r4
+
+	mov.l	dbcmdcnt_reg, r4
+	mov.l	dbcmdcnt_data1, r1
+	mov.l	r1, @r4
+
+	mov.l	dbrfpdn0_reg, r4
+	mov.l	dbrfpdn0_data1, r1
+	mov.l	r1, @r4
+#else
+	/* SBSC: disable power down and put in self-refresh mode */
 	mov.l	1f, r4
 	mov.l	2f, r1
 	mov.l	@r4, r2
@@ -35,16 +90,9 @@ ENTRY(sh_mobile_standby)
 	mov.l   3f, r3
 	and	r3, r2
 	mov.l	r2, @r4
+#endif
 
 skip_set_sf:
-	tst	#SUSP_SH_SLEEP, r0
-	bt	test_standby
-
-	/* set mode to "sleep mode" */
-	bra	do_sleep
-	 mov	#0x00, r1
-
-test_standby:
 	tst	#SUSP_SH_STANDBY, r0
 	bt	test_rstandby
 
@@ -62,62 +110,135 @@ test_rstandby:
 
 test_ustandby:
 	tst	#SUSP_SH_USTANDBY, r0
-	bt	done_sleep
+	bt	force_sleep
 
 	/* set mode to "u-standby mode" */
-	mov	#0x10, r1
+	bra	do_sleep
+	 mov	#0x10, r1
+
+force_sleep:
 
-	/* fall-through */
+	/* set mode to "sleep mode" */
+	mov	#0x00, r1
 
 do_sleep:
 	/* setup and enter selected standby mode */
 	mov.l	5f, r4
 	mov.l	r1, @r4
+again:
 	sleep
+	bra	again
+	 nop
+
+restore_jump_vbr:
+	/* setup spc with return address to c code */
+	mov.l	saved_spc, k0
+	ldc	k0, spc
+
+	/* restore vbr */
+	mov.l	saved_vbr, k0
+	ldc	k0, vbr
+
+	/* setup ssr with saved sr */
+	mov.l	saved_sr, k0
+	ldc	k0, ssr
+
+	/* get mode flags */
+	mov.l	saved_mode, k0
 
 done_sleep:
 	/* reset standby mode to sleep mode */
-	mov.l	5f, r4
-	mov	#0x00, r1
-	mov.l	r1, @r4
+	mov.l	5f, k4
+	mov	#0x00, k1
+	mov.l	k1, @k4
 
-	tst	#SUSP_SH_SF, r0
+	tst	#SUSP_SH_SF, k0
 	bt	skip_restore_sf
 
-	/* SDRAM: set auto-refresh mode */
-	mov.l	1f, r4
-	mov.l	@r4, r2
-	mov.l   4f, r3
-	and	r3, r2
-	mov.l	r2, @r4
-	mov.l	6f, r4
-	mov.l	7f, r1
-	mov.l	8f, r2
-	mov.l	@r4, r3
-	mov	#-1, r4
-	add	r4, r3
-	or	r2, r3
-	mov.l	r3, @r1
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+	/* DBSC: put memory in auto-refresh mode */
+	mov.l	dbrfpdn0_reg, k4
+	mov.l	dbrfpdn0_data0, k1
+	mov.l	k1, @k4
+
+	nop /* sleep 140 ns */
+	nop
+	nop
+	nop
+
+	mov.l	dbcmdcnt_reg, k4
+	mov.l	dbcmdcnt_data0, k1
+	mov.l	k1, @k4
+
+	mov.l	dbcmdcnt_reg, k4
+	mov.l	dbcmdcnt_data1, k1
+	mov.l	k1, @k4
+
+	mov.l	dben_reg, k4
+	mov.l	dben_data1, k1
+	mov.l	k1, @k4
+
+	mov.l	dbrfpdn0_reg, k4
+	mov.l	dbrfpdn0_data2, k1
+	mov.l	k1, @k4
+#else
+	/* SBSC: set auto-refresh mode */
+	mov.l	1f, k4
+	mov.l	@k4, k0
+	mov.l   4f, k1
+	and	k1, k0
+	mov.l	k0, @k4
+	mov.l	6f, k4
+	mov.l	8f, k0
+	mov.l	@k4, k1
+	mov	#-1, k4
+	add	k4, k1
+	or	k1, k0
+	mov.l	7f, k1
+	mov.l	k0, @k1
+#endif
 skip_restore_sf:
-	rts
+	/* jump to vbr vector */
+	mov.l	saved_vbr, k0
+	mov.l	offset_vbr, k4
+	add	k4, k0
+	jmp	@k0
 	 nop
 
 	.balign 4
+saved_mode:	.long	0
+saved_spc:	.long	0
+saved_sr:	.long	0
+saved_vbr:	.long	0
+offset_vbr:	.long	0x600
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+dben_reg:	.long	0xfd000010 /* DBEN */
+dben_data0:	.long	0
+dben_data1:	.long	1
+dbrfpdn0_reg:	.long	0xfd000040 /* DBRFPDN0 */
+dbrfpdn0_data0:	.long	0
+dbrfpdn0_data1:	.long	1
+dbrfpdn0_data2:	.long	0x00010000
+dbcmdcnt_reg:	.long	0xfd000014 /* DBCMDCNT */
+dbcmdcnt_data0:	.long	2
+dbcmdcnt_data1:	.long	4
+#else
 1:	.long	0xfe400008 /* SDCR0 */
 2:	.long	0x00000400
 3:	.long	0xffff7fff
 4:	.long	0xfffffbff
+#endif
 5:	.long	0xa4150020 /* STBCR */
 6:	.long   0xfe40001c /* RTCOR */
 7:	.long   0xfe400018 /* RTCNT */
 8:	.long   0xa55a0000
 
+
 /* interrupt vector @ 0x600 */
 	.balign 	0x400,0,0x400
 	.long	0xdeadbeef
 	.balign 	0x200,0,0x200
-	/* sh7722 will end up here in sleep mode */
-	rte
+	bra	restore_jump_vbr
 	 nop
 sh_mobile_standby_end:
 
diff --git a/arch/sh/kernel/dwarf.c b/arch/sh/kernel/dwarf.c
index db02136..5fd6e60 100644
--- a/arch/sh/kernel/dwarf.c
+++ b/arch/sh/kernel/dwarf.c
@@ -330,7 +330,6 @@ struct dwarf_fde *dwarf_lookup_fde(unsigned long pc)
  *	@fde: the FDE for this function
  *	@frame: the instructions calculate the CFA for this frame
  *	@pc: the program counter of the address we're interested in
- *	@define_ra: keep executing insns until the return addr reg is defined?
  *
  *	Execute the Call Frame instruction sequence starting at
  *	@insn_start and ending at @insn_end. The instructions describe
@@ -342,36 +341,17 @@ static int dwarf_cfa_execute_insns(unsigned char *insn_start,
 				   struct dwarf_cie *cie,
 				   struct dwarf_fde *fde,
 				   struct dwarf_frame *frame,
-				   unsigned long pc,
-				   bool define_ra)
+				   unsigned long pc)
 {
 	unsigned char insn;
 	unsigned char *current_insn;
 	unsigned int count, delta, reg, expr_len, offset;
-	bool seen_ra_reg;
 
 	current_insn = insn_start;
 
-	/*
-	 * If we're executing instructions for the dwarf_unwind_stack()
-	 * FDE we need to keep executing instructions until the value of
-	 * DWARF_ARCH_RA_REG is defined. See the comment in
-	 * dwarf_unwind_stack() for more details.
-	 */
-	if (define_ra)
-		seen_ra_reg = false;
-	else
-		seen_ra_reg = true;
-
-	while (current_insn < insn_end && (frame->pc <= pc || !seen_ra_reg) ) {
+	while (current_insn < insn_end && frame->pc <= pc) {
 		insn = __raw_readb(current_insn++);
 
-		if (!seen_ra_reg) {
-			if (frame->num_regs >= DWARF_ARCH_RA_REG &&
-			    frame->regs[DWARF_ARCH_RA_REG].flags)
-				seen_ra_reg = true;
-		}
-
 		/*
 		 * Firstly, handle the opcodes that embed their operands
 		 * in the instructions.
@@ -484,6 +464,19 @@ static int dwarf_cfa_execute_insns(unsigned char *insn_start,
 			frame->regs[reg].flags |= DWARF_REG_OFFSET;
 			frame->regs[reg].addr = offset;
 			break;
+		case DW_CFA_GNU_args_size:
+			count = dwarf_read_uleb128(current_insn, &offset);
+			current_insn += count;
+			break;
+		case DW_CFA_GNU_negative_offset_extended:
+			count = dwarf_read_uleb128(current_insn, &reg);
+			current_insn += count;
+			count = dwarf_read_uleb128(current_insn, &offset);
+			offset *= cie->data_alignment_factor;
+			dwarf_frame_alloc_regs(frame, reg);
+			frame->regs[reg].flags |= DWARF_REG_OFFSET;
+			frame->regs[reg].addr = -offset;
+			break;
 		default:
 			pr_debug("unhandled DWARF instruction 0x%x\n", insn);
 			break;
@@ -510,26 +503,17 @@ struct dwarf_frame *dwarf_unwind_stack(unsigned long pc,
 	struct dwarf_fde *fde;
 	unsigned long addr;
 	int i, offset;
-	bool define_ra = false;
 
 	/*
 	 * If this is the first invocation of this recursive function we
 	 * need get the contents of a physical register to get the CFA
 	 * in order to begin the virtual unwinding of the stack.
 	 *
-	 * Setting "define_ra" to true indictates that we want
-	 * dwarf_cfa_execute_insns() to continue executing instructions
-	 * until we know how to calculate the value of DWARF_ARCH_RA_REG
-	 * (which we need in order to kick off the whole unwinding
-	 * process).
-	 *
 	 * NOTE: the return address is guaranteed to be setup by the
 	 * time this function makes its first function call.
 	 */
-	if (!pc && !prev) {
-		pc = (unsigned long)&dwarf_unwind_stack;
-		define_ra = true;
-	}
+	if (!pc && !prev)
+		pc = (unsigned long)current_text_addr();
 
 	frame = kzalloc(sizeof(*frame), GFP_ATOMIC);
 	if (!frame)
@@ -565,11 +549,11 @@ struct dwarf_frame *dwarf_unwind_stack(unsigned long pc,
 	/* CIE initial instructions */
 	dwarf_cfa_execute_insns(cie->initial_instructions,
 				cie->instructions_end, cie, fde,
-				frame, pc, false);
+				frame, pc);
 
 	/* FDE instructions */
 	dwarf_cfa_execute_insns(fde->instructions, fde->end, cie,
-				fde, frame, pc, define_ra);
+				fde, frame, pc);
 
 	/* Calculate the CFA */
 	switch (frame->flags) {
diff --git a/arch/sh/kernel/entry-common.S b/arch/sh/kernel/entry-common.S
index e63178f..7004776 100644
--- a/arch/sh/kernel/entry-common.S
+++ b/arch/sh/kernel/entry-common.S
@@ -77,15 +77,6 @@ ENTRY(ret_from_irq)
 	!
 	mov	#OFF_SR, r0
 	mov.l	@(r0,r15), r0	! get status register
-
-	shlr2	r0
-	and	#0x3c, r0
-	cmp/eq	#0x3c, r0
-	bt	9f
-	TRACE_IRQS_ON
-9:
-	mov	#OFF_SR, r0
-	mov.l	@(r0,r15), r0	! get status register
 	shll	r0
 	shll	r0		! kernel space?
 	get_current_thread_info r8, r0
@@ -96,6 +87,7 @@ ENTRY(ret_from_irq)
 	 nop
 ENTRY(resume_kernel)
 	cli
+	TRACE_IRQS_OFF
 	mov.l	@(TI_PRE_COUNT,r8), r0	! current_thread_info->preempt_count
 	tst	r0, r0
 	bf	noresched
@@ -213,12 +205,25 @@ syscall_trace_entry:
 	 mov.l	r0, @(OFF_R0,r15)	! Return value
 
 __restore_all:
-	mov.l	1f, r0
+	mov	#OFF_SR, r0
+	mov.l	@(r0,r15), r0	! get status register
+
+	shlr2	r0
+	and	#0x3c, r0
+	cmp/eq	#0x3c, r0
+	bt	1f
+	TRACE_IRQS_ON
+	bra	2f
+	 nop
+1:
+	TRACE_IRQS_OFF
+2:
+	mov.l	3f, r0
 	jmp	@r0
 	 nop
 
 	.align	2
-1:	.long	restore_all
+3:	.long	restore_all
 
 	.align	2
 syscall_badsys:			! Bad syscall number
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 738bdc6..13ffa5d 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -24,6 +24,7 @@ config X86
 	select HAVE_UNSTABLE_SCHED_CLOCK
 	select HAVE_IDE
 	select HAVE_OPROFILE
+	select HAVE_PERF_COUNTERS if (!M386 && !M486)
 	select HAVE_IOREMAP_PROT
 	select HAVE_KPROBES
 	select ARCH_WANT_OPTIONAL_GPIOLIB
@@ -742,7 +743,6 @@ config X86_UP_IOAPIC
 config X86_LOCAL_APIC
 	def_bool y
 	depends on X86_64 || SMP || X86_32_NON_STANDARD || X86_UP_APIC
-	select HAVE_PERF_COUNTERS if (!M386 && !M486)
 
 config X86_IO_APIC
 	def_bool y
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index e2485b0..63fddcd 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -400,6 +400,13 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c)
 		level = cpuid_eax(1);
 		if((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58)
 			set_cpu_cap(c, X86_FEATURE_REP_GOOD);
+
+		/*
+		 * Some BIOSes incorrectly force this feature, but only K8
+		 * revision D (model = 0x14) and later actually support it.
+		 */
+		if (c->x86_model < 0x14)
+			clear_cpu_cap(c, X86_FEATURE_LAHF_LM);
 	}
 	if (c->x86 == 0x10 || c->x86 == 0x11)
 		set_cpu_cap(c, X86_FEATURE_REP_GOOD);
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index f1961c0..5ce60a8 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -59,7 +59,30 @@ void __init setup_cpu_local_masks(void)
 	alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask);
 }
 
-static const struct cpu_dev *this_cpu __cpuinitdata;
+static void __cpuinit default_init(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_X86_64
+	display_cacheinfo(c);
+#else
+	/* Not much we can do here... */
+	/* Check if at least it has cpuid */
+	if (c->cpuid_level == -1) {
+		/* No cpuid. It must be an ancient CPU */
+		if (c->x86 == 4)
+			strcpy(c->x86_model_id, "486");
+		else if (c->x86 == 3)
+			strcpy(c->x86_model_id, "386");
+	}
+#endif
+}
+
+static const struct cpu_dev __cpuinitconst default_cpu = {
+	.c_init		= default_init,
+	.c_vendor	= "Unknown",
+	.c_x86_vendor	= X86_VENDOR_UNKNOWN,
+};
+
+static const struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
 
 DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
 #ifdef CONFIG_X86_64
@@ -332,29 +355,6 @@ void switch_to_new_gdt(int cpu)
 
 static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {};
 
-static void __cpuinit default_init(struct cpuinfo_x86 *c)
-{
-#ifdef CONFIG_X86_64
-	display_cacheinfo(c);
-#else
-	/* Not much we can do here... */
-	/* Check if at least it has cpuid */
-	if (c->cpuid_level == -1) {
-		/* No cpuid. It must be an ancient CPU */
-		if (c->x86 == 4)
-			strcpy(c->x86_model_id, "486");
-		else if (c->x86 == 3)
-			strcpy(c->x86_model_id, "386");
-	}
-#endif
-}
-
-static const struct cpu_dev __cpuinitconst default_cpu = {
-	.c_init	= default_init,
-	.c_vendor = "Unknown",
-	.c_x86_vendor = X86_VENDOR_UNKNOWN,
-};
-
 static void __cpuinit get_model_name(struct cpuinfo_x86 *c)
 {
 	unsigned int *v;
diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
index bff8dd1..8bc64cf 100644
--- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
+++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
@@ -36,6 +36,7 @@
 
 static DEFINE_PER_CPU(__u64, next_check) = INITIAL_JIFFIES;
 static DEFINE_PER_CPU(unsigned long, thermal_throttle_count);
+static DEFINE_PER_CPU(bool, thermal_throttle_active);
 
 static atomic_t therm_throt_en		= ATOMIC_INIT(0);
 
@@ -96,24 +97,27 @@ static int therm_throt_process(int curr)
 {
 	unsigned int cpu = smp_processor_id();
 	__u64 tmp_jiffs = get_jiffies_64();
+	bool was_throttled = __get_cpu_var(thermal_throttle_active);
+	bool is_throttled = __get_cpu_var(thermal_throttle_active) = curr;
 
-	if (curr)
+	if (is_throttled)
 		__get_cpu_var(thermal_throttle_count)++;
 
-	if (time_before64(tmp_jiffs, __get_cpu_var(next_check)))
+	if (!(was_throttled ^ is_throttled) &&
+	    time_before64(tmp_jiffs, __get_cpu_var(next_check)))
 		return 0;
 
 	__get_cpu_var(next_check) = tmp_jiffs + CHECK_INTERVAL;
 
 	/* if we just entered the thermal event */
-	if (curr) {
+	if (is_throttled) {
 		printk(KERN_CRIT "CPU%d: Temperature above threshold, "
-		       "cpu clock throttled (total events = %lu)\n", cpu,
-		       __get_cpu_var(thermal_throttle_count));
+		       "cpu clock throttled (total events = %lu)\n",
+		       cpu, __get_cpu_var(thermal_throttle_count));
 
 		add_taint(TAINT_MACHINE_CHECK);
-	} else {
-		printk(KERN_CRIT "CPU%d: Temperature/speed normal\n", cpu);
+	} else if (was_throttled) {
+		printk(KERN_INFO "CPU%d: Temperature/speed normal\n", cpu);
 	}
 
 	return 1;
diff --git a/arch/x86/kernel/cpu/perf_counter.c b/arch/x86/kernel/cpu/perf_counter.c
index a7aa8f9..900332b 100644
--- a/arch/x86/kernel/cpu/perf_counter.c
+++ b/arch/x86/kernel/cpu/perf_counter.c
@@ -55,6 +55,7 @@ struct x86_pmu {
 	int		num_counters_fixed;
 	int		counter_bits;
 	u64		counter_mask;
+	int		apic;
 	u64		max_period;
 	u64		intel_ctrl;
 };
@@ -72,8 +73,8 @@ static const u64 p6_perfmon_event_map[] =
 {
   [PERF_COUNT_HW_CPU_CYCLES]		= 0x0079,
   [PERF_COUNT_HW_INSTRUCTIONS]		= 0x00c0,
-  [PERF_COUNT_HW_CACHE_REFERENCES]	= 0x0000,
-  [PERF_COUNT_HW_CACHE_MISSES]		= 0x0000,
+  [PERF_COUNT_HW_CACHE_REFERENCES]	= 0x0f2e,
+  [PERF_COUNT_HW_CACHE_MISSES]		= 0x012e,
   [PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= 0x00c4,
   [PERF_COUNT_HW_BRANCH_MISSES]		= 0x00c5,
   [PERF_COUNT_HW_BUS_CYCLES]		= 0x0062,
@@ -613,6 +614,7 @@ static DEFINE_MUTEX(pmc_reserve_mutex);
 
 static bool reserve_pmc_hardware(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	int i;
 
 	if (nmi_watchdog == NMI_LOCAL_APIC)
@@ -627,9 +629,11 @@ static bool reserve_pmc_hardware(void)
 		if (!reserve_evntsel_nmi(x86_pmu.eventsel + i))
 			goto eventsel_fail;
 	}
+#endif
 
 	return true;
 
+#ifdef CONFIG_X86_LOCAL_APIC
 eventsel_fail:
 	for (i--; i >= 0; i--)
 		release_evntsel_nmi(x86_pmu.eventsel + i);
@@ -644,10 +648,12 @@ perfctr_fail:
 		enable_lapic_nmi_watchdog();
 
 	return false;
+#endif
 }
 
 static void release_pmc_hardware(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	int i;
 
 	for (i = 0; i < x86_pmu.num_counters; i++) {
@@ -657,6 +663,7 @@ static void release_pmc_hardware(void)
 
 	if (nmi_watchdog == NMI_LOCAL_APIC)
 		enable_lapic_nmi_watchdog();
+#endif
 }
 
 static void hw_perf_counter_destroy(struct perf_counter *counter)
@@ -748,6 +755,15 @@ static int __hw_perf_counter_init(struct perf_counter *counter)
 		hwc->sample_period = x86_pmu.max_period;
 		hwc->last_period = hwc->sample_period;
 		atomic64_set(&hwc->period_left, hwc->sample_period);
+	} else {
+		/*
+		 * If we have a PMU initialized but no APIC
+		 * interrupts, we cannot sample hardware
+		 * counters (user-space has to fall back and
+		 * sample via a hrtimer based software counter):
+		 */
+		if (!x86_pmu.apic)
+			return -EOPNOTSUPP;
 	}
 
 	counter->destroy = hw_perf_counter_destroy;
@@ -1449,18 +1465,22 @@ void smp_perf_pending_interrupt(struct pt_regs *regs)
 
 void set_perf_counter_pending(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	apic->send_IPI_self(LOCAL_PENDING_VECTOR);
+#endif
 }
 
 void perf_counters_lapic_init(void)
 {
-	if (!x86_pmu_initialized())
+#ifdef CONFIG_X86_LOCAL_APIC
+	if (!x86_pmu.apic || !x86_pmu_initialized())
 		return;
 
 	/*
 	 * Always use NMI for PMU
 	 */
 	apic_write(APIC_LVTPC, APIC_DM_NMI);
+#endif
 }
 
 static int __kprobes
@@ -1484,7 +1504,9 @@ perf_counter_nmi_handler(struct notifier_block *self,
 
 	regs = args->regs;
 
+#ifdef CONFIG_X86_LOCAL_APIC
 	apic_write(APIC_LVTPC, APIC_DM_NMI);
+#endif
 	/*
 	 * Can't rely on the handled return value to say it was our NMI, two
 	 * counters could trigger 'simultaneously' raising two back-to-back NMIs.
@@ -1515,6 +1537,7 @@ static struct x86_pmu p6_pmu = {
 	.event_map		= p6_pmu_event_map,
 	.raw_event		= p6_pmu_raw_event,
 	.max_events		= ARRAY_SIZE(p6_perfmon_event_map),
+	.apic			= 1,
 	.max_period		= (1ULL << 31) - 1,
 	.version		= 0,
 	.num_counters		= 2,
@@ -1541,6 +1564,7 @@ static struct x86_pmu intel_pmu = {
 	.event_map		= intel_pmu_event_map,
 	.raw_event		= intel_pmu_raw_event,
 	.max_events		= ARRAY_SIZE(intel_perfmon_event_map),
+	.apic			= 1,
 	/*
 	 * Intel PMCs cannot be accessed sanely above 32 bit width,
 	 * so we install an artificial 1<<31 period regardless of
@@ -1564,6 +1588,7 @@ static struct x86_pmu amd_pmu = {
 	.num_counters		= 4,
 	.counter_bits		= 48,
 	.counter_mask		= (1ULL << 48) - 1,
+	.apic			= 1,
 	/* use highest bit to detect overflow */
 	.max_period		= (1ULL << 47) - 1,
 };
@@ -1589,13 +1614,14 @@ static int p6_pmu_init(void)
 		return -ENODEV;
 	}
 
+	x86_pmu = p6_pmu;
+
 	if (!cpu_has_apic) {
-		pr_info("no Local APIC, try rebooting with lapic");
-		return -ENODEV;
+		pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n");
+		pr_info("no hardware sampling interrupt available.\n");
+		x86_pmu.apic = 0;
 	}
 
-	x86_pmu				= p6_pmu;
-
 	return 0;
 }
 
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index 9eb8976..a06e8d1 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -418,20 +418,20 @@ static int __init set_pci_reboot(const struct dmi_system_id *d)
 }
 
 static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
-	{	/* Handle problems with rebooting on Apple MacBook5,2 */
+	{	/* Handle problems with rebooting on Apple MacBook5 */
 		.callback = set_pci_reboot,
-		.ident = "Apple MacBook",
+		.ident = "Apple MacBook5",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5,2"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5"),
 		},
 	},
-	{	/* Handle problems with rebooting on Apple MacBookPro5,1 */
+	{	/* Handle problems with rebooting on Apple MacBookPro5 */
 		.callback = set_pci_reboot,
-		.ident = "Apple MacBookPro5,1",
+		.ident = "Apple MacBookPro5",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5,1"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5"),
 		},
 	},
 	{ }