diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index 6db513674050..9908d477ccd9 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -37,6 +37,9 @@ config SPARC64
 	def_bool 64BIT
 	select ARCH_SUPPORTS_MSI
 	select HAVE_FUNCTION_TRACER
+	select HAVE_FUNCTION_GRAPH_TRACER
+	select HAVE_FUNCTION_GRAPH_FP_TEST
+	select HAVE_FUNCTION_TRACE_MCOUNT_TEST
 	select HAVE_KRETPROBES
 	select HAVE_KPROBES
 	select HAVE_LMB
diff --git a/arch/sparc/Kconfig.debug b/arch/sparc/Kconfig.debug
index 9d3c889718ac..1b4a831565f9 100644
--- a/arch/sparc/Kconfig.debug
+++ b/arch/sparc/Kconfig.debug
@@ -19,13 +19,10 @@ config DEBUG_DCFLUSH
 	bool "D-cache flush debugging"
 	depends on SPARC64 && DEBUG_KERNEL
 
-config STACK_DEBUG
-	bool "Stack Overflow Detection Support"
-
 config MCOUNT
 	bool
 	depends on SPARC64
-	depends on STACK_DEBUG || FUNCTION_TRACER
+	depends on FUNCTION_TRACER
 	default y
 
 config FRAME_POINTER
diff --git a/arch/sparc/include/asm/cpudata_64.h b/arch/sparc/include/asm/cpudata_64.h
index 926397d345ff..050ef35b9dcf 100644
--- a/arch/sparc/include/asm/cpudata_64.h
+++ b/arch/sparc/include/asm/cpudata_64.h
@@ -17,7 +17,7 @@ typedef struct {
 	unsigned int	__nmi_count;
 	unsigned long	clock_tick;	/* %tick's per second */
 	unsigned long	__pad;
-	unsigned int	__pad1;
+	unsigned int	irq0_irqs;
 	unsigned int	__pad2;
 
 	/* Dcache line 2, rarely used */
diff --git a/arch/sparc/include/asm/irqflags_64.h b/arch/sparc/include/asm/irqflags_64.h
index 8b49bf920df3..bfa1ea45b4cd 100644
--- a/arch/sparc/include/asm/irqflags_64.h
+++ b/arch/sparc/include/asm/irqflags_64.h
@@ -76,9 +76,26 @@ static inline int raw_irqs_disabled(void)
  */
 static inline unsigned long __raw_local_irq_save(void)
 {
-	unsigned long flags = __raw_local_save_flags();
+	unsigned long flags, tmp;
 
-	raw_local_irq_disable();
+	/* Disable interrupts to PIL_NORMAL_MAX unless we already
+	 * are using PIL_NMI, in which case PIL_NMI is retained.
+	 *
+	 * The only values we ever program into the %pil are 0,
+	 * PIL_NORMAL_MAX and PIL_NMI.
+	 *
+	 * Since PIL_NMI is the largest %pil value and all bits are
+	 * set in it (0xf), it doesn't matter what PIL_NORMAL_MAX
+	 * actually is.
+	 */
+	__asm__ __volatile__(
+		"rdpr	%%pil, %0\n\t"
+		"or	%0, %2, %1\n\t"
+		"wrpr	%1, 0x0, %%pil"
+		: "=r" (flags), "=r" (tmp)
+		: "i" (PIL_NORMAL_MAX)
+		: "memory"
+	);
 
 	return flags;
 }
diff --git a/arch/sparc/kernel/Makefile b/arch/sparc/kernel/Makefile
index c6316142db4e..0c2dc1f24a9a 100644
--- a/arch/sparc/kernel/Makefile
+++ b/arch/sparc/kernel/Makefile
@@ -13,6 +13,14 @@ extra-y     += init_task.o
 CPPFLAGS_vmlinux.lds := -Usparc -m$(BITS)
 extra-y              += vmlinux.lds
 
+ifdef CONFIG_FUNCTION_TRACER
+# Do not profile debug and lowlevel utilities
+CFLAGS_REMOVE_ftrace.o := -pg
+CFLAGS_REMOVE_time_$(BITS).o := -pg
+CFLAGS_REMOVE_perf_event.o := -pg
+CFLAGS_REMOVE_pcr.o := -pg
+endif
+
 obj-$(CONFIG_SPARC32)   += entry.o wof.o wuf.o
 obj-$(CONFIG_SPARC32)   += etrap_32.o
 obj-$(CONFIG_SPARC32)   += rtrap_32.o
@@ -85,7 +93,7 @@ obj-$(CONFIG_KGDB)        += kgdb_$(BITS).o
 
 
 obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o
-CFLAGS_REMOVE_ftrace.o := -pg
+obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
 
 obj-$(CONFIG_EARLYFB) += btext.o
 obj-$(CONFIG_STACKTRACE)     += stacktrace.o
diff --git a/arch/sparc/kernel/ftrace.c b/arch/sparc/kernel/ftrace.c
index 9103a56b39e8..03ab022e51c5 100644
--- a/arch/sparc/kernel/ftrace.c
+++ b/arch/sparc/kernel/ftrace.c
@@ -13,7 +13,7 @@ static const u32 ftrace_nop = 0x01000000;
 
 static u32 ftrace_call_replace(unsigned long ip, unsigned long addr)
 {
-	static u32 call;
+	u32 call;
 	s32 off;
 
 	off = ((s32)addr - (s32)ip);
@@ -91,3 +91,61 @@ int __init ftrace_dyn_arch_init(void *data)
 	return 0;
 }
 #endif
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+
+#ifdef CONFIG_DYNAMIC_FTRACE
+extern void ftrace_graph_call(void);
+
+int ftrace_enable_ftrace_graph_caller(void)
+{
+	unsigned long ip = (unsigned long)(&ftrace_graph_call);
+	u32 old, new;
+
+	old = *(u32 *) &ftrace_graph_call;
+	new = ftrace_call_replace(ip, (unsigned long) &ftrace_graph_caller);
+	return ftrace_modify_code(ip, old, new);
+}
+
+int ftrace_disable_ftrace_graph_caller(void)
+{
+	unsigned long ip = (unsigned long)(&ftrace_graph_call);
+	u32 old, new;
+
+	old = *(u32 *) &ftrace_graph_call;
+	new = ftrace_call_replace(ip, (unsigned long) &ftrace_stub);
+
+	return ftrace_modify_code(ip, old, new);
+}
+
+#endif /* !CONFIG_DYNAMIC_FTRACE */
+
+/*
+ * Hook the return address and push it in the stack of return addrs
+ * in current thread info.
+ */
+unsigned long prepare_ftrace_return(unsigned long parent,
+				    unsigned long self_addr,
+				    unsigned long frame_pointer)
+{
+	unsigned long return_hooker = (unsigned long) &return_to_handler;
+	struct ftrace_graph_ent trace;
+
+	if (unlikely(atomic_read(&current->tracing_graph_pause)))
+		return parent + 8UL;
+
+	if (ftrace_push_return_trace(parent, self_addr, &trace.depth,
+				     frame_pointer) == -EBUSY)
+		return parent + 8UL;
+
+	trace.func = self_addr;
+
+	/* Only trace if the calling function expects to */
+	if (!ftrace_graph_entry(&trace)) {
+		current->curr_ret_stack--;
+		return parent + 8UL;
+	}
+
+	return return_hooker;
+}
+#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c
index e1cbdb94d97b..454ce3a25273 100644
--- a/arch/sparc/kernel/irq_64.c
+++ b/arch/sparc/kernel/irq_64.c
@@ -20,6 +20,7 @@
 #include <linux/delay.h>
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
+#include <linux/ftrace.h>
 #include <linux/irq.h>
 
 #include <asm/ptrace.h>
@@ -647,6 +648,14 @@ unsigned int sun4v_build_virq(u32 devhandle, unsigned int devino)
 	bucket = kzalloc(sizeof(struct ino_bucket), GFP_ATOMIC);
 	if (unlikely(!bucket))
 		return 0;
+
+	/* The only reference we store to the IRQ bucket is
+	 * by physical address which kmemleak can't see, tell
+	 * it that this object explicitly is not a leak and
+	 * should be scanned.
+	 */
+	kmemleak_not_leak(bucket);
+
 	__flush_dcache_range((unsigned long) bucket,
 			     ((unsigned long) bucket +
 			      sizeof(struct ino_bucket)));
@@ -721,7 +730,7 @@ static __attribute__((always_inline)) void restore_hardirq_stack(void *orig_sp)
 	__asm__ __volatile__("mov %0, %%sp" : : "r" (orig_sp));
 }
 
-void handler_irq(int irq, struct pt_regs *regs)
+void __irq_entry handler_irq(int irq, struct pt_regs *regs)
 {
 	unsigned long pstate, bucket_pa;
 	struct pt_regs *old_regs;
diff --git a/arch/sparc/kernel/kgdb_64.c b/arch/sparc/kernel/kgdb_64.c
index f5a0fd490b59..0a2bd0f99fc1 100644
--- a/arch/sparc/kernel/kgdb_64.c
+++ b/arch/sparc/kernel/kgdb_64.c
@@ -5,6 +5,7 @@
 
 #include <linux/kgdb.h>
 #include <linux/kdebug.h>
+#include <linux/ftrace.h>
 
 #include <asm/kdebug.h>
 #include <asm/ptrace.h>
@@ -108,7 +109,7 @@ void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs)
 }
 
 #ifdef CONFIG_SMP
-void smp_kgdb_capture_client(int irq, struct pt_regs *regs)
+void __irq_entry smp_kgdb_capture_client(int irq, struct pt_regs *regs)
 {
 	unsigned long flags;
 
diff --git a/arch/sparc/kernel/nmi.c b/arch/sparc/kernel/nmi.c
index b287b62c7ea3..75a3d1a25356 100644
--- a/arch/sparc/kernel/nmi.c
+++ b/arch/sparc/kernel/nmi.c
@@ -92,7 +92,6 @@ static void die_nmi(const char *str, struct pt_regs *regs, int do_panic)
 notrace __kprobes void perfctr_irq(int irq, struct pt_regs *regs)
 {
 	unsigned int sum, touched = 0;
-	int cpu = smp_processor_id();
 
 	clear_softint(1 << irq);
 
@@ -106,7 +105,7 @@ notrace __kprobes void perfctr_irq(int irq, struct pt_regs *regs)
 	else
 		pcr_ops->write(PCR_PIC_PRIV);
 
-	sum = kstat_irqs_cpu(0, cpu);
+	sum = local_cpu_data().irq0_irqs;
 	if (__get_cpu_var(nmi_touch)) {
 		__get_cpu_var(nmi_touch) = 0;
 		touched = 1;
diff --git a/arch/sparc/kernel/pci_common.c b/arch/sparc/kernel/pci_common.c
index b775658a927d..8a000583b5cf 100644
--- a/arch/sparc/kernel/pci_common.c
+++ b/arch/sparc/kernel/pci_common.c
@@ -371,14 +371,19 @@ static void pci_register_iommu_region(struct pci_pbm_info *pbm)
 		struct resource *rp = kzalloc(sizeof(*rp), GFP_KERNEL);
 
 		if (!rp) {
-			prom_printf("Cannot allocate IOMMU resource.\n");
-			prom_halt();
+			pr_info("%s: Cannot allocate IOMMU resource.\n",
+				pbm->name);
+			return;
 		}
 		rp->name = "IOMMU";
 		rp->start = pbm->mem_space.start + (unsigned long) vdma[0];
 		rp->end = rp->start + (unsigned long) vdma[1] - 1UL;
 		rp->flags = IORESOURCE_BUSY;
-		request_resource(&pbm->mem_space, rp);
+		if (request_resource(&pbm->mem_space, rp)) {
+			pr_info("%s: Unable to request IOMMU resource.\n",
+				pbm->name);
+			kfree(rp);
+		}
 	}
 }
 
diff --git a/arch/sparc/kernel/pcr.c b/arch/sparc/kernel/pcr.c
index 2d94e7a03af5..c4a6a50b4849 100644
--- a/arch/sparc/kernel/pcr.c
+++ b/arch/sparc/kernel/pcr.c
@@ -8,6 +8,7 @@
 #include <linux/irq.h>
 
 #include <linux/perf_event.h>
+#include <linux/ftrace.h>
 
 #include <asm/pil.h>
 #include <asm/pcr.h>
@@ -34,7 +35,7 @@ unsigned int picl_shift;
  * Therefore in such situations we defer the work by signalling
  * a lower level cpu IRQ.
  */
-void deferred_pcr_work_irq(int irq, struct pt_regs *regs)
+void __irq_entry deferred_pcr_work_irq(int irq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs;
 
diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c
index 4c5334528109..b6a2b8f47040 100644
--- a/arch/sparc/kernel/smp_64.c
+++ b/arch/sparc/kernel/smp_64.c
@@ -22,6 +22,7 @@
 #include <linux/profile.h>
 #include <linux/bootmem.h>
 #include <linux/vmalloc.h>
+#include <linux/ftrace.h>
 #include <linux/cpu.h>
 #include <linux/slab.h>
 
@@ -823,13 +824,13 @@ void arch_send_call_function_single_ipi(int cpu)
 		      &cpumask_of_cpu(cpu));
 }
 
-void smp_call_function_client(int irq, struct pt_regs *regs)
+void __irq_entry smp_call_function_client(int irq, struct pt_regs *regs)
 {
 	clear_softint(1 << irq);
 	generic_smp_call_function_interrupt();
 }
 
-void smp_call_function_single_client(int irq, struct pt_regs *regs)
+void __irq_entry smp_call_function_single_client(int irq, struct pt_regs *regs)
 {
 	clear_softint(1 << irq);
 	generic_smp_call_function_single_interrupt();
@@ -965,7 +966,7 @@ void flush_dcache_page_all(struct mm_struct *mm, struct page *page)
 	put_cpu();
 }
 
-void smp_new_mmu_context_version_client(int irq, struct pt_regs *regs)
+void __irq_entry smp_new_mmu_context_version_client(int irq, struct pt_regs *regs)
 {
 	struct mm_struct *mm;
 	unsigned long flags;
@@ -1149,7 +1150,7 @@ void smp_release(void)
  */
 extern void prom_world(int);
 
-void smp_penguin_jailcell(int irq, struct pt_regs *regs)
+void __irq_entry smp_penguin_jailcell(int irq, struct pt_regs *regs)
 {
 	clear_softint(1 << irq);
 
@@ -1365,7 +1366,7 @@ void smp_send_reschedule(int cpu)
 		      &cpumask_of_cpu(cpu));
 }
 
-void smp_receive_signal_client(int irq, struct pt_regs *regs)
+void __irq_entry smp_receive_signal_client(int irq, struct pt_regs *regs)
 {
 	clear_softint(1 << irq);
 }
diff --git a/arch/sparc/kernel/time_64.c b/arch/sparc/kernel/time_64.c
index 67e165102885..c7bbe6cf7b85 100644
--- a/arch/sparc/kernel/time_64.c
+++ b/arch/sparc/kernel/time_64.c
@@ -35,6 +35,7 @@
 #include <linux/clocksource.h>
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
+#include <linux/ftrace.h>
 
 #include <asm/oplib.h>
 #include <asm/timer.h>
@@ -717,7 +718,7 @@ static struct clock_event_device sparc64_clockevent = {
 };
 static DEFINE_PER_CPU(struct clock_event_device, sparc64_events);
 
-void timer_interrupt(int irq, struct pt_regs *regs)
+void __irq_entry timer_interrupt(int irq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs = set_irq_regs(regs);
 	unsigned long tick_mask = tick_ops->softint_mask;
@@ -728,6 +729,7 @@ void timer_interrupt(int irq, struct pt_regs *regs)
 
 	irq_enter();
 
+	local_cpu_data().irq0_irqs++;
 	kstat_incr_irqs_this_cpu(0, irq_to_desc(0));
 
 	if (unlikely(!evt->event_handler)) {
diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c
index 837dfc2390d6..9da57f032983 100644
--- a/arch/sparc/kernel/traps_64.c
+++ b/arch/sparc/kernel/traps_64.c
@@ -2203,27 +2203,6 @@ void dump_stack(void)
 
 EXPORT_SYMBOL(dump_stack);
 
-static inline int is_kernel_stack(struct task_struct *task,
-				  struct reg_window *rw)
-{
-	unsigned long rw_addr = (unsigned long) rw;
-	unsigned long thread_base, thread_end;
-
-	if (rw_addr < PAGE_OFFSET) {
-		if (task != &init_task)
-			return 0;
-	}
-
-	thread_base = (unsigned long) task_stack_page(task);
-	thread_end = thread_base + sizeof(union thread_union);
-	if (rw_addr >= thread_base &&
-	    rw_addr < thread_end &&
-	    !(rw_addr & 0x7UL))
-		return 1;
-
-	return 0;
-}
-
 static inline struct reg_window *kernel_stack_up(struct reg_window *rw)
 {
 	unsigned long fp = rw->ins[6];
@@ -2252,6 +2231,7 @@ void die_if_kernel(char *str, struct pt_regs *regs)
 	show_regs(regs);
 	add_taint(TAINT_DIE);
 	if (regs->tstate & TSTATE_PRIV) {
+		struct thread_info *tp = current_thread_info();
 		struct reg_window *rw = (struct reg_window *)
 			(regs->u_regs[UREG_FP] + STACK_BIAS);
 
@@ -2259,8 +2239,8 @@ void die_if_kernel(char *str, struct pt_regs *regs)
 		 * find some badly aligned kernel stack.
 		 */
 		while (rw &&
-		       count++ < 30&&
-		       is_kernel_stack(current, rw)) {
+		       count++ < 30 &&
+		       kstack_valid(tp, (unsigned long) rw)) {
 			printk("Caller[%016lx]: %pS\n", rw->ins[7],
 			       (void *) rw->ins[7]);
 
diff --git a/arch/sparc/kernel/vmlinux.lds.S b/arch/sparc/kernel/vmlinux.lds.S
index 4e5992593967..0c1e6783657f 100644
--- a/arch/sparc/kernel/vmlinux.lds.S
+++ b/arch/sparc/kernel/vmlinux.lds.S
@@ -46,11 +46,16 @@ SECTIONS
 		SCHED_TEXT
 		LOCK_TEXT
 		KPROBES_TEXT
+		IRQENTRY_TEXT
 		*(.gnu.warning)
 	} = 0
 	_etext = .;
 
 	RO_DATA(PAGE_SIZE)
+
+	/* Start of data section */
+	_sdata = .;
+
 	.data1 : {
 		*(.data1)
 	}
diff --git a/arch/sparc/lib/mcount.S b/arch/sparc/lib/mcount.S
index 24b8b12deed2..3753e3c6e176 100644
--- a/arch/sparc/lib/mcount.S
+++ b/arch/sparc/lib/mcount.S
@@ -7,26 +7,11 @@
 
 #include <linux/linkage.h>
 
-#include <asm/ptrace.h>
-#include <asm/thread_info.h>
-
 /*
  * This is the main variant and is called by C code.  GCC's -pg option
  * automatically instruments every C function with a call to this.
  */
 
-#ifdef CONFIG_STACK_DEBUG
-
-#define OVSTACKSIZE	4096		/* lets hope this is enough */
-
-	.data
-	.align		8
-panicstring:
-	.asciz		"Stack overflow\n"
-	.align		8
-ovstack:
-	.skip		OVSTACKSIZE
-#endif
 	.text
 	.align		32
 	.globl		_mcount
@@ -35,84 +20,48 @@ ovstack:
 	.type		mcount,#function
 _mcount:
 mcount:
-#ifdef CONFIG_STACK_DEBUG
-	/*
-	 * Check whether %sp is dangerously low.
-	 */
-	ldub		[%g6 + TI_FPDEPTH], %g1
-	srl		%g1, 1, %g3
-	add		%g3, 1, %g3
-	sllx		%g3, 8, %g3			! each fpregs frame is 256b
-	add		%g3, 192, %g3
-	add		%g6, %g3, %g3			! where does task_struct+frame end?
-	sub		%g3, STACK_BIAS, %g3
-	cmp		%sp, %g3
-	bg,pt		%xcc, 1f
-	 nop
-	lduh		[%g6 + TI_CPU], %g1
-	sethi		%hi(hardirq_stack), %g3
-	or		%g3, %lo(hardirq_stack), %g3
-	sllx		%g1, 3, %g1
-	ldx		[%g3 + %g1], %g7
-	sub		%g7, STACK_BIAS, %g7
-	cmp		%sp, %g7
-	bleu,pt		%xcc, 2f
-	 sethi		%hi(THREAD_SIZE), %g3
-	add		%g7, %g3, %g7
-	cmp		%sp, %g7
-	blu,pn		%xcc, 1f
-2:	 sethi		%hi(softirq_stack), %g3
-	or		%g3, %lo(softirq_stack), %g3
-	ldx		[%g3 + %g1], %g7
-	sub		%g7, STACK_BIAS, %g7
-	cmp		%sp, %g7
-	bleu,pt		%xcc, 3f
-	 sethi		%hi(THREAD_SIZE), %g3
-	add		%g7, %g3, %g7
-	cmp		%sp, %g7
-	blu,pn		%xcc, 1f
-	 nop
-	/* If we are already on ovstack, don't hop onto it
-	 * again, we are already trying to output the stack overflow
-	 * message.
-	 */
-3:	sethi		%hi(ovstack), %g7		! cant move to panic stack fast enough
-	 or		%g7, %lo(ovstack), %g7
-	add		%g7, OVSTACKSIZE, %g3
-	sub		%g3, STACK_BIAS + 192, %g3
-	sub		%g7, STACK_BIAS, %g7
-	cmp		%sp, %g7
-	blu,pn		%xcc, 2f
-	 cmp		%sp, %g3
-	bleu,pn		%xcc, 1f
-	 nop
-2:	mov		%g3, %sp
-	sethi		%hi(panicstring), %g3
-	call		prom_printf
-	 or		%g3, %lo(panicstring), %o0
-	call		prom_halt
-	 nop
-1:
-#endif
 #ifdef CONFIG_FUNCTION_TRACER
 #ifdef CONFIG_DYNAMIC_FTRACE
-	mov		%o7, %o0
-	.globl		mcount_call
-mcount_call:
-	call		ftrace_stub
-	 mov		%o0, %o7
+	/* Do nothing, the retl/nop below is all we need.  */
 #else
-	sethi		%hi(ftrace_trace_function), %g1
+	sethi		%hi(function_trace_stop), %g1
+	lduw		[%g1 + %lo(function_trace_stop)], %g2
+	brnz,pn		%g2, 2f
+	 sethi		%hi(ftrace_trace_function), %g1
 	sethi		%hi(ftrace_stub), %g2
 	ldx		[%g1 + %lo(ftrace_trace_function)], %g1
 	or		%g2, %lo(ftrace_stub), %g2
 	cmp		%g1, %g2
 	be,pn		%icc, 1f
-	 mov		%i7, %o1
-	jmpl		%g1, %g0
-	 mov		%o7, %o0
+	 mov		%i7, %g3
+	save		%sp, -128, %sp
+	mov		%g3, %o1
+	jmpl		%g1, %o7
+	 mov		%i7, %o0
+	ret
+	 restore
 	/* not reached */
 1:
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+	sethi		%hi(ftrace_graph_return), %g1
+	ldx		[%g1 + %lo(ftrace_graph_return)], %g3
+	cmp		%g2, %g3
+	bne,pn		%xcc, 5f
+	 sethi		%hi(ftrace_graph_entry_stub), %g2
+	sethi		%hi(ftrace_graph_entry), %g1
+	or		%g2, %lo(ftrace_graph_entry_stub), %g2
+	ldx		[%g1 + %lo(ftrace_graph_entry)], %g1
+	cmp		%g1, %g2
+	be,pt		%xcc, 2f
+	 nop
+5:	mov		%i7, %g2
+	mov		%fp, %g3
+	save		%sp, -128, %sp
+	mov		%g2, %l0
+	ba,pt		%xcc, ftrace_graph_caller
+	 mov		%g3, %l1
+#endif
+2:
 #endif
 #endif
 	retl
@@ -131,14 +80,50 @@ ftrace_stub:
 	.globl		ftrace_caller
 	.type		ftrace_caller,#function
 ftrace_caller:
-	mov		%i7, %o1
-	mov		%o7, %o0
+	sethi		%hi(function_trace_stop), %g1
+	mov		%i7, %g2
+	lduw		[%g1 + %lo(function_trace_stop)], %g1
+	brnz,pn		%g1, ftrace_stub
+	 mov		%fp, %g3
+	save		%sp, -128, %sp
+	mov		%g2, %o1
+	mov		%g2, %l0
+	mov		%g3, %l1
 	.globl		ftrace_call
 ftrace_call:
 	call		ftrace_stub
-	 mov		%o0, %o7
-	retl
+	 mov		%i7, %o0
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+	.globl		ftrace_graph_call
+ftrace_graph_call:
+	call		ftrace_stub
 	 nop
+#endif
+	ret
+	 restore
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+	.size		ftrace_graph_call,.-ftrace_graph_call
+#endif
+	.size		ftrace_call,.-ftrace_call
 	.size		ftrace_caller,.-ftrace_caller
 #endif
 #endif
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+ENTRY(ftrace_graph_caller)
+	mov		%l0, %o0
+	mov		%i7, %o1
+	call		prepare_ftrace_return
+	 mov		%l1, %o2
+	ret
+	 restore	%o0, -8, %i7
+END(ftrace_graph_caller)
+
+ENTRY(return_to_handler)
+	save		%sp, -128, %sp
+	call		ftrace_return_to_handler
+	 mov		%fp, %o0
+	jmpl		%o0 + 8, %g0
+	 restore
+END(return_to_handler)
+#endif
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index ff017108700d..935248bdbc47 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -356,7 +356,7 @@ config SLUB_STATS
 config DEBUG_KMEMLEAK
 	bool "Kernel memory leak detector"
 	depends on DEBUG_KERNEL && EXPERIMENTAL && !MEMORY_HOTPLUG && \
-		(X86 || ARM || PPC || S390 || SUPERH || MICROBLAZE)
+		(X86 || ARM || PPC || S390 || SPARC64 || SUPERH || MICROBLAZE)
 
 	select DEBUG_FS if SYSFS
 	select STACKTRACE if STACKTRACE_SUPPORT