Merge branch 'linus' into x86/cleanups

Merge reason: We were on an older pre-rc1 base, move to almost-rc2.

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

29 files changed, 2200 insertions, 596 deletions

diff --git a/tools/perf/CREDITS b/tools/perf/CREDITS
new file mode 100644
index 000000000000..c2ddcb3acbd0
--- /dev/null
+++ b/tools/perf/CREDITS
@@ -0,0 +1,30 @@
+Most of the infrastructure that 'perf' uses here has been reused
+from the Git project, as of version:
+
+    66996ec: Sync with 1.6.2.4
+
+Here is an (incomplete!) list of main contributors to those files
+in util/* and elsewhere:
+
+ Alex Riesen
+ Christian Couder
+ Dmitry Potapov
+ Jeff King
+ Johannes Schindelin
+ Johannes Sixt
+ Junio C Hamano
+ Linus Torvalds
+ Matthias Kestenholz
+ Michal Ostrowski
+ Miklos Vajna
+ Petr Baudis
+ Pierre Habouzit
+ René Scharfe
+ Samuel Tardieu
+ Shawn O. Pearce
+ Steffen Prohaska
+ Steve Haslam
+
+Thanks guys!
+
+The full history of the files can be found in the upstream Git commits.
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index 52d3fc6846a9..8aa3f8c88707 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -13,13 +13,25 @@ SYNOPSIS
 DESCRIPTION
 -----------
 This command displays the performance counter profile information recorded
-via perf report.
+via perf record.
 
 OPTIONS
 -------
 -i::
 --input=::
         Input file name. (default: perf.data)
+-d::
+--dsos=::
+	Only consider symbols in these dsos. CSV that understands
+	file://filename entries.
+-C::
+--comms=::
+	Only consider symbols in these comms. CSV that understands
+	file://filename entries.
+-S::
+--symbols=::
+	Only consider these symbols. CSV that understands
+	file://filename entries.
 
 SEE ALSO
 --------
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index c368a72721d7..0d74346d21ab 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -8,8 +8,8 @@ perf-stat - Run a command and gather performance counter statistics
 SYNOPSIS
 --------
 [verse]
-'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] <command>
-'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] -- <command> [<options>]
+'perf stat' [-e <EVENT> | --event=EVENT] [-S] [-a] <command>
+'perf stat' [-e <EVENT> | --event=EVENT] [-S] [-a] -- <command> [<options>]
 
 DESCRIPTION
 -----------
@@ -40,7 +40,7 @@ OPTIONS
 -a::
         system-wide collection
 
--l::
+-S::
         scale counter values
 
 EXAMPLES
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 0cbd5d6874ec..9c6d0ae3708e 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -157,10 +157,15 @@ uname_R := $(shell sh -c 'uname -r 2>/dev/null || echo not')
 uname_P := $(shell sh -c 'uname -p 2>/dev/null || echo not')
 uname_V := $(shell sh -c 'uname -v 2>/dev/null || echo not')
 
+# If we're on a 64-bit kernel, use -m64
+ifneq ($(patsubst %64,%,$(uname_M)),$(uname_M))
+  M64 := -m64
+endif
+
 # CFLAGS and LDFLAGS are for the users to override from the command line.
 
-CFLAGS = -ggdb3 -Wall -Werror -Wstrict-prototypes -Wmissing-declarations -Wmissing-prototypes -std=gnu99 -Wdeclaration-after-statement -O6
-LDFLAGS = -lpthread -lrt -lelf
+CFLAGS = $(M64) -ggdb3 -Wall -Wstrict-prototypes -Wmissing-declarations -Wmissing-prototypes -std=gnu99 -Wdeclaration-after-statement -Werror -O6
+LDFLAGS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS)
 ALL_LDFLAGS = $(LDFLAGS)
 STRIP ?= strip
@@ -285,6 +290,7 @@ LIB_FILE=libperf.a
 
 LIB_H += ../../include/linux/perf_counter.h
 LIB_H += perf.h
+LIB_H += util/types.h
 LIB_H += util/list.h
 LIB_H += util/rbtree.h
 LIB_H += util/levenshtein.h
@@ -295,6 +301,7 @@ LIB_H += util/util.h
 LIB_H += util/help.h
 LIB_H += util/strbuf.h
 LIB_H += util/string.h
+LIB_H += util/strlist.h
 LIB_H += util/run-command.h
 LIB_H += util/sigchain.h
 LIB_H += util/symbol.h
@@ -316,12 +323,15 @@ LIB_OBJS += util/run-command.o
 LIB_OBJS += util/quote.o
 LIB_OBJS += util/strbuf.o
 LIB_OBJS += util/string.o
+LIB_OBJS += util/strlist.o
 LIB_OBJS += util/usage.o
 LIB_OBJS += util/wrapper.o
 LIB_OBJS += util/sigchain.o
 LIB_OBJS += util/symbol.o
 LIB_OBJS += util/color.o
 LIB_OBJS += util/pager.o
+LIB_OBJS += util/header.o
+LIB_OBJS += util/callchain.o
 
 BUILTIN_OBJS += builtin-annotate.o
 BUILTIN_OBJS += builtin-help.o
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index b1ed5f766cb3..722c0f54e549 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -25,6 +25,10 @@
 #define SHOW_USER	2
 #define SHOW_HV		4
 
+#define MIN_GREEN		0.5
+#define MIN_RED		5.0
+
+
 static char		const *input_name = "perf.data";
 static char		*vmlinux = "vmlinux";
 
@@ -39,40 +43,42 @@ static int		dump_trace = 0;
 
 static int		verbose;
 
+static int		print_line;
+
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
 struct ip_event {
 	struct perf_event_header header;
-	__u64 ip;
-	__u32 pid, tid;
+	u64 ip;
+	u32 pid, tid;
 };
 
 struct mmap_event {
 	struct perf_event_header header;
-	__u32 pid, tid;
-	__u64 start;
-	__u64 len;
-	__u64 pgoff;
+	u32 pid, tid;
+	u64 start;
+	u64 len;
+	u64 pgoff;
 	char filename[PATH_MAX];
 };
 
 struct comm_event {
 	struct perf_event_header header;
-	__u32 pid, tid;
+	u32 pid, tid;
 	char comm[16];
 };
 
 struct fork_event {
 	struct perf_event_header header;
-	__u32 pid, ppid;
+	u32 pid, ppid;
 };
 
 struct period_event {
 	struct perf_event_header header;
-	__u64 time;
-	__u64 id;
-	__u64 sample_period;
+	u64 time;
+	u64 id;
+	u64 sample_period;
 };
 
 typedef union event_union {
@@ -84,6 +90,13 @@ typedef union event_union {
 	struct period_event		period;
 } event_t;
 
+
+struct sym_ext {
+	struct rb_node	node;
+	double		percent;
+	char		*path;
+};
+
 static LIST_HEAD(dsos);
 static struct dso *kernel_dso;
 static struct dso *vdso;
@@ -145,7 +158,7 @@ static void dsos__fprintf(FILE *fp)
 		dso__fprintf(pos, fp);
 }
 
-static struct symbol *vdso__find_symbol(struct dso *dso, __u64 ip)
+static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip)
 {
 	return dso__find_symbol(kernel_dso, ip);
 }
@@ -178,19 +191,19 @@ static int load_kernel(void)
 
 struct map {
 	struct list_head node;
-	__u64	 start;
-	__u64	 end;
-	__u64	 pgoff;
-	__u64	 (*map_ip)(struct map *, __u64);
+	u64	 start;
+	u64	 end;
+	u64	 pgoff;
+	u64	 (*map_ip)(struct map *, u64);
 	struct dso	 *dso;
 };
 
-static __u64 map__map_ip(struct map *map, __u64 ip)
+static u64 map__map_ip(struct map *map, u64 ip)
 {
 	return ip - map->start + map->pgoff;
 }
 
-static __u64 vdso__map_ip(struct map *map, __u64 ip)
+static u64 vdso__map_ip(struct map *map, u64 ip)
 {
 	return ip;
 }
@@ -373,7 +386,7 @@ static int thread__fork(struct thread *self, struct thread *parent)
 	return 0;
 }
 
-static struct map *thread__find_map(struct thread *self, __u64 ip)
+static struct map *thread__find_map(struct thread *self, u64 ip)
 {
 	struct map *pos;
 
@@ -414,7 +427,7 @@ struct hist_entry {
 	struct map	 *map;
 	struct dso	 *dso;
 	struct symbol	 *sym;
-	__u64	 ip;
+	u64	 ip;
 	char		 level;
 
 	uint32_t	 count;
@@ -519,7 +532,7 @@ sort__dso_print(FILE *fp, struct hist_entry *self)
 	if (self->dso)
 		return fprintf(fp, "%-25s", self->dso->name);
 
-	return fprintf(fp, "%016llx         ", (__u64)self->ip);
+	return fprintf(fp, "%016llx         ", (u64)self->ip);
 }
 
 static struct sort_entry sort_dso = {
@@ -533,7 +546,7 @@ static struct sort_entry sort_dso = {
 static int64_t
 sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-	__u64 ip_l, ip_r;
+	u64 ip_l, ip_r;
 
 	if (left->sym == right->sym)
 		return 0;
@@ -550,13 +563,13 @@ sort__sym_print(FILE *fp, struct hist_entry *self)
 	size_t ret = 0;
 
 	if (verbose)
-		ret += fprintf(fp, "%#018llx  ", (__u64)self->ip);
+		ret += fprintf(fp, "%#018llx  ", (u64)self->ip);
 
 	if (self->sym) {
 		ret += fprintf(fp, "[%c] %s",
 			self->dso == kernel_dso ? 'k' : '.', self->sym->name);
 	} else {
-		ret += fprintf(fp, "%#016llx", (__u64)self->ip);
+		ret += fprintf(fp, "%#016llx", (u64)self->ip);
 	}
 
 	return ret;
@@ -647,7 +660,7 @@ hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
 /*
  * collect histogram counts
  */
-static void hist_hit(struct hist_entry *he, __u64 ip)
+static void hist_hit(struct hist_entry *he, u64 ip)
 {
 	unsigned int sym_size, offset;
 	struct symbol *sym = he->sym;
@@ -676,7 +689,7 @@ static void hist_hit(struct hist_entry *he, __u64 ip)
 
 static int
 hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
-		struct symbol *sym, __u64 ip, char level)
+		struct symbol *sym, u64 ip, char level)
 {
 	struct rb_node **p = &hist.rb_node;
 	struct rb_node *parent = NULL;
@@ -842,13 +855,13 @@ static unsigned long total = 0,
 		     total_unknown = 0;
 
 static int
-process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
 	char level;
 	int show = 0;
 	struct dso *dso = NULL;
 	struct thread *thread = threads__findnew(event->ip.pid);
-	__u64 ip = event->ip.ip;
+	u64 ip = event->ip.ip;
 	struct map *map = NULL;
 
 	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
@@ -1000,10 +1013,10 @@ process_period_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	if (event->header.misc & PERF_EVENT_MISC_OVERFLOW)
-		return process_overflow_event(event, offset, head);
-
 	switch (event->header.type) {
+	case PERF_EVENT_SAMPLE:
+		return process_sample_event(event, offset, head);
+
 	case PERF_EVENT_MMAP:
 		return process_mmap_event(event, offset, head);
 
@@ -1030,13 +1043,33 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
+static char *get_color(double percent)
+{
+	char *color = PERF_COLOR_NORMAL;
+
+	/*
+	 * We color high-overhead entries in red, mid-overhead
+	 * entries in green - and keep the low overhead places
+	 * normal:
+	 */
+	if (percent >= MIN_RED)
+		color = PERF_COLOR_RED;
+	else {
+		if (percent > MIN_GREEN)
+			color = PERF_COLOR_GREEN;
+	}
+	return color;
+}
+
 static int
-parse_line(FILE *file, struct symbol *sym, __u64 start, __u64 len)
+parse_line(FILE *file, struct symbol *sym, u64 start, u64 len)
 {
 	char *line = NULL, *tmp, *tmp2;
+	static const char *prev_line;
+	static const char *prev_color;
 	unsigned int offset;
 	size_t line_len;
-	__u64 line_ip;
+	u64 line_ip;
 	int ret;
 	char *c;
 
@@ -1073,27 +1106,36 @@ parse_line(FILE *file, struct symbol *sym, __u64 start, __u64 len)
 	}
 
 	if (line_ip != -1) {
+		const char *path = NULL;
 		unsigned int hits = 0;
 		double percent = 0.0;
-		char *color = PERF_COLOR_NORMAL;
+		char *color;
+		struct sym_ext *sym_ext = sym->priv;
 
 		offset = line_ip - start;
 		if (offset < len)
 			hits = sym->hist[offset];
 
-		if (sym->hist_sum)
+		if (offset < len && sym_ext) {
+			path = sym_ext[offset].path;
+			percent = sym_ext[offset].percent;
+		} else if (sym->hist_sum)
 			percent = 100.0 * hits / sym->hist_sum;
 
+		color = get_color(percent);
+
 		/*
-		 * We color high-overhead entries in red, mid-overhead
-		 * entries in green - and keep the low overhead places
-		 * normal:
+		 * Also color the filename and line if needed, with
+		 * the same color than the percentage. Don't print it
+		 * twice for close colored ip with the same filename:line
 		 */
-		if (percent >= 5.0)
-			color = PERF_COLOR_RED;
-		else {
-			if (percent > 0.5)
-				color = PERF_COLOR_GREEN;
+		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);
@@ -1109,10 +1151,125 @@ parse_line(FILE *file, struct symbol *sym, __u64 start, __u64 len)
 	return 0;
 }
 
+static struct rb_root root_sym_ext;
+
+static void insert_source_line(struct sym_ext *sym_ext)
+{
+	struct sym_ext *iter;
+	struct rb_node **p = &root_sym_ext.rb_node;
+	struct rb_node *parent = NULL;
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct sym_ext, node);
+
+		if (sym_ext->percent > iter->percent)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&sym_ext->node, parent, p);
+	rb_insert_color(&sym_ext->node, &root_sym_ext);
+}
+
+static void free_source_line(struct symbol *sym, int len)
+{
+	struct sym_ext *sym_ext = sym->priv;
+	int i;
+
+	if (!sym_ext)
+		return;
+
+	for (i = 0; i < len; i++)
+		free(sym_ext[i].path);
+	free(sym_ext);
+
+	sym->priv = NULL;
+	root_sym_ext = RB_ROOT;
+}
+
+/* Get the filename:line for the colored entries */
+static void
+get_source_line(struct symbol *sym, u64 start, int len, char *filename)
+{
+	int i;
+	char cmd[PATH_MAX * 2];
+	struct sym_ext *sym_ext;
+
+	if (!sym->hist_sum)
+		return;
+
+	sym->priv = calloc(len, sizeof(struct sym_ext));
+	if (!sym->priv)
+		return;
+
+	sym_ext = sym->priv;
+
+	for (i = 0; i < len; i++) {
+		char *path = NULL;
+		size_t line_len;
+		u64 offset;
+		FILE *fp;
+
+		sym_ext[i].percent = 100.0 * sym->hist[i] / sym->hist_sum;
+		if (sym_ext[i].percent <= 0.5)
+			continue;
+
+		offset = start + i;
+		sprintf(cmd, "addr2line -e %s %016llx", filename, offset);
+		fp = popen(cmd, "r");
+		if (!fp)
+			continue;
+
+		if (getline(&path, &line_len, fp) < 0 || !line_len)
+			goto next;
+
+		sym_ext[i].path = malloc(sizeof(char) * line_len + 1);
+		if (!sym_ext[i].path)
+			goto next;
+
+		strcpy(sym_ext[i].path, path);
+		insert_source_line(&sym_ext[i]);
+
+	next:
+		pclose(fp);
+	}
+}
+
+static void print_summary(char *filename)
+{
+	struct sym_ext *sym_ext;
+	struct rb_node *node;
+
+	printf("\nSorted summary for file %s\n", filename);
+	printf("----------------------------------------------\n\n");
+
+	if (RB_EMPTY_ROOT(&root_sym_ext)) {
+		printf(" Nothing higher than %1.1f%%\n", MIN_GREEN);
+		return;
+	}
+
+	node = rb_first(&root_sym_ext);
+	while (node) {
+		double percent;
+		char *color;
+		char *path;
+
+		sym_ext = rb_entry(node, struct sym_ext, node);
+		percent = sym_ext->percent;
+		color = get_color(percent);
+		path = sym_ext->path;
+
+		color_fprintf(stdout, color, " %7.2f %s", percent, path);
+		node = rb_next(node);
+	}
+}
+
 static void annotate_sym(struct dso *dso, struct symbol *sym)
 {
 	char *filename = dso->name;
-	__u64 start, end, len;
+	u64 start, end, len;
 	char command[PATH_MAX*2];
 	FILE *file;
 
@@ -1121,13 +1278,6 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 	if (dso == kernel_dso)
 		filename = vmlinux;
 
-	printf("\n------------------------------------------------\n");
-	printf(" Percent |	Source code & Disassembly of %s\n", filename);
-	printf("------------------------------------------------\n");
-
-	if (verbose >= 2)
-		printf("annotating [%p] %30s : [%p] %30s\n", dso, dso->name, sym, sym->name);
-
 	start = sym->obj_start;
 	if (!start)
 		start = sym->start;
@@ -1135,7 +1285,19 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 	end = start + sym->end - sym->start + 1;
 	len = sym->end - sym->start;
 
-	sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s", (__u64)start, (__u64)end, filename);
+	if (print_line) {
+		get_source_line(sym, start, len, filename);
+		print_summary(filename);
+	}
+
+	printf("\n\n------------------------------------------------\n");
+	printf(" Percent |	Source code & Disassembly of %s\n", filename);
+	printf("------------------------------------------------\n");
+
+	if (verbose >= 2)
+		printf("annotating [%p] %30s : [%p] %30s\n", dso, dso->name, sym, sym->name);
+
+	sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s", (u64)start, (u64)end, filename);
 
 	if (verbose >= 3)
 		printf("doing: %s\n", command);
@@ -1150,6 +1312,8 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 	}
 
 	pclose(file);
+	if (print_line)
+		free_source_line(sym, len);
 }
 
 static void find_annotations(void)
@@ -1308,6 +1472,8 @@ static const struct option options[] = {
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
 	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
+	OPT_BOOLEAN('l', "print-line", &print_line,
+		    "print matching source lines (may be slow)"),
 	OPT_END()
 };
 
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 0f5771f615da..d18546f37d7c 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -14,6 +14,8 @@
 #include "util/parse-events.h"
 #include "util/string.h"
 
+#include "util/header.h"
+
 #include <unistd.h>
 #include <sched.h>
 
@@ -37,33 +39,40 @@ static pid_t			target_pid			= -1;
 static int			inherit				= 1;
 static int			force				= 0;
 static int			append_file			= 0;
+static int			call_graph			= 0;
 static int			verbose				= 0;
+static int			inherit_stat			= 0;
+static int			no_samples			= 0;
 
 static long			samples;
 static struct timeval		last_read;
 static struct timeval		this_read;
 
-static __u64			bytes_written;
+static u64			bytes_written;
 
 static struct pollfd		event_array[MAX_NR_CPUS * MAX_COUNTERS];
 
 static int			nr_poll;
 static int			nr_cpu;
 
+static int			file_new = 1;
+
+struct perf_header		*header;
+
 struct mmap_event {
 	struct perf_event_header	header;
-	__u32				pid;
-	__u32				tid;
-	__u64				start;
-	__u64				len;
-	__u64				pgoff;
+	u32				pid;
+	u32				tid;
+	u64				start;
+	u64				len;
+	u64				pgoff;
 	char				filename[PATH_MAX];
 };
 
 struct comm_event {
 	struct perf_event_header	header;
-	__u32				pid;
-	__u32				tid;
+	u32				pid;
+	u32				tid;
 	char				comm[16];
 };
 
@@ -77,10 +86,10 @@ struct mmap_data {
 
 static struct mmap_data		mmap_array[MAX_NR_CPUS][MAX_COUNTERS];
 
-static unsigned int mmap_read_head(struct mmap_data *md)
+static unsigned long mmap_read_head(struct mmap_data *md)
 {
 	struct perf_counter_mmap_page *pc = md->base;
-	int head;
+	long head;
 
 	head = pc->data_head;
 	rmb();
@@ -88,6 +97,32 @@ static unsigned int mmap_read_head(struct mmap_data *md)
 	return head;
 }
 
+static void mmap_write_tail(struct mmap_data *md, unsigned long tail)
+{
+	struct perf_counter_mmap_page *pc = md->base;
+
+	/*
+	 * ensure all reads are done before we write the tail out.
+	 */
+	/* mb(); */
+	pc->data_tail = tail;
+}
+
+static void write_output(void *buf, size_t size)
+{
+	while (size) {
+		int ret = write(output, buf, size);
+
+		if (ret < 0)
+			die("failed to write");
+
+		size -= ret;
+		buf += ret;
+
+		bytes_written += ret;
+	}
+}
+
 static void mmap_read(struct mmap_data *md)
 {
 	unsigned int head = mmap_read_head(md);
@@ -108,7 +143,7 @@ static void mmap_read(struct mmap_data *md)
 	 * In either case, truncate and restart at head.
 	 */
 	diff = head - old;
-	if (diff > md->mask / 2 || diff < 0) {
+	if (diff < 0) {
 		struct timeval iv;
 		unsigned long msecs;
 
@@ -136,36 +171,17 @@ static void mmap_read(struct mmap_data *md)
 		size = md->mask + 1 - (old & md->mask);
 		old += size;
 
-		while (size) {
-			int ret = write(output, buf, size);
-
-			if (ret < 0)
-				die("failed to write");
-
-			size -= ret;
-			buf += ret;
-
-			bytes_written += ret;
-		}
+		write_output(buf, size);
 	}
 
 	buf = &data[old & md->mask];
 	size = head - old;
 	old += size;
 
-	while (size) {
-		int ret = write(output, buf, size);
-
-		if (ret < 0)
-			die("failed to write");
-
-		size -= ret;
-		buf += ret;
-
-		bytes_written += ret;
-	}
+	write_output(buf, size);
 
 	md->prev = old;
+	mmap_write_tail(md, old);
 }
 
 static volatile int done = 0;
@@ -191,7 +207,7 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 	struct comm_event comm_ev;
 	char filename[PATH_MAX];
 	char bf[BUFSIZ];
-	int fd, ret;
+	int fd;
 	size_t size;
 	char *field, *sep;
 	DIR *tasks;
@@ -201,8 +217,12 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 
 	fd = open(filename, O_RDONLY);
 	if (fd < 0) {
-		fprintf(stderr, "couldn't open %s\n", filename);
-		exit(EXIT_FAILURE);
+		/*
+		 * We raced with a task exiting - just return:
+		 */
+		if (verbose)
+			fprintf(stderr, "couldn't open %s\n", filename);
+		return;
 	}
 	if (read(fd, bf, sizeof(bf)) < 0) {
 		fprintf(stderr, "couldn't read %s\n", filename);
@@ -223,17 +243,13 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 
 	comm_ev.pid = pid;
 	comm_ev.header.type = PERF_EVENT_COMM;
-	size = ALIGN(size, sizeof(__u64));
+	size = ALIGN(size, sizeof(u64));
 	comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size);
 
 	if (!full) {
 		comm_ev.tid = pid;
 
-		ret = write(output, &comm_ev, comm_ev.header.size);
-		if (ret < 0) {
-			perror("failed to write");
-			exit(-1);
-		}
+		write_output(&comm_ev, comm_ev.header.size);
 		return;
 	}
 
@@ -248,11 +264,7 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 
 		comm_ev.tid = pid;
 
-		ret = write(output, &comm_ev, comm_ev.header.size);
-		if (ret < 0) {
-			perror("failed to write");
-			exit(-1);
-		}
+		write_output(&comm_ev, comm_ev.header.size);
 	}
 	closedir(tasks);
 	return;
@@ -272,8 +284,12 @@ static void pid_synthesize_mmap_samples(pid_t pid)
 
 	fp = fopen(filename, "r");
 	if (fp == NULL) {
-		fprintf(stderr, "couldn't open %s\n", filename);
-		exit(EXIT_FAILURE);
+		/*
+		 * We raced with a task exiting - just return:
+		 */
+		if (verbose)
+			fprintf(stderr, "couldn't open %s\n", filename);
+		return;
 	}
 	while (1) {
 		char bf[BUFSIZ], *pbf = bf;
@@ -295,33 +311,29 @@ static void pid_synthesize_mmap_samples(pid_t pid)
 			continue;
 		pbf += n + 3;
 		if (*pbf == 'x') { /* vm_exec */
-			char *execname = strrchr(bf, ' ');
+			char *execname = strchr(bf, '/');
 
-			if (execname == NULL || execname[1] != '/')
+			if (execname == NULL)
 				continue;
 
-			execname += 1;
 			size = strlen(execname);
 			execname[size - 1] = '\0'; /* Remove \n */
 			memcpy(mmap_ev.filename, execname, size);
-			size = ALIGN(size, sizeof(__u64));
+			size = ALIGN(size, sizeof(u64));
 			mmap_ev.len -= mmap_ev.start;
 			mmap_ev.header.size = (sizeof(mmap_ev) -
 					       (sizeof(mmap_ev.filename) - size));
 			mmap_ev.pid = pid;
 			mmap_ev.tid = pid;
 
-			if (write(output, &mmap_ev, mmap_ev.header.size) < 0) {
-				perror("failed to write");
-				exit(-1);
-			}
+			write_output(&mmap_ev, mmap_ev.header.size);
 		}
 	}
 
 	fclose(fp);
 }
 
-static void synthesize_samples(void)
+static void synthesize_all(void)
 {
 	DIR *proc;
 	struct dirent dirent, *next;
@@ -345,24 +357,58 @@ static void synthesize_samples(void)
 
 static int group_fd;
 
+static struct perf_header_attr *get_header_attr(struct perf_counter_attr *a, int nr)
+{
+	struct perf_header_attr *h_attr;
+
+	if (nr < header->attrs) {
+		h_attr = header->attr[nr];
+	} else {
+		h_attr = perf_header_attr__new(a);
+		perf_header__add_attr(header, h_attr);
+	}
+
+	return h_attr;
+}
+
 static void create_counter(int counter, int cpu, pid_t pid)
 {
 	struct perf_counter_attr *attr = attrs + counter;
-	int track = 1;
+	struct perf_header_attr *h_attr;
+	int track = !counter; /* only the first counter needs these */
+	struct {
+		u64 count;
+		u64 time_enabled;
+		u64 time_running;
+		u64 id;
+	} read_data;
+
+	attr->read_format	= PERF_FORMAT_TOTAL_TIME_ENABLED |
+				  PERF_FORMAT_TOTAL_TIME_RUNNING |
+				  PERF_FORMAT_ID;
 
 	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+
 	if (freq) {
 		attr->sample_type	|= PERF_SAMPLE_PERIOD;
 		attr->freq		= 1;
 		attr->sample_freq	= freq;
 	}
+
+	if (no_samples)
+		attr->sample_freq = 0;
+
+	if (inherit_stat)
+		attr->inherit_stat = 1;
+
+	if (call_graph)
+		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
+
 	attr->mmap		= track;
 	attr->comm		= track;
 	attr->inherit		= (cpu < 0) && inherit;
 	attr->disabled		= 1;
 
-	track = 0; /* only the first counter needs these */
-
 try_again:
 	fd[nr_cpu][counter] = sys_perf_counter_open(attr, pid, cpu, group_fd, 0);
 
@@ -393,6 +439,22 @@ try_again:
 		exit(-1);
 	}
 
+	h_attr = get_header_attr(attr, counter);
+
+	if (!file_new) {
+		if (memcmp(&h_attr->attr, attr, sizeof(*attr))) {
+			fprintf(stderr, "incompatible append\n");
+			exit(-1);
+		}
+	}
+
+	if (read(fd[nr_cpu][counter], &read_data, sizeof(read_data)) == -1) {
+		perror("Unable to read perf file descriptor\n");
+		exit(-1);
+	}
+
+	perf_header_attr__add_id(h_attr, read_data.id);
+
 	assert(fd[nr_cpu][counter] >= 0);
 	fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK);
 
@@ -410,7 +472,7 @@ try_again:
 	mmap_array[nr_cpu][counter].prev = 0;
 	mmap_array[nr_cpu][counter].mask = mmap_pages*page_size - 1;
 	mmap_array[nr_cpu][counter].base = mmap(NULL, (mmap_pages+1)*page_size,
-			PROT_READ, MAP_SHARED, fd[nr_cpu][counter], 0);
+			PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter], 0);
 	if (mmap_array[nr_cpu][counter].base == MAP_FAILED) {
 		error("failed to mmap with %d (%s)\n", errno, strerror(errno));
 		exit(-1);
@@ -423,11 +485,6 @@ static void open_counters(int cpu, pid_t pid)
 {
 	int counter;
 
-	if (pid > 0) {
-		pid_synthesize_comm_event(pid, 0);
-		pid_synthesize_mmap_samples(pid);
-	}
-
 	group_fd = -1;
 	for (counter = 0; counter < nr_counters; counter++)
 		create_counter(counter, cpu, pid);
@@ -435,11 +492,18 @@ static void open_counters(int cpu, pid_t pid)
 	nr_cpu++;
 }
 
+static void atexit_header(void)
+{
+	header->data_size += bytes_written;
+
+	perf_header__write(header, output);
+}
+
 static int __cmd_record(int argc, const char **argv)
 {
 	int i, counter;
 	struct stat st;
-	pid_t pid;
+	pid_t pid = 0;
 	int flags;
 	int ret;
 
@@ -448,6 +512,10 @@ static int __cmd_record(int argc, const char **argv)
 	assert(nr_cpus <= MAX_NR_CPUS);
 	assert(nr_cpus >= 0);
 
+	atexit(sig_atexit);
+	signal(SIGCHLD, sig_handler);
+	signal(SIGINT, sig_handler);
+
 	if (!stat(output_name, &st) && !force && !append_file) {
 		fprintf(stderr, "Error, output file %s exists, use -A to append or -f to overwrite.\n",
 				output_name);
@@ -456,7 +524,7 @@ static int __cmd_record(int argc, const char **argv)
 
 	flags = O_CREAT|O_RDWR;
 	if (append_file)
-		flags |= O_APPEND;
+		file_new = 0;
 	else
 		flags |= O_TRUNC;
 
@@ -466,14 +534,30 @@ static int __cmd_record(int argc, const char **argv)
 		exit(-1);
 	}
 
+	if (!file_new)
+		header = perf_header__read(output);
+	else
+		header = perf_header__new();
+
+	atexit(atexit_header);
+
 	if (!system_wide) {
-		open_counters(-1, target_pid != -1 ? target_pid : getpid());
+		pid = target_pid;
+		if (pid == -1)
+			pid = getpid();
+
+		open_counters(-1, pid);
 	} else for (i = 0; i < nr_cpus; i++)
 		open_counters(i, target_pid);
 
-	atexit(sig_atexit);
-	signal(SIGCHLD, sig_handler);
-	signal(SIGINT, sig_handler);
+	if (file_new)
+		perf_header__write(header, output);
+
+	if (!system_wide) {
+		pid_synthesize_comm_event(pid, 0);
+		pid_synthesize_mmap_samples(pid);
+	} else
+		synthesize_all();
 
 	if (target_pid == -1 && argc) {
 		pid = fork();
@@ -498,10 +582,7 @@ static int __cmd_record(int argc, const char **argv)
 		}
 	}
 
-	if (system_wide)
-		synthesize_samples();
-
-	while (!done) {
+	for (;;) {
 		int hits = samples;
 
 		for (i = 0; i < nr_cpu; i++) {
@@ -509,8 +590,11 @@ static int __cmd_record(int argc, const char **argv)
 				mmap_read(&mmap_array[i][counter]);
 		}
 
-		if (hits == samples)
+		if (hits == samples) {
+			if (done)
+				break;
 			ret = poll(event_array, nr_poll, 100);
+		}
 	}
 
 	/*
@@ -555,8 +639,14 @@ static const struct option options[] = {
 		    "profile at this frequency"),
 	OPT_INTEGER('m', "mmap-pages", &mmap_pages,
 		    "number of mmap data pages"),
+	OPT_BOOLEAN('g', "call-graph", &call_graph,
+		    "do call-graph (stack chain/backtrace) recording"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
+	OPT_BOOLEAN('s', "stat", &inherit_stat,
+		    "per thread counts"),
+	OPT_BOOLEAN('n', "no-samples", &no_samples,
+		    "don't sample"),
 	OPT_END()
 };
 
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 82fa93b4db99..135b7837e6bf 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -15,8 +15,11 @@
 #include "util/rbtree.h"
 #include "util/symbol.h"
 #include "util/string.h"
+#include "util/callchain.h"
+#include "util/strlist.h"
 
 #include "perf.h"
+#include "util/header.h"
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
@@ -30,51 +33,78 @@ static char		*vmlinux = NULL;
 
 static char		default_sort_order[] = "comm,dso";
 static char		*sort_order = default_sort_order;
+static char		*dso_list_str, *comm_list_str, *sym_list_str;
+static struct strlist	*dso_list, *comm_list, *sym_list;
 
 static int		input;
 static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
 static int		dump_trace = 0;
 #define dprintf(x...)	do { if (dump_trace) printf(x); } while (0)
+#define cdprintf(x...)	do { if (dump_trace) color_fprintf(stdout, color, x); } while (0)
 
 static int		verbose;
+#define eprintf(x...)	do { if (verbose) fprintf(stderr, x); } while (0)
+
 static int		full_paths;
 
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
+static char		default_parent_pattern[] = "^sys_|^do_page_fault";
+static char		*parent_pattern = default_parent_pattern;
+static regex_t		parent_regex;
+
+static int		exclude_other = 1;
+static int		callchain;
+
+static u64		sample_type;
+
 struct ip_event {
 	struct perf_event_header header;
-	__u64 ip;
-	__u32 pid, tid;
-	__u64 period;
+	u64 ip;
+	u32 pid, tid;
+	unsigned char __more_data[];
 };
 
 struct mmap_event {
 	struct perf_event_header header;
-	__u32 pid, tid;
-	__u64 start;
-	__u64 len;
-	__u64 pgoff;
+	u32 pid, tid;
+	u64 start;
+	u64 len;
+	u64 pgoff;
 	char filename[PATH_MAX];
 };
 
 struct comm_event {
 	struct perf_event_header header;
-	__u32 pid, tid;
+	u32 pid, tid;
 	char comm[16];
 };
 
 struct fork_event {
 	struct perf_event_header header;
-	__u32 pid, ppid;
+	u32 pid, ppid;
 };
 
 struct period_event {
 	struct perf_event_header header;
-	__u64 time;
-	__u64 id;
-	__u64 sample_period;
+	u64 time;
+	u64 id;
+	u64 sample_period;
+};
+
+struct lost_event {
+	struct perf_event_header header;
+	u64 id;
+	u64 lost;
+};
+
+struct read_event {
+	struct perf_event_header header;
+	u32 pid,tid;
+	u64 value;
+	u64 format[3];
 };
 
 typedef union event_union {
@@ -84,6 +114,8 @@ typedef union event_union {
 	struct comm_event		comm;
 	struct fork_event		fork;
 	struct period_event		period;
+	struct lost_event		lost;
+	struct read_event		read;
 } event_t;
 
 static LIST_HEAD(dsos);
@@ -119,15 +151,11 @@ static struct dso *dsos__findnew(const char *name)
 
 	nr = dso__load(dso, NULL, verbose);
 	if (nr < 0) {
-		if (verbose)
-			fprintf(stderr, "Failed to open: %s\n", name);
+		eprintf("Failed to open: %s\n", name);
 		goto out_delete_dso;
 	}
-	if (!nr && verbose) {
-		fprintf(stderr,
-		"No symbols found in: %s, maybe install a debug package?\n",
-				name);
-	}
+	if (!nr)
+		eprintf("No symbols found in: %s, maybe install a debug package?\n", name);
 
 	dsos__add(dso);
 
@@ -146,7 +174,7 @@ static void dsos__fprintf(FILE *fp)
 		dso__fprintf(pos, fp);
 }
 
-static struct symbol *vdso__find_symbol(struct dso *dso, __u64 ip)
+static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip)
 {
 	return dso__find_symbol(kernel_dso, ip);
 }
@@ -193,26 +221,26 @@ static int strcommon(const char *pathname)
 
 struct map {
 	struct list_head node;
-	__u64	 start;
-	__u64	 end;
-	__u64	 pgoff;
-	__u64	 (*map_ip)(struct map *, __u64);
+	u64	 start;
+	u64	 end;
+	u64	 pgoff;
+	u64	 (*map_ip)(struct map *, u64);
 	struct dso	 *dso;
 };
 
-static __u64 map__map_ip(struct map *map, __u64 ip)
+static u64 map__map_ip(struct map *map, u64 ip)
 {
 	return ip - map->start + map->pgoff;
 }
 
-static __u64 vdso__map_ip(struct map *map, __u64 ip)
+static u64 vdso__map_ip(struct map *map, u64 ip)
 {
 	return ip;
 }
 
 static inline int is_anon_memory(const char *filename)
 {
-     return strcmp(filename, "//anon") == 0;
+	return strcmp(filename, "//anon") == 0;
 }
 
 static struct map *map__new(struct mmap_event *event)
@@ -383,9 +411,27 @@ static void thread__insert_map(struct thread *self, struct map *map)
 
 	list_for_each_entry_safe(pos, tmp, &self->maps, node) {
 		if (map__overlap(pos, map)) {
-			list_del_init(&pos->node);
-			/* XXX leaks dsos */
-			free(pos);
+			if (verbose >= 2) {
+				printf("overlapping maps:\n");
+				map__fprintf(map, stdout);
+				map__fprintf(pos, stdout);
+			}
+
+			if (map->start <= pos->start && map->end > pos->start)
+				pos->start = map->end;
+
+			if (map->end >= pos->end && map->start < pos->end)
+				pos->end = map->start;
+
+			if (verbose >= 2) {
+				printf("after collision:\n");
+				map__fprintf(pos, stdout);
+			}
+
+			if (pos->start >= pos->end) {
+				list_del_init(&pos->node);
+				free(pos);
+			}
 		}
 	}
 
@@ -412,7 +458,7 @@ static int thread__fork(struct thread *self, struct thread *parent)
 	return 0;
 }
 
-static struct map *thread__find_map(struct thread *self, __u64 ip)
+static struct map *thread__find_map(struct thread *self, u64 ip)
 {
 	struct map *pos;
 
@@ -447,16 +493,19 @@ static size_t threads__fprintf(FILE *fp)
 static struct rb_root hist;
 
 struct hist_entry {
-	struct rb_node	 rb_node;
-
-	struct thread	 *thread;
-	struct map	 *map;
-	struct dso	 *dso;
-	struct symbol	 *sym;
-	__u64		 ip;
-	char		 level;
-
-	__u64		 count;
+	struct rb_node		rb_node;
+
+	struct thread		*thread;
+	struct map		*map;
+	struct dso		*dso;
+	struct symbol		*sym;
+	struct symbol		*parent;
+	u64			ip;
+	char			level;
+	struct callchain_node	callchain;
+	struct rb_root		sorted_chain;
+
+	u64			count;
 };
 
 /*
@@ -473,6 +522,16 @@ struct sort_entry {
 	size_t	(*print)(FILE *fp, struct hist_entry *);
 };
 
+static int64_t cmp_null(void *l, void *r)
+{
+	if (!l && !r)
+		return 0;
+	else if (!l)
+		return -1;
+	else
+		return 1;
+}
+
 /* --sort pid */
 
 static int64_t
@@ -507,14 +566,8 @@ sort__comm_collapse(struct hist_entry *left, struct hist_entry *right)
 	char *comm_l = left->thread->comm;
 	char *comm_r = right->thread->comm;
 
-	if (!comm_l || !comm_r) {
-		if (!comm_l && !comm_r)
-			return 0;
-		else if (!comm_l)
-			return -1;
-		else
-			return 1;
-	}
+	if (!comm_l || !comm_r)
+		return cmp_null(comm_l, comm_r);
 
 	return strcmp(comm_l, comm_r);
 }
@@ -540,14 +593,8 @@ sort__dso_cmp(struct hist_entry *left, struct hist_entry *right)
 	struct dso *dso_l = left->dso;
 	struct dso *dso_r = right->dso;
 
-	if (!dso_l || !dso_r) {
-		if (!dso_l && !dso_r)
-			return 0;
-		else if (!dso_l)
-			return -1;
-		else
-			return 1;
-	}
+	if (!dso_l || !dso_r)
+		return cmp_null(dso_l, dso_r);
 
 	return strcmp(dso_l->name, dso_r->name);
 }
@@ -558,7 +605,7 @@ sort__dso_print(FILE *fp, struct hist_entry *self)
 	if (self->dso)
 		return fprintf(fp, "%-25s", self->dso->name);
 
-	return fprintf(fp, "%016llx         ", (__u64)self->ip);
+	return fprintf(fp, "%016llx         ", (u64)self->ip);
 }
 
 static struct sort_entry sort_dso = {
@@ -572,7 +619,7 @@ static struct sort_entry sort_dso = {
 static int64_t
 sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
 {
-	__u64 ip_l, ip_r;
+	u64 ip_l, ip_r;
 
 	if (left->sym == right->sym)
 		return 0;
@@ -589,13 +636,13 @@ sort__sym_print(FILE *fp, struct hist_entry *self)
 	size_t ret = 0;
 
 	if (verbose)
-		ret += fprintf(fp, "%#018llx  ", (__u64)self->ip);
+		ret += fprintf(fp, "%#018llx  ", (u64)self->ip);
 
 	if (self->sym) {
 		ret += fprintf(fp, "[%c] %s",
 			self->dso == kernel_dso ? 'k' : '.', self->sym->name);
 	} else {
-		ret += fprintf(fp, "%#016llx", (__u64)self->ip);
+		ret += fprintf(fp, "%#016llx", (u64)self->ip);
 	}
 
 	return ret;
@@ -607,7 +654,38 @@ static struct sort_entry sort_sym = {
 	.print	= sort__sym_print,
 };
 
+/* --sort parent */
+
+static int64_t
+sort__parent_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct symbol *sym_l = left->parent;
+	struct symbol *sym_r = right->parent;
+
+	if (!sym_l || !sym_r)
+		return cmp_null(sym_l, sym_r);
+
+	return strcmp(sym_l->name, sym_r->name);
+}
+
+static size_t
+sort__parent_print(FILE *fp, struct hist_entry *self)
+{
+	size_t ret = 0;
+
+	ret += fprintf(fp, "%-20s", self->parent ? self->parent->name : "[other]");
+
+	return ret;
+}
+
+static struct sort_entry sort_parent = {
+	.header = "Parent symbol       ",
+	.cmp	= sort__parent_cmp,
+	.print	= sort__parent_print,
+};
+
 static int sort__need_collapse = 0;
+static int sort__has_parent = 0;
 
 struct sort_dimension {
 	char			*name;
@@ -620,6 +698,7 @@ static struct sort_dimension sort_dimensions[] = {
 	{ .name = "comm",	.entry = &sort_comm,	},
 	{ .name = "dso",	.entry = &sort_dso,	},
 	{ .name = "symbol",	.entry = &sort_sym,	},
+	{ .name = "parent",	.entry = &sort_parent,	},
 };
 
 static LIST_HEAD(hist_entry__sort_list);
@@ -640,6 +719,19 @@ static int sort_dimension__add(char *tok)
 		if (sd->entry->collapse)
 			sort__need_collapse = 1;
 
+		if (sd->entry == &sort_parent) {
+			int ret = regcomp(&parent_regex, parent_pattern, REG_EXTENDED);
+			if (ret) {
+				char err[BUFSIZ];
+
+				regerror(ret, &parent_regex, err, sizeof(err));
+				fprintf(stderr, "Invalid regex: %s\n%s",
+					parent_pattern, err);
+				exit(-1);
+			}
+			sort__has_parent = 1;
+		}
+
 		list_add_tail(&sd->entry->list, &hist_entry__sort_list);
 		sd->taken = 1;
 
@@ -684,11 +776,56 @@ hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
 }
 
 static size_t
-hist_entry__fprintf(FILE *fp, struct hist_entry *self, __u64 total_samples)
+callchain__fprintf(FILE *fp, struct callchain_node *self, u64 total_samples)
+{
+	struct callchain_list *chain;
+	size_t ret = 0;
+
+	if (!self)
+		return 0;
+
+	ret += callchain__fprintf(fp, self->parent, total_samples);
+
+
+	list_for_each_entry(chain, &self->val, list)
+		ret += fprintf(fp, "                %p\n", (void *)chain->ip);
+
+	return ret;
+}
+
+static size_t
+hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
+			      u64 total_samples)
+{
+	struct rb_node *rb_node;
+	struct callchain_node *chain;
+	size_t ret = 0;
+
+	rb_node = rb_first(&self->sorted_chain);
+	while (rb_node) {
+		double percent;
+
+		chain = rb_entry(rb_node, struct callchain_node, rb_node);
+		percent = chain->hit * 100.0 / total_samples;
+		ret += fprintf(fp, "           %6.2f%%\n", percent);
+		ret += callchain__fprintf(fp, chain, total_samples);
+		ret += fprintf(fp, "\n");
+		rb_node = rb_next(rb_node);
+	}
+
+	return ret;
+}
+
+
+static size_t
+hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
 {
 	struct sort_entry *se;
 	size_t ret;
 
+	if (exclude_other && !self->parent)
+		return 0;
+
 	if (total_samples) {
 		double percent = self->count * 100.0 / total_samples;
 		char *color = PERF_COLOR_NORMAL;
@@ -711,22 +848,89 @@ hist_entry__fprintf(FILE *fp, struct hist_entry *self, __u64 total_samples)
 		ret = fprintf(fp, "%12Ld ", self->count);
 
 	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		if (exclude_other && (se == &sort_parent))
+			continue;
+
 		fprintf(fp, "  ");
 		ret += se->print(fp, self);
 	}
 
 	ret += fprintf(fp, "\n");
 
+	if (callchain)
+		hist_entry_callchain__fprintf(fp, self, total_samples);
+
 	return ret;
 }
 
 /*
+ *
+ */
+
+static struct symbol *
+resolve_symbol(struct thread *thread, struct map **mapp,
+	       struct dso **dsop, u64 *ipp)
+{
+	struct dso *dso = dsop ? *dsop : NULL;
+	struct map *map = mapp ? *mapp : NULL;
+	u64 ip = *ipp;
+
+	if (!thread)
+		return NULL;
+
+	if (dso)
+		goto got_dso;
+
+	if (map)
+		goto got_map;
+
+	map = thread__find_map(thread, ip);
+	if (map != NULL) {
+		if (mapp)
+			*mapp = map;
+got_map:
+		ip = map->map_ip(map, ip);
+
+		dso = map->dso;
+	} else {
+		/*
+		 * If this is outside of all known maps,
+		 * and is a negative address, try to look it
+		 * up in the kernel dso, as it might be a
+		 * vsyscall (which executes in user-mode):
+		 */
+		if ((long long)ip < 0)
+		dso = kernel_dso;
+	}
+	dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
+	dprintf(" ...... map: %Lx -> %Lx\n", *ipp, ip);
+	*ipp  = ip;
+
+	if (dsop)
+		*dsop = dso;
+
+	if (!dso)
+		return NULL;
+got_dso:
+	return dso->find_symbol(dso, ip);
+}
+
+static int call__match(struct symbol *sym)
+{
+	if (sym->name && !regexec(&parent_regex, sym->name, 0, NULL, 0))
+		return 1;
+
+	return 0;
+}
+
+/*
  * collect histogram counts
  */
 
 static int
 hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
-		struct symbol *sym, __u64 ip, char level, __u64 count)
+		struct symbol *sym, u64 ip, struct ip_callchain *chain,
+		char level, u64 count)
 {
 	struct rb_node **p = &hist.rb_node;
 	struct rb_node *parent = NULL;
@@ -739,9 +943,42 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
 		.ip	= ip,
 		.level	= level,
 		.count	= count,
+		.parent = NULL,
+		.sorted_chain = RB_ROOT
 	};
 	int cmp;
 
+	if (sort__has_parent && chain) {
+		u64 context = PERF_CONTEXT_MAX;
+		int i;
+
+		for (i = 0; i < chain->nr; i++) {
+			u64 ip = chain->ips[i];
+			struct dso *dso = NULL;
+			struct symbol *sym;
+
+			if (ip >= PERF_CONTEXT_MAX) {
+				context = ip;
+				continue;
+			}
+
+			switch (context) {
+			case PERF_CONTEXT_KERNEL:
+				dso = kernel_dso;
+				break;
+			default:
+				break;
+			}
+
+			sym = resolve_symbol(thread, NULL, &dso, &ip);
+
+			if (sym && call__match(sym)) {
+				entry.parent = sym;
+				break;
+			}
+		}
+	}
+
 	while (*p != NULL) {
 		parent = *p;
 		he = rb_entry(parent, struct hist_entry, rb_node);
@@ -750,6 +987,8 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
 
 		if (!cmp) {
 			he->count += count;
+			if (callchain)
+				append_chain(&he->callchain, chain);
 			return 0;
 		}
 
@@ -763,6 +1002,10 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
 	if (!he)
 		return -ENOMEM;
 	*he = entry;
+	if (callchain) {
+		callchain_init(&he->callchain);
+		append_chain(&he->callchain, chain);
+	}
 	rb_link_node(&he->rb_node, parent, p);
 	rb_insert_color(&he->rb_node, &hist);
 
@@ -839,6 +1082,9 @@ static void output__insert_entry(struct hist_entry *he)
 	struct rb_node *parent = NULL;
 	struct hist_entry *iter;
 
+	if (callchain)
+		sort_chain_to_rbtree(&he->sorted_chain, &he->callchain);
+
 	while (*p != NULL) {
 		parent = *p;
 		iter = rb_entry(parent, struct hist_entry, rb_node);
@@ -873,7 +1119,7 @@ static void output__resort(void)
 	}
 }
 
-static size_t output__fprintf(FILE *fp, __u64 total_samples)
+static size_t output__fprintf(FILE *fp, u64 total_samples)
 {
 	struct hist_entry *pos;
 	struct sort_entry *se;
@@ -882,18 +1128,24 @@ static size_t output__fprintf(FILE *fp, __u64 total_samples)
 
 	fprintf(fp, "\n");
 	fprintf(fp, "#\n");
-	fprintf(fp, "# (%Ld samples)\n", (__u64)total_samples);
+	fprintf(fp, "# (%Ld samples)\n", (u64)total_samples);
 	fprintf(fp, "#\n");
 
 	fprintf(fp, "# Overhead");
-	list_for_each_entry(se, &hist_entry__sort_list, list)
+	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		if (exclude_other && (se == &sort_parent))
+			continue;
 		fprintf(fp, "  %s", se->header);
+	}
 	fprintf(fp, "\n");
 
 	fprintf(fp, "# ........");
 	list_for_each_entry(se, &hist_entry__sort_list, list) {
 		int i;
 
+		if (exclude_other && (se == &sort_parent))
+			continue;
+
 		fprintf(fp, "  ");
 		for (i = 0; i < strlen(se->header); i++)
 			fprintf(fp, ".");
@@ -907,7 +1159,8 @@ static size_t output__fprintf(FILE *fp, __u64 total_samples)
 		ret += hist_entry__fprintf(fp, pos, total_samples);
 	}
 
-	if (!strcmp(sort_order, default_sort_order)) {
+	if (sort_order == default_sort_order &&
+			parent_pattern == default_parent_pattern) {
 		fprintf(fp, "#\n");
 		fprintf(fp, "# (For more details, try: perf report --sort comm,dso,symbol)\n");
 		fprintf(fp, "#\n");
@@ -932,23 +1185,41 @@ static unsigned long total = 0,
 		     total_mmap = 0,
 		     total_comm = 0,
 		     total_fork = 0,
-		     total_unknown = 0;
+		     total_unknown = 0,
+		     total_lost = 0;
+
+static int validate_chain(struct ip_callchain *chain, event_t *event)
+{
+	unsigned int chain_size;
+
+	chain_size = event->header.size;
+	chain_size -= (unsigned long)&event->ip.__more_data - (unsigned long)event;
+
+	if (chain->nr*sizeof(u64) > chain_size)
+		return -1;
+
+	return 0;
+}
 
 static int
-process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
 	char level;
 	int show = 0;
 	struct dso *dso = NULL;
 	struct thread *thread = threads__findnew(event->ip.pid);
-	__u64 ip = event->ip.ip;
-	__u64 period = 1;
+	u64 ip = event->ip.ip;
+	u64 period = 1;
 	struct map *map = NULL;
+	void *more_data = event->ip.__more_data;
+	struct ip_callchain *chain = NULL;
 
-	if (event->header.type & PERF_SAMPLE_PERIOD)
-		period = event->ip.period;
+	if (sample_type & PERF_SAMPLE_PERIOD) {
+		period = *(u64 *)more_data;
+		more_data += sizeof(u64);
+	}
 
-	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p period: %Ld\n",
+	dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d: %p period: %Ld\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->header.misc,
@@ -956,14 +1227,35 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
 		(void *)(long)ip,
 		(long long)period);
 
+	if (sample_type & PERF_SAMPLE_CALLCHAIN) {
+		int i;
+
+		chain = (void *)more_data;
+
+		dprintf("... chain: nr:%Lu\n", chain->nr);
+
+		if (validate_chain(chain, event) < 0) {
+			eprintf("call-chain problem with event, skipping it.\n");
+			return 0;
+		}
+
+		if (dump_trace) {
+			for (i = 0; i < chain->nr; i++)
+				dprintf("..... %2d: %016Lx\n", i, chain->ips[i]);
+		}
+	}
+
 	dprintf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 
 	if (thread == NULL) {
-		fprintf(stderr, "problem processing %d event, skipping it.\n",
+		eprintf("problem processing %d event, skipping it.\n",
 			event->header.type);
 		return -1;
 	}
 
+	if (comm_list && !strlist__has_entry(comm_list, thread->comm))
+		return 0;
+
 	if (event->header.misc & PERF_EVENT_MISC_KERNEL) {
 		show = SHOW_KERNEL;
 		level = 'k';
@@ -977,22 +1269,6 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
 		show = SHOW_USER;
 		level = '.';
 
-		map = thread__find_map(thread, ip);
-		if (map != NULL) {
-			ip = map->map_ip(map, ip);
-			dso = map->dso;
-		} else {
-			/*
-			 * If this is outside of all known maps,
-			 * and is a negative address, try to look it
-			 * up in the kernel dso, as it might be a
-			 * vsyscall (which executes in user-mode):
-			 */
-			if ((long long)ip < 0)
-				dso = kernel_dso;
-		}
-		dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
-
 	} else {
 		show = SHOW_HV;
 		level = 'H';
@@ -1000,14 +1276,16 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
 	}
 
 	if (show & show_mask) {
-		struct symbol *sym = NULL;
+		struct symbol *sym = resolve_symbol(thread, &map, &dso, &ip);
 
-		if (dso)
-			sym = dso->find_symbol(dso, ip);
+		if (dso_list && dso && dso->name && !strlist__has_entry(dso_list, dso->name))
+			return 0;
+
+		if (sym_list && sym && !strlist__has_entry(sym_list, sym->name))
+			return 0;
 
-		if (hist_entry__add(thread, map, dso, sym, ip, level, period)) {
-			fprintf(stderr,
-		"problem incrementing symbol count, skipping event\n");
+		if (hist_entry__add(thread, map, dso, sym, ip, chain, level, period)) {
+			eprintf("problem incrementing symbol count, skipping event\n");
 			return -1;
 		}
 	}
@@ -1096,12 +1374,77 @@ process_period_event(event_t *event, unsigned long offset, unsigned long head)
 }
 
 static int
+process_lost_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	dprintf("%p [%p]: PERF_EVENT_LOST: id:%Ld: lost:%Ld\n",
+		(void *)(offset + head),
+		(void *)(long)(event->header.size),
+		event->lost.id,
+		event->lost.lost);
+
+	total_lost += event->lost.lost;
+
+	return 0;
+}
+
+static void trace_event(event_t *event)
+{
+	unsigned char *raw_event = (void *)event;
+	char *color = PERF_COLOR_BLUE;
+	int i, j;
+
+	if (!dump_trace)
+		return;
+
+	dprintf(".");
+	cdprintf("\n. ... raw event: size %d bytes\n", event->header.size);
+
+	for (i = 0; i < event->header.size; i++) {
+		if ((i & 15) == 0) {
+			dprintf(".");
+			cdprintf("  %04x: ", i);
+		}
+
+		cdprintf(" %02x", raw_event[i]);
+
+		if (((i & 15) == 15) || i == event->header.size-1) {
+			cdprintf("  ");
+			for (j = 0; j < 15-(i & 15); j++)
+				cdprintf("   ");
+			for (j = 0; j < (i & 15); j++) {
+				if (isprint(raw_event[i-15+j]))
+					cdprintf("%c", raw_event[i-15+j]);
+				else
+					cdprintf(".");
+			}
+			cdprintf("\n");
+		}
+	}
+	dprintf(".\n");
+}
+
+static int
+process_read_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	dprintf("%p [%p]: PERF_EVENT_READ: %d %d %Lu\n",
+			(void *)(offset + head),
+			(void *)(long)(event->header.size),
+			event->read.pid,
+			event->read.tid,
+			event->read.value);
+
+	return 0;
+}
+
+static int
 process_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	if (event->header.misc & PERF_EVENT_MISC_OVERFLOW)
-		return process_overflow_event(event, offset, head);
+	trace_event(event);
 
 	switch (event->header.type) {
+	case PERF_EVENT_SAMPLE:
+		return process_sample_event(event, offset, head);
+
 	case PERF_EVENT_MMAP:
 		return process_mmap_event(event, offset, head);
 
@@ -1113,6 +1456,13 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 
 	case PERF_EVENT_PERIOD:
 		return process_period_event(event, offset, head);
+
+	case PERF_EVENT_LOST:
+		return process_lost_event(event, offset, head);
+
+	case PERF_EVENT_READ:
+		return process_read_event(event, offset, head);
+
 	/*
 	 * We dont process them right now but they are fine:
 	 */
@@ -1128,11 +1478,30 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
+static struct perf_header	*header;
+
+static u64 perf_header__sample_type(void)
+{
+	u64 sample_type = 0;
+	int i;
+
+	for (i = 0; i < header->attrs; i++) {
+		struct perf_header_attr *attr = header->attr[i];
+
+		if (!sample_type)
+			sample_type = attr->attr.sample_type;
+		else if (sample_type != attr->attr.sample_type)
+			die("non matching sample_type");
+	}
+
+	return sample_type;
+}
+
 static int __cmd_report(void)
 {
 	int ret, rc = EXIT_FAILURE;
 	unsigned long offset = 0;
-	unsigned long head = 0;
+	unsigned long head, shift;
 	struct stat stat;
 	event_t *event;
 	uint32_t size;
@@ -1160,6 +1529,16 @@ static int __cmd_report(void)
 		exit(0);
 	}
 
+	header = perf_header__read(input);
+	head = header->data_offset;
+
+	sample_type = perf_header__sample_type();
+
+	if (sort__has_parent && !(sample_type & PERF_SAMPLE_CALLCHAIN)) {
+		fprintf(stderr, "selected --sort parent, but no callchain data\n");
+		exit(-1);
+	}
+
 	if (load_kernel() < 0) {
 		perror("failed to load kernel symbols");
 		return EXIT_FAILURE;
@@ -1175,6 +1554,11 @@ static int __cmd_report(void)
 		cwd = NULL;
 		cwdlen = 0;
 	}
+
+	shift = page_size * (head / page_size);
+	offset += shift;
+	head -= shift;
+
 remap:
 	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
 			   MAP_SHARED, input, offset);
@@ -1191,9 +1575,10 @@ more:
 		size = 8;
 
 	if (head + event->header.size >= page_size * mmap_window) {
-		unsigned long shift = page_size * (head / page_size);
 		int ret;
 
+		shift = page_size * (head / page_size);
+
 		ret = munmap(buf, page_size * mmap_window);
 		assert(ret == 0);
 
@@ -1204,7 +1589,7 @@ more:
 
 	size = event->header.size;
 
-	dprintf("%p [%p]: event: %d\n",
+	dprintf("\n%p [%p]: event: %d\n",
 			(void *)(offset + head),
 			(void *)(long)event->header.size,
 			event->header.type);
@@ -1231,9 +1616,13 @@ more:
 
 	head += size;
 
+	if (offset + head >= header->data_offset + header->data_size)
+		goto done;
+
 	if (offset + head < stat.st_size)
 		goto more;
 
+done:
 	rc = EXIT_SUCCESS;
 	close(input);
 
@@ -1241,6 +1630,7 @@ more:
 	dprintf("    mmap events: %10ld\n", total_mmap);
 	dprintf("    comm events: %10ld\n", total_comm);
 	dprintf("    fork events: %10ld\n", total_fork);
+	dprintf("    lost events: %10ld\n", total_lost);
 	dprintf(" unknown events: %10ld\n", total_unknown);
 
 	if (dump_trace)
@@ -1273,9 +1663,20 @@ static const struct option options[] = {
 		    "dump raw trace in ASCII"),
 	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
 	OPT_STRING('s', "sort", &sort_order, "key[,key2...]",
-		   "sort by key(s): pid, comm, dso, symbol. Default: pid,symbol"),
+		   "sort by key(s): pid, comm, dso, symbol, parent"),
 	OPT_BOOLEAN('P', "full-paths", &full_paths,
 		    "Don't shorten the pathnames taking into account the cwd"),
+	OPT_STRING('p', "parent", &parent_pattern, "regex",
+		   "regex filter to identify parent, see: '--sort parent'"),
+	OPT_BOOLEAN('x', "exclude-other", &exclude_other,
+		    "Only display entries with parent-match"),
+	OPT_BOOLEAN('c', "callchain", &callchain, "Display callchains"),
+	OPT_STRING('d', "dsos", &dso_list_str, "dso[,dso...]",
+		   "only consider symbols in these dsos"),
+	OPT_STRING('C', "comms", &comm_list_str, "comm[,comm...]",
+		   "only consider symbols in these comms"),
+	OPT_STRING('S', "symbols", &sym_list_str, "symbol[,symbol...]",
+		   "only consider these symbols"),
 	OPT_END()
 };
 
@@ -1294,6 +1695,19 @@ static void setup_sorting(void)
 	free(str);
 }
 
+static void setup_list(struct strlist **list, const char *list_str,
+		       const char *list_name)
+{
+	if (list_str) {
+		*list = strlist__new(true, list_str);
+		if (!*list) {
+			fprintf(stderr, "problems parsing %s list\n",
+				list_name);
+			exit(129);
+		}
+	}
+}
+
 int cmd_report(int argc, const char **argv, const char *prefix)
 {
 	symbol__init();
@@ -1304,12 +1718,21 @@ int cmd_report(int argc, const char **argv, const char *prefix)
 
 	setup_sorting();
 
+	if (parent_pattern != default_parent_pattern)
+		sort_dimension__add("parent");
+	else
+		exclude_other = 0;
+
 	/*
 	 * Any (unrecognized) arguments left?
 	 */
 	if (argc)
 		usage_with_options(report_usage, options);
 
+	setup_list(&dso_list, dso_list_str, "dso");
+	setup_list(&comm_list, comm_list_str, "comm");
+	setup_list(&sym_list, sym_list_str, "symbol");
+
 	setup_pager();
 
 	return __cmd_report();
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index c43e4a97dc42..2e03524a1de0 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -32,6 +32,7 @@
  *   Wu Fengguang <fengguang.wu@intel.com>
  *   Mike Galbraith <efault@gmx.de>
  *   Paul Mackerras <paulus@samba.org>
+ *   Jaswinder Singh Rajput <jaswinder@kernel.org>
  *
  * Released under the GPL v2. (and only v2, not any later version)
  */
@@ -43,8 +44,9 @@
 #include "util/parse-events.h"
 
 #include <sys/prctl.h>
+#include <math.h>
 
-static struct perf_counter_attr default_attrs[MAX_COUNTERS] = {
+static struct perf_counter_attr default_attrs[] = {
 
   { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK	},
   { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES},
@@ -58,35 +60,46 @@ static struct perf_counter_attr default_attrs[MAX_COUNTERS] = {
 
 };
 
+#define MAX_RUN			100
+
 static int			system_wide			=  0;
-static int			inherit				=  1;
 static int			verbose				=  0;
+static int			nr_cpus				=  0;
+static int			run_idx				=  0;
+
+static int			run_count			=  1;
+static int			inherit				=  1;
+static int			scale				=  1;
+static int			target_pid			= -1;
+static int			null_run			=  0;
 
 static int			fd[MAX_NR_CPUS][MAX_COUNTERS];
 
-static int			target_pid			= -1;
-static int			nr_cpus				=  0;
-static unsigned int		page_size;
+static u64			runtime_nsecs[MAX_RUN];
+static u64			walltime_nsecs[MAX_RUN];
+static u64			runtime_cycles[MAX_RUN];
 
-static int			scale				=  1;
+static u64			event_res[MAX_RUN][MAX_COUNTERS][3];
+static u64			event_scaled[MAX_RUN][MAX_COUNTERS];
 
-static const unsigned int default_count[] = {
-	1000000,
-	1000000,
-	  10000,
-	  10000,
-	1000000,
-	  10000,
-};
+static u64			event_res_avg[MAX_COUNTERS][3];
+static u64			event_res_noise[MAX_COUNTERS][3];
+
+static u64			event_scaled_avg[MAX_COUNTERS];
 
-static __u64			event_res[MAX_COUNTERS][3];
-static __u64			event_scaled[MAX_COUNTERS];
+static u64			runtime_nsecs_avg;
+static u64			runtime_nsecs_noise;
 
-static __u64			runtime_nsecs;
-static __u64			walltime_nsecs;
-static __u64			runtime_cycles;
+static u64			walltime_nsecs_avg;
+static u64			walltime_nsecs_noise;
 
-static void create_perf_stat_counter(int counter)
+static u64			runtime_cycles_avg;
+static u64			runtime_cycles_noise;
+
+#define ERR_PERF_OPEN \
+"Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n"
+
+static void create_perf_stat_counter(int counter, int pid)
 {
 	struct perf_counter_attr *attr = attrs + counter;
 
@@ -96,20 +109,21 @@ static void create_perf_stat_counter(int counter)
 
 	if (system_wide) {
 		int cpu;
-		for (cpu = 0; cpu < nr_cpus; cpu ++) {
+		for (cpu = 0; cpu < nr_cpus; cpu++) {
 			fd[cpu][counter] = sys_perf_counter_open(attr, -1, cpu, -1, 0);
-			if (fd[cpu][counter] < 0 && verbose) {
-				printf("Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n", counter, fd[cpu][counter], strerror(errno));
-			}
+			if (fd[cpu][counter] < 0 && verbose)
+				fprintf(stderr, ERR_PERF_OPEN, counter,
+					fd[cpu][counter], strerror(errno));
 		}
 	} else {
-		attr->inherit	= inherit;
-		attr->disabled	= 1;
-
-		fd[0][counter] = sys_perf_counter_open(attr, 0, -1, -1, 0);
-		if (fd[0][counter] < 0 && verbose) {
-			printf("Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n", counter, fd[0][counter], strerror(errno));
-		}
+		attr->inherit	     = inherit;
+		attr->disabled	     = 1;
+		attr->enable_on_exec = 1;
+
+		fd[0][counter] = sys_perf_counter_open(attr, pid, -1, -1, 0);
+		if (fd[0][counter] < 0 && verbose)
+			fprintf(stderr, ERR_PERF_OPEN, counter,
+				fd[0][counter], strerror(errno));
 	}
 }
 
@@ -135,22 +149,24 @@ static inline int nsec_counter(int counter)
  */
 static void read_counter(int counter)
 {
-	__u64 *count, single_count[3];
+	u64 *count, single_count[3];
 	ssize_t res;
 	int cpu, nv;
 	int scaled;
 
-	count = event_res[counter];
+	count = event_res[run_idx][counter];
 
 	count[0] = count[1] = count[2] = 0;
 
 	nv = scale ? 3 : 1;
-	for (cpu = 0; cpu < nr_cpus; cpu ++) {
+	for (cpu = 0; cpu < nr_cpus; cpu++) {
 		if (fd[cpu][counter] < 0)
 			continue;
 
-		res = read(fd[cpu][counter], single_count, nv * sizeof(__u64));
-		assert(res == nv * sizeof(__u64));
+		res = read(fd[cpu][counter], single_count, nv * sizeof(u64));
+		assert(res == nv * sizeof(u64));
+		close(fd[cpu][counter]);
+		fd[cpu][counter] = -1;
 
 		count[0] += single_count[0];
 		if (scale) {
@@ -162,13 +178,13 @@ static void read_counter(int counter)
 	scaled = 0;
 	if (scale) {
 		if (count[2] == 0) {
-			event_scaled[counter] = -1;
+			event_scaled[run_idx][counter] = -1;
 			count[0] = 0;
 			return;
 		}
 
 		if (count[2] < count[1]) {
-			event_scaled[counter] = 1;
+			event_scaled[run_idx][counter] = 1;
 			count[0] = (unsigned long long)
 				((double)count[0] * count[1] / count[2] + 0.5);
 		}
@@ -178,10 +194,129 @@ static void read_counter(int counter)
 	 */
 	if (attrs[counter].type == PERF_TYPE_SOFTWARE &&
 		attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK)
-		runtime_nsecs = count[0];
+		runtime_nsecs[run_idx] = count[0];
 	if (attrs[counter].type == PERF_TYPE_HARDWARE &&
 		attrs[counter].config == PERF_COUNT_HW_CPU_CYCLES)
-		runtime_cycles = count[0];
+		runtime_cycles[run_idx] = count[0];
+}
+
+static int run_perf_stat(int argc, const char **argv)
+{
+	unsigned long long t0, t1;
+	int status = 0;
+	int counter;
+	int pid;
+	int child_ready_pipe[2], go_pipe[2];
+	char buf;
+
+	if (!system_wide)
+		nr_cpus = 1;
+
+	if (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0) {
+		perror("failed to create pipes");
+		exit(1);
+	}
+
+	if ((pid = fork()) < 0)
+		perror("failed to fork");
+
+	if (!pid) {
+		close(child_ready_pipe[0]);
+		close(go_pipe[1]);
+		fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC);
+
+		/*
+		 * Do a dummy execvp to get the PLT entry resolved,
+		 * so we avoid the resolver overhead on the real
+		 * execvp call.
+		 */
+		execvp("", (char **)argv);
+
+		/*
+		 * Tell the parent we're ready to go
+		 */
+		close(child_ready_pipe[1]);
+
+		/*
+		 * Wait until the parent tells us to go.
+		 */
+		read(go_pipe[0], &buf, 1);
+
+		execvp(argv[0], (char **)argv);
+
+		perror(argv[0]);
+		exit(-1);
+	}
+
+	/*
+	 * Wait for the child to be ready to exec.
+	 */
+	close(child_ready_pipe[1]);
+	close(go_pipe[0]);
+	read(child_ready_pipe[0], &buf, 1);
+	close(child_ready_pipe[0]);
+
+	for (counter = 0; counter < nr_counters; counter++)
+		create_perf_stat_counter(counter, pid);
+
+	/*
+	 * Enable counters and exec the command:
+	 */
+	t0 = rdclock();
+
+	close(go_pipe[1]);
+	wait(&status);
+
+	t1 = rdclock();
+
+	walltime_nsecs[run_idx] = t1 - t0;
+
+	for (counter = 0; counter < nr_counters; counter++)
+		read_counter(counter);
+
+	return WEXITSTATUS(status);
+}
+
+static void print_noise(u64 *count, u64 *noise)
+{
+	if (run_count > 1)
+		fprintf(stderr, "   ( +- %7.3f%% )",
+			(double)noise[0]/(count[0]+1)*100.0);
+}
+
+static void nsec_printout(int counter, u64 *count, u64 *noise)
+{
+	double msecs = (double)count[0] / 1000000;
+
+	fprintf(stderr, " %14.6f  %-24s", msecs, event_name(counter));
+
+	if (attrs[counter].type == PERF_TYPE_SOFTWARE &&
+		attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) {
+
+		if (walltime_nsecs_avg)
+			fprintf(stderr, " # %10.3f CPUs ",
+				(double)count[0] / (double)walltime_nsecs_avg);
+	}
+	print_noise(count, noise);
+}
+
+static void abs_printout(int counter, u64 *count, u64 *noise)
+{
+	fprintf(stderr, " %14Ld  %-24s", count[0], event_name(counter));
+
+	if (runtime_cycles_avg &&
+		attrs[counter].type == PERF_TYPE_HARDWARE &&
+			attrs[counter].config == PERF_COUNT_HW_INSTRUCTIONS) {
+
+		fprintf(stderr, " # %10.3f IPC  ",
+			(double)count[0] / (double)runtime_cycles_avg);
+	} else {
+		if (runtime_nsecs_avg) {
+			fprintf(stderr, " # %10.3f M/sec",
+				(double)count[0]/runtime_nsecs_avg*1000.0);
+		}
+	}
+	print_noise(count, noise);
 }
 
 /*
@@ -189,87 +324,123 @@ static void read_counter(int counter)
  */
 static void print_counter(int counter)
 {
-	__u64 *count;
+	u64 *count, *noise;
 	int scaled;
 
-	count = event_res[counter];
-	scaled = event_scaled[counter];
+	count = event_res_avg[counter];
+	noise = event_res_noise[counter];
+	scaled = event_scaled_avg[counter];
 
 	if (scaled == -1) {
-		fprintf(stderr, " %14s  %-20s\n",
+		fprintf(stderr, " %14s  %-24s\n",
 			"<not counted>", event_name(counter));
 		return;
 	}
 
-	if (nsec_counter(counter)) {
-		double msecs = (double)count[0] / 1000000;
-
-		fprintf(stderr, " %14.6f  %-20s",
-			msecs, event_name(counter));
-		if (attrs[counter].type == PERF_TYPE_SOFTWARE &&
-			attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) {
+	if (nsec_counter(counter))
+		nsec_printout(counter, count, noise);
+	else
+		abs_printout(counter, count, noise);
 
-			if (walltime_nsecs)
-				fprintf(stderr, " # %11.3f CPU utilization factor",
-					(double)count[0] / (double)walltime_nsecs);
-		}
-	} else {
-		fprintf(stderr, " %14Ld  %-20s",
-			count[0], event_name(counter));
-		if (runtime_nsecs)
-			fprintf(stderr, " # %11.3f M/sec",
-				(double)count[0]/runtime_nsecs*1000.0);
-		if (runtime_cycles &&
-			attrs[counter].type == PERF_TYPE_HARDWARE &&
-				attrs[counter].config == PERF_COUNT_HW_INSTRUCTIONS) {
-
-			fprintf(stderr, " # %1.3f per cycle",
-				(double)count[0] / (double)runtime_cycles);
-		}
-	}
 	if (scaled)
 		fprintf(stderr, "  (scaled from %.2f%%)",
 			(double) count[2] / count[1] * 100);
+
 	fprintf(stderr, "\n");
 }
 
-static int do_perf_stat(int argc, const char **argv)
+/*
+ * normalize_noise noise values down to stddev:
+ */
+static void normalize_noise(u64 *val)
 {
-	unsigned long long t0, t1;
-	int counter;
-	int status;
-	int pid;
-	int i;
+	double res;
 
-	if (!system_wide)
-		nr_cpus = 1;
+	res = (double)*val / (run_count * sqrt((double)run_count));
 
-	for (counter = 0; counter < nr_counters; counter++)
-		create_perf_stat_counter(counter);
+	*val = (u64)res;
+}
 
-	/*
-	 * Enable counters and exec the command:
-	 */
-	t0 = rdclock();
-	prctl(PR_TASK_PERF_COUNTERS_ENABLE);
+static void update_avg(const char *name, int idx, u64 *avg, u64 *val)
+{
+	*avg += *val;
 
-	if ((pid = fork()) < 0)
-		perror("failed to fork");
+	if (verbose > 1)
+		fprintf(stderr, "debug: %20s[%d]: %Ld\n", name, idx, *val);
+}
+/*
+ * Calculate the averages and noises:
+ */
+static void calc_avg(void)
+{
+	int i, j;
+
+	if (verbose > 1)
+		fprintf(stderr, "\n");
+
+	for (i = 0; i < run_count; i++) {
+		update_avg("runtime", 0, &runtime_nsecs_avg, runtime_nsecs + i);
+		update_avg("walltime", 0, &walltime_nsecs_avg, walltime_nsecs + i);
+		update_avg("runtime_cycles", 0, &runtime_cycles_avg, runtime_cycles + i);
+
+		for (j = 0; j < nr_counters; j++) {
+			update_avg("counter/0", j,
+				event_res_avg[j]+0, event_res[i][j]+0);
+			update_avg("counter/1", j,
+				event_res_avg[j]+1, event_res[i][j]+1);
+			update_avg("counter/2", j,
+				event_res_avg[j]+2, event_res[i][j]+2);
+			if (event_scaled[i][j] != -1)
+				update_avg("scaled", j,
+					event_scaled_avg + j, event_scaled[i]+j);
+			else
+				event_scaled_avg[j] = -1;
+		}
+	}
+	runtime_nsecs_avg /= run_count;
+	walltime_nsecs_avg /= run_count;
+	runtime_cycles_avg /= run_count;
+
+	for (j = 0; j < nr_counters; j++) {
+		event_res_avg[j][0] /= run_count;
+		event_res_avg[j][1] /= run_count;
+		event_res_avg[j][2] /= run_count;
+	}
 
-	if (!pid) {
-		if (execvp(argv[0], (char **)argv)) {
-			perror(argv[0]);
-			exit(-1);
+	for (i = 0; i < run_count; i++) {
+		runtime_nsecs_noise +=
+			abs((s64)(runtime_nsecs[i] - runtime_nsecs_avg));
+		walltime_nsecs_noise +=
+			abs((s64)(walltime_nsecs[i] - walltime_nsecs_avg));
+		runtime_cycles_noise +=
+			abs((s64)(runtime_cycles[i] - runtime_cycles_avg));
+
+		for (j = 0; j < nr_counters; j++) {
+			event_res_noise[j][0] +=
+				abs((s64)(event_res[i][j][0] - event_res_avg[j][0]));
+			event_res_noise[j][1] +=
+				abs((s64)(event_res[i][j][1] - event_res_avg[j][1]));
+			event_res_noise[j][2] +=
+				abs((s64)(event_res[i][j][2] - event_res_avg[j][2]));
 		}
 	}
 
-	while (wait(&status) >= 0)
-		;
+	normalize_noise(&runtime_nsecs_noise);
+	normalize_noise(&walltime_nsecs_noise);
+	normalize_noise(&runtime_cycles_noise);
 
-	prctl(PR_TASK_PERF_COUNTERS_DISABLE);
-	t1 = rdclock();
+	for (j = 0; j < nr_counters; j++) {
+		normalize_noise(&event_res_noise[j][0]);
+		normalize_noise(&event_res_noise[j][1]);
+		normalize_noise(&event_res_noise[j][2]);
+	}
+}
 
-	walltime_nsecs = t1 - t0;
+static void print_stat(int argc, const char **argv)
+{
+	int i, counter;
+
+	calc_avg();
 
 	fflush(stdout);
 
@@ -279,22 +450,22 @@ static int do_perf_stat(int argc, const char **argv)
 	for (i = 1; i < argc; i++)
 		fprintf(stderr, " %s", argv[i]);
 
-	fprintf(stderr, "\':\n");
-	fprintf(stderr, "\n");
-
-	for (counter = 0; counter < nr_counters; counter++)
-		read_counter(counter);
+	fprintf(stderr, "\'");
+	if (run_count > 1)
+		fprintf(stderr, " (%d runs)", run_count);
+	fprintf(stderr, ":\n\n");
 
 	for (counter = 0; counter < nr_counters; counter++)
 		print_counter(counter);
 
-
 	fprintf(stderr, "\n");
-	fprintf(stderr, " Wall-clock time elapsed: %12.6f msecs\n",
-			(double)(t1-t0)/1e6);
-	fprintf(stderr, "\n");
-
-	return 0;
+	fprintf(stderr, " %14.9f  seconds time elapsed",
+			(double)walltime_nsecs_avg/1e9);
+	if (run_count > 1) {
+		fprintf(stderr, "   ( +- %7.3f%% )",
+			100.0*(double)walltime_nsecs_noise/(double)walltime_nsecs_avg);
+	}
+	fprintf(stderr, "\n\n");
 }
 
 static volatile int signr = -1;
@@ -327,26 +498,33 @@ static const struct option options[] = {
 	OPT_INTEGER('p', "pid", &target_pid,
 		    "stat events on existing pid"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
-			    "system-wide collection from all CPUs"),
+		    "system-wide collection from all CPUs"),
 	OPT_BOOLEAN('S', "scale", &scale,
-			    "scale/normalize counters"),
+		    "scale/normalize counters"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
+	OPT_INTEGER('r', "repeat", &run_count,
+		    "repeat command and print average + stddev (max: 100)"),
+	OPT_BOOLEAN('n', "null", &null_run,
+		    "null run - dont start any counters"),
 	OPT_END()
 };
 
 int cmd_stat(int argc, const char **argv, const char *prefix)
 {
-	page_size = sysconf(_SC_PAGE_SIZE);
-
-	memcpy(attrs, default_attrs, sizeof(attrs));
+	int status;
 
 	argc = parse_options(argc, argv, options, stat_usage, 0);
 	if (!argc)
 		usage_with_options(stat_usage, options);
+	if (run_count <= 0 || run_count > MAX_RUN)
+		usage_with_options(stat_usage, options);
 
-	if (!nr_counters)
-		nr_counters = 8;
+	/* Set attrs and nr_counters if no event is selected and !null_run */
+	if (!null_run && !nr_counters) {
+		memcpy(attrs, default_attrs, sizeof(default_attrs));
+		nr_counters = ARRAY_SIZE(default_attrs);
+	}
 
 	nr_cpus = sysconf(_SC_NPROCESSORS_ONLN);
 	assert(nr_cpus <= MAX_NR_CPUS);
@@ -363,5 +541,14 @@ int cmd_stat(int argc, const char **argv, const char *prefix)
 	signal(SIGALRM, skip_signal);
 	signal(SIGABRT, skip_signal);
 
-	return do_perf_stat(argc, argv);
+	status = 0;
+	for (run_idx = 0; run_idx < run_count; run_idx++) {
+		if (run_count != 1 && verbose)
+			fprintf(stderr, "[ perf stat: executing run #%d ... ]\n", run_idx + 1);
+		status = run_perf_stat(argc, argv);
+	}
+
+	print_stat(argc, argv);
+
+	return status;
 }
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index fe338d3c5d7e..cf0d21f1ae10 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -54,7 +54,7 @@ static int			system_wide			=  0;
 
 static int			default_interval		= 100000;
 
-static __u64			count_filter			=  5;
+static u64			count_filter			=  5;
 static int			print_entries			= 15;
 
 static int			target_pid			= -1;
@@ -79,8 +79,8 @@ static int			dump_symtab;
  * Symbols
  */
 
-static __u64			min_ip;
-static __u64			max_ip = -1ll;
+static u64			min_ip;
+static u64			max_ip = -1ll;
 
 struct sym_entry {
 	struct rb_node		rb_node;
@@ -194,7 +194,7 @@ static void print_sym_table(void)
 		100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec)));
 
 	if (nr_counters == 1) {
-		printf("%Ld", attrs[0].sample_period);
+		printf("%Ld", (u64)attrs[0].sample_period);
 		if (freq)
 			printf("Hz ");
 		else
@@ -372,7 +372,7 @@ out_delete_dso:
 /*
  * Binary search in the histogram table and record the hit:
  */
-static void record_ip(__u64 ip, int counter)
+static void record_ip(u64 ip, int counter)
 {
 	struct symbol *sym = dso__find_symbol(kernel_dso, ip);
 
@@ -392,11 +392,11 @@ static void record_ip(__u64 ip, int counter)
 	samples--;
 }
 
-static void process_event(__u64 ip, int counter)
+static void process_event(u64 ip, int counter, int user)
 {
 	samples++;
 
-	if (ip < min_ip || ip > max_ip) {
+	if (user) {
 		userspace_samples++;
 		return;
 	}
@@ -463,15 +463,15 @@ static void mmap_read_counter(struct mmap_data *md)
 	for (; old != head;) {
 		struct ip_event {
 			struct perf_event_header header;
-			__u64 ip;
-			__u32 pid, target_pid;
+			u64 ip;
+			u32 pid, target_pid;
 		};
 		struct mmap_event {
 			struct perf_event_header header;
-			__u32 pid, target_pid;
-			__u64 start;
-			__u64 len;
-			__u64 pgoff;
+			u32 pid, target_pid;
+			u64 start;
+			u64 len;
+			u64 pgoff;
 			char filename[PATH_MAX];
 		};
 
@@ -509,9 +509,10 @@ static void mmap_read_counter(struct mmap_data *md)
 
 		old += size;
 
-		if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) {
-			if (event->header.type & PERF_SAMPLE_IP)
-				process_event(event->ip.ip, md->counter);
+		if (event->header.type == PERF_EVENT_SAMPLE) {
+			int user =
+	(event->header.misc & PERF_EVENT_MISC_CPUMODE_MASK) == PERF_EVENT_MISC_USER;
+			process_event(event->ip.ip, md->counter, user);
 		}
 	}
 
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 87a1aca4a424..8f729aedc1a3 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -13,12 +13,29 @@
 #define cpu_relax()	asm volatile ("" ::: "memory");
 #endif
 
+#ifdef __s390__
+#include "../../arch/s390/include/asm/unistd.h"
+#define rmb()		asm volatile("bcr 15,0" ::: "memory")
+#define cpu_relax()	asm volatile("" ::: "memory");
+#endif
+
+#ifdef __sh__
+#include "../../arch/sh/include/asm/unistd.h"
+#if defined(__SH4A__) || defined(__SH5__)
+# define rmb()		asm volatile("synco" ::: "memory")
+#else
+# define rmb()		asm volatile("" ::: "memory")
+#endif
+#define cpu_relax()	asm volatile("" ::: "memory")
+#endif
+
 #include <time.h>
 #include <unistd.h>
 #include <sys/types.h>
 #include <sys/syscall.h>
 
 #include "../../include/linux/perf_counter.h"
+#include "util/types.h"
 
 /*
  * prctl(PR_TASK_PERF_COUNTERS_DISABLE) will (cheaply) disable all
@@ -65,4 +82,9 @@ sys_perf_counter_open(struct perf_counter_attr *attr,
 #define MAX_COUNTERS			256
 #define MAX_NR_CPUS			256
 
+struct ip_callchain {
+	u64 nr;
+	u64 ips[0];
+};
+
 #endif
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
new file mode 100644
index 000000000000..ad3c28578961
--- /dev/null
+++ b/tools/perf/util/callchain.c
@@ -0,0 +1,174 @@
+/*
+ * Copyright (C) 2009, Frederic Weisbecker <fweisbec@gmail.com>
+ *
+ * Handle the callchains from the stream in an ad-hoc radix tree and then
+ * sort them in an rbtree.
+ *
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <stdbool.h>
+#include <errno.h>
+
+#include "callchain.h"
+
+
+static void 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;
+
+	while (*p) {
+		parent = *p;
+		rnode = rb_entry(parent, struct callchain_node, rb_node);
+
+		if (rnode->hit < chain->hit)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&chain->rb_node, parent, p);
+	rb_insert_color(&chain->rb_node, root);
+}
+
+/*
+ * Once we get every callchains from the stream, we can now
+ * sort them by hit
+ */
+void sort_chain_to_rbtree(struct rb_root *rb_root, struct callchain_node *node)
+{
+	struct callchain_node *child;
+
+	list_for_each_entry(child, &node->children, brothers)
+		sort_chain_to_rbtree(rb_root, child);
+
+	if (node->hit)
+		rb_insert_callchain(rb_root, node);
+}
+
+static struct callchain_node *create_child(struct callchain_node *parent)
+{
+	struct callchain_node *new;
+
+	new = malloc(sizeof(*new));
+	if (!new) {
+		perror("not enough memory to create child for code path tree");
+		return NULL;
+	}
+	new->parent = parent;
+	INIT_LIST_HEAD(&new->children);
+	INIT_LIST_HEAD(&new->val);
+	list_add_tail(&new->brothers, &parent->children);
+
+	return new;
+}
+
+static void
+fill_node(struct callchain_node *node, struct ip_callchain *chain, int start)
+{
+	int i;
+
+	for (i = start; i < chain->nr; i++) {
+		struct callchain_list *call;
+
+		call = malloc(sizeof(*chain));
+		if (!call) {
+			perror("not enough memory for the code path tree");
+			return;
+		}
+		call->ip = chain->ips[i];
+		list_add_tail(&call->list, &node->val);
+	}
+	node->val_nr = i - start;
+}
+
+static void add_child(struct callchain_node *parent, struct ip_callchain *chain)
+{
+	struct callchain_node *new;
+
+	new = create_child(parent);
+	fill_node(new, chain, parent->val_nr);
+
+	new->hit = 1;
+}
+
+static void
+split_add_child(struct callchain_node *parent, struct ip_callchain *chain,
+		struct callchain_list *to_split, int idx)
+{
+	struct callchain_node *new;
+
+	/* split */
+	new = create_child(parent);
+	list_move_tail(&to_split->list, &new->val);
+	new->hit = parent->hit;
+	parent->hit = 0;
+	parent->val_nr = idx;
+
+	/* create the new one */
+	add_child(parent, chain);
+}
+
+static int
+__append_chain(struct callchain_node *root, struct ip_callchain *chain,
+		int start);
+
+static int
+__append_chain_children(struct callchain_node *root, struct ip_callchain *chain)
+{
+	struct callchain_node *rnode;
+
+	/* lookup in childrens */
+	list_for_each_entry(rnode, &root->children, brothers) {
+		int ret = __append_chain(rnode, chain, root->val_nr);
+		if (!ret)
+			return 0;
+	}
+	return -1;
+}
+
+static int
+__append_chain(struct callchain_node *root, struct ip_callchain *chain,
+		int start)
+{
+	struct callchain_list *cnode;
+	int i = start;
+	bool found = false;
+
+	/* lookup in the current node */
+	list_for_each_entry(cnode, &root->val, list) {
+		if (cnode->ip != chain->ips[i++])
+			break;
+		if (!found)
+			found = true;
+		if (i == chain->nr)
+			break;
+	}
+
+	/* matches not, relay on the parent */
+	if (!found)
+		return -1;
+
+	/* we match only a part of the node. Split it and add the new chain */
+	if (i < root->val_nr) {
+		split_add_child(root, chain, cnode, i);
+		return 0;
+	}
+
+	/* we match 100% of the path, increment the hit */
+	if (i == root->val_nr) {
+		root->hit++;
+		return 0;
+	}
+
+	return __append_chain_children(root, chain);
+}
+
+void append_chain(struct callchain_node *root, struct ip_callchain *chain)
+{
+	if (__append_chain_children(root, chain) == -1)
+		add_child(root, chain);
+}
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
new file mode 100644
index 000000000000..fa1cd2f71fd3
--- /dev/null
+++ b/tools/perf/util/callchain.h
@@ -0,0 +1,33 @@
+#ifndef __PERF_CALLCHAIN_H
+#define __PERF_CALLCHAIN_H
+
+#include "../perf.h"
+#include "list.h"
+#include "rbtree.h"
+
+
+struct callchain_node {
+	struct callchain_node	*parent;
+	struct list_head	brothers;
+	struct list_head 	children;
+	struct list_head 	val;
+	struct rb_node		rb_node;
+	int			val_nr;
+	int			hit;
+};
+
+struct callchain_list {
+	unsigned long		ip;
+	struct list_head	list;
+};
+
+static inline void callchain_init(struct callchain_node *node)
+{
+	INIT_LIST_HEAD(&node->brothers);
+	INIT_LIST_HEAD(&node->children);
+	INIT_LIST_HEAD(&node->val);
+}
+
+void append_chain(struct callchain_node *root, struct ip_callchain *chain);
+void sort_chain_to_rbtree(struct rb_root *rb_root, struct callchain_node *node);
+#endif
diff --git a/tools/perf/util/ctype.c b/tools/perf/util/ctype.c
index b90ec004f29c..0b791bd346bc 100644
--- a/tools/perf/util/ctype.c
+++ b/tools/perf/util/ctype.c
@@ -11,16 +11,21 @@ enum {
 	D = GIT_DIGIT,
 	G = GIT_GLOB_SPECIAL,	/* *, ?, [, \\ */
 	R = GIT_REGEX_SPECIAL,	/* $, (, ), +, ., ^, {, | * */
+	P = GIT_PRINT_EXTRA,	/* printable - alpha - digit - glob - regex */
+
+	PS = GIT_SPACE | GIT_PRINT_EXTRA,
 };
 
 unsigned char sane_ctype[256] = {
+/*	0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F			    */
+
 	0, 0, 0, 0, 0, 0, 0, 0, 0, S, S, 0, 0, S, 0, 0,		/*   0.. 15 */
 	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,		/*  16.. 31 */
-	S, 0, 0, 0, R, 0, 0, 0, R, R, G, R, 0, 0, R, 0,		/*  32.. 47 */
-	D, D, D, D, D, D, D, D, D, D, 0, 0, 0, 0, 0, G,		/*  48.. 63 */
-	0, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A,		/*  64.. 79 */
-	A, A, A, A, A, A, A, A, A, A, A, G, G, 0, R, 0,		/*  80.. 95 */
-	0, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A,		/*  96..111 */
-	A, A, A, A, A, A, A, A, A, A, A, R, R, 0, 0, 0,		/* 112..127 */
+	PS,P, P, P, R, P, P, P, R, R, G, R, P, P, R, P,		/*  32.. 47 */
+	D, D, D, D, D, D, D, D, D, D, P, P, P, P, P, G,		/*  48.. 63 */
+	P, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A,		/*  64.. 79 */
+	A, A, A, A, A, A, A, A, A, A, A, G, G, P, R, P,		/*  80.. 95 */
+	P, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A,		/*  96..111 */
+	A, A, A, A, A, A, A, A, A, A, A, R, R, P, P, 0,		/* 112..127 */
 	/* Nothing in the 128.. range */
 };
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
new file mode 100644
index 000000000000..450384b3bbe5
--- /dev/null
+++ b/tools/perf/util/header.c
@@ -0,0 +1,242 @@
+#include <sys/types.h>
+#include <unistd.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "util.h"
+#include "header.h"
+
+/*
+ *
+ */
+
+struct perf_header_attr *perf_header_attr__new(struct perf_counter_attr *attr)
+{
+	struct perf_header_attr *self = malloc(sizeof(*self));
+
+	if (!self)
+		die("nomem");
+
+	self->attr = *attr;
+	self->ids = 0;
+	self->size = 1;
+	self->id = malloc(sizeof(u64));
+
+	if (!self->id)
+		die("nomem");
+
+	return self;
+}
+
+void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
+{
+	int pos = self->ids;
+
+	self->ids++;
+	if (self->ids > self->size) {
+		self->size *= 2;
+		self->id = realloc(self->id, self->size * sizeof(u64));
+		if (!self->id)
+			die("nomem");
+	}
+	self->id[pos] = id;
+}
+
+/*
+ *
+ */
+
+struct perf_header *perf_header__new(void)
+{
+	struct perf_header *self = malloc(sizeof(*self));
+
+	if (!self)
+		die("nomem");
+
+	self->frozen = 0;
+
+	self->attrs = 0;
+	self->size = 1;
+	self->attr = malloc(sizeof(void *));
+
+	if (!self->attr)
+		die("nomem");
+
+	self->data_offset = 0;
+	self->data_size = 0;
+
+	return self;
+}
+
+void perf_header__add_attr(struct perf_header *self,
+			   struct perf_header_attr *attr)
+{
+	int pos = self->attrs;
+
+	if (self->frozen)
+		die("frozen");
+
+	self->attrs++;
+	if (self->attrs > self->size) {
+		self->size *= 2;
+		self->attr = realloc(self->attr, self->size * sizeof(void *));
+		if (!self->attr)
+			die("nomem");
+	}
+	self->attr[pos] = attr;
+}
+
+static const char *__perf_magic = "PERFFILE";
+
+#define PERF_MAGIC	(*(u64 *)__perf_magic)
+
+struct perf_file_section {
+	u64 offset;
+	u64 size;
+};
+
+struct perf_file_attr {
+	struct perf_counter_attr	attr;
+	struct perf_file_section	ids;
+};
+
+struct perf_file_header {
+	u64				magic;
+	u64				size;
+	u64				attr_size;
+	struct perf_file_section	attrs;
+	struct perf_file_section	data;
+};
+
+static void do_write(int fd, void *buf, size_t size)
+{
+	while (size) {
+		int ret = write(fd, buf, size);
+
+		if (ret < 0)
+			die("failed to write");
+
+		size -= ret;
+		buf += ret;
+	}
+}
+
+void perf_header__write(struct perf_header *self, int fd)
+{
+	struct perf_file_header f_header;
+	struct perf_file_attr   f_attr;
+	struct perf_header_attr	*attr;
+	int i;
+
+	lseek(fd, sizeof(f_header), SEEK_SET);
+
+
+	for (i = 0; i < self->attrs; i++) {
+		attr = self->attr[i];
+
+		attr->id_offset = lseek(fd, 0, SEEK_CUR);
+		do_write(fd, attr->id, attr->ids * sizeof(u64));
+	}
+
+
+	self->attr_offset = lseek(fd, 0, SEEK_CUR);
+
+	for (i = 0; i < self->attrs; i++) {
+		attr = self->attr[i];
+
+		f_attr = (struct perf_file_attr){
+			.attr = attr->attr,
+			.ids  = {
+				.offset = attr->id_offset,
+				.size   = attr->ids * sizeof(u64),
+			}
+		};
+		do_write(fd, &f_attr, sizeof(f_attr));
+	}
+
+
+	self->data_offset = lseek(fd, 0, SEEK_CUR);
+
+	f_header = (struct perf_file_header){
+		.magic	   = PERF_MAGIC,
+		.size	   = sizeof(f_header),
+		.attr_size = sizeof(f_attr),
+		.attrs = {
+			.offset = self->attr_offset,
+			.size   = self->attrs * sizeof(f_attr),
+		},
+		.data = {
+			.offset = self->data_offset,
+			.size	= self->data_size,
+		},
+	};
+
+	lseek(fd, 0, SEEK_SET);
+	do_write(fd, &f_header, sizeof(f_header));
+	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+
+	self->frozen = 1;
+}
+
+static void do_read(int fd, void *buf, size_t size)
+{
+	while (size) {
+		int ret = read(fd, buf, size);
+
+		if (ret < 0)
+			die("failed to read");
+
+		size -= ret;
+		buf += ret;
+	}
+}
+
+struct perf_header *perf_header__read(int fd)
+{
+	struct perf_header	*self = perf_header__new();
+	struct perf_file_header f_header;
+	struct perf_file_attr	f_attr;
+	u64			f_id;
+
+	int nr_attrs, nr_ids, i, j;
+
+	lseek(fd, 0, SEEK_SET);
+	do_read(fd, &f_header, sizeof(f_header));
+
+	if (f_header.magic	!= PERF_MAGIC		||
+	    f_header.size	!= sizeof(f_header)	||
+	    f_header.attr_size	!= sizeof(f_attr))
+		die("incompatible file format");
+
+	nr_attrs = f_header.attrs.size / sizeof(f_attr);
+	lseek(fd, f_header.attrs.offset, SEEK_SET);
+
+	for (i = 0; i < nr_attrs; i++) {
+		struct perf_header_attr *attr;
+		off_t tmp = lseek(fd, 0, SEEK_CUR);
+
+		do_read(fd, &f_attr, sizeof(f_attr));
+
+		attr = perf_header_attr__new(&f_attr.attr);
+
+		nr_ids = f_attr.ids.size / sizeof(u64);
+		lseek(fd, f_attr.ids.offset, SEEK_SET);
+
+		for (j = 0; j < nr_ids; j++) {
+			do_read(fd, &f_id, sizeof(f_id));
+
+			perf_header_attr__add_id(attr, f_id);
+		}
+		perf_header__add_attr(self, attr);
+		lseek(fd, tmp, SEEK_SET);
+	}
+
+	self->data_offset = f_header.data.offset;
+	self->data_size   = f_header.data.size;
+
+	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+
+	self->frozen = 1;
+
+	return self;
+}
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
new file mode 100644
index 000000000000..b5ef53ad4c7a
--- /dev/null
+++ b/tools/perf/util/header.h
@@ -0,0 +1,37 @@
+#ifndef _PERF_HEADER_H
+#define _PERF_HEADER_H
+
+#include "../../../include/linux/perf_counter.h"
+#include <sys/types.h>
+#include "types.h"
+
+struct perf_header_attr {
+	struct perf_counter_attr attr;
+	int ids, size;
+	u64 *id;
+	off_t id_offset;
+};
+
+struct perf_header {
+	int frozen;
+	int attrs, size;
+	struct perf_header_attr **attr;
+	off_t attr_offset;
+	u64 data_offset;
+	u64 data_size;
+};
+
+struct perf_header *perf_header__read(int fd);
+void perf_header__write(struct perf_header *self, int fd);
+
+void perf_header__add_attr(struct perf_header *self,
+			   struct perf_header_attr *attr);
+
+struct perf_header_attr *
+perf_header_attr__new(struct perf_counter_attr *attr);
+void perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
+
+
+struct perf_header *perf_header__new(void);
+
+#endif /* _PERF_HEADER_H */
diff --git a/tools/perf/util/help.c b/tools/perf/util/help.c
index 6653f7dd1d78..17a00e0df2c4 100644
--- a/tools/perf/util/help.c
+++ b/tools/perf/util/help.c
@@ -126,21 +126,6 @@ static int is_executable(const char *name)
 	    !S_ISREG(st.st_mode))
 		return 0;
 
-#ifdef __MINGW32__
-	/* cannot trust the executable bit, peek into the file instead */
-	char buf[3] = { 0 };
-	int n;
-	int fd = open(name, O_RDONLY);
-	st.st_mode &= ~S_IXUSR;
-	if (fd >= 0) {
-		n = read(fd, buf, 2);
-		if (n == 2)
-			/* DOS executables start with "MZ" */
-			if (!strcmp(buf, "#!") || !strcmp(buf, "MZ"))
-				st.st_mode |= S_IXUSR;
-		close(fd);
-	}
-#endif
 	return st.st_mode & S_IXUSR;
 }
 
diff --git a/tools/perf/util/pager.c b/tools/perf/util/pager.c
index a28bccae5458..1915de20dcac 100644
--- a/tools/perf/util/pager.c
+++ b/tools/perf/util/pager.c
@@ -9,7 +9,6 @@
 
 static int spawned_pager;
 
-#ifndef __MINGW32__
 static void pager_preexec(void)
 {
 	/*
@@ -24,7 +23,6 @@ static void pager_preexec(void)
 
 	setenv("LESS", "FRSX", 0);
 }
-#endif
 
 static const char *pager_argv[] = { "sh", "-c", NULL, NULL };
 static struct child_process pager_process;
@@ -70,9 +68,8 @@ void setup_pager(void)
 	pager_argv[2] = pager;
 	pager_process.argv = pager_argv;
 	pager_process.in = -1;
-#ifndef __MINGW32__
 	pager_process.preexec_cb = pager_preexec;
-#endif
+
 	if (start_command(&pager_process))
 		return;
 
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 5a72586e1df0..4d042f104cdc 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -13,35 +13,31 @@ int					nr_counters;
 struct perf_counter_attr		attrs[MAX_COUNTERS];
 
 struct event_symbol {
-	__u8	type;
-	__u64	config;
+	u8	type;
+	u64	config;
 	char	*symbol;
+	char	*alias;
 };
 
-#define C(x, y) .type = PERF_TYPE_##x, .config = PERF_COUNT_##y
-#define CR(x, y) .type = PERF_TYPE_##x, .config = y
+#define CHW(x) .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_##x
+#define CSW(x) .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_##x
 
 static struct event_symbol event_symbols[] = {
-  { C(HARDWARE, HW_CPU_CYCLES),		"cpu-cycles",		},
-  { C(HARDWARE, HW_CPU_CYCLES),		"cycles",		},
-  { C(HARDWARE, HW_INSTRUCTIONS),	"instructions",		},
-  { C(HARDWARE, HW_CACHE_REFERENCES),	"cache-references",	},
-  { C(HARDWARE, HW_CACHE_MISSES),	"cache-misses",		},
-  { C(HARDWARE, HW_BRANCH_INSTRUCTIONS),"branch-instructions",	},
-  { C(HARDWARE, HW_BRANCH_INSTRUCTIONS),"branches",		},
-  { C(HARDWARE, HW_BRANCH_MISSES),	"branch-misses",	},
-  { C(HARDWARE, HW_BUS_CYCLES),		"bus-cycles",		},
-
-  { C(SOFTWARE, SW_CPU_CLOCK),		"cpu-clock",		},
-  { C(SOFTWARE, SW_TASK_CLOCK),		"task-clock",		},
-  { C(SOFTWARE, SW_PAGE_FAULTS),	"page-faults",		},
-  { C(SOFTWARE, SW_PAGE_FAULTS),	"faults",		},
-  { C(SOFTWARE, SW_PAGE_FAULTS_MIN),	"minor-faults",		},
-  { C(SOFTWARE, SW_PAGE_FAULTS_MAJ),	"major-faults",		},
-  { C(SOFTWARE, SW_CONTEXT_SWITCHES),	"context-switches",	},
-  { C(SOFTWARE, SW_CONTEXT_SWITCHES),	"cs",			},
-  { C(SOFTWARE, SW_CPU_MIGRATIONS),	"cpu-migrations",	},
-  { C(SOFTWARE, SW_CPU_MIGRATIONS),	"migrations",		},
+  { CHW(CPU_CYCLES),		"cpu-cycles",		"cycles"	},
+  { CHW(INSTRUCTIONS),		"instructions",		""		},
+  { CHW(CACHE_REFERENCES),	"cache-references",	""		},
+  { CHW(CACHE_MISSES),		"cache-misses",		""		},
+  { CHW(BRANCH_INSTRUCTIONS),	"branch-instructions",	"branches"	},
+  { CHW(BRANCH_MISSES),		"branch-misses",	""		},
+  { CHW(BUS_CYCLES),		"bus-cycles",		""		},
+
+  { CSW(CPU_CLOCK),		"cpu-clock",		""		},
+  { CSW(TASK_CLOCK),		"task-clock",		""		},
+  { CSW(PAGE_FAULTS),		"page-faults",		"faults"	},
+  { CSW(PAGE_FAULTS_MIN),	"minor-faults",		""		},
+  { CSW(PAGE_FAULTS_MAJ),	"major-faults",		""		},
+  { CSW(CONTEXT_SWITCHES),	"context-switches",	"cs"		},
+  { CSW(CPU_MIGRATIONS),	"cpu-migrations",	"migrations"	},
 };
 
 #define __PERF_COUNTER_FIELD(config, name) \
@@ -63,8 +59,8 @@ static char *hw_event_names[] = {
 };
 
 static char *sw_event_names[] = {
-	"cpu-clock-ticks",
-	"task-clock-ticks",
+	"cpu-clock-msecs",
+	"task-clock-msecs",
 	"page-faults",
 	"context-switches",
 	"CPU-migrations",
@@ -74,29 +70,73 @@ static char *sw_event_names[] = {
 
 #define MAX_ALIASES 8
 
-static char *hw_cache [][MAX_ALIASES] = {
-	{ "L1-data"		, "l1-d", "l1d"					},
-	{ "L1-instruction"	, "l1-i", "l1i"					},
-	{ "L2"			, "l2"						},
-	{ "Data-TLB"		, "dtlb", "d-tlb"				},
-	{ "Instruction-TLB"	, "itlb", "i-tlb"				},
-	{ "Branch"		, "bpu" , "btb", "bpc"				},
+static char *hw_cache[][MAX_ALIASES] = {
+ { "L1-d$",	"l1-d",		"l1d",		"L1-data",		},
+ { "L1-i$",	"l1-i",		"l1i",		"L1-instruction",	},
+ { "LLC",	"L2"							},
+ { "dTLB",	"d-tlb",	"Data-TLB",				},
+ { "iTLB",	"i-tlb",	"Instruction-TLB",			},
+ { "branch",	"branches",	"bpu",		"btb",		"bpc",	},
 };
 
-static char *hw_cache_op [][MAX_ALIASES] = {
-	{ "Load"		, "read"					},
-	{ "Store"		, "write"					},
-	{ "Prefetch"		, "speculative-read", "speculative-load"	},
+static char *hw_cache_op[][MAX_ALIASES] = {
+ { "load",	"loads",	"read",					},
+ { "store",	"stores",	"write",				},
+ { "prefetch",	"prefetches",	"speculative-read", "speculative-load",	},
 };
 
-static char *hw_cache_result [][MAX_ALIASES] = {
-	{ "Reference"		, "ops", "access"				},
-	{ "Miss"								},
+static char *hw_cache_result[][MAX_ALIASES] = {
+ { "refs",	"Reference",	"ops",		"access",		},
+ { "misses",	"miss",							},
 };
 
+#define C(x)		PERF_COUNT_HW_CACHE_##x
+#define CACHE_READ	(1 << C(OP_READ))
+#define CACHE_WRITE	(1 << C(OP_WRITE))
+#define CACHE_PREFETCH	(1 << C(OP_PREFETCH))
+#define COP(x)		(1 << x)
+
+/*
+ * cache operartion stat
+ * L1I : Read and prefetch only
+ * ITLB and BPU : Read-only
+ */
+static unsigned long hw_cache_stat[C(MAX)] = {
+ [C(L1D)]	= (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(L1I)]	= (CACHE_READ | CACHE_PREFETCH),
+ [C(LL)]	= (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(DTLB)]	= (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(ITLB)]	= (CACHE_READ),
+ [C(BPU)]	= (CACHE_READ),
+};
+
+static int is_cache_op_valid(u8 cache_type, u8 cache_op)
+{
+	if (hw_cache_stat[cache_type] & COP(cache_op))
+		return 1;	/* valid */
+	else
+		return 0;	/* invalid */
+}
+
+static char *event_cache_name(u8 cache_type, u8 cache_op, u8 cache_result)
+{
+	static char name[50];
+
+	if (cache_result) {
+		sprintf(name, "%s-%s-%s", hw_cache[cache_type][0],
+			hw_cache_op[cache_op][0],
+			hw_cache_result[cache_result][0]);
+	} else {
+		sprintf(name, "%s-%s", hw_cache[cache_type][0],
+			hw_cache_op[cache_op][1]);
+	}
+
+	return name;
+}
+
 char *event_name(int counter)
 {
-	__u64 config = attrs[counter].config;
+	u64 config = attrs[counter].config;
 	int type = attrs[counter].type;
 	static char buf[32];
 
@@ -112,8 +152,7 @@ char *event_name(int counter)
 		return "unknown-hardware";
 
 	case PERF_TYPE_HW_CACHE: {
-		__u8 cache_type, cache_op, cache_result;
-		static char name[100];
+		u8 cache_type, cache_op, cache_result;
 
 		cache_type   = (config >>  0) & 0xff;
 		if (cache_type > PERF_COUNT_HW_CACHE_MAX)
@@ -127,12 +166,10 @@ char *event_name(int counter)
 		if (cache_result > PERF_COUNT_HW_CACHE_RESULT_MAX)
 			return "unknown-ext-hardware-cache-result";
 
-		sprintf(name, "%s-Cache-%s-%ses",
-			hw_cache[cache_type][0],
-			hw_cache_op[cache_op][0],
-			hw_cache_result[cache_result][0]);
+		if (!is_cache_op_valid(cache_type, cache_op))
+			return "invalid-cache";
 
-		return name;
+		return event_cache_name(cache_type, cache_op, cache_result);
 	}
 
 	case PERF_TYPE_SOFTWARE:
@@ -163,7 +200,8 @@ static int parse_aliases(const char *str, char *names[][MAX_ALIASES], int size)
 	return -1;
 }
 
-static int parse_generic_hw_symbols(const char *str, struct perf_counter_attr *attr)
+static int
+parse_generic_hw_symbols(const char *str, struct perf_counter_attr *attr)
 {
 	int cache_type = -1, cache_op = 0, cache_result = 0;
 
@@ -182,6 +220,9 @@ static int parse_generic_hw_symbols(const char *str, struct perf_counter_attr *a
 	if (cache_op == -1)
 		cache_op = PERF_COUNT_HW_CACHE_OP_READ;
 
+	if (!is_cache_op_valid(cache_type, cache_op))
+		return -EINVAL;
+
 	cache_result = parse_aliases(str, hw_cache_result,
 					PERF_COUNT_HW_CACHE_RESULT_MAX);
 	/*
@@ -196,13 +237,26 @@ static int parse_generic_hw_symbols(const char *str, struct perf_counter_attr *a
 	return 0;
 }
 
+static int check_events(const char *str, unsigned int i)
+{
+	if (!strncmp(str, event_symbols[i].symbol,
+		     strlen(event_symbols[i].symbol)))
+		return 1;
+
+	if (strlen(event_symbols[i].alias))
+		if (!strncmp(str, event_symbols[i].alias,
+			     strlen(event_symbols[i].alias)))
+			return 1;
+	return 0;
+}
+
 /*
  * Each event can have multiple symbolic names.
  * Symbolic names are (almost) exactly matched.
  */
 static int parse_event_symbols(const char *str, struct perf_counter_attr *attr)
 {
-	__u64 config, id;
+	u64 config, id;
 	int type;
 	unsigned int i;
 	const char *sep, *pstr;
@@ -235,9 +289,7 @@ static int parse_event_symbols(const char *str, struct perf_counter_attr *attr)
 	}
 
 	for (i = 0; i < ARRAY_SIZE(event_symbols); i++) {
-		if (!strncmp(str, event_symbols[i].symbol,
-			     strlen(event_symbols[i].symbol))) {
-
+		if (check_events(str, i)) {
 			attr->type = event_symbols[i].type;
 			attr->config = event_symbols[i].config;
 
@@ -289,6 +341,7 @@ void print_events(void)
 {
 	struct event_symbol *syms = event_symbols;
 	unsigned int i, type, prev_type = -1;
+	char name[40];
 
 	fprintf(stderr, "\n");
 	fprintf(stderr, "List of pre-defined events (to be used in -e):\n");
@@ -301,14 +354,18 @@ void print_events(void)
 		if (type != prev_type)
 			fprintf(stderr, "\n");
 
-		fprintf(stderr, "  %-30s [%s]\n", syms->symbol,
+		if (strlen(syms->alias))
+			sprintf(name, "%s OR %s", syms->symbol, syms->alias);
+		else
+			strcpy(name, syms->symbol);
+		fprintf(stderr, "  %-40s [%s]\n", name,
 			event_type_descriptors[type]);
 
 		prev_type = type;
 	}
 
 	fprintf(stderr, "\n");
-	fprintf(stderr, "  %-30s [raw hardware event descriptor]\n",
+	fprintf(stderr, "  %-40s [raw hardware event descriptor]\n",
 		"rNNN");
 	fprintf(stderr, "\n");
 
diff --git a/tools/perf/util/run-command.c b/tools/perf/util/run-command.c
index b2f5e854f40a..a3935343091a 100644
--- a/tools/perf/util/run-command.c
+++ b/tools/perf/util/run-command.c
@@ -65,7 +65,6 @@ int start_command(struct child_process *cmd)
 		cmd->err = fderr[0];
 	}
 
-#ifndef __MINGW32__
 	fflush(NULL);
 	cmd->pid = fork();
 	if (!cmd->pid) {
@@ -118,71 +117,6 @@ int start_command(struct child_process *cmd)
 		}
 		exit(127);
 	}
-#else
-	int s0 = -1, s1 = -1, s2 = -1;	/* backups of stdin, stdout, stderr */
-	const char **sargv = cmd->argv;
-	char **env = environ;
-
-	if (cmd->no_stdin) {
-		s0 = dup(0);
-		dup_devnull(0);
-	} else if (need_in) {
-		s0 = dup(0);
-		dup2(fdin[0], 0);
-	} else if (cmd->in) {
-		s0 = dup(0);
-		dup2(cmd->in, 0);
-	}
-
-	if (cmd->no_stderr) {
-		s2 = dup(2);
-		dup_devnull(2);
-	} else if (need_err) {
-		s2 = dup(2);
-		dup2(fderr[1], 2);
-	}
-
-	if (cmd->no_stdout) {
-		s1 = dup(1);
-		dup_devnull(1);
-	} else if (cmd->stdout_to_stderr) {
-		s1 = dup(1);
-		dup2(2, 1);
-	} else if (need_out) {
-		s1 = dup(1);
-		dup2(fdout[1], 1);
-	} else if (cmd->out > 1) {
-		s1 = dup(1);
-		dup2(cmd->out, 1);
-	}
-
-	if (cmd->dir)
-		die("chdir in start_command() not implemented");
-	if (cmd->env) {
-		env = copy_environ();
-		for (; *cmd->env; cmd->env++)
-			env = env_setenv(env, *cmd->env);
-	}
-
-	if (cmd->perf_cmd) {
-		cmd->argv = prepare_perf_cmd(cmd->argv);
-	}
-
-	cmd->pid = mingw_spawnvpe(cmd->argv[0], cmd->argv, env);
-
-	if (cmd->env)
-		free_environ(env);
-	if (cmd->perf_cmd)
-		free(cmd->argv);
-
-	cmd->argv = sargv;
-	if (s0 >= 0)
-		dup2(s0, 0), close(s0);
-	if (s1 >= 0)
-		dup2(s1, 1), close(s1);
-	if (s2 >= 0)
-		dup2(s2, 2), close(s2);
-#endif
 
 	if (cmd->pid < 0) {
 		int err = errno;
@@ -288,14 +222,6 @@ int run_command_v_opt_cd_env(const char **argv, int opt, const char *dir, const
 	return run_command(&cmd);
 }
 
-#ifdef __MINGW32__
-static __stdcall unsigned run_thread(void *data)
-{
-	struct async *async = data;
-	return async->proc(async->fd_for_proc, async->data);
-}
-#endif
-
 int start_async(struct async *async)
 {
 	int pipe_out[2];
@@ -304,7 +230,6 @@ int start_async(struct async *async)
 		return error("cannot create pipe: %s", strerror(errno));
 	async->out = pipe_out[0];
 
-#ifndef __MINGW32__
 	/* Flush stdio before fork() to avoid cloning buffers */
 	fflush(NULL);
 
@@ -319,33 +244,17 @@ int start_async(struct async *async)
 		exit(!!async->proc(pipe_out[1], async->data));
 	}
 	close(pipe_out[1]);
-#else
-	async->fd_for_proc = pipe_out[1];
-	async->tid = (HANDLE) _beginthreadex(NULL, 0, run_thread, async, 0, NULL);
-	if (!async->tid) {
-		error("cannot create thread: %s", strerror(errno));
-		close_pair(pipe_out);
-		return -1;
-	}
-#endif
+
 	return 0;
 }
 
 int finish_async(struct async *async)
 {
-#ifndef __MINGW32__
 	int ret = 0;
 
 	if (wait_or_whine(async->pid))
 		ret = error("waitpid (async) failed");
-#else
-	DWORD ret = 0;
-	if (WaitForSingleObject(async->tid, INFINITE) != WAIT_OBJECT_0)
-		ret = error("waiting for thread failed: %lu", GetLastError());
-	else if (!GetExitCodeThread(async->tid, &ret))
-		ret = error("cannot get thread exit code: %lu", GetLastError());
-	CloseHandle(async->tid);
-#endif
+
 	return ret;
 }
 
diff --git a/tools/perf/util/run-command.h b/tools/perf/util/run-command.h
index 328289f23669..cc1837deba88 100644
--- a/tools/perf/util/run-command.h
+++ b/tools/perf/util/run-command.h
@@ -79,12 +79,7 @@ struct async {
 	int (*proc)(int fd, void *data);
 	void *data;
 	int out;	/* caller reads from here and closes it */
-#ifndef __MINGW32__
 	pid_t pid;
-#else
-	HANDLE tid;
-	int fd_for_proc;
-#endif
 };
 
 int start_async(struct async *async);
diff --git a/tools/perf/util/strbuf.c b/tools/perf/util/strbuf.c
index eaba09306802..464e7ca898cf 100644
--- a/tools/perf/util/strbuf.c
+++ b/tools/perf/util/strbuf.c
@@ -259,7 +259,7 @@ size_t strbuf_fread(struct strbuf *sb, size_t size, FILE *f)
 	res = fread(sb->buf + sb->len, 1, size, f);
 	if (res > 0)
 		strbuf_setlen(sb, sb->len + res);
-	else if (res < 0 && oldalloc == 0)
+	else if (oldalloc == 0)
 		strbuf_release(sb);
 	return res;
 }
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index ec33c0c7f4e2..c93eca9a7be3 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -15,7 +15,7 @@ static int hex(char ch)
  * While we find nice hex chars, build a long_val.
  * Return number of chars processed.
  */
-int hex2u64(const char *ptr, __u64 *long_val)
+int hex2u64(const char *ptr, u64 *long_val)
 {
 	const char *p = ptr;
 	*long_val = 0;
diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h
index 72812c1c9a7a..3dca2f654cd0 100644
--- a/tools/perf/util/string.h
+++ b/tools/perf/util/string.h
@@ -1,8 +1,8 @@
 #ifndef _PERF_STRING_H_
 #define _PERF_STRING_H_
 
-#include <linux/types.h>
+#include "types.h"
 
-int hex2u64(const char *ptr, __u64 *val);
+int hex2u64(const char *ptr, u64 *val);
 
 #endif
diff --git a/tools/perf/util/strlist.c b/tools/perf/util/strlist.c
new file mode 100644
index 000000000000..025a78edfffe
--- /dev/null
+++ b/tools/perf/util/strlist.c
@@ -0,0 +1,184 @@
+/*
+ * (c) 2009 Arnaldo Carvalho de Melo <acme@redhat.com>
+ *
+ * Licensed under the GPLv2.
+ */
+
+#include "strlist.h"
+#include <errno.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+static struct str_node *str_node__new(const char *s, bool dupstr)
+{
+	struct str_node *self = malloc(sizeof(*self));
+
+	if (self != NULL) {
+		if (dupstr) {
+			s = strdup(s);
+			if (s == NULL)
+				goto out_delete;
+		}
+		self->s = s;
+	}
+
+	return self;
+
+out_delete:
+	free(self);
+	return NULL;
+}
+
+static void str_node__delete(struct str_node *self, bool dupstr)
+{
+	if (dupstr)
+		free((void *)self->s);
+	free(self);
+}
+
+int strlist__add(struct strlist *self, const char *new_entry)
+{
+	struct rb_node **p = &self->entries.rb_node;
+	struct rb_node *parent = NULL;
+	struct str_node *sn;
+
+	while (*p != NULL) {
+		int rc;
+
+		parent = *p;
+		sn = rb_entry(parent, struct str_node, rb_node);
+		rc = strcmp(sn->s, new_entry);
+
+		if (rc > 0)
+			p = &(*p)->rb_left;
+		else if (rc < 0)
+			p = &(*p)->rb_right;
+		else
+			return -EEXIST;
+	}
+
+	sn = str_node__new(new_entry, self->dupstr);
+	if (sn == NULL)
+		return -ENOMEM;
+
+	rb_link_node(&sn->rb_node, parent, p);
+	rb_insert_color(&sn->rb_node, &self->entries);
+
+	return 0;
+}
+
+int strlist__load(struct strlist *self, const char *filename)
+{
+	char entry[1024];
+	int err;
+	FILE *fp = fopen(filename, "r");
+
+	if (fp == NULL)
+		return errno;
+
+	while (fgets(entry, sizeof(entry), fp) != NULL) {
+		const size_t len = strlen(entry);
+
+		if (len == 0)
+			continue;
+		entry[len - 1] = '\0';
+
+		err = strlist__add(self, entry);
+		if (err != 0)
+			goto out;
+	}
+
+	err = 0;
+out:
+	fclose(fp);
+	return err;
+}
+
+void strlist__remove(struct strlist *self, struct str_node *sn)
+{
+	rb_erase(&sn->rb_node, &self->entries);
+	str_node__delete(sn, self->dupstr);
+}
+
+bool strlist__has_entry(struct strlist *self, const char *entry)
+{
+	struct rb_node **p = &self->entries.rb_node;
+	struct rb_node *parent = NULL;
+
+	while (*p != NULL) {
+		struct str_node *sn;
+		int rc;
+
+		parent = *p;
+		sn = rb_entry(parent, struct str_node, rb_node);
+		rc = strcmp(sn->s, entry);
+
+		if (rc > 0)
+			p = &(*p)->rb_left;
+		else if (rc < 0)
+			p = &(*p)->rb_right;
+		else
+			return true;
+	}
+
+	return false;
+}
+
+static int strlist__parse_list_entry(struct strlist *self, const char *s)
+{
+	if (strncmp(s, "file://", 7) == 0)
+		return strlist__load(self, s + 7);
+
+	return strlist__add(self, s);
+}
+
+int strlist__parse_list(struct strlist *self, const char *s)
+{
+	char *sep;
+	int err;
+
+	while ((sep = strchr(s, ',')) != NULL) {
+		*sep = '\0';
+		err = strlist__parse_list_entry(self, s);
+		*sep = ',';
+		if (err != 0)
+			return err;
+		s = sep + 1;
+	}
+
+	return *s ? strlist__parse_list_entry(self, s) : 0;
+}
+
+struct strlist *strlist__new(bool dupstr, const char *slist)
+{
+	struct strlist *self = malloc(sizeof(*self));
+
+	if (self != NULL) {
+		self->entries = RB_ROOT;
+		self->dupstr = dupstr;
+		if (slist && strlist__parse_list(self, slist) != 0)
+			goto out_error;
+	}
+
+	return self;
+out_error:
+	free(self);
+	return NULL;
+}
+
+void strlist__delete(struct strlist *self)
+{
+	if (self != NULL) {
+		struct str_node *pos;
+		struct rb_node *next = rb_first(&self->entries);
+
+		while (next) {
+			pos = rb_entry(next, struct str_node, rb_node);
+			next = rb_next(&pos->rb_node);
+			strlist__remove(self, pos);
+		}
+		self->entries = RB_ROOT;
+		free(self);
+	}
+}
diff --git a/tools/perf/util/strlist.h b/tools/perf/util/strlist.h
new file mode 100644
index 000000000000..2fb117fb4b67
--- /dev/null
+++ b/tools/perf/util/strlist.h
@@ -0,0 +1,32 @@
+#ifndef STRLIST_H_
+#define STRLIST_H_
+
+#include "rbtree.h"
+#include <stdbool.h>
+
+struct str_node {
+	struct rb_node rb_node;
+	const char     *s;
+};
+
+struct strlist {
+	struct rb_root entries;
+	bool dupstr;
+};
+
+struct strlist *strlist__new(bool dupstr, const char *slist);
+void strlist__delete(struct strlist *self);
+
+void strlist__remove(struct strlist *self, struct str_node *sn);
+int strlist__load(struct strlist *self, const char *filename);
+int strlist__add(struct strlist *self, const char *str);
+
+bool strlist__has_entry(struct strlist *self, const char *entry);
+
+static inline bool strlist__empty(const struct strlist *self)
+{
+	return rb_first(&self->entries) == NULL;
+}
+
+int strlist__parse_list(struct strlist *self, const char *s);
+#endif /* STRLIST_H_ */
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 49a55f813712..78c2efde01b7 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -9,9 +9,9 @@
 
 const char *sym_hist_filter;
 
-static struct symbol *symbol__new(__u64 start, __u64 len,
+static struct symbol *symbol__new(u64 start, u64 len,
 				  const char *name, unsigned int priv_size,
-				  __u64 obj_start, int verbose)
+				  u64 obj_start, int verbose)
 {
 	size_t namelen = strlen(name) + 1;
 	struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen);
@@ -21,14 +21,14 @@ static struct symbol *symbol__new(__u64 start, __u64 len,
 
 	if (verbose >= 2)
 		printf("new symbol: %016Lx [%08lx]: %s, hist: %p, obj_start: %p\n",
-			(__u64)start, (unsigned long)len, name, self->hist, (void *)(unsigned long)obj_start);
+			(u64)start, (unsigned long)len, name, self->hist, (void *)(unsigned long)obj_start);
 
 	self->obj_start= obj_start;
 	self->hist = NULL;
 	self->hist_sum = 0;
 
 	if (sym_hist_filter && !strcmp(name, sym_hist_filter))
-		self->hist = calloc(sizeof(__u64), len);
+		self->hist = calloc(sizeof(u64), len);
 
 	if (priv_size) {
 		memset(self, 0, priv_size);
@@ -89,7 +89,7 @@ static void dso__insert_symbol(struct dso *self, struct symbol *sym)
 {
 	struct rb_node **p = &self->syms.rb_node;
 	struct rb_node *parent = NULL;
-	const __u64 ip = sym->start;
+	const u64 ip = sym->start;
 	struct symbol *s;
 
 	while (*p != NULL) {
@@ -104,7 +104,7 @@ static void dso__insert_symbol(struct dso *self, struct symbol *sym)
 	rb_insert_color(&sym->rb_node, &self->syms);
 }
 
-struct symbol *dso__find_symbol(struct dso *self, __u64 ip)
+struct symbol *dso__find_symbol(struct dso *self, u64 ip)
 {
 	struct rb_node *n;
 
@@ -151,7 +151,7 @@ static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int verb
 		goto out_failure;
 
 	while (!feof(file)) {
-		__u64 start;
+		u64 start;
 		struct symbol *sym;
 		int line_len, len;
 		char symbol_type;
@@ -232,7 +232,7 @@ static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int verb
 		goto out_failure;
 
 	while (!feof(file)) {
-		__u64 start, size;
+		u64 start, size;
 		struct symbol *sym;
 		int line_len, len;
 
@@ -353,7 +353,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, Elf *elf,
 {
 	uint32_t nr_rel_entries, idx;
 	GElf_Sym sym;
-	__u64 plt_offset;
+	u64 plt_offset;
 	GElf_Shdr shdr_plt;
 	struct symbol *f;
 	GElf_Shdr shdr_rel_plt;
@@ -520,10 +520,12 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 	nr_syms = shdr.sh_size / shdr.sh_entsize;
 
 	memset(&sym, 0, sizeof(sym));
-
+	self->prelinked = elf_section_by_name(elf, &ehdr, &shdr,
+					      ".gnu.prelink_undo",
+					      NULL) != NULL;
 	elf_symtab__for_each_symbol(syms, nr_syms, index, sym) {
 		struct symbol *f;
-		__u64 obj_start;
+		u64 obj_start;
 
 		if (!elf_sym__is_function(&sym))
 			continue;
@@ -535,7 +537,13 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 		gelf_getshdr(sec, &shdr);
 		obj_start = sym.st_value;
 
-		sym.st_value -= shdr.sh_addr - shdr.sh_offset;
+		if (self->prelinked) {
+			if (verbose >= 2)
+				printf("adjusting symbol: st_value: %Lx sh_addr: %Lx sh_offset: %Lx\n",
+					(u64)sym.st_value, (u64)shdr.sh_addr, (u64)shdr.sh_offset);
+
+			sym.st_value -= shdr.sh_addr - shdr.sh_offset;
+		}
 
 		f = symbol__new(sym.st_value, sym.st_size,
 				elf_sym__name(&sym, symstrs),
@@ -569,6 +577,8 @@ int dso__load(struct dso *self, symbol_filter_t filter, int verbose)
 	if (!name)
 		return -1;
 
+	self->prelinked = 0;
+
 	if (strncmp(self->name, "/tmp/perf-", 10) == 0)
 		return dso__load_perf_map(self, filter, verbose);
 
@@ -629,7 +639,7 @@ int dso__load_kernel(struct dso *self, const char *vmlinux,
 	if (vmlinux)
 		err = dso__load_vmlinux(self, vmlinux, filter, verbose);
 
-	if (err)
+	if (err < 0)
 		err = dso__load_kallsyms(self, filter, verbose);
 
 	return err;
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 0d1292bd8270..2c48ace8203b 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -2,24 +2,27 @@
 #define _PERF_SYMBOL_ 1
 
 #include <linux/types.h>
+#include "types.h"
 #include "list.h"
 #include "rbtree.h"
 
 struct symbol {
 	struct rb_node	rb_node;
-	__u64		start;
-	__u64		end;
-	__u64		obj_start;
-	__u64		hist_sum;
-	__u64		*hist;
+	u64		start;
+	u64		end;
+	u64		obj_start;
+	u64		hist_sum;
+	u64		*hist;
+	void		*priv;
 	char		name[0];
 };
 
 struct dso {
 	struct list_head node;
 	struct rb_root	 syms;
+	struct symbol    *(*find_symbol)(struct dso *, u64 ip);
 	unsigned int	 sym_priv_size;
-	struct symbol    *(*find_symbol)(struct dso *, __u64 ip);
+	unsigned char	 prelinked;
 	char		 name[0];
 };
 
@@ -35,7 +38,7 @@ static inline void *dso__sym_priv(struct dso *self, struct symbol *sym)
 	return ((void *)sym) - self->sym_priv_size;
 }
 
-struct symbol *dso__find_symbol(struct dso *self, __u64 ip);
+struct symbol *dso__find_symbol(struct dso *self, u64 ip);
 
 int dso__load_kernel(struct dso *self, const char *vmlinux,
 		     symbol_filter_t filter, int verbose);
diff --git a/tools/perf/util/types.h b/tools/perf/util/types.h
new file mode 100644
index 000000000000..5e75f9005940
--- /dev/null
+++ b/tools/perf/util/types.h
@@ -0,0 +1,17 @@
+#ifndef _PERF_TYPES_H
+#define _PERF_TYPES_H
+
+/*
+ * We define u64 as unsigned long long for every architecture
+ * so that we can print it with %Lx without getting warnings.
+ */
+typedef unsigned long long u64;
+typedef signed long long   s64;
+typedef unsigned int	   u32;
+typedef signed int	   s32;
+typedef unsigned short	   u16;
+typedef signed short	   s16;
+typedef unsigned char	   u8;
+typedef signed char	   s8;
+
+#endif /* _PERF_TYPES_H */
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 76590a16c271..b4be6071c105 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -67,7 +67,6 @@
 #include <assert.h>
 #include <regex.h>
 #include <utime.h>
-#ifndef __MINGW32__
 #include <sys/wait.h>
 #include <sys/poll.h>
 #include <sys/socket.h>
@@ -81,30 +80,11 @@
 #include <netdb.h>
 #include <pwd.h>
 #include <inttypes.h>
-#if defined(__CYGWIN__)
-#undef _XOPEN_SOURCE
-#include <grp.h>
-#define _XOPEN_SOURCE 600
-#include "compat/cygwin.h"
-#else
-#undef _ALL_SOURCE /* AIX 5.3L defines a struct list with _ALL_SOURCE. */
-#include <grp.h>
-#define _ALL_SOURCE 1
-#endif
-#else 	/* __MINGW32__ */
-/* pull in Windows compatibility stuff */
-#include "compat/mingw.h"
-#endif	/* __MINGW32__ */
 
 #ifndef NO_ICONV
 #include <iconv.h>
 #endif
 
-#ifndef NO_OPENSSL
-#include <openssl/ssl.h>
-#include <openssl/err.h>
-#endif
-
 /* On most systems <limits.h> would have given us this, but
  * not on some systems (e.g. GNU/Hurd).
  */
@@ -332,17 +312,20 @@ static inline int has_extension(const char *filename, const char *ext)
 #undef tolower
 #undef toupper
 extern unsigned char sane_ctype[256];
-#define GIT_SPACE 0x01
-#define GIT_DIGIT 0x02
-#define GIT_ALPHA 0x04
-#define GIT_GLOB_SPECIAL 0x08
-#define GIT_REGEX_SPECIAL 0x10
+#define GIT_SPACE		0x01
+#define GIT_DIGIT		0x02
+#define GIT_ALPHA		0x04
+#define GIT_GLOB_SPECIAL	0x08
+#define GIT_REGEX_SPECIAL	0x10
+#define GIT_PRINT_EXTRA		0x20
+#define GIT_PRINT		0x3E
 #define sane_istest(x,mask) ((sane_ctype[(unsigned char)(x)] & (mask)) != 0)
 #define isascii(x) (((x) & ~0x7f) == 0)
 #define isspace(x) sane_istest(x,GIT_SPACE)
 #define isdigit(x) sane_istest(x,GIT_DIGIT)
 #define isalpha(x) sane_istest(x,GIT_ALPHA)
 #define isalnum(x) sane_istest(x,GIT_ALPHA | GIT_DIGIT)
+#define isprint(x) sane_istest(x,GIT_PRINT)
 #define is_glob_special(x) sane_istest(x,GIT_GLOB_SPECIAL)
 #define is_regex_special(x) sane_istest(x,GIT_GLOB_SPECIAL | GIT_REGEX_SPECIAL)
 #define tolower(x) sane_case((unsigned char)(x), 0x20)