ARC changes for 4.1-rc1:

- perf fixes/improvements
 - miscll cleanups
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQIcBAABAgAGBQJVOHi0AAoJEGnX8d3iisJe/LEP/0CZi32TvcLZks0d/euGUjrQ
 izY3/acjuCiBXTIXpEgbe7qikUgNUTuJjyXN4BCrwsjQ1diFt2v9/iH3bh/iNFMm
 s0Ykq2SW5T/owKSOOqpf7lPdFkuitogITnjZIF85kMQczzGDVz6FGj16UQ9jDdsK
 rxb21QBkyywBwh3usWZtV2DM8UGjnjWWV7mO9rJR1bl7BVJ9vnUU4pD6TDUnC+mx
 np+tqJAQ0GpXrXeZRjLYY2VlrWmj29//yZcOrAwhQpz6cSMa4IaIaD516AOY1ybo
 UOiepa6B9+rpXdKhM6+CQ564ga5S/lsldnPK3IQyrV3fzBG65BBWHysowGiGXMeo
 kzkiBMMnnxh1Xg74yn4WTeSHMIdQ8ilXi6zrewvQC7jvsvHGLF+rJ+75GNOeYzaV
 eK9dKAzsxvPpE23OknK/oLO9nVu+nxvllCWXobkASTwdzjKLj8csmUvVJbuy/hH+
 C6KSuVwuNXU7SBWvqJGdH6Iv0acZaueM32B8zkQ2d89ZDHRCLBLRq4vkhoKIJ8u4
 rm3ZUwn+TyXlO/RY7pr2tDUb9dsVx/qxGiFvow+o73jA1XMgEp60RpmgesK32mis
 veTeuplmq+I7ZjPyuHCWvSTM9nRXNSdCm2WIxirGliFzKbUIqouhondA076J9eU2
 AkCgh2TPG5h1SLTl/7Cs
 =No6A
 -----END PGP SIGNATURE-----

Merge tag 'arc-4.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc

Pull ARC updates from Vineet Gupta:

 - perf fixes/improvements

 - misc cleanups

* tag 'arc-4.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: perf: don't add code for impossible case
  ARC: perf: Rename DT binding to not confuse with power mgmt
  ARC: perf: add user space attribution in callchains
  ARC: perf: Add kernel callchain support
  ARC: perf: support cache hit/miss ratio
  ARC: perf: Add some comments/debug stuff
  ARC: perf: make @arc_pmu static global
  ARC: mem init spring cleaning - No functional changes
  ARC: Fix RTT boot printing
  ARC: fold __builtin_constant_p() into test_bit()
  ARC: rename unhandled exception handler
  ARC: cosmetic: Remove unused ECR bitfield masks
  ARC: Fix WRITE_BCR
  ARC: [nsimosci] Update defconfig
  arc: copy_thread(): rename 'arg' argument to 'kthread_arg'

Documentation/devicetree/bindings/arc/pct.txt

@@ -0,0 +1,20 @@
+* ARC Performance Counters
+
+The ARC700 can be configured with a pipeline performance monitor for counting
+CPU and cache events like cache misses and hits. Like conventional PCT there
+are 100+ hardware conditions dynamically mapped to upto 32 counters
+
+Note that:
+ * The ARC 700 PCT does not support interrupts; although HW events may be
+   counted, the HW events themselves cannot serve as a trigger for a sample.
+
+Required properties:
+
+- compatible : should contain
+	"snps,arc700-pct"
+
+Example:
+
+pmu {
+        compatible = "snps,arc700-pct";
+};

Documentation/devicetree/bindings/arc/pmu.txt

@@ -1,24 +0,0 @@
-* ARC Performance Monitor Unit
-
-The ARC 700 can be configured with a pipeline performance monitor for counting
-CPU and cache events like cache misses and hits.
-
-Note that:
- * ARC 700 refers to a family of ARC processor cores;
-   - There is only one type of PMU available for the whole family;
-   - The PMU may support different sets of events; supported events are probed
-     at boot time, as required by the reference manual.
-
- * The ARC 700 PMU does not support interrupts; although HW events may be
-   counted, the HW events themselves cannot serve as a trigger for a sample.
-
-Required properties:
-
-- compatible : should contain
-	"snps,arc700-pmu"
-
-Example:
-
-pmu {
-        compatible = "snps,arc700-pmu";
-};

arch/arc/boot/dts/angel4.dts

@@ -64,7 +64,7 @@
 		};
 
 		arcpmu0: pmu {
-			compatible = "snps,arc700-pmu";
+			compatible = "snps,arc700-pct";
 		};
 	};
 };

arch/arc/configs/nsimosci_defconfig

@@ -2,6 +2,9 @@ CONFIG_CROSS_COMPILE="arc-linux-uclibc-"
 # CONFIG_LOCALVERSION_AUTO is not set
 CONFIG_DEFAULT_HOSTNAME="ARCLinux"
 # CONFIG_SWAP is not set
+CONFIG_SYSVIPC=y
+# CONFIG_CROSS_MEMORY_ATTACH is not set
+CONFIG_NO_HZ=y
 CONFIG_HIGH_RES_TIMERS=y
 CONFIG_IKCONFIG=y
 CONFIG_IKCONFIG_PROC=y
@@ -9,7 +12,7 @@ CONFIG_NAMESPACES=y
 # CONFIG_UTS_NS is not set
 # CONFIG_PID_NS is not set
 CONFIG_BLK_DEV_INITRD=y
-CONFIG_INITRAMFS_SOURCE="../arc_initramfs"
+CONFIG_INITRAMFS_SOURCE="../arc_initramfs/"
 CONFIG_KALLSYMS_ALL=y
 CONFIG_EMBEDDED=y
 # CONFIG_SLUB_DEBUG is not set
@@ -21,12 +24,9 @@ CONFIG_MODULES=y
 # CONFIG_IOSCHED_DEADLINE is not set
 # CONFIG_IOSCHED_CFQ is not set
 CONFIG_ARC_PLAT_FPGA_LEGACY=y
-# CONFIG_ARC_IDE is not set
-# CONFIG_ARCTANGENT_EMAC is not set
 # CONFIG_ARC_HAS_RTSC is not set
 CONFIG_ARC_BUILTIN_DTB_NAME="nsimosci"
 # CONFIG_COMPACTION is not set
-# CONFIG_CROSS_MEMORY_ATTACH is not set
 CONFIG_NET=y
 CONFIG_PACKET=y
 CONFIG_UNIX=y
@@ -39,23 +39,23 @@ CONFIG_INET=y
 # CONFIG_FIRMWARE_IN_KERNEL is not set
 # CONFIG_BLK_DEV is not set
 CONFIG_NETDEVICES=y
-# CONFIG_INPUT_MOUSEDEV_PSAUX is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+CONFIG_INPUT_EVDEV=y
 # CONFIG_MOUSE_PS2_ALPS is not set
 # CONFIG_MOUSE_PS2_LOGIPS2PP is not set
 # CONFIG_MOUSE_PS2_SYNAPTICS is not set
+# CONFIG_MOUSE_PS2_CYPRESS is not set
 # CONFIG_MOUSE_PS2_TRACKPOINT is not set
 CONFIG_MOUSE_PS2_TOUCHKIT=y
-# CONFIG_SERIO_I8042 is not set
 # CONFIG_SERIO_SERPORT is not set
 CONFIG_SERIO_ARC_PS2=y
 # CONFIG_LEGACY_PTYS is not set
 # CONFIG_DEVKMEM is not set
 CONFIG_SERIAL_8250=y
 CONFIG_SERIAL_8250_CONSOLE=y
-CONFIG_SERIAL_8250_DW=y
+CONFIG_SERIAL_8250_NR_UARTS=1
+CONFIG_SERIAL_8250_RUNTIME_UARTS=1
 CONFIG_SERIAL_OF_PLATFORM=y
-CONFIG_SERIAL_ARC=y
-CONFIG_SERIAL_ARC_CONSOLE=y
 # CONFIG_HW_RANDOM is not set
 # CONFIG_HWMON is not set
 CONFIG_FB=y
@@ -72,4 +72,3 @@ CONFIG_TMPFS=y
 CONFIG_NFS_FS=y
 # CONFIG_ENABLE_WARN_DEPRECATED is not set
 # CONFIG_ENABLE_MUST_CHECK is not set
-CONFIG_XZ_DEC=y

arch/arc/include/asm/arcregs.h

@@ -30,6 +30,7 @@
 #define ARC_REG_D_UNCACH_BCR	0x6A
 #define ARC_REG_BPU_BCR		0xc0
 #define ARC_REG_ISA_CFG_BCR	0xc1
+#define ARC_REG_RTT_BCR		0xF2
 #define ARC_REG_SMART_BCR	0xFF
 
 /* status32 Bits Positions */
@@ -50,11 +51,7 @@
  * [15: 8] = Exception Cause Code
  * [ 7: 0] = Exception Parameters (for certain types only)
  */
-#define ECR_VEC_MASK			0xff0000
+#define ECR_V_MEM_ERR			0x01
-#define ECR_CODE_MASK			0x00ff00
-#define ECR_PARAM_MASK			0x0000ff
-
-/* Exception Cause Vector Values */
 #define ECR_V_INSN_ERR			0x02
 #define ECR_V_MACH_CHK			0x20
 #define ECR_V_ITLB_MISS			0x21
@@ -62,7 +59,8 @@
 #define ECR_V_PROTV			0x23
 #define ECR_V_TRAP			0x25
 
-/* Protection Violation Exception Cause Code Values */
+/* DTLB Miss and Protection Violation Cause Codes */
+
 #define ECR_C_PROTV_INST_FETCH		0x00
 #define ECR_C_PROTV_LOAD		0x01
 #define ECR_C_PROTV_STORE		0x02
@@ -173,11 +171,11 @@
 	}						\
 }
 
-#define WRITE_BCR(reg, into)				\
+#define WRITE_AUX(reg, into)				\
 {							\
 	unsigned int tmp;				\
 	if (sizeof(tmp) == sizeof(into)) {		\
-		tmp = (*(unsigned int *)(into));	\
+		tmp = (*(unsigned int *)&(into));	\
 		write_aux_reg(reg, tmp);		\
 	} else  {					\
 		extern void bogus_undefined(void);	\

arch/arc/include/asm/bitops.h

@@ -32,6 +32,20 @@ static inline void set_bit(unsigned long nr, volatile unsigned long *m)
 
 	m += nr >> 5;
 
+	/*
+	 * ARC ISA micro-optimization:
+	 *
+	 * Instructions dealing with bitpos only consider lower 5 bits (0-31)
+	 * e.g (x << 33) is handled like (x << 1) by ASL instruction
+	 *  (mem pointer still needs adjustment to point to next word)
+	 *
+	 * Hence the masking to clamp @nr arg can be elided in general.
+	 *
+	 * However if @nr is a constant (above assumed it in a register),
+	 * and greater than 31, gcc can optimize away (x << 33) to 0,
+	 * as overflow, given the 32-bit ISA. Thus masking needs to be done
+	 * for constant @nr, but no code is generated due to const prop.
+	 */
 	if (__builtin_constant_p(nr))
 		nr &= 0x1f;
 
@@ -374,29 +388,20 @@ __test_and_change_bit(unsigned long nr, volatile unsigned long *m)
  * This routine doesn't need to be atomic.
  */
 static inline int
-__constant_test_bit(unsigned int nr, const volatile unsigned long *addr)
+test_bit(unsigned int nr, const volatile unsigned long *addr)
-{
-	return ((1UL << (nr & 31)) &
-		(((const volatile unsigned int *)addr)[nr >> 5])) != 0;
-}
-
-static inline int
-__test_bit(unsigned int nr, const volatile unsigned long *addr)
 {
 	unsigned long mask;
 
 	addr += nr >> 5;
 
-	/* ARC700 only considers 5 bits in bit-fiddling insn */
+	if (__builtin_constant_p(nr))
+		nr &= 0x1f;
+
 	mask = 1 << nr;
 
 	return ((mask & *addr) != 0);
 }
 
-#define test_bit(nr, addr)	(__builtin_constant_p(nr) ? \
-					__constant_test_bit((nr), (addr)) : \
-					__test_bit((nr), (addr)))
-
 /*
  * Count the number of zeros, starting from MSB
  * Helper for fls( ) friends

arch/arc/include/asm/perf_event.h

@@ -54,29 +54,13 @@ struct arc_reg_cc_build {
 #define PERF_COUNT_ARC_BPOK	(PERF_COUNT_HW_MAX + 3)
 #define PERF_COUNT_ARC_EDTLB	(PERF_COUNT_HW_MAX + 4)
 #define PERF_COUNT_ARC_EITLB	(PERF_COUNT_HW_MAX + 5)
-#define PERF_COUNT_ARC_HW_MAX	(PERF_COUNT_HW_MAX + 6)
+#define PERF_COUNT_ARC_LDC	(PERF_COUNT_HW_MAX + 6)
+#define PERF_COUNT_ARC_STC	(PERF_COUNT_HW_MAX + 7)
+
+#define PERF_COUNT_ARC_HW_MAX	(PERF_COUNT_HW_MAX + 8)
 
 /*
- * The "generalized" performance events seem to really be a copy
+ * Some ARC pct quirks:
- * of the available events on x86 processors; the mapping to ARC
- * events is not always possible 1-to-1. Fortunately, there doesn't
- * seem to be an exact definition for these events, so we can cheat
- * a bit where necessary.
- *
- * In particular, the following PERF events may behave a bit differently
- * compared to other architectures:
- *
- * PERF_COUNT_HW_CPU_CYCLES
- *	Cycles not in halted state
- *
- * PERF_COUNT_HW_REF_CPU_CYCLES
- *	Reference cycles not in halted state, same as PERF_COUNT_HW_CPU_CYCLES
- *	for now as we don't do Dynamic Voltage/Frequency Scaling (yet)
- *
- * PERF_COUNT_HW_BUS_CYCLES
- *	Unclear what this means, Intel uses 0x013c, which according to
- *	their datasheet means "unhalted reference cycles". It sounds similar
- *	to PERF_COUNT_HW_REF_CPU_CYCLES, and we use the same counter for it.
  *
  * PERF_COUNT_HW_STALLED_CYCLES_BACKEND
  * PERF_COUNT_HW_STALLED_CYCLES_FRONTEND
@@ -91,21 +75,38 @@ struct arc_reg_cc_build {
  *	Note that I$ cache misses aren't counted by either of the two!
  */
 
+/*
+ * ARC PCT has hardware conditions with fixed "names" but variable "indexes"
+ * (based on a specific RTL build)
+ * Below is the static map between perf generic/arc specific event_id and
+ * h/w condition names.
+ * At the time of probe, we loop thru each index and find it's name to
+ * complete the mapping of perf event_id to h/w index as latter is needed
+ * to program the counter really
+ */
 static const char * const arc_pmu_ev_hw_map[] = {
+	/* count cycles */
 	[PERF_COUNT_HW_CPU_CYCLES] = "crun",
 	[PERF_COUNT_HW_REF_CPU_CYCLES] = "crun",
 	[PERF_COUNT_HW_BUS_CYCLES] = "crun",
-	[PERF_COUNT_HW_INSTRUCTIONS] = "iall",
+
-	[PERF_COUNT_HW_BRANCH_MISSES] = "bpfail",
-	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = "ijmp",
 	[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = "bflush",
 	[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = "bstall",
-	[PERF_COUNT_ARC_DCLM] = "dclm",
+
-	[PERF_COUNT_ARC_DCSM] = "dcsm",
+	/* counts condition */
-	[PERF_COUNT_ARC_ICM] = "icm",
+	[PERF_COUNT_HW_INSTRUCTIONS] = "iall",
-	[PERF_COUNT_ARC_BPOK] = "bpok",
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = "ijmp",
-	[PERF_COUNT_ARC_EDTLB] = "edtlb",
+	[PERF_COUNT_ARC_BPOK]         = "bpok",	  /* NP-NT, PT-T, PNT-NT */
-	[PERF_COUNT_ARC_EITLB] = "eitlb",
+	[PERF_COUNT_HW_BRANCH_MISSES] = "bpfail", /* NP-T, PT-NT, PNT-T */
+
+	[PERF_COUNT_ARC_LDC] = "imemrdc",	/* Instr: mem read cached */
+	[PERF_COUNT_ARC_STC] = "imemwrc",	/* Instr: mem write cached */
+
+	[PERF_COUNT_ARC_DCLM] = "dclm",		/* D-cache Load Miss */
+	[PERF_COUNT_ARC_DCSM] = "dcsm",		/* D-cache Store Miss */
+	[PERF_COUNT_ARC_ICM] = "icm",		/* I-cache Miss */
+	[PERF_COUNT_ARC_EDTLB] = "edtlb",	/* D-TLB Miss */
+	[PERF_COUNT_ARC_EITLB] = "eitlb",	/* I-TLB Miss */
 };
 
 #define C(_x)			PERF_COUNT_HW_CACHE_##_x
@@ -114,11 +115,11 @@ static const char * const arc_pmu_ev_hw_map[] = {
 static const unsigned arc_pmu_cache_map[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = {
 	[C(L1D)] = {
 		[C(OP_READ)] = {
-			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_ACCESS)]	= PERF_COUNT_ARC_LDC,
 			[C(RESULT_MISS)]	= PERF_COUNT_ARC_DCLM,
 		},
 		[C(OP_WRITE)] = {
-			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_ACCESS)]	= PERF_COUNT_ARC_STC,
 			[C(RESULT_MISS)]	= PERF_COUNT_ARC_DCSM,
 		},
 		[C(OP_PREFETCH)] = {
@@ -128,7 +129,7 @@ static const unsigned arc_pmu_cache_map[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = {
 	},
 	[C(L1I)] = {
 		[C(OP_READ)] = {
-			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_ACCESS)]	= PERF_COUNT_HW_INSTRUCTIONS,
 			[C(RESULT_MISS)]	= PERF_COUNT_ARC_ICM,
 		},
 		[C(OP_WRITE)] = {
@@ -156,9 +157,10 @@ static const unsigned arc_pmu_cache_map[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = {
 	},
 	[C(DTLB)] = {
 		[C(OP_READ)] = {
-			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_ACCESS)]	= PERF_COUNT_ARC_LDC,
 			[C(RESULT_MISS)]	= PERF_COUNT_ARC_EDTLB,
 		},
+			/* DTLB LD/ST Miss not segregated by h/w*/
 		[C(OP_WRITE)] = {
 			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
 			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,

arch/arc/kernel/perf_event.c

@@ -16,6 +16,7 @@
 #include <linux/perf_event.h>
 #include <linux/platform_device.h>
 #include <asm/arcregs.h>
+#include <asm/stacktrace.h>
 
 struct arc_pmu {
 	struct pmu	pmu;
@@ -25,6 +26,46 @@ struct arc_pmu {
 	int		ev_hw_idx[PERF_COUNT_ARC_HW_MAX];
 };
 
+struct arc_callchain_trace {
+	int depth;
+	void *perf_stuff;
+};
+
+static int callchain_trace(unsigned int addr, void *data)
+{
+	struct arc_callchain_trace *ctrl = data;
+	struct perf_callchain_entry *entry = ctrl->perf_stuff;
+	perf_callchain_store(entry, addr);
+
+	if (ctrl->depth++ < 3)
+		return 0;
+
+	return -1;
+}
+
+void
+perf_callchain_kernel(struct perf_callchain_entry *entry, struct pt_regs *regs)
+{
+	struct arc_callchain_trace ctrl = {
+		.depth = 0,
+		.perf_stuff = entry,
+	};
+
+	arc_unwind_core(NULL, regs, callchain_trace, &ctrl);
+}
+
+void
+perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs)
+{
+	/*
+	 * User stack can't be unwound trivially with kernel dwarf unwinder
+	 * So for now just record the user PC
+	 */
+	perf_callchain_store(entry, instruction_pointer(regs));
+}
+
+static struct arc_pmu *arc_pmu;
+
 /* read counter #idx; note that counter# != event# on ARC! */
 static uint64_t arc_pmu_read_counter(int idx)
 {
@@ -47,7 +88,6 @@ static uint64_t arc_pmu_read_counter(int idx)
 static void arc_perf_event_update(struct perf_event *event,
 				  struct hw_perf_event *hwc, int idx)
 {
-	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
 	uint64_t prev_raw_count, new_raw_count;
 	int64_t delta;
 
@@ -89,13 +129,16 @@ static int arc_pmu_cache_event(u64 config)
 	if (ret == CACHE_OP_UNSUPPORTED)
 		return -ENOENT;
 
+	pr_debug("init cache event: type/op/result %d/%d/%d with h/w %d \'%s\'\n",
+		 cache_type, cache_op, cache_result, ret,
+		 arc_pmu_ev_hw_map[ret]);
+
 	return ret;
 }
 
 /* initializes hw_perf_event structure if event is supported */
 static int arc_pmu_event_init(struct perf_event *event)
 {
-	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
 	struct hw_perf_event *hwc = &event->hw;
 	int ret;
 
@@ -106,8 +149,9 @@ static int arc_pmu_event_init(struct perf_event *event)
 		if (arc_pmu->ev_hw_idx[event->attr.config] < 0)
 			return -ENOENT;
 		hwc->config = arc_pmu->ev_hw_idx[event->attr.config];
-		pr_debug("initializing event %d with cfg %d\n",
+		pr_debug("init event %d with h/w %d \'%s\'\n",
-			 (int) event->attr.config, (int) hwc->config);
+			 (int) event->attr.config, (int) hwc->config,
+			 arc_pmu_ev_hw_map[event->attr.config]);
 		return 0;
 	case PERF_TYPE_HW_CACHE:
 		ret = arc_pmu_cache_event(event->attr.config);
@@ -183,8 +227,6 @@ static void arc_pmu_stop(struct perf_event *event, int flags)
 
 static void arc_pmu_del(struct perf_event *event, int flags)
 {
-	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
-
 	arc_pmu_stop(event, PERF_EF_UPDATE);
 	__clear_bit(event->hw.idx, arc_pmu->used_mask);
 
@@ -194,7 +236,6 @@ static void arc_pmu_del(struct perf_event *event, int flags)
 /* allocate hardware counter and optionally start counting */
 static int arc_pmu_add(struct perf_event *event, int flags)
 {
-	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
 	struct hw_perf_event *hwc = &event->hw;
 	int idx = hwc->idx;
 
@@ -247,10 +288,7 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 	BUG_ON(pct_bcr.c > ARC_PMU_MAX_HWEVENTS);
 
 	READ_BCR(ARC_REG_CC_BUILD, cc_bcr);
-	if (!cc_bcr.v) {
+	BUG_ON(!cc_bcr.v); /* Counters exist but No countable conditions ? */
-		pr_err("Performance counters exist, but no countable conditions?\n");
-		return -ENODEV;
-	}
 
 	arc_pmu = devm_kzalloc(&pdev->dev, sizeof(struct arc_pmu), GFP_KERNEL);
 	if (!arc_pmu)
@@ -263,19 +301,22 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 		arc_pmu->n_counters, arc_pmu->counter_size, cc_bcr.c);
 
 	cc_name.str[8] = 0;
-	for (i = 0; i < PERF_COUNT_HW_MAX; i++)
+	for (i = 0; i < PERF_COUNT_ARC_HW_MAX; i++)
 		arc_pmu->ev_hw_idx[i] = -1;
 
+	/* loop thru all available h/w condition indexes */
 	for (j = 0; j < cc_bcr.c; j++) {
 		write_aux_reg(ARC_REG_CC_INDEX, j);
 		cc_name.indiv.word0 = read_aux_reg(ARC_REG_CC_NAME0);
 		cc_name.indiv.word1 = read_aux_reg(ARC_REG_CC_NAME1);
+
+		/* See if it has been mapped to a perf event_id */
 		for (i = 0; i < ARRAY_SIZE(arc_pmu_ev_hw_map); i++) {
 			if (arc_pmu_ev_hw_map[i] &&
 			    !strcmp(arc_pmu_ev_hw_map[i], cc_name.str) &&
 			    strlen(arc_pmu_ev_hw_map[i])) {
-				pr_debug("mapping %d to idx %d with name %s\n",
+				pr_debug("mapping perf event %2d to h/w event \'%8s\' (idx %d)\n",
-					 i, j, cc_name.str);
+					 i, cc_name.str, j);
 				arc_pmu->ev_hw_idx[i] = j;
 			}
 		}
@@ -302,7 +343,7 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 
 #ifdef CONFIG_OF
 static const struct of_device_id arc_pmu_match[] = {
-	{ .compatible = "snps,arc700-pmu" },
+	{ .compatible = "snps,arc700-pct" },
 	{},
 };
 MODULE_DEVICE_TABLE(of, arc_pmu_match);
@@ -310,7 +351,7 @@ MODULE_DEVICE_TABLE(of, arc_pmu_match);
 
 static struct platform_driver arc_pmu_driver = {
 	.driver	= {
-		.name		= "arc700-pmu",
+		.name		= "arc700-pct",
 		.of_match_table = of_match_ptr(arc_pmu_match),
 	},
 	.probe		= arc_pmu_device_probe,

arch/arc/kernel/process.c

@@ -49,7 +49,10 @@ void arch_cpu_idle(void)
 
 asmlinkage void ret_from_fork(void);
 
-/* Layout of Child kernel mode stack as setup at the end of this function is
+/*
+ * Copy architecture-specific thread state
+ *
+ * Layout of Child kernel mode stack as setup at the end of this function is
  *
  * |     ...        |
  * |     ...        |
@@ -81,7 +84,7 @@ asmlinkage void ret_from_fork(void);
  * ------------------  <===== END of PAGE
  */
 int copy_thread(unsigned long clone_flags,
-		unsigned long usp, unsigned long arg,
+		unsigned long usp, unsigned long kthread_arg,
 		struct task_struct *p)
 {
 	struct pt_regs *c_regs;        /* child's pt_regs */
@@ -112,7 +115,7 @@ int copy_thread(unsigned long clone_flags,
 	if (unlikely(p->flags & PF_KTHREAD)) {
 		memset(c_regs, 0, sizeof(struct pt_regs));
 
-		c_callee->r13 = arg; /* argument to kernel thread */
+		c_callee->r13 = kthread_arg;
 		c_callee->r14 = usp;  /* function */
 
 		return 0;

arch/arc/kernel/setup.c

@@ -120,7 +120,10 @@ static void read_arc_build_cfg_regs(void)
 	READ_BCR(ARC_REG_SMART_BCR, bcr);
 	cpu->extn.smart = bcr.ver ? 1 : 0;
 
-	cpu->extn.debug = cpu->extn.ap | cpu->extn.smart;
+	READ_BCR(ARC_REG_RTT_BCR, bcr);
+	cpu->extn.rtt = bcr.ver ? 1 : 0;
+
+	cpu->extn.debug = cpu->extn.ap | cpu->extn.smart | cpu->extn.rtt;
 }
 
 static const struct cpuinfo_data arc_cpu_tbl[] = {

arch/arc/kernel/traps.c

@@ -42,7 +42,7 @@ void die(const char *str, struct pt_regs *regs, unsigned long address)
  *  -for kernel, chk if due to copy_(to|from)_user, otherwise die()
  */
 static noinline int
-handle_exception(const char *str, struct pt_regs *regs, siginfo_t *info)
+unhandled_exception(const char *str, struct pt_regs *regs, siginfo_t *info)
 {
 	if (user_mode(regs)) {
 		struct task_struct *tsk = current;
@@ -71,7 +71,7 @@ int name(unsigned long address, struct pt_regs *regs) \
 		.si_code  = sicode,		\
 		.si_addr = (void __user *)address,	\
 	};					\
-	return handle_exception(str, regs, &info);\
+	return unhandled_exception(str, regs, &info);\
 }
 
 /*

arch/arc/mm/init.c

@@ -71,7 +71,7 @@ early_param("initrd", early_initrd);
  */
 void __init setup_arch_memory(void)
 {
-	unsigned long zones_size[MAX_NR_ZONES] = { 0, 0 };
+	unsigned long zones_size[MAX_NR_ZONES];
 	unsigned long end_mem = CONFIG_LINUX_LINK_BASE + arc_mem_sz;
 
 	init_mm.start_code = (unsigned long)_text;
@@ -90,7 +90,7 @@ void __init setup_arch_memory(void)
 	/*------------- externs in mm need setting up ---------------*/
 
 	/* first page of system - kernel .vector starts here */
-	min_low_pfn = PFN_DOWN(CONFIG_LINUX_LINK_BASE);
+	min_low_pfn = ARCH_PFN_OFFSET;
 
 	/* Last usable page of low mem (no HIGHMEM yet for ARC port) */
 	max_low_pfn = max_pfn = PFN_DOWN(end_mem);
@@ -111,7 +111,7 @@ void __init setup_arch_memory(void)
 
 	/*-------------- node setup --------------------------------*/
 	memset(zones_size, 0, sizeof(zones_size));
-	zones_size[ZONE_NORMAL] = max_low_pfn - min_low_pfn;
+	zones_size[ZONE_NORMAL] = max_mapnr;
 
 	/*
 	 * We can't use the helper free_area_init(zones[]) because it uses
@@ -123,6 +123,8 @@ void __init setup_arch_memory(void)
 			    zones_size,		/* num pages per zone */
 			    min_low_pfn,	/* first pfn of node */
 			    NULL);		/* NO holes */
+
+	high_memory = (void *)end_mem;
 }
 
 /*
@@ -133,7 +135,6 @@ void __init setup_arch_memory(void)
  */
 void __init mem_init(void)
 {
-	high_memory = (void *)(CONFIG_LINUX_LINK_BASE + arc_mem_sz);
 	free_all_bootmem();
 	mem_init_print_info(NULL);
 }