25 files changed, 210 insertions, 278 deletions
diff --git a/arch/arc/boot/dts/angel4.dts b/arch/arc/boot/dts/angel4.dts
index bcf662d21a57..5bb2fdaca02f 100644
--- a/arch/arc/boot/dts/angel4.dts
+++ b/arch/arc/boot/dts/angel4.dts
@@ -17,7 +17,7 @@
 	interrupt-parent = <&intc>;
 
 	chosen {
-		bootargs = "console=ttyARC0,115200n8";
+		bootargs = "console=ttyARC0,115200n8 earlyprintk=ttyARC0";
 	};
 
 	aliases {
diff --git a/arch/arc/include/asm/atomic.h b/arch/arc/include/asm/atomic.h
index 03e494f695d1..83f03ca6caf6 100644
--- a/arch/arc/include/asm/atomic.h
+++ b/arch/arc/include/asm/atomic.h
@@ -190,11 +190,6 @@ static inline void atomic_clear_mask(unsigned long mask, unsigned long *addr)
 
 #endif /* !CONFIG_ARC_HAS_LLSC */
 
-#define smp_mb__before_atomic_dec()	barrier()
-#define smp_mb__after_atomic_dec()	barrier()
-#define smp_mb__before_atomic_inc()	barrier()
-#define smp_mb__after_atomic_inc()	barrier()
-
 /**
  * __atomic_add_unless - add unless the number is a given value
  * @v: pointer of type atomic_t
diff --git a/arch/arc/include/asm/bitops.h b/arch/arc/include/asm/bitops.h
index 647a83a8e756..ebc0cf3164dc 100644
--- a/arch/arc/include/asm/bitops.h
+++ b/arch/arc/include/asm/bitops.h
@@ -19,6 +19,7 @@
 
 #include <linux/types.h>
 #include <linux/compiler.h>
+#include <asm/barrier.h>
 
 /*
  * Hardware assisted read-modify-write using ARC700 LLOCK/SCOND insns.
@@ -496,10 +497,6 @@ static inline __attribute__ ((const)) int __ffs(unsigned long word)
  */
 #define ffz(x)	__ffs(~(x))
 
-/* TODO does this affect uni-processor code */
-#define smp_mb__before_clear_bit()  barrier()
-#define smp_mb__after_clear_bit()   barrier()
-
 #include <asm-generic/bitops/hweight.h>
 #include <asm-generic/bitops/fls64.h>
 #include <asm-generic/bitops/sched.h>
diff --git a/arch/arc/include/asm/cache.h b/arch/arc/include/asm/cache.h
index 2fd3162ec4df..b3c750979aa1 100644
--- a/arch/arc/include/asm/cache.h
+++ b/arch/arc/include/asm/cache.h
@@ -55,4 +55,31 @@ extern void read_decode_cache_bcr(void);
 
 #endif	/* !__ASSEMBLY__ */
 
+/* Instruction cache related Auxiliary registers */
+#define ARC_REG_IC_BCR		0x77	/* Build Config reg */
+#define ARC_REG_IC_IVIC		0x10
+#define ARC_REG_IC_CTRL		0x11
+#define ARC_REG_IC_IVIL		0x19
+#if defined(CONFIG_ARC_MMU_V3)
+#define ARC_REG_IC_PTAG		0x1E
+#endif
+
+/* Bit val in IC_CTRL */
+#define IC_CTRL_CACHE_DISABLE   0x1
+
+/* Data cache related Auxiliary registers */
+#define ARC_REG_DC_BCR		0x72	/* Build Config reg */
+#define ARC_REG_DC_IVDC		0x47
+#define ARC_REG_DC_CTRL		0x48
+#define ARC_REG_DC_IVDL		0x4A
+#define ARC_REG_DC_FLSH		0x4B
+#define ARC_REG_DC_FLDL		0x4C
+#if defined(CONFIG_ARC_MMU_V3)
+#define ARC_REG_DC_PTAG		0x5C
+#endif
+
+/* Bit val in DC_CTRL */
+#define DC_CTRL_INV_MODE_FLUSH  0x40
+#define DC_CTRL_FLUSH_STATUS    0x100
+
 #endif /* _ASM_CACHE_H */
diff --git a/arch/arc/include/asm/irq.h b/arch/arc/include/asm/irq.h
index 291a70db68b8..fb4efb648971 100644
--- a/arch/arc/include/asm/irq.h
+++ b/arch/arc/include/asm/irq.h
@@ -19,8 +19,6 @@
 #include <asm-generic/irq.h>
 
 extern void arc_init_IRQ(void);
-extern int get_hw_config_num_irq(void);
-
-void arc_local_timer_setup(unsigned int cpu);
+void arc_local_timer_setup(void);
 
 #endif
diff --git a/arch/arc/include/asm/processor.h b/arch/arc/include/asm/processor.h
index 15334ab66b56..d99f9b37cd15 100644
--- a/arch/arc/include/asm/processor.h
+++ b/arch/arc/include/asm/processor.h
@@ -18,7 +18,6 @@
 
 #ifndef __ASSEMBLY__
 
-#include <asm/arcregs.h>	/* for STATUS_E1_MASK et all */
 #include <asm/ptrace.h>
 
 /* Arch specific stuff which needs to be saved per task.
@@ -41,15 +40,13 @@ struct thread_struct {
 /* Forward declaration, a strange C thing */
 struct task_struct;
 
-/*
- * Return saved PC of a blocked thread.
- */
+/* Return saved PC of a blocked thread  */
 unsigned long thread_saved_pc(struct task_struct *t);
 
 #define task_pt_regs(p) \
 	((struct pt_regs *)(THREAD_SIZE + (void *)task_stack_page(p)) - 1)
 
-/* Free all resources held by a thread. */
+/* Free all resources held by a thread */
 #define release_thread(thread) do { } while (0)
 
 /* Prepare to copy thread state - unlazy all lazy status */
@@ -82,26 +79,8 @@ unsigned long thread_saved_pc(struct task_struct *t);
 #define KSTK_BLINK(tsk) KSTK_REG(tsk, 4)
 #define KSTK_FP(tsk)    KSTK_REG(tsk, 0)
 
-/*
- * Do necessary setup to start up a newly executed thread.
- *
- * E1,E2 so that Interrupts are enabled in user mode
- * L set, so Loop inhibited to begin with
- * lp_start and lp_end seeded with bogus non-zero values so to easily catch
- * the ARC700 sr to lp_start hardware bug
- */
-#define start_thread(_regs, _pc, _usp)				\
-do {								\
-	set_fs(USER_DS); /* reads from user space */		\
-	(_regs)->ret = (_pc);					\
-	/* Interrupts enabled in User Mode */			\
-	(_regs)->status32 = STATUS_U_MASK | STATUS_L_MASK	\
-		| STATUS_E1_MASK | STATUS_E2_MASK;		\
-	(_regs)->sp = (_usp);					\
-	/* bogus seed values for debugging */			\
-	(_regs)->lp_start = 0x10;				\
-	(_regs)->lp_end = 0x80;					\
-} while (0)
+extern void start_thread(struct pt_regs * regs, unsigned long pc,
+			 unsigned long usp);
 
 extern unsigned int get_wchan(struct task_struct *p);
 
diff --git a/arch/arc/include/asm/sections.h b/arch/arc/include/asm/sections.h
index 764f1e3ba752..09db952e14bd 100644
--- a/arch/arc/include/asm/sections.h
+++ b/arch/arc/include/asm/sections.h
@@ -12,6 +12,5 @@
 #include <asm-generic/sections.h>
 
 extern char __arc_dccm_base[];
-extern char __dtb_start[];
 
 #endif
diff --git a/arch/arc/include/uapi/asm/Kbuild b/arch/arc/include/uapi/asm/Kbuild
index 18fefaea73fd..f50d02df78d5 100644
--- a/arch/arc/include/uapi/asm/Kbuild
+++ b/arch/arc/include/uapi/asm/Kbuild
@@ -2,11 +2,4 @@
 include include/uapi/asm-generic/Kbuild.asm
 header-y += elf.h
 header-y += page.h
-header-y += setup.h
-header-y += byteorder.h
 header-y += cachectl.h
-header-y += ptrace.h
-header-y += sigcontext.h
-header-y += signal.h
-header-y += swab.h
-header-y += unistd.h
diff --git a/arch/arc/include/uapi/asm/ptrace.h b/arch/arc/include/uapi/asm/ptrace.h
index 2618cc13ba75..76a7739aab1c 100644
--- a/arch/arc/include/uapi/asm/ptrace.h
+++ b/arch/arc/include/uapi/asm/ptrace.h
@@ -11,6 +11,7 @@
 #ifndef _UAPI__ASM_ARC_PTRACE_H
 #define _UAPI__ASM_ARC_PTRACE_H
 
+#define PTRACE_GET_THREAD_AREA	25
 
 #ifndef __ASSEMBLY__
 /*
diff --git a/arch/arc/kernel/ctx_sw_asm.S b/arch/arc/kernel/ctx_sw_asm.S
index 2ff0347a2fd7..e248594097e7 100644
--- a/arch/arc/kernel/ctx_sw_asm.S
+++ b/arch/arc/kernel/ctx_sw_asm.S
@@ -10,9 +10,9 @@
  *  -This is the more "natural" hand written assembler
  */
 
+#include <linux/linkage.h>
 #include <asm/entry.h>       /* For the SAVE_* macros */
 #include <asm/asm-offsets.h>
-#include <asm/linkage.h>
 
 #define KSP_WORD_OFF 	((TASK_THREAD + THREAD_KSP) / 4)
 
diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index b6dc4e21fd32..fffdb5e41b20 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -41,8 +41,8 @@ const struct machine_desc * __init setup_machine_fdt(void *dt)
 {
 	const struct machine_desc *mdesc;
 	unsigned long dt_root;
-	void *clk;
-	unsigned long len;
+	const void *clk;
+	int len;
 
 	if (!early_init_dt_scan(dt))
 		return NULL;
diff --git a/arch/arc/kernel/entry.S b/arch/arc/kernel/entry.S
index 29b82adbf0b4..83a046a7cd06 100644
--- a/arch/arc/kernel/entry.S
+++ b/arch/arc/kernel/entry.S
@@ -156,7 +156,7 @@ ARCFP_DATA int1_saved_reg
 int1_saved_reg:
 	.zero 4
 
-/* Each Interrupt level needs it's own scratch */
+/* Each Interrupt level needs its own scratch */
 #ifdef CONFIG_ARC_COMPACT_IRQ_LEVELS
 
 ARCFP_DATA int2_saved_reg
@@ -473,7 +473,7 @@ trap_with_param:
 	lr  r0, [efa]
 	mov r1, sp
 
-	; Now that we have read EFA, its safe to do "fake" rtie
+	; Now that we have read EFA, it is safe to do "fake" rtie
 	;   and get out of CPU exception mode
 	FAKE_RET_FROM_EXCPN r11
 
@@ -678,9 +678,9 @@ not_exception:
 	brne r9, event_IRQ2, 149f
 
 	;------------------------------------------------------------------
-	; if L2 IRQ interrupted a L1 ISR,  we'd disbaled preemption earlier
-	; so that sched doesnt move to new task, causing L1 to be delayed
-	; undeterministically. Now that we've achieved that, lets reset
+	; if L2 IRQ interrupted an L1 ISR,  we'd disabled preemption earlier
+	; so that sched doesn't move to new task, causing L1 to be delayed
+	; undeterministically. Now that we've achieved that, let's reset
 	; things to what they were, before returning from L2 context
 	;----------------------------------------------------------------
 
@@ -736,7 +736,7 @@ ENTRY(ret_from_fork)
 	; put last task in scheduler queue
 	bl   @schedule_tail
 
-	; If kernel thread, jump to it's entry-point
+	; If kernel thread, jump to its entry-point
 	ld   r9, [sp, PT_status32]
 	brne r9, 0, 1f
 
diff --git a/arch/arc/kernel/head.S b/arch/arc/kernel/head.S
index 4ad04915dc6b..4d2481bd8b98 100644
--- a/arch/arc/kernel/head.S
+++ b/arch/arc/kernel/head.S
@@ -12,10 +12,42 @@
  *      to skip certain things during boot on simulator
  */
 
+#include <linux/linkage.h>
 #include <asm/asm-offsets.h>
 #include <asm/entry.h>
-#include <linux/linkage.h>
 #include <asm/arcregs.h>
+#include <asm/cache.h>
+
+.macro CPU_EARLY_SETUP
+
+	; Setting up Vectror Table (in case exception happens in early boot
+	sr	@_int_vec_base_lds, [AUX_INTR_VEC_BASE]
+
+	; Disable I-cache/D-cache if kernel so configured
+	lr	r5, [ARC_REG_IC_BCR]
+	breq    r5, 0, 1f		; I$ doesn't exist
+	lr	r5, [ARC_REG_IC_CTRL]
+#ifdef CONFIG_ARC_HAS_ICACHE
+	bclr	r5, r5, 0		; 0 - Enable, 1 is Disable
+#else
+	bset	r5, r5, 0		; I$ exists, but is not used
+#endif
+	sr	r5, [ARC_REG_IC_CTRL]
+
+1:
+	lr	r5, [ARC_REG_DC_BCR]
+	breq    r5, 0, 1f		; D$ doesn't exist
+	lr	r5, [ARC_REG_DC_CTRL]
+	bclr	r5, r5, 6		; Invalidate (discard w/o wback)
+#ifdef CONFIG_ARC_HAS_DCACHE
+	bclr	r5, r5, 0		; Enable (+Inv)
+#else
+	bset	r5, r5, 0		; Disable (+Inv)
+#endif
+	sr	r5, [ARC_REG_DC_CTRL]
+
+1:
+.endm
 
 	.cpu A7
 
@@ -27,7 +59,7 @@ stext:
 	; Don't clobber r0-r2 yet. It might have bootloader provided info
 	;-------------------------------------------------------------------
 
-	sr	@_int_vec_base_lds, [AUX_INTR_VEC_BASE]
+	CPU_EARLY_SETUP
 
 #ifdef CONFIG_SMP
 	; Ensure Boot (Master) proceeds. Others wait in platform dependent way
@@ -45,10 +77,11 @@ stext:
 	; Clear BSS before updating any globals
 	; XXX: use ZOL here
 	mov	r5, __bss_start
-	mov	r6, __bss_stop
+	sub	r6, __bss_stop, r5
+	lsr.f	lp_count, r6, 2
+	lpnz	1f
+	st.ab   0, [r5, 4]
 1:
-	st.ab   0, [r5,4]
-	brlt    r5, r6, 1b
 
 	; Uboot - kernel ABI
 	;    r0 = [0] No uboot interaction, [1] cmdline in r2, [2] DTB in r2
@@ -90,7 +123,7 @@ stext:
 
 first_lines_of_secondary:
 
-	sr	@_int_vec_base_lds, [AUX_INTR_VEC_BASE]
+	CPU_EARLY_SETUP
 
 	; setup per-cpu idle task as "current" on this CPU
 	ld	r0, [@secondary_idle_tsk]
diff --git a/arch/arc/kernel/irq.c b/arch/arc/kernel/irq.c
index a4b141ee9a6a..7d653c0d0773 100644
--- a/arch/arc/kernel/irq.c
+++ b/arch/arc/kernel/irq.c
@@ -150,24 +150,6 @@ void arch_do_IRQ(unsigned int irq, struct pt_regs *regs)
 	set_irq_regs(old_regs);
 }
 
-int get_hw_config_num_irq(void)
-{
-	uint32_t val = read_aux_reg(ARC_REG_VECBASE_BCR);
-
-	switch (val & 0x03) {
-	case 0:
-		return 16;
-	case 1:
-		return 32;
-	case 2:
-		return 8;
-	default:
-		return 0;
-	}
-
-	return 0;
-}
-
 /*
  * arch_local_irq_enable - Enable interrupts.
  *
diff --git a/arch/arc/kernel/process.c b/arch/arc/kernel/process.c
index 07a3a968fe49..fdd89715d2d3 100644
--- a/arch/arc/kernel/process.c
+++ b/arch/arc/kernel/process.c
@@ -151,6 +151,29 @@ int copy_thread(unsigned long clone_flags,
 }
 
 /*
+ * Do necessary setup to start up a new user task
+ */
+void start_thread(struct pt_regs * regs, unsigned long pc, unsigned long usp)
+{
+	set_fs(USER_DS); /* user space */
+
+	regs->sp = usp;
+	regs->ret = pc;
+
+	/*
+	 * [U]ser Mode bit set
+	 * [L] ZOL loop inhibited to begin with - cleared by a LP insn
+	 * Interrupts enabled
+	 */
+	regs->status32 = STATUS_U_MASK | STATUS_L_MASK |
+			 STATUS_E1_MASK | STATUS_E2_MASK;
+
+	/* bogus seed values for debugging */
+	regs->lp_start = 0x10;
+	regs->lp_end = 0x80;
+}
+
+/*
  * Some archs flush debug and FPU info here
  */
 void flush_thread(void)
diff --git a/arch/arc/kernel/ptrace.c b/arch/arc/kernel/ptrace.c
index 5d76706139dd..13b3ffb27a38 100644
--- a/arch/arc/kernel/ptrace.c
+++ b/arch/arc/kernel/ptrace.c
@@ -146,6 +146,10 @@ long arch_ptrace(struct task_struct *child, long request,
 	pr_debug("REQ=%ld: ADDR =0x%lx, DATA=0x%lx)\n", request, addr, data);
 
 	switch (request) {
+	case PTRACE_GET_THREAD_AREA:
+		ret = put_user(task_thread_info(child)->thr_ptr,
+			       (unsigned long __user *)data);
+		break;
 	default:
 		ret = ptrace_request(child, request, addr, data);
 		break;
diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c
index 40859e5619f9..c802bb500602 100644
--- a/arch/arc/kernel/smp.c
+++ b/arch/arc/kernel/smp.c
@@ -138,7 +138,7 @@ void start_kernel_secondary(void)
 	if (machine_desc->init_smp)
 		machine_desc->init_smp(smp_processor_id());
 
-	arc_local_timer_setup(cpu);
+	arc_local_timer_setup();
 
 	local_irq_enable();
 	preempt_disable();
@@ -337,8 +337,19 @@ irqreturn_t do_IPI(int irq, void *dev_id)
  * API called by platform code to hookup arch-common ISR to their IPI IRQ
  */
 static DEFINE_PER_CPU(int, ipi_dev);
+
+static struct irqaction arc_ipi_irq = {
+        .name    = "IPI Interrupt",
+        .flags   = IRQF_PERCPU,
+        .handler = do_IPI,
+};
+
 int smp_ipi_irq_setup(int cpu, int irq)
 {
-	int *dev_id = &per_cpu(ipi_dev, smp_processor_id());
-	return request_percpu_irq(irq, do_IPI, "IPI Interrupt", dev_id);
+	if (!cpu)
+		return setup_irq(irq, &arc_ipi_irq);
+	else
+		arch_unmask_irq(irq);
+
+	return 0;
 }
diff --git a/arch/arc/kernel/time.c b/arch/arc/kernel/time.c
index 71c42521c77f..36c2aa99436f 100644
--- a/arch/arc/kernel/time.c
+++ b/arch/arc/kernel/time.c
@@ -219,12 +219,13 @@ static struct irqaction arc_timer_irq = {
 /*
  * Setup the local event timer for @cpu
  */
-void arc_local_timer_setup(unsigned int cpu)
+void arc_local_timer_setup()
 {
-	struct clock_event_device *clk = &per_cpu(arc_clockevent_device, cpu);
+	struct clock_event_device *evt = this_cpu_ptr(&arc_clockevent_device);
+	int cpu = smp_processor_id();
 
-	clk->cpumask = cpumask_of(cpu);
-	clockevents_config_and_register(clk, arc_get_core_freq(),
+	evt->cpumask = cpumask_of(cpu);
+	clockevents_config_and_register(evt, arc_get_core_freq(),
 					0, ARC_TIMER_MAX);
 
 	/*
@@ -261,7 +262,7 @@ void __init time_init(void)
 		clocksource_register_hz(&arc_counter, arc_get_core_freq());
 
 	/* sets up the periodic event timer */
-	arc_local_timer_setup(smp_processor_id());
+	arc_local_timer_setup();
 
 	if (machine_desc->init_time)
 		machine_desc->init_time();
diff --git a/arch/arc/kernel/troubleshoot.c b/arch/arc/kernel/troubleshoot.c
index 73a7450ee622..1badf9b84b51 100644
--- a/arch/arc/kernel/troubleshoot.c
+++ b/arch/arc/kernel/troubleshoot.c
@@ -86,12 +86,13 @@ static void show_faulting_vma(unsigned long address, char *buf)
 	unsigned long ino = 0;
 	dev_t dev = 0;
 	char *nm = buf;
+	struct mm_struct *active_mm = current->active_mm;
 
 	/* can't use print_vma_addr() yet as it doesn't check for
 	 * non-inclusive vma
 	 */
-
-	vma = find_vma(current->active_mm, address);
+	down_read(&active_mm->mmap_sem);
+	vma = find_vma(active_mm, address);
 
 	/* check against the find_vma( ) behaviour which returns the next VMA
 	 * if the container VMA is not found
@@ -110,9 +111,10 @@ static void show_faulting_vma(unsigned long address, char *buf)
 			vma->vm_start < TASK_UNMAPPED_BASE ?
 				address : address - vma->vm_start,
 			nm, vma->vm_start, vma->vm_end);
-	} else {
+	} else
 		pr_info("    @No matching VMA found\n");
-	}
+
+	up_read(&active_mm->mmap_sem);
 }
 
 static void show_ecr_verbose(struct pt_regs *regs)
diff --git a/arch/arc/kernel/vmlinux.lds.S b/arch/arc/kernel/vmlinux.lds.S
index 2555f5886af6..dd35bde39f69 100644
--- a/arch/arc/kernel/vmlinux.lds.S
+++ b/arch/arc/kernel/vmlinux.lds.S
@@ -116,7 +116,7 @@ SECTIONS
 
 	_edata = .;
 
-	BSS_SECTION(0, 0, 0)
+	BSS_SECTION(4, 4, 4)
 
 #ifdef CONFIG_ARC_DW2_UNWIND
 	. = ALIGN(PAGE_SIZE);
diff --git a/arch/arc/mm/cache_arc700.c b/arch/arc/mm/cache_arc700.c
index 89edf7961a2f..353b202c37c9 100644
--- a/arch/arc/mm/cache_arc700.c
+++ b/arch/arc/mm/cache_arc700.c
@@ -73,33 +73,6 @@
 #include <asm/cachectl.h>
 #include <asm/setup.h>
 
-/* Instruction cache related Auxiliary registers */
-#define ARC_REG_IC_BCR		0x77	/* Build Config reg */
-#define ARC_REG_IC_IVIC		0x10
-#define ARC_REG_IC_CTRL		0x11
-#define ARC_REG_IC_IVIL		0x19
-#if (CONFIG_ARC_MMU_VER > 2)
-#define ARC_REG_IC_PTAG		0x1E
-#endif
-
-/* Bit val in IC_CTRL */
-#define IC_CTRL_CACHE_DISABLE   0x1
-
-/* Data cache related Auxiliary registers */
-#define ARC_REG_DC_BCR		0x72	/* Build Config reg */
-#define ARC_REG_DC_IVDC		0x47
-#define ARC_REG_DC_CTRL		0x48
-#define ARC_REG_DC_IVDL		0x4A
-#define ARC_REG_DC_FLSH		0x4B
-#define ARC_REG_DC_FLDL		0x4C
-#if (CONFIG_ARC_MMU_VER > 2)
-#define ARC_REG_DC_PTAG		0x5C
-#endif
-
-/* Bit val in DC_CTRL */
-#define DC_CTRL_INV_MODE_FLUSH  0x40
-#define DC_CTRL_FLUSH_STATUS    0x100
-
 char *arc_cache_mumbojumbo(int c, char *buf, int len)
 {
 	int n = 0;
@@ -168,72 +141,43 @@ void read_decode_cache_bcr(void)
  */
 void arc_cache_init(void)
 {
-	unsigned int cpu = smp_processor_id();
-	struct cpuinfo_arc_cache *ic = &cpuinfo_arc700[cpu].icache;
-	struct cpuinfo_arc_cache *dc = &cpuinfo_arc700[cpu].dcache;
-	unsigned int dcache_does_alias, temp;
+	unsigned int __maybe_unused cpu = smp_processor_id();
+	struct cpuinfo_arc_cache __maybe_unused *ic, __maybe_unused *dc;
 	char str[256];
 
 	printk(arc_cache_mumbojumbo(0, str, sizeof(str)));
 
-	if (!ic->ver)
-		goto chk_dc;
-
 #ifdef CONFIG_ARC_HAS_ICACHE
-	/* 1. Confirm some of I-cache params which Linux assumes */
-	if (ic->line_len != L1_CACHE_BYTES)
-		panic("Cache H/W doesn't match kernel Config");
-
-	if (ic->ver != CONFIG_ARC_MMU_VER)
-		panic("Cache ver doesn't match MMU ver\n");
-#endif
-
-	/* Enable/disable I-Cache */
-	temp = read_aux_reg(ARC_REG_IC_CTRL);
-
-#ifdef CONFIG_ARC_HAS_ICACHE
-	temp &= ~IC_CTRL_CACHE_DISABLE;
-#else
-	temp |= IC_CTRL_CACHE_DISABLE;
+	ic = &cpuinfo_arc700[cpu].icache;
+	if (ic->ver) {
+		if (ic->line_len != L1_CACHE_BYTES)
+			panic("ICache line [%d] != kernel Config [%d]",
+			      ic->line_len, L1_CACHE_BYTES);
+
+		if (ic->ver != CONFIG_ARC_MMU_VER)
+			panic("Cache ver [%d] doesn't match MMU ver [%d]\n",
+			      ic->ver, CONFIG_ARC_MMU_VER);
+	}
 #endif
 
-	write_aux_reg(ARC_REG_IC_CTRL, temp);
-
-chk_dc:
-	if (!dc->ver)
-		return;
-
 #ifdef CONFIG_ARC_HAS_DCACHE
-	if (dc->line_len != L1_CACHE_BYTES)
-		panic("Cache H/W doesn't match kernel Config");
+	dc = &cpuinfo_arc700[cpu].dcache;
+	if (dc->ver) {
+		unsigned int dcache_does_alias;
 
-	/* check for D-Cache aliasing */
-	dcache_does_alias = (dc->sz / dc->assoc) > PAGE_SIZE;
+		if (dc->line_len != L1_CACHE_BYTES)
+			panic("DCache line [%d] != kernel Config [%d]",
+			      dc->line_len, L1_CACHE_BYTES);
 
-	if (dcache_does_alias && !cache_is_vipt_aliasing())
-		panic("Enable CONFIG_ARC_CACHE_VIPT_ALIASING\n");
-	else if (!dcache_does_alias && cache_is_vipt_aliasing())
-		panic("Don't need CONFIG_ARC_CACHE_VIPT_ALIASING\n");
-#endif
-
-	/* Set the default Invalidate Mode to "simpy discard dirty lines"
-	 *  as this is more frequent then flush before invalidate
-	 * Ofcourse we toggle this default behviour when desired
-	 */
-	temp = read_aux_reg(ARC_REG_DC_CTRL);
-	temp &= ~DC_CTRL_INV_MODE_FLUSH;
+		/* check for D-Cache aliasing */
+		dcache_does_alias = (dc->sz / dc->assoc) > PAGE_SIZE;
 
-#ifdef CONFIG_ARC_HAS_DCACHE
-	/* Enable D-Cache: Clear Bit 0 */
-	write_aux_reg(ARC_REG_DC_CTRL, temp & ~IC_CTRL_CACHE_DISABLE);
-#else
-	/* Flush D cache */
-	write_aux_reg(ARC_REG_DC_FLSH, 0x1);
-	/* Disable D cache */
-	write_aux_reg(ARC_REG_DC_CTRL, temp | IC_CTRL_CACHE_DISABLE);
+		if (dcache_does_alias && !cache_is_vipt_aliasing())
+			panic("Enable CONFIG_ARC_CACHE_VIPT_ALIASING\n");
+		else if (!dcache_does_alias && cache_is_vipt_aliasing())
+			panic("Don't need CONFIG_ARC_CACHE_VIPT_ALIASING\n");
+	}
 #endif
-
-	return;
 }
 
 #define OP_INV		0x1
@@ -253,12 +197,16 @@ static inline void __cache_line_loop(unsigned long paddr, unsigned long vaddr,
 
 	if (cacheop == OP_INV_IC) {
 		aux_cmd = ARC_REG_IC_IVIL;
+#if (CONFIG_ARC_MMU_VER > 2)
 		aux_tag = ARC_REG_IC_PTAG;
+#endif
 	}
 	else {
 		/* d$ cmd: INV (discard or wback-n-discard) OR FLUSH (wback) */
 		aux_cmd = cacheop & OP_INV ? ARC_REG_DC_IVDL : ARC_REG_DC_FLDL;
+#if (CONFIG_ARC_MMU_VER > 2)
 		aux_tag = ARC_REG_DC_PTAG;
+#endif
 	}
 
 	/* Ensure we properly floor/ceil the non-line aligned/sized requests
@@ -441,7 +389,7 @@ static inline void __dc_line_op(unsigned long paddr, unsigned long vaddr,
 /***********************************************************
  * Machine specific helper for per line I-Cache invalidate.
  */
-static void __ic_line_inv_vaddr(unsigned long paddr, unsigned long vaddr,
+static void __ic_line_inv_vaddr_local(unsigned long paddr, unsigned long vaddr,
 				unsigned long sz)
 {
 	unsigned long flags;
@@ -457,6 +405,23 @@ static inline void __ic_entire_inv(void)
 	read_aux_reg(ARC_REG_IC_CTRL);	/* blocks */
 }
 
+struct ic_line_inv_vaddr_ipi {
+	unsigned long paddr, vaddr;
+	int sz;
+};
+
+static void __ic_line_inv_vaddr_helper(void *info)
+{
+        struct ic_line_inv_vaddr_ipi *ic_inv = (struct ic_line_inv_vaddr_ipi*) info;
+        __ic_line_inv_vaddr_local(ic_inv->paddr, ic_inv->vaddr, ic_inv->sz);
+}
+
+static void __ic_line_inv_vaddr(unsigned long paddr, unsigned long vaddr,
+				unsigned long sz)
+{
+	struct ic_line_inv_vaddr_ipi ic_inv = { paddr, vaddr , sz};
+	on_each_cpu(__ic_line_inv_vaddr_helper, &ic_inv, 1);
+}
 #else
 
 #define __ic_entire_inv()
@@ -605,12 +570,8 @@ void flush_icache_range(unsigned long kstart, unsigned long kend)
  */
 void __sync_icache_dcache(unsigned long paddr, unsigned long vaddr, int len)
 {
-	unsigned long flags;
-
-	local_irq_save(flags);
-	__ic_line_inv_vaddr(paddr, vaddr, len);
 	__dc_line_op(paddr, vaddr, len, OP_FLUSH_N_INV);
-	local_irq_restore(flags);
+	__ic_line_inv_vaddr(paddr, vaddr, len);
 }
 
 /* wrapper to compile time eliminate alignment checks in flush loop */
diff --git a/arch/arc/plat-arcfpga/Kconfig b/arch/arc/plat-arcfpga/Kconfig
index 33058aa40e77..e27bb5cc3c1e 100644
--- a/arch/arc/plat-arcfpga/Kconfig
+++ b/arch/arc/plat-arcfpga/Kconfig
@@ -48,36 +48,4 @@ config ARC_SERIAL_BAUD
 	help
 	  Baud rate for the ARC UART
 
-menuconfig ARC_HAS_BVCI_LAT_UNIT
-	bool "BVCI Bus Latency Unit"
-	depends on ARC_BOARD_ML509 || ARC_BOARD_ANGEL4
-	help
-	  IP to add artificial latency to BVCI Bus Based FPGA builds.
-	  The default latency (even worst case) for FPGA is non-realistic
-	  (~10 SDRAM, ~5 SSRAM).
-
-config BVCI_LAT_UNITS
-	hex "Latency Unit(s) Bitmap"
-	default "0x0"
-	depends on ARC_HAS_BVCI_LAT_UNIT
-	help
-	  There are multiple Latency Units corresponding to the many
-	  interfaces of the system bus arbiter (both CPU side as well as
-	  the peripheral side).
-	  To add latency to ALL memory transaction, choose Unit 0, otherwise
-	  for finer grainer - interface wise latency, specify a bitmap (1 bit
-	  per unit) of all units. e.g. 1,2,12 will be 0x1003
-
-	  Unit  0 - System Arb and Mem Controller
-	  Unit  1 - I$ and System Bus
-	  Unit  2 - D$ and System Bus
-	  ..
-	  Unit 12 - IDE Disk controller and System Bus
-
-config BVCI_LAT_CYCLES
-	int "Latency Value in cycles"
-	range 0 63
-	default "30"
-	depends on ARC_HAS_BVCI_LAT_UNIT
-
 endif
diff --git a/arch/arc/plat-arcfpga/Makefile b/arch/arc/plat-arcfpga/Makefile
index a44e22ebc1b7..4d1bddc34b5b 100644
--- a/arch/arc/plat-arcfpga/Makefile
+++ b/arch/arc/plat-arcfpga/Makefile
@@ -9,4 +9,4 @@
 KBUILD_CFLAGS	+= -Iarch/arc/plat-arcfpga/include
 
 obj-y := platform.o irq.o
-obj-$(CONFIG_SMP)		+= smp.o
+obj-$(CONFIG_ISS_SMP_EXTN)		+= smp.o
diff --git a/arch/arc/plat-arcfpga/platform.c b/arch/arc/plat-arcfpga/platform.c
index 19b76b61f44b..61c7e5997387 100644
--- a/arch/arc/plat-arcfpga/platform.c
+++ b/arch/arc/plat-arcfpga/platform.c
@@ -22,59 +22,6 @@
 #include <plat/smp.h>
 #include <plat/irq.h>
 
-/*-----------------------BVCI Latency Unit -----------------------------*/
-
-#ifdef CONFIG_ARC_HAS_BVCI_LAT_UNIT
-
-int lat_cycles = CONFIG_BVCI_LAT_CYCLES;
-
-/* BVCI Bus Profiler: Latency Unit */
-static void __init setup_bvci_lat_unit(void)
-{
-#define MAX_BVCI_UNITS 12
-
-	unsigned int i;
-	unsigned int *base = (unsigned int *)BVCI_LAT_UNIT_BASE;
-	const unsigned long units_req = CONFIG_BVCI_LAT_UNITS;
-	const unsigned int REG_UNIT = 21;
-	const unsigned int REG_VAL = 22;
-
-	/*
-	 * There are multiple Latency Units corresponding to the many
-	 * interfaces of the system bus arbiter (both CPU side as well as
-	 * the peripheral side).
-	 *
-	 * Unit  0 - System Arb and Mem Controller - adds latency to all
-	 *	    memory trasactions
-	 * Unit  1 - I$ and System Bus
-	 * Unit  2 - D$ and System Bus
-	 * ..
-	 * Unit 12 - IDE Disk controller and System Bus
-	 *
-	 * The programmers model requires writing to lat_unit reg first
-	 * and then the latency value (cycles) to lat_value reg
-	 */
-
-	if (CONFIG_BVCI_LAT_UNITS == 0) {
-		writel(0, base + REG_UNIT);
-		writel(lat_cycles, base + REG_VAL);
-		pr_info("BVCI Latency for all Memory Transactions %d cycles\n",
-			lat_cycles);
-	} else {
-		for_each_set_bit(i, &units_req, MAX_BVCI_UNITS) {
-			writel(i + 1, base + REG_UNIT); /* loop is 0 based */
-			writel(lat_cycles, base + REG_VAL);
-			pr_info("BVCI Latency for Unit[%d] = %d cycles\n",
-				(i + 1), lat_cycles);
-		}
-	}
-}
-#else
-static void __init setup_bvci_lat_unit(void)
-{
-}
-#endif
-
 /*----------------------- Platform Devices -----------------------------*/
 
 #if IS_ENABLED(CONFIG_SERIAL_ARC)
@@ -132,16 +79,11 @@ static void arc_fpga_serial_init(void)
 				   ARRAY_SIZE(fpga_early_devs));
 
 	/*
-	 * ARC console driver registers itself as an early platform driver
-	 * of class "earlyprintk".
-	 * Install it here, followed by probe of devices.
-	 * The installation here doesn't require earlyprintk in command line
-	 * To do so however, replace the lines below with
-	 *	parse_early_param();
-	 *	early_platform_driver_probe("earlyprintk", 1, 1);
-	 *						      ^^
+	 * ARC console driver registers (build time) as an early platform driver
+	 * of class "earlyprintk". However it needs explicit cmdline toggle
+	 * "earlyprintk=ttyARC0" to be successfuly runtime registered.
+	 * Otherwise the early probe below fails to find the driver
 	 */
-	early_platform_driver_register_all("earlyprintk");
 	early_platform_driver_probe("earlyprintk", 1, 0);
 
 	/*
@@ -165,11 +107,9 @@ static void __init plat_fpga_early_init(void)
 {
 	pr_info("[plat-arcfpga]: registering early dev resources\n");
 
-	setup_bvci_lat_unit();
-
 	arc_fpga_serial_init();
 
-#ifdef CONFIG_SMP
+#ifdef CONFIG_ISS_SMP_EXTN
 	iss_model_init_early_smp();
 #endif
 }
@@ -211,7 +151,7 @@ MACHINE_START(ANGEL4, "angel4")
 	.init_early	= plat_fpga_early_init,
 	.init_machine	= plat_fpga_populate_dev,
 	.init_irq	= plat_fpga_init_IRQ,
-#ifdef CONFIG_SMP
+#ifdef CONFIG_ISS_SMP_EXTN
 	.init_smp	= iss_model_init_smp,
 #endif
 MACHINE_END
diff --git a/arch/arc/plat-arcfpga/smp.c b/arch/arc/plat-arcfpga/smp.c
index 8a12741f5f7a..92bad9122077 100644
--- a/arch/arc/plat-arcfpga/smp.c
+++ b/arch/arc/plat-arcfpga/smp.c
@@ -42,6 +42,24 @@ static void iss_model_smp_wakeup_cpu(int cpu, unsigned long pc)
 
 }
 
+static inline int get_hw_config_num_irq(void)
+{
+	uint32_t val = read_aux_reg(ARC_REG_VECBASE_BCR);
+
+	switch (val & 0x03) {
+	case 0:
+		return 16;
+	case 1:
+		return 32;
+	case 2:
+		return 8;
+	default:
+		return 0;
+	}
+
+	return 0;
+}
+
 /*
  * Any SMP specific init any CPU does when it comes up.
  * Here we setup the CPU to enable Inter-Processor-Interrupts