Merge branch 'perf/core' into perf/probes

Conflicts:
	tools/perf/Makefile

Merge reason:

 - fix the conflict
 - pick up the pr_*() infrastructure to queue up dependent patch

Signed-off-by: Ingo Molnar <mingo@elte.hu>

67 files changed, 3281 insertions, 2920 deletions

diff --git a/tools/perf/Documentation/perf-timechart.txt b/tools/perf/Documentation/perf-timechart.txt
index a7910099d6f..4b1788355ec 100644
--- a/tools/perf/Documentation/perf-timechart.txt
+++ b/tools/perf/Documentation/perf-timechart.txt
@@ -31,9 +31,12 @@ OPTIONS
 -w::
 --width=::
         Select the width of the SVG file (default: 1000)
--p::
+-P::
 --power-only::
         Only output the CPU power section of the diagram
+-p::
+--process::
+        Select the processes to display, by name or PID
 
 
 SEE ALSO
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 1811a7015f9..147e3cf035d 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -201,7 +201,14 @@ EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement
 
-CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
+ifeq ("$(origin DEBUG)", "command line")
+  PERF_DEBUG = $(DEBUG)
+endif
+ifndef PERF_DEBUG
+  CFLAGS_OPTIMIZE = -O6
+endif
+
+CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
 LDFLAGS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS)
 ALL_LDFLAGS = $(LDFLAGS)
@@ -329,8 +336,26 @@ LIB_H += ../../include/linux/perf_event.h
 LIB_H += ../../include/linux/rbtree.h
 LIB_H += ../../include/linux/list.h
 LIB_H += ../../include/linux/stringify.h
+LIB_H += util/include/linux/bitmap.h
+LIB_H += util/include/linux/bitops.h
+LIB_H += util/include/linux/compiler.h
+LIB_H += util/include/linux/ctype.h
+LIB_H += util/include/linux/kernel.h
 LIB_H += util/include/linux/list.h
+LIB_H += util/include/linux/module.h
+LIB_H += util/include/linux/poison.h
+LIB_H += util/include/linux/prefetch.h
+LIB_H += util/include/linux/rbtree.h
+LIB_H += util/include/linux/string.h
+LIB_H += util/include/linux/types.h
+LIB_H += util/include/asm/asm-offsets.h
+LIB_H += util/include/asm/bitops.h
+LIB_H += util/include/asm/byteorder.h
+LIB_H += util/include/asm/swab.h
+LIB_H += util/include/asm/system.h
+LIB_H += util/include/asm/uaccess.h
 LIB_H += perf.h
+LIB_H += util/event.h
 LIB_H += util/types.h
 LIB_H += util/levenshtein.h
 LIB_H += util/parse-options.h
@@ -344,9 +369,12 @@ LIB_H += util/strlist.h
 LIB_H += util/run-command.h
 LIB_H += util/sigchain.h
 LIB_H += util/symbol.h
-LIB_H += util/module.h
 LIB_H += util/color.h
 LIB_H += util/values.h
+LIB_H += util/sort.h
+LIB_H += util/hist.h
+LIB_H += util/thread.h
+LIB_H += util/data_map.h
 
 LIB_OBJS += util/abspath.o
 LIB_OBJS += util/alias.o
@@ -360,6 +388,9 @@ LIB_OBJS += util/parse-options.o
 LIB_OBJS += util/parse-events.o
 LIB_OBJS += util/path.o
 LIB_OBJS += util/rbtree.o
+LIB_OBJS += util/bitmap.o
+LIB_OBJS += util/hweight.o
+LIB_OBJS += util/find_next_bit.o
 LIB_OBJS += util/run-command.o
 LIB_OBJS += util/quote.o
 LIB_OBJS += util/strbuf.o
@@ -369,7 +400,6 @@ LIB_OBJS += util/usage.o
 LIB_OBJS += util/wrapper.o
 LIB_OBJS += util/sigchain.o
 LIB_OBJS += util/symbol.o
-LIB_OBJS += util/module.o
 LIB_OBJS += util/color.o
 LIB_OBJS += util/pager.o
 LIB_OBJS += util/header.o
@@ -382,6 +412,9 @@ LIB_OBJS += util/trace-event-parse.o
 LIB_OBJS += util/trace-event-read.o
 LIB_OBJS += util/trace-event-info.o
 LIB_OBJS += util/svghelper.o
+LIB_OBJS += util/sort.o
+LIB_OBJS += util/hist.o
+LIB_OBJS += util/data_map.o
 
 BUILTIN_OBJS += builtin-annotate.o
 BUILTIN_OBJS += builtin-help.o
@@ -424,8 +457,12 @@ ifeq ($(uname_S),Darwin)
 	PTHREAD_LIBS =
 endif
 
+ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
+	msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]);
+endif
+
 ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
-	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]);
+	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel);
 endif
 
 ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
@@ -795,6 +832,19 @@ util/config.o: util/config.c PERF-CFLAGS
 util/rbtree.o: ../../lib/rbtree.c PERF-CFLAGS
 	$(QUIET_CC)$(CC) -o util/rbtree.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
 
+# some perf warning policies can't fit to lib/bitmap.c, eg: it warns about variable shadowing
+# from <string.h> that comes from kernel headers wrapping.
+KBITMAP_FLAGS=`echo $(ALL_CFLAGS) | sed s/-Wshadow// | sed s/-Wswitch-default// | sed s/-Wextra//`
+
+util/bitmap.o: ../../lib/bitmap.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/bitmap.o -c $(KBITMAP_FLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/hweight.o: ../../lib/hweight.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/hweight.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/find_next_bit.o: ../../lib/find_next_bit.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/find_next_bit.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
 perf-%$X: %.o $(PERFLIBS)
 	$(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS)
 
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 1ec74161581..6d63c2eea2c 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -22,15 +22,13 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 #include "util/thread.h"
+#include "util/sort.h"
+#include "util/hist.h"
 
 static char		const *input_name = "perf.data";
 
-static char		default_sort_order[] = "comm,symbol";
-static char		*sort_order = default_sort_order;
-
 static int		force;
 static int		input;
-static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
 static int		full_paths;
 
@@ -39,9 +37,10 @@ static int		print_line;
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
-static struct rb_root	threads;
-static struct thread	*last_match;
-
+struct sym_hist {
+	u64		sum;
+	u64		ip[0];
+};
 
 struct sym_ext {
 	struct rb_node	node;
@@ -49,247 +48,33 @@ struct sym_ext {
 	char		*path;
 };
 
-/*
- * histogram, sorted on item, collects counts
- */
-
-static struct rb_root hist;
-
-struct hist_entry {
-	struct rb_node	 rb_node;
-
-	struct thread	 *thread;
-	struct map	 *map;
-	struct dso	 *dso;
-	struct symbol	 *sym;
-	u64	 ip;
-	char		 level;
-
-	uint32_t	 count;
-};
-
-/*
- * configurable sorting bits
- */
-
-struct sort_entry {
-	struct list_head list;
-
-	const char *header;
-
-	int64_t (*cmp)(struct hist_entry *, struct hist_entry *);
-	int64_t (*collapse)(struct hist_entry *, struct hist_entry *);
-	size_t	(*print)(FILE *fp, struct hist_entry *);
-};
-
-/* --sort pid */
-
-static int64_t
-sort__thread_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	return right->thread->pid - left->thread->pid;
-}
-
-static size_t
-sort__thread_print(FILE *fp, struct hist_entry *self)
-{
-	return fprintf(fp, "%16s:%5d", self->thread->comm ?: "", self->thread->pid);
-}
-
-static struct sort_entry sort_thread = {
-	.header = "         Command:  Pid",
-	.cmp	= sort__thread_cmp,
-	.print	= sort__thread_print,
-};
-
-/* --sort comm */
-
-static int64_t
-sort__comm_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	return right->thread->pid - left->thread->pid;
-}
-
-static int64_t
-sort__comm_collapse(struct hist_entry *left, struct hist_entry *right)
-{
-	char *comm_l = left->thread->comm;
-	char *comm_r = right->thread->comm;
-
-	if (!comm_l || !comm_r) {
-		if (!comm_l && !comm_r)
-			return 0;
-		else if (!comm_l)
-			return -1;
-		else
-			return 1;
-	}
-
-	return strcmp(comm_l, comm_r);
-}
-
-static size_t
-sort__comm_print(FILE *fp, struct hist_entry *self)
-{
-	return fprintf(fp, "%16s", self->thread->comm);
-}
-
-static struct sort_entry sort_comm = {
-	.header		= "         Command",
-	.cmp		= sort__comm_cmp,
-	.collapse	= sort__comm_collapse,
-	.print		= sort__comm_print,
-};
-
-/* --sort dso */
-
-static int64_t
-sort__dso_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	struct dso *dso_l = left->dso;
-	struct dso *dso_r = right->dso;
-
-	if (!dso_l || !dso_r) {
-		if (!dso_l && !dso_r)
-			return 0;
-		else if (!dso_l)
-			return -1;
-		else
-			return 1;
-	}
-
-	return strcmp(dso_l->name, dso_r->name);
-}
-
-static size_t
-sort__dso_print(FILE *fp, struct hist_entry *self)
-{
-	if (self->dso)
-		return fprintf(fp, "%-25s", self->dso->name);
-
-	return fprintf(fp, "%016llx         ", (u64)self->ip);
-}
-
-static struct sort_entry sort_dso = {
-	.header = "Shared Object            ",
-	.cmp	= sort__dso_cmp,
-	.print	= sort__dso_print,
-};
-
-/* --sort symbol */
-
-static int64_t
-sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	u64 ip_l, ip_r;
-
-	if (left->sym == right->sym)
-		return 0;
-
-	ip_l = left->sym ? left->sym->start : left->ip;
-	ip_r = right->sym ? right->sym->start : right->ip;
-
-	return (int64_t)(ip_r - ip_l);
-}
-
-static size_t
-sort__sym_print(FILE *fp, struct hist_entry *self)
-{
-	size_t ret = 0;
-
-	if (verbose)
-		ret += fprintf(fp, "%#018llx  ", (u64)self->ip);
-
-	if (self->sym) {
-		ret += fprintf(fp, "[%c] %s",
-			self->dso == kernel_dso ? 'k' : '.', self->sym->name);
-	} else {
-		ret += fprintf(fp, "%#016llx", (u64)self->ip);
-	}
-
-	return ret;
-}
-
-static struct sort_entry sort_sym = {
-	.header = "Symbol",
-	.cmp	= sort__sym_cmp,
-	.print	= sort__sym_print,
-};
-
-static int sort__need_collapse = 0;
-
-struct sort_dimension {
-	const char		*name;
-	struct sort_entry	*entry;
-	int			taken;
-};
-
-static struct sort_dimension sort_dimensions[] = {
-	{ .name = "pid",	.entry = &sort_thread,	},
-	{ .name = "comm",	.entry = &sort_comm,	},
-	{ .name = "dso",	.entry = &sort_dso,	},
-	{ .name = "symbol",	.entry = &sort_sym,	},
+struct sym_priv {
+	struct sym_hist	*hist;
+	struct sym_ext	*ext;
 };
 
-static LIST_HEAD(hist_entry__sort_list);
+static const char *sym_hist_filter;
 
-static int sort_dimension__add(char *tok)
+static int symbol_filter(struct map *map, struct symbol *sym)
 {
-	unsigned int i;
-
-	for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) {
-		struct sort_dimension *sd = &sort_dimensions[i];
-
-		if (sd->taken)
-			continue;
-
-		if (strncasecmp(tok, sd->name, strlen(tok)))
-			continue;
-
-		if (sd->entry->collapse)
-			sort__need_collapse = 1;
-
-		list_add_tail(&sd->entry->list, &hist_entry__sort_list);
-		sd->taken = 1;
+	if (sym_hist_filter == NULL ||
+	    strcmp(sym->name, sym_hist_filter) == 0) {
+		struct sym_priv *priv = dso__sym_priv(map->dso, sym);
+		const int size = (sizeof(*priv->hist) +
+				  (sym->end - sym->start) * sizeof(u64));
 
+		priv->hist = malloc(size);
+		if (priv->hist)
+			memset(priv->hist, 0, size);
 		return 0;
 	}
-
-	return -ESRCH;
-}
-
-static int64_t
-hist_entry__cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	struct sort_entry *se;
-	int64_t cmp = 0;
-
-	list_for_each_entry(se, &hist_entry__sort_list, list) {
-		cmp = se->cmp(left, right);
-		if (cmp)
-			break;
-	}
-
-	return cmp;
-}
-
-static int64_t
-hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
-{
-	struct sort_entry *se;
-	int64_t cmp = 0;
-
-	list_for_each_entry(se, &hist_entry__sort_list, list) {
-		int64_t (*f)(struct hist_entry *, struct hist_entry *);
-
-		f = se->collapse ?: se->cmp;
-
-		cmp = f(left, right);
-		if (cmp)
-			break;
-	}
-
-	return cmp;
+	/*
+	 * FIXME: We should really filter it out, as we don't want to go thru symbols
+	 * we're not interested, and if a DSO ends up with no symbols, delete it too,
+	 * but right now the kernel loading routines in symbol.c bail out if no symbols
+	 * are found, fix it later.
+	 */
+	return 0;
 }
 
 /*
@@ -299,196 +84,60 @@ static void hist_hit(struct hist_entry *he, u64 ip)
 {
 	unsigned int sym_size, offset;
 	struct symbol *sym = he->sym;
+	struct sym_priv *priv;
+	struct sym_hist *h;
 
 	he->count++;
 
-	if (!sym || !sym->hist)
+	if (!sym || !he->map)
+		return;
+
+	priv = dso__sym_priv(he->map->dso, sym);
+	if (!priv->hist)
 		return;
 
 	sym_size = sym->end - sym->start;
 	offset = ip - sym->start;
 
+	if (verbose)
+		fprintf(stderr, "%s: ip=%Lx\n", __func__,
+			he->map->unmap_ip(he->map, ip));
+
 	if (offset >= sym_size)
 		return;
 
-	sym->hist_sum++;
-	sym->hist[offset]++;
+	h = priv->hist;
+	h->sum++;
+	h->ip[offset]++;
 
 	if (verbose >= 3)
 		printf("%p %s: count++ [ip: %p, %08Lx] => %Ld\n",
 			(void *)(unsigned long)he->sym->start,
 			he->sym->name,
 			(void *)(unsigned long)ip, ip - he->sym->start,
-			sym->hist[offset]);
+			h->ip[offset]);
 }
 
-static int
-hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
-		struct symbol *sym, u64 ip, char level)
+static int hist_entry__add(struct thread *thread, struct map *map,
+			   struct symbol *sym, u64 ip, u64 count, char level)
 {
-	struct rb_node **p = &hist.rb_node;
-	struct rb_node *parent = NULL;
-	struct hist_entry *he;
-	struct hist_entry entry = {
-		.thread	= thread,
-		.map	= map,
-		.dso	= dso,
-		.sym	= sym,
-		.ip	= ip,
-		.level	= level,
-		.count	= 1,
-	};
-	int cmp;
-
-	while (*p != NULL) {
-		parent = *p;
-		he = rb_entry(parent, struct hist_entry, rb_node);
-
-		cmp = hist_entry__cmp(&entry, he);
-
-		if (!cmp) {
-			hist_hit(he, ip);
-
-			return 0;
-		}
-
-		if (cmp < 0)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	he = malloc(sizeof(*he));
-	if (!he)
+	bool hit;
+	struct hist_entry *he = __hist_entry__add(thread, map, sym, NULL, ip,
+						  count, level, &hit);
+	if (he == NULL)
 		return -ENOMEM;
-	*he = entry;
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &hist);
-
+	hist_hit(he, ip);
 	return 0;
 }
 
-static void hist_entry__free(struct hist_entry *he)
-{
-	free(he);
-}
-
-/*
- * collapse the histogram
- */
-
-static struct rb_root collapse_hists;
-
-static void collapse__insert_entry(struct hist_entry *he)
-{
-	struct rb_node **p = &collapse_hists.rb_node;
-	struct rb_node *parent = NULL;
-	struct hist_entry *iter;
-	int64_t cmp;
-
-	while (*p != NULL) {
-		parent = *p;
-		iter = rb_entry(parent, struct hist_entry, rb_node);
-
-		cmp = hist_entry__collapse(iter, he);
-
-		if (!cmp) {
-			iter->count += he->count;
-			hist_entry__free(he);
-			return;
-		}
-
-		if (cmp < 0)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &collapse_hists);
-}
-
-static void collapse__resort(void)
-{
-	struct rb_node *next;
-	struct hist_entry *n;
-
-	if (!sort__need_collapse)
-		return;
-
-	next = rb_first(&hist);
-	while (next) {
-		n = rb_entry(next, struct hist_entry, rb_node);
-		next = rb_next(&n->rb_node);
-
-		rb_erase(&n->rb_node, &hist);
-		collapse__insert_entry(n);
-	}
-}
-
-/*
- * reverse the map, sort on count.
- */
-
-static struct rb_root output_hists;
-
-static void output__insert_entry(struct hist_entry *he)
-{
-	struct rb_node **p = &output_hists.rb_node;
-	struct rb_node *parent = NULL;
-	struct hist_entry *iter;
-
-	while (*p != NULL) {
-		parent = *p;
-		iter = rb_entry(parent, struct hist_entry, rb_node);
-
-		if (he->count > iter->count)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &output_hists);
-}
-
-static void output__resort(void)
-{
-	struct rb_node *next;
-	struct hist_entry *n;
-	struct rb_root *tree = &hist;
-
-	if (sort__need_collapse)
-		tree = &collapse_hists;
-
-	next = rb_first(tree);
-
-	while (next) {
-		n = rb_entry(next, struct hist_entry, rb_node);
-		next = rb_next(&n->rb_node);
-
-		rb_erase(&n->rb_node, tree);
-		output__insert_entry(n);
-	}
-}
-
-static unsigned long total = 0,
-		     total_mmap = 0,
-		     total_comm = 0,
-		     total_fork = 0,
-		     total_unknown = 0;
-
 static int
 process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
 	char level;
-	int show = 0;
-	struct dso *dso = NULL;
-	struct thread *thread;
 	u64 ip = event->ip.ip;
 	struct map *map = NULL;
-
-	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+	struct symbol *sym = NULL;
+	struct thread *thread = threads__findnew(event->ip.pid);
 
 	dump_printf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
 		(void *)(offset + head),
@@ -497,60 +146,53 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		event->ip.pid,
 		(void *)(long)ip);
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
-
 	if (thread == NULL) {
 		fprintf(stderr, "problem processing %d event, skipping it.\n",
 			event->header.type);
 		return -1;
 	}
 
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+
 	if (event->header.misc & PERF_RECORD_MISC_KERNEL) {
-		show = SHOW_KERNEL;
 		level = 'k';
-
-		dso = kernel_dso;
-
-		dump_printf(" ...... dso: %s\n", dso->name);
-
+		sym = kernel_maps__find_symbol(ip, &map);
+		dump_printf(" ...... dso: %s\n",
+			    map ? map->dso->long_name : "<not found>");
 	} else if (event->header.misc & PERF_RECORD_MISC_USER) {
-
-		show = SHOW_USER;
 		level = '.';
-
 		map = thread__find_map(thread, ip);
 		if (map != NULL) {
+got_map:
 			ip = map->map_ip(map, ip);
-			dso = map->dso;
+			sym = map->dso->find_symbol(map->dso, ip);
 		} else {
 			/*
 			 * If this is outside of all known maps,
 			 * and is a negative address, try to look it
 			 * up in the kernel dso, as it might be a
-			 * vsyscall (which executes in user-mode):
+			 * vsyscall or vdso (which executes in user-mode).
+			 *
+			 * XXX This is nasty, we should have a symbol list in
+			 * the "[vdso]" dso, but for now lets use the old
+			 * trick of looking in the whole kernel symbol list.
 			 */
-			if ((long long)ip < 0)
-				dso = kernel_dso;
+			if ((long long)ip < 0) {
+				map = kernel_map;
+				goto got_map;
+			}
 		}
-		dump_printf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
-
+		dump_printf(" ...... dso: %s\n",
+			    map ? map->dso->long_name : "<not found>");
 	} else {
-		show = SHOW_HV;
 		level = 'H';
 		dump_printf(" ...... dso: [hypervisor]\n");
 	}
 
-	if (show & show_mask) {
-		struct symbol *sym = NULL;
-
-		if (dso)
-			sym = dso->find_symbol(dso, ip);
-
-		if (hist_entry__add(thread, map, dso, sym, ip, level)) {
-			fprintf(stderr,
-		"problem incrementing symbol count, skipping event\n");
-			return -1;
-		}
+	if (hist_entry__add(thread, map, sym, ip, 1, level)) {
+		fprintf(stderr, "problem incrementing symbol count, "
+				"skipping event\n");
+		return -1;
 	}
 	total++;
 
@@ -560,10 +202,9 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-	struct map *map = map__new(&event->mmap, NULL, 0);
-
-	thread = threads__findnew(event->mmap.pid, &threads, &last_match);
+	struct map *map = map__new(&event->mmap, NULL, 0,
+				   sizeof(struct sym_priv), symbol_filter);
+	struct thread *thread = threads__findnew(event->mmap.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_MMAP %d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
@@ -588,9 +229,8 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
+	struct thread *thread = threads__findnew(event->comm.pid);
 
-	thread = threads__findnew(event->comm.pid, &threads, &last_match);
 	dump_printf("%p [%p]: PERF_RECORD_COMM: %s:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -609,11 +249,9 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-	struct thread *parent;
+	struct thread *thread = threads__findnew(event->fork.pid);
+	struct thread *parent = threads__findnew(event->fork.ppid);
 
-	thread = threads__findnew(event->fork.pid, &threads, &last_match);
-	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
 	dump_printf("%p [%p]: PERF_RECORD_FORK: %d:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -665,14 +303,15 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
-static int
-parse_line(FILE *file, struct symbol *sym, u64 start, u64 len)
+static int parse_line(FILE *file, struct hist_entry *he, u64 len)
 {
+	struct symbol *sym = he->sym;
 	char *line = NULL, *tmp, *tmp2;
 	static const char *prev_line;
 	static const char *prev_color;
 	unsigned int offset;
 	size_t line_len;
+	u64 start;
 	s64 line_ip;
 	int ret;
 	char *c;
@@ -709,22 +348,26 @@ parse_line(FILE *file, struct symbol *sym, u64 start, u64 len)
 			line_ip = -1;
 	}
 
+	start = he->map->unmap_ip(he->map, sym->start);
+
 	if (line_ip != -1) {
 		const char *path = NULL;
 		unsigned int hits = 0;
 		double percent = 0.0;
 		const char *color;
-		struct sym_ext *sym_ext = sym->priv;
+		struct sym_priv *priv = dso__sym_priv(he->map->dso, sym);
+		struct sym_ext *sym_ext = priv->ext;
+		struct sym_hist *h = priv->hist;
 
 		offset = line_ip - start;
 		if (offset < len)
-			hits = sym->hist[offset];
+			hits = h->ip[offset];
 
 		if (offset < len && sym_ext) {
 			path = sym_ext[offset].path;
 			percent = sym_ext[offset].percent;
-		} else if (sym->hist_sum)
-			percent = 100.0 * hits / sym->hist_sum;
+		} else if (h->sum)
+			percent = 100.0 * hits / h->sum;
 
 		color = get_percent_color(percent);
 
@@ -777,9 +420,10 @@ static void insert_source_line(struct sym_ext *sym_ext)
 	rb_insert_color(&sym_ext->node, &root_sym_ext);
 }
 
-static void free_source_line(struct symbol *sym, int len)
+static void free_source_line(struct hist_entry *he, int len)
 {
-	struct sym_ext *sym_ext = sym->priv;
+	struct sym_priv *priv = dso__sym_priv(he->map->dso, he->sym);
+	struct sym_ext *sym_ext = priv->ext;
 	int i;
 
 	if (!sym_ext)
@@ -789,26 +433,30 @@ static void free_source_line(struct symbol *sym, int len)
 		free(sym_ext[i].path);
 	free(sym_ext);
 
-	sym->priv = NULL;
+	priv->ext = NULL;
 	root_sym_ext = RB_ROOT;
 }
 
 /* Get the filename:line for the colored entries */
 static void
-get_source_line(struct symbol *sym, u64 start, int len, const char *filename)
+get_source_line(struct hist_entry *he, int len, const char *filename)
 {
+	struct symbol *sym = he->sym;
+	u64 start;
 	int i;
 	char cmd[PATH_MAX * 2];
 	struct sym_ext *sym_ext;
+	struct sym_priv *priv = dso__sym_priv(he->map->dso, sym);
+	struct sym_hist *h = priv->hist;
 
-	if (!sym->hist_sum)
+	if (!h->sum)
 		return;
 
-	sym->priv = calloc(len, sizeof(struct sym_ext));
-	if (!sym->priv)
+	sym_ext = priv->ext = calloc(len, sizeof(struct sym_ext));
+	if (!priv->ext)
 		return;
 
-	sym_ext = sym->priv;
+	start = he->map->unmap_ip(he->map, sym->start);
 
 	for (i = 0; i < len; i++) {
 		char *path = NULL;
@@ -816,7 +464,7 @@ get_source_line(struct symbol *sym, u64 start, int len, const char *filename)
 		u64 offset;
 		FILE *fp;
 
-		sym_ext[i].percent = 100.0 * sym->hist[i] / sym->hist_sum;
+		sym_ext[i].percent = 100.0 * h->ip[i] / h->sum;
 		if (sym_ext[i].percent <= 0.5)
 			continue;
 
@@ -870,33 +518,34 @@ static void print_summary(const char *filename)
 	}
 }
 
-static void annotate_sym(struct dso *dso, struct symbol *sym)
+static void annotate_sym(struct hist_entry *he)
 {
-	const char *filename = dso->name, *d_filename;
-	u64 start, end, len;
+	struct map *map = he->map;
+	struct dso *dso = map->dso;
+	struct symbol *sym = he->sym;
+	const char *filename = dso->long_name, *d_filename;
+	u64 len;
 	char command[PATH_MAX*2];
 	FILE *file;
 
 	if (!filename)
 		return;
-	if (sym->module)
-		filename = sym->module->path;
-	else if (dso == kernel_dso)
-		filename = vmlinux_name;
-
-	start = sym->obj_start;
-	if (!start)
-		start = sym->start;
+
+	if (verbose)
+		fprintf(stderr, "%s: filename=%s, sym=%s, start=%Lx, end=%Lx\n",
+			__func__, filename, sym->name,
+			map->unmap_ip(map, sym->start),
+			map->unmap_ip(map, sym->end));
+
 	if (full_paths)
 		d_filename = filename;
 	else
 		d_filename = basename(filename);
 
-	end = start + sym->end - sym->start + 1;
 	len = sym->end - sym->start;
 
 	if (print_line) {
-		get_source_line(sym, start, len, filename);
+		get_source_line(he, len, filename);
 		print_summary(filename);
 	}
 
@@ -905,10 +554,12 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 	printf("------------------------------------------------\n");
 
 	if (verbose >= 2)
-		printf("annotating [%p] %30s : [%p] %30s\n", dso, dso->name, sym, sym->name);
+		printf("annotating [%p] %30s : [%p] %30s\n",
+		       dso, dso->long_name, sym, sym->name);
 
 	sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s|grep -v %s",
-			(u64)start, (u64)end, filename, filename);
+		map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end),
+		filename, filename);
 
 	if (verbose >= 3)
 		printf("doing: %s\n", command);
@@ -918,35 +569,38 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 		return;
 
 	while (!feof(file)) {
-		if (parse_line(file, sym, start, len) < 0)
+		if (parse_line(file, he, len) < 0)
 			break;
 	}
 
 	pclose(file);
 	if (print_line)
-		free_source_line(sym, len);
+		free_source_line(he, len);
 }
 
 static void find_annotations(void)
 {
 	struct rb_node *nd;
-	struct dso *dso;
-	int count = 0;
 
-	list_for_each_entry(dso, &dsos, node) {
+	for (nd = rb_first(&output_hists); nd; nd = rb_next(nd)) {
+		struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node);
+		struct sym_priv *priv;
 
-		for (nd = rb_first(&dso->syms); nd; nd = rb_next(nd)) {
-			struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+		if (he->sym == NULL)
+			continue;
 
-			if (sym->hist) {
-				annotate_sym(dso, sym);
-				count++;
-			}
-		}
-	}
+		priv = dso__sym_priv(he->map->dso, he->sym);
+		if (priv->hist == NULL)
+			continue;
 
-	if (!count)
-		printf(" Error: symbol '%s' not present amongst the samples.\n", sym_hist_filter);
+		annotate_sym(he);
+		/*
+		 * Since we have a hist_entry per IP for the same symbol, free
+		 * he->sym->hist to signal we already processed this symbol.
+		 */
+		free(priv->hist);
+		priv->hist = NULL;
+	}
 }
 
 static int __cmd_annotate(void)
@@ -959,7 +613,7 @@ static int __cmd_annotate(void)
 	uint32_t size;
 	char *buf;
 
-	register_idle_thread(&threads, &last_match);
+	register_idle_thread();
 
 	input = open(input_name, O_RDONLY);
 	if (input < 0) {
@@ -983,7 +637,7 @@ static int __cmd_annotate(void)
 		exit(0);
 	}
 
-	if (load_kernel() < 0) {
+	if (load_kernel(sizeof(struct sym_priv), symbol_filter) < 0) {
 		perror("failed to load kernel symbols");
 		return EXIT_FAILURE;
 	}
@@ -1059,14 +713,14 @@ more:
 	if (dump_trace)
 		return 0;
 
-	if (verbose >= 3)
-		threads__fprintf(stdout, &threads);
+	if (verbose > 3)
+		threads__fprintf(stdout);
 
-	if (verbose >= 2)
+	if (verbose > 2)
 		dsos__fprintf(stdout);
 
 	collapse__resort();
-	output__resort();
+	output__resort(total);
 
 	find_annotations();
 
@@ -1134,10 +788,13 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used)
 		sym_hist_filter = argv[0];
 	}
 
-	if (!sym_hist_filter)
-		usage_with_options(annotate_usage, options);
-
 	setup_pager();
 
+	if (field_sep && *field_sep == '.') {
+		fputs("'.' is the only non valid --field-separator argument\n",
+				stderr);
+		exit(129);
+	}
+
 	return __cmd_annotate();
 }
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 3eeef339c78..ac5ddfff445 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -17,55 +17,51 @@
 #include "util/header.h"
 #include "util/event.h"
 #include "util/debug.h"
-#include "util/trace-event.h"
 
 #include <unistd.h>
 #include <sched.h>
 
-#define ALIGN(x, a)		__ALIGN_MASK(x, (typeof(x))(a)-1)
-#define __ALIGN_MASK(x, mask)	(((x)+(mask))&~(mask))
-
 static int			fd[MAX_NR_CPUS][MAX_COUNTERS];
 
-static long			default_interval		= 100000;
+static long			default_interval		=      0;
 
-static int			nr_cpus				= 0;
+static int			nr_cpus				=      0;
 static unsigned int		page_size;
-static unsigned int		mmap_pages			= 128;
-static int			freq				= 0;
+static unsigned int		mmap_pages			=    128;
+static int			freq				=   1000;
 static int			output;
 static const char		*output_name			= "perf.data";
-static int			group				= 0;
-static unsigned int		realtime_prio			= 0;
-static int			raw_samples			= 0;
-static int			system_wide			= 0;
-static int			profile_cpu			= -1;
-static pid_t			target_pid			= -1;
-static pid_t			child_pid			= -1;
-static int			inherit				= 1;
-static int			force				= 0;
-static int			append_file			= 0;
-static int			call_graph			= 0;
-static int			inherit_stat			= 0;
-static int			no_samples			= 0;
-static int			sample_address			= 0;
-static int			multiplex			= 0;
-static int			multiplex_fd			= -1;
-
-static long			samples;
+static int			group				=      0;
+static unsigned int		realtime_prio			=      0;
+static int			raw_samples			=      0;
+static int			system_wide			=      0;
+static int			profile_cpu			=     -1;
+static pid_t			target_pid			=     -1;
+static pid_t			child_pid			=     -1;
+static int			inherit				=      1;
+static int			force				=      0;
+static int			append_file			=      0;
+static int			call_graph			=      0;
+static int			inherit_stat			=      0;
+static int			no_samples			=      0;
+static int			sample_address			=      0;
+static int			multiplex			=      0;
+static int			multiplex_fd			=     -1;
+
+static long			samples				=      0;
 static struct timeval		last_read;
 static struct timeval		this_read;
 
-static u64			bytes_written;
+static u64			bytes_written			=      0;
 
 static struct pollfd		event_array[MAX_NR_CPUS * MAX_COUNTERS];
 
-static int			nr_poll;
-static int			nr_cpu;
+static int			nr_poll				=      0;
+static int			nr_cpu				=      0;
 
-static int			file_new = 1;
+static int			file_new			=      1;
 
-struct perf_header		*header;
+struct perf_header		*header				=   NULL;
 
 struct mmap_data {
 	int			counter;
@@ -375,9 +371,11 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n
 
 static void create_counter(int counter, int cpu, pid_t pid)
 {
+	char *filter = filters[counter];
 	struct perf_event_attr *attr = attrs + counter;
 	struct perf_header_attr *h_attr;
 	int track = !counter; /* only the first counter needs these */
+	int ret;
 	struct {
 		u64 count;
 		u64 time_enabled;
@@ -480,7 +478,6 @@ try_again:
 		multiplex_fd = fd[nr_cpu][counter];
 
 	if (multiplex && fd[nr_cpu][counter] != multiplex_fd) {
-		int ret;
 
 		ret = ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_SET_OUTPUT, multiplex_fd);
 		assert(ret != -1);
@@ -500,6 +497,16 @@ try_again:
 		}
 	}
 
+	if (filter != NULL) {
+		ret = ioctl(fd[nr_cpu][counter],
+			    PERF_EVENT_IOC_SET_FILTER, filter);
+		if (ret) {
+			error("failed to set filter with %d (%s)\n", errno,
+			      strerror(errno));
+			exit(-1);
+		}
+	}
+
 	ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_ENABLE);
 }
 
@@ -566,17 +573,17 @@ static int __cmd_record(int argc, const char **argv)
 	else
 		header = perf_header__new();
 
-
 	if (raw_samples) {
-		read_tracing_data(attrs, nr_counters);
+		perf_header__feat_trace_info(header);
 	} else {
 		for (i = 0; i < nr_counters; i++) {
 			if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
-				read_tracing_data(attrs, nr_counters);
+				perf_header__feat_trace_info(header);
 				break;
 			}
 		}
 	}
+
 	atexit(atexit_header);
 
 	if (!system_wide) {
@@ -623,7 +630,7 @@ static int __cmd_record(int argc, const char **argv)
 
 		param.sched_priority = realtime_prio;
 		if (sched_setscheduler(0, SCHED_FIFO, &param)) {
-			printf("Could not set realtime priority.\n");
+			pr_err("Could not set realtime priority.\n");
 			exit(-1);
 		}
 	}
@@ -677,6 +684,8 @@ static const struct option options[] = {
 	OPT_CALLBACK('e', "event", NULL, "event",
 		     "event selector. use 'perf list' to list available events",
 		     parse_events),
+	OPT_CALLBACK(0, "filter", NULL, "filter",
+		     "event filter", parse_filter),
 	OPT_INTEGER('p', "pid", &target_pid,
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
@@ -731,6 +740,18 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		attrs[0].config = PERF_COUNT_HW_CPU_CYCLES;
 	}
 
+	/*
+	 * User specified count overrides default frequency.
+	 */
+	if (default_interval)
+		freq = 0;
+	else if (freq) {
+		default_interval = freq;
+	} else {
+		fprintf(stderr, "frequency and count are zero, aborting\n");
+		exit(EXIT_FAILURE);
+	}
+
 	for (counter = 0; counter < nr_counters; counter++) {
 		if (attrs[counter].sample_period)
 			continue;
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 19669c20088..b3d814b5455 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -26,20 +26,18 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 
+#include "util/data_map.h"
 #include "util/thread.h"
+#include "util/sort.h"
+#include "util/hist.h"
 
 static char		const *input_name = "perf.data";
 
-static char		default_sort_order[] = "comm,dso,symbol";
-static char		*sort_order = default_sort_order;
 static char		*dso_list_str, *comm_list_str, *sym_list_str,
 			*col_width_list_str;
 static struct strlist	*dso_list, *comm_list, *sym_list;
-static char		*field_sep;
 
 static int		force;
-static int		input;
-static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
 static int		full_paths;
 static int		show_nr_samples;
@@ -50,374 +48,39 @@ static struct perf_read_values	show_threads_values;
 static char		default_pretty_printing_style[] = "normal";
 static char		*pretty_printing_style = default_pretty_printing_style;
 
-static unsigned long	page_size;
-static unsigned long	mmap_window = 32;
-
-static char		default_parent_pattern[] = "^sys_|^do_page_fault";
-static char		*parent_pattern = default_parent_pattern;
-static regex_t		parent_regex;
-
 static int		exclude_other = 1;
 
 static char		callchain_default_opt[] = "fractal,0.5";
 
-static int		callchain;
-
-static char		__cwd[PATH_MAX];
-static char		*cwd = __cwd;
+static char		*cwd;
 static int		cwdlen;
 
-static struct rb_root	threads;
-static struct thread	*last_match;
-
 static struct perf_header *header;
 
-static
-struct callchain_param	callchain_param = {
-	.mode	= CHAIN_GRAPH_REL,
-	.min_percent = 0.5
-};
-
 static u64		sample_type;
 
-static int repsep_fprintf(FILE *fp, const char *fmt, ...)
-{
-	int n;
-	va_list ap;
-
-	va_start(ap, fmt);
-	if (!field_sep)
-		n = vfprintf(fp, fmt, ap);
-	else {
-		char *bf = NULL;
-		n = vasprintf(&bf, fmt, ap);
-		if (n > 0) {
-			char *sep = bf;
-
-			while (1) {
-				sep = strchr(sep, *field_sep);
-				if (sep == NULL)
-					break;
-				*sep = '.';
-			}
-		}
-		fputs(bf, fp);
-		free(bf);
-	}
-	va_end(ap);
-	return n;
-}
-
-static unsigned int dsos__col_width,
-		    comms__col_width,
-		    threads__col_width;
-
-/*
- * histogram, sorted on item, collects counts
- */
-
-static struct rb_root hist;
-
-struct hist_entry {
-	struct rb_node		rb_node;
-
-	struct thread		*thread;
-	struct map		*map;
-	struct dso		*dso;
-	struct symbol		*sym;
-	struct symbol		*parent;
-	u64			ip;
-	char			level;
-	struct callchain_node	callchain;
-	struct rb_root		sorted_chain;
-
-	u64			count;
-};
-
-/*
- * configurable sorting bits
- */
-
-struct sort_entry {
-	struct list_head list;
-
-	const char *header;
-
-	int64_t (*cmp)(struct hist_entry *, struct hist_entry *);
-	int64_t (*collapse)(struct hist_entry *, struct hist_entry *);
-	size_t	(*print)(FILE *fp, struct hist_entry *, unsigned int width);
-	unsigned int *width;
-	bool	elide;
-};
-
-static int64_t cmp_null(void *l, void *r)
-{
-	if (!l && !r)
-		return 0;
-	else if (!l)
-		return -1;
-	else
-		return 1;
-}
-
-/* --sort pid */
-
-static int64_t
-sort__thread_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	return right->thread->pid - left->thread->pid;
-}
-
-static size_t
-sort__thread_print(FILE *fp, struct hist_entry *self, unsigned int width)
-{
-	return repsep_fprintf(fp, "%*s:%5d", width - 6,
-			      self->thread->comm ?: "", self->thread->pid);
-}
-
-static struct sort_entry sort_thread = {
-	.header = "Command:  Pid",
-	.cmp	= sort__thread_cmp,
-	.print	= sort__thread_print,
-	.width	= &threads__col_width,
-};
-
-/* --sort comm */
-
-static int64_t
-sort__comm_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	return right->thread->pid - left->thread->pid;
-}
-
-static int64_t
-sort__comm_collapse(struct hist_entry *left, struct hist_entry *right)
-{
-	char *comm_l = left->thread->comm;
-	char *comm_r = right->thread->comm;
-
-	if (!comm_l || !comm_r)
-		return cmp_null(comm_l, comm_r);
-
-	return strcmp(comm_l, comm_r);
-}
-
-static size_t
-sort__comm_print(FILE *fp, struct hist_entry *self, unsigned int width)
-{
-	return repsep_fprintf(fp, "%*s", width, self->thread->comm);
-}
-
-static struct sort_entry sort_comm = {
-	.header		= "Command",
-	.cmp		= sort__comm_cmp,
-	.collapse	= sort__comm_collapse,
-	.print		= sort__comm_print,
-	.width		= &comms__col_width,
-};
-
-/* --sort dso */
-
-static int64_t
-sort__dso_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	struct dso *dso_l = left->dso;
-	struct dso *dso_r = right->dso;
-
-	if (!dso_l || !dso_r)
-		return cmp_null(dso_l, dso_r);
-
-	return strcmp(dso_l->name, dso_r->name);
-}
-
-static size_t
-sort__dso_print(FILE *fp, struct hist_entry *self, unsigned int width)
-{
-	if (self->dso)
-		return repsep_fprintf(fp, "%-*s", width, self->dso->name);
-
-	return repsep_fprintf(fp, "%*llx", width, (u64)self->ip);
-}
-
-static struct sort_entry sort_dso = {
-	.header = "Shared Object",
-	.cmp	= sort__dso_cmp,
-	.print	= sort__dso_print,
-	.width	= &dsos__col_width,
-};
-
-/* --sort symbol */
-
-static int64_t
-sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	u64 ip_l, ip_r;
-
-	if (left->sym == right->sym)
-		return 0;
-
-	ip_l = left->sym ? left->sym->start : left->ip;
-	ip_r = right->sym ? right->sym->start : right->ip;
-
-	return (int64_t)(ip_r - ip_l);
-}
 
 static size_t
-sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used)
+callchain__fprintf_left_margin(FILE *fp, int left_margin)
 {
-	size_t ret = 0;
+	int i;
+	int ret;
 
-	if (verbose)
-		ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip,
-				      dso__symtab_origin(self->dso));
+	ret = fprintf(fp, "            ");
 
-	ret += repsep_fprintf(fp, "[%c] ", self->level);
-	if (self->sym) {
-		ret += repsep_fprintf(fp, "%s", self->sym->name);
-
-		if (self->sym->module)
-			ret += repsep_fprintf(fp, "\t[%s]",
-					     self->sym->module->name);
-	} else {
-		ret += repsep_fprintf(fp, "%#016llx", (u64)self->ip);
-	}
+	for (i = 0; i < left_margin; i++)
+		ret += fprintf(fp, " ");
 
 	return ret;
 }
 
-static struct sort_entry sort_sym = {
-	.header = "Symbol",
-	.cmp	= sort__sym_cmp,
-	.print	= sort__sym_print,
-};
-
-/* --sort parent */
-
-static int64_t
-sort__parent_cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	struct symbol *sym_l = left->parent;
-	struct symbol *sym_r = right->parent;
-
-	if (!sym_l || !sym_r)
-		return cmp_null(sym_l, sym_r);
-
-	return strcmp(sym_l->name, sym_r->name);
-}
-
-static size_t
-sort__parent_print(FILE *fp, struct hist_entry *self, unsigned int width)
-{
-	return repsep_fprintf(fp, "%-*s", width,
-			      self->parent ? self->parent->name : "[other]");
-}
-
-static unsigned int parent_symbol__col_width;
-
-static struct sort_entry sort_parent = {
-	.header = "Parent symbol",
-	.cmp	= sort__parent_cmp,
-	.print	= sort__parent_print,
-	.width	= &parent_symbol__col_width,
-};
-
-static int sort__need_collapse = 0;
-static int sort__has_parent = 0;
-
-struct sort_dimension {
-	const char		*name;
-	struct sort_entry	*entry;
-	int			taken;
-};
-
-static struct sort_dimension sort_dimensions[] = {
-	{ .name = "pid",	.entry = &sort_thread,	},
-	{ .name = "comm",	.entry = &sort_comm,	},
-	{ .name = "dso",	.entry = &sort_dso,	},
-	{ .name = "symbol",	.entry = &sort_sym,	},
-	{ .name = "parent",	.entry = &sort_parent,	},
-};
-
-static LIST_HEAD(hist_entry__sort_list);
-
-static int sort_dimension__add(const char *tok)
-{
-	unsigned int i;
-
-	for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) {
-		struct sort_dimension *sd = &sort_dimensions[i];
-
-		if (sd->taken)
-			continue;
-
-		if (strncasecmp(tok, sd->name, strlen(tok)))
-			continue;
-
-		if (sd->entry->collapse)
-			sort__need_collapse = 1;
-
-		if (sd->entry == &sort_parent) {
-			int ret = regcomp(&parent_regex, parent_pattern, REG_EXTENDED);
-			if (ret) {
-				char err[BUFSIZ];
-
-				regerror(ret, &parent_regex, err, sizeof(err));
-				fprintf(stderr, "Invalid regex: %s\n%s",
-					parent_pattern, err);
-				exit(-1);
-			}
-			sort__has_parent = 1;
-		}
-
-		list_add_tail(&sd->entry->list, &hist_entry__sort_list);
-		sd->taken = 1;
-
-		return 0;
-	}
-
-	return -ESRCH;
-}
-
-static int64_t
-hist_entry__cmp(struct hist_entry *left, struct hist_entry *right)
-{
-	struct sort_entry *se;
-	int64_t cmp = 0;
-
-	list_for_each_entry(se, &hist_entry__sort_list, list) {
-		cmp = se->cmp(left, right);
-		if (cmp)
-			break;
-	}
-
-	return cmp;
-}
-
-static int64_t
-hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
-{
-	struct sort_entry *se;
-	int64_t cmp = 0;
-
-	list_for_each_entry(se, &hist_entry__sort_list, list) {
-		int64_t (*f)(struct hist_entry *, struct hist_entry *);
-
-		f = se->collapse ?: se->cmp;
-
-		cmp = f(left, right);
-		if (cmp)
-			break;
-	}
-
-	return cmp;
-}
-
-static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask)
+static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask,
+					  int left_margin)
 {
 	int i;
 	size_t ret = 0;
 
-	ret += fprintf(fp, "%s", "                ");
+	ret += callchain__fprintf_left_margin(fp, left_margin);
 
 	for (i = 0; i < depth; i++)
 		if (depth_mask & (1 << i))
@@ -432,12 +95,12 @@ static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask)
 static size_t
 ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, int depth,
 		       int depth_mask, int count, u64 total_samples,
-		       int hits)
+		       int hits, int left_margin)
 {
 	int i;
 	size_t ret = 0;
 
-	ret += fprintf(fp, "%s", "                ");
+	ret += callchain__fprintf_left_margin(fp, left_margin);
 	for (i = 0; i < depth; i++) {
 		if (depth_mask & (1 << i))
 			ret += fprintf(fp, "|");
@@ -475,8 +138,9 @@ static void init_rem_hits(void)
 }
 
 static size_t
-callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
-			u64 total_samples, int depth, int depth_mask)
+__callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
+			   u64 total_samples, int depth, int depth_mask,
+			   int left_margin)
 {
 	struct rb_node *node, *next;
 	struct callchain_node *child;
@@ -517,7 +181,8 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 		 * But we keep the older depth mask for the line seperator
 		 * to keep the level link until we reach the last child
 		 */
-		ret += ipchain__fprintf_graph_line(fp, depth, depth_mask);
+		ret += ipchain__fprintf_graph_line(fp, depth, depth_mask,
+						   left_margin);
 		i = 0;
 		list_for_each_entry(chain, &child->val, list) {
 			if (chain->ip >= PERF_CONTEXT_MAX)
@@ -525,11 +190,13 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 			ret += ipchain__fprintf_graph(fp, chain, depth,
 						      new_depth_mask, i++,
 						      new_total,
-						      cumul);
+						      cumul,
+						      left_margin);
 		}
-		ret += callchain__fprintf_graph(fp, child, new_total,
-						depth + 1,
-						new_depth_mask | (1 << depth));
+		ret += __callchain__fprintf_graph(fp, child, new_total,
+						  depth + 1,
+						  new_depth_mask | (1 << depth),
+						  left_margin);
 		node = next;
 	}
 
@@ -543,12 +210,51 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 
 		ret += ipchain__fprintf_graph(fp, &rem_hits, depth,
 					      new_depth_mask, 0, new_total,
-					      remaining);
+					      remaining, left_margin);
 	}
 
 	return ret;
 }
 
+
+static size_t
+callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
+			 u64 total_samples, int left_margin)
+{
+	struct callchain_list *chain;
+	bool printed = false;
+	int i = 0;
+	int ret = 0;
+
+	list_for_each_entry(chain, &self->val, list) {
+		if (chain->ip >= PERF_CONTEXT_MAX)
+			continue;
+
+		if (!i++ && sort__first_dimension == SORT_SYM)
+			continue;
+
+		if (!printed) {
+			ret += callchain__fprintf_left_margin(fp, left_margin);
+			ret += fprintf(fp, "|\n");
+			ret += callchain__fprintf_left_margin(fp, left_margin);
+			ret += fprintf(fp, "---");
+
+			left_margin += 3;
+			printed = true;
+		} else
+			ret += callchain__fprintf_left_margin(fp, left_margin);
+
+		if (chain->sym)
+			ret += fprintf(fp, " %s\n", chain->sym->name);
+		else
+			ret += fprintf(fp, " %p\n", (void *)(long)chain->ip);
+	}
+
+	ret += __callchain__fprintf_graph(fp, self, total_samples, 1, 1, left_margin);
+
+	return ret;
+}
+
 static size_t
 callchain__fprintf_flat(FILE *fp, struct callchain_node *self,
 			u64 total_samples)
@@ -577,7 +283,7 @@ callchain__fprintf_flat(FILE *fp, struct callchain_node *self,
 
 static size_t
 hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
-			      u64 total_samples)
+			      u64 total_samples, int left_margin)
 {
 	struct rb_node *rb_node;
 	struct callchain_node *chain;
@@ -597,8 +303,8 @@ hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
 			break;
 		case CHAIN_GRAPH_ABS: /* Falldown */
 		case CHAIN_GRAPH_REL:
-			ret += callchain__fprintf_graph(fp, chain,
-							total_samples, 1, 1);
+			ret += callchain__fprintf_graph(fp, chain, total_samples,
+							left_margin);
 		case CHAIN_NONE:
 		default:
 			break;
@@ -610,7 +316,6 @@ hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
 	return ret;
 }
 
-
 static size_t
 hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
 {
@@ -644,8 +349,19 @@ hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
 
 	ret += fprintf(fp, "\n");
 
-	if (callchain)
-		hist_entry_callchain__fprintf(fp, self, total_samples);
+	if (callchain) {
+		int left_margin = 0;
+
+		if (sort__first_dimension == SORT_COMM) {
+			se = list_first_entry(&hist_entry__sort_list, typeof(*se),
+						list);
+			left_margin = se->width ? *se->width : 0;
+			left_margin -= thread__comm_len(self->thread);
+		}
+
+		hist_entry_callchain__fprintf(fp, self, total_samples,
+					      left_margin);
+	}
 
 	return ret;
 }
@@ -695,22 +411,17 @@ static int thread__set_comm_adjust(struct thread *self, const char *comm)
 
 
 static struct symbol *
-resolve_symbol(struct thread *thread, struct map **mapp,
-	       struct dso **dsop, u64 *ipp)
+resolve_symbol(struct thread *thread, struct map **mapp, u64 *ipp)
 {
-	struct dso *dso = dsop ? *dsop : NULL;
 	struct map *map = mapp ? *mapp : NULL;
 	u64 ip = *ipp;
 
-	if (!thread)
-		return NULL;
-
-	if (dso)
-		goto got_dso;
-
 	if (map)
 		goto got_map;
 
+	if (!thread)
+		return NULL;
+
 	map = thread__find_map(thread, ip);
 	if (map != NULL) {
 		/*
@@ -725,29 +436,26 @@ resolve_symbol(struct thread *thread, struct map **mapp,
 			*mapp = map;
 got_map:
 		ip = map->map_ip(map, ip);
-
-		dso = map->dso;
 	} else {
 		/*
 		 * If this is outside of all known maps,
 		 * and is a negative address, try to look it
 		 * up in the kernel dso, as it might be a
-		 * vsyscall (which executes in user-mode):
+		 * vsyscall or vdso (which executes in user-mode).
+		 *
+		 * XXX This is nasty, we should have a symbol list in
+		 * the "[vdso]" dso, but for now lets use the old
+		 * trick of looking in the whole kernel symbol list.
 		 */
 		if ((long long)ip < 0)
-		dso = kernel_dso;
+			return kernel_maps__find_symbol(ip, mapp);
 	}
-	dump_printf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
+	dump_printf(" ...... dso: %s\n",
+		    map ? map->dso->long_name : "<not found>");
 	dump_printf(" ...... map: %Lx -> %Lx\n", *ipp, ip);
 	*ipp  = ip;
 
-	if (dsop)
-		*dsop = dso;
-
-	if (!dso)
-		return NULL;
-got_dso:
-	return dso->find_symbol(dso, ip);
+	return map ? map->dso->find_symbol(map->dso, ip) : NULL;
 }
 
 static int call__match(struct symbol *sym)
@@ -758,9 +466,9 @@ static int call__match(struct symbol *sym)
 	return 0;
 }
 
-static struct symbol **
-resolve_callchain(struct thread *thread, struct map *map __used,
-		    struct ip_callchain *chain, struct hist_entry *entry)
+static struct symbol **resolve_callchain(struct thread *thread, struct map *map,
+					 struct ip_callchain *chain,
+					 struct symbol **parent)
 {
 	u64 context = PERF_CONTEXT_MAX;
 	struct symbol **syms = NULL;
@@ -776,8 +484,7 @@ resolve_callchain(struct thread *thread, struct map *map __used,
 
 	for (i = 0; i < chain->nr; i++) {
 		u64 ip = chain->ips[i];
-		struct dso *dso = NULL;
-		struct symbol *sym;
+		struct symbol *sym = NULL;
 
 		if (ip >= PERF_CONTEXT_MAX) {
 			context = ip;
@@ -786,21 +493,18 @@ resolve_callchain(struct thread *thread, struct map *map __used,
 
 		switch (context) {
 		case PERF_CONTEXT_HV:
-			dso = hypervisor_dso;
 			break;
 		case PERF_CONTEXT_KERNEL:
-			dso = kernel_dso;
+			sym = kernel_maps__find_symbol(ip, &map);
 			break;
 		default:
+			sym = resolve_symbol(thread, &map, &ip);
 			break;
 		}
 
-		sym = resolve_symbol(thread, NULL, &dso, &ip);
-
 		if (sym) {
-			if (sort__has_parent && call__match(sym) &&
-			    !entry->parent)
-				entry->parent = sym;
+			if (sort__has_parent && !*parent && call__match(sym))
+				*parent = sym;
 			if (!callchain)
 				break;
 			syms[i] = sym;
@@ -815,177 +519,35 @@ resolve_callchain(struct thread *thread, struct map *map __used,
  */
 
 static int
-hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
+hist_entry__add(struct thread *thread, struct map *map,
 		struct symbol *sym, u64 ip, struct ip_callchain *chain,
 		char level, u64 count)
 {
-	struct rb_node **p = &hist.rb_node;
-	struct rb_node *parent = NULL;
+	struct symbol **syms = NULL, *parent = NULL;
+	bool hit;
 	struct hist_entry *he;
-	struct symbol **syms = NULL;
-	struct hist_entry entry = {
-		.thread	= thread,
-		.map	= map,
-		.dso	= dso,
-		.sym	= sym,
-		.ip	= ip,
-		.level	= level,
-		.count	= count,
-		.parent = NULL,
-		.sorted_chain = RB_ROOT
-	};
-	int cmp;
 
 	if ((sort__has_parent || callchain) && chain)
-		syms = resolve_callchain(thread, map, chain, &entry);
-
-	while (*p != NULL) {
-		parent = *p;
-		he = rb_entry(parent, struct hist_entry, rb_node);
-
-		cmp = hist_entry__cmp(&entry, he);
+		syms = resolve_callchain(thread, map, chain, &parent);
 
-		if (!cmp) {
-			he->count += count;
-			if (callchain) {
-				append_chain(&he->callchain, chain, syms);
-				free(syms);
-			}
-			return 0;
-		}
+	he = __hist_entry__add(thread, map, sym, parent,
+			       ip, count, level, &hit);
+	if (he == NULL)
+		return -ENOMEM;
 
-		if (cmp < 0)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
+	if (hit)
+		he->count += count;
 
-	he = malloc(sizeof(*he));
-	if (!he)
-		return -ENOMEM;
-	*he = entry;
 	if (callchain) {
-		callchain_init(&he->callchain);
+		if (!hit)
+			callchain_init(&he->callchain);
 		append_chain(&he->callchain, chain, syms);
 		free(syms);
 	}
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &hist);
 
 	return 0;
 }
 
-static void hist_entry__free(struct hist_entry *he)
-{
-	free(he);
-}
-
-/*
- * collapse the histogram
- */
-
-static struct rb_root collapse_hists;
-
-static void collapse__insert_entry(struct hist_entry *he)
-{
-	struct rb_node **p = &collapse_hists.rb_node;
-	struct rb_node *parent = NULL;
-	struct hist_entry *iter;
-	int64_t cmp;
-
-	while (*p != NULL) {
-		parent = *p;
-		iter = rb_entry(parent, struct hist_entry, rb_node);
-
-		cmp = hist_entry__collapse(iter, he);
-
-		if (!cmp) {
-			iter->count += he->count;
-			hist_entry__free(he);
-			return;
-		}
-
-		if (cmp < 0)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &collapse_hists);
-}
-
-static void collapse__resort(void)
-{
-	struct rb_node *next;
-	struct hist_entry *n;
-
-	if (!sort__need_collapse)
-		return;
-
-	next = rb_first(&hist);
-	while (next) {
-		n = rb_entry(next, struct hist_entry, rb_node);
-		next = rb_next(&n->rb_node);
-
-		rb_erase(&n->rb_node, &hist);
-		collapse__insert_entry(n);
-	}
-}
-
-/*
- * reverse the map, sort on count.
- */
-
-static struct rb_root output_hists;
-
-static void output__insert_entry(struct hist_entry *he, u64 min_callchain_hits)
-{
-	struct rb_node **p = &output_hists.rb_node;
-	struct rb_node *parent = NULL;
-	struct hist_entry *iter;
-
-	if (callchain)
-		callchain_param.sort(&he->sorted_chain, &he->callchain,
-				      min_callchain_hits, &callchain_param);
-
-	while (*p != NULL) {
-		parent = *p;
-		iter = rb_entry(parent, struct hist_entry, rb_node);
-
-		if (he->count > iter->count)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	rb_link_node(&he->rb_node, parent, p);
-	rb_insert_color(&he->rb_node, &output_hists);
-}
-
-static void output__resort(u64 total_samples)
-{
-	struct rb_node *next;
-	struct hist_entry *n;
-	struct rb_root *tree = &hist;
-	u64 min_callchain_hits;
-
-	min_callchain_hits = total_samples * (callchain_param.min_percent / 100);
-
-	if (sort__need_collapse)
-		tree = &collapse_hists;
-
-	next = rb_first(tree);
-
-	while (next) {
-		n = rb_entry(next, struct hist_entry, rb_node);
-		next = rb_next(&n->rb_node);
-
-		rb_erase(&n->rb_node, tree);
-		output__insert_entry(n, min_callchain_hits);
-	}
-}
-
 static size_t output__fprintf(FILE *fp, u64 total_samples)
 {
 	struct hist_entry *pos;
@@ -1080,13 +642,6 @@ print_entries:
 	return ret;
 }
 
-static unsigned long total = 0,
-		     total_mmap = 0,
-		     total_comm = 0,
-		     total_fork = 0,
-		     total_unknown = 0,
-		     total_lost = 0;
-
 static int validate_chain(struct ip_callchain *chain, event_t *event)
 {
 	unsigned int chain_size;
@@ -1104,17 +659,14 @@ static int
 process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
 	char level;
-	int show = 0;
-	struct dso *dso = NULL;
-	struct thread *thread;
+	struct symbol *sym = NULL;
 	u64 ip = event->ip.ip;
 	u64 period = 1;
 	struct map *map = NULL;
 	void *more_data = event->ip.__more_data;
 	struct ip_callchain *chain = NULL;
 	int cpumode;
-
-	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->ip.pid);
 
 	if (sample_type & PERF_SAMPLE_PERIOD) {
 		period = *(u64 *)more_data;
@@ -1137,7 +689,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		dump_printf("... chain: nr:%Lu\n", chain->nr);
 
 		if (validate_chain(chain, event) < 0) {
-			eprintf("call-chain problem with event, skipping it.\n");
+			pr_debug("call-chain problem with event, "
+				 "skipping it.\n");
 			return 0;
 		}
 
@@ -1147,56 +700,49 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		}
 	}
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
-
 	if (thread == NULL) {
-		eprintf("problem processing %d event, skipping it.\n",
+		pr_debug("problem processing %d event, skipping it.\n",
 			event->header.type);
 		return -1;
 	}
 
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+
 	if (comm_list && !strlist__has_entry(comm_list, thread->comm))
 		return 0;
 
 	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
 
 	if (cpumode == PERF_RECORD_MISC_KERNEL) {
-		show = SHOW_KERNEL;
 		level = 'k';
-
-		dso = kernel_dso;
-
-		dump_printf(" ...... dso: %s\n", dso->name);
-
+		sym = kernel_maps__find_symbol(ip, &map);
+		dump_printf(" ...... dso: %s\n",
+			    map ? map->dso->long_name : "<not found>");
 	} else if (cpumode == PERF_RECORD_MISC_USER) {
-
-		show = SHOW_USER;
 		level = '.';
+		sym = resolve_symbol(thread, &map, &ip);
 
 	} else {
-		show = SHOW_HV;
 		level = 'H';
-
-		dso = hypervisor_dso;
-
 		dump_printf(" ...... dso: [hypervisor]\n");
 	}
 
-	if (show & show_mask) {
-		struct symbol *sym = resolve_symbol(thread, &map, &dso, &ip);
-
-		if (dso_list && (!dso || !dso->name ||
-				 !strlist__has_entry(dso_list, dso->name)))
-			return 0;
+	if (dso_list &&
+	    (!map || !map->dso ||
+	     !(strlist__has_entry(dso_list, map->dso->short_name) ||
+	       (map->dso->short_name != map->dso->long_name &&
+		strlist__has_entry(dso_list, map->dso->long_name)))))
+		return 0;
 
-		if (sym_list && (!sym || !strlist__has_entry(sym_list, sym->name)))
-			return 0;
+	if (sym_list && sym && !strlist__has_entry(sym_list, sym->name))
+		return 0;
 
-		if (hist_entry__add(thread, map, dso, sym, ip, chain, level, period)) {
-			eprintf("problem incrementing symbol count, skipping event\n");
-			return -1;
-		}
+	if (hist_entry__add(thread, map, sym, ip,
+			    chain, level, period)) {
+		pr_debug("problem incrementing symbol count, skipping event\n");
+		return -1;
 	}
+
 	total += period;
 
 	return 0;
@@ -1205,10 +751,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-	struct map *map = map__new(&event->mmap, cwd, cwdlen);
-
-	thread = threads__findnew(event->mmap.pid, &threads, &last_match);
+	struct map *map = map__new(&event->mmap, cwd, cwdlen, 0, NULL);
+	struct thread *thread = threads__findnew(event->mmap.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
@@ -1234,9 +778,7 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-
-	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->comm.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_COMM: %s:%d\n",
 		(void *)(offset + head),
@@ -1256,11 +798,8 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_task_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-	struct thread *parent;
-
-	thread = threads__findnew(event->fork.pid, &threads, &last_match);
-	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->fork.pid);
+	struct thread *parent = threads__findnew(event->fork.ppid);
 
 	dump_printf("%p [%p]: PERF_RECORD_%s: (%d:%d):(%d:%d)\n",
 		(void *)(offset + head),
@@ -1331,216 +870,79 @@ process_read_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
-static int
-process_event(event_t *event, unsigned long offset, unsigned long head)
-{
-	trace_event(event);
-
-	switch (event->header.type) {
-	case PERF_RECORD_SAMPLE:
-		return process_sample_event(event, offset, head);
-
-	case PERF_RECORD_MMAP:
-		return process_mmap_event(event, offset, head);
-
-	case PERF_RECORD_COMM:
-		return process_comm_event(event, offset, head);
-
-	case PERF_RECORD_FORK:
-	case PERF_RECORD_EXIT:
-		return process_task_event(event, offset, head);
-
-	case PERF_RECORD_LOST:
-		return process_lost_event(event, offset, head);
-
-	case PERF_RECORD_READ:
-		return process_read_event(event, offset, head);
-
-	/*
-	 * We dont process them right now but they are fine:
-	 */
-
-	case PERF_RECORD_THROTTLE:
-	case PERF_RECORD_UNTHROTTLE:
-		return 0;
-
-	default:
-		return -1;
-	}
-
-	return 0;
-}
-
-static int __cmd_report(void)
+static int sample_type_check(u64 type)
 {
-	int ret, rc = EXIT_FAILURE;
-	unsigned long offset = 0;
-	unsigned long head, shift;
-	struct stat input_stat;
-	struct thread *idle;
-	event_t *event;
-	uint32_t size;
-	char *buf;
-
-	idle = register_idle_thread(&threads, &last_match);
-	thread__comm_adjust(idle);
-
-	if (show_threads)
-		perf_read_values_init(&show_threads_values);
-
-	input = open(input_name, O_RDONLY);
-	if (input < 0) {
-		fprintf(stderr, " failed to open file: %s", input_name);
-		if (!strcmp(input_name, "perf.data"))
-			fprintf(stderr, "  (try 'perf record' first)");
-		fprintf(stderr, "\n");
-		exit(-1);
-	}
-
-	ret = fstat(input, &input_stat);
-	if (ret < 0) {
-		perror("failed to stat file");
-		exit(-1);
-	}
-
-	if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
-		fprintf(stderr, "file: %s not owned by current user or root\n", input_name);
-		exit(-1);
-	}
-
-	if (!input_stat.st_size) {
-		fprintf(stderr, "zero-sized file, nothing to do!\n");
-		exit(0);
-	}
-
-	header = perf_header__read(input);
-	head = header->data_offset;
-
-	sample_type = perf_header__sample_type(header);
+	sample_type = type;
 
 	if (!(sample_type & PERF_SAMPLE_CALLCHAIN)) {
 		if (sort__has_parent) {
 			fprintf(stderr, "selected --sort parent, but no"
 					" callchain data. Did you call"
 					" perf record without -g?\n");
-			exit(-1);
+			return -1;
 		}
 		if (callchain) {
 			fprintf(stderr, "selected -g but no callchain data."
 					" Did you call perf record without"
 					" -g?\n");
-			exit(-1);
+			return -1;
 		}
 	} else if (callchain_param.mode != CHAIN_NONE && !callchain) {
 			callchain = 1;
 			if (register_callchain_param(&callchain_param) < 0) {
 				fprintf(stderr, "Can't register callchain"
 						" params\n");
-				exit(-1);
+				return -1;
 			}
 	}
 
-	if (load_kernel() < 0) {
-		perror("failed to load kernel symbols");
-		return EXIT_FAILURE;
-	}
-
-	if (!full_paths) {
-		if (getcwd(__cwd, sizeof(__cwd)) == NULL) {
-			perror("failed to get the current directory");
-			return EXIT_FAILURE;
-		}
-		cwdlen = strlen(cwd);
-	} else {
-		cwd = NULL;
-		cwdlen = 0;
-	}
-
-	shift = page_size * (head / page_size);
-	offset += shift;
-	head -= shift;
-
-remap:
-	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
-			   MAP_SHARED, input, offset);
-	if (buf == MAP_FAILED) {
-		perror("failed to mmap file");
-		exit(-1);
-	}
-
-more:
-	event = (event_t *)(buf + head);
-
-	size = event->header.size;
-	if (!size)
-		size = 8;
-
-	if (head + event->header.size >= page_size * mmap_window) {
-		int munmap_ret;
-
-		shift = page_size * (head / page_size);
-
-		munmap_ret = munmap(buf, page_size * mmap_window);
-		assert(munmap_ret == 0);
-
-		offset += shift;
-		head -= shift;
-		goto remap;
-	}
-
-	size = event->header.size;
-
-	dump_printf("\n%p [%p]: event: %d\n",
-			(void *)(offset + head),
-			(void *)(long)event->header.size,
-			event->header.type);
-
-	if (!size || process_event(event, offset, head) < 0) {
-
-		dump_printf("%p [%p]: skipping unknown header type: %d\n",
-			(void *)(offset + head),
-			(void *)(long)(event->header.size),
-			event->header.type);
-
-		total_unknown++;
+	return 0;
+}
 
-		/*
-		 * assume we lost track of the stream, check alignment, and
-		 * increment a single u64 in the hope to catch on again 'soon'.
-		 */
+static struct perf_file_handler file_handler = {
+	.process_sample_event	= process_sample_event,
+	.process_mmap_event	= process_mmap_event,
+	.process_comm_event	= process_comm_event,
+	.process_exit_event	= process_task_event,
+	.process_fork_event	= process_task_event,
+	.process_lost_event	= process_lost_event,
+	.process_read_event	= process_read_event,
+	.sample_type_check	= sample_type_check,
+};
 
-		if (unlikely(head & 7))
-			head &= ~7ULL;
 
-		size = 8;
-	}
+static int __cmd_report(void)
+{
+	struct thread *idle;
+	int ret;
 
-	head += size;
+	idle = register_idle_thread();
+	thread__comm_adjust(idle);
 
-	if (offset + head >= header->data_offset + header->data_size)
-		goto done;
+	if (show_threads)
+		perf_read_values_init(&show_threads_values);
 
-	if (offset + head < (unsigned long)input_stat.st_size)
-		goto more;
+	register_perf_file_handler(&file_handler);
 
-done:
-	rc = EXIT_SUCCESS;
-	close(input);
+	ret = mmap_dispatch_perf_file(&header, input_name, force, full_paths,
+				      &cwdlen, &cwd);
+	if (ret)
+		return ret;
 
 	dump_printf("      IP events: %10ld\n", total);
 	dump_printf("    mmap events: %10ld\n", total_mmap);
 	dump_printf("    comm events: %10ld\n", total_comm);
 	dump_printf("    fork events: %10ld\n", total_fork);
 	dump_printf("    lost events: %10ld\n", total_lost);
-	dump_printf(" unknown events: %10ld\n", total_unknown);
+	dump_printf(" unknown events: %10ld\n", file_handler.total_unknown);
 
 	if (dump_trace)
 		return 0;
 
-	if (verbose >= 3)
-		threads__fprintf(stdout, &threads);
+	if (verbose > 3)
+		threads__fprintf(stdout);
 
-	if (verbose >= 2)
+	if (verbose > 2)
 		dsos__fprintf(stdout);
 
 	collapse__resort();
@@ -1550,7 +952,7 @@ done:
 	if (show_threads)
 		perf_read_values_destroy(&show_threads_values);
 
-	return rc;
+	return ret;
 }
 
 static int
@@ -1606,7 +1008,8 @@ setup:
 	return 0;
 }
 
-static const char * const report_usage[] = {
+//static const char * const report_usage[] = {
+const char * const report_usage[] = {
 	"perf report [<options>] <command>",
 	NULL
 };
@@ -1692,8 +1095,6 @@ int cmd_report(int argc, const char **argv, const char *prefix __used)
 {
 	symbol__init();
 
-	page_size = getpagesize();
-
 	argc = parse_options(argc, argv, options, report_usage, 0);
 
 	setup_sorting();
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index ce2d5be4f30..9a48d9626be 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -11,6 +11,7 @@
 #include "util/trace-event.h"
 
 #include "util/debug.h"
+#include "util/data_map.h"
 
 #include <sys/types.h>
 #include <sys/prctl.h>
@@ -20,26 +21,23 @@
 #include <math.h>
 
 static char			const *input_name = "perf.data";
-static int			input;
-static unsigned long		page_size;
-static unsigned long		mmap_window = 32;
 
 static unsigned long		total_comm = 0;
 
-static struct rb_root		threads;
-static struct thread		*last_match;
-
 static struct perf_header	*header;
 static u64			sample_type;
 
 static char			default_sort_order[] = "avg, max, switch, runtime";
 static char			*sort_order = default_sort_order;
 
+static int			profile_cpu = -1;
+
+static char			*cwd;
+static int			cwdlen;
+
 #define PR_SET_NAME		15               /* Set process name */
 #define MAX_CPUS		4096
 
-#define BUG_ON(x)		assert(!(x))
-
 static u64			run_measurement_overhead;
 static u64			sleep_measurement_overhead;
 
@@ -74,6 +72,7 @@ enum sched_event_type {
 	SCHED_EVENT_RUN,
 	SCHED_EVENT_SLEEP,
 	SCHED_EVENT_WAKEUP,
+	SCHED_EVENT_MIGRATION,
 };
 
 struct sched_atom {
@@ -398,6 +397,8 @@ process_sched_event(struct task_desc *this_task __used, struct sched_atom *atom)
 				ret = sem_post(atom->wait_sem);
 			BUG_ON(ret);
 			break;
+		case SCHED_EVENT_MIGRATION:
+			break;
 		default:
 			BUG_ON(1);
 	}
@@ -635,9 +636,7 @@ static void test_calibrations(void)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-
-	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->comm.tid);
 
 	dump_printf("%p [%p]: perf_event_comm: %s:%d\n",
 		(void *)(offset + head),
@@ -745,6 +744,22 @@ struct trace_fork_event {
 	u32 child_pid;
 };
 
+struct trace_migrate_task_event {
+	u32 size;
+
+	u16 common_type;
+	u8 common_flags;
+	u8 common_preempt_count;
+	u32 common_pid;
+	u32 common_tgid;
+
+	char comm[16];
+	u32 pid;
+
+	u32 prio;
+	u32 cpu;
+};
+
 struct trace_sched_handler {
 	void (*switch_event)(struct trace_switch_event *,
 			     struct event *,
@@ -769,6 +784,12 @@ struct trace_sched_handler {
 			   int cpu,
 			   u64 timestamp,
 			   struct thread *thread);
+
+	void (*migrate_task_event)(struct trace_migrate_task_event *,
+			   struct event *,
+			   int cpu,
+			   u64 timestamp,
+			   struct thread *thread);
 };
 
 
@@ -1058,8 +1079,8 @@ latency_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %Ld < 0 ?\n", delta);
 
 
-	sched_out = threads__findnew(switch_event->prev_pid, &threads, &last_match);
-	sched_in = threads__findnew(switch_event->next_pid, &threads, &last_match);
+	sched_out = threads__findnew(switch_event->prev_pid);
+	sched_in = threads__findnew(switch_event->next_pid);
 
 	out_events = thread_atoms_search(&atom_root, sched_out, &cmp_pid);
 	if (!out_events) {
@@ -1092,13 +1113,10 @@ latency_runtime_event(struct trace_runtime_event *runtime_event,
 		     u64 timestamp,
 		     struct thread *this_thread __used)
 {
-	struct work_atoms *atoms;
-	struct thread *thread;
+	struct thread *thread = threads__findnew(runtime_event->pid);
+	struct work_atoms *atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
 
 	BUG_ON(cpu >= MAX_CPUS || cpu < 0);
-
-	thread = threads__findnew(runtime_event->pid, &threads, &last_match);
-	atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(thread);
 		atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
@@ -1125,7 +1143,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 	if (!wakeup_event->success)
 		return;
 
-	wakee = threads__findnew(wakeup_event->pid, &threads, &last_match);
+	wakee = threads__findnew(wakeup_event->pid);
 	atoms = thread_atoms_search(&atom_root, wakee, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(wakee);
@@ -1139,7 +1157,12 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 
 	atom = list_entry(atoms->work_list.prev, struct work_atom, list);
 
-	if (atom->state != THREAD_SLEEPING)
+	/*
+	 * You WILL be missing events if you've recorded only
+	 * one CPU, or are only looking at only one, so don't
+	 * make useless noise.
+	 */
+	if (profile_cpu == -1 && atom->state != THREAD_SLEEPING)
 		nr_state_machine_bugs++;
 
 	nr_timestamps++;
@@ -1152,11 +1175,51 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 	atom->wake_up_time = timestamp;
 }
 
+static void
+latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event,
+		     struct event *__event __used,
+		     int cpu __used,
+		     u64 timestamp,
+		     struct thread *thread __used)
+{
+	struct work_atoms *atoms;
+	struct work_atom *atom;
+	struct thread *migrant;
+
+	/*
+	 * Only need to worry about migration when profiling one CPU.
+	 */
+	if (profile_cpu == -1)
+		return;
+
+	migrant = threads__findnew(migrate_task_event->pid);
+	atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid);
+	if (!atoms) {
+		thread_atoms_insert(migrant);
+		register_pid(migrant->pid, migrant->comm);
+		atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid);
+		if (!atoms)
+			die("migration-event: Internal tree error");
+		add_sched_out_event(atoms, 'R', timestamp);
+	}
+
+	BUG_ON(list_empty(&atoms->work_list));
+
+	atom = list_entry(atoms->work_list.prev, struct work_atom, list);
+	atom->sched_in_time = atom->sched_out_time = atom->wake_up_time = timestamp;
+
+	nr_timestamps++;
+
+	if (atom->sched_out_time > timestamp)
+		nr_unordered_timestamps++;
+}
+
 static struct trace_sched_handler lat_ops  = {
 	.wakeup_event		= latency_wakeup_event,
 	.switch_event		= latency_switch_event,
 	.runtime_event		= latency_runtime_event,
 	.fork_event		= latency_fork_event,
+	.migrate_task_event	= latency_migrate_task_event,
 };
 
 static void output_lat_thread(struct work_atoms *work_list)
@@ -1385,8 +1448,8 @@ map_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %Ld < 0 ?\n", delta);
 
 
-	sched_out = threads__findnew(switch_event->prev_pid, &threads, &last_match);
-	sched_in = threads__findnew(switch_event->next_pid, &threads, &last_match);
+	sched_out = threads__findnew(switch_event->prev_pid);
+	sched_in = threads__findnew(switch_event->next_pid);
 
 	curr_thread[this_cpu] = sched_in;
 
@@ -1517,6 +1580,26 @@ process_sched_exit_event(struct event *event,
 }
 
 static void
+process_sched_migrate_task_event(struct raw_event_sample *raw,
+			   struct event *event,
+			   int cpu __used,
+			   u64 timestamp __used,
+			   struct thread *thread __used)
+{
+	struct trace_migrate_task_event migrate_task_event;
+
+	FILL_COMMON_FIELDS(migrate_task_event, event, raw->data);
+
+	FILL_ARRAY(migrate_task_event, comm, event, raw->data);
+	FILL_FIELD(migrate_task_event, pid, event, raw->data);
+	FILL_FIELD(migrate_task_event, prio, event, raw->data);
+	FILL_FIELD(migrate_task_event, cpu, event, raw->data);
+
+	if (trace_handler->migrate_task_event)
+		trace_handler->migrate_task_event(&migrate_task_event, event, cpu, timestamp, thread);
+}
+
+static void
 process_raw_event(event_t *raw_event __used, void *more_data,
 		  int cpu, u64 timestamp, struct thread *thread)
 {
@@ -1539,23 +1622,24 @@ process_raw_event(event_t *raw_event __used, void *more_data,
 		process_sched_fork_event(raw, event, cpu, timestamp, thread);
 	if (!strcmp(event->name, "sched_process_exit"))
 		process_sched_exit_event(event, cpu, timestamp, thread);
+	if (!strcmp(event->name, "sched_migrate_task"))
+		process_sched_migrate_task_event(raw, event, cpu, timestamp, thread);
 }
 
 static int
 process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	char level;
-	int show = 0;
-	struct dso *dso = NULL;
 	struct thread *thread;
 	u64 ip = event->ip.ip;
 	u64 timestamp = -1;
 	u32 cpu = -1;
 	u64 period = 1;
 	void *more_data = event->ip.__more_data;
-	int cpumode;
 
-	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+	if (!(sample_type & PERF_SAMPLE_RAW))
+		return 0;
+
+	thread = threads__findnew(event->ip.pid);
 
 	if (sample_type & PERF_SAMPLE_TIME) {
 		timestamp = *(u64 *)more_data;
@@ -1581,169 +1665,60 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		(void *)(long)ip,
 		(long long)period);
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
-
 	if (thread == NULL) {
-		eprintf("problem processing %d event, skipping it.\n",
-			event->header.type);
+		pr_debug("problem processing %d event, skipping it.\n",
+			 event->header.type);
 		return -1;
 	}
 
-	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-
-	if (cpumode == PERF_RECORD_MISC_KERNEL) {
-		show = SHOW_KERNEL;
-		level = 'k';
-
-		dso = kernel_dso;
-
-		dump_printf(" ...... dso: %s\n", dso->name);
-
-	} else if (cpumode == PERF_RECORD_MISC_USER) {
-
-		show = SHOW_USER;
-		level = '.';
-
-	} else {
-		show = SHOW_HV;
-		level = 'H';
-
-		dso = hypervisor_dso;
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 
-		dump_printf(" ...... dso: [hypervisor]\n");
-	}
+	if (profile_cpu != -1 && profile_cpu != (int) cpu)
+		return 0;
 
-	if (sample_type & PERF_SAMPLE_RAW)
-		process_raw_event(event, more_data, cpu, timestamp, thread);
+	process_raw_event(event, more_data, cpu, timestamp, thread);
 
 	return 0;
 }
 
 static int
-process_event(event_t *event, unsigned long offset, unsigned long head)
+process_lost_event(event_t *event __used,
+		   unsigned long offset __used,
+		   unsigned long head __used)
 {
-	trace_event(event);
+	nr_lost_chunks++;
+	nr_lost_events += event->lost.lost;
 
-	nr_events++;
-	switch (event->header.type) {
-	case PERF_RECORD_MMAP:
-		return 0;
-	case PERF_RECORD_LOST:
-		nr_lost_chunks++;
-		nr_lost_events += event->lost.lost;
-		return 0;
-
-	case PERF_RECORD_COMM:
-		return process_comm_event(event, offset, head);
-
-	case PERF_RECORD_EXIT ... PERF_RECORD_READ:
-		return 0;
+	return 0;
+}
 
-	case PERF_RECORD_SAMPLE:
-		return process_sample_event(event, offset, head);
+static int sample_type_check(u64 type)
+{
+	sample_type = type;
 
-	case PERF_RECORD_MAX:
-	default:
+	if (!(sample_type & PERF_SAMPLE_RAW)) {
+		fprintf(stderr,
+			"No trace sample to read. Did you call perf record "
+			"without -R?");
 		return -1;
 	}
 
 	return 0;
 }
 
+static struct perf_file_handler file_handler = {
+	.process_sample_event	= process_sample_event,
+	.process_comm_event	= process_comm_event,
+	.process_lost_event	= process_lost_event,
+	.sample_type_check	= sample_type_check,
+};
+
 static int read_events(void)
 {
-	int ret, rc = EXIT_FAILURE;
-	unsigned long offset = 0;
-	unsigned long head = 0;
-	struct stat perf_stat;
-	event_t *event;
-	uint32_t size;
-	char *buf;
-
-	trace_report();
-	register_idle_thread(&threads, &last_match);
-
-	input = open(input_name, O_RDONLY);
-	if (input < 0) {
-		perror("failed to open file");
-		exit(-1);
-	}
-
-	ret = fstat(input, &perf_stat);
-	if (ret < 0) {
-		perror("failed to stat file");
-		exit(-1);
-	}
-
-	if (!perf_stat.st_size) {
-		fprintf(stderr, "zero-sized file, nothing to do!\n");
-		exit(0);
-	}
-	header = perf_header__read(input);
-	head = header->data_offset;
-	sample_type = perf_header__sample_type(header);
-
-	if (!(sample_type & PERF_SAMPLE_RAW))
-		die("No trace sample to read. Did you call perf record "
-		    "without -R?");
-
-	if (load_kernel() < 0) {
-		perror("failed to load kernel symbols");
-		return EXIT_FAILURE;
-	}
-
-remap:
-	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
-			   MAP_SHARED, input, offset);
-	if (buf == MAP_FAILED) {
-		perror("failed to mmap file");
-		exit(-1);
-	}
-
-more:
-	event = (event_t *)(buf + head);
-
-	size = event->header.size;
-	if (!size)
-		size = 8;
-
-	if (head + event->header.size >= page_size * mmap_window) {
-		unsigned long shift = page_size * (head / page_size);
-		int res;
-
-		res = munmap(buf, page_size * mmap_window);
-		assert(res == 0);
-
-		offset += shift;
-		head -= shift;
-		goto remap;
-	}
-
-	size = event->header.size;
-
-
-	if (!size || process_event(event, offset, head) < 0) {
-
-		/*
-		 * assume we lost track of the stream, check alignment, and
-		 * increment a single u64 in the hope to catch on again 'soon'.
-		 */
-
-		if (unlikely(head & 7))
-			head &= ~7ULL;
-
-		size = 8;
-	}
-
-	head += size;
-
-	if (offset + head < (unsigned long)perf_stat.st_size)
-		goto more;
-
-	rc = EXIT_SUCCESS;
-	close(input);
+	register_idle_thread();
+	register_perf_file_handler(&file_handler);
 
-	return rc;
+	return mmap_dispatch_perf_file(&header, input_name, 0, 0, &cwdlen, &cwd);
 }
 
 static void print_bad_events(void)
@@ -1883,6 +1858,8 @@ static const struct option latency_options[] = {
 		   "sort by key(s): runtime, switch, avg, max"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
+	OPT_INTEGER('C', "CPU", &profile_cpu,
+		    "CPU to profile on"),
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
 	OPT_END()
@@ -1961,7 +1938,6 @@ static int __cmd_record(int argc, const char **argv)
 int cmd_sched(int argc, const char **argv, const char *prefix __used)
 {
 	symbol__init();
-	page_size = getpagesize();
 
 	argc = parse_options(argc, argv, sched_options, sched_usage,
 			     PARSE_OPT_STOP_AT_NON_OPTION);
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 3db31e7bf17..c6df3770b87 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -50,15 +50,17 @@
 
 static struct perf_event_attr default_attrs[] = {
 
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK	},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS	},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS	},
-
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES	},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS	},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES	},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK		},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES	},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS		},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS		},
+
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS	},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES	},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES		},
 
 };
 
@@ -125,6 +127,7 @@ struct stats			event_res_stats[MAX_COUNTERS][3];
 struct stats			runtime_nsecs_stats;
 struct stats			walltime_nsecs_stats;
 struct stats			runtime_cycles_stats;
+struct stats			runtime_branches_stats;
 
 #define MATCH_EVENT(t, c, counter)			\
 	(attrs[counter].type == PERF_TYPE_##t &&	\
@@ -235,6 +238,8 @@ static void read_counter(int counter)
 		update_stats(&runtime_nsecs_stats, count[0]);
 	if (MATCH_EVENT(HARDWARE, HW_CPU_CYCLES, counter))
 		update_stats(&runtime_cycles_stats, count[0]);
+	if (MATCH_EVENT(HARDWARE, HW_BRANCH_INSTRUCTIONS, counter))
+		update_stats(&runtime_branches_stats, count[0]);
 }
 
 static int run_perf_stat(int argc __used, const char **argv)
@@ -352,6 +357,14 @@ static void abs_printout(int counter, double avg)
 			ratio = avg / total;
 
 		fprintf(stderr, " # %10.3f IPC  ", ratio);
+	} else if (MATCH_EVENT(HARDWARE, HW_BRANCH_MISSES, counter)) {
+		total = avg_stats(&runtime_branches_stats);
+
+		if (total)
+			ratio = avg * 100 / total;
+
+		fprintf(stderr, " # %10.3f %%    ", ratio);
+
 	} else {
 		total = avg_stats(&runtime_nsecs_stats);
 
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 702d8fe58fb..0a2f22261c3 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -153,6 +153,17 @@ static struct wake_event     *wake_events;
 
 struct sample_wrapper *all_samples;
 
+
+struct process_filter;
+struct process_filter {
+	char 			*name;
+	int  			pid;
+	struct process_filter 	*next;
+};
+
+static struct process_filter *process_filter;
+
+
 static struct per_pid *find_create_pid(int pid)
 {
 	struct per_pid *cursor = all_data;
@@ -763,21 +774,42 @@ static void draw_wakeups(void)
 				c = p->all;
 				while (c) {
 					if (c->Y && c->start_time <= we->time && c->end_time >= we->time) {
-						if (p->pid == we->waker) {
+						if (p->pid == we->waker && !from) {
 							from = c->Y;
-							task_from = c->comm;
+							task_from = strdup(c->comm);
 						}
-						if (p->pid == we->wakee) {
+						if (p->pid == we->wakee && !to) {
 							to = c->Y;
-							task_to = c->comm;
+							task_to = strdup(c->comm);
 						}
 					}
 					c = c->next;
 				}
+				c = p->all;
+				while (c) {
+					if (p->pid == we->waker && !from) {
+						from = c->Y;
+						task_from = strdup(c->comm);
+					}
+					if (p->pid == we->wakee && !to) {
+						to = c->Y;
+						task_to = strdup(c->comm);
+					}
+					c = c->next;
+				}
 			}
 			p = p->next;
 		}
 
+		if (!task_from) {
+			task_from = malloc(40);
+			sprintf(task_from, "[%i]", we->waker);
+		}
+		if (!task_to) {
+			task_to = malloc(40);
+			sprintf(task_to, "[%i]", we->wakee);
+		}
+
 		if (we->waker == -1)
 			svg_interrupt(we->time, to);
 		else if (from && to && abs(from - to) == 1)
@@ -785,6 +817,9 @@ static void draw_wakeups(void)
 		else
 			svg_partial_wakeline(we->time, from, task_from, to, task_to);
 		we = we->next;
+
+		free(task_from);
+		free(task_to);
 	}
 }
 
@@ -858,12 +893,89 @@ static void draw_process_bars(void)
 	}
 }
 
+static void add_process_filter(const char *string)
+{
+	struct process_filter *filt;
+	int pid;
+
+	pid = strtoull(string, NULL, 10);
+	filt = malloc(sizeof(struct process_filter));
+	if (!filt)
+		return;
+
+	filt->name = strdup(string);
+	filt->pid  = pid;
+	filt->next = process_filter;
+
+	process_filter = filt;
+}
+
+static int passes_filter(struct per_pid *p, struct per_pidcomm *c)
+{
+	struct process_filter *filt;
+	if (!process_filter)
+		return 1;
+
+	filt = process_filter;
+	while (filt) {
+		if (filt->pid && p->pid == filt->pid)
+			return 1;
+		if (strcmp(filt->name, c->comm) == 0)
+			return 1;
+		filt = filt->next;
+	}
+	return 0;
+}
+
+static int determine_display_tasks_filtered(void)
+{
+	struct per_pid *p;
+	struct per_pidcomm *c;
+	int count = 0;
+
+	p = all_data;
+	while (p) {
+		p->display = 0;
+		if (p->start_time == 1)
+			p->start_time = first_time;
+
+		/* no exit marker, task kept running to the end */
+		if (p->end_time == 0)
+			p->end_time = last_time;
+
+		c = p->all;
+
+		while (c) {
+			c->display = 0;
+
+			if (c->start_time == 1)
+				c->start_time = first_time;
+
+			if (passes_filter(p, c)) {
+				c->display = 1;
+				p->display = 1;
+				count++;
+			}
+
+			if (c->end_time == 0)
+				c->end_time = last_time;
+
+			c = c->next;
+		}
+		p = p->next;
+	}
+	return count;
+}
+
 static int determine_display_tasks(u64 threshold)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
 	int count = 0;
 
+	if (process_filter)
+		return determine_display_tasks_filtered();
+
 	p = all_data;
 	while (p) {
 		p->display = 0;
@@ -1050,12 +1162,10 @@ more:
 	size = event->header.size;
 
 	if (!size || process_event(event) < 0) {
-
-		printf("%p [%p]: skipping unknown header type: %d\n",
-			(void *)(offset + head),
-			(void *)(long)(event->header.size),
-			event->header.type);
-
+		pr_warning("%p [%p]: skipping unknown header type: %d\n",
+			   (void *)(offset + head),
+			   (void *)(long)(event->header.size),
+			   event->header.type);
 		/*
 		 * assume we lost track of the stream, check alignment, and
 		 * increment a single u64 in the hope to catch on again 'soon'.
@@ -1088,7 +1198,8 @@ done:
 
 	write_svg_file(output_name);
 
-	printf("Written %2.1f seconds of trace to %s.\n", (last_time - first_time) / 1000000000.0, output_name);
+	pr_info("Written %2.1f seconds of trace to %s.\n",
+		(last_time - first_time) / 1000000000.0, output_name);
 
 	return rc;
 }
@@ -1129,6 +1240,14 @@ static int __cmd_record(int argc, const char **argv)
 	return cmd_record(i, rec_argv, NULL);
 }
 
+static int
+parse_process(const struct option *opt __used, const char *arg, int __used unset)
+{
+	if (arg)
+		add_process_filter(arg);
+	return 0;
+}
+
 static const struct option options[] = {
 	OPT_STRING('i', "input", &input_name, "file",
 		    "input file name"),
@@ -1136,8 +1255,11 @@ static const struct option options[] = {
 		    "output file name"),
 	OPT_INTEGER('w', "width", &svg_page_width,
 		    "page width"),
-	OPT_BOOLEAN('p', "power-only", &power_only,
+	OPT_BOOLEAN('P', "power-only", &power_only,
 		    "output power data only"),
+	OPT_CALLBACK('p', "process", NULL, "process",
+		      "process selector. Pass a pid or process name.",
+		       parse_process),
 	OPT_END()
 };
 
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index 37512e93623..4a9fe228be2 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -22,6 +22,7 @@
 
 #include "util/symbol.h"
 #include "util/color.h"
+#include "util/thread.h"
 #include "util/util.h"
 #include <linux/rbtree.h>
 #include "util/parse-options.h"
@@ -54,26 +55,26 @@
 
 static int			fd[MAX_NR_CPUS][MAX_COUNTERS];
 
-static int			system_wide			=  0;
+static int			system_wide			=      0;
 
-static int			default_interval		= 100000;
+static int			default_interval		=      0;
 
-static int			count_filter			=  5;
-static int			print_entries			= 15;
+static int			count_filter			=      5;
+static int			print_entries			=     15;
 
-static int			target_pid			= -1;
-static int			inherit				=  0;
-static int			profile_cpu			= -1;
-static int			nr_cpus				=  0;
-static unsigned int		realtime_prio			=  0;
-static int			group				=  0;
+static int			target_pid			=     -1;
+static int			inherit				=      0;
+static int			profile_cpu			=     -1;
+static int			nr_cpus				=      0;
+static unsigned int		realtime_prio			=      0;
+static int			group				=      0;
 static unsigned int		page_size;
-static unsigned int		mmap_pages			= 16;
-static int			freq				=  0;
+static unsigned int		mmap_pages			=     16;
+static int			freq				=   1000; /* 1 KHz */
 
-static int			delay_secs			=  2;
-static int			zero;
-static int			dump_symtab;
+static int			delay_secs			=      2;
+static int			zero                            =      0;
+static int			dump_symtab                     =      0;
 
 /*
  * Source
@@ -86,19 +87,16 @@ struct source_line {
 	struct source_line	*next;
 };
 
-static char			*sym_filter			=  NULL;
-struct sym_entry		*sym_filter_entry		=  NULL;
-static int			sym_pcnt_filter			=  5;
-static int			sym_counter			=  0;
-static int			display_weighted		= -1;
+static char			*sym_filter			=   NULL;
+struct sym_entry		*sym_filter_entry		=   NULL;
+static int			sym_pcnt_filter			=      5;
+static int			sym_counter			=      0;
+static int			display_weighted		=     -1;
 
 /*
  * Symbols
  */
 
-static u64			min_ip;
-static u64			max_ip = -1ll;
-
 struct sym_entry {
 	struct rb_node		rb_node;
 	struct list_head	node;
@@ -106,6 +104,7 @@ struct sym_entry {
 	unsigned long		snap_count;
 	double			weight;
 	int			skip;
+	struct map		*map;
 	struct source_line	*source;
 	struct source_line	*lines;
 	struct source_line	**lines_tail;
@@ -119,12 +118,11 @@ struct sym_entry {
 static void parse_source(struct sym_entry *syme)
 {
 	struct symbol *sym;
-	struct module *module;
-	struct section *section = NULL;
+	struct map *map;
 	FILE *file;
 	char command[PATH_MAX*2];
-	const char *path = vmlinux_name;
-	u64 start, end, len;
+	const char *path;
+	u64 len;
 
 	if (!syme)
 		return;
@@ -135,27 +133,16 @@ static void parse_source(struct sym_entry *syme)
 	}
 
 	sym = (struct symbol *)(syme + 1);
-	module = sym->module;
-
-	if (module)
-		path = module->path;
-	if (!path)
-		return;
-
-	start = sym->obj_start;
-	if (!start)
-		start = sym->start;
+	map = syme->map;
+	path = map->dso->long_name;
 
-	if (module) {
-		section = module->sections->find_section(module->sections, ".text");
-		if (section)
-			start -= section->vma;
-	}
-
-	end = start + sym->end - sym->start + 1;
 	len = sym->end - sym->start;
 
-	sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s", start, end, path);
+	sprintf(command,
+		"objdump --start-address=0x%016Lx "
+			 "--stop-address=0x%016Lx -dS %s",
+		map->unmap_ip(map, sym->start),
+		map->unmap_ip(map, sym->end), path);
 
 	file = popen(command, "r");
 	if (!file)
@@ -187,13 +174,11 @@ static void parse_source(struct sym_entry *syme)
 
 		if (strlen(src->line)>8 && src->line[8] == ':') {
 			src->eip = strtoull(src->line, NULL, 16);
-			if (section)
-				src->eip += section->vma;
+			src->eip = map->unmap_ip(map, src->eip);
 		}
 		if (strlen(src->line)>8 && src->line[16] == ':') {
 			src->eip = strtoull(src->line, NULL, 16);
-			if (section)
-				src->eip += section->vma;
+			src->eip = map->unmap_ip(map, src->eip);
 		}
 	}
 	pclose(file);
@@ -245,16 +230,9 @@ static void lookup_sym_source(struct sym_entry *syme)
 	struct symbol *symbol = (struct symbol *)(syme + 1);
 	struct source_line *line;
 	char pattern[PATH_MAX];
-	char *idx;
 
 	sprintf(pattern, "<%s>:", symbol->name);
 
-	if (symbol->module) {
-		idx = strstr(pattern, "\t");
-		if (idx)
-			*idx = 0;
-	}
-
 	pthread_mutex_lock(&syme->source_lock);
 	for (line = syme->lines; line; line = line->next) {
 		if (strstr(line->line, pattern)) {
@@ -516,8 +494,8 @@ static void print_sym_table(void)
 		if (verbose)
 			printf(" - %016llx", sym->start);
 		printf(" : %s", sym->name);
-		if (sym->module)
-			printf("\t[%s]", sym->module->name);
+		if (syme->map->dso->name[0] == '[')
+			printf(" \t%s", syme->map->dso->name);
 		printf("\n");
 	}
 }
@@ -686,6 +664,8 @@ static void handle_keypress(int c)
 	switch (c) {
 		case 'd':
 			prompt_integer(&delay_secs, "Enter display delay");
+			if (delay_secs < 1)
+				delay_secs = 1;
 			break;
 		case 'e':
 			prompt_integer(&print_entries, "Enter display entries (lines)");
@@ -788,7 +768,7 @@ static const char *skip_symbols[] = {
 	NULL
 };
 
-static int symbol_filter(struct dso *self, struct symbol *sym)
+static int symbol_filter(struct map *map, struct symbol *sym)
 {
 	struct sym_entry *syme;
 	const char *name = sym->name;
@@ -810,7 +790,8 @@ static int symbol_filter(struct dso *self, struct symbol *sym)
 	    strstr(name, "_text_end"))
 		return 1;
 
-	syme = dso__sym_priv(self, sym);
+	syme = dso__sym_priv(map->dso, sym);
+	syme->map = map;
 	pthread_mutex_init(&syme->source_lock, NULL);
 	if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter))
 		sym_filter_entry = syme;
@@ -827,34 +808,14 @@ static int symbol_filter(struct dso *self, struct symbol *sym)
 
 static int parse_symbols(void)
 {
-	struct rb_node *node;
-	struct symbol  *sym;
-	int use_modules = vmlinux_name ? 1 : 0;
-
-	kernel_dso = dso__new("[kernel]", sizeof(struct sym_entry));
-	if (kernel_dso == NULL)
+	if (dsos__load_kernel(vmlinux_name, sizeof(struct sym_entry),
+			      symbol_filter, 1) <= 0)
 		return -1;
 
-	if (dso__load_kernel(kernel_dso, vmlinux_name, symbol_filter, verbose, use_modules) <= 0)
-		goto out_delete_dso;
-
-	node = rb_first(&kernel_dso->syms);
-	sym = rb_entry(node, struct symbol, rb_node);
-	min_ip = sym->start;
-
-	node = rb_last(&kernel_dso->syms);
-	sym = rb_entry(node, struct symbol, rb_node);
-	max_ip = sym->end;
-
 	if (dump_symtab)
-		dso__fprintf(kernel_dso, stderr);
+		dsos__fprintf(stderr);
 
 	return 0;
-
-out_delete_dso:
-	dso__delete(kernel_dso);
-	kernel_dso = NULL;
-	return -1;
 }
 
 /*
@@ -862,10 +823,11 @@ out_delete_dso:
  */
 static void record_ip(u64 ip, int counter)
 {
-	struct symbol *sym = dso__find_symbol(kernel_dso, ip);
+	struct map *map;
+	struct symbol *sym = kernel_maps__find_symbol(ip, &map);
 
 	if (sym != NULL) {
-		struct sym_entry *syme = dso__sym_priv(kernel_dso, sym);
+		struct sym_entry *syme = dso__sym_priv(map->dso, sym);
 
 		if (!syme->skip) {
 			syme->count[counter]++;
@@ -911,8 +873,6 @@ static unsigned int mmap_read_head(struct mmap_data *md)
 	return head;
 }
 
-struct timeval last_read, this_read;
-
 static void mmap_read_counter(struct mmap_data *md)
 {
 	unsigned int head = mmap_read_head(md);
@@ -920,8 +880,6 @@ static void mmap_read_counter(struct mmap_data *md)
 	unsigned char *data = md->base + page_size;
 	int diff;
 
-	gettimeofday(&this_read, NULL);
-
 	/*
 	 * If we're further behind than half the buffer, there's a chance
 	 * the writer will bite our tail and mess up the samples under us.
@@ -932,14 +890,7 @@ static void mmap_read_counter(struct mmap_data *md)
 	 */
 	diff = head - old;
 	if (diff > md->mask / 2 || diff < 0) {
-		struct timeval iv;
-		unsigned long msecs;
-
-		timersub(&this_read, &last_read, &iv);
-		msecs = iv.tv_sec*1000 + iv.tv_usec/1000;
-
-		fprintf(stderr, "WARNING: failed to keep up with mmap data."
-				"  Last read %lu msecs ago.\n", msecs);
+		fprintf(stderr, "WARNING: failed to keep up with mmap data.\n");
 
 		/*
 		 * head points to a known good entry, start there.
@@ -947,8 +898,6 @@ static void mmap_read_counter(struct mmap_data *md)
 		old = head;
 	}
 
-	last_read = this_read;
-
 	for (; old != head;) {
 		event_t *event = (event_t *)&data[old & md->mask];
 
@@ -1016,7 +965,13 @@ static void start_counter(int i, int counter)
 	attr = attrs + counter;
 
 	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
-	attr->freq		= freq;
+
+	if (freq) {
+		attr->sample_type	|= PERF_SAMPLE_PERIOD;
+		attr->freq		= 1;
+		attr->sample_freq	= freq;
+	}
+
 	attr->inherit		= (cpu < 0) && inherit;
 
 try_again:
@@ -1171,11 +1126,6 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 	if (argc)
 		usage_with_options(top_usage, options);
 
-	if (freq) {
-		default_interval = freq;
-		freq = 1;
-	}
-
 	/* CPU and PID are mutually exclusive */
 	if (target_pid != -1 && profile_cpu != -1) {
 		printf("WARNING: PID switch overriding CPU\n");
@@ -1192,6 +1142,19 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 	parse_symbols();
 	parse_source(sym_filter_entry);
 
+
+	/*
+	 * User specified count overrides default frequency.
+	 */
+	if (default_interval)
+		freq = 0;
+	else if (freq) {
+		default_interval = freq;
+	} else {
+		fprintf(stderr, "frequency and count are zero, aborting\n");
+		exit(EXIT_FAILURE);
+	}
+
 	/*
 	 * Fill in the ones not specifically initialized via -c:
 	 */
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 0c5e4f72f2b..e566bbe3f22 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -12,28 +12,24 @@
 #include "util/debug.h"
 
 #include "util/trace-event.h"
+#include "util/data_map.h"
 
 static char		const *input_name = "perf.data";
-static int		input;
-static unsigned long	page_size;
-static unsigned long	mmap_window = 32;
 
 static unsigned long	total = 0;
 static unsigned long	total_comm = 0;
 
-static struct rb_root	threads;
-static struct thread	*last_match;
-
 static struct perf_header *header;
 static u64		sample_type;
 
+static char		*cwd;
+static int		cwdlen;
+
 
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-
-	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->comm.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_COMM: %s:%d\n",
 		(void *)(offset + head),
@@ -53,18 +49,12 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	char level;
-	int show = 0;
-	struct dso *dso = NULL;
-	struct thread *thread;
 	u64 ip = event->ip.ip;
 	u64 timestamp = -1;
 	u32 cpu = -1;
 	u64 period = 1;
 	void *more_data = event->ip.__more_data;
-	int cpumode;
-
-	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->ip.pid);
 
 	if (sample_type & PERF_SAMPLE_TIME) {
 		timestamp = *(u64 *)more_data;
@@ -90,37 +80,13 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		(void *)(long)ip,
 		(long long)period);
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
-
 	if (thread == NULL) {
-		eprintf("problem processing %d event, skipping it.\n",
-			event->header.type);
+		pr_debug("problem processing %d event, skipping it.\n",
+			 event->header.type);
 		return -1;
 	}
 
-	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-
-	if (cpumode == PERF_RECORD_MISC_KERNEL) {
-		show = SHOW_KERNEL;
-		level = 'k';
-
-		dso = kernel_dso;
-
-		dump_printf(" ...... dso: %s\n", dso->name);
-
-	} else if (cpumode == PERF_RECORD_MISC_USER) {
-
-		show = SHOW_USER;
-		level = '.';
-
-	} else {
-		show = SHOW_HV;
-		level = 'H';
-
-		dso = hypervisor_dso;
-
-		dump_printf(" ...... dso: [hypervisor]\n");
-	}
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 
 	if (sample_type & PERF_SAMPLE_RAW) {
 		struct {
@@ -140,121 +106,32 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
-static int
-process_event(event_t *event, unsigned long offset, unsigned long head)
+static int sample_type_check(u64 type)
 {
-	trace_event(event);
-
-	switch (event->header.type) {
-	case PERF_RECORD_MMAP ... PERF_RECORD_LOST:
-		return 0;
-
-	case PERF_RECORD_COMM:
-		return process_comm_event(event, offset, head);
+	sample_type = type;
 
-	case PERF_RECORD_EXIT ... PERF_RECORD_READ:
-		return 0;
-
-	case PERF_RECORD_SAMPLE:
-		return process_sample_event(event, offset, head);
-
-	case PERF_RECORD_MAX:
-	default:
+	if (!(sample_type & PERF_SAMPLE_RAW)) {
+		fprintf(stderr,
+			"No trace sample to read. Did you call perf record "
+			"without -R?");
 		return -1;
 	}
 
 	return 0;
 }
 
+static struct perf_file_handler file_handler = {
+	.process_sample_event	= process_sample_event,
+	.process_comm_event	= process_comm_event,
+	.sample_type_check	= sample_type_check,
+};
+
 static int __cmd_trace(void)
 {
-	int ret, rc = EXIT_FAILURE;
-	unsigned long offset = 0;
-	unsigned long head = 0;
-	struct stat perf_stat;
-	event_t *event;
-	uint32_t size;
-	char *buf;
-
-	trace_report();
-	register_idle_thread(&threads, &last_match);
-
-	input = open(input_name, O_RDONLY);
-	if (input < 0) {
-		perror("failed to open file");
-		exit(-1);
-	}
-
-	ret = fstat(input, &perf_stat);
-	if (ret < 0) {
-		perror("failed to stat file");
-		exit(-1);
-	}
-
-	if (!perf_stat.st_size) {
-		fprintf(stderr, "zero-sized file, nothing to do!\n");
-		exit(0);
-	}
-	header = perf_header__read(input);
-	head = header->data_offset;
-	sample_type = perf_header__sample_type(header);
-
-	if (!(sample_type & PERF_SAMPLE_RAW))
-		die("No trace sample to read. Did you call perf record "
-		    "without -R?");
-
-	if (load_kernel() < 0) {
-		perror("failed to load kernel symbols");
-		return EXIT_FAILURE;
-	}
-
-remap:
-	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
-			   MAP_SHARED, input, offset);
-	if (buf == MAP_FAILED) {
-		perror("failed to mmap file");
-		exit(-1);
-	}
-
-more:
-	event = (event_t *)(buf + head);
-
-	if (head + event->header.size >= page_size * mmap_window) {
-		unsigned long shift = page_size * (head / page_size);
-		int res;
-
-		res = munmap(buf, page_size * mmap_window);
-		assert(res == 0);
-
-		offset += shift;
-		head -= shift;
-		goto remap;
-	}
-
-	size = event->header.size;
-
-	if (!size || process_event(event, offset, head) < 0) {
-
-		/*
-		 * assume we lost track of the stream, check alignment, and
-		 * increment a single u64 in the hope to catch on again 'soon'.
-		 */
-
-		if (unlikely(head & 7))
-			head &= ~7ULL;
-
-		size = 8;
-	}
-
-	head += size;
-
-	if (offset + head < (unsigned long)perf_stat.st_size)
-		goto more;
-
-	rc = EXIT_SUCCESS;
-	close(input);
+	register_idle_thread();
+	register_perf_file_handler(&file_handler);
 
-	return rc;
+	return mmap_dispatch_perf_file(&header, input_name, 0, 0, &cwdlen, &cwd);
 }
 
 static const char * const annotate_usage[] = {
@@ -267,13 +144,14 @@ static const struct option options[] = {
 		    "dump raw trace in ASCII"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
+	OPT_BOOLEAN('l', "latency", &latency_format,
+		    "show latency attributes (irqs/preemption disabled, etc)"),
 	OPT_END()
 };
 
 int cmd_trace(int argc, const char **argv, const char *prefix __used)
 {
 	symbol__init();
-	page_size = getpagesize();
 
 	argc = parse_options(argc, argv, options, annotate_usage, 0);
 	if (argc) {
diff --git a/tools/perf/perf.c b/tools/perf/perf.c
index c570d177d5c..9cafe546326 100644
--- a/tools/perf/perf.c
+++ b/tools/perf/perf.c
@@ -89,8 +89,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 		/*
 		 * Check remaining flags.
 		 */
-		if (!prefixcmp(cmd, "--exec-path")) {
-			cmd += 11;
+		if (!prefixcmp(cmd, CMD_EXEC_PATH)) {
+			cmd += strlen(CMD_EXEC_PATH);
 			if (*cmd == '=')
 				perf_set_argv_exec_path(cmd + 1);
 			else {
@@ -117,8 +117,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 			(*argv)++;
 			(*argc)--;
 			handled++;
-		} else if (!prefixcmp(cmd, "--perf-dir=")) {
-			setenv(PERF_DIR_ENVIRONMENT, cmd + 10, 1);
+		} else if (!prefixcmp(cmd, CMD_PERF_DIR)) {
+			setenv(PERF_DIR_ENVIRONMENT, cmd + strlen(CMD_PERF_DIR), 1);
 			if (envchanged)
 				*envchanged = 1;
 		} else if (!strcmp(cmd, "--work-tree")) {
@@ -131,8 +131,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 				*envchanged = 1;
 			(*argv)++;
 			(*argc)--;
-		} else if (!prefixcmp(cmd, "--work-tree=")) {
-			setenv(PERF_WORK_TREE_ENVIRONMENT, cmd + 12, 1);
+		} else if (!prefixcmp(cmd, CMD_WORK_TREE)) {
+			setenv(PERF_WORK_TREE_ENVIRONMENT, cmd + strlen(CMD_WORK_TREE), 1);
 			if (envchanged)
 				*envchanged = 1;
 		} else if (!strcmp(cmd, "--debugfs-dir")) {
@@ -146,8 +146,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 				*envchanged = 1;
 			(*argv)++;
 			(*argc)--;
-		} else if (!prefixcmp(cmd, "--debugfs-dir=")) {
-			strncpy(debugfs_mntpt, cmd + 14, MAXPATHLEN);
+		} else if (!prefixcmp(cmd, CMD_DEBUGFS_DIR)) {
+			strncpy(debugfs_mntpt, cmd + strlen(CMD_DEBUGFS_DIR), MAXPATHLEN);
 			debugfs_mntpt[MAXPATHLEN - 1] = '\0';
 			if (envchanged)
 				*envchanged = 1;
diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN
index c561d1538c0..54552a00a11 100755
--- a/tools/perf/util/PERF-VERSION-GEN
+++ b/tools/perf/util/PERF-VERSION-GEN
@@ -1,7 +1,7 @@
 #!/bin/sh
 
 GVF=PERF-VERSION-FILE
-DEF_VER=v0.0.1.PERF
+DEF_VER=v0.0.2.PERF
 
 LF='
 '
diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h
index 6f8ea9d210b..918eb376abe 100644
--- a/tools/perf/util/cache.h
+++ b/tools/perf/util/cache.h
@@ -1,10 +1,15 @@
-#ifndef CACHE_H
-#define CACHE_H
+#ifndef __PERF_CACHE_H
+#define __PERF_CACHE_H
 
 #include "util.h"
 #include "strbuf.h"
 #include "../perf.h"
 
+#define CMD_EXEC_PATH "--exec-path"
+#define CMD_PERF_DIR "--perf-dir="
+#define CMD_WORK_TREE "--work-tree="
+#define CMD_DEBUGFS_DIR "--debugfs-dir="
+
 #define PERF_DIR_ENVIRONMENT "PERF_DIR"
 #define PERF_WORK_TREE_ENVIRONMENT "PERF_WORK_TREE"
 #define DEFAULT_PERF_DIR_ENVIRONMENT ".perf"
@@ -117,4 +122,4 @@ extern char *perf_pathdup(const char *fmt, ...)
 
 extern size_t strlcpy(char *dest, const char *src, size_t size);
 
-#endif /* CACHE_H */
+#endif /* __PERF_CACHE_H */
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index 3b8380f1b47..b3b71258272 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -206,7 +206,7 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain,
 	}
 	node->val_nr = chain->nr - start;
 	if (!node->val_nr)
-		printf("Warning: empty node in callchain tree\n");
+		pr_warning("Warning: empty node in callchain tree\n");
 }
 
 static void
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
index 43cf3ea9e08..ad4626de4c2 100644
--- a/tools/perf/util/callchain.h
+++ b/tools/perf/util/callchain.h
@@ -58,4 +58,4 @@ static inline u64 cumul_hits(struct callchain_node *node)
 int register_callchain_param(struct callchain_param *param);
 void append_chain(struct callchain_node *root, struct ip_callchain *chain,
 		  struct symbol **syms);
-#endif
+#endif	/* __PERF_CALLCHAIN_H */
diff --git a/tools/perf/util/color.h b/tools/perf/util/color.h
index 58d597564b9..24e8809210b 100644
--- a/tools/perf/util/color.h
+++ b/tools/perf/util/color.h
@@ -1,5 +1,5 @@
-#ifndef COLOR_H
-#define COLOR_H
+#ifndef __PERF_COLOR_H
+#define __PERF_COLOR_H
 
 /* "\033[1;38;5;2xx;48;5;2xxm\0" is 23 bytes */
 #define COLOR_MAXLEN 24
@@ -39,4 +39,4 @@ int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *bu
 int percent_color_fprintf(FILE *fp, const char *fmt, double percent);
 const char *get_percent_color(double percent);
 
-#endif /* COLOR_H */
+#endif /* __PERF_COLOR_H */
diff --git a/tools/perf/util/data_map.c b/tools/perf/util/data_map.c
new file mode 100644
index 00000000000..18accb8fee4
--- /dev/null
+++ b/tools/perf/util/data_map.c
@@ -0,0 +1,222 @@
+#include "data_map.h"
+#include "symbol.h"
+#include "util.h"
+#include "debug.h"
+
+
+static struct perf_file_handler *curr_handler;
+static unsigned long	mmap_window = 32;
+static char		__cwd[PATH_MAX];
+
+static int
+process_event_stub(event_t *event __used,
+		   unsigned long offset __used,
+		   unsigned long head __used)
+{
+	return 0;
+}
+
+void register_perf_file_handler(struct perf_file_handler *handler)
+{
+	if (!handler->process_sample_event)
+		handler->process_sample_event = process_event_stub;
+	if (!handler->process_mmap_event)
+		handler->process_mmap_event = process_event_stub;
+	if (!handler->process_comm_event)
+		handler->process_comm_event = process_event_stub;
+	if (!handler->process_fork_event)
+		handler->process_fork_event = process_event_stub;
+	if (!handler->process_exit_event)
+		handler->process_exit_event = process_event_stub;
+	if (!handler->process_lost_event)
+		handler->process_lost_event = process_event_stub;
+	if (!handler->process_read_event)
+		handler->process_read_event = process_event_stub;
+	if (!handler->process_throttle_event)
+		handler->process_throttle_event = process_event_stub;
+	if (!handler->process_unthrottle_event)
+		handler->process_unthrottle_event = process_event_stub;
+
+	curr_handler = handler;
+}
+
+static int
+process_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	trace_event(event);
+
+	switch (event->header.type) {
+	case PERF_RECORD_SAMPLE:
+		return curr_handler->process_sample_event(event, offset, head);
+	case PERF_RECORD_MMAP:
+		return curr_handler->process_mmap_event(event, offset, head);
+	case PERF_RECORD_COMM:
+		return curr_handler->process_comm_event(event, offset, head);
+	case PERF_RECORD_FORK:
+		return curr_handler->process_fork_event(event, offset, head);
+	case PERF_RECORD_EXIT:
+		return curr_handler->process_exit_event(event, offset, head);
+	case PERF_RECORD_LOST:
+		return curr_handler->process_lost_event(event, offset, head);
+	case PERF_RECORD_READ:
+		return curr_handler->process_read_event(event, offset, head);
+	case PERF_RECORD_THROTTLE:
+		return curr_handler->process_throttle_event(event, offset, head);
+	case PERF_RECORD_UNTHROTTLE:
+		return curr_handler->process_unthrottle_event(event, offset, head);
+	default:
+		curr_handler->total_unknown++;
+		return -1;
+	}
+}
+
+int mmap_dispatch_perf_file(struct perf_header **pheader,
+			    const char *input_name,
+			    int force,
+			    int full_paths,
+			    int *cwdlen,
+			    char **cwd)
+{
+	int ret, rc = EXIT_FAILURE;
+	struct perf_header *header;
+	unsigned long head, shift;
+	unsigned long offset = 0;
+	struct stat input_stat;
+	size_t	page_size;
+	u64 sample_type;
+	event_t *event;
+	uint32_t size;
+	int input;
+	char *buf;
+
+	if (!curr_handler)
+		die("Forgot to register perf file handler");
+
+	page_size = getpagesize();
+
+	input = open(input_name, O_RDONLY);
+	if (input < 0) {
+		fprintf(stderr, " failed to open file: %s", input_name);
+		if (!strcmp(input_name, "perf.data"))
+			fprintf(stderr, "  (try 'perf record' first)");
+		fprintf(stderr, "\n");
+		exit(-1);
+	}
+
+	ret = fstat(input, &input_stat);
+	if (ret < 0) {
+		perror("failed to stat file");
+		exit(-1);
+	}
+
+	if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
+		fprintf(stderr, "file: %s not owned by current user or root\n",
+			input_name);
+		exit(-1);
+	}
+
+	if (!input_stat.st_size) {
+		fprintf(stderr, "zero-sized file, nothing to do!\n");
+		exit(0);
+	}
+
+	*pheader = perf_header__read(input);
+	header = *pheader;
+	head = header->data_offset;
+
+	sample_type = perf_header__sample_type(header);
+
+	if (curr_handler->sample_type_check)
+		if (curr_handler->sample_type_check(sample_type) < 0)
+			exit(-1);
+
+	if (load_kernel(0, NULL) < 0) {
+		perror("failed to load kernel symbols");
+		return EXIT_FAILURE;
+	}
+
+	if (!full_paths) {
+		if (getcwd(__cwd, sizeof(__cwd)) == NULL) {
+			perror("failed to get the current directory");
+			return EXIT_FAILURE;
+		}
+		*cwd = __cwd;
+		*cwdlen = strlen(*cwd);
+	} else {
+		*cwd = NULL;
+		*cwdlen = 0;
+	}
+
+	shift = page_size * (head / page_size);
+	offset += shift;
+	head -= shift;
+
+remap:
+	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
+			   MAP_SHARED, input, offset);
+	if (buf == MAP_FAILED) {
+		perror("failed to mmap file");
+		exit(-1);
+	}
+
+more:
+	event = (event_t *)(buf + head);
+
+	size = event->header.size;
+	if (!size)
+		size = 8;
+
+	if (head + event->header.size >= page_size * mmap_window) {
+		int munmap_ret;
+
+		shift = page_size * (head / page_size);
+
+		munmap_ret = munmap(buf, page_size * mmap_window);
+		assert(munmap_ret == 0);
+
+		offset += shift;
+		head -= shift;
+		goto remap;
+	}
+
+	size = event->header.size;
+
+	dump_printf("\n%p [%p]: event: %d\n",
+			(void *)(offset + head),
+			(void *)(long)event->header.size,
+			event->header.type);
+
+	if (!size || process_event(event, offset, head) < 0) {
+
+		dump_printf("%p [%p]: skipping unknown header type: %d\n",
+			(void *)(offset + head),
+			(void *)(long)(event->header.size),
+			event->header.type);
+
+		/*
+		 * assume we lost track of the stream, check alignment, and
+		 * increment a single u64 in the hope to catch on again 'soon'.
+		 */
+
+		if (unlikely(head & 7))
+			head &= ~7ULL;
+
+		size = 8;
+	}
+
+	head += size;
+
+	if (offset + head >= header->data_offset + header->data_size)
+		goto done;
+
+	if (offset + head < (unsigned long)input_stat.st_size)
+		goto more;
+
+done:
+	rc = EXIT_SUCCESS;
+	close(input);
+
+	return rc;
+}
+
+
diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h
new file mode 100644
index 00000000000..716d1053b07
--- /dev/null
+++ b/tools/perf/util/data_map.h
@@ -0,0 +1,31 @@
+#ifndef __PERF_DATAMAP_H
+#define __PERF_DATAMAP_H
+
+#include "event.h"
+#include "header.h"
+
+typedef int (*event_type_handler_t)(event_t *, unsigned long, unsigned long);
+
+struct perf_file_handler {
+	event_type_handler_t	process_sample_event;
+	event_type_handler_t	process_mmap_event;
+	event_type_handler_t	process_comm_event;
+	event_type_handler_t	process_fork_event;
+	event_type_handler_t	process_exit_event;
+	event_type_handler_t	process_lost_event;
+	event_type_handler_t	process_read_event;
+	event_type_handler_t	process_throttle_event;
+	event_type_handler_t	process_unthrottle_event;
+	int			(*sample_type_check)(u64 sample_type);
+	unsigned long		total_unknown;
+};
+
+void register_perf_file_handler(struct perf_file_handler *handler);
+int mmap_dispatch_perf_file(struct perf_header **pheader,
+			    const char *input_name,
+			    int force,
+			    int full_paths,
+			    int *cwdlen,
+			    char **cwd);
+
+#endif
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index e8ca98fe0bd..28d520d5a1f 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -13,12 +13,12 @@
 int verbose = 0;
 int dump_trace = 0;
 
-int eprintf(const char *fmt, ...)
+int eprintf(int level, const char *fmt, ...)
 {
 	va_list args;
 	int ret = 0;
 
-	if (verbose) {
+	if (verbose >= level) {
 		va_start(args, fmt);
 		ret = vfprintf(stderr, fmt, args);
 		va_end(args);
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index 437eea58ce4..e8b18a1f87a 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -1,8 +1,13 @@
 /* For debugging general purposes */
+#ifndef __PERF_DEBUG_H
+#define __PERF_DEBUG_H
 
 extern int verbose;
 extern int dump_trace;
 
-int eprintf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+int eprintf(int level,
+	    const char *fmt, ...) __attribute__((format(printf, 2, 3)));
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
 void trace_event(event_t *event);
+
+#endif	/* __PERF_DEBUG_H */
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index 2c9c26d6ded..d972b4b0d38 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -1,14 +1,10 @@
 #ifndef __PERF_RECORD_H
 #define __PERF_RECORD_H
+
 #include "../perf.h"
 #include "util.h"
 #include <linux/list.h>
-
-enum {
-	SHOW_KERNEL	= 1,
-	SHOW_USER	= 2,
-	SHOW_HV		= 4,
-};
+#include <linux/rbtree.h>
 
 /*
  * PERF_SAMPLE_IP | PERF_SAMPLE_TID | *
@@ -78,11 +74,15 @@ typedef union event_union {
 } event_t;
 
 struct map {
-	struct list_head	node;
+	union {
+		struct rb_node	rb_node;
+		struct list_head node;
+	};
 	u64			start;
 	u64			end;
 	u64			pgoff;
 	u64			(*map_ip)(struct map *, u64);
+	u64			(*unmap_ip)(struct map *, u64);
 	struct dso		*dso;
 };
 
@@ -91,14 +91,24 @@ static inline u64 map__map_ip(struct map *map, u64 ip)
 	return ip - map->start + map->pgoff;
 }
 
-static inline u64 vdso__map_ip(struct map *map __used, u64 ip)
+static inline u64 map__unmap_ip(struct map *map, u64 ip)
+{
+	return ip + map->start - map->pgoff;
+}
+
+static inline u64 identity__map_ip(struct map *map __used, u64 ip)
 {
 	return ip;
 }
 
-struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen);
+struct symbol;
+
+typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym);
+
+struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen,
+		     unsigned int sym_priv_size, symbol_filter_t filter);
 struct map *map__clone(struct map *self);
 int map__overlap(struct map *l, struct map *r);
 size_t map__fprintf(struct map *self, FILE *fp);
 
-#endif
+#endif /* __PERF_RECORD_H */
diff --git a/tools/perf/util/exec_cmd.h b/tools/perf/util/exec_cmd.h
index effe25eb154..31647ac92ed 100644
--- a/tools/perf/util/exec_cmd.h
+++ b/tools/perf/util/exec_cmd.h
@@ -1,5 +1,5 @@
-#ifndef PERF_EXEC_CMD_H
-#define PERF_EXEC_CMD_H
+#ifndef __PERF_EXEC_CMD_H
+#define __PERF_EXEC_CMD_H
 
 extern void perf_set_argv_exec_path(const char *exec_path);
 extern const char *perf_extract_argv0_path(const char *path);
@@ -10,4 +10,4 @@ extern int execv_perf_cmd(const char **argv); /* NULL terminated */
 extern int execl_perf_cmd(const char *cmd, ...);
 extern const char *system_path(const char *path);
 
-#endif /* PERF_EXEC_CMD_H */
+#endif /* __PERF_EXEC_CMD_H */
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index e306857b2c2..7d26659b806 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -5,6 +5,8 @@
 
 #include "util.h"
 #include "header.h"
+#include "../perf.h"
+#include "trace-event.h"
 
 /*
  * Create new perf.data header attribute:
@@ -46,23 +48,17 @@ void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
  */
 struct perf_header *perf_header__new(void)
 {
-	struct perf_header *self = malloc(sizeof(*self));
+	struct perf_header *self = calloc(sizeof(*self), 1);
 
 	if (!self)
 		die("nomem");
 
-	self->frozen = 0;
-
-	self->attrs = 0;
 	self->size = 1;
 	self->attr = malloc(sizeof(void *));
 
 	if (!self->attr)
 		die("nomem");
 
-	self->data_offset = 0;
-	self->data_size = 0;
-
 	return self;
 }
 
@@ -97,7 +93,7 @@ static struct perf_trace_event_type *events;
 void perf_header__push_event(u64 id, const char *name)
 {
 	if (strlen(name) > MAX_EVENT_NAME)
-		printf("Event %s will be truncated\n", name);
+		pr_warning("Event %s will be truncated\n", name);
 
 	if (!events) {
 		events = malloc(sizeof(struct perf_trace_event_type));
@@ -145,8 +141,14 @@ struct perf_file_header {
 	struct perf_file_section	attrs;
 	struct perf_file_section	data;
 	struct perf_file_section	event_types;
+	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
 };
 
+void perf_header__feat_trace_info(struct perf_header *header)
+{
+	set_bit(HEADER_TRACE_INFO, header->adds_features);
+}
+
 static void do_write(int fd, void *buf, size_t size)
 {
 	while (size) {
@@ -160,6 +162,32 @@ static void do_write(int fd, void *buf, size_t size)
 	}
 }
 
+static void perf_header__adds_write(struct perf_header *self, int fd)
+{
+	struct perf_file_section trace_sec;
+	u64 cur_offset = lseek(fd, 0, SEEK_CUR);
+	unsigned long *feat_mask = self->adds_features;
+
+	if (test_bit(HEADER_TRACE_INFO, feat_mask)) {
+		/* Write trace info */
+		trace_sec.offset = lseek(fd, sizeof(trace_sec), SEEK_CUR);
+		read_tracing_data(fd, attrs, nr_counters);
+		trace_sec.size = lseek(fd, 0, SEEK_CUR) - trace_sec.offset;
+
+		/* Write trace info headers */
+		lseek(fd, cur_offset, SEEK_SET);
+		do_write(fd, &trace_sec, sizeof(trace_sec));
+
+		/*
+		 * Update cur_offset. So that other (future)
+		 * features can set their own infos in this place. But if we are
+		 * the only feature, at least that seeks to the place the data
+		 * should begin.
+		 */
+		cur_offset = lseek(fd, trace_sec.offset + trace_sec.size, SEEK_SET);
+	}
+};
+
 void perf_header__write(struct perf_header *self, int fd)
 {
 	struct perf_file_header f_header;
@@ -198,6 +226,7 @@ void perf_header__write(struct perf_header *self, int fd)
 	if (events)
 		do_write(fd, events, self->event_size);
 
+	perf_header__adds_write(self, fd);
 
 	self->data_offset = lseek(fd, 0, SEEK_CUR);
 
@@ -219,6 +248,8 @@ void perf_header__write(struct perf_header *self, int fd)
 		},
 	};
 
+	memcpy(&f_header.adds_features, &self->adds_features, sizeof(self->adds_features));
+
 	lseek(fd, 0, SEEK_SET);
 	do_write(fd, &f_header, sizeof(f_header));
 	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
@@ -241,6 +272,20 @@ static void do_read(int fd, void *buf, size_t size)
 	}
 }
 
+static void perf_header__adds_read(struct perf_header *self, int fd)
+{
+	const unsigned long *feat_mask = self->adds_features;
+
+	if (test_bit(HEADER_TRACE_INFO, feat_mask)) {
+		struct perf_file_section trace_sec;
+
+		do_read(fd, &trace_sec, sizeof(trace_sec));
+		lseek(fd, trace_sec.offset, SEEK_SET);
+		trace_report(fd);
+		lseek(fd, trace_sec.offset + trace_sec.size, SEEK_SET);
+	}
+};
+
 struct perf_header *perf_header__read(int fd)
 {
 	struct perf_header	*self = perf_header__new();
@@ -254,10 +299,16 @@ struct perf_header *perf_header__read(int fd)
 	do_read(fd, &f_header, sizeof(f_header));
 
 	if (f_header.magic	!= PERF_MAGIC		||
-	    f_header.size	!= sizeof(f_header)	||
 	    f_header.attr_size	!= sizeof(f_attr))
 		die("incompatible file format");
 
+	if (f_header.size != sizeof(f_header)) {
+		/* Support the previous format */
+		if (f_header.size == offsetof(typeof(f_header), adds_features))
+			bitmap_zero(f_header.adds_features, HEADER_FEAT_BITS);
+		else
+			die("incompatible file format");
+	}
 	nr_attrs = f_header.attrs.size / sizeof(f_attr);
 	lseek(fd, f_header.attrs.offset, SEEK_SET);
 
@@ -290,6 +341,11 @@ struct perf_header *perf_header__read(int fd)
 		do_read(fd, events, f_header.event_types.size);
 		event_count =  f_header.event_types.size / sizeof(struct perf_trace_event_type);
 	}
+
+	memcpy(&self->adds_features, &f_header.adds_features, sizeof(f_header.adds_features));
+
+	perf_header__adds_read(self, fd);
+
 	self->event_offset = f_header.event_types.offset;
 	self->event_size   = f_header.event_types.size;
 
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
index a0761bc7863..2ea9dfb1236 100644
--- a/tools/perf/util/header.h
+++ b/tools/perf/util/header.h
@@ -1,10 +1,12 @@
-#ifndef _PERF_HEADER_H
-#define _PERF_HEADER_H
+#ifndef __PERF_HEADER_H
+#define __PERF_HEADER_H
 
 #include "../../../include/linux/perf_event.h"
 #include <sys/types.h>
 #include "types.h"
 
+#include <linux/bitmap.h>
+
 struct perf_header_attr {
 	struct perf_event_attr attr;
 	int ids, size;
@@ -12,15 +14,20 @@ struct perf_header_attr {
 	off_t id_offset;
 };
 
+#define HEADER_TRACE_INFO			1
+
+#define HEADER_FEAT_BITS			256
+
 struct perf_header {
-	int frozen;
-	int attrs, size;
+	int			frozen;
+	int			attrs, size;
 	struct perf_header_attr **attr;
-	s64 attr_offset;
-	u64 data_offset;
-	u64 data_size;
-	u64 event_offset;
-	u64 event_size;
+	s64			attr_offset;
+	u64			data_offset;
+	u64			data_size;
+	u64			event_offset;
+	u64			event_size;
+	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
 };
 
 struct perf_header *perf_header__read(int fd);
@@ -40,8 +47,8 @@ void perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
 u64 perf_header__sample_type(struct perf_header *header);
 struct perf_event_attr *
 perf_header__find_attr(u64 id, struct perf_header *header);
-
+void perf_header__feat_trace_info(struct perf_header *header);
 
 struct perf_header *perf_header__new(void);
 
-#endif /* _PERF_HEADER_H */
+#endif /* __PERF_HEADER_H */
diff --git a/tools/perf/util/help.h b/tools/perf/util/help.h
index 7128783637b..7f5c6dedd71 100644
--- a/tools/perf/util/help.h
+++ b/tools/perf/util/help.h
@@ -1,5 +1,5 @@
-#ifndef HELP_H
-#define HELP_H
+#ifndef __PERF_HELP_H
+#define __PERF_HELP_H
 
 struct cmdnames {
 	size_t alloc;
@@ -26,4 +26,4 @@ int is_in_cmdlist(struct cmdnames *c, const char *s);
 void list_commands(const char *title, struct cmdnames *main_cmds,
 		   struct cmdnames *other_cmds);
 
-#endif /* HELP_H */
+#endif /* __PERF_HELP_H */
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
new file mode 100644
index 00000000000..7393a02fd8d
--- /dev/null
+++ b/tools/perf/util/hist.c
@@ -0,0 +1,210 @@
+#include "hist.h"
+
+struct rb_root hist;
+struct rb_root collapse_hists;
+struct rb_root output_hists;
+int callchain;
+
+struct callchain_param	callchain_param = {
+	.mode	= CHAIN_GRAPH_REL,
+	.min_percent = 0.5
+};
+
+unsigned long total;
+unsigned long total_mmap;
+unsigned long total_comm;
+unsigned long total_fork;
+unsigned long total_unknown;
+unsigned long total_lost;
+
+/*
+ * histogram, sorted on item, collects counts
+ */
+
+struct hist_entry *__hist_entry__add(struct thread *thread, struct map *map,
+				     struct symbol *sym,
+				     struct symbol *sym_parent,
+				     u64 ip, u64 count, char level, bool *hit)
+{
+	struct rb_node **p = &hist.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *he;
+	struct hist_entry entry = {
+		.thread	= thread,
+		.map	= map,
+		.sym	= sym,
+		.ip	= ip,
+		.level	= level,
+		.count	= count,
+		.parent = sym_parent,
+	};
+	int cmp;
+
+	while (*p != NULL) {
+		parent = *p;
+		he = rb_entry(parent, struct hist_entry, rb_node);
+
+		cmp = hist_entry__cmp(&entry, he);
+
+		if (!cmp) {
+			*hit = true;
+			return he;
+		}
+
+		if (cmp < 0)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	he = malloc(sizeof(*he));
+	if (!he)
+		return NULL;
+	*he = entry;
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &hist);
+	*hit = false;
+	return he;
+}
+
+int64_t
+hist_entry__cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct sort_entry *se;
+	int64_t cmp = 0;
+
+	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		cmp = se->cmp(left, right);
+		if (cmp)
+			break;
+	}
+
+	return cmp;
+}
+
+int64_t
+hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
+{
+	struct sort_entry *se;
+	int64_t cmp = 0;
+
+	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		int64_t (*f)(struct hist_entry *, struct hist_entry *);
+
+		f = se->collapse ?: se->cmp;
+
+		cmp = f(left, right);
+		if (cmp)
+			break;
+	}
+
+	return cmp;
+}
+
+void hist_entry__free(struct hist_entry *he)
+{
+	free(he);
+}
+
+/*
+ * collapse the histogram
+ */
+
+void collapse__insert_entry(struct hist_entry *he)
+{
+	struct rb_node **p = &collapse_hists.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *iter;
+	int64_t cmp;
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct hist_entry, rb_node);
+
+		cmp = hist_entry__collapse(iter, he);
+
+		if (!cmp) {
+			iter->count += he->count;
+			hist_entry__free(he);
+			return;
+		}
+
+		if (cmp < 0)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &collapse_hists);
+}
+
+void collapse__resort(void)
+{
+	struct rb_node *next;
+	struct hist_entry *n;
+
+	if (!sort__need_collapse)
+		return;
+
+	next = rb_first(&hist);
+	while (next) {
+		n = rb_entry(next, struct hist_entry, rb_node);
+		next = rb_next(&n->rb_node);
+
+		rb_erase(&n->rb_node, &hist);
+		collapse__insert_entry(n);
+	}
+}
+
+/*
+ * reverse the map, sort on count.
+ */
+
+void output__insert_entry(struct hist_entry *he, u64 min_callchain_hits)
+{
+	struct rb_node **p = &output_hists.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *iter;
+
+	if (callchain)
+		callchain_param.sort(&he->sorted_chain, &he->callchain,
+				      min_callchain_hits, &callchain_param);
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct hist_entry, rb_node);
+
+		if (he->count > iter->count)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &output_hists);
+}
+
+void output__resort(u64 total_samples)
+{
+	struct rb_node *next;
+	struct hist_entry *n;
+	struct rb_root *tree = &hist;
+	u64 min_callchain_hits;
+
+	min_callchain_hits =
+		total_samples * (callchain_param.min_percent / 100);
+
+	if (sort__need_collapse)
+		tree = &collapse_hists;
+
+	next = rb_first(tree);
+
+	while (next) {
+		n = rb_entry(next, struct hist_entry, rb_node);
+		next = rb_next(&n->rb_node);
+
+		rb_erase(&n->rb_node, tree);
+		output__insert_entry(n, min_callchain_hits);
+	}
+}
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
new file mode 100644
index 00000000000..ac2149c559b
--- /dev/null
+++ b/tools/perf/util/hist.h
@@ -0,0 +1,50 @@
+#ifndef __PERF_HIST_H
+#define __PERF_HIST_H
+#include "../builtin.h"
+
+#include "util.h"
+
+#include "color.h"
+#include <linux/list.h>
+#include "cache.h"
+#include <linux/rbtree.h>
+#include "symbol.h"
+#include "string.h"
+#include "callchain.h"
+#include "strlist.h"
+#include "values.h"
+
+#include "../perf.h"
+#include "debug.h"
+#include "header.h"
+
+#include "parse-options.h"
+#include "parse-events.h"
+
+#include "thread.h"
+#include "sort.h"
+
+extern struct rb_root hist;
+extern struct rb_root collapse_hists;
+extern struct rb_root output_hists;
+extern int callchain;
+extern struct callchain_param callchain_param;
+extern unsigned long total;
+extern unsigned long total_mmap;
+extern unsigned long total_comm;
+extern unsigned long total_fork;
+extern unsigned long total_unknown;
+extern unsigned long total_lost;
+
+struct hist_entry *__hist_entry__add(struct thread *thread, struct map *map,
+				     struct symbol *sym, struct symbol *parent,
+				     u64 ip, u64 count, char level, bool *hit);
+extern int64_t hist_entry__cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t hist_entry__collapse(struct hist_entry *, struct hist_entry *);
+extern void hist_entry__free(struct hist_entry *);
+extern void collapse__insert_entry(struct hist_entry *);
+extern void collapse__resort(void);
+extern void output__insert_entry(struct hist_entry *, u64);
+extern void output__resort(u64);
+
+#endif	/* __PERF_HIST_H */
diff --git a/tools/perf/util/include/asm/asm-offsets.h b/tools/perf/util/include/asm/asm-offsets.h
new file mode 100644
index 00000000000..ed538942523
--- /dev/null
+++ b/tools/perf/util/include/asm/asm-offsets.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/bitops.h b/tools/perf/util/include/asm/bitops.h
new file mode 100644
index 00000000000..58e9817ffae
--- /dev/null
+++ b/tools/perf/util/include/asm/bitops.h
@@ -0,0 +1,18 @@
+#ifndef _PERF_ASM_BITOPS_H_
+#define _PERF_ASM_BITOPS_H_
+
+#include <sys/types.h>
+#include "../../types.h"
+#include <linux/compiler.h>
+
+/* CHECKME: Not sure both always match */
+#define BITS_PER_LONG	__WORDSIZE
+
+#include "../../../../include/asm-generic/bitops/__fls.h"
+#include "../../../../include/asm-generic/bitops/fls.h"
+#include "../../../../include/asm-generic/bitops/fls64.h"
+#include "../../../../include/asm-generic/bitops/__ffs.h"
+#include "../../../../include/asm-generic/bitops/ffz.h"
+#include "../../../../include/asm-generic/bitops/hweight.h"
+
+#endif
diff --git a/tools/perf/util/include/asm/byteorder.h b/tools/perf/util/include/asm/byteorder.h
new file mode 100644
index 00000000000..b722abe3a62
--- /dev/null
+++ b/tools/perf/util/include/asm/byteorder.h
@@ -0,0 +1,2 @@
+#include <asm/types.h>
+#include "../../../../include/linux/swab.h"
diff --git a/tools/perf/util/include/asm/swab.h b/tools/perf/util/include/asm/swab.h
new file mode 100644
index 00000000000..ed538942523
--- /dev/null
+++ b/tools/perf/util/include/asm/swab.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/uaccess.h b/tools/perf/util/include/asm/uaccess.h
new file mode 100644
index 00000000000..d0f72b8fcc3
--- /dev/null
+++ b/tools/perf/util/include/asm/uaccess.h
@@ -0,0 +1,14 @@
+#ifndef _PERF_ASM_UACCESS_H_
+#define _PERF_ASM_UACCESS_H_
+
+#define __get_user(src, dest)						\
+({									\
+	(src) = *dest;							\
+	0;								\
+})
+
+#define get_user	__get_user
+
+#define access_ok(type, addr, size)	1
+
+#endif
diff --git a/tools/perf/util/include/linux/bitmap.h b/tools/perf/util/include/linux/bitmap.h
new file mode 100644
index 00000000000..821c1033bcc
--- /dev/null
+++ b/tools/perf/util/include/linux/bitmap.h
@@ -0,0 +1,2 @@
+#include "../../../../include/linux/bitmap.h"
+#include "../../../../include/asm-generic/bitops/find.h"
diff --git a/tools/perf/util/include/linux/bitops.h b/tools/perf/util/include/linux/bitops.h
new file mode 100644
index 00000000000..ace57c36d1d
--- /dev/null
+++ b/tools/perf/util/include/linux/bitops.h
@@ -0,0 +1,27 @@
+#ifndef _PERF_LINUX_BITOPS_H_
+#define _PERF_LINUX_BITOPS_H_
+
+#define __KERNEL__
+
+#define CONFIG_GENERIC_FIND_NEXT_BIT
+#define CONFIG_GENERIC_FIND_FIRST_BIT
+#include "../../../../include/linux/bitops.h"
+
+static inline void set_bit(int nr, unsigned long *addr)
+{
+	addr[nr / BITS_PER_LONG] |= 1UL << (nr % BITS_PER_LONG);
+}
+
+static __always_inline int test_bit(unsigned int nr, const unsigned long *addr)
+{
+	return ((1UL << (nr % BITS_PER_LONG)) &
+		(((unsigned long *)addr)[nr / BITS_PER_LONG])) != 0;
+}
+
+unsigned long generic_find_next_zero_le_bit(const unsigned long *addr, unsigned
+		long size, unsigned long offset);
+
+unsigned long generic_find_next_le_bit(const unsigned long *addr, unsigned
+		long size, unsigned long offset);
+
+#endif
diff --git a/tools/perf/util/include/linux/compiler.h b/tools/perf/util/include/linux/compiler.h
new file mode 100644
index 00000000000..dfb0713ed47
--- /dev/null
+++ b/tools/perf/util/include/linux/compiler.h
@@ -0,0 +1,10 @@
+#ifndef _PERF_LINUX_COMPILER_H_
+#define _PERF_LINUX_COMPILER_H_
+
+#ifndef __always_inline
+#define __always_inline	inline
+#endif
+#define __user
+#define __attribute_const__
+
+#endif
diff --git a/tools/perf/util/include/linux/ctype.h b/tools/perf/util/include/linux/ctype.h
new file mode 100644
index 00000000000..bae5783282e
--- /dev/null
+++ b/tools/perf/util/include/linux/ctype.h
@@ -0,0 +1 @@
+#include "../../../../include/linux/ctype.h"
diff --git a/tools/perf/util/include/linux/kernel.h b/tools/perf/util/include/linux/kernel.h
index a6b87390cb5..21c0274c02f 100644
--- a/tools/perf/util/include/linux/kernel.h
+++ b/tools/perf/util/include/linux/kernel.h
@@ -1,6 +1,16 @@
 #ifndef PERF_LINUX_KERNEL_H_
 #define PERF_LINUX_KERNEL_H_
 
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <assert.h>
+
+#define DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d))
+
+#define ALIGN(x,a)		__ALIGN_MASK(x,(typeof(x))(a)-1)
+#define __ALIGN_MASK(x,mask)	(((x)+(mask))&~(mask))
+
 #ifndef offsetof
 #define offsetof(TYPE, MEMBER) ((size_t) &((TYPE *)0)->MEMBER)
 #endif
@@ -26,4 +36,70 @@
 	_max1 > _max2 ? _max1 : _max2; })
 #endif
 
+#ifndef min
+#define min(x, y) ({				\
+	typeof(x) _min1 = (x);			\
+	typeof(y) _min2 = (y);			\
+	(void) (&_min1 == &_min2);		\
+	_min1 < _min2 ? _min1 : _min2; })
+#endif
+
+#ifndef BUG_ON
+#define BUG_ON(cond) assert(!(cond))
+#endif
+
+/*
+ * Both need more care to handle endianness
+ * (Don't use bitmap_copy_le() for now)
+ */
+#define cpu_to_le64(x)	(x)
+#define cpu_to_le32(x)	(x)
+
+static inline int
+vscnprintf(char *buf, size_t size, const char *fmt, va_list args)
+{
+	int i;
+	ssize_t ssize = size;
+
+	i = vsnprintf(buf, size, fmt, args);
+
+	return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline int scnprintf(char * buf, size_t size, const char * fmt, ...)
+{
+	va_list args;
+	ssize_t ssize = size;
+	int i;
+
+	va_start(args, fmt);
+	i = vsnprintf(buf, size, fmt, args);
+	va_end(args);
+
+	return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline unsigned long
+simple_strtoul(const char *nptr, char **endptr, int base)
+{
+	return strtoul(nptr, endptr, base);
+}
+
+#ifndef pr_fmt
+#define pr_fmt(fmt) fmt
+#endif
+
+#define pr_err(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_warning(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_info(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_debug(fmt, ...) \
+	eprintf(1, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debugN(n, fmt, ...) \
+	eprintf(n, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug3(fmt, ...) pr_debugN(3, pr_fmt(fmt), ##__VA_ARGS__)
+
 #endif
diff --git a/tools/perf/util/include/linux/string.h b/tools/perf/util/include/linux/string.h
new file mode 100644
index 00000000000..3b2f5900276
--- /dev/null
+++ b/tools/perf/util/include/linux/string.h
@@ -0,0 +1 @@
+#include <string.h>
diff --git a/tools/perf/util/include/linux/types.h b/tools/perf/util/include/linux/types.h
new file mode 100644
index 00000000000..196862a81a2
--- /dev/null
+++ b/tools/perf/util/include/linux/types.h
@@ -0,0 +1,9 @@
+#ifndef _PERF_LINUX_TYPES_H_
+#define _PERF_LINUX_TYPES_H_
+
+#include <asm/types.h>
+
+#define DECLARE_BITMAP(name,bits) \
+	unsigned long name[BITS_TO_LONGS(bits)]
+
+#endif
diff --git a/tools/perf/util/levenshtein.h b/tools/perf/util/levenshtein.h
index 0173abeef52..b0fcb6d8a88 100644
--- a/tools/perf/util/levenshtein.h
+++ b/tools/perf/util/levenshtein.h
@@ -1,8 +1,8 @@
-#ifndef LEVENSHTEIN_H
-#define LEVENSHTEIN_H
+#ifndef __PERF_LEVENSHTEIN_H
+#define __PERF_LEVENSHTEIN_H
 
 int levenshtein(const char *string1, const char *string2,
 	int swap_penalty, int substition_penalty,
 	int insertion_penalty, int deletion_penalty);
 
-#endif
+#endif /* __PERF_LEVENSHTEIN_H */
diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index 804e0238273..c1c55682534 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -3,6 +3,7 @@
 #include <stdlib.h>
 #include <string.h>
 #include <stdio.h>
+#include "debug.h"
 
 static inline int is_anon_memory(const char *filename)
 {
@@ -19,7 +20,8 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 	return n;
 }
 
- struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen)
+struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen,
+		     unsigned int sym_priv_size, symbol_filter_t filter)
 {
 	struct map *self = malloc(sizeof(*self));
 
@@ -27,6 +29,7 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 		const char *filename = event->filename;
 		char newfilename[PATH_MAX];
 		int anon;
+		bool new_dso;
 
 		if (cwd) {
 			int n = strcommon(filename, cwd, cwdlen);
@@ -49,14 +52,29 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 		self->end   = event->start + event->len;
 		self->pgoff = event->pgoff;
 
-		self->dso = dsos__findnew(filename);
+		self->dso = dsos__findnew(filename, sym_priv_size, &new_dso);
 		if (self->dso == NULL)
 			goto out_delete;
 
+		if (new_dso) {
+			int nr = dso__load(self->dso, self, filter);
+
+			if (nr < 0)
+				pr_warning("Failed to open %s, continuing "
+					   "without symbols\n",
+					   self->dso->long_name);
+			else if (nr == 0)
+				pr_warning("No symbols found in %s, maybe "
+					   "install a debug package?\n",
+					   self->dso->long_name);
+		}
+
 		if (self->dso == vdso || anon)
-			self->map_ip = vdso__map_ip;
-		else
+			self->map_ip = self->unmap_ip = identity__map_ip;
+		else {
 			self->map_ip = map__map_ip;
+			self->unmap_ip = map__unmap_ip;
+		}
 	}
 	return self;
 out_delete:
diff --git a/tools/perf/util/module.c b/tools/perf/util/module.c
deleted file mode 100644
index 0d8c85defcd..00000000000
--- a/tools/perf/util/module.c
+++ /dev/null
@@ -1,545 +0,0 @@
-#include "util.h"
-#include "../perf.h"
-#include "string.h"
-#include "module.h"
-
-#include <libelf.h>
-#include <libgen.h>
-#include <gelf.h>
-#include <elf.h>
-#include <dirent.h>
-#include <sys/utsname.h>
-
-static unsigned int crc32(const char *p, unsigned int len)
-{
-	int i;
-	unsigned int crc = 0;
-
-	while (len--) {
-		crc ^= *p++;
-		for (i = 0; i < 8; i++)
-			crc = (crc >> 1) ^ ((crc & 1) ? 0xedb88320 : 0);
-	}
-	return crc;
-}
-
-/* module section methods */
-
-struct sec_dso *sec_dso__new_dso(const char *name)
-{
-	struct sec_dso *self = malloc(sizeof(*self) + strlen(name) + 1);
-
-	if (self != NULL) {
-		strcpy(self->name, name);
-		self->secs = RB_ROOT;
-		self->find_section = sec_dso__find_section;
-	}
-
-	return self;
-}
-
-static void sec_dso__delete_section(struct section *self)
-{
-	free(((void *)self));
-}
-
-void sec_dso__delete_sections(struct sec_dso *self)
-{
-	struct section *pos;
-	struct rb_node *next = rb_first(&self->secs);
-
-	while (next) {
-		pos = rb_entry(next, struct section, rb_node);
-		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &self->secs);
-		sec_dso__delete_section(pos);
-	}
-}
-
-void sec_dso__delete_self(struct sec_dso *self)
-{
-	sec_dso__delete_sections(self);
-	free(self);
-}
-
-static void sec_dso__insert_section(struct sec_dso *self, struct section *sec)
-{
-	struct rb_node **p = &self->secs.rb_node;
-	struct rb_node *parent = NULL;
-	const u64 hash = sec->hash;
-	struct section *s;
-
-	while (*p != NULL) {
-		parent = *p;
-		s = rb_entry(parent, struct section, rb_node);
-		if (hash < s->hash)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-	rb_link_node(&sec->rb_node, parent, p);
-	rb_insert_color(&sec->rb_node, &self->secs);
-}
-
-struct section *sec_dso__find_section(struct sec_dso *self, const char *name)
-{
-	struct rb_node *n;
-	u64 hash;
-	int len;
-
-	if (self == NULL)
-		return NULL;
-
-	len = strlen(name);
-	hash = crc32(name, len);
-
-	n = self->secs.rb_node;
-
-	while (n) {
-		struct section *s = rb_entry(n, struct section, rb_node);
-
-		if (hash < s->hash)
-			n = n->rb_left;
-		else if (hash > s->hash)
-			n = n->rb_right;
-		else {
-			if (!strcmp(name, s->name))
-				return s;
-			else
-				n = rb_next(&s->rb_node);
-		}
-	}
-
-	return NULL;
-}
-
-static size_t sec_dso__fprintf_section(struct section *self, FILE *fp)
-{
-	return fprintf(fp, "name:%s vma:%llx path:%s\n",
-		       self->name, self->vma, self->path);
-}
-
-size_t sec_dso__fprintf(struct sec_dso *self, FILE *fp)
-{
-	size_t ret = fprintf(fp, "dso: %s\n", self->name);
-
-	struct rb_node *nd;
-	for (nd = rb_first(&self->secs); nd; nd = rb_next(nd)) {
-		struct section *pos = rb_entry(nd, struct section, rb_node);
-		ret += sec_dso__fprintf_section(pos, fp);
-	}
-
-	return ret;
-}
-
-static struct section *section__new(const char *name, const char *path)
-{
-	struct section *self = calloc(1, sizeof(*self));
-
-	if (!self)
-		goto out_failure;
-
-	self->name = calloc(1, strlen(name) + 1);
-	if (!self->name)
-		goto out_failure;
-
-	self->path = calloc(1, strlen(path) + 1);
-	if (!self->path)
-		goto out_failure;
-
-	strcpy(self->name, name);
-	strcpy(self->path, path);
-	self->hash = crc32(self->name, strlen(name));
-
-	return self;
-
-out_failure:
-	if (self) {
-		if (self->name)
-			free(self->name);
-		if (self->path)
-			free(self->path);
-		free(self);
-	}
-
-	return NULL;
-}
-
-/* module methods */
-
-struct mod_dso *mod_dso__new_dso(const char *name)
-{
-	struct mod_dso *self = malloc(sizeof(*self) + strlen(name) + 1);
-
-	if (self != NULL) {
-		strcpy(self->name, name);
-		self->mods = RB_ROOT;
-		self->find_module = mod_dso__find_module;
-	}
-
-	return self;
-}
-
-static void mod_dso__delete_module(struct module *self)
-{
-	free(((void *)self));
-}
-
-void mod_dso__delete_modules(struct mod_dso *self)
-{
-	struct module *pos;
-	struct rb_node *next = rb_first(&self->mods);
-
-	while (next) {
-		pos = rb_entry(next, struct module, rb_node);
-		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &self->mods);
-		mod_dso__delete_module(pos);
-	}
-}
-
-void mod_dso__delete_self(struct mod_dso *self)
-{
-	mod_dso__delete_modules(self);
-	free(self);
-}
-
-static void mod_dso__insert_module(struct mod_dso *self, struct module *mod)
-{
-	struct rb_node **p = &self->mods.rb_node;
-	struct rb_node *parent = NULL;
-	const u64 hash = mod->hash;
-	struct module *m;
-
-	while (*p != NULL) {
-		parent = *p;
-		m = rb_entry(parent, struct module, rb_node);
-		if (hash < m->hash)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-	rb_link_node(&mod->rb_node, parent, p);
-	rb_insert_color(&mod->rb_node, &self->mods);
-}
-
-struct module *mod_dso__find_module(struct mod_dso *self, const char *name)
-{
-	struct rb_node *n;
-	u64 hash;
-	int len;
-
-	if (self == NULL)
-		return NULL;
-
-	len = strlen(name);
-	hash = crc32(name, len);
-
-	n = self->mods.rb_node;
-
-	while (n) {
-		struct module *m = rb_entry(n, struct module, rb_node);
-
-		if (hash < m->hash)
-			n = n->rb_left;
-		else if (hash > m->hash)
-			n = n->rb_right;
-		else {
-			if (!strcmp(name, m->name))
-				return m;
-			else
-				n = rb_next(&m->rb_node);
-		}
-	}
-
-	return NULL;
-}
-
-static size_t mod_dso__fprintf_module(struct module *self, FILE *fp)
-{
-	return fprintf(fp, "name:%s path:%s\n", self->name, self->path);
-}
-
-size_t mod_dso__fprintf(struct mod_dso *self, FILE *fp)
-{
-	struct rb_node *nd;
-	size_t ret;
-
-	ret = fprintf(fp, "dso: %s\n", self->name);
-
-	for (nd = rb_first(&self->mods); nd; nd = rb_next(nd)) {
-		struct module *pos = rb_entry(nd, struct module, rb_node);
-
-		ret += mod_dso__fprintf_module(pos, fp);
-	}
-
-	return ret;
-}
-
-static struct module *module__new(const char *name, const char *path)
-{
-	struct module *self = calloc(1, sizeof(*self));
-
-	if (!self)
-		goto out_failure;
-
-	self->name = calloc(1, strlen(name) + 1);
-	if (!self->name)
-		goto out_failure;
-
-	self->path = calloc(1, strlen(path) + 1);
-	if (!self->path)
-		goto out_failure;
-
-	strcpy(self->name, name);
-	strcpy(self->path, path);
-	self->hash = crc32(self->name, strlen(name));
-
-	return self;
-
-out_failure:
-	if (self) {
-		if (self->name)
-			free(self->name);
-		if (self->path)
-			free(self->path);
-		free(self);
-	}
-
-	return NULL;
-}
-
-static int mod_dso__load_sections(struct module *mod)
-{
-	int count = 0, path_len;
-	struct dirent *entry;
-	char *line = NULL;
-	char *dir_path;
-	DIR *dir;
-	size_t n;
-
-	path_len = strlen("/sys/module/");
-	path_len += strlen(mod->name);
-	path_len += strlen("/sections/");
-
-	dir_path = calloc(1, path_len + 1);
-	if (dir_path == NULL)
-		goto out_failure;
-
-	strcat(dir_path, "/sys/module/");
-	strcat(dir_path, mod->name);
-	strcat(dir_path, "/sections/");
-
-	dir = opendir(dir_path);
-	if (dir == NULL)
-		goto out_free;
-
-	while ((entry = readdir(dir))) {
-		struct section *section;
-		char *path, *vma;
-		int line_len;
-		FILE *file;
-
-		if (!strcmp(".", entry->d_name) || !strcmp("..", entry->d_name))
-			continue;
-
-		path = calloc(1, path_len + strlen(entry->d_name) + 1);
-		if (path == NULL)
-			break;
-		strcat(path, dir_path);
-		strcat(path, entry->d_name);
-
-		file = fopen(path, "r");
-		if (file == NULL) {
-			free(path);
-			break;
-		}
-
-		line_len = getline(&line, &n, file);
-		if (line_len < 0) {
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		if (!line) {
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		line[--line_len] = '\0'; /* \n */
-
-		vma = strstr(line, "0x");
-		if (!vma) {
-			free(path);
-			fclose(file);
-			break;
-		}
-		vma += 2;
-
-		section = section__new(entry->d_name, path);
-		if (!section) {
-			fprintf(stderr, "load_sections: allocation error\n");
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		hex2u64(vma, &section->vma);
-		sec_dso__insert_section(mod->sections, section);
-
-		free(path);
-		fclose(file);
-		count++;
-	}
-
-	closedir(dir);
-	free(line);
-	free(dir_path);
-
-	return count;
-
-out_free:
-	free(dir_path);
-
-out_failure:
-	return count;
-}
-
-static int mod_dso__load_module_paths(struct mod_dso *self)
-{
-	struct utsname uts;
-	int count = 0, len, err = -1;
-	char *line = NULL;
-	FILE *file;
-	char *dpath, *dir;
-	size_t n;
-
-	if (uname(&uts) < 0)
-		return err;
-
-	len = strlen("/lib/modules/");
-	len += strlen(uts.release);
-	len += strlen("/modules.dep");
-
-	dpath = calloc(1, len + 1);
-	if (dpath == NULL)
-		return err;
-
-	strcat(dpath, "/lib/modules/");
-	strcat(dpath, uts.release);
-	strcat(dpath, "/modules.dep");
-
-	file = fopen(dpath, "r");
-	if (file == NULL)
-		goto out_failure;
-
-	dir = dirname(dpath);
-	if (!dir)
-		goto out_failure;
-	strcat(dir, "/");
-
-	while (!feof(file)) {
-		struct module *module;
-		char *name, *path, *tmp;
-		FILE *modfile;
-		int line_len;
-
-		line_len = getline(&line, &n, file);
-		if (line_len < 0)
-			break;
-
-		if (!line)
-			break;
-
-		line[--line_len] = '\0'; /* \n */
-
-		path = strchr(line, ':');
-		if (!path)
-			break;
-		*path = '\0';
-
-		path = strdup(line);
-		if (!path)
-			break;
-
-		if (!strstr(path, dir)) {
-			if (strncmp(path, "kernel/", 7))
-				break;
-
-			free(path);
-			path = calloc(1, strlen(dir) + strlen(line) + 1);
-			if (!path)
-				break;
-			strcat(path, dir);
-			strcat(path, line);
-		}
-
-		modfile = fopen(path, "r");
-		if (modfile == NULL)
-			break;
-		fclose(modfile);
-
-		name = strdup(path);
-		if (!name)
-			break;
-
-		name = strtok(name, "/");
-		tmp = name;
-
-		while (tmp) {
-			tmp = strtok(NULL, "/");
-			if (tmp)
-				name = tmp;
-		}
-
-		name = strsep(&name, ".");
-		if (!name)
-			break;
-
-		/* Quirk: replace '-' with '_' in all modules */
-		for (len = strlen(name); len; len--) {
-			if (*(name+len) == '-')
-				*(name+len) = '_';
-		}
-
-		module = module__new(name, path);
-		if (!module)
-			break;
-		mod_dso__insert_module(self, module);
-
-		module->sections = sec_dso__new_dso("sections");
-		if (!module->sections)
-			break;
-
-		module->active = mod_dso__load_sections(module);
-
-		if (module->active > 0)
-			count++;
-	}
-
-	if (feof(file))
-		err = count;
-	else
-		fprintf(stderr, "load_module_paths: modules.dep parsing failure!\n");
-
-out_failure:
-	if (dpath)
-		free(dpath);
-	if (file)
-		fclose(file);
-	if (line)
-		free(line);
-
-	return err;
-}
-
-int mod_dso__load_modules(struct mod_dso *dso)
-{
-	int err;
-
-	err = mod_dso__load_module_paths(dso);
-
-	return err;
-}
diff --git a/tools/perf/util/module.h b/tools/perf/util/module.h
deleted file mode 100644
index 8a592ef641c..00000000000
--- a/tools/perf/util/module.h
+++ /dev/null
@@ -1,53 +0,0 @@
-#ifndef _PERF_MODULE_
-#define _PERF_MODULE_ 1
-
-#include <linux/types.h>
-#include "../types.h"
-#include <linux/list.h>
-#include <linux/rbtree.h>
-
-struct section {
-	struct rb_node	rb_node;
-	u64		hash;
-	u64		vma;
-	char		*name;
-	char		*path;
-};
-
-struct sec_dso {
-	struct list_head node;
-	struct rb_root	 secs;
-	struct section    *(*find_section)(struct sec_dso *, const char *name);
-	char		 name[0];
-};
-
-struct module {
-	struct rb_node	rb_node;
-	u64		hash;
-	char		*name;
-	char		*path;
-	struct sec_dso	*sections;
-	int		active;
-};
-
-struct mod_dso {
-	struct list_head node;
-	struct rb_root	 mods;
-	struct module    *(*find_module)(struct mod_dso *, const char *name);
-	char		 name[0];
-};
-
-struct sec_dso *sec_dso__new_dso(const char *name);
-void sec_dso__delete_sections(struct sec_dso *self);
-void sec_dso__delete_self(struct sec_dso *self);
-size_t sec_dso__fprintf(struct sec_dso *self, FILE *fp);
-struct section *sec_dso__find_section(struct sec_dso *self, const char *name);
-
-struct mod_dso *mod_dso__new_dso(const char *name);
-void mod_dso__delete_modules(struct mod_dso *self);
-void mod_dso__delete_self(struct mod_dso *self);
-size_t mod_dso__fprintf(struct mod_dso *self, FILE *fp);
-struct module *mod_dso__find_module(struct mod_dso *self, const char *name);
-int mod_dso__load_modules(struct mod_dso *dso);
-
-#endif /* _PERF_MODULE_ */
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 8cfb48cbbea..b097570e962 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -8,9 +8,10 @@
 #include "cache.h"
 #include "header.h"
 
-int					nr_counters;
+int				nr_counters;
 
 struct perf_event_attr		attrs[MAX_COUNTERS];
+char				*filters[MAX_COUNTERS];
 
 struct event_symbol {
 	u8		type;
@@ -708,7 +709,6 @@ static void store_event_type(const char *orgname)
 	perf_header__push_event(id, orgname);
 }
 
-
 int parse_events(const struct option *opt __used, const char *str, int unset __used)
 {
 	struct perf_event_attr attr;
@@ -745,6 +745,28 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u
 	return 0;
 }
 
+int parse_filter(const struct option *opt __used, const char *str,
+		 int unset __used)
+{
+	int i = nr_counters - 1;
+	int len = strlen(str);
+
+	if (i < 0 || attrs[i].type != PERF_TYPE_TRACEPOINT) {
+		fprintf(stderr,
+			"-F option should follow a -e tracepoint option\n");
+		return -1;
+	}
+
+	filters[i] = malloc(len + 1);
+	if (!filters[i]) {
+		fprintf(stderr, "not enough memory to hold filter string\n");
+		return -1;
+	}
+	strcpy(filters[i], str);
+
+	return 0;
+}
+
 static const char * const event_type_descriptors[] = {
 	"",
 	"Hardware event",
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index 30c60811284..b8c1f64bc93 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -1,5 +1,5 @@
-#ifndef _PARSE_EVENTS_H
-#define _PARSE_EVENTS_H
+#ifndef __PERF_PARSE_EVENTS_H
+#define __PERF_PARSE_EVENTS_H
 /*
  * Parse symbolic events/counts passed in as options:
  */
@@ -17,11 +17,13 @@ extern struct tracepoint_path *tracepoint_id_to_path(u64 config);
 extern int			nr_counters;
 
 extern struct perf_event_attr attrs[MAX_COUNTERS];
+extern char *filters[MAX_COUNTERS];
 
 extern const char *event_name(int ctr);
 extern const char *__event_name(int type, u64 config);
 
 extern int parse_events(const struct option *opt, const char *str, int unset);
+extern int parse_filter(const struct option *opt, const char *str, int unset);
 
 #define EVENTS_HELP_MAX (128*1024)
 
@@ -31,4 +33,4 @@ extern char debugfs_path[];
 extern int valid_debugfs_mount(const char *debugfs);
 
 
-#endif /* _PARSE_EVENTS_H */
+#endif /* __PERF_PARSE_EVENTS_H */
diff --git a/tools/perf/util/parse-options.h b/tools/perf/util/parse-options.h
index 2ee248ff27e..948805af43c 100644
--- a/tools/perf/util/parse-options.h
+++ b/tools/perf/util/parse-options.h
@@ -1,5 +1,5 @@
-#ifndef PARSE_OPTIONS_H
-#define PARSE_OPTIONS_H
+#ifndef __PERF_PARSE_OPTIONS_H
+#define __PERF_PARSE_OPTIONS_H
 
 enum parse_opt_type {
 	/* special types */
@@ -174,4 +174,4 @@ extern int parse_opt_verbosity_cb(const struct option *, const char *, int);
 
 extern const char *parse_options_fix_filename(const char *prefix, const char *file);
 
-#endif
+#endif /* __PERF_PARSE_OPTIONS_H */
diff --git a/tools/perf/util/quote.h b/tools/perf/util/quote.h
index a5454a1d1c1..b6a01973391 100644
--- a/tools/perf/util/quote.h
+++ b/tools/perf/util/quote.h
@@ -1,5 +1,5 @@
-#ifndef QUOTE_H
-#define QUOTE_H
+#ifndef __PERF_QUOTE_H
+#define __PERF_QUOTE_H
 
 #include <stddef.h>
 #include <stdio.h>
@@ -65,4 +65,4 @@ extern void perl_quote_print(FILE *stream, const char *src);
 extern void python_quote_print(FILE *stream, const char *src);
 extern void tcl_quote_print(FILE *stream, const char *src);
 
-#endif
+#endif /* __PERF_QUOTE_H */
diff --git a/tools/perf/util/run-command.h b/tools/perf/util/run-command.h
index cc1837deba8..d79028727ce 100644
--- a/tools/perf/util/run-command.h
+++ b/tools/perf/util/run-command.h
@@ -1,5 +1,5 @@
-#ifndef RUN_COMMAND_H
-#define RUN_COMMAND_H
+#ifndef __PERF_RUN_COMMAND_H
+#define __PERF_RUN_COMMAND_H
 
 enum {
 	ERR_RUN_COMMAND_FORK = 10000,
@@ -85,4 +85,4 @@ struct async {
 int start_async(struct async *async);
 int finish_async(struct async *async);
 
-#endif
+#endif /* __PERF_RUN_COMMAND_H */
diff --git a/tools/perf/util/sigchain.h b/tools/perf/util/sigchain.h
index 618083bce0c..1a53c11265f 100644
--- a/tools/perf/util/sigchain.h
+++ b/tools/perf/util/sigchain.h
@@ -1,5 +1,5 @@
-#ifndef SIGCHAIN_H
-#define SIGCHAIN_H
+#ifndef __PERF_SIGCHAIN_H
+#define __PERF_SIGCHAIN_H
 
 typedef void (*sigchain_fun)(int);
 
@@ -8,4 +8,4 @@ int sigchain_pop(int sig);
 
 void sigchain_push_common(sigchain_fun f);
 
-#endif /* SIGCHAIN_H */
+#endif /* __PERF_SIGCHAIN_H */
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
new file mode 100644
index 00000000000..b490354d1b2
--- /dev/null
+++ b/tools/perf/util/sort.c
@@ -0,0 +1,290 @@
+#include "sort.h"
+
+regex_t		parent_regex;
+char		default_parent_pattern[] = "^sys_|^do_page_fault";
+char		*parent_pattern = default_parent_pattern;
+char		default_sort_order[] = "comm,dso,symbol";
+char		*sort_order = default_sort_order;
+int		sort__need_collapse = 0;
+int		sort__has_parent = 0;
+
+enum sort_type	sort__first_dimension;
+
+unsigned int dsos__col_width;
+unsigned int comms__col_width;
+unsigned int threads__col_width;
+static unsigned int parent_symbol__col_width;
+char * field_sep;
+
+LIST_HEAD(hist_entry__sort_list);
+
+struct sort_entry sort_thread = {
+	.header = "Command:  Pid",
+	.cmp	= sort__thread_cmp,
+	.print	= sort__thread_print,
+	.width	= &threads__col_width,
+};
+
+struct sort_entry sort_comm = {
+	.header		= "Command",
+	.cmp		= sort__comm_cmp,
+	.collapse	= sort__comm_collapse,
+	.print		= sort__comm_print,
+	.width		= &comms__col_width,
+};
+
+struct sort_entry sort_dso = {
+	.header = "Shared Object",
+	.cmp	= sort__dso_cmp,
+	.print	= sort__dso_print,
+	.width	= &dsos__col_width,
+};
+
+struct sort_entry sort_sym = {
+	.header = "Symbol",
+	.cmp	= sort__sym_cmp,
+	.print	= sort__sym_print,
+};
+
+struct sort_entry sort_parent = {
+	.header = "Parent symbol",
+	.cmp	= sort__parent_cmp,
+	.print	= sort__parent_print,
+	.width	= &parent_symbol__col_width,
+};
+
+struct sort_dimension {
+	const char		*name;
+	struct sort_entry	*entry;
+	int			taken;
+};
+
+static struct sort_dimension sort_dimensions[] = {
+	{ .name = "pid",	.entry = &sort_thread,	},
+	{ .name = "comm",	.entry = &sort_comm,	},
+	{ .name = "dso",	.entry = &sort_dso,	},
+	{ .name = "symbol",	.entry = &sort_sym,	},
+	{ .name = "parent",	.entry = &sort_parent,	},
+};
+
+int64_t cmp_null(void *l, void *r)
+{
+	if (!l && !r)
+		return 0;
+	else if (!l)
+		return -1;
+	else
+		return 1;
+}
+
+/* --sort pid */
+
+int64_t
+sort__thread_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	return right->thread->pid - left->thread->pid;
+}
+
+int repsep_fprintf(FILE *fp, const char *fmt, ...)
+{
+	int n;
+	va_list ap;
+
+	va_start(ap, fmt);
+	if (!field_sep)
+		n = vfprintf(fp, fmt, ap);
+	else {
+		char *bf = NULL;
+		n = vasprintf(&bf, fmt, ap);
+		if (n > 0) {
+			char *sep = bf;
+
+			while (1) {
+				sep = strchr(sep, *field_sep);
+				if (sep == NULL)
+					break;
+				*sep = '.';
+			}
+		}
+		fputs(bf, fp);
+		free(bf);
+	}
+	va_end(ap);
+	return n;
+}
+
+size_t
+sort__thread_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%*s:%5d", width - 6,
+			      self->thread->comm ?: "", self->thread->pid);
+}
+
+size_t
+sort__comm_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%*s", width, self->thread->comm);
+}
+
+/* --sort dso */
+
+int64_t
+sort__dso_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct dso *dso_l = left->map ? left->map->dso : NULL;
+	struct dso *dso_r = right->map ? right->map->dso : NULL;
+	const char *dso_name_l, *dso_name_r;
+
+	if (!dso_l || !dso_r)
+		return cmp_null(dso_l, dso_r);
+
+	if (verbose) {
+		dso_name_l = dso_l->long_name;
+		dso_name_r = dso_r->long_name;
+	} else {
+		dso_name_l = dso_l->short_name;
+		dso_name_r = dso_r->short_name;
+	}
+
+	return strcmp(dso_name_l, dso_name_r);
+}
+
+size_t
+sort__dso_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	if (self->map && self->map->dso) {
+		const char *dso_name = !verbose ? self->map->dso->short_name :
+						  self->map->dso->long_name;
+		return repsep_fprintf(fp, "%-*s", width, dso_name);
+	}
+
+	return repsep_fprintf(fp, "%*llx", width, (u64)self->ip);
+}
+
+/* --sort symbol */
+
+int64_t
+sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	u64 ip_l, ip_r;
+
+	if (left->sym == right->sym)
+		return 0;
+
+	ip_l = left->sym ? left->sym->start : left->ip;
+	ip_r = right->sym ? right->sym->start : right->ip;
+
+	return (int64_t)(ip_r - ip_l);
+}
+
+
+size_t
+sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used)
+{
+	size_t ret = 0;
+
+	if (verbose) {
+		char o = self->map ? dso__symtab_origin(self->map->dso) : '!';
+		ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip, o);
+	}
+
+	ret += repsep_fprintf(fp, "[%c] ", self->level);
+	if (self->sym)
+		ret += repsep_fprintf(fp, "%s", self->sym->name);
+	else
+		ret += repsep_fprintf(fp, "%#016llx", (u64)self->ip);
+
+	return ret;
+}
+
+/* --sort comm */
+
+int64_t
+sort__comm_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	return right->thread->pid - left->thread->pid;
+}
+
+int64_t
+sort__comm_collapse(struct hist_entry *left, struct hist_entry *right)
+{
+	char *comm_l = left->thread->comm;
+	char *comm_r = right->thread->comm;
+
+	if (!comm_l || !comm_r)
+		return cmp_null(comm_l, comm_r);
+
+	return strcmp(comm_l, comm_r);
+}
+
+/* --sort parent */
+
+int64_t
+sort__parent_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct symbol *sym_l = left->parent;
+	struct symbol *sym_r = right->parent;
+
+	if (!sym_l || !sym_r)
+		return cmp_null(sym_l, sym_r);
+
+	return strcmp(sym_l->name, sym_r->name);
+}
+
+size_t
+sort__parent_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%-*s", width,
+			      self->parent ? self->parent->name : "[other]");
+}
+
+int sort_dimension__add(const char *tok)
+{
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) {
+		struct sort_dimension *sd = &sort_dimensions[i];
+
+		if (sd->taken)
+			continue;
+
+		if (strncasecmp(tok, sd->name, strlen(tok)))
+			continue;
+
+		if (sd->entry->collapse)
+			sort__need_collapse = 1;
+
+		if (sd->entry == &sort_parent) {
+			int ret = regcomp(&parent_regex, parent_pattern, REG_EXTENDED);
+			if (ret) {
+				char err[BUFSIZ];
+
+				regerror(ret, &parent_regex, err, sizeof(err));
+				fprintf(stderr, "Invalid regex: %s\n%s",
+					parent_pattern, err);
+				exit(-1);
+			}
+			sort__has_parent = 1;
+		}
+
+		if (list_empty(&hist_entry__sort_list)) {
+			if (!strcmp(sd->name, "pid"))
+				sort__first_dimension = SORT_PID;
+			else if (!strcmp(sd->name, "comm"))
+				sort__first_dimension = SORT_COMM;
+			else if (!strcmp(sd->name, "dso"))
+				sort__first_dimension = SORT_DSO;
+			else if (!strcmp(sd->name, "symbol"))
+				sort__first_dimension = SORT_SYM;
+			else if (!strcmp(sd->name, "parent"))
+				sort__first_dimension = SORT_PARENT;
+		}
+
+		list_add_tail(&sd->entry->list, &hist_entry__sort_list);
+		sd->taken = 1;
+
+		return 0;
+	}
+
+	return -ESRCH;
+}
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
new file mode 100644
index 00000000000..333e664ff45
--- /dev/null
+++ b/tools/perf/util/sort.h
@@ -0,0 +1,99 @@
+#ifndef __PERF_SORT_H
+#define __PERF_SORT_H
+#include "../builtin.h"
+
+#include "util.h"
+
+#include "color.h"
+#include <linux/list.h>
+#include "cache.h"
+#include <linux/rbtree.h>
+#include "symbol.h"
+#include "string.h"
+#include "callchain.h"
+#include "strlist.h"
+#include "values.h"
+
+#include "../perf.h"
+#include "debug.h"
+#include "header.h"
+
+#include "parse-options.h"
+#include "parse-events.h"
+
+#include "thread.h"
+#include "sort.h"
+
+extern regex_t parent_regex;
+extern char *sort_order;
+extern char default_parent_pattern[];
+extern char *parent_pattern;
+extern char default_sort_order[];
+extern int sort__need_collapse;
+extern int sort__has_parent;
+extern char *field_sep;
+extern struct sort_entry sort_comm;
+extern struct sort_entry sort_dso;
+extern struct sort_entry sort_sym;
+extern struct sort_entry sort_parent;
+extern unsigned int dsos__col_width;
+extern unsigned int comms__col_width;
+extern unsigned int threads__col_width;
+extern enum sort_type sort__first_dimension;
+
+struct hist_entry {
+	struct rb_node		rb_node;
+	u64			count;
+	struct thread		*thread;
+	struct map		*map;
+	struct symbol		*sym;
+	u64			ip;
+	char			level;
+	struct symbol		*parent;
+	struct callchain_node	callchain;
+	struct rb_root		sorted_chain;
+};
+
+enum sort_type {
+	SORT_PID,
+	SORT_COMM,
+	SORT_DSO,
+	SORT_SYM,
+	SORT_PARENT
+};
+
+/*
+ * configurable sorting bits
+ */
+
+struct sort_entry {
+	struct list_head list;
+
+	const char *header;
+
+	int64_t (*cmp)(struct hist_entry *, struct hist_entry *);
+	int64_t (*collapse)(struct hist_entry *, struct hist_entry *);
+	size_t	(*print)(FILE *fp, struct hist_entry *, unsigned int width);
+	unsigned int *width;
+	bool	elide;
+};
+
+extern struct sort_entry sort_thread;
+extern struct list_head hist_entry__sort_list;
+
+extern int repsep_fprintf(FILE *fp, const char *fmt, ...);
+extern size_t sort__thread_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__comm_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__dso_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__sym_print(FILE *, struct hist_entry *, unsigned int __used);
+extern int64_t cmp_null(void *, void *);
+extern int64_t sort__thread_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__comm_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__comm_collapse(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__dso_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__sym_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__parent_cmp(struct hist_entry *, struct hist_entry *);
+extern size_t sort__parent_print(FILE *, struct hist_entry *, unsigned int);
+extern int sort_dimension__add(const char *);
+
+#endif	/* __PERF_SORT_H */
diff --git a/tools/perf/util/strbuf.h b/tools/perf/util/strbuf.h
index d2aa86c014c..a3d121d6c83 100644
--- a/tools/perf/util/strbuf.h
+++ b/tools/perf/util/strbuf.h
@@ -1,5 +1,5 @@
-#ifndef STRBUF_H
-#define STRBUF_H
+#ifndef __PERF_STRBUF_H
+#define __PERF_STRBUF_H
 
 /*
  * Strbuf's can be use in many ways: as a byte array, or to store arbitrary
@@ -134,4 +134,4 @@ extern int launch_editor(const char *path, struct strbuf *buffer, const char *co
 extern int strbuf_branchname(struct strbuf *sb, const char *name);
 extern int strbuf_check_branch_ref(struct strbuf *sb, const char *name);
 
-#endif /* STRBUF_H */
+#endif /* __PERF_STRBUF_H */
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index c93eca9a7be..04743d3e903 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -1,3 +1,4 @@
+#include <string.h>
 #include "string.h"
 
 static int hex(char ch)
@@ -32,3 +33,13 @@ int hex2u64(const char *ptr, u64 *long_val)
 
 	return p - ptr;
 }
+
+char *strxfrchar(char *s, char from, char to)
+{
+	char *p = s;
+
+	while ((p = strchr(p, from)) != NULL)
+		*p++ = to;
+
+	return s;
+}
diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h
index bf39dfadfd2..2c84bf65ba0 100644
--- a/tools/perf/util/string.h
+++ b/tools/perf/util/string.h
@@ -1,11 +1,12 @@
-#ifndef _PERF_STRING_H_
-#define _PERF_STRING_H_
+#ifndef __PERF_STRING_H_
+#define __PERF_STRING_H_
 
 #include "types.h"
 
 int hex2u64(const char *ptr, u64 *val);
+char *strxfrchar(char *s, char from, char to);
 
 #define _STR(x) #x
 #define STR(x) _STR(x)
 
-#endif
+#endif /* __PERF_STRING_H */
diff --git a/tools/perf/util/strlist.h b/tools/perf/util/strlist.h
index 921818e44a5..cb4659306d7 100644
--- a/tools/perf/util/strlist.h
+++ b/tools/perf/util/strlist.h
@@ -1,5 +1,5 @@
-#ifndef STRLIST_H_
-#define STRLIST_H_
+#ifndef __PERF_STRLIST_H
+#define __PERF_STRLIST_H
 
 #include <linux/rbtree.h>
 #include <stdbool.h>
@@ -36,4 +36,4 @@ static inline unsigned int strlist__nr_entries(const struct strlist *self)
 }
 
 int strlist__parse_list(struct strlist *self, const char *s);
-#endif /* STRLIST_H_ */
+#endif /* __PERF_STRLIST_H */
diff --git a/tools/perf/util/svghelper.c b/tools/perf/util/svghelper.c
index 856655d8b0b..b3637db025a 100644
--- a/tools/perf/util/svghelper.c
+++ b/tools/perf/util/svghelper.c
@@ -103,7 +103,7 @@ void open_svg(const char *filename, int cpus, int rows, u64 start, u64 end)
 	fprintf(svgfile, "      rect.process2 { fill:rgb(180,180,180); fill-opacity:0.9; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.sample   { fill:rgb(  0,  0,255); fill-opacity:0.8; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.blocked  { fill:rgb(255,  0,  0); fill-opacity:0.5; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
-	fprintf(svgfile, "      rect.waiting  { fill:rgb(214,214,  0); fill-opacity:0.3; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
+	fprintf(svgfile, "      rect.waiting  { fill:rgb(224,214,  0); fill-opacity:0.8; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.WAITING  { fill:rgb(255,214, 48); fill-opacity:0.6; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.cpu      { fill:rgb(192,192,192); fill-opacity:0.2; stroke-width:0.5; stroke:rgb(128,128,128); } \n");
 	fprintf(svgfile, "      rect.pstate   { fill:rgb(128,128,128); fill-opacity:0.8; stroke-width:0; } \n");
diff --git a/tools/perf/util/svghelper.h b/tools/perf/util/svghelper.h
index cd93195aedb..e0781989cc3 100644
--- a/tools/perf/util/svghelper.h
+++ b/tools/perf/util/svghelper.h
@@ -1,5 +1,5 @@
-#ifndef _INCLUDE_GUARD_SVG_HELPER_
-#define _INCLUDE_GUARD_SVG_HELPER_
+#ifndef __PERF_SVGHELPER_H
+#define __PERF_SVGHELPER_H
 
 #include "types.h"
 
@@ -25,4 +25,4 @@ extern void svg_close(void);
 
 extern int svg_page_width;
 
-#endif
+#endif /* __PERF_SVGHELPER_H */
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 47ea0609a76..8f0208ce237 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -2,14 +2,14 @@
 #include "../perf.h"
 #include "string.h"
 #include "symbol.h"
+#include "thread.h"
 
 #include "debug.h"
 
 #include <libelf.h>
 #include <gelf.h>
 #include <elf.h>
-
-const char *sym_hist_filter;
+#include <sys/utsname.h>
 
 enum dso_origin {
 	DSO__ORIG_KERNEL = 0,
@@ -18,12 +18,65 @@ enum dso_origin {
 	DSO__ORIG_UBUNTU,
 	DSO__ORIG_BUILDID,
 	DSO__ORIG_DSO,
+	DSO__ORIG_KMODULE,
 	DSO__ORIG_NOT_FOUND,
 };
 
-static struct symbol *symbol__new(u64 start, u64 len,
-				  const char *name, unsigned int priv_size,
-				  u64 obj_start, int v)
+static void dsos__add(struct dso *dso);
+static struct dso *dsos__find(const char *name);
+static struct map *map__new2(u64 start, struct dso *dso);
+static void kernel_maps__insert(struct map *map);
+
+static struct rb_root kernel_maps;
+
+static void dso__fixup_sym_end(struct dso *self)
+{
+	struct rb_node *nd, *prevnd = rb_first(&self->syms);
+	struct symbol *curr, *prev;
+
+	if (prevnd == NULL)
+		return;
+
+	curr = rb_entry(prevnd, struct symbol, rb_node);
+
+	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
+		prev = curr;
+		curr = rb_entry(nd, struct symbol, rb_node);
+
+		if (prev->end == prev->start)
+			prev->end = curr->start - 1;
+	}
+
+	/* Last entry */
+	if (curr->end == curr->start)
+		curr->end = roundup(curr->start, 4096);
+}
+
+static void kernel_maps__fixup_end(void)
+{
+	struct map *prev, *curr;
+	struct rb_node *nd, *prevnd = rb_first(&kernel_maps);
+
+	if (prevnd == NULL)
+		return;
+
+	curr = rb_entry(prevnd, struct map, rb_node);
+
+	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
+		prev = curr;
+		curr = rb_entry(nd, struct map, rb_node);
+		prev->end = curr->start - 1;
+	}
+
+	nd = rb_last(&curr->dso->syms);
+	if (nd) {
+		struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+		curr->end = sym->end;
+	}
+}
+
+static struct symbol *symbol__new(u64 start, u64 len, const char *name,
+				  unsigned int priv_size)
 {
 	size_t namelen = strlen(name) + 1;
 	struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen);
@@ -31,23 +84,15 @@ static struct symbol *symbol__new(u64 start, u64 len,
 	if (!self)
 		return NULL;
 
-	if (v >= 2)
-		printf("new symbol: %016Lx [%08lx]: %s, hist: %p, obj_start: %p\n",
-			(u64)start, (unsigned long)len, name, self->hist, (void *)(unsigned long)obj_start);
-
-	self->obj_start= obj_start;
-	self->hist = NULL;
-	self->hist_sum = 0;
-
-	if (sym_hist_filter && !strcmp(name, sym_hist_filter))
-		self->hist = calloc(sizeof(u64), len);
-
 	if (priv_size) {
 		memset(self, 0, priv_size);
 		self = ((void *)self) + priv_size;
 	}
 	self->start = start;
 	self->end   = len ? start + len - 1 : start;
+
+	pr_debug3("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end);
+
 	memcpy(self->name, name, namelen);
 
 	return self;
@@ -60,12 +105,8 @@ static void symbol__delete(struct symbol *self, unsigned int priv_size)
 
 static size_t symbol__fprintf(struct symbol *self, FILE *fp)
 {
-	if (!self->module)
-		return fprintf(fp, " %llx-%llx %s\n",
+	return fprintf(fp, " %llx-%llx %s\n",
 		       self->start, self->end, self->name);
-	else
-		return fprintf(fp, " %llx-%llx %s \t[%s]\n",
-		       self->start, self->end, self->name, self->module->name);
 }
 
 struct dso *dso__new(const char *name, unsigned int sym_priv_size)
@@ -74,6 +115,8 @@ struct dso *dso__new(const char *name, unsigned int sym_priv_size)
 
 	if (self != NULL) {
 		strcpy(self->name, name);
+		self->long_name = self->name;
+		self->short_name = self->name;
 		self->syms = RB_ROOT;
 		self->sym_priv_size = sym_priv_size;
 		self->find_symbol = dso__find_symbol;
@@ -100,6 +143,8 @@ static void dso__delete_symbols(struct dso *self)
 void dso__delete(struct dso *self)
 {
 	dso__delete_symbols(self);
+	if (self->long_name != self->name)
+		free(self->long_name);
 	free(self);
 }
 
@@ -147,7 +192,7 @@ struct symbol *dso__find_symbol(struct dso *self, u64 ip)
 
 size_t dso__fprintf(struct dso *self, FILE *fp)
 {
-	size_t ret = fprintf(fp, "dso: %s\n", self->name);
+	size_t ret = fprintf(fp, "dso: %s\n", self->short_name);
 
 	struct rb_node *nd;
 	for (nd = rb_first(&self->syms); nd; nd = rb_next(nd)) {
@@ -158,13 +203,16 @@ size_t dso__fprintf(struct dso *self, FILE *fp)
 	return ret;
 }
 
-static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
+/*
+ * Loads the function entries in /proc/kallsyms into kernel_map->dso,
+ * so that we can in the next step set the symbol ->end address and then
+ * call kernel_maps__split_kallsyms.
+ */
+static int kernel_maps__load_all_kallsyms(void)
 {
-	struct rb_node *nd, *prevnd;
 	char *line = NULL;
 	size_t n;
 	FILE *file = fopen("/proc/kallsyms", "r");
-	int count = 0;
 
 	if (file == NULL)
 		goto out_failure;
@@ -174,6 +222,7 @@ static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
 		struct symbol *sym;
 		int line_len, len;
 		char symbol_type;
+		char *symbol_name;
 
 		line_len = getline(&line, &n, file);
 		if (line_len < 0)
@@ -196,44 +245,24 @@ static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
 		 */
 		if (symbol_type != 'T' && symbol_type != 'W')
 			continue;
+
+		symbol_name = line + len + 2;
 		/*
-		 * Well fix up the end later, when we have all sorted.
+		 * Will fix up the end later, when we have all symbols sorted.
 		 */
-		sym = symbol__new(start, 0xdead, line + len + 2,
-				  self->sym_priv_size, 0, v);
+		sym = symbol__new(start, 0, symbol_name,
+				  kernel_map->dso->sym_priv_size);
 
 		if (sym == NULL)
 			goto out_delete_line;
 
-		if (filter && filter(self, sym))
-			symbol__delete(sym, self->sym_priv_size);
-		else {
-			dso__insert_symbol(self, sym);
-			count++;
-		}
-	}
-
-	/*
-	 * Now that we have all sorted out, just set the ->end of all
-	 * symbols
-	 */
-	prevnd = rb_first(&self->syms);
-
-	if (prevnd == NULL)
-		goto out_delete_line;
-
-	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
-		struct symbol *prev = rb_entry(prevnd, struct symbol, rb_node),
-			      *curr = rb_entry(nd, struct symbol, rb_node);
-
-		prev->end = curr->start - 1;
-		prevnd = nd;
+		dso__insert_symbol(kernel_map->dso, sym);
 	}
 
 	free(line);
 	fclose(file);
 
-	return count;
+	return 0;
 
 out_delete_line:
 	free(line);
@@ -241,14 +270,124 @@ out_failure:
 	return -1;
 }
 
-static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int v)
+/*
+ * Split the symbols into maps, making sure there are no overlaps, i.e. the
+ * kernel range is broken in several maps, named [kernel].N, as we don't have
+ * the original ELF section names vmlinux have.
+ */
+static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
+{
+	struct map *map = kernel_map;
+	struct symbol *pos;
+	int count = 0;
+	struct rb_node *next = rb_first(&kernel_map->dso->syms);
+	int kernel_range = 0;
+
+	while (next) {
+		char *module;
+
+		pos = rb_entry(next, struct symbol, rb_node);
+		next = rb_next(&pos->rb_node);
+
+		module = strchr(pos->name, '\t');
+		if (module) {
+			if (!use_modules)
+				goto delete_symbol;
+
+			*module++ = '\0';
+
+			if (strcmp(map->dso->name, module)) {
+				map = kernel_maps__find_by_dso_name(module);
+				if (!map) {
+					pr_err("/proc/{kallsyms,modules} "
+					       "inconsistency!\n");
+					return -1;
+				}
+			}
+			/*
+			 * So that we look just like we get from .ko files,
+			 * i.e. not prelinked, relative to map->start.
+			 */
+			pos->start = map->map_ip(map, pos->start);
+			pos->end   = map->map_ip(map, pos->end);
+		} else if (map != kernel_map) {
+			char dso_name[PATH_MAX];
+			struct dso *dso;
+
+			snprintf(dso_name, sizeof(dso_name), "[kernel].%d",
+				 kernel_range++);
+
+			dso = dso__new(dso_name,
+				       kernel_map->dso->sym_priv_size);
+			if (dso == NULL)
+				return -1;
+
+			map = map__new2(pos->start, dso);
+			if (map == NULL) {
+				dso__delete(dso);
+				return -1;
+			}
+
+			map->map_ip = map->unmap_ip = identity__map_ip;
+			kernel_maps__insert(map);
+			++kernel_range;
+		}
+
+		if (filter && filter(map, pos)) {
+delete_symbol:
+			rb_erase(&pos->rb_node, &kernel_map->dso->syms);
+			symbol__delete(pos, kernel_map->dso->sym_priv_size);
+		} else {
+			if (map != kernel_map) {
+				rb_erase(&pos->rb_node, &kernel_map->dso->syms);
+				dso__insert_symbol(map->dso, pos);
+			}
+			count++;
+		}
+	}
+
+	return count;
+}
+
+
+static int kernel_maps__load_kallsyms(symbol_filter_t filter, int use_modules)
+{
+	if (kernel_maps__load_all_kallsyms())
+		return -1;
+
+	dso__fixup_sym_end(kernel_map->dso);
+
+	return kernel_maps__split_kallsyms(filter, use_modules);
+}
+
+static size_t kernel_maps__fprintf(FILE *fp)
+{
+	size_t printed = fprintf(fp, "Kernel maps:\n");
+	struct rb_node *nd;
+
+	for (nd = rb_first(&kernel_maps); nd; nd = rb_next(nd)) {
+		struct map *pos = rb_entry(nd, struct map, rb_node);
+
+		printed += fprintf(fp, "Map:");
+		printed += map__fprintf(pos, fp);
+		if (verbose > 1) {
+			printed += dso__fprintf(pos->dso, fp);
+			printed += fprintf(fp, "--\n");
+		}
+	}
+
+	return printed + fprintf(fp, "END kernel maps\n");
+}
+
+static int dso__load_perf_map(struct dso *self, struct map *map,
+			      symbol_filter_t filter)
 {
 	char *line = NULL;
 	size_t n;
 	FILE *file;
 	int nr_syms = 0;
 
-	file = fopen(self->name, "r");
+	file = fopen(self->long_name, "r");
 	if (file == NULL)
 		goto out_failure;
 
@@ -279,12 +418,12 @@ static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int v)
 			continue;
 
 		sym = symbol__new(start, size, line + len,
-				  self->sym_priv_size, start, v);
+				  self->sym_priv_size);
 
 		if (sym == NULL)
 			goto out_delete_line;
 
-		if (filter && filter(self, sym))
+		if (filter && filter(map, sym))
 			symbol__delete(sym, self->sym_priv_size);
 		else {
 			dso__insert_symbol(self, sym);
@@ -393,7 +532,7 @@ static Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep,
  * And always look at the original dso, not at debuginfo packages, that
  * have the PLT data stripped out (shdr_rel_plt.sh_type == SHT_NOBITS).
  */
-static int dso__synthesize_plt_symbols(struct  dso *self, int v)
+static int dso__synthesize_plt_symbols(struct  dso *self)
 {
 	uint32_t nr_rel_entries, idx;
 	GElf_Sym sym;
@@ -409,7 +548,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 	Elf *elf;
 	int nr = 0, symidx, fd, err = 0;
 
-	fd = open(self->name, O_RDONLY);
+	fd = open(self->long_name, O_RDONLY);
 	if (fd < 0)
 		goto out;
 
@@ -477,7 +616,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size, 0, v);
+					sympltname, self->sym_priv_size);
 			if (!f)
 				goto out_elf_end;
 
@@ -495,7 +634,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size, 0, v);
+					sympltname, self->sym_priv_size);
 			if (!f)
 				goto out_elf_end;
 
@@ -513,14 +652,18 @@ out_close:
 	if (err == 0)
 		return nr;
 out:
-	fprintf(stderr, "%s: problems reading %s PLT info.\n",
-		__func__, self->name);
+	pr_warning("%s: problems reading %s PLT info.\n",
+		   __func__, self->long_name);
 	return 0;
 }
 
-static int dso__load_sym(struct dso *self, int fd, const char *name,
-			 symbol_filter_t filter, int v, struct module *mod)
+static int dso__load_sym(struct dso *self, struct map *map, const char *name,
+			 int fd, symbol_filter_t filter, int kernel,
+			 int kmodule)
 {
+	struct map *curr_map = map;
+	struct dso *curr_dso = self;
+	size_t dso_name_len = strlen(self->short_name);
 	Elf_Data *symstrs, *secstrs;
 	uint32_t nr_syms;
 	int err = -1;
@@ -531,19 +674,16 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 	GElf_Sym sym;
 	Elf_Scn *sec, *sec_strndx;
 	Elf *elf;
-	int nr = 0, kernel = !strcmp("[kernel]", self->name);
+	int nr = 0;
 
 	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
 	if (elf == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot read %s ELF file.\n",
-				__func__, name);
+		pr_err("%s: cannot read %s ELF file.\n", __func__, name);
 		goto out_close;
 	}
 
 	if (gelf_getehdr(elf, &ehdr) == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+		pr_err("%s: cannot get elf header.\n", __func__);
 		goto out_elf_end;
 	}
 
@@ -587,9 +727,7 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 	elf_symtab__for_each_symbol(syms, nr_syms, idx, sym) {
 		struct symbol *f;
 		const char *elf_name;
-		char *demangled;
-		u64 obj_start;
-		struct section *section = NULL;
+		char *demangled = NULL;
 		int is_label = elf_sym__is_label(&sym);
 		const char *section_name;
 
@@ -605,52 +743,85 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 		if (is_label && !elf_sec__is_text(&shdr, secstrs))
 			continue;
 
+		elf_name = elf_sym__name(&sym, symstrs);
 		section_name = elf_sec__name(&shdr, secstrs);
-		obj_start = sym.st_value;
 
-		if (self->adjust_symbols) {
-			if (v >= 2)
-				printf("adjusting symbol: st_value: %Lx sh_addr: %Lx sh_offset: %Lx\n",
-					(u64)sym.st_value, (u64)shdr.sh_addr, (u64)shdr.sh_offset);
+		if (kernel || kmodule) {
+			char dso_name[PATH_MAX];
 
-			sym.st_value -= shdr.sh_addr - shdr.sh_offset;
-		}
+			if (strcmp(section_name,
+				   curr_dso->short_name + dso_name_len) == 0)
+				goto new_symbol;
 
-		if (mod) {
-			section = mod->sections->find_section(mod->sections, section_name);
-			if (section)
-				sym.st_value += section->vma;
-			else {
-				fprintf(stderr, "dso__load_sym() module %s lookup of %s failed\n",
-					mod->name, section_name);
-				goto out_elf_end;
+			if (strcmp(section_name, ".text") == 0) {
+				curr_map = map;
+				curr_dso = self;
+				goto new_symbol;
 			}
+
+			snprintf(dso_name, sizeof(dso_name),
+				 "%s%s", self->short_name, section_name);
+
+			curr_map = kernel_maps__find_by_dso_name(dso_name);
+			if (curr_map == NULL) {
+				u64 start = sym.st_value;
+
+				if (kmodule)
+					start += map->start + shdr.sh_offset;
+
+				curr_dso = dso__new(dso_name, self->sym_priv_size);
+				if (curr_dso == NULL)
+					goto out_elf_end;
+				curr_map = map__new2(start, curr_dso);
+				if (curr_map == NULL) {
+					dso__delete(curr_dso);
+					goto out_elf_end;
+				}
+				curr_map->map_ip = identity__map_ip;
+				curr_map->unmap_ip = identity__map_ip;
+				curr_dso->origin = DSO__ORIG_KERNEL;
+				kernel_maps__insert(curr_map);
+				dsos__add(curr_dso);
+			} else
+				curr_dso = curr_map->dso;
+
+			goto new_symbol;
+		}
+
+		if (curr_dso->adjust_symbols) {
+			pr_debug2("adjusting symbol: st_value: %Lx sh_addr: "
+				  "%Lx sh_offset: %Lx\n", (u64)sym.st_value,
+				  (u64)shdr.sh_addr, (u64)shdr.sh_offset);
+			sym.st_value -= shdr.sh_addr - shdr.sh_offset;
 		}
 		/*
 		 * We need to figure out if the object was created from C++ sources
 		 * DWARF DW_compile_unit has this, but we don't always have access
 		 * to it...
 		 */
-		elf_name = elf_sym__name(&sym, symstrs);
 		demangled = bfd_demangle(NULL, elf_name, DMGL_PARAMS | DMGL_ANSI);
 		if (demangled != NULL)
 			elf_name = demangled;
-
+new_symbol:
 		f = symbol__new(sym.st_value, sym.st_size, elf_name,
-				self->sym_priv_size, obj_start, v);
+				curr_dso->sym_priv_size);
 		free(demangled);
 		if (!f)
 			goto out_elf_end;
 
-		if (filter && filter(self, f))
-			symbol__delete(f, self->sym_priv_size);
+		if (filter && filter(curr_map, f))
+			symbol__delete(f, curr_dso->sym_priv_size);
 		else {
-			f->module = mod;
-			dso__insert_symbol(self, f);
+			dso__insert_symbol(curr_dso, f);
 			nr++;
 		}
 	}
 
+	/*
+	 * For misannotated, zeroed, ASM function sizes.
+	 */
+	if (nr > 0)
+		dso__fixup_sym_end(self);
 	err = nr;
 out_elf_end:
 	elf_end(elf);
@@ -660,7 +831,7 @@ out_close:
 
 #define BUILD_ID_SIZE 128
 
-static char *dso__read_build_id(struct dso *self, int v)
+static char *dso__read_build_id(struct dso *self)
 {
 	int i;
 	GElf_Ehdr ehdr;
@@ -670,22 +841,20 @@ static char *dso__read_build_id(struct dso *self, int v)
 	char *build_id = NULL, *bid;
 	unsigned char *raw;
 	Elf *elf;
-	int fd = open(self->name, O_RDONLY);
+	int fd = open(self->long_name, O_RDONLY);
 
 	if (fd < 0)
 		goto out;
 
 	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
 	if (elf == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot read %s ELF file.\n",
-				__func__, self->name);
+		pr_err("%s: cannot read %s ELF file.\n", __func__,
+		       self->long_name);
 		goto out_close;
 	}
 
 	if (gelf_getehdr(elf, &ehdr) == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+		pr_err("%s: cannot get elf header.\n", __func__);
 		goto out_elf_end;
 	}
 
@@ -707,8 +876,7 @@ static char *dso__read_build_id(struct dso *self, int v)
 		++raw;
 		bid += 2;
 	}
-	if (v >= 2)
-		printf("%s(%s): %s\n", __func__, self->name, build_id);
+	pr_debug2("%s(%s): %s\n", __func__, self->long_name, build_id);
 out_elf_end:
 	elf_end(elf);
 out_close:
@@ -726,6 +894,7 @@ char dso__symtab_origin(const struct dso *self)
 		[DSO__ORIG_UBUNTU] =   'u',
 		[DSO__ORIG_BUILDID] =  'b',
 		[DSO__ORIG_DSO] =      'd',
+		[DSO__ORIG_KMODULE] =  'K',
 	};
 
 	if (self == NULL || self->origin == DSO__ORIG_NOT_FOUND)
@@ -733,7 +902,7 @@ char dso__symtab_origin(const struct dso *self)
 	return origin[self->origin];
 }
 
-int dso__load(struct dso *self, symbol_filter_t filter, int v)
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 {
 	int size = PATH_MAX;
 	char *name = malloc(size), *build_id = NULL;
@@ -746,7 +915,7 @@ int dso__load(struct dso *self, symbol_filter_t filter, int v)
 	self->adjust_symbols = 0;
 
 	if (strncmp(self->name, "/tmp/perf-", 10) == 0) {
-		ret = dso__load_perf_map(self, filter, v);
+		ret = dso__load_perf_map(self, map, filter);
 		self->origin = ret > 0 ? DSO__ORIG_JAVA_JIT :
 					 DSO__ORIG_NOT_FOUND;
 		return ret;
@@ -759,13 +928,15 @@ more:
 		self->origin++;
 		switch (self->origin) {
 		case DSO__ORIG_FEDORA:
-			snprintf(name, size, "/usr/lib/debug%s.debug", self->name);
+			snprintf(name, size, "/usr/lib/debug%s.debug",
+				 self->long_name);
 			break;
 		case DSO__ORIG_UBUNTU:
-			snprintf(name, size, "/usr/lib/debug%s", self->name);
+			snprintf(name, size, "/usr/lib/debug%s",
+				 self->long_name);
 			break;
 		case DSO__ORIG_BUILDID:
-			build_id = dso__read_build_id(self, v);
+			build_id = dso__read_build_id(self);
 			if (build_id != NULL) {
 				snprintf(name, size,
 					 "/usr/lib/debug/.build-id/%.2s/%s.debug",
@@ -776,7 +947,7 @@ more:
 			self->origin++;
 			/* Fall thru */
 		case DSO__ORIG_DSO:
-			snprintf(name, size, "%s", self->name);
+			snprintf(name, size, "%s", self->long_name);
 			break;
 
 		default:
@@ -786,7 +957,7 @@ more:
 		fd = open(name, O_RDONLY);
 	} while (fd < 0);
 
-	ret = dso__load_sym(self, fd, name, filter, v, NULL);
+	ret = dso__load_sym(self, map, name, fd, filter, 0, 0);
 	close(fd);
 
 	/*
@@ -796,7 +967,7 @@ more:
 		goto more;
 
 	if (ret > 0) {
-		int nr_plt = dso__synthesize_plt_symbols(self, v);
+		int nr_plt = dso__synthesize_plt_symbols(self);
 		if (nr_plt > 0)
 			ret += nr_plt;
 	}
@@ -807,137 +978,321 @@ out:
 	return ret;
 }
 
-static int dso__load_module(struct dso *self, struct mod_dso *mods, const char *name,
-			     symbol_filter_t filter, int v)
+struct map *kernel_map;
+
+static void kernel_maps__insert(struct map *map)
 {
-	struct module *mod = mod_dso__find_module(mods, name);
-	int err = 0, fd;
+	maps__insert(&kernel_maps, map);
+}
 
-	if (mod == NULL || !mod->active)
-		return err;
+struct symbol *kernel_maps__find_symbol(u64 ip, struct map **mapp)
+{
+	struct map *map = maps__find(&kernel_maps, ip);
 
-	fd = open(mod->path, O_RDONLY);
+	if (mapp)
+		*mapp = map;
 
-	if (fd < 0)
+	if (map) {
+		ip = map->map_ip(map, ip);
+		return map->dso->find_symbol(map->dso, ip);
+	}
+
+	return NULL;
+}
+
+struct map *kernel_maps__find_by_dso_name(const char *name)
+{
+	struct rb_node *nd;
+
+	for (nd = rb_first(&kernel_maps); nd; nd = rb_next(nd)) {
+		struct map *map = rb_entry(nd, struct map, rb_node);
+
+		if (map->dso && strcmp(map->dso->name, name) == 0)
+			return map;
+	}
+
+	return NULL;
+}
+
+static int dso__load_module_sym(struct dso *self, struct map *map,
+				symbol_filter_t filter)
+{
+	int err = 0, fd = open(self->long_name, O_RDONLY);
+
+	if (fd < 0) {
+		pr_err("%s: cannot open %s\n", __func__, self->long_name);
 		return err;
+	}
 
-	err = dso__load_sym(self, fd, name, filter, v, mod);
+	err = dso__load_sym(self, map, self->long_name, fd, filter, 0, 1);
 	close(fd);
 
 	return err;
 }
 
-int dso__load_modules(struct dso *self, symbol_filter_t filter, int v)
+static int dsos__load_modules_sym_dir(char *dirname, symbol_filter_t filter)
 {
-	struct mod_dso *mods = mod_dso__new_dso("modules");
-	struct module *pos;
-	struct rb_node *next;
-	int err, count = 0;
+	struct dirent *dent;
+	int nr_symbols = 0, err;
+	DIR *dir = opendir(dirname);
 
-	err = mod_dso__load_modules(mods);
+	if (!dir) {
+		pr_err("%s: cannot open %s dir\n", __func__, dirname);
+		return -1;
+	}
 
-	if (err <= 0)
-		return err;
+	while ((dent = readdir(dir)) != NULL) {
+		char path[PATH_MAX];
+
+		if (dent->d_type == DT_DIR) {
+			if (!strcmp(dent->d_name, ".") ||
+			    !strcmp(dent->d_name, ".."))
+				continue;
+
+			snprintf(path, sizeof(path), "%s/%s",
+				 dirname, dent->d_name);
+			err = dsos__load_modules_sym_dir(path, filter);
+			if (err < 0)
+				goto failure;
+		} else {
+			char *dot = strrchr(dent->d_name, '.'),
+			     dso_name[PATH_MAX];
+			struct map *map;
+			struct rb_node *last;
+
+			if (dot == NULL || strcmp(dot, ".ko"))
+				continue;
+			snprintf(dso_name, sizeof(dso_name), "[%.*s]",
+				 (int)(dot - dent->d_name), dent->d_name);
+
+			strxfrchar(dso_name, '-', '_');
+			map = kernel_maps__find_by_dso_name(dso_name);
+			if (map == NULL)
+				continue;
+
+			snprintf(path, sizeof(path), "%s/%s",
+				 dirname, dent->d_name);
+
+			map->dso->long_name = strdup(path);
+			if (map->dso->long_name == NULL)
+				goto failure;
+
+			err = dso__load_module_sym(map->dso, map, filter);
+			if (err < 0)
+				goto failure;
+			last = rb_last(&map->dso->syms);
+			if (last) {
+				struct symbol *sym;
+				/*
+				 * We do this here as well, even having the
+				 * symbol size found in the symtab because
+				 * misannotated ASM symbols may have the size
+				 * set to zero.
+				 */
+				dso__fixup_sym_end(map->dso);
+
+				sym = rb_entry(last, struct symbol, rb_node);
+				map->end = map->start + sym->end;
+			}
+		}
+		nr_symbols += err;
+	}
 
-	/*
-	 * Iterate over modules, and load active symbols.
-	 */
-	next = rb_first(&mods->mods);
-	while (next) {
-		pos = rb_entry(next, struct module, rb_node);
-		err = dso__load_module(self, mods, pos->name, filter, v);
+	return nr_symbols;
+failure:
+	closedir(dir);
+	return -1;
+}
 
-		if (err < 0)
-			break;
+static int dsos__load_modules_sym(symbol_filter_t filter)
+{
+	struct utsname uts;
+	char modules_path[PATH_MAX];
 
-		next = rb_next(&pos->rb_node);
-		count += err;
-	}
+	if (uname(&uts) < 0)
+		return -1;
 
-	if (err < 0) {
-		mod_dso__delete_modules(mods);
-		mod_dso__delete_self(mods);
-		return err;
-	}
+	snprintf(modules_path, sizeof(modules_path), "/lib/modules/%s/kernel",
+		 uts.release);
 
-	return count;
+	return dsos__load_modules_sym_dir(modules_path, filter);
 }
 
-static inline void dso__fill_symbol_holes(struct dso *self)
+/*
+ * Constructor variant for modules (where we know from /proc/modules where
+ * they are loaded) and for vmlinux, where only after we load all the
+ * symbols we'll know where it starts and ends.
+ */
+static struct map *map__new2(u64 start, struct dso *dso)
 {
-	struct symbol *prev = NULL;
-	struct rb_node *nd;
+	struct map *self = malloc(sizeof(*self));
 
-	for (nd = rb_last(&self->syms); nd; nd = rb_prev(nd)) {
-		struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
+	if (self != NULL) {
+		self->start = start;
+		/*
+		 * Will be filled after we load all the symbols
+		 */
+		self->end = 0;
 
-		if (prev) {
-			u64 hole = 0;
-			int alias = pos->start == prev->start;
+		self->pgoff = 0;
+		self->dso = dso;
+		self->map_ip = map__map_ip;
+		self->unmap_ip = map__unmap_ip;
+		RB_CLEAR_NODE(&self->rb_node);
+	}
+	return self;
+}
 
-			if (!alias)
-				hole = prev->start - pos->end - 1;
+static int dsos__load_modules(unsigned int sym_priv_size)
+{
+	char *line = NULL;
+	size_t n;
+	FILE *file = fopen("/proc/modules", "r");
+	struct map *map;
 
-			if (hole || alias) {
-				if (alias)
-					pos->end = prev->end;
-				else if (hole)
-					pos->end = prev->start - 1;
-			}
+	if (file == NULL)
+		return -1;
+
+	while (!feof(file)) {
+		char name[PATH_MAX];
+		u64 start;
+		struct dso *dso;
+		char *sep;
+		int line_len;
+
+		line_len = getline(&line, &n, file);
+		if (line_len < 0)
+			break;
+
+		if (!line)
+			goto out_failure;
+
+		line[--line_len] = '\0'; /* \n */
+
+		sep = strrchr(line, 'x');
+		if (sep == NULL)
+			continue;
+
+		hex2u64(sep + 1, &start);
+
+		sep = strchr(line, ' ');
+		if (sep == NULL)
+			continue;
+
+		*sep = '\0';
+
+		snprintf(name, sizeof(name), "[%s]", line);
+		dso = dso__new(name, sym_priv_size);
+
+		if (dso == NULL)
+			goto out_delete_line;
+
+		map = map__new2(start, dso);
+		if (map == NULL) {
+			dso__delete(dso);
+			goto out_delete_line;
 		}
-		prev = pos;
+
+		dso->origin = DSO__ORIG_KMODULE;
+		kernel_maps__insert(map);
+		dsos__add(dso);
 	}
+
+	free(line);
+	fclose(file);
+
+	return 0;
+
+out_delete_line:
+	free(line);
+out_failure:
+	return -1;
 }
 
-static int dso__load_vmlinux(struct dso *self, const char *vmlinux,
-			     symbol_filter_t filter, int v)
+static int dso__load_vmlinux(struct dso *self, struct map *map,
+			     const char *vmlinux, symbol_filter_t filter)
 {
 	int err, fd = open(vmlinux, O_RDONLY);
 
 	if (fd < 0)
 		return -1;
 
-	err = dso__load_sym(self, fd, vmlinux, filter, v, NULL);
-
-	if (err > 0)
-		dso__fill_symbol_holes(self);
+	err = dso__load_sym(self, map, self->long_name, fd, filter, 1, 0);
 
 	close(fd);
 
 	return err;
 }
 
-int dso__load_kernel(struct dso *self, const char *vmlinux,
-		     symbol_filter_t filter, int v, int use_modules)
+int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
+		      symbol_filter_t filter, int use_modules)
 {
 	int err = -1;
+	struct dso *dso = dso__new(vmlinux, sym_priv_size);
+
+	if (dso == NULL)
+		return -1;
+
+	dso->short_name = "[kernel]";
+	kernel_map = map__new2(0, dso);
+	if (kernel_map == NULL)
+		goto out_delete_dso;
+
+	kernel_map->map_ip = kernel_map->unmap_ip = identity__map_ip;
+
+	if (use_modules && dsos__load_modules(sym_priv_size) < 0) {
+		pr_warning("Failed to load list of modules in use! "
+			   "Continuing...\n");
+		use_modules = 0;
+	}
 
 	if (vmlinux) {
-		err = dso__load_vmlinux(self, vmlinux, filter, v);
+		err = dso__load_vmlinux(dso, kernel_map, vmlinux, filter);
 		if (err > 0 && use_modules) {
-			int syms = dso__load_modules(self, filter, v);
+			int syms = dsos__load_modules_sym(filter);
 
-			if (syms < 0) {
-				fprintf(stderr, "dso__load_modules failed!\n");
-				return syms;
-			}
-			err += syms;
+			if (syms < 0)
+				pr_warning("Failed to read module symbols!"
+					   " Continuing...\n");
+			else
+				err += syms;
 		}
 	}
 
 	if (err <= 0)
-		err = dso__load_kallsyms(self, filter, v);
+		err = kernel_maps__load_kallsyms(filter, use_modules);
+
+	if (err > 0) {
+		struct rb_node *node = rb_first(&dso->syms);
+		struct symbol *sym = rb_entry(node, struct symbol, rb_node);
+
+		kernel_map->start = sym->start;
+		node = rb_last(&dso->syms);
+		sym = rb_entry(node, struct symbol, rb_node);
+		kernel_map->end = sym->end;
+
+		dso->origin = DSO__ORIG_KERNEL;
+		kernel_maps__insert(kernel_map);
+		/*
+		 * Now that we have all sorted out, just set the ->end of all
+		 * maps:
+		 */
+		kernel_maps__fixup_end();
+		dsos__add(dso);
 
-	if (err > 0)
-		self->origin = DSO__ORIG_KERNEL;
+		if (verbose)
+			kernel_maps__fprintf(stderr);
+	}
 
 	return err;
+
+out_delete_dso:
+	dso__delete(dso);
+	return -1;
 }
 
 LIST_HEAD(dsos);
-struct dso	*kernel_dso;
 struct dso	*vdso;
-struct dso	*hypervisor_dso;
 
 const char	*vmlinux_name = "vmlinux";
 int		modules;
@@ -957,33 +1312,21 @@ static struct dso *dsos__find(const char *name)
 	return NULL;
 }
 
-struct dso *dsos__findnew(const char *name)
+struct dso *dsos__findnew(const char *name, unsigned int sym_priv_size,
+			  bool *is_new)
 {
 	struct dso *dso = dsos__find(name);
-	int nr;
 
-	if (dso)
-		return dso;
-
-	dso = dso__new(name, 0);
-	if (!dso)
-		goto out_delete_dso;
-
-	nr = dso__load(dso, NULL, verbose);
-	if (nr < 0) {
-		eprintf("Failed to open: %s\n", name);
-		goto out_delete_dso;
-	}
-	if (!nr)
-		eprintf("No symbols found in: %s, maybe install a debug package?\n", name);
-
-	dsos__add(dso);
+	if (!dso) {
+		dso = dso__new(name, sym_priv_size);
+		if (dso) {
+			dsos__add(dso);
+			*is_new = true;
+		}
+	} else
+		*is_new = false;
 
 	return dso;
-
-out_delete_dso:
-	dso__delete(dso);
-	return NULL;
 }
 
 void dsos__fprintf(FILE *fp)
@@ -994,43 +1337,21 @@ void dsos__fprintf(FILE *fp)
 		dso__fprintf(pos, fp);
 }
 
-static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip)
-{
-	return dso__find_symbol(dso, ip);
-}
-
-int load_kernel(void)
+int load_kernel(unsigned int sym_priv_size, symbol_filter_t filter)
 {
-	int err;
-
-	kernel_dso = dso__new("[kernel]", 0);
-	if (!kernel_dso)
+	if (dsos__load_kernel(vmlinux_name, sym_priv_size, filter,
+			      modules) <= 0)
 		return -1;
 
-	err = dso__load_kernel(kernel_dso, vmlinux_name, NULL, verbose, modules);
-	if (err <= 0) {
-		dso__delete(kernel_dso);
-		kernel_dso = NULL;
-	} else
-		dsos__add(kernel_dso);
-
 	vdso = dso__new("[vdso]", 0);
 	if (!vdso)
 		return -1;
 
-	vdso->find_symbol = vdso__find_symbol;
-
 	dsos__add(vdso);
 
-	hypervisor_dso = dso__new("[hypervisor]", 0);
-	if (!hypervisor_dso)
-		return -1;
-	dsos__add(hypervisor_dso);
-
-	return err;
+	return 0;
 }
 
-
 void symbol__init(void)
 {
 	elf_version(EV_CURRENT);
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 6e849071640..77b7b3e4241 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -1,11 +1,11 @@
-#ifndef _PERF_SYMBOL_
-#define _PERF_SYMBOL_ 1
+#ifndef __PERF_SYMBOL
+#define __PERF_SYMBOL 1
 
 #include <linux/types.h>
+#include <stdbool.h>
 #include "types.h"
 #include <linux/list.h>
 #include <linux/rbtree.h>
-#include "module.h"
 #include "event.h"
 
 #ifdef HAVE_CPLUS_DEMANGLE
@@ -36,11 +36,6 @@ struct symbol {
 	struct rb_node	rb_node;
 	u64		start;
 	u64		end;
-	u64		obj_start;
-	u64		hist_sum;
-	u64		*hist;
-	struct module	*module;
-	void		*priv;
 	char		name[0];
 };
 
@@ -52,13 +47,11 @@ struct dso {
 	unsigned char	 adjust_symbols;
 	unsigned char	 slen_calculated;
 	unsigned char	 origin;
+	const char	 *short_name;
+	char	 	 *long_name;
 	char		 name[0];
 };
 
-extern const char *sym_hist_filter;
-
-typedef int (*symbol_filter_t)(struct dso *self, struct symbol *sym);
-
 struct dso *dso__new(const char *name, unsigned int sym_priv_size);
 void dso__delete(struct dso *self);
 
@@ -69,24 +62,23 @@ static inline void *dso__sym_priv(struct dso *self, struct symbol *sym)
 
 struct symbol *dso__find_symbol(struct dso *self, u64 ip);
 
-int dso__load_kernel(struct dso *self, const char *vmlinux,
-		     symbol_filter_t filter, int verbose, int modules);
-int dso__load_modules(struct dso *self, symbol_filter_t filter, int verbose);
-int dso__load(struct dso *self, symbol_filter_t filter, int verbose);
-struct dso *dsos__findnew(const char *name);
+int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
+		      symbol_filter_t filter, int modules);
+struct dso *dsos__findnew(const char *name, unsigned int sym_priv_size,
+			  bool *is_new);
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter);
 void dsos__fprintf(FILE *fp);
 
 size_t dso__fprintf(struct dso *self, FILE *fp);
 char dso__symtab_origin(const struct dso *self);
 
-int load_kernel(void);
+int load_kernel(unsigned int sym_priv_size, symbol_filter_t filter);
 
 void symbol__init(void);
 
 extern struct list_head dsos;
-extern struct dso *kernel_dso;
+extern struct map *kernel_map;
 extern struct dso *vdso;
-extern struct dso *hypervisor_dso;
 extern const char *vmlinux_name;
 extern int   modules;
-#endif /* _PERF_SYMBOL_ */
+#endif /* __PERF_SYMBOL */
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index 45efb5db0d1..0f6d78c9863 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -6,6 +6,9 @@
 #include "util.h"
 #include "debug.h"
 
+static struct rb_root threads;
+static struct thread *last_match;
+
 static struct thread *thread__new(pid_t pid)
 {
 	struct thread *self = calloc(1, sizeof(*self));
@@ -15,7 +18,8 @@ static struct thread *thread__new(pid_t pid)
 		self->comm = malloc(32);
 		if (self->comm)
 			snprintf(self->comm, 32, ":%d", self->pid);
-		INIT_LIST_HEAD(&self->maps);
+		self->maps = RB_ROOT;
+		INIT_LIST_HEAD(&self->removed_maps);
 	}
 
 	return self;
@@ -29,21 +33,40 @@ int thread__set_comm(struct thread *self, const char *comm)
 	return self->comm ? 0 : -ENOMEM;
 }
 
+int thread__comm_len(struct thread *self)
+{
+	if (!self->comm_len) {
+		if (!self->comm)
+			return 0;
+		self->comm_len = strlen(self->comm);
+	}
+
+	return self->comm_len;
+}
+
 static size_t thread__fprintf(struct thread *self, FILE *fp)
 {
+	struct rb_node *nd;
 	struct map *pos;
-	size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm);
+	size_t ret = fprintf(fp, "Thread %d %s\nCurrent maps:\n",
+			     self->pid, self->comm);
 
-	list_for_each_entry(pos, &self->maps, node)
+	for (nd = rb_first(&self->maps); nd; nd = rb_next(nd)) {
+		pos = rb_entry(nd, struct map, rb_node);
+		ret += map__fprintf(pos, fp);
+	}
+
+	ret = fprintf(fp, "Removed maps:\n");
+
+	list_for_each_entry(pos, &self->removed_maps, node)
 		ret += map__fprintf(pos, fp);
 
 	return ret;
 }
 
-struct thread *
-threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
+struct thread *threads__findnew(pid_t pid)
 {
-	struct rb_node **p = &threads->rb_node;
+	struct rb_node **p = &threads.rb_node;
 	struct rb_node *parent = NULL;
 	struct thread *th;
 
@@ -52,15 +75,15 @@ threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
 	 * so most of the time we dont have to look up
 	 * the full rbtree:
 	 */
-	if (*last_match && (*last_match)->pid == pid)
-		return *last_match;
+	if (last_match && last_match->pid == pid)
+		return last_match;
 
 	while (*p != NULL) {
 		parent = *p;
 		th = rb_entry(parent, struct thread, rb_node);
 
 		if (th->pid == pid) {
-			*last_match = th;
+			last_match = th;
 			return th;
 		}
 
@@ -73,17 +96,16 @@ threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
 	th = thread__new(pid);
 	if (th != NULL) {
 		rb_link_node(&th->rb_node, parent, p);
-		rb_insert_color(&th->rb_node, threads);
-		*last_match = th;
+		rb_insert_color(&th->rb_node, &threads);
+		last_match = th;
 	}
 
 	return th;
 }
 
-struct thread *
-register_idle_thread(struct rb_root *threads, struct thread **last_match)
+struct thread *register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0, threads, last_match);
+	struct thread *thread = threads__findnew(0);
 
 	if (!thread || thread__set_comm(thread, "swapper")) {
 		fprintf(stderr, "problem inserting idle task.\n");
@@ -93,42 +115,82 @@ register_idle_thread(struct rb_root *threads, struct thread **last_match)
 	return thread;
 }
 
-void thread__insert_map(struct thread *self, struct map *map)
+static void thread__remove_overlappings(struct thread *self, struct map *map)
 {
-	struct map *pos, *tmp;
-
-	list_for_each_entry_safe(pos, tmp, &self->maps, node) {
-		if (map__overlap(pos, map)) {
-			if (verbose >= 2) {
-				printf("overlapping maps:\n");
-				map__fprintf(map, stdout);
-				map__fprintf(pos, stdout);
-			}
-
-			if (map->start <= pos->start && map->end > pos->start)
-				pos->start = map->end;
-
-			if (map->end >= pos->end && map->start < pos->end)
-				pos->end = map->start;
-
-			if (verbose >= 2) {
-				printf("after collision:\n");
-				map__fprintf(pos, stdout);
-			}
-
-			if (pos->start >= pos->end) {
-				list_del_init(&pos->node);
-				free(pos);
-			}
+	struct rb_node *next = rb_first(&self->maps);
+
+	while (next) {
+		struct map *pos = rb_entry(next, struct map, rb_node);
+		next = rb_next(&pos->rb_node);
+
+		if (!map__overlap(pos, map))
+			continue;
+
+		if (verbose >= 2) {
+			fputs("overlapping maps:\n", stderr);
+			map__fprintf(map, stderr);
+			map__fprintf(pos, stderr);
 		}
+
+		rb_erase(&pos->rb_node, &self->maps);
+		/*
+		 * We may have references to this map, for instance in some
+		 * hist_entry instances, so just move them to a separate
+		 * list.
+		 */
+		list_add_tail(&pos->node, &self->removed_maps);
+	}
+}
+
+void maps__insert(struct rb_root *maps, struct map *map)
+{
+	struct rb_node **p = &maps->rb_node;
+	struct rb_node *parent = NULL;
+	const u64 ip = map->start;
+	struct map *m;
+
+	while (*p != NULL) {
+		parent = *p;
+		m = rb_entry(parent, struct map, rb_node);
+		if (ip < m->start)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
 	}
 
-	list_add_tail(&map->node, &self->maps);
+	rb_link_node(&map->rb_node, parent, p);
+	rb_insert_color(&map->rb_node, maps);
+}
+
+struct map *maps__find(struct rb_root *maps, u64 ip)
+{
+	struct rb_node **p = &maps->rb_node;
+	struct rb_node *parent = NULL;
+	struct map *m;
+
+	while (*p != NULL) {
+		parent = *p;
+		m = rb_entry(parent, struct map, rb_node);
+		if (ip < m->start)
+			p = &(*p)->rb_left;
+		else if (ip > m->end)
+			p = &(*p)->rb_right;
+		else
+			return m;
+	}
+
+	return NULL;
+}
+
+void thread__insert_map(struct thread *self, struct map *map)
+{
+	thread__remove_overlappings(self, map);
+	maps__insert(&self->maps, map);
 }
 
 int thread__fork(struct thread *self, struct thread *parent)
 {
-	struct map *map;
+	struct rb_node *nd;
 
 	if (self->comm)
 		free(self->comm);
@@ -136,7 +198,8 @@ int thread__fork(struct thread *self, struct thread *parent)
 	if (!self->comm)
 		return -ENOMEM;
 
-	list_for_each_entry(map, &parent->maps, node) {
+	for (nd = rb_first(&parent->maps); nd; nd = rb_next(nd)) {
+		struct map *map = rb_entry(nd, struct map, rb_node);
 		struct map *new = map__clone(map);
 		if (!new)
 			return -ENOMEM;
@@ -146,26 +209,12 @@ int thread__fork(struct thread *self, struct thread *parent)
 	return 0;
 }
 
-struct map *thread__find_map(struct thread *self, u64 ip)
-{
-	struct map *pos;
-
-	if (self == NULL)
-		return NULL;
-
-	list_for_each_entry(pos, &self->maps, node)
-		if (ip >= pos->start && ip <= pos->end)
-			return pos;
-
-	return NULL;
-}
-
-size_t threads__fprintf(FILE *fp, struct rb_root *threads)
+size_t threads__fprintf(FILE *fp)
 {
 	size_t ret = 0;
 	struct rb_node *nd;
 
-	for (nd = rb_first(threads); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&threads); nd; nd = rb_next(nd)) {
 		struct thread *pos = rb_entry(nd, struct thread, rb_node);
 
 		ret += thread__fprintf(pos, fp);
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
index 32aea3c1c2a..53addd77ce8 100644
--- a/tools/perf/util/thread.h
+++ b/tools/perf/util/thread.h
@@ -1,22 +1,37 @@
+#ifndef __PERF_THREAD_H
+#define __PERF_THREAD_H
+
 #include <linux/rbtree.h>
-#include <linux/list.h>
 #include <unistd.h>
 #include "symbol.h"
 
 struct thread {
 	struct rb_node		rb_node;
-	struct list_head	maps;
+	struct rb_root		maps;
+	struct list_head	removed_maps;
 	pid_t			pid;
 	char			shortname[3];
 	char			*comm;
+	int			comm_len;
 };
 
 int thread__set_comm(struct thread *self, const char *comm);
-struct thread *
-threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match);
-struct thread *
-register_idle_thread(struct rb_root *threads, struct thread **last_match);
+int thread__comm_len(struct thread *self);
+struct thread *threads__findnew(pid_t pid);
+struct thread *register_idle_thread(void);
 void thread__insert_map(struct thread *self, struct map *map);
 int thread__fork(struct thread *self, struct thread *parent);
-struct map *thread__find_map(struct thread *self, u64 ip);
-size_t threads__fprintf(FILE *fp, struct rb_root *threads);
+size_t threads__fprintf(FILE *fp);
+
+void maps__insert(struct rb_root *maps, struct map *map);
+struct map *maps__find(struct rb_root *maps, u64 ip);
+
+struct symbol *kernel_maps__find_symbol(const u64 ip, struct map **mapp);
+struct map *kernel_maps__find_by_dso_name(const char *name);
+
+static inline struct map *thread__find_map(struct thread *self, u64 ip)
+{
+	return self ? maps__find(&self->maps, ip) : NULL;
+}
+
+#endif	/* __PERF_THREAD_H */
diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c
index af4b0573b37..831052d4b4f 100644
--- a/tools/perf/util/trace-event-info.c
+++ b/tools/perf/util/trace-event-info.c
@@ -496,14 +496,12 @@ get_tracepoints_path(struct perf_event_attr *pattrs, int nb_events)
 
 	return path.next;
 }
-void read_tracing_data(struct perf_event_attr *pattrs, int nb_events)
+void read_tracing_data(int fd, struct perf_event_attr *pattrs, int nb_events)
 {
 	char buf[BUFSIZ];
 	struct tracepoint_path *tps;
 
-	output_fd = open(output_file, O_WRONLY | O_CREAT | O_TRUNC | O_LARGEFILE, 0644);
-	if (output_fd < 0)
-		die("creating file '%s'", output_file);
+	output_fd = fd;
 
 	buf[0] = 23;
 	buf[1] = 8;
diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c
index 55c9659a56e..eae56050308 100644
--- a/tools/perf/util/trace-event-parse.c
+++ b/tools/perf/util/trace-event-parse.c
@@ -40,6 +40,8 @@ int header_page_size_size;
 int header_page_data_offset;
 int header_page_data_size;
 
+int latency_format;
+
 static char *input_buf;
 static unsigned long long input_buf_ptr;
 static unsigned long long input_buf_siz;
@@ -284,18 +286,19 @@ void parse_ftrace_printk(char *file, unsigned int size __unused)
 	char *line;
 	char *next = NULL;
 	char *addr_str;
-	int ret;
 	int i;
 
 	line = strtok_r(file, "\n", &next);
 	while (line) {
+		addr_str = strsep(&line, ":");
+		if (!line) {
+			warning("error parsing print strings");
+			break;
+		}
 		item = malloc_or_die(sizeof(*item));
-		ret = sscanf(line, "%as : %as",
-			     (float *)(void *)&addr_str, /* workaround gcc warning */
-			     (float *)(void *)&item->printk);
 		item->addr = strtoull(addr_str, NULL, 16);
-		free(addr_str);
-
+		/* fmt still has a space, skip it */
+		item->printk = strdup(line+1);
 		item->next = list;
 		list = item;
 		line = strtok_r(NULL, "\n", &next);
@@ -522,7 +525,10 @@ static enum event_type __read_token(char **tok)
 			last_ch = ch;
 			ch = __read_char();
 			buf[i++] = ch;
-		} while (ch != quote_ch && last_ch != '\\');
+			/* the '\' '\' will cancel itself */
+			if (ch == '\\' && last_ch == '\\')
+				last_ch = 0;
+		} while (ch != quote_ch || last_ch == '\\');
 		/* remove the last quote */
 		i--;
 		goto out;
@@ -610,7 +616,7 @@ static enum event_type read_token_item(char **tok)
 static int test_type(enum event_type type, enum event_type expect)
 {
 	if (type != expect) {
-		die("Error: expected type %d but read %d",
+		warning("Error: expected type %d but read %d",
 		    expect, type);
 		return -1;
 	}
@@ -621,13 +627,13 @@ static int test_type_token(enum event_type type, char *token,
 		    enum event_type expect, const char *expect_tok)
 {
 	if (type != expect) {
-		die("Error: expected type %d but read %d",
+		warning("Error: expected type %d but read %d",
 		    expect, type);
 		return -1;
 	}
 
 	if (strcmp(token, expect_tok) != 0) {
-		die("Error: expected '%s' but read '%s'",
+		warning("Error: expected '%s' but read '%s'",
 		    expect_tok, token);
 		return -1;
 	}
@@ -665,7 +671,7 @@ static int __read_expected(enum event_type expect, const char *str, int newline_
 
 	free_token(token);
 
-	return 0;
+	return ret;
 }
 
 static int read_expected(enum event_type expect, const char *str)
@@ -682,10 +688,10 @@ static char *event_read_name(void)
 {
 	char *token;
 
-	if (read_expected(EVENT_ITEM, (char *)"name") < 0)
+	if (read_expected(EVENT_ITEM, "name") < 0)
 		return NULL;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return NULL;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -703,10 +709,10 @@ static int event_read_id(void)
 	char *token;
 	int id;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"ID") < 0)
+	if (read_expected_item(EVENT_ITEM, "ID") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -721,6 +727,24 @@ static int event_read_id(void)
 	return -1;
 }
 
+static int field_is_string(struct format_field *field)
+{
+	if ((field->flags & FIELD_IS_ARRAY) &&
+	    (!strstr(field->type, "char") || !strstr(field->type, "u8") ||
+	     !strstr(field->type, "s8")))
+		return 1;
+
+	return 0;
+}
+
+static int field_is_dynamic(struct format_field *field)
+{
+	if (!strcmp(field->type, "__data_loc"))
+		return 1;
+
+	return 0;
+}
+
 static int event_read_fields(struct event *event, struct format_field **fields)
 {
 	struct format_field *field = NULL;
@@ -738,7 +762,7 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 
 		count++;
 
-		if (test_type_token(type, token, EVENT_ITEM, (char *)"field"))
+		if (test_type_token(type, token, EVENT_ITEM, "field"))
 			goto fail;
 		free_token(token);
 
@@ -753,7 +777,7 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 			type = read_token(&token);
 		}
 
-		if (test_type_token(type, token, EVENT_OP, (char *)":") < 0)
+		if (test_type_token(type, token, EVENT_OP, ":") < 0)
 			return -1;
 
 		if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -865,14 +889,20 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 			free(brackets);
 		}
 
-		if (test_type_token(type, token,  EVENT_OP, (char *)";"))
+		if (field_is_string(field)) {
+			field->flags |= FIELD_IS_STRING;
+			if (field_is_dynamic(field))
+				field->flags |= FIELD_IS_DYNAMIC;
+		}
+
+		if (test_type_token(type, token,  EVENT_OP, ";"))
 			goto fail;
 		free_token(token);
 
-		if (read_expected(EVENT_ITEM, (char *)"offset") < 0)
+		if (read_expected(EVENT_ITEM, "offset") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_OP, (char *)":") < 0)
+		if (read_expected(EVENT_OP, ":") < 0)
 			goto fail_expect;
 
 		if (read_expect_type(EVENT_ITEM, &token))
@@ -880,13 +910,13 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 		field->offset = strtoul(token, NULL, 0);
 		free_token(token);
 
-		if (read_expected(EVENT_OP, (char *)";") < 0)
+		if (read_expected(EVENT_OP, ";") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_ITEM, (char *)"size") < 0)
+		if (read_expected(EVENT_ITEM, "size") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_OP, (char *)":") < 0)
+		if (read_expected(EVENT_OP, ":") < 0)
 			goto fail_expect;
 
 		if (read_expect_type(EVENT_ITEM, &token))
@@ -894,11 +924,33 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 		field->size = strtoul(token, NULL, 0);
 		free_token(token);
 
-		if (read_expected(EVENT_OP, (char *)";") < 0)
+		if (read_expected(EVENT_OP, ";") < 0)
 			goto fail_expect;
 
-		if (read_expect_type(EVENT_NEWLINE, &token) < 0)
-			goto fail;
+		type = read_token(&token);
+		if (type != EVENT_NEWLINE) {
+			/* newer versions of the kernel have a "signed" type */
+			if (test_type_token(type, token, EVENT_ITEM, "signed"))
+				goto fail;
+
+			free_token(token);
+
+			if (read_expected(EVENT_OP, ":") < 0)
+				goto fail_expect;
+
+			if (read_expect_type(EVENT_ITEM, &token))
+				goto fail;
+
+			/* add signed type */
+
+			free_token(token);
+			if (read_expected(EVENT_OP, ";") < 0)
+				goto fail_expect;
+
+			if (read_expect_type(EVENT_NEWLINE, &token))
+				goto fail;
+		}
+
 		free_token(token);
 
 		*fields = field;
@@ -921,10 +973,10 @@ static int event_read_format(struct event *event)
 	char *token;
 	int ret;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"format") < 0)
+	if (read_expected_item(EVENT_ITEM, "format") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_NEWLINE, &token))
@@ -984,7 +1036,7 @@ process_cond(struct event *event, struct print_arg *top, char **tok)
 
 	*tok = NULL;
 	type = process_arg(event, left, &token);
-	if (test_type_token(type, token, EVENT_OP, (char *)":"))
+	if (test_type_token(type, token, EVENT_OP, ":"))
 		goto out_free;
 
 	arg->op.op = token;
@@ -1004,6 +1056,35 @@ out_free:
 	return EVENT_ERROR;
 }
 
+static enum event_type
+process_array(struct event *event, struct print_arg *top, char **tok)
+{
+	struct print_arg *arg;
+	enum event_type type;
+	char *token = NULL;
+
+	arg = malloc_or_die(sizeof(*arg));
+	memset(arg, 0, sizeof(*arg));
+
+	*tok = NULL;
+	type = process_arg(event, arg, &token);
+	if (test_type_token(type, token, EVENT_OP, "]"))
+		goto out_free;
+
+	top->op.right = arg;
+
+	free_token(token);
+	type = read_token_item(&token);
+	*tok = token;
+
+	return type;
+
+out_free:
+	free_token(*tok);
+	free_arg(arg);
+	return EVENT_ERROR;
+}
+
 static int get_op_prio(char *op)
 {
 	if (!op[1]) {
@@ -1128,6 +1209,8 @@ process_op(struct event *event, struct print_arg *arg, char **tok)
 		   strcmp(token, "*") == 0 ||
 		   strcmp(token, "^") == 0 ||
 		   strcmp(token, "/") == 0 ||
+		   strcmp(token, "<") == 0 ||
+		   strcmp(token, ">") == 0 ||
 		   strcmp(token, "==") == 0 ||
 		   strcmp(token, "!=") == 0) {
 
@@ -1144,17 +1227,46 @@ process_op(struct event *event, struct print_arg *arg, char **tok)
 
 		right = malloc_or_die(sizeof(*right));
 
-		type = process_arg(event, right, tok);
+		type = read_token_item(&token);
+		*tok = token;
+
+		/* could just be a type pointer */
+		if ((strcmp(arg->op.op, "*") == 0) &&
+		    type == EVENT_DELIM && (strcmp(token, ")") == 0)) {
+			if (left->type != PRINT_ATOM)
+				die("bad pointer type");
+			left->atom.atom = realloc(left->atom.atom,
+					    sizeof(left->atom.atom) + 3);
+			strcat(left->atom.atom, " *");
+			*arg = *left;
+			free(arg);
+
+			return type;
+		}
+
+		type = process_arg_token(event, right, tok, type);
 
 		arg->op.right = right;
 
+	} else if (strcmp(token, "[") == 0) {
+
+		left = malloc_or_die(sizeof(*left));
+		*left = *arg;
+
+		arg->type = PRINT_OP;
+		arg->op.op = token;
+		arg->op.left = left;
+
+		arg->op.prio = 0;
+		type = process_array(event, arg, tok);
+
 	} else {
-		die("unknown op '%s'", token);
+		warning("unknown op '%s'", token);
+		event->flags |= EVENT_FL_FAILED;
 		/* the arg is now the left side */
 		return EVENT_NONE;
 	}
 
-
 	if (type == EVENT_OP) {
 		int prio;
 
@@ -1178,7 +1290,7 @@ process_entry(struct event *event __unused, struct print_arg *arg,
 	char *field;
 	char *token;
 
-	if (read_expected(EVENT_OP, (char *)"->") < 0)
+	if (read_expected(EVENT_OP, "->") < 0)
 		return EVENT_ERROR;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -1338,14 +1450,14 @@ process_fields(struct event *event, struct print_flag_sym **list, char **tok)
 	do {
 		free_token(token);
 		type = read_token_item(&token);
-		if (test_type_token(type, token, EVENT_OP, (char *)"{"))
+		if (test_type_token(type, token, EVENT_OP, "{"))
 			break;
 
 		arg = malloc_or_die(sizeof(*arg));
 
 		free_token(token);
 		type = process_arg(event, arg, &token);
-		if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+		if (test_type_token(type, token, EVENT_DELIM, ","))
 			goto out_free;
 
 		field = malloc_or_die(sizeof(*field));
@@ -1356,7 +1468,7 @@ process_fields(struct event *event, struct print_flag_sym **list, char **tok)
 
 		free_token(token);
 		type = process_arg(event, arg, &token);
-		if (test_type_token(type, token, EVENT_OP, (char *)"}"))
+		if (test_type_token(type, token, EVENT_OP, "}"))
 			goto out_free;
 
 		value = arg_eval(arg);
@@ -1391,13 +1503,13 @@ process_flags(struct event *event, struct print_arg *arg, char **tok)
 	memset(arg, 0, sizeof(*arg));
 	arg->type = PRINT_FLAGS;
 
-	if (read_expected_item(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected_item(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	field = malloc_or_die(sizeof(*field));
 
 	type = process_arg(event, field, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	arg->flags.field = field;
@@ -1408,11 +1520,11 @@ process_flags(struct event *event, struct print_arg *arg, char **tok)
 		type = read_token_item(&token);
 	}
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	type = process_fields(event, &arg->flags.flags, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")"))
+	if (test_type_token(type, token, EVENT_DELIM, ")"))
 		goto out_free;
 
 	free_token(token);
@@ -1434,19 +1546,19 @@ process_symbols(struct event *event, struct print_arg *arg, char **tok)
 	memset(arg, 0, sizeof(*arg));
 	arg->type = PRINT_SYMBOL;
 
-	if (read_expected_item(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected_item(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	field = malloc_or_die(sizeof(*field));
 
 	type = process_arg(event, field, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	arg->symbol.field = field;
 
 	type = process_fields(event, &arg->symbol.symbols, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")"))
+	if (test_type_token(type, token, EVENT_DELIM, ")"))
 		goto out_free;
 
 	free_token(token);
@@ -1463,7 +1575,6 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 {
 	struct print_arg *item_arg;
 	enum event_type type;
-	int ptr_cast = 0;
 	char *token;
 
 	type = process_arg(event, arg, &token);
@@ -1471,28 +1582,13 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 	if (type == EVENT_ERROR)
 		return EVENT_ERROR;
 
-	if (type == EVENT_OP) {
-		/* handle the ptr casts */
-		if (!strcmp(token, "*")) {
-			/*
-			 * FIXME: should we zapp whitespaces before ')' ?
-			 * (may require a peek_token_item())
-			 */
-			if (__peek_char() == ')') {
-				ptr_cast = 1;
-				free_token(token);
-				type = read_token_item(&token);
-			}
-		}
-		if (!ptr_cast) {
-			type = process_op(event, arg, &token);
+	if (type == EVENT_OP)
+		type = process_op(event, arg, &token);
 
-			if (type == EVENT_ERROR)
-				return EVENT_ERROR;
-		}
-	}
+	if (type == EVENT_ERROR)
+		return EVENT_ERROR;
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")")) {
+	if (test_type_token(type, token, EVENT_DELIM, ")")) {
 		free_token(token);
 		return EVENT_ERROR;
 	}
@@ -1516,13 +1612,6 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 		item_arg = malloc_or_die(sizeof(*item_arg));
 
 		arg->type = PRINT_TYPE;
-		if (ptr_cast) {
-			char *old = arg->atom.atom;
-
-			arg->atom.atom = malloc_or_die(strlen(old + 3));
-			sprintf(arg->atom.atom, "%s *", old);
-			free(old);
-		}
 		arg->typecast.type = arg->atom.atom;
 		arg->typecast.item = item_arg;
 		type = process_arg_token(event, item_arg, &token, type);
@@ -1540,7 +1629,7 @@ process_str(struct event *event __unused, struct print_arg *arg, char **tok)
 	enum event_type type;
 	char *token;
 
-	if (read_expected(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -1550,7 +1639,7 @@ process_str(struct event *event __unused, struct print_arg *arg, char **tok)
 	arg->string.string = token;
 	arg->string.offset = -1;
 
-	if (read_expected(EVENT_DELIM, (char *)")") < 0)
+	if (read_expected(EVENT_DELIM, ")") < 0)
 		return EVENT_ERROR;
 
 	type = read_token(&token);
@@ -1637,12 +1726,18 @@ process_arg_token(struct event *event, struct print_arg *arg,
 
 static int event_read_print_args(struct event *event, struct print_arg **list)
 {
-	enum event_type type;
+	enum event_type type = EVENT_ERROR;
 	struct print_arg *arg;
 	char *token;
 	int args = 0;
 
 	do {
+		if (type == EVENT_NEWLINE) {
+			free_token(token);
+			type = read_token_item(&token);
+			continue;
+		}
+
 		arg = malloc_or_die(sizeof(*arg));
 		memset(arg, 0, sizeof(*arg));
 
@@ -1683,18 +1778,19 @@ static int event_read_print(struct event *event)
 	char *token;
 	int ret;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"print") < 0)
+	if (read_expected_item(EVENT_ITEM, "print") < 0)
 		return -1;
 
-	if (read_expected(EVENT_ITEM, (char *)"fmt") < 0)
+	if (read_expected(EVENT_ITEM, "fmt") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_DQUOTE, &token) < 0)
 		goto fail;
 
+ concat:
 	event->print_fmt.format = token;
 	event->print_fmt.args = NULL;
 
@@ -1704,7 +1800,22 @@ static int event_read_print(struct event *event)
 	if (type == EVENT_NONE)
 		return 0;
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	/* Handle concatination of print lines */
+	if (type == EVENT_DQUOTE) {
+		char *cat;
+
+		cat = malloc_or_die(strlen(event->print_fmt.format) +
+				    strlen(token) + 1);
+		strcpy(cat, event->print_fmt.format);
+		strcat(cat, token);
+		free_token(token);
+		free_token(event->print_fmt.format);
+		event->print_fmt.format = NULL;
+		token = cat;
+		goto concat;
+	}
+
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto fail;
 
 	free_token(token);
@@ -1713,7 +1824,7 @@ static int event_read_print(struct event *event)
 	if (ret < 0)
 		return -1;
 
-	return 0;
+	return ret;
 
  fail:
 	free_token(token);
@@ -1822,37 +1933,67 @@ static int get_common_info(const char *type, int *offset, int *size)
 	return 0;
 }
 
-int trace_parse_common_type(void *data)
+static int __parse_common(void *data, int *size, int *offset,
+			  const char *name)
 {
-	static int type_offset;
-	static int type_size;
 	int ret;
 
-	if (!type_size) {
-		ret = get_common_info("common_type",
-				      &type_offset,
-				      &type_size);
+	if (!*size) {
+		ret = get_common_info(name, offset, size);
 		if (ret < 0)
 			return ret;
 	}
-	return read_size(data + type_offset, type_size);
+	return read_size(data + *offset, *size);
+}
+
+int trace_parse_common_type(void *data)
+{
+	static int type_offset;
+	static int type_size;
+
+	return __parse_common(data, &type_size, &type_offset,
+			      "common_type");
 }
 
 static int parse_common_pid(void *data)
 {
 	static int pid_offset;
 	static int pid_size;
+
+	return __parse_common(data, &pid_size, &pid_offset,
+			      "common_pid");
+}
+
+static int parse_common_pc(void *data)
+{
+	static int pc_offset;
+	static int pc_size;
+
+	return __parse_common(data, &pc_size, &pc_offset,
+			      "common_preempt_count");
+}
+
+static int parse_common_flags(void *data)
+{
+	static int flags_offset;
+	static int flags_size;
+
+	return __parse_common(data, &flags_size, &flags_offset,
+			      "common_flags");
+}
+
+static int parse_common_lock_depth(void *data)
+{
+	static int ld_offset;
+	static int ld_size;
 	int ret;
 
-	if (!pid_size) {
-		ret = get_common_info("common_pid",
-				      &pid_offset,
-				      &pid_size);
-		if (ret < 0)
-			return ret;
-	}
+	ret = __parse_common(data, &ld_size, &ld_offset,
+			     "common_lock_depth");
+	if (ret < 0)
+		return -1;
 
-	return read_size(data + pid_offset, pid_size);
+	return ret;
 }
 
 struct event *trace_find_event(int id)
@@ -1871,6 +2012,7 @@ static unsigned long long eval_num_arg(void *data, int size,
 {
 	unsigned long long val = 0;
 	unsigned long long left, right;
+	struct print_arg *larg;
 
 	switch (arg->type) {
 	case PRINT_NULL:
@@ -1897,6 +2039,26 @@ static unsigned long long eval_num_arg(void *data, int size,
 		return 0;
 		break;
 	case PRINT_OP:
+		if (strcmp(arg->op.op, "[") == 0) {
+			/*
+			 * Arrays are special, since we don't want
+			 * to read the arg as is.
+			 */
+			if (arg->op.left->type != PRINT_FIELD)
+				goto default_op; /* oops, all bets off */
+			larg = arg->op.left;
+			if (!larg->field.field) {
+				larg->field.field =
+					find_any_field(event, larg->field.name);
+				if (!larg->field.field)
+					die("field %s not found", larg->field.name);
+			}
+			right = eval_num_arg(data, size, event, arg->op.right);
+			val = read_size(data + larg->field.field->offset +
+					right * long_size, long_size);
+			break;
+		}
+ default_op:
 		left = eval_num_arg(data, size, event, arg->op.left);
 		right = eval_num_arg(data, size, event, arg->op.right);
 		switch (arg->op.op[0]) {
@@ -1947,6 +2109,12 @@ static unsigned long long eval_num_arg(void *data, int size,
 				die("unknown op '%s'", arg->op.op);
 			val = left == right;
 			break;
+		case '-':
+			val = left - right;
+			break;
+		case '+':
+			val = left + right;
+			break;
 		default:
 			die("unknown op '%s'", arg->op.op);
 		}
@@ -2145,8 +2313,9 @@ static struct print_arg *make_bprint_args(char *fmt, void *data, int size, struc
 			case 'u':
 			case 'x':
 			case 'i':
-				bptr = (void *)(((unsigned long)bptr + (long_size - 1)) &
-						~(long_size - 1));
+				/* the pointers are always 4 bytes aligned */
+				bptr = (void *)(((unsigned long)bptr + 3) &
+						~3);
 				switch (ls) {
 				case 0:
 				case 1:
@@ -2270,7 +2439,27 @@ static void pretty_print(void *data, int size, struct event *event)
 
 	for (; *ptr; ptr++) {
 		ls = 0;
-		if (*ptr == '%') {
+		if (*ptr == '\\') {
+			ptr++;
+			switch (*ptr) {
+			case 'n':
+				printf("\n");
+				break;
+			case 't':
+				printf("\t");
+				break;
+			case 'r':
+				printf("\r");
+				break;
+			case '\\':
+				printf("\\");
+				break;
+			default:
+				printf("%c", *ptr);
+				break;
+			}
+
+		} else if (*ptr == '%') {
 			saveptr = ptr;
 			show_func = 0;
  cont_process:
@@ -2377,6 +2566,41 @@ static inline int log10_cpu(int nb)
 	return 1;
 }
 
+static void print_lat_fmt(void *data, int size __unused)
+{
+	unsigned int lat_flags;
+	unsigned int pc;
+	int lock_depth;
+	int hardirq;
+	int softirq;
+
+	lat_flags = parse_common_flags(data);
+	pc = parse_common_pc(data);
+	lock_depth = parse_common_lock_depth(data);
+
+	hardirq = lat_flags & TRACE_FLAG_HARDIRQ;
+	softirq = lat_flags & TRACE_FLAG_SOFTIRQ;
+
+	printf("%c%c%c",
+	       (lat_flags & TRACE_FLAG_IRQS_OFF) ? 'd' :
+	       (lat_flags & TRACE_FLAG_IRQS_NOSUPPORT) ?
+	       'X' : '.',
+	       (lat_flags & TRACE_FLAG_NEED_RESCHED) ?
+	       'N' : '.',
+	       (hardirq && softirq) ? 'H' :
+	       hardirq ? 'h' : softirq ? 's' : '.');
+
+	if (pc)
+		printf("%x", pc);
+	else
+		printf(".");
+
+	if (lock_depth < 0)
+		printf(".");
+	else
+		printf("%d", lock_depth);
+}
+
 /* taken from Linux, written by Frederic Weisbecker */
 static void print_graph_cpu(int cpu)
 {
@@ -2620,6 +2844,11 @@ pretty_print_func_ent(void *data, int size, struct event *event,
 
 	printf(" | ");
 
+	if (latency_format) {
+		print_lat_fmt(data, size);
+		printf(" | ");
+	}
+
 	field = find_field(event, "func");
 	if (!field)
 		die("function entry does not have func field");
@@ -2663,6 +2892,11 @@ pretty_print_func_ret(void *data, int size __unused, struct event *event,
 
 	printf(" | ");
 
+	if (latency_format) {
+		print_lat_fmt(data, size);
+		printf(" | ");
+	}
+
 	field = find_field(event, "rettime");
 	if (!field)
 		die("can't find rettime in return graph");
@@ -2724,7 +2958,7 @@ void print_event(int cpu, void *data, int size, unsigned long long nsecs,
 
 	event = trace_find_event(type);
 	if (!event) {
-		printf("ug! no event found for type %d\n", type);
+		warning("ug! no event found for type %d", type);
 		return;
 	}
 
@@ -2734,9 +2968,20 @@ void print_event(int cpu, void *data, int size, unsigned long long nsecs,
 		return pretty_print_func_graph(data, size, event, cpu,
 					       pid, comm, secs, usecs);
 
-	printf("%16s-%-5d [%03d] %5lu.%09Lu: %s: ",
-	       comm, pid,  cpu,
-	       secs, nsecs, event->name);
+	if (latency_format) {
+		printf("%8.8s-%-5d %3d",
+		       comm, pid, cpu);
+		print_lat_fmt(data, size);
+	} else
+		printf("%16s-%-5d [%03d]", comm, pid,  cpu);
+
+	printf(" %5lu.%06lu: %s: ", secs, usecs, event->name);
+
+	if (event->flags & EVENT_FL_FAILED) {
+		printf("EVENT '%s' FAILED TO PARSE\n",
+		       event->name);
+		return;
+	}
 
 	pretty_print(data, size, event);
 	printf("\n");
@@ -2807,46 +3052,71 @@ static void print_args(struct print_arg *args)
 	}
 }
 
-static void parse_header_field(char *type,
+static void parse_header_field(const char *field,
 			       int *offset, int *size)
 {
 	char *token;
+	int type;
 
-	if (read_expected(EVENT_ITEM, (char *)"field") < 0)
+	if (read_expected(EVENT_ITEM, "field") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
+
 	/* type */
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	free_token(token);
 
-	if (read_expected(EVENT_ITEM, type) < 0)
+	if (read_expected(EVENT_ITEM, field) < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)";") < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
-	if (read_expected(EVENT_ITEM, (char *)"offset") < 0)
+	if (read_expected(EVENT_ITEM, "offset") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	*offset = atoi(token);
 	free_token(token);
-	if (read_expected(EVENT_OP, (char *)";") < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
-	if (read_expected(EVENT_ITEM, (char *)"size") < 0)
+	if (read_expected(EVENT_ITEM, "size") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	*size = atoi(token);
 	free_token(token);
-	if (read_expected(EVENT_OP, (char *)";") < 0)
-		return;
-	if (read_expect_type(EVENT_NEWLINE, &token) < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
+	type = read_token(&token);
+	if (type != EVENT_NEWLINE) {
+		/* newer versions of the kernel have a "signed" type */
+		if (type != EVENT_ITEM)
+			goto fail;
+
+		if (strcmp(token, "signed") != 0)
+			goto fail;
+
+		free_token(token);
+
+		if (read_expected(EVENT_OP, ":") < 0)
+			return;
+
+		if (read_expect_type(EVENT_ITEM, &token))
+			goto fail;
+
+		free_token(token);
+		if (read_expected(EVENT_OP, ";") < 0)
+			return;
+
+		if (read_expect_type(EVENT_NEWLINE, &token))
+			goto fail;
+	}
+ fail:
 	free_token(token);
 }
 
@@ -2854,11 +3124,11 @@ int parse_header_page(char *buf, unsigned long size)
 {
 	init_input_buf(buf, size);
 
-	parse_header_field((char *)"timestamp", &header_page_ts_offset,
+	parse_header_field("timestamp", &header_page_ts_offset,
 			   &header_page_ts_size);
-	parse_header_field((char *)"commit", &header_page_size_offset,
+	parse_header_field("commit", &header_page_size_offset,
 			   &header_page_size_size);
-	parse_header_field((char *)"data", &header_page_data_offset,
+	parse_header_field("data", &header_page_data_offset,
 			   &header_page_data_size);
 
 	return 0;
@@ -2909,6 +3179,9 @@ int parse_ftrace_file(char *buf, unsigned long size)
 	if (ret < 0)
 		die("failed to read ftrace event print fmt");
 
+	/* New ftrace handles args */
+	if (ret > 0)
+		return 0;
 	/*
 	 * The arguments for ftrace files are parsed by the fields.
 	 * Set up the fields as their arguments.
@@ -2926,7 +3199,7 @@ int parse_ftrace_file(char *buf, unsigned long size)
 	return 0;
 }
 
-int parse_event_file(char *buf, unsigned long size, char *system__unused __unused)
+int parse_event_file(char *buf, unsigned long size, char *sys)
 {
 	struct event *event;
 	int ret;
@@ -2946,12 +3219,18 @@ int parse_event_file(char *buf, unsigned long size, char *system__unused __unuse
 		die("failed to read event id");
 
 	ret = event_read_format(event);
-	if (ret < 0)
-		die("failed to read event format");
+	if (ret < 0) {
+		warning("failed to read event format for %s", event->name);
+		goto event_failed;
+	}
 
 	ret = event_read_print(event);
-	if (ret < 0)
-		die("failed to read event print fmt");
+	if (ret < 0) {
+		warning("failed to read event print fmt for %s", event->name);
+		goto event_failed;
+	}
+
+	event->system = strdup(sys);
 
 #define PRINT_ARGS 0
 	if (PRINT_ARGS && event->print_fmt.args)
@@ -2959,6 +3238,12 @@ int parse_event_file(char *buf, unsigned long size, char *system__unused __unuse
 
 	add_event(event);
 	return 0;
+
+ event_failed:
+	event->flags |= EVENT_FL_FAILED;
+	/* still add it even if it failed */
+	add_event(event);
+	return -1;
 }
 
 void parse_set_info(int nr_cpus, int long_sz)
diff --git a/tools/perf/util/trace-event-read.c b/tools/perf/util/trace-event-read.c
index 1b5c847d2c2..44292e06cca 100644
--- a/tools/perf/util/trace-event-read.c
+++ b/tools/perf/util/trace-event-read.c
@@ -458,9 +458,8 @@ struct record *trace_read_data(int cpu)
 	return data;
 }
 
-void trace_report(void)
+void trace_report(int fd)
 {
-	const char *input_file = "trace.info";
 	char buf[BUFSIZ];
 	char test[] = { 23, 8, 68 };
 	char *version;
@@ -468,9 +467,7 @@ void trace_report(void)
 	int show_funcs = 0;
 	int show_printk = 0;
 
-	input_fd = open(input_file, O_RDONLY);
-	if (input_fd < 0)
-		die("opening '%s'\n", input_file);
+	input_fd = fd;
 
 	read_or_die(buf, 3);
 	if (memcmp(buf, test, 3) != 0)
diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h
index 693f815c942..f6637c2fa1f 100644
--- a/tools/perf/util/trace-event.h
+++ b/tools/perf/util/trace-event.h
@@ -1,5 +1,5 @@
-#ifndef _TRACE_EVENTS_H
-#define _TRACE_EVENTS_H
+#ifndef __PERF_TRACE_EVENTS_H
+#define __PERF_TRACE_EVENTS_H
 
 #include "parse-events.h"
 
@@ -26,6 +26,9 @@ enum {
 enum format_flags {
 	FIELD_IS_ARRAY		= 1,
 	FIELD_IS_POINTER	= 2,
+	FIELD_IS_SIGNED		= 4,
+	FIELD_IS_STRING		= 8,
+	FIELD_IS_DYNAMIC	= 16,
 };
 
 struct format_field {
@@ -132,15 +135,18 @@ struct event {
 	int			flags;
 	struct format		format;
 	struct print_fmt	print_fmt;
+	char			*system;
 };
 
 enum {
-	EVENT_FL_ISFTRACE	= 1,
-	EVENT_FL_ISPRINT	= 2,
-	EVENT_FL_ISBPRINT	= 4,
-	EVENT_FL_ISFUNC		= 8,
-	EVENT_FL_ISFUNCENT	= 16,
-	EVENT_FL_ISFUNCRET	= 32,
+	EVENT_FL_ISFTRACE	= 0x01,
+	EVENT_FL_ISPRINT	= 0x02,
+	EVENT_FL_ISBPRINT	= 0x04,
+	EVENT_FL_ISFUNC		= 0x08,
+	EVENT_FL_ISFUNCENT	= 0x10,
+	EVENT_FL_ISFUNCRET	= 0x20,
+
+	EVENT_FL_FAILED		= 0x80000000
 };
 
 struct record {
@@ -154,7 +160,7 @@ struct record *trace_read_data(int cpu);
 
 void parse_set_info(int nr_cpus, int long_sz);
 
-void trace_report(void);
+void trace_report(int fd);
 
 void *malloc_or_die(unsigned int size);
 
@@ -166,7 +172,7 @@ void print_funcs(void);
 void print_printk(void);
 
 int parse_ftrace_file(char *buf, unsigned long size);
-int parse_event_file(char *buf, unsigned long size, char *system);
+int parse_event_file(char *buf, unsigned long size, char *sys);
 void print_event(int cpu, void *data, int size, unsigned long long nsecs,
 		  char *comm);
 
@@ -233,6 +239,8 @@ extern int header_page_size_size;
 extern int header_page_data_offset;
 extern int header_page_data_size;
 
+extern int latency_format;
+
 int parse_header_page(char *buf, unsigned long size);
 int trace_parse_common_type(void *data);
 struct event *trace_find_event(int id);
@@ -240,6 +248,15 @@ unsigned long long
 raw_field_value(struct event *event, const char *name, void *data);
 void *raw_field_ptr(struct event *event, const char *name, void *data);
 
-void read_tracing_data(struct perf_event_attr *pattrs, int nb_events);
+void read_tracing_data(int fd, struct perf_event_attr *pattrs, int nb_events);
+
+/* taken from kernel/trace/trace.h */
+enum trace_flag_type {
+	TRACE_FLAG_IRQS_OFF		= 0x01,
+	TRACE_FLAG_IRQS_NOSUPPORT	= 0x02,
+	TRACE_FLAG_NEED_RESCHED		= 0x04,
+	TRACE_FLAG_HARDIRQ		= 0x08,
+	TRACE_FLAG_SOFTIRQ		= 0x10,
+};
 
-#endif /* _TRACE_EVENTS_H */
+#endif /* __PERF_TRACE_EVENTS_H */
diff --git a/tools/perf/util/types.h b/tools/perf/util/types.h
index 5e75f900594..7d6b8331f89 100644
--- a/tools/perf/util/types.h
+++ b/tools/perf/util/types.h
@@ -1,5 +1,5 @@
-#ifndef _PERF_TYPES_H
-#define _PERF_TYPES_H
+#ifndef __PERF_TYPES_H
+#define __PERF_TYPES_H
 
 /*
  * We define u64 as unsigned long long for every architecture
@@ -14,4 +14,4 @@ typedef signed short	   s16;
 typedef unsigned char	   u8;
 typedef signed char	   s8;
 
-#endif /* _PERF_TYPES_H */
+#endif /* __PERF_TYPES_H */
diff --git a/tools/perf/util/values.h b/tools/perf/util/values.h
index cadf8cf2a59..2fa967e1a88 100644
--- a/tools/perf/util/values.h
+++ b/tools/perf/util/values.h
@@ -1,5 +1,5 @@
-#ifndef _PERF_VALUES_H
-#define _PERF_VALUES_H
+#ifndef __PERF_VALUES_H
+#define __PERF_VALUES_H
 
 #include "types.h"
 
@@ -24,4 +24,4 @@ void perf_read_values_add_value(struct perf_read_values *values,
 void perf_read_values_display(FILE *fp, struct perf_read_values *values,
 			      int raw);
 
-#endif /* _PERF_VALUES_H */
+#endif /* __PERF_VALUES_H */