Merge tag 'csky-for-linus-4.21' of git://github.com/c-sky/csky-linux

Pull arch/csky updates from Guo Ren:
 "Here are three main features (cpu_hotplug, basic ftrace, basic perf)
  and some bugfixes:

  Features:
   - Add CPU-hotplug support for SMP
   - Add ftrace with function trace and function graph trace
   - Add Perf support
   - Add EM_CSKY_OLD 39
   - optimize kernel panic print.
   - remove syscall_exit_work

  Bugfixes:
   - fix abiv2 mmap(... O_SYNC) failure
   - fix gdb coredump error
   - remove vdsp implement for kernel
   - fix qemu failure to bootup sometimes
   - fix ftrace call-graph panic
   - fix device tree node reference leak
   - remove meaningless header-y
   - fix save hi,lo,dspcr regs in switch_stack
   - remove unused members in processor.h"

* tag 'csky-for-linus-4.21' of git://github.com/c-sky/csky-linux:
  csky: Add perf support for C-SKY
  csky: Add EM_CSKY_OLD 39
  clocksource/drivers/c-sky: fixup ftrace call-graph panic
  csky: ftrace call graph supported.
  csky: basic ftrace supported
  csky: remove unused members in processor.h
  csky: optimize kernel panic print.
  csky: stacktrace supported.
  csky: CPU-hotplug supported for SMP
  clocksource/drivers/c-sky: fixup qemu fail to bootup sometimes.
  csky: fixup save hi,lo,dspcr regs in switch_stack.
  csky: remove syscall_exit_work
  csky: fixup remove vdsp implement for kernel.
  csky: bugfix gdb coredump error.
  csky: fixup abiv2 mmap(... O_SYNC) failed.
  csky: define syscall_get_arch()
  elf-em.h: add EM_CSKY
  csky: remove meaningless header-y
  csky: Don't leak device tree node reference

12 files changed, 1265 insertions, 144 deletions

diff --git a/arch/csky/kernel/Makefile b/arch/csky/kernel/Makefile
index 4422de756cde..484e6d3a3647 100644
--- a/arch/csky/kernel/Makefile
+++ b/arch/csky/kernel/Makefile
@@ -6,3 +6,10 @@ obj-y += process.o cpu-probe.o ptrace.o dumpstack.o
 
 obj-$(CONFIG_MODULES)			+= module.o
 obj-$(CONFIG_SMP)			+= smp.o
+obj-$(CONFIG_FUNCTION_TRACER)		+= ftrace.o
+obj-$(CONFIG_STACKTRACE)		+= stacktrace.o
+obj-$(CONFIG_CSKY_PMU_V1)		+= perf_event.o
+
+ifdef CONFIG_FUNCTION_TRACER
+CFLAGS_REMOVE_ftrace.o = $(CC_FLAGS_FTRACE)
+endif
diff --git a/arch/csky/kernel/asm-offsets.c b/arch/csky/kernel/asm-offsets.c
index 8d3ed811321f..9b48b1b1a61b 100644
--- a/arch/csky/kernel/asm-offsets.c
+++ b/arch/csky/kernel/asm-offsets.c
@@ -20,12 +20,9 @@ int main(void)
 	/* offsets into the thread struct */
 	DEFINE(THREAD_KSP,        offsetof(struct thread_struct, ksp));
 	DEFINE(THREAD_SR,         offsetof(struct thread_struct, sr));
-	DEFINE(THREAD_ESP0,       offsetof(struct thread_struct, esp0));
 	DEFINE(THREAD_FESR,       offsetof(struct thread_struct, user_fp.fesr));
 	DEFINE(THREAD_FCR,        offsetof(struct thread_struct, user_fp.fcr));
 	DEFINE(THREAD_FPREG,      offsetof(struct thread_struct, user_fp.vr));
-	DEFINE(THREAD_DSPHI,      offsetof(struct thread_struct, hi));
-	DEFINE(THREAD_DSPLO,      offsetof(struct thread_struct, lo));
 
 	/* offsets into the thread_info struct */
 	DEFINE(TINFO_FLAGS,       offsetof(struct thread_info, flags));
diff --git a/arch/csky/kernel/dumpstack.c b/arch/csky/kernel/dumpstack.c
index a9a03ac57ec5..659253e9989c 100644
--- a/arch/csky/kernel/dumpstack.c
+++ b/arch/csky/kernel/dumpstack.c
@@ -7,60 +7,39 @@ int kstack_depth_to_print = 48;
 
 void show_trace(unsigned long *stack)
 {
-	unsigned long *endstack;
+	unsigned long *stack_end;
+	unsigned long *stack_start;
+	unsigned long *fp;
 	unsigned long addr;
-	int i;
 
-	pr_info("Call Trace:\n");
-	addr = (unsigned long)stack + THREAD_SIZE - 1;
-	endstack = (unsigned long *)(addr & -THREAD_SIZE);
-	i = 0;
-	while (stack + 1 <= endstack) {
-		addr = *stack++;
-		/*
-		 * If the address is either in the text segment of the
-		 * kernel, or in the region which contains vmalloc'ed
-		 * memory, it *may* be the address of a calling
-		 * routine; if so, print it so that someone tracing
-		 * down the cause of the crash will be able to figure
-		 * out the call path that was taken.
-		 */
-		if (__kernel_text_address(addr)) {
-#ifndef CONFIG_KALLSYMS
-			if (i % 5 == 0)
-				pr_cont("\n       ");
+	addr = (unsigned long) stack & THREAD_MASK;
+	stack_start = (unsigned long *) addr;
+	stack_end = (unsigned long *) (addr + THREAD_SIZE);
+
+	fp = stack;
+	pr_info("\nCall Trace:");
+
+	while (fp > stack_start && fp < stack_end) {
+#ifdef CONFIG_STACKTRACE
+		addr	= fp[1];
+		fp	= (unsigned long *) fp[0];
+#else
+		addr	= *fp++;
 #endif
-			pr_cont(" [<%08lx>] %pS\n", addr, (void *)addr);
-			i++;
-		}
+		if (__kernel_text_address(addr))
+			pr_cont("\n[<%08lx>] %pS", addr, (void *)addr);
 	}
 	pr_cont("\n");
 }
 
 void show_stack(struct task_struct *task, unsigned long *stack)
 {
-	unsigned long *p;
-	unsigned long *endstack;
-	int i;
-
 	if (!stack) {
 		if (task)
-			stack = (unsigned long *)task->thread.esp0;
+			stack = (unsigned long *)thread_saved_fp(task);
 		else
 			stack = (unsigned long *)&stack;
 	}
-	endstack = (unsigned long *)
-		(((unsigned long)stack + THREAD_SIZE - 1) & -THREAD_SIZE);
 
-	pr_info("Stack from %08lx:", (unsigned long)stack);
-	p = stack;
-	for (i = 0; i < kstack_depth_to_print; i++) {
-		if (p + 1 > endstack)
-			break;
-		if (i % 8 == 0)
-			pr_cont("\n       ");
-		pr_cont(" %08lx", *p++);
-	}
-	pr_cont("\n");
 	show_trace(stack);
 }
diff --git a/arch/csky/kernel/entry.S b/arch/csky/kernel/entry.S
index 79f92b8606c8..5137ed9062bd 100644
--- a/arch/csky/kernel/entry.S
+++ b/arch/csky/kernel/entry.S
@@ -122,16 +122,6 @@ ENTRY(csky_systemcall)
 
 	psrset  ee, ie
 
-	/* Stack frame for syscall, origin call set_esp0 */
-	mov     r12, sp
-
-	bmaski  r11, 13
-	andn    r12, r11
-	bgeni   r11, 9
-	addi    r11, 32
-	addu    r12, r11
-	st      sp, (r12, 0)
-
 	lrw     r11, __NR_syscalls
 	cmphs   syscallid, r11		/* Check nr of syscall */
 	bt      ret_from_exception
@@ -183,18 +173,10 @@ ENTRY(csky_systemcall)
 #endif
 	stw	a0, (sp, LSAVE_A0)	/* Save return value */
 
-	movi	a0, 1			/* leave system call */
-	mov	a1, sp			/* sp = pt_regs pointer */
-	jbsr	syscall_trace
-
-syscall_exit_work:
-	ld	syscallid, (sp, LSAVE_PSR)
-	btsti	syscallid, 31
-	bt	2f
-
-	jmpi	resume_userspace
-
-2:      RESTORE_ALL
+	movi    a0, 1                   /* leave system call */
+	mov     a1, sp                  /* right now, sp --> pt_regs */
+	jbsr    syscall_trace
+	br	ret_from_exception
 
 ENTRY(ret_from_kernel_thread)
 	jbsr	schedule_tail
@@ -238,8 +220,6 @@ resume_userspace:
 1:  RESTORE_ALL
 
 exit_work:
-	mov	a0, sp			/* Stack address is arg[0] */
-	jbsr	set_esp0		/* Call C level */
 	btsti	r8, TIF_NEED_RESCHED
 	bt	work_resched
 	/* If thread_info->flag is empty, RESTORE_ALL */
@@ -354,34 +334,12 @@ ENTRY(__switch_to)
 
 	stw	sp, (a3, THREAD_KSP)
 
-#ifdef CONFIG_CPU_HAS_HILO
-	lrw	r10, THREAD_DSPHI
-	add	r10, a3
-	mfhi	r6
-	mflo	r7
-	stw	r6, (r10, 0)		/* THREAD_DSPHI */
-	stw	r7, (r10, 4)		/* THREAD_DSPLO */
-	mfcr	r6, cr14
-	stw	r6, (r10, 8)		/* THREAD_DSPCSR */
-#endif
-
 	/* Set up next process to run */
 	lrw	a3, TASK_THREAD
 	addu	a3, a1
 
 	ldw	sp, (a3, THREAD_KSP)	/* Set next kernel sp */
 
-#ifdef CONFIG_CPU_HAS_HILO
-	lrw	r10, THREAD_DSPHI
-	add	r10, a3
-	ldw	r6, (r10, 8)		/* THREAD_DSPCSR */
-	mtcr	r6, cr14
-	ldw	r6, (r10, 0)		/* THREAD_DSPHI */
-	ldw	r7, (r10, 4)		/* THREAD_DSPLO */
-	mthi	r6
-	mtlo	r7
-#endif
-
 	ldw	a2, (a3, THREAD_SR)	/* Set next PSR */
 	mtcr	a2, psr
 
diff --git a/arch/csky/kernel/ftrace.c b/arch/csky/kernel/ftrace.c
new file mode 100644
index 000000000000..274c431f1810
--- /dev/null
+++ b/arch/csky/kernel/ftrace.c
@@ -0,0 +1,50 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2018 Hangzhou C-SKY Microsystems co.,ltd.
+
+#include <linux/ftrace.h>
+#include <linux/uaccess.h>
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr,
+			   unsigned long frame_pointer)
+{
+	unsigned long return_hooker = (unsigned long)&return_to_handler;
+	unsigned long old;
+
+	if (unlikely(atomic_read(&current->tracing_graph_pause)))
+		return;
+
+	old = *parent;
+
+	if (!function_graph_enter(old, self_addr,
+			*(unsigned long *)frame_pointer, parent)) {
+		/*
+		 * For csky-gcc function has sub-call:
+		 * subi	sp,	sp, 8
+		 * stw	r8,	(sp, 0)
+		 * mov	r8,	sp
+		 * st.w r15,	(sp, 0x4)
+		 * push	r15
+		 * jl	_mcount
+		 * We only need set *parent for resume
+		 *
+		 * For csky-gcc function has no sub-call:
+		 * subi	sp,	sp, 4
+		 * stw	r8,	(sp, 0)
+		 * mov	r8,	sp
+		 * push	r15
+		 * jl	_mcount
+		 * We need set *parent and *(frame_pointer + 4) for resume,
+		 * because lr is resumed twice.
+		 */
+		*parent = return_hooker;
+		frame_pointer += 4;
+		if (*(unsigned long *)frame_pointer == old)
+			*(unsigned long *)frame_pointer = return_hooker;
+	}
+}
+#endif
+
+/* _mcount is defined in abi's mcount.S */
+extern void _mcount(void);
+EXPORT_SYMBOL(_mcount);
diff --git a/arch/csky/kernel/perf_event.c b/arch/csky/kernel/perf_event.c
new file mode 100644
index 000000000000..376c972f5f37
--- /dev/null
+++ b/arch/csky/kernel/perf_event.c
@@ -0,0 +1,1031 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2018 Hangzhou C-SKY Microsystems co.,ltd.
+
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/perf_event.h>
+#include <linux/platform_device.h>
+
+#define CSKY_PMU_MAX_EVENTS 32
+
+#define HPCR		"<0, 0x0>"	/* PMU Control reg */
+#define HPCNTENR	"<0, 0x4>"	/* Count Enable reg */
+
+static uint64_t (*hw_raw_read_mapping[CSKY_PMU_MAX_EVENTS])(void);
+static void (*hw_raw_write_mapping[CSKY_PMU_MAX_EVENTS])(uint64_t val);
+
+struct csky_pmu_t {
+	struct pmu	pmu;
+	uint32_t	hpcr;
+} csky_pmu;
+
+#define cprgr(reg)				\
+({						\
+	unsigned int tmp;			\
+	asm volatile("cprgr %0, "reg"\n"	\
+		     : "=r"(tmp)		\
+		     :				\
+		     : "memory");		\
+	tmp;					\
+})
+
+#define cpwgr(reg, val)		\
+({				\
+	asm volatile(		\
+	"cpwgr %0, "reg"\n"	\
+	:			\
+	: "r"(val)		\
+	: "memory");		\
+})
+
+#define cprcr(reg)				\
+({						\
+	unsigned int tmp;			\
+	asm volatile("cprcr %0, "reg"\n"	\
+		     : "=r"(tmp)		\
+		     :				\
+		     : "memory");		\
+	tmp;					\
+})
+
+#define cpwcr(reg, val)		\
+({				\
+	asm volatile(		\
+	"cpwcr %0, "reg"\n"	\
+	:			\
+	: "r"(val)		\
+	: "memory");		\
+})
+
+/* cycle counter */
+static uint64_t csky_pmu_read_cc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x3>");
+		lo  = cprgr("<0, 0x2>");
+		hi  = cprgr("<0, 0x3>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_cc(uint64_t val)
+{
+	cpwgr("<0, 0x2>", (uint32_t)  val);
+	cpwgr("<0, 0x3>", (uint32_t) (val >> 32));
+}
+
+/* instruction counter */
+static uint64_t csky_pmu_read_ic(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x5>");
+		lo  = cprgr("<0, 0x4>");
+		hi  = cprgr("<0, 0x5>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_ic(uint64_t val)
+{
+	cpwgr("<0, 0x4>", (uint32_t)  val);
+	cpwgr("<0, 0x5>", (uint32_t) (val >> 32));
+}
+
+/* l1 icache access counter */
+static uint64_t csky_pmu_read_icac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x7>");
+		lo  = cprgr("<0, 0x6>");
+		hi  = cprgr("<0, 0x7>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_icac(uint64_t val)
+{
+	cpwgr("<0, 0x6>", (uint32_t)  val);
+	cpwgr("<0, 0x7>", (uint32_t) (val >> 32));
+}
+
+/* l1 icache miss counter */
+static uint64_t csky_pmu_read_icmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x9>");
+		lo  = cprgr("<0, 0x8>");
+		hi  = cprgr("<0, 0x9>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_icmc(uint64_t val)
+{
+	cpwgr("<0, 0x8>", (uint32_t)  val);
+	cpwgr("<0, 0x9>", (uint32_t) (val >> 32));
+}
+
+/* l1 dcache access counter */
+static uint64_t csky_pmu_read_dcac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0xb>");
+		lo  = cprgr("<0, 0xa>");
+		hi  = cprgr("<0, 0xb>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcac(uint64_t val)
+{
+	cpwgr("<0, 0xa>", (uint32_t)  val);
+	cpwgr("<0, 0xb>", (uint32_t) (val >> 32));
+}
+
+/* l1 dcache miss counter */
+static uint64_t csky_pmu_read_dcmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0xd>");
+		lo  = cprgr("<0, 0xc>");
+		hi  = cprgr("<0, 0xd>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcmc(uint64_t val)
+{
+	cpwgr("<0, 0xc>", (uint32_t)  val);
+	cpwgr("<0, 0xd>", (uint32_t) (val >> 32));
+}
+
+/* l2 cache access counter */
+static uint64_t csky_pmu_read_l2ac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0xf>");
+		lo  = cprgr("<0, 0xe>");
+		hi  = cprgr("<0, 0xf>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2ac(uint64_t val)
+{
+	cpwgr("<0, 0xe>", (uint32_t)  val);
+	cpwgr("<0, 0xf>", (uint32_t) (val >> 32));
+}
+
+/* l2 cache miss counter */
+static uint64_t csky_pmu_read_l2mc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x11>");
+		lo  = cprgr("<0, 0x10>");
+		hi  = cprgr("<0, 0x11>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2mc(uint64_t val)
+{
+	cpwgr("<0, 0x10>", (uint32_t)  val);
+	cpwgr("<0, 0x11>", (uint32_t) (val >> 32));
+}
+
+/* I-UTLB miss counter */
+static uint64_t csky_pmu_read_iutlbmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x15>");
+		lo  = cprgr("<0, 0x14>");
+		hi  = cprgr("<0, 0x15>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_iutlbmc(uint64_t val)
+{
+	cpwgr("<0, 0x14>", (uint32_t)  val);
+	cpwgr("<0, 0x15>", (uint32_t) (val >> 32));
+}
+
+/* D-UTLB miss counter */
+static uint64_t csky_pmu_read_dutlbmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x17>");
+		lo  = cprgr("<0, 0x16>");
+		hi  = cprgr("<0, 0x17>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dutlbmc(uint64_t val)
+{
+	cpwgr("<0, 0x16>", (uint32_t)  val);
+	cpwgr("<0, 0x17>", (uint32_t) (val >> 32));
+}
+
+/* JTLB miss counter */
+static uint64_t csky_pmu_read_jtlbmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x19>");
+		lo  = cprgr("<0, 0x18>");
+		hi  = cprgr("<0, 0x19>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_jtlbmc(uint64_t val)
+{
+	cpwgr("<0, 0x18>", (uint32_t)  val);
+	cpwgr("<0, 0x19>", (uint32_t) (val >> 32));
+}
+
+/* software counter */
+static uint64_t csky_pmu_read_softc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x1b>");
+		lo  = cprgr("<0, 0x1a>");
+		hi  = cprgr("<0, 0x1b>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_softc(uint64_t val)
+{
+	cpwgr("<0, 0x1a>", (uint32_t)  val);
+	cpwgr("<0, 0x1b>", (uint32_t) (val >> 32));
+}
+
+/* conditional branch mispredict counter */
+static uint64_t csky_pmu_read_cbmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x1d>");
+		lo  = cprgr("<0, 0x1c>");
+		hi  = cprgr("<0, 0x1d>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_cbmc(uint64_t val)
+{
+	cpwgr("<0, 0x1c>", (uint32_t)  val);
+	cpwgr("<0, 0x1d>", (uint32_t) (val >> 32));
+}
+
+/* conditional branch instruction counter */
+static uint64_t csky_pmu_read_cbic(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x1f>");
+		lo  = cprgr("<0, 0x1e>");
+		hi  = cprgr("<0, 0x1f>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_cbic(uint64_t val)
+{
+	cpwgr("<0, 0x1e>", (uint32_t)  val);
+	cpwgr("<0, 0x1f>", (uint32_t) (val >> 32));
+}
+
+/* indirect branch mispredict counter */
+static uint64_t csky_pmu_read_ibmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x21>");
+		lo  = cprgr("<0, 0x20>");
+		hi  = cprgr("<0, 0x21>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_ibmc(uint64_t val)
+{
+	cpwgr("<0, 0x20>", (uint32_t)  val);
+	cpwgr("<0, 0x21>", (uint32_t) (val >> 32));
+}
+
+/* indirect branch instruction counter */
+static uint64_t csky_pmu_read_ibic(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x23>");
+		lo  = cprgr("<0, 0x22>");
+		hi  = cprgr("<0, 0x23>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_ibic(uint64_t val)
+{
+	cpwgr("<0, 0x22>", (uint32_t)  val);
+	cpwgr("<0, 0x23>", (uint32_t) (val >> 32));
+}
+
+/* LSU spec fail counter */
+static uint64_t csky_pmu_read_lsfc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x25>");
+		lo  = cprgr("<0, 0x24>");
+		hi  = cprgr("<0, 0x25>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_lsfc(uint64_t val)
+{
+	cpwgr("<0, 0x24>", (uint32_t)  val);
+	cpwgr("<0, 0x25>", (uint32_t) (val >> 32));
+}
+
+/* store instruction counter */
+static uint64_t csky_pmu_read_sic(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x27>");
+		lo  = cprgr("<0, 0x26>");
+		hi  = cprgr("<0, 0x27>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_sic(uint64_t val)
+{
+	cpwgr("<0, 0x26>", (uint32_t)  val);
+	cpwgr("<0, 0x27>", (uint32_t) (val >> 32));
+}
+
+/* dcache read access counter */
+static uint64_t csky_pmu_read_dcrac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x29>");
+		lo  = cprgr("<0, 0x28>");
+		hi  = cprgr("<0, 0x29>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcrac(uint64_t val)
+{
+	cpwgr("<0, 0x28>", (uint32_t)  val);
+	cpwgr("<0, 0x29>", (uint32_t) (val >> 32));
+}
+
+/* dcache read miss counter */
+static uint64_t csky_pmu_read_dcrmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x2b>");
+		lo  = cprgr("<0, 0x2a>");
+		hi  = cprgr("<0, 0x2b>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcrmc(uint64_t val)
+{
+	cpwgr("<0, 0x2a>", (uint32_t)  val);
+	cpwgr("<0, 0x2b>", (uint32_t) (val >> 32));
+}
+
+/* dcache write access counter */
+static uint64_t csky_pmu_read_dcwac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x2d>");
+		lo  = cprgr("<0, 0x2c>");
+		hi  = cprgr("<0, 0x2d>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcwac(uint64_t val)
+{
+	cpwgr("<0, 0x2c>", (uint32_t)  val);
+	cpwgr("<0, 0x2d>", (uint32_t) (val >> 32));
+}
+
+/* dcache write miss counter */
+static uint64_t csky_pmu_read_dcwmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x2f>");
+		lo  = cprgr("<0, 0x2e>");
+		hi  = cprgr("<0, 0x2f>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_dcwmc(uint64_t val)
+{
+	cpwgr("<0, 0x2e>", (uint32_t)  val);
+	cpwgr("<0, 0x2f>", (uint32_t) (val >> 32));
+}
+
+/* l2cache read access counter */
+static uint64_t csky_pmu_read_l2rac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x31>");
+		lo  = cprgr("<0, 0x30>");
+		hi  = cprgr("<0, 0x31>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2rac(uint64_t val)
+{
+	cpwgr("<0, 0x30>", (uint32_t)  val);
+	cpwgr("<0, 0x31>", (uint32_t) (val >> 32));
+}
+
+/* l2cache read miss counter */
+static uint64_t csky_pmu_read_l2rmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x33>");
+		lo  = cprgr("<0, 0x32>");
+		hi  = cprgr("<0, 0x33>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2rmc(uint64_t val)
+{
+	cpwgr("<0, 0x32>", (uint32_t)  val);
+	cpwgr("<0, 0x33>", (uint32_t) (val >> 32));
+}
+
+/* l2cache write access counter */
+static uint64_t csky_pmu_read_l2wac(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x35>");
+		lo  = cprgr("<0, 0x34>");
+		hi  = cprgr("<0, 0x35>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2wac(uint64_t val)
+{
+	cpwgr("<0, 0x34>", (uint32_t)  val);
+	cpwgr("<0, 0x35>", (uint32_t) (val >> 32));
+}
+
+/* l2cache write miss counter */
+static uint64_t csky_pmu_read_l2wmc(void)
+{
+	uint32_t lo, hi, tmp;
+	uint64_t result;
+
+	do {
+		tmp = cprgr("<0, 0x37>");
+		lo  = cprgr("<0, 0x36>");
+		hi  = cprgr("<0, 0x37>");
+	} while (hi != tmp);
+
+	result = (uint64_t) (hi) << 32;
+	result |= lo;
+
+	return result;
+}
+
+static void csky_pmu_write_l2wmc(uint64_t val)
+{
+	cpwgr("<0, 0x36>", (uint32_t)  val);
+	cpwgr("<0, 0x37>", (uint32_t) (val >> 32));
+}
+
+#define HW_OP_UNSUPPORTED	0xffff
+static const int csky_pmu_hw_map[PERF_COUNT_HW_MAX] = {
+	[PERF_COUNT_HW_CPU_CYCLES]		= 0x1,
+	[PERF_COUNT_HW_INSTRUCTIONS]		= 0x2,
+	[PERF_COUNT_HW_CACHE_REFERENCES]	= HW_OP_UNSUPPORTED,
+	[PERF_COUNT_HW_CACHE_MISSES]		= HW_OP_UNSUPPORTED,
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= 0xf,
+	[PERF_COUNT_HW_BRANCH_MISSES]		= 0xe,
+	[PERF_COUNT_HW_BUS_CYCLES]		= HW_OP_UNSUPPORTED,
+	[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND]	= HW_OP_UNSUPPORTED,
+	[PERF_COUNT_HW_STALLED_CYCLES_BACKEND]	= HW_OP_UNSUPPORTED,
+	[PERF_COUNT_HW_REF_CPU_CYCLES]		= HW_OP_UNSUPPORTED,
+};
+
+#define C(_x)			PERF_COUNT_HW_CACHE_##_x
+#define CACHE_OP_UNSUPPORTED	0xffff
+static const int csky_pmu_cache_map[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = {
+	[C(L1D)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= 0x14,
+			[C(RESULT_MISS)]	= 0x15,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= 0x16,
+			[C(RESULT_MISS)]	= 0x17,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= 0x5,
+			[C(RESULT_MISS)]	= 0x6,
+		},
+	},
+	[C(L1I)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= 0x3,
+			[C(RESULT_MISS)]	= 0x4,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(LL)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= 0x18,
+			[C(RESULT_MISS)]	= 0x19,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= 0x1a,
+			[C(RESULT_MISS)]	= 0x1b,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= 0x7,
+			[C(RESULT_MISS)]	= 0x8,
+		},
+	},
+	[C(DTLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= 0x5,
+			[C(RESULT_MISS)]	= 0xb,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(ITLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= 0x3,
+			[C(RESULT_MISS)]	= 0xa,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(BPU)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(NODE)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+};
+
+static void csky_perf_event_update(struct perf_event *event,
+				   struct hw_perf_event *hwc)
+{
+	uint64_t prev_raw_count = local64_read(&hwc->prev_count);
+	uint64_t new_raw_count = hw_raw_read_mapping[hwc->idx]();
+	int64_t delta = new_raw_count - prev_raw_count;
+
+	/*
+	 * We aren't afraid of hwc->prev_count changing beneath our feet
+	 * because there's no way for us to re-enter this function anytime.
+	 */
+	local64_set(&hwc->prev_count, new_raw_count);
+	local64_add(delta, &event->count);
+	local64_sub(delta, &hwc->period_left);
+}
+
+static void csky_pmu_read(struct perf_event *event)
+{
+	csky_perf_event_update(event, &event->hw);
+}
+
+static int csky_pmu_cache_event(u64 config)
+{
+	unsigned int cache_type, cache_op, cache_result;
+
+	cache_type	= (config >>  0) & 0xff;
+	cache_op	= (config >>  8) & 0xff;
+	cache_result	= (config >> 16) & 0xff;
+
+	if (cache_type >= PERF_COUNT_HW_CACHE_MAX)
+		return -EINVAL;
+	if (cache_op >= PERF_COUNT_HW_CACHE_OP_MAX)
+		return -EINVAL;
+	if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
+		return -EINVAL;
+
+	return csky_pmu_cache_map[cache_type][cache_op][cache_result];
+}
+
+static int csky_pmu_event_init(struct perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int ret;
+
+	if (event->attr.exclude_user)
+		csky_pmu.hpcr = BIT(2);
+	else if (event->attr.exclude_kernel)
+		csky_pmu.hpcr = BIT(3);
+	else
+		csky_pmu.hpcr = BIT(2) | BIT(3);
+
+	csky_pmu.hpcr |= BIT(1) | BIT(0);
+
+	switch (event->attr.type) {
+	case PERF_TYPE_HARDWARE:
+		if (event->attr.config >= PERF_COUNT_HW_MAX)
+			return -ENOENT;
+		ret = csky_pmu_hw_map[event->attr.config];
+		if (ret == HW_OP_UNSUPPORTED)
+			return -ENOENT;
+		hwc->idx = ret;
+		return 0;
+	case PERF_TYPE_HW_CACHE:
+		ret = csky_pmu_cache_event(event->attr.config);
+		if (ret == CACHE_OP_UNSUPPORTED)
+			return -ENOENT;
+		hwc->idx = ret;
+		return 0;
+	case PERF_TYPE_RAW:
+		if (hw_raw_read_mapping[event->attr.config] == NULL)
+			return -ENOENT;
+		hwc->idx = event->attr.config;
+		return 0;
+	default:
+		return -ENOENT;
+	}
+}
+
+/* starts all counters */
+static void csky_pmu_enable(struct pmu *pmu)
+{
+	cpwcr(HPCR, csky_pmu.hpcr);
+}
+
+/* stops all counters */
+static void csky_pmu_disable(struct pmu *pmu)
+{
+	cpwcr(HPCR, BIT(1));
+}
+
+static void csky_pmu_start(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (WARN_ON_ONCE(idx == -1))
+		return;
+
+	if (flags & PERF_EF_RELOAD)
+		WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE));
+
+	hwc->state = 0;
+
+	cpwcr(HPCNTENR, BIT(idx) | cprcr(HPCNTENR));
+}
+
+static void csky_pmu_stop(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (!(event->hw.state & PERF_HES_STOPPED)) {
+		cpwcr(HPCNTENR, ~BIT(idx) & cprcr(HPCNTENR));
+		event->hw.state |= PERF_HES_STOPPED;
+	}
+
+	if ((flags & PERF_EF_UPDATE) &&
+	    !(event->hw.state & PERF_HES_UPTODATE)) {
+		csky_perf_event_update(event, &event->hw);
+		event->hw.state |= PERF_HES_UPTODATE;
+	}
+}
+
+static void csky_pmu_del(struct perf_event *event, int flags)
+{
+	csky_pmu_stop(event, PERF_EF_UPDATE);
+
+	perf_event_update_userpage(event);
+}
+
+/* allocate hardware counter and optionally start counting */
+static int csky_pmu_add(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+
+	local64_set(&hwc->prev_count, 0);
+
+	if (hw_raw_write_mapping[hwc->idx] != NULL)
+		hw_raw_write_mapping[hwc->idx](0);
+
+	hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+	if (flags & PERF_EF_START)
+		csky_pmu_start(event, PERF_EF_RELOAD);
+
+	perf_event_update_userpage(event);
+
+	return 0;
+}
+
+int __init init_hw_perf_events(void)
+{
+	csky_pmu.pmu = (struct pmu) {
+		.pmu_enable	= csky_pmu_enable,
+		.pmu_disable	= csky_pmu_disable,
+		.event_init	= csky_pmu_event_init,
+		.add		= csky_pmu_add,
+		.del		= csky_pmu_del,
+		.start		= csky_pmu_start,
+		.stop		= csky_pmu_stop,
+		.read		= csky_pmu_read,
+	};
+
+	memset((void *)hw_raw_read_mapping, 0,
+		sizeof(hw_raw_read_mapping[CSKY_PMU_MAX_EVENTS]));
+
+	hw_raw_read_mapping[0x1]  = csky_pmu_read_cc;
+	hw_raw_read_mapping[0x2]  = csky_pmu_read_ic;
+	hw_raw_read_mapping[0x3]  = csky_pmu_read_icac;
+	hw_raw_read_mapping[0x4]  = csky_pmu_read_icmc;
+	hw_raw_read_mapping[0x5]  = csky_pmu_read_dcac;
+	hw_raw_read_mapping[0x6]  = csky_pmu_read_dcmc;
+	hw_raw_read_mapping[0x7]  = csky_pmu_read_l2ac;
+	hw_raw_read_mapping[0x8]  = csky_pmu_read_l2mc;
+	hw_raw_read_mapping[0xa]  = csky_pmu_read_iutlbmc;
+	hw_raw_read_mapping[0xb]  = csky_pmu_read_dutlbmc;
+	hw_raw_read_mapping[0xc]  = csky_pmu_read_jtlbmc;
+	hw_raw_read_mapping[0xd]  = csky_pmu_read_softc;
+	hw_raw_read_mapping[0xe]  = csky_pmu_read_cbmc;
+	hw_raw_read_mapping[0xf]  = csky_pmu_read_cbic;
+	hw_raw_read_mapping[0x10] = csky_pmu_read_ibmc;
+	hw_raw_read_mapping[0x11] = csky_pmu_read_ibic;
+	hw_raw_read_mapping[0x12] = csky_pmu_read_lsfc;
+	hw_raw_read_mapping[0x13] = csky_pmu_read_sic;
+	hw_raw_read_mapping[0x14] = csky_pmu_read_dcrac;
+	hw_raw_read_mapping[0x15] = csky_pmu_read_dcrmc;
+	hw_raw_read_mapping[0x16] = csky_pmu_read_dcwac;
+	hw_raw_read_mapping[0x17] = csky_pmu_read_dcwmc;
+	hw_raw_read_mapping[0x18] = csky_pmu_read_l2rac;
+	hw_raw_read_mapping[0x19] = csky_pmu_read_l2rmc;
+	hw_raw_read_mapping[0x1a] = csky_pmu_read_l2wac;
+	hw_raw_read_mapping[0x1b] = csky_pmu_read_l2wmc;
+
+	memset((void *)hw_raw_write_mapping, 0,
+		sizeof(hw_raw_write_mapping[CSKY_PMU_MAX_EVENTS]));
+
+	hw_raw_write_mapping[0x1]  = csky_pmu_write_cc;
+	hw_raw_write_mapping[0x2]  = csky_pmu_write_ic;
+	hw_raw_write_mapping[0x3]  = csky_pmu_write_icac;
+	hw_raw_write_mapping[0x4]  = csky_pmu_write_icmc;
+	hw_raw_write_mapping[0x5]  = csky_pmu_write_dcac;
+	hw_raw_write_mapping[0x6]  = csky_pmu_write_dcmc;
+	hw_raw_write_mapping[0x7]  = csky_pmu_write_l2ac;
+	hw_raw_write_mapping[0x8]  = csky_pmu_write_l2mc;
+	hw_raw_write_mapping[0xa]  = csky_pmu_write_iutlbmc;
+	hw_raw_write_mapping[0xb]  = csky_pmu_write_dutlbmc;
+	hw_raw_write_mapping[0xc]  = csky_pmu_write_jtlbmc;
+	hw_raw_write_mapping[0xd]  = csky_pmu_write_softc;
+	hw_raw_write_mapping[0xe]  = csky_pmu_write_cbmc;
+	hw_raw_write_mapping[0xf]  = csky_pmu_write_cbic;
+	hw_raw_write_mapping[0x10] = csky_pmu_write_ibmc;
+	hw_raw_write_mapping[0x11] = csky_pmu_write_ibic;
+	hw_raw_write_mapping[0x12] = csky_pmu_write_lsfc;
+	hw_raw_write_mapping[0x13] = csky_pmu_write_sic;
+	hw_raw_write_mapping[0x14] = csky_pmu_write_dcrac;
+	hw_raw_write_mapping[0x15] = csky_pmu_write_dcrmc;
+	hw_raw_write_mapping[0x16] = csky_pmu_write_dcwac;
+	hw_raw_write_mapping[0x17] = csky_pmu_write_dcwmc;
+	hw_raw_write_mapping[0x18] = csky_pmu_write_l2rac;
+	hw_raw_write_mapping[0x19] = csky_pmu_write_l2rmc;
+	hw_raw_write_mapping[0x1a] = csky_pmu_write_l2wac;
+	hw_raw_write_mapping[0x1b] = csky_pmu_write_l2wmc;
+
+	csky_pmu.pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT;
+
+	cpwcr(HPCR, BIT(31) | BIT(30) | BIT(1));
+
+	return perf_pmu_register(&csky_pmu.pmu, "cpu", PERF_TYPE_RAW);
+}
+arch_initcall(init_hw_perf_events);
diff --git a/arch/csky/kernel/process.c b/arch/csky/kernel/process.c
index 8ed20028b160..e555740c0be5 100644
--- a/arch/csky/kernel/process.c
+++ b/arch/csky/kernel/process.c
@@ -93,26 +93,31 @@ int dump_task_regs(struct task_struct *tsk, elf_gregset_t *pr_regs)
 
 unsigned long get_wchan(struct task_struct *p)
 {
-	unsigned long esp, pc;
-	unsigned long stack_page;
+	unsigned long lr;
+	unsigned long *fp, *stack_start, *stack_end;
 	int count = 0;
 
 	if (!p || p == current || p->state == TASK_RUNNING)
 		return 0;
 
-	stack_page = (unsigned long)p;
-	esp = p->thread.esp0;
+	stack_start = (unsigned long *)end_of_stack(p);
+	stack_end = (unsigned long *)(task_stack_page(p) + THREAD_SIZE);
+
+	fp = (unsigned long *) thread_saved_fp(p);
 	do {
-		if (esp < stack_page+sizeof(struct task_struct) ||
-		    esp >= 8184+stack_page)
+		if (fp < stack_start || fp > stack_end)
 			return 0;
-		/*FIXME: There's may be error here!*/
-		pc = ((unsigned long *)esp)[1];
-		/* FIXME: This depends on the order of these functions. */
-		if (!in_sched_functions(pc))
-			return pc;
-		esp = *(unsigned long *) esp;
+#ifdef CONFIG_STACKTRACE
+		lr = fp[1];
+		fp = (unsigned long *)fp[0];
+#else
+		lr = *fp++;
+#endif
+		if (!in_sched_functions(lr) &&
+		    __kernel_text_address(lr))
+			return lr;
 	} while (count++ < 16);
+
 	return 0;
 }
 EXPORT_SYMBOL(get_wchan);
diff --git a/arch/csky/kernel/ptrace.c b/arch/csky/kernel/ptrace.c
index 34b30257298f..57f1afe19a52 100644
--- a/arch/csky/kernel/ptrace.c
+++ b/arch/csky/kernel/ptrace.c
@@ -50,15 +50,11 @@ static void singlestep_enable(struct task_struct *tsk)
  */
 void user_enable_single_step(struct task_struct *child)
 {
-	if (child->thread.esp0 == 0)
-		return;
 	singlestep_enable(child);
 }
 
 void user_disable_single_step(struct task_struct *child)
 {
-	if (child->thread.esp0 == 0)
-		return;
 	singlestep_disable(child);
 }
 
@@ -95,7 +91,9 @@ static int gpr_set(struct task_struct *target,
 		return ret;
 
 	regs.sr = task_pt_regs(target)->sr;
-
+#ifdef CONFIG_CPU_HAS_HILO
+	regs.dcsr = task_pt_regs(target)->dcsr;
+#endif
 	task_thread_info(target)->tp_value = regs.tls;
 
 	*task_pt_regs(target) = regs;
@@ -239,6 +237,7 @@ asmlinkage void syscall_trace(int why, struct pt_regs *regs)
 	regs->regs[SYSTRACE_SAVENUM] = saved_why;
 }
 
+extern void show_stack(struct task_struct *task, unsigned long *stack);
 void show_regs(struct pt_regs *fp)
 {
 	unsigned long   *sp;
@@ -261,35 +260,37 @@ void show_regs(struct pt_regs *fp)
 		       (int) (((unsigned long) current) + 2 * PAGE_SIZE));
 	}
 
-	pr_info("PC: 0x%08lx\n", (long)fp->pc);
+	pr_info("PC: 0x%08lx (%pS)\n", (long)fp->pc, (void *)fp->pc);
+	pr_info("LR: 0x%08lx (%pS)\n", (long)fp->lr, (void *)fp->lr);
+	pr_info("SP: 0x%08lx\n", (long)fp);
 	pr_info("orig_a0: 0x%08lx\n", fp->orig_a0);
 	pr_info("PSR: 0x%08lx\n", (long)fp->sr);
 
-	pr_info("a0: 0x%08lx  a1: 0x%08lx  a2: 0x%08lx  a3: 0x%08lx\n",
-	       fp->a0, fp->a1, fp->a2, fp->a3);
+	pr_info(" a0: 0x%08lx   a1: 0x%08lx   a2: 0x%08lx   a3: 0x%08lx\n",
+		fp->a0, fp->a1, fp->a2, fp->a3);
 #if defined(__CSKYABIV2__)
-	pr_info("r4: 0x%08lx  r5: 0x%08lx    r6: 0x%08lx    r7: 0x%08lx\n",
+	pr_info(" r4: 0x%08lx   r5: 0x%08lx   r6: 0x%08lx   r7: 0x%08lx\n",
 		fp->regs[0], fp->regs[1], fp->regs[2], fp->regs[3]);
-	pr_info("r8: 0x%08lx  r9: 0x%08lx   r10: 0x%08lx   r11: 0x%08lx\n",
+	pr_info(" r8: 0x%08lx   r9: 0x%08lx  r10: 0x%08lx  r11: 0x%08lx\n",
 		fp->regs[4], fp->regs[5], fp->regs[6], fp->regs[7]);
-	pr_info("r12 0x%08lx  r13: 0x%08lx   r15: 0x%08lx\n",
+	pr_info("r12: 0x%08lx  r13: 0x%08lx  r15: 0x%08lx\n",
 		fp->regs[8], fp->regs[9], fp->lr);
-	pr_info("r16:0x%08lx   r17: 0x%08lx   r18: 0x%08lx    r19: 0x%08lx\n",
+	pr_info("r16: 0x%08lx  r17: 0x%08lx  r18: 0x%08lx  r19: 0x%08lx\n",
 		fp->exregs[0], fp->exregs[1], fp->exregs[2], fp->exregs[3]);
-	pr_info("r20 0x%08lx   r21: 0x%08lx   r22: 0x%08lx    r23: 0x%08lx\n",
+	pr_info("r20: 0x%08lx  r21: 0x%08lx  r22: 0x%08lx  r23: 0x%08lx\n",
 		fp->exregs[4], fp->exregs[5], fp->exregs[6], fp->exregs[7]);
-	pr_info("r24 0x%08lx   r25: 0x%08lx   r26: 0x%08lx    r27: 0x%08lx\n",
+	pr_info("r24: 0x%08lx  r25: 0x%08lx  r26: 0x%08lx  r27: 0x%08lx\n",
 		fp->exregs[8], fp->exregs[9], fp->exregs[10], fp->exregs[11]);
-	pr_info("r28 0x%08lx   r29: 0x%08lx   r30: 0x%08lx    tls: 0x%08lx\n",
+	pr_info("r28: 0x%08lx  r29: 0x%08lx  r30: 0x%08lx  tls: 0x%08lx\n",
 		fp->exregs[12], fp->exregs[13], fp->exregs[14], fp->tls);
-	pr_info("hi 0x%08lx    lo: 0x%08lx\n",
+	pr_info(" hi: 0x%08lx   lo: 0x%08lx\n",
 		fp->rhi, fp->rlo);
 #else
-	pr_info("r6: 0x%08lx   r7: 0x%08lx   r8: 0x%08lx   r9: 0x%08lx\n",
+	pr_info(" r6: 0x%08lx   r7: 0x%08lx   r8: 0x%08lx   r9: 0x%08lx\n",
 		fp->regs[0], fp->regs[1], fp->regs[2], fp->regs[3]);
-	pr_info("r10: 0x%08lx   r11: 0x%08lx   r12: 0x%08lx   r13: 0x%08lx\n",
+	pr_info("r10: 0x%08lx  r11: 0x%08lx  r12: 0x%08lx  r13: 0x%08lx\n",
 		fp->regs[4], fp->regs[5], fp->regs[6], fp->regs[7]);
-	pr_info("r14 0x%08lx   r1: 0x%08lx   r15: 0x%08lx\n",
+	pr_info("r14: 0x%08lx   r1: 0x%08lx  r15: 0x%08lx\n",
 		fp->regs[8], fp->regs[9], fp->lr);
 #endif
 
@@ -311,4 +312,7 @@ void show_regs(struct pt_regs *fp)
 		pr_cont("%08x ", (int) *sp++);
 	}
 	pr_cont("\n");
+
+	show_stack(NULL, (unsigned long *)fp->regs[4]);
+	return;
 }
diff --git a/arch/csky/kernel/signal.c b/arch/csky/kernel/signal.c
index 9967c10eee2b..207a891479d2 100644
--- a/arch/csky/kernel/signal.c
+++ b/arch/csky/kernel/signal.c
@@ -238,8 +238,6 @@ static void do_signal(struct pt_regs *regs, int syscall)
 	if (!user_mode(regs))
 		return;
 
-	current->thread.esp0 = (unsigned long)regs;
-
 	/*
 	 * If we were from a system call, check for system call restarting...
 	 */
diff --git a/arch/csky/kernel/smp.c b/arch/csky/kernel/smp.c
index 36ebaf9834e1..ddc4dd79f282 100644
--- a/arch/csky/kernel/smp.c
+++ b/arch/csky/kernel/smp.c
@@ -16,6 +16,7 @@
 #include <linux/of.h>
 #include <linux/sched/task_stack.h>
 #include <linux/sched/mm.h>
+#include <linux/sched/hotplug.h>
 #include <asm/irq.h>
 #include <asm/traps.h>
 #include <asm/sections.h>
@@ -112,12 +113,8 @@ void __init smp_prepare_cpus(unsigned int max_cpus)
 {
 }
 
-static void __init enable_smp_ipi(void)
-{
-	enable_percpu_irq(ipi_irq, 0);
-}
-
 static int ipi_dummy_dev;
+
 void __init setup_smp_ipi(void)
 {
 	int rc;
@@ -130,7 +127,7 @@ void __init setup_smp_ipi(void)
 	if (rc)
 		panic("%s IRQ request failed\n", __func__);
 
-	enable_smp_ipi();
+	enable_percpu_irq(ipi_irq, 0);
 }
 
 void __init setup_smp(void)
@@ -138,7 +135,7 @@ void __init setup_smp(void)
 	struct device_node *node = NULL;
 	int cpu;
 
-	while ((node = of_find_node_by_type(node, "cpu"))) {
+	for_each_of_cpu_node(node) {
 		if (!of_device_is_available(node))
 			continue;
 
@@ -161,12 +158,10 @@ volatile unsigned int secondary_stack;
 
 int __cpu_up(unsigned int cpu, struct task_struct *tidle)
 {
-	unsigned int tmp;
-
-	secondary_stack = (unsigned int)tidle->stack + THREAD_SIZE;
+	unsigned long mask = 1 << cpu;
 
+	secondary_stack = (unsigned int)tidle->stack + THREAD_SIZE - 8;
 	secondary_hint = mfcr("cr31");
-
 	secondary_ccr  = mfcr("cr18");
 
 	/*
@@ -176,10 +171,13 @@ int __cpu_up(unsigned int cpu, struct task_struct *tidle)
 	 */
 	mtcr("cr17", 0x22);
 
-	/* Enable cpu in SMP reset ctrl reg */
-	tmp = mfcr("cr<29, 0>");
-	tmp |= 1 << cpu;
-	mtcr("cr<29, 0>", tmp);
+	if (mask & mfcr("cr<29, 0>")) {
+		send_arch_ipi(cpumask_of(cpu));
+	} else {
+		/* Enable cpu in SMP reset ctrl reg */
+		mask |= mfcr("cr<29, 0>");
+		mtcr("cr<29, 0>", mask);
+	}
 
 	/* Wait for the cpu online */
 	while (!cpu_online(cpu));
@@ -219,7 +217,7 @@ void csky_start_secondary(void)
 	init_fpu();
 #endif
 
-	enable_smp_ipi();
+	enable_percpu_irq(ipi_irq, 0);
 
 	mmget(mm);
 	mmgrab(mm);
@@ -235,3 +233,46 @@ void csky_start_secondary(void)
 	preempt_disable();
 	cpu_startup_entry(CPUHP_AP_ONLINE_IDLE);
 }
+
+#ifdef CONFIG_HOTPLUG_CPU
+int __cpu_disable(void)
+{
+	unsigned int cpu = smp_processor_id();
+
+	set_cpu_online(cpu, false);
+
+	irq_migrate_all_off_this_cpu();
+
+	clear_tasks_mm_cpumask(cpu);
+
+	return 0;
+}
+
+void __cpu_die(unsigned int cpu)
+{
+	if (!cpu_wait_death(cpu, 5)) {
+		pr_crit("CPU%u: shutdown failed\n", cpu);
+		return;
+	}
+	pr_notice("CPU%u: shutdown\n", cpu);
+}
+
+void arch_cpu_idle_dead(void)
+{
+	idle_task_exit();
+
+	cpu_report_death();
+
+	while (!secondary_stack)
+		arch_cpu_idle();
+
+	local_irq_disable();
+
+	asm volatile(
+		"mov	sp, %0\n"
+		"mov	r8, %0\n"
+		"jmpi	csky_start_secondary"
+		:
+		: "r" (secondary_stack));
+}
+#endif
diff --git a/arch/csky/kernel/stacktrace.c b/arch/csky/kernel/stacktrace.c
new file mode 100644
index 000000000000..fec777a643f1
--- /dev/null
+++ b/arch/csky/kernel/stacktrace.c
@@ -0,0 +1,57 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (C) 2018 Hangzhou C-SKY Microsystems co.,ltd. */
+
+#include <linux/sched/debug.h>
+#include <linux/sched/task_stack.h>
+#include <linux/stacktrace.h>
+#include <linux/ftrace.h>
+
+void save_stack_trace(struct stack_trace *trace)
+{
+	save_stack_trace_tsk(current, trace);
+}
+EXPORT_SYMBOL_GPL(save_stack_trace);
+
+void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
+{
+	unsigned long *fp, *stack_start, *stack_end;
+	unsigned long addr;
+	int skip = trace->skip;
+	int savesched;
+	int graph_idx = 0;
+
+	if (tsk == current) {
+		asm volatile("mov %0, r8\n":"=r"(fp));
+		savesched = 1;
+	} else {
+		fp = (unsigned long *)thread_saved_fp(tsk);
+		savesched = 0;
+	}
+
+	addr = (unsigned long) fp & THREAD_MASK;
+	stack_start = (unsigned long *) addr;
+	stack_end = (unsigned long *) (addr + THREAD_SIZE);
+
+	while (fp > stack_start && fp < stack_end) {
+		unsigned long lpp, fpp;
+
+		fpp = fp[0];
+		lpp = fp[1];
+		if (!__kernel_text_address(lpp))
+			break;
+		else
+			lpp = ftrace_graph_ret_addr(tsk, &graph_idx, lpp, NULL);
+
+		if (savesched || !in_sched_functions(lpp)) {
+			if (skip) {
+				skip--;
+			} else {
+				trace->entries[trace->nr_entries++] = lpp;
+				if (trace->nr_entries >= trace->max_entries)
+					break;
+			}
+		}
+		fp = (unsigned long *)fpp;
+	}
+}
+EXPORT_SYMBOL_GPL(save_stack_trace_tsk);
diff --git a/arch/csky/kernel/traps.c b/arch/csky/kernel/traps.c
index a8368ed43517..f487a9b996ae 100644
--- a/arch/csky/kernel/traps.c
+++ b/arch/csky/kernel/traps.c
@@ -106,7 +106,6 @@ void buserr(struct pt_regs *regs)
 	pr_err("User mode Bus Error\n");
 	show_regs(regs);
 
-	current->thread.esp0 = (unsigned long) regs;
 	force_sig_fault(SIGSEGV, 0, (void __user *)regs->pc, current);
 }
 
@@ -162,8 +161,3 @@ asmlinkage void trap_c(struct pt_regs *regs)
 	}
 	send_sig(sig, current, 0);
 }
-
-asmlinkage void set_esp0(unsigned long ssp)
-{
-	current->thread.esp0 = ssp;
-}