Merge branch 'master' into for-linus

69 files changed, 6330 insertions, 1442 deletions

diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN
index c561d1538c03..54552a00a117 100755
--- a/tools/perf/util/PERF-VERSION-GEN
+++ b/tools/perf/util/PERF-VERSION-GEN
@@ -1,7 +1,7 @@
 #!/bin/sh
 
 GVF=PERF-VERSION-FILE
-DEF_VER=v0.0.1.PERF
+DEF_VER=v0.0.2.PERF
 
 LF='
 '
diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h
index 6f8ea9d210b6..918eb376abe3 100644
--- a/tools/perf/util/cache.h
+++ b/tools/perf/util/cache.h
@@ -1,10 +1,15 @@
-#ifndef CACHE_H
-#define CACHE_H
+#ifndef __PERF_CACHE_H
+#define __PERF_CACHE_H
 
 #include "util.h"
 #include "strbuf.h"
 #include "../perf.h"
 
+#define CMD_EXEC_PATH "--exec-path"
+#define CMD_PERF_DIR "--perf-dir="
+#define CMD_WORK_TREE "--work-tree="
+#define CMD_DEBUGFS_DIR "--debugfs-dir="
+
 #define PERF_DIR_ENVIRONMENT "PERF_DIR"
 #define PERF_WORK_TREE_ENVIRONMENT "PERF_WORK_TREE"
 #define DEFAULT_PERF_DIR_ENVIRONMENT ".perf"
@@ -117,4 +122,4 @@ extern char *perf_pathdup(const char *fmt, ...)
 
 extern size_t strlcpy(char *dest, const char *src, size_t size);
 
-#endif /* CACHE_H */
+#endif /* __PERF_CACHE_H */
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index 3b8380f1b478..b3b71258272a 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -206,7 +206,7 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain,
 	}
 	node->val_nr = chain->nr - start;
 	if (!node->val_nr)
-		printf("Warning: empty node in callchain tree\n");
+		pr_warning("Warning: empty node in callchain tree\n");
 }
 
 static void
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
index 43cf3ea9e088..ad4626de4c2b 100644
--- a/tools/perf/util/callchain.h
+++ b/tools/perf/util/callchain.h
@@ -58,4 +58,4 @@ static inline u64 cumul_hits(struct callchain_node *node)
 int register_callchain_param(struct callchain_param *param);
 void append_chain(struct callchain_node *root, struct ip_callchain *chain,
 		  struct symbol **syms);
-#endif
+#endif	/* __PERF_CALLCHAIN_H */
diff --git a/tools/perf/util/color.h b/tools/perf/util/color.h
index 58d597564b99..24e8809210bb 100644
--- a/tools/perf/util/color.h
+++ b/tools/perf/util/color.h
@@ -1,5 +1,5 @@
-#ifndef COLOR_H
-#define COLOR_H
+#ifndef __PERF_COLOR_H
+#define __PERF_COLOR_H
 
 /* "\033[1;38;5;2xx;48;5;2xxm\0" is 23 bytes */
 #define COLOR_MAXLEN 24
@@ -39,4 +39,4 @@ int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *bu
 int percent_color_fprintf(FILE *fp, const char *fmt, double percent);
 const char *get_percent_color(double percent);
 
-#endif /* COLOR_H */
+#endif /* __PERF_COLOR_H */
diff --git a/tools/perf/util/ctype.c b/tools/perf/util/ctype.c
index 0b791bd346bc..35073621e5de 100644
--- a/tools/perf/util/ctype.c
+++ b/tools/perf/util/ctype.c
@@ -29,3 +29,11 @@ unsigned char sane_ctype[256] = {
 	A, A, A, A, A, A, A, A, A, A, A, R, R, P, P, 0,		/* 112..127 */
 	/* Nothing in the 128.. range */
 };
+
+const char *graph_line =
+	"_____________________________________________________________________"
+	"_____________________________________________________________________";
+const char *graph_dotted_line =
+	"---------------------------------------------------------------------"
+	"---------------------------------------------------------------------"
+	"---------------------------------------------------------------------";
diff --git a/tools/perf/util/data_map.c b/tools/perf/util/data_map.c
new file mode 100644
index 000000000000..ca0bedf637c2
--- /dev/null
+++ b/tools/perf/util/data_map.c
@@ -0,0 +1,291 @@
+#include "data_map.h"
+#include "symbol.h"
+#include "util.h"
+#include "debug.h"
+
+
+static struct perf_file_handler *curr_handler;
+static unsigned long	mmap_window = 32;
+static char		__cwd[PATH_MAX];
+
+static int process_event_stub(event_t *event __used)
+{
+	dump_printf(": unhandled!\n");
+	return 0;
+}
+
+void register_perf_file_handler(struct perf_file_handler *handler)
+{
+	if (!handler->process_sample_event)
+		handler->process_sample_event = process_event_stub;
+	if (!handler->process_mmap_event)
+		handler->process_mmap_event = process_event_stub;
+	if (!handler->process_comm_event)
+		handler->process_comm_event = process_event_stub;
+	if (!handler->process_fork_event)
+		handler->process_fork_event = process_event_stub;
+	if (!handler->process_exit_event)
+		handler->process_exit_event = process_event_stub;
+	if (!handler->process_lost_event)
+		handler->process_lost_event = process_event_stub;
+	if (!handler->process_read_event)
+		handler->process_read_event = process_event_stub;
+	if (!handler->process_throttle_event)
+		handler->process_throttle_event = process_event_stub;
+	if (!handler->process_unthrottle_event)
+		handler->process_unthrottle_event = process_event_stub;
+
+	curr_handler = handler;
+}
+
+static const char *event__name[] = {
+	[0]			 = "TOTAL",
+	[PERF_RECORD_MMAP]	 = "MMAP",
+	[PERF_RECORD_LOST]	 = "LOST",
+	[PERF_RECORD_COMM]	 = "COMM",
+	[PERF_RECORD_EXIT]	 = "EXIT",
+	[PERF_RECORD_THROTTLE]	 = "THROTTLE",
+	[PERF_RECORD_UNTHROTTLE] = "UNTHROTTLE",
+	[PERF_RECORD_FORK]	 = "FORK",
+	[PERF_RECORD_READ]	 = "READ",
+	[PERF_RECORD_SAMPLE]	 = "SAMPLE",
+};
+
+unsigned long event__total[PERF_RECORD_MAX];
+
+void event__print_totals(void)
+{
+	int i;
+	for (i = 0; i < PERF_RECORD_MAX; ++i)
+		pr_info("%10s events: %10ld\n",
+			event__name[i], event__total[i]);
+}
+
+static int
+process_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	trace_event(event);
+
+	if (event->header.type < PERF_RECORD_MAX) {
+		dump_printf("%p [%p]: PERF_RECORD_%s",
+			    (void *)(offset + head),
+			    (void *)(long)(event->header.size),
+			    event__name[event->header.type]);
+		++event__total[0];
+		++event__total[event->header.type];
+	}
+
+	switch (event->header.type) {
+	case PERF_RECORD_SAMPLE:
+		return curr_handler->process_sample_event(event);
+	case PERF_RECORD_MMAP:
+		return curr_handler->process_mmap_event(event);
+	case PERF_RECORD_COMM:
+		return curr_handler->process_comm_event(event);
+	case PERF_RECORD_FORK:
+		return curr_handler->process_fork_event(event);
+	case PERF_RECORD_EXIT:
+		return curr_handler->process_exit_event(event);
+	case PERF_RECORD_LOST:
+		return curr_handler->process_lost_event(event);
+	case PERF_RECORD_READ:
+		return curr_handler->process_read_event(event);
+	case PERF_RECORD_THROTTLE:
+		return curr_handler->process_throttle_event(event);
+	case PERF_RECORD_UNTHROTTLE:
+		return curr_handler->process_unthrottle_event(event);
+	default:
+		curr_handler->total_unknown++;
+		return -1;
+	}
+}
+
+int perf_header__read_build_ids(int input, off_t offset, off_t size)
+{
+	struct build_id_event bev;
+	char filename[PATH_MAX];
+	off_t limit = offset + size;
+	int err = -1;
+
+	while (offset < limit) {
+		struct dso *dso;
+		ssize_t len;
+
+		if (read(input, &bev, sizeof(bev)) != sizeof(bev))
+			goto out;
+
+		len = bev.header.size - sizeof(bev);
+		if (read(input, filename, len) != len)
+			goto out;
+
+		dso = dsos__findnew(filename);
+		if (dso != NULL)
+			dso__set_build_id(dso, &bev.build_id);
+
+		offset += bev.header.size;
+	}
+	err = 0;
+out:
+	return err;
+}
+
+int mmap_dispatch_perf_file(struct perf_header **pheader,
+			    const char *input_name,
+			    int force,
+			    int full_paths,
+			    int *cwdlen,
+			    char **cwd)
+{
+	int err;
+	struct perf_header *header;
+	unsigned long head, shift;
+	unsigned long offset = 0;
+	struct stat input_stat;
+	size_t	page_size;
+	u64 sample_type;
+	event_t *event;
+	uint32_t size;
+	int input;
+	char *buf;
+
+	if (curr_handler == NULL) {
+		pr_debug("Forgot to register perf file handler\n");
+		return -EINVAL;
+	}
+
+	page_size = getpagesize();
+
+	input = open(input_name, O_RDONLY);
+	if (input < 0) {
+		pr_err("Failed to open file: %s", input_name);
+		if (!strcmp(input_name, "perf.data"))
+			pr_err("  (try 'perf record' first)");
+		pr_err("\n");
+		return -errno;
+	}
+
+	if (fstat(input, &input_stat) < 0) {
+		pr_err("failed to stat file");
+		err = -errno;
+		goto out_close;
+	}
+
+	err = -EACCES;
+	if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
+		pr_err("file: %s not owned by current user or root\n",
+			input_name);
+		goto out_close;
+	}
+
+	if (input_stat.st_size == 0) {
+		pr_info("zero-sized file, nothing to do!\n");
+		goto done;
+	}
+
+	err = -ENOMEM;
+	header = perf_header__new();
+	if (header == NULL)
+		goto out_close;
+
+	err = perf_header__read(header, input);
+	if (err < 0)
+		goto out_delete;
+	*pheader = header;
+	head = header->data_offset;
+
+	sample_type = perf_header__sample_type(header);
+
+	err = -EINVAL;
+	if (curr_handler->sample_type_check &&
+	    curr_handler->sample_type_check(sample_type) < 0)
+		goto out_delete;
+
+	if (!full_paths) {
+		if (getcwd(__cwd, sizeof(__cwd)) == NULL) {
+			pr_err("failed to get the current directory\n");
+			err = -errno;
+			goto out_delete;
+		}
+		*cwd = __cwd;
+		*cwdlen = strlen(*cwd);
+	} else {
+		*cwd = NULL;
+		*cwdlen = 0;
+	}
+
+	shift = page_size * (head / page_size);
+	offset += shift;
+	head -= shift;
+
+remap:
+	buf = mmap(NULL, page_size * mmap_window, PROT_READ,
+		   MAP_SHARED, input, offset);
+	if (buf == MAP_FAILED) {
+		pr_err("failed to mmap file\n");
+		err = -errno;
+		goto out_delete;
+	}
+
+more:
+	event = (event_t *)(buf + head);
+
+	size = event->header.size;
+	if (!size)
+		size = 8;
+
+	if (head + event->header.size >= page_size * mmap_window) {
+		int munmap_ret;
+
+		shift = page_size * (head / page_size);
+
+		munmap_ret = munmap(buf, page_size * mmap_window);
+		assert(munmap_ret == 0);
+
+		offset += shift;
+		head -= shift;
+		goto remap;
+	}
+
+	size = event->header.size;
+
+	dump_printf("\n%p [%p]: event: %d\n",
+			(void *)(offset + head),
+			(void *)(long)event->header.size,
+			event->header.type);
+
+	if (!size || process_event(event, offset, head) < 0) {
+
+		dump_printf("%p [%p]: skipping unknown header type: %d\n",
+			(void *)(offset + head),
+			(void *)(long)(event->header.size),
+			event->header.type);
+
+		/*
+		 * assume we lost track of the stream, check alignment, and
+		 * increment a single u64 in the hope to catch on again 'soon'.
+		 */
+
+		if (unlikely(head & 7))
+			head &= ~7ULL;
+
+		size = 8;
+	}
+
+	head += size;
+
+	if (offset + head >= header->data_offset + header->data_size)
+		goto done;
+
+	if (offset + head < (unsigned long)input_stat.st_size)
+		goto more;
+
+done:
+	err = 0;
+out_close:
+	close(input);
+
+	return err;
+out_delete:
+	perf_header__delete(header);
+	goto out_close;
+}
diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h
new file mode 100644
index 000000000000..3180ff7e3633
--- /dev/null
+++ b/tools/perf/util/data_map.h
@@ -0,0 +1,32 @@
+#ifndef __PERF_DATAMAP_H
+#define __PERF_DATAMAP_H
+
+#include "event.h"
+#include "header.h"
+
+typedef int (*event_type_handler_t)(event_t *);
+
+struct perf_file_handler {
+	event_type_handler_t	process_sample_event;
+	event_type_handler_t	process_mmap_event;
+	event_type_handler_t	process_comm_event;
+	event_type_handler_t	process_fork_event;
+	event_type_handler_t	process_exit_event;
+	event_type_handler_t	process_lost_event;
+	event_type_handler_t	process_read_event;
+	event_type_handler_t	process_throttle_event;
+	event_type_handler_t	process_unthrottle_event;
+	int			(*sample_type_check)(u64 sample_type);
+	unsigned long		total_unknown;
+};
+
+void register_perf_file_handler(struct perf_file_handler *handler);
+int mmap_dispatch_perf_file(struct perf_header **pheader,
+			    const char *input_name,
+			    int force,
+			    int full_paths,
+			    int *cwdlen,
+			    char **cwd);
+int perf_header__read_build_ids(int input, off_t offset, off_t file_size);
+
+#endif
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index e8ca98fe0bd4..28d520d5a1fb 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -13,12 +13,12 @@
 int verbose = 0;
 int dump_trace = 0;
 
-int eprintf(const char *fmt, ...)
+int eprintf(int level, const char *fmt, ...)
 {
 	va_list args;
 	int ret = 0;
 
-	if (verbose) {
+	if (verbose >= level) {
 		va_start(args, fmt);
 		ret = vfprintf(stderr, fmt, args);
 		va_end(args);
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index 437eea58ce40..c6c24c522dea 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -1,8 +1,15 @@
 /* For debugging general purposes */
+#ifndef __PERF_DEBUG_H
+#define __PERF_DEBUG_H
+
+#include "event.h"
 
 extern int verbose;
 extern int dump_trace;
 
-int eprintf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+int eprintf(int level,
+	    const char *fmt, ...) __attribute__((format(printf, 2, 3)));
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
 void trace_event(event_t *event);
+
+#endif	/* __PERF_DEBUG_H */
diff --git a/tools/perf/util/debugfs.c b/tools/perf/util/debugfs.c
new file mode 100644
index 000000000000..06b73ee02c49
--- /dev/null
+++ b/tools/perf/util/debugfs.c
@@ -0,0 +1,241 @@
+#include "util.h"
+#include "debugfs.h"
+#include "cache.h"
+
+static int debugfs_premounted;
+static char debugfs_mountpoint[MAX_PATH+1];
+
+static const char *debugfs_known_mountpoints[] = {
+	"/sys/kernel/debug/",
+	"/debug/",
+	0,
+};
+
+/* use this to force a umount */
+void debugfs_force_cleanup(void)
+{
+	debugfs_find_mountpoint();
+	debugfs_premounted = 0;
+	debugfs_umount();
+}
+
+/* construct a full path to a debugfs element */
+int debugfs_make_path(const char *element, char *buffer, int size)
+{
+	int len;
+
+	if (strlen(debugfs_mountpoint) == 0) {
+		buffer[0] = '\0';
+		return -1;
+	}
+
+	len = strlen(debugfs_mountpoint) + strlen(element) + 1;
+	if (len >= size)
+		return len+1;
+
+	snprintf(buffer, size-1, "%s/%s", debugfs_mountpoint, element);
+	return 0;
+}
+
+static int debugfs_found;
+
+/* find the path to the mounted debugfs */
+const char *debugfs_find_mountpoint(void)
+{
+	const char **ptr;
+	char type[100];
+	FILE *fp;
+
+	if (debugfs_found)
+		return (const char *) debugfs_mountpoint;
+
+	ptr = debugfs_known_mountpoints;
+	while (*ptr) {
+		if (debugfs_valid_mountpoint(*ptr) == 0) {
+			debugfs_found = 1;
+			strcpy(debugfs_mountpoint, *ptr);
+			return debugfs_mountpoint;
+		}
+		ptr++;
+	}
+
+	/* give up and parse /proc/mounts */
+	fp = fopen("/proc/mounts", "r");
+	if (fp == NULL)
+		die("Can't open /proc/mounts for read");
+
+	while (fscanf(fp, "%*s %"
+		      STR(MAX_PATH)
+		      "s %99s %*s %*d %*d\n",
+		      debugfs_mountpoint, type) == 2) {
+		if (strcmp(type, "debugfs") == 0)
+			break;
+	}
+	fclose(fp);
+
+	if (strcmp(type, "debugfs") != 0)
+		return NULL;
+
+	debugfs_found = 1;
+
+	return debugfs_mountpoint;
+}
+
+/* verify that a mountpoint is actually a debugfs instance */
+
+int debugfs_valid_mountpoint(const char *debugfs)
+{
+	struct statfs st_fs;
+
+	if (statfs(debugfs, &st_fs) < 0)
+		return -ENOENT;
+	else if (st_fs.f_type != (long) DEBUGFS_MAGIC)
+		return -ENOENT;
+
+	return 0;
+}
+
+
+int debugfs_valid_entry(const char *path)
+{
+	struct stat st;
+
+	if (stat(path, &st))
+		return -errno;
+
+	return 0;
+}
+
+/* mount the debugfs somewhere */
+
+int debugfs_mount(const char *mountpoint)
+{
+	char mountcmd[128];
+
+	/* see if it's already mounted */
+	if (debugfs_find_mountpoint()) {
+		debugfs_premounted = 1;
+		return 0;
+	}
+
+	/* if not mounted and no argument */
+	if (mountpoint == NULL) {
+		/* see if environment variable set */
+		mountpoint = getenv(PERF_DEBUGFS_ENVIRONMENT);
+		/* if no environment variable, use default */
+		if (mountpoint == NULL)
+			mountpoint = "/sys/kernel/debug";
+	}
+
+	/* save the mountpoint */
+	strncpy(debugfs_mountpoint, mountpoint, sizeof(debugfs_mountpoint));
+
+	/* mount it */
+	snprintf(mountcmd, sizeof(mountcmd),
+		 "/bin/mount -t debugfs debugfs %s", mountpoint);
+	return system(mountcmd);
+}
+
+/* umount the debugfs */
+
+int debugfs_umount(void)
+{
+	char umountcmd[128];
+	int ret;
+
+	/* if it was already mounted, leave it */
+	if (debugfs_premounted)
+		return 0;
+
+	/* make sure it's a valid mount point */
+	ret = debugfs_valid_mountpoint(debugfs_mountpoint);
+	if (ret)
+		return ret;
+
+	snprintf(umountcmd, sizeof(umountcmd),
+		 "/bin/umount %s", debugfs_mountpoint);
+	return system(umountcmd);
+}
+
+int debugfs_write(const char *entry, const char *value)
+{
+	char path[MAX_PATH+1];
+	int ret, count;
+	int fd;
+
+	/* construct the path */
+	snprintf(path, sizeof(path), "%s/%s", debugfs_mountpoint, entry);
+
+	/* verify that it exists */
+	ret = debugfs_valid_entry(path);
+	if (ret)
+		return ret;
+
+	/* get how many chars we're going to write */
+	count = strlen(value);
+
+	/* open the debugfs entry */
+	fd = open(path, O_RDWR);
+	if (fd < 0)
+		return -errno;
+
+	while (count > 0) {
+		/* write it */
+		ret = write(fd, value, count);
+		if (ret <= 0) {
+			if (ret == EAGAIN)
+				continue;
+			close(fd);
+			return -errno;
+		}
+		count -= ret;
+	}
+
+	/* close it */
+	close(fd);
+
+	/* return success */
+	return 0;
+}
+
+/*
+ * read a debugfs entry
+ * returns the number of chars read or a negative errno
+ */
+int debugfs_read(const char *entry, char *buffer, size_t size)
+{
+	char path[MAX_PATH+1];
+	int ret;
+	int fd;
+
+	/* construct the path */
+	snprintf(path, sizeof(path), "%s/%s", debugfs_mountpoint, entry);
+
+	/* verify that it exists */
+	ret = debugfs_valid_entry(path);
+	if (ret)
+		return ret;
+
+	/* open the debugfs entry */
+	fd = open(path, O_RDONLY);
+	if (fd < 0)
+		return -errno;
+
+	do {
+		/* read it */
+		ret = read(fd, buffer, size);
+		if (ret == 0) {
+			close(fd);
+			return EOF;
+		}
+	} while (ret < 0 && errno == EAGAIN);
+
+	/* close it */
+	close(fd);
+
+	/* make *sure* there's a null character at the end */
+	buffer[ret] = '\0';
+
+	/* return the number of chars read */
+	return ret;
+}
diff --git a/tools/perf/util/debugfs.h b/tools/perf/util/debugfs.h
new file mode 100644
index 000000000000..3cd14f9ae784
--- /dev/null
+++ b/tools/perf/util/debugfs.h
@@ -0,0 +1,25 @@
+#ifndef __DEBUGFS_H__
+#define __DEBUGFS_H__
+
+#include <sys/mount.h>
+
+#ifndef MAX_PATH
+# define MAX_PATH 256
+#endif
+
+#ifndef STR
+# define _STR(x) #x
+# define STR(x) _STR(x)
+#endif
+
+extern const char *debugfs_find_mountpoint(void);
+extern int debugfs_valid_mountpoint(const char *debugfs);
+extern int debugfs_valid_entry(const char *path);
+extern int debugfs_mount(const char *mountpoint);
+extern int debugfs_umount(void);
+extern int debugfs_write(const char *entry, const char *value);
+extern int debugfs_read(const char *entry, char *buffer, size_t size);
+extern void debugfs_force_cleanup(void);
+extern int debugfs_make_path(const char *element, char *buffer, int size);
+
+#endif /* __DEBUGFS_H__ */
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
new file mode 100644
index 000000000000..414b89d1bde9
--- /dev/null
+++ b/tools/perf/util/event.c
@@ -0,0 +1,312 @@
+#include <linux/types.h>
+#include "event.h"
+#include "debug.h"
+#include "string.h"
+#include "thread.h"
+
+static pid_t event__synthesize_comm(pid_t pid, int full,
+				    int (*process)(event_t *event))
+{
+	event_t ev;
+	char filename[PATH_MAX];
+	char bf[BUFSIZ];
+	FILE *fp;
+	size_t size = 0;
+	DIR *tasks;
+	struct dirent dirent, *next;
+	pid_t tgid = 0;
+
+	snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
+
+	fp = fopen(filename, "r");
+	if (fp == NULL) {
+out_race:
+		/*
+		 * We raced with a task exiting - just return:
+		 */
+		pr_debug("couldn't open %s\n", filename);
+		return 0;
+	}
+
+	memset(&ev.comm, 0, sizeof(ev.comm));
+	while (!ev.comm.comm[0] || !ev.comm.pid) {
+		if (fgets(bf, sizeof(bf), fp) == NULL)
+			goto out_failure;
+
+		if (memcmp(bf, "Name:", 5) == 0) {
+			char *name = bf + 5;
+			while (*name && isspace(*name))
+				++name;
+			size = strlen(name) - 1;
+			memcpy(ev.comm.comm, name, size++);
+		} else if (memcmp(bf, "Tgid:", 5) == 0) {
+			char *tgids = bf + 5;
+			while (*tgids && isspace(*tgids))
+				++tgids;
+			tgid = ev.comm.pid = atoi(tgids);
+		}
+	}
+
+	ev.comm.header.type = PERF_RECORD_COMM;
+	size = ALIGN(size, sizeof(u64));
+	ev.comm.header.size = sizeof(ev.comm) - (sizeof(ev.comm.comm) - size);
+
+	if (!full) {
+		ev.comm.tid = pid;
+
+		process(&ev);
+		goto out_fclose;
+	}
+
+	snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
+
+	tasks = opendir(filename);
+	if (tasks == NULL)
+		goto out_race;
+
+	while (!readdir_r(tasks, &dirent, &next) && next) {
+		char *end;
+		pid = strtol(dirent.d_name, &end, 10);
+		if (*end)
+			continue;
+
+		ev.comm.tid = pid;
+
+		process(&ev);
+	}
+	closedir(tasks);
+
+out_fclose:
+	fclose(fp);
+	return tgid;
+
+out_failure:
+	pr_warning("couldn't get COMM and pgid, malformed %s\n", filename);
+	return -1;
+}
+
+static int event__synthesize_mmap_events(pid_t pid, pid_t tgid,
+					 int (*process)(event_t *event))
+{
+	char filename[PATH_MAX];
+	FILE *fp;
+
+	snprintf(filename, sizeof(filename), "/proc/%d/maps", pid);
+
+	fp = fopen(filename, "r");
+	if (fp == NULL) {
+		/*
+		 * We raced with a task exiting - just return:
+		 */
+		pr_debug("couldn't open %s\n", filename);
+		return -1;
+	}
+
+	while (1) {
+		char bf[BUFSIZ], *pbf = bf;
+		event_t ev = {
+			.header = { .type = PERF_RECORD_MMAP },
+		};
+		int n;
+		size_t size;
+		if (fgets(bf, sizeof(bf), fp) == NULL)
+			break;
+
+		/* 00400000-0040c000 r-xp 00000000 fd:01 41038  /bin/cat */
+		n = hex2u64(pbf, &ev.mmap.start);
+		if (n < 0)
+			continue;
+		pbf += n + 1;
+		n = hex2u64(pbf, &ev.mmap.len);
+		if (n < 0)
+			continue;
+		pbf += n + 3;
+		if (*pbf == 'x') { /* vm_exec */
+			char *execname = strchr(bf, '/');
+
+			/* Catch VDSO */
+			if (execname == NULL)
+				execname = strstr(bf, "[vdso]");
+
+			if (execname == NULL)
+				continue;
+
+			size = strlen(execname);
+			execname[size - 1] = '\0'; /* Remove \n */
+			memcpy(ev.mmap.filename, execname, size);
+			size = ALIGN(size, sizeof(u64));
+			ev.mmap.len -= ev.mmap.start;
+			ev.mmap.header.size = (sizeof(ev.mmap) -
+					       (sizeof(ev.mmap.filename) - size));
+			ev.mmap.pid = tgid;
+			ev.mmap.tid = pid;
+
+			process(&ev);
+		}
+	}
+
+	fclose(fp);
+	return 0;
+}
+
+int event__synthesize_thread(pid_t pid, int (*process)(event_t *event))
+{
+	pid_t tgid = event__synthesize_comm(pid, 1, process);
+	if (tgid == -1)
+		return -1;
+	return event__synthesize_mmap_events(pid, tgid, process);
+}
+
+void event__synthesize_threads(int (*process)(event_t *event))
+{
+	DIR *proc;
+	struct dirent dirent, *next;
+
+	proc = opendir("/proc");
+
+	while (!readdir_r(proc, &dirent, &next) && next) {
+		char *end;
+		pid_t pid = strtol(dirent.d_name, &end, 10);
+
+		if (*end) /* only interested in proper numerical dirents */
+			continue;
+
+		event__synthesize_thread(pid, process);
+	}
+
+	closedir(proc);
+}
+
+char *event__cwd;
+int  event__cwdlen;
+
+struct events_stats event__stats;
+
+int event__process_comm(event_t *self)
+{
+	struct thread *thread = threads__findnew(self->comm.pid);
+
+	dump_printf(": %s:%d\n", self->comm.comm, self->comm.pid);
+
+	if (thread == NULL || thread__set_comm(thread, self->comm.comm)) {
+		dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
+		return -1;
+	}
+
+	return 0;
+}
+
+int event__process_lost(event_t *self)
+{
+	dump_printf(": id:%Ld: lost:%Ld\n", self->lost.id, self->lost.lost);
+	event__stats.lost += self->lost.lost;
+	return 0;
+}
+
+int event__process_mmap(event_t *self)
+{
+	struct thread *thread = threads__findnew(self->mmap.pid);
+	struct map *map = map__new(&self->mmap, MAP__FUNCTION,
+				   event__cwd, event__cwdlen);
+
+	dump_printf(" %d/%d: [%p(%p) @ %p]: %s\n",
+		    self->mmap.pid, self->mmap.tid,
+		    (void *)(long)self->mmap.start,
+		    (void *)(long)self->mmap.len,
+		    (void *)(long)self->mmap.pgoff,
+		    self->mmap.filename);
+
+	if (thread == NULL || map == NULL)
+		dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n");
+	else
+		thread__insert_map(thread, map);
+
+	return 0;
+}
+
+int event__process_task(event_t *self)
+{
+	struct thread *thread = threads__findnew(self->fork.pid);
+	struct thread *parent = threads__findnew(self->fork.ppid);
+
+	dump_printf("(%d:%d):(%d:%d)\n", self->fork.pid, self->fork.tid,
+		    self->fork.ppid, self->fork.ptid);
+	/*
+	 * A thread clone will have the same PID for both parent and child.
+	 */
+	if (thread == parent)
+		return 0;
+
+	if (self->header.type == PERF_RECORD_EXIT)
+		return 0;
+
+	if (thread == NULL || parent == NULL ||
+	    thread__fork(thread, parent) < 0) {
+		dump_printf("problem processing PERF_RECORD_FORK, skipping event.\n");
+		return -1;
+	}
+
+	return 0;
+}
+
+void thread__find_addr_location(struct thread *self, u8 cpumode,
+				enum map_type type, u64 addr,
+				struct addr_location *al,
+				symbol_filter_t filter)
+{
+	struct thread *thread = al->thread = self;
+
+	al->addr = addr;
+
+	if (cpumode & PERF_RECORD_MISC_KERNEL) {
+		al->level = 'k';
+		thread = kthread;
+	} else if (cpumode & PERF_RECORD_MISC_USER)
+		al->level = '.';
+	else {
+		al->level = 'H';
+		al->map = NULL;
+		al->sym = NULL;
+		return;
+	}
+try_again:
+	al->map = thread__find_map(thread, type, al->addr);
+	if (al->map == NULL) {
+		/*
+		 * 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 or vdso (which executes in user-mode).
+		 *
+		 * XXX This is nasty, we should have a symbol list in the
+		 * "[vdso]" dso, but for now lets use the old trick of looking
+		 * in the whole kernel symbol list.
+		 */
+		if ((long long)al->addr < 0 && thread != kthread) {
+			thread = kthread;
+			goto try_again;
+		}
+		al->sym = NULL;
+	} else {
+		al->addr = al->map->map_ip(al->map, al->addr);
+		al->sym = map__find_symbol(al->map, al->addr, filter);
+	}
+}
+
+int event__preprocess_sample(const event_t *self, struct addr_location *al,
+			     symbol_filter_t filter)
+{
+	u8 cpumode = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+	struct thread *thread = threads__findnew(self->ip.pid);
+
+	if (thread == NULL)
+		return -1;
+
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+
+	thread__find_addr_location(thread, cpumode, MAP__FUNCTION,
+				   self->ip.ip, al, filter);
+	dump_printf(" ...... dso: %s\n",
+		    al->map ? al->map->dso->long_name :
+			al->level == 'H' ? "[hypervisor]" : "<not found>");
+	return 0;
+}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index 2c9c26d6ded0..a4cc8105cf67 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -1,14 +1,10 @@
 #ifndef __PERF_RECORD_H
 #define __PERF_RECORD_H
+
 #include "../perf.h"
 #include "util.h"
 #include <linux/list.h>
-
-enum {
-	SHOW_KERNEL	= 1,
-	SHOW_USER	= 2,
-	SHOW_HV		= 4,
-};
+#include <linux/rbtree.h>
 
 /*
  * PERF_SAMPLE_IP | PERF_SAMPLE_TID | *
@@ -65,6 +61,13 @@ struct sample_event{
 	u64 array[];
 };
 
+#define BUILD_ID_SIZE 20
+
+struct build_id_event {
+	struct perf_event_header header;
+	u8			 build_id[ALIGN(BUILD_ID_SIZE, sizeof(u64))];
+	char			 filename[];
+};
 
 typedef union event_union {
 	struct perf_event_header	header;
@@ -77,12 +80,30 @@ typedef union event_union {
 	struct sample_event		sample;
 } event_t;
 
+struct events_stats {
+	unsigned long total;
+	unsigned long lost;
+};
+
+void event__print_totals(void);
+
+enum map_type {
+	MAP__FUNCTION = 0,
+
+	MAP__NR_TYPES,
+};
+
 struct map {
-	struct list_head	node;
+	union {
+		struct rb_node	rb_node;
+		struct list_head node;
+	};
 	u64			start;
 	u64			end;
+	enum map_type		type;
 	u64			pgoff;
 	u64			(*map_ip)(struct map *, u64);
+	u64			(*unmap_ip)(struct map *, u64);
 	struct dso		*dso;
 };
 
@@ -91,14 +112,48 @@ static inline u64 map__map_ip(struct map *map, u64 ip)
 	return ip - map->start + map->pgoff;
 }
 
-static inline u64 vdso__map_ip(struct map *map __used, u64 ip)
+static inline u64 map__unmap_ip(struct map *map, u64 ip)
+{
+	return ip + map->start - map->pgoff;
+}
+
+static inline u64 identity__map_ip(struct map *map __used, u64 ip)
 {
 	return ip;
 }
 
-struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen);
+struct symbol;
+
+typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym);
+
+void map__init(struct map *self, enum map_type type,
+	       u64 start, u64 end, u64 pgoff, struct dso *dso);
+struct map *map__new(struct mmap_event *event, enum map_type,
+		     char *cwd, int cwdlen);
+void map__delete(struct map *self);
 struct map *map__clone(struct map *self);
 int map__overlap(struct map *l, struct map *r);
 size_t map__fprintf(struct map *self, FILE *fp);
+struct symbol *map__find_symbol(struct map *self, u64 addr,
+				symbol_filter_t filter);
+void map__fixup_start(struct map *self);
+void map__fixup_end(struct map *self);
+
+int event__synthesize_thread(pid_t pid, int (*process)(event_t *event));
+void event__synthesize_threads(int (*process)(event_t *event));
+
+extern char *event__cwd;
+extern int  event__cwdlen;
+extern struct events_stats event__stats;
+extern unsigned long event__total[PERF_RECORD_MAX];
+
+int event__process_comm(event_t *self);
+int event__process_lost(event_t *self);
+int event__process_mmap(event_t *self);
+int event__process_task(event_t *self);
+
+struct addr_location;
+int event__preprocess_sample(const event_t *self, struct addr_location *al,
+			     symbol_filter_t filter);
 
-#endif
+#endif /* __PERF_RECORD_H */
diff --git a/tools/perf/util/exec_cmd.h b/tools/perf/util/exec_cmd.h
index effe25eb1545..31647ac92ed1 100644
--- a/tools/perf/util/exec_cmd.h
+++ b/tools/perf/util/exec_cmd.h
@@ -1,5 +1,5 @@
-#ifndef PERF_EXEC_CMD_H
-#define PERF_EXEC_CMD_H
+#ifndef __PERF_EXEC_CMD_H
+#define __PERF_EXEC_CMD_H
 
 extern void perf_set_argv_exec_path(const char *exec_path);
 extern const char *perf_extract_argv0_path(const char *path);
@@ -10,4 +10,4 @@ extern int execv_perf_cmd(const char **argv); /* NULL terminated */
 extern int execl_perf_cmd(const char *cmd, ...);
 extern const char *system_path(const char *path);
 
-#endif /* PERF_EXEC_CMD_H */
+#endif /* __PERF_EXEC_CMD_H */
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index e306857b2c2b..4805e6dfd23c 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -2,9 +2,15 @@
 #include <unistd.h>
 #include <stdio.h>
 #include <stdlib.h>
+#include <linux/list.h>
 
 #include "util.h"
 #include "header.h"
+#include "../perf.h"
+#include "trace-event.h"
+#include "symbol.h"
+#include "data_map.h"
+#include "debug.h"
 
 /*
  * Create new perf.data header attribute:
@@ -13,32 +19,43 @@ struct perf_header_attr *perf_header_attr__new(struct perf_event_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");
+	if (self != NULL) {
+		self->attr = *attr;
+		self->ids  = 0;
+		self->size = 1;
+		self->id   = malloc(sizeof(u64));
+		if (self->id == NULL) {
+			free(self);
+			self = NULL;
+		}
+	}
 
 	return self;
 }
 
-void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
+void perf_header_attr__delete(struct perf_header_attr *self)
+{
+	free(self->id);
+	free(self);
+}
+
+int perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
 {
 	int pos = self->ids;
 
 	self->ids++;
 	if (self->ids > self->size) {
-		self->size *= 2;
-		self->id = realloc(self->id, self->size * sizeof(u64));
-		if (!self->id)
-			die("nomem");
+		int nsize = self->size * 2;
+		u64 *nid = realloc(self->id, nsize * sizeof(u64));
+
+		if (nid == NULL)
+			return -1;
+
+		self->size = nsize;
+		self->id = nid;
 	}
 	self->id[pos] = id;
+	return 0;
 }
 
 /*
@@ -46,42 +63,52 @@ void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
  */
 struct perf_header *perf_header__new(void)
 {
-	struct perf_header *self = malloc(sizeof(*self));
+	struct perf_header *self = zalloc(sizeof(*self));
 
-	if (!self)
-		die("nomem");
+	if (self != NULL) {
+		self->size = 1;
+		self->attr = malloc(sizeof(void *));
 
-	self->frozen = 0;
+		if (self->attr == NULL) {
+			free(self);
+			self = NULL;
+		}
+	}
 
-	self->attrs = 0;
-	self->size = 1;
-	self->attr = malloc(sizeof(void *));
+	return self;
+}
 
-	if (!self->attr)
-		die("nomem");
+void perf_header__delete(struct perf_header *self)
+{
+	int i;
 
-	self->data_offset = 0;
-	self->data_size = 0;
+	for (i = 0; i < self->attrs; ++i)
+		perf_header_attr__delete(self->attr[i]);
 
-	return self;
+	free(self->attr);
+	free(self);
 }
 
-void perf_header__add_attr(struct perf_header *self,
-			   struct perf_header_attr *attr)
+int perf_header__add_attr(struct perf_header *self,
+			  struct perf_header_attr *attr)
 {
-	int pos = self->attrs;
-
 	if (self->frozen)
-		die("frozen");
+		return -1;
 
-	self->attrs++;
-	if (self->attrs > self->size) {
-		self->size *= 2;
-		self->attr = realloc(self->attr, self->size * sizeof(void *));
-		if (!self->attr)
-			die("nomem");
+	if (self->attrs == self->size) {
+		int nsize = self->size * 2;
+		struct perf_header_attr **nattr;
+
+		nattr = realloc(self->attr, nsize * sizeof(void *));
+		if (nattr == NULL)
+			return -1;
+
+		self->size = nsize;
+		self->attr = nattr;
 	}
-	self->attr[pos] = attr;
+
+	self->attr[self->attrs++] = attr;
+	return 0;
 }
 
 #define MAX_EVENT_NAME 64
@@ -97,7 +124,7 @@ static struct perf_trace_event_type *events;
 void perf_header__push_event(u64 id, const char *name)
 {
 	if (strlen(name) > MAX_EVENT_NAME)
-		printf("Event %s will be truncated\n", name);
+		pr_warning("Event %s will be truncated\n", name);
 
 	if (!events) {
 		events = malloc(sizeof(struct perf_trace_event_type));
@@ -128,44 +155,137 @@ 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_event_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;
-	struct perf_file_section	event_types;
-};
+void perf_header__set_feat(struct perf_header *self, int feat)
+{
+	set_bit(feat, self->adds_features);
+}
 
-static void do_write(int fd, void *buf, size_t size)
+bool perf_header__has_feat(const struct perf_header *self, int feat)
+{
+	return test_bit(feat, self->adds_features);
+}
+
+static int do_write(int fd, const void *buf, size_t size)
 {
 	while (size) {
 		int ret = write(fd, buf, size);
 
 		if (ret < 0)
-			die("failed to write");
+			return -errno;
 
 		size -= ret;
 		buf += ret;
 	}
+
+	return 0;
+}
+
+static int __dsos__write_buildid_table(struct list_head *head, int fd)
+{
+	struct dso *pos;
+
+	list_for_each_entry(pos, head, node) {
+		int err;
+		struct build_id_event b;
+		size_t len;
+
+		if (!pos->has_build_id)
+			continue;
+		len = pos->long_name_len + 1;
+		len = ALIGN(len, 64);
+		memset(&b, 0, sizeof(b));
+		memcpy(&b.build_id, pos->build_id, sizeof(pos->build_id));
+		b.header.size = sizeof(b) + len;
+		err = do_write(fd, &b, sizeof(b));
+		if (err < 0)
+			return err;
+		err = do_write(fd, pos->long_name, len);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
 }
 
-void perf_header__write(struct perf_header *self, int fd)
+static int dsos__write_buildid_table(int fd)
+{
+	int err = __dsos__write_buildid_table(&dsos__kernel, fd);
+	if (err == 0)
+		err = __dsos__write_buildid_table(&dsos__user, fd);
+	return err;
+}
+
+static int perf_header__adds_write(struct perf_header *self, int fd)
+{
+	int nr_sections;
+	struct perf_file_section *feat_sec;
+	int sec_size;
+	u64 sec_start;
+	int idx = 0, err;
+
+	if (dsos__read_build_ids())
+		perf_header__set_feat(self, HEADER_BUILD_ID);
+
+	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	if (!nr_sections)
+		return 0;
+
+	feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+	if (feat_sec == NULL)
+		return -ENOMEM;
+
+	sec_size = sizeof(*feat_sec) * nr_sections;
+
+	sec_start = self->data_offset + self->data_size;
+	lseek(fd, sec_start + sec_size, SEEK_SET);
+
+	if (perf_header__has_feat(self, HEADER_TRACE_INFO)) {
+		struct perf_file_section *trace_sec;
+
+		trace_sec = &feat_sec[idx++];
+
+		/* Write trace info */
+		trace_sec->offset = lseek(fd, 0, SEEK_CUR);
+		read_tracing_data(fd, attrs, nr_counters);
+		trace_sec->size = lseek(fd, 0, SEEK_CUR) - trace_sec->offset;
+	}
+
+
+	if (perf_header__has_feat(self, HEADER_BUILD_ID)) {
+		struct perf_file_section *buildid_sec;
+
+		buildid_sec = &feat_sec[idx++];
+
+		/* Write build-ids */
+		buildid_sec->offset = lseek(fd, 0, SEEK_CUR);
+		err = dsos__write_buildid_table(fd);
+		if (err < 0) {
+			pr_debug("failed to write buildid table\n");
+			goto out_free;
+		}
+		buildid_sec->size = lseek(fd, 0, SEEK_CUR) - buildid_sec->offset;
+	}
+
+	lseek(fd, sec_start, SEEK_SET);
+	err = do_write(fd, feat_sec, sec_size);
+	if (err < 0)
+		pr_debug("failed to write feature section\n");
+out_free:
+	free(feat_sec);
+	return err;
+}
+
+int perf_header__write(struct perf_header *self, int fd, bool at_exit)
 {
 	struct perf_file_header f_header;
 	struct perf_file_attr   f_attr;
 	struct perf_header_attr	*attr;
-	int i;
+	int i, err;
 
 	lseek(fd, sizeof(f_header), SEEK_SET);
 
@@ -174,7 +294,11 @@ void perf_header__write(struct perf_header *self, int fd)
 		attr = self->attr[i];
 
 		attr->id_offset = lseek(fd, 0, SEEK_CUR);
-		do_write(fd, attr->id, attr->ids * sizeof(u64));
+		err = do_write(fd, attr->id, attr->ids * sizeof(u64));
+		if (err < 0) {
+			pr_debug("failed to write perf header\n");
+			return err;
+		}
 	}
 
 
@@ -190,17 +314,31 @@ void perf_header__write(struct perf_header *self, int fd)
 				.size   = attr->ids * sizeof(u64),
 			}
 		};
-		do_write(fd, &f_attr, sizeof(f_attr));
+		err = do_write(fd, &f_attr, sizeof(f_attr));
+		if (err < 0) {
+			pr_debug("failed to write perf header attribute\n");
+			return err;
+		}
 	}
 
 	self->event_offset = lseek(fd, 0, SEEK_CUR);
 	self->event_size = event_count * sizeof(struct perf_trace_event_type);
-	if (events)
-		do_write(fd, events, self->event_size);
-
+	if (events) {
+		err = do_write(fd, events, self->event_size);
+		if (err < 0) {
+			pr_debug("failed to write perf header events\n");
+			return err;
+		}
+	}
 
 	self->data_offset = lseek(fd, 0, SEEK_CUR);
 
+	if (at_exit) {
+		err = perf_header__adds_write(self, fd);
+		if (err < 0)
+			return err;
+	}
+
 	f_header = (struct perf_file_header){
 		.magic	   = PERF_MAGIC,
 		.size	   = sizeof(f_header),
@@ -219,11 +357,18 @@ void perf_header__write(struct perf_header *self, int fd)
 		},
 	};
 
+	memcpy(&f_header.adds_features, &self->adds_features, sizeof(self->adds_features));
+
 	lseek(fd, 0, SEEK_SET);
-	do_write(fd, &f_header, sizeof(f_header));
+	err = do_write(fd, &f_header, sizeof(f_header));
+	if (err < 0) {
+		pr_debug("failed to write perf header\n");
+		return err;
+	}
 	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
 
 	self->frozen = 1;
+	return 0;
 }
 
 static void do_read(int fd, void *buf, size_t size)
@@ -241,22 +386,109 @@ static void do_read(int fd, void *buf, size_t size)
 	}
 }
 
-struct perf_header *perf_header__read(int fd)
+int perf_header__process_sections(struct perf_header *self, int fd,
+				  int (*process)(struct perf_file_section *self,
+						 int feat, int fd))
+{
+	struct perf_file_section *feat_sec;
+	int nr_sections;
+	int sec_size;
+	int idx = 0;
+	int err = 0, feat = 1;
+
+	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	if (!nr_sections)
+		return 0;
+
+	feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+	if (!feat_sec)
+		return -1;
+
+	sec_size = sizeof(*feat_sec) * nr_sections;
+
+	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+
+	do_read(fd, feat_sec, sec_size);
+
+	while (idx < nr_sections && feat < HEADER_LAST_FEATURE) {
+		if (perf_header__has_feat(self, feat)) {
+			struct perf_file_section *sec = &feat_sec[idx++];
+
+			err = process(sec, feat, fd);
+			if (err < 0)
+				break;
+		}
+		++feat;
+	}
+
+	free(feat_sec);
+	return err;
+};
+
+int perf_file_header__read(struct perf_file_header *self,
+			   struct perf_header *ph, int fd)
+{
+	lseek(fd, 0, SEEK_SET);
+	do_read(fd, self, sizeof(*self));
+
+	if (self->magic     != PERF_MAGIC ||
+	    self->attr_size != sizeof(struct perf_file_attr))
+		return -1;
+
+	if (self->size != sizeof(*self)) {
+		/* Support the previous format */
+		if (self->size == offsetof(typeof(*self), adds_features))
+			bitmap_zero(self->adds_features, HEADER_FEAT_BITS);
+		else
+			return -1;
+	}
+
+	memcpy(&ph->adds_features, &self->adds_features,
+	       sizeof(self->adds_features));
+
+	ph->event_offset = self->event_types.offset;
+	ph->event_size	 = self->event_types.size;
+	ph->data_offset	 = self->data.offset;
+	ph->data_size	 = self->data.size;
+	return 0;
+}
+
+static int perf_file_section__process(struct perf_file_section *self,
+				      int feat, int fd)
+{
+	if (lseek(fd, self->offset, SEEK_SET) < 0) {
+		pr_debug("Failed to lseek to %Ld offset for feature %d, "
+			 "continuing...\n", self->offset, feat);
+		return 0;
+	}
+
+	switch (feat) {
+	case HEADER_TRACE_INFO:
+		trace_report(fd);
+		break;
+
+	case HEADER_BUILD_ID:
+		if (perf_header__read_build_ids(fd, self->offset, self->size))
+			pr_debug("Failed to read buildids, continuing...\n");
+		break;
+	default:
+		pr_debug("unknown feature %d, continuing...\n", feat);
+	}
+
+	return 0;
+}
+
+int perf_header__read(struct perf_header *self, 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");
+	if (perf_file_header__read(&f_header, self, fd) < 0) {
+		pr_debug("incompatible file format\n");
+		return -EINVAL;
+	}
 
 	nr_attrs = f_header.attrs.size / sizeof(f_attr);
 	lseek(fd, f_header.attrs.offset, SEEK_SET);
@@ -269,6 +501,8 @@ struct perf_header *perf_header__read(int fd)
 		tmp = lseek(fd, 0, SEEK_CUR);
 
 		attr = perf_header_attr__new(&f_attr.attr);
+		if (attr == NULL)
+			 return -ENOMEM;
 
 		nr_ids = f_attr.ids.size / sizeof(u64);
 		lseek(fd, f_attr.ids.offset, SEEK_SET);
@@ -276,31 +510,34 @@ struct perf_header *perf_header__read(int fd)
 		for (j = 0; j < nr_ids; j++) {
 			do_read(fd, &f_id, sizeof(f_id));
 
-			perf_header_attr__add_id(attr, f_id);
+			if (perf_header_attr__add_id(attr, f_id) < 0) {
+				perf_header_attr__delete(attr);
+				return -ENOMEM;
+			}
 		}
-		perf_header__add_attr(self, attr);
+		if (perf_header__add_attr(self, attr) < 0) {
+			perf_header_attr__delete(attr);
+			return -ENOMEM;
+		}
+
 		lseek(fd, tmp, SEEK_SET);
 	}
 
 	if (f_header.event_types.size) {
 		lseek(fd, f_header.event_types.offset, SEEK_SET);
 		events = malloc(f_header.event_types.size);
-		if (!events)
-			die("nomem");
+		if (events == NULL)
+			return -ENOMEM;
 		do_read(fd, events, f_header.event_types.size);
 		event_count =  f_header.event_types.size / sizeof(struct perf_trace_event_type);
 	}
-	self->event_offset = f_header.event_types.offset;
-	self->event_size   = f_header.event_types.size;
 
-	self->data_offset = f_header.data.offset;
-	self->data_size   = f_header.data.size;
+	perf_header__process_sections(self, fd, perf_file_section__process);
 
 	lseek(fd, self->data_offset, SEEK_SET);
 
 	self->frozen = 1;
-
-	return self;
+	return 0;
 }
 
 u64 perf_header__sample_type(struct perf_header *header)
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
index a0761bc7863c..d1dbe2b79c42 100644
--- a/tools/perf/util/header.h
+++ b/tools/perf/util/header.h
@@ -1,10 +1,13 @@
-#ifndef _PERF_HEADER_H
-#define _PERF_HEADER_H
+#ifndef __PERF_HEADER_H
+#define __PERF_HEADER_H
 
 #include "../../../include/linux/perf_event.h"
 #include <sys/types.h>
+#include <stdbool.h>
 #include "types.h"
 
+#include <linux/bitmap.h>
+
 struct perf_header_attr {
 	struct perf_event_attr attr;
 	int ids, size;
@@ -12,36 +15,71 @@ struct perf_header_attr {
 	off_t id_offset;
 };
 
+enum {
+	HEADER_TRACE_INFO = 1,
+	HEADER_BUILD_ID,
+	HEADER_LAST_FEATURE,
+};
+
+#define HEADER_FEAT_BITS			256
+
+struct perf_file_section {
+	u64 offset;
+	u64 size;
+};
+
+struct perf_file_header {
+	u64				magic;
+	u64				size;
+	u64				attr_size;
+	struct perf_file_section	attrs;
+	struct perf_file_section	data;
+	struct perf_file_section	event_types;
+	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
+};
+
+struct perf_header;
+
+int perf_file_header__read(struct perf_file_header *self,
+			   struct perf_header *ph, int fd);
+
 struct perf_header {
-	int frozen;
-	int attrs, size;
+	int			frozen;
+	int			attrs, size;
 	struct perf_header_attr **attr;
-	s64 attr_offset;
-	u64 data_offset;
-	u64 data_size;
-	u64 event_offset;
-	u64 event_size;
+	s64			attr_offset;
+	u64			data_offset;
+	u64			data_size;
+	u64			event_offset;
+	u64			event_size;
+	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
 };
 
-struct perf_header *perf_header__read(int fd);
-void perf_header__write(struct perf_header *self, int fd);
+struct perf_header *perf_header__new(void);
+void perf_header__delete(struct perf_header *self);
 
-void perf_header__add_attr(struct perf_header *self,
-			   struct perf_header_attr *attr);
+int perf_header__read(struct perf_header *self, int fd);
+int perf_header__write(struct perf_header *self, int fd, bool at_exit);
+
+int perf_header__add_attr(struct perf_header *self,
+			  struct perf_header_attr *attr);
 
 void perf_header__push_event(u64 id, const char *name);
 char *perf_header__find_event(u64 id);
 
+struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr);
+void perf_header_attr__delete(struct perf_header_attr *self);
 
-struct perf_header_attr *
-perf_header_attr__new(struct perf_event_attr *attr);
-void perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
+int perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
 
 u64 perf_header__sample_type(struct perf_header *header);
 struct perf_event_attr *
 perf_header__find_attr(u64 id, struct perf_header *header);
+void perf_header__set_feat(struct perf_header *self, int feat);
+bool perf_header__has_feat(const struct perf_header *self, int feat);
 
+int perf_header__process_sections(struct perf_header *self, int fd,
+				  int (*process)(struct perf_file_section *self,
+						 int feat, int fd));
 
-struct perf_header *perf_header__new(void);
-
-#endif /* _PERF_HEADER_H */
+#endif /* __PERF_HEADER_H */
diff --git a/tools/perf/util/help.h b/tools/perf/util/help.h
index 7128783637b4..7f5c6dedd714 100644
--- a/tools/perf/util/help.h
+++ b/tools/perf/util/help.h
@@ -1,5 +1,5 @@
-#ifndef HELP_H
-#define HELP_H
+#ifndef __PERF_HELP_H
+#define __PERF_HELP_H
 
 struct cmdnames {
 	size_t alloc;
@@ -26,4 +26,4 @@ int is_in_cmdlist(struct cmdnames *c, const char *s);
 void list_commands(const char *title, struct cmdnames *main_cmds,
 		   struct cmdnames *other_cmds);
 
-#endif /* HELP_H */
+#endif /* __PERF_HELP_H */
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
new file mode 100644
index 000000000000..0ebf6ee16caa
--- /dev/null
+++ b/tools/perf/util/hist.c
@@ -0,0 +1,202 @@
+#include "hist.h"
+
+struct rb_root hist;
+struct rb_root collapse_hists;
+struct rb_root output_hists;
+int callchain;
+
+struct callchain_param	callchain_param = {
+	.mode	= CHAIN_GRAPH_REL,
+	.min_percent = 0.5
+};
+
+/*
+ * histogram, sorted on item, collects counts
+ */
+
+struct hist_entry *__hist_entry__add(struct addr_location *al,
+				     struct symbol *sym_parent,
+				     u64 count, bool *hit)
+{
+	struct rb_node **p = &hist.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *he;
+	struct hist_entry entry = {
+		.thread	= al->thread,
+		.map	= al->map,
+		.sym	= al->sym,
+		.ip	= al->addr,
+		.level	= al->level,
+		.count	= count,
+		.parent = sym_parent,
+	};
+	int cmp;
+
+	while (*p != NULL) {
+		parent = *p;
+		he = rb_entry(parent, struct hist_entry, rb_node);
+
+		cmp = hist_entry__cmp(&entry, he);
+
+		if (!cmp) {
+			*hit = true;
+			return he;
+		}
+
+		if (cmp < 0)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	he = malloc(sizeof(*he));
+	if (!he)
+		return NULL;
+	*he = entry;
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &hist);
+	*hit = false;
+	return he;
+}
+
+int64_t
+hist_entry__cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct sort_entry *se;
+	int64_t cmp = 0;
+
+	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		cmp = se->cmp(left, right);
+		if (cmp)
+			break;
+	}
+
+	return cmp;
+}
+
+int64_t
+hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
+{
+	struct sort_entry *se;
+	int64_t cmp = 0;
+
+	list_for_each_entry(se, &hist_entry__sort_list, list) {
+		int64_t (*f)(struct hist_entry *, struct hist_entry *);
+
+		f = se->collapse ?: se->cmp;
+
+		cmp = f(left, right);
+		if (cmp)
+			break;
+	}
+
+	return cmp;
+}
+
+void hist_entry__free(struct hist_entry *he)
+{
+	free(he);
+}
+
+/*
+ * collapse the histogram
+ */
+
+void collapse__insert_entry(struct hist_entry *he)
+{
+	struct rb_node **p = &collapse_hists.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *iter;
+	int64_t cmp;
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct hist_entry, rb_node);
+
+		cmp = hist_entry__collapse(iter, he);
+
+		if (!cmp) {
+			iter->count += he->count;
+			hist_entry__free(he);
+			return;
+		}
+
+		if (cmp < 0)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &collapse_hists);
+}
+
+void collapse__resort(void)
+{
+	struct rb_node *next;
+	struct hist_entry *n;
+
+	if (!sort__need_collapse)
+		return;
+
+	next = rb_first(&hist);
+	while (next) {
+		n = rb_entry(next, struct hist_entry, rb_node);
+		next = rb_next(&n->rb_node);
+
+		rb_erase(&n->rb_node, &hist);
+		collapse__insert_entry(n);
+	}
+}
+
+/*
+ * reverse the map, sort on count.
+ */
+
+void output__insert_entry(struct hist_entry *he, u64 min_callchain_hits)
+{
+	struct rb_node **p = &output_hists.rb_node;
+	struct rb_node *parent = NULL;
+	struct hist_entry *iter;
+
+	if (callchain)
+		callchain_param.sort(&he->sorted_chain, &he->callchain,
+				      min_callchain_hits, &callchain_param);
+
+	while (*p != NULL) {
+		parent = *p;
+		iter = rb_entry(parent, struct hist_entry, rb_node);
+
+		if (he->count > iter->count)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&he->rb_node, parent, p);
+	rb_insert_color(&he->rb_node, &output_hists);
+}
+
+void output__resort(u64 total_samples)
+{
+	struct rb_node *next;
+	struct hist_entry *n;
+	struct rb_root *tree = &hist;
+	u64 min_callchain_hits;
+
+	min_callchain_hits =
+		total_samples * (callchain_param.min_percent / 100);
+
+	if (sort__need_collapse)
+		tree = &collapse_hists;
+
+	next = rb_first(tree);
+
+	while (next) {
+		n = rb_entry(next, struct hist_entry, rb_node);
+		next = rb_next(&n->rb_node);
+
+		rb_erase(&n->rb_node, tree);
+		output__insert_entry(n, min_callchain_hits);
+	}
+}
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
new file mode 100644
index 000000000000..3020db0c9292
--- /dev/null
+++ b/tools/perf/util/hist.h
@@ -0,0 +1,50 @@
+#ifndef __PERF_HIST_H
+#define __PERF_HIST_H
+#include "../builtin.h"
+
+#include "util.h"
+
+#include "color.h"
+#include <linux/list.h>
+#include "cache.h"
+#include <linux/rbtree.h>
+#include "symbol.h"
+#include "string.h"
+#include "callchain.h"
+#include "strlist.h"
+#include "values.h"
+
+#include "../perf.h"
+#include "debug.h"
+#include "header.h"
+
+#include "parse-options.h"
+#include "parse-events.h"
+
+#include "thread.h"
+#include "sort.h"
+
+extern struct rb_root hist;
+extern struct rb_root collapse_hists;
+extern struct rb_root output_hists;
+extern int callchain;
+extern struct callchain_param callchain_param;
+extern unsigned long total;
+extern unsigned long total_mmap;
+extern unsigned long total_comm;
+extern unsigned long total_fork;
+extern unsigned long total_unknown;
+extern unsigned long total_lost;
+
+struct hist_entry *__hist_entry__add(struct addr_location *al,
+				     struct symbol *parent,
+				     u64 count, bool *hit);
+extern int64_t hist_entry__cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t hist_entry__collapse(struct hist_entry *, struct hist_entry *);
+extern void hist_entry__free(struct hist_entry *);
+extern void collapse__insert_entry(struct hist_entry *);
+extern void collapse__resort(void);
+extern void output__insert_entry(struct hist_entry *, u64);
+extern void output__resort(u64);
+
+#endif	/* __PERF_HIST_H */
diff --git a/tools/perf/util/include/asm/asm-offsets.h b/tools/perf/util/include/asm/asm-offsets.h
new file mode 100644
index 000000000000..ed538942523d
--- /dev/null
+++ b/tools/perf/util/include/asm/asm-offsets.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/bitops.h b/tools/perf/util/include/asm/bitops.h
new file mode 100644
index 000000000000..58e9817ffae0
--- /dev/null
+++ b/tools/perf/util/include/asm/bitops.h
@@ -0,0 +1,18 @@
+#ifndef _PERF_ASM_BITOPS_H_
+#define _PERF_ASM_BITOPS_H_
+
+#include <sys/types.h>
+#include "../../types.h"
+#include <linux/compiler.h>
+
+/* CHECKME: Not sure both always match */
+#define BITS_PER_LONG	__WORDSIZE
+
+#include "../../../../include/asm-generic/bitops/__fls.h"
+#include "../../../../include/asm-generic/bitops/fls.h"
+#include "../../../../include/asm-generic/bitops/fls64.h"
+#include "../../../../include/asm-generic/bitops/__ffs.h"
+#include "../../../../include/asm-generic/bitops/ffz.h"
+#include "../../../../include/asm-generic/bitops/hweight.h"
+
+#endif
diff --git a/tools/perf/util/include/asm/bug.h b/tools/perf/util/include/asm/bug.h
new file mode 100644
index 000000000000..7fcc6810adc2
--- /dev/null
+++ b/tools/perf/util/include/asm/bug.h
@@ -0,0 +1,22 @@
+#ifndef _PERF_ASM_GENERIC_BUG_H
+#define _PERF_ASM_GENERIC_BUG_H
+
+#define __WARN_printf(arg...)	do { fprintf(stderr, arg); } while (0)
+
+#define WARN(condition, format...) ({		\
+	int __ret_warn_on = !!(condition);	\
+	if (unlikely(__ret_warn_on))		\
+		__WARN_printf(format);		\
+	unlikely(__ret_warn_on);		\
+})
+
+#define WARN_ONCE(condition, format...)	({	\
+	static int __warned;			\
+	int __ret_warn_once = !!(condition);	\
+						\
+	if (unlikely(__ret_warn_once))		\
+		if (WARN(!__warned, format)) 	\
+			__warned = 1;		\
+	unlikely(__ret_warn_once);		\
+})
+#endif
diff --git a/tools/perf/util/include/asm/byteorder.h b/tools/perf/util/include/asm/byteorder.h
new file mode 100644
index 000000000000..b722abe3a626
--- /dev/null
+++ b/tools/perf/util/include/asm/byteorder.h
@@ -0,0 +1,2 @@
+#include <asm/types.h>
+#include "../../../../include/linux/swab.h"
diff --git a/tools/perf/util/include/asm/swab.h b/tools/perf/util/include/asm/swab.h
new file mode 100644
index 000000000000..ed538942523d
--- /dev/null
+++ b/tools/perf/util/include/asm/swab.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/uaccess.h b/tools/perf/util/include/asm/uaccess.h
new file mode 100644
index 000000000000..d0f72b8fcc35
--- /dev/null
+++ b/tools/perf/util/include/asm/uaccess.h
@@ -0,0 +1,14 @@
+#ifndef _PERF_ASM_UACCESS_H_
+#define _PERF_ASM_UACCESS_H_
+
+#define __get_user(src, dest)						\
+({									\
+	(src) = *dest;							\
+	0;								\
+})
+
+#define get_user	__get_user
+
+#define access_ok(type, addr, size)	1
+
+#endif
diff --git a/tools/perf/util/include/linux/bitmap.h b/tools/perf/util/include/linux/bitmap.h
new file mode 100644
index 000000000000..94507639a8c4
--- /dev/null
+++ b/tools/perf/util/include/linux/bitmap.h
@@ -0,0 +1,3 @@
+#include "../../../../include/linux/bitmap.h"
+#include "../../../../include/asm-generic/bitops/find.h"
+#include <linux/errno.h>
diff --git a/tools/perf/util/include/linux/bitops.h b/tools/perf/util/include/linux/bitops.h
new file mode 100644
index 000000000000..8d63116e9435
--- /dev/null
+++ b/tools/perf/util/include/linux/bitops.h
@@ -0,0 +1,29 @@
+#ifndef _PERF_LINUX_BITOPS_H_
+#define _PERF_LINUX_BITOPS_H_
+
+#define __KERNEL__
+
+#define CONFIG_GENERIC_FIND_NEXT_BIT
+#define CONFIG_GENERIC_FIND_FIRST_BIT
+#include "../../../../include/linux/bitops.h"
+
+#undef __KERNEL__
+
+static inline void set_bit(int nr, unsigned long *addr)
+{
+	addr[nr / BITS_PER_LONG] |= 1UL << (nr % BITS_PER_LONG);
+}
+
+static __always_inline int test_bit(unsigned int nr, const unsigned long *addr)
+{
+	return ((1UL << (nr % BITS_PER_LONG)) &
+		(((unsigned long *)addr)[nr / BITS_PER_LONG])) != 0;
+}
+
+unsigned long generic_find_next_zero_le_bit(const unsigned long *addr, unsigned
+		long size, unsigned long offset);
+
+unsigned long generic_find_next_le_bit(const unsigned long *addr, unsigned
+		long size, unsigned long offset);
+
+#endif
diff --git a/tools/perf/util/include/linux/compiler.h b/tools/perf/util/include/linux/compiler.h
new file mode 100644
index 000000000000..dfb0713ed47f
--- /dev/null
+++ b/tools/perf/util/include/linux/compiler.h
@@ -0,0 +1,10 @@
+#ifndef _PERF_LINUX_COMPILER_H_
+#define _PERF_LINUX_COMPILER_H_
+
+#ifndef __always_inline
+#define __always_inline	inline
+#endif
+#define __user
+#define __attribute_const__
+
+#endif
diff --git a/tools/perf/util/include/linux/ctype.h b/tools/perf/util/include/linux/ctype.h
new file mode 100644
index 000000000000..a53d4ee1e0b7
--- /dev/null
+++ b/tools/perf/util/include/linux/ctype.h
@@ -0,0 +1 @@
+#include "../util.h"
diff --git a/tools/perf/util/include/linux/kernel.h b/tools/perf/util/include/linux/kernel.h
index a6b87390cb52..21c0274c02fa 100644
--- a/tools/perf/util/include/linux/kernel.h
+++ b/tools/perf/util/include/linux/kernel.h
@@ -1,6 +1,16 @@
 #ifndef PERF_LINUX_KERNEL_H_
 #define PERF_LINUX_KERNEL_H_
 
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <assert.h>
+
+#define DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d))
+
+#define ALIGN(x,a)		__ALIGN_MASK(x,(typeof(x))(a)-1)
+#define __ALIGN_MASK(x,mask)	(((x)+(mask))&~(mask))
+
 #ifndef offsetof
 #define offsetof(TYPE, MEMBER) ((size_t) &((TYPE *)0)->MEMBER)
 #endif
@@ -26,4 +36,70 @@
 	_max1 > _max2 ? _max1 : _max2; })
 #endif
 
+#ifndef min
+#define min(x, y) ({				\
+	typeof(x) _min1 = (x);			\
+	typeof(y) _min2 = (y);			\
+	(void) (&_min1 == &_min2);		\
+	_min1 < _min2 ? _min1 : _min2; })
+#endif
+
+#ifndef BUG_ON
+#define BUG_ON(cond) assert(!(cond))
+#endif
+
+/*
+ * Both need more care to handle endianness
+ * (Don't use bitmap_copy_le() for now)
+ */
+#define cpu_to_le64(x)	(x)
+#define cpu_to_le32(x)	(x)
+
+static inline int
+vscnprintf(char *buf, size_t size, const char *fmt, va_list args)
+{
+	int i;
+	ssize_t ssize = size;
+
+	i = vsnprintf(buf, size, fmt, args);
+
+	return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline int scnprintf(char * buf, size_t size, const char * fmt, ...)
+{
+	va_list args;
+	ssize_t ssize = size;
+	int i;
+
+	va_start(args, fmt);
+	i = vsnprintf(buf, size, fmt, args);
+	va_end(args);
+
+	return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline unsigned long
+simple_strtoul(const char *nptr, char **endptr, int base)
+{
+	return strtoul(nptr, endptr, base);
+}
+
+#ifndef pr_fmt
+#define pr_fmt(fmt) fmt
+#endif
+
+#define pr_err(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_warning(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_info(fmt, ...) \
+	do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_debug(fmt, ...) \
+	eprintf(1, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debugN(n, fmt, ...) \
+	eprintf(n, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug3(fmt, ...) pr_debugN(3, pr_fmt(fmt), ##__VA_ARGS__)
+
 #endif
diff --git a/tools/perf/util/include/linux/string.h b/tools/perf/util/include/linux/string.h
new file mode 100644
index 000000000000..3b2f5900276f
--- /dev/null
+++ b/tools/perf/util/include/linux/string.h
@@ -0,0 +1 @@
+#include <string.h>
diff --git a/tools/perf/util/include/linux/types.h b/tools/perf/util/include/linux/types.h
new file mode 100644
index 000000000000..196862a81a21
--- /dev/null
+++ b/tools/perf/util/include/linux/types.h
@@ -0,0 +1,9 @@
+#ifndef _PERF_LINUX_TYPES_H_
+#define _PERF_LINUX_TYPES_H_
+
+#include <asm/types.h>
+
+#define DECLARE_BITMAP(name,bits) \
+	unsigned long name[BITS_TO_LONGS(bits)]
+
+#endif
diff --git a/tools/perf/util/levenshtein.h b/tools/perf/util/levenshtein.h
index 0173abeef52c..b0fcb6d8a881 100644
--- a/tools/perf/util/levenshtein.h
+++ b/tools/perf/util/levenshtein.h
@@ -1,8 +1,8 @@
-#ifndef LEVENSHTEIN_H
-#define LEVENSHTEIN_H
+#ifndef __PERF_LEVENSHTEIN_H
+#define __PERF_LEVENSHTEIN_H
 
 int levenshtein(const char *string1, const char *string2,
 	int swap_penalty, int substition_penalty,
 	int insertion_penalty, int deletion_penalty);
 
-#endif
+#endif /* __PERF_LEVENSHTEIN_H */
diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index 804e02382739..69f94fe9db20 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -3,6 +3,7 @@
 #include <stdlib.h>
 #include <string.h>
 #include <stdio.h>
+#include "debug.h"
 
 static inline int is_anon_memory(const char *filename)
 {
@@ -19,13 +20,28 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 	return n;
 }
 
- struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen)
+void map__init(struct map *self, enum map_type type,
+	       u64 start, u64 end, u64 pgoff, struct dso *dso)
+{
+	self->type     = type;
+	self->start    = start;
+	self->end      = end;
+	self->pgoff    = pgoff;
+	self->dso      = dso;
+	self->map_ip   = map__map_ip;
+	self->unmap_ip = map__unmap_ip;
+	RB_CLEAR_NODE(&self->rb_node);
+}
+
+struct map *map__new(struct mmap_event *event, enum map_type type,
+		     char *cwd, int cwdlen)
 {
 	struct map *self = malloc(sizeof(*self));
 
 	if (self != NULL) {
 		const char *filename = event->filename;
 		char newfilename[PATH_MAX];
+		struct dso *dso;
 		int anon;
 
 		if (cwd) {
@@ -45,18 +61,15 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 			filename = newfilename;
 		}
 
-		self->start = event->start;
-		self->end   = event->start + event->len;
-		self->pgoff = event->pgoff;
-
-		self->dso = dsos__findnew(filename);
-		if (self->dso == NULL)
+		dso = dsos__findnew(filename);
+		if (dso == NULL)
 			goto out_delete;
 
+		map__init(self, type, event->start, event->start + event->len,
+			  event->pgoff, dso);
+
 		if (self->dso == vdso || anon)
-			self->map_ip = vdso__map_ip;
-		else
-			self->map_ip = map__map_ip;
+			self->map_ip = self->unmap_ip = identity__map_ip;
 	}
 	return self;
 out_delete:
@@ -64,6 +77,72 @@ out_delete:
 	return NULL;
 }
 
+void map__delete(struct map *self)
+{
+	free(self);
+}
+
+void map__fixup_start(struct map *self)
+{
+	struct rb_root *symbols = &self->dso->symbols[self->type];
+	struct rb_node *nd = rb_first(symbols);
+	if (nd != NULL) {
+		struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+		self->start = sym->start;
+	}
+}
+
+void map__fixup_end(struct map *self)
+{
+	struct rb_root *symbols = &self->dso->symbols[self->type];
+	struct rb_node *nd = rb_last(symbols);
+	if (nd != NULL) {
+		struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+		self->end = sym->end;
+	}
+}
+
+#define DSO__DELETED "(deleted)"
+
+struct symbol *map__find_symbol(struct map *self, u64 addr,
+				symbol_filter_t filter)
+{
+	if (!dso__loaded(self->dso, self->type)) {
+		int nr = dso__load(self->dso, self, filter);
+
+		if (nr < 0) {
+			if (self->dso->has_build_id) {
+				char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+				build_id__sprintf(self->dso->build_id,
+						  sizeof(self->dso->build_id),
+						  sbuild_id);
+				pr_warning("%s with build id %s not found",
+					   self->dso->long_name, sbuild_id);
+			} else
+				pr_warning("Failed to open %s",
+					   self->dso->long_name);
+			pr_warning(", continuing without symbols\n");
+			return NULL;
+		} else if (nr == 0) {
+			const char *name = self->dso->long_name;
+			const size_t len = strlen(name);
+			const size_t real_len = len - sizeof(DSO__DELETED);
+
+			if (len > sizeof(DSO__DELETED) &&
+			    strcmp(name + real_len + 1, DSO__DELETED) == 0) {
+				pr_warning("%.*s was updated, restart the long running apps that use it!\n",
+					   (int)real_len, name);
+			} else {
+				pr_warning("no symbols found in %s, maybe install a debug package?\n", name);
+			}
+			return NULL;
+		}
+	}
+
+	return self->dso->find_symbol(self->dso, self->type, addr);
+}
+
 struct map *map__clone(struct map *self)
 {
 	struct map *map = malloc(sizeof(*self));
diff --git a/tools/perf/util/module.c b/tools/perf/util/module.c
deleted file mode 100644
index 0d8c85defcd2..000000000000
--- a/tools/perf/util/module.c
+++ /dev/null
@@ -1,545 +0,0 @@
-#include "util.h"
-#include "../perf.h"
-#include "string.h"
-#include "module.h"
-
-#include <libelf.h>
-#include <libgen.h>
-#include <gelf.h>
-#include <elf.h>
-#include <dirent.h>
-#include <sys/utsname.h>
-
-static unsigned int crc32(const char *p, unsigned int len)
-{
-	int i;
-	unsigned int crc = 0;
-
-	while (len--) {
-		crc ^= *p++;
-		for (i = 0; i < 8; i++)
-			crc = (crc >> 1) ^ ((crc & 1) ? 0xedb88320 : 0);
-	}
-	return crc;
-}
-
-/* module section methods */
-
-struct sec_dso *sec_dso__new_dso(const char *name)
-{
-	struct sec_dso *self = malloc(sizeof(*self) + strlen(name) + 1);
-
-	if (self != NULL) {
-		strcpy(self->name, name);
-		self->secs = RB_ROOT;
-		self->find_section = sec_dso__find_section;
-	}
-
-	return self;
-}
-
-static void sec_dso__delete_section(struct section *self)
-{
-	free(((void *)self));
-}
-
-void sec_dso__delete_sections(struct sec_dso *self)
-{
-	struct section *pos;
-	struct rb_node *next = rb_first(&self->secs);
-
-	while (next) {
-		pos = rb_entry(next, struct section, rb_node);
-		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &self->secs);
-		sec_dso__delete_section(pos);
-	}
-}
-
-void sec_dso__delete_self(struct sec_dso *self)
-{
-	sec_dso__delete_sections(self);
-	free(self);
-}
-
-static void sec_dso__insert_section(struct sec_dso *self, struct section *sec)
-{
-	struct rb_node **p = &self->secs.rb_node;
-	struct rb_node *parent = NULL;
-	const u64 hash = sec->hash;
-	struct section *s;
-
-	while (*p != NULL) {
-		parent = *p;
-		s = rb_entry(parent, struct section, rb_node);
-		if (hash < s->hash)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-	rb_link_node(&sec->rb_node, parent, p);
-	rb_insert_color(&sec->rb_node, &self->secs);
-}
-
-struct section *sec_dso__find_section(struct sec_dso *self, const char *name)
-{
-	struct rb_node *n;
-	u64 hash;
-	int len;
-
-	if (self == NULL)
-		return NULL;
-
-	len = strlen(name);
-	hash = crc32(name, len);
-
-	n = self->secs.rb_node;
-
-	while (n) {
-		struct section *s = rb_entry(n, struct section, rb_node);
-
-		if (hash < s->hash)
-			n = n->rb_left;
-		else if (hash > s->hash)
-			n = n->rb_right;
-		else {
-			if (!strcmp(name, s->name))
-				return s;
-			else
-				n = rb_next(&s->rb_node);
-		}
-	}
-
-	return NULL;
-}
-
-static size_t sec_dso__fprintf_section(struct section *self, FILE *fp)
-{
-	return fprintf(fp, "name:%s vma:%llx path:%s\n",
-		       self->name, self->vma, self->path);
-}
-
-size_t sec_dso__fprintf(struct sec_dso *self, FILE *fp)
-{
-	size_t ret = fprintf(fp, "dso: %s\n", self->name);
-
-	struct rb_node *nd;
-	for (nd = rb_first(&self->secs); nd; nd = rb_next(nd)) {
-		struct section *pos = rb_entry(nd, struct section, rb_node);
-		ret += sec_dso__fprintf_section(pos, fp);
-	}
-
-	return ret;
-}
-
-static struct section *section__new(const char *name, const char *path)
-{
-	struct section *self = calloc(1, sizeof(*self));
-
-	if (!self)
-		goto out_failure;
-
-	self->name = calloc(1, strlen(name) + 1);
-	if (!self->name)
-		goto out_failure;
-
-	self->path = calloc(1, strlen(path) + 1);
-	if (!self->path)
-		goto out_failure;
-
-	strcpy(self->name, name);
-	strcpy(self->path, path);
-	self->hash = crc32(self->name, strlen(name));
-
-	return self;
-
-out_failure:
-	if (self) {
-		if (self->name)
-			free(self->name);
-		if (self->path)
-			free(self->path);
-		free(self);
-	}
-
-	return NULL;
-}
-
-/* module methods */
-
-struct mod_dso *mod_dso__new_dso(const char *name)
-{
-	struct mod_dso *self = malloc(sizeof(*self) + strlen(name) + 1);
-
-	if (self != NULL) {
-		strcpy(self->name, name);
-		self->mods = RB_ROOT;
-		self->find_module = mod_dso__find_module;
-	}
-
-	return self;
-}
-
-static void mod_dso__delete_module(struct module *self)
-{
-	free(((void *)self));
-}
-
-void mod_dso__delete_modules(struct mod_dso *self)
-{
-	struct module *pos;
-	struct rb_node *next = rb_first(&self->mods);
-
-	while (next) {
-		pos = rb_entry(next, struct module, rb_node);
-		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &self->mods);
-		mod_dso__delete_module(pos);
-	}
-}
-
-void mod_dso__delete_self(struct mod_dso *self)
-{
-	mod_dso__delete_modules(self);
-	free(self);
-}
-
-static void mod_dso__insert_module(struct mod_dso *self, struct module *mod)
-{
-	struct rb_node **p = &self->mods.rb_node;
-	struct rb_node *parent = NULL;
-	const u64 hash = mod->hash;
-	struct module *m;
-
-	while (*p != NULL) {
-		parent = *p;
-		m = rb_entry(parent, struct module, rb_node);
-		if (hash < m->hash)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-	rb_link_node(&mod->rb_node, parent, p);
-	rb_insert_color(&mod->rb_node, &self->mods);
-}
-
-struct module *mod_dso__find_module(struct mod_dso *self, const char *name)
-{
-	struct rb_node *n;
-	u64 hash;
-	int len;
-
-	if (self == NULL)
-		return NULL;
-
-	len = strlen(name);
-	hash = crc32(name, len);
-
-	n = self->mods.rb_node;
-
-	while (n) {
-		struct module *m = rb_entry(n, struct module, rb_node);
-
-		if (hash < m->hash)
-			n = n->rb_left;
-		else if (hash > m->hash)
-			n = n->rb_right;
-		else {
-			if (!strcmp(name, m->name))
-				return m;
-			else
-				n = rb_next(&m->rb_node);
-		}
-	}
-
-	return NULL;
-}
-
-static size_t mod_dso__fprintf_module(struct module *self, FILE *fp)
-{
-	return fprintf(fp, "name:%s path:%s\n", self->name, self->path);
-}
-
-size_t mod_dso__fprintf(struct mod_dso *self, FILE *fp)
-{
-	struct rb_node *nd;
-	size_t ret;
-
-	ret = fprintf(fp, "dso: %s\n", self->name);
-
-	for (nd = rb_first(&self->mods); nd; nd = rb_next(nd)) {
-		struct module *pos = rb_entry(nd, struct module, rb_node);
-
-		ret += mod_dso__fprintf_module(pos, fp);
-	}
-
-	return ret;
-}
-
-static struct module *module__new(const char *name, const char *path)
-{
-	struct module *self = calloc(1, sizeof(*self));
-
-	if (!self)
-		goto out_failure;
-
-	self->name = calloc(1, strlen(name) + 1);
-	if (!self->name)
-		goto out_failure;
-
-	self->path = calloc(1, strlen(path) + 1);
-	if (!self->path)
-		goto out_failure;
-
-	strcpy(self->name, name);
-	strcpy(self->path, path);
-	self->hash = crc32(self->name, strlen(name));
-
-	return self;
-
-out_failure:
-	if (self) {
-		if (self->name)
-			free(self->name);
-		if (self->path)
-			free(self->path);
-		free(self);
-	}
-
-	return NULL;
-}
-
-static int mod_dso__load_sections(struct module *mod)
-{
-	int count = 0, path_len;
-	struct dirent *entry;
-	char *line = NULL;
-	char *dir_path;
-	DIR *dir;
-	size_t n;
-
-	path_len = strlen("/sys/module/");
-	path_len += strlen(mod->name);
-	path_len += strlen("/sections/");
-
-	dir_path = calloc(1, path_len + 1);
-	if (dir_path == NULL)
-		goto out_failure;
-
-	strcat(dir_path, "/sys/module/");
-	strcat(dir_path, mod->name);
-	strcat(dir_path, "/sections/");
-
-	dir = opendir(dir_path);
-	if (dir == NULL)
-		goto out_free;
-
-	while ((entry = readdir(dir))) {
-		struct section *section;
-		char *path, *vma;
-		int line_len;
-		FILE *file;
-
-		if (!strcmp(".", entry->d_name) || !strcmp("..", entry->d_name))
-			continue;
-
-		path = calloc(1, path_len + strlen(entry->d_name) + 1);
-		if (path == NULL)
-			break;
-		strcat(path, dir_path);
-		strcat(path, entry->d_name);
-
-		file = fopen(path, "r");
-		if (file == NULL) {
-			free(path);
-			break;
-		}
-
-		line_len = getline(&line, &n, file);
-		if (line_len < 0) {
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		if (!line) {
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		line[--line_len] = '\0'; /* \n */
-
-		vma = strstr(line, "0x");
-		if (!vma) {
-			free(path);
-			fclose(file);
-			break;
-		}
-		vma += 2;
-
-		section = section__new(entry->d_name, path);
-		if (!section) {
-			fprintf(stderr, "load_sections: allocation error\n");
-			free(path);
-			fclose(file);
-			break;
-		}
-
-		hex2u64(vma, &section->vma);
-		sec_dso__insert_section(mod->sections, section);
-
-		free(path);
-		fclose(file);
-		count++;
-	}
-
-	closedir(dir);
-	free(line);
-	free(dir_path);
-
-	return count;
-
-out_free:
-	free(dir_path);
-
-out_failure:
-	return count;
-}
-
-static int mod_dso__load_module_paths(struct mod_dso *self)
-{
-	struct utsname uts;
-	int count = 0, len, err = -1;
-	char *line = NULL;
-	FILE *file;
-	char *dpath, *dir;
-	size_t n;
-
-	if (uname(&uts) < 0)
-		return err;
-
-	len = strlen("/lib/modules/");
-	len += strlen(uts.release);
-	len += strlen("/modules.dep");
-
-	dpath = calloc(1, len + 1);
-	if (dpath == NULL)
-		return err;
-
-	strcat(dpath, "/lib/modules/");
-	strcat(dpath, uts.release);
-	strcat(dpath, "/modules.dep");
-
-	file = fopen(dpath, "r");
-	if (file == NULL)
-		goto out_failure;
-
-	dir = dirname(dpath);
-	if (!dir)
-		goto out_failure;
-	strcat(dir, "/");
-
-	while (!feof(file)) {
-		struct module *module;
-		char *name, *path, *tmp;
-		FILE *modfile;
-		int line_len;
-
-		line_len = getline(&line, &n, file);
-		if (line_len < 0)
-			break;
-
-		if (!line)
-			break;
-
-		line[--line_len] = '\0'; /* \n */
-
-		path = strchr(line, ':');
-		if (!path)
-			break;
-		*path = '\0';
-
-		path = strdup(line);
-		if (!path)
-			break;
-
-		if (!strstr(path, dir)) {
-			if (strncmp(path, "kernel/", 7))
-				break;
-
-			free(path);
-			path = calloc(1, strlen(dir) + strlen(line) + 1);
-			if (!path)
-				break;
-			strcat(path, dir);
-			strcat(path, line);
-		}
-
-		modfile = fopen(path, "r");
-		if (modfile == NULL)
-			break;
-		fclose(modfile);
-
-		name = strdup(path);
-		if (!name)
-			break;
-
-		name = strtok(name, "/");
-		tmp = name;
-
-		while (tmp) {
-			tmp = strtok(NULL, "/");
-			if (tmp)
-				name = tmp;
-		}
-
-		name = strsep(&name, ".");
-		if (!name)
-			break;
-
-		/* Quirk: replace '-' with '_' in all modules */
-		for (len = strlen(name); len; len--) {
-			if (*(name+len) == '-')
-				*(name+len) = '_';
-		}
-
-		module = module__new(name, path);
-		if (!module)
-			break;
-		mod_dso__insert_module(self, module);
-
-		module->sections = sec_dso__new_dso("sections");
-		if (!module->sections)
-			break;
-
-		module->active = mod_dso__load_sections(module);
-
-		if (module->active > 0)
-			count++;
-	}
-
-	if (feof(file))
-		err = count;
-	else
-		fprintf(stderr, "load_module_paths: modules.dep parsing failure!\n");
-
-out_failure:
-	if (dpath)
-		free(dpath);
-	if (file)
-		fclose(file);
-	if (line)
-		free(line);
-
-	return err;
-}
-
-int mod_dso__load_modules(struct mod_dso *dso)
-{
-	int err;
-
-	err = mod_dso__load_module_paths(dso);
-
-	return err;
-}
diff --git a/tools/perf/util/module.h b/tools/perf/util/module.h
deleted file mode 100644
index 8a592ef641ca..000000000000
--- a/tools/perf/util/module.h
+++ /dev/null
@@ -1,53 +0,0 @@
-#ifndef _PERF_MODULE_
-#define _PERF_MODULE_ 1
-
-#include <linux/types.h>
-#include "../types.h"
-#include <linux/list.h>
-#include <linux/rbtree.h>
-
-struct section {
-	struct rb_node	rb_node;
-	u64		hash;
-	u64		vma;
-	char		*name;
-	char		*path;
-};
-
-struct sec_dso {
-	struct list_head node;
-	struct rb_root	 secs;
-	struct section    *(*find_section)(struct sec_dso *, const char *name);
-	char		 name[0];
-};
-
-struct module {
-	struct rb_node	rb_node;
-	u64		hash;
-	char		*name;
-	char		*path;
-	struct sec_dso	*sections;
-	int		active;
-};
-
-struct mod_dso {
-	struct list_head node;
-	struct rb_root	 mods;
-	struct module    *(*find_module)(struct mod_dso *, const char *name);
-	char		 name[0];
-};
-
-struct sec_dso *sec_dso__new_dso(const char *name);
-void sec_dso__delete_sections(struct sec_dso *self);
-void sec_dso__delete_self(struct sec_dso *self);
-size_t sec_dso__fprintf(struct sec_dso *self, FILE *fp);
-struct section *sec_dso__find_section(struct sec_dso *self, const char *name);
-
-struct mod_dso *mod_dso__new_dso(const char *name);
-void mod_dso__delete_modules(struct mod_dso *self);
-void mod_dso__delete_self(struct mod_dso *self);
-size_t mod_dso__fprintf(struct mod_dso *self, FILE *fp);
-struct module *mod_dso__find_module(struct mod_dso *self, const char *name);
-int mod_dso__load_modules(struct mod_dso *dso);
-
-#endif /* _PERF_MODULE_ */
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 8cfb48cbbea0..9e5dbd66d34d 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -1,4 +1,4 @@
-
+#include "../../../include/linux/hw_breakpoint.h"
 #include "util.h"
 #include "../perf.h"
 #include "parse-options.h"
@@ -7,10 +7,12 @@
 #include "string.h"
 #include "cache.h"
 #include "header.h"
+#include "debugfs.h"
 
-int					nr_counters;
+int				nr_counters;
 
 struct perf_event_attr		attrs[MAX_COUNTERS];
+char				*filters[MAX_COUNTERS];
 
 struct event_symbol {
 	u8		type;
@@ -46,6 +48,8 @@ static struct event_symbol event_symbols[] = {
   { CSW(PAGE_FAULTS_MAJ),	"major-faults",		""		},
   { CSW(CONTEXT_SWITCHES),	"context-switches",	"cs"		},
   { CSW(CPU_MIGRATIONS),	"cpu-migrations",	"migrations"	},
+  { CSW(ALIGNMENT_FAULTS),	"alignment-faults",	""		},
+  { CSW(EMULATION_FAULTS),	"emulation-faults",	""		},
 };
 
 #define __PERF_EVENT_FIELD(config, name) \
@@ -74,6 +78,8 @@ static const char *sw_event_names[] = {
 	"CPU-migrations",
 	"minor-faults",
 	"major-faults",
+	"alignment-faults",
+	"emulation-faults",
 };
 
 #define MAX_ALIASES 8
@@ -148,16 +154,6 @@ static int tp_event_has_id(struct dirent *sys_dir, struct dirent *evt_dir)
 
 #define MAX_EVENT_LENGTH 512
 
-int valid_debugfs_mount(const char *debugfs)
-{
-	struct statfs st_fs;
-
-	if (statfs(debugfs, &st_fs) < 0)
-		return -ENOENT;
-	else if (st_fs.f_type != (long) DEBUGFS_MAGIC)
-		return -ENOENT;
-	return 0;
-}
 
 struct tracepoint_path *tracepoint_id_to_path(u64 config)
 {
@@ -170,7 +166,7 @@ struct tracepoint_path *tracepoint_id_to_path(u64 config)
 	char evt_path[MAXPATHLEN];
 	char dir_path[MAXPATHLEN];
 
-	if (valid_debugfs_mount(debugfs_path))
+	if (debugfs_valid_mountpoint(debugfs_path))
 		return NULL;
 
 	sys_dir = opendir(debugfs_path);
@@ -201,7 +197,7 @@ struct tracepoint_path *tracepoint_id_to_path(u64 config)
 			if (id == config) {
 				closedir(evt_dir);
 				closedir(sys_dir);
-				path = calloc(1, sizeof(path));
+				path = zalloc(sizeof(path));
 				path->system = malloc(MAX_EVENT_LENGTH);
 				if (!path->system) {
 					free(path);
@@ -509,7 +505,7 @@ static enum event_result parse_tracepoint_event(const char **strp,
 	char sys_name[MAX_EVENT_LENGTH];
 	unsigned int sys_length, evt_length;
 
-	if (valid_debugfs_mount(debugfs_path))
+	if (debugfs_valid_mountpoint(debugfs_path))
 		return 0;
 
 	evt_name = strchr(*strp, ':');
@@ -544,6 +540,81 @@ static enum event_result parse_tracepoint_event(const char **strp,
 						     attr, strp);
 }
 
+static enum event_result
+parse_breakpoint_type(const char *type, const char **strp,
+		      struct perf_event_attr *attr)
+{
+	int i;
+
+	for (i = 0; i < 3; i++) {
+		if (!type[i])
+			break;
+
+		switch (type[i]) {
+		case 'r':
+			attr->bp_type |= HW_BREAKPOINT_R;
+			break;
+		case 'w':
+			attr->bp_type |= HW_BREAKPOINT_W;
+			break;
+		case 'x':
+			attr->bp_type |= HW_BREAKPOINT_X;
+			break;
+		default:
+			return EVT_FAILED;
+		}
+	}
+	if (!attr->bp_type) /* Default */
+		attr->bp_type = HW_BREAKPOINT_R | HW_BREAKPOINT_W;
+
+	*strp = type + i;
+
+	return EVT_HANDLED;
+}
+
+static enum event_result
+parse_breakpoint_event(const char **strp, struct perf_event_attr *attr)
+{
+	const char *target;
+	const char *type;
+	char *endaddr;
+	u64 addr;
+	enum event_result err;
+
+	target = strchr(*strp, ':');
+	if (!target)
+		return EVT_FAILED;
+
+	if (strncmp(*strp, "mem", target - *strp) != 0)
+		return EVT_FAILED;
+
+	target++;
+
+	addr = strtoull(target, &endaddr, 0);
+	if (target == endaddr)
+		return EVT_FAILED;
+
+	attr->bp_addr = addr;
+	*strp = endaddr;
+
+	type = strchr(target, ':');
+
+	/* If no type is defined, just rw as default */
+	if (!type) {
+		attr->bp_type = HW_BREAKPOINT_R | HW_BREAKPOINT_W;
+	} else {
+		err = parse_breakpoint_type(++type, strp, attr);
+		if (err == EVT_FAILED)
+			return EVT_FAILED;
+	}
+
+	/* We should find a nice way to override the access type */
+	attr->bp_len = HW_BREAKPOINT_LEN_4;
+	attr->type = PERF_TYPE_BREAKPOINT;
+
+	return EVT_HANDLED;
+}
+
 static int check_events(const char *str, unsigned int i)
 {
 	int n;
@@ -677,6 +748,12 @@ parse_event_symbols(const char **str, struct perf_event_attr *attr)
 	if (ret != EVT_FAILED)
 		goto modifier;
 
+	ret = parse_breakpoint_event(str, attr);
+	if (ret != EVT_FAILED)
+		goto modifier;
+
+	fprintf(stderr, "invalid or unsupported event: '%s'\n", *str);
+	fprintf(stderr, "Run 'perf list' for a list of valid events\n");
 	return EVT_FAILED;
 
 modifier:
@@ -708,7 +785,6 @@ static void store_event_type(const char *orgname)
 	perf_header__push_event(id, orgname);
 }
 
-
 int parse_events(const struct option *opt __used, const char *str, int unset __used)
 {
 	struct perf_event_attr attr;
@@ -745,6 +821,28 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u
 	return 0;
 }
 
+int parse_filter(const struct option *opt __used, const char *str,
+		 int unset __used)
+{
+	int i = nr_counters - 1;
+	int len = strlen(str);
+
+	if (i < 0 || attrs[i].type != PERF_TYPE_TRACEPOINT) {
+		fprintf(stderr,
+			"-F option should follow a -e tracepoint option\n");
+		return -1;
+	}
+
+	filters[i] = malloc(len + 1);
+	if (!filters[i]) {
+		fprintf(stderr, "not enough memory to hold filter string\n");
+		return -1;
+	}
+	strcpy(filters[i], str);
+
+	return 0;
+}
+
 static const char * const event_type_descriptors[] = {
 	"",
 	"Hardware event",
@@ -764,7 +862,7 @@ static void print_tracepoint_events(void)
 	char evt_path[MAXPATHLEN];
 	char dir_path[MAXPATHLEN];
 
-	if (valid_debugfs_mount(debugfs_path))
+	if (debugfs_valid_mountpoint(debugfs_path))
 		return;
 
 	sys_dir = opendir(debugfs_path);
@@ -782,7 +880,7 @@ static void print_tracepoint_events(void)
 		for_each_event(sys_dirent, evt_dir, evt_dirent, evt_next) {
 			snprintf(evt_path, MAXPATHLEN, "%s:%s",
 				 sys_dirent.d_name, evt_dirent.d_name);
-			fprintf(stderr, "  %-42s [%s]\n", evt_path,
+			printf("  %-42s [%s]\n", evt_path,
 				event_type_descriptors[PERF_TYPE_TRACEPOINT+1]);
 		}
 		closedir(evt_dir);
@@ -799,8 +897,8 @@ void print_events(void)
 	unsigned int i, type, op, prev_type = -1;
 	char name[40];
 
-	fprintf(stderr, "\n");
-	fprintf(stderr, "List of pre-defined events (to be used in -e):\n");
+	printf("\n");
+	printf("List of pre-defined events (to be used in -e):\n");
 
 	for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) {
 		type = syms->type + 1;
@@ -808,19 +906,19 @@ void print_events(void)
 			type = 0;
 
 		if (type != prev_type)
-			fprintf(stderr, "\n");
+			printf("\n");
 
 		if (strlen(syms->alias))
 			sprintf(name, "%s OR %s", syms->symbol, syms->alias);
 		else
 			strcpy(name, syms->symbol);
-		fprintf(stderr, "  %-42s [%s]\n", name,
+		printf("  %-42s [%s]\n", name,
 			event_type_descriptors[type]);
 
 		prev_type = type;
 	}
 
-	fprintf(stderr, "\n");
+	printf("\n");
 	for (type = 0; type < PERF_COUNT_HW_CACHE_MAX; type++) {
 		for (op = 0; op < PERF_COUNT_HW_CACHE_OP_MAX; op++) {
 			/* skip invalid cache type */
@@ -828,17 +926,20 @@ void print_events(void)
 				continue;
 
 			for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
-				fprintf(stderr, "  %-42s [%s]\n",
+				printf("  %-42s [%s]\n",
 					event_cache_name(type, op, i),
 					event_type_descriptors[4]);
 			}
 		}
 	}
 
-	fprintf(stderr, "\n");
-	fprintf(stderr, "  %-42s [raw hardware event descriptor]\n",
+	printf("\n");
+	printf("  %-42s [raw hardware event descriptor]\n",
 		"rNNN");
-	fprintf(stderr, "\n");
+	printf("\n");
+
+	printf("  %-42s [hardware breakpoint]\n", "mem:<addr>[:access]");
+	printf("\n");
 
 	print_tracepoint_events();
 
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index 30c608112845..b8c1f64bc935 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -1,5 +1,5 @@
-#ifndef _PARSE_EVENTS_H
-#define _PARSE_EVENTS_H
+#ifndef __PERF_PARSE_EVENTS_H
+#define __PERF_PARSE_EVENTS_H
 /*
  * Parse symbolic events/counts passed in as options:
  */
@@ -17,11 +17,13 @@ extern struct tracepoint_path *tracepoint_id_to_path(u64 config);
 extern int			nr_counters;
 
 extern struct perf_event_attr attrs[MAX_COUNTERS];
+extern char *filters[MAX_COUNTERS];
 
 extern const char *event_name(int ctr);
 extern const char *__event_name(int type, u64 config);
 
 extern int parse_events(const struct option *opt, const char *str, int unset);
+extern int parse_filter(const struct option *opt, const char *str, int unset);
 
 #define EVENTS_HELP_MAX (128*1024)
 
@@ -31,4 +33,4 @@ extern char debugfs_path[];
 extern int valid_debugfs_mount(const char *debugfs);
 
 
-#endif /* _PARSE_EVENTS_H */
+#endif /* __PERF_PARSE_EVENTS_H */
diff --git a/tools/perf/util/parse-options.h b/tools/perf/util/parse-options.h
index 2ee248ff27e5..948805af43c2 100644
--- a/tools/perf/util/parse-options.h
+++ b/tools/perf/util/parse-options.h
@@ -1,5 +1,5 @@
-#ifndef PARSE_OPTIONS_H
-#define PARSE_OPTIONS_H
+#ifndef __PERF_PARSE_OPTIONS_H
+#define __PERF_PARSE_OPTIONS_H
 
 enum parse_opt_type {
 	/* special types */
@@ -174,4 +174,4 @@ extern int parse_opt_verbosity_cb(const struct option *, const char *, int);
 
 extern const char *parse_options_fix_filename(const char *prefix, const char *file);
 
-#endif
+#endif /* __PERF_PARSE_OPTIONS_H */
diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
new file mode 100644
index 000000000000..cd7fbda5e2a5
--- /dev/null
+++ b/tools/perf/util/probe-event.c
@@ -0,0 +1,484 @@
+/*
+ * probe-event.c : perf-probe definition to kprobe_events format converter
+ *
+ * Written by Masami Hiramatsu <mhiramat@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ */
+
+#define _GNU_SOURCE
+#include <sys/utsname.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+
+#undef _GNU_SOURCE
+#include "event.h"
+#include "string.h"
+#include "strlist.h"
+#include "debug.h"
+#include "parse-events.h"  /* For debugfs_path */
+#include "probe-event.h"
+
+#define MAX_CMDLEN 256
+#define MAX_PROBE_ARGS 128
+#define PERFPROBE_GROUP "probe"
+
+#define semantic_error(msg ...) die("Semantic error :" msg)
+
+/* If there is no space to write, returns -E2BIG. */
+static int e_snprintf(char *str, size_t size, const char *format, ...)
+{
+	int ret;
+	va_list ap;
+	va_start(ap, format);
+	ret = vsnprintf(str, size, format, ap);
+	va_end(ap);
+	if (ret >= (int)size)
+		ret = -E2BIG;
+	return ret;
+}
+
+/* Parse probepoint definition. */
+static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp)
+{
+	char *ptr, *tmp;
+	char c, nc = 0;
+	/*
+	 * <Syntax>
+	 * perf probe SRC:LN
+	 * perf probe FUNC[+OFFS|%return][@SRC]
+	 */
+
+	ptr = strpbrk(arg, ":+@%");
+	if (ptr) {
+		nc = *ptr;
+		*ptr++ = '\0';
+	}
+
+	/* Check arg is function or file and copy it */
+	if (strchr(arg, '.'))	/* File */
+		pp->file = strdup(arg);
+	else			/* Function */
+		pp->function = strdup(arg);
+	DIE_IF(pp->file == NULL && pp->function == NULL);
+
+	/* Parse other options */
+	while (ptr) {
+		arg = ptr;
+		c = nc;
+		ptr = strpbrk(arg, ":+@%");
+		if (ptr) {
+			nc = *ptr;
+			*ptr++ = '\0';
+		}
+		switch (c) {
+		case ':':	/* Line number */
+			pp->line = strtoul(arg, &tmp, 0);
+			if (*tmp != '\0')
+				semantic_error("There is non-digit charactor"
+						" in line number.");
+			break;
+		case '+':	/* Byte offset from a symbol */
+			pp->offset = strtoul(arg, &tmp, 0);
+			if (*tmp != '\0')
+				semantic_error("There is non-digit charactor"
+						" in offset.");
+			break;
+		case '@':	/* File name */
+			if (pp->file)
+				semantic_error("SRC@SRC is not allowed.");
+			pp->file = strdup(arg);
+			DIE_IF(pp->file == NULL);
+			if (ptr)
+				semantic_error("@SRC must be the last "
+					       "option.");
+			break;
+		case '%':	/* Probe places */
+			if (strcmp(arg, "return") == 0) {
+				pp->retprobe = 1;
+			} else	/* Others not supported yet */
+				semantic_error("%%%s is not supported.", arg);
+			break;
+		default:
+			DIE_IF("Program has a bug.");
+			break;
+		}
+	}
+
+	/* Exclusion check */
+	if (pp->line && pp->offset)
+		semantic_error("Offset can't be used with line number.");
+
+	if (!pp->line && pp->file && !pp->function)
+		semantic_error("File always requires line number.");
+
+	if (pp->offset && !pp->function)
+		semantic_error("Offset requires an entry function.");
+
+	if (pp->retprobe && !pp->function)
+		semantic_error("Return probe requires an entry function.");
+
+	if ((pp->offset || pp->line) && pp->retprobe)
+		semantic_error("Offset/Line can't be used with return probe.");
+
+	pr_debug("symbol:%s file:%s line:%d offset:%d, return:%d\n",
+		 pp->function, pp->file, pp->line, pp->offset, pp->retprobe);
+}
+
+/* Parse perf-probe event definition */
+int parse_perf_probe_event(const char *str, struct probe_point *pp)
+{
+	char **argv;
+	int argc, i, need_dwarf = 0;
+
+	argv = argv_split(str, &argc);
+	if (!argv)
+		die("argv_split failed.");
+	if (argc > MAX_PROBE_ARGS + 1)
+		semantic_error("Too many arguments");
+
+	/* Parse probe point */
+	parse_perf_probe_probepoint(argv[0], pp);
+	if (pp->file || pp->line)
+		need_dwarf = 1;
+
+	/* Copy arguments and ensure return probe has no C argument */
+	pp->nr_args = argc - 1;
+	pp->args = zalloc(sizeof(char *) * pp->nr_args);
+	for (i = 0; i < pp->nr_args; i++) {
+		pp->args[i] = strdup(argv[i + 1]);
+		if (!pp->args[i])
+			die("Failed to copy argument.");
+		if (is_c_varname(pp->args[i])) {
+			if (pp->retprobe)
+				semantic_error("You can't specify local"
+						" variable for kretprobe");
+			need_dwarf = 1;
+		}
+	}
+
+	argv_free(argv);
+	return need_dwarf;
+}
+
+/* Parse kprobe_events event into struct probe_point */
+void parse_trace_kprobe_event(const char *str, char **group, char **event,
+			      struct probe_point *pp)
+{
+	char pr;
+	char *p;
+	int ret, i, argc;
+	char **argv;
+
+	pr_debug("Parsing kprobe_events: %s\n", str);
+	argv = argv_split(str, &argc);
+	if (!argv)
+		die("argv_split failed.");
+	if (argc < 2)
+		semantic_error("Too less arguments.");
+
+	/* Scan event and group name. */
+	ret = sscanf(argv[0], "%c:%a[^/ \t]/%a[^ \t]",
+		     &pr, (float *)(void *)group, (float *)(void *)event);
+	if (ret != 3)
+		semantic_error("Failed to parse event name: %s", argv[0]);
+	pr_debug("Group:%s Event:%s probe:%c\n", *group, *event, pr);
+
+	if (!pp)
+		goto end;
+
+	pp->retprobe = (pr == 'r');
+
+	/* Scan function name and offset */
+	ret = sscanf(argv[1], "%a[^+]+%d", (float *)(void *)&pp->function, &pp->offset);
+	if (ret == 1)
+		pp->offset = 0;
+
+	/* kprobe_events doesn't have this information */
+	pp->line = 0;
+	pp->file = NULL;
+
+	pp->nr_args = argc - 2;
+	pp->args = zalloc(sizeof(char *) * pp->nr_args);
+	for (i = 0; i < pp->nr_args; i++) {
+		p = strchr(argv[i + 2], '=');
+		if (p)	/* We don't need which register is assigned. */
+			*p = '\0';
+		pp->args[i] = strdup(argv[i + 2]);
+		if (!pp->args[i])
+			die("Failed to copy argument.");
+	}
+
+end:
+	argv_free(argv);
+}
+
+int synthesize_perf_probe_event(struct probe_point *pp)
+{
+	char *buf;
+	char offs[64] = "", line[64] = "";
+	int i, len, ret;
+
+	pp->probes[0] = buf = zalloc(MAX_CMDLEN);
+	if (!buf)
+		die("Failed to allocate memory by zalloc.");
+	if (pp->offset) {
+		ret = e_snprintf(offs, 64, "+%d", pp->offset);
+		if (ret <= 0)
+			goto error;
+	}
+	if (pp->line) {
+		ret = e_snprintf(line, 64, ":%d", pp->line);
+		if (ret <= 0)
+			goto error;
+	}
+
+	if (pp->function)
+		ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s", pp->function,
+				 offs, pp->retprobe ? "%return" : "", line);
+	else
+		ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s", pp->file, line);
+	if (ret <= 0)
+		goto error;
+	len = ret;
+
+	for (i = 0; i < pp->nr_args; i++) {
+		ret = e_snprintf(&buf[len], MAX_CMDLEN - len, " %s",
+				 pp->args[i]);
+		if (ret <= 0)
+			goto error;
+		len += ret;
+	}
+	pp->found = 1;
+
+	return pp->found;
+error:
+	free(pp->probes[0]);
+
+	return ret;
+}
+
+int synthesize_trace_kprobe_event(struct probe_point *pp)
+{
+	char *buf;
+	int i, len, ret;
+
+	pp->probes[0] = buf = zalloc(MAX_CMDLEN);
+	if (!buf)
+		die("Failed to allocate memory by zalloc.");
+	ret = e_snprintf(buf, MAX_CMDLEN, "%s+%d", pp->function, pp->offset);
+	if (ret <= 0)
+		goto error;
+	len = ret;
+
+	for (i = 0; i < pp->nr_args; i++) {
+		ret = e_snprintf(&buf[len], MAX_CMDLEN - len, " %s",
+				 pp->args[i]);
+		if (ret <= 0)
+			goto error;
+		len += ret;
+	}
+	pp->found = 1;
+
+	return pp->found;
+error:
+	free(pp->probes[0]);
+
+	return ret;
+}
+
+static int open_kprobe_events(int flags, int mode)
+{
+	char buf[PATH_MAX];
+	int ret;
+
+	ret = e_snprintf(buf, PATH_MAX, "%s/../kprobe_events", debugfs_path);
+	if (ret < 0)
+		die("Failed to make kprobe_events path.");
+
+	ret = open(buf, flags, mode);
+	if (ret < 0) {
+		if (errno == ENOENT)
+			die("kprobe_events file does not exist -"
+			    " please rebuild with CONFIG_KPROBE_TRACER.");
+		else
+			die("Could not open kprobe_events file: %s",
+			    strerror(errno));
+	}
+	return ret;
+}
+
+/* Get raw string list of current kprobe_events */
+static struct strlist *get_trace_kprobe_event_rawlist(int fd)
+{
+	int ret, idx;
+	FILE *fp;
+	char buf[MAX_CMDLEN];
+	char *p;
+	struct strlist *sl;
+
+	sl = strlist__new(true, NULL);
+
+	fp = fdopen(dup(fd), "r");
+	while (!feof(fp)) {
+		p = fgets(buf, MAX_CMDLEN, fp);
+		if (!p)
+			break;
+
+		idx = strlen(p) - 1;
+		if (p[idx] == '\n')
+			p[idx] = '\0';
+		ret = strlist__add(sl, buf);
+		if (ret < 0)
+			die("strlist__add failed: %s", strerror(-ret));
+	}
+	fclose(fp);
+
+	return sl;
+}
+
+/* Free and zero clear probe_point */
+static void clear_probe_point(struct probe_point *pp)
+{
+	int i;
+
+	if (pp->function)
+		free(pp->function);
+	if (pp->file)
+		free(pp->file);
+	for (i = 0; i < pp->nr_args; i++)
+		free(pp->args[i]);
+	if (pp->args)
+		free(pp->args);
+	for (i = 0; i < pp->found; i++)
+		free(pp->probes[i]);
+	memset(pp, 0, sizeof(pp));
+}
+
+/* List up current perf-probe events */
+void show_perf_probe_events(void)
+{
+	unsigned int i;
+	int fd;
+	char *group, *event;
+	struct probe_point pp;
+	struct strlist *rawlist;
+	struct str_node *ent;
+
+	fd = open_kprobe_events(O_RDONLY, 0);
+	rawlist = get_trace_kprobe_event_rawlist(fd);
+	close(fd);
+
+	for (i = 0; i < strlist__nr_entries(rawlist); i++) {
+		ent = strlist__entry(rawlist, i);
+		parse_trace_kprobe_event(ent->s, &group, &event, &pp);
+		synthesize_perf_probe_event(&pp);
+		printf("[%s:%s]\t%s\n", group, event, pp.probes[0]);
+		free(group);
+		free(event);
+		clear_probe_point(&pp);
+	}
+
+	strlist__delete(rawlist);
+}
+
+/* Get current perf-probe event names */
+static struct strlist *get_perf_event_names(int fd)
+{
+	unsigned int i;
+	char *group, *event;
+	struct strlist *sl, *rawlist;
+	struct str_node *ent;
+
+	rawlist = get_trace_kprobe_event_rawlist(fd);
+
+	sl = strlist__new(false, NULL);
+	for (i = 0; i < strlist__nr_entries(rawlist); i++) {
+		ent = strlist__entry(rawlist, i);
+		parse_trace_kprobe_event(ent->s, &group, &event, NULL);
+		strlist__add(sl, event);
+		free(group);
+	}
+
+	strlist__delete(rawlist);
+
+	return sl;
+}
+
+static int write_trace_kprobe_event(int fd, const char *buf)
+{
+	int ret;
+
+	ret = write(fd, buf, strlen(buf));
+	if (ret <= 0)
+		die("Failed to create event.");
+	else
+		printf("Added new event: %s\n", buf);
+
+	return ret;
+}
+
+static void get_new_event_name(char *buf, size_t len, const char *base,
+			       struct strlist *namelist)
+{
+	int i, ret;
+	for (i = 0; i < MAX_EVENT_INDEX; i++) {
+		ret = e_snprintf(buf, len, "%s_%d", base, i);
+		if (ret < 0)
+			die("snprintf() failed: %s", strerror(-ret));
+		if (!strlist__has_entry(namelist, buf))
+			break;
+	}
+	if (i == MAX_EVENT_INDEX)
+		die("Too many events are on the same function.");
+}
+
+void add_trace_kprobe_events(struct probe_point *probes, int nr_probes)
+{
+	int i, j, fd;
+	struct probe_point *pp;
+	char buf[MAX_CMDLEN];
+	char event[64];
+	struct strlist *namelist;
+
+	fd = open_kprobe_events(O_RDWR, O_APPEND);
+	/* Get current event names */
+	namelist = get_perf_event_names(fd);
+
+	for (j = 0; j < nr_probes; j++) {
+		pp = probes + j;
+		for (i = 0; i < pp->found; i++) {
+			/* Get an unused new event name */
+			get_new_event_name(event, 64, pp->function, namelist);
+			snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s\n",
+				 pp->retprobe ? 'r' : 'p',
+				 PERFPROBE_GROUP, event,
+				 pp->probes[i]);
+			write_trace_kprobe_event(fd, buf);
+			/* Add added event name to namelist */
+			strlist__add(namelist, event);
+		}
+	}
+	close(fd);
+}
diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h
new file mode 100644
index 000000000000..0c6fe56fe38a
--- /dev/null
+++ b/tools/perf/util/probe-event.h
@@ -0,0 +1,18 @@
+#ifndef _PROBE_EVENT_H
+#define _PROBE_EVENT_H
+
+#include "probe-finder.h"
+#include "strlist.h"
+
+extern int parse_perf_probe_event(const char *str, struct probe_point *pp);
+extern int synthesize_perf_probe_event(struct probe_point *pp);
+extern void parse_trace_kprobe_event(const char *str, char **group,
+				     char **event, struct probe_point *pp);
+extern int synthesize_trace_kprobe_event(struct probe_point *pp);
+extern void add_trace_kprobe_events(struct probe_point *probes, int nr_probes);
+extern void show_perf_probe_events(void);
+
+/* Maximum index number of event-name postfix */
+#define MAX_EVENT_INDEX	1024
+
+#endif /*_PROBE_EVENT_H */
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
new file mode 100644
index 000000000000..293cdfc1b8ca
--- /dev/null
+++ b/tools/perf/util/probe-finder.c
@@ -0,0 +1,732 @@
+/*
+ * probe-finder.c : C expression to kprobe event converter
+ *
+ * Written by Masami Hiramatsu <mhiramat@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ */
+
+#include <sys/utsname.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <getopt.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+#include <ctype.h>
+
+#include "event.h"
+#include "debug.h"
+#include "util.h"
+#include "probe-finder.h"
+
+
+/* Dwarf_Die Linkage to parent Die */
+struct die_link {
+	struct die_link *parent;	/* Parent die */
+	Dwarf_Die die;			/* Current die */
+};
+
+static Dwarf_Debug __dw_debug;
+static Dwarf_Error __dw_error;
+
+/*
+ * Generic dwarf analysis helpers
+ */
+
+#define X86_32_MAX_REGS 8
+const char *x86_32_regs_table[X86_32_MAX_REGS] = {
+	"%ax",
+	"%cx",
+	"%dx",
+	"%bx",
+	"$stack",	/* Stack address instead of %sp */
+	"%bp",
+	"%si",
+	"%di",
+};
+
+#define X86_64_MAX_REGS 16
+const char *x86_64_regs_table[X86_64_MAX_REGS] = {
+	"%ax",
+	"%dx",
+	"%cx",
+	"%bx",
+	"%si",
+	"%di",
+	"%bp",
+	"%sp",
+	"%r8",
+	"%r9",
+	"%r10",
+	"%r11",
+	"%r12",
+	"%r13",
+	"%r14",
+	"%r15",
+};
+
+/* TODO: switching by dwarf address size */
+#ifdef __x86_64__
+#define ARCH_MAX_REGS X86_64_MAX_REGS
+#define arch_regs_table x86_64_regs_table
+#else
+#define ARCH_MAX_REGS X86_32_MAX_REGS
+#define arch_regs_table x86_32_regs_table
+#endif
+
+/* Return architecture dependent register string (for kprobe-tracer) */
+static const char *get_arch_regstr(unsigned int n)
+{
+	return (n <= ARCH_MAX_REGS) ? arch_regs_table[n] : NULL;
+}
+
+/*
+ * Compare the tail of two strings.
+ * Return 0 if whole of either string is same as another's tail part.
+ */
+static int strtailcmp(const char *s1, const char *s2)
+{
+	int i1 = strlen(s1);
+	int i2 = strlen(s2);
+	while (--i1 > 0 && --i2 > 0) {
+		if (s1[i1] != s2[i2])
+			return s1[i1] - s2[i2];
+	}
+	return 0;
+}
+
+/* Find the fileno of the target file. */
+static Dwarf_Unsigned cu_find_fileno(Dwarf_Die cu_die, const char *fname)
+{
+	Dwarf_Signed cnt, i;
+	Dwarf_Unsigned found = 0;
+	char **srcs;
+	int ret;
+
+	if (!fname)
+		return 0;
+
+	ret = dwarf_srcfiles(cu_die, &srcs, &cnt, &__dw_error);
+	if (ret == DW_DLV_OK) {
+		for (i = 0; i < cnt && !found; i++) {
+			if (strtailcmp(srcs[i], fname) == 0)
+				found = i + 1;
+			dwarf_dealloc(__dw_debug, srcs[i], DW_DLA_STRING);
+		}
+		for (; i < cnt; i++)
+			dwarf_dealloc(__dw_debug, srcs[i], DW_DLA_STRING);
+		dwarf_dealloc(__dw_debug, srcs, DW_DLA_LIST);
+	}
+	if (found)
+		pr_debug("found fno: %d\n", (int)found);
+	return found;
+}
+
+/* Compare diename and tname */
+static int die_compare_name(Dwarf_Die dw_die, const char *tname)
+{
+	char *name;
+	int ret;
+	ret = dwarf_diename(dw_die, &name, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_OK) {
+		ret = strcmp(tname, name);
+		dwarf_dealloc(__dw_debug, name, DW_DLA_STRING);
+	} else
+		ret = -1;
+	return ret;
+}
+
+/* Check the address is in the subprogram(function). */
+static int die_within_subprogram(Dwarf_Die sp_die, Dwarf_Addr addr,
+				 Dwarf_Signed *offs)
+{
+	Dwarf_Addr lopc, hipc;
+	int ret;
+
+	/* TODO: check ranges */
+	ret = dwarf_lowpc(sp_die, &lopc, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_NO_ENTRY)
+		return 0;
+	ret = dwarf_highpc(sp_die, &hipc, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	if (lopc <= addr && addr < hipc) {
+		*offs = addr - lopc;
+		return 1;
+	} else
+		return 0;
+}
+
+/* Check the die is inlined function */
+static Dwarf_Bool die_inlined_subprogram(Dwarf_Die dw_die)
+{
+	/* TODO: check strictly */
+	Dwarf_Bool inl;
+	int ret;
+
+	ret = dwarf_hasattr(dw_die, DW_AT_inline, &inl, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	return inl;
+}
+
+/* Get the offset of abstruct_origin */
+static Dwarf_Off die_get_abstract_origin(Dwarf_Die dw_die)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Off cu_offs;
+	int ret;
+
+	ret = dwarf_attr(dw_die, DW_AT_abstract_origin, &attr, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	ret = dwarf_formref(attr, &cu_offs, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+	return cu_offs;
+}
+
+/* Get entry pc(or low pc, 1st entry of ranges)  of the die */
+static Dwarf_Addr die_get_entrypc(Dwarf_Die dw_die)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Addr addr;
+	Dwarf_Off offs;
+	Dwarf_Ranges *ranges;
+	Dwarf_Signed cnt;
+	int ret;
+
+	/* Try to get entry pc */
+	ret = dwarf_attr(dw_die, DW_AT_entry_pc, &attr, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_OK) {
+		ret = dwarf_formaddr(attr, &addr, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+		dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+		return addr;
+	}
+
+	/* Try to get low pc */
+	ret = dwarf_lowpc(dw_die, &addr, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_OK)
+		return addr;
+
+	/* Try to get ranges */
+	ret = dwarf_attr(dw_die, DW_AT_ranges, &attr, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	ret = dwarf_formref(attr, &offs, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	ret = dwarf_get_ranges(__dw_debug, offs, &ranges, &cnt, NULL,
+				&__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	addr = ranges[0].dwr_addr1;
+	dwarf_ranges_dealloc(__dw_debug, ranges, cnt);
+	return addr;
+}
+
+/*
+ * Search a Die from Die tree.
+ * Note: cur_link->die should be deallocated in this function.
+ */
+static int __search_die_tree(struct die_link *cur_link,
+			     int (*die_cb)(struct die_link *, void *),
+			     void *data)
+{
+	Dwarf_Die new_die;
+	struct die_link new_link;
+	int ret;
+
+	if (!die_cb)
+		return 0;
+
+	/* Check current die */
+	while (!(ret = die_cb(cur_link, data))) {
+		/* Check child die */
+		ret = dwarf_child(cur_link->die, &new_die, &__dw_error);
+		DIE_IF(ret == DW_DLV_ERROR);
+		if (ret == DW_DLV_OK) {
+			new_link.parent = cur_link;
+			new_link.die = new_die;
+			ret = __search_die_tree(&new_link, die_cb, data);
+			if (ret)
+				break;
+		}
+
+		/* Move to next sibling */
+		ret = dwarf_siblingof(__dw_debug, cur_link->die, &new_die,
+				      &__dw_error);
+		DIE_IF(ret == DW_DLV_ERROR);
+		dwarf_dealloc(__dw_debug, cur_link->die, DW_DLA_DIE);
+		cur_link->die = new_die;
+		if (ret == DW_DLV_NO_ENTRY)
+			return 0;
+	}
+	dwarf_dealloc(__dw_debug, cur_link->die, DW_DLA_DIE);
+	return ret;
+}
+
+/* Search a die in its children's die tree */
+static int search_die_from_children(Dwarf_Die parent_die,
+				    int (*die_cb)(struct die_link *, void *),
+				    void *data)
+{
+	struct die_link new_link;
+	int ret;
+
+	new_link.parent = NULL;
+	ret = dwarf_child(parent_die, &new_link.die, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_OK)
+		return __search_die_tree(&new_link, die_cb, data);
+	else
+		return 0;
+}
+
+/* Find a locdesc corresponding to the address */
+static int attr_get_locdesc(Dwarf_Attribute attr, Dwarf_Locdesc *desc,
+			    Dwarf_Addr addr)
+{
+	Dwarf_Signed lcnt;
+	Dwarf_Locdesc **llbuf;
+	int ret, i;
+
+	ret = dwarf_loclist_n(attr, &llbuf, &lcnt, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	ret = DW_DLV_NO_ENTRY;
+	for (i = 0; i < lcnt; ++i) {
+		if (llbuf[i]->ld_lopc <= addr &&
+		    llbuf[i]->ld_hipc > addr) {
+			memcpy(desc, llbuf[i], sizeof(Dwarf_Locdesc));
+			desc->ld_s =
+				malloc(sizeof(Dwarf_Loc) * llbuf[i]->ld_cents);
+			DIE_IF(desc->ld_s == NULL);
+			memcpy(desc->ld_s, llbuf[i]->ld_s,
+				sizeof(Dwarf_Loc) * llbuf[i]->ld_cents);
+			ret = DW_DLV_OK;
+			break;
+		}
+		dwarf_dealloc(__dw_debug, llbuf[i]->ld_s, DW_DLA_LOC_BLOCK);
+		dwarf_dealloc(__dw_debug, llbuf[i], DW_DLA_LOCDESC);
+	}
+	/* Releasing loop */
+	for (; i < lcnt; ++i) {
+		dwarf_dealloc(__dw_debug, llbuf[i]->ld_s, DW_DLA_LOC_BLOCK);
+		dwarf_dealloc(__dw_debug, llbuf[i], DW_DLA_LOCDESC);
+	}
+	dwarf_dealloc(__dw_debug, llbuf, DW_DLA_LIST);
+	return ret;
+}
+
+/* Get decl_file attribute value (file number) */
+static Dwarf_Unsigned die_get_decl_file(Dwarf_Die sp_die)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Unsigned fno;
+	int ret;
+
+	ret = dwarf_attr(sp_die, DW_AT_decl_file, &attr, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_formudata(attr, &fno, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+	return fno;
+}
+
+/* Get decl_line attribute value (line number) */
+static Dwarf_Unsigned die_get_decl_line(Dwarf_Die sp_die)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Unsigned lno;
+	int ret;
+
+	ret = dwarf_attr(sp_die, DW_AT_decl_line, &attr, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_formudata(attr, &lno, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+	return lno;
+}
+
+/*
+ * Probe finder related functions
+ */
+
+/* Show a location */
+static void show_location(Dwarf_Loc *loc, struct probe_finder *pf)
+{
+	Dwarf_Small op;
+	Dwarf_Unsigned regn;
+	Dwarf_Signed offs;
+	int deref = 0, ret;
+	const char *regs;
+
+	op = loc->lr_atom;
+
+	/* If this is based on frame buffer, set the offset */
+	if (op == DW_OP_fbreg) {
+		deref = 1;
+		offs = (Dwarf_Signed)loc->lr_number;
+		op = pf->fbloc.ld_s[0].lr_atom;
+		loc = &pf->fbloc.ld_s[0];
+	} else
+		offs = 0;
+
+	if (op >= DW_OP_breg0 && op <= DW_OP_breg31) {
+		regn = op - DW_OP_breg0;
+		offs += (Dwarf_Signed)loc->lr_number;
+		deref = 1;
+	} else if (op >= DW_OP_reg0 && op <= DW_OP_reg31) {
+		regn = op - DW_OP_reg0;
+	} else if (op == DW_OP_bregx) {
+		regn = loc->lr_number;
+		offs += (Dwarf_Signed)loc->lr_number2;
+		deref = 1;
+	} else if (op == DW_OP_regx) {
+		regn = loc->lr_number;
+	} else
+		die("Dwarf_OP %d is not supported.\n", op);
+
+	regs = get_arch_regstr(regn);
+	if (!regs)
+		die("%lld exceeds max register number.\n", regn);
+
+	if (deref)
+		ret = snprintf(pf->buf, pf->len,
+				 " %s=%+lld(%s)", pf->var, offs, regs);
+	else
+		ret = snprintf(pf->buf, pf->len, " %s=%s", pf->var, regs);
+	DIE_IF(ret < 0);
+	DIE_IF(ret >= pf->len);
+}
+
+/* Show a variables in kprobe event format */
+static void show_variable(Dwarf_Die vr_die, struct probe_finder *pf)
+{
+	Dwarf_Attribute attr;
+	Dwarf_Locdesc ld;
+	int ret;
+
+	ret = dwarf_attr(vr_die, DW_AT_location, &attr, &__dw_error);
+	if (ret != DW_DLV_OK)
+		goto error;
+	ret = attr_get_locdesc(attr, &ld, (pf->addr - pf->cu_base));
+	if (ret != DW_DLV_OK)
+		goto error;
+	/* TODO? */
+	DIE_IF(ld.ld_cents != 1);
+	show_location(&ld.ld_s[0], pf);
+	free(ld.ld_s);
+	dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+	return ;
+error:
+	die("Failed to find the location of %s at this address.\n"
+	    " Perhaps, it has been optimized out.\n", pf->var);
+}
+
+static int variable_callback(struct die_link *dlink, void *data)
+{
+	struct probe_finder *pf = (struct probe_finder *)data;
+	Dwarf_Half tag;
+	int ret;
+
+	ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if ((tag == DW_TAG_formal_parameter ||
+	     tag == DW_TAG_variable) &&
+	    (die_compare_name(dlink->die, pf->var) == 0)) {
+		show_variable(dlink->die, pf);
+		return 1;
+	}
+	/* TODO: Support struct members and arrays */
+	return 0;
+}
+
+/* Find a variable in a subprogram die */
+static void find_variable(Dwarf_Die sp_die, struct probe_finder *pf)
+{
+	int ret;
+
+	if (!is_c_varname(pf->var)) {
+		/* Output raw parameters */
+		ret = snprintf(pf->buf, pf->len, " %s", pf->var);
+		DIE_IF(ret < 0);
+		DIE_IF(ret >= pf->len);
+		return ;
+	}
+
+	pr_debug("Searching '%s' variable in context.\n", pf->var);
+	/* Search child die for local variables and parameters. */
+	ret = search_die_from_children(sp_die, variable_callback, pf);
+	if (!ret)
+		die("Failed to find '%s' in this function.\n", pf->var);
+}
+
+/* Get a frame base on the address */
+static void get_current_frame_base(Dwarf_Die sp_die, struct probe_finder *pf)
+{
+	Dwarf_Attribute attr;
+	int ret;
+
+	ret = dwarf_attr(sp_die, DW_AT_frame_base, &attr, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+	ret = attr_get_locdesc(attr, &pf->fbloc, (pf->addr - pf->cu_base));
+	DIE_IF(ret != DW_DLV_OK);
+	dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+}
+
+static void free_current_frame_base(struct probe_finder *pf)
+{
+	free(pf->fbloc.ld_s);
+	memset(&pf->fbloc, 0, sizeof(Dwarf_Locdesc));
+}
+
+/* Show a probe point to output buffer */
+static void show_probepoint(Dwarf_Die sp_die, Dwarf_Signed offs,
+			    struct probe_finder *pf)
+{
+	struct probe_point *pp = pf->pp;
+	char *name;
+	char tmp[MAX_PROBE_BUFFER];
+	int ret, i, len;
+
+	/* Output name of probe point */
+	ret = dwarf_diename(sp_die, &name, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (ret == DW_DLV_OK) {
+		ret = snprintf(tmp, MAX_PROBE_BUFFER, "%s+%u", name,
+				(unsigned int)offs);
+		/* Copy the function name if possible */
+		if (!pp->function) {
+			pp->function = strdup(name);
+			pp->offset = offs;
+		}
+		dwarf_dealloc(__dw_debug, name, DW_DLA_STRING);
+	} else {
+		/* This function has no name. */
+		ret = snprintf(tmp, MAX_PROBE_BUFFER, "0x%llx", pf->addr);
+		if (!pp->function) {
+			/* TODO: Use _stext */
+			pp->function = strdup("");
+			pp->offset = (int)pf->addr;
+		}
+	}
+	DIE_IF(ret < 0);
+	DIE_IF(ret >= MAX_PROBE_BUFFER);
+	len = ret;
+	pr_debug("Probe point found: %s\n", tmp);
+
+	/* Find each argument */
+	get_current_frame_base(sp_die, pf);
+	for (i = 0; i < pp->nr_args; i++) {
+		pf->var = pp->args[i];
+		pf->buf = &tmp[len];
+		pf->len = MAX_PROBE_BUFFER - len;
+		find_variable(sp_die, pf);
+		len += strlen(pf->buf);
+	}
+	free_current_frame_base(pf);
+
+	pp->probes[pp->found] = strdup(tmp);
+	pp->found++;
+}
+
+static int probeaddr_callback(struct die_link *dlink, void *data)
+{
+	struct probe_finder *pf = (struct probe_finder *)data;
+	Dwarf_Half tag;
+	Dwarf_Signed offs;
+	int ret;
+
+	ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	/* Check the address is in this subprogram */
+	if (tag == DW_TAG_subprogram &&
+	    die_within_subprogram(dlink->die, pf->addr, &offs)) {
+		show_probepoint(dlink->die, offs, pf);
+		return 1;
+	}
+	return 0;
+}
+
+/* Find probe point from its line number */
+static void find_by_line(struct probe_finder *pf)
+{
+	Dwarf_Signed cnt, i, clm;
+	Dwarf_Line *lines;
+	Dwarf_Unsigned lineno = 0;
+	Dwarf_Addr addr;
+	Dwarf_Unsigned fno;
+	int ret;
+
+	ret = dwarf_srclines(pf->cu_die, &lines, &cnt, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+
+	for (i = 0; i < cnt; i++) {
+		ret = dwarf_line_srcfileno(lines[i], &fno, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+		if (fno != pf->fno)
+			continue;
+
+		ret = dwarf_lineno(lines[i], &lineno, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+		if (lineno != pf->lno)
+			continue;
+
+		ret = dwarf_lineoff(lines[i], &clm, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+
+		ret = dwarf_lineaddr(lines[i], &addr, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+		pr_debug("Probe line found: line[%d]:%u,%d addr:0x%llx\n",
+			 (int)i, (unsigned)lineno, (int)clm, addr);
+		pf->addr = addr;
+		/* Search a real subprogram including this line, */
+		ret = search_die_from_children(pf->cu_die,
+					       probeaddr_callback, pf);
+		if (ret == 0)
+			die("Probe point is not found in subprograms.\n");
+		/* Continuing, because target line might be inlined. */
+	}
+	dwarf_srclines_dealloc(__dw_debug, lines, cnt);
+}
+
+/* Search function from function name */
+static int probefunc_callback(struct die_link *dlink, void *data)
+{
+	struct probe_finder *pf = (struct probe_finder *)data;
+	struct probe_point *pp = pf->pp;
+	struct die_link *lk;
+	Dwarf_Signed offs;
+	Dwarf_Half tag;
+	int ret;
+
+	ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+	DIE_IF(ret == DW_DLV_ERROR);
+	if (tag == DW_TAG_subprogram) {
+		if (die_compare_name(dlink->die, pp->function) == 0) {
+			if (pp->line) {	/* Function relative line */
+				pf->fno = die_get_decl_file(dlink->die);
+				pf->lno = die_get_decl_line(dlink->die)
+					 + pp->line;
+				find_by_line(pf);
+				return 1;
+			}
+			if (die_inlined_subprogram(dlink->die)) {
+				/* Inlined function, save it. */
+				ret = dwarf_die_CU_offset(dlink->die,
+							  &pf->inl_offs,
+							  &__dw_error);
+				DIE_IF(ret != DW_DLV_OK);
+				pr_debug("inline definition offset %lld\n",
+					 pf->inl_offs);
+				return 0;	/* Continue to search */
+			}
+			/* Get probe address */
+			pf->addr = die_get_entrypc(dlink->die);
+			pf->addr += pp->offset;
+			/* TODO: Check the address in this function */
+			show_probepoint(dlink->die, pp->offset, pf);
+			return 1; /* Exit; no same symbol in this CU. */
+		}
+	} else if (tag == DW_TAG_inlined_subroutine && pf->inl_offs) {
+		if (die_get_abstract_origin(dlink->die) == pf->inl_offs) {
+			/* Get probe address */
+			pf->addr = die_get_entrypc(dlink->die);
+			pf->addr += pp->offset;
+			pr_debug("found inline addr: 0x%llx\n", pf->addr);
+			/* Inlined function. Get a real subprogram */
+			for (lk = dlink->parent; lk != NULL; lk = lk->parent) {
+				tag = 0;
+				dwarf_tag(lk->die, &tag, &__dw_error);
+				DIE_IF(ret == DW_DLV_ERROR);
+				if (tag == DW_TAG_subprogram &&
+				    !die_inlined_subprogram(lk->die))
+					goto found;
+			}
+			die("Failed to find real subprogram.\n");
+found:
+			/* Get offset from subprogram */
+			ret = die_within_subprogram(lk->die, pf->addr, &offs);
+			DIE_IF(!ret);
+			show_probepoint(lk->die, offs, pf);
+			/* Continue to search */
+		}
+	}
+	return 0;
+}
+
+static void find_by_func(struct probe_finder *pf)
+{
+	search_die_from_children(pf->cu_die, probefunc_callback, pf);
+}
+
+/* Find a probe point */
+int find_probepoint(int fd, struct probe_point *pp)
+{
+	Dwarf_Half addr_size = 0;
+	Dwarf_Unsigned next_cuh = 0;
+	int cu_number = 0, ret;
+	struct probe_finder pf = {.pp = pp};
+
+	ret = dwarf_init(fd, DW_DLC_READ, 0, 0, &__dw_debug, &__dw_error);
+	if (ret != DW_DLV_OK) {
+		pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO.\n");
+		return -ENOENT;
+	}
+
+	pp->found = 0;
+	while (++cu_number) {
+		/* Search CU (Compilation Unit) */
+		ret = dwarf_next_cu_header(__dw_debug, NULL, NULL, NULL,
+			&addr_size, &next_cuh, &__dw_error);
+		DIE_IF(ret == DW_DLV_ERROR);
+		if (ret == DW_DLV_NO_ENTRY)
+			break;
+
+		/* Get the DIE(Debugging Information Entry) of this CU */
+		ret = dwarf_siblingof(__dw_debug, 0, &pf.cu_die, &__dw_error);
+		DIE_IF(ret != DW_DLV_OK);
+
+		/* Check if target file is included. */
+		if (pp->file)
+			pf.fno = cu_find_fileno(pf.cu_die, pp->file);
+
+		if (!pp->file || pf.fno) {
+			/* Save CU base address (for frame_base) */
+			ret = dwarf_lowpc(pf.cu_die, &pf.cu_base, &__dw_error);
+			DIE_IF(ret == DW_DLV_ERROR);
+			if (ret == DW_DLV_NO_ENTRY)
+				pf.cu_base = 0;
+			if (pp->function)
+				find_by_func(&pf);
+			else {
+				pf.lno = pp->line;
+				find_by_line(&pf);
+			}
+		}
+		dwarf_dealloc(__dw_debug, pf.cu_die, DW_DLA_DIE);
+	}
+	ret = dwarf_finish(__dw_debug, &__dw_error);
+	DIE_IF(ret != DW_DLV_OK);
+
+	return pp->found;
+}
+
diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h
new file mode 100644
index 000000000000..bdebca6697d2
--- /dev/null
+++ b/tools/perf/util/probe-finder.h
@@ -0,0 +1,57 @@
+#ifndef _PROBE_FINDER_H
+#define _PROBE_FINDER_H
+
+#define MAX_PATH_LEN 256
+#define MAX_PROBE_BUFFER 1024
+#define MAX_PROBES 128
+
+static inline int is_c_varname(const char *name)
+{
+	/* TODO */
+	return isalpha(name[0]) || name[0] == '_';
+}
+
+struct probe_point {
+	/* Inputs */
+	char	*file;		/* File name */
+	int	line;		/* Line number */
+
+	char	*function;	/* Function name */
+	int	offset;		/* Offset bytes */
+
+	int	nr_args;	/* Number of arguments */
+	char	**args;		/* Arguments */
+
+	int	retprobe;	/* Return probe */
+
+	/* Output */
+	int	found;		/* Number of found probe points */
+	char	*probes[MAX_PROBES];	/* Output buffers (will be allocated)*/
+};
+
+#ifndef NO_LIBDWARF
+extern int find_probepoint(int fd, struct probe_point *pp);
+
+#include <libdwarf/dwarf.h>
+#include <libdwarf/libdwarf.h>
+
+struct probe_finder {
+	struct probe_point	*pp;	/* Target probe point */
+
+	/* For function searching */
+	Dwarf_Addr	addr;		/* Address */
+	Dwarf_Unsigned	fno;		/* File number */
+	Dwarf_Unsigned	lno;		/* Line number */
+	Dwarf_Off	inl_offs;	/* Inline offset */
+	Dwarf_Die	cu_die;		/* Current CU */
+
+	/* For variable searching */
+	Dwarf_Addr	cu_base;	/* Current CU base address */
+	Dwarf_Locdesc	fbloc;		/* Location of Current Frame Base */
+	const char	*var;		/* Current variable name */
+	char		*buf;		/* Current output buffer */
+	int		len;		/* Length of output buffer */
+};
+#endif /* NO_LIBDWARF */
+
+#endif /*_PROBE_FINDER_H */
diff --git a/tools/perf/util/quote.h b/tools/perf/util/quote.h
index a5454a1d1c13..b6a019733919 100644
--- a/tools/perf/util/quote.h
+++ b/tools/perf/util/quote.h
@@ -1,5 +1,5 @@
-#ifndef QUOTE_H
-#define QUOTE_H
+#ifndef __PERF_QUOTE_H
+#define __PERF_QUOTE_H
 
 #include <stddef.h>
 #include <stdio.h>
@@ -65,4 +65,4 @@ extern void perl_quote_print(FILE *stream, const char *src);
 extern void python_quote_print(FILE *stream, const char *src);
 extern void tcl_quote_print(FILE *stream, const char *src);
 
-#endif
+#endif /* __PERF_QUOTE_H */
diff --git a/tools/perf/util/run-command.h b/tools/perf/util/run-command.h
index cc1837deba88..d79028727ce2 100644
--- a/tools/perf/util/run-command.h
+++ b/tools/perf/util/run-command.h
@@ -1,5 +1,5 @@
-#ifndef RUN_COMMAND_H
-#define RUN_COMMAND_H
+#ifndef __PERF_RUN_COMMAND_H
+#define __PERF_RUN_COMMAND_H
 
 enum {
 	ERR_RUN_COMMAND_FORK = 10000,
@@ -85,4 +85,4 @@ struct async {
 int start_async(struct async *async);
 int finish_async(struct async *async);
 
-#endif
+#endif /* __PERF_RUN_COMMAND_H */
diff --git a/tools/perf/util/sigchain.h b/tools/perf/util/sigchain.h
index 618083bce0c6..1a53c11265fd 100644
--- a/tools/perf/util/sigchain.h
+++ b/tools/perf/util/sigchain.h
@@ -1,5 +1,5 @@
-#ifndef SIGCHAIN_H
-#define SIGCHAIN_H
+#ifndef __PERF_SIGCHAIN_H
+#define __PERF_SIGCHAIN_H
 
 typedef void (*sigchain_fun)(int);
 
@@ -8,4 +8,4 @@ int sigchain_pop(int sig);
 
 void sigchain_push_common(sigchain_fun f);
 
-#endif /* SIGCHAIN_H */
+#endif /* __PERF_SIGCHAIN_H */
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
new file mode 100644
index 000000000000..b490354d1b23
--- /dev/null
+++ b/tools/perf/util/sort.c
@@ -0,0 +1,290 @@
+#include "sort.h"
+
+regex_t		parent_regex;
+char		default_parent_pattern[] = "^sys_|^do_page_fault";
+char		*parent_pattern = default_parent_pattern;
+char		default_sort_order[] = "comm,dso,symbol";
+char		*sort_order = default_sort_order;
+int		sort__need_collapse = 0;
+int		sort__has_parent = 0;
+
+enum sort_type	sort__first_dimension;
+
+unsigned int dsos__col_width;
+unsigned int comms__col_width;
+unsigned int threads__col_width;
+static unsigned int parent_symbol__col_width;
+char * field_sep;
+
+LIST_HEAD(hist_entry__sort_list);
+
+struct sort_entry sort_thread = {
+	.header = "Command:  Pid",
+	.cmp	= sort__thread_cmp,
+	.print	= sort__thread_print,
+	.width	= &threads__col_width,
+};
+
+struct sort_entry sort_comm = {
+	.header		= "Command",
+	.cmp		= sort__comm_cmp,
+	.collapse	= sort__comm_collapse,
+	.print		= sort__comm_print,
+	.width		= &comms__col_width,
+};
+
+struct sort_entry sort_dso = {
+	.header = "Shared Object",
+	.cmp	= sort__dso_cmp,
+	.print	= sort__dso_print,
+	.width	= &dsos__col_width,
+};
+
+struct sort_entry sort_sym = {
+	.header = "Symbol",
+	.cmp	= sort__sym_cmp,
+	.print	= sort__sym_print,
+};
+
+struct sort_entry sort_parent = {
+	.header = "Parent symbol",
+	.cmp	= sort__parent_cmp,
+	.print	= sort__parent_print,
+	.width	= &parent_symbol__col_width,
+};
+
+struct sort_dimension {
+	const char		*name;
+	struct sort_entry	*entry;
+	int			taken;
+};
+
+static struct sort_dimension sort_dimensions[] = {
+	{ .name = "pid",	.entry = &sort_thread,	},
+	{ .name = "comm",	.entry = &sort_comm,	},
+	{ .name = "dso",	.entry = &sort_dso,	},
+	{ .name = "symbol",	.entry = &sort_sym,	},
+	{ .name = "parent",	.entry = &sort_parent,	},
+};
+
+int64_t cmp_null(void *l, void *r)
+{
+	if (!l && !r)
+		return 0;
+	else if (!l)
+		return -1;
+	else
+		return 1;
+}
+
+/* --sort pid */
+
+int64_t
+sort__thread_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	return right->thread->pid - left->thread->pid;
+}
+
+int repsep_fprintf(FILE *fp, const char *fmt, ...)
+{
+	int n;
+	va_list ap;
+
+	va_start(ap, fmt);
+	if (!field_sep)
+		n = vfprintf(fp, fmt, ap);
+	else {
+		char *bf = NULL;
+		n = vasprintf(&bf, fmt, ap);
+		if (n > 0) {
+			char *sep = bf;
+
+			while (1) {
+				sep = strchr(sep, *field_sep);
+				if (sep == NULL)
+					break;
+				*sep = '.';
+			}
+		}
+		fputs(bf, fp);
+		free(bf);
+	}
+	va_end(ap);
+	return n;
+}
+
+size_t
+sort__thread_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%*s:%5d", width - 6,
+			      self->thread->comm ?: "", self->thread->pid);
+}
+
+size_t
+sort__comm_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%*s", width, self->thread->comm);
+}
+
+/* --sort dso */
+
+int64_t
+sort__dso_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct dso *dso_l = left->map ? left->map->dso : NULL;
+	struct dso *dso_r = right->map ? right->map->dso : NULL;
+	const char *dso_name_l, *dso_name_r;
+
+	if (!dso_l || !dso_r)
+		return cmp_null(dso_l, dso_r);
+
+	if (verbose) {
+		dso_name_l = dso_l->long_name;
+		dso_name_r = dso_r->long_name;
+	} else {
+		dso_name_l = dso_l->short_name;
+		dso_name_r = dso_r->short_name;
+	}
+
+	return strcmp(dso_name_l, dso_name_r);
+}
+
+size_t
+sort__dso_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	if (self->map && self->map->dso) {
+		const char *dso_name = !verbose ? self->map->dso->short_name :
+						  self->map->dso->long_name;
+		return repsep_fprintf(fp, "%-*s", width, dso_name);
+	}
+
+	return repsep_fprintf(fp, "%*llx", width, (u64)self->ip);
+}
+
+/* --sort symbol */
+
+int64_t
+sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	u64 ip_l, ip_r;
+
+	if (left->sym == right->sym)
+		return 0;
+
+	ip_l = left->sym ? left->sym->start : left->ip;
+	ip_r = right->sym ? right->sym->start : right->ip;
+
+	return (int64_t)(ip_r - ip_l);
+}
+
+
+size_t
+sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used)
+{
+	size_t ret = 0;
+
+	if (verbose) {
+		char o = self->map ? dso__symtab_origin(self->map->dso) : '!';
+		ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip, o);
+	}
+
+	ret += repsep_fprintf(fp, "[%c] ", self->level);
+	if (self->sym)
+		ret += repsep_fprintf(fp, "%s", self->sym->name);
+	else
+		ret += repsep_fprintf(fp, "%#016llx", (u64)self->ip);
+
+	return ret;
+}
+
+/* --sort comm */
+
+int64_t
+sort__comm_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	return right->thread->pid - left->thread->pid;
+}
+
+int64_t
+sort__comm_collapse(struct hist_entry *left, struct hist_entry *right)
+{
+	char *comm_l = left->thread->comm;
+	char *comm_r = right->thread->comm;
+
+	if (!comm_l || !comm_r)
+		return cmp_null(comm_l, comm_r);
+
+	return strcmp(comm_l, comm_r);
+}
+
+/* --sort parent */
+
+int64_t
+sort__parent_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	struct symbol *sym_l = left->parent;
+	struct symbol *sym_r = right->parent;
+
+	if (!sym_l || !sym_r)
+		return cmp_null(sym_l, sym_r);
+
+	return strcmp(sym_l->name, sym_r->name);
+}
+
+size_t
+sort__parent_print(FILE *fp, struct hist_entry *self, unsigned int width)
+{
+	return repsep_fprintf(fp, "%-*s", width,
+			      self->parent ? self->parent->name : "[other]");
+}
+
+int sort_dimension__add(const char *tok)
+{
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) {
+		struct sort_dimension *sd = &sort_dimensions[i];
+
+		if (sd->taken)
+			continue;
+
+		if (strncasecmp(tok, sd->name, strlen(tok)))
+			continue;
+
+		if (sd->entry->collapse)
+			sort__need_collapse = 1;
+
+		if (sd->entry == &sort_parent) {
+			int ret = regcomp(&parent_regex, parent_pattern, REG_EXTENDED);
+			if (ret) {
+				char err[BUFSIZ];
+
+				regerror(ret, &parent_regex, err, sizeof(err));
+				fprintf(stderr, "Invalid regex: %s\n%s",
+					parent_pattern, err);
+				exit(-1);
+			}
+			sort__has_parent = 1;
+		}
+
+		if (list_empty(&hist_entry__sort_list)) {
+			if (!strcmp(sd->name, "pid"))
+				sort__first_dimension = SORT_PID;
+			else if (!strcmp(sd->name, "comm"))
+				sort__first_dimension = SORT_COMM;
+			else if (!strcmp(sd->name, "dso"))
+				sort__first_dimension = SORT_DSO;
+			else if (!strcmp(sd->name, "symbol"))
+				sort__first_dimension = SORT_SYM;
+			else if (!strcmp(sd->name, "parent"))
+				sort__first_dimension = SORT_PARENT;
+		}
+
+		list_add_tail(&sd->entry->list, &hist_entry__sort_list);
+		sd->taken = 1;
+
+		return 0;
+	}
+
+	return -ESRCH;
+}
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
new file mode 100644
index 000000000000..333e664ff45f
--- /dev/null
+++ b/tools/perf/util/sort.h
@@ -0,0 +1,99 @@
+#ifndef __PERF_SORT_H
+#define __PERF_SORT_H
+#include "../builtin.h"
+
+#include "util.h"
+
+#include "color.h"
+#include <linux/list.h>
+#include "cache.h"
+#include <linux/rbtree.h>
+#include "symbol.h"
+#include "string.h"
+#include "callchain.h"
+#include "strlist.h"
+#include "values.h"
+
+#include "../perf.h"
+#include "debug.h"
+#include "header.h"
+
+#include "parse-options.h"
+#include "parse-events.h"
+
+#include "thread.h"
+#include "sort.h"
+
+extern regex_t parent_regex;
+extern char *sort_order;
+extern char default_parent_pattern[];
+extern char *parent_pattern;
+extern char default_sort_order[];
+extern int sort__need_collapse;
+extern int sort__has_parent;
+extern char *field_sep;
+extern struct sort_entry sort_comm;
+extern struct sort_entry sort_dso;
+extern struct sort_entry sort_sym;
+extern struct sort_entry sort_parent;
+extern unsigned int dsos__col_width;
+extern unsigned int comms__col_width;
+extern unsigned int threads__col_width;
+extern enum sort_type sort__first_dimension;
+
+struct hist_entry {
+	struct rb_node		rb_node;
+	u64			count;
+	struct thread		*thread;
+	struct map		*map;
+	struct symbol		*sym;
+	u64			ip;
+	char			level;
+	struct symbol		*parent;
+	struct callchain_node	callchain;
+	struct rb_root		sorted_chain;
+};
+
+enum sort_type {
+	SORT_PID,
+	SORT_COMM,
+	SORT_DSO,
+	SORT_SYM,
+	SORT_PARENT
+};
+
+/*
+ * configurable sorting bits
+ */
+
+struct sort_entry {
+	struct list_head list;
+
+	const char *header;
+
+	int64_t (*cmp)(struct hist_entry *, struct hist_entry *);
+	int64_t (*collapse)(struct hist_entry *, struct hist_entry *);
+	size_t	(*print)(FILE *fp, struct hist_entry *, unsigned int width);
+	unsigned int *width;
+	bool	elide;
+};
+
+extern struct sort_entry sort_thread;
+extern struct list_head hist_entry__sort_list;
+
+extern int repsep_fprintf(FILE *fp, const char *fmt, ...);
+extern size_t sort__thread_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__comm_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__dso_print(FILE *, struct hist_entry *, unsigned int);
+extern size_t sort__sym_print(FILE *, struct hist_entry *, unsigned int __used);
+extern int64_t cmp_null(void *, void *);
+extern int64_t sort__thread_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__comm_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__comm_collapse(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__dso_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__sym_cmp(struct hist_entry *, struct hist_entry *);
+extern int64_t sort__parent_cmp(struct hist_entry *, struct hist_entry *);
+extern size_t sort__parent_print(FILE *, struct hist_entry *, unsigned int);
+extern int sort_dimension__add(const char *);
+
+#endif	/* __PERF_SORT_H */
diff --git a/tools/perf/util/strbuf.h b/tools/perf/util/strbuf.h
index d2aa86c014c1..a3d121d6c83e 100644
--- a/tools/perf/util/strbuf.h
+++ b/tools/perf/util/strbuf.h
@@ -1,5 +1,5 @@
-#ifndef STRBUF_H
-#define STRBUF_H
+#ifndef __PERF_STRBUF_H
+#define __PERF_STRBUF_H
 
 /*
  * Strbuf's can be use in many ways: as a byte array, or to store arbitrary
@@ -134,4 +134,4 @@ extern int launch_editor(const char *path, struct strbuf *buffer, const char *co
 extern int strbuf_branchname(struct strbuf *sb, const char *name);
 extern int strbuf_check_branch_ref(struct strbuf *sb, const char *name);
 
-#endif /* STRBUF_H */
+#endif /* __PERF_STRBUF_H */
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index c93eca9a7be3..f24a8cc933d5 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -1,4 +1,5 @@
 #include "string.h"
+#include "util.h"
 
 static int hex(char ch)
 {
@@ -32,3 +33,196 @@ int hex2u64(const char *ptr, u64 *long_val)
 
 	return p - ptr;
 }
+
+char *strxfrchar(char *s, char from, char to)
+{
+	char *p = s;
+
+	while ((p = strchr(p, from)) != NULL)
+		*p++ = to;
+
+	return s;
+}
+
+#define K 1024LL
+/*
+ * perf_atoll()
+ * Parse (\d+)(b|B|kb|KB|mb|MB|gb|GB|tb|TB) (e.g. "256MB")
+ * and return its numeric value
+ */
+s64 perf_atoll(const char *str)
+{
+	unsigned int i;
+	s64 length = -1, unit = 1;
+
+	if (!isdigit(str[0]))
+		goto out_err;
+
+	for (i = 1; i < strlen(str); i++) {
+		switch (str[i]) {
+		case 'B':
+		case 'b':
+			break;
+		case 'K':
+			if (str[i + 1] != 'B')
+				goto out_err;
+			else
+				goto kilo;
+		case 'k':
+			if (str[i + 1] != 'b')
+				goto out_err;
+kilo:
+			unit = K;
+			break;
+		case 'M':
+			if (str[i + 1] != 'B')
+				goto out_err;
+			else
+				goto mega;
+		case 'm':
+			if (str[i + 1] != 'b')
+				goto out_err;
+mega:
+			unit = K * K;
+			break;
+		case 'G':
+			if (str[i + 1] != 'B')
+				goto out_err;
+			else
+				goto giga;
+		case 'g':
+			if (str[i + 1] != 'b')
+				goto out_err;
+giga:
+			unit = K * K * K;
+			break;
+		case 'T':
+			if (str[i + 1] != 'B')
+				goto out_err;
+			else
+				goto tera;
+		case 't':
+			if (str[i + 1] != 'b')
+				goto out_err;
+tera:
+			unit = K * K * K * K;
+			break;
+		case '\0':	/* only specified figures */
+			unit = 1;
+			break;
+		default:
+			if (!isdigit(str[i]))
+				goto out_err;
+			break;
+		}
+	}
+
+	length = atoll(str) * unit;
+	goto out;
+
+out_err:
+	length = -1;
+out:
+	return length;
+}
+
+/*
+ * Helper function for splitting a string into an argv-like array.
+ * originaly copied from lib/argv_split.c
+ */
+static const char *skip_sep(const char *cp)
+{
+	while (*cp && isspace(*cp))
+		cp++;
+
+	return cp;
+}
+
+static const char *skip_arg(const char *cp)
+{
+	while (*cp && !isspace(*cp))
+		cp++;
+
+	return cp;
+}
+
+static int count_argc(const char *str)
+{
+	int count = 0;
+
+	while (*str) {
+		str = skip_sep(str);
+		if (*str) {
+			count++;
+			str = skip_arg(str);
+		}
+	}
+
+	return count;
+}
+
+/**
+ * argv_free - free an argv
+ * @argv - the argument vector to be freed
+ *
+ * Frees an argv and the strings it points to.
+ */
+void argv_free(char **argv)
+{
+	char **p;
+	for (p = argv; *p; p++)
+		free(*p);
+
+	free(argv);
+}
+
+/**
+ * argv_split - split a string at whitespace, returning an argv
+ * @str: the string to be split
+ * @argcp: returned argument count
+ *
+ * Returns an array of pointers to strings which are split out from
+ * @str.  This is performed by strictly splitting on white-space; no
+ * quote processing is performed.  Multiple whitespace characters are
+ * considered to be a single argument separator.  The returned array
+ * is always NULL-terminated.  Returns NULL on memory allocation
+ * failure.
+ */
+char **argv_split(const char *str, int *argcp)
+{
+	int argc = count_argc(str);
+	char **argv = zalloc(sizeof(*argv) * (argc+1));
+	char **argvp;
+
+	if (argv == NULL)
+		goto out;
+
+	if (argcp)
+		*argcp = argc;
+
+	argvp = argv;
+
+	while (*str) {
+		str = skip_sep(str);
+
+		if (*str) {
+			const char *p = str;
+			char *t;
+
+			str = skip_arg(str);
+
+			t = strndup(p, str-p);
+			if (t == NULL)
+				goto fail;
+			*argvp++ = t;
+		}
+	}
+	*argvp = NULL;
+
+out:
+	return argv;
+
+fail:
+	argv_free(argv);
+	return NULL;
+}
diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h
index bf39dfadfd24..bfecec265a1a 100644
--- a/tools/perf/util/string.h
+++ b/tools/perf/util/string.h
@@ -1,11 +1,15 @@
-#ifndef _PERF_STRING_H_
-#define _PERF_STRING_H_
+#ifndef __PERF_STRING_H_
+#define __PERF_STRING_H_
 
 #include "types.h"
 
 int hex2u64(const char *ptr, u64 *val);
+char *strxfrchar(char *s, char from, char to);
+s64 perf_atoll(const char *str);
+char **argv_split(const char *str, int *argcp);
+void argv_free(char **argv);
 
 #define _STR(x) #x
 #define STR(x) _STR(x)
 
-#endif
+#endif /* __PERF_STRING_H */
diff --git a/tools/perf/util/strlist.h b/tools/perf/util/strlist.h
index 921818e44a54..cb4659306d7b 100644
--- a/tools/perf/util/strlist.h
+++ b/tools/perf/util/strlist.h
@@ -1,5 +1,5 @@
-#ifndef STRLIST_H_
-#define STRLIST_H_
+#ifndef __PERF_STRLIST_H
+#define __PERF_STRLIST_H
 
 #include <linux/rbtree.h>
 #include <stdbool.h>
@@ -36,4 +36,4 @@ static inline unsigned int strlist__nr_entries(const struct strlist *self)
 }
 
 int strlist__parse_list(struct strlist *self, const char *s);
-#endif /* STRLIST_H_ */
+#endif /* __PERF_STRLIST_H */
diff --git a/tools/perf/util/svghelper.c b/tools/perf/util/svghelper.c
index 856655d8b0b8..b3637db025a2 100644
--- a/tools/perf/util/svghelper.c
+++ b/tools/perf/util/svghelper.c
@@ -103,7 +103,7 @@ void open_svg(const char *filename, int cpus, int rows, u64 start, u64 end)
 	fprintf(svgfile, "      rect.process2 { fill:rgb(180,180,180); fill-opacity:0.9; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.sample   { fill:rgb(  0,  0,255); fill-opacity:0.8; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.blocked  { fill:rgb(255,  0,  0); fill-opacity:0.5; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
-	fprintf(svgfile, "      rect.waiting  { fill:rgb(214,214,  0); fill-opacity:0.3; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
+	fprintf(svgfile, "      rect.waiting  { fill:rgb(224,214,  0); fill-opacity:0.8; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.WAITING  { fill:rgb(255,214, 48); fill-opacity:0.6; stroke-width:0;   stroke:rgb(  0,  0,  0); } \n");
 	fprintf(svgfile, "      rect.cpu      { fill:rgb(192,192,192); fill-opacity:0.2; stroke-width:0.5; stroke:rgb(128,128,128); } \n");
 	fprintf(svgfile, "      rect.pstate   { fill:rgb(128,128,128); fill-opacity:0.8; stroke-width:0; } \n");
diff --git a/tools/perf/util/svghelper.h b/tools/perf/util/svghelper.h
index cd93195aedb3..e0781989cc31 100644
--- a/tools/perf/util/svghelper.h
+++ b/tools/perf/util/svghelper.h
@@ -1,5 +1,5 @@
-#ifndef _INCLUDE_GUARD_SVG_HELPER_
-#define _INCLUDE_GUARD_SVG_HELPER_
+#ifndef __PERF_SVGHELPER_H
+#define __PERF_SVGHELPER_H
 
 #include "types.h"
 
@@ -25,4 +25,4 @@ extern void svg_close(void);
 
 extern int svg_page_width;
 
-#endif
+#endif /* __PERF_SVGHELPER_H */
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 47ea0609a760..fffcb937cdcb 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -2,14 +2,20 @@
 #include "../perf.h"
 #include "string.h"
 #include "symbol.h"
+#include "thread.h"
 
 #include "debug.h"
 
+#include <asm/bug.h>
 #include <libelf.h>
 #include <gelf.h>
 #include <elf.h>
+#include <limits.h>
+#include <sys/utsname.h>
 
-const char *sym_hist_filter;
+#ifndef NT_GNU_BUILD_ID
+#define NT_GNU_BUILD_ID 3
+#endif
 
 enum dso_origin {
 	DSO__ORIG_KERNEL = 0,
@@ -18,94 +24,189 @@ enum dso_origin {
 	DSO__ORIG_UBUNTU,
 	DSO__ORIG_BUILDID,
 	DSO__ORIG_DSO,
+	DSO__ORIG_KMODULE,
 	DSO__ORIG_NOT_FOUND,
 };
 
-static struct symbol *symbol__new(u64 start, u64 len,
-				  const char *name, unsigned int priv_size,
-				  u64 obj_start, int v)
+static void dsos__add(struct list_head *head, struct dso *dso);
+static struct map *thread__find_map_by_name(struct thread *self, char *name);
+static struct map *map__new2(u64 start, struct dso *dso, enum map_type type);
+struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr);
+static int dso__load_kernel_sym(struct dso *self, struct map *map,
+				struct thread *thread, symbol_filter_t filter);
+unsigned int symbol__priv_size;
+static int vmlinux_path__nr_entries;
+static char **vmlinux_path;
+
+static struct symbol_conf symbol_conf__defaults = {
+	.use_modules	  = true,
+	.try_vmlinux_path = true,
+};
+
+static struct thread kthread_mem;
+struct thread *kthread = &kthread_mem;
+
+bool dso__loaded(const struct dso *self, enum map_type type)
 {
-	size_t namelen = strlen(name) + 1;
-	struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen);
+	return self->loaded & (1 << type);
+}
 
-	if (!self)
-		return NULL;
+static void dso__set_loaded(struct dso *self, enum map_type type)
+{
+	self->loaded |= (1 << type);
+}
 
-	if (v >= 2)
-		printf("new symbol: %016Lx [%08lx]: %s, hist: %p, obj_start: %p\n",
-			(u64)start, (unsigned long)len, name, self->hist, (void *)(unsigned long)obj_start);
+static void symbols__fixup_end(struct rb_root *self)
+{
+	struct rb_node *nd, *prevnd = rb_first(self);
+	struct symbol *curr, *prev;
+
+	if (prevnd == NULL)
+		return;
 
-	self->obj_start= obj_start;
-	self->hist = NULL;
-	self->hist_sum = 0;
+	curr = rb_entry(prevnd, struct symbol, rb_node);
 
-	if (sym_hist_filter && !strcmp(name, sym_hist_filter))
-		self->hist = calloc(sizeof(u64), len);
+	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
+		prev = curr;
+		curr = rb_entry(nd, struct symbol, rb_node);
 
-	if (priv_size) {
-		memset(self, 0, priv_size);
-		self = ((void *)self) + priv_size;
+		if (prev->end == prev->start)
+			prev->end = curr->start - 1;
 	}
+
+	/* Last entry */
+	if (curr->end == curr->start)
+		curr->end = roundup(curr->start, 4096);
+}
+
+static void __thread__fixup_maps_end(struct thread *self, enum map_type type)
+{
+	struct map *prev, *curr;
+	struct rb_node *nd, *prevnd = rb_first(&self->maps[type]);
+
+	if (prevnd == NULL)
+		return;
+
+	curr = rb_entry(prevnd, struct map, rb_node);
+
+	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
+		prev = curr;
+		curr = rb_entry(nd, struct map, rb_node);
+		prev->end = curr->start - 1;
+	}
+
+	/*
+	 * We still haven't the actual symbols, so guess the
+	 * last map final address.
+	 */
+	curr->end = ~0UL;
+}
+
+static void thread__fixup_maps_end(struct thread *self)
+{
+	int i;
+	for (i = 0; i < MAP__NR_TYPES; ++i)
+		__thread__fixup_maps_end(self, i);
+}
+
+static struct symbol *symbol__new(u64 start, u64 len, const char *name)
+{
+	size_t namelen = strlen(name) + 1;
+	struct symbol *self = zalloc(symbol__priv_size +
+				     sizeof(*self) + namelen);
+	if (self == NULL)
+		return NULL;
+
+	if (symbol__priv_size)
+		self = ((void *)self) + symbol__priv_size;
+
 	self->start = start;
 	self->end   = len ? start + len - 1 : start;
+
+	pr_debug3("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end);
+
 	memcpy(self->name, name, namelen);
 
 	return self;
 }
 
-static void symbol__delete(struct symbol *self, unsigned int priv_size)
+static void symbol__delete(struct symbol *self)
 {
-	free(((void *)self) - priv_size);
+	free(((void *)self) - symbol__priv_size);
 }
 
 static size_t symbol__fprintf(struct symbol *self, FILE *fp)
 {
-	if (!self->module)
-		return fprintf(fp, " %llx-%llx %s\n",
+	return fprintf(fp, " %llx-%llx %s\n",
 		       self->start, self->end, self->name);
-	else
-		return fprintf(fp, " %llx-%llx %s \t[%s]\n",
-		       self->start, self->end, self->name, self->module->name);
 }
 
-struct dso *dso__new(const char *name, unsigned int sym_priv_size)
+static void dso__set_long_name(struct dso *self, char *name)
+{
+	if (name == NULL)
+		return;
+	self->long_name = name;
+	self->long_name_len = strlen(name);
+}
+
+static void dso__set_basename(struct dso *self)
+{
+	self->short_name = basename(self->long_name);
+}
+
+struct dso *dso__new(const char *name)
 {
 	struct dso *self = malloc(sizeof(*self) + strlen(name) + 1);
 
 	if (self != NULL) {
+		int i;
 		strcpy(self->name, name);
-		self->syms = RB_ROOT;
-		self->sym_priv_size = sym_priv_size;
+		dso__set_long_name(self, self->name);
+		self->short_name = self->name;
+		for (i = 0; i < MAP__NR_TYPES; ++i)
+			self->symbols[i] = RB_ROOT;
 		self->find_symbol = dso__find_symbol;
 		self->slen_calculated = 0;
 		self->origin = DSO__ORIG_NOT_FOUND;
+		self->loaded = 0;
+		self->has_build_id = 0;
 	}
 
 	return self;
 }
 
-static void dso__delete_symbols(struct dso *self)
+static void symbols__delete(struct rb_root *self)
 {
 	struct symbol *pos;
-	struct rb_node *next = rb_first(&self->syms);
+	struct rb_node *next = rb_first(self);
 
 	while (next) {
 		pos = rb_entry(next, struct symbol, rb_node);
 		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &self->syms);
-		symbol__delete(pos, self->sym_priv_size);
+		rb_erase(&pos->rb_node, self);
+		symbol__delete(pos);
 	}
 }
 
 void dso__delete(struct dso *self)
 {
-	dso__delete_symbols(self);
+	int i;
+	for (i = 0; i < MAP__NR_TYPES; ++i)
+		symbols__delete(&self->symbols[i]);
+	if (self->long_name != self->name)
+		free(self->long_name);
 	free(self);
 }
 
-static void dso__insert_symbol(struct dso *self, struct symbol *sym)
+void dso__set_build_id(struct dso *self, void *build_id)
 {
-	struct rb_node **p = &self->syms.rb_node;
+	memcpy(self->build_id, build_id, sizeof(self->build_id));
+	self->has_build_id = 1;
+}
+
+static void symbols__insert(struct rb_root *self, struct symbol *sym)
+{
+	struct rb_node **p = &self->rb_node;
 	struct rb_node *parent = NULL;
 	const u64 ip = sym->start;
 	struct symbol *s;
@@ -119,17 +220,17 @@ static void dso__insert_symbol(struct dso *self, struct symbol *sym)
 			p = &(*p)->rb_right;
 	}
 	rb_link_node(&sym->rb_node, parent, p);
-	rb_insert_color(&sym->rb_node, &self->syms);
+	rb_insert_color(&sym->rb_node, self);
 }
 
-struct symbol *dso__find_symbol(struct dso *self, u64 ip)
+static struct symbol *symbols__find(struct rb_root *self, u64 ip)
 {
 	struct rb_node *n;
 
 	if (self == NULL)
 		return NULL;
 
-	n = self->syms.rb_node;
+	n = self->rb_node;
 
 	while (n) {
 		struct symbol *s = rb_entry(n, struct symbol, rb_node);
@@ -145,12 +246,42 @@ struct symbol *dso__find_symbol(struct dso *self, u64 ip)
 	return NULL;
 }
 
-size_t dso__fprintf(struct dso *self, FILE *fp)
+struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr)
 {
-	size_t ret = fprintf(fp, "dso: %s\n", self->name);
+	return symbols__find(&self->symbols[type], addr);
+}
+
+int build_id__sprintf(u8 *self, int len, char *bf)
+{
+	char *bid = bf;
+	u8 *raw = self;
+	int i;
+
+	for (i = 0; i < len; ++i) {
+		sprintf(bid, "%02x", *raw);
+		++raw;
+		bid += 2;
+	}
+
+	return raw - self;
+}
+
+size_t dso__fprintf_buildid(struct dso *self, FILE *fp)
+{
+	char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+	build_id__sprintf(self->build_id, sizeof(self->build_id), sbuild_id);
+	return fprintf(fp, "%s", sbuild_id);
+}
 
+size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp)
+{
 	struct rb_node *nd;
-	for (nd = rb_first(&self->syms); nd; nd = rb_next(nd)) {
+	size_t ret = fprintf(fp, "dso: %s (", self->short_name);
+
+	ret += dso__fprintf_buildid(self, fp);
+	ret += fprintf(fp, ")\n");
+	for (nd = rb_first(&self->symbols[type]); nd; nd = rb_next(nd)) {
 		struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
 		ret += symbol__fprintf(pos, fp);
 	}
@@ -158,13 +289,17 @@ size_t dso__fprintf(struct dso *self, FILE *fp)
 	return ret;
 }
 
-static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
+/*
+ * Loads the function entries in /proc/kallsyms into kernel_map->dso,
+ * so that we can in the next step set the symbol ->end address and then
+ * call kernel_maps__split_kallsyms.
+ */
+static int dso__load_all_kallsyms(struct dso *self, struct map *map)
 {
-	struct rb_node *nd, *prevnd;
 	char *line = NULL;
 	size_t n;
+	struct rb_root *root = &self->symbols[map->type];
 	FILE *file = fopen("/proc/kallsyms", "r");
-	int count = 0;
 
 	if (file == NULL)
 		goto out_failure;
@@ -174,6 +309,7 @@ static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
 		struct symbol *sym;
 		int line_len, len;
 		char symbol_type;
+		char *symbol_name;
 
 		line_len = getline(&line, &n, file);
 		if (line_len < 0)
@@ -196,44 +332,26 @@ static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int v)
 		 */
 		if (symbol_type != 'T' && symbol_type != 'W')
 			continue;
+
+		symbol_name = line + len + 2;
 		/*
-		 * Well fix up the end later, when we have all sorted.
+		 * Will fix up the end later, when we have all symbols sorted.
 		 */
-		sym = symbol__new(start, 0xdead, line + len + 2,
-				  self->sym_priv_size, 0, v);
+		sym = symbol__new(start, 0, symbol_name);
 
 		if (sym == NULL)
 			goto out_delete_line;
-
-		if (filter && filter(self, sym))
-			symbol__delete(sym, self->sym_priv_size);
-		else {
-			dso__insert_symbol(self, sym);
-			count++;
-		}
-	}
-
-	/*
-	 * Now that we have all sorted out, just set the ->end of all
-	 * symbols
-	 */
-	prevnd = rb_first(&self->syms);
-
-	if (prevnd == NULL)
-		goto out_delete_line;
-
-	for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) {
-		struct symbol *prev = rb_entry(prevnd, struct symbol, rb_node),
-			      *curr = rb_entry(nd, struct symbol, rb_node);
-
-		prev->end = curr->start - 1;
-		prevnd = nd;
+		/*
+		 * We will pass the symbols to the filter later, in
+		 * map__split_kallsyms, when we have split the maps per module
+		 */
+		symbols__insert(root, sym);
 	}
 
 	free(line);
 	fclose(file);
 
-	return count;
+	return 0;
 
 out_delete_line:
 	free(line);
@@ -241,14 +359,114 @@ out_failure:
 	return -1;
 }
 
-static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int v)
+/*
+ * Split the symbols into maps, making sure there are no overlaps, i.e. the
+ * kernel range is broken in several maps, named [kernel].N, as we don't have
+ * the original ELF section names vmlinux have.
+ */
+static int dso__split_kallsyms(struct dso *self, struct map *map, struct thread *thread,
+			       symbol_filter_t filter)
+{
+	struct map *curr_map = map;
+	struct symbol *pos;
+	int count = 0;
+	struct rb_root *root = &self->symbols[map->type];
+	struct rb_node *next = rb_first(root);
+	int kernel_range = 0;
+
+	while (next) {
+		char *module;
+
+		pos = rb_entry(next, struct symbol, rb_node);
+		next = rb_next(&pos->rb_node);
+
+		module = strchr(pos->name, '\t');
+		if (module) {
+			if (!thread->use_modules)
+				goto discard_symbol;
+
+			*module++ = '\0';
+
+			if (strcmp(self->name, module)) {
+				curr_map = thread__find_map_by_name(thread, module);
+				if (curr_map == NULL) {
+					pr_debug("/proc/{kallsyms,modules} "
+					         "inconsistency!\n");
+					return -1;
+				}
+			}
+			/*
+			 * So that we look just like we get from .ko files,
+			 * i.e. not prelinked, relative to map->start.
+			 */
+			pos->start = curr_map->map_ip(curr_map, pos->start);
+			pos->end   = curr_map->map_ip(curr_map, pos->end);
+		} else if (curr_map != map) {
+			char dso_name[PATH_MAX];
+			struct dso *dso;
+
+			snprintf(dso_name, sizeof(dso_name), "[kernel].%d",
+				 kernel_range++);
+
+			dso = dso__new(dso_name);
+			if (dso == NULL)
+				return -1;
+
+			curr_map = map__new2(pos->start, dso, map->type);
+			if (map == NULL) {
+				dso__delete(dso);
+				return -1;
+			}
+
+			curr_map->map_ip = curr_map->unmap_ip = identity__map_ip;
+			__thread__insert_map(thread, curr_map);
+			++kernel_range;
+		}
+
+		if (filter && filter(curr_map, pos)) {
+discard_symbol:		rb_erase(&pos->rb_node, root);
+			symbol__delete(pos);
+		} else {
+			if (curr_map != map) {
+				rb_erase(&pos->rb_node, root);
+				symbols__insert(&curr_map->dso->symbols[curr_map->type], pos);
+			}
+			count++;
+		}
+	}
+
+	return count;
+}
+
+
+static int dso__load_kallsyms(struct dso *self, struct map *map,
+			      struct thread *thread, symbol_filter_t filter)
+{
+	if (dso__load_all_kallsyms(self, map) < 0)
+		return -1;
+
+	symbols__fixup_end(&self->symbols[map->type]);
+	self->origin = DSO__ORIG_KERNEL;
+
+	return dso__split_kallsyms(self, map, thread, filter);
+}
+
+size_t kernel_maps__fprintf(FILE *fp)
+{
+	size_t printed = fprintf(fp, "Kernel maps:\n");
+	printed += thread__fprintf_maps(kthread, fp);
+	return printed + fprintf(fp, "END kernel maps\n");
+}
+
+static int dso__load_perf_map(struct dso *self, struct map *map,
+			      symbol_filter_t filter)
 {
 	char *line = NULL;
 	size_t n;
 	FILE *file;
 	int nr_syms = 0;
 
-	file = fopen(self->name, "r");
+	file = fopen(self->long_name, "r");
 	if (file == NULL)
 		goto out_failure;
 
@@ -278,16 +496,15 @@ static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int v)
 		if (len + 2 >= line_len)
 			continue;
 
-		sym = symbol__new(start, size, line + len,
-				  self->sym_priv_size, start, v);
+		sym = symbol__new(start, size, line + len);
 
 		if (sym == NULL)
 			goto out_delete_line;
 
-		if (filter && filter(self, sym))
-			symbol__delete(sym, self->sym_priv_size);
+		if (filter && filter(map, sym))
+			symbol__delete(sym);
 		else {
-			dso__insert_symbol(self, sym);
+			symbols__insert(&self->symbols[map->type], sym);
 			nr_syms++;
 		}
 	}
@@ -393,7 +610,8 @@ static Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep,
  * And always look at the original dso, not at debuginfo packages, that
  * have the PLT data stripped out (shdr_rel_plt.sh_type == SHT_NOBITS).
  */
-static int dso__synthesize_plt_symbols(struct  dso *self, int v)
+static int dso__synthesize_plt_symbols(struct  dso *self, struct map *map,
+				       symbol_filter_t filter)
 {
 	uint32_t nr_rel_entries, idx;
 	GElf_Sym sym;
@@ -409,11 +627,11 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 	Elf *elf;
 	int nr = 0, symidx, fd, err = 0;
 
-	fd = open(self->name, O_RDONLY);
+	fd = open(self->long_name, O_RDONLY);
 	if (fd < 0)
 		goto out;
 
-	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+	elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
 	if (elf == NULL)
 		goto out_close;
 
@@ -477,12 +695,16 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size, 0, v);
+					sympltname);
 			if (!f)
 				goto out_elf_end;
 
-			dso__insert_symbol(self, f);
-			++nr;
+			if (filter && filter(map, f))
+				symbol__delete(f);
+			else {
+				symbols__insert(&self->symbols[map->type], f);
+				++nr;
+			}
 		}
 	} else if (shdr_rel_plt.sh_type == SHT_REL) {
 		GElf_Rel pos_mem, *pos;
@@ -495,12 +717,16 @@ static int dso__synthesize_plt_symbols(struct  dso *self, int v)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size, 0, v);
+					sympltname);
 			if (!f)
 				goto out_elf_end;
 
-			dso__insert_symbol(self, f);
-			++nr;
+			if (filter && filter(map, f))
+				symbol__delete(f);
+			else {
+				symbols__insert(&self->symbols[map->type], f);
+				++nr;
+			}
 		}
 	}
 
@@ -513,14 +739,18 @@ out_close:
 	if (err == 0)
 		return nr;
 out:
-	fprintf(stderr, "%s: problems reading %s PLT info.\n",
-		__func__, self->name);
+	pr_warning("%s: problems reading %s PLT info.\n",
+		   __func__, self->long_name);
 	return 0;
 }
 
-static int dso__load_sym(struct dso *self, int fd, const char *name,
-			 symbol_filter_t filter, int v, struct module *mod)
+static int dso__load_sym(struct dso *self, struct map *map,
+			 struct thread *thread, const char *name, int fd,
+			 symbol_filter_t filter, int kernel, int kmodule)
 {
+	struct map *curr_map = map;
+	struct dso *curr_dso = self;
+	size_t dso_name_len = strlen(self->short_name);
 	Elf_Data *symstrs, *secstrs;
 	uint32_t nr_syms;
 	int err = -1;
@@ -531,19 +761,16 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 	GElf_Sym sym;
 	Elf_Scn *sec, *sec_strndx;
 	Elf *elf;
-	int nr = 0, kernel = !strcmp("[kernel]", self->name);
+	int nr = 0;
 
-	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+	elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
 	if (elf == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot read %s ELF file.\n",
-				__func__, name);
+		pr_err("%s: cannot read %s ELF file.\n", __func__, name);
 		goto out_close;
 	}
 
 	if (gelf_getehdr(elf, &ehdr) == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+		pr_err("%s: cannot get elf header.\n", __func__);
 		goto out_elf_end;
 	}
 
@@ -587,9 +814,7 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 	elf_symtab__for_each_symbol(syms, nr_syms, idx, sym) {
 		struct symbol *f;
 		const char *elf_name;
-		char *demangled;
-		u64 obj_start;
-		struct section *section = NULL;
+		char *demangled = NULL;
 		int is_label = elf_sym__is_label(&sym);
 		const char *section_name;
 
@@ -605,52 +830,85 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 		if (is_label && !elf_sec__is_text(&shdr, secstrs))
 			continue;
 
+		elf_name = elf_sym__name(&sym, symstrs);
 		section_name = elf_sec__name(&shdr, secstrs);
-		obj_start = sym.st_value;
 
-		if (self->adjust_symbols) {
-			if (v >= 2)
-				printf("adjusting symbol: st_value: %Lx sh_addr: %Lx sh_offset: %Lx\n",
-					(u64)sym.st_value, (u64)shdr.sh_addr, (u64)shdr.sh_offset);
+		if (kernel || kmodule) {
+			char dso_name[PATH_MAX];
 
-			sym.st_value -= shdr.sh_addr - shdr.sh_offset;
-		}
+			if (strcmp(section_name,
+				   curr_dso->short_name + dso_name_len) == 0)
+				goto new_symbol;
 
-		if (mod) {
-			section = mod->sections->find_section(mod->sections, section_name);
-			if (section)
-				sym.st_value += section->vma;
-			else {
-				fprintf(stderr, "dso__load_sym() module %s lookup of %s failed\n",
-					mod->name, section_name);
-				goto out_elf_end;
+			if (strcmp(section_name, ".text") == 0) {
+				curr_map = map;
+				curr_dso = self;
+				goto new_symbol;
 			}
+
+			snprintf(dso_name, sizeof(dso_name),
+				 "%s%s", self->short_name, section_name);
+
+			curr_map = thread__find_map_by_name(thread, dso_name);
+			if (curr_map == NULL) {
+				u64 start = sym.st_value;
+
+				if (kmodule)
+					start += map->start + shdr.sh_offset;
+
+				curr_dso = dso__new(dso_name);
+				if (curr_dso == NULL)
+					goto out_elf_end;
+				curr_map = map__new2(start, curr_dso,
+						     MAP__FUNCTION);
+				if (curr_map == NULL) {
+					dso__delete(curr_dso);
+					goto out_elf_end;
+				}
+				curr_map->map_ip = identity__map_ip;
+				curr_map->unmap_ip = identity__map_ip;
+				curr_dso->origin = DSO__ORIG_KERNEL;
+				__thread__insert_map(kthread, curr_map);
+				dsos__add(&dsos__kernel, curr_dso);
+			} else
+				curr_dso = curr_map->dso;
+
+			goto new_symbol;
+		}
+
+		if (curr_dso->adjust_symbols) {
+			pr_debug2("adjusting symbol: st_value: %Lx sh_addr: "
+				  "%Lx sh_offset: %Lx\n", (u64)sym.st_value,
+				  (u64)shdr.sh_addr, (u64)shdr.sh_offset);
+			sym.st_value -= shdr.sh_addr - shdr.sh_offset;
 		}
 		/*
 		 * We need to figure out if the object was created from C++ sources
 		 * DWARF DW_compile_unit has this, but we don't always have access
 		 * to it...
 		 */
-		elf_name = elf_sym__name(&sym, symstrs);
 		demangled = bfd_demangle(NULL, elf_name, DMGL_PARAMS | DMGL_ANSI);
 		if (demangled != NULL)
 			elf_name = demangled;
-
-		f = symbol__new(sym.st_value, sym.st_size, elf_name,
-				self->sym_priv_size, obj_start, v);
+new_symbol:
+		f = symbol__new(sym.st_value, sym.st_size, elf_name);
 		free(demangled);
 		if (!f)
 			goto out_elf_end;
 
-		if (filter && filter(self, f))
-			symbol__delete(f, self->sym_priv_size);
+		if (filter && filter(curr_map, f))
+			symbol__delete(f);
 		else {
-			f->module = mod;
-			dso__insert_symbol(self, f);
+			symbols__insert(&curr_dso->symbols[curr_map->type], f);
 			nr++;
 		}
 	}
 
+	/*
+	 * For misannotated, zeroed, ASM function sizes.
+	 */
+	if (nr > 0)
+		symbols__fixup_end(&self->symbols[map->type]);
 	err = nr;
 out_elf_end:
 	elf_end(elf);
@@ -658,63 +916,153 @@ out_close:
 	return err;
 }
 
-#define BUILD_ID_SIZE 128
+static bool dso__build_id_equal(const struct dso *self, u8 *build_id)
+{
+	return memcmp(self->build_id, build_id, sizeof(self->build_id)) == 0;
+}
 
-static char *dso__read_build_id(struct dso *self, int v)
+static bool __dsos__read_build_ids(struct list_head *head)
 {
-	int i;
+	bool have_build_id = false;
+	struct dso *pos;
+
+	list_for_each_entry(pos, head, node)
+		if (filename__read_build_id(pos->long_name, pos->build_id,
+					    sizeof(pos->build_id)) > 0) {
+			have_build_id	  = true;
+			pos->has_build_id = true;
+		}
+
+	return have_build_id;
+}
+
+bool dsos__read_build_ids(void)
+{
+	return __dsos__read_build_ids(&dsos__kernel) ||
+	       __dsos__read_build_ids(&dsos__user);
+}
+
+/*
+ * Align offset to 4 bytes as needed for note name and descriptor data.
+ */
+#define NOTE_ALIGN(n) (((n) + 3) & -4U)
+
+int filename__read_build_id(const char *filename, void *bf, size_t size)
+{
+	int fd, err = -1;
 	GElf_Ehdr ehdr;
 	GElf_Shdr shdr;
-	Elf_Data *build_id_data;
+	Elf_Data *data;
 	Elf_Scn *sec;
-	char *build_id = NULL, *bid;
-	unsigned char *raw;
+	Elf_Kind ek;
+	void *ptr;
 	Elf *elf;
-	int fd = open(self->name, O_RDONLY);
 
+	if (size < BUILD_ID_SIZE)
+		goto out;
+
+	fd = open(filename, O_RDONLY);
 	if (fd < 0)
 		goto out;
 
-	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+	elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
 	if (elf == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot read %s ELF file.\n",
-				__func__, self->name);
+		pr_debug2("%s: cannot read %s ELF file.\n", __func__, filename);
 		goto out_close;
 	}
 
+	ek = elf_kind(elf);
+	if (ek != ELF_K_ELF)
+		goto out_elf_end;
+
 	if (gelf_getehdr(elf, &ehdr) == NULL) {
-		if (v)
-			fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+		pr_err("%s: cannot get elf header.\n", __func__);
 		goto out_elf_end;
 	}
 
-	sec = elf_section_by_name(elf, &ehdr, &shdr, ".note.gnu.build-id", NULL);
-	if (sec == NULL)
-		goto out_elf_end;
+	sec = elf_section_by_name(elf, &ehdr, &shdr,
+				  ".note.gnu.build-id", NULL);
+	if (sec == NULL) {
+		sec = elf_section_by_name(elf, &ehdr, &shdr,
+					  ".notes", NULL);
+		if (sec == NULL)
+			goto out_elf_end;
+	}
 
-	build_id_data = elf_getdata(sec, NULL);
-	if (build_id_data == NULL)
-		goto out_elf_end;
-	build_id = malloc(BUILD_ID_SIZE);
-	if (build_id == NULL)
+	data = elf_getdata(sec, NULL);
+	if (data == NULL)
 		goto out_elf_end;
-	raw = build_id_data->d_buf + 16;
-	bid = build_id;
 
-	for (i = 0; i < 20; ++i) {
-		sprintf(bid, "%02x", *raw);
-		++raw;
-		bid += 2;
+	ptr = data->d_buf;
+	while (ptr < (data->d_buf + data->d_size)) {
+		GElf_Nhdr *nhdr = ptr;
+		int namesz = NOTE_ALIGN(nhdr->n_namesz),
+		    descsz = NOTE_ALIGN(nhdr->n_descsz);
+		const char *name;
+
+		ptr += sizeof(*nhdr);
+		name = ptr;
+		ptr += namesz;
+		if (nhdr->n_type == NT_GNU_BUILD_ID &&
+		    nhdr->n_namesz == sizeof("GNU")) {
+			if (memcmp(name, "GNU", sizeof("GNU")) == 0) {
+				memcpy(bf, ptr, BUILD_ID_SIZE);
+				err = BUILD_ID_SIZE;
+				break;
+			}
+		}
+		ptr += descsz;
 	}
-	if (v >= 2)
-		printf("%s(%s): %s\n", __func__, self->name, build_id);
 out_elf_end:
 	elf_end(elf);
 out_close:
 	close(fd);
 out:
-	return build_id;
+	return err;
+}
+
+int sysfs__read_build_id(const char *filename, void *build_id, size_t size)
+{
+	int fd, err = -1;
+
+	if (size < BUILD_ID_SIZE)
+		goto out;
+
+	fd = open(filename, O_RDONLY);
+	if (fd < 0)
+		goto out;
+
+	while (1) {
+		char bf[BUFSIZ];
+		GElf_Nhdr nhdr;
+		int namesz, descsz;
+
+		if (read(fd, &nhdr, sizeof(nhdr)) != sizeof(nhdr))
+			break;
+
+		namesz = NOTE_ALIGN(nhdr.n_namesz);
+		descsz = NOTE_ALIGN(nhdr.n_descsz);
+		if (nhdr.n_type == NT_GNU_BUILD_ID &&
+		    nhdr.n_namesz == sizeof("GNU")) {
+			if (read(fd, bf, namesz) != namesz)
+				break;
+			if (memcmp(bf, "GNU", sizeof("GNU")) == 0) {
+				if (read(fd, build_id,
+				    BUILD_ID_SIZE) == BUILD_ID_SIZE) {
+					err = 0;
+					break;
+				}
+			} else if (read(fd, bf, descsz) != descsz)
+				break;
+		} else {
+			int n = namesz + descsz;
+			if (read(fd, bf, n) != n)
+				break;
+		}
+	}
+	close(fd);
+out:
+	return err;
 }
 
 char dso__symtab_origin(const struct dso *self)
@@ -726,6 +1074,7 @@ char dso__symtab_origin(const struct dso *self)
 		[DSO__ORIG_UBUNTU] =   'u',
 		[DSO__ORIG_BUILDID] =  'b',
 		[DSO__ORIG_DSO] =      'd',
+		[DSO__ORIG_KMODULE] =  'K',
 	};
 
 	if (self == NULL || self->origin == DSO__ORIG_NOT_FOUND)
@@ -733,20 +1082,27 @@ char dso__symtab_origin(const struct dso *self)
 	return origin[self->origin];
 }
 
-int dso__load(struct dso *self, symbol_filter_t filter, int v)
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 {
 	int size = PATH_MAX;
-	char *name = malloc(size), *build_id = NULL;
+	char *name;
+	u8 build_id[BUILD_ID_SIZE];
 	int ret = -1;
 	int fd;
 
+	dso__set_loaded(self, map->type);
+
+	if (self->kernel)
+		return dso__load_kernel_sym(self, map, kthread, filter);
+
+	name = malloc(size);
 	if (!name)
 		return -1;
 
 	self->adjust_symbols = 0;
 
 	if (strncmp(self->name, "/tmp/perf-", 10) == 0) {
-		ret = dso__load_perf_map(self, filter, v);
+		ret = dso__load_perf_map(self, map, filter);
 		self->origin = ret > 0 ? DSO__ORIG_JAVA_JIT :
 					 DSO__ORIG_NOT_FOUND;
 		return ret;
@@ -759,34 +1115,50 @@ more:
 		self->origin++;
 		switch (self->origin) {
 		case DSO__ORIG_FEDORA:
-			snprintf(name, size, "/usr/lib/debug%s.debug", self->name);
+			snprintf(name, size, "/usr/lib/debug%s.debug",
+				 self->long_name);
 			break;
 		case DSO__ORIG_UBUNTU:
-			snprintf(name, size, "/usr/lib/debug%s", self->name);
+			snprintf(name, size, "/usr/lib/debug%s",
+				 self->long_name);
 			break;
 		case DSO__ORIG_BUILDID:
-			build_id = dso__read_build_id(self, v);
-			if (build_id != NULL) {
+			if (filename__read_build_id(self->long_name, build_id,
+						    sizeof(build_id))) {
+				char build_id_hex[BUILD_ID_SIZE * 2 + 1];
+
+				build_id__sprintf(build_id, sizeof(build_id),
+						  build_id_hex);
 				snprintf(name, size,
 					 "/usr/lib/debug/.build-id/%.2s/%s.debug",
-					build_id, build_id + 2);
-				free(build_id);
+					build_id_hex, build_id_hex + 2);
+				if (self->has_build_id)
+					goto compare_build_id;
 				break;
 			}
 			self->origin++;
 			/* Fall thru */
 		case DSO__ORIG_DSO:
-			snprintf(name, size, "%s", self->name);
+			snprintf(name, size, "%s", self->long_name);
 			break;
 
 		default:
 			goto out;
 		}
 
+		if (self->has_build_id) {
+			if (filename__read_build_id(name, build_id,
+						    sizeof(build_id)) < 0)
+				goto more;
+compare_build_id:
+			if (!dso__build_id_equal(self, build_id))
+				goto more;
+		}
+
 		fd = open(name, O_RDONLY);
 	} while (fd < 0);
 
-	ret = dso__load_sym(self, fd, name, filter, v, NULL);
+	ret = dso__load_sym(self, map, NULL, name, fd, filter, 0, 0);
 	close(fd);
 
 	/*
@@ -796,7 +1168,7 @@ more:
 		goto more;
 
 	if (ret > 0) {
-		int nr_plt = dso__synthesize_plt_symbols(self, v);
+		int nr_plt = dso__synthesize_plt_symbols(self, map, filter);
 		if (nr_plt > 0)
 			ret += nr_plt;
 	}
@@ -807,151 +1179,279 @@ out:
 	return ret;
 }
 
-static int dso__load_module(struct dso *self, struct mod_dso *mods, const char *name,
-			     symbol_filter_t filter, int v)
+static struct map *thread__find_map_by_name(struct thread *self, char *name)
 {
-	struct module *mod = mod_dso__find_module(mods, name);
-	int err = 0, fd;
+	struct rb_node *nd;
 
-	if (mod == NULL || !mod->active)
-		return err;
+	for (nd = rb_first(&self->maps[MAP__FUNCTION]); nd; nd = rb_next(nd)) {
+		struct map *map = rb_entry(nd, struct map, rb_node);
 
-	fd = open(mod->path, O_RDONLY);
+		if (map->dso && strcmp(map->dso->name, name) == 0)
+			return map;
+	}
 
-	if (fd < 0)
-		return err;
+	return NULL;
+}
 
-	err = dso__load_sym(self, fd, name, filter, v, mod);
-	close(fd);
+static int dsos__set_modules_path_dir(char *dirname)
+{
+	struct dirent *dent;
+	DIR *dir = opendir(dirname);
 
-	return err;
+	if (!dir) {
+		pr_debug("%s: cannot open %s dir\n", __func__, dirname);
+		return -1;
+	}
+
+	while ((dent = readdir(dir)) != NULL) {
+		char path[PATH_MAX];
+
+		if (dent->d_type == DT_DIR) {
+			if (!strcmp(dent->d_name, ".") ||
+			    !strcmp(dent->d_name, ".."))
+				continue;
+
+			snprintf(path, sizeof(path), "%s/%s",
+				 dirname, dent->d_name);
+			if (dsos__set_modules_path_dir(path) < 0)
+				goto failure;
+		} else {
+			char *dot = strrchr(dent->d_name, '.'),
+			     dso_name[PATH_MAX];
+			struct map *map;
+			char *long_name;
+
+			if (dot == NULL || strcmp(dot, ".ko"))
+				continue;
+			snprintf(dso_name, sizeof(dso_name), "[%.*s]",
+				 (int)(dot - dent->d_name), dent->d_name);
+
+			strxfrchar(dso_name, '-', '_');
+			map = thread__find_map_by_name(kthread, dso_name);
+			if (map == NULL)
+				continue;
+
+			snprintf(path, sizeof(path), "%s/%s",
+				 dirname, dent->d_name);
+
+			long_name = strdup(path);
+			if (long_name == NULL)
+				goto failure;
+			dso__set_long_name(map->dso, long_name);
+		}
+	}
+
+	return 0;
+failure:
+	closedir(dir);
+	return -1;
 }
 
-int dso__load_modules(struct dso *self, symbol_filter_t filter, int v)
+static int dsos__set_modules_path(void)
 {
-	struct mod_dso *mods = mod_dso__new_dso("modules");
-	struct module *pos;
-	struct rb_node *next;
-	int err, count = 0;
+	struct utsname uts;
+	char modules_path[PATH_MAX];
 
-	err = mod_dso__load_modules(mods);
-
-	if (err <= 0)
-		return err;
+	if (uname(&uts) < 0)
+		return -1;
 
-	/*
-	 * Iterate over modules, and load active symbols.
-	 */
-	next = rb_first(&mods->mods);
-	while (next) {
-		pos = rb_entry(next, struct module, rb_node);
-		err = dso__load_module(self, mods, pos->name, filter, v);
+	snprintf(modules_path, sizeof(modules_path), "/lib/modules/%s/kernel",
+		 uts.release);
 
-		if (err < 0)
-			break;
+	return dsos__set_modules_path_dir(modules_path);
+}
 
-		next = rb_next(&pos->rb_node);
-		count += err;
-	}
+/*
+ * Constructor variant for modules (where we know from /proc/modules where
+ * they are loaded) and for vmlinux, where only after we load all the
+ * symbols we'll know where it starts and ends.
+ */
+static struct map *map__new2(u64 start, struct dso *dso, enum map_type type)
+{
+	struct map *self = malloc(sizeof(*self));
 
-	if (err < 0) {
-		mod_dso__delete_modules(mods);
-		mod_dso__delete_self(mods);
-		return err;
+	if (self != NULL) {
+		/*
+		 * ->end will be filled after we load all the symbols
+		 */
+		map__init(self, type, start, 0, 0, dso);
 	}
 
-	return count;
+	return self;
 }
 
-static inline void dso__fill_symbol_holes(struct dso *self)
+static int thread__create_module_maps(struct thread *self)
 {
-	struct symbol *prev = NULL;
-	struct rb_node *nd;
+	char *line = NULL;
+	size_t n;
+	FILE *file = fopen("/proc/modules", "r");
+	struct map *map;
 
-	for (nd = rb_last(&self->syms); nd; nd = rb_prev(nd)) {
-		struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
+	if (file == NULL)
+		return -1;
 
-		if (prev) {
-			u64 hole = 0;
-			int alias = pos->start == prev->start;
+	while (!feof(file)) {
+		char name[PATH_MAX];
+		u64 start;
+		struct dso *dso;
+		char *sep;
+		int line_len;
 
-			if (!alias)
-				hole = prev->start - pos->end - 1;
+		line_len = getline(&line, &n, file);
+		if (line_len < 0)
+			break;
 
-			if (hole || alias) {
-				if (alias)
-					pos->end = prev->end;
-				else if (hole)
-					pos->end = prev->start - 1;
-			}
+		if (!line)
+			goto out_failure;
+
+		line[--line_len] = '\0'; /* \n */
+
+		sep = strrchr(line, 'x');
+		if (sep == NULL)
+			continue;
+
+		hex2u64(sep + 1, &start);
+
+		sep = strchr(line, ' ');
+		if (sep == NULL)
+			continue;
+
+		*sep = '\0';
+
+		snprintf(name, sizeof(name), "[%s]", line);
+		dso = dso__new(name);
+
+		if (dso == NULL)
+			goto out_delete_line;
+
+		map = map__new2(start, dso, MAP__FUNCTION);
+		if (map == NULL) {
+			dso__delete(dso);
+			goto out_delete_line;
 		}
-		prev = pos;
+
+		snprintf(name, sizeof(name),
+			 "/sys/module/%s/notes/.note.gnu.build-id", line);
+		if (sysfs__read_build_id(name, dso->build_id,
+					 sizeof(dso->build_id)) == 0)
+			dso->has_build_id = true;
+
+		dso->origin = DSO__ORIG_KMODULE;
+		__thread__insert_map(self, map);
+		dsos__add(&dsos__kernel, dso);
 	}
+
+	free(line);
+	fclose(file);
+
+	return dsos__set_modules_path();
+
+out_delete_line:
+	free(line);
+out_failure:
+	return -1;
 }
 
-static int dso__load_vmlinux(struct dso *self, const char *vmlinux,
-			     symbol_filter_t filter, int v)
+static int dso__load_vmlinux(struct dso *self, struct map *map, struct thread *thread,
+			     const char *vmlinux, symbol_filter_t filter)
 {
-	int err, fd = open(vmlinux, O_RDONLY);
+	int err = -1, fd;
 
-	if (fd < 0)
-		return -1;
+	if (self->has_build_id) {
+		u8 build_id[BUILD_ID_SIZE];
 
-	err = dso__load_sym(self, fd, vmlinux, filter, v, NULL);
+		if (filename__read_build_id(vmlinux, build_id,
+					    sizeof(build_id)) < 0) {
+			pr_debug("No build_id in %s, ignoring it\n", vmlinux);
+			return -1;
+		}
+		if (!dso__build_id_equal(self, build_id)) {
+			char expected_build_id[BUILD_ID_SIZE * 2 + 1],
+			     vmlinux_build_id[BUILD_ID_SIZE * 2 + 1];
+
+			build_id__sprintf(self->build_id,
+					  sizeof(self->build_id),
+					  expected_build_id);
+			build_id__sprintf(build_id, sizeof(build_id),
+					  vmlinux_build_id);
+			pr_debug("build_id in %s is %s while expected is %s, "
+				 "ignoring it\n", vmlinux, vmlinux_build_id,
+				 expected_build_id);
+			return -1;
+		}
+	}
 
-	if (err > 0)
-		dso__fill_symbol_holes(self);
+	fd = open(vmlinux, O_RDONLY);
+	if (fd < 0)
+		return -1;
 
+	dso__set_loaded(self, map->type);
+	err = dso__load_sym(self, map, thread, self->long_name, fd, filter, 1, 0);
 	close(fd);
 
 	return err;
 }
 
-int dso__load_kernel(struct dso *self, const char *vmlinux,
-		     symbol_filter_t filter, int v, int use_modules)
+static int dso__load_kernel_sym(struct dso *self, struct map *map,
+				struct thread *thread, symbol_filter_t filter)
 {
-	int err = -1;
-
-	if (vmlinux) {
-		err = dso__load_vmlinux(self, vmlinux, filter, v);
-		if (err > 0 && use_modules) {
-			int syms = dso__load_modules(self, filter, v);
-
-			if (syms < 0) {
-				fprintf(stderr, "dso__load_modules failed!\n");
-				return syms;
+	int err;
+	bool is_kallsyms;
+
+	if (vmlinux_path != NULL) {
+		int i;
+		pr_debug("Looking at the vmlinux_path (%d entries long)\n",
+			 vmlinux_path__nr_entries);
+		for (i = 0; i < vmlinux_path__nr_entries; ++i) {
+			err = dso__load_vmlinux(self, map, thread,
+						vmlinux_path[i], filter);
+			if (err > 0) {
+				pr_debug("Using %s for symbols\n",
+					 vmlinux_path[i]);
+				dso__set_long_name(self,
+						   strdup(vmlinux_path[i]));
+				goto out_fixup;
 			}
-			err += syms;
 		}
 	}
 
-	if (err <= 0)
-		err = dso__load_kallsyms(self, filter, v);
+	is_kallsyms = self->long_name[0] == '[';
+	if (is_kallsyms)
+		goto do_kallsyms;
 
-	if (err > 0)
-		self->origin = DSO__ORIG_KERNEL;
+	err = dso__load_vmlinux(self, map, thread, self->long_name, filter);
+	if (err <= 0) {
+		pr_info("The file %s cannot be used, "
+			"trying to use /proc/kallsyms...", self->long_name);
+do_kallsyms:
+		err = dso__load_kallsyms(self, map, thread, filter);
+		if (err > 0 && !is_kallsyms)
+                        dso__set_long_name(self, strdup("[kernel.kallsyms]"));
+	}
+
+	if (err > 0) {
+out_fixup:
+		map__fixup_start(map);
+		map__fixup_end(map);
+	}
 
 	return err;
 }
 
-LIST_HEAD(dsos);
-struct dso	*kernel_dso;
-struct dso	*vdso;
-struct dso	*hypervisor_dso;
-
-const char	*vmlinux_name = "vmlinux";
-int		modules;
+LIST_HEAD(dsos__user);
+LIST_HEAD(dsos__kernel);
+struct dso *vdso;
 
-static void dsos__add(struct dso *dso)
+static void dsos__add(struct list_head *head, struct dso *dso)
 {
-	list_add_tail(&dso->node, &dsos);
+	list_add_tail(&dso->node, head);
 }
 
-static struct dso *dsos__find(const char *name)
+static struct dso *dsos__find(struct list_head *head, const char *name)
 {
 	struct dso *pos;
 
-	list_for_each_entry(pos, &dsos, node)
+	list_for_each_entry(pos, head, node)
 		if (strcmp(pos->name, name) == 0)
 			return pos;
 	return NULL;
@@ -959,79 +1459,170 @@ static struct dso *dsos__find(const char *name)
 
 struct dso *dsos__findnew(const char *name)
 {
-	struct dso *dso = dsos__find(name);
-	int nr;
-
-	if (dso)
-		return dso;
-
-	dso = dso__new(name, 0);
-	if (!dso)
-		goto out_delete_dso;
+	struct dso *dso = dsos__find(&dsos__user, name);
 
-	nr = dso__load(dso, NULL, verbose);
-	if (nr < 0) {
-		eprintf("Failed to open: %s\n", name);
-		goto out_delete_dso;
+	if (!dso) {
+		dso = dso__new(name);
+		if (dso != NULL) {
+			dsos__add(&dsos__user, dso);
+			dso__set_basename(dso);
+		}
 	}
-	if (!nr)
-		eprintf("No symbols found in: %s, maybe install a debug package?\n", name);
-
-	dsos__add(dso);
 
 	return dso;
+}
 
-out_delete_dso:
-	dso__delete(dso);
-	return NULL;
+static void __dsos__fprintf(struct list_head *head, FILE *fp)
+{
+	struct dso *pos;
+
+	list_for_each_entry(pos, head, node) {
+		int i;
+		for (i = 0; i < MAP__NR_TYPES; ++i)
+			dso__fprintf(pos, i, fp);
+	}
 }
 
 void dsos__fprintf(FILE *fp)
 {
+	__dsos__fprintf(&dsos__kernel, fp);
+	__dsos__fprintf(&dsos__user, fp);
+}
+
+static size_t __dsos__fprintf_buildid(struct list_head *head, FILE *fp)
+{
 	struct dso *pos;
+	size_t ret = 0;
 
-	list_for_each_entry(pos, &dsos, node)
-		dso__fprintf(pos, fp);
+	list_for_each_entry(pos, head, node) {
+		ret += dso__fprintf_buildid(pos, fp);
+		ret += fprintf(fp, " %s\n", pos->long_name);
+	}
+	return ret;
 }
 
-static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip)
+size_t dsos__fprintf_buildid(FILE *fp)
 {
-	return dso__find_symbol(dso, ip);
+	return (__dsos__fprintf_buildid(&dsos__kernel, fp) +
+		__dsos__fprintf_buildid(&dsos__user, fp));
 }
 
-int load_kernel(void)
+static int thread__create_kernel_map(struct thread *self, const char *vmlinux)
 {
-	int err;
+	struct map *kmap;
+	struct dso *kernel = dso__new(vmlinux ?: "[kernel.kallsyms]");
 
-	kernel_dso = dso__new("[kernel]", 0);
-	if (!kernel_dso)
+	if (kernel == NULL)
 		return -1;
 
-	err = dso__load_kernel(kernel_dso, vmlinux_name, NULL, verbose, modules);
-	if (err <= 0) {
-		dso__delete(kernel_dso);
-		kernel_dso = NULL;
-	} else
-		dsos__add(kernel_dso);
+	kmap = map__new2(0, kernel, MAP__FUNCTION);
+	if (kmap == NULL)
+		goto out_delete_kernel_dso;
 
-	vdso = dso__new("[vdso]", 0);
-	if (!vdso)
-		return -1;
+	kmap->map_ip	   = kmap->unmap_ip = identity__map_ip;
+	kernel->short_name = "[kernel]";
+	kernel->kernel	   = 1;
 
-	vdso->find_symbol = vdso__find_symbol;
+	vdso = dso__new("[vdso]");
+	if (vdso == NULL)
+		goto out_delete_kernel_map;
+	dso__set_loaded(vdso, MAP__FUNCTION);
 
-	dsos__add(vdso);
+	if (sysfs__read_build_id("/sys/kernel/notes", kernel->build_id,
+				 sizeof(kernel->build_id)) == 0)
+		kernel->has_build_id = true;
 
-	hypervisor_dso = dso__new("[hypervisor]", 0);
-	if (!hypervisor_dso)
-		return -1;
-	dsos__add(hypervisor_dso);
+	__thread__insert_map(self, kmap);
+	dsos__add(&dsos__kernel, kernel);
+	dsos__add(&dsos__user, vdso);
 
-	return err;
+	return 0;
+
+out_delete_kernel_map:
+	map__delete(kmap);
+out_delete_kernel_dso:
+	dso__delete(kernel);
+	return -1;
+}
+
+static void vmlinux_path__exit(void)
+{
+	while (--vmlinux_path__nr_entries >= 0) {
+		free(vmlinux_path[vmlinux_path__nr_entries]);
+		vmlinux_path[vmlinux_path__nr_entries] = NULL;
+	}
+
+	free(vmlinux_path);
+	vmlinux_path = NULL;
 }
 
+static int vmlinux_path__init(void)
+{
+	struct utsname uts;
+	char bf[PATH_MAX];
+
+	if (uname(&uts) < 0)
+		return -1;
+
+	vmlinux_path = malloc(sizeof(char *) * 5);
+	if (vmlinux_path == NULL)
+		return -1;
+
+	vmlinux_path[vmlinux_path__nr_entries] = strdup("vmlinux");
+	if (vmlinux_path[vmlinux_path__nr_entries] == NULL)
+		goto out_fail;
+	++vmlinux_path__nr_entries;
+	vmlinux_path[vmlinux_path__nr_entries] = strdup("/boot/vmlinux");
+	if (vmlinux_path[vmlinux_path__nr_entries] == NULL)
+		goto out_fail;
+	++vmlinux_path__nr_entries;
+	snprintf(bf, sizeof(bf), "/boot/vmlinux-%s", uts.release);
+	vmlinux_path[vmlinux_path__nr_entries] = strdup(bf);
+	if (vmlinux_path[vmlinux_path__nr_entries] == NULL)
+		goto out_fail;
+	++vmlinux_path__nr_entries;
+	snprintf(bf, sizeof(bf), "/lib/modules/%s/build/vmlinux", uts.release);
+	vmlinux_path[vmlinux_path__nr_entries] = strdup(bf);
+	if (vmlinux_path[vmlinux_path__nr_entries] == NULL)
+		goto out_fail;
+	++vmlinux_path__nr_entries;
+	snprintf(bf, sizeof(bf), "/usr/lib/debug/lib/modules/%s/vmlinux",
+		 uts.release);
+	vmlinux_path[vmlinux_path__nr_entries] = strdup(bf);
+	if (vmlinux_path[vmlinux_path__nr_entries] == NULL)
+		goto out_fail;
+	++vmlinux_path__nr_entries;
+
+	return 0;
+
+out_fail:
+	vmlinux_path__exit();
+	return -1;
+}
 
-void symbol__init(void)
+int symbol__init(struct symbol_conf *conf)
 {
+	const struct symbol_conf *pconf = conf ?: &symbol_conf__defaults;
+
 	elf_version(EV_CURRENT);
+	symbol__priv_size = pconf->priv_size;
+	thread__init(kthread, 0);
+
+	if (pconf->try_vmlinux_path && vmlinux_path__init() < 0)
+		return -1;
+
+	if (thread__create_kernel_map(kthread, pconf->vmlinux_name) < 0) {
+		vmlinux_path__exit();
+		return -1;
+	}
+
+	kthread->use_modules = pconf->use_modules;
+	if (pconf->use_modules && thread__create_module_maps(kthread) < 0)
+		pr_debug("Failed to load list of modules in use, "
+			 "continuing...\n");
+	/*
+	 * Now that we have all the maps created, just set the ->end of them:
+	 */
+	thread__fixup_maps_end(kthread);
+	return 0;
 }
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 6e8490716408..17003efa0b39 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -1,11 +1,11 @@
-#ifndef _PERF_SYMBOL_
-#define _PERF_SYMBOL_ 1
+#ifndef __PERF_SYMBOL
+#define __PERF_SYMBOL 1
 
 #include <linux/types.h>
+#include <stdbool.h>
 #include "types.h"
 #include <linux/list.h>
 #include <linux/rbtree.h>
-#include "module.h"
 #include "event.h"
 
 #ifdef HAVE_CPLUS_DEMANGLE
@@ -27,6 +27,16 @@ static inline char *bfd_demangle(void __used *v, const char __used *c,
 #endif
 #endif
 
+/*
+ * libelf 0.8.x and earlier do not support ELF_C_READ_MMAP;
+ * for newer versions we can use mmap to reduce memory usage:
+ */
+#ifdef LIBELF_NO_MMAP
+# define PERF_ELF_C_READ_MMAP ELF_C_READ
+#else
+# define PERF_ELF_C_READ_MMAP ELF_C_READ_MMAP
+#endif
+
 #ifndef DMGL_PARAMS
 #define DMGL_PARAMS      (1 << 0)       /* Include function args */
 #define DMGL_ANSI        (1 << 1)       /* Include const, volatile, etc */
@@ -36,57 +46,75 @@ struct symbol {
 	struct rb_node	rb_node;
 	u64		start;
 	u64		end;
-	u64		obj_start;
-	u64		hist_sum;
-	u64		*hist;
-	struct module	*module;
-	void		*priv;
 	char		name[0];
 };
 
+struct symbol_conf {
+	unsigned short	priv_size;
+	bool		try_vmlinux_path,
+			use_modules;
+	const char	*vmlinux_name;
+};
+
+extern unsigned int symbol__priv_size;
+
+static inline void *symbol__priv(struct symbol *self)
+{
+	return ((void *)self) - symbol__priv_size;
+}
+
+struct addr_location {
+	struct thread *thread;
+	struct map    *map;
+	struct symbol *sym;
+	u64	      addr;
+	char	      level;
+};
+
 struct dso {
 	struct list_head node;
-	struct rb_root	 syms;
-	struct symbol    *(*find_symbol)(struct dso *, u64 ip);
-	unsigned int	 sym_priv_size;
-	unsigned char	 adjust_symbols;
-	unsigned char	 slen_calculated;
+	struct rb_root	 symbols[MAP__NR_TYPES];
+	struct symbol    *(*find_symbol)(struct dso *self,
+					 enum map_type type, u64 addr);
+	u8		 adjust_symbols:1;
+	u8		 slen_calculated:1;
+	u8		 has_build_id:1;
+	u8		 kernel:1;
 	unsigned char	 origin;
+	u8		 loaded;
+	u8		 build_id[BUILD_ID_SIZE];
+	u16		 long_name_len;
+	const char	 *short_name;
+	char	 	 *long_name;
 	char		 name[0];
 };
 
-extern const char *sym_hist_filter;
-
-typedef int (*symbol_filter_t)(struct dso *self, struct symbol *sym);
-
-struct dso *dso__new(const char *name, unsigned int sym_priv_size);
+struct dso *dso__new(const char *name);
 void dso__delete(struct dso *self);
 
-static inline void *dso__sym_priv(struct dso *self, struct symbol *sym)
-{
-	return ((void *)sym) - self->sym_priv_size;
-}
+bool dso__loaded(const struct dso *self, enum map_type type);
 
-struct symbol *dso__find_symbol(struct dso *self, u64 ip);
-
-int dso__load_kernel(struct dso *self, const char *vmlinux,
-		     symbol_filter_t filter, int verbose, int modules);
-int dso__load_modules(struct dso *self, symbol_filter_t filter, int verbose);
-int dso__load(struct dso *self, symbol_filter_t filter, int verbose);
 struct dso *dsos__findnew(const char *name);
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter);
 void dsos__fprintf(FILE *fp);
+size_t dsos__fprintf_buildid(FILE *fp);
 
-size_t dso__fprintf(struct dso *self, FILE *fp);
+size_t dso__fprintf_buildid(struct dso *self, FILE *fp);
+size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp);
 char dso__symtab_origin(const struct dso *self);
+void dso__set_build_id(struct dso *self, void *build_id);
+
+int filename__read_build_id(const char *filename, void *bf, size_t size);
+int sysfs__read_build_id(const char *filename, void *bf, size_t size);
+bool dsos__read_build_ids(void);
+int build_id__sprintf(u8 *self, int len, char *bf);
 
-int load_kernel(void);
+size_t kernel_maps__fprintf(FILE *fp);
 
-void symbol__init(void);
+int symbol__init(struct symbol_conf *conf);
 
-extern struct list_head dsos;
-extern struct dso *kernel_dso;
+struct thread;
+struct thread *kthread;
+extern struct list_head dsos__user, dsos__kernel;
 extern struct dso *vdso;
-extern struct dso *hypervisor_dso;
-extern const char *vmlinux_name;
-extern int   modules;
-#endif /* _PERF_SYMBOL_ */
+#endif /* __PERF_SYMBOL */
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index 45efb5db0d19..603f5610861b 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -6,16 +6,29 @@
 #include "util.h"
 #include "debug.h"
 
+static struct rb_root threads;
+static struct thread *last_match;
+
+void thread__init(struct thread *self, pid_t pid)
+{
+	int i;
+	self->pid = pid;
+	self->comm = NULL;
+	for (i = 0; i < MAP__NR_TYPES; ++i) {
+		self->maps[i] = RB_ROOT;
+		INIT_LIST_HEAD(&self->removed_maps[i]);
+	}
+}
+
 static struct thread *thread__new(pid_t pid)
 {
-	struct thread *self = calloc(1, sizeof(*self));
+	struct thread *self = zalloc(sizeof(*self));
 
 	if (self != NULL) {
-		self->pid = pid;
+		thread__init(self, pid);
 		self->comm = malloc(32);
 		if (self->comm)
 			snprintf(self->comm, 32, ":%d", self->pid);
-		INIT_LIST_HEAD(&self->maps);
 	}
 
 	return self;
@@ -29,21 +42,84 @@ int thread__set_comm(struct thread *self, const char *comm)
 	return self->comm ? 0 : -ENOMEM;
 }
 
-static size_t thread__fprintf(struct thread *self, FILE *fp)
+int thread__comm_len(struct thread *self)
+{
+	if (!self->comm_len) {
+		if (!self->comm)
+			return 0;
+		self->comm_len = strlen(self->comm);
+	}
+
+	return self->comm_len;
+}
+
+static const char *map_type__name[MAP__NR_TYPES] = {
+	[MAP__FUNCTION] = "Functions",
+};
+
+static size_t __thread__fprintf_maps(struct thread *self,
+				     enum map_type type, FILE *fp)
+{
+	size_t printed = fprintf(fp, "%s:\n", map_type__name[type]);
+	struct rb_node *nd;
+
+	for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) {
+		struct map *pos = rb_entry(nd, struct map, rb_node);
+		printed += fprintf(fp, "Map:");
+		printed += map__fprintf(pos, fp);
+		if (verbose > 1) {
+			printed += dso__fprintf(pos->dso, type, fp);
+			printed += fprintf(fp, "--\n");
+		}
+	}
+
+	return printed;
+}
+
+size_t thread__fprintf_maps(struct thread *self, FILE *fp)
+{
+	size_t printed = 0, i;
+	for (i = 0; i < MAP__NR_TYPES; ++i)
+		printed += __thread__fprintf_maps(self, i, fp);
+	return printed;
+}
+
+static size_t __thread__fprintf_removed_maps(struct thread *self,
+					     enum map_type type, FILE *fp)
 {
 	struct map *pos;
-	size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm);
+	size_t printed = 0;
+
+	list_for_each_entry(pos, &self->removed_maps[type], node) {
+		printed += fprintf(fp, "Map:");
+		printed += map__fprintf(pos, fp);
+		if (verbose > 1) {
+			printed += dso__fprintf(pos->dso, type, fp);
+			printed += fprintf(fp, "--\n");
+		}
+	}
+	return printed;
+}
 
-	list_for_each_entry(pos, &self->maps, node)
-		ret += map__fprintf(pos, fp);
+static size_t thread__fprintf_removed_maps(struct thread *self, FILE *fp)
+{
+	size_t printed = 0, i;
+	for (i = 0; i < MAP__NR_TYPES; ++i)
+		printed += __thread__fprintf_removed_maps(self, i, fp);
+	return printed;
+}
 
-	return ret;
+static size_t thread__fprintf(struct thread *self, FILE *fp)
+{
+	size_t printed = fprintf(fp, "Thread %d %s\n", self->pid, self->comm);
+	printed += thread__fprintf_removed_maps(self, fp);
+	printed += fprintf(fp, "Removed maps:\n");
+	return printed + thread__fprintf_removed_maps(self, fp);
 }
 
-struct thread *
-threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
+struct thread *threads__findnew(pid_t pid)
 {
-	struct rb_node **p = &threads->rb_node;
+	struct rb_node **p = &threads.rb_node;
 	struct rb_node *parent = NULL;
 	struct thread *th;
 
@@ -52,15 +128,15 @@ threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
 	 * so most of the time we dont have to look up
 	 * the full rbtree:
 	 */
-	if (*last_match && (*last_match)->pid == pid)
-		return *last_match;
+	if (last_match && last_match->pid == pid)
+		return last_match;
 
 	while (*p != NULL) {
 		parent = *p;
 		th = rb_entry(parent, struct thread, rb_node);
 
 		if (th->pid == pid) {
-			*last_match = th;
+			last_match = th;
 			return th;
 		}
 
@@ -73,17 +149,16 @@ threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
 	th = thread__new(pid);
 	if (th != NULL) {
 		rb_link_node(&th->rb_node, parent, p);
-		rb_insert_color(&th->rb_node, threads);
-		*last_match = th;
+		rb_insert_color(&th->rb_node, &threads);
+		last_match = th;
 	}
 
 	return th;
 }
 
-struct thread *
-register_idle_thread(struct rb_root *threads, struct thread **last_match)
+struct thread *register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0, threads, last_match);
+	struct thread *thread = threads__findnew(0);
 
 	if (!thread || thread__set_comm(thread, "swapper")) {
 		fprintf(stderr, "problem inserting idle task.\n");
@@ -93,79 +168,116 @@ register_idle_thread(struct rb_root *threads, struct thread **last_match)
 	return thread;
 }
 
-void thread__insert_map(struct thread *self, struct map *map)
+static void thread__remove_overlappings(struct thread *self, struct map *map)
 {
-	struct map *pos, *tmp;
+	struct rb_root *root = &self->maps[map->type];
+	struct rb_node *next = rb_first(root);
 
-	list_for_each_entry_safe(pos, tmp, &self->maps, node) {
-		if (map__overlap(pos, map)) {
-			if (verbose >= 2) {
-				printf("overlapping maps:\n");
-				map__fprintf(map, stdout);
-				map__fprintf(pos, stdout);
-			}
+	while (next) {
+		struct map *pos = rb_entry(next, struct map, rb_node);
+		next = rb_next(&pos->rb_node);
 
-			if (map->start <= pos->start && map->end > pos->start)
-				pos->start = map->end;
+		if (!map__overlap(pos, map))
+			continue;
 
-			if (map->end >= pos->end && map->start < pos->end)
-				pos->end = map->start;
+		if (verbose >= 2) {
+			fputs("overlapping maps:\n", stderr);
+			map__fprintf(map, stderr);
+			map__fprintf(pos, stderr);
+		}
 
-			if (verbose >= 2) {
-				printf("after collision:\n");
-				map__fprintf(pos, stdout);
-			}
+		rb_erase(&pos->rb_node, root);
+		/*
+		 * We may have references to this map, for instance in some
+		 * hist_entry instances, so just move them to a separate
+		 * list.
+		 */
+		list_add_tail(&pos->node, &self->removed_maps[map->type]);
+	}
+}
 
-			if (pos->start >= pos->end) {
-				list_del_init(&pos->node);
-				free(pos);
-			}
-		}
+void maps__insert(struct rb_root *maps, struct map *map)
+{
+	struct rb_node **p = &maps->rb_node;
+	struct rb_node *parent = NULL;
+	const u64 ip = map->start;
+	struct map *m;
+
+	while (*p != NULL) {
+		parent = *p;
+		m = rb_entry(parent, struct map, rb_node);
+		if (ip < m->start)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
 	}
 
-	list_add_tail(&map->node, &self->maps);
+	rb_link_node(&map->rb_node, parent, p);
+	rb_insert_color(&map->rb_node, maps);
 }
 
-int thread__fork(struct thread *self, struct thread *parent)
+struct map *maps__find(struct rb_root *maps, u64 ip)
 {
-	struct map *map;
+	struct rb_node **p = &maps->rb_node;
+	struct rb_node *parent = NULL;
+	struct map *m;
 
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(parent->comm);
-	if (!self->comm)
-		return -ENOMEM;
+	while (*p != NULL) {
+		parent = *p;
+		m = rb_entry(parent, struct map, rb_node);
+		if (ip < m->start)
+			p = &(*p)->rb_left;
+		else if (ip > m->end)
+			p = &(*p)->rb_right;
+		else
+			return m;
+	}
+
+	return NULL;
+}
+
+void thread__insert_map(struct thread *self, struct map *map)
+{
+	thread__remove_overlappings(self, map);
+	maps__insert(&self->maps[map->type], map);
+}
 
-	list_for_each_entry(map, &parent->maps, node) {
+static int thread__clone_maps(struct thread *self, struct thread *parent,
+			      enum map_type type)
+{
+	struct rb_node *nd;
+	for (nd = rb_first(&parent->maps[type]); nd; nd = rb_next(nd)) {
+		struct map *map = rb_entry(nd, struct map, rb_node);
 		struct map *new = map__clone(map);
-		if (!new)
+		if (new == NULL)
 			return -ENOMEM;
 		thread__insert_map(self, new);
 	}
-
 	return 0;
 }
 
-struct map *thread__find_map(struct thread *self, u64 ip)
+int thread__fork(struct thread *self, struct thread *parent)
 {
-	struct map *pos;
+	int i;
 
-	if (self == NULL)
-		return NULL;
-
-	list_for_each_entry(pos, &self->maps, node)
-		if (ip >= pos->start && ip <= pos->end)
-			return pos;
+	if (self->comm)
+		free(self->comm);
+	self->comm = strdup(parent->comm);
+	if (!self->comm)
+		return -ENOMEM;
 
-	return NULL;
+	for (i = 0; i < MAP__NR_TYPES; ++i)
+		if (thread__clone_maps(self, parent, i) < 0)
+			return -ENOMEM;
+	return 0;
 }
 
-size_t threads__fprintf(FILE *fp, struct rb_root *threads)
+size_t threads__fprintf(FILE *fp)
 {
 	size_t ret = 0;
 	struct rb_node *nd;
 
-	for (nd = rb_first(threads); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&threads); nd; nd = rb_next(nd)) {
 		struct thread *pos = rb_entry(nd, struct thread, rb_node);
 
 		ret += thread__fprintf(pos, fp);
@@ -173,3 +285,15 @@ size_t threads__fprintf(FILE *fp, struct rb_root *threads)
 
 	return ret;
 }
+
+struct symbol *thread__find_symbol(struct thread *self,
+				   enum map_type type, u64 addr,
+				   symbol_filter_t filter)
+{
+	struct map *map = thread__find_map(self, type, addr);
+
+	if (map != NULL)
+		return map__find_symbol(map, map->map_ip(map, addr), filter);
+
+	return NULL;
+}
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
index 32aea3c1c2ad..686d6e914d9e 100644
--- a/tools/perf/util/thread.h
+++ b/tools/perf/util/thread.h
@@ -1,22 +1,56 @@
+#ifndef __PERF_THREAD_H
+#define __PERF_THREAD_H
+
 #include <linux/rbtree.h>
-#include <linux/list.h>
 #include <unistd.h>
 #include "symbol.h"
 
 struct thread {
 	struct rb_node		rb_node;
-	struct list_head	maps;
+	struct rb_root		maps[MAP__NR_TYPES];
+	struct list_head	removed_maps[MAP__NR_TYPES];
 	pid_t			pid;
+	bool			use_modules;
 	char			shortname[3];
 	char			*comm;
+	int			comm_len;
 };
 
+void thread__init(struct thread *self, pid_t pid);
 int thread__set_comm(struct thread *self, const char *comm);
-struct thread *
-threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match);
-struct thread *
-register_idle_thread(struct rb_root *threads, struct thread **last_match);
+int thread__comm_len(struct thread *self);
+struct thread *threads__findnew(pid_t pid);
+struct thread *register_idle_thread(void);
 void thread__insert_map(struct thread *self, struct map *map);
 int thread__fork(struct thread *self, struct thread *parent);
-struct map *thread__find_map(struct thread *self, u64 ip);
-size_t threads__fprintf(FILE *fp, struct rb_root *threads);
+size_t thread__fprintf_maps(struct thread *self, FILE *fp);
+size_t threads__fprintf(FILE *fp);
+
+void maps__insert(struct rb_root *maps, struct map *map);
+struct map *maps__find(struct rb_root *maps, u64 addr);
+
+static inline struct map *thread__find_map(struct thread *self,
+					   enum map_type type, u64 addr)
+{
+	return self ? maps__find(&self->maps[type], addr) : NULL;
+}
+
+static inline void __thread__insert_map(struct thread *self, struct map *map)
+{
+	 maps__insert(&self->maps[map->type], map);
+}
+
+void thread__find_addr_location(struct thread *self, u8 cpumode,
+				enum map_type type, u64 addr,
+				struct addr_location *al,
+				symbol_filter_t filter);
+struct symbol *thread__find_symbol(struct thread *self,
+				   enum map_type type, u64 addr,
+				   symbol_filter_t filter);
+
+static inline struct symbol *
+thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter)
+{
+	return thread__find_symbol(self, MAP__FUNCTION, addr, filter);
+}
+#endif	/* __PERF_THREAD_H */
diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c
index af4b0573b37f..cace35595530 100644
--- a/tools/perf/util/trace-event-info.c
+++ b/tools/perf/util/trace-event-info.c
@@ -33,11 +33,11 @@
 #include <ctype.h>
 #include <errno.h>
 #include <stdbool.h>
+#include <linux/kernel.h>
 
 #include "../perf.h"
 #include "trace-event.h"
 
-
 #define VERSION "0.5"
 
 #define _STR(x) #x
@@ -483,27 +483,33 @@ static struct tracepoint_path *
 get_tracepoints_path(struct perf_event_attr *pattrs, int nb_events)
 {
 	struct tracepoint_path path, *ppath = &path;
-	int i;
+	int i, nr_tracepoints = 0;
 
 	for (i = 0; i < nb_events; i++) {
 		if (pattrs[i].type != PERF_TYPE_TRACEPOINT)
 			continue;
+		++nr_tracepoints;
 		ppath->next = tracepoint_id_to_path(pattrs[i].config);
 		if (!ppath->next)
 			die("%s\n", "No memory to alloc tracepoints list");
 		ppath = ppath->next;
 	}
 
-	return path.next;
+	return nr_tracepoints > 0 ? path.next : NULL;
 }
-void read_tracing_data(struct perf_event_attr *pattrs, int nb_events)
+
+int read_tracing_data(int fd, struct perf_event_attr *pattrs, int nb_events)
 {
 	char buf[BUFSIZ];
-	struct tracepoint_path *tps;
+	struct tracepoint_path *tps = get_tracepoints_path(pattrs, nb_events);
+
+	/*
+	 * What? No tracepoints? No sense writing anything here, bail out.
+	 */
+	if (tps == NULL)
+		return -1;
 
-	output_fd = open(output_file, O_WRONLY | O_CREAT | O_TRUNC | O_LARGEFILE, 0644);
-	if (output_fd < 0)
-		die("creating file '%s'", output_file);
+	output_fd = fd;
 
 	buf[0] = 23;
 	buf[1] = 8;
@@ -530,11 +536,11 @@ void read_tracing_data(struct perf_event_attr *pattrs, int nb_events)
 	page_size = getpagesize();
 	write_or_die(&page_size, 4);
 
-	tps = get_tracepoints_path(pattrs, nb_events);
-
 	read_header_files();
 	read_ftrace_files(tps);
 	read_event_files(tps);
 	read_proc_kallsyms();
 	read_ftrace_printk();
+
+	return 0;
 }
diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c
index 55c9659a56e2..0302405aa2ca 100644
--- a/tools/perf/util/trace-event-parse.c
+++ b/tools/perf/util/trace-event-parse.c
@@ -40,12 +40,19 @@ int header_page_size_size;
 int header_page_data_offset;
 int header_page_data_size;
 
+int latency_format;
+
 static char *input_buf;
 static unsigned long long input_buf_ptr;
 static unsigned long long input_buf_siz;
 
 static int cpus;
 static int long_size;
+static int is_flag_field;
+static int is_symbolic_field;
+
+static struct format_field *
+find_any_field(struct event *event, const char *name);
 
 static void init_input_buf(char *buf, unsigned long long size)
 {
@@ -284,18 +291,19 @@ void parse_ftrace_printk(char *file, unsigned int size __unused)
 	char *line;
 	char *next = NULL;
 	char *addr_str;
-	int ret;
 	int i;
 
 	line = strtok_r(file, "\n", &next);
 	while (line) {
+		addr_str = strsep(&line, ":");
+		if (!line) {
+			warning("error parsing print strings");
+			break;
+		}
 		item = malloc_or_die(sizeof(*item));
-		ret = sscanf(line, "%as : %as",
-			     (float *)(void *)&addr_str, /* workaround gcc warning */
-			     (float *)(void *)&item->printk);
 		item->addr = strtoull(addr_str, NULL, 16);
-		free(addr_str);
-
+		/* fmt still has a space, skip it */
+		item->printk = strdup(line+1);
 		item->next = list;
 		list = item;
 		line = strtok_r(NULL, "\n", &next);
@@ -522,7 +530,10 @@ static enum event_type __read_token(char **tok)
 			last_ch = ch;
 			ch = __read_char();
 			buf[i++] = ch;
-		} while (ch != quote_ch && last_ch != '\\');
+			/* the '\' '\' will cancel itself */
+			if (ch == '\\' && last_ch == '\\')
+				last_ch = 0;
+		} while (ch != quote_ch || last_ch == '\\');
 		/* remove the last quote */
 		i--;
 		goto out;
@@ -610,7 +621,7 @@ static enum event_type read_token_item(char **tok)
 static int test_type(enum event_type type, enum event_type expect)
 {
 	if (type != expect) {
-		die("Error: expected type %d but read %d",
+		warning("Error: expected type %d but read %d",
 		    expect, type);
 		return -1;
 	}
@@ -621,13 +632,13 @@ static int test_type_token(enum event_type type, char *token,
 		    enum event_type expect, const char *expect_tok)
 {
 	if (type != expect) {
-		die("Error: expected type %d but read %d",
+		warning("Error: expected type %d but read %d",
 		    expect, type);
 		return -1;
 	}
 
 	if (strcmp(token, expect_tok) != 0) {
-		die("Error: expected '%s' but read '%s'",
+		warning("Error: expected '%s' but read '%s'",
 		    expect_tok, token);
 		return -1;
 	}
@@ -665,7 +676,7 @@ static int __read_expected(enum event_type expect, const char *str, int newline_
 
 	free_token(token);
 
-	return 0;
+	return ret;
 }
 
 static int read_expected(enum event_type expect, const char *str)
@@ -682,10 +693,10 @@ static char *event_read_name(void)
 {
 	char *token;
 
-	if (read_expected(EVENT_ITEM, (char *)"name") < 0)
+	if (read_expected(EVENT_ITEM, "name") < 0)
 		return NULL;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return NULL;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -703,10 +714,10 @@ static int event_read_id(void)
 	char *token;
 	int id;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"ID") < 0)
+	if (read_expected_item(EVENT_ITEM, "ID") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -721,6 +732,24 @@ static int event_read_id(void)
 	return -1;
 }
 
+static int field_is_string(struct format_field *field)
+{
+	if ((field->flags & FIELD_IS_ARRAY) &&
+	    (!strstr(field->type, "char") || !strstr(field->type, "u8") ||
+	     !strstr(field->type, "s8")))
+		return 1;
+
+	return 0;
+}
+
+static int field_is_dynamic(struct format_field *field)
+{
+	if (!strcmp(field->type, "__data_loc"))
+		return 1;
+
+	return 0;
+}
+
 static int event_read_fields(struct event *event, struct format_field **fields)
 {
 	struct format_field *field = NULL;
@@ -738,7 +767,7 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 
 		count++;
 
-		if (test_type_token(type, token, EVENT_ITEM, (char *)"field"))
+		if (test_type_token(type, token, EVENT_ITEM, "field"))
 			goto fail;
 		free_token(token);
 
@@ -753,7 +782,7 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 			type = read_token(&token);
 		}
 
-		if (test_type_token(type, token, EVENT_OP, (char *)":") < 0)
+		if (test_type_token(type, token, EVENT_OP, ":") < 0)
 			return -1;
 
 		if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -865,14 +894,20 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 			free(brackets);
 		}
 
-		if (test_type_token(type, token,  EVENT_OP, (char *)";"))
+		if (field_is_string(field)) {
+			field->flags |= FIELD_IS_STRING;
+			if (field_is_dynamic(field))
+				field->flags |= FIELD_IS_DYNAMIC;
+		}
+
+		if (test_type_token(type, token,  EVENT_OP, ";"))
 			goto fail;
 		free_token(token);
 
-		if (read_expected(EVENT_ITEM, (char *)"offset") < 0)
+		if (read_expected(EVENT_ITEM, "offset") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_OP, (char *)":") < 0)
+		if (read_expected(EVENT_OP, ":") < 0)
 			goto fail_expect;
 
 		if (read_expect_type(EVENT_ITEM, &token))
@@ -880,13 +915,13 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 		field->offset = strtoul(token, NULL, 0);
 		free_token(token);
 
-		if (read_expected(EVENT_OP, (char *)";") < 0)
+		if (read_expected(EVENT_OP, ";") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_ITEM, (char *)"size") < 0)
+		if (read_expected(EVENT_ITEM, "size") < 0)
 			goto fail_expect;
 
-		if (read_expected(EVENT_OP, (char *)":") < 0)
+		if (read_expected(EVENT_OP, ":") < 0)
 			goto fail_expect;
 
 		if (read_expect_type(EVENT_ITEM, &token))
@@ -894,11 +929,34 @@ static int event_read_fields(struct event *event, struct format_field **fields)
 		field->size = strtoul(token, NULL, 0);
 		free_token(token);
 
-		if (read_expected(EVENT_OP, (char *)";") < 0)
+		if (read_expected(EVENT_OP, ";") < 0)
 			goto fail_expect;
 
-		if (read_expect_type(EVENT_NEWLINE, &token) < 0)
-			goto fail;
+		type = read_token(&token);
+		if (type != EVENT_NEWLINE) {
+			/* newer versions of the kernel have a "signed" type */
+			if (test_type_token(type, token, EVENT_ITEM, "signed"))
+				goto fail;
+
+			free_token(token);
+
+			if (read_expected(EVENT_OP, ":") < 0)
+				goto fail_expect;
+
+			if (read_expect_type(EVENT_ITEM, &token))
+				goto fail;
+
+			if (strtoul(token, NULL, 0))
+				field->flags |= FIELD_IS_SIGNED;
+
+			free_token(token);
+			if (read_expected(EVENT_OP, ";") < 0)
+				goto fail_expect;
+
+			if (read_expect_type(EVENT_NEWLINE, &token))
+				goto fail;
+		}
+
 		free_token(token);
 
 		*fields = field;
@@ -921,10 +979,10 @@ static int event_read_format(struct event *event)
 	char *token;
 	int ret;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"format") < 0)
+	if (read_expected_item(EVENT_ITEM, "format") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_NEWLINE, &token))
@@ -984,7 +1042,7 @@ process_cond(struct event *event, struct print_arg *top, char **tok)
 
 	*tok = NULL;
 	type = process_arg(event, left, &token);
-	if (test_type_token(type, token, EVENT_OP, (char *)":"))
+	if (test_type_token(type, token, EVENT_OP, ":"))
 		goto out_free;
 
 	arg->op.op = token;
@@ -1004,6 +1062,35 @@ out_free:
 	return EVENT_ERROR;
 }
 
+static enum event_type
+process_array(struct event *event, struct print_arg *top, char **tok)
+{
+	struct print_arg *arg;
+	enum event_type type;
+	char *token = NULL;
+
+	arg = malloc_or_die(sizeof(*arg));
+	memset(arg, 0, sizeof(*arg));
+
+	*tok = NULL;
+	type = process_arg(event, arg, &token);
+	if (test_type_token(type, token, EVENT_OP, "]"))
+		goto out_free;
+
+	top->op.right = arg;
+
+	free_token(token);
+	type = read_token_item(&token);
+	*tok = token;
+
+	return type;
+
+out_free:
+	free_token(*tok);
+	free_arg(arg);
+	return EVENT_ERROR;
+}
+
 static int get_op_prio(char *op)
 {
 	if (!op[1]) {
@@ -1128,6 +1215,8 @@ process_op(struct event *event, struct print_arg *arg, char **tok)
 		   strcmp(token, "*") == 0 ||
 		   strcmp(token, "^") == 0 ||
 		   strcmp(token, "/") == 0 ||
+		   strcmp(token, "<") == 0 ||
+		   strcmp(token, ">") == 0 ||
 		   strcmp(token, "==") == 0 ||
 		   strcmp(token, "!=") == 0) {
 
@@ -1144,17 +1233,46 @@ process_op(struct event *event, struct print_arg *arg, char **tok)
 
 		right = malloc_or_die(sizeof(*right));
 
-		type = process_arg(event, right, tok);
+		type = read_token_item(&token);
+		*tok = token;
+
+		/* could just be a type pointer */
+		if ((strcmp(arg->op.op, "*") == 0) &&
+		    type == EVENT_DELIM && (strcmp(token, ")") == 0)) {
+			if (left->type != PRINT_ATOM)
+				die("bad pointer type");
+			left->atom.atom = realloc(left->atom.atom,
+					    sizeof(left->atom.atom) + 3);
+			strcat(left->atom.atom, " *");
+			*arg = *left;
+			free(arg);
+
+			return type;
+		}
+
+		type = process_arg_token(event, right, tok, type);
 
 		arg->op.right = right;
 
+	} else if (strcmp(token, "[") == 0) {
+
+		left = malloc_or_die(sizeof(*left));
+		*left = *arg;
+
+		arg->type = PRINT_OP;
+		arg->op.op = token;
+		arg->op.left = left;
+
+		arg->op.prio = 0;
+		type = process_array(event, arg, tok);
+
 	} else {
-		die("unknown op '%s'", token);
+		warning("unknown op '%s'", token);
+		event->flags |= EVENT_FL_FAILED;
 		/* the arg is now the left side */
 		return EVENT_NONE;
 	}
 
-
 	if (type == EVENT_OP) {
 		int prio;
 
@@ -1178,7 +1296,7 @@ process_entry(struct event *event __unused, struct print_arg *arg,
 	char *field;
 	char *token;
 
-	if (read_expected(EVENT_OP, (char *)"->") < 0)
+	if (read_expected(EVENT_OP, "->") < 0)
 		return EVENT_ERROR;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -1188,6 +1306,16 @@ process_entry(struct event *event __unused, struct print_arg *arg,
 	arg->type = PRINT_FIELD;
 	arg->field.name = field;
 
+	if (is_flag_field) {
+		arg->field.field = find_any_field(event, arg->field.name);
+		arg->field.field->flags |= FIELD_IS_FLAG;
+		is_flag_field = 0;
+	} else if (is_symbolic_field) {
+		arg->field.field = find_any_field(event, arg->field.name);
+		arg->field.field->flags |= FIELD_IS_SYMBOLIC;
+		is_symbolic_field = 0;
+	}
+
 	type = read_token(&token);
 	*tok = token;
 
@@ -1338,14 +1466,14 @@ process_fields(struct event *event, struct print_flag_sym **list, char **tok)
 	do {
 		free_token(token);
 		type = read_token_item(&token);
-		if (test_type_token(type, token, EVENT_OP, (char *)"{"))
+		if (test_type_token(type, token, EVENT_OP, "{"))
 			break;
 
 		arg = malloc_or_die(sizeof(*arg));
 
 		free_token(token);
 		type = process_arg(event, arg, &token);
-		if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+		if (test_type_token(type, token, EVENT_DELIM, ","))
 			goto out_free;
 
 		field = malloc_or_die(sizeof(*field));
@@ -1356,7 +1484,7 @@ process_fields(struct event *event, struct print_flag_sym **list, char **tok)
 
 		free_token(token);
 		type = process_arg(event, arg, &token);
-		if (test_type_token(type, token, EVENT_OP, (char *)"}"))
+		if (test_type_token(type, token, EVENT_OP, "}"))
 			goto out_free;
 
 		value = arg_eval(arg);
@@ -1391,13 +1519,13 @@ process_flags(struct event *event, struct print_arg *arg, char **tok)
 	memset(arg, 0, sizeof(*arg));
 	arg->type = PRINT_FLAGS;
 
-	if (read_expected_item(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected_item(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	field = malloc_or_die(sizeof(*field));
 
 	type = process_arg(event, field, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	arg->flags.field = field;
@@ -1408,11 +1536,11 @@ process_flags(struct event *event, struct print_arg *arg, char **tok)
 		type = read_token_item(&token);
 	}
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	type = process_fields(event, &arg->flags.flags, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")"))
+	if (test_type_token(type, token, EVENT_DELIM, ")"))
 		goto out_free;
 
 	free_token(token);
@@ -1434,19 +1562,19 @@ process_symbols(struct event *event, struct print_arg *arg, char **tok)
 	memset(arg, 0, sizeof(*arg));
 	arg->type = PRINT_SYMBOL;
 
-	if (read_expected_item(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected_item(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	field = malloc_or_die(sizeof(*field));
 
 	type = process_arg(event, field, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto out_free;
 
 	arg->symbol.field = field;
 
 	type = process_fields(event, &arg->symbol.symbols, &token);
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")"))
+	if (test_type_token(type, token, EVENT_DELIM, ")"))
 		goto out_free;
 
 	free_token(token);
@@ -1463,7 +1591,6 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 {
 	struct print_arg *item_arg;
 	enum event_type type;
-	int ptr_cast = 0;
 	char *token;
 
 	type = process_arg(event, arg, &token);
@@ -1471,28 +1598,13 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 	if (type == EVENT_ERROR)
 		return EVENT_ERROR;
 
-	if (type == EVENT_OP) {
-		/* handle the ptr casts */
-		if (!strcmp(token, "*")) {
-			/*
-			 * FIXME: should we zapp whitespaces before ')' ?
-			 * (may require a peek_token_item())
-			 */
-			if (__peek_char() == ')') {
-				ptr_cast = 1;
-				free_token(token);
-				type = read_token_item(&token);
-			}
-		}
-		if (!ptr_cast) {
-			type = process_op(event, arg, &token);
+	if (type == EVENT_OP)
+		type = process_op(event, arg, &token);
 
-			if (type == EVENT_ERROR)
-				return EVENT_ERROR;
-		}
-	}
+	if (type == EVENT_ERROR)
+		return EVENT_ERROR;
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)")")) {
+	if (test_type_token(type, token, EVENT_DELIM, ")")) {
 		free_token(token);
 		return EVENT_ERROR;
 	}
@@ -1516,13 +1628,6 @@ process_paren(struct event *event, struct print_arg *arg, char **tok)
 		item_arg = malloc_or_die(sizeof(*item_arg));
 
 		arg->type = PRINT_TYPE;
-		if (ptr_cast) {
-			char *old = arg->atom.atom;
-
-			arg->atom.atom = malloc_or_die(strlen(old + 3));
-			sprintf(arg->atom.atom, "%s *", old);
-			free(old);
-		}
 		arg->typecast.type = arg->atom.atom;
 		arg->typecast.item = item_arg;
 		type = process_arg_token(event, item_arg, &token, type);
@@ -1540,7 +1645,7 @@ process_str(struct event *event __unused, struct print_arg *arg, char **tok)
 	enum event_type type;
 	char *token;
 
-	if (read_expected(EVENT_DELIM, (char *)"(") < 0)
+	if (read_expected(EVENT_DELIM, "(") < 0)
 		return EVENT_ERROR;
 
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -1550,7 +1655,7 @@ process_str(struct event *event __unused, struct print_arg *arg, char **tok)
 	arg->string.string = token;
 	arg->string.offset = -1;
 
-	if (read_expected(EVENT_DELIM, (char *)")") < 0)
+	if (read_expected(EVENT_DELIM, ")") < 0)
 		return EVENT_ERROR;
 
 	type = read_token(&token);
@@ -1578,9 +1683,11 @@ process_arg_token(struct event *event, struct print_arg *arg,
 			type = process_entry(event, arg, &token);
 		} else if (strcmp(token, "__print_flags") == 0) {
 			free_token(token);
+			is_flag_field = 1;
 			type = process_flags(event, arg, &token);
 		} else if (strcmp(token, "__print_symbolic") == 0) {
 			free_token(token);
+			is_symbolic_field = 1;
 			type = process_symbols(event, arg, &token);
 		} else if (strcmp(token, "__get_str") == 0) {
 			free_token(token);
@@ -1637,12 +1744,18 @@ process_arg_token(struct event *event, struct print_arg *arg,
 
 static int event_read_print_args(struct event *event, struct print_arg **list)
 {
-	enum event_type type;
+	enum event_type type = EVENT_ERROR;
 	struct print_arg *arg;
 	char *token;
 	int args = 0;
 
 	do {
+		if (type == EVENT_NEWLINE) {
+			free_token(token);
+			type = read_token_item(&token);
+			continue;
+		}
+
 		arg = malloc_or_die(sizeof(*arg));
 		memset(arg, 0, sizeof(*arg));
 
@@ -1683,18 +1796,19 @@ static int event_read_print(struct event *event)
 	char *token;
 	int ret;
 
-	if (read_expected_item(EVENT_ITEM, (char *)"print") < 0)
+	if (read_expected_item(EVENT_ITEM, "print") < 0)
 		return -1;
 
-	if (read_expected(EVENT_ITEM, (char *)"fmt") < 0)
+	if (read_expected(EVENT_ITEM, "fmt") < 0)
 		return -1;
 
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return -1;
 
 	if (read_expect_type(EVENT_DQUOTE, &token) < 0)
 		goto fail;
 
+ concat:
 	event->print_fmt.format = token;
 	event->print_fmt.args = NULL;
 
@@ -1704,7 +1818,22 @@ static int event_read_print(struct event *event)
 	if (type == EVENT_NONE)
 		return 0;
 
-	if (test_type_token(type, token, EVENT_DELIM, (char *)","))
+	/* Handle concatination of print lines */
+	if (type == EVENT_DQUOTE) {
+		char *cat;
+
+		cat = malloc_or_die(strlen(event->print_fmt.format) +
+				    strlen(token) + 1);
+		strcpy(cat, event->print_fmt.format);
+		strcat(cat, token);
+		free_token(token);
+		free_token(event->print_fmt.format);
+		event->print_fmt.format = NULL;
+		token = cat;
+		goto concat;
+	}
+
+	if (test_type_token(type, token, EVENT_DELIM, ","))
 		goto fail;
 
 	free_token(token);
@@ -1713,7 +1842,7 @@ static int event_read_print(struct event *event)
 	if (ret < 0)
 		return -1;
 
-	return 0;
+	return ret;
 
  fail:
 	free_token(token);
@@ -1759,7 +1888,7 @@ find_any_field(struct event *event, const char *name)
 	return find_field(event, name);
 }
 
-static unsigned long long read_size(void *ptr, int size)
+unsigned long long read_size(void *ptr, int size)
 {
 	switch (size) {
 	case 1:
@@ -1822,37 +1951,67 @@ static int get_common_info(const char *type, int *offset, int *size)
 	return 0;
 }
 
-int trace_parse_common_type(void *data)
+static int __parse_common(void *data, int *size, int *offset,
+			  const char *name)
 {
-	static int type_offset;
-	static int type_size;
 	int ret;
 
-	if (!type_size) {
-		ret = get_common_info("common_type",
-				      &type_offset,
-				      &type_size);
+	if (!*size) {
+		ret = get_common_info(name, offset, size);
 		if (ret < 0)
 			return ret;
 	}
-	return read_size(data + type_offset, type_size);
+	return read_size(data + *offset, *size);
+}
+
+int trace_parse_common_type(void *data)
+{
+	static int type_offset;
+	static int type_size;
+
+	return __parse_common(data, &type_size, &type_offset,
+			      "common_type");
 }
 
-static int parse_common_pid(void *data)
+int trace_parse_common_pid(void *data)
 {
 	static int pid_offset;
 	static int pid_size;
+
+	return __parse_common(data, &pid_size, &pid_offset,
+			      "common_pid");
+}
+
+int parse_common_pc(void *data)
+{
+	static int pc_offset;
+	static int pc_size;
+
+	return __parse_common(data, &pc_size, &pc_offset,
+			      "common_preempt_count");
+}
+
+int parse_common_flags(void *data)
+{
+	static int flags_offset;
+	static int flags_size;
+
+	return __parse_common(data, &flags_size, &flags_offset,
+			      "common_flags");
+}
+
+int parse_common_lock_depth(void *data)
+{
+	static int ld_offset;
+	static int ld_size;
 	int ret;
 
-	if (!pid_size) {
-		ret = get_common_info("common_pid",
-				      &pid_offset,
-				      &pid_size);
-		if (ret < 0)
-			return ret;
-	}
+	ret = __parse_common(data, &ld_size, &ld_offset,
+			     "common_lock_depth");
+	if (ret < 0)
+		return -1;
 
-	return read_size(data + pid_offset, pid_size);
+	return ret;
 }
 
 struct event *trace_find_event(int id)
@@ -1866,11 +2025,20 @@ struct event *trace_find_event(int id)
 	return event;
 }
 
+struct event *trace_find_next_event(struct event *event)
+{
+	if (!event)
+		return event_list;
+
+	return event->next;
+}
+
 static unsigned long long eval_num_arg(void *data, int size,
 				   struct event *event, struct print_arg *arg)
 {
 	unsigned long long val = 0;
 	unsigned long long left, right;
+	struct print_arg *larg;
 
 	switch (arg->type) {
 	case PRINT_NULL:
@@ -1897,6 +2065,26 @@ static unsigned long long eval_num_arg(void *data, int size,
 		return 0;
 		break;
 	case PRINT_OP:
+		if (strcmp(arg->op.op, "[") == 0) {
+			/*
+			 * Arrays are special, since we don't want
+			 * to read the arg as is.
+			 */
+			if (arg->op.left->type != PRINT_FIELD)
+				goto default_op; /* oops, all bets off */
+			larg = arg->op.left;
+			if (!larg->field.field) {
+				larg->field.field =
+					find_any_field(event, larg->field.name);
+				if (!larg->field.field)
+					die("field %s not found", larg->field.name);
+			}
+			right = eval_num_arg(data, size, event, arg->op.right);
+			val = read_size(data + larg->field.field->offset +
+					right * long_size, long_size);
+			break;
+		}
+ default_op:
 		left = eval_num_arg(data, size, event, arg->op.left);
 		right = eval_num_arg(data, size, event, arg->op.right);
 		switch (arg->op.op[0]) {
@@ -1947,6 +2135,12 @@ static unsigned long long eval_num_arg(void *data, int size,
 				die("unknown op '%s'", arg->op.op);
 			val = left == right;
 			break;
+		case '-':
+			val = left - right;
+			break;
+		case '+':
+			val = left + right;
+			break;
 		default:
 			die("unknown op '%s'", arg->op.op);
 		}
@@ -1978,7 +2172,7 @@ static const struct flag flags[] = {
 	{ "HRTIMER_RESTART", 1 },
 };
 
-static unsigned long long eval_flag(const char *flag)
+unsigned long long eval_flag(const char *flag)
 {
 	int i;
 
@@ -2145,8 +2339,9 @@ static struct print_arg *make_bprint_args(char *fmt, void *data, int size, struc
 			case 'u':
 			case 'x':
 			case 'i':
-				bptr = (void *)(((unsigned long)bptr + (long_size - 1)) &
-						~(long_size - 1));
+				/* the pointers are always 4 bytes aligned */
+				bptr = (void *)(((unsigned long)bptr + 3) &
+						~3);
 				switch (ls) {
 				case 0:
 				case 1:
@@ -2270,7 +2465,27 @@ static void pretty_print(void *data, int size, struct event *event)
 
 	for (; *ptr; ptr++) {
 		ls = 0;
-		if (*ptr == '%') {
+		if (*ptr == '\\') {
+			ptr++;
+			switch (*ptr) {
+			case 'n':
+				printf("\n");
+				break;
+			case 't':
+				printf("\t");
+				break;
+			case 'r':
+				printf("\r");
+				break;
+			case '\\':
+				printf("\\");
+				break;
+			default:
+				printf("%c", *ptr);
+				break;
+			}
+
+		} else if (*ptr == '%') {
 			saveptr = ptr;
 			show_func = 0;
  cont_process:
@@ -2377,6 +2592,41 @@ static inline int log10_cpu(int nb)
 	return 1;
 }
 
+static void print_lat_fmt(void *data, int size __unused)
+{
+	unsigned int lat_flags;
+	unsigned int pc;
+	int lock_depth;
+	int hardirq;
+	int softirq;
+
+	lat_flags = parse_common_flags(data);
+	pc = parse_common_pc(data);
+	lock_depth = parse_common_lock_depth(data);
+
+	hardirq = lat_flags & TRACE_FLAG_HARDIRQ;
+	softirq = lat_flags & TRACE_FLAG_SOFTIRQ;
+
+	printf("%c%c%c",
+	       (lat_flags & TRACE_FLAG_IRQS_OFF) ? 'd' :
+	       (lat_flags & TRACE_FLAG_IRQS_NOSUPPORT) ?
+	       'X' : '.',
+	       (lat_flags & TRACE_FLAG_NEED_RESCHED) ?
+	       'N' : '.',
+	       (hardirq && softirq) ? 'H' :
+	       hardirq ? 'h' : softirq ? 's' : '.');
+
+	if (pc)
+		printf("%x", pc);
+	else
+		printf(".");
+
+	if (lock_depth < 0)
+		printf(".");
+	else
+		printf("%d", lock_depth);
+}
+
 /* taken from Linux, written by Frederic Weisbecker */
 static void print_graph_cpu(int cpu)
 {
@@ -2452,7 +2702,7 @@ get_return_for_leaf(int cpu, int cur_pid, unsigned long long cur_func,
 	if (!(event->flags & EVENT_FL_ISFUNCRET))
 		return NULL;
 
-	pid = parse_common_pid(next->data);
+	pid = trace_parse_common_pid(next->data);
 	field = find_field(event, "func");
 	if (!field)
 		die("function return does not have field func");
@@ -2620,6 +2870,11 @@ pretty_print_func_ent(void *data, int size, struct event *event,
 
 	printf(" | ");
 
+	if (latency_format) {
+		print_lat_fmt(data, size);
+		printf(" | ");
+	}
+
 	field = find_field(event, "func");
 	if (!field)
 		die("function entry does not have func field");
@@ -2663,6 +2918,11 @@ pretty_print_func_ret(void *data, int size __unused, struct event *event,
 
 	printf(" | ");
 
+	if (latency_format) {
+		print_lat_fmt(data, size);
+		printf(" | ");
+	}
+
 	field = find_field(event, "rettime");
 	if (!field)
 		die("can't find rettime in return graph");
@@ -2724,19 +2984,30 @@ void print_event(int cpu, void *data, int size, unsigned long long nsecs,
 
 	event = trace_find_event(type);
 	if (!event) {
-		printf("ug! no event found for type %d\n", type);
+		warning("ug! no event found for type %d", type);
 		return;
 	}
 
-	pid = parse_common_pid(data);
+	pid = trace_parse_common_pid(data);
 
 	if (event->flags & (EVENT_FL_ISFUNCENT | EVENT_FL_ISFUNCRET))
 		return pretty_print_func_graph(data, size, event, cpu,
 					       pid, comm, secs, usecs);
 
-	printf("%16s-%-5d [%03d] %5lu.%09Lu: %s: ",
-	       comm, pid,  cpu,
-	       secs, nsecs, event->name);
+	if (latency_format) {
+		printf("%8.8s-%-5d %3d",
+		       comm, pid, cpu);
+		print_lat_fmt(data, size);
+	} else
+		printf("%16s-%-5d [%03d]", comm, pid,  cpu);
+
+	printf(" %5lu.%06lu: %s: ", secs, usecs, event->name);
+
+	if (event->flags & EVENT_FL_FAILED) {
+		printf("EVENT '%s' FAILED TO PARSE\n",
+		       event->name);
+		return;
+	}
 
 	pretty_print(data, size, event);
 	printf("\n");
@@ -2807,46 +3078,71 @@ static void print_args(struct print_arg *args)
 	}
 }
 
-static void parse_header_field(char *type,
+static void parse_header_field(const char *field,
 			       int *offset, int *size)
 {
 	char *token;
+	int type;
 
-	if (read_expected(EVENT_ITEM, (char *)"field") < 0)
+	if (read_expected(EVENT_ITEM, "field") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
+
 	/* type */
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	free_token(token);
 
-	if (read_expected(EVENT_ITEM, type) < 0)
+	if (read_expected(EVENT_ITEM, field) < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)";") < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
-	if (read_expected(EVENT_ITEM, (char *)"offset") < 0)
+	if (read_expected(EVENT_ITEM, "offset") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	*offset = atoi(token);
 	free_token(token);
-	if (read_expected(EVENT_OP, (char *)";") < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
-	if (read_expected(EVENT_ITEM, (char *)"size") < 0)
+	if (read_expected(EVENT_ITEM, "size") < 0)
 		return;
-	if (read_expected(EVENT_OP, (char *)":") < 0)
+	if (read_expected(EVENT_OP, ":") < 0)
 		return;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
-		return;
+		goto fail;
 	*size = atoi(token);
 	free_token(token);
-	if (read_expected(EVENT_OP, (char *)";") < 0)
-		return;
-	if (read_expect_type(EVENT_NEWLINE, &token) < 0)
+	if (read_expected(EVENT_OP, ";") < 0)
 		return;
+	type = read_token(&token);
+	if (type != EVENT_NEWLINE) {
+		/* newer versions of the kernel have a "signed" type */
+		if (type != EVENT_ITEM)
+			goto fail;
+
+		if (strcmp(token, "signed") != 0)
+			goto fail;
+
+		free_token(token);
+
+		if (read_expected(EVENT_OP, ":") < 0)
+			return;
+
+		if (read_expect_type(EVENT_ITEM, &token))
+			goto fail;
+
+		free_token(token);
+		if (read_expected(EVENT_OP, ";") < 0)
+			return;
+
+		if (read_expect_type(EVENT_NEWLINE, &token))
+			goto fail;
+	}
+ fail:
 	free_token(token);
 }
 
@@ -2854,11 +3150,11 @@ int parse_header_page(char *buf, unsigned long size)
 {
 	init_input_buf(buf, size);
 
-	parse_header_field((char *)"timestamp", &header_page_ts_offset,
+	parse_header_field("timestamp", &header_page_ts_offset,
 			   &header_page_ts_size);
-	parse_header_field((char *)"commit", &header_page_size_offset,
+	parse_header_field("commit", &header_page_size_offset,
 			   &header_page_size_size);
-	parse_header_field((char *)"data", &header_page_data_offset,
+	parse_header_field("data", &header_page_data_offset,
 			   &header_page_data_size);
 
 	return 0;
@@ -2909,6 +3205,9 @@ int parse_ftrace_file(char *buf, unsigned long size)
 	if (ret < 0)
 		die("failed to read ftrace event print fmt");
 
+	/* New ftrace handles args */
+	if (ret > 0)
+		return 0;
 	/*
 	 * The arguments for ftrace files are parsed by the fields.
 	 * Set up the fields as their arguments.
@@ -2926,7 +3225,7 @@ int parse_ftrace_file(char *buf, unsigned long size)
 	return 0;
 }
 
-int parse_event_file(char *buf, unsigned long size, char *system__unused __unused)
+int parse_event_file(char *buf, unsigned long size, char *sys)
 {
 	struct event *event;
 	int ret;
@@ -2946,12 +3245,18 @@ int parse_event_file(char *buf, unsigned long size, char *system__unused __unuse
 		die("failed to read event id");
 
 	ret = event_read_format(event);
-	if (ret < 0)
-		die("failed to read event format");
+	if (ret < 0) {
+		warning("failed to read event format for %s", event->name);
+		goto event_failed;
+	}
 
 	ret = event_read_print(event);
-	if (ret < 0)
-		die("failed to read event print fmt");
+	if (ret < 0) {
+		warning("failed to read event print fmt for %s", event->name);
+		goto event_failed;
+	}
+
+	event->system = strdup(sys);
 
 #define PRINT_ARGS 0
 	if (PRINT_ARGS && event->print_fmt.args)
@@ -2959,6 +3264,12 @@ int parse_event_file(char *buf, unsigned long size, char *system__unused __unuse
 
 	add_event(event);
 	return 0;
+
+ event_failed:
+	event->flags |= EVENT_FL_FAILED;
+	/* still add it even if it failed */
+	add_event(event);
+	return -1;
 }
 
 void parse_set_info(int nr_cpus, int long_sz)
diff --git a/tools/perf/util/trace-event-perl.c b/tools/perf/util/trace-event-perl.c
new file mode 100644
index 000000000000..51e833fd58c3
--- /dev/null
+++ b/tools/perf/util/trace-event-perl.c
@@ -0,0 +1,598 @@
+/*
+ * trace-event-perl.  Feed perf trace events to an embedded Perl interpreter.
+ *
+ * Copyright (C) 2009 Tom Zanussi <tzanussi@gmail.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <ctype.h>
+#include <errno.h>
+
+#include "../perf.h"
+#include "util.h"
+#include "trace-event.h"
+#include "trace-event-perl.h"
+
+void xs_init(pTHX);
+
+void boot_Perf__Trace__Context(pTHX_ CV *cv);
+void boot_DynaLoader(pTHX_ CV *cv);
+
+void xs_init(pTHX)
+{
+	const char *file = __FILE__;
+	dXSUB_SYS;
+
+	newXS("Perf::Trace::Context::bootstrap", boot_Perf__Trace__Context,
+	      file);
+	newXS("DynaLoader::boot_DynaLoader", boot_DynaLoader, file);
+}
+
+INTERP my_perl;
+
+#define FTRACE_MAX_EVENT				\
+	((1 << (sizeof(unsigned short) * 8)) - 1)
+
+struct event *events[FTRACE_MAX_EVENT];
+
+static struct scripting_context *scripting_context;
+
+static char *cur_field_name;
+static int zero_flag_atom;
+
+static void define_symbolic_value(const char *ev_name,
+				  const char *field_name,
+				  const char *field_value,
+				  const char *field_str)
+{
+	unsigned long long value;
+	dSP;
+
+	value = eval_flag(field_value);
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+
+	XPUSHs(sv_2mortal(newSVpv(ev_name, 0)));
+	XPUSHs(sv_2mortal(newSVpv(field_name, 0)));
+	XPUSHs(sv_2mortal(newSVuv(value)));
+	XPUSHs(sv_2mortal(newSVpv(field_str, 0)));
+
+	PUTBACK;
+	if (get_cv("main::define_symbolic_value", 0))
+		call_pv("main::define_symbolic_value", G_SCALAR);
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void define_symbolic_values(struct print_flag_sym *field,
+				   const char *ev_name,
+				   const char *field_name)
+{
+	define_symbolic_value(ev_name, field_name, field->value, field->str);
+	if (field->next)
+		define_symbolic_values(field->next, ev_name, field_name);
+}
+
+static void define_symbolic_field(const char *ev_name,
+				  const char *field_name)
+{
+	dSP;
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+
+	XPUSHs(sv_2mortal(newSVpv(ev_name, 0)));
+	XPUSHs(sv_2mortal(newSVpv(field_name, 0)));
+
+	PUTBACK;
+	if (get_cv("main::define_symbolic_field", 0))
+		call_pv("main::define_symbolic_field", G_SCALAR);
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void define_flag_value(const char *ev_name,
+			      const char *field_name,
+			      const char *field_value,
+			      const char *field_str)
+{
+	unsigned long long value;
+	dSP;
+
+	value = eval_flag(field_value);
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+
+	XPUSHs(sv_2mortal(newSVpv(ev_name, 0)));
+	XPUSHs(sv_2mortal(newSVpv(field_name, 0)));
+	XPUSHs(sv_2mortal(newSVuv(value)));
+	XPUSHs(sv_2mortal(newSVpv(field_str, 0)));
+
+	PUTBACK;
+	if (get_cv("main::define_flag_value", 0))
+		call_pv("main::define_flag_value", G_SCALAR);
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void define_flag_values(struct print_flag_sym *field,
+			       const char *ev_name,
+			       const char *field_name)
+{
+	define_flag_value(ev_name, field_name, field->value, field->str);
+	if (field->next)
+		define_flag_values(field->next, ev_name, field_name);
+}
+
+static void define_flag_field(const char *ev_name,
+			      const char *field_name,
+			      const char *delim)
+{
+	dSP;
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+
+	XPUSHs(sv_2mortal(newSVpv(ev_name, 0)));
+	XPUSHs(sv_2mortal(newSVpv(field_name, 0)));
+	XPUSHs(sv_2mortal(newSVpv(delim, 0)));
+
+	PUTBACK;
+	if (get_cv("main::define_flag_field", 0))
+		call_pv("main::define_flag_field", G_SCALAR);
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void define_event_symbols(struct event *event,
+				 const char *ev_name,
+				 struct print_arg *args)
+{
+	switch (args->type) {
+	case PRINT_NULL:
+		break;
+	case PRINT_ATOM:
+		define_flag_value(ev_name, cur_field_name, "0",
+				  args->atom.atom);
+		zero_flag_atom = 0;
+		break;
+	case PRINT_FIELD:
+		if (cur_field_name)
+			free(cur_field_name);
+		cur_field_name = strdup(args->field.name);
+		break;
+	case PRINT_FLAGS:
+		define_event_symbols(event, ev_name, args->flags.field);
+		define_flag_field(ev_name, cur_field_name, args->flags.delim);
+		define_flag_values(args->flags.flags, ev_name, cur_field_name);
+		break;
+	case PRINT_SYMBOL:
+		define_event_symbols(event, ev_name, args->symbol.field);
+		define_symbolic_field(ev_name, cur_field_name);
+		define_symbolic_values(args->symbol.symbols, ev_name,
+				       cur_field_name);
+		break;
+	case PRINT_STRING:
+		break;
+	case PRINT_TYPE:
+		define_event_symbols(event, ev_name, args->typecast.item);
+		break;
+	case PRINT_OP:
+		if (strcmp(args->op.op, ":") == 0)
+			zero_flag_atom = 1;
+		define_event_symbols(event, ev_name, args->op.left);
+		define_event_symbols(event, ev_name, args->op.right);
+		break;
+	default:
+		/* we should warn... */
+		return;
+	}
+
+	if (args->next)
+		define_event_symbols(event, ev_name, args->next);
+}
+
+static inline struct event *find_cache_event(int type)
+{
+	static char ev_name[256];
+	struct event *event;
+
+	if (events[type])
+		return events[type];
+
+	events[type] = event = trace_find_event(type);
+	if (!event)
+		return NULL;
+
+	sprintf(ev_name, "%s::%s", event->system, event->name);
+
+	define_event_symbols(event, ev_name, event->print_fmt.args);
+
+	return event;
+}
+
+int common_pc(struct scripting_context *context)
+{
+	int pc;
+
+	pc = parse_common_pc(context->event_data);
+
+	return pc;
+}
+
+int common_flags(struct scripting_context *context)
+{
+	int flags;
+
+	flags = parse_common_flags(context->event_data);
+
+	return flags;
+}
+
+int common_lock_depth(struct scripting_context *context)
+{
+	int lock_depth;
+
+	lock_depth = parse_common_lock_depth(context->event_data);
+
+	return lock_depth;
+}
+
+static void perl_process_event(int cpu, void *data,
+			       int size __attribute((unused)),
+			       unsigned long long nsecs, char *comm)
+{
+	struct format_field *field;
+	static char handler[256];
+	unsigned long long val;
+	unsigned long s, ns;
+	struct event *event;
+	int type;
+	int pid;
+
+	dSP;
+
+	type = trace_parse_common_type(data);
+
+	event = find_cache_event(type);
+	if (!event)
+		die("ug! no event found for type %d", type);
+
+	pid = trace_parse_common_pid(data);
+
+	sprintf(handler, "%s::%s", event->system, event->name);
+
+	s = nsecs / NSECS_PER_SEC;
+	ns = nsecs - s * NSECS_PER_SEC;
+
+	scripting_context->event_data = data;
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+
+	XPUSHs(sv_2mortal(newSVpv(handler, 0)));
+	XPUSHs(sv_2mortal(newSViv(PTR2IV(scripting_context))));
+	XPUSHs(sv_2mortal(newSVuv(cpu)));
+	XPUSHs(sv_2mortal(newSVuv(s)));
+	XPUSHs(sv_2mortal(newSVuv(ns)));
+	XPUSHs(sv_2mortal(newSViv(pid)));
+	XPUSHs(sv_2mortal(newSVpv(comm, 0)));
+
+	/* common fields other than pid can be accessed via xsub fns */
+
+	for (field = event->format.fields; field; field = field->next) {
+		if (field->flags & FIELD_IS_STRING) {
+			int offset;
+			if (field->flags & FIELD_IS_DYNAMIC) {
+				offset = *(int *)(data + field->offset);
+				offset &= 0xffff;
+			} else
+				offset = field->offset;
+			XPUSHs(sv_2mortal(newSVpv((char *)data + offset, 0)));
+		} else { /* FIELD_IS_NUMERIC */
+			val = read_size(data + field->offset, field->size);
+			if (field->flags & FIELD_IS_SIGNED) {
+				XPUSHs(sv_2mortal(newSViv(val)));
+			} else {
+				XPUSHs(sv_2mortal(newSVuv(val)));
+			}
+		}
+	}
+
+	PUTBACK;
+
+	if (get_cv(handler, 0))
+		call_pv(handler, G_SCALAR);
+	else if (get_cv("main::trace_unhandled", 0)) {
+		XPUSHs(sv_2mortal(newSVpv(handler, 0)));
+		XPUSHs(sv_2mortal(newSViv(PTR2IV(scripting_context))));
+		XPUSHs(sv_2mortal(newSVuv(cpu)));
+		XPUSHs(sv_2mortal(newSVuv(nsecs)));
+		XPUSHs(sv_2mortal(newSViv(pid)));
+		XPUSHs(sv_2mortal(newSVpv(comm, 0)));
+		call_pv("main::trace_unhandled", G_SCALAR);
+	}
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void run_start_sub(void)
+{
+	dSP; /* access to Perl stack */
+	PUSHMARK(SP);
+
+	if (get_cv("main::trace_begin", 0))
+		call_pv("main::trace_begin", G_DISCARD | G_NOARGS);
+}
+
+/*
+ * Start trace script
+ */
+static int perl_start_script(const char *script)
+{
+	const char *command_line[2] = { "", NULL };
+
+	command_line[1] = script;
+
+	my_perl = perl_alloc();
+	perl_construct(my_perl);
+
+	if (perl_parse(my_perl, xs_init, 2, (char **)command_line,
+		       (char **)NULL))
+		return -1;
+
+	perl_run(my_perl);
+	if (SvTRUE(ERRSV))
+		return -1;
+
+	run_start_sub();
+
+	fprintf(stderr, "perf trace started with Perl script %s\n\n", script);
+
+	return 0;
+}
+
+/*
+ * Stop trace script
+ */
+static int perl_stop_script(void)
+{
+	dSP; /* access to Perl stack */
+	PUSHMARK(SP);
+
+	if (get_cv("main::trace_end", 0))
+		call_pv("main::trace_end", G_DISCARD | G_NOARGS);
+
+	perl_destruct(my_perl);
+	perl_free(my_perl);
+
+	fprintf(stderr, "\nperf trace Perl script stopped\n");
+
+	return 0;
+}
+
+static int perl_generate_script(const char *outfile)
+{
+	struct event *event = NULL;
+	struct format_field *f;
+	char fname[PATH_MAX];
+	int not_first, count;
+	FILE *ofp;
+
+	sprintf(fname, "%s.pl", outfile);
+	ofp = fopen(fname, "w");
+	if (ofp == NULL) {
+		fprintf(stderr, "couldn't open %s\n", fname);
+		return -1;
+	}
+
+	fprintf(ofp, "# perf trace event handlers, "
+		"generated by perf trace -g perl\n");
+
+	fprintf(ofp, "# Licensed under the terms of the GNU GPL"
+		" License version 2\n\n");
+
+	fprintf(ofp, "# The common_* event handler fields are the most useful "
+		"fields common to\n");
+
+	fprintf(ofp, "# all events.  They don't necessarily correspond to "
+		"the 'common_*' fields\n");
+
+	fprintf(ofp, "# in the format files.  Those fields not available as "
+		"handler params can\n");
+
+	fprintf(ofp, "# be retrieved using Perl functions of the form "
+		"common_*($context).\n");
+
+	fprintf(ofp, "# See Context.pm for the list of available "
+		"functions.\n\n");
+
+	fprintf(ofp, "use lib \"$ENV{'PERF_EXEC_PATH'}/scripts/perl/"
+		"Perf-Trace-Util/lib\";\n");
+
+	fprintf(ofp, "use lib \"./Perf-Trace-Util/lib\";\n");
+	fprintf(ofp, "use Perf::Trace::Core;\n");
+	fprintf(ofp, "use Perf::Trace::Context;\n");
+	fprintf(ofp, "use Perf::Trace::Util;\n\n");
+
+	fprintf(ofp, "sub trace_begin\n{\n\t# optional\n}\n\n");
+	fprintf(ofp, "sub trace_end\n{\n\t# optional\n}\n\n");
+
+	while ((event = trace_find_next_event(event))) {
+		fprintf(ofp, "sub %s::%s\n{\n", event->system, event->name);
+		fprintf(ofp, "\tmy (");
+
+		fprintf(ofp, "$event_name, ");
+		fprintf(ofp, "$context, ");
+		fprintf(ofp, "$common_cpu, ");
+		fprintf(ofp, "$common_secs, ");
+		fprintf(ofp, "$common_nsecs,\n");
+		fprintf(ofp, "\t    $common_pid, ");
+		fprintf(ofp, "$common_comm,\n\t    ");
+
+		not_first = 0;
+		count = 0;
+
+		for (f = event->format.fields; f; f = f->next) {
+			if (not_first++)
+				fprintf(ofp, ", ");
+			if (++count % 5 == 0)
+				fprintf(ofp, "\n\t    ");
+
+			fprintf(ofp, "$%s", f->name);
+		}
+		fprintf(ofp, ") = @_;\n\n");
+
+		fprintf(ofp, "\tprint_header($event_name, $common_cpu, "
+			"$common_secs, $common_nsecs,\n\t             "
+			"$common_pid, $common_comm);\n\n");
+
+		fprintf(ofp, "\tprintf(\"");
+
+		not_first = 0;
+		count = 0;
+
+		for (f = event->format.fields; f; f = f->next) {
+			if (not_first++)
+				fprintf(ofp, ", ");
+			if (count && count % 4 == 0) {
+				fprintf(ofp, "\".\n\t       \"");
+			}
+			count++;
+
+			fprintf(ofp, "%s=", f->name);
+			if (f->flags & FIELD_IS_STRING ||
+			    f->flags & FIELD_IS_FLAG ||
+			    f->flags & FIELD_IS_SYMBOLIC)
+				fprintf(ofp, "%%s");
+			else if (f->flags & FIELD_IS_SIGNED)
+				fprintf(ofp, "%%d");
+			else
+				fprintf(ofp, "%%u");
+		}
+
+		fprintf(ofp, "\\n\",\n\t       ");
+
+		not_first = 0;
+		count = 0;
+
+		for (f = event->format.fields; f; f = f->next) {
+			if (not_first++)
+				fprintf(ofp, ", ");
+
+			if (++count % 5 == 0)
+				fprintf(ofp, "\n\t       ");
+
+			if (f->flags & FIELD_IS_FLAG) {
+				if ((count - 1) % 5 != 0) {
+					fprintf(ofp, "\n\t       ");
+					count = 4;
+				}
+				fprintf(ofp, "flag_str(\"");
+				fprintf(ofp, "%s::%s\", ", event->system,
+					event->name);
+				fprintf(ofp, "\"%s\", $%s)", f->name,
+					f->name);
+			} else if (f->flags & FIELD_IS_SYMBOLIC) {
+				if ((count - 1) % 5 != 0) {
+					fprintf(ofp, "\n\t       ");
+					count = 4;
+				}
+				fprintf(ofp, "symbol_str(\"");
+				fprintf(ofp, "%s::%s\", ", event->system,
+					event->name);
+				fprintf(ofp, "\"%s\", $%s)", f->name,
+					f->name);
+			} else
+				fprintf(ofp, "$%s", f->name);
+		}
+
+		fprintf(ofp, ");\n");
+		fprintf(ofp, "}\n\n");
+	}
+
+	fprintf(ofp, "sub trace_unhandled\n{\n\tmy ($event_name, $context, "
+		"$common_cpu, $common_secs, $common_nsecs,\n\t    "
+		"$common_pid, $common_comm) = @_;\n\n");
+
+	fprintf(ofp, "\tprint_header($event_name, $common_cpu, "
+		"$common_secs, $common_nsecs,\n\t             $common_pid, "
+		"$common_comm);\n}\n\n");
+
+	fprintf(ofp, "sub print_header\n{\n"
+		"\tmy ($event_name, $cpu, $secs, $nsecs, $pid, $comm) = @_;\n\n"
+		"\tprintf(\"%%-20s %%5u %%05u.%%09u %%8u %%-20s \",\n\t       "
+		"$event_name, $cpu, $secs, $nsecs, $pid, $comm);\n}");
+
+	fclose(ofp);
+
+	fprintf(stderr, "generated Perl script: %s\n", fname);
+
+	return 0;
+}
+
+struct scripting_ops perl_scripting_ops = {
+	.name = "Perl",
+	.start_script = perl_start_script,
+	.stop_script = perl_stop_script,
+	.process_event = perl_process_event,
+	.generate_script = perl_generate_script,
+};
+
+#ifdef NO_LIBPERL
+void setup_perl_scripting(void)
+{
+	fprintf(stderr, "Perl scripting not supported."
+		"  Install libperl and rebuild perf to enable it.  e.g. "
+		"apt-get install libperl-dev (ubuntu), yum install "
+		"perl-ExtUtils-Embed (Fedora), etc.\n");
+}
+#else
+void setup_perl_scripting(void)
+{
+	int err;
+	err = script_spec_register("Perl", &perl_scripting_ops);
+	if (err)
+		die("error registering Perl script extension");
+
+	err = script_spec_register("pl", &perl_scripting_ops);
+	if (err)
+		die("error registering pl script extension");
+
+	scripting_context = malloc(sizeof(struct scripting_context));
+}
+#endif
diff --git a/tools/perf/util/trace-event-perl.h b/tools/perf/util/trace-event-perl.h
new file mode 100644
index 000000000000..8fe0d866fe1a
--- /dev/null
+++ b/tools/perf/util/trace-event-perl.h
@@ -0,0 +1,51 @@
+#ifndef __PERF_TRACE_EVENT_PERL_H
+#define __PERF_TRACE_EVENT_PERL_H
+#ifdef NO_LIBPERL
+typedef int INTERP;
+#define dSP
+#define ENTER
+#define SAVETMPS
+#define PUTBACK
+#define SPAGAIN
+#define FREETMPS
+#define LEAVE
+#define SP
+#define ERRSV
+#define G_SCALAR		(0)
+#define G_DISCARD		(0)
+#define G_NOARGS		(0)
+#define PUSHMARK(a)
+#define SvTRUE(a)		(0)
+#define XPUSHs(s)
+#define sv_2mortal(a)
+#define newSVpv(a,b)
+#define newSVuv(a)
+#define newSViv(a)
+#define get_cv(a,b)		(0)
+#define call_pv(a,b)		(0)
+#define perl_alloc()		(0)
+#define perl_construct(a)	(0)
+#define perl_parse(a,b,c,d,e)	(0)
+#define perl_run(a)		(0)
+#define perl_destruct(a)	(0)
+#define perl_free(a)		(0)
+#define pTHX			void
+#define CV			void
+#define dXSUB_SYS
+#define pTHX_
+static inline void newXS(const char *a, void *b, const char *c) {}
+#else
+#include <EXTERN.h>
+#include <perl.h>
+typedef PerlInterpreter * INTERP;
+#endif
+
+struct scripting_context {
+	void *event_data;
+};
+
+int common_pc(struct scripting_context *context);
+int common_flags(struct scripting_context *context);
+int common_lock_depth(struct scripting_context *context);
+
+#endif /* __PERF_TRACE_EVENT_PERL_H */
diff --git a/tools/perf/util/trace-event-read.c b/tools/perf/util/trace-event-read.c
index 1b5c847d2c22..342dfdd43f87 100644
--- a/tools/perf/util/trace-event-read.c
+++ b/tools/perf/util/trace-event-read.c
@@ -458,9 +458,8 @@ struct record *trace_read_data(int cpu)
 	return data;
 }
 
-void trace_report(void)
+void trace_report(int fd)
 {
-	const char *input_file = "trace.info";
 	char buf[BUFSIZ];
 	char test[] = { 23, 8, 68 };
 	char *version;
@@ -468,17 +467,15 @@ void trace_report(void)
 	int show_funcs = 0;
 	int show_printk = 0;
 
-	input_fd = open(input_file, O_RDONLY);
-	if (input_fd < 0)
-		die("opening '%s'\n", input_file);
+	input_fd = fd;
 
 	read_or_die(buf, 3);
 	if (memcmp(buf, test, 3) != 0)
-		die("not an trace data file");
+		die("no trace data in the file");
 
 	read_or_die(buf, 7);
 	if (memcmp(buf, "tracing", 7) != 0)
-		die("not a trace file (missing tracing)");
+		die("not a trace file (missing 'tracing' tag)");
 
 	version = read_string();
 	if (show_version)
diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h
index 693f815c9429..81698d5e6503 100644
--- a/tools/perf/util/trace-event.h
+++ b/tools/perf/util/trace-event.h
@@ -1,5 +1,5 @@
-#ifndef _TRACE_EVENTS_H
-#define _TRACE_EVENTS_H
+#ifndef __PERF_TRACE_EVENTS_H
+#define __PERF_TRACE_EVENTS_H
 
 #include "parse-events.h"
 
@@ -26,6 +26,11 @@ enum {
 enum format_flags {
 	FIELD_IS_ARRAY		= 1,
 	FIELD_IS_POINTER	= 2,
+	FIELD_IS_SIGNED		= 4,
+	FIELD_IS_STRING		= 8,
+	FIELD_IS_DYNAMIC	= 16,
+	FIELD_IS_FLAG		= 32,
+	FIELD_IS_SYMBOLIC	= 64,
 };
 
 struct format_field {
@@ -132,15 +137,18 @@ struct event {
 	int			flags;
 	struct format		format;
 	struct print_fmt	print_fmt;
+	char			*system;
 };
 
 enum {
-	EVENT_FL_ISFTRACE	= 1,
-	EVENT_FL_ISPRINT	= 2,
-	EVENT_FL_ISBPRINT	= 4,
-	EVENT_FL_ISFUNC		= 8,
-	EVENT_FL_ISFUNCENT	= 16,
-	EVENT_FL_ISFUNCRET	= 32,
+	EVENT_FL_ISFTRACE	= 0x01,
+	EVENT_FL_ISPRINT	= 0x02,
+	EVENT_FL_ISBPRINT	= 0x04,
+	EVENT_FL_ISFUNC		= 0x08,
+	EVENT_FL_ISFUNCENT	= 0x10,
+	EVENT_FL_ISFUNCRET	= 0x20,
+
+	EVENT_FL_FAILED		= 0x80000000
 };
 
 struct record {
@@ -154,7 +162,7 @@ struct record *trace_read_data(int cpu);
 
 void parse_set_info(int nr_cpus, int long_sz);
 
-void trace_report(void);
+void trace_report(int fd);
 
 void *malloc_or_die(unsigned int size);
 
@@ -166,7 +174,7 @@ void print_funcs(void);
 void print_printk(void);
 
 int parse_ftrace_file(char *buf, unsigned long size);
-int parse_event_file(char *buf, unsigned long size, char *system);
+int parse_event_file(char *buf, unsigned long size, char *sys);
 void print_event(int cpu, void *data, int size, unsigned long long nsecs,
 		  char *comm);
 
@@ -233,13 +241,45 @@ extern int header_page_size_size;
 extern int header_page_data_offset;
 extern int header_page_data_size;
 
+extern int latency_format;
+
 int parse_header_page(char *buf, unsigned long size);
 int trace_parse_common_type(void *data);
+int trace_parse_common_pid(void *data);
+int parse_common_pc(void *data);
+int parse_common_flags(void *data);
+int parse_common_lock_depth(void *data);
 struct event *trace_find_event(int id);
+struct event *trace_find_next_event(struct event *event);
+unsigned long long read_size(void *ptr, int size);
 unsigned long long
 raw_field_value(struct event *event, const char *name, void *data);
 void *raw_field_ptr(struct event *event, const char *name, void *data);
+unsigned long long eval_flag(const char *flag);
+
+int read_tracing_data(int fd, struct perf_event_attr *pattrs, int nb_events);
+
+/* taken from kernel/trace/trace.h */
+enum trace_flag_type {
+	TRACE_FLAG_IRQS_OFF		= 0x01,
+	TRACE_FLAG_IRQS_NOSUPPORT	= 0x02,
+	TRACE_FLAG_NEED_RESCHED		= 0x04,
+	TRACE_FLAG_HARDIRQ		= 0x08,
+	TRACE_FLAG_SOFTIRQ		= 0x10,
+};
+
+struct scripting_ops {
+	const char *name;
+	int (*start_script) (const char *);
+	int (*stop_script) (void);
+	void (*process_event) (int cpu, void *data, int size,
+			       unsigned long long nsecs, char *comm);
+	int (*generate_script) (const char *outfile);
+};
+
+int script_spec_register(const char *spec, struct scripting_ops *ops);
 
-void read_tracing_data(struct perf_event_attr *pattrs, int nb_events);
+extern struct scripting_ops perl_scripting_ops;
+void setup_perl_scripting(void);
 
-#endif /* _TRACE_EVENTS_H */
+#endif /* __PERF_TRACE_EVENTS_H */
diff --git a/tools/perf/util/types.h b/tools/perf/util/types.h
index 5e75f9005940..7d6b8331f898 100644
--- a/tools/perf/util/types.h
+++ b/tools/perf/util/types.h
@@ -1,5 +1,5 @@
-#ifndef _PERF_TYPES_H
-#define _PERF_TYPES_H
+#ifndef __PERF_TYPES_H
+#define __PERF_TYPES_H
 
 /*
  * We define u64 as unsigned long long for every architecture
@@ -14,4 +14,4 @@ typedef signed short	   s16;
 typedef unsigned char	   u8;
 typedef signed char	   s8;
 
-#endif /* _PERF_TYPES_H */
+#endif /* __PERF_TYPES_H */
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 9de2329dd44d..c673d8825883 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -84,6 +84,9 @@
 #include <iconv.h>
 #endif
 
+extern const char *graph_line;
+extern const char *graph_dotted_line;
+
 /* On most systems <limits.h> would have given us this, but
  * not on some systems (e.g. GNU/Hurd).
  */
@@ -134,6 +137,15 @@ extern void die(const char *err, ...) NORETURN __attribute__((format (printf, 1,
 extern int error(const char *err, ...) __attribute__((format (printf, 1, 2)));
 extern void warning(const char *err, ...) __attribute__((format (printf, 1, 2)));
 
+#include "../../../include/linux/stringify.h"
+
+#define DIE_IF(cnd)	\
+	do { if (cnd)	\
+		die(" at (" __FILE__ ":" __stringify(__LINE__) "): "	\
+		    __stringify(cnd) "\n");				\
+	} while (0)
+
+
 extern void set_die_routine(void (*routine)(const char *err, va_list params) NORETURN);
 
 extern int prefixcmp(const char *str, const char *prefix);
@@ -278,17 +290,15 @@ static inline char *gitstrchrnul(const char *s, int c)
  * Wrappers:
  */
 extern char *xstrdup(const char *str);
-extern void *xmalloc(size_t size);
+extern void *xmalloc(size_t size) __attribute__((weak));
 extern void *xmemdupz(const void *data, size_t len);
 extern char *xstrndup(const char *str, size_t len);
-extern void *xrealloc(void *ptr, size_t size);
-extern void *xcalloc(size_t nmemb, size_t size);
-extern void *xmmap(void *start, size_t length, int prot, int flags, int fd, off_t offset);
-extern ssize_t xread(int fd, void *buf, size_t len);
-extern ssize_t xwrite(int fd, const void *buf, size_t len);
-extern int xdup(int fd);
-extern FILE *xfdopen(int fd, const char *mode);
-extern int xmkstemp(char *template);
+extern void *xrealloc(void *ptr, size_t size) __attribute__((weak));
+
+static inline void *zalloc(size_t size)
+{
+	return calloc(1, size);
+}
 
 static inline size_t xsize_t(off_t len)
 {
@@ -306,6 +316,7 @@ static inline int has_extension(const char *filename, const char *ext)
 #undef isascii
 #undef isspace
 #undef isdigit
+#undef isxdigit
 #undef isalpha
 #undef isprint
 #undef isalnum
@@ -323,6 +334,8 @@ extern unsigned char sane_ctype[256];
 #define isascii(x) (((x) & ~0x7f) == 0)
 #define isspace(x) sane_istest(x,GIT_SPACE)
 #define isdigit(x) sane_istest(x,GIT_DIGIT)
+#define isxdigit(x)	\
+	(sane_istest(toupper(x), GIT_ALPHA | GIT_DIGIT) && toupper(x) < 'G')
 #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)
diff --git a/tools/perf/util/values.h b/tools/perf/util/values.h
index cadf8cf2a590..2fa967e1a88a 100644
--- a/tools/perf/util/values.h
+++ b/tools/perf/util/values.h
@@ -1,5 +1,5 @@
-#ifndef _PERF_VALUES_H
-#define _PERF_VALUES_H
+#ifndef __PERF_VALUES_H
+#define __PERF_VALUES_H
 
 #include "types.h"
 
@@ -24,4 +24,4 @@ void perf_read_values_add_value(struct perf_read_values *values,
 void perf_read_values_display(FILE *fp, struct perf_read_values *values,
 			      int raw);
 
-#endif /* _PERF_VALUES_H */
+#endif /* __PERF_VALUES_H */
diff --git a/tools/perf/util/wrapper.c b/tools/perf/util/wrapper.c
index 4574ac28396f..bf44ca85d23b 100644
--- a/tools/perf/util/wrapper.c
+++ b/tools/perf/util/wrapper.c
@@ -79,43 +79,12 @@ void *xrealloc(void *ptr, size_t size)
 	return ret;
 }
 
-void *xcalloc(size_t nmemb, size_t size)
-{
-	void *ret = calloc(nmemb, size);
-	if (!ret && (!nmemb || !size))
-		ret = calloc(1, 1);
-	if (!ret) {
-		release_pack_memory(nmemb * size, -1);
-		ret = calloc(nmemb, size);
-		if (!ret && (!nmemb || !size))
-			ret = calloc(1, 1);
-		if (!ret)
-			die("Out of memory, calloc failed");
-	}
-	return ret;
-}
-
-void *xmmap(void *start, size_t length,
-	int prot, int flags, int fd, off_t offset)
-{
-	void *ret = mmap(start, length, prot, flags, fd, offset);
-	if (ret == MAP_FAILED) {
-		if (!length)
-			return NULL;
-		release_pack_memory(length, fd);
-		ret = mmap(start, length, prot, flags, fd, offset);
-		if (ret == MAP_FAILED)
-			die("Out of memory? mmap failed: %s", strerror(errno));
-	}
-	return ret;
-}
-
 /*
  * xread() is the same a read(), but it automatically restarts read()
  * operations with a recoverable error (EAGAIN and EINTR). xread()
  * DOES NOT GUARANTEE that "len" bytes is read even if the data is available.
  */
-ssize_t xread(int fd, void *buf, size_t len)
+static ssize_t xread(int fd, void *buf, size_t len)
 {
 	ssize_t nr;
 	while (1) {
@@ -131,7 +100,7 @@ ssize_t xread(int fd, void *buf, size_t len)
  * operations with a recoverable error (EAGAIN and EINTR). xwrite() DOES NOT
  * GUARANTEE that "len" bytes is written even if the operation is successful.
  */
-ssize_t xwrite(int fd, const void *buf, size_t len)
+static ssize_t xwrite(int fd, const void *buf, size_t len)
 {
 	ssize_t nr;
 	while (1) {
@@ -179,29 +148,3 @@ ssize_t write_in_full(int fd, const void *buf, size_t count)
 
 	return total;
 }
-
-int xdup(int fd)
-{
-	int ret = dup(fd);
-	if (ret < 0)
-		die("dup failed: %s", strerror(errno));
-	return ret;
-}
-
-FILE *xfdopen(int fd, const char *mode)
-{
-	FILE *stream = fdopen(fd, mode);
-	if (stream == NULL)
-		die("Out of memory? fdopen failed: %s", strerror(errno));
-	return stream;
-}
-
-int xmkstemp(char *template)
-{
-	int fd;
-
-	fd = mkstemp(template);
-	if (fd < 0)
-		die("Unable to create temporary file: %s", strerror(errno));
-	return fd;
-}