Merge tag 'arc-5.0-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc

Pull ARC architecture updates from Vineet Gupta:

 - Perf support for raw events

 - boot log printing: return stack, action points

 - fix memset to avoid prefetchw bleeding past end of buffer

 - do_page_fault fix for mmap_sem held while returning to userspace

 - other misc fixes

* tag 'arc-5.0-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARCv2: lib: memeset: fix doing prefetchw outside of buffer
  ARC: mm: do_page_fault fixes #1: relinquish mmap_sem if signal arrives while handle_mm_fault
  ARC: show_regs: lockdep: re-enable preemption
  ARC: show_regs: lockdep: avoid page allocator...
  ARC: perf: avoid kernel killing where it is possible
  ARC: perf: move HW events mapping to separate function
  ARC: perf: introduce Kernel PMU events support
  ARC: perf: trivial code cleanup
  ARC: perf: map generic branches to correct hardware condition
  ARC: adjust memblock_reserve of kernel memory
  arc: remove redundant kernel-space generic-y
  ARC: fix __ffs return value to avoid build warnings
  ARC: boot log: print Action point details
  ARCv2: boot log: BPU return stack depth

arch/arc/include/asm/Kbuild

@@ -3,23 +3,19 @@ generic-y += bugs.h
 generic-y += compat.h
 generic-y += device.h
 generic-y += div64.h
-generic-y += dma-mapping.h
 generic-y += emergency-restart.h
 generic-y += extable.h
-generic-y += fb.h
 generic-y += ftrace.h
 generic-y += hardirq.h
 generic-y += hw_irq.h
 generic-y += irq_regs.h
 generic-y += irq_work.h
-generic-y += kmap_types.h
 generic-y += local.h
 generic-y += local64.h
 generic-y += mcs_spinlock.h
 generic-y += mm-arch-hooks.h
 generic-y += msi.h
 generic-y += parport.h
-generic-y += pci.h
 generic-y += percpu.h
 generic-y += preempt.h
 generic-y += topology.h

arch/arc/include/asm/arcregs.h

@@ -216,6 +216,14 @@ struct bcr_fp_arcv2 {
 #endif
 };
 
+struct bcr_actionpoint {
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	unsigned int pad:21, min:1, num:2, ver:8;
+#else
+	unsigned int ver:8, num:2, min:1, pad:21;
+#endif
+};
+
 #include <soc/arc/timers.h>
 
 struct bcr_bpu_arcompact {
@@ -283,7 +291,7 @@ struct cpuinfo_arc_cache {
 };
 
 struct cpuinfo_arc_bpu {
-	unsigned int ver, full, num_cache, num_pred;
+	unsigned int ver, full, num_cache, num_pred, ret_stk;
 };
 
 struct cpuinfo_arc_ccm {
@@ -302,7 +310,7 @@ struct cpuinfo_arc {
 	struct {
 		unsigned int swap:1, norm:1, minmax:1, barrel:1, crc:1, swape:1, pad1:2,
 			     fpu_sp:1, fpu_dp:1, dual:1, dual_enb:1, pad2:4,
-			     debug:1, ap:1, smart:1, rtt:1, pad3:4,
+			     ap_num:4, ap_full:1, smart:1, rtt:1, pad3:1,
 			     timer0:1, timer1:1, rtc:1, gfrc:1, pad4:4;
 	} extn;
 	struct bcr_mpy extn_mpy;

arch/arc/include/asm/bitops.h

@@ -340,7 +340,7 @@ static inline __attribute__ ((const)) int __fls(unsigned long x)
 /*
  * __ffs: Similar to ffs, but zero based (0-31)
  */
-static inline __attribute__ ((const)) int __ffs(unsigned long word)
+static inline __attribute__ ((const)) unsigned long __ffs(unsigned long word)
 {
 	if (!word)
 		return word;
@@ -400,9 +400,9 @@ static inline __attribute__ ((const)) int ffs(unsigned long x)
 /*
  * __ffs: Similar to ffs, but zero based (0-31)
  */
-static inline __attribute__ ((const)) int __ffs(unsigned long x)
+static inline __attribute__ ((const)) unsigned long __ffs(unsigned long x)
 {
-	int n;
+	unsigned long n;
 
 	asm volatile(
 	"	ffs.f	%0, %1		\n"  /* 0:31; 31(Z) if src 0 */

arch/arc/include/asm/perf_event.h

@@ -103,7 +103,8 @@ static const char * const arc_pmu_ev_hw_map[] = {
 
 	/* counts condition */
 	[PERF_COUNT_HW_INSTRUCTIONS] = "iall",
-	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = "ijmp", /* Excludes ZOL jumps */
+	/* All jump instructions that are taken */
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = "ijmptak",
 	[PERF_COUNT_ARC_BPOK]         = "bpok",	  /* NP-NT, PT-T, PNT-NT */
 #ifdef CONFIG_ISA_ARCV2
 	[PERF_COUNT_HW_BRANCH_MISSES] = "bpmp",

arch/arc/kernel/perf_event.c

-/*
- * Linux performance counter support for ARC700 series
- *
- * Copyright (C) 2013-2015 Synopsys, Inc. (www.synopsys.com)
- *
- * This code is inspired by the perf support of various other architectures.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- */
+// SPDX-License-Identifier: GPL-2.0+
+//
+// Linux performance counter support for ARC CPUs.
+// This code is inspired by the perf support of various other architectures.
+//
+// Copyright (C) 2013-2018 Synopsys, Inc. (www.synopsys.com)
+
 #include <linux/errno.h>
 #include <linux/interrupt.h>
 #include <linux/module.h>
@@ -19,12 +14,31 @@
 #include <asm/arcregs.h>
 #include <asm/stacktrace.h>
 
+/* HW holds 8 symbols + one for null terminator */
+#define ARCPMU_EVENT_NAME_LEN	9
+
+enum arc_pmu_attr_groups {
+	ARCPMU_ATTR_GR_EVENTS,
+	ARCPMU_ATTR_GR_FORMATS,
+	ARCPMU_NR_ATTR_GR
+};
+
+struct arc_pmu_raw_event_entry {
+	char name[ARCPMU_EVENT_NAME_LEN];
+};
+
 struct arc_pmu {
 	struct pmu	pmu;
 	unsigned int	irq;
 	int		n_counters;
+	int		n_events;
 	u64		max_period;
 	int		ev_hw_idx[PERF_COUNT_ARC_HW_MAX];
+
+	struct arc_pmu_raw_event_entry	*raw_entry;
+	struct attribute		**attrs;
+	struct perf_pmu_events_attr	*attr;
+	const struct attribute_group	*attr_groups[ARCPMU_NR_ATTR_GR + 1];
 };
 
 struct arc_pmu_cpu {
@@ -49,6 +63,7 @@ static int callchain_trace(unsigned int addr, void *data)
 {
 	struct arc_callchain_trace *ctrl = data;
 	struct perf_callchain_entry_ctx *entry = ctrl->perf_stuff;
+
 	perf_callchain_store(entry, addr);
 
 	if (ctrl->depth++ < 3)
@@ -57,8 +72,8 @@ static int callchain_trace(unsigned int addr, void *data)
 	return -1;
 }
 
-void
-perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs)
+void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
+			   struct pt_regs *regs)
 {
 	struct arc_callchain_trace ctrl = {
 		.depth = 0,
@@ -68,8 +83,8 @@ perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *re
 	arc_unwind_core(NULL, regs, callchain_trace, &ctrl);
 }
 
-void
-perf_callchain_user(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs)
+void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
+			 struct pt_regs *regs)
 {
 	/*
 	 * User stack can't be unwound trivially with kernel dwarf unwinder
@@ -82,10 +97,10 @@ static struct arc_pmu *arc_pmu;
 static DEFINE_PER_CPU(struct arc_pmu_cpu, arc_pmu_cpu);
 
 /* read counter #idx; note that counter# != event# on ARC! */
-static uint64_t arc_pmu_read_counter(int idx)
+static u64 arc_pmu_read_counter(int idx)
 {
-	uint32_t tmp;
-	uint64_t result;
+	u32 tmp;
+	u64 result;
 
 	/*
 	 * ARC supports making 'snapshots' of the counters, so we don't
@@ -94,7 +109,7 @@ static uint64_t arc_pmu_read_counter(int idx)
 	write_aux_reg(ARC_REG_PCT_INDEX, idx);
 	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
 	write_aux_reg(ARC_REG_PCT_CONTROL, tmp | ARC_REG_PCT_CONTROL_SN);
-	result = (uint64_t) (read_aux_reg(ARC_REG_PCT_SNAPH)) << 32;
+	result = (u64) (read_aux_reg(ARC_REG_PCT_SNAPH)) << 32;
 	result |= read_aux_reg(ARC_REG_PCT_SNAPL);
 
 	return result;
@@ -103,9 +118,9 @@ 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)
 {
-	uint64_t prev_raw_count = local64_read(&hwc->prev_count);
-	uint64_t new_raw_count = arc_pmu_read_counter(idx);
-	int64_t delta = new_raw_count - prev_raw_count;
+	u64 prev_raw_count = local64_read(&hwc->prev_count);
+	u64 new_raw_count = arc_pmu_read_counter(idx);
+	s64 delta = new_raw_count - prev_raw_count;
 
 	/*
 	 * We aren't afraid of hwc->prev_count changing beneath our feet
@@ -155,7 +170,7 @@ static int arc_pmu_event_init(struct perf_event *event)
 	int ret;
 
 	if (!is_sampling_event(event)) {
-		hwc->sample_period  = arc_pmu->max_period;
+		hwc->sample_period = arc_pmu->max_period;
 		hwc->last_period = hwc->sample_period;
 		local64_set(&hwc->period_left, hwc->sample_period);
 	}
@@ -192,6 +207,18 @@ static int arc_pmu_event_init(struct perf_event *event)
 		pr_debug("init cache event with h/w %08x \'%s\'\n",
 			 (int)hwc->config, arc_pmu_ev_hw_map[ret]);
 		return 0;
+
+	case PERF_TYPE_RAW:
+		if (event->attr.config >= arc_pmu->n_events)
+			return -ENOENT;
+
+		hwc->config |= event->attr.config;
+		pr_debug("init raw event with idx %lld \'%s\'\n",
+			 event->attr.config,
+			 arc_pmu->raw_entry[event->attr.config].name);
+
+		return 0;
+
 	default:
 		return -ENOENT;
 	}
@@ -200,7 +227,7 @@ static int arc_pmu_event_init(struct perf_event *event)
 /* starts all counters */
 static void arc_pmu_enable(struct pmu *pmu)
 {
-	uint32_t tmp;
+	u32 tmp;
 	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
 	write_aux_reg(ARC_REG_PCT_CONTROL, (tmp & 0xffff0000) | 0x1);
 }
@@ -208,7 +235,7 @@ static void arc_pmu_enable(struct pmu *pmu)
 /* stops all counters */
 static void arc_pmu_disable(struct pmu *pmu)
 {
-	uint32_t tmp;
+	u32 tmp;
 	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
 	write_aux_reg(ARC_REG_PCT_CONTROL, (tmp & 0xffff0000) | 0x0);
 }
@@ -228,7 +255,7 @@ static int arc_pmu_event_set_period(struct perf_event *event)
 		local64_set(&hwc->period_left, left);
 		hwc->last_period = period;
 		overflow = 1;
-	} else	if (unlikely(left <= 0)) {
+	} else if (unlikely(left <= 0)) {
 		/* left underflowed by less than period. */
 		left += period;
 		local64_set(&hwc->period_left, left);
@@ -246,8 +273,8 @@ static int arc_pmu_event_set_period(struct perf_event *event)
 	write_aux_reg(ARC_REG_PCT_INDEX, idx);
 
 	/* Write value */
-	write_aux_reg(ARC_REG_PCT_COUNTL, (u32)value);
-	write_aux_reg(ARC_REG_PCT_COUNTH, (value >> 32));
+	write_aux_reg(ARC_REG_PCT_COUNTL, lower_32_bits(value));
+	write_aux_reg(ARC_REG_PCT_COUNTH, upper_32_bits(value));
 
 	perf_event_update_userpage(event);
 
@@ -277,7 +304,7 @@ static void arc_pmu_start(struct perf_event *event, int flags)
 	/* Enable interrupt for this counter */
 	if (is_sampling_event(event))
 		write_aux_reg(ARC_REG_PCT_INT_CTRL,
-			      read_aux_reg(ARC_REG_PCT_INT_CTRL) | (1 << idx));
+			      read_aux_reg(ARC_REG_PCT_INT_CTRL) | BIT(idx));
 
 	/* enable ARC pmu here */
 	write_aux_reg(ARC_REG_PCT_INDEX, idx);		/* counter # */
@@ -295,9 +322,9 @@ static void arc_pmu_stop(struct perf_event *event, int flags)
 		 * Reset interrupt flag by writing of 1. This is required
 		 * to make sure pending interrupt was not left.
 		 */
-		write_aux_reg(ARC_REG_PCT_INT_ACT, 1 << idx);
+		write_aux_reg(ARC_REG_PCT_INT_ACT, BIT(idx));
 		write_aux_reg(ARC_REG_PCT_INT_CTRL,
-			      read_aux_reg(ARC_REG_PCT_INT_CTRL) & ~(1 << idx));
+			      read_aux_reg(ARC_REG_PCT_INT_CTRL) & ~BIT(idx));
 	}
 
 	if (!(event->hw.state & PERF_HES_STOPPED)) {
@@ -349,9 +376,10 @@ static int arc_pmu_add(struct perf_event *event, int flags)
 
 	if (is_sampling_event(event)) {
 		/* Mimic full counter overflow as other arches do */
-		write_aux_reg(ARC_REG_PCT_INT_CNTL, (u32)arc_pmu->max_period);
+		write_aux_reg(ARC_REG_PCT_INT_CNTL,
+			      lower_32_bits(arc_pmu->max_period));
 		write_aux_reg(ARC_REG_PCT_INT_CNTH,
-			      (arc_pmu->max_period >> 32));
+			      upper_32_bits(arc_pmu->max_period));
 	}
 
 	write_aux_reg(ARC_REG_PCT_CONFIG, 0);
@@ -392,7 +420,7 @@ static irqreturn_t arc_pmu_intr(int irq, void *dev)
 		idx = __ffs(active_ints);
 
 		/* Reset interrupt flag by writing of 1 */
-		write_aux_reg(ARC_REG_PCT_INT_ACT, 1 << idx);
+		write_aux_reg(ARC_REG_PCT_INT_ACT, BIT(idx));
 
 		/*
 		 * On reset of "interrupt active" bit corresponding
@@ -400,7 +428,7 @@ static irqreturn_t arc_pmu_intr(int irq, void *dev)
 		 * Now we need to re-enable interrupt for the counter.
 		 */
 		write_aux_reg(ARC_REG_PCT_INT_CTRL,
-			read_aux_reg(ARC_REG_PCT_INT_CTRL) | (1 << idx));
+			read_aux_reg(ARC_REG_PCT_INT_CTRL) | BIT(idx));
 
 		event = pmu_cpu->act_counter[idx];
 		hwc = &event->hw;
@@ -414,7 +442,7 @@ static irqreturn_t arc_pmu_intr(int irq, void *dev)
 				arc_pmu_stop(event, 0);
 		}
 
-		active_ints &= ~(1U << idx);
+		active_ints &= ~BIT(idx);
 	} while (active_ints);
 
 done:
@@ -441,19 +469,108 @@ static void arc_cpu_pmu_irq_init(void *data)
 	write_aux_reg(ARC_REG_PCT_INT_ACT, 0xffffffff);
 }
 
+/* Event field occupies the bottom 15 bits of our config field */
+PMU_FORMAT_ATTR(event, "config:0-14");
+static struct attribute *arc_pmu_format_attrs[] = {
+	&format_attr_event.attr,
+	NULL,
+};
+
+static struct attribute_group arc_pmu_format_attr_gr = {
+	.name = "format",
+	.attrs = arc_pmu_format_attrs,
+};
+
+static ssize_t arc_pmu_events_sysfs_show(struct device *dev,
+					 struct device_attribute *attr,
+					 char *page)
+{
+	struct perf_pmu_events_attr *pmu_attr;
+
+	pmu_attr = container_of(attr, struct perf_pmu_events_attr, attr);
+	return sprintf(page, "event=0x%04llx\n", pmu_attr->id);
+}
+
+/*
+ * We don't add attrs here as we don't have pre-defined list of perf events.
+ * We will generate and add attrs dynamically in probe() after we read HW
+ * configuration.
+ */
+static struct attribute_group arc_pmu_events_attr_gr = {
+	.name = "events",
+};
+
+static void arc_pmu_add_raw_event_attr(int j, char *str)
+{
+	memmove(arc_pmu->raw_entry[j].name, str, ARCPMU_EVENT_NAME_LEN - 1);
+	arc_pmu->attr[j].attr.attr.name = arc_pmu->raw_entry[j].name;
+	arc_pmu->attr[j].attr.attr.mode = VERIFY_OCTAL_PERMISSIONS(0444);
+	arc_pmu->attr[j].attr.show = arc_pmu_events_sysfs_show;
+	arc_pmu->attr[j].id = j;
+	arc_pmu->attrs[j] = &(arc_pmu->attr[j].attr.attr);
+}
+
+static int arc_pmu_raw_alloc(struct device *dev)
+{
+	arc_pmu->attr = devm_kmalloc_array(dev, arc_pmu->n_events + 1,
+		sizeof(*arc_pmu->attr), GFP_KERNEL | __GFP_ZERO);
+	if (!arc_pmu->attr)
+		return -ENOMEM;
+
+	arc_pmu->attrs = devm_kmalloc_array(dev, arc_pmu->n_events + 1,
+		sizeof(*arc_pmu->attrs), GFP_KERNEL | __GFP_ZERO);
+	if (!arc_pmu->attrs)
+		return -ENOMEM;
+
+	arc_pmu->raw_entry = devm_kmalloc_array(dev, arc_pmu->n_events,
+		sizeof(*arc_pmu->raw_entry), GFP_KERNEL | __GFP_ZERO);
+	if (!arc_pmu->raw_entry)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static inline bool event_in_hw_event_map(int i, char *name)
+{
+	if (!arc_pmu_ev_hw_map[i])
+		return false;
+
+	if (!strlen(arc_pmu_ev_hw_map[i]))
+		return false;
+
+	if (strcmp(arc_pmu_ev_hw_map[i], name))
+		return false;
+
+	return true;
+}
+
+static void arc_pmu_map_hw_event(int j, char *str)
+{
+	int i;
+
+	/* See if HW condition has been mapped to a perf event_id */
+	for (i = 0; i < ARRAY_SIZE(arc_pmu_ev_hw_map); i++) {
+		if (event_in_hw_event_map(i, str)) {
+			pr_debug("mapping perf event %2d to h/w event \'%8s\' (idx %d)\n",
+				 i, str, j);
+			arc_pmu->ev_hw_idx[i] = j;
+		}
+	}
+}
+
 static int arc_pmu_device_probe(struct platform_device *pdev)
 {
 	struct arc_reg_pct_build pct_bcr;
 	struct arc_reg_cc_build cc_bcr;
-	int i, j, has_interrupts;
+	int i, has_interrupts;
 	int counter_size;	/* in bits */
 
 	union cc_name {
 		struct {
-			uint32_t word0, word1;
+			u32 word0, word1;
 			char sentinel;
 		} indiv;
-		char str[9];
+		char str[ARCPMU_EVENT_NAME_LEN];
 	} cc_name;
 
 
@@ -463,15 +580,22 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 	BUILD_BUG_ON(ARC_PERF_MAX_COUNTERS > 32);
-	BUG_ON(pct_bcr.c > ARC_PERF_MAX_COUNTERS);
+	if (WARN_ON(pct_bcr.c > ARC_PERF_MAX_COUNTERS))
+		return -EINVAL;
 
 	READ_BCR(ARC_REG_CC_BUILD, cc_bcr);
-	BUG_ON(!cc_bcr.v); /* Counters exist but No countable conditions ? */
+	if (WARN(!cc_bcr.v, "Counters exist but No countable conditions?"))
+		return -EINVAL;
 
 	arc_pmu = devm_kzalloc(&pdev->dev, sizeof(struct arc_pmu), GFP_KERNEL);
 	if (!arc_pmu)
 		return -ENOMEM;
 
+	arc_pmu->n_events = cc_bcr.c;
+
+	if (arc_pmu_raw_alloc(&pdev->dev))
+		return -ENOMEM;
+
 	has_interrupts = is_isa_arcv2() ? pct_bcr.i : 0;
 
 	arc_pmu->n_counters = pct_bcr.c;
@@ -481,30 +605,26 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 
 	pr_info("ARC perf\t: %d counters (%d bits), %d conditions%s\n",
 		arc_pmu->n_counters, counter_size, cc_bcr.c,
-		has_interrupts ? ", [overflow IRQ support]":"");
+		has_interrupts ? ", [overflow IRQ support]" : "");
 
-	cc_name.str[8] = 0;
+	cc_name.str[ARCPMU_EVENT_NAME_LEN - 1] = 0;
 	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);
+	for (i = 0; i < cc_bcr.c; i++) {
+		write_aux_reg(ARC_REG_CC_INDEX, i);
 		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 perf event %2d to h/w event \'%8s\' (idx %d)\n",
-					 i, cc_name.str, j);
-				arc_pmu->ev_hw_idx[i] = j;
-			}
-		}
+		arc_pmu_map_hw_event(i, cc_name.str);
+		arc_pmu_add_raw_event_attr(i, cc_name.str);
 	}
 
+	arc_pmu_events_attr_gr.attrs = arc_pmu->attrs;
+	arc_pmu->attr_groups[ARCPMU_ATTR_GR_EVENTS] = &arc_pmu_events_attr_gr;
+	arc_pmu->attr_groups[ARCPMU_ATTR_GR_FORMATS] = &arc_pmu_format_attr_gr;
+
 	arc_pmu->pmu = (struct pmu) {
 		.pmu_enable	= arc_pmu_enable,
 		.pmu_disable	= arc_pmu_disable,
@@ -514,6 +634,7 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 		.start		= arc_pmu_start,
 		.stop		= arc_pmu_stop,
 		.read		= arc_pmu_read,
+		.attr_groups	= arc_pmu->attr_groups,
 	};
 
 	if (has_interrupts) {
@@ -535,17 +656,19 @@ static int arc_pmu_device_probe(struct platform_device *pdev)
 	} else
 		arc_pmu->pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT;
 
-	return perf_pmu_register(&arc_pmu->pmu, pdev->name, PERF_TYPE_RAW);
+	/*
+	 * perf parser doesn't really like '-' symbol in events name, so let's
+	 * use '_' in arc pct name as it goes to kernel PMU event prefix.
+	 */
+	return perf_pmu_register(&arc_pmu->pmu, "arc_pct", PERF_TYPE_RAW);
 }
 
-#ifdef CONFIG_OF
 static const struct of_device_id arc_pmu_match[] = {
 	{ .compatible = "snps,arc700-pct" },
 	{ .compatible = "snps,archs-pct" },
 	{},
 };
 MODULE_DEVICE_TABLE(of, arc_pmu_match);
-#endif
 
 static struct platform_driver arc_pmu_driver = {
 	.driver	= {

arch/arc/kernel/setup.c

@@ -123,6 +123,7 @@ static void read_arc_build_cfg_regs(void)
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[smp_processor_id()];
 	const struct id_to_str *tbl;
 	struct bcr_isa_arcv2 isa;
+	struct bcr_actionpoint ap;
 
 	FIX_PTR(cpu);
 
@@ -195,6 +196,7 @@ static void read_arc_build_cfg_regs(void)
 		cpu->bpu.full = bpu.ft;
 		cpu->bpu.num_cache = 256 << bpu.bce;
 		cpu->bpu.num_pred = 2048 << bpu.pte;
+		cpu->bpu.ret_stk = 4 << bpu.rse;
 
 		if (cpu->core.family >= 0x54) {
 			unsigned int exec_ctrl;
@@ -207,8 +209,11 @@ static void read_arc_build_cfg_regs(void)
 		}
 	}
 
-	READ_BCR(ARC_REG_AP_BCR, bcr);
-	cpu->extn.ap = bcr.ver ? 1 : 0;
+	READ_BCR(ARC_REG_AP_BCR, ap);
+	if (ap.ver) {
+		cpu->extn.ap_num = 2 << ap.num;
+		cpu->extn.ap_full = !!ap.min;
+	}
 
 	READ_BCR(ARC_REG_SMART_BCR, bcr);
 	cpu->extn.smart = bcr.ver ? 1 : 0;
@@ -216,8 +221,6 @@ static void read_arc_build_cfg_regs(void)
 	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;
-
 	READ_BCR(ARC_REG_ISA_CFG_BCR, isa);
 
 	/* some hacks for lack of feature BCR info in old ARC700 cores */
@@ -299,10 +302,10 @@ static char *arc_cpu_mumbojumbo(int cpu_id, char *buf, int len)
 
 	if (cpu->bpu.ver)
 		n += scnprintf(buf + n, len - n,
-			      "BPU\t\t: %s%s match, cache:%d, Predict Table:%d",
+			      "BPU\t\t: %s%s match, cache:%d, Predict Table:%d Return stk: %d",
 			      IS_AVAIL1(cpu->bpu.full, "full"),
 			      IS_AVAIL1(!cpu->bpu.full, "partial"),
-			      cpu->bpu.num_cache, cpu->bpu.num_pred);
+			      cpu->bpu.num_cache, cpu->bpu.num_pred, cpu->bpu.ret_stk);
 
 	if (is_isa_arcv2()) {
 		struct bcr_lpb lpb;
@@ -336,11 +339,17 @@ static char *arc_extn_mumbojumbo(int cpu_id, char *buf, int len)
 			       IS_AVAIL1(cpu->extn.fpu_sp, "SP "),
 			       IS_AVAIL1(cpu->extn.fpu_dp, "DP "));
 
-	if (cpu->extn.debug)
-		n += scnprintf(buf + n, len - n, "DEBUG\t\t: %s%s%s\n",
-			       IS_AVAIL1(cpu->extn.ap, "ActionPoint "),
+	if (cpu->extn.ap_num | cpu->extn.smart | cpu->extn.rtt) {
+		n += scnprintf(buf + n, len - n, "DEBUG\t\t: %s%s",
 			       IS_AVAIL1(cpu->extn.smart, "smaRT "),
 			       IS_AVAIL1(cpu->extn.rtt, "RTT "));
+		if (cpu->extn.ap_num) {
+			n += scnprintf(buf + n, len - n, "ActionPoint %d/%s",
+				       cpu->extn.ap_num,
+				       cpu->extn.ap_full ? "full":"min");
+		}
+		n += scnprintf(buf + n, len - n, "\n");
+	}
 
 	if (cpu->dccm.sz || cpu->iccm.sz)
 		n += scnprintf(buf + n, len - n, "Extn [CCM]\t: DCCM @ %x, %d KB / ICCM: @ %x, %d KB\n",

arch/arc/kernel/troubleshoot.c

@@ -18,6 +18,8 @@
 #include <asm/arcregs.h>
 #include <asm/irqflags.h>
 
+#define ARC_PATH_MAX	256
+
 /*
  * Common routine to print scratch regs (r0-r12) or callee regs (r13-r25)
  *   -Prints 3 regs per line and a CR.
@@ -58,11 +60,12 @@ static void show_callee_regs(struct callee_regs *cregs)
 	print_reg_file(&(cregs->r13), 13);
 }
 
-static void print_task_path_n_nm(struct task_struct *tsk, char *buf)
+static void print_task_path_n_nm(struct task_struct *tsk)
 {
 	char *path_nm = NULL;
 	struct mm_struct *mm;
 	struct file *exe_file;
+	char buf[ARC_PATH_MAX];
 
 	mm = get_task_mm(tsk);
 	if (!mm)
@@ -72,7 +75,7 @@ static void print_task_path_n_nm(struct task_struct *tsk, char *buf)
 	mmput(mm);
 
 	if (exe_file) {
-		path_nm = file_path(exe_file, buf, 255);
+		path_nm = file_path(exe_file, buf, ARC_PATH_MAX-1);
 		fput(exe_file);
 	}
 
@@ -80,10 +83,9 @@ static void print_task_path_n_nm(struct task_struct *tsk, char *buf)
 	pr_info("Path: %s\n", !IS_ERR(path_nm) ? path_nm : "?");
 }
 
-static void show_faulting_vma(unsigned long address, char *buf)
+static void show_faulting_vma(unsigned long address)
 {
 	struct vm_area_struct *vma;
-	char *nm = buf;
 	struct mm_struct *active_mm = current->active_mm;
 
 	/* can't use print_vma_addr() yet as it doesn't check for
@@ -96,8 +98,11 @@ static void show_faulting_vma(unsigned long address, char *buf)
 	 * if the container VMA is not found
 	 */
 	if (vma && (vma->vm_start <= address)) {
+		char buf[ARC_PATH_MAX];
+		char *nm = "?";
+
 		if (vma->vm_file) {
-			nm = file_path(vma->vm_file, buf, PAGE_SIZE - 1);
+			nm = file_path(vma->vm_file, buf, ARC_PATH_MAX-1);
 			if (IS_ERR(nm))
 				nm = "?";
 		}
@@ -173,13 +178,14 @@ void show_regs(struct pt_regs *regs)
 {
 	struct task_struct *tsk = current;
 	struct callee_regs *cregs;
-	char *buf;
 
-	buf = (char *)__get_free_page(GFP_KERNEL);
-	if (!buf)
-		return;
+	/*
+	 * generic code calls us with preemption disabled, but some calls
+	 * here could sleep, so re-enable to avoid lockdep splat
+	 */
+	preempt_enable();
 
-	print_task_path_n_nm(tsk, buf);
+	print_task_path_n_nm(tsk);
 	show_regs_print_info(KERN_INFO);
 
 	show_ecr_verbose(regs);
@@ -189,7 +195,7 @@ void show_regs(struct pt_regs *regs)
 		(void *)regs->blink, (void *)regs->ret);
 
 	if (user_mode(regs))
-		show_faulting_vma(regs->ret, buf); /* faulting code, not data */
+		show_faulting_vma(regs->ret); /* faulting code, not data */
 
 	pr_info("[STAT32]: 0x%08lx", regs->status32);
 
@@ -222,7 +228,7 @@ void show_regs(struct pt_regs *regs)
 	if (cregs)
 		show_callee_regs(cregs);
 
-	free_page((unsigned long)buf);
+	preempt_disable();
 }
 
 void show_kernel_fault_diag(const char *str, struct pt_regs *regs,

arch/arc/lib/memset-archs.S

@@ -7,11 +7,39 @@
  */
 
 #include <linux/linkage.h>
+#include <asm/cache.h>
 
-#undef PREALLOC_NOT_AVAIL
+/*
+ * The memset implementation below is optimized to use prefetchw and prealloc
+ * instruction in case of CPU with 64B L1 data cache line (L1_CACHE_SHIFT == 6)
+ * If you want to implement optimized memset for other possible L1 data cache
+ * line lengths (32B and 128B) you should rewrite code carefully checking
+ * we don't call any prefetchw/prealloc instruction for L1 cache lines which
+ * don't belongs to memset area.
+ */
+
+#if L1_CACHE_SHIFT == 6
+
+.macro PREALLOC_INSTR	reg, off
+	prealloc	[\reg, \off]
+.endm
+
+.macro PREFETCHW_INSTR	reg, off
+	prefetchw	[\reg, \off]
+.endm
+
+#else
+
+.macro PREALLOC_INSTR
+.endm
+
+.macro PREFETCHW_INSTR
+.endm
+
+#endif
 
 ENTRY_CFI(memset)
-	prefetchw [r0]		; Prefetch the write location
+	PREFETCHW_INSTR	r0, 0	; Prefetch the first write location
 	mov.f	0, r2
 ;;; if size is zero
 	jz.d	[blink]
@@ -48,11 +76,8 @@ ENTRY_CFI(memset)
 
 	lpnz	@.Lset64bytes
 	;; LOOP START
-#ifdef PREALLOC_NOT_AVAIL
-	prefetchw [r3, 64]	;Prefetch the next write location
-#else
-	prealloc  [r3, 64]
-#endif
+	PREALLOC_INSTR	r3, 64	; alloc next line w/o fetching
+
 #ifdef CONFIG_ARC_HAS_LL64
 	std.ab	r4, [r3, 8]
 	std.ab	r4, [r3, 8]
@@ -85,7 +110,6 @@ ENTRY_CFI(memset)
 	lsr.f	lp_count, r2, 5 ;Last remaining  max 124 bytes
 	lpnz	.Lset32bytes
 	;; LOOP START
-	prefetchw   [r3, 32]	;Prefetch the next write location
 #ifdef CONFIG_ARC_HAS_LL64
 	std.ab	r4, [r3, 8]
 	std.ab	r4, [r3, 8]

arch/arc/mm/fault.c

@@ -141,12 +141,17 @@ void do_page_fault(unsigned long address, struct pt_regs *regs)
 	 */
 	fault = handle_mm_fault(vma, address, flags);
 
-	/* If Pagefault was interrupted by SIGKILL, exit page fault "early" */
 	if (fatal_signal_pending(current)) {
-		if ((fault & VM_FAULT_ERROR) && !(fault & VM_FAULT_RETRY))
-			up_read(&mm->mmap_sem);
-		if (user_mode(regs))
+
+		/*
+		 * if fault retry, mmap_sem already relinquished by core mm
+		 * so OK to return to user mode (with signal handled first)
+		 */
+		if (fault & VM_FAULT_RETRY) {
+			if (!user_mode(regs))
+				goto no_context;
 			return;
+		}
 	}
 
 	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);

arch/arc/mm/init.c

@@ -119,7 +119,8 @@ void __init setup_arch_memory(void)
 	 */
 
 	memblock_add_node(low_mem_start, low_mem_sz, 0);
-	memblock_reserve(low_mem_start, __pa(_end) - low_mem_start);
+	memblock_reserve(CONFIG_LINUX_LINK_BASE,
+			 __pa(_end) - CONFIG_LINUX_LINK_BASE);
 
 #ifdef CONFIG_BLK_DEV_INITRD
 	if (phys_initrd_size) {