Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (184 commits)
  perf probe: Clean up probe_point_lazy_walker() return value
  tracing: Fix irqoff selftest expanding max buffer
  tracing: Align 4 byte ints together in struct tracer
  tracing: Export trace_set_clr_event()
  tracing: Explain about unstable clock on resume with ring buffer warning
  ftrace/graph: Trace function entry before updating index
  ftrace: Add .ref.text as one of the safe areas to trace
  tracing: Adjust conditional expression latency formatting.
  tracing: Fix event alignment: skb:kfree_skb
  tracing: Fix event alignment: mce:mce_record
  tracing: Fix event alignment: kvm:kvm_hv_hypercall
  tracing: Fix event alignment: module:module_request
  tracing: Fix event alignment: ftrace:context_switch and ftrace:wakeup
  tracing: Remove lock_depth from event entry
  perf header: Stop using 'self'
  perf session: Use evlist/evsel for managing perf.data attributes
  perf top: Don't let events to eat up whole header line
  perf top: Fix events overflow in top command
  ring-buffer: Remove unused #include <linux/trace_irq.h>
  tracing: Add an 'overwrite' trace_option.
  ...

57 files changed, 5072 insertions, 1821 deletions

diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
new file mode 100644
index 000000000000..0d0830c98cd7
--- /dev/null
+++ b/tools/perf/util/annotate.c
@@ -0,0 +1,605 @@
+/*
+ * Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Parts came from builtin-annotate.c, see those files for further
+ * copyright notes.
+ *
+ * Released under the GPL v2. (and only v2, not any later version)
+ */
+
+#include "util.h"
+#include "build-id.h"
+#include "color.h"
+#include "cache.h"
+#include "symbol.h"
+#include "debug.h"
+#include "annotate.h"
+#include <pthread.h>
+
+int symbol__annotate_init(struct map *map __used, struct symbol *sym)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	pthread_mutex_init(&notes->lock, NULL);
+	return 0;
+}
+
+int symbol__alloc_hist(struct symbol *sym, int nevents)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	size_t sizeof_sym_hist = (sizeof(struct sym_hist) +
+				  (sym->end - sym->start) * sizeof(u64));
+
+	notes->src = zalloc(sizeof(*notes->src) + nevents * sizeof_sym_hist);
+	if (notes->src == NULL)
+		return -1;
+	notes->src->sizeof_sym_hist = sizeof_sym_hist;
+	notes->src->nr_histograms   = nevents;
+	INIT_LIST_HEAD(&notes->src->source);
+	return 0;
+}
+
+void symbol__annotate_zero_histograms(struct symbol *sym)
+{
+	struct annotation *notes = symbol__annotation(sym);
+
+	pthread_mutex_lock(&notes->lock);
+	if (notes->src != NULL)
+		memset(notes->src->histograms, 0,
+		       notes->src->nr_histograms * notes->src->sizeof_sym_hist);
+	pthread_mutex_unlock(&notes->lock);
+}
+
+int symbol__inc_addr_samples(struct symbol *sym, struct map *map,
+			     int evidx, u64 addr)
+{
+	unsigned offset;
+	struct annotation *notes;
+	struct sym_hist *h;
+
+	notes = symbol__annotation(sym);
+	if (notes->src == NULL)
+		return -ENOMEM;
+
+	pr_debug3("%s: addr=%#" PRIx64 "\n", __func__, map->unmap_ip(map, addr));
+
+	if (addr >= sym->end)
+		return 0;
+
+	offset = addr - sym->start;
+	h = annotation__histogram(notes, evidx);
+	h->sum++;
+	h->addr[offset]++;
+
+	pr_debug3("%#" PRIx64 " %s: period++ [addr: %#" PRIx64 ", %#" PRIx64
+		  ", evidx=%d] => %" PRIu64 "\n", sym->start, sym->name,
+		  addr, addr - sym->start, evidx, h->addr[offset]);
+	return 0;
+}
+
+static struct objdump_line *objdump_line__new(s64 offset, char *line, size_t privsize)
+{
+	struct objdump_line *self = malloc(sizeof(*self) + privsize);
+
+	if (self != NULL) {
+		self->offset = offset;
+		self->line = line;
+	}
+
+	return self;
+}
+
+void objdump_line__free(struct objdump_line *self)
+{
+	free(self->line);
+	free(self);
+}
+
+static void objdump__add_line(struct list_head *head, struct objdump_line *line)
+{
+	list_add_tail(&line->node, head);
+}
+
+struct objdump_line *objdump__get_next_ip_line(struct list_head *head,
+					       struct objdump_line *pos)
+{
+	list_for_each_entry_continue(pos, head, node)
+		if (pos->offset >= 0)
+			return pos;
+
+	return NULL;
+}
+
+static int objdump_line__print(struct objdump_line *oline, struct symbol *sym,
+			       int evidx, u64 len, int min_pcnt,
+			       int printed, int max_lines,
+			       struct objdump_line *queue)
+{
+	static const char *prev_line;
+	static const char *prev_color;
+
+	if (oline->offset != -1) {
+		const char *path = NULL;
+		unsigned int hits = 0;
+		double percent = 0.0;
+		const char *color;
+		struct annotation *notes = symbol__annotation(sym);
+		struct source_line *src_line = notes->src->lines;
+		struct sym_hist *h = annotation__histogram(notes, evidx);
+		s64 offset = oline->offset;
+		struct objdump_line *next;
+
+		next = objdump__get_next_ip_line(&notes->src->source, oline);
+
+		while (offset < (s64)len &&
+		       (next == NULL || offset < next->offset)) {
+			if (src_line) {
+				if (path == NULL)
+					path = src_line[offset].path;
+				percent += src_line[offset].percent;
+			} else
+				hits += h->addr[offset];
+
+			++offset;
+		}
+
+		if (src_line == NULL && h->sum)
+			percent = 100.0 * hits / h->sum;
+
+		if (percent < min_pcnt)
+			return -1;
+
+		if (max_lines && printed >= max_lines)
+			return 1;
+
+		if (queue != NULL) {
+			list_for_each_entry_from(queue, &notes->src->source, node) {
+				if (queue == oline)
+					break;
+				objdump_line__print(queue, sym, evidx, len,
+						    0, 0, 1, NULL);
+			}
+		}
+
+		color = get_percent_color(percent);
+
+		/*
+		 * Also color the filename and line if needed, with
+		 * the same color than the percentage. Don't print it
+		 * twice for close colored addr with the same filename:line
+		 */
+		if (path) {
+			if (!prev_line || strcmp(prev_line, path)
+				       || color != prev_color) {
+				color_fprintf(stdout, color, " %s", path);
+				prev_line = path;
+				prev_color = color;
+			}
+		}
+
+		color_fprintf(stdout, color, " %7.2f", percent);
+		printf(" :	");
+		color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", oline->line);
+	} else if (max_lines && printed >= max_lines)
+		return 1;
+	else {
+		if (queue)
+			return -1;
+
+		if (!*oline->line)
+			printf("         :\n");
+		else
+			printf("         :	%s\n", oline->line);
+	}
+
+	return 0;
+}
+
+static int symbol__parse_objdump_line(struct symbol *sym, struct map *map,
+				      FILE *file, size_t privsize)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	struct objdump_line *objdump_line;
+	char *line = NULL, *tmp, *tmp2, *c;
+	size_t line_len;
+	s64 line_ip, offset = -1;
+
+	if (getline(&line, &line_len, file) < 0)
+		return -1;
+
+	if (!line)
+		return -1;
+
+	while (line_len != 0 && isspace(line[line_len - 1]))
+		line[--line_len] = '\0';
+
+	c = strchr(line, '\n');
+	if (c)
+		*c = 0;
+
+	line_ip = -1;
+
+	/*
+	 * Strip leading spaces:
+	 */
+	tmp = line;
+	while (*tmp) {
+		if (*tmp != ' ')
+			break;
+		tmp++;
+	}
+
+	if (*tmp) {
+		/*
+		 * Parse hexa addresses followed by ':'
+		 */
+		line_ip = strtoull(tmp, &tmp2, 16);
+		if (*tmp2 != ':' || tmp == tmp2 || tmp2[1] == '\0')
+			line_ip = -1;
+	}
+
+	if (line_ip != -1) {
+		u64 start = map__rip_2objdump(map, sym->start),
+		    end = map__rip_2objdump(map, sym->end);
+
+		offset = line_ip - start;
+		if (offset < 0 || (u64)line_ip > end)
+			offset = -1;
+	}
+
+	objdump_line = objdump_line__new(offset, line, privsize);
+	if (objdump_line == NULL) {
+		free(line);
+		return -1;
+	}
+	objdump__add_line(&notes->src->source, objdump_line);
+
+	return 0;
+}
+
+int symbol__annotate(struct symbol *sym, struct map *map, size_t privsize)
+{
+	struct dso *dso = map->dso;
+	char *filename = dso__build_id_filename(dso, NULL, 0);
+	bool free_filename = true;
+	char command[PATH_MAX * 2];
+	FILE *file;
+	int err = 0;
+	char symfs_filename[PATH_MAX];
+
+	if (filename) {
+		snprintf(symfs_filename, sizeof(symfs_filename), "%s%s",
+			 symbol_conf.symfs, filename);
+	}
+
+	if (filename == NULL) {
+		if (dso->has_build_id) {
+			pr_err("Can't annotate %s: not enough memory\n",
+			       sym->name);
+			return -ENOMEM;
+		}
+		goto fallback;
+	} else if (readlink(symfs_filename, command, sizeof(command)) < 0 ||
+		   strstr(command, "[kernel.kallsyms]") ||
+		   access(symfs_filename, R_OK)) {
+		free(filename);
+fallback:
+		/*
+		 * If we don't have build-ids or the build-id file isn't in the
+		 * cache, or is just a kallsyms file, well, lets hope that this
+		 * DSO is the same as when 'perf record' ran.
+		 */
+		filename = dso->long_name;
+		snprintf(symfs_filename, sizeof(symfs_filename), "%s%s",
+			 symbol_conf.symfs, filename);
+		free_filename = false;
+	}
+
+	if (dso->origin == DSO__ORIG_KERNEL) {
+		char bf[BUILD_ID_SIZE * 2 + 16] = " with build id ";
+		char *build_id_msg = NULL;
+
+		if (dso->annotate_warned)
+			goto out_free_filename;
+
+		if (dso->has_build_id) {
+			build_id__sprintf(dso->build_id,
+					  sizeof(dso->build_id), bf + 15);
+			build_id_msg = bf;
+		}
+		err = -ENOENT;
+		dso->annotate_warned = 1;
+		pr_err("Can't annotate %s: No vmlinux file%s was found in the "
+		       "path.\nPlease use 'perf buildid-cache -av vmlinux' or "
+		       "--vmlinux vmlinux.\n",
+		       sym->name, build_id_msg ?: "");
+		goto out_free_filename;
+	}
+
+	pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__,
+		 filename, sym->name, map->unmap_ip(map, sym->start),
+		 map->unmap_ip(map, sym->end));
+
+	pr_debug("annotating [%p] %30s : [%p] %30s\n",
+		 dso, dso->long_name, sym, sym->name);
+
+	snprintf(command, sizeof(command),
+		 "objdump --start-address=0x%016" PRIx64
+		 " --stop-address=0x%016" PRIx64 " -dS -C %s|grep -v %s|expand",
+		 map__rip_2objdump(map, sym->start),
+		 map__rip_2objdump(map, sym->end),
+		 symfs_filename, filename);
+
+	pr_debug("Executing: %s\n", command);
+
+	file = popen(command, "r");
+	if (!file)
+		goto out_free_filename;
+
+	while (!feof(file))
+		if (symbol__parse_objdump_line(sym, map, file, privsize) < 0)
+			break;
+
+	pclose(file);
+out_free_filename:
+	if (free_filename)
+		free(filename);
+	return err;
+}
+
+static void insert_source_line(struct rb_root *root, struct source_line *src_line)
+{
+	struct source_line *iter;
+	struct rb_node **p = &root->rb_node;
+	struct rb_node *parent = NULL;
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct source_line, node);
+
+		if (src_line->percent > iter->percent)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&src_line->node, parent, p);
+	rb_insert_color(&src_line->node, root);
+}
+
+static void symbol__free_source_line(struct symbol *sym, int len)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	struct source_line *src_line = notes->src->lines;
+	int i;
+
+	for (i = 0; i < len; i++)
+		free(src_line[i].path);
+
+	free(src_line);
+	notes->src->lines = NULL;
+}
+
+/* Get the filename:line for the colored entries */
+static int symbol__get_source_line(struct symbol *sym, struct map *map,
+				   int evidx, struct rb_root *root, int len,
+				   const char *filename)
+{
+	u64 start;
+	int i;
+	char cmd[PATH_MAX * 2];
+	struct source_line *src_line;
+	struct annotation *notes = symbol__annotation(sym);
+	struct sym_hist *h = annotation__histogram(notes, evidx);
+
+	if (!h->sum)
+		return 0;
+
+	src_line = notes->src->lines = calloc(len, sizeof(struct source_line));
+	if (!notes->src->lines)
+		return -1;
+
+	start = map->unmap_ip(map, sym->start);
+
+	for (i = 0; i < len; i++) {
+		char *path = NULL;
+		size_t line_len;
+		u64 offset;
+		FILE *fp;
+
+		src_line[i].percent = 100.0 * h->addr[i] / h->sum;
+		if (src_line[i].percent <= 0.5)
+			continue;
+
+		offset = start + i;
+		sprintf(cmd, "addr2line -e %s %016" PRIx64, filename, offset);
+		fp = popen(cmd, "r");
+		if (!fp)
+			continue;
+
+		if (getline(&path, &line_len, fp) < 0 || !line_len)
+			goto next;
+
+		src_line[i].path = malloc(sizeof(char) * line_len + 1);
+		if (!src_line[i].path)
+			goto next;
+
+		strcpy(src_line[i].path, path);
+		insert_source_line(root, &src_line[i]);
+
+	next:
+		pclose(fp);
+	}
+
+	return 0;
+}
+
+static void print_summary(struct rb_root *root, const char *filename)
+{
+	struct source_line *src_line;
+	struct rb_node *node;
+
+	printf("\nSorted summary for file %s\n", filename);
+	printf("----------------------------------------------\n\n");
+
+	if (RB_EMPTY_ROOT(root)) {
+		printf(" Nothing higher than %1.1f%%\n", MIN_GREEN);
+		return;
+	}
+
+	node = rb_first(root);
+	while (node) {
+		double percent;
+		const char *color;
+		char *path;
+
+		src_line = rb_entry(node, struct source_line, node);
+		percent = src_line->percent;
+		color = get_percent_color(percent);
+		path = src_line->path;
+
+		color_fprintf(stdout, color, " %7.2f %s", percent, path);
+		node = rb_next(node);
+	}
+}
+
+static void symbol__annotate_hits(struct symbol *sym, int evidx)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	struct sym_hist *h = annotation__histogram(notes, evidx);
+	u64 len = sym->end - sym->start, offset;
+
+	for (offset = 0; offset < len; ++offset)
+		if (h->addr[offset] != 0)
+			printf("%*" PRIx64 ": %" PRIu64 "\n", BITS_PER_LONG / 2,
+			       sym->start + offset, h->addr[offset]);
+	printf("%*s: %" PRIu64 "\n", BITS_PER_LONG / 2, "h->sum", h->sum);
+}
+
+int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx,
+			    bool full_paths, int min_pcnt, int max_lines,
+			    int context)
+{
+	struct dso *dso = map->dso;
+	const char *filename = dso->long_name, *d_filename;
+	struct annotation *notes = symbol__annotation(sym);
+	struct objdump_line *pos, *queue = NULL;
+	int printed = 2, queue_len = 0;
+	int more = 0;
+	u64 len;
+
+	if (full_paths)
+		d_filename = filename;
+	else
+		d_filename = basename(filename);
+
+	len = sym->end - sym->start;
+
+	printf(" Percent |	Source code & Disassembly of %s\n", d_filename);
+	printf("------------------------------------------------\n");
+
+	if (verbose)
+		symbol__annotate_hits(sym, evidx);
+
+	list_for_each_entry(pos, &notes->src->source, node) {
+		if (context && queue == NULL) {
+			queue = pos;
+			queue_len = 0;
+		}
+
+		switch (objdump_line__print(pos, sym, evidx, len, min_pcnt,
+					    printed, max_lines, queue)) {
+		case 0:
+			++printed;
+			if (context) {
+				printed += queue_len;
+				queue = NULL;
+				queue_len = 0;
+			}
+			break;
+		case 1:
+			/* filtered by max_lines */
+			++more;
+			break;
+		case -1:
+		default:
+			/*
+			 * Filtered by min_pcnt or non IP lines when
+			 * context != 0
+			 */
+			if (!context)
+				break;
+			if (queue_len == context)
+				queue = list_entry(queue->node.next, typeof(*queue), node);
+			else
+				++queue_len;
+			break;
+		}
+	}
+
+	return more;
+}
+
+void symbol__annotate_zero_histogram(struct symbol *sym, int evidx)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	struct sym_hist *h = annotation__histogram(notes, evidx);
+
+	memset(h, 0, notes->src->sizeof_sym_hist);
+}
+
+void symbol__annotate_decay_histogram(struct symbol *sym, int evidx)
+{
+	struct annotation *notes = symbol__annotation(sym);
+	struct sym_hist *h = annotation__histogram(notes, evidx);
+	struct objdump_line *pos;
+	int len = sym->end - sym->start;
+
+	h->sum = 0;
+
+	list_for_each_entry(pos, &notes->src->source, node) {
+		if (pos->offset != -1 && pos->offset < len) {
+			h->addr[pos->offset] = h->addr[pos->offset] * 7 / 8;
+			h->sum += h->addr[pos->offset];
+		}
+	}
+}
+
+void objdump_line_list__purge(struct list_head *head)
+{
+	struct objdump_line *pos, *n;
+
+	list_for_each_entry_safe(pos, n, head, node) {
+		list_del(&pos->node);
+		objdump_line__free(pos);
+	}
+}
+
+int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx,
+			 bool print_lines, bool full_paths, int min_pcnt,
+			 int max_lines)
+{
+	struct dso *dso = map->dso;
+	const char *filename = dso->long_name;
+	struct rb_root source_line = RB_ROOT;
+	u64 len;
+
+	if (symbol__annotate(sym, map, 0) < 0)
+		return -1;
+
+	len = sym->end - sym->start;
+
+	if (print_lines) {
+		symbol__get_source_line(sym, map, evidx, &source_line,
+					len, filename);
+		print_summary(&source_line, filename);
+	}
+
+	symbol__annotate_printf(sym, map, evidx, full_paths,
+				min_pcnt, max_lines, 0);
+	if (print_lines)
+		symbol__free_source_line(sym, len);
+
+	objdump_line_list__purge(&symbol__annotation(sym)->src->source);
+
+	return 0;
+}
diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h
new file mode 100644
index 000000000000..c2c286896801
--- /dev/null
+++ b/tools/perf/util/annotate.h
@@ -0,0 +1,103 @@
+#ifndef __PERF_ANNOTATE_H
+#define __PERF_ANNOTATE_H
+
+#include <stdbool.h>
+#include "types.h"
+#include "symbol.h"
+#include <linux/list.h>
+#include <linux/rbtree.h>
+
+struct objdump_line {
+	struct list_head node;
+	s64		 offset;
+	char		 *line;
+};
+
+void objdump_line__free(struct objdump_line *self);
+struct objdump_line *objdump__get_next_ip_line(struct list_head *head,
+					       struct objdump_line *pos);
+
+struct sym_hist {
+	u64		sum;
+	u64		addr[0];
+};
+
+struct source_line {
+	struct rb_node	node;
+	double		percent;
+	char		*path;
+};
+
+/** struct annotated_source - symbols with hits have this attached as in sannotation
+ *
+ * @histogram: Array of addr hit histograms per event being monitored
+ * @lines: If 'print_lines' is specified, per source code line percentages
+ * @source: source parsed from objdump -dS
+ *
+ * lines is allocated, percentages calculated and all sorted by percentage
+ * when the annotation is about to be presented, so the percentages are for
+ * one of the entries in the histogram array, i.e. for the event/counter being
+ * presented. It is deallocated right after symbol__{tui,tty,etc}_annotate
+ * returns.
+ */
+struct annotated_source {
+	struct list_head   source;
+	struct source_line *lines;
+	int    		   nr_histograms;
+	int    		   sizeof_sym_hist;
+	struct sym_hist	   histograms[0];
+};
+
+struct annotation {
+	pthread_mutex_t		lock;
+	struct annotated_source *src;
+};
+
+struct sannotation {
+	struct annotation annotation;
+	struct symbol	  symbol;
+};
+
+static inline struct sym_hist *annotation__histogram(struct annotation *notes, int idx)
+{
+	return (((void *)&notes->src->histograms) +
+	 	(notes->src->sizeof_sym_hist * idx));
+}
+
+static inline struct annotation *symbol__annotation(struct symbol *sym)
+{
+	struct sannotation *a = container_of(sym, struct sannotation, symbol);
+	return &a->annotation;
+}
+
+int symbol__inc_addr_samples(struct symbol *sym, struct map *map,
+			     int evidx, u64 addr);
+int symbol__alloc_hist(struct symbol *sym, int nevents);
+void symbol__annotate_zero_histograms(struct symbol *sym);
+
+int symbol__annotate(struct symbol *sym, struct map *map, size_t privsize);
+int symbol__annotate_init(struct map *map __used, struct symbol *sym);
+int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx,
+			    bool full_paths, int min_pcnt, int max_lines,
+			    int context);
+void symbol__annotate_zero_histogram(struct symbol *sym, int evidx);
+void symbol__annotate_decay_histogram(struct symbol *sym, int evidx);
+void objdump_line_list__purge(struct list_head *head);
+
+int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx,
+			 bool print_lines, bool full_paths, int min_pcnt,
+			 int max_lines);
+
+#ifdef NO_NEWT_SUPPORT
+static inline int symbol__tui_annotate(struct symbol *sym __used,
+				       struct map *map __used,
+				       int evidx __used, int refresh __used)
+{
+	return 0;
+}
+#else
+int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx,
+			 int refresh);
+#endif
+
+#endif	/* __PERF_ANNOTATE_H */
diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c
index deffb8c96071..31f934af9861 100644
--- a/tools/perf/util/build-id.c
+++ b/tools/perf/util/build-id.c
@@ -14,8 +14,8 @@
 #include <linux/kernel.h>
 #include "debug.h"
 
-static int build_id__mark_dso_hit(event_t *event,
-				  struct sample_data *sample __used,
+static int build_id__mark_dso_hit(union perf_event *event,
+				  struct perf_sample *sample __used,
 				  struct perf_session *session)
 {
 	struct addr_location al;
@@ -37,13 +37,14 @@ static int build_id__mark_dso_hit(event_t *event,
 	return 0;
 }
 
-static int event__exit_del_thread(event_t *self, struct sample_data *sample __used,
-				  struct perf_session *session)
+static int perf_event__exit_del_thread(union perf_event *event,
+				       struct perf_sample *sample __used,
+				       struct perf_session *session)
 {
-	struct thread *thread = perf_session__findnew(session, self->fork.tid);
+	struct thread *thread = perf_session__findnew(session, event->fork.tid);
 
-	dump_printf("(%d:%d):(%d:%d)\n", self->fork.pid, self->fork.tid,
-		    self->fork.ppid, self->fork.ptid);
+	dump_printf("(%d:%d):(%d:%d)\n", event->fork.pid, event->fork.tid,
+		    event->fork.ppid, event->fork.ptid);
 
 	if (thread) {
 		rb_erase(&thread->rb_node, &session->threads);
@@ -56,9 +57,9 @@ static int event__exit_del_thread(event_t *self, struct sample_data *sample __us
 
 struct perf_event_ops build_id__mark_dso_hit_ops = {
 	.sample	= build_id__mark_dso_hit,
-	.mmap	= event__process_mmap,
-	.fork	= event__process_task,
-	.exit	= event__exit_del_thread,
+	.mmap	= perf_event__process_mmap,
+	.fork	= perf_event__process_task,
+	.exit	= perf_event__exit_del_thread,
 };
 
 char *dso__build_id_filename(struct dso *self, char *bf, size_t size)
diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h
index a7729797fd96..fc5e5a09d5b9 100644
--- a/tools/perf/util/cache.h
+++ b/tools/perf/util/cache.h
@@ -34,13 +34,14 @@ extern int pager_use_color;
 extern int use_browser;
 
 #ifdef NO_NEWT_SUPPORT
-static inline void setup_browser(void)
+static inline void setup_browser(bool fallback_to_pager)
 {
-	setup_pager();
+	if (fallback_to_pager)
+		setup_pager();
 }
 static inline void exit_browser(bool wait_for_ok __used) {}
 #else
-void setup_browser(void);
+void setup_browser(bool fallback_to_pager);
 void exit_browser(bool wait_for_ok);
 #endif
 
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index e12d539417b2..9f7106a8d9a4 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2009-2010, Frederic Weisbecker <fweisbec@gmail.com>
+ * Copyright (C) 2009-2011, Frederic Weisbecker <fweisbec@gmail.com>
  *
  * Handle the callchains from the stream in an ad-hoc radix tree and then
  * sort them in an rbtree.
@@ -18,7 +18,8 @@
 #include "util.h"
 #include "callchain.h"
 
-bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event)
+bool ip_callchain__valid(struct ip_callchain *chain,
+			 const union perf_event *event)
 {
 	unsigned int chain_size = event->header.size;
 	chain_size -= (unsigned long)&event->ip.__more_data - (unsigned long)event;
@@ -26,10 +27,10 @@ bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event)
 }
 
 #define chain_for_each_child(child, parent)	\
-	list_for_each_entry(child, &parent->children, brothers)
+	list_for_each_entry(child, &parent->children, siblings)
 
 #define chain_for_each_child_safe(child, next, parent)	\
-	list_for_each_entry_safe(child, next, &parent->children, brothers)
+	list_for_each_entry_safe(child, next, &parent->children, siblings)
 
 static void
 rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
@@ -38,14 +39,14 @@ rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
 	struct callchain_node *rnode;
-	u64 chain_cumul = cumul_hits(chain);
+	u64 chain_cumul = callchain_cumul_hits(chain);
 
 	while (*p) {
 		u64 rnode_cumul;
 
 		parent = *p;
 		rnode = rb_entry(parent, struct callchain_node, rb_node);
-		rnode_cumul = cumul_hits(rnode);
+		rnode_cumul = callchain_cumul_hits(rnode);
 
 		switch (mode) {
 		case CHAIN_FLAT:
@@ -104,7 +105,7 @@ static void __sort_chain_graph_abs(struct callchain_node *node,
 
 	chain_for_each_child(child, node) {
 		__sort_chain_graph_abs(child, min_hit);
-		if (cumul_hits(child) >= min_hit)
+		if (callchain_cumul_hits(child) >= min_hit)
 			rb_insert_callchain(&node->rb_root, child,
 					    CHAIN_GRAPH_ABS);
 	}
@@ -129,7 +130,7 @@ static void __sort_chain_graph_rel(struct callchain_node *node,
 
 	chain_for_each_child(child, node) {
 		__sort_chain_graph_rel(child, min_percent);
-		if (cumul_hits(child) >= min_hit)
+		if (callchain_cumul_hits(child) >= min_hit)
 			rb_insert_callchain(&node->rb_root, child,
 					    CHAIN_GRAPH_REL);
 	}
@@ -143,7 +144,7 @@ sort_chain_graph_rel(struct rb_root *rb_root, struct callchain_root *chain_root,
 	rb_root->rb_node = chain_root->node.rb_root.rb_node;
 }
 
-int register_callchain_param(struct callchain_param *param)
+int callchain_register_param(struct callchain_param *param)
 {
 	switch (param->mode) {
 	case CHAIN_GRAPH_ABS:
@@ -189,32 +190,27 @@ create_child(struct callchain_node *parent, bool inherit_children)
 		chain_for_each_child(next, new)
 			next->parent = new;
 	}
-	list_add_tail(&new->brothers, &parent->children);
+	list_add_tail(&new->siblings, &parent->children);
 
 	return new;
 }
 
 
-struct resolved_ip {
-	u64		  ip;
-	struct map_symbol ms;
-};
-
-struct resolved_chain {
-	u64			nr;
-	struct resolved_ip	ips[0];
-};
-
-
 /*
  * Fill the node with callchain values
  */
 static void
-fill_node(struct callchain_node *node, struct resolved_chain *chain, int start)
+fill_node(struct callchain_node *node, struct callchain_cursor *cursor)
 {
-	unsigned int i;
+	struct callchain_cursor_node *cursor_node;
+
+	node->val_nr = cursor->nr - cursor->pos;
+	if (!node->val_nr)
+		pr_warning("Warning: empty node in callchain tree\n");
 
-	for (i = start; i < chain->nr; i++) {
+	cursor_node = callchain_cursor_current(cursor);
+
+	while (cursor_node) {
 		struct callchain_list *call;
 
 		call = zalloc(sizeof(*call));
@@ -222,23 +218,25 @@ fill_node(struct callchain_node *node, struct resolved_chain *chain, int start)
 			perror("not enough memory for the code path tree");
 			return;
 		}
-		call->ip = chain->ips[i].ip;
-		call->ms = chain->ips[i].ms;
+		call->ip = cursor_node->ip;
+		call->ms.sym = cursor_node->sym;
+		call->ms.map = cursor_node->map;
 		list_add_tail(&call->list, &node->val);
+
+		callchain_cursor_advance(cursor);
+		cursor_node = callchain_cursor_current(cursor);
 	}
-	node->val_nr = chain->nr - start;
-	if (!node->val_nr)
-		pr_warning("Warning: empty node in callchain tree\n");
 }
 
 static void
-add_child(struct callchain_node *parent, struct resolved_chain *chain,
-	  int start, u64 period)
+add_child(struct callchain_node *parent,
+	  struct callchain_cursor *cursor,
+	  u64 period)
 {
 	struct callchain_node *new;
 
 	new = create_child(parent, false);
-	fill_node(new, chain, start);
+	fill_node(new, cursor);
 
 	new->children_hit = 0;
 	new->hit = period;
@@ -250,9 +248,10 @@ add_child(struct callchain_node *parent, struct resolved_chain *chain,
  * Then create another child to host the given callchain of new branch
  */
 static void
-split_add_child(struct callchain_node *parent, struct resolved_chain *chain,
-		struct callchain_list *to_split, int idx_parents, int idx_local,
-		u64 period)
+split_add_child(struct callchain_node *parent,
+		struct callchain_cursor *cursor,
+		struct callchain_list *to_split,
+		u64 idx_parents, u64 idx_local, u64 period)
 {
 	struct callchain_node *new;
 	struct list_head *old_tail;
@@ -272,14 +271,14 @@ split_add_child(struct callchain_node *parent, struct resolved_chain *chain,
 	/* split the hits */
 	new->hit = parent->hit;
 	new->children_hit = parent->children_hit;
-	parent->children_hit = cumul_hits(new);
+	parent->children_hit = callchain_cumul_hits(new);
 	new->val_nr = parent->val_nr - idx_local;
 	parent->val_nr = idx_local;
 
 	/* create a new child for the new branch if any */
-	if (idx_total < chain->nr) {
+	if (idx_total < cursor->nr) {
 		parent->hit = 0;
-		add_child(parent, chain, idx_total, period);
+		add_child(parent, cursor, period);
 		parent->children_hit += period;
 	} else {
 		parent->hit = period;
@@ -287,36 +286,41 @@ split_add_child(struct callchain_node *parent, struct resolved_chain *chain,
 }
 
 static int
-append_chain(struct callchain_node *root, struct resolved_chain *chain,
-	     unsigned int start, u64 period);
+append_chain(struct callchain_node *root,
+	     struct callchain_cursor *cursor,
+	     u64 period);
 
 static void
-append_chain_children(struct callchain_node *root, struct resolved_chain *chain,
-		      unsigned int start, u64 period)
+append_chain_children(struct callchain_node *root,
+		      struct callchain_cursor *cursor,
+		      u64 period)
 {
 	struct callchain_node *rnode;
 
 	/* lookup in childrens */
 	chain_for_each_child(rnode, root) {
-		unsigned int ret = append_chain(rnode, chain, start, period);
+		unsigned int ret = append_chain(rnode, cursor, period);
 
 		if (!ret)
 			goto inc_children_hit;
 	}
 	/* nothing in children, add to the current node */
-	add_child(root, chain, start, period);
+	add_child(root, cursor, period);
 
 inc_children_hit:
 	root->children_hit += period;
 }
 
 static int
-append_chain(struct callchain_node *root, struct resolved_chain *chain,
-	     unsigned int start, u64 period)
+append_chain(struct callchain_node *root,
+	     struct callchain_cursor *cursor,
+	     u64 period)
 {
+	struct callchain_cursor_node *curr_snap = cursor->curr;
 	struct callchain_list *cnode;
-	unsigned int i = start;
+	u64 start = cursor->pos;
 	bool found = false;
+	u64 matches;
 
 	/*
 	 * Lookup in the current node
@@ -324,141 +328,134 @@ append_chain(struct callchain_node *root, struct resolved_chain *chain,
 	 * anywhere inside a function.
 	 */
 	list_for_each_entry(cnode, &root->val, list) {
+		struct callchain_cursor_node *node;
 		struct symbol *sym;
 
-		if (i == chain->nr)
+		node = callchain_cursor_current(cursor);
+		if (!node)
 			break;
 
-		sym = chain->ips[i].ms.sym;
+		sym = node->sym;
 
 		if (cnode->ms.sym && sym) {
 			if (cnode->ms.sym->start != sym->start)
 				break;
-		} else if (cnode->ip != chain->ips[i].ip)
+		} else if (cnode->ip != node->ip)
 			break;
 
 		if (!found)
 			found = true;
-		i++;
+
+		callchain_cursor_advance(cursor);
 	}
 
 	/* matches not, relay on the parent */
-	if (!found)
+	if (!found) {
+		cursor->curr = curr_snap;
+		cursor->pos = start;
 		return -1;
+	}
+
+	matches = cursor->pos - start;
 
 	/* we match only a part of the node. Split it and add the new chain */
-	if (i - start < root->val_nr) {
-		split_add_child(root, chain, cnode, start, i - start, period);
+	if (matches < root->val_nr) {
+		split_add_child(root, cursor, cnode, start, matches, period);
 		return 0;
 	}
 
 	/* we match 100% of the path, increment the hit */
-	if (i - start == root->val_nr && i == chain->nr) {
+	if (matches == root->val_nr && cursor->pos == cursor->nr) {
 		root->hit += period;
 		return 0;
 	}
 
 	/* We match the node and still have a part remaining */
-	append_chain_children(root, chain, i, period);
+	append_chain_children(root, cursor, period);
 
 	return 0;
 }
 
-static void filter_context(struct ip_callchain *old, struct resolved_chain *new,
-			   struct map_symbol *syms)
-{
-	int i, j = 0;
-
-	for (i = 0; i < (int)old->nr; i++) {
-		if (old->ips[i] >= PERF_CONTEXT_MAX)
-			continue;
-
-		new->ips[j].ip = old->ips[i];
-		new->ips[j].ms = syms[i];
-		j++;
-	}
-
-	new->nr = j;
-}
-
-
-int callchain_append(struct callchain_root *root, struct ip_callchain *chain,
-		     struct map_symbol *syms, u64 period)
+int callchain_append(struct callchain_root *root,
+		     struct callchain_cursor *cursor,
+		     u64 period)
 {
-	struct resolved_chain *filtered;
-
-	if (!chain->nr)
+	if (!cursor->nr)
 		return 0;
 
-	filtered = zalloc(sizeof(*filtered) +
-			  chain->nr * sizeof(struct resolved_ip));
-	if (!filtered)
-		return -ENOMEM;
-
-	filter_context(chain, filtered, syms);
-
-	if (!filtered->nr)
-		goto end;
+	callchain_cursor_commit(cursor);
 
-	append_chain_children(&root->node, filtered, 0, period);
+	append_chain_children(&root->node, cursor, period);
 
-	if (filtered->nr > root->max_depth)
-		root->max_depth = filtered->nr;
-end:
-	free(filtered);
+	if (cursor->nr > root->max_depth)
+		root->max_depth = cursor->nr;
 
 	return 0;
 }
 
 static int
-merge_chain_branch(struct callchain_node *dst, struct callchain_node *src,
-		   struct resolved_chain *chain)
+merge_chain_branch(struct callchain_cursor *cursor,
+		   struct callchain_node *dst, struct callchain_node *src)
 {
+	struct callchain_cursor_node **old_last = cursor->last;
 	struct callchain_node *child, *next_child;
 	struct callchain_list *list, *next_list;
-	int old_pos = chain->nr;
+	int old_pos = cursor->nr;
 	int err = 0;
 
 	list_for_each_entry_safe(list, next_list, &src->val, list) {
-		chain->ips[chain->nr].ip = list->ip;
-		chain->ips[chain->nr].ms = list->ms;
-		chain->nr++;
+		callchain_cursor_append(cursor, list->ip,
+					list->ms.map, list->ms.sym);
 		list_del(&list->list);
 		free(list);
 	}
 
-	if (src->hit)
-		append_chain_children(dst, chain, 0, src->hit);
+	if (src->hit) {
+		callchain_cursor_commit(cursor);
+		append_chain_children(dst, cursor, src->hit);
+	}
 
 	chain_for_each_child_safe(child, next_child, src) {
-		err = merge_chain_branch(dst, child, chain);
+		err = merge_chain_branch(cursor, dst, child);
 		if (err)
 			break;
 
-		list_del(&child->brothers);
+		list_del(&child->siblings);
 		free(child);
 	}
 
-	chain->nr = old_pos;
+	cursor->nr = old_pos;
+	cursor->last = old_last;
 
 	return err;
 }
 
-int callchain_merge(struct callchain_root *dst, struct callchain_root *src)
+int callchain_merge(struct callchain_cursor *cursor,
+		    struct callchain_root *dst, struct callchain_root *src)
+{
+	return merge_chain_branch(cursor, &dst->node, &src->node);
+}
+
+int callchain_cursor_append(struct callchain_cursor *cursor,
+			    u64 ip, struct map *map, struct symbol *sym)
 {
-	struct resolved_chain *chain;
-	int err;
+	struct callchain_cursor_node *node = *cursor->last;
 
-	chain = malloc(sizeof(*chain) +
-		       src->max_depth * sizeof(struct resolved_ip));
-	if (!chain)
-		return -ENOMEM;
+	if (!node) {
+		node = calloc(sizeof(*node), 1);
+		if (!node)
+			return -ENOMEM;
 
-	chain->nr = 0;
+		*cursor->last = node;
+	}
 
-	err = merge_chain_branch(&dst->node, &src->node, chain);
+	node->ip = ip;
+	node->map = map;
+	node->sym = sym;
 
-	free(chain);
+	cursor->nr++;
 
-	return err;
+	cursor->last = &node->next;
+
+	return 0;
 }
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
index c15fb8c24ad2..1a79df9f739f 100644
--- a/tools/perf/util/callchain.h
+++ b/tools/perf/util/callchain.h
@@ -16,7 +16,7 @@ enum chain_mode {
 
 struct callchain_node {
 	struct callchain_node	*parent;
-	struct list_head	brothers;
+	struct list_head	siblings;
 	struct list_head	children;
 	struct list_head	val;
 	struct rb_node		rb_node; /* to sort nodes in an rbtree */
@@ -49,9 +49,30 @@ struct callchain_list {
 	struct list_head	list;
 };
 
+/*
+ * A callchain cursor is a single linked list that
+ * let one feed a callchain progressively.
+ * It keeps persitent allocated entries to minimize
+ * allocations.
+ */
+struct callchain_cursor_node {
+	u64				ip;
+	struct map			*map;
+	struct symbol			*sym;
+	struct callchain_cursor_node	*next;
+};
+
+struct callchain_cursor {
+	u64				nr;
+	struct callchain_cursor_node	*first;
+	struct callchain_cursor_node	**last;
+	u64				pos;
+	struct callchain_cursor_node	*curr;
+};
+
 static inline void callchain_init(struct callchain_root *root)
 {
-	INIT_LIST_HEAD(&root->node.brothers);
+	INIT_LIST_HEAD(&root->node.siblings);
 	INIT_LIST_HEAD(&root->node.children);
 	INIT_LIST_HEAD(&root->node.val);
 
@@ -61,15 +82,54 @@ static inline void callchain_init(struct callchain_root *root)
 	root->max_depth = 0;
 }
 
-static inline u64 cumul_hits(struct callchain_node *node)
+static inline u64 callchain_cumul_hits(struct callchain_node *node)
 {
 	return node->hit + node->children_hit;
 }
 
-int register_callchain_param(struct callchain_param *param);
-int callchain_append(struct callchain_root *root, struct ip_callchain *chain,
-		     struct map_symbol *syms, u64 period);
-int callchain_merge(struct callchain_root *dst, struct callchain_root *src);
+int callchain_register_param(struct callchain_param *param);
+int callchain_append(struct callchain_root *root,
+		     struct callchain_cursor *cursor,
+		     u64 period);
+
+int callchain_merge(struct callchain_cursor *cursor,
+		    struct callchain_root *dst, struct callchain_root *src);
+
+bool ip_callchain__valid(struct ip_callchain *chain,
+			 const union perf_event *event);
+/*
+ * Initialize a cursor before adding entries inside, but keep
+ * the previously allocated entries as a cache.
+ */
+static inline void callchain_cursor_reset(struct callchain_cursor *cursor)
+{
+	cursor->nr = 0;
+	cursor->last = &cursor->first;
+}
+
+int callchain_cursor_append(struct callchain_cursor *cursor, u64 ip,
+			    struct map *map, struct symbol *sym);
 
-bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event);
+/* Close a cursor writing session. Initialize for the reader */
+static inline void callchain_cursor_commit(struct callchain_cursor *cursor)
+{
+	cursor->curr = cursor->first;
+	cursor->pos = 0;
+}
+
+/* Cursor reading iteration helpers */
+static inline struct callchain_cursor_node *
+callchain_cursor_current(struct callchain_cursor *cursor)
+{
+	if (cursor->pos == cursor->nr)
+		return NULL;
+
+	return cursor->curr;
+}
+
+static inline void callchain_cursor_advance(struct callchain_cursor *cursor)
+{
+	cursor->curr = cursor->curr->next;
+	cursor->pos++;
+}
 #endif	/* __PERF_CALLCHAIN_H */
diff --git a/tools/perf/util/cgroup.c b/tools/perf/util/cgroup.c
new file mode 100644
index 000000000000..9fea75535221
--- /dev/null
+++ b/tools/perf/util/cgroup.c
@@ -0,0 +1,178 @@
+#include "util.h"
+#include "../perf.h"
+#include "parse-options.h"
+#include "evsel.h"
+#include "cgroup.h"
+#include "debugfs.h" /* MAX_PATH, STR() */
+#include "evlist.h"
+
+int nr_cgroups;
+
+static int
+cgroupfs_find_mountpoint(char *buf, size_t maxlen)
+{
+	FILE *fp;
+	char mountpoint[MAX_PATH+1], tokens[MAX_PATH+1], type[MAX_PATH+1];
+	char *token, *saved_ptr;
+	int found = 0;
+
+	fp = fopen("/proc/mounts", "r");
+	if (!fp)
+		return -1;
+
+	/*
+	 * in order to handle split hierarchy, we need to scan /proc/mounts
+	 * and inspect every cgroupfs mount point to find one that has
+	 * perf_event subsystem
+	 */
+	while (fscanf(fp, "%*s %"STR(MAX_PATH)"s %"STR(MAX_PATH)"s %"
+				STR(MAX_PATH)"s %*d %*d\n",
+				mountpoint, type, tokens) == 3) {
+
+		if (!strcmp(type, "cgroup")) {
+
+			token = strtok_r(tokens, ",", &saved_ptr);
+
+			while (token != NULL) {
+				if (!strcmp(token, "perf_event")) {
+					found = 1;
+					break;
+				}
+				token = strtok_r(NULL, ",", &saved_ptr);
+			}
+		}
+		if (found)
+			break;
+	}
+	fclose(fp);
+	if (!found)
+		return -1;
+
+	if (strlen(mountpoint) < maxlen) {
+		strcpy(buf, mountpoint);
+		return 0;
+	}
+	return -1;
+}
+
+static int open_cgroup(char *name)
+{
+	char path[MAX_PATH+1];
+	char mnt[MAX_PATH+1];
+	int fd;
+
+
+	if (cgroupfs_find_mountpoint(mnt, MAX_PATH+1))
+		return -1;
+
+	snprintf(path, MAX_PATH, "%s/%s", mnt, name);
+
+	fd = open(path, O_RDONLY);
+	if (fd == -1)
+		fprintf(stderr, "no access to cgroup %s\n", path);
+
+	return fd;
+}
+
+static int add_cgroup(struct perf_evlist *evlist, char *str)
+{
+	struct perf_evsel *counter;
+	struct cgroup_sel *cgrp = NULL;
+	int n;
+	/*
+	 * check if cgrp is already defined, if so we reuse it
+	 */
+	list_for_each_entry(counter, &evlist->entries, node) {
+		cgrp = counter->cgrp;
+		if (!cgrp)
+			continue;
+		if (!strcmp(cgrp->name, str))
+			break;
+
+		cgrp = NULL;
+	}
+
+	if (!cgrp) {
+		cgrp = zalloc(sizeof(*cgrp));
+		if (!cgrp)
+			return -1;
+
+		cgrp->name = str;
+
+		cgrp->fd = open_cgroup(str);
+		if (cgrp->fd == -1) {
+			free(cgrp);
+			return -1;
+		}
+	}
+
+	/*
+	 * find corresponding event
+	 * if add cgroup N, then need to find event N
+	 */
+	n = 0;
+	list_for_each_entry(counter, &evlist->entries, node) {
+		if (n == nr_cgroups)
+			goto found;
+		n++;
+	}
+	if (cgrp->refcnt == 0)
+		free(cgrp);
+
+	return -1;
+found:
+	cgrp->refcnt++;
+	counter->cgrp = cgrp;
+	return 0;
+}
+
+void close_cgroup(struct cgroup_sel *cgrp)
+{
+	if (!cgrp)
+		return;
+
+	/* XXX: not reentrant */
+	if (--cgrp->refcnt == 0) {
+		close(cgrp->fd);
+		free(cgrp->name);
+		free(cgrp);
+	}
+}
+
+int parse_cgroups(const struct option *opt __used, const char *str,
+		  int unset __used)
+{
+	struct perf_evlist *evlist = *(struct perf_evlist **)opt->value;
+	const char *p, *e, *eos = str + strlen(str);
+	char *s;
+	int ret;
+
+	if (list_empty(&evlist->entries)) {
+		fprintf(stderr, "must define events before cgroups\n");
+		return -1;
+	}
+
+	for (;;) {
+		p = strchr(str, ',');
+		e = p ? p : eos;
+
+		/* allow empty cgroups, i.e., skip */
+		if (e - str) {
+			/* termination added */
+			s = strndup(str, e - str);
+			if (!s)
+				return -1;
+			ret = add_cgroup(evlist, s);
+			if (ret) {
+				free(s);
+				return -1;
+			}
+		}
+		/* nr_cgroups is increased een for empty cgroups */
+		nr_cgroups++;
+		if (!p)
+			break;
+		str = p+1;
+	}
+	return 0;
+}
diff --git a/tools/perf/util/cgroup.h b/tools/perf/util/cgroup.h
new file mode 100644
index 000000000000..89acd6debdc5
--- /dev/null
+++ b/tools/perf/util/cgroup.h
@@ -0,0 +1,17 @@
+#ifndef __CGROUP_H__
+#define __CGROUP_H__
+
+struct option;
+
+struct cgroup_sel {
+	char *name;
+	int fd;
+	int refcnt;
+};
+
+
+extern int nr_cgroups; /* number of explicit cgroups defined */
+extern void close_cgroup(struct cgroup_sel *cgrp);
+extern int parse_cgroups(const struct option *opt, const char *str, int unset);
+
+#endif /* __CGROUP_H__ */
diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index 3ccaa1043383..6893eec693ab 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -177,3 +177,8 @@ struct cpu_map *cpu_map__dummy_new(void)
 
 	return cpus;
 }
+
+void cpu_map__delete(struct cpu_map *map)
+{
+	free(map);
+}
diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h
index f7a4f42f6307..072c0a374794 100644
--- a/tools/perf/util/cpumap.h
+++ b/tools/perf/util/cpumap.h
@@ -8,6 +8,6 @@ struct cpu_map {
 
 struct cpu_map *cpu_map__new(const char *cpu_list);
 struct cpu_map *cpu_map__dummy_new(void);
-void *cpu_map__delete(struct cpu_map *map);
+void cpu_map__delete(struct cpu_map *map);
 
 #endif /* __PERF_CPUMAP_H */
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index 01bbe8ecec3f..d4536a9e0d8c 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -57,7 +57,7 @@ void ui__warning(const char *format, ...)
 }
 #endif
 
-void trace_event(event_t *event)
+void trace_event(union perf_event *event)
 {
 	unsigned char *raw_event = (void *)event;
 	const char *color = PERF_COLOR_BLUE;
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index ca35fd66b5df..93516cf4682c 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -9,7 +9,7 @@ extern int verbose;
 extern bool quiet, dump_trace;
 
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
-void trace_event(event_t *event);
+void trace_event(union perf_event *event);
 
 struct ui_progress;
 
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index 50d0a931497a..2b15c362ef56 100644
--- a/tools/perf/util/event.c
+++ b/tools/perf/util/event.c
@@ -6,8 +6,9 @@
 #include "string.h"
 #include "strlist.h"
 #include "thread.h"
+#include "thread_map.h"
 
-static const char *event__name[] = {
+static const char *perf_event__names[] = {
 	[0]			 = "TOTAL",
 	[PERF_RECORD_MMAP]	 = "MMAP",
 	[PERF_RECORD_LOST]	 = "LOST",
@@ -25,16 +26,16 @@ static const char *event__name[] = {
 	[PERF_RECORD_FINISHED_ROUND]	 = "FINISHED_ROUND",
 };
 
-const char *event__get_event_name(unsigned int id)
+const char *perf_event__name(unsigned int id)
 {
-	if (id >= ARRAY_SIZE(event__name))
+	if (id >= ARRAY_SIZE(perf_event__names))
 		return "INVALID";
-	if (!event__name[id])
+	if (!perf_event__names[id])
 		return "UNKNOWN";
-	return event__name[id];
+	return perf_event__names[id];
 }
 
-static struct sample_data synth_sample = {
+static struct perf_sample synth_sample = {
 	.pid	   = -1,
 	.tid	   = -1,
 	.time	   = -1,
@@ -43,9 +44,9 @@ static struct sample_data synth_sample = {
 	.period	   = 1,
 };
 
-static pid_t event__synthesize_comm(event_t *event, pid_t pid, int full,
-				    event__handler_t process,
-				    struct perf_session *session)
+static pid_t perf_event__synthesize_comm(union perf_event *event, pid_t pid,
+					 int full, perf_event__handler_t process,
+					 struct perf_session *session)
 {
 	char filename[PATH_MAX];
 	char bf[BUFSIZ];
@@ -126,9 +127,10 @@ out:
 	return tgid;
 }
 
-static int event__synthesize_mmap_events(event_t *event, pid_t pid, pid_t tgid,
-					 event__handler_t process,
-					 struct perf_session *session)
+static int perf_event__synthesize_mmap_events(union perf_event *event,
+					      pid_t pid, pid_t tgid,
+					      perf_event__handler_t process,
+					      struct perf_session *session)
 {
 	char filename[PATH_MAX];
 	FILE *fp;
@@ -199,14 +201,14 @@ static int event__synthesize_mmap_events(event_t *event, pid_t pid, pid_t tgid,
 	return 0;
 }
 
-int event__synthesize_modules(event__handler_t process,
-			      struct perf_session *session,
-			      struct machine *machine)
+int perf_event__synthesize_modules(perf_event__handler_t process,
+				   struct perf_session *session,
+				   struct machine *machine)
 {
 	struct rb_node *nd;
 	struct map_groups *kmaps = &machine->kmaps;
-	event_t *event = zalloc(sizeof(event->mmap) + session->id_hdr_size);
-
+	union perf_event *event = zalloc((sizeof(event->mmap) +
+					  session->id_hdr_size));
 	if (event == NULL) {
 		pr_debug("Not enough memory synthesizing mmap event "
 			 "for kernel modules\n");
@@ -251,23 +253,24 @@ int event__synthesize_modules(event__handler_t process,
 	return 0;
 }
 
-static int __event__synthesize_thread(event_t *comm_event, event_t *mmap_event,
-				      pid_t pid, event__handler_t process,
+static int __event__synthesize_thread(union perf_event *comm_event,
+				      union perf_event *mmap_event,
+				      pid_t pid, perf_event__handler_t process,
 				      struct perf_session *session)
 {
-	pid_t tgid = event__synthesize_comm(comm_event, pid, 1, process,
+	pid_t tgid = perf_event__synthesize_comm(comm_event, pid, 1, process,
 					    session);
 	if (tgid == -1)
 		return -1;
-	return event__synthesize_mmap_events(mmap_event, pid, tgid,
+	return perf_event__synthesize_mmap_events(mmap_event, pid, tgid,
 					     process, session);
 }
 
-int event__synthesize_thread_map(struct thread_map *threads,
-				 event__handler_t process,
-				 struct perf_session *session)
+int perf_event__synthesize_thread_map(struct thread_map *threads,
+				      perf_event__handler_t process,
+				      struct perf_session *session)
 {
-	event_t *comm_event, *mmap_event;
+	union perf_event *comm_event, *mmap_event;
 	int err = -1, thread;
 
 	comm_event = malloc(sizeof(comm_event->comm) + session->id_hdr_size);
@@ -294,12 +297,12 @@ out:
 	return err;
 }
 
-int event__synthesize_threads(event__handler_t process,
-			      struct perf_session *session)
+int perf_event__synthesize_threads(perf_event__handler_t process,
+				   struct perf_session *session)
 {
 	DIR *proc;
 	struct dirent dirent, *next;
-	event_t *comm_event, *mmap_event;
+	union perf_event *comm_event, *mmap_event;
 	int err = -1;
 
 	comm_event = malloc(sizeof(comm_event->comm) + session->id_hdr_size);
@@ -357,10 +360,10 @@ static int find_symbol_cb(void *arg, const char *name, char type,
 	return 1;
 }
 
-int event__synthesize_kernel_mmap(event__handler_t process,
-				  struct perf_session *session,
-				  struct machine *machine,
-				  const char *symbol_name)
+int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
+				       struct perf_session *session,
+				       struct machine *machine,
+				       const char *symbol_name)
 {
 	size_t size;
 	const char *filename, *mmap_name;
@@ -374,8 +377,8 @@ int event__synthesize_kernel_mmap(event__handler_t process,
 	 * kernels.
 	 */
 	struct process_symbol_args args = { .name = symbol_name, };
-	event_t *event = zalloc(sizeof(event->mmap) + session->id_hdr_size);
-
+	union perf_event *event = zalloc((sizeof(event->mmap) +
+					  session->id_hdr_size));
 	if (event == NULL) {
 		pr_debug("Not enough memory synthesizing mmap event "
 			 "for kernel modules\n");
@@ -421,42 +424,15 @@ int event__synthesize_kernel_mmap(event__handler_t process,
 	return err;
 }
 
-static void thread__comm_adjust(struct thread *self, struct hists *hists)
-{
-	char *comm = self->comm;
-
-	if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
-	    (!symbol_conf.comm_list ||
-	     strlist__has_entry(symbol_conf.comm_list, comm))) {
-		u16 slen = strlen(comm);
-
-		if (hists__new_col_len(hists, HISTC_COMM, slen))
-			hists__set_col_len(hists, HISTC_THREAD, slen + 6);
-	}
-}
-
-static int thread__set_comm_adjust(struct thread *self, const char *comm,
-				   struct hists *hists)
+int perf_event__process_comm(union perf_event *event,
+			     struct perf_sample *sample __used,
+			     struct perf_session *session)
 {
-	int ret = thread__set_comm(self, comm);
-
-	if (ret)
-		return ret;
-
-	thread__comm_adjust(self, hists);
+	struct thread *thread = perf_session__findnew(session, event->comm.tid);
 
-	return 0;
-}
+	dump_printf(": %s:%d\n", event->comm.comm, event->comm.tid);
 
-int event__process_comm(event_t *self, struct sample_data *sample __used,
-			struct perf_session *session)
-{
-	struct thread *thread = perf_session__findnew(session, self->comm.tid);
-
-	dump_printf(": %s:%d\n", self->comm.comm, self->comm.tid);
-
-	if (thread == NULL || thread__set_comm_adjust(thread, self->comm.comm,
-						      &session->hists)) {
+	if (thread == NULL || thread__set_comm(thread, event->comm.comm)) {
 		dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
 		return -1;
 	}
@@ -464,19 +440,21 @@ int event__process_comm(event_t *self, struct sample_data *sample __used,
 	return 0;
 }
 
-int event__process_lost(event_t *self, struct sample_data *sample __used,
-			struct perf_session *session)
+int perf_event__process_lost(union perf_event *event,
+			     struct perf_sample *sample __used,
+			     struct perf_session *session)
 {
 	dump_printf(": id:%" PRIu64 ": lost:%" PRIu64 "\n",
-		    self->lost.id, self->lost.lost);
-	session->hists.stats.total_lost += self->lost.lost;
+		    event->lost.id, event->lost.lost);
+	session->hists.stats.total_lost += event->lost.lost;
 	return 0;
 }
 
-static void event_set_kernel_mmap_len(struct map **maps, event_t *self)
+static void perf_event__set_kernel_mmap_len(union perf_event *event,
+					    struct map **maps)
 {
-	maps[MAP__FUNCTION]->start = self->mmap.start;
-	maps[MAP__FUNCTION]->end   = self->mmap.start + self->mmap.len;
+	maps[MAP__FUNCTION]->start = event->mmap.start;
+	maps[MAP__FUNCTION]->end   = event->mmap.start + event->mmap.len;
 	/*
 	 * Be a bit paranoid here, some perf.data file came with
 	 * a zero sized synthesized MMAP event for the kernel.
@@ -485,8 +463,8 @@ static void event_set_kernel_mmap_len(struct map **maps, event_t *self)
 		maps[MAP__FUNCTION]->end = ~0ULL;
 }
 
-static int event__process_kernel_mmap(event_t *self,
-			struct perf_session *session)
+static int perf_event__process_kernel_mmap(union perf_event *event,
+					   struct perf_session *session)
 {
 	struct map *map;
 	char kmmap_prefix[PATH_MAX];
@@ -494,9 +472,9 @@ static int event__process_kernel_mmap(event_t *self,
 	enum dso_kernel_type kernel_type;
 	bool is_kernel_mmap;
 
-	machine = perf_session__findnew_machine(session, self->mmap.pid);
+	machine = perf_session__findnew_machine(session, event->mmap.pid);
 	if (!machine) {
-		pr_err("Can't find id %d's machine\n", self->mmap.pid);
+		pr_err("Can't find id %d's machine\n", event->mmap.pid);
 		goto out_problem;
 	}
 
@@ -506,17 +484,17 @@ static int event__process_kernel_mmap(event_t *self,
 	else
 		kernel_type = DSO_TYPE_GUEST_KERNEL;
 
-	is_kernel_mmap = memcmp(self->mmap.filename,
+	is_kernel_mmap = memcmp(event->mmap.filename,
 				kmmap_prefix,
 				strlen(kmmap_prefix)) == 0;
-	if (self->mmap.filename[0] == '/' ||
-	    (!is_kernel_mmap && self->mmap.filename[0] == '[')) {
+	if (event->mmap.filename[0] == '/' ||
+	    (!is_kernel_mmap && event->mmap.filename[0] == '[')) {
 
 		char short_module_name[1024];
 		char *name, *dot;
 
-		if (self->mmap.filename[0] == '/') {
-			name = strrchr(self->mmap.filename, '/');
+		if (event->mmap.filename[0] == '/') {
+			name = strrchr(event->mmap.filename, '/');
 			if (name == NULL)
 				goto out_problem;
 
@@ -528,10 +506,10 @@ static int event__process_kernel_mmap(event_t *self,
 					"[%.*s]", (int)(dot - name), name);
 			strxfrchar(short_module_name, '-', '_');
 		} else
-			strcpy(short_module_name, self->mmap.filename);
+			strcpy(short_module_name, event->mmap.filename);
 
-		map = machine__new_module(machine, self->mmap.start,
-					  self->mmap.filename);
+		map = machine__new_module(machine, event->mmap.start,
+					  event->mmap.filename);
 		if (map == NULL)
 			goto out_problem;
 
@@ -541,9 +519,9 @@ static int event__process_kernel_mmap(event_t *self,
 
 		map->dso->short_name = name;
 		map->dso->sname_alloc = 1;
-		map->end = map->start + self->mmap.len;
+		map->end = map->start + event->mmap.len;
 	} else if (is_kernel_mmap) {
-		const char *symbol_name = (self->mmap.filename +
+		const char *symbol_name = (event->mmap.filename +
 				strlen(kmmap_prefix));
 		/*
 		 * Should be there already, from the build-id table in
@@ -558,10 +536,10 @@ static int event__process_kernel_mmap(event_t *self,
 		if (__machine__create_kernel_maps(machine, kernel) < 0)
 			goto out_problem;
 
-		event_set_kernel_mmap_len(machine->vmlinux_maps, self);
+		perf_event__set_kernel_mmap_len(event, machine->vmlinux_maps);
 		perf_session__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps,
 							 symbol_name,
-							 self->mmap.pgoff);
+							 event->mmap.pgoff);
 		if (machine__is_default_guest(machine)) {
 			/*
 			 * preload dso of guest kernel and modules
@@ -575,22 +553,23 @@ out_problem:
 	return -1;
 }
 
-int event__process_mmap(event_t *self, struct sample_data *sample __used,
-			struct perf_session *session)
+int perf_event__process_mmap(union perf_event *event,
+			     struct perf_sample *sample __used,
+			     struct perf_session *session)
 {
 	struct machine *machine;
 	struct thread *thread;
 	struct map *map;
-	u8 cpumode = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
 	int ret = 0;
 
 	dump_printf(" %d/%d: [%#" PRIx64 "(%#" PRIx64 ") @ %#" PRIx64 "]: %s\n",
-			self->mmap.pid, self->mmap.tid, self->mmap.start,
-			self->mmap.len, self->mmap.pgoff, self->mmap.filename);
+			event->mmap.pid, event->mmap.tid, event->mmap.start,
+			event->mmap.len, event->mmap.pgoff, event->mmap.filename);
 
 	if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL ||
 	    cpumode == PERF_RECORD_MISC_KERNEL) {
-		ret = event__process_kernel_mmap(self, session);
+		ret = perf_event__process_kernel_mmap(event, session);
 		if (ret < 0)
 			goto out_problem;
 		return 0;
@@ -599,12 +578,12 @@ int event__process_mmap(event_t *self, struct sample_data *sample __used,
 	machine = perf_session__find_host_machine(session);
 	if (machine == NULL)
 		goto out_problem;
-	thread = perf_session__findnew(session, self->mmap.pid);
+	thread = perf_session__findnew(session, event->mmap.pid);
 	if (thread == NULL)
 		goto out_problem;
-	map = map__new(&machine->user_dsos, self->mmap.start,
-			self->mmap.len, self->mmap.pgoff,
-			self->mmap.pid, self->mmap.filename,
+	map = map__new(&machine->user_dsos, event->mmap.start,
+			event->mmap.len, event->mmap.pgoff,
+			event->mmap.pid, event->mmap.filename,
 			MAP__FUNCTION);
 	if (map == NULL)
 		goto out_problem;
@@ -617,16 +596,17 @@ out_problem:
 	return 0;
 }
 
-int event__process_task(event_t *self, struct sample_data *sample __used,
-			struct perf_session *session)
+int perf_event__process_task(union perf_event *event,
+			     struct perf_sample *sample __used,
+			     struct perf_session *session)
 {
-	struct thread *thread = perf_session__findnew(session, self->fork.tid);
-	struct thread *parent = perf_session__findnew(session, self->fork.ptid);
+	struct thread *thread = perf_session__findnew(session, event->fork.tid);
+	struct thread *parent = perf_session__findnew(session, event->fork.ptid);
 
-	dump_printf("(%d:%d):(%d:%d)\n", self->fork.pid, self->fork.tid,
-		    self->fork.ppid, self->fork.ptid);
+	dump_printf("(%d:%d):(%d:%d)\n", event->fork.pid, event->fork.tid,
+		    event->fork.ppid, event->fork.ptid);
 
-	if (self->header.type == PERF_RECORD_EXIT) {
+	if (event->header.type == PERF_RECORD_EXIT) {
 		perf_session__remove_thread(session, thread);
 		return 0;
 	}
@@ -640,20 +620,22 @@ int event__process_task(event_t *self, struct sample_data *sample __used,
 	return 0;
 }
 
-int event__process(event_t *event, struct sample_data *sample,
-		   struct perf_session *session)
+int perf_event__process(union perf_event *event, struct perf_sample *sample,
+			struct perf_session *session)
 {
 	switch (event->header.type) {
 	case PERF_RECORD_COMM:
-		event__process_comm(event, sample, session);
+		perf_event__process_comm(event, sample, session);
 		break;
 	case PERF_RECORD_MMAP:
-		event__process_mmap(event, sample, session);
+		perf_event__process_mmap(event, sample, session);
 		break;
 	case PERF_RECORD_FORK:
 	case PERF_RECORD_EXIT:
-		event__process_task(event, sample, session);
+		perf_event__process_task(event, sample, session);
 		break;
+	case PERF_RECORD_LOST:
+		perf_event__process_lost(event, sample, session);
 	default:
 		break;
 	}
@@ -750,24 +732,14 @@ void thread__find_addr_location(struct thread *self,
 		al->sym = NULL;
 }
 
-static void dso__calc_col_width(struct dso *self, struct hists *hists)
-{
-	if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
-	    (!symbol_conf.dso_list ||
-	     strlist__has_entry(symbol_conf.dso_list, self->name))) {
-		u16 slen = dso__name_len(self);
-		hists__new_col_len(hists, HISTC_DSO, slen);
-	}
-
-	self->slen_calculated = 1;
-}
-
-int event__preprocess_sample(const event_t *self, struct perf_session *session,
-			     struct addr_location *al, struct sample_data *data,
-			     symbol_filter_t filter)
+int perf_event__preprocess_sample(const union perf_event *event,
+				  struct perf_session *session,
+				  struct addr_location *al,
+				  struct perf_sample *sample,
+				  symbol_filter_t filter)
 {
-	u8 cpumode = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-	struct thread *thread = perf_session__findnew(session, self->ip.pid);
+	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+	struct thread *thread = perf_session__findnew(session, event->ip.pid);
 
 	if (thread == NULL)
 		return -1;
@@ -789,12 +761,12 @@ int event__preprocess_sample(const event_t *self, struct perf_session *session,
 		machine__create_kernel_maps(&session->host_machine);
 
 	thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION,
-			      self->ip.pid, self->ip.ip, al);
+			      event->ip.pid, event->ip.ip, al);
 	dump_printf(" ...... dso: %s\n",
 		    al->map ? al->map->dso->long_name :
 			al->level == 'H' ? "[hypervisor]" : "<not found>");
 	al->sym = NULL;
-	al->cpu = data->cpu;
+	al->cpu = sample->cpu;
 
 	if (al->map) {
 		if (symbol_conf.dso_list &&
@@ -805,23 +777,8 @@ int event__preprocess_sample(const event_t *self, struct perf_session *session,
 			strlist__has_entry(symbol_conf.dso_list,
 					   al->map->dso->long_name)))))
 			goto out_filtered;
-		/*
-		 * We have to do this here as we may have a dso with no symbol
-		 * hit that has a name longer than the ones with symbols
-		 * sampled.
-		 */
-		if (!sort_dso.elide && !al->map->dso->slen_calculated)
-			dso__calc_col_width(al->map->dso, &session->hists);
 
 		al->sym = map__find_symbol(al->map, al->addr, filter);
-	} else {
-		const unsigned int unresolved_col_width = BITS_PER_LONG / 4;
-
-		if (hists__col_len(&session->hists, HISTC_DSO) < unresolved_col_width &&
-		    !symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
-		    !symbol_conf.dso_list)
-			hists__set_col_len(&session->hists, HISTC_DSO,
-					   unresolved_col_width);
 	}
 
 	if (symbol_conf.sym_list && al->sym &&
@@ -834,128 +791,3 @@ out_filtered:
 	al->filtered = true;
 	return 0;
 }
-
-static int event__parse_id_sample(const event_t *event,
-				  struct perf_session *session,
-				  struct sample_data *sample)
-{
-	const u64 *array;
-	u64 type;
-
-	sample->cpu = sample->pid = sample->tid = -1;
-	sample->stream_id = sample->id = sample->time = -1ULL;
-
-	if (!session->sample_id_all)
-		return 0;
-
-	array = event->sample.array;
-	array += ((event->header.size -
-		   sizeof(event->header)) / sizeof(u64)) - 1;
-	type = session->sample_type;
-
-	if (type & PERF_SAMPLE_CPU) {
-		u32 *p = (u32 *)array;
-		sample->cpu = *p;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_STREAM_ID) {
-		sample->stream_id = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_ID) {
-		sample->id = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_TIME) {
-		sample->time = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_TID) {
-		u32 *p = (u32 *)array;
-		sample->pid = p[0];
-		sample->tid = p[1];
-	}
-
-	return 0;
-}
-
-int event__parse_sample(const event_t *event, struct perf_session *session,
-			struct sample_data *data)
-{
-	const u64 *array;
-	u64 type;
-
-	if (event->header.type != PERF_RECORD_SAMPLE)
-		return event__parse_id_sample(event, session, data);
-
-	array = event->sample.array;
-	type = session->sample_type;
-
-	if (type & PERF_SAMPLE_IP) {
-		data->ip = event->ip.ip;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_TID) {
-		u32 *p = (u32 *)array;
-		data->pid = p[0];
-		data->tid = p[1];
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_TIME) {
-		data->time = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_ADDR) {
-		data->addr = *array;
-		array++;
-	}
-
-	data->id = -1ULL;
-	if (type & PERF_SAMPLE_ID) {
-		data->id = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_STREAM_ID) {
-		data->stream_id = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_CPU) {
-		u32 *p = (u32 *)array;
-		data->cpu = *p;
-		array++;
-	} else
-		data->cpu = -1;
-
-	if (type & PERF_SAMPLE_PERIOD) {
-		data->period = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_READ) {
-		pr_debug("PERF_SAMPLE_READ is unsuported for now\n");
-		return -1;
-	}
-
-	if (type & PERF_SAMPLE_CALLCHAIN) {
-		data->callchain = (struct ip_callchain *)array;
-		array += 1 + data->callchain->nr;
-	}
-
-	if (type & PERF_SAMPLE_RAW) {
-		u32 *p = (u32 *)array;
-		data->raw_size = *p;
-		p++;
-		data->raw_data = p;
-	}
-
-	return 0;
-}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index cc7b52f9b492..9c35170fb379 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -61,7 +61,7 @@ struct sample_event {
 	u64 array[];
 };
 
-struct sample_data {
+struct perf_sample {
 	u64 ip;
 	u32 pid, tid;
 	u64 time;
@@ -117,7 +117,7 @@ struct tracing_data_event {
 	u32 size;
 };
 
-typedef union event_union {
+union perf_event {
 	struct perf_event_header	header;
 	struct ip_event			ip;
 	struct mmap_event		mmap;
@@ -130,50 +130,54 @@ typedef union event_union {
 	struct event_type_event		event_type;
 	struct tracing_data_event	tracing_data;
 	struct build_id_event		build_id;
-} event_t;
+};
 
-void event__print_totals(void);
+void perf_event__print_totals(void);
 
 struct perf_session;
 struct thread_map;
 
-typedef int (*event__handler_synth_t)(event_t *event, 
+typedef int (*perf_event__handler_synth_t)(union perf_event *event, 
+					   struct perf_session *session);
+typedef int (*perf_event__handler_t)(union perf_event *event,
+				     struct perf_sample *sample,
 				      struct perf_session *session);
-typedef int (*event__handler_t)(event_t *event, struct sample_data *sample,
-				struct perf_session *session);
-
-int event__synthesize_thread_map(struct thread_map *threads,
-				 event__handler_t process,
-				 struct perf_session *session);
-int event__synthesize_threads(event__handler_t process,
-			      struct perf_session *session);
-int event__synthesize_kernel_mmap(event__handler_t process,
-				struct perf_session *session,
-				struct machine *machine,
-				const char *symbol_name);
-
-int event__synthesize_modules(event__handler_t process,
-			      struct perf_session *session,
-			      struct machine *machine);
-
-int event__process_comm(event_t *self, struct sample_data *sample,
-			struct perf_session *session);
-int event__process_lost(event_t *self, struct sample_data *sample,
-			struct perf_session *session);
-int event__process_mmap(event_t *self, struct sample_data *sample,
-			struct perf_session *session);
-int event__process_task(event_t *self, struct sample_data *sample,
+
+int perf_event__synthesize_thread_map(struct thread_map *threads,
+				      perf_event__handler_t process,
+				      struct perf_session *session);
+int perf_event__synthesize_threads(perf_event__handler_t process,
+				   struct perf_session *session);
+int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
+				       struct perf_session *session,
+				       struct machine *machine,
+				       const char *symbol_name);
+
+int perf_event__synthesize_modules(perf_event__handler_t process,
+				   struct perf_session *session,
+				   struct machine *machine);
+
+int perf_event__process_comm(union perf_event *event, struct perf_sample *sample,
+			     struct perf_session *session);
+int perf_event__process_lost(union perf_event *event, struct perf_sample *sample,
+			     struct perf_session *session);
+int perf_event__process_mmap(union perf_event *event, struct perf_sample *sample,
+			     struct perf_session *session);
+int perf_event__process_task(union perf_event *event, struct perf_sample *sample,
+			     struct perf_session *session);
+int perf_event__process(union perf_event *event, struct perf_sample *sample,
 			struct perf_session *session);
-int event__process(event_t *event, struct sample_data *sample,
-		   struct perf_session *session);
 
 struct addr_location;
-int event__preprocess_sample(const event_t *self, struct perf_session *session,
-			     struct addr_location *al, struct sample_data *data,
-			     symbol_filter_t filter);
-int event__parse_sample(const event_t *event, struct perf_session *session,
-			struct sample_data *sample);
+int perf_event__preprocess_sample(const union perf_event *self,
+				  struct perf_session *session,
+				  struct addr_location *al,
+				  struct perf_sample *sample,
+				  symbol_filter_t filter);
+
+const char *perf_event__name(unsigned int id);
 
-const char *event__get_event_name(unsigned int id);
+int perf_event__parse_sample(const union perf_event *event, u64 type,
+			     bool sample_id_all, struct perf_sample *sample);
 
 #endif /* __PERF_RECORD_H */
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
new file mode 100644
index 000000000000..d852cefa20de
--- /dev/null
+++ b/tools/perf/util/evlist.c
@@ -0,0 +1,394 @@
+/*
+ * Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Parts came from builtin-{top,stat,record}.c, see those files for further
+ * copyright notes.
+ *
+ * Released under the GPL v2. (and only v2, not any later version)
+ */
+#include <poll.h>
+#include "cpumap.h"
+#include "thread_map.h"
+#include "evlist.h"
+#include "evsel.h"
+#include "util.h"
+
+#include <sys/mman.h>
+
+#include <linux/bitops.h>
+#include <linux/hash.h>
+
+#define FD(e, x, y) (*(int *)xyarray__entry(e->fd, x, y))
+#define SID(e, x, y) xyarray__entry(e->sample_id, x, y)
+
+void perf_evlist__init(struct perf_evlist *evlist, struct cpu_map *cpus,
+		       struct thread_map *threads)
+{
+	int i;
+
+	for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i)
+		INIT_HLIST_HEAD(&evlist->heads[i]);
+	INIT_LIST_HEAD(&evlist->entries);
+	perf_evlist__set_maps(evlist, cpus, threads);
+}
+
+struct perf_evlist *perf_evlist__new(struct cpu_map *cpus,
+				     struct thread_map *threads)
+{
+	struct perf_evlist *evlist = zalloc(sizeof(*evlist));
+
+	if (evlist != NULL)
+		perf_evlist__init(evlist, cpus, threads);
+
+	return evlist;
+}
+
+static void perf_evlist__purge(struct perf_evlist *evlist)
+{
+	struct perf_evsel *pos, *n;
+
+	list_for_each_entry_safe(pos, n, &evlist->entries, node) {
+		list_del_init(&pos->node);
+		perf_evsel__delete(pos);
+	}
+
+	evlist->nr_entries = 0;
+}
+
+void perf_evlist__exit(struct perf_evlist *evlist)
+{
+	free(evlist->mmap);
+	free(evlist->pollfd);
+	evlist->mmap = NULL;
+	evlist->pollfd = NULL;
+}
+
+void perf_evlist__delete(struct perf_evlist *evlist)
+{
+	perf_evlist__purge(evlist);
+	perf_evlist__exit(evlist);
+	free(evlist);
+}
+
+void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry)
+{
+	list_add_tail(&entry->node, &evlist->entries);
+	++evlist->nr_entries;
+}
+
+int perf_evlist__add_default(struct perf_evlist *evlist)
+{
+	struct perf_event_attr attr = {
+		.type = PERF_TYPE_HARDWARE,
+		.config = PERF_COUNT_HW_CPU_CYCLES,
+	};
+	struct perf_evsel *evsel = perf_evsel__new(&attr, 0);
+
+	if (evsel == NULL)
+		return -ENOMEM;
+
+	perf_evlist__add(evlist, evsel);
+	return 0;
+}
+
+int perf_evlist__alloc_pollfd(struct perf_evlist *evlist)
+{
+	int nfds = evlist->cpus->nr * evlist->threads->nr * evlist->nr_entries;
+	evlist->pollfd = malloc(sizeof(struct pollfd) * nfds);
+	return evlist->pollfd != NULL ? 0 : -ENOMEM;
+}
+
+void perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd)
+{
+	fcntl(fd, F_SETFL, O_NONBLOCK);
+	evlist->pollfd[evlist->nr_fds].fd = fd;
+	evlist->pollfd[evlist->nr_fds].events = POLLIN;
+	evlist->nr_fds++;
+}
+
+static void perf_evlist__id_hash(struct perf_evlist *evlist,
+				 struct perf_evsel *evsel,
+				 int cpu, int thread, u64 id)
+{
+	int hash;
+	struct perf_sample_id *sid = SID(evsel, cpu, thread);
+
+	sid->id = id;
+	sid->evsel = evsel;
+	hash = hash_64(sid->id, PERF_EVLIST__HLIST_BITS);
+	hlist_add_head(&sid->node, &evlist->heads[hash]);
+}
+
+void perf_evlist__id_add(struct perf_evlist *evlist, struct perf_evsel *evsel,
+			 int cpu, int thread, u64 id)
+{
+	perf_evlist__id_hash(evlist, evsel, cpu, thread, id);
+	evsel->id[evsel->ids++] = id;
+}
+
+static int perf_evlist__id_add_fd(struct perf_evlist *evlist,
+				  struct perf_evsel *evsel,
+				  int cpu, int thread, int fd)
+{
+	u64 read_data[4] = { 0, };
+	int id_idx = 1; /* The first entry is the counter value */
+
+	if (!(evsel->attr.read_format & PERF_FORMAT_ID) ||
+	    read(fd, &read_data, sizeof(read_data)) == -1)
+		return -1;
+
+	if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
+		++id_idx;
+	if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING)
+		++id_idx;
+
+	perf_evlist__id_add(evlist, evsel, cpu, thread, read_data[id_idx]);
+	return 0;
+}
+
+struct perf_evsel *perf_evlist__id2evsel(struct perf_evlist *evlist, u64 id)
+{
+	struct hlist_head *head;
+	struct hlist_node *pos;
+	struct perf_sample_id *sid;
+	int hash;
+
+	if (evlist->nr_entries == 1)
+		return list_entry(evlist->entries.next, struct perf_evsel, node);
+
+	hash = hash_64(id, PERF_EVLIST__HLIST_BITS);
+	head = &evlist->heads[hash];
+
+	hlist_for_each_entry(sid, pos, head, node)
+		if (sid->id == id)
+			return sid->evsel;
+	return NULL;
+}
+
+union perf_event *perf_evlist__read_on_cpu(struct perf_evlist *evlist, int cpu)
+{
+	/* XXX Move this to perf.c, making it generally available */
+	unsigned int page_size = sysconf(_SC_PAGE_SIZE);
+	struct perf_mmap *md = &evlist->mmap[cpu];
+	unsigned int head = perf_mmap__read_head(md);
+	unsigned int old = md->prev;
+	unsigned char *data = md->base + page_size;
+	union perf_event *event = NULL;
+
+	if (evlist->overwrite) {
+		/*
+		 * 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.
+		 *
+		 * If we somehow ended up ahead of the head, we got messed up.
+		 *
+		 * In either case, truncate and restart at head.
+		 */
+		int diff = head - old;
+		if (diff > md->mask / 2 || diff < 0) {
+			fprintf(stderr, "WARNING: failed to keep up with mmap data.\n");
+
+			/*
+			 * head points to a known good entry, start there.
+			 */
+			old = head;
+		}
+	}
+
+	if (old != head) {
+		size_t size;
+
+		event = (union perf_event *)&data[old & md->mask];
+		size = event->header.size;
+
+		/*
+		 * Event straddles the mmap boundary -- header should always
+		 * be inside due to u64 alignment of output.
+		 */
+		if ((old & md->mask) + size != ((old + size) & md->mask)) {
+			unsigned int offset = old;
+			unsigned int len = min(sizeof(*event), size), cpy;
+			void *dst = &evlist->event_copy;
+
+			do {
+				cpy = min(md->mask + 1 - (offset & md->mask), len);
+				memcpy(dst, &data[offset & md->mask], cpy);
+				offset += cpy;
+				dst += cpy;
+				len -= cpy;
+			} while (len);
+
+			event = &evlist->event_copy;
+		}
+
+		old += size;
+	}
+
+	md->prev = old;
+
+	if (!evlist->overwrite)
+		perf_mmap__write_tail(md, old);
+
+	return event;
+}
+
+void perf_evlist__munmap(struct perf_evlist *evlist)
+{
+	int cpu;
+
+	for (cpu = 0; cpu < evlist->cpus->nr; cpu++) {
+		if (evlist->mmap[cpu].base != NULL) {
+			munmap(evlist->mmap[cpu].base, evlist->mmap_len);
+			evlist->mmap[cpu].base = NULL;
+		}
+	}
+}
+
+int perf_evlist__alloc_mmap(struct perf_evlist *evlist)
+{
+	evlist->mmap = zalloc(evlist->cpus->nr * sizeof(struct perf_mmap));
+	return evlist->mmap != NULL ? 0 : -ENOMEM;
+}
+
+static int __perf_evlist__mmap(struct perf_evlist *evlist, int cpu, int prot,
+			       int mask, int fd)
+{
+	evlist->mmap[cpu].prev = 0;
+	evlist->mmap[cpu].mask = mask;
+	evlist->mmap[cpu].base = mmap(NULL, evlist->mmap_len, prot,
+				      MAP_SHARED, fd, 0);
+	if (evlist->mmap[cpu].base == MAP_FAILED)
+		return -1;
+
+	perf_evlist__add_pollfd(evlist, fd);
+	return 0;
+}
+
+/** perf_evlist__mmap - Create per cpu maps to receive events
+ *
+ * @evlist - list of events
+ * @pages - map length in pages
+ * @overwrite - overwrite older events?
+ *
+ * If overwrite is false the user needs to signal event consuption using:
+ *
+ *	struct perf_mmap *m = &evlist->mmap[cpu];
+ *	unsigned int head = perf_mmap__read_head(m);
+ *
+ *	perf_mmap__write_tail(m, head)
+ *
+ * Using perf_evlist__read_on_cpu does this automatically.
+ */
+int perf_evlist__mmap(struct perf_evlist *evlist, int pages, bool overwrite)
+{
+	unsigned int page_size = sysconf(_SC_PAGE_SIZE);
+	int mask = pages * page_size - 1, cpu;
+	struct perf_evsel *first_evsel, *evsel;
+	const struct cpu_map *cpus = evlist->cpus;
+	const struct thread_map *threads = evlist->threads;
+	int thread, prot = PROT_READ | (overwrite ? 0 : PROT_WRITE);
+
+	if (evlist->mmap == NULL && perf_evlist__alloc_mmap(evlist) < 0)
+		return -ENOMEM;
+
+	if (evlist->pollfd == NULL && perf_evlist__alloc_pollfd(evlist) < 0)
+		return -ENOMEM;
+
+	evlist->overwrite = overwrite;
+	evlist->mmap_len = (pages + 1) * page_size;
+	first_evsel = list_entry(evlist->entries.next, struct perf_evsel, node);
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		if ((evsel->attr.read_format & PERF_FORMAT_ID) &&
+		    evsel->sample_id == NULL &&
+		    perf_evsel__alloc_id(evsel, cpus->nr, threads->nr) < 0)
+			return -ENOMEM;
+
+		for (cpu = 0; cpu < cpus->nr; cpu++) {
+			for (thread = 0; thread < threads->nr; thread++) {
+				int fd = FD(evsel, cpu, thread);
+
+				if (evsel->idx || thread) {
+					if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT,
+						  FD(first_evsel, cpu, 0)) != 0)
+						goto out_unmap;
+				} else if (__perf_evlist__mmap(evlist, cpu, prot, mask, fd) < 0)
+					goto out_unmap;
+
+				if ((evsel->attr.read_format & PERF_FORMAT_ID) &&
+				    perf_evlist__id_add_fd(evlist, evsel, cpu, thread, fd) < 0)
+					goto out_unmap;
+			}
+		}
+	}
+
+	return 0;
+
+out_unmap:
+	for (cpu = 0; cpu < cpus->nr; cpu++) {
+		if (evlist->mmap[cpu].base != NULL) {
+			munmap(evlist->mmap[cpu].base, evlist->mmap_len);
+			evlist->mmap[cpu].base = NULL;
+		}
+	}
+	return -1;
+}
+
+int perf_evlist__create_maps(struct perf_evlist *evlist, pid_t target_pid,
+			     pid_t target_tid, const char *cpu_list)
+{
+	evlist->threads = thread_map__new(target_pid, target_tid);
+
+	if (evlist->threads == NULL)
+		return -1;
+
+	if (target_tid != -1)
+		evlist->cpus = cpu_map__dummy_new();
+	else
+		evlist->cpus = cpu_map__new(cpu_list);
+
+	if (evlist->cpus == NULL)
+		goto out_delete_threads;
+
+	return 0;
+
+out_delete_threads:
+	thread_map__delete(evlist->threads);
+	return -1;
+}
+
+void perf_evlist__delete_maps(struct perf_evlist *evlist)
+{
+	cpu_map__delete(evlist->cpus);
+	thread_map__delete(evlist->threads);
+	evlist->cpus	= NULL;
+	evlist->threads = NULL;
+}
+
+int perf_evlist__set_filters(struct perf_evlist *evlist)
+{
+	const struct thread_map *threads = evlist->threads;
+	const struct cpu_map *cpus = evlist->cpus;
+	struct perf_evsel *evsel;
+	char *filter;
+	int thread;
+	int cpu;
+	int err;
+	int fd;
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		filter = evsel->filter;
+		if (!filter)
+			continue;
+		for (cpu = 0; cpu < cpus->nr; cpu++) {
+			for (thread = 0; thread < threads->nr; thread++) {
+				fd = FD(evsel, cpu, thread);
+				err = ioctl(fd, PERF_EVENT_IOC_SET_FILTER, filter);
+				if (err)
+					return err;
+			}
+		}
+	}
+
+	return 0;
+}
diff --git a/tools/perf/util/evlist.h b/tools/perf/util/evlist.h
new file mode 100644
index 000000000000..8b1cb7a4c5f1
--- /dev/null
+++ b/tools/perf/util/evlist.h
@@ -0,0 +1,68 @@
+#ifndef __PERF_EVLIST_H
+#define __PERF_EVLIST_H 1
+
+#include <linux/list.h>
+#include "../perf.h"
+#include "event.h"
+
+struct pollfd;
+struct thread_map;
+struct cpu_map;
+
+#define PERF_EVLIST__HLIST_BITS 8
+#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS)
+
+struct perf_evlist {
+	struct list_head entries;
+	struct hlist_head heads[PERF_EVLIST__HLIST_SIZE];
+	int		 nr_entries;
+	int		 nr_fds;
+	int		 mmap_len;
+	bool		 overwrite;
+	union perf_event event_copy;
+	struct perf_mmap *mmap;
+	struct pollfd	 *pollfd;
+	struct thread_map *threads;
+	struct cpu_map	  *cpus;
+};
+
+struct perf_evsel;
+
+struct perf_evlist *perf_evlist__new(struct cpu_map *cpus,
+				     struct thread_map *threads);
+void perf_evlist__init(struct perf_evlist *evlist, struct cpu_map *cpus,
+		       struct thread_map *threads);
+void perf_evlist__exit(struct perf_evlist *evlist);
+void perf_evlist__delete(struct perf_evlist *evlist);
+
+void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry);
+int perf_evlist__add_default(struct perf_evlist *evlist);
+
+void perf_evlist__id_add(struct perf_evlist *evlist, struct perf_evsel *evsel,
+			 int cpu, int thread, u64 id);
+
+int perf_evlist__alloc_pollfd(struct perf_evlist *evlist);
+void perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd);
+
+struct perf_evsel *perf_evlist__id2evsel(struct perf_evlist *evlist, u64 id);
+
+union perf_event *perf_evlist__read_on_cpu(struct perf_evlist *self, int cpu);
+
+int perf_evlist__alloc_mmap(struct perf_evlist *evlist);
+int perf_evlist__mmap(struct perf_evlist *evlist, int pages, bool overwrite);
+void perf_evlist__munmap(struct perf_evlist *evlist);
+
+static inline void perf_evlist__set_maps(struct perf_evlist *evlist,
+					 struct cpu_map *cpus,
+					 struct thread_map *threads)
+{
+	evlist->cpus	= cpus;
+	evlist->threads	= threads;
+}
+
+int perf_evlist__create_maps(struct perf_evlist *evlist, pid_t target_pid,
+			     pid_t target_tid, const char *cpu_list);
+void perf_evlist__delete_maps(struct perf_evlist *evlist);
+int perf_evlist__set_filters(struct perf_evlist *evlist);
+
+#endif /* __PERF_EVLIST_H */
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index d8575d31ee6c..662596afd7f1 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -1,20 +1,34 @@
+/*
+ * Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Parts came from builtin-{top,stat,record}.c, see those files for further
+ * copyright notes.
+ *
+ * Released under the GPL v2. (and only v2, not any later version)
+ */
+
 #include "evsel.h"
-#include "../perf.h"
+#include "evlist.h"
 #include "util.h"
 #include "cpumap.h"
-#include "thread.h"
+#include "thread_map.h"
 
 #define FD(e, x, y) (*(int *)xyarray__entry(e->fd, x, y))
 
+void perf_evsel__init(struct perf_evsel *evsel,
+		      struct perf_event_attr *attr, int idx)
+{
+	evsel->idx	   = idx;
+	evsel->attr	   = *attr;
+	INIT_LIST_HEAD(&evsel->node);
+}
+
 struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx)
 {
 	struct perf_evsel *evsel = zalloc(sizeof(*evsel));
 
-	if (evsel != NULL) {
-		evsel->idx	   = idx;
-		evsel->attr	   = *attr;
-		INIT_LIST_HEAD(&evsel->node);
-	}
+	if (evsel != NULL)
+		perf_evsel__init(evsel, attr, idx);
 
 	return evsel;
 }
@@ -25,6 +39,22 @@ int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads)
 	return evsel->fd != NULL ? 0 : -ENOMEM;
 }
 
+int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads)
+{
+	evsel->sample_id = xyarray__new(ncpus, nthreads, sizeof(struct perf_sample_id));
+	if (evsel->sample_id == NULL)
+		return -ENOMEM;
+
+	evsel->id = zalloc(ncpus * nthreads * sizeof(u64));
+	if (evsel->id == NULL) {
+		xyarray__delete(evsel->sample_id);
+		evsel->sample_id = NULL;
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
 int perf_evsel__alloc_counts(struct perf_evsel *evsel, int ncpus)
 {
 	evsel->counts = zalloc((sizeof(*evsel->counts) +
@@ -38,6 +68,14 @@ void perf_evsel__free_fd(struct perf_evsel *evsel)
 	evsel->fd = NULL;
 }
 
+void perf_evsel__free_id(struct perf_evsel *evsel)
+{
+	xyarray__delete(evsel->sample_id);
+	evsel->sample_id = NULL;
+	free(evsel->id);
+	evsel->id = NULL;
+}
+
 void perf_evsel__close_fd(struct perf_evsel *evsel, int ncpus, int nthreads)
 {
 	int cpu, thread;
@@ -49,10 +87,19 @@ void perf_evsel__close_fd(struct perf_evsel *evsel, int ncpus, int nthreads)
 		}
 }
 
-void perf_evsel__delete(struct perf_evsel *evsel)
+void perf_evsel__exit(struct perf_evsel *evsel)
 {
 	assert(list_empty(&evsel->node));
 	xyarray__delete(evsel->fd);
+	xyarray__delete(evsel->sample_id);
+	free(evsel->id);
+}
+
+void perf_evsel__delete(struct perf_evsel *evsel)
+{
+	perf_evsel__exit(evsel);
+	close_cgroup(evsel->cgrp);
+	free(evsel->name);
 	free(evsel);
 }
 
@@ -128,21 +175,51 @@ int __perf_evsel__read(struct perf_evsel *evsel,
 }
 
 static int __perf_evsel__open(struct perf_evsel *evsel, struct cpu_map *cpus,
-			      struct thread_map *threads)
+			      struct thread_map *threads, bool group, bool inherit)
 {
 	int cpu, thread;
+	unsigned long flags = 0;
+	int pid = -1;
 
 	if (evsel->fd == NULL &&
 	    perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0)
 		return -1;
 
+	if (evsel->cgrp) {
+		flags = PERF_FLAG_PID_CGROUP;
+		pid = evsel->cgrp->fd;
+	}
+
 	for (cpu = 0; cpu < cpus->nr; cpu++) {
+		int group_fd = -1;
+		/*
+		 * Don't allow mmap() of inherited per-task counters. This
+		 * would create a performance issue due to all children writing
+		 * to the same buffer.
+		 *
+		 * FIXME:
+		 * Proper fix is not to pass 'inherit' to perf_evsel__open*,
+		 * but a 'flags' parameter, with 'group' folded there as well,
+		 * then introduce a PERF_O_{MMAP,GROUP,INHERIT} enum, and if
+		 * O_MMAP is set, emit a warning if cpu < 0 and O_INHERIT is
+		 * set. Lets go for the minimal fix first tho.
+		 */
+		evsel->attr.inherit = (cpus->map[cpu] >= 0) && inherit;
+
 		for (thread = 0; thread < threads->nr; thread++) {
+
+			if (!evsel->cgrp)
+				pid = threads->map[thread];
+
 			FD(evsel, cpu, thread) = sys_perf_event_open(&evsel->attr,
-								     threads->map[thread],
-								     cpus->map[cpu], -1, 0);
+								     pid,
+								     cpus->map[cpu],
+								     group_fd, flags);
 			if (FD(evsel, cpu, thread) < 0)
 				goto out_close;
+
+			if (group && group_fd == -1)
+				group_fd = FD(evsel, cpu, thread);
 		}
 	}
 
@@ -175,10 +252,9 @@ static struct {
 	.threads = { -1, },
 };
 
-int perf_evsel__open(struct perf_evsel *evsel,
-		     struct cpu_map *cpus, struct thread_map *threads)
+int perf_evsel__open(struct perf_evsel *evsel, struct cpu_map *cpus,
+		     struct thread_map *threads, bool group, bool inherit)
 {
-
 	if (cpus == NULL) {
 		/* Work around old compiler warnings about strict aliasing */
 		cpus = &empty_cpu_map.map;
@@ -187,15 +263,135 @@ int perf_evsel__open(struct perf_evsel *evsel,
 	if (threads == NULL)
 		threads = &empty_thread_map.map;
 
-	return __perf_evsel__open(evsel, cpus, threads);
+	return __perf_evsel__open(evsel, cpus, threads, group, inherit);
 }
 
-int perf_evsel__open_per_cpu(struct perf_evsel *evsel, struct cpu_map *cpus)
+int perf_evsel__open_per_cpu(struct perf_evsel *evsel,
+			     struct cpu_map *cpus, bool group, bool inherit)
 {
-	return __perf_evsel__open(evsel, cpus, &empty_thread_map.map);
+	return __perf_evsel__open(evsel, cpus, &empty_thread_map.map, group, inherit);
+}
+
+int perf_evsel__open_per_thread(struct perf_evsel *evsel,
+				struct thread_map *threads, bool group, bool inherit)
+{
+	return __perf_evsel__open(evsel, &empty_cpu_map.map, threads, group, inherit);
+}
+
+static int perf_event__parse_id_sample(const union perf_event *event, u64 type,
+				       struct perf_sample *sample)
+{
+	const u64 *array = event->sample.array;
+
+	array += ((event->header.size -
+		   sizeof(event->header)) / sizeof(u64)) - 1;
+
+	if (type & PERF_SAMPLE_CPU) {
+		u32 *p = (u32 *)array;
+		sample->cpu = *p;
+		array--;
+	}
+
+	if (type & PERF_SAMPLE_STREAM_ID) {
+		sample->stream_id = *array;
+		array--;
+	}
+
+	if (type & PERF_SAMPLE_ID) {
+		sample->id = *array;
+		array--;
+	}
+
+	if (type & PERF_SAMPLE_TIME) {
+		sample->time = *array;
+		array--;
+	}
+
+	if (type & PERF_SAMPLE_TID) {
+		u32 *p = (u32 *)array;
+		sample->pid = p[0];
+		sample->tid = p[1];
+	}
+
+	return 0;
 }
 
-int perf_evsel__open_per_thread(struct perf_evsel *evsel, struct thread_map *threads)
+int perf_event__parse_sample(const union perf_event *event, u64 type,
+			     bool sample_id_all, struct perf_sample *data)
 {
-	return __perf_evsel__open(evsel, &empty_cpu_map.map, threads);
+	const u64 *array;
+
+	data->cpu = data->pid = data->tid = -1;
+	data->stream_id = data->id = data->time = -1ULL;
+
+	if (event->header.type != PERF_RECORD_SAMPLE) {
+		if (!sample_id_all)
+			return 0;
+		return perf_event__parse_id_sample(event, type, data);
+	}
+
+	array = event->sample.array;
+
+	if (type & PERF_SAMPLE_IP) {
+		data->ip = event->ip.ip;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_TID) {
+		u32 *p = (u32 *)array;
+		data->pid = p[0];
+		data->tid = p[1];
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_TIME) {
+		data->time = *array;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_ADDR) {
+		data->addr = *array;
+		array++;
+	}
+
+	data->id = -1ULL;
+	if (type & PERF_SAMPLE_ID) {
+		data->id = *array;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_STREAM_ID) {
+		data->stream_id = *array;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_CPU) {
+		u32 *p = (u32 *)array;
+		data->cpu = *p;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_PERIOD) {
+		data->period = *array;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_READ) {
+		fprintf(stderr, "PERF_SAMPLE_READ is unsuported for now\n");
+		return -1;
+	}
+
+	if (type & PERF_SAMPLE_CALLCHAIN) {
+		data->callchain = (struct ip_callchain *)array;
+		array += 1 + data->callchain->nr;
+	}
+
+	if (type & PERF_SAMPLE_RAW) {
+		u32 *p = (u32 *)array;
+		data->raw_size = *p;
+		p++;
+		data->raw_data = p;
+	}
+
+	return 0;
 }
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index b2d755fe88a5..6710ab538342 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -6,6 +6,8 @@
 #include "../../../include/linux/perf_event.h"
 #include "types.h"
 #include "xyarray.h"
+#include "cgroup.h"
+#include "hist.h"
  
 struct perf_counts_values {
 	union {
@@ -24,31 +26,66 @@ struct perf_counts {
 	struct perf_counts_values cpu[];
 };
 
+struct perf_evsel;
+
+/*
+ * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are
+ * more than one entry in the evlist.
+ */
+struct perf_sample_id {
+	struct hlist_node 	node;
+	u64		 	id;
+	struct perf_evsel	*evsel;
+};
+
+/** struct perf_evsel - event selector
+ *
+ * @name - Can be set to retain the original event name passed by the user,
+ *         so that when showing results in tools such as 'perf stat', we
+ *         show the name used, not some alias.
+ */
 struct perf_evsel {
 	struct list_head	node;
 	struct perf_event_attr	attr;
 	char			*filter;
 	struct xyarray		*fd;
+	struct xyarray		*sample_id;
+	u64			*id;
 	struct perf_counts	*counts;
 	int			idx;
-	void			*priv;
+	int			ids;
+	struct hists		hists;
+	char			*name;
+	union {
+		void		*priv;
+		off_t		id_offset;
+	};
+	struct cgroup_sel	*cgrp;
 };
 
 struct cpu_map;
 struct thread_map;
+struct perf_evlist;
 
 struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx);
+void perf_evsel__init(struct perf_evsel *evsel,
+		      struct perf_event_attr *attr, int idx);
+void perf_evsel__exit(struct perf_evsel *evsel);
 void perf_evsel__delete(struct perf_evsel *evsel);
 
 int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
+int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads);
 int perf_evsel__alloc_counts(struct perf_evsel *evsel, int ncpus);
 void perf_evsel__free_fd(struct perf_evsel *evsel);
+void perf_evsel__free_id(struct perf_evsel *evsel);
 void perf_evsel__close_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
 
-int perf_evsel__open_per_cpu(struct perf_evsel *evsel, struct cpu_map *cpus);
-int perf_evsel__open_per_thread(struct perf_evsel *evsel, struct thread_map *threads);
-int perf_evsel__open(struct perf_evsel *evsel, 
-		     struct cpu_map *cpus, struct thread_map *threads);
+int perf_evsel__open_per_cpu(struct perf_evsel *evsel,
+			     struct cpu_map *cpus, bool group, bool inherit);
+int perf_evsel__open_per_thread(struct perf_evsel *evsel,
+				struct thread_map *threads, bool group, bool inherit);
+int perf_evsel__open(struct perf_evsel *evsel, struct cpu_map *cpus,
+		     struct thread_map *threads, bool group, bool inherit);
 
 #define perf_evsel__match(evsel, t, c)		\
 	(evsel->attr.type == PERF_TYPE_##t &&	\
diff --git a/tools/perf/util/exec_cmd.c b/tools/perf/util/exec_cmd.c
index 67eeff571568..7adf4ad15d8f 100644
--- a/tools/perf/util/exec_cmd.c
+++ b/tools/perf/util/exec_cmd.c
@@ -11,31 +11,12 @@ static const char *argv0_path;
 
 const char *system_path(const char *path)
 {
-#ifdef RUNTIME_PREFIX
-	static const char *prefix;
-#else
 	static const char *prefix = PREFIX;
-#endif
 	struct strbuf d = STRBUF_INIT;
 
 	if (is_absolute_path(path))
 		return path;
 
-#ifdef RUNTIME_PREFIX
-	assert(argv0_path);
-	assert(is_absolute_path(argv0_path));
-
-	if (!prefix &&
-	    !(prefix = strip_path_suffix(argv0_path, PERF_EXEC_PATH)) &&
-	    !(prefix = strip_path_suffix(argv0_path, BINDIR)) &&
-	    !(prefix = strip_path_suffix(argv0_path, "perf"))) {
-		prefix = PREFIX;
-		fprintf(stderr, "RUNTIME_PREFIX requested, "
-				"but prefix computation failed.  "
-				"Using static fallback '%s'.\n", prefix);
-	}
-#endif
-
 	strbuf_addf(&d, "%s/%s", prefix, path);
 	path = strbuf_detach(&d, NULL);
 	return path;
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 0866bcdb5e8e..e5230c0ef95b 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -8,6 +8,8 @@
 #include <linux/list.h>
 #include <linux/kernel.h>
 
+#include "evlist.h"
+#include "evsel.h"
 #include "util.h"
 #include "header.h"
 #include "../perf.h"
@@ -18,89 +20,6 @@
 
 static bool no_buildid_cache = false;
 
-/*
- * Create new perf.data header attribute:
- */
-struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr)
-{
-	struct perf_header_attr *self = malloc(sizeof(*self));
-
-	if (self != NULL) {
-		self->attr = *attr;
-		self->ids  = 0;
-		self->size = 1;
-		self->id   = malloc(sizeof(u64));
-		if (self->id == NULL) {
-			free(self);
-			self = NULL;
-		}
-	}
-
-	return self;
-}
-
-void perf_header_attr__delete(struct perf_header_attr *self)
-{
-	free(self->id);
-	free(self);
-}
-
-int perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
-{
-	int pos = self->ids;
-
-	self->ids++;
-	if (self->ids > self->size) {
-		int nsize = self->size * 2;
-		u64 *nid = realloc(self->id, nsize * sizeof(u64));
-
-		if (nid == NULL)
-			return -1;
-
-		self->size = nsize;
-		self->id = nid;
-	}
-	self->id[pos] = id;
-	return 0;
-}
-
-int perf_header__init(struct perf_header *self)
-{
-	self->size = 1;
-	self->attr = malloc(sizeof(void *));
-	return self->attr == NULL ? -ENOMEM : 0;
-}
-
-void perf_header__exit(struct perf_header *self)
-{
-	int i;
-	for (i = 0; i < self->attrs; ++i)
-                perf_header_attr__delete(self->attr[i]);
-	free(self->attr);
-}
-
-int perf_header__add_attr(struct perf_header *self,
-			  struct perf_header_attr *attr)
-{
-	if (self->frozen)
-		return -1;
-
-	if (self->attrs == self->size) {
-		int nsize = self->size * 2;
-		struct perf_header_attr **nattr;
-
-		nattr = realloc(self->attr, nsize * sizeof(void *));
-		if (nattr == NULL)
-			return -1;
-
-		self->size = nsize;
-		self->attr = nattr;
-	}
-
-	self->attr[self->attrs++] = attr;
-	return 0;
-}
-
 static int event_count;
 static struct perf_trace_event_type *events;
 
@@ -147,19 +66,19 @@ struct perf_file_attr {
 	struct perf_file_section	ids;
 };
 
-void perf_header__set_feat(struct perf_header *self, int feat)
+void perf_header__set_feat(struct perf_header *header, int feat)
 {
-	set_bit(feat, self->adds_features);
+	set_bit(feat, header->adds_features);
 }
 
-void perf_header__clear_feat(struct perf_header *self, int feat)
+void perf_header__clear_feat(struct perf_header *header, int feat)
 {
-	clear_bit(feat, self->adds_features);
+	clear_bit(feat, header->adds_features);
 }
 
-bool perf_header__has_feat(const struct perf_header *self, int feat)
+bool perf_header__has_feat(const struct perf_header *header, int feat)
 {
-	return test_bit(feat, self->adds_features);
+	return test_bit(feat, header->adds_features);
 }
 
 static int do_write(int fd, const void *buf, size_t size)
@@ -228,22 +147,22 @@ static int __dsos__write_buildid_table(struct list_head *head, pid_t pid,
 	return 0;
 }
 
-static int machine__write_buildid_table(struct machine *self, int fd)
+static int machine__write_buildid_table(struct machine *machine, int fd)
 {
 	int err;
 	u16 kmisc = PERF_RECORD_MISC_KERNEL,
 	    umisc = PERF_RECORD_MISC_USER;
 
-	if (!machine__is_host(self)) {
+	if (!machine__is_host(machine)) {
 		kmisc = PERF_RECORD_MISC_GUEST_KERNEL;
 		umisc = PERF_RECORD_MISC_GUEST_USER;
 	}
 
-	err = __dsos__write_buildid_table(&self->kernel_dsos, self->pid,
+	err = __dsos__write_buildid_table(&machine->kernel_dsos, machine->pid,
 					  kmisc, fd);
 	if (err == 0)
-		err = __dsos__write_buildid_table(&self->user_dsos,
-						  self->pid, umisc, fd);
+		err = __dsos__write_buildid_table(&machine->user_dsos,
+						  machine->pid, umisc, fd);
 	return err;
 }
 
@@ -366,12 +285,12 @@ out_free:
 	return err;
 }
 
-static int dso__cache_build_id(struct dso *self, const char *debugdir)
+static int dso__cache_build_id(struct dso *dso, const char *debugdir)
 {
-	bool is_kallsyms = self->kernel && self->long_name[0] != '/';
+	bool is_kallsyms = dso->kernel && dso->long_name[0] != '/';
 
-	return build_id_cache__add_b(self->build_id, sizeof(self->build_id),
-				     self->long_name, debugdir, is_kallsyms);
+	return build_id_cache__add_b(dso->build_id, sizeof(dso->build_id),
+				     dso->long_name, debugdir, is_kallsyms);
 }
 
 static int __dsos__cache_build_ids(struct list_head *head, const char *debugdir)
@@ -386,14 +305,14 @@ static int __dsos__cache_build_ids(struct list_head *head, const char *debugdir)
 	return err;
 }
 
-static int machine__cache_build_ids(struct machine *self, const char *debugdir)
+static int machine__cache_build_ids(struct machine *machine, const char *debugdir)
 {
-	int ret = __dsos__cache_build_ids(&self->kernel_dsos, debugdir);
-	ret |= __dsos__cache_build_ids(&self->user_dsos, debugdir);
+	int ret = __dsos__cache_build_ids(&machine->kernel_dsos, debugdir);
+	ret |= __dsos__cache_build_ids(&machine->user_dsos, debugdir);
 	return ret;
 }
 
-static int perf_session__cache_build_ids(struct perf_session *self)
+static int perf_session__cache_build_ids(struct perf_session *session)
 {
 	struct rb_node *nd;
 	int ret;
@@ -404,28 +323,28 @@ static int perf_session__cache_build_ids(struct perf_session *self)
 	if (mkdir(debugdir, 0755) != 0 && errno != EEXIST)
 		return -1;
 
-	ret = machine__cache_build_ids(&self->host_machine, debugdir);
+	ret = machine__cache_build_ids(&session->host_machine, debugdir);
 
-	for (nd = rb_first(&self->machines); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret |= machine__cache_build_ids(pos, debugdir);
 	}
 	return ret ? -1 : 0;
 }
 
-static bool machine__read_build_ids(struct machine *self, bool with_hits)
+static bool machine__read_build_ids(struct machine *machine, bool with_hits)
 {
-	bool ret = __dsos__read_build_ids(&self->kernel_dsos, with_hits);
-	ret |= __dsos__read_build_ids(&self->user_dsos, with_hits);
+	bool ret = __dsos__read_build_ids(&machine->kernel_dsos, with_hits);
+	ret |= __dsos__read_build_ids(&machine->user_dsos, with_hits);
 	return ret;
 }
 
-static bool perf_session__read_build_ids(struct perf_session *self, bool with_hits)
+static bool perf_session__read_build_ids(struct perf_session *session, bool with_hits)
 {
 	struct rb_node *nd;
-	bool ret = machine__read_build_ids(&self->host_machine, with_hits);
+	bool ret = machine__read_build_ids(&session->host_machine, with_hits);
 
-	for (nd = rb_first(&self->machines); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret |= machine__read_build_ids(pos, with_hits);
 	}
@@ -433,7 +352,8 @@ static bool perf_session__read_build_ids(struct perf_session *self, bool with_hi
 	return ret;
 }
 
-static int perf_header__adds_write(struct perf_header *self, int fd)
+static int perf_header__adds_write(struct perf_header *header,
+				   struct perf_evlist *evlist, int fd)
 {
 	int nr_sections;
 	struct perf_session *session;
@@ -442,13 +362,13 @@ static int perf_header__adds_write(struct perf_header *self, int fd)
 	u64 sec_start;
 	int idx = 0, err;
 
-	session = container_of(self, struct perf_session, header);
+	session = container_of(header, struct perf_session, header);
 
-	if (perf_header__has_feat(self, HEADER_BUILD_ID &&
+	if (perf_header__has_feat(header, HEADER_BUILD_ID &&
 	    !perf_session__read_build_ids(session, true)))
-		perf_header__clear_feat(self, HEADER_BUILD_ID);
+		perf_header__clear_feat(header, HEADER_BUILD_ID);
 
-	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
 	if (!nr_sections)
 		return 0;
 
@@ -458,28 +378,28 @@ static int perf_header__adds_write(struct perf_header *self, int fd)
 
 	sec_size = sizeof(*feat_sec) * nr_sections;
 
-	sec_start = self->data_offset + self->data_size;
+	sec_start = header->data_offset + header->data_size;
 	lseek(fd, sec_start + sec_size, SEEK_SET);
 
-	if (perf_header__has_feat(self, HEADER_TRACE_INFO)) {
+	if (perf_header__has_feat(header, HEADER_TRACE_INFO)) {
 		struct perf_file_section *trace_sec;
 
 		trace_sec = &feat_sec[idx++];
 
 		/* Write trace info */
 		trace_sec->offset = lseek(fd, 0, SEEK_CUR);
-		read_tracing_data(fd, &evsel_list);
+		read_tracing_data(fd, &evlist->entries);
 		trace_sec->size = lseek(fd, 0, SEEK_CUR) - trace_sec->offset;
 	}
 
-	if (perf_header__has_feat(self, HEADER_BUILD_ID)) {
+	if (perf_header__has_feat(header, HEADER_BUILD_ID)) {
 		struct perf_file_section *buildid_sec;
 
 		buildid_sec = &feat_sec[idx++];
 
 		/* Write build-ids */
 		buildid_sec->offset = lseek(fd, 0, SEEK_CUR);
-		err = dsos__write_buildid_table(self, fd);
+		err = dsos__write_buildid_table(header, fd);
 		if (err < 0) {
 			pr_debug("failed to write buildid table\n");
 			goto out_free;
@@ -518,32 +438,41 @@ int perf_header__write_pipe(int fd)
 	return 0;
 }
 
-int perf_header__write(struct perf_header *self, int fd, bool at_exit)
+int perf_session__write_header(struct perf_session *session,
+			       struct perf_evlist *evlist,
+			       int fd, bool at_exit)
 {
 	struct perf_file_header f_header;
 	struct perf_file_attr   f_attr;
-	struct perf_header_attr	*attr;
-	int i, err;
+	struct perf_header *header = &session->header;
+	struct perf_evsel *attr, *pair = NULL;
+	int err;
 
 	lseek(fd, sizeof(f_header), SEEK_SET);
 
-	for (i = 0; i < self->attrs; i++) {
-		attr = self->attr[i];
+	if (session->evlist != evlist)
+		pair = list_entry(session->evlist->entries.next, struct perf_evsel, node);
 
+	list_for_each_entry(attr, &evlist->entries, node) {
 		attr->id_offset = lseek(fd, 0, SEEK_CUR);
 		err = do_write(fd, attr->id, attr->ids * sizeof(u64));
 		if (err < 0) {
+out_err_write:
 			pr_debug("failed to write perf header\n");
 			return err;
 		}
+		if (session->evlist != evlist) {
+			err = do_write(fd, pair->id, pair->ids * sizeof(u64));
+			if (err < 0)
+				goto out_err_write;
+			attr->ids += pair->ids;
+			pair = list_entry(pair->node.next, struct perf_evsel, node);
+		}
 	}
 
+	header->attr_offset = lseek(fd, 0, SEEK_CUR);
 
-	self->attr_offset = lseek(fd, 0, SEEK_CUR);
-
-	for (i = 0; i < self->attrs; i++) {
-		attr = self->attr[i];
-
+	list_for_each_entry(attr, &evlist->entries, node) {
 		f_attr = (struct perf_file_attr){
 			.attr = attr->attr,
 			.ids  = {
@@ -558,20 +487,20 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit)
 		}
 	}
 
-	self->event_offset = lseek(fd, 0, SEEK_CUR);
-	self->event_size = event_count * sizeof(struct perf_trace_event_type);
+	header->event_offset = lseek(fd, 0, SEEK_CUR);
+	header->event_size = event_count * sizeof(struct perf_trace_event_type);
 	if (events) {
-		err = do_write(fd, events, self->event_size);
+		err = do_write(fd, events, header->event_size);
 		if (err < 0) {
 			pr_debug("failed to write perf header events\n");
 			return err;
 		}
 	}
 
-	self->data_offset = lseek(fd, 0, SEEK_CUR);
+	header->data_offset = lseek(fd, 0, SEEK_CUR);
 
 	if (at_exit) {
-		err = perf_header__adds_write(self, fd);
+		err = perf_header__adds_write(header, evlist, fd);
 		if (err < 0)
 			return err;
 	}
@@ -581,20 +510,20 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit)
 		.size	   = sizeof(f_header),
 		.attr_size = sizeof(f_attr),
 		.attrs = {
-			.offset = self->attr_offset,
-			.size   = self->attrs * sizeof(f_attr),
+			.offset = header->attr_offset,
+			.size   = evlist->nr_entries * sizeof(f_attr),
 		},
 		.data = {
-			.offset = self->data_offset,
-			.size	= self->data_size,
+			.offset = header->data_offset,
+			.size	= header->data_size,
 		},
 		.event_types = {
-			.offset = self->event_offset,
-			.size	= self->event_size,
+			.offset = header->event_offset,
+			.size	= header->event_size,
 		},
 	};
 
-	memcpy(&f_header.adds_features, &self->adds_features, sizeof(self->adds_features));
+	memcpy(&f_header.adds_features, &header->adds_features, sizeof(header->adds_features));
 
 	lseek(fd, 0, SEEK_SET);
 	err = do_write(fd, &f_header, sizeof(f_header));
@@ -602,26 +531,26 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit)
 		pr_debug("failed to write perf header\n");
 		return err;
 	}
-	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+	lseek(fd, header->data_offset + header->data_size, SEEK_SET);
 
-	self->frozen = 1;
+	header->frozen = 1;
 	return 0;
 }
 
-static int perf_header__getbuffer64(struct perf_header *self,
+static int perf_header__getbuffer64(struct perf_header *header,
 				    int fd, void *buf, size_t size)
 {
 	if (readn(fd, buf, size) <= 0)
 		return -1;
 
-	if (self->needs_swap)
+	if (header->needs_swap)
 		mem_bswap_64(buf, size);
 
 	return 0;
 }
 
-int perf_header__process_sections(struct perf_header *self, int fd,
-				  int (*process)(struct perf_file_section *self,
+int perf_header__process_sections(struct perf_header *header, int fd,
+				  int (*process)(struct perf_file_section *section,
 						 struct perf_header *ph,
 						 int feat, int fd))
 {
@@ -631,7 +560,7 @@ int perf_header__process_sections(struct perf_header *self, int fd,
 	int idx = 0;
 	int err = -1, feat = 1;
 
-	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
 	if (!nr_sections)
 		return 0;
 
@@ -641,17 +570,17 @@ int perf_header__process_sections(struct perf_header *self, int fd,
 
 	sec_size = sizeof(*feat_sec) * nr_sections;
 
-	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+	lseek(fd, header->data_offset + header->data_size, SEEK_SET);
 
-	if (perf_header__getbuffer64(self, fd, feat_sec, sec_size))
+	if (perf_header__getbuffer64(header, fd, feat_sec, sec_size))
 		goto out_free;
 
 	err = 0;
 	while (idx < nr_sections && feat < HEADER_LAST_FEATURE) {
-		if (perf_header__has_feat(self, feat)) {
+		if (perf_header__has_feat(header, feat)) {
 			struct perf_file_section *sec = &feat_sec[idx++];
 
-			err = process(sec, self, feat, fd);
+			err = process(sec, header, feat, fd);
 			if (err < 0)
 				break;
 		}
@@ -662,35 +591,35 @@ out_free:
 	return err;
 }
 
-int perf_file_header__read(struct perf_file_header *self,
+int perf_file_header__read(struct perf_file_header *header,
 			   struct perf_header *ph, int fd)
 {
 	lseek(fd, 0, SEEK_SET);
 
-	if (readn(fd, self, sizeof(*self)) <= 0 ||
-	    memcmp(&self->magic, __perf_magic, sizeof(self->magic)))
+	if (readn(fd, header, sizeof(*header)) <= 0 ||
+	    memcmp(&header->magic, __perf_magic, sizeof(header->magic)))
 		return -1;
 
-	if (self->attr_size != sizeof(struct perf_file_attr)) {
-		u64 attr_size = bswap_64(self->attr_size);
+	if (header->attr_size != sizeof(struct perf_file_attr)) {
+		u64 attr_size = bswap_64(header->attr_size);
 
 		if (attr_size != sizeof(struct perf_file_attr))
 			return -1;
 
-		mem_bswap_64(self, offsetof(struct perf_file_header,
+		mem_bswap_64(header, offsetof(struct perf_file_header,
 					    adds_features));
 		ph->needs_swap = true;
 	}
 
-	if (self->size != sizeof(*self)) {
+	if (header->size != sizeof(*header)) {
 		/* Support the previous format */
-		if (self->size == offsetof(typeof(*self), adds_features))
-			bitmap_zero(self->adds_features, HEADER_FEAT_BITS);
+		if (header->size == offsetof(typeof(*header), adds_features))
+			bitmap_zero(header->adds_features, HEADER_FEAT_BITS);
 		else
 			return -1;
 	}
 
-	memcpy(&ph->adds_features, &self->adds_features,
+	memcpy(&ph->adds_features, &header->adds_features,
 	       sizeof(ph->adds_features));
 	/*
 	 * FIXME: hack that assumes that if we need swap the perf.data file
@@ -704,10 +633,10 @@ int perf_file_header__read(struct perf_file_header *self,
 		perf_header__set_feat(ph, HEADER_BUILD_ID);
 	}
 
-	ph->event_offset = self->event_types.offset;
-	ph->event_size   = self->event_types.size;
-	ph->data_offset  = self->data.offset;
-	ph->data_size	 = self->data.size;
+	ph->event_offset = header->event_types.offset;
+	ph->event_size   = header->event_types.size;
+	ph->data_offset  = header->data.offset;
+	ph->data_size	 = header->data.size;
 	return 0;
 }
 
@@ -766,11 +695,10 @@ out:
 	return err;
 }
 
-static int perf_header__read_build_ids(struct perf_header *self,
-			int input, u64 offset, u64 size)
+static int perf_header__read_build_ids(struct perf_header *header,
+				       int input, u64 offset, u64 size)
 {
-	struct perf_session *session = container_of(self,
-			struct perf_session, header);
+	struct perf_session *session = container_of(header, struct perf_session, header);
 	struct build_id_event bev;
 	char filename[PATH_MAX];
 	u64 limit = offset + size;
@@ -782,7 +710,7 @@ static int perf_header__read_build_ids(struct perf_header *self,
 		if (read(input, &bev, sizeof(bev)) != sizeof(bev))
 			goto out;
 
-		if (self->needs_swap)
+		if (header->needs_swap)
 			perf_event_header__bswap(&bev.header);
 
 		len = bev.header.size - sizeof(bev);
@@ -798,13 +726,13 @@ out:
 	return err;
 }
 
-static int perf_file_section__process(struct perf_file_section *self,
+static int perf_file_section__process(struct perf_file_section *section,
 				      struct perf_header *ph,
 				      int feat, int fd)
 {
-	if (lseek(fd, self->offset, SEEK_SET) == (off_t)-1) {
+	if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) {
 		pr_debug("Failed to lseek to %" PRIu64 " offset for feature "
-			  "%d, continuing...\n", self->offset, feat);
+			  "%d, continuing...\n", section->offset, feat);
 		return 0;
 	}
 
@@ -814,7 +742,7 @@ static int perf_file_section__process(struct perf_file_section *self,
 		break;
 
 	case HEADER_BUILD_ID:
-		if (perf_header__read_build_ids(ph, fd, self->offset, self->size))
+		if (perf_header__read_build_ids(ph, fd, section->offset, section->size))
 			pr_debug("Failed to read buildids, continuing...\n");
 		break;
 	default:
@@ -824,21 +752,21 @@ static int perf_file_section__process(struct perf_file_section *self,
 	return 0;
 }
 
-static int perf_file_header__read_pipe(struct perf_pipe_file_header *self,
+static int perf_file_header__read_pipe(struct perf_pipe_file_header *header,
 				       struct perf_header *ph, int fd,
 				       bool repipe)
 {
-	if (readn(fd, self, sizeof(*self)) <= 0 ||
-	    memcmp(&self->magic, __perf_magic, sizeof(self->magic)))
+	if (readn(fd, header, sizeof(*header)) <= 0 ||
+	    memcmp(&header->magic, __perf_magic, sizeof(header->magic)))
 		return -1;
 
-	if (repipe && do_write(STDOUT_FILENO, self, sizeof(*self)) < 0)
+	if (repipe && do_write(STDOUT_FILENO, header, sizeof(*header)) < 0)
 		return -1;
 
-	if (self->size != sizeof(*self)) {
-		u64 size = bswap_64(self->size);
+	if (header->size != sizeof(*header)) {
+		u64 size = bswap_64(header->size);
 
-		if (size != sizeof(*self))
+		if (size != sizeof(*header))
 			return -1;
 
 		ph->needs_swap = true;
@@ -849,10 +777,10 @@ static int perf_file_header__read_pipe(struct perf_pipe_file_header *self,
 
 static int perf_header__read_pipe(struct perf_session *session, int fd)
 {
-	struct perf_header *self = &session->header;
+	struct perf_header *header = &session->header;
 	struct perf_pipe_file_header f_header;
 
-	if (perf_file_header__read_pipe(&f_header, self, fd,
+	if (perf_file_header__read_pipe(&f_header, header, fd,
 					session->repipe) < 0) {
 		pr_debug("incompatible file format\n");
 		return -EINVAL;
@@ -863,18 +791,22 @@ static int perf_header__read_pipe(struct perf_session *session, int fd)
 	return 0;
 }
 
-int perf_header__read(struct perf_session *session, int fd)
+int perf_session__read_header(struct perf_session *session, int fd)
 {
-	struct perf_header *self = &session->header;
+	struct perf_header *header = &session->header;
 	struct perf_file_header	f_header;
 	struct perf_file_attr	f_attr;
 	u64			f_id;
 	int nr_attrs, nr_ids, i, j;
 
+	session->evlist = perf_evlist__new(NULL, NULL);
+	if (session->evlist == NULL)
+		return -ENOMEM;
+
 	if (session->fd_pipe)
 		return perf_header__read_pipe(session, fd);
 
-	if (perf_file_header__read(&f_header, self, fd) < 0) {
+	if (perf_file_header__read(&f_header, header, fd) < 0) {
 		pr_debug("incompatible file format\n");
 		return -EINVAL;
 	}
@@ -883,33 +815,39 @@ int perf_header__read(struct perf_session *session, int fd)
 	lseek(fd, f_header.attrs.offset, SEEK_SET);
 
 	for (i = 0; i < nr_attrs; i++) {
-		struct perf_header_attr *attr;
+		struct perf_evsel *evsel;
 		off_t tmp;
 
-		if (perf_header__getbuffer64(self, fd, &f_attr, sizeof(f_attr)))
+		if (perf_header__getbuffer64(header, fd, &f_attr, sizeof(f_attr)))
 			goto out_errno;
 
 		tmp = lseek(fd, 0, SEEK_CUR);
+		evsel = perf_evsel__new(&f_attr.attr, i);
 
-		attr = perf_header_attr__new(&f_attr.attr);
-		if (attr == NULL)
-			 return -ENOMEM;
+		if (evsel == NULL)
+			goto out_delete_evlist;
+		/*
+		 * Do it before so that if perf_evsel__alloc_id fails, this
+		 * entry gets purged too at perf_evlist__delete().
+		 */
+		perf_evlist__add(session->evlist, evsel);
 
 		nr_ids = f_attr.ids.size / sizeof(u64);
+		/*
+		 * We don't have the cpu and thread maps on the header, so
+		 * for allocating the perf_sample_id table we fake 1 cpu and
+		 * hattr->ids threads.
+		 */
+		if (perf_evsel__alloc_id(evsel, 1, nr_ids))
+			goto out_delete_evlist;
+
 		lseek(fd, f_attr.ids.offset, SEEK_SET);
 
 		for (j = 0; j < nr_ids; j++) {
-			if (perf_header__getbuffer64(self, fd, &f_id, sizeof(f_id)))
+			if (perf_header__getbuffer64(header, fd, &f_id, sizeof(f_id)))
 				goto out_errno;
 
-			if (perf_header_attr__add_id(attr, f_id) < 0) {
-				perf_header_attr__delete(attr);
-				return -ENOMEM;
-			}
-		}
-		if (perf_header__add_attr(self, attr) < 0) {
-			perf_header_attr__delete(attr);
-			return -ENOMEM;
+			perf_evlist__id_add(session->evlist, evsel, 0, j, f_id);
 		}
 
 		lseek(fd, tmp, SEEK_SET);
@@ -920,93 +858,63 @@ int perf_header__read(struct perf_session *session, int fd)
 		events = malloc(f_header.event_types.size);
 		if (events == NULL)
 			return -ENOMEM;
-		if (perf_header__getbuffer64(self, fd, events,
+		if (perf_header__getbuffer64(header, fd, events,
 					     f_header.event_types.size))
 			goto out_errno;
 		event_count =  f_header.event_types.size / sizeof(struct perf_trace_event_type);
 	}
 
-	perf_header__process_sections(self, fd, perf_file_section__process);
+	perf_header__process_sections(header, fd, perf_file_section__process);
 
-	lseek(fd, self->data_offset, SEEK_SET);
+	lseek(fd, header->data_offset, SEEK_SET);
 
-	self->frozen = 1;
+	header->frozen = 1;
 	return 0;
 out_errno:
 	return -errno;
+
+out_delete_evlist:
+	perf_evlist__delete(session->evlist);
+	session->evlist = NULL;
+	return -ENOMEM;
 }
 
-u64 perf_header__sample_type(struct perf_header *header)
+u64 perf_evlist__sample_type(struct perf_evlist *evlist)
 {
+	struct perf_evsel *pos;
 	u64 type = 0;
-	int i;
-
-	for (i = 0; i < header->attrs; i++) {
-		struct perf_header_attr *attr = header->attr[i];
 
+	list_for_each_entry(pos, &evlist->entries, node) {
 		if (!type)
-			type = attr->attr.sample_type;
-		else if (type != attr->attr.sample_type)
+			type = pos->attr.sample_type;
+		else if (type != pos->attr.sample_type)
 			die("non matching sample_type");
 	}
 
 	return type;
 }
 
-bool perf_header__sample_id_all(const struct perf_header *header)
+bool perf_evlist__sample_id_all(const struct perf_evlist *evlist)
 {
 	bool value = false, first = true;
-	int i;
-
-	for (i = 0; i < header->attrs; i++) {
-		struct perf_header_attr *attr = header->attr[i];
+	struct perf_evsel *pos;
 
+	list_for_each_entry(pos, &evlist->entries, node) {
 		if (first) {
-			value = attr->attr.sample_id_all;
+			value = pos->attr.sample_id_all;
 			first = false;
-		} else if (value != attr->attr.sample_id_all)
+		} else if (value != pos->attr.sample_id_all)
 			die("non matching sample_id_all");
 	}
 
 	return value;
 }
 
-struct perf_event_attr *
-perf_header__find_attr(u64 id, struct perf_header *header)
-{
-	int i;
-
-	/*
-	 * We set id to -1 if the data file doesn't contain sample
-	 * ids. This can happen when the data file contains one type
-	 * of event and in that case, the header can still store the
-	 * event attribute information. Check for this and avoid
-	 * walking through the entire list of ids which may be large.
-	 */
-	if (id == -1ULL) {
-		if (header->attrs > 0)
-			return &header->attr[0]->attr;
-		return NULL;
-	}
-
-	for (i = 0; i < header->attrs; i++) {
-		struct perf_header_attr *attr = header->attr[i];
-		int j;
-
-		for (j = 0; j < attr->ids; j++) {
-			if (attr->id[j] == id)
-				return &attr->attr;
-		}
-	}
-
-	return NULL;
-}
-
-int event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
-			   event__handler_t process,
-			   struct perf_session *session)
+int perf_event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
+				perf_event__handler_t process,
+				struct perf_session *session)
 {
-	event_t *ev;
+	union perf_event *ev;
 	size_t size;
 	int err;
 
@@ -1033,17 +941,15 @@ int event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
 	return err;
 }
 
-int event__synthesize_attrs(struct perf_header *self, event__handler_t process,
-			    struct perf_session *session)
+int perf_session__synthesize_attrs(struct perf_session *session,
+				   perf_event__handler_t process)
 {
-	struct perf_header_attr	*attr;
-	int i, err = 0;
-
-	for (i = 0; i < self->attrs; i++) {
-		attr = self->attr[i];
+	struct perf_evsel *attr;
+	int err = 0;
 
-		err = event__synthesize_attr(&attr->attr, attr->ids, attr->id,
-					     process, session);
+	list_for_each_entry(attr, &session->evlist->entries, node) {
+		err = perf_event__synthesize_attr(&attr->attr, attr->ids,
+						  attr->id, process, session);
 		if (err) {
 			pr_debug("failed to create perf header attribute\n");
 			return err;
@@ -1053,29 +959,39 @@ int event__synthesize_attrs(struct perf_header *self, event__handler_t process,
 	return err;
 }
 
-int event__process_attr(event_t *self, struct perf_session *session)
+int perf_event__process_attr(union perf_event *event,
+			     struct perf_session *session)
 {
-	struct perf_header_attr *attr;
 	unsigned int i, ids, n_ids;
+	struct perf_evsel *evsel;
 
-	attr = perf_header_attr__new(&self->attr.attr);
-	if (attr == NULL)
+	if (session->evlist == NULL) {
+		session->evlist = perf_evlist__new(NULL, NULL);
+		if (session->evlist == NULL)
+			return -ENOMEM;
+	}
+
+	evsel = perf_evsel__new(&event->attr.attr,
+				session->evlist->nr_entries);
+	if (evsel == NULL)
 		return -ENOMEM;
 
-	ids = self->header.size;
-	ids -= (void *)&self->attr.id - (void *)self;
+	perf_evlist__add(session->evlist, evsel);
+
+	ids = event->header.size;
+	ids -= (void *)&event->attr.id - (void *)event;
 	n_ids = ids / sizeof(u64);
+	/*
+	 * We don't have the cpu and thread maps on the header, so
+	 * for allocating the perf_sample_id table we fake 1 cpu and
+	 * hattr->ids threads.
+	 */
+	if (perf_evsel__alloc_id(evsel, 1, n_ids))
+		return -ENOMEM;
 
 	for (i = 0; i < n_ids; i++) {
-		if (perf_header_attr__add_id(attr, self->attr.id[i]) < 0) {
-			perf_header_attr__delete(attr);
-			return -ENOMEM;
-		}
-	}
-
-	if (perf_header__add_attr(&session->header, attr) < 0) {
-		perf_header_attr__delete(attr);
-		return -ENOMEM;
+		perf_evlist__id_add(session->evlist, evsel, 0, i,
+				    event->attr.id[i]);
 	}
 
 	perf_session__update_sample_type(session);
@@ -1083,11 +999,11 @@ int event__process_attr(event_t *self, struct perf_session *session)
 	return 0;
 }
 
-int event__synthesize_event_type(u64 event_id, char *name,
-				 event__handler_t process,
-				 struct perf_session *session)
+int perf_event__synthesize_event_type(u64 event_id, char *name,
+				      perf_event__handler_t process,
+				      struct perf_session *session)
 {
-	event_t ev;
+	union perf_event ev;
 	size_t size = 0;
 	int err = 0;
 
@@ -1108,8 +1024,8 @@ int event__synthesize_event_type(u64 event_id, char *name,
 	return err;
 }
 
-int event__synthesize_event_types(event__handler_t process,
-				  struct perf_session *session)
+int perf_event__synthesize_event_types(perf_event__handler_t process,
+				       struct perf_session *session)
 {
 	struct perf_trace_event_type *type;
 	int i, err = 0;
@@ -1117,8 +1033,9 @@ int event__synthesize_event_types(event__handler_t process,
 	for (i = 0; i < event_count; i++) {
 		type = &events[i];
 
-		err = event__synthesize_event_type(type->event_id, type->name,
-						   process, session);
+		err = perf_event__synthesize_event_type(type->event_id,
+							type->name, process,
+							session);
 		if (err) {
 			pr_debug("failed to create perf header event type\n");
 			return err;
@@ -1128,28 +1045,28 @@ int event__synthesize_event_types(event__handler_t process,
 	return err;
 }
 
-int event__process_event_type(event_t *self,
-			      struct perf_session *session __unused)
+int perf_event__process_event_type(union perf_event *event,
+				   struct perf_session *session __unused)
 {
-	if (perf_header__push_event(self->event_type.event_type.event_id,
-				    self->event_type.event_type.name) < 0)
+	if (perf_header__push_event(event->event_type.event_type.event_id,
+				    event->event_type.event_type.name) < 0)
 		return -ENOMEM;
 
 	return 0;
 }
 
-int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
-				   event__handler_t process,
+int perf_event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
+					 perf_event__handler_t process,
 				   struct perf_session *session __unused)
 {
-	event_t ev;
+	union perf_event ev;
 	ssize_t size = 0, aligned_size = 0, padding;
-	int err = 0;
+	int err __used = 0;
 
 	memset(&ev, 0, sizeof(ev));
 
 	ev.tracing_data.header.type = PERF_RECORD_HEADER_TRACING_DATA;
-	size = read_tracing_data_size(fd, pattrs);
+	size = read_tracing_data_size(fd, &evlist->entries);
 	if (size <= 0)
 		return size;
 	aligned_size = ALIGN(size, sizeof(u64));
@@ -1159,16 +1076,16 @@ int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
 
 	process(&ev, NULL, session);
 
-	err = read_tracing_data(fd, pattrs);
+	err = read_tracing_data(fd, &evlist->entries);
 	write_padded(fd, NULL, 0, padding);
 
 	return aligned_size;
 }
 
-int event__process_tracing_data(event_t *self,
-				struct perf_session *session)
+int perf_event__process_tracing_data(union perf_event *event,
+				     struct perf_session *session)
 {
-	ssize_t size_read, padding, size = self->tracing_data.size;
+	ssize_t size_read, padding, size = event->tracing_data.size;
 	off_t offset = lseek(session->fd, 0, SEEK_CUR);
 	char buf[BUFSIZ];
 
@@ -1194,12 +1111,12 @@ int event__process_tracing_data(event_t *self,
 	return size_read + padding;
 }
 
-int event__synthesize_build_id(struct dso *pos, u16 misc,
-			       event__handler_t process,
-			       struct machine *machine,
-			       struct perf_session *session)
+int perf_event__synthesize_build_id(struct dso *pos, u16 misc,
+				    perf_event__handler_t process,
+				    struct machine *machine,
+				    struct perf_session *session)
 {
-	event_t ev;
+	union perf_event ev;
 	size_t len;
 	int err = 0;
 
@@ -1222,11 +1139,11 @@ int event__synthesize_build_id(struct dso *pos, u16 misc,
 	return err;
 }
 
-int event__process_build_id(event_t *self,
-			    struct perf_session *session)
+int perf_event__process_build_id(union perf_event *event,
+				 struct perf_session *session)
 {
-	__event_process_build_id(&self->build_id,
-				 self->build_id.filename,
+	__event_process_build_id(&event->build_id,
+				 event->build_id.filename,
 				 session);
 	return 0;
 }
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
index 33f16be7b72f..456661d7f10e 100644
--- a/tools/perf/util/header.h
+++ b/tools/perf/util/header.h
@@ -9,13 +9,6 @@
 
 #include <linux/bitmap.h>
 
-struct perf_header_attr {
-	struct perf_event_attr attr;
-	int ids, size;
-	u64 *id;
-	off_t id_offset;
-};
-
 enum {
 	HEADER_TRACE_INFO = 1,
 	HEADER_BUILD_ID,
@@ -46,14 +39,12 @@ struct perf_pipe_file_header {
 
 struct perf_header;
 
-int perf_file_header__read(struct perf_file_header *self,
+int perf_file_header__read(struct perf_file_header *header,
 			   struct perf_header *ph, int fd);
 
 struct perf_header {
 	int			frozen;
-	int			attrs, size;
 	bool			needs_swap;
-	struct perf_header_attr **attr;
 	s64			attr_offset;
 	u64			data_offset;
 	u64			data_size;
@@ -62,34 +53,25 @@ struct perf_header {
 	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
 };
 
-int perf_header__init(struct perf_header *self);
-void perf_header__exit(struct perf_header *self);
+struct perf_evlist;
 
-int perf_header__read(struct perf_session *session, int fd);
-int perf_header__write(struct perf_header *self, int fd, bool at_exit);
+int perf_session__read_header(struct perf_session *session, int fd);
+int perf_session__write_header(struct perf_session *session,
+			       struct perf_evlist *evlist,
+			       int fd, bool at_exit);
 int perf_header__write_pipe(int fd);
 
-int perf_header__add_attr(struct perf_header *self,
-			  struct perf_header_attr *attr);
-
 int perf_header__push_event(u64 id, const char *name);
 char *perf_header__find_event(u64 id);
 
-struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr);
-void perf_header_attr__delete(struct perf_header_attr *self);
+u64 perf_evlist__sample_type(struct perf_evlist *evlist);
+bool perf_evlist__sample_id_all(const struct perf_evlist *evlist);
+void perf_header__set_feat(struct perf_header *header, int feat);
+void perf_header__clear_feat(struct perf_header *header, int feat);
+bool perf_header__has_feat(const struct perf_header *header, int feat);
 
-int perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
-
-u64 perf_header__sample_type(struct perf_header *header);
-bool perf_header__sample_id_all(const struct perf_header *header);
-struct perf_event_attr *
-perf_header__find_attr(u64 id, struct perf_header *header);
-void perf_header__set_feat(struct perf_header *self, int feat);
-void perf_header__clear_feat(struct perf_header *self, int feat);
-bool perf_header__has_feat(const struct perf_header *self, int feat);
-
-int perf_header__process_sections(struct perf_header *self, int fd,
-				  int (*process)(struct perf_file_section *self,
+int perf_header__process_sections(struct perf_header *header, int fd,
+				  int (*process)(struct perf_file_section *section,
 						 struct perf_header *ph,
 						 int feat, int fd));
 
@@ -97,32 +79,31 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
 			  const char *name, bool is_kallsyms);
 int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir);
 
-int event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
-			   event__handler_t process,
-			   struct perf_session *session);
-int event__synthesize_attrs(struct perf_header *self,
-			    event__handler_t process,
-			    struct perf_session *session);
-int event__process_attr(event_t *self, struct perf_session *session);
-
-int event__synthesize_event_type(u64 event_id, char *name,
-				 event__handler_t process,
-				 struct perf_session *session);
-int event__synthesize_event_types(event__handler_t process,
-				  struct perf_session *session);
-int event__process_event_type(event_t *self,
-			      struct perf_session *session);
-
-int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
-				   event__handler_t process,
-				   struct perf_session *session);
-int event__process_tracing_data(event_t *self,
+int perf_event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
+				perf_event__handler_t process,
 				struct perf_session *session);
+int perf_session__synthesize_attrs(struct perf_session *session,
+				   perf_event__handler_t process);
+int perf_event__process_attr(union perf_event *event, struct perf_session *session);
+
+int perf_event__synthesize_event_type(u64 event_id, char *name,
+				      perf_event__handler_t process,
+				      struct perf_session *session);
+int perf_event__synthesize_event_types(perf_event__handler_t process,
+				       struct perf_session *session);
+int perf_event__process_event_type(union perf_event *event,
+				   struct perf_session *session);
 
-int event__synthesize_build_id(struct dso *pos, u16 misc,
-			       event__handler_t process,
-			       struct machine *machine,
-			       struct perf_session *session);
-int event__process_build_id(event_t *self, struct perf_session *session);
-
+int perf_event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
+					perf_event__handler_t process,
+					struct perf_session *session);
+int perf_event__process_tracing_data(union perf_event *event,
+				     struct perf_session *session);
+
+int perf_event__synthesize_build_id(struct dso *pos, u16 misc,
+				    perf_event__handler_t process,
+				    struct machine *machine,
+				    struct perf_session *session);
+int perf_event__process_build_id(union perf_event *event,
+				 struct perf_session *session);
 #endif /* __PERF_HEADER_H */
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
index df51560f16f7..627a02e03c57 100644
--- a/tools/perf/util/hist.c
+++ b/tools/perf/util/hist.c
@@ -1,3 +1,4 @@
+#include "annotate.h"
 #include "util.h"
 #include "build-id.h"
 #include "hist.h"
@@ -49,6 +50,15 @@ static void hists__calc_col_len(struct hists *self, struct hist_entry *h)
 
 	if (h->ms.sym)
 		hists__new_col_len(self, HISTC_SYMBOL, h->ms.sym->namelen);
+	else {
+		const unsigned int unresolved_col_width = BITS_PER_LONG / 4;
+
+		if (hists__col_len(self, HISTC_DSO) < unresolved_col_width &&
+		    !symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
+		    !symbol_conf.dso_list)
+			hists__set_col_len(self, HISTC_DSO,
+					   unresolved_col_width);
+	}
 
 	len = thread__comm_len(h->thread);
 	if (hists__new_col_len(self, HISTC_COMM, len))
@@ -211,7 +221,9 @@ void hist_entry__free(struct hist_entry *he)
  * collapse the histogram
  */
 
-static bool collapse__insert_entry(struct rb_root *root, struct hist_entry *he)
+static bool hists__collapse_insert_entry(struct hists *self,
+					 struct rb_root *root,
+					 struct hist_entry *he)
 {
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
@@ -226,8 +238,11 @@ static bool collapse__insert_entry(struct rb_root *root, struct hist_entry *he)
 
 		if (!cmp) {
 			iter->period += he->period;
-			if (symbol_conf.use_callchain)
-				callchain_merge(iter->callchain, he->callchain);
+			if (symbol_conf.use_callchain) {
+				callchain_cursor_reset(&self->callchain_cursor);
+				callchain_merge(&self->callchain_cursor, iter->callchain,
+						he->callchain);
+			}
 			hist_entry__free(he);
 			return false;
 		}
@@ -262,7 +277,7 @@ void hists__collapse_resort(struct hists *self)
 		next = rb_next(&n->rb_node);
 
 		rb_erase(&n->rb_node, &self->entries);
-		if (collapse__insert_entry(&tmp, n))
+		if (hists__collapse_insert_entry(self, &tmp, n))
 			hists__inc_nr_entries(self, n);
 	}
 
@@ -425,7 +440,7 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 		u64 cumul;
 
 		child = rb_entry(node, struct callchain_node, rb_node);
-		cumul = cumul_hits(child);
+		cumul = callchain_cumul_hits(child);
 		remaining -= cumul;
 
 		/*
@@ -947,225 +962,14 @@ void hists__filter_by_thread(struct hists *self, const struct thread *thread)
 	}
 }
 
-static int symbol__alloc_hist(struct symbol *self)
-{
-	struct sym_priv *priv = symbol__priv(self);
-	const int size = (sizeof(*priv->hist) +
-			  (self->end - self->start) * sizeof(u64));
-
-	priv->hist = zalloc(size);
-	return priv->hist == NULL ? -1 : 0;
-}
-
-int hist_entry__inc_addr_samples(struct hist_entry *self, u64 ip)
-{
-	unsigned int sym_size, offset;
-	struct symbol *sym = self->ms.sym;
-	struct sym_priv *priv;
-	struct sym_hist *h;
-
-	if (!sym || !self->ms.map)
-		return 0;
-
-	priv = symbol__priv(sym);
-	if (priv->hist == NULL && symbol__alloc_hist(sym) < 0)
-		return -ENOMEM;
-
-	sym_size = sym->end - sym->start;
-	offset = ip - sym->start;
-
-	pr_debug3("%s: ip=%#" PRIx64 "\n", __func__, self->ms.map->unmap_ip(self->ms.map, ip));
-
-	if (offset >= sym_size)
-		return 0;
-
-	h = priv->hist;
-	h->sum++;
-	h->ip[offset]++;
-
-	pr_debug3("%#" PRIx64 " %s: period++ [ip: %#" PRIx64 ", %#" PRIx64
-		  "] => %" PRIu64 "\n", self->ms.sym->start, self->ms.sym->name,
-		  ip, ip - self->ms.sym->start, h->ip[offset]);
-	return 0;
-}
-
-static struct objdump_line *objdump_line__new(s64 offset, char *line, size_t privsize)
-{
-	struct objdump_line *self = malloc(sizeof(*self) + privsize);
-
-	if (self != NULL) {
-		self->offset = offset;
-		self->line = line;
-	}
-
-	return self;
-}
-
-void objdump_line__free(struct objdump_line *self)
-{
-	free(self->line);
-	free(self);
-}
-
-static void objdump__add_line(struct list_head *head, struct objdump_line *line)
-{
-	list_add_tail(&line->node, head);
-}
-
-struct objdump_line *objdump__get_next_ip_line(struct list_head *head,
-					       struct objdump_line *pos)
-{
-	list_for_each_entry_continue(pos, head, node)
-		if (pos->offset >= 0)
-			return pos;
-
-	return NULL;
-}
-
-static int hist_entry__parse_objdump_line(struct hist_entry *self, FILE *file,
-					  struct list_head *head, size_t privsize)
+int hist_entry__inc_addr_samples(struct hist_entry *he, int evidx, u64 ip)
 {
-	struct symbol *sym = self->ms.sym;
-	struct objdump_line *objdump_line;
-	char *line = NULL, *tmp, *tmp2, *c;
-	size_t line_len;
-	s64 line_ip, offset = -1;
-
-	if (getline(&line, &line_len, file) < 0)
-		return -1;
-
-	if (!line)
-		return -1;
-
-	while (line_len != 0 && isspace(line[line_len - 1]))
-		line[--line_len] = '\0';
-
-	c = strchr(line, '\n');
-	if (c)
-		*c = 0;
-
-	line_ip = -1;
-
-	/*
-	 * Strip leading spaces:
-	 */
-	tmp = line;
-	while (*tmp) {
-		if (*tmp != ' ')
-			break;
-		tmp++;
-	}
-
-	if (*tmp) {
-		/*
-		 * Parse hexa addresses followed by ':'
-		 */
-		line_ip = strtoull(tmp, &tmp2, 16);
-		if (*tmp2 != ':' || tmp == tmp2 || tmp2[1] == '\0')
-			line_ip = -1;
-	}
-
-	if (line_ip != -1) {
-		u64 start = map__rip_2objdump(self->ms.map, sym->start),
-		    end = map__rip_2objdump(self->ms.map, sym->end);
-
-		offset = line_ip - start;
-		if (offset < 0 || (u64)line_ip > end)
-			offset = -1;
-	}
-
-	objdump_line = objdump_line__new(offset, line, privsize);
-	if (objdump_line == NULL) {
-		free(line);
-		return -1;
-	}
-	objdump__add_line(head, objdump_line);
-
-	return 0;
+	return symbol__inc_addr_samples(he->ms.sym, he->ms.map, evidx, ip);
 }
 
-int hist_entry__annotate(struct hist_entry *self, struct list_head *head,
-			 size_t privsize)
+int hist_entry__annotate(struct hist_entry *he, size_t privsize)
 {
-	struct symbol *sym = self->ms.sym;
-	struct map *map = self->ms.map;
-	struct dso *dso = map->dso;
-	char *filename = dso__build_id_filename(dso, NULL, 0);
-	bool free_filename = true;
-	char command[PATH_MAX * 2];
-	FILE *file;
-	int err = 0;
-	u64 len;
-	char symfs_filename[PATH_MAX];
-
-	if (filename) {
-		snprintf(symfs_filename, sizeof(symfs_filename), "%s%s",
-			 symbol_conf.symfs, filename);
-	}
-
-	if (filename == NULL) {
-		if (dso->has_build_id) {
-			pr_err("Can't annotate %s: not enough memory\n",
-			       sym->name);
-			return -ENOMEM;
-		}
-		goto fallback;
-	} else if (readlink(symfs_filename, command, sizeof(command)) < 0 ||
-		   strstr(command, "[kernel.kallsyms]") ||
-		   access(symfs_filename, R_OK)) {
-		free(filename);
-fallback:
-		/*
-		 * If we don't have build-ids or the build-id file isn't in the
-		 * cache, or is just a kallsyms file, well, lets hope that this
-		 * DSO is the same as when 'perf record' ran.
-		 */
-		filename = dso->long_name;
-		snprintf(symfs_filename, sizeof(symfs_filename), "%s%s",
-			 symbol_conf.symfs, filename);
-		free_filename = false;
-	}
-
-	if (dso->origin == DSO__ORIG_KERNEL) {
-		if (dso->annotate_warned)
-			goto out_free_filename;
-		err = -ENOENT;
-		dso->annotate_warned = 1;
-		pr_err("Can't annotate %s: No vmlinux file was found in the "
-		       "path\n", sym->name);
-		goto out_free_filename;
-	}
-
-	pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__,
-		 filename, sym->name, map->unmap_ip(map, sym->start),
-		 map->unmap_ip(map, sym->end));
-
-	len = sym->end - sym->start;
-
-	pr_debug("annotating [%p] %30s : [%p] %30s\n",
-		 dso, dso->long_name, sym, sym->name);
-
-	snprintf(command, sizeof(command),
-		 "objdump --start-address=0x%016" PRIx64 " --stop-address=0x%016" PRIx64 " -dS -C %s|grep -v %s|expand",
-		 map__rip_2objdump(map, sym->start),
-		 map__rip_2objdump(map, sym->end),
-		 symfs_filename, filename);
-
-	pr_debug("Executing: %s\n", command);
-
-	file = popen(command, "r");
-	if (!file)
-		goto out_free_filename;
-
-	while (!feof(file))
-		if (hist_entry__parse_objdump_line(self, file, head, privsize) < 0)
-			break;
-
-	pclose(file);
-out_free_filename:
-	if (free_filename)
-		free(filename);
-	return err;
+	return symbol__annotate(he->ms.sym, he->ms.map, privsize);
 }
 
 void hists__inc_nr_events(struct hists *self, u32 type)
@@ -1180,8 +984,12 @@ size_t hists__fprintf_nr_events(struct hists *self, FILE *fp)
 	size_t ret = 0;
 
 	for (i = 0; i < PERF_RECORD_HEADER_MAX; ++i) {
-		const char *name = event__get_event_name(i);
+		const char *name;
+
+		if (self->stats.nr_events[i] == 0)
+			continue;
 
+		name = perf_event__name(i);
 		if (!strcmp(name, "UNKNOWN"))
 			continue;
 
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index ee789856a8c9..cb6858a2f9a3 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -9,33 +9,6 @@ extern struct callchain_param callchain_param;
 struct hist_entry;
 struct addr_location;
 struct symbol;
-struct rb_root;
-
-struct objdump_line {
-	struct list_head node;
-	s64		 offset;
-	char		 *line;
-};
-
-void objdump_line__free(struct objdump_line *self);
-struct objdump_line *objdump__get_next_ip_line(struct list_head *head,
-					       struct objdump_line *pos);
-
-struct sym_hist {
-	u64		sum;
-	u64		ip[0];
-};
-
-struct sym_ext {
-	struct rb_node	node;
-	double		percent;
-	char		*path;
-};
-
-struct sym_priv {
-	struct sym_hist	*hist;
-	struct sym_ext	*ext;
-};
 
 /*
  * The kernel collects the number of events it couldn't send in a stretch and
@@ -69,14 +42,13 @@ enum hist_column {
 };
 
 struct hists {
-	struct rb_node		rb_node;
 	struct rb_root		entries;
 	u64			nr_entries;
 	struct events_stats	stats;
-	u64			config;
 	u64			event_stream;
-	u32			type;
 	u16			col_len[HISTC_NR_COLS];
+	/* Best would be to reuse the session callchain cursor */
+	struct callchain_cursor	callchain_cursor;
 };
 
 struct hist_entry *__hists__add_entry(struct hists *self,
@@ -102,9 +74,8 @@ size_t hists__fprintf_nr_events(struct hists *self, FILE *fp);
 size_t hists__fprintf(struct hists *self, struct hists *pair,
 		      bool show_displacement, FILE *fp);
 
-int hist_entry__inc_addr_samples(struct hist_entry *self, u64 ip);
-int hist_entry__annotate(struct hist_entry *self, struct list_head *head,
-			 size_t privsize);
+int hist_entry__inc_addr_samples(struct hist_entry *self, int evidx, u64 addr);
+int hist_entry__annotate(struct hist_entry *self, size_t privsize);
 
 void hists__filter_by_dso(struct hists *self, const struct dso *dso);
 void hists__filter_by_thread(struct hists *self, const struct thread *thread);
@@ -113,21 +84,18 @@ u16 hists__col_len(struct hists *self, enum hist_column col);
 void hists__set_col_len(struct hists *self, enum hist_column col, u16 len);
 bool hists__new_col_len(struct hists *self, enum hist_column col, u16 len);
 
-#ifdef NO_NEWT_SUPPORT
-static inline int hists__browse(struct hists *self __used,
-				const char *helpline __used,
-				const char *ev_name __used)
-{
-	return 0;
-}
+struct perf_evlist;
 
-static inline int hists__tui_browse_tree(struct rb_root *self __used,
-					 const char *help __used)
+#ifdef NO_NEWT_SUPPORT
+static inline
+int perf_evlist__tui_browse_hists(struct perf_evlist *evlist __used,
+				  const char *help __used)
 {
 	return 0;
 }
 
-static inline int hist_entry__tui_annotate(struct hist_entry *self __used)
+static inline int hist_entry__tui_annotate(struct hist_entry *self __used,
+					   int evidx __used)
 {
 	return 0;
 }
@@ -135,14 +103,12 @@ static inline int hist_entry__tui_annotate(struct hist_entry *self __used)
 #define KEY_RIGHT -2
 #else
 #include <newt.h>
-int hists__browse(struct hists *self, const char *helpline,
-		  const char *ev_name);
-int hist_entry__tui_annotate(struct hist_entry *self);
+int hist_entry__tui_annotate(struct hist_entry *self, int evidx);
 
 #define KEY_LEFT NEWT_KEY_LEFT
 #define KEY_RIGHT NEWT_KEY_RIGHT
 
-int hists__tui_browse_tree(struct rb_root *self, const char *help);
+int perf_evlist__tui_browse_hists(struct perf_evlist *evlist, const char *help);
 #endif
 
 unsigned int hists__sort_list_width(struct hists *self);
diff --git a/tools/perf/util/include/linux/list.h b/tools/perf/util/include/linux/list.h
index f5ca26e53fbb..356c7e467b83 100644
--- a/tools/perf/util/include/linux/list.h
+++ b/tools/perf/util/include/linux/list.h
@@ -1,3 +1,4 @@
+#include <linux/kernel.h>
 #include "../../../../include/linux/list.h"
 
 #ifndef PERF_LIST_H
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 135f69baf966..54a7e2634d58 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -1,6 +1,7 @@
 #include "../../../include/linux/hw_breakpoint.h"
 #include "util.h"
 #include "../perf.h"
+#include "evlist.h"
 #include "evsel.h"
 #include "parse-options.h"
 #include "parse-events.h"
@@ -11,10 +12,6 @@
 #include "header.h"
 #include "debugfs.h"
 
-int				nr_counters;
-
-LIST_HEAD(evsel_list);
-
 struct event_symbol {
 	u8		type;
 	u64		config;
@@ -271,6 +268,9 @@ const char *event_name(struct perf_evsel *evsel)
 	u64 config = evsel->attr.config;
 	int type = evsel->attr.type;
 
+	if (evsel->name)
+		return evsel->name;
+
 	return __event_name(type, config);
 }
 
@@ -449,8 +449,8 @@ parse_single_tracepoint_event(char *sys_name,
 /* sys + ':' + event + ':' + flags*/
 #define MAX_EVOPT_LEN	(MAX_EVENT_LENGTH * 2 + 2 + 128)
 static enum event_result
-parse_multiple_tracepoint_event(char *sys_name, const char *evt_exp,
-				char *flags)
+parse_multiple_tracepoint_event(const struct option *opt, char *sys_name,
+				const char *evt_exp, char *flags)
 {
 	char evt_path[MAXPATHLEN];
 	struct dirent *evt_ent;
@@ -483,15 +483,16 @@ parse_multiple_tracepoint_event(char *sys_name, const char *evt_exp,
 		if (len < 0)
 			return EVT_FAILED;
 
-		if (parse_events(NULL, event_opt, 0))
+		if (parse_events(opt, event_opt, 0))
 			return EVT_FAILED;
 	}
 
 	return EVT_HANDLED_ALL;
 }
 
-static enum event_result parse_tracepoint_event(const char **strp,
-				    struct perf_event_attr *attr)
+static enum event_result
+parse_tracepoint_event(const struct option *opt, const char **strp,
+		       struct perf_event_attr *attr)
 {
 	const char *evt_name;
 	char *flags = NULL, *comma_loc;
@@ -530,7 +531,7 @@ static enum event_result parse_tracepoint_event(const char **strp,
 		return EVT_FAILED;
 	if (strpbrk(evt_name, "*?")) {
 		*strp += strlen(sys_name) + evt_length + 1; /* 1 == the ':' */
-		return parse_multiple_tracepoint_event(sys_name, evt_name,
+		return parse_multiple_tracepoint_event(opt, sys_name, evt_name,
 						       flags);
 	} else {
 		return parse_single_tracepoint_event(sys_name, evt_name,
@@ -740,11 +741,12 @@ parse_event_modifier(const char **strp, struct perf_event_attr *attr)
  * Symbolic names are (almost) exactly matched.
  */
 static enum event_result
-parse_event_symbols(const char **str, struct perf_event_attr *attr)
+parse_event_symbols(const struct option *opt, const char **str,
+		    struct perf_event_attr *attr)
 {
 	enum event_result ret;
 
-	ret = parse_tracepoint_event(str, attr);
+	ret = parse_tracepoint_event(opt, str, attr);
 	if (ret != EVT_FAILED)
 		goto modifier;
 
@@ -778,14 +780,17 @@ modifier:
 	return ret;
 }
 
-int parse_events(const struct option *opt __used, const char *str, int unset __used)
+int parse_events(const struct option *opt, const char *str, int unset __used)
 {
+	struct perf_evlist *evlist = *(struct perf_evlist **)opt->value;
 	struct perf_event_attr attr;
 	enum event_result ret;
+	const char *ostr;
 
 	for (;;) {
+		ostr = str;
 		memset(&attr, 0, sizeof(attr));
-		ret = parse_event_symbols(&str, &attr);
+		ret = parse_event_symbols(opt, &str, &attr);
 		if (ret == EVT_FAILED)
 			return -1;
 
@@ -794,12 +799,15 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u
 
 		if (ret != EVT_HANDLED_ALL) {
 			struct perf_evsel *evsel;
-			evsel = perf_evsel__new(&attr,
-						nr_counters);
+			evsel = perf_evsel__new(&attr, evlist->nr_entries);
 			if (evsel == NULL)
 				return -1;
-			list_add_tail(&evsel->node, &evsel_list);
-			++nr_counters;
+			perf_evlist__add(evlist, evsel);
+
+			evsel->name = calloc(str - ostr + 1, 1);
+			if (!evsel->name)
+				return -1;
+			strncpy(evsel->name, ostr, str - ostr);
 		}
 
 		if (*str == 0)
@@ -813,13 +821,14 @@ 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 parse_filter(const struct option *opt, const char *str,
 		 int unset __used)
 {
+	struct perf_evlist *evlist = *(struct perf_evlist **)opt->value;
 	struct perf_evsel *last = NULL;
 
-	if (!list_empty(&evsel_list))
-		last = list_entry(evsel_list.prev, struct perf_evsel, node);
+	if (evlist->nr_entries > 0)
+		last = list_entry(evlist->entries.prev, struct perf_evsel, node);
 
 	if (last == NULL || last->attr.type != PERF_TYPE_TRACEPOINT) {
 		fprintf(stderr,
@@ -849,7 +858,7 @@ static const char * const event_type_descriptors[] = {
  * Print the events from <debugfs_mount_point>/tracing/events
  */
 
-static void print_tracepoint_events(void)
+void print_tracepoint_events(const char *subsys_glob, const char *event_glob)
 {
 	DIR *sys_dir, *evt_dir;
 	struct dirent *sys_next, *evt_next, sys_dirent, evt_dirent;
@@ -864,6 +873,9 @@ static void print_tracepoint_events(void)
 		return;
 
 	for_each_subsystem(sys_dir, sys_dirent, sys_next) {
+		if (subsys_glob != NULL && 
+		    !strglobmatch(sys_dirent.d_name, subsys_glob))
+			continue;
 
 		snprintf(dir_path, MAXPATHLEN, "%s/%s", debugfs_path,
 			 sys_dirent.d_name);
@@ -872,6 +884,10 @@ static void print_tracepoint_events(void)
 			continue;
 
 		for_each_event(sys_dirent, evt_dir, evt_dirent, evt_next) {
+			if (event_glob != NULL && 
+			    !strglobmatch(evt_dirent.d_name, event_glob))
+				continue;
+
 			snprintf(evt_path, MAXPATHLEN, "%s:%s",
 				 sys_dirent.d_name, evt_dirent.d_name);
 			printf("  %-42s [%s]\n", evt_path,
@@ -923,13 +939,61 @@ int is_valid_tracepoint(const char *event_string)
 	return 0;
 }
 
+void print_events_type(u8 type)
+{
+	struct event_symbol *syms = event_symbols;
+	unsigned int i;
+	char name[64];
+
+	for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) {
+		if (type != syms->type)
+			continue;
+
+		if (strlen(syms->alias))
+			snprintf(name, sizeof(name),  "%s OR %s",
+				 syms->symbol, syms->alias);
+		else
+			snprintf(name, sizeof(name), "%s", syms->symbol);
+
+		printf("  %-42s [%s]\n", name,
+			event_type_descriptors[type]);
+	}
+}
+
+int print_hwcache_events(const char *event_glob)
+{
+	unsigned int type, op, i, printed = 0;
+
+	for (type = 0; type < PERF_COUNT_HW_CACHE_MAX; type++) {
+		for (op = 0; op < PERF_COUNT_HW_CACHE_OP_MAX; op++) {
+			/* skip invalid cache type */
+			if (!is_cache_op_valid(type, op))
+				continue;
+
+			for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
+				char *name = event_cache_name(type, op, i);
+
+				if (event_glob != NULL && 
+				    !strglobmatch(name, event_glob))
+					continue;
+
+				printf("  %-42s [%s]\n", name,
+					event_type_descriptors[PERF_TYPE_HW_CACHE]);
+				++printed;
+			}
+		}
+	}
+
+	return printed;
+}
+
 /*
  * Print the help text for the event symbols:
  */
-void print_events(void)
+void print_events(const char *event_glob)
 {
 	struct event_symbol *syms = event_symbols;
-	unsigned int i, type, op, prev_type = -1;
+	unsigned int i, type, prev_type = -1, printed = 0, ntypes_printed = 0;
 	char name[40];
 
 	printf("\n");
@@ -938,8 +1002,16 @@ void print_events(void)
 	for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) {
 		type = syms->type;
 
-		if (type != prev_type)
+		if (type != prev_type && printed) {
 			printf("\n");
+			printed = 0;
+			ntypes_printed++;
+		}
+
+		if (event_glob != NULL && 
+		    !(strglobmatch(syms->symbol, event_glob) ||
+		      (syms->alias && strglobmatch(syms->alias, event_glob))))
+			continue;
 
 		if (strlen(syms->alias))
 			sprintf(name, "%s OR %s", syms->symbol, syms->alias);
@@ -949,22 +1021,17 @@ void print_events(void)
 			event_type_descriptors[type]);
 
 		prev_type = type;
+		++printed;
 	}
 
-	printf("\n");
-	for (type = 0; type < PERF_COUNT_HW_CACHE_MAX; type++) {
-		for (op = 0; op < PERF_COUNT_HW_CACHE_OP_MAX; op++) {
-			/* skip invalid cache type */
-			if (!is_cache_op_valid(type, op))
-				continue;
-
-			for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
-				printf("  %-42s [%s]\n",
-					event_cache_name(type, op, i),
-					event_type_descriptors[PERF_TYPE_HW_CACHE]);
-			}
-		}
+	if (ntypes_printed) {
+		printed = 0;
+		printf("\n");
 	}
+	print_hwcache_events(event_glob);
+
+	if (event_glob != NULL)
+		return;
 
 	printf("\n");
 	printf("  %-42s [%s]\n",
@@ -977,37 +1044,7 @@ void print_events(void)
 			event_type_descriptors[PERF_TYPE_BREAKPOINT]);
 	printf("\n");
 
-	print_tracepoint_events();
+	print_tracepoint_events(NULL, NULL);
 
 	exit(129);
 }
-
-int perf_evsel_list__create_default(void)
-{
-	struct perf_evsel *evsel;
-	struct perf_event_attr attr;
-
-	memset(&attr, 0, sizeof(attr));
-	attr.type = PERF_TYPE_HARDWARE;
-	attr.config = PERF_COUNT_HW_CPU_CYCLES;
-
-	evsel = perf_evsel__new(&attr, 0);
-
-	if (evsel == NULL)
-		return -ENOMEM;
-
-	list_add(&evsel->node, &evsel_list);
-	++nr_counters;
-	return 0;
-}
-
-void perf_evsel_list__delete(void)
-{
-	struct perf_evsel *pos, *n;
-
-	list_for_each_entry_safe(pos, n, &evsel_list, node) {
-		list_del_init(&pos->node);
-		perf_evsel__delete(pos);
-	}
-	nr_counters = 0;
-}
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index 458e3ecf17af..212f88e07a9c 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -9,11 +9,6 @@
 struct list_head;
 struct perf_evsel;
 
-extern struct list_head evsel_list;
-
-int perf_evsel_list__create_default(void);
-void perf_evsel_list__delete(void);
-
 struct option;
 
 struct tracepoint_path {
@@ -25,8 +20,6 @@ struct tracepoint_path {
 extern struct tracepoint_path *tracepoint_id_to_path(u64 config);
 extern bool have_tracepoints(struct list_head *evlist);
 
-extern int			nr_counters;
-
 const char *event_name(struct perf_evsel *event);
 extern const char *__event_name(int type, u64 config);
 
@@ -35,7 +28,10 @@ extern int parse_filter(const struct option *opt, const char *str, int unset);
 
 #define EVENTS_HELP_MAX (128*1024)
 
-extern void print_events(void);
+void print_events(const char *event_glob);
+void print_events_type(u8 type);
+void print_tracepoint_events(const char *subsys_glob, const char *event_glob);
+int print_hwcache_events(const char *event_glob);
 extern int is_valid_tracepoint(const char *event_string);
 
 extern char debugfs_path[];
diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
index 6e29d9c9dccc..5ddee66020a7 100644
--- a/tools/perf/util/probe-event.c
+++ b/tools/perf/util/probe-event.c
@@ -31,6 +31,7 @@
 #include <string.h>
 #include <stdarg.h>
 #include <limits.h>
+#include <elf.h>
 
 #undef _GNU_SOURCE
 #include "util.h"
@@ -111,7 +112,25 @@ static struct symbol *__find_kernel_function_by_name(const char *name,
 						     NULL);
 }
 
-const char *kernel_get_module_path(const char *module)
+static struct map *kernel_get_module_map(const char *module)
+{
+	struct rb_node *nd;
+	struct map_groups *grp = &machine.kmaps;
+
+	if (!module)
+		module = "kernel";
+
+	for (nd = rb_first(&grp->maps[MAP__FUNCTION]); nd; nd = rb_next(nd)) {
+		struct map *pos = rb_entry(nd, struct map, rb_node);
+		if (strncmp(pos->dso->short_name + 1, module,
+			    pos->dso->short_name_len - 2) == 0) {
+			return pos;
+		}
+	}
+	return NULL;
+}
+
+static struct dso *kernel_get_module_dso(const char *module)
 {
 	struct dso *dso;
 	struct map *map;
@@ -141,7 +160,13 @@ const char *kernel_get_module_path(const char *module)
 		}
 	}
 found:
-	return dso->long_name;
+	return dso;
+}
+
+const char *kernel_get_module_path(const char *module)
+{
+	struct dso *dso = kernel_get_module_dso(module);
+	return (dso) ? dso->long_name : NULL;
 }
 
 #ifdef DWARF_SUPPORT
@@ -384,7 +409,7 @@ int show_line_range(struct line_range *lr, const char *module)
 	setup_pager();
 
 	if (lr->function)
-		fprintf(stdout, "<%s:%d>\n", lr->function,
+		fprintf(stdout, "<%s@%s:%d>\n", lr->function, lr->path,
 			lr->start - lr->offset);
 	else
 		fprintf(stdout, "<%s:%d>\n", lr->path, lr->start);
@@ -426,12 +451,14 @@ end:
 }
 
 static int show_available_vars_at(int fd, struct perf_probe_event *pev,
-				  int max_vls, bool externs)
+				  int max_vls, struct strfilter *_filter,
+				  bool externs)
 {
 	char *buf;
-	int ret, i;
+	int ret, i, nvars;
 	struct str_node *node;
 	struct variable_list *vls = NULL, *vl;
+	const char *var;
 
 	buf = synthesize_perf_probe_point(&pev->point);
 	if (!buf)
@@ -439,36 +466,45 @@ static int show_available_vars_at(int fd, struct perf_probe_event *pev,
 	pr_debug("Searching variables at %s\n", buf);
 
 	ret = find_available_vars_at(fd, pev, &vls, max_vls, externs);
-	if (ret > 0) {
-		/* Some variables were found */
-		fprintf(stdout, "Available variables at %s\n", buf);
-		for (i = 0; i < ret; i++) {
-			vl = &vls[i];
-			/*
-			 * A probe point might be converted to
-			 * several trace points.
-			 */
-			fprintf(stdout, "\t@<%s+%lu>\n", vl->point.symbol,
-				vl->point.offset);
-			free(vl->point.symbol);
-			if (vl->vars) {
-				strlist__for_each(node, vl->vars)
+	if (ret <= 0) {
+		pr_err("Failed to find variables at %s (%d)\n", buf, ret);
+		goto end;
+	}
+	/* Some variables are found */
+	fprintf(stdout, "Available variables at %s\n", buf);
+	for (i = 0; i < ret; i++) {
+		vl = &vls[i];
+		/*
+		 * A probe point might be converted to
+		 * several trace points.
+		 */
+		fprintf(stdout, "\t@<%s+%lu>\n", vl->point.symbol,
+			vl->point.offset);
+		free(vl->point.symbol);
+		nvars = 0;
+		if (vl->vars) {
+			strlist__for_each(node, vl->vars) {
+				var = strchr(node->s, '\t') + 1;
+				if (strfilter__compare(_filter, var)) {
 					fprintf(stdout, "\t\t%s\n", node->s);
-				strlist__delete(vl->vars);
-			} else
-				fprintf(stdout, "(No variables)\n");
+					nvars++;
+				}
+			}
+			strlist__delete(vl->vars);
 		}
-		free(vls);
-	} else
-		pr_err("Failed to find variables at %s (%d)\n", buf, ret);
-
+		if (nvars == 0)
+			fprintf(stdout, "\t\t(No matched variables)\n");
+	}
+	free(vls);
+end:
 	free(buf);
 	return ret;
 }
 
 /* Show available variables on given probe point */
 int show_available_vars(struct perf_probe_event *pevs, int npevs,
-			int max_vls, const char *module, bool externs)
+			int max_vls, const char *module,
+			struct strfilter *_filter, bool externs)
 {
 	int i, fd, ret = 0;
 
@@ -485,7 +521,8 @@ int show_available_vars(struct perf_probe_event *pevs, int npevs,
 	setup_pager();
 
 	for (i = 0; i < npevs && ret >= 0; i++)
-		ret = show_available_vars_at(fd, &pevs[i], max_vls, externs);
+		ret = show_available_vars_at(fd, &pevs[i], max_vls, _filter,
+					     externs);
 
 	close(fd);
 	return ret;
@@ -531,7 +568,9 @@ int show_line_range(struct line_range *lr __unused, const char *module __unused)
 
 int show_available_vars(struct perf_probe_event *pevs __unused,
 			int npevs __unused, int max_vls __unused,
-			const char *module __unused, bool externs __unused)
+			const char *module __unused,
+			struct strfilter *filter __unused,
+			bool externs __unused)
 {
 	pr_warning("Debuginfo-analysis is not supported.\n");
 	return -ENOSYS;
@@ -556,11 +595,11 @@ static int parse_line_num(char **ptr, int *val, const char *what)
  * The line range syntax is described by:
  *
  *         SRC[:SLN[+NUM|-ELN]]
- *         FNC[:SLN[+NUM|-ELN]]
+ *         FNC[@SRC][:SLN[+NUM|-ELN]]
  */
 int parse_line_range_desc(const char *arg, struct line_range *lr)
 {
-	char *range, *name = strdup(arg);
+	char *range, *file, *name = strdup(arg);
 	int err;
 
 	if (!name)
@@ -610,7 +649,16 @@ int parse_line_range_desc(const char *arg, struct line_range *lr)
 		}
 	}
 
-	if (strchr(name, '.'))
+	file = strchr(name, '@');
+	if (file) {
+		*file = '\0';
+		lr->file = strdup(++file);
+		if (lr->file == NULL) {
+			err = -ENOMEM;
+			goto err;
+		}
+		lr->function = name;
+	} else if (strchr(name, '.'))
 		lr->file = name;
 	else
 		lr->function = name;
@@ -1784,9 +1832,12 @@ int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
 	}
 
 	/* Loop 2: add all events */
-	for (i = 0; i < npevs && ret >= 0; i++)
+	for (i = 0; i < npevs; i++) {
 		ret = __add_probe_trace_events(pkgs[i].pev, pkgs[i].tevs,
 						pkgs[i].ntevs, force_add);
+		if (ret < 0)
+			break;
+	}
 end:
 	/* Loop 3: cleanup and free trace events  */
 	for (i = 0; i < npevs; i++) {
@@ -1912,4 +1963,46 @@ int del_perf_probe_events(struct strlist *dellist)
 
 	return ret;
 }
+/* TODO: don't use a global variable for filter ... */
+static struct strfilter *available_func_filter;
+
+/*
+ * If a symbol corresponds to a function with global binding and
+ * matches filter return 0. For all others return 1.
+ */
+static int filter_available_functions(struct map *map __unused,
+				      struct symbol *sym)
+{
+	if (sym->binding == STB_GLOBAL &&
+	    strfilter__compare(available_func_filter, sym->name))
+		return 0;
+	return 1;
+}
+
+int show_available_funcs(const char *module, struct strfilter *_filter)
+{
+	struct map *map;
+	int ret;
+
+	setup_pager();
+
+	ret = init_vmlinux();
+	if (ret < 0)
+		return ret;
 
+	map = kernel_get_module_map(module);
+	if (!map) {
+		pr_err("Failed to find %s map.\n", (module) ? : "kernel");
+		return -EINVAL;
+	}
+	available_func_filter = _filter;
+	if (map__load(map, filter_available_functions)) {
+		pr_err("Failed to load map.\n");
+		return -EINVAL;
+	}
+	if (!dso__sorted_by_name(map->dso, map->type))
+		dso__sort_by_name(map->dso, map->type);
+
+	dso__fprintf_symbols_by_name(map->dso, map->type, stdout);
+	return 0;
+}
diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h
index 5accbedfea37..3434fc9d79d5 100644
--- a/tools/perf/util/probe-event.h
+++ b/tools/perf/util/probe-event.h
@@ -3,6 +3,7 @@
 
 #include <stdbool.h>
 #include "strlist.h"
+#include "strfilter.h"
 
 extern bool probe_event_dry_run;
 
@@ -126,7 +127,8 @@ extern int show_perf_probe_events(void);
 extern int show_line_range(struct line_range *lr, const char *module);
 extern int show_available_vars(struct perf_probe_event *pevs, int npevs,
 			       int max_probe_points, const char *module,
-			       bool externs);
+			       struct strfilter *filter, bool externs);
+extern int show_available_funcs(const char *module, struct strfilter *filter);
 
 
 /* Maximum index number of event-name postfix */
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index ab83b6ac5d65..194f9e2a3285 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -33,6 +33,7 @@
 #include <ctype.h>
 #include <dwarf-regs.h>
 
+#include <linux/bitops.h>
 #include "event.h"
 #include "debug.h"
 #include "util.h"
@@ -280,6 +281,19 @@ static bool die_compare_name(Dwarf_Die *dw_die, const char *tname)
 	return name ? (strcmp(tname, name) == 0) : false;
 }
 
+/* Get callsite line number of inline-function instance */
+static int die_get_call_lineno(Dwarf_Die *in_die)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Word ret;
+
+	if (!dwarf_attr(in_die, DW_AT_call_line, &attr))
+		return -ENOENT;
+
+	dwarf_formudata(&attr, &ret);
+	return (int)ret;
+}
+
 /* Get type die */
 static Dwarf_Die *die_get_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
 {
@@ -320,13 +334,23 @@ static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
 	return vr_die;
 }
 
-static bool die_is_signed_type(Dwarf_Die *tp_die)
+static int die_get_attr_udata(Dwarf_Die *tp_die, unsigned int attr_name,
+			      Dwarf_Word *result)
 {
 	Dwarf_Attribute attr;
+
+	if (dwarf_attr(tp_die, attr_name, &attr) == NULL ||
+	    dwarf_formudata(&attr, result) != 0)
+		return -ENOENT;
+
+	return 0;
+}
+
+static bool die_is_signed_type(Dwarf_Die *tp_die)
+{
 	Dwarf_Word ret;
 
-	if (dwarf_attr(tp_die, DW_AT_encoding, &attr) == NULL ||
-	    dwarf_formudata(&attr, &ret) != 0)
+	if (die_get_attr_udata(tp_die, DW_AT_encoding, &ret))
 		return false;
 
 	return (ret == DW_ATE_signed_char || ret == DW_ATE_signed ||
@@ -335,11 +359,29 @@ static bool die_is_signed_type(Dwarf_Die *tp_die)
 
 static int die_get_byte_size(Dwarf_Die *tp_die)
 {
-	Dwarf_Attribute attr;
 	Dwarf_Word ret;
 
-	if (dwarf_attr(tp_die, DW_AT_byte_size, &attr) == NULL ||
-	    dwarf_formudata(&attr, &ret) != 0)
+	if (die_get_attr_udata(tp_die, DW_AT_byte_size, &ret))
+		return 0;
+
+	return (int)ret;
+}
+
+static int die_get_bit_size(Dwarf_Die *tp_die)
+{
+	Dwarf_Word ret;
+
+	if (die_get_attr_udata(tp_die, DW_AT_bit_size, &ret))
+		return 0;
+
+	return (int)ret;
+}
+
+static int die_get_bit_offset(Dwarf_Die *tp_die)
+{
+	Dwarf_Word ret;
+
+	if (die_get_attr_udata(tp_die, DW_AT_bit_offset, &ret))
 		return 0;
 
 	return (int)ret;
@@ -458,6 +500,151 @@ static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr,
 	return die_find_child(sp_die, __die_find_inline_cb, &addr, die_mem);
 }
 
+/* Walker on lines (Note: line number will not be sorted) */
+typedef int (* line_walk_handler_t) (const char *fname, int lineno,
+				     Dwarf_Addr addr, void *data);
+
+struct __line_walk_param {
+	const char *fname;
+	line_walk_handler_t handler;
+	void *data;
+	int retval;
+};
+
+static int __die_walk_funclines_cb(Dwarf_Die *in_die, void *data)
+{
+	struct __line_walk_param *lw = data;
+	Dwarf_Addr addr;
+	int lineno;
+
+	if (dwarf_tag(in_die) == DW_TAG_inlined_subroutine) {
+		lineno = die_get_call_lineno(in_die);
+		if (lineno > 0 && dwarf_entrypc(in_die, &addr) == 0) {
+			lw->retval = lw->handler(lw->fname, lineno, addr,
+						 lw->data);
+			if (lw->retval != 0)
+				return DIE_FIND_CB_FOUND;
+		}
+	}
+	return DIE_FIND_CB_SIBLING;
+}
+
+/* Walk on lines of blocks included in given DIE */
+static int __die_walk_funclines(Dwarf_Die *sp_die,
+				line_walk_handler_t handler, void *data)
+{
+	struct __line_walk_param lw = {
+		.handler = handler,
+		.data = data,
+		.retval = 0,
+	};
+	Dwarf_Die die_mem;
+	Dwarf_Addr addr;
+	int lineno;
+
+	/* Handle function declaration line */
+	lw.fname = dwarf_decl_file(sp_die);
+	if (lw.fname && dwarf_decl_line(sp_die, &lineno) == 0 &&
+	    dwarf_entrypc(sp_die, &addr) == 0) {
+		lw.retval = handler(lw.fname, lineno, addr, data);
+		if (lw.retval != 0)
+			goto done;
+	}
+	die_find_child(sp_die, __die_walk_funclines_cb, &lw, &die_mem);
+done:
+	return lw.retval;
+}
+
+static int __die_walk_culines_cb(Dwarf_Die *sp_die, void *data)
+{
+	struct __line_walk_param *lw = data;
+
+	lw->retval = __die_walk_funclines(sp_die, lw->handler, lw->data);
+	if (lw->retval != 0)
+		return DWARF_CB_ABORT;
+
+	return DWARF_CB_OK;
+}
+
+/*
+ * Walk on lines inside given PDIE. If the PDIE is subprogram, walk only on
+ * the lines inside the subprogram, otherwise PDIE must be a CU DIE.
+ */
+static int die_walk_lines(Dwarf_Die *pdie, line_walk_handler_t handler,
+			  void *data)
+{
+	Dwarf_Lines *lines;
+	Dwarf_Line *line;
+	Dwarf_Addr addr;
+	const char *fname;
+	int lineno, ret = 0;
+	Dwarf_Die die_mem, *cu_die;
+	size_t nlines, i;
+
+	/* Get the CU die */
+	if (dwarf_tag(pdie) == DW_TAG_subprogram)
+		cu_die = dwarf_diecu(pdie, &die_mem, NULL, NULL);
+	else
+		cu_die = pdie;
+	if (!cu_die) {
+		pr_debug2("Failed to get CU from subprogram\n");
+		return -EINVAL;
+	}
+
+	/* Get lines list in the CU */
+	if (dwarf_getsrclines(cu_die, &lines, &nlines) != 0) {
+		pr_debug2("Failed to get source lines on this CU.\n");
+		return -ENOENT;
+	}
+	pr_debug2("Get %zd lines from this CU\n", nlines);
+
+	/* Walk on the lines on lines list */
+	for (i = 0; i < nlines; i++) {
+		line = dwarf_onesrcline(lines, i);
+		if (line == NULL ||
+		    dwarf_lineno(line, &lineno) != 0 ||
+		    dwarf_lineaddr(line, &addr) != 0) {
+			pr_debug2("Failed to get line info. "
+				  "Possible error in debuginfo.\n");
+			continue;
+		}
+		/* Filter lines based on address */
+		if (pdie != cu_die)
+			/*
+			 * Address filtering
+			 * The line is included in given function, and
+			 * no inline block includes it.
+			 */
+			if (!dwarf_haspc(pdie, addr) ||
+			    die_find_inlinefunc(pdie, addr, &die_mem))
+				continue;
+		/* Get source line */
+		fname = dwarf_linesrc(line, NULL, NULL);
+
+		ret = handler(fname, lineno, addr, data);
+		if (ret != 0)
+			return ret;
+	}
+
+	/*
+	 * Dwarf lines doesn't include function declarations and inlined
+	 * subroutines. We have to check functions list or given function.
+	 */
+	if (pdie != cu_die)
+		ret = __die_walk_funclines(pdie, handler, data);
+	else {
+		struct __line_walk_param param = {
+			.handler = handler,
+			.data = data,
+			.retval = 0,
+		};
+		dwarf_getfuncs(cu_die, __die_walk_culines_cb, &param, 0);
+		ret = param.retval;
+	}
+
+	return ret;
+}
+
 struct __find_variable_param {
 	const char *name;
 	Dwarf_Addr addr;
@@ -669,6 +856,8 @@ static_var:
 	return 0;
 }
 
+#define BYTES_TO_BITS(nb)	((nb) * BITS_PER_LONG / sizeof(long))
+
 static int convert_variable_type(Dwarf_Die *vr_die,
 				 struct probe_trace_arg *tvar,
 				 const char *cast)
@@ -685,6 +874,14 @@ static int convert_variable_type(Dwarf_Die *vr_die,
 		return (tvar->type == NULL) ? -ENOMEM : 0;
 	}
 
+	if (die_get_bit_size(vr_die) != 0) {
+		/* This is a bitfield */
+		ret = snprintf(buf, 16, "b%d@%d/%zd", die_get_bit_size(vr_die),
+				die_get_bit_offset(vr_die),
+				BYTES_TO_BITS(die_get_byte_size(vr_die)));
+		goto formatted;
+	}
+
 	if (die_get_real_type(vr_die, &type) == NULL) {
 		pr_warning("Failed to get a type information of %s.\n",
 			   dwarf_diename(vr_die));
@@ -729,29 +926,31 @@ static int convert_variable_type(Dwarf_Die *vr_die,
 		return (tvar->type == NULL) ? -ENOMEM : 0;
 	}
 
-	ret = die_get_byte_size(&type) * 8;
-	if (ret) {
-		/* Check the bitwidth */
-		if (ret > MAX_BASIC_TYPE_BITS) {
-			pr_info("%s exceeds max-bitwidth."
-				" Cut down to %d bits.\n",
-				dwarf_diename(&type), MAX_BASIC_TYPE_BITS);
-			ret = MAX_BASIC_TYPE_BITS;
-		}
+	ret = BYTES_TO_BITS(die_get_byte_size(&type));
+	if (!ret)
+		/* No size ... try to use default type */
+		return 0;
 
-		ret = snprintf(buf, 16, "%c%d",
-			       die_is_signed_type(&type) ? 's' : 'u', ret);
-		if (ret < 0 || ret >= 16) {
-			if (ret >= 16)
-				ret = -E2BIG;
-			pr_warning("Failed to convert variable type: %s\n",
-				   strerror(-ret));
-			return ret;
-		}
-		tvar->type = strdup(buf);
-		if (tvar->type == NULL)
-			return -ENOMEM;
+	/* Check the bitwidth */
+	if (ret > MAX_BASIC_TYPE_BITS) {
+		pr_info("%s exceeds max-bitwidth. Cut down to %d bits.\n",
+			dwarf_diename(&type), MAX_BASIC_TYPE_BITS);
+		ret = MAX_BASIC_TYPE_BITS;
+	}
+	ret = snprintf(buf, 16, "%c%d",
+		       die_is_signed_type(&type) ? 's' : 'u', ret);
+
+formatted:
+	if (ret < 0 || ret >= 16) {
+		if (ret >= 16)
+			ret = -E2BIG;
+		pr_warning("Failed to convert variable type: %s\n",
+			   strerror(-ret));
+		return ret;
 	}
+	tvar->type = strdup(buf);
+	if (tvar->type == NULL)
+		return -ENOMEM;
 	return 0;
 }
 
@@ -1050,157 +1249,102 @@ static int call_probe_finder(Dwarf_Die *sp_die, struct probe_finder *pf)
 	return ret;
 }
 
-/* Find probe point from its line number */
-static int find_probe_point_by_line(struct probe_finder *pf)
+static int probe_point_line_walker(const char *fname, int lineno,
+				   Dwarf_Addr addr, void *data)
 {
-	Dwarf_Lines *lines;
-	Dwarf_Line *line;
-	size_t nlines, i;
-	Dwarf_Addr addr;
-	int lineno;
-	int ret = 0;
-
-	if (dwarf_getsrclines(&pf->cu_die, &lines, &nlines) != 0) {
-		pr_warning("No source lines found.\n");
-		return -ENOENT;
-	}
+	struct probe_finder *pf = data;
+	int ret;
 
-	for (i = 0; i < nlines && ret == 0; i++) {
-		line = dwarf_onesrcline(lines, i);
-		if (dwarf_lineno(line, &lineno) != 0 ||
-		    lineno != pf->lno)
-			continue;
+	if (lineno != pf->lno || strtailcmp(fname, pf->fname) != 0)
+		return 0;
 
-		/* TODO: Get fileno from line, but how? */
-		if (strtailcmp(dwarf_linesrc(line, NULL, NULL), pf->fname) != 0)
-			continue;
+	pf->addr = addr;
+	ret = call_probe_finder(NULL, pf);
 
-		if (dwarf_lineaddr(line, &addr) != 0) {
-			pr_warning("Failed to get the address of the line.\n");
-			return -ENOENT;
-		}
-		pr_debug("Probe line found: line[%d]:%d addr:0x%jx\n",
-			 (int)i, lineno, (uintmax_t)addr);
-		pf->addr = addr;
+	/* Continue if no error, because the line will be in inline function */
+	return ret < 0 ? ret : 0;
+}
 
-		ret = call_probe_finder(NULL, pf);
-		/* Continuing, because target line might be inlined. */
-	}
-	return ret;
+/* Find probe point from its line number */
+static int find_probe_point_by_line(struct probe_finder *pf)
+{
+	return die_walk_lines(&pf->cu_die, probe_point_line_walker, pf);
 }
 
 /* Find lines which match lazy pattern */
 static int find_lazy_match_lines(struct list_head *head,
 				 const char *fname, const char *pat)
 {
-	char *fbuf, *p1, *p2;
-	int fd, line, nlines = -1;
-	struct stat st;
-
-	fd = open(fname, O_RDONLY);
-	if (fd < 0) {
-		pr_warning("Failed to open %s: %s\n", fname, strerror(-fd));
+	FILE *fp;
+	char *line = NULL;
+	size_t line_len;
+	ssize_t len;
+	int count = 0, linenum = 1;
+
+	fp = fopen(fname, "r");
+	if (!fp) {
+		pr_warning("Failed to open %s: %s\n", fname, strerror(errno));
 		return -errno;
 	}
 
-	if (fstat(fd, &st) < 0) {
-		pr_warning("Failed to get the size of %s: %s\n",
-			   fname, strerror(errno));
-		nlines = -errno;
-		goto out_close;
-	}
-
-	nlines = -ENOMEM;
-	fbuf = malloc(st.st_size + 2);
-	if (fbuf == NULL)
-		goto out_close;
-	if (read(fd, fbuf, st.st_size) < 0) {
-		pr_warning("Failed to read %s: %s\n", fname, strerror(errno));
-		nlines = -errno;
-		goto out_free_fbuf;
-	}
-	fbuf[st.st_size] = '\n';	/* Dummy line */
-	fbuf[st.st_size + 1] = '\0';
-	p1 = fbuf;
-	line = 1;
-	nlines = 0;
-	while ((p2 = strchr(p1, '\n')) != NULL) {
-		*p2 = '\0';
-		if (strlazymatch(p1, pat)) {
-			line_list__add_line(head, line);
-			nlines++;
+	while ((len = getline(&line, &line_len, fp)) > 0) {
+
+		if (line[len - 1] == '\n')
+			line[len - 1] = '\0';
+
+		if (strlazymatch(line, pat)) {
+			line_list__add_line(head, linenum);
+			count++;
 		}
-		line++;
-		p1 = p2 + 1;
+		linenum++;
 	}
-out_free_fbuf:
-	free(fbuf);
-out_close:
-	close(fd);
-	return nlines;
+
+	if (ferror(fp))
+		count = -errno;
+	free(line);
+	fclose(fp);
+
+	if (count == 0)
+		pr_debug("No matched lines found in %s.\n", fname);
+	return count;
+}
+
+static int probe_point_lazy_walker(const char *fname, int lineno,
+				   Dwarf_Addr addr, void *data)
+{
+	struct probe_finder *pf = data;
+	int ret;
+
+	if (!line_list__has_line(&pf->lcache, lineno) ||
+	    strtailcmp(fname, pf->fname) != 0)
+		return 0;
+
+	pr_debug("Probe line found: line:%d addr:0x%llx\n",
+		 lineno, (unsigned long long)addr);
+	pf->addr = addr;
+	ret = call_probe_finder(NULL, pf);
+
+	/*
+	 * Continue if no error, because the lazy pattern will match
+	 * to other lines
+	 */
+	return ret < 0 ? ret : 0;
 }
 
 /* Find probe points from lazy pattern  */
 static int find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf)
 {
-	Dwarf_Lines *lines;
-	Dwarf_Line *line;
-	size_t nlines, i;
-	Dwarf_Addr addr;
-	Dwarf_Die die_mem;
-	int lineno;
 	int ret = 0;
 
 	if (list_empty(&pf->lcache)) {
 		/* Matching lazy line pattern */
 		ret = find_lazy_match_lines(&pf->lcache, pf->fname,
 					    pf->pev->point.lazy_line);
-		if (ret == 0) {
-			pr_debug("No matched lines found in %s.\n", pf->fname);
-			return 0;
-		} else if (ret < 0)
+		if (ret <= 0)
 			return ret;
 	}
 
-	if (dwarf_getsrclines(&pf->cu_die, &lines, &nlines) != 0) {
-		pr_warning("No source lines found.\n");
-		return -ENOENT;
-	}
-
-	for (i = 0; i < nlines && ret >= 0; i++) {
-		line = dwarf_onesrcline(lines, i);
-
-		if (dwarf_lineno(line, &lineno) != 0 ||
-		    !line_list__has_line(&pf->lcache, lineno))
-			continue;
-
-		/* TODO: Get fileno from line, but how? */
-		if (strtailcmp(dwarf_linesrc(line, NULL, NULL), pf->fname) != 0)
-			continue;
-
-		if (dwarf_lineaddr(line, &addr) != 0) {
-			pr_debug("Failed to get the address of line %d.\n",
-				 lineno);
-			continue;
-		}
-		if (sp_die) {
-			/* Address filtering 1: does sp_die include addr? */
-			if (!dwarf_haspc(sp_die, addr))
-				continue;
-			/* Address filtering 2: No child include addr? */
-			if (die_find_inlinefunc(sp_die, addr, &die_mem))
-				continue;
-		}
-
-		pr_debug("Probe line found: line[%d]:%d addr:0x%llx\n",
-			 (int)i, lineno, (unsigned long long)addr);
-		pf->addr = addr;
-
-		ret = call_probe_finder(sp_die, pf);
-		/* Continuing, because target line might be inlined. */
-	}
-	/* TODO: deallocate lines, but how? */
-	return ret;
+	return die_walk_lines(sp_die, probe_point_lazy_walker, pf);
 }
 
 /* Callback parameter with return value */
@@ -1318,8 +1462,7 @@ static int find_probes(int fd, struct probe_finder *pf)
 	off = 0;
 	line_list__init(&pf->lcache);
 	/* Loop on CUs (Compilation Unit) */
-	while (!dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL) &&
-	       ret >= 0) {
+	while (!dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL)) {
 		/* Get the DIE(Debugging Information Entry) of this CU */
 		diep = dwarf_offdie(dbg, off + cuhl, &pf->cu_die);
 		if (!diep)
@@ -1340,6 +1483,8 @@ static int find_probes(int fd, struct probe_finder *pf)
 				pf->lno = pp->line;
 				ret = find_probe_point_by_line(pf);
 			}
+			if (ret < 0)
+				break;
 		}
 		off = noff;
 	}
@@ -1644,91 +1789,28 @@ static int line_range_add_line(const char *src, unsigned int lineno,
 	return line_list__add_line(&lr->line_list, lineno);
 }
 
-/* Search function declaration lines */
-static int line_range_funcdecl_cb(Dwarf_Die *sp_die, void *data)
+static int line_range_walk_cb(const char *fname, int lineno,
+			      Dwarf_Addr addr __used,
+			      void *data)
 {
-	struct dwarf_callback_param *param = data;
-	struct line_finder *lf = param->data;
-	const char *src;
-	int lineno;
+	struct line_finder *lf = data;
 
-	src = dwarf_decl_file(sp_die);
-	if (src && strtailcmp(src, lf->fname) != 0)
-		return DWARF_CB_OK;
-
-	if (dwarf_decl_line(sp_die, &lineno) != 0 ||
+	if ((strtailcmp(fname, lf->fname) != 0) ||
 	    (lf->lno_s > lineno || lf->lno_e < lineno))
-		return DWARF_CB_OK;
+		return 0;
 
-	param->retval = line_range_add_line(src, lineno, lf->lr);
-	if (param->retval < 0)
-		return DWARF_CB_ABORT;
-	return DWARF_CB_OK;
-}
+	if (line_range_add_line(fname, lineno, lf->lr) < 0)
+		return -EINVAL;
 
-static int find_line_range_func_decl_lines(struct line_finder *lf)
-{
-	struct dwarf_callback_param param = {.data = (void *)lf, .retval = 0};
-	dwarf_getfuncs(&lf->cu_die, line_range_funcdecl_cb, &param, 0);
-	return param.retval;
+	return 0;
 }
 
 /* Find line range from its line number */
 static int find_line_range_by_line(Dwarf_Die *sp_die, struct line_finder *lf)
 {
-	Dwarf_Lines *lines;
-	Dwarf_Line *line;
-	size_t nlines, i;
-	Dwarf_Addr addr;
-	int lineno, ret = 0;
-	const char *src;
-	Dwarf_Die die_mem;
-
-	line_list__init(&lf->lr->line_list);
-	if (dwarf_getsrclines(&lf->cu_die, &lines, &nlines) != 0) {
-		pr_warning("No source lines found.\n");
-		return -ENOENT;
-	}
-
-	/* Search probable lines on lines list */
-	for (i = 0; i < nlines; i++) {
-		line = dwarf_onesrcline(lines, i);
-		if (dwarf_lineno(line, &lineno) != 0 ||
-		    (lf->lno_s > lineno || lf->lno_e < lineno))
-			continue;
-
-		if (sp_die) {
-			/* Address filtering 1: does sp_die include addr? */
-			if (dwarf_lineaddr(line, &addr) != 0 ||
-			    !dwarf_haspc(sp_die, addr))
-				continue;
-
-			/* Address filtering 2: No child include addr? */
-			if (die_find_inlinefunc(sp_die, addr, &die_mem))
-				continue;
-		}
-
-		/* TODO: Get fileno from line, but how? */
-		src = dwarf_linesrc(line, NULL, NULL);
-		if (strtailcmp(src, lf->fname) != 0)
-			continue;
-
-		ret = line_range_add_line(src, lineno, lf->lr);
-		if (ret < 0)
-			return ret;
-	}
+	int ret;
 
-	/*
-	 * Dwarf lines doesn't include function declarations. We have to
-	 * check functions list or given function.
-	 */
-	if (sp_die) {
-		src = dwarf_decl_file(sp_die);
-		if (src && dwarf_decl_line(sp_die, &lineno) == 0 &&
-		    (lf->lno_s <= lineno && lf->lno_e >= lineno))
-			ret = line_range_add_line(src, lineno, lf->lr);
-	} else
-		ret = find_line_range_func_decl_lines(lf);
+	ret = die_walk_lines(sp_die ?: &lf->cu_die, line_range_walk_cb, lf);
 
 	/* Update status */
 	if (ret >= 0)
@@ -1758,9 +1840,6 @@ static int line_range_search_cb(Dwarf_Die *sp_die, void *data)
 	struct line_finder *lf = param->data;
 	struct line_range *lr = lf->lr;
 
-	pr_debug("find (%llx) %s\n",
-		 (unsigned long long)dwarf_dieoffset(sp_die),
-		 dwarf_diename(sp_die));
 	if (dwarf_tag(sp_die) == DW_TAG_subprogram &&
 	    die_compare_name(sp_die, lr->function)) {
 		lf->fname = dwarf_decl_file(sp_die);
diff --git a/tools/perf/util/python.c b/tools/perf/util/python.c
new file mode 100644
index 000000000000..a9f2d7e1204d
--- /dev/null
+++ b/tools/perf/util/python.c
@@ -0,0 +1,896 @@
+#include <Python.h>
+#include <structmember.h>
+#include <inttypes.h>
+#include <poll.h>
+#include "evlist.h"
+#include "evsel.h"
+#include "event.h"
+#include "cpumap.h"
+#include "thread_map.h"
+
+/* Define PyVarObject_HEAD_INIT for python 2.5 */
+#ifndef PyVarObject_HEAD_INIT
+# define PyVarObject_HEAD_INIT(type, size) PyObject_HEAD_INIT(type) size,
+#endif
+
+struct throttle_event {
+	struct perf_event_header header;
+	u64			 time;
+	u64			 id;
+	u64			 stream_id;
+};
+
+PyMODINIT_FUNC initperf(void);
+
+#define member_def(type, member, ptype, help) \
+	{ #member, ptype, \
+	  offsetof(struct pyrf_event, event) + offsetof(struct type, member), \
+	  0, help }
+
+#define sample_member_def(name, member, ptype, help) \
+	{ #name, ptype, \
+	  offsetof(struct pyrf_event, sample) + offsetof(struct perf_sample, member), \
+	  0, help }
+
+struct pyrf_event {
+	PyObject_HEAD
+	struct perf_sample sample;
+	union perf_event   event;
+};
+
+#define sample_members \
+	sample_member_def(sample_ip, ip, T_ULONGLONG, "event type"),			 \
+	sample_member_def(sample_pid, pid, T_INT, "event pid"),			 \
+	sample_member_def(sample_tid, tid, T_INT, "event tid"),			 \
+	sample_member_def(sample_time, time, T_ULONGLONG, "event timestamp"),		 \
+	sample_member_def(sample_addr, addr, T_ULONGLONG, "event addr"),		 \
+	sample_member_def(sample_id, id, T_ULONGLONG, "event id"),			 \
+	sample_member_def(sample_stream_id, stream_id, T_ULONGLONG, "event stream id"), \
+	sample_member_def(sample_period, period, T_ULONGLONG, "event period"),		 \
+	sample_member_def(sample_cpu, cpu, T_UINT, "event cpu"),
+
+static char pyrf_mmap_event__doc[] = PyDoc_STR("perf mmap event object.");
+
+static PyMemberDef pyrf_mmap_event__members[] = {
+	sample_members
+	member_def(perf_event_header, type, T_UINT, "event type"),
+	member_def(mmap_event, pid, T_UINT, "event pid"),
+	member_def(mmap_event, tid, T_UINT, "event tid"),
+	member_def(mmap_event, start, T_ULONGLONG, "start of the map"),
+	member_def(mmap_event, len, T_ULONGLONG, "map length"),
+	member_def(mmap_event, pgoff, T_ULONGLONG, "page offset"),
+	member_def(mmap_event, filename, T_STRING_INPLACE, "backing store"),
+	{ .name = NULL, },
+};
+
+static PyObject *pyrf_mmap_event__repr(struct pyrf_event *pevent)
+{
+	PyObject *ret;
+	char *s;
+
+	if (asprintf(&s, "{ type: mmap, pid: %u, tid: %u, start: %#" PRIx64 ", "
+			 "length: %#" PRIx64 ", offset: %#" PRIx64 ", "
+			 "filename: %s }",
+		     pevent->event.mmap.pid, pevent->event.mmap.tid,
+		     pevent->event.mmap.start, pevent->event.mmap.len,
+		     pevent->event.mmap.pgoff, pevent->event.mmap.filename) < 0) {
+		ret = PyErr_NoMemory();
+	} else {
+		ret = PyString_FromString(s);
+		free(s);
+	}
+	return ret;
+}
+
+static PyTypeObject pyrf_mmap_event__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.mmap_event",
+	.tp_basicsize	= sizeof(struct pyrf_event),
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_mmap_event__doc,
+	.tp_members	= pyrf_mmap_event__members,
+	.tp_repr	= (reprfunc)pyrf_mmap_event__repr,
+};
+
+static char pyrf_task_event__doc[] = PyDoc_STR("perf task (fork/exit) event object.");
+
+static PyMemberDef pyrf_task_event__members[] = {
+	sample_members
+	member_def(perf_event_header, type, T_UINT, "event type"),
+	member_def(fork_event, pid, T_UINT, "event pid"),
+	member_def(fork_event, ppid, T_UINT, "event ppid"),
+	member_def(fork_event, tid, T_UINT, "event tid"),
+	member_def(fork_event, ptid, T_UINT, "event ptid"),
+	member_def(fork_event, time, T_ULONGLONG, "timestamp"),
+	{ .name = NULL, },
+};
+
+static PyObject *pyrf_task_event__repr(struct pyrf_event *pevent)
+{
+	return PyString_FromFormat("{ type: %s, pid: %u, ppid: %u, tid: %u, "
+				   "ptid: %u, time: %" PRIu64 "}",
+				   pevent->event.header.type == PERF_RECORD_FORK ? "fork" : "exit",
+				   pevent->event.fork.pid,
+				   pevent->event.fork.ppid,
+				   pevent->event.fork.tid,
+				   pevent->event.fork.ptid,
+				   pevent->event.fork.time);
+}
+
+static PyTypeObject pyrf_task_event__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.task_event",
+	.tp_basicsize	= sizeof(struct pyrf_event),
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_task_event__doc,
+	.tp_members	= pyrf_task_event__members,
+	.tp_repr	= (reprfunc)pyrf_task_event__repr,
+};
+
+static char pyrf_comm_event__doc[] = PyDoc_STR("perf comm event object.");
+
+static PyMemberDef pyrf_comm_event__members[] = {
+	sample_members
+	member_def(perf_event_header, type, T_UINT, "event type"),
+	member_def(comm_event, pid, T_UINT, "event pid"),
+	member_def(comm_event, tid, T_UINT, "event tid"),
+	member_def(comm_event, comm, T_STRING_INPLACE, "process name"),
+	{ .name = NULL, },
+};
+
+static PyObject *pyrf_comm_event__repr(struct pyrf_event *pevent)
+{
+	return PyString_FromFormat("{ type: comm, pid: %u, tid: %u, comm: %s }",
+				   pevent->event.comm.pid,
+				   pevent->event.comm.tid,
+				   pevent->event.comm.comm);
+}
+
+static PyTypeObject pyrf_comm_event__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.comm_event",
+	.tp_basicsize	= sizeof(struct pyrf_event),
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_comm_event__doc,
+	.tp_members	= pyrf_comm_event__members,
+	.tp_repr	= (reprfunc)pyrf_comm_event__repr,
+};
+
+static char pyrf_throttle_event__doc[] = PyDoc_STR("perf throttle event object.");
+
+static PyMemberDef pyrf_throttle_event__members[] = {
+	sample_members
+	member_def(perf_event_header, type, T_UINT, "event type"),
+	member_def(throttle_event, time, T_ULONGLONG, "timestamp"),
+	member_def(throttle_event, id, T_ULONGLONG, "event id"),
+	member_def(throttle_event, stream_id, T_ULONGLONG, "event stream id"),
+	{ .name = NULL, },
+};
+
+static PyObject *pyrf_throttle_event__repr(struct pyrf_event *pevent)
+{
+	struct throttle_event *te = (struct throttle_event *)(&pevent->event.header + 1);
+
+	return PyString_FromFormat("{ type: %sthrottle, time: %" PRIu64 ", id: %" PRIu64
+				   ", stream_id: %" PRIu64 " }",
+				   pevent->event.header.type == PERF_RECORD_THROTTLE ? "" : "un",
+				   te->time, te->id, te->stream_id);
+}
+
+static PyTypeObject pyrf_throttle_event__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.throttle_event",
+	.tp_basicsize	= sizeof(struct pyrf_event),
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_throttle_event__doc,
+	.tp_members	= pyrf_throttle_event__members,
+	.tp_repr	= (reprfunc)pyrf_throttle_event__repr,
+};
+
+static int pyrf_event__setup_types(void)
+{
+	int err;
+	pyrf_mmap_event__type.tp_new =
+	pyrf_task_event__type.tp_new =
+	pyrf_comm_event__type.tp_new =
+	pyrf_throttle_event__type.tp_new = PyType_GenericNew;
+	err = PyType_Ready(&pyrf_mmap_event__type);
+	if (err < 0)
+		goto out;
+	err = PyType_Ready(&pyrf_task_event__type);
+	if (err < 0)
+		goto out;
+	err = PyType_Ready(&pyrf_comm_event__type);
+	if (err < 0)
+		goto out;
+	err = PyType_Ready(&pyrf_throttle_event__type);
+	if (err < 0)
+		goto out;
+out:
+	return err;
+}
+
+static PyTypeObject *pyrf_event__type[] = {
+	[PERF_RECORD_MMAP]	 = &pyrf_mmap_event__type,
+	[PERF_RECORD_LOST]	 = &pyrf_mmap_event__type,
+	[PERF_RECORD_COMM]	 = &pyrf_comm_event__type,
+	[PERF_RECORD_EXIT]	 = &pyrf_task_event__type,
+	[PERF_RECORD_THROTTLE]	 = &pyrf_throttle_event__type,
+	[PERF_RECORD_UNTHROTTLE] = &pyrf_throttle_event__type,
+	[PERF_RECORD_FORK]	 = &pyrf_task_event__type,
+	[PERF_RECORD_READ]	 = &pyrf_mmap_event__type,
+	[PERF_RECORD_SAMPLE]	 = &pyrf_mmap_event__type,
+};
+
+static PyObject *pyrf_event__new(union perf_event *event)
+{
+	struct pyrf_event *pevent;
+	PyTypeObject *ptype;
+
+	if (event->header.type < PERF_RECORD_MMAP ||
+	    event->header.type > PERF_RECORD_SAMPLE)
+		return NULL;
+
+	ptype = pyrf_event__type[event->header.type];
+	pevent = PyObject_New(struct pyrf_event, ptype);
+	if (pevent != NULL)
+		memcpy(&pevent->event, event, event->header.size);
+	return (PyObject *)pevent;
+}
+
+struct pyrf_cpu_map {
+	PyObject_HEAD
+
+	struct cpu_map *cpus;
+};
+
+static int pyrf_cpu_map__init(struct pyrf_cpu_map *pcpus,
+			      PyObject *args, PyObject *kwargs)
+{
+	static char *kwlist[] = { "cpustr", NULL, NULL, };
+	char *cpustr = NULL;
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|s",
+					 kwlist, &cpustr))
+		return -1;
+
+	pcpus->cpus = cpu_map__new(cpustr);
+	if (pcpus->cpus == NULL)
+		return -1;
+	return 0;
+}
+
+static void pyrf_cpu_map__delete(struct pyrf_cpu_map *pcpus)
+{
+	cpu_map__delete(pcpus->cpus);
+	pcpus->ob_type->tp_free((PyObject*)pcpus);
+}
+
+static Py_ssize_t pyrf_cpu_map__length(PyObject *obj)
+{
+	struct pyrf_cpu_map *pcpus = (void *)obj;
+
+	return pcpus->cpus->nr;
+}
+
+static PyObject *pyrf_cpu_map__item(PyObject *obj, Py_ssize_t i)
+{
+	struct pyrf_cpu_map *pcpus = (void *)obj;
+
+	if (i >= pcpus->cpus->nr)
+		return NULL;
+
+	return Py_BuildValue("i", pcpus->cpus->map[i]);
+}
+
+static PySequenceMethods pyrf_cpu_map__sequence_methods = {
+	.sq_length = pyrf_cpu_map__length,
+	.sq_item   = pyrf_cpu_map__item,
+};
+
+static char pyrf_cpu_map__doc[] = PyDoc_STR("cpu map object.");
+
+static PyTypeObject pyrf_cpu_map__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.cpu_map",
+	.tp_basicsize	= sizeof(struct pyrf_cpu_map),
+	.tp_dealloc	= (destructor)pyrf_cpu_map__delete,
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_cpu_map__doc,
+	.tp_as_sequence	= &pyrf_cpu_map__sequence_methods,
+	.tp_init	= (initproc)pyrf_cpu_map__init,
+};
+
+static int pyrf_cpu_map__setup_types(void)
+{
+	pyrf_cpu_map__type.tp_new = PyType_GenericNew;
+	return PyType_Ready(&pyrf_cpu_map__type);
+}
+
+struct pyrf_thread_map {
+	PyObject_HEAD
+
+	struct thread_map *threads;
+};
+
+static int pyrf_thread_map__init(struct pyrf_thread_map *pthreads,
+				 PyObject *args, PyObject *kwargs)
+{
+	static char *kwlist[] = { "pid", "tid", NULL, NULL, };
+	int pid = -1, tid = -1;
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|ii",
+					 kwlist, &pid, &tid))
+		return -1;
+
+	pthreads->threads = thread_map__new(pid, tid);
+	if (pthreads->threads == NULL)
+		return -1;
+	return 0;
+}
+
+static void pyrf_thread_map__delete(struct pyrf_thread_map *pthreads)
+{
+	thread_map__delete(pthreads->threads);
+	pthreads->ob_type->tp_free((PyObject*)pthreads);
+}
+
+static Py_ssize_t pyrf_thread_map__length(PyObject *obj)
+{
+	struct pyrf_thread_map *pthreads = (void *)obj;
+
+	return pthreads->threads->nr;
+}
+
+static PyObject *pyrf_thread_map__item(PyObject *obj, Py_ssize_t i)
+{
+	struct pyrf_thread_map *pthreads = (void *)obj;
+
+	if (i >= pthreads->threads->nr)
+		return NULL;
+
+	return Py_BuildValue("i", pthreads->threads->map[i]);
+}
+
+static PySequenceMethods pyrf_thread_map__sequence_methods = {
+	.sq_length = pyrf_thread_map__length,
+	.sq_item   = pyrf_thread_map__item,
+};
+
+static char pyrf_thread_map__doc[] = PyDoc_STR("thread map object.");
+
+static PyTypeObject pyrf_thread_map__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.thread_map",
+	.tp_basicsize	= sizeof(struct pyrf_thread_map),
+	.tp_dealloc	= (destructor)pyrf_thread_map__delete,
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_thread_map__doc,
+	.tp_as_sequence	= &pyrf_thread_map__sequence_methods,
+	.tp_init	= (initproc)pyrf_thread_map__init,
+};
+
+static int pyrf_thread_map__setup_types(void)
+{
+	pyrf_thread_map__type.tp_new = PyType_GenericNew;
+	return PyType_Ready(&pyrf_thread_map__type);
+}
+
+struct pyrf_evsel {
+	PyObject_HEAD
+
+	struct perf_evsel evsel;
+};
+
+static int pyrf_evsel__init(struct pyrf_evsel *pevsel,
+			    PyObject *args, PyObject *kwargs)
+{
+	struct perf_event_attr attr = {
+		.type = PERF_TYPE_HARDWARE,
+		.config = PERF_COUNT_HW_CPU_CYCLES,
+		.sample_type = PERF_SAMPLE_PERIOD | PERF_SAMPLE_TID,
+	};
+	static char *kwlist[] = {
+		"type",
+		"config",
+		"sample_freq",
+		"sample_period",
+		"sample_type",
+		"read_format",
+		"disabled",
+		"inherit",
+		"pinned",
+		"exclusive",
+		"exclude_user",
+		"exclude_kernel",
+		"exclude_hv",
+		"exclude_idle",
+		"mmap",
+		"comm",
+		"freq",
+		"inherit_stat",
+		"enable_on_exec",
+		"task",
+		"watermark",
+		"precise_ip",
+		"mmap_data",
+		"sample_id_all",
+		"wakeup_events",
+		"bp_type",
+		"bp_addr",
+		"bp_len", NULL, NULL, };
+	u64 sample_period = 0;
+	u32 disabled = 0,
+	    inherit = 0,
+	    pinned = 0,
+	    exclusive = 0,
+	    exclude_user = 0,
+	    exclude_kernel = 0,
+	    exclude_hv = 0,
+	    exclude_idle = 0,
+	    mmap = 0,
+	    comm = 0,
+	    freq = 1,
+	    inherit_stat = 0,
+	    enable_on_exec = 0,
+	    task = 0,
+	    watermark = 0,
+	    precise_ip = 0,
+	    mmap_data = 0,
+	    sample_id_all = 1;
+	int idx = 0;
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs,
+					 "|iKiKKiiiiiiiiiiiiiiiiiiiiiKK", kwlist,
+					 &attr.type, &attr.config, &attr.sample_freq,
+					 &sample_period, &attr.sample_type,
+					 &attr.read_format, &disabled, &inherit,
+					 &pinned, &exclusive, &exclude_user,
+					 &exclude_kernel, &exclude_hv, &exclude_idle,
+					 &mmap, &comm, &freq, &inherit_stat,
+					 &enable_on_exec, &task, &watermark,
+					 &precise_ip, &mmap_data, &sample_id_all,
+					 &attr.wakeup_events, &attr.bp_type,
+					 &attr.bp_addr, &attr.bp_len, &idx))
+		return -1;
+
+	/* union... */
+	if (sample_period != 0) {
+		if (attr.sample_freq != 0)
+			return -1; /* FIXME: throw right exception */
+		attr.sample_period = sample_period;
+	}
+
+	/* Bitfields */
+	attr.disabled	    = disabled;
+	attr.inherit	    = inherit;
+	attr.pinned	    = pinned;
+	attr.exclusive	    = exclusive;
+	attr.exclude_user   = exclude_user;
+	attr.exclude_kernel = exclude_kernel;
+	attr.exclude_hv	    = exclude_hv;
+	attr.exclude_idle   = exclude_idle;
+	attr.mmap	    = mmap;
+	attr.comm	    = comm;
+	attr.freq	    = freq;
+	attr.inherit_stat   = inherit_stat;
+	attr.enable_on_exec = enable_on_exec;
+	attr.task	    = task;
+	attr.watermark	    = watermark;
+	attr.precise_ip	    = precise_ip;
+	attr.mmap_data	    = mmap_data;
+	attr.sample_id_all  = sample_id_all;
+
+	perf_evsel__init(&pevsel->evsel, &attr, idx);
+	return 0;
+}
+
+static void pyrf_evsel__delete(struct pyrf_evsel *pevsel)
+{
+	perf_evsel__exit(&pevsel->evsel);
+	pevsel->ob_type->tp_free((PyObject*)pevsel);
+}
+
+static PyObject *pyrf_evsel__open(struct pyrf_evsel *pevsel,
+				  PyObject *args, PyObject *kwargs)
+{
+	struct perf_evsel *evsel = &pevsel->evsel;
+	struct cpu_map *cpus = NULL;
+	struct thread_map *threads = NULL;
+	PyObject *pcpus = NULL, *pthreads = NULL;
+	int group = 0, overwrite = 0;
+	static char *kwlist[] = {"cpus", "threads", "group", "overwrite", NULL, NULL};
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|OOii", kwlist,
+					 &pcpus, &pthreads, &group, &overwrite))
+		return NULL;
+
+	if (pthreads != NULL)
+		threads = ((struct pyrf_thread_map *)pthreads)->threads;
+
+	if (pcpus != NULL)
+		cpus = ((struct pyrf_cpu_map *)pcpus)->cpus;
+
+	if (perf_evsel__open(evsel, cpus, threads, group, overwrite) < 0) {
+		PyErr_SetFromErrno(PyExc_OSError);
+		return NULL;
+	}
+
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
+static PyMethodDef pyrf_evsel__methods[] = {
+	{
+		.ml_name  = "open",
+		.ml_meth  = (PyCFunction)pyrf_evsel__open,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("open the event selector file descriptor table.")
+	},
+	{ .ml_name = NULL, }
+};
+
+static char pyrf_evsel__doc[] = PyDoc_STR("perf event selector list object.");
+
+static PyTypeObject pyrf_evsel__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.evsel",
+	.tp_basicsize	= sizeof(struct pyrf_evsel),
+	.tp_dealloc	= (destructor)pyrf_evsel__delete,
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_doc		= pyrf_evsel__doc,
+	.tp_methods	= pyrf_evsel__methods,
+	.tp_init	= (initproc)pyrf_evsel__init,
+};
+
+static int pyrf_evsel__setup_types(void)
+{
+	pyrf_evsel__type.tp_new = PyType_GenericNew;
+	return PyType_Ready(&pyrf_evsel__type);
+}
+
+struct pyrf_evlist {
+	PyObject_HEAD
+
+	struct perf_evlist evlist;
+};
+
+static int pyrf_evlist__init(struct pyrf_evlist *pevlist,
+			     PyObject *args, PyObject *kwargs __used)
+{
+	PyObject *pcpus = NULL, *pthreads = NULL;
+	struct cpu_map *cpus;
+	struct thread_map *threads;
+
+	if (!PyArg_ParseTuple(args, "OO", &pcpus, &pthreads))
+		return -1;
+
+	threads = ((struct pyrf_thread_map *)pthreads)->threads;
+	cpus = ((struct pyrf_cpu_map *)pcpus)->cpus;
+	perf_evlist__init(&pevlist->evlist, cpus, threads);
+	return 0;
+}
+
+static void pyrf_evlist__delete(struct pyrf_evlist *pevlist)
+{
+	perf_evlist__exit(&pevlist->evlist);
+	pevlist->ob_type->tp_free((PyObject*)pevlist);
+}
+
+static PyObject *pyrf_evlist__mmap(struct pyrf_evlist *pevlist,
+				   PyObject *args, PyObject *kwargs)
+{
+	struct perf_evlist *evlist = &pevlist->evlist;
+	static char *kwlist[] = {"pages", "overwrite",
+				  NULL, NULL};
+	int pages = 128, overwrite = false;
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|ii", kwlist,
+					 &pages, &overwrite))
+		return NULL;
+
+	if (perf_evlist__mmap(evlist, pages, overwrite) < 0) {
+		PyErr_SetFromErrno(PyExc_OSError);
+		return NULL;
+	}
+
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
+static PyObject *pyrf_evlist__poll(struct pyrf_evlist *pevlist,
+				   PyObject *args, PyObject *kwargs)
+{
+	struct perf_evlist *evlist = &pevlist->evlist;
+	static char *kwlist[] = {"timeout", NULL, NULL};
+	int timeout = -1, n;
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|i", kwlist, &timeout))
+		return NULL;
+
+	n = poll(evlist->pollfd, evlist->nr_fds, timeout);
+	if (n < 0) {
+		PyErr_SetFromErrno(PyExc_OSError);
+		return NULL;
+	}
+
+	return Py_BuildValue("i", n);
+}
+
+static PyObject *pyrf_evlist__get_pollfd(struct pyrf_evlist *pevlist,
+					 PyObject *args __used, PyObject *kwargs __used)
+{
+	struct perf_evlist *evlist = &pevlist->evlist;
+        PyObject *list = PyList_New(0);
+	int i;
+
+	for (i = 0; i < evlist->nr_fds; ++i) {
+		PyObject *file;
+		FILE *fp = fdopen(evlist->pollfd[i].fd, "r");
+
+		if (fp == NULL)
+			goto free_list;
+
+		file = PyFile_FromFile(fp, "perf", "r", NULL);
+		if (file == NULL)
+			goto free_list;
+
+		if (PyList_Append(list, file) != 0) {
+			Py_DECREF(file);
+			goto free_list;
+		}
+			
+		Py_DECREF(file);
+	}
+
+	return list;
+free_list:
+	return PyErr_NoMemory();
+}
+
+
+static PyObject *pyrf_evlist__add(struct pyrf_evlist *pevlist,
+				  PyObject *args, PyObject *kwargs __used)
+{
+	struct perf_evlist *evlist = &pevlist->evlist;
+	PyObject *pevsel;
+	struct perf_evsel *evsel;
+
+	if (!PyArg_ParseTuple(args, "O", &pevsel))
+		return NULL;
+
+	Py_INCREF(pevsel);
+	evsel = &((struct pyrf_evsel *)pevsel)->evsel;
+	evsel->idx = evlist->nr_entries;
+	perf_evlist__add(evlist, evsel);
+
+	return Py_BuildValue("i", evlist->nr_entries);
+}
+
+static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,
+					  PyObject *args, PyObject *kwargs)
+{
+	struct perf_evlist *evlist = &pevlist->evlist;
+	union perf_event *event;
+	int sample_id_all = 1, cpu;
+	static char *kwlist[] = {"sample_id_all", NULL, NULL};
+
+	if (!PyArg_ParseTupleAndKeywords(args, kwargs, "i|i", kwlist,
+					 &cpu, &sample_id_all))
+		return NULL;
+
+	event = perf_evlist__read_on_cpu(evlist, cpu);
+	if (event != NULL) {
+		struct perf_evsel *first;
+		PyObject *pyevent = pyrf_event__new(event);
+		struct pyrf_event *pevent = (struct pyrf_event *)pyevent;
+
+		if (pyevent == NULL)
+			return PyErr_NoMemory();
+
+		first = list_entry(evlist->entries.next, struct perf_evsel, node);
+		perf_event__parse_sample(event, first->attr.sample_type, sample_id_all,
+					 &pevent->sample);
+		return pyevent;
+	}
+
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
+static PyMethodDef pyrf_evlist__methods[] = {
+	{
+		.ml_name  = "mmap",
+		.ml_meth  = (PyCFunction)pyrf_evlist__mmap,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("mmap the file descriptor table.")
+	},
+	{
+		.ml_name  = "poll",
+		.ml_meth  = (PyCFunction)pyrf_evlist__poll,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("poll the file descriptor table.")
+	},
+	{
+		.ml_name  = "get_pollfd",
+		.ml_meth  = (PyCFunction)pyrf_evlist__get_pollfd,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("get the poll file descriptor table.")
+	},
+	{
+		.ml_name  = "add",
+		.ml_meth  = (PyCFunction)pyrf_evlist__add,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("adds an event selector to the list.")
+	},
+	{
+		.ml_name  = "read_on_cpu",
+		.ml_meth  = (PyCFunction)pyrf_evlist__read_on_cpu,
+		.ml_flags = METH_VARARGS | METH_KEYWORDS,
+		.ml_doc	  = PyDoc_STR("reads an event.")
+	},
+	{ .ml_name = NULL, }
+};
+
+static Py_ssize_t pyrf_evlist__length(PyObject *obj)
+{
+	struct pyrf_evlist *pevlist = (void *)obj;
+
+	return pevlist->evlist.nr_entries;
+}
+
+static PyObject *pyrf_evlist__item(PyObject *obj, Py_ssize_t i)
+{
+	struct pyrf_evlist *pevlist = (void *)obj;
+	struct perf_evsel *pos;
+
+	if (i >= pevlist->evlist.nr_entries)
+		return NULL;
+
+	list_for_each_entry(pos, &pevlist->evlist.entries, node)
+		if (i-- == 0)
+			break;
+
+	return Py_BuildValue("O", container_of(pos, struct pyrf_evsel, evsel));
+}
+
+static PySequenceMethods pyrf_evlist__sequence_methods = {
+	.sq_length = pyrf_evlist__length,
+	.sq_item   = pyrf_evlist__item,
+};
+
+static char pyrf_evlist__doc[] = PyDoc_STR("perf event selector list object.");
+
+static PyTypeObject pyrf_evlist__type = {
+	PyVarObject_HEAD_INIT(NULL, 0)
+	.tp_name	= "perf.evlist",
+	.tp_basicsize	= sizeof(struct pyrf_evlist),
+	.tp_dealloc	= (destructor)pyrf_evlist__delete,
+	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+	.tp_as_sequence	= &pyrf_evlist__sequence_methods,
+	.tp_doc		= pyrf_evlist__doc,
+	.tp_methods	= pyrf_evlist__methods,
+	.tp_init	= (initproc)pyrf_evlist__init,
+};
+
+static int pyrf_evlist__setup_types(void)
+{
+	pyrf_evlist__type.tp_new = PyType_GenericNew;
+	return PyType_Ready(&pyrf_evlist__type);
+}
+
+static struct {
+	const char *name;
+	int	    value;
+} perf__constants[] = {
+	{ "TYPE_HARDWARE",   PERF_TYPE_HARDWARE },
+	{ "TYPE_SOFTWARE",   PERF_TYPE_SOFTWARE },
+	{ "TYPE_TRACEPOINT", PERF_TYPE_TRACEPOINT },
+	{ "TYPE_HW_CACHE",   PERF_TYPE_HW_CACHE },
+	{ "TYPE_RAW",	     PERF_TYPE_RAW },
+	{ "TYPE_BREAKPOINT", PERF_TYPE_BREAKPOINT },
+
+	{ "COUNT_HW_CPU_CYCLES",	  PERF_COUNT_HW_CPU_CYCLES },
+	{ "COUNT_HW_INSTRUCTIONS",	  PERF_COUNT_HW_INSTRUCTIONS },
+	{ "COUNT_HW_CACHE_REFERENCES",	  PERF_COUNT_HW_CACHE_REFERENCES },
+	{ "COUNT_HW_CACHE_MISSES",	  PERF_COUNT_HW_CACHE_MISSES },
+	{ "COUNT_HW_BRANCH_INSTRUCTIONS", PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
+	{ "COUNT_HW_BRANCH_MISSES",	  PERF_COUNT_HW_BRANCH_MISSES },
+	{ "COUNT_HW_BUS_CYCLES",	  PERF_COUNT_HW_BUS_CYCLES },
+	{ "COUNT_HW_CACHE_L1D",		  PERF_COUNT_HW_CACHE_L1D },
+	{ "COUNT_HW_CACHE_L1I",		  PERF_COUNT_HW_CACHE_L1I },
+	{ "COUNT_HW_CACHE_LL",	  	  PERF_COUNT_HW_CACHE_LL },
+	{ "COUNT_HW_CACHE_DTLB",	  PERF_COUNT_HW_CACHE_DTLB },
+	{ "COUNT_HW_CACHE_ITLB",	  PERF_COUNT_HW_CACHE_ITLB },
+	{ "COUNT_HW_CACHE_BPU",		  PERF_COUNT_HW_CACHE_BPU },
+	{ "COUNT_HW_CACHE_OP_READ",	  PERF_COUNT_HW_CACHE_OP_READ },
+	{ "COUNT_HW_CACHE_OP_WRITE",	  PERF_COUNT_HW_CACHE_OP_WRITE },
+	{ "COUNT_HW_CACHE_OP_PREFETCH",	  PERF_COUNT_HW_CACHE_OP_PREFETCH },
+	{ "COUNT_HW_CACHE_RESULT_ACCESS", PERF_COUNT_HW_CACHE_RESULT_ACCESS },
+	{ "COUNT_HW_CACHE_RESULT_MISS",   PERF_COUNT_HW_CACHE_RESULT_MISS },
+
+	{ "COUNT_SW_CPU_CLOCK",	       PERF_COUNT_SW_CPU_CLOCK },
+	{ "COUNT_SW_TASK_CLOCK",       PERF_COUNT_SW_TASK_CLOCK },
+	{ "COUNT_SW_PAGE_FAULTS",      PERF_COUNT_SW_PAGE_FAULTS },
+	{ "COUNT_SW_CONTEXT_SWITCHES", PERF_COUNT_SW_CONTEXT_SWITCHES },
+	{ "COUNT_SW_CPU_MIGRATIONS",   PERF_COUNT_SW_CPU_MIGRATIONS },
+	{ "COUNT_SW_PAGE_FAULTS_MIN",  PERF_COUNT_SW_PAGE_FAULTS_MIN },
+	{ "COUNT_SW_PAGE_FAULTS_MAJ",  PERF_COUNT_SW_PAGE_FAULTS_MAJ },
+	{ "COUNT_SW_ALIGNMENT_FAULTS", PERF_COUNT_SW_ALIGNMENT_FAULTS },
+	{ "COUNT_SW_EMULATION_FAULTS", PERF_COUNT_SW_EMULATION_FAULTS },
+
+	{ "SAMPLE_IP",	      PERF_SAMPLE_IP },
+	{ "SAMPLE_TID",	      PERF_SAMPLE_TID },
+	{ "SAMPLE_TIME",      PERF_SAMPLE_TIME },
+	{ "SAMPLE_ADDR",      PERF_SAMPLE_ADDR },
+	{ "SAMPLE_READ",      PERF_SAMPLE_READ },
+	{ "SAMPLE_CALLCHAIN", PERF_SAMPLE_CALLCHAIN },
+	{ "SAMPLE_ID",	      PERF_SAMPLE_ID },
+	{ "SAMPLE_CPU",	      PERF_SAMPLE_CPU },
+	{ "SAMPLE_PERIOD",    PERF_SAMPLE_PERIOD },
+	{ "SAMPLE_STREAM_ID", PERF_SAMPLE_STREAM_ID },
+	{ "SAMPLE_RAW",	      PERF_SAMPLE_RAW },
+
+	{ "FORMAT_TOTAL_TIME_ENABLED", PERF_FORMAT_TOTAL_TIME_ENABLED },
+	{ "FORMAT_TOTAL_TIME_RUNNING", PERF_FORMAT_TOTAL_TIME_RUNNING },
+	{ "FORMAT_ID",		       PERF_FORMAT_ID },
+	{ "FORMAT_GROUP",	       PERF_FORMAT_GROUP },
+
+	{ "RECORD_MMAP",       PERF_RECORD_MMAP },
+	{ "RECORD_LOST",       PERF_RECORD_LOST },
+	{ "RECORD_COMM",       PERF_RECORD_COMM },
+	{ "RECORD_EXIT",       PERF_RECORD_EXIT },
+	{ "RECORD_THROTTLE",   PERF_RECORD_THROTTLE },
+	{ "RECORD_UNTHROTTLE", PERF_RECORD_UNTHROTTLE },
+	{ "RECORD_FORK",       PERF_RECORD_FORK },
+	{ "RECORD_READ",       PERF_RECORD_READ },
+	{ "RECORD_SAMPLE",     PERF_RECORD_SAMPLE },
+	{ .name = NULL, },
+};
+
+static PyMethodDef perf__methods[] = {
+	{ .ml_name = NULL, }
+};
+
+PyMODINIT_FUNC initperf(void)
+{
+	PyObject *obj;
+	int i;
+	PyObject *dict, *module = Py_InitModule("perf", perf__methods);
+
+	if (module == NULL ||
+	    pyrf_event__setup_types() < 0 ||
+	    pyrf_evlist__setup_types() < 0 ||
+	    pyrf_evsel__setup_types() < 0 ||
+	    pyrf_thread_map__setup_types() < 0 ||
+	    pyrf_cpu_map__setup_types() < 0)
+		return;
+
+	Py_INCREF(&pyrf_evlist__type);
+	PyModule_AddObject(module, "evlist", (PyObject*)&pyrf_evlist__type);
+
+	Py_INCREF(&pyrf_evsel__type);
+	PyModule_AddObject(module, "evsel", (PyObject*)&pyrf_evsel__type);
+
+	Py_INCREF(&pyrf_thread_map__type);
+	PyModule_AddObject(module, "thread_map", (PyObject*)&pyrf_thread_map__type);
+
+	Py_INCREF(&pyrf_cpu_map__type);
+	PyModule_AddObject(module, "cpu_map", (PyObject*)&pyrf_cpu_map__type);
+
+	dict = PyModule_GetDict(module);
+	if (dict == NULL)
+		goto error;
+
+	for (i = 0; perf__constants[i].name != NULL; i++) {
+		obj = PyInt_FromLong(perf__constants[i].value);
+		if (obj == NULL)
+			goto error;
+		PyDict_SetItemString(dict, perf__constants[i].name, obj);
+		Py_DECREF(obj);
+	}
+
+error:
+	if (PyErr_Occurred())
+		PyErr_SetString(PyExc_ImportError, "perf: Init failed!");
+}
diff --git a/tools/perf/util/scripting-engines/trace-event-python.c b/tools/perf/util/scripting-engines/trace-event-python.c
index c6d99334bdfa..2040b8538527 100644
--- a/tools/perf/util/scripting-engines/trace-event-python.c
+++ b/tools/perf/util/scripting-engines/trace-event-python.c
@@ -248,8 +248,7 @@ static void python_process_event(int cpu, void *data,
 	context = PyCObject_FromVoidPtr(scripting_context, NULL);
 
 	PyTuple_SetItem(t, n++, PyString_FromString(handler_name));
-	PyTuple_SetItem(t, n++,
-			PyCObject_FromVoidPtr(scripting_context, NULL));
+	PyTuple_SetItem(t, n++, context);
 
 	if (handler) {
 		PyTuple_SetItem(t, n++, PyInt_FromLong(cpu));
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 105f00bfd555..f26639fa0fb3 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -7,6 +7,8 @@
 #include <sys/types.h>
 #include <sys/mman.h>
 
+#include "evlist.h"
+#include "evsel.h"
 #include "session.h"
 #include "sort.h"
 #include "util.h"
@@ -19,7 +21,7 @@ static int perf_session__open(struct perf_session *self, bool force)
 		self->fd_pipe = true;
 		self->fd = STDIN_FILENO;
 
-		if (perf_header__read(self, self->fd) < 0)
+		if (perf_session__read_header(self, self->fd) < 0)
 			pr_err("incompatible file format");
 
 		return 0;
@@ -51,7 +53,7 @@ static int perf_session__open(struct perf_session *self, bool force)
 		goto out_close;
 	}
 
-	if (perf_header__read(self, self->fd) < 0) {
+	if (perf_session__read_header(self, self->fd) < 0) {
 		pr_err("incompatible file format");
 		goto out_close;
 	}
@@ -67,7 +69,7 @@ out_close:
 
 static void perf_session__id_header_size(struct perf_session *session)
 {
-       struct sample_data *data;
+       struct perf_sample *data;
        u64 sample_type = session->sample_type;
        u16 size = 0;
 
@@ -92,21 +94,10 @@ out:
        session->id_hdr_size = size;
 }
 
-void perf_session__set_sample_id_all(struct perf_session *session, bool value)
-{
-	session->sample_id_all = value;
-	perf_session__id_header_size(session);
-}
-
-void perf_session__set_sample_type(struct perf_session *session, u64 type)
-{
-	session->sample_type = type;
-}
-
 void perf_session__update_sample_type(struct perf_session *self)
 {
-	self->sample_type = perf_header__sample_type(&self->header);
-	self->sample_id_all = perf_header__sample_id_all(&self->header);
+	self->sample_type = perf_evlist__sample_type(self->evlist);
+	self->sample_id_all = perf_evlist__sample_id_all(self->evlist);
 	perf_session__id_header_size(self);
 }
 
@@ -135,13 +126,9 @@ struct perf_session *perf_session__new(const char *filename, int mode,
 	if (self == NULL)
 		goto out;
 
-	if (perf_header__init(&self->header) < 0)
-		goto out_free;
-
 	memcpy(self->filename, filename, len);
 	self->threads = RB_ROOT;
 	INIT_LIST_HEAD(&self->dead_threads);
-	self->hists_tree = RB_ROOT;
 	self->last_match = NULL;
 	/*
 	 * On 64bit we can mmap the data file in one go. No need for tiny mmap
@@ -162,17 +149,16 @@ struct perf_session *perf_session__new(const char *filename, int mode,
 	if (mode == O_RDONLY) {
 		if (perf_session__open(self, force) < 0)
 			goto out_delete;
+		perf_session__update_sample_type(self);
 	} else if (mode == O_WRONLY) {
 		/*
 		 * In O_RDONLY mode this will be performed when reading the
-		 * kernel MMAP event, in event__process_mmap().
+		 * kernel MMAP event, in perf_event__process_mmap().
 		 */
 		if (perf_session__create_kernel_maps(self) < 0)
 			goto out_delete;
 	}
 
-	perf_session__update_sample_type(self);
-
 	if (ops && ops->ordering_requires_timestamps &&
 	    ops->ordered_samples && !self->sample_id_all) {
 		dump_printf("WARNING: No sample_id_all support, falling back to unordered processing\n");
@@ -181,9 +167,6 @@ struct perf_session *perf_session__new(const char *filename, int mode,
 
 out:
 	return self;
-out_free:
-	free(self);
-	return NULL;
 out_delete:
 	perf_session__delete(self);
 	return NULL;
@@ -214,7 +197,6 @@ static void perf_session__delete_threads(struct perf_session *self)
 
 void perf_session__delete(struct perf_session *self)
 {
-	perf_header__exit(&self->header);
 	perf_session__destroy_kernel_maps(self);
 	perf_session__delete_dead_threads(self);
 	perf_session__delete_threads(self);
@@ -242,17 +224,16 @@ static bool symbol__match_parent_regex(struct symbol *sym)
 	return 0;
 }
 
-struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
-						   struct thread *thread,
-						   struct ip_callchain *chain,
-						   struct symbol **parent)
+int perf_session__resolve_callchain(struct perf_session *self,
+				    struct thread *thread,
+				    struct ip_callchain *chain,
+				    struct symbol **parent)
 {
 	u8 cpumode = PERF_RECORD_MISC_USER;
 	unsigned int i;
-	struct map_symbol *syms = calloc(chain->nr, sizeof(*syms));
+	int err;
 
-	if (!syms)
-		return NULL;
+	callchain_cursor_reset(&self->callchain_cursor);
 
 	for (i = 0; i < chain->nr; i++) {
 		u64 ip = chain->ips[i];
@@ -281,30 +262,33 @@ struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
 				*parent = al.sym;
 			if (!symbol_conf.use_callchain)
 				break;
-			syms[i].map = al.map;
-			syms[i].sym = al.sym;
 		}
+
+		err = callchain_cursor_append(&self->callchain_cursor,
+					      ip, al.map, al.sym);
+		if (err)
+			return err;
 	}
 
-	return syms;
+	return 0;
 }
 
-static int process_event_synth_stub(event_t *event __used,
+static int process_event_synth_stub(union perf_event *event __used,
 				    struct perf_session *session __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_event_stub(event_t *event __used,
-			      struct sample_data *sample __used,
+static int process_event_stub(union perf_event *event __used,
+			      struct perf_sample *sample __used,
 			      struct perf_session *session __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_finished_round_stub(event_t *event __used,
+static int process_finished_round_stub(union perf_event *event __used,
 				       struct perf_session *session __used,
 				       struct perf_event_ops *ops __used)
 {
@@ -312,7 +296,7 @@ static int process_finished_round_stub(event_t *event __used,
 	return 0;
 }
 
-static int process_finished_round(event_t *event,
+static int process_finished_round(union perf_event *event,
 				  struct perf_session *session,
 				  struct perf_event_ops *ops);
 
@@ -329,7 +313,7 @@ static void perf_event_ops__fill_defaults(struct perf_event_ops *handler)
 	if (handler->exit == NULL)
 		handler->exit = process_event_stub;
 	if (handler->lost == NULL)
-		handler->lost = event__process_lost;
+		handler->lost = perf_event__process_lost;
 	if (handler->read == NULL)
 		handler->read = process_event_stub;
 	if (handler->throttle == NULL)
@@ -363,98 +347,98 @@ void mem_bswap_64(void *src, int byte_size)
 	}
 }
 
-static void event__all64_swap(event_t *self)
+static void perf_event__all64_swap(union perf_event *event)
 {
-	struct perf_event_header *hdr = &self->header;
-	mem_bswap_64(hdr + 1, self->header.size - sizeof(*hdr));
+	struct perf_event_header *hdr = &event->header;
+	mem_bswap_64(hdr + 1, event->header.size - sizeof(*hdr));
 }
 
-static void event__comm_swap(event_t *self)
+static void perf_event__comm_swap(union perf_event *event)
 {
-	self->comm.pid = bswap_32(self->comm.pid);
-	self->comm.tid = bswap_32(self->comm.tid);
+	event->comm.pid = bswap_32(event->comm.pid);
+	event->comm.tid = bswap_32(event->comm.tid);
 }
 
-static void event__mmap_swap(event_t *self)
+static void perf_event__mmap_swap(union perf_event *event)
 {
-	self->mmap.pid	 = bswap_32(self->mmap.pid);
-	self->mmap.tid	 = bswap_32(self->mmap.tid);
-	self->mmap.start = bswap_64(self->mmap.start);
-	self->mmap.len	 = bswap_64(self->mmap.len);
-	self->mmap.pgoff = bswap_64(self->mmap.pgoff);
+	event->mmap.pid	  = bswap_32(event->mmap.pid);
+	event->mmap.tid	  = bswap_32(event->mmap.tid);
+	event->mmap.start = bswap_64(event->mmap.start);
+	event->mmap.len	  = bswap_64(event->mmap.len);
+	event->mmap.pgoff = bswap_64(event->mmap.pgoff);
 }
 
-static void event__task_swap(event_t *self)
+static void perf_event__task_swap(union perf_event *event)
 {
-	self->fork.pid	= bswap_32(self->fork.pid);
-	self->fork.tid	= bswap_32(self->fork.tid);
-	self->fork.ppid	= bswap_32(self->fork.ppid);
-	self->fork.ptid	= bswap_32(self->fork.ptid);
-	self->fork.time	= bswap_64(self->fork.time);
+	event->fork.pid	 = bswap_32(event->fork.pid);
+	event->fork.tid	 = bswap_32(event->fork.tid);
+	event->fork.ppid = bswap_32(event->fork.ppid);
+	event->fork.ptid = bswap_32(event->fork.ptid);
+	event->fork.time = bswap_64(event->fork.time);
 }
 
-static void event__read_swap(event_t *self)
+static void perf_event__read_swap(union perf_event *event)
 {
-	self->read.pid		= bswap_32(self->read.pid);
-	self->read.tid		= bswap_32(self->read.tid);
-	self->read.value	= bswap_64(self->read.value);
-	self->read.time_enabled	= bswap_64(self->read.time_enabled);
-	self->read.time_running	= bswap_64(self->read.time_running);
-	self->read.id		= bswap_64(self->read.id);
+	event->read.pid		 = bswap_32(event->read.pid);
+	event->read.tid		 = bswap_32(event->read.tid);
+	event->read.value	 = bswap_64(event->read.value);
+	event->read.time_enabled = bswap_64(event->read.time_enabled);
+	event->read.time_running = bswap_64(event->read.time_running);
+	event->read.id		 = bswap_64(event->read.id);
 }
 
-static void event__attr_swap(event_t *self)
+static void perf_event__attr_swap(union perf_event *event)
 {
 	size_t size;
 
-	self->attr.attr.type		= bswap_32(self->attr.attr.type);
-	self->attr.attr.size		= bswap_32(self->attr.attr.size);
-	self->attr.attr.config		= bswap_64(self->attr.attr.config);
-	self->attr.attr.sample_period	= bswap_64(self->attr.attr.sample_period);
-	self->attr.attr.sample_type	= bswap_64(self->attr.attr.sample_type);
-	self->attr.attr.read_format	= bswap_64(self->attr.attr.read_format);
-	self->attr.attr.wakeup_events	= bswap_32(self->attr.attr.wakeup_events);
-	self->attr.attr.bp_type		= bswap_32(self->attr.attr.bp_type);
-	self->attr.attr.bp_addr		= bswap_64(self->attr.attr.bp_addr);
-	self->attr.attr.bp_len		= bswap_64(self->attr.attr.bp_len);
-
-	size = self->header.size;
-	size -= (void *)&self->attr.id - (void *)self;
-	mem_bswap_64(self->attr.id, size);
+	event->attr.attr.type		= bswap_32(event->attr.attr.type);
+	event->attr.attr.size		= bswap_32(event->attr.attr.size);
+	event->attr.attr.config		= bswap_64(event->attr.attr.config);
+	event->attr.attr.sample_period	= bswap_64(event->attr.attr.sample_period);
+	event->attr.attr.sample_type	= bswap_64(event->attr.attr.sample_type);
+	event->attr.attr.read_format	= bswap_64(event->attr.attr.read_format);
+	event->attr.attr.wakeup_events	= bswap_32(event->attr.attr.wakeup_events);
+	event->attr.attr.bp_type	= bswap_32(event->attr.attr.bp_type);
+	event->attr.attr.bp_addr	= bswap_64(event->attr.attr.bp_addr);
+	event->attr.attr.bp_len		= bswap_64(event->attr.attr.bp_len);
+
+	size = event->header.size;
+	size -= (void *)&event->attr.id - (void *)event;
+	mem_bswap_64(event->attr.id, size);
 }
 
-static void event__event_type_swap(event_t *self)
+static void perf_event__event_type_swap(union perf_event *event)
 {
-	self->event_type.event_type.event_id =
-		bswap_64(self->event_type.event_type.event_id);
+	event->event_type.event_type.event_id =
+		bswap_64(event->event_type.event_type.event_id);
 }
 
-static void event__tracing_data_swap(event_t *self)
+static void perf_event__tracing_data_swap(union perf_event *event)
 {
-	self->tracing_data.size = bswap_32(self->tracing_data.size);
+	event->tracing_data.size = bswap_32(event->tracing_data.size);
 }
 
-typedef void (*event__swap_op)(event_t *self);
-
-static event__swap_op event__swap_ops[] = {
-	[PERF_RECORD_MMAP]   = event__mmap_swap,
-	[PERF_RECORD_COMM]   = event__comm_swap,
-	[PERF_RECORD_FORK]   = event__task_swap,
-	[PERF_RECORD_EXIT]   = event__task_swap,
-	[PERF_RECORD_LOST]   = event__all64_swap,
-	[PERF_RECORD_READ]   = event__read_swap,
-	[PERF_RECORD_SAMPLE] = event__all64_swap,
-	[PERF_RECORD_HEADER_ATTR]   = event__attr_swap,
-	[PERF_RECORD_HEADER_EVENT_TYPE]   = event__event_type_swap,
-	[PERF_RECORD_HEADER_TRACING_DATA]   = event__tracing_data_swap,
-	[PERF_RECORD_HEADER_BUILD_ID]   = NULL,
-	[PERF_RECORD_HEADER_MAX]    = NULL,
+typedef void (*perf_event__swap_op)(union perf_event *event);
+
+static perf_event__swap_op perf_event__swap_ops[] = {
+	[PERF_RECORD_MMAP]		  = perf_event__mmap_swap,
+	[PERF_RECORD_COMM]		  = perf_event__comm_swap,
+	[PERF_RECORD_FORK]		  = perf_event__task_swap,
+	[PERF_RECORD_EXIT]		  = perf_event__task_swap,
+	[PERF_RECORD_LOST]		  = perf_event__all64_swap,
+	[PERF_RECORD_READ]		  = perf_event__read_swap,
+	[PERF_RECORD_SAMPLE]		  = perf_event__all64_swap,
+	[PERF_RECORD_HEADER_ATTR]	  = perf_event__attr_swap,
+	[PERF_RECORD_HEADER_EVENT_TYPE]	  = perf_event__event_type_swap,
+	[PERF_RECORD_HEADER_TRACING_DATA] = perf_event__tracing_data_swap,
+	[PERF_RECORD_HEADER_BUILD_ID]	  = NULL,
+	[PERF_RECORD_HEADER_MAX]	  = NULL,
 };
 
 struct sample_queue {
 	u64			timestamp;
 	u64			file_offset;
-	event_t			*event;
+	union perf_event	*event;
 	struct list_head	list;
 };
 
@@ -472,8 +456,8 @@ static void perf_session_free_sample_buffers(struct perf_session *session)
 }
 
 static int perf_session_deliver_event(struct perf_session *session,
-				      event_t *event,
-				      struct sample_data *sample,
+				      union perf_event *event,
+				      struct perf_sample *sample,
 				      struct perf_event_ops *ops,
 				      u64 file_offset);
 
@@ -483,7 +467,7 @@ static void flush_sample_queue(struct perf_session *s,
 	struct ordered_samples *os = &s->ordered_samples;
 	struct list_head *head = &os->samples;
 	struct sample_queue *tmp, *iter;
-	struct sample_data sample;
+	struct perf_sample sample;
 	u64 limit = os->next_flush;
 	u64 last_ts = os->last_sample ? os->last_sample->timestamp : 0ULL;
 
@@ -494,7 +478,7 @@ static void flush_sample_queue(struct perf_session *s,
 		if (iter->timestamp > limit)
 			break;
 
-		event__parse_sample(iter->event, s, &sample);
+		perf_session__parse_sample(s, iter->event, &sample);
 		perf_session_deliver_event(s, iter->event, &sample, ops,
 					   iter->file_offset);
 
@@ -550,7 +534,7 @@ static void flush_sample_queue(struct perf_session *s,
  *      Flush every events below timestamp 7
  *      etc...
  */
-static int process_finished_round(event_t *event __used,
+static int process_finished_round(union perf_event *event __used,
 				  struct perf_session *session,
 				  struct perf_event_ops *ops)
 {
@@ -607,12 +591,12 @@ static void __queue_event(struct sample_queue *new, struct perf_session *s)
 
 #define MAX_SAMPLE_BUFFER	(64 * 1024 / sizeof(struct sample_queue))
 
-static int perf_session_queue_event(struct perf_session *s, event_t *event,
-				    struct sample_data *data, u64 file_offset)
+static int perf_session_queue_event(struct perf_session *s, union perf_event *event,
+				    struct perf_sample *sample, u64 file_offset)
 {
 	struct ordered_samples *os = &s->ordered_samples;
 	struct list_head *sc = &os->sample_cache;
-	u64 timestamp = data->time;
+	u64 timestamp = sample->time;
 	struct sample_queue *new;
 
 	if (!timestamp || timestamp == ~0ULL)
@@ -648,7 +632,7 @@ static int perf_session_queue_event(struct perf_session *s, event_t *event,
 	return 0;
 }
 
-static void callchain__printf(struct sample_data *sample)
+static void callchain__printf(struct perf_sample *sample)
 {
 	unsigned int i;
 
@@ -660,8 +644,8 @@ static void callchain__printf(struct sample_data *sample)
 }
 
 static void perf_session__print_tstamp(struct perf_session *session,
-				       event_t *event,
-				       struct sample_data *sample)
+				       union perf_event *event,
+				       struct perf_sample *sample)
 {
 	if (event->header.type != PERF_RECORD_SAMPLE &&
 	    !session->sample_id_all) {
@@ -676,8 +660,8 @@ static void perf_session__print_tstamp(struct perf_session *session,
 		printf("%" PRIu64 " ", sample->time);
 }
 
-static void dump_event(struct perf_session *session, event_t *event,
-		       u64 file_offset, struct sample_data *sample)
+static void dump_event(struct perf_session *session, union perf_event *event,
+		       u64 file_offset, struct perf_sample *sample)
 {
 	if (!dump_trace)
 		return;
@@ -691,11 +675,11 @@ static void dump_event(struct perf_session *session, event_t *event,
 		perf_session__print_tstamp(session, event, sample);
 
 	printf("%#" PRIx64 " [%#x]: PERF_RECORD_%s", file_offset,
-	       event->header.size, event__get_event_name(event->header.type));
+	       event->header.size, perf_event__name(event->header.type));
 }
 
-static void dump_sample(struct perf_session *session, event_t *event,
-			struct sample_data *sample)
+static void dump_sample(struct perf_session *session, union perf_event *event,
+			struct perf_sample *sample)
 {
 	if (!dump_trace)
 		return;
@@ -709,8 +693,8 @@ static void dump_sample(struct perf_session *session, event_t *event,
 }
 
 static int perf_session_deliver_event(struct perf_session *session,
-				      event_t *event,
-				      struct sample_data *sample,
+				      union perf_event *event,
+				      struct perf_sample *sample,
 				      struct perf_event_ops *ops,
 				      u64 file_offset)
 {
@@ -743,7 +727,7 @@ static int perf_session_deliver_event(struct perf_session *session,
 }
 
 static int perf_session__preprocess_sample(struct perf_session *session,
-					   event_t *event, struct sample_data *sample)
+					   union perf_event *event, struct perf_sample *sample)
 {
 	if (event->header.type != PERF_RECORD_SAMPLE ||
 	    !(session->sample_type & PERF_SAMPLE_CALLCHAIN))
@@ -758,7 +742,7 @@ static int perf_session__preprocess_sample(struct perf_session *session,
 	return 0;
 }
 
-static int perf_session__process_user_event(struct perf_session *session, event_t *event,
+static int perf_session__process_user_event(struct perf_session *session, union perf_event *event,
 					    struct perf_event_ops *ops, u64 file_offset)
 {
 	dump_event(session, event, file_offset, NULL);
@@ -783,15 +767,16 @@ static int perf_session__process_user_event(struct perf_session *session, event_
 }
 
 static int perf_session__process_event(struct perf_session *session,
-				       event_t *event,
+				       union perf_event *event,
 				       struct perf_event_ops *ops,
 				       u64 file_offset)
 {
-	struct sample_data sample;
+	struct perf_sample sample;
 	int ret;
 
-	if (session->header.needs_swap && event__swap_ops[event->header.type])
-		event__swap_ops[event->header.type](event);
+	if (session->header.needs_swap &&
+	    perf_event__swap_ops[event->header.type])
+		perf_event__swap_ops[event->header.type](event);
 
 	if (event->header.type >= PERF_RECORD_HEADER_MAX)
 		return -EINVAL;
@@ -804,7 +789,7 @@ static int perf_session__process_event(struct perf_session *session,
 	/*
 	 * For all kernel events we get the sample data
 	 */
-	event__parse_sample(event, session, &sample);
+	perf_session__parse_sample(session, event, &sample);
 
 	/* Preprocess sample records - precheck callchains */
 	if (perf_session__preprocess_sample(session, event, &sample))
@@ -843,7 +828,7 @@ static struct thread *perf_session__register_idle_thread(struct perf_session *se
 static void perf_session__warn_about_errors(const struct perf_session *session,
 					    const struct perf_event_ops *ops)
 {
-	if (ops->lost == event__process_lost &&
+	if (ops->lost == perf_event__process_lost &&
 	    session->hists.stats.total_lost != 0) {
 		ui__warning("Processed %" PRIu64 " events and LOST %" PRIu64
 			    "!\n\nCheck IO/CPU overload!\n\n",
@@ -875,7 +860,7 @@ volatile int session_done;
 static int __perf_session__process_pipe_events(struct perf_session *self,
 					       struct perf_event_ops *ops)
 {
-	event_t event;
+	union perf_event event;
 	uint32_t size;
 	int skip = 0;
 	u64 head;
@@ -956,7 +941,7 @@ int __perf_session__process_events(struct perf_session *session,
 	struct ui_progress *progress;
 	size_t	page_size, mmap_size;
 	char *buf, *mmaps[8];
-	event_t *event;
+	union perf_event *event;
 	uint32_t size;
 
 	perf_event_ops__fill_defaults(ops);
@@ -1001,7 +986,7 @@ remap:
 	file_pos = file_offset + head;
 
 more:
-	event = (event_t *)(buf + head);
+	event = (union perf_event *)(buf + head);
 
 	if (session->header.needs_swap)
 		perf_event_header__bswap(&event->header);
@@ -1134,3 +1119,18 @@ size_t perf_session__fprintf_dsos_buildid(struct perf_session *self, FILE *fp,
 	size_t ret = machine__fprintf_dsos_buildid(&self->host_machine, fp, with_hits);
 	return ret + machines__fprintf_dsos_buildid(&self->machines, fp, with_hits);
 }
+
+size_t perf_session__fprintf_nr_events(struct perf_session *session, FILE *fp)
+{
+	struct perf_evsel *pos;
+	size_t ret = fprintf(fp, "Aggregated stats:\n");
+
+	ret += hists__fprintf_nr_events(&session->hists, fp);
+
+	list_for_each_entry(pos, &session->evlist->entries, node) {
+		ret += fprintf(fp, "%s stats:\n", event_name(pos));
+		ret += hists__fprintf_nr_events(&pos->hists, fp);
+	}
+
+	return ret;
+}
diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h
index decd83f274fd..b5b148b0aaca 100644
--- a/tools/perf/util/session.h
+++ b/tools/perf/util/session.h
@@ -34,12 +34,12 @@ struct perf_session {
 	struct thread		*last_match;
 	struct machine		host_machine;
 	struct rb_root		machines;
-	struct rb_root		hists_tree;
+	struct perf_evlist	*evlist;
 	/*
-	 * FIXME: should point to the first entry in hists_tree and
-	 *        be a hists instance. Right now its only 'report'
-	 *        that is using ->hists_tree while all the rest use
-	 *        ->hists.
+	 * FIXME: Need to split this up further, we need global
+	 *	  stats + per event stats. 'perf diff' also needs
+	 *	  to properly support multiple events in a single
+	 *	  perf.data file.
 	 */
 	struct hists		hists;
 	u64			sample_type;
@@ -51,15 +51,17 @@ struct perf_session {
 	int			cwdlen;
 	char			*cwd;
 	struct ordered_samples	ordered_samples;
-	char filename[0];
+	struct callchain_cursor	callchain_cursor;
+	char			filename[0];
 };
 
 struct perf_event_ops;
 
-typedef int (*event_op)(event_t *self, struct sample_data *sample,
+typedef int (*event_op)(union perf_event *self, struct perf_sample *sample,
 			struct perf_session *session);
-typedef int (*event_synth_op)(event_t *self, struct perf_session *session);
-typedef int (*event_op2)(event_t *self, struct perf_session *session,
+typedef int (*event_synth_op)(union perf_event *self,
+			      struct perf_session *session);
+typedef int (*event_op2)(union perf_event *self, struct perf_session *session,
 			 struct perf_event_ops *ops);
 
 struct perf_event_ops {
@@ -94,10 +96,10 @@ int __perf_session__process_events(struct perf_session *self,
 int perf_session__process_events(struct perf_session *self,
 				 struct perf_event_ops *event_ops);
 
-struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
-						   struct thread *thread,
-						   struct ip_callchain *chain,
-						   struct symbol **parent);
+int perf_session__resolve_callchain(struct perf_session *self,
+				    struct thread *thread,
+				    struct ip_callchain *chain,
+				    struct symbol **parent);
 
 bool perf_session__has_traces(struct perf_session *self, const char *msg);
 
@@ -110,8 +112,6 @@ void mem_bswap_64(void *src, int byte_size);
 int perf_session__create_kernel_maps(struct perf_session *self);
 
 void perf_session__update_sample_type(struct perf_session *self);
-void perf_session__set_sample_id_all(struct perf_session *session, bool value);
-void perf_session__set_sample_type(struct perf_session *session, u64 type);
 void perf_session__remove_thread(struct perf_session *self, struct thread *th);
 
 static inline
@@ -149,9 +149,14 @@ size_t perf_session__fprintf_dsos(struct perf_session *self, FILE *fp);
 size_t perf_session__fprintf_dsos_buildid(struct perf_session *self,
 					  FILE *fp, bool with_hits);
 
-static inline
-size_t perf_session__fprintf_nr_events(struct perf_session *self, FILE *fp)
+size_t perf_session__fprintf_nr_events(struct perf_session *session, FILE *fp);
+
+static inline int perf_session__parse_sample(struct perf_session *session,
+					     const union perf_event *event,
+					     struct perf_sample *sample)
 {
-	return hists__fprintf_nr_events(&self->hists, fp);
+	return perf_event__parse_sample(event, session->sample_type,
+					session->sample_id_all, sample);
 }
+
 #endif /* __PERF_SESSION_H */
diff --git a/tools/perf/util/setup.py b/tools/perf/util/setup.py
new file mode 100644
index 000000000000..e24ffadb20b2
--- /dev/null
+++ b/tools/perf/util/setup.py
@@ -0,0 +1,19 @@
+#!/usr/bin/python2
+
+from distutils.core import setup, Extension
+
+perf = Extension('perf',
+		  sources = ['util/python.c', 'util/ctype.c', 'util/evlist.c',
+			     'util/evsel.c', 'util/cpumap.c', 'util/thread_map.c',
+			     'util/util.c', 'util/xyarray.c', 'util/cgroup.c'],
+		  include_dirs = ['util/include'],
+		  extra_compile_args = ['-fno-strict-aliasing', '-Wno-write-strings'])
+
+setup(name='perf',
+      version='0.1',
+      description='Interface with the Linux profiling infrastructure',
+      author='Arnaldo Carvalho de Melo',
+      author_email='acme@redhat.com',
+      license='GPLv2',
+      url='http://perf.wiki.kernel.org',
+      ext_modules=[perf])
diff --git a/tools/perf/util/strfilter.c b/tools/perf/util/strfilter.c
new file mode 100644
index 000000000000..834c8ebfe38e
--- /dev/null
+++ b/tools/perf/util/strfilter.c
@@ -0,0 +1,199 @@
+#include "util.h"
+#include "string.h"
+#include "strfilter.h"
+
+/* Operators */
+static const char *OP_and	= "&";	/* Logical AND */
+static const char *OP_or	= "|";	/* Logical OR */
+static const char *OP_not	= "!";	/* Logical NOT */
+
+#define is_operator(c)	((c) == '|' || (c) == '&' || (c) == '!')
+#define is_separator(c)	(is_operator(c) || (c) == '(' || (c) == ')')
+
+static void strfilter_node__delete(struct strfilter_node *self)
+{
+	if (self) {
+		if (self->p && !is_operator(*self->p))
+			free((char *)self->p);
+		strfilter_node__delete(self->l);
+		strfilter_node__delete(self->r);
+		free(self);
+	}
+}
+
+void strfilter__delete(struct strfilter *self)
+{
+	if (self) {
+		strfilter_node__delete(self->root);
+		free(self);
+	}
+}
+
+static const char *get_token(const char *s, const char **e)
+{
+	const char *p;
+
+	while (isspace(*s))	/* Skip spaces */
+		s++;
+
+	if (*s == '\0') {
+		p = s;
+		goto end;
+	}
+
+	p = s + 1;
+	if (!is_separator(*s)) {
+		/* End search */
+retry:
+		while (*p && !is_separator(*p) && !isspace(*p))
+			p++;
+		/* Escape and special case: '!' is also used in glob pattern */
+		if (*(p - 1) == '\\' || (*p == '!' && *(p - 1) == '[')) {
+			p++;
+			goto retry;
+		}
+	}
+end:
+	*e = p;
+	return s;
+}
+
+static struct strfilter_node *strfilter_node__alloc(const char *op,
+						    struct strfilter_node *l,
+						    struct strfilter_node *r)
+{
+	struct strfilter_node *ret = zalloc(sizeof(struct strfilter_node));
+
+	if (ret) {
+		ret->p = op;
+		ret->l = l;
+		ret->r = r;
+	}
+
+	return ret;
+}
+
+static struct strfilter_node *strfilter_node__new(const char *s,
+						  const char **ep)
+{
+	struct strfilter_node root, *cur, *last_op;
+	const char *e;
+
+	if (!s)
+		return NULL;
+
+	memset(&root, 0, sizeof(root));
+	last_op = cur = &root;
+
+	s = get_token(s, &e);
+	while (*s != '\0' && *s != ')') {
+		switch (*s) {
+		case '&':	/* Exchg last OP->r with AND */
+			if (!cur->r || !last_op->r)
+				goto error;
+			cur = strfilter_node__alloc(OP_and, last_op->r, NULL);
+			if (!cur)
+				goto nomem;
+			last_op->r = cur;
+			last_op = cur;
+			break;
+		case '|':	/* Exchg the root with OR */
+			if (!cur->r || !root.r)
+				goto error;
+			cur = strfilter_node__alloc(OP_or, root.r, NULL);
+			if (!cur)
+				goto nomem;
+			root.r = cur;
+			last_op = cur;
+			break;
+		case '!':	/* Add NOT as a leaf node */
+			if (cur->r)
+				goto error;
+			cur->r = strfilter_node__alloc(OP_not, NULL, NULL);
+			if (!cur->r)
+				goto nomem;
+			cur = cur->r;
+			break;
+		case '(':	/* Recursively parses inside the parenthesis */
+			if (cur->r)
+				goto error;
+			cur->r = strfilter_node__new(s + 1, &s);
+			if (!s)
+				goto nomem;
+			if (!cur->r || *s != ')')
+				goto error;
+			e = s + 1;
+			break;
+		default:
+			if (cur->r)
+				goto error;
+			cur->r = strfilter_node__alloc(NULL, NULL, NULL);
+			if (!cur->r)
+				goto nomem;
+			cur->r->p = strndup(s, e - s);
+			if (!cur->r->p)
+				goto nomem;
+		}
+		s = get_token(e, &e);
+	}
+	if (!cur->r)
+		goto error;
+	*ep = s;
+	return root.r;
+nomem:
+	s = NULL;
+error:
+	*ep = s;
+	strfilter_node__delete(root.r);
+	return NULL;
+}
+
+/*
+ * Parse filter rule and return new strfilter.
+ * Return NULL if fail, and *ep == NULL if memory allocation failed.
+ */
+struct strfilter *strfilter__new(const char *rules, const char **err)
+{
+	struct strfilter *ret = zalloc(sizeof(struct strfilter));
+	const char *ep = NULL;
+
+	if (ret)
+		ret->root = strfilter_node__new(rules, &ep);
+
+	if (!ret || !ret->root || *ep != '\0') {
+		if (err)
+			*err = ep;
+		strfilter__delete(ret);
+		ret = NULL;
+	}
+
+	return ret;
+}
+
+static bool strfilter_node__compare(struct strfilter_node *self,
+				    const char *str)
+{
+	if (!self || !self->p)
+		return false;
+
+	switch (*self->p) {
+	case '|':	/* OR */
+		return strfilter_node__compare(self->l, str) ||
+			strfilter_node__compare(self->r, str);
+	case '&':	/* AND */
+		return strfilter_node__compare(self->l, str) &&
+			strfilter_node__compare(self->r, str);
+	case '!':	/* NOT */
+		return !strfilter_node__compare(self->r, str);
+	default:
+		return strglobmatch(str, self->p);
+	}
+}
+
+/* Return true if STR matches the filter rules */
+bool strfilter__compare(struct strfilter *self, const char *str)
+{
+	if (!self)
+		return false;
+	return strfilter_node__compare(self->root, str);
+}
diff --git a/tools/perf/util/strfilter.h b/tools/perf/util/strfilter.h
new file mode 100644
index 000000000000..00f58a7506de
--- /dev/null
+++ b/tools/perf/util/strfilter.h
@@ -0,0 +1,48 @@
+#ifndef __PERF_STRFILTER_H
+#define __PERF_STRFILTER_H
+/* General purpose glob matching filter */
+
+#include <linux/list.h>
+#include <stdbool.h>
+
+/* A node of string filter */
+struct strfilter_node {
+	struct strfilter_node *l;	/* Tree left branche (for &,|) */
+	struct strfilter_node *r;	/* Tree right branche (for !,&,|) */
+	const char *p;		/* Operator or rule */
+};
+
+/* String filter */
+struct strfilter {
+	struct strfilter_node *root;
+};
+
+/**
+ * strfilter__new - Create a new string filter
+ * @rules: Filter rule, which is a combination of glob expressions.
+ * @err: Pointer which points an error detected on @rules
+ *
+ * Parse @rules and return new strfilter. Return NULL if an error detected.
+ * In that case, *@err will indicate where it is detected, and *@err is NULL
+ * if a memory allocation is failed.
+ */
+struct strfilter *strfilter__new(const char *rules, const char **err);
+
+/**
+ * strfilter__compare - compare given string and a string filter
+ * @self: String filter
+ * @str: target string
+ *
+ * Compare @str and @self. Return true if the str match the rule
+ */
+bool strfilter__compare(struct strfilter *self, const char *str);
+
+/**
+ * strfilter__delete - delete a string filter
+ * @self: String filter to delete
+ *
+ * Delete @self.
+ */
+void strfilter__delete(struct strfilter *self);
+
+#endif
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index b1bf490aff88..00014e32c288 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -207,7 +207,6 @@ struct dso *dso__new(const char *name)
 		dso__set_short_name(self, self->name);
 		for (i = 0; i < MAP__NR_TYPES; ++i)
 			self->symbols[i] = self->symbol_names[i] = RB_ROOT;
-		self->slen_calculated = 0;
 		self->origin = DSO__ORIG_NOT_FOUND;
 		self->loaded = 0;
 		self->sorted_by_name = 0;
@@ -1525,8 +1524,8 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 			     symbol_conf.symfs, self->long_name);
 			break;
 		case DSO__ORIG_GUEST_KMODULE:
-			if (map->groups && map->groups->machine)
-				root_dir = map->groups->machine->root_dir;
+			if (map->groups && machine)
+				root_dir = machine->root_dir;
 			else
 				root_dir = "";
 			snprintf(name, size, "%s%s%s", symbol_conf.symfs,
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 670cd1c88f54..4d7ed09fe332 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -132,7 +132,6 @@ struct dso {
 	struct rb_root	 symbol_names[MAP__NR_TYPES];
 	enum dso_kernel_type	kernel;
 	u8		 adjust_symbols:1;
-	u8		 slen_calculated:1;
 	u8		 has_build_id:1;
 	u8		 hit:1;
 	u8		 annotate_warned:1;
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index 00f4eade2e3e..d5d3b22250f3 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -7,61 +7,6 @@
 #include "util.h"
 #include "debug.h"
 
-/* Skip "." and ".." directories */
-static int filter(const struct dirent *dir)
-{
-	if (dir->d_name[0] == '.')
-		return 0;
-	else
-		return 1;
-}
-
-struct thread_map *thread_map__new_by_pid(pid_t pid)
-{
-	struct thread_map *threads;
-	char name[256];
-	int items;
-	struct dirent **namelist = NULL;
-	int i;
-
-	sprintf(name, "/proc/%d/task", pid);
-	items = scandir(name, &namelist, filter, NULL);
-	if (items <= 0)
-                return NULL;
-
-	threads = malloc(sizeof(*threads) + sizeof(pid_t) * items);
-	if (threads != NULL) {
-		for (i = 0; i < items; i++)
-			threads->map[i] = atoi(namelist[i]->d_name);
-		threads->nr = items;
-	}
-
-	for (i=0; i<items; i++)
-		free(namelist[i]);
-	free(namelist);
-
-	return threads;
-}
-
-struct thread_map *thread_map__new_by_tid(pid_t tid)
-{
-	struct thread_map *threads = malloc(sizeof(*threads) + sizeof(pid_t));
-
-	if (threads != NULL) {
-		threads->map[0] = tid;
-		threads->nr	= 1;
-	}
-
-	return threads;
-}
-
-struct thread_map *thread_map__new(pid_t pid, pid_t tid)
-{
-	if (pid != -1)
-		return thread_map__new_by_pid(pid);
-	return thread_map__new_by_tid(tid);
-}
-
 static struct thread *thread__new(pid_t pid)
 {
 	struct thread *self = zalloc(sizeof(*self));
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
index d7574101054a..e5f2401c1b5e 100644
--- a/tools/perf/util/thread.h
+++ b/tools/perf/util/thread.h
@@ -18,24 +18,10 @@ struct thread {
 	int			comm_len;
 };
 
-struct thread_map {
-	int nr;
-	int map[];
-};
-
 struct perf_session;
 
 void thread__delete(struct thread *self);
 
-struct thread_map *thread_map__new_by_pid(pid_t pid);
-struct thread_map *thread_map__new_by_tid(pid_t tid);
-struct thread_map *thread_map__new(pid_t pid, pid_t tid);
-
-static inline void thread_map__delete(struct thread_map *threads)
-{
-	free(threads);
-}
-
 int thread__set_comm(struct thread *self, const char *comm);
 int thread__comm_len(struct thread *self);
 struct thread *perf_session__findnew(struct perf_session *self, pid_t pid);
diff --git a/tools/perf/util/thread_map.c b/tools/perf/util/thread_map.c
new file mode 100644
index 000000000000..a5df131b77c3
--- /dev/null
+++ b/tools/perf/util/thread_map.c
@@ -0,0 +1,64 @@
+#include <dirent.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include "thread_map.h"
+
+/* Skip "." and ".." directories */
+static int filter(const struct dirent *dir)
+{
+	if (dir->d_name[0] == '.')
+		return 0;
+	else
+		return 1;
+}
+
+struct thread_map *thread_map__new_by_pid(pid_t pid)
+{
+	struct thread_map *threads;
+	char name[256];
+	int items;
+	struct dirent **namelist = NULL;
+	int i;
+
+	sprintf(name, "/proc/%d/task", pid);
+	items = scandir(name, &namelist, filter, NULL);
+	if (items <= 0)
+                return NULL;
+
+	threads = malloc(sizeof(*threads) + sizeof(pid_t) * items);
+	if (threads != NULL) {
+		for (i = 0; i < items; i++)
+			threads->map[i] = atoi(namelist[i]->d_name);
+		threads->nr = items;
+	}
+
+	for (i=0; i<items; i++)
+		free(namelist[i]);
+	free(namelist);
+
+	return threads;
+}
+
+struct thread_map *thread_map__new_by_tid(pid_t tid)
+{
+	struct thread_map *threads = malloc(sizeof(*threads) + sizeof(pid_t));
+
+	if (threads != NULL) {
+		threads->map[0] = tid;
+		threads->nr	= 1;
+	}
+
+	return threads;
+}
+
+struct thread_map *thread_map__new(pid_t pid, pid_t tid)
+{
+	if (pid != -1)
+		return thread_map__new_by_pid(pid);
+	return thread_map__new_by_tid(tid);
+}
+
+void thread_map__delete(struct thread_map *threads)
+{
+	free(threads);
+}
diff --git a/tools/perf/util/thread_map.h b/tools/perf/util/thread_map.h
new file mode 100644
index 000000000000..3cb907311409
--- /dev/null
+++ b/tools/perf/util/thread_map.h
@@ -0,0 +1,15 @@
+#ifndef __PERF_THREAD_MAP_H
+#define __PERF_THREAD_MAP_H
+
+#include <sys/types.h>
+
+struct thread_map {
+	int nr;
+	int map[];
+};
+
+struct thread_map *thread_map__new_by_pid(pid_t pid);
+struct thread_map *thread_map__new_by_tid(pid_t tid);
+struct thread_map *thread_map__new(pid_t pid, pid_t tid);
+void thread_map__delete(struct thread_map *threads);
+#endif	/* __PERF_THREAD_MAP_H */
diff --git a/tools/perf/util/top.c b/tools/perf/util/top.c
new file mode 100644
index 000000000000..75cfe4d45119
--- /dev/null
+++ b/tools/perf/util/top.c
@@ -0,0 +1,238 @@
+/*
+ * Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Refactored from builtin-top.c, see that files for further copyright notes.
+ *
+ * Released under the GPL v2. (and only v2, not any later version)
+ */
+
+#include "cpumap.h"
+#include "event.h"
+#include "evlist.h"
+#include "evsel.h"
+#include "parse-events.h"
+#include "symbol.h"
+#include "top.h"
+#include <inttypes.h>
+
+/*
+ * Ordering weight: count-1 * count-2 * ... / count-n
+ */
+static double sym_weight(const struct sym_entry *sym, struct perf_top *top)
+{
+	double weight = sym->snap_count;
+	int counter;
+
+	if (!top->display_weighted)
+		return weight;
+
+	for (counter = 1; counter < top->evlist->nr_entries - 1; counter++)
+		weight *= sym->count[counter];
+
+	weight /= (sym->count[counter] + 1);
+
+	return weight;
+}
+
+static void perf_top__remove_active_sym(struct perf_top *top, struct sym_entry *syme)
+{
+	pthread_mutex_lock(&top->active_symbols_lock);
+	list_del_init(&syme->node);
+	pthread_mutex_unlock(&top->active_symbols_lock);
+}
+
+static void rb_insert_active_sym(struct rb_root *tree, struct sym_entry *se)
+{
+	struct rb_node **p = &tree->rb_node;
+	struct rb_node *parent = NULL;
+	struct sym_entry *iter;
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct sym_entry, rb_node);
+
+		if (se->weight > iter->weight)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&se->rb_node, parent, p);
+	rb_insert_color(&se->rb_node, tree);
+}
+
+#define SNPRINTF(buf, size, fmt, args...) \
+({ \
+	size_t r = snprintf(buf, size, fmt, ## args); \
+	r > size ?  size : r; \
+})
+
+size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size)
+{
+	struct perf_evsel *counter;
+	float samples_per_sec = top->samples / top->delay_secs;
+	float ksamples_per_sec = top->kernel_samples / top->delay_secs;
+	float esamples_percent = (100.0 * top->exact_samples) / top->samples;
+	size_t ret = 0;
+
+	if (!perf_guest) {
+		ret = SNPRINTF(bf, size,
+			       "   PerfTop:%8.0f irqs/sec  kernel:%4.1f%%"
+			       "  exact: %4.1f%% [", samples_per_sec,
+			       100.0 - (100.0 * ((samples_per_sec - ksamples_per_sec) /
+					samples_per_sec)),
+				esamples_percent);
+	} else {
+		float us_samples_per_sec = top->us_samples / top->delay_secs;
+		float guest_kernel_samples_per_sec = top->guest_kernel_samples / top->delay_secs;
+		float guest_us_samples_per_sec = top->guest_us_samples / top->delay_secs;
+
+		ret = SNPRINTF(bf, size,
+			       "   PerfTop:%8.0f irqs/sec  kernel:%4.1f%% us:%4.1f%%"
+			       " guest kernel:%4.1f%% guest us:%4.1f%%"
+			       " exact: %4.1f%% [", samples_per_sec,
+			       100.0 - (100.0 * ((samples_per_sec - ksamples_per_sec) /
+						 samples_per_sec)),
+			       100.0 - (100.0 * ((samples_per_sec - us_samples_per_sec) /
+						 samples_per_sec)),
+			       100.0 - (100.0 * ((samples_per_sec -
+						  guest_kernel_samples_per_sec) /
+						 samples_per_sec)),
+			       100.0 - (100.0 * ((samples_per_sec -
+						  guest_us_samples_per_sec) /
+						 samples_per_sec)),
+			       esamples_percent);
+	}
+
+	if (top->evlist->nr_entries == 1 || !top->display_weighted) {
+		struct perf_evsel *first;
+		first = list_entry(top->evlist->entries.next, struct perf_evsel, node);
+		ret += SNPRINTF(bf + ret, size - ret, "%" PRIu64 "%s ",
+				(uint64_t)first->attr.sample_period,
+				top->freq ? "Hz" : "");
+	}
+
+	if (!top->display_weighted) {
+		ret += SNPRINTF(bf + ret, size - ret, "%s",
+				event_name(top->sym_evsel));
+	} else {
+		/*
+		 * Don't let events eat all the space. Leaving 30 bytes
+		 * for the rest should be enough.
+		 */
+		size_t last_pos = size - 30;
+
+		list_for_each_entry(counter, &top->evlist->entries, node) {
+			ret += SNPRINTF(bf + ret, size - ret, "%s%s",
+					counter->idx ? "/" : "",
+					event_name(counter));
+			if (ret > last_pos) {
+				sprintf(bf + last_pos - 3, "..");
+				ret = last_pos - 1;
+				break;
+			}
+		}
+	}
+
+	ret += SNPRINTF(bf + ret, size - ret, "], ");
+
+	if (top->target_pid != -1)
+		ret += SNPRINTF(bf + ret, size - ret, " (target_pid: %d",
+				top->target_pid);
+	else if (top->target_tid != -1)
+		ret += SNPRINTF(bf + ret, size - ret, " (target_tid: %d",
+				top->target_tid);
+	else
+		ret += SNPRINTF(bf + ret, size - ret, " (all");
+
+	if (top->cpu_list)
+		ret += SNPRINTF(bf + ret, size - ret, ", CPU%s: %s)",
+				top->evlist->cpus->nr > 1 ? "s" : "", top->cpu_list);
+	else {
+		if (top->target_tid != -1)
+			ret += SNPRINTF(bf + ret, size - ret, ")");
+		else
+			ret += SNPRINTF(bf + ret, size - ret, ", %d CPU%s)",
+					top->evlist->cpus->nr,
+					top->evlist->cpus->nr > 1 ? "s" : "");
+	}
+
+	return ret;
+}
+
+void perf_top__reset_sample_counters(struct perf_top *top)
+{
+	top->samples = top->us_samples = top->kernel_samples =
+	top->exact_samples = top->guest_kernel_samples =
+	top->guest_us_samples = 0;
+}
+
+float perf_top__decay_samples(struct perf_top *top, struct rb_root *root)
+{
+	struct sym_entry *syme, *n;
+	float sum_ksamples = 0.0;
+	int snap = !top->display_weighted ? top->sym_counter : 0, j;
+
+	/* Sort the active symbols */
+	pthread_mutex_lock(&top->active_symbols_lock);
+	syme = list_entry(top->active_symbols.next, struct sym_entry, node);
+	pthread_mutex_unlock(&top->active_symbols_lock);
+
+	top->rb_entries = 0;
+	list_for_each_entry_safe_from(syme, n, &top->active_symbols, node) {
+		syme->snap_count = syme->count[snap];
+		if (syme->snap_count != 0) {
+
+			if ((top->hide_user_symbols &&
+			     syme->origin == PERF_RECORD_MISC_USER) ||
+			    (top->hide_kernel_symbols &&
+			     syme->origin == PERF_RECORD_MISC_KERNEL)) {
+				perf_top__remove_active_sym(top, syme);
+				continue;
+			}
+			syme->weight = sym_weight(syme, top);
+
+			if ((int)syme->snap_count >= top->count_filter) {
+				rb_insert_active_sym(root, syme);
+				++top->rb_entries;
+			}
+			sum_ksamples += syme->snap_count;
+
+			for (j = 0; j < top->evlist->nr_entries; j++)
+				syme->count[j] = top->zero ? 0 : syme->count[j] * 7 / 8;
+		} else
+			perf_top__remove_active_sym(top, syme);
+	}
+
+	return sum_ksamples;
+}
+
+/*
+ * Find the longest symbol name that will be displayed
+ */
+void perf_top__find_widths(struct perf_top *top, struct rb_root *root,
+			   int *dso_width, int *dso_short_width, int *sym_width)
+{
+	struct rb_node *nd;
+	int printed = 0;
+
+	*sym_width = *dso_width = *dso_short_width = 0;
+
+	for (nd = rb_first(root); nd; nd = rb_next(nd)) {
+		struct sym_entry *syme = rb_entry(nd, struct sym_entry, rb_node);
+		struct symbol *sym = sym_entry__symbol(syme);
+
+		if (++printed > top->print_entries ||
+		    (int)syme->snap_count < top->count_filter)
+			continue;
+
+		if (syme->map->dso->long_name_len > *dso_width)
+			*dso_width = syme->map->dso->long_name_len;
+
+		if (syme->map->dso->short_name_len > *dso_short_width)
+			*dso_short_width = syme->map->dso->short_name_len;
+
+		if (sym->namelen > *sym_width)
+			*sym_width = sym->namelen;
+	}
+}
diff --git a/tools/perf/util/top.h b/tools/perf/util/top.h
new file mode 100644
index 000000000000..96d1cb78af01
--- /dev/null
+++ b/tools/perf/util/top.h
@@ -0,0 +1,66 @@
+#ifndef __PERF_TOP_H
+#define __PERF_TOP_H 1
+
+#include "types.h"
+#include "../perf.h"
+#include <stddef.h>
+#include <pthread.h>
+#include <linux/list.h>
+#include <linux/rbtree.h>
+
+struct perf_evlist;
+struct perf_evsel;
+
+struct sym_entry {
+	struct rb_node		rb_node;
+	struct list_head	node;
+	unsigned long		snap_count;
+	double			weight;
+	int			skip;
+	u8			origin;
+	struct map		*map;
+	unsigned long		count[0];
+};
+
+static inline struct symbol *sym_entry__symbol(struct sym_entry *self)
+{
+       return ((void *)self) + symbol_conf.priv_size;
+}
+
+struct perf_top {
+	struct perf_evlist *evlist;
+	/*
+	 * Symbols will be added here in perf_event__process_sample and will
+	 * get out after decayed.
+	 */
+	struct list_head   active_symbols;
+	pthread_mutex_t	   active_symbols_lock;
+	pthread_cond_t	   active_symbols_cond;
+	u64		   samples;
+	u64		   kernel_samples, us_samples;
+	u64		   exact_samples;
+	u64		   guest_us_samples, guest_kernel_samples;
+	int		   print_entries, count_filter, delay_secs;
+	int		   display_weighted, freq, rb_entries, sym_counter;
+	pid_t		   target_pid, target_tid;
+	bool		   hide_kernel_symbols, hide_user_symbols, zero;
+	const char	   *cpu_list;
+	struct sym_entry   *sym_filter_entry;
+	struct perf_evsel  *sym_evsel;
+};
+
+size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size);
+void perf_top__reset_sample_counters(struct perf_top *top);
+float perf_top__decay_samples(struct perf_top *top, struct rb_root *root);
+void perf_top__find_widths(struct perf_top *top, struct rb_root *root,
+			   int *dso_width, int *dso_short_width, int *sym_width);
+
+#ifdef NO_NEWT_SUPPORT
+static inline int perf_top__tui_browser(struct perf_top *top __used)
+{
+	return 0;
+}
+#else
+int perf_top__tui_browser(struct perf_top *top);
+#endif
+#endif /* __PERF_TOP_H */
diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c
index 73a02223c629..d8e622dd738a 100644
--- a/tools/perf/util/trace-event-parse.c
+++ b/tools/perf/util/trace-event-parse.c
@@ -153,7 +153,7 @@ void parse_proc_kallsyms(char *file, unsigned int size __unused)
 	char *next = NULL;
 	char *addr_str;
 	char ch;
-	int ret;
+	int ret __used;
 	int i;
 
 	line = strtok_r(file, "\n", &next);
diff --git a/tools/perf/util/ui/browser.c b/tools/perf/util/ui/browser.c
index 8bc010edca25..611219f80680 100644
--- a/tools/perf/util/ui/browser.c
+++ b/tools/perf/util/ui/browser.c
@@ -1,4 +1,5 @@
 #include "libslang.h"
+#include "ui.h"
 #include <linux/compiler.h>
 #include <linux/list.h>
 #include <linux/rbtree.h>
@@ -156,6 +157,20 @@ void ui_browser__add_exit_keys(struct ui_browser *self, int keys[])
 	}
 }
 
+void __ui_browser__show_title(struct ui_browser *browser, const char *title)
+{
+	SLsmg_gotorc(0, 0);
+	ui_browser__set_color(browser, NEWT_COLORSET_ROOT);
+	slsmg_write_nstring(title, browser->width);
+}
+
+void ui_browser__show_title(struct ui_browser *browser, const char *title)
+{
+	pthread_mutex_lock(&ui__lock);
+	__ui_browser__show_title(browser, title);
+	pthread_mutex_unlock(&ui__lock);
+}
+
 int ui_browser__show(struct ui_browser *self, const char *title,
 		     const char *helpline, ...)
 {
@@ -178,9 +193,8 @@ int ui_browser__show(struct ui_browser *self, const char *title,
 	if (self->sb == NULL)
 		return -1;
 
-	SLsmg_gotorc(0, 0);
-	ui_browser__set_color(self, NEWT_COLORSET_ROOT);
-	slsmg_write_nstring(title, self->width);
+	pthread_mutex_lock(&ui__lock);
+	__ui_browser__show_title(self, title);
 
 	ui_browser__add_exit_keys(self, keys);
 	newtFormAddComponent(self->form, self->sb);
@@ -188,25 +202,30 @@ int ui_browser__show(struct ui_browser *self, const char *title,
 	va_start(ap, helpline);
 	ui_helpline__vpush(helpline, ap);
 	va_end(ap);
+	pthread_mutex_unlock(&ui__lock);
 	return 0;
 }
 
 void ui_browser__hide(struct ui_browser *self)
 {
+	pthread_mutex_lock(&ui__lock);
 	newtFormDestroy(self->form);
 	self->form = NULL;
 	ui_helpline__pop();
+	pthread_mutex_unlock(&ui__lock);
 }
 
 int ui_browser__refresh(struct ui_browser *self)
 {
 	int row;
 
+	pthread_mutex_lock(&ui__lock);
 	newtScrollbarSet(self->sb, self->index, self->nr_entries - 1);
 	row = self->refresh(self);
 	ui_browser__set_color(self, HE_COLORSET_NORMAL);
 	SLsmg_fill_region(self->y + row, self->x,
 			  self->height - row, self->width, ' ');
+	pthread_mutex_unlock(&ui__lock);
 
 	return 0;
 }
diff --git a/tools/perf/util/ui/browser.h b/tools/perf/util/ui/browser.h
index 0dc7e4da36f5..fc63dda10910 100644
--- a/tools/perf/util/ui/browser.h
+++ b/tools/perf/util/ui/browser.h
@@ -24,7 +24,6 @@ struct ui_browser {
 	u32	      nr_entries;
 };
 
-
 void ui_browser__set_color(struct ui_browser *self, int color);
 void ui_browser__set_percent_color(struct ui_browser *self,
 				   double percent, bool current);
@@ -35,6 +34,8 @@ void ui_browser__reset_index(struct ui_browser *self);
 void ui_browser__gotorc(struct ui_browser *self, int y, int x);
 void ui_browser__add_exit_key(struct ui_browser *self, int key);
 void ui_browser__add_exit_keys(struct ui_browser *self, int keys[]);
+void __ui_browser__show_title(struct ui_browser *browser, const char *title);
+void ui_browser__show_title(struct ui_browser *browser, const char *title);
 int ui_browser__show(struct ui_browser *self, const char *title,
 		     const char *helpline, ...);
 void ui_browser__hide(struct ui_browser *self);
diff --git a/tools/perf/util/ui/browsers/annotate.c b/tools/perf/util/ui/browsers/annotate.c
index 82b78f99251b..8c17a8730e4a 100644
--- a/tools/perf/util/ui/browsers/annotate.c
+++ b/tools/perf/util/ui/browsers/annotate.c
@@ -1,9 +1,12 @@
 #include "../browser.h"
 #include "../helpline.h"
 #include "../libslang.h"
+#include "../../annotate.h"
 #include "../../hist.h"
 #include "../../sort.h"
 #include "../../symbol.h"
+#include "../../annotate.h"
+#include <pthread.h>
 
 static void ui__error_window(const char *fmt, ...)
 {
@@ -42,8 +45,6 @@ static void annotate_browser__write(struct ui_browser *self, void *entry, int ro
 		struct objdump_line_rb_node *olrb = objdump_line__rb(ol);
 		ui_browser__set_percent_color(self, olrb->percent, current_entry);
 		slsmg_printf(" %7.2f ", olrb->percent);
-		if (!current_entry)
-			ui_browser__set_color(self, HE_COLORSET_CODE);
 	} else {
 		ui_browser__set_percent_color(self, 0, current_entry);
 		slsmg_write_nstring(" ", 9);
@@ -55,35 +56,40 @@ static void annotate_browser__write(struct ui_browser *self, void *entry, int ro
 		slsmg_write_nstring(" ", width - 18);
 	else
 		slsmg_write_nstring(ol->line, width - 18);
+
+	if (!current_entry)
+		ui_browser__set_color(self, HE_COLORSET_CODE);
 }
 
 static double objdump_line__calc_percent(struct objdump_line *self,
-					 struct list_head *head,
-					 struct symbol *sym)
+					 struct symbol *sym, int evidx)
 {
 	double percent = 0.0;
 
 	if (self->offset != -1) {
 		int len = sym->end - sym->start;
 		unsigned int hits = 0;
-		struct sym_priv *priv = symbol__priv(sym);
-		struct sym_ext *sym_ext = priv->ext;
-		struct sym_hist *h = priv->hist;
+		struct annotation *notes = symbol__annotation(sym);
+		struct source_line *src_line = notes->src->lines;
+		struct sym_hist *h = annotation__histogram(notes, evidx);
 		s64 offset = self->offset;
-		struct objdump_line *next = objdump__get_next_ip_line(head, self);
-
+		struct objdump_line *next;
 
+		next = objdump__get_next_ip_line(&notes->src->source, self);
 		while (offset < (s64)len &&
 		       (next == NULL || offset < next->offset)) {
-			if (sym_ext) {
-				percent += sym_ext[offset].percent;
+			if (src_line) {
+				percent += src_line[offset].percent;
 			} else
-				hits += h->ip[offset];
+				hits += h->addr[offset];
 
 			++offset;
 		}
-
-		if (sym_ext == NULL && h->sum)
+		/*
+ 		 * If the percentage wasn't already calculated in
+ 		 * symbol__get_source_line, do it now:
+ 		 */
+		if (src_line == NULL && h->sum)
 			percent = 100.0 * hits / h->sum;
 	}
 
@@ -133,103 +139,161 @@ static void annotate_browser__set_top(struct annotate_browser *self,
 	self->curr_hot = nd;
 }
 
-static int annotate_browser__run(struct annotate_browser *self)
+static void annotate_browser__calc_percent(struct annotate_browser *browser,
+					   int evidx)
 {
-	struct rb_node *nd;
-	struct hist_entry *he = self->b.priv;
-	int key;
+	struct symbol *sym = browser->b.priv;
+	struct annotation *notes = symbol__annotation(sym);
+	struct objdump_line *pos;
 
-	if (ui_browser__show(&self->b, he->ms.sym->name,
-			     "<-, -> or ESC: exit, TAB/shift+TAB: cycle thru samples") < 0)
-		return -1;
+	browser->entries = RB_ROOT;
+
+	pthread_mutex_lock(&notes->lock);
+
+	list_for_each_entry(pos, &notes->src->source, node) {
+		struct objdump_line_rb_node *rbpos = objdump_line__rb(pos);
+		rbpos->percent = objdump_line__calc_percent(pos, sym, evidx);
+		if (rbpos->percent < 0.01) {
+			RB_CLEAR_NODE(&rbpos->rb_node);
+			continue;
+		}
+		objdump__insert_line(&browser->entries, rbpos);
+	}
+	pthread_mutex_unlock(&notes->lock);
+
+	browser->curr_hot = rb_last(&browser->entries);
+}
+
+static int annotate_browser__run(struct annotate_browser *self, int evidx,
+				 int refresh)
+{
+	struct rb_node *nd = NULL;
+	struct symbol *sym = self->b.priv;
 	/*
-	 * To allow builtin-annotate to cycle thru multiple symbols by
+	 * RIGHT To allow builtin-annotate to cycle thru multiple symbols by
 	 * examining the exit key for this function.
 	 */
-	ui_browser__add_exit_key(&self->b, NEWT_KEY_RIGHT);
+	int exit_keys[] = { 'H', NEWT_KEY_TAB, NEWT_KEY_UNTAB,
+			    NEWT_KEY_RIGHT, 0 };
+	int key;
+
+	if (ui_browser__show(&self->b, sym->name,
+			     "<-, -> or ESC: exit, TAB/shift+TAB: "
+			     "cycle hottest lines, H: Hottest") < 0)
+		return -1;
+
+	ui_browser__add_exit_keys(&self->b, exit_keys);
+	annotate_browser__calc_percent(self, evidx);
+
+	if (self->curr_hot)
+		annotate_browser__set_top(self, self->curr_hot);
 
 	nd = self->curr_hot;
-	if (nd) {
-		int tabs[] = { NEWT_KEY_TAB, NEWT_KEY_UNTAB, 0 };
-		ui_browser__add_exit_keys(&self->b, tabs);
-	}
+
+	if (refresh != 0)
+		newtFormSetTimer(self->b.form, refresh);
 
 	while (1) {
 		key = ui_browser__run(&self->b);
 
+		if (refresh != 0) {
+			annotate_browser__calc_percent(self, evidx);
+			/*
+			 * Current line focus got out of the list of most active
+			 * lines, NULL it so that if TAB|UNTAB is pressed, we
+			 * move to curr_hot (current hottest line).
+			 */
+			if (nd != NULL && RB_EMPTY_NODE(nd))
+				nd = NULL;
+		}
+
 		switch (key) {
+		case -1:
+			/*
+ 			 * FIXME we need to check if it was
+ 			 * es.reason == NEWT_EXIT_TIMER
+ 			 */
+			if (refresh != 0)
+				symbol__annotate_decay_histogram(sym, evidx);
+			continue;
 		case NEWT_KEY_TAB:
-			nd = rb_prev(nd);
-			if (nd == NULL)
-				nd = rb_last(&self->entries);
-			annotate_browser__set_top(self, nd);
+			if (nd != NULL) {
+				nd = rb_prev(nd);
+				if (nd == NULL)
+					nd = rb_last(&self->entries);
+			} else
+				nd = self->curr_hot;
 			break;
 		case NEWT_KEY_UNTAB:
-			nd = rb_next(nd);
-			if (nd == NULL)
-				nd = rb_first(&self->entries);
-			annotate_browser__set_top(self, nd);
+			if (nd != NULL)
+				nd = rb_next(nd);
+				if (nd == NULL)
+					nd = rb_first(&self->entries);
+			else
+				nd = self->curr_hot;
+			break;
+		case 'H':
+			nd = self->curr_hot;
 			break;
 		default:
 			goto out;
 		}
+
+		if (nd != NULL)
+			annotate_browser__set_top(self, nd);
 	}
 out:
 	ui_browser__hide(&self->b);
 	return key;
 }
 
-int hist_entry__tui_annotate(struct hist_entry *self)
+int hist_entry__tui_annotate(struct hist_entry *he, int evidx)
+{
+	return symbol__tui_annotate(he->ms.sym, he->ms.map, evidx, 0);
+}
+
+int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx,
+			 int refresh)
 {
 	struct objdump_line *pos, *n;
-	struct objdump_line_rb_node *rbpos;
-	LIST_HEAD(head);
+	struct annotation *notes = symbol__annotation(sym);
 	struct annotate_browser browser = {
 		.b = {
-			.entries = &head,
+			.entries = &notes->src->source,
 			.refresh = ui_browser__list_head_refresh,
 			.seek	 = ui_browser__list_head_seek,
 			.write	 = annotate_browser__write,
-			.priv	 = self,
+			.priv	 = sym,
 		},
 	};
 	int ret;
 
-	if (self->ms.sym == NULL)
+	if (sym == NULL)
 		return -1;
 
-	if (self->ms.map->dso->annotate_warned)
+	if (map->dso->annotate_warned)
 		return -1;
 
-	if (hist_entry__annotate(self, &head, sizeof(*rbpos)) < 0) {
+	if (symbol__annotate(sym, map, sizeof(struct objdump_line_rb_node)) < 0) {
 		ui__error_window(ui_helpline__last_msg);
 		return -1;
 	}
 
 	ui_helpline__push("Press <- or ESC to exit");
 
-	list_for_each_entry(pos, &head, node) {
+	list_for_each_entry(pos, &notes->src->source, node) {
+		struct objdump_line_rb_node *rbpos;
 		size_t line_len = strlen(pos->line);
+
 		if (browser.b.width < line_len)
 			browser.b.width = line_len;
 		rbpos = objdump_line__rb(pos);
 		rbpos->idx = browser.b.nr_entries++;
-		rbpos->percent = objdump_line__calc_percent(pos, &head, self->ms.sym);
-		if (rbpos->percent < 0.01)
-			continue;
-		objdump__insert_line(&browser.entries, rbpos);
 	}
 
-	/*
-	 * Position the browser at the hottest line.
-	 */
-	browser.curr_hot = rb_last(&browser.entries);
-	if (browser.curr_hot)
-		annotate_browser__set_top(&browser, browser.curr_hot);
-
 	browser.b.width += 18; /* Percentage */
-	ret = annotate_browser__run(&browser);
-	list_for_each_entry_safe(pos, n, &head, node) {
+	ret = annotate_browser__run(&browser, evidx, refresh);
+	list_for_each_entry_safe(pos, n, &notes->src->source, node) {
 		list_del(&pos->node);
 		objdump_line__free(pos);
 	}
diff --git a/tools/perf/util/ui/browsers/hists.c b/tools/perf/util/ui/browsers/hists.c
index 60c463c16028..798efdca3ead 100644
--- a/tools/perf/util/ui/browsers/hists.c
+++ b/tools/perf/util/ui/browsers/hists.c
@@ -7,6 +7,8 @@
 #include <newt.h>
 #include <linux/rbtree.h>
 
+#include "../../evsel.h"
+#include "../../evlist.h"
 #include "../../hist.h"
 #include "../../pstack.h"
 #include "../../sort.h"
@@ -292,7 +294,8 @@ static int hist_browser__run(struct hist_browser *self, const char *title)
 {
 	int key;
 	int exit_keys[] = { 'a', '?', 'h', 'C', 'd', 'D', 'E', 't',
-			    NEWT_KEY_ENTER, NEWT_KEY_RIGHT, NEWT_KEY_LEFT, 0, };
+			    NEWT_KEY_ENTER, NEWT_KEY_RIGHT, NEWT_KEY_LEFT,
+			    NEWT_KEY_TAB, NEWT_KEY_UNTAB, 0, };
 
 	self->b.entries = &self->hists->entries;
 	self->b.nr_entries = self->hists->nr_entries;
@@ -377,7 +380,7 @@ static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self,
 	while (node) {
 		struct callchain_node *child = rb_entry(node, struct callchain_node, rb_node);
 		struct rb_node *next = rb_next(node);
-		u64 cumul = cumul_hits(child);
+		u64 cumul = callchain_cumul_hits(child);
 		struct callchain_list *chain;
 		char folded_sign = ' ';
 		int first = true;
@@ -638,6 +641,9 @@ static void ui_browser__hists_seek(struct ui_browser *self,
 	struct rb_node *nd;
 	bool first = true;
 
+	if (self->nr_entries == 0)
+		return;
+
 	switch (whence) {
 	case SEEK_SET:
 		nd = hists__filter_entries(rb_first(self->entries));
@@ -797,8 +803,11 @@ static int hists__browser_title(struct hists *self, char *bf, size_t size,
 	return printed;
 }
 
-int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
+static int perf_evsel__hists_browse(struct perf_evsel *evsel,
+				    const char *helpline, const char *ev_name,
+				    bool left_exits)
 {
+	struct hists *self = &evsel->hists;
 	struct hist_browser *browser = hist_browser__new(self);
 	struct pstack *fstack;
 	const struct thread *thread_filter = NULL;
@@ -818,8 +827,8 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 	hists__browser_title(self, msg, sizeof(msg), ev_name,
 			     dso_filter, thread_filter);
 	while (1) {
-		const struct thread *thread;
-		const struct dso *dso;
+		const struct thread *thread = NULL;
+		const struct dso *dso = NULL;
 		char *options[16];
 		int nr_options = 0, choice = 0, i,
 		    annotate = -2, zoom_dso = -2, zoom_thread = -2,
@@ -827,8 +836,10 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 
 		key = hist_browser__run(browser, msg);
 
-		thread = hist_browser__selected_thread(browser);
-		dso = browser->selection->map ? browser->selection->map->dso : NULL;
+		if (browser->he_selection != NULL) {
+			thread = hist_browser__selected_thread(browser);
+			dso = browser->selection->map ? browser->selection->map->dso : NULL;
+		}
 
 		switch (key) {
 		case NEWT_KEY_TAB:
@@ -839,7 +850,8 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 			 */
 			goto out_free_stack;
 		case 'a':
-			if (browser->selection->map == NULL &&
+			if (browser->selection == NULL ||
+			    browser->selection->map == NULL ||
 			    browser->selection->map->dso->annotate_warned)
 				continue;
 			goto do_annotate;
@@ -858,6 +870,7 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 					"E         Expand all callchains\n"
 					"d         Zoom into current DSO\n"
 					"t         Zoom into current Thread\n"
+					"TAB/UNTAB Switch events\n"
 					"q/CTRL+C  Exit browser");
 			continue;
 		case NEWT_KEY_ENTER:
@@ -867,8 +880,14 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 		case NEWT_KEY_LEFT: {
 			const void *top;
 
-			if (pstack__empty(fstack))
+			if (pstack__empty(fstack)) {
+				/*
+				 * Go back to the perf_evsel_menu__run or other user
+				 */
+				if (left_exits)
+					goto out_free_stack;
 				continue;
+			}
 			top = pstack__pop(fstack);
 			if (top == &dso_filter)
 				goto zoom_out_dso;
@@ -877,14 +896,16 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 			continue;
 		}
 		case NEWT_KEY_ESCAPE:
-			if (!ui__dialog_yesno("Do you really want to exit?"))
+			if (!left_exits &&
+			    !ui__dialog_yesno("Do you really want to exit?"))
 				continue;
 			/* Fall thru */
 		default:
 			goto out_free_stack;
 		}
 
-		if (browser->selection->sym != NULL &&
+		if (browser->selection != NULL &&
+		    browser->selection->sym != NULL &&
 		    !browser->selection->map->dso->annotate_warned &&
 		    asprintf(&options[nr_options], "Annotate %s",
 			     browser->selection->sym->name) > 0)
@@ -903,7 +924,8 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 			     (dso->kernel ? "the Kernel" : dso->short_name)) > 0)
 			zoom_dso = nr_options++;
 
-		if (browser->selection->map != NULL &&
+		if (browser->selection != NULL &&
+		    browser->selection->map != NULL &&
 		    asprintf(&options[nr_options], "Browse map details") > 0)
 			browse_map = nr_options++;
 
@@ -923,19 +945,11 @@ int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
 		if (choice == annotate) {
 			struct hist_entry *he;
 do_annotate:
-			if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) {
-				browser->selection->map->dso->annotate_warned = 1;
-				ui_helpline__puts("No vmlinux file found, can't "
-						 "annotate with just a "
-						 "kallsyms file");
-				continue;
-			}
-
 			he = hist_browser__selected_entry(browser);
 			if (he == NULL)
 				continue;
 
-			hist_entry__tui_annotate(he);
+			hist_entry__tui_annotate(he, evsel->idx);
 		} else if (choice == browse_map)
 			map__browse(browser->selection->map);
 		else if (choice == zoom_dso) {
@@ -984,30 +998,141 @@ out:
 	return key;
 }
 
-int hists__tui_browse_tree(struct rb_root *self, const char *help)
+struct perf_evsel_menu {
+	struct ui_browser b;
+	struct perf_evsel *selection;
+};
+
+static void perf_evsel_menu__write(struct ui_browser *browser,
+				   void *entry, int row)
+{
+	struct perf_evsel_menu *menu = container_of(browser,
+						    struct perf_evsel_menu, b);
+	struct perf_evsel *evsel = list_entry(entry, struct perf_evsel, node);
+	bool current_entry = ui_browser__is_current_entry(browser, row);
+	unsigned long nr_events = evsel->hists.stats.nr_events[PERF_RECORD_SAMPLE];
+	const char *ev_name = event_name(evsel);
+	char bf[256], unit;
+
+	ui_browser__set_color(browser, current_entry ? HE_COLORSET_SELECTED :
+						       HE_COLORSET_NORMAL);
+
+	nr_events = convert_unit(nr_events, &unit);
+	snprintf(bf, sizeof(bf), "%lu%c%s%s", nr_events,
+		 unit, unit == ' ' ? "" : " ", ev_name);
+	slsmg_write_nstring(bf, browser->width);
+
+	if (current_entry)
+		menu->selection = evsel;
+}
+
+static int perf_evsel_menu__run(struct perf_evsel_menu *menu, const char *help)
 {
-	struct rb_node *first = rb_first(self), *nd = first, *next;
-	int key = 0;
+	int exit_keys[] = { NEWT_KEY_ENTER, NEWT_KEY_RIGHT, 0, };
+	struct perf_evlist *evlist = menu->b.priv;
+	struct perf_evsel *pos;
+	const char *ev_name, *title = "Available samples";
+	int key;
+
+	if (ui_browser__show(&menu->b, title,
+			     "ESC: exit, ENTER|->: Browse histograms") < 0)
+		return -1;
+
+	ui_browser__add_exit_keys(&menu->b, exit_keys);
 
-	while (nd) {
-		struct hists *hists = rb_entry(nd, struct hists, rb_node);
-		const char *ev_name = __event_name(hists->type, hists->config);
+	while (1) {
+		key = ui_browser__run(&menu->b);
 
-		key = hists__browse(hists, help, ev_name);
 		switch (key) {
-		case NEWT_KEY_TAB:
-			next = rb_next(nd);
-			if (next)
-				nd = next;
+		case NEWT_KEY_RIGHT:
+		case NEWT_KEY_ENTER:
+			if (!menu->selection)
+				continue;
+			pos = menu->selection;
+browse_hists:
+			ev_name = event_name(pos);
+			key = perf_evsel__hists_browse(pos, help, ev_name, true);
+			ui_browser__show_title(&menu->b, title);
 			break;
-		case NEWT_KEY_UNTAB:
-			if (nd == first)
+		case NEWT_KEY_LEFT:
+			continue;
+		case NEWT_KEY_ESCAPE:
+			if (!ui__dialog_yesno("Do you really want to exit?"))
 				continue;
-			nd = rb_prev(nd);
+			/* Fall thru */
+		default:
+			goto out;
+		}
+
+		switch (key) {
+		case NEWT_KEY_TAB:
+			if (pos->node.next == &evlist->entries)
+				pos = list_entry(evlist->entries.next, struct perf_evsel, node);
+			else
+				pos = list_entry(pos->node.next, struct perf_evsel, node);
+			goto browse_hists;
+		case NEWT_KEY_UNTAB:
+			if (pos->node.prev == &evlist->entries)
+				pos = list_entry(evlist->entries.prev, struct perf_evsel, node);
+			else
+				pos = list_entry(pos->node.prev, struct perf_evsel, node);
+			goto browse_hists;
+		case 'q':
+		case CTRL('c'):
+			goto out;
 		default:
-			return key;
+			break;
 		}
 	}
 
+out:
+	ui_browser__hide(&menu->b);
 	return key;
 }
+
+static int __perf_evlist__tui_browse_hists(struct perf_evlist *evlist,
+					   const char *help)
+{
+	struct perf_evsel *pos;
+	struct perf_evsel_menu menu = {
+		.b = {
+			.entries    = &evlist->entries,
+			.refresh    = ui_browser__list_head_refresh,
+			.seek	    = ui_browser__list_head_seek,
+			.write	    = perf_evsel_menu__write,
+			.nr_entries = evlist->nr_entries,
+			.priv	    = evlist,
+		},
+	};
+
+	ui_helpline__push("Press ESC to exit");
+
+	list_for_each_entry(pos, &evlist->entries, node) {
+		const char *ev_name = event_name(pos);
+		size_t line_len = strlen(ev_name) + 7;
+
+		if (menu.b.width < line_len)
+			menu.b.width = line_len;
+		/*
+		 * Cache the evsel name, tracepoints have a _high_ cost per
+		 * event_name() call.
+		 */
+		if (pos->name == NULL)
+			pos->name = strdup(ev_name);
+	}
+
+	return perf_evsel_menu__run(&menu, help);
+}
+
+int perf_evlist__tui_browse_hists(struct perf_evlist *evlist, const char *help)
+{
+
+	if (evlist->nr_entries == 1) {
+		struct perf_evsel *first = list_entry(evlist->entries.next,
+						      struct perf_evsel, node);
+		const char *ev_name = event_name(first);
+		return perf_evsel__hists_browse(first, help, ev_name, false);
+	}
+
+	return __perf_evlist__tui_browse_hists(evlist, help);
+}
diff --git a/tools/perf/util/ui/browsers/map.c b/tools/perf/util/ui/browsers/map.c
index e5158369106e..8462bffe20bc 100644
--- a/tools/perf/util/ui/browsers/map.c
+++ b/tools/perf/util/ui/browsers/map.c
@@ -41,7 +41,7 @@ static int ui_entry__read(const char *title, char *bf, size_t size, int width)
 out_free_form:
 	newtPopWindow();
 	newtFormDestroy(form);
-	return 0;
+	return err;
 }
 
 struct map_browser {
diff --git a/tools/perf/util/ui/browsers/top.c b/tools/perf/util/ui/browsers/top.c
new file mode 100644
index 000000000000..5a06538532af
--- /dev/null
+++ b/tools/perf/util/ui/browsers/top.c
@@ -0,0 +1,213 @@
+/*
+ * Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Parts came from builtin-{top,stat,record}.c, see those files for further
+ * copyright notes.
+ *
+ * Released under the GPL v2. (and only v2, not any later version)
+ */
+#include "../browser.h"
+#include "../../annotate.h"
+#include "../helpline.h"
+#include "../libslang.h"
+#include "../util.h"
+#include "../../evlist.h"
+#include "../../hist.h"
+#include "../../sort.h"
+#include "../../symbol.h"
+#include "../../top.h"
+
+struct perf_top_browser {
+	struct ui_browser b;
+	struct rb_root	  root;
+	struct sym_entry  *selection;
+	float		  sum_ksamples;
+	int		  dso_width;
+	int		  dso_short_width;
+	int		  sym_width;
+};
+
+static void perf_top_browser__write(struct ui_browser *browser, void *entry, int row)
+{
+	struct perf_top_browser *top_browser = container_of(browser, struct perf_top_browser, b);
+	struct sym_entry *syme = rb_entry(entry, struct sym_entry, rb_node);
+	bool current_entry = ui_browser__is_current_entry(browser, row);
+	struct symbol *symbol = sym_entry__symbol(syme);
+	struct perf_top *top = browser->priv;
+	int width = browser->width;
+	double pcnt;
+
+	pcnt = 100.0 - (100.0 * ((top_browser->sum_ksamples - syme->snap_count) /
+				 top_browser->sum_ksamples));
+	ui_browser__set_percent_color(browser, pcnt, current_entry);
+
+	if (top->evlist->nr_entries == 1 || !top->display_weighted) {
+		slsmg_printf("%20.2f ", syme->weight);
+		width -= 24;
+	} else {
+		slsmg_printf("%9.1f %10ld ", syme->weight, syme->snap_count);
+		width -= 23;
+	}
+
+	slsmg_printf("%4.1f%%", pcnt);
+	width -= 7;
+
+	if (verbose) {
+		slsmg_printf(" %016" PRIx64, symbol->start);
+		width -= 17;
+	}
+
+	slsmg_printf(" %-*.*s ", top_browser->sym_width, top_browser->sym_width,
+		     symbol->name);
+	width -= top_browser->sym_width;
+	slsmg_write_nstring(width >= syme->map->dso->long_name_len ?
+				syme->map->dso->long_name :
+				syme->map->dso->short_name, width);
+
+	if (current_entry)
+		top_browser->selection = syme;
+}
+
+static void perf_top_browser__update_rb_tree(struct perf_top_browser *browser)
+{
+	struct perf_top *top = browser->b.priv;
+	u64 top_idx = browser->b.top_idx;
+
+	browser->root = RB_ROOT;
+	browser->b.top = NULL;
+	browser->sum_ksamples = perf_top__decay_samples(top, &browser->root);
+	/*
+ 	 * No active symbols
+ 	 */
+	if (top->rb_entries == 0)
+		return;
+
+	perf_top__find_widths(top, &browser->root, &browser->dso_width,
+			      &browser->dso_short_width,
+                              &browser->sym_width);
+	if (browser->sym_width + browser->dso_width > browser->b.width - 29) {
+		browser->dso_width = browser->dso_short_width;
+		if (browser->sym_width + browser->dso_width > browser->b.width - 29)
+			browser->sym_width = browser->b.width - browser->dso_width - 29;
+	}
+
+	/*
+	 * Adjust the ui_browser indexes since the entries in the browser->root
+	 * rb_tree may have changed, then seek it from start, so that we get a
+	 * possible new top of the screen.
+ 	 */
+	browser->b.nr_entries = top->rb_entries;
+
+	if (top_idx >= browser->b.nr_entries) {
+		if (browser->b.height >= browser->b.nr_entries)
+			top_idx = browser->b.nr_entries - browser->b.height;
+		else
+			top_idx = 0;
+	}
+
+	if (browser->b.index >= top_idx + browser->b.height)
+		browser->b.index = top_idx + browser->b.index - browser->b.top_idx;
+
+	if (browser->b.index >= browser->b.nr_entries)
+		browser->b.index = browser->b.nr_entries - 1;
+
+	browser->b.top_idx = top_idx;
+	browser->b.seek(&browser->b, top_idx, SEEK_SET);
+}
+
+static void perf_top_browser__annotate(struct perf_top_browser *browser)
+{
+	struct sym_entry *syme = browser->selection;
+	struct symbol *sym = sym_entry__symbol(syme);
+	struct annotation *notes = symbol__annotation(sym);
+	struct perf_top *top = browser->b.priv;
+
+	if (notes->src != NULL)
+		goto do_annotation;
+
+	pthread_mutex_lock(&notes->lock);
+
+	top->sym_filter_entry = NULL;
+
+	if (symbol__alloc_hist(sym, top->evlist->nr_entries) < 0) {
+		pr_err("Not enough memory for annotating '%s' symbol!\n",
+		       sym->name);
+		pthread_mutex_unlock(&notes->lock);
+		return;
+	}
+
+	top->sym_filter_entry = syme;
+
+	pthread_mutex_unlock(&notes->lock);
+do_annotation:
+	symbol__tui_annotate(sym, syme->map, 0, top->delay_secs * 1000);
+}
+
+static int perf_top_browser__run(struct perf_top_browser *browser)
+{
+	int key;
+	char title[160];
+	struct perf_top *top = browser->b.priv;
+	int delay_msecs = top->delay_secs * 1000;
+	int exit_keys[] = { 'a', NEWT_KEY_ENTER, NEWT_KEY_RIGHT, 0, };
+
+	perf_top_browser__update_rb_tree(browser);
+        perf_top__header_snprintf(top, title, sizeof(title));
+        perf_top__reset_sample_counters(top);
+
+	if (ui_browser__show(&browser->b, title,
+			     "ESC: exit, ENTER|->|a: Live Annotate") < 0)
+		return -1;
+
+	newtFormSetTimer(browser->b.form, delay_msecs);
+	ui_browser__add_exit_keys(&browser->b, exit_keys);
+
+	while (1) {
+		key = ui_browser__run(&browser->b);
+
+		switch (key) {
+		case -1:
+			/* FIXME we need to check if it was es.reason == NEWT_EXIT_TIMER */
+			perf_top_browser__update_rb_tree(browser);
+			perf_top__header_snprintf(top, title, sizeof(title));
+			perf_top__reset_sample_counters(top);
+			ui_browser__set_color(&browser->b, NEWT_COLORSET_ROOT);
+			SLsmg_gotorc(0, 0);
+			slsmg_write_nstring(title, browser->b.width);
+			break;
+		case 'a':
+		case NEWT_KEY_RIGHT:
+		case NEWT_KEY_ENTER:
+			if (browser->selection)
+				perf_top_browser__annotate(browser);
+			break;
+		case NEWT_KEY_LEFT:
+			continue;
+		case NEWT_KEY_ESCAPE:
+			if (!ui__dialog_yesno("Do you really want to exit?"))
+				continue;
+			/* Fall thru */
+		default:
+			goto out;
+		}
+	}
+out:
+	ui_browser__hide(&browser->b);
+	return key;
+}
+
+int perf_top__tui_browser(struct perf_top *top)
+{
+	struct perf_top_browser browser = {
+		.b = {
+			.entries = &browser.root,
+			.refresh = ui_browser__rb_tree_refresh,
+			.seek	 = ui_browser__rb_tree_seek,
+			.write	 = perf_top_browser__write,
+			.priv	 = top,
+		},
+	};
+
+	ui_helpline__push("Press <- or ESC to exit");
+	return perf_top_browser__run(&browser);
+}
diff --git a/tools/perf/util/ui/helpline.c b/tools/perf/util/ui/helpline.c
index 8d79daa4458a..f36d2ff509ed 100644
--- a/tools/perf/util/ui/helpline.c
+++ b/tools/perf/util/ui/helpline.c
@@ -5,6 +5,7 @@
 
 #include "../debug.h"
 #include "helpline.h"
+#include "ui.h"
 
 void ui_helpline__pop(void)
 {
@@ -55,7 +56,8 @@ int ui_helpline__show_help(const char *format, va_list ap)
 	int ret;
 	static int backlog;
 
-        ret = vsnprintf(ui_helpline__last_msg + backlog,
+	pthread_mutex_lock(&ui__lock);
+	ret = vsnprintf(ui_helpline__last_msg + backlog,
 			sizeof(ui_helpline__last_msg) - backlog, format, ap);
 	backlog += ret;
 
@@ -64,6 +66,7 @@ int ui_helpline__show_help(const char *format, va_list ap)
 		newtRefresh();
 		backlog = 0;
 	}
+	pthread_mutex_unlock(&ui__lock);
 
 	return ret;
 }
diff --git a/tools/perf/util/ui/libslang.h b/tools/perf/util/ui/libslang.h
index 5623da8e8080..2b63e1c9b181 100644
--- a/tools/perf/util/ui/libslang.h
+++ b/tools/perf/util/ui/libslang.h
@@ -13,11 +13,11 @@
 
 #if SLANG_VERSION < 20104
 #define slsmg_printf(msg, args...) \
-	SLsmg_printf((char *)msg, ##args)
+	SLsmg_printf((char *)(msg), ##args)
 #define slsmg_write_nstring(msg, len) \
-	SLsmg_write_nstring((char *)msg, len)
+	SLsmg_write_nstring((char *)(msg), len)
 #define sltt_set_color(obj, name, fg, bg) \
-	SLtt_set_color(obj,(char *)name, (char *)fg, (char *)bg)
+	SLtt_set_color(obj,(char *)(name), (char *)(fg), (char *)(bg))
 #else
 #define slsmg_printf SLsmg_printf
 #define slsmg_write_nstring SLsmg_write_nstring
diff --git a/tools/perf/util/ui/setup.c b/tools/perf/util/ui/setup.c
index 662085032eb7..ee46d671db59 100644
--- a/tools/perf/util/ui/setup.c
+++ b/tools/perf/util/ui/setup.c
@@ -6,6 +6,9 @@
 #include "../debug.h"
 #include "browser.h"
 #include "helpline.h"
+#include "ui.h"
+
+pthread_mutex_t ui__lock = PTHREAD_MUTEX_INITIALIZER;
 
 static void newt_suspend(void *d __used)
 {
@@ -14,11 +17,12 @@ static void newt_suspend(void *d __used)
 	newtResume();
 }
 
-void setup_browser(void)
+void setup_browser(bool fallback_to_pager)
 {
 	if (!isatty(1) || !use_browser || dump_trace) {
 		use_browser = 0;
-		setup_pager();
+		if (fallback_to_pager)
+			setup_pager();
 		return;
 	}
 
diff --git a/tools/perf/util/ui/ui.h b/tools/perf/util/ui/ui.h
new file mode 100644
index 000000000000..d264e059c829
--- /dev/null
+++ b/tools/perf/util/ui/ui.h
@@ -0,0 +1,8 @@
+#ifndef _PERF_UI_H_
+#define _PERF_UI_H_ 1
+
+#include <pthread.h>
+
+extern pthread_mutex_t ui__lock;
+
+#endif /* _PERF_UI_H_ */
diff --git a/tools/perf/util/ui/util.c b/tools/perf/util/ui/util.c
index 7b5a8926624e..fdf1fc8f08bc 100644
--- a/tools/perf/util/ui/util.c
+++ b/tools/perf/util/ui/util.c
@@ -9,6 +9,7 @@
 #include "../debug.h"
 #include "browser.h"
 #include "helpline.h"
+#include "ui.h"
 #include "util.h"
 
 static void newt_form__set_exit_keys(newtComponent self)
@@ -118,10 +119,12 @@ void ui__warning(const char *format, ...)
 	va_list args;
 
 	va_start(args, format);
-	if (use_browser > 0)
+	if (use_browser > 0) {
+		pthread_mutex_lock(&ui__lock);
 		newtWinMessagev((char *)warning_str, (char *)ok,
 				(char *)format, args);
-	else
+		pthread_mutex_unlock(&ui__lock);
+	} else
 		vfprintf(stderr, format, args);
 	va_end(args);
 }
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index e833f26f3bfc..fc784284ac8b 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -70,9 +70,7 @@
 #include <sys/poll.h>
 #include <sys/socket.h>
 #include <sys/ioctl.h>
-#ifndef NO_SYS_SELECT_H
 #include <sys/select.h>
-#endif
 #include <netinet/in.h>
 #include <netinet/tcp.h>
 #include <arpa/inet.h>
@@ -83,10 +81,6 @@
 #include "types.h"
 #include <sys/ttydefaults.h>
 
-#ifndef NO_ICONV
-#include <iconv.h>
-#endif
-
 extern const char *graph_line;
 extern const char *graph_dotted_line;
 extern char buildid_dir[];
@@ -236,26 +230,6 @@ static inline int sane_case(int x, int high)
 	return x;
 }
 
-#ifndef DIR_HAS_BSD_GROUP_SEMANTICS
-# define FORCE_DIR_SET_GID S_ISGID
-#else
-# define FORCE_DIR_SET_GID 0
-#endif
-
-#ifdef NO_NSEC
-#undef USE_NSEC
-#define ST_CTIME_NSEC(st) 0
-#define ST_MTIME_NSEC(st) 0
-#else
-#ifdef USE_ST_TIMESPEC
-#define ST_CTIME_NSEC(st) ((unsigned int)((st).st_ctimespec.tv_nsec))
-#define ST_MTIME_NSEC(st) ((unsigned int)((st).st_mtimespec.tv_nsec))
-#else
-#define ST_CTIME_NSEC(st) ((unsigned int)((st).st_ctim.tv_nsec))
-#define ST_MTIME_NSEC(st) ((unsigned int)((st).st_mtim.tv_nsec))
-#endif
-#endif
-
 int mkdir_p(char *path, mode_t mode);
 int copyfile(const char *from, const char *to);