Merge branch 'perf/core' into perf/probes

Resolved merge conflict in tools/perf/Makefile

Merge reason: we want to queue up a dependent patch.

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

41 files changed, 2413 insertions, 567 deletions

diff --git a/tools/perf/Documentation/perf-bench.txt b/tools/perf/Documentation/perf-bench.txt
new file mode 100644
index 0000000..ae525ac
--- /dev/null
+++ b/tools/perf/Documentation/perf-bench.txt
@@ -0,0 +1,120 @@
+perf-bench(1)
+============
+
+NAME
+----
+perf-bench - General framework for benchmark suites
+
+SYNOPSIS
+--------
+[verse]
+'perf bench' [<common options>] <subsystem> <suite> [<options>]
+
+DESCRIPTION
+-----------
+This 'perf bench' command is general framework for benchmark suites.
+
+COMMON OPTIONS
+--------------
+-f::
+--format=::
+Specify format style.
+Current available format styles are,
+
+'default'::
+Default style. This is mainly for human reading.
+---------------------
+% perf bench sched pipe                      # with no style specify
+(executing 1000000 pipe operations between two tasks)
+        Total time:5.855 sec
+                5.855061 usecs/op
+		170792 ops/sec
+---------------------
+
+'simple'::
+This simple style is friendly for automated
+processing by scripts.
+---------------------
+% perf bench --format=simple sched pipe      # specified simple
+5.988
+---------------------
+
+SUBSYSTEM
+---------
+
+'sched'::
+	Scheduler and IPC mechanisms.
+
+SUITES FOR 'sched'
+~~~~~~~~~~~~~~~~~~
+*messaging*::
+Suite for evaluating performance of scheduler and IPC mechanisms.
+Based on hackbench by Rusty Russell.
+
+Options of *pipe*
+^^^^^^^^^^^^^^^^^
+-p::
+--pipe::
+Use pipe() instead of socketpair()
+
+-t::
+--thread::
+Be multi thread instead of multi process
+
+-g::
+--group=::
+Specify number of groups
+
+-l::
+--loop=::
+Specify number of loops
+
+Example of *messaging*
+^^^^^^^^^^^^^^^^^^^^^^
+
+---------------------
+% perf bench sched messaging                 # run with default
+options (20 sender and receiver processes per group)
+(10 groups == 400 processes run)
+
+      Total time:0.308 sec
+
+% perf bench sched messaging -t -g 20        # be multi-thread,with 20 groups
+(20 sender and receiver threads per group)
+(20 groups == 800 threads run)
+
+      Total time:0.582 sec
+---------------------
+
+*pipe*::
+Suite for pipe() system call.
+Based on pipe-test-1m.c by Ingo Molnar.
+
+Options of *pipe*
+^^^^^^^^^^^^^^^^^
+-l::
+--loop=::
+Specify number of loops.
+
+Example of *pipe*
+^^^^^^^^^^^^^^^^^
+
+---------------------
+% perf bench sched pipe
+(executing 1000000 pipe operations between two tasks)
+
+        Total time:8.091 sec
+                8.091833 usecs/op
+                123581 ops/sec
+
+% perf bench sched pipe -l 1000              # loop 1000
+(executing 1000 pipe operations between two tasks)
+
+        Total time:0.016 sec
+                16.948000 usecs/op
+                59004 ops/sec
+---------------------
+
+SEE ALSO
+--------
+linkperf:perf[1]
diff --git a/tools/perf/Documentation/perf-buildid-list.txt b/tools/perf/Documentation/perf-buildid-list.txt
new file mode 100644
index 0000000..01b642c
--- /dev/null
+++ b/tools/perf/Documentation/perf-buildid-list.txt
@@ -0,0 +1,34 @@
+perf-buildid-list(1)
+====================
+
+NAME
+----
+perf-buildid-list - List the buildids in a perf.data file
+
+SYNOPSIS
+--------
+[verse]
+'perf buildid-list <options>'
+
+DESCRIPTION
+-----------
+This command displays the buildids found in a perf.data file, so that other
+tools can be used to fetch packages with matching symbol tables for use by
+perf report.
+
+OPTIONS
+-------
+-i::
+--input=::
+        Input file name. (default: perf.data)
+-f::
+--force::
+	Don't do ownership validation.
+-v::
+--verbose::
+	Be more verbose.
+
+SEE ALSO
+--------
+linkperf:perf-record[1], linkperf:perf-top[1],
+linkperf:perf-report[1]
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index 59f0b84..9dccb18 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -24,11 +24,11 @@ OPTIONS
 --dsos=::
 	Only consider symbols in these dsos. CSV that understands
 	file://filename entries.
--n
---show-nr-samples
+-n::
+--show-nr-samples::
 	Show the number of samples for each symbol
--T
---threads
+-T::
+--threads::
 	Show per-thread event counters
 -C::
 --comms=::
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 147e3cf..3dbb5c5 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -177,8 +177,7 @@ endif
 # Include saner warnings here, which can catch bugs:
 #
 
-EXTRA_WARNINGS := -Wcast-align
-EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat
+EXTRA_WARNINGS := -Wformat
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-security
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-y2k
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wshadow
@@ -208,7 +207,7 @@ ifndef PERF_DEBUG
   CFLAGS_OPTIMIZE = -O6
 endif
 
-CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
+CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
 LDFLAGS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS)
 ALL_LDFLAGS = $(LDFLAGS)
@@ -260,6 +259,9 @@ PTHREAD_LIBS = -lpthread
 # explicitly what architecture to check for. Fix this up for yours..
 SPARSE_FLAGS = -D__BIG_ENDIAN__ -D__powerpc__
 
+ifeq ($(shell sh -c "echo 'int foo(void) {char X[2]; return 3;}' | $(CC) -x c -c -Werror -fstack-protector-all - -o /dev/null >/dev/null 2>&1 && echo y"), y)
+  CFLAGS := $(CFLAGS) -fstack-protector-all
+endif
 
 
 ### --- END CONFIGURATION SECTION ---
@@ -355,6 +357,7 @@ LIB_H += util/include/asm/swab.h
 LIB_H += util/include/asm/system.h
 LIB_H += util/include/asm/uaccess.h
 LIB_H += perf.h
+LIB_H += util/debugfs.h
 LIB_H += util/event.h
 LIB_H += util/types.h
 LIB_H += util/levenshtein.h
@@ -380,7 +383,9 @@ LIB_OBJS += util/abspath.o
 LIB_OBJS += util/alias.o
 LIB_OBJS += util/config.o
 LIB_OBJS += util/ctype.o
+LIB_OBJS += util/debugfs.o
 LIB_OBJS += util/environment.o
+LIB_OBJS += util/event.o
 LIB_OBJS += util/exec_cmd.o
 LIB_OBJS += util/help.o
 LIB_OBJS += util/levenshtein.o
@@ -417,8 +422,16 @@ LIB_OBJS += util/hist.o
 LIB_OBJS += util/data_map.o
 
 BUILTIN_OBJS += builtin-annotate.o
+
+BUILTIN_OBJS += builtin-bench.o
+
+# Benchmark modules
+BUILTIN_OBJS += bench/sched-messaging.o
+BUILTIN_OBJS += bench/sched-pipe.o
+
 BUILTIN_OBJS += builtin-help.o
 BUILTIN_OBJS += builtin-sched.o
+BUILTIN_OBJS += builtin-buildid-list.o
 BUILTIN_OBJS += builtin-list.o
 BUILTIN_OBJS += builtin-record.o
 BUILTIN_OBJS += builtin-report.o
@@ -457,12 +470,16 @@ ifeq ($(uname_S),Darwin)
 	PTHREAD_LIBS =
 endif
 
+ifeq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
 ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
 	msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]);
 endif
 
-ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
-	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel);
+	ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
+		BASIC_CFLAGS += -DLIBELF_NO_MMAP
+	endif
+else
+	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]);
 endif
 
 ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
diff --git a/tools/perf/bench/bench.h b/tools/perf/bench/bench.h
new file mode 100644
index 0000000..9fbd8d7
--- /dev/null
+++ b/tools/perf/bench/bench.h
@@ -0,0 +1,16 @@
+#ifndef BENCH_H
+#define BENCH_H
+
+extern int bench_sched_messaging(int argc, const char **argv, const char *prefix);
+extern int bench_sched_pipe(int argc, const char **argv, const char *prefix);
+
+#define BENCH_FORMAT_DEFAULT_STR	"default"
+#define BENCH_FORMAT_DEFAULT		0
+#define BENCH_FORMAT_SIMPLE_STR		"simple"
+#define BENCH_FORMAT_SIMPLE		1
+
+#define BENCH_FORMAT_UNKNOWN		-1
+
+extern int bench_format;
+
+#endif
diff --git a/tools/perf/bench/sched-messaging.c b/tools/perf/bench/sched-messaging.c
new file mode 100644
index 0000000..605a2a9
--- /dev/null
+++ b/tools/perf/bench/sched-messaging.c
@@ -0,0 +1,336 @@
+/*
+ *
+ * builtin-bench-messaging.c
+ *
+ * messaging: Benchmark for scheduler and IPC mechanisms
+ *
+ * Based on hackbench by Rusty Russell <rusty@rustcorp.com.au>
+ * Ported to perf by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/parse-options.h"
+#include "../builtin.h"
+#include "bench.h"
+
+/* Test groups of 20 processes spraying to 20 receivers */
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <sys/time.h>
+#include <sys/poll.h>
+#include <limits.h>
+
+#define DATASIZE 100
+
+static int use_pipes = 0;
+static unsigned int loops = 100;
+static unsigned int thread_mode = 0;
+static unsigned int num_groups = 10;
+
+struct sender_context {
+	unsigned int num_fds;
+	int ready_out;
+	int wakefd;
+	int out_fds[0];
+};
+
+struct receiver_context {
+	unsigned int num_packets;
+	int in_fds[2];
+	int ready_out;
+	int wakefd;
+};
+
+static void barf(const char *msg)
+{
+	fprintf(stderr, "%s (error: %s)\n", msg, strerror(errno));
+	exit(1);
+}
+
+static void fdpair(int fds[2])
+{
+	if (use_pipes) {
+		if (pipe(fds) == 0)
+			return;
+	} else {
+		if (socketpair(AF_UNIX, SOCK_STREAM, 0, fds) == 0)
+			return;
+	}
+
+	barf(use_pipes ? "pipe()" : "socketpair()");
+}
+
+/* Block until we're ready to go */
+static void ready(int ready_out, int wakefd)
+{
+	char dummy;
+	struct pollfd pollfd = { .fd = wakefd, .events = POLLIN };
+
+	/* Tell them we're ready. */
+	if (write(ready_out, &dummy, 1) != 1)
+		barf("CLIENT: ready write");
+
+	/* Wait for "GO" signal */
+	if (poll(&pollfd, 1, -1) != 1)
+		barf("poll");
+}
+
+/* Sender sprays loops messages down each file descriptor */
+static void *sender(struct sender_context *ctx)
+{
+	char data[DATASIZE];
+	unsigned int i, j;
+
+	ready(ctx->ready_out, ctx->wakefd);
+
+	/* Now pump to every receiver. */
+	for (i = 0; i < loops; i++) {
+		for (j = 0; j < ctx->num_fds; j++) {
+			int ret, done = 0;
+
+again:
+			ret = write(ctx->out_fds[j], data + done,
+				    sizeof(data)-done);
+			if (ret < 0)
+				barf("SENDER: write");
+			done += ret;
+			if (done < DATASIZE)
+				goto again;
+		}
+	}
+
+	return NULL;
+}
+
+
+/* One receiver per fd */
+static void *receiver(struct receiver_context* ctx)
+{
+	unsigned int i;
+
+	if (!thread_mode)
+		close(ctx->in_fds[1]);
+
+	/* Wait for start... */
+	ready(ctx->ready_out, ctx->wakefd);
+
+	/* Receive them all */
+	for (i = 0; i < ctx->num_packets; i++) {
+		char data[DATASIZE];
+		int ret, done = 0;
+
+again:
+		ret = read(ctx->in_fds[0], data + done, DATASIZE - done);
+		if (ret < 0)
+			barf("SERVER: read");
+		done += ret;
+		if (done < DATASIZE)
+			goto again;
+	}
+
+	return NULL;
+}
+
+static pthread_t create_worker(void *ctx, void *(*func)(void *))
+{
+	pthread_attr_t attr;
+	pthread_t childid;
+	int err;
+
+	if (!thread_mode) {
+		/* process mode */
+		/* Fork the receiver. */
+		switch (fork()) {
+		case -1:
+			barf("fork()");
+			break;
+		case 0:
+			(*func) (ctx);
+			exit(0);
+			break;
+		default:
+			break;
+		}
+
+		return (pthread_t)0;
+	}
+
+	if (pthread_attr_init(&attr) != 0)
+		barf("pthread_attr_init:");
+
+#ifndef __ia64__
+	if (pthread_attr_setstacksize(&attr, PTHREAD_STACK_MIN) != 0)
+		barf("pthread_attr_setstacksize");
+#endif
+
+	err = pthread_create(&childid, &attr, func, ctx);
+	if (err != 0) {
+		fprintf(stderr, "pthread_create failed: %s (%d)\n",
+			strerror(err), err);
+		exit(-1);
+	}
+	return childid;
+}
+
+static void reap_worker(pthread_t id)
+{
+	int proc_status;
+	void *thread_status;
+
+	if (!thread_mode) {
+		/* process mode */
+		wait(&proc_status);
+		if (!WIFEXITED(proc_status))
+			exit(1);
+	} else {
+		pthread_join(id, &thread_status);
+	}
+}
+
+/* One group of senders and receivers */
+static unsigned int group(pthread_t *pth,
+		unsigned int num_fds,
+		int ready_out,
+		int wakefd)
+{
+	unsigned int i;
+	struct sender_context *snd_ctx = malloc(sizeof(struct sender_context)
+			+ num_fds * sizeof(int));
+
+	if (!snd_ctx)
+		barf("malloc()");
+
+	for (i = 0; i < num_fds; i++) {
+		int fds[2];
+		struct receiver_context *ctx = malloc(sizeof(*ctx));
+
+		if (!ctx)
+			barf("malloc()");
+
+
+		/* Create the pipe between client and server */
+		fdpair(fds);
+
+		ctx->num_packets = num_fds * loops;
+		ctx->in_fds[0] = fds[0];
+		ctx->in_fds[1] = fds[1];
+		ctx->ready_out = ready_out;
+		ctx->wakefd = wakefd;
+
+		pth[i] = create_worker(ctx, (void *)receiver);
+
+		snd_ctx->out_fds[i] = fds[1];
+		if (!thread_mode)
+			close(fds[0]);
+	}
+
+	/* Now we have all the fds, fork the senders */
+	for (i = 0; i < num_fds; i++) {
+		snd_ctx->ready_out = ready_out;
+		snd_ctx->wakefd = wakefd;
+		snd_ctx->num_fds = num_fds;
+
+		pth[num_fds+i] = create_worker(snd_ctx, (void *)sender);
+	}
+
+	/* Close the fds we have left */
+	if (!thread_mode)
+		for (i = 0; i < num_fds; i++)
+			close(snd_ctx->out_fds[i]);
+
+	/* Return number of children to reap */
+	return num_fds * 2;
+}
+
+static const struct option options[] = {
+	OPT_BOOLEAN('p', "pipe", &use_pipes,
+		    "Use pipe() instead of socketpair()"),
+	OPT_BOOLEAN('t', "thread", &thread_mode,
+		    "Be multi thread instead of multi process"),
+	OPT_INTEGER('g', "group", &num_groups,
+		    "Specify number of groups"),
+	OPT_INTEGER('l', "loop", &loops,
+		    "Specify number of loops"),
+	OPT_END()
+};
+
+static const char * const bench_sched_message_usage[] = {
+	"perf bench sched messaging <options>",
+	NULL
+};
+
+int bench_sched_messaging(int argc, const char **argv,
+		    const char *prefix __used)
+{
+	unsigned int i, total_children;
+	struct timeval start, stop, diff;
+	unsigned int num_fds = 20;
+	int readyfds[2], wakefds[2];
+	char dummy;
+	pthread_t *pth_tab;
+
+	argc = parse_options(argc, argv, options,
+			     bench_sched_message_usage, 0);
+
+	pth_tab = malloc(num_fds * 2 * num_groups * sizeof(pthread_t));
+	if (!pth_tab)
+		barf("main:malloc()");
+
+	fdpair(readyfds);
+	fdpair(wakefds);
+
+	total_children = 0;
+	for (i = 0; i < num_groups; i++)
+		total_children += group(pth_tab+total_children, num_fds,
+					readyfds[1], wakefds[0]);
+
+	/* Wait for everyone to be ready */
+	for (i = 0; i < total_children; i++)
+		if (read(readyfds[0], &dummy, 1) != 1)
+			barf("Reading for readyfds");
+
+	gettimeofday(&start, NULL);
+
+	/* Kick them off */
+	if (write(wakefds[1], &dummy, 1) != 1)
+		barf("Writing to start them");
+
+	/* Reap them all */
+	for (i = 0; i < total_children; i++)
+		reap_worker(pth_tab[i]);
+
+	gettimeofday(&stop, NULL);
+
+	timersub(&stop, &start, &diff);
+
+	switch (bench_format) {
+	case BENCH_FORMAT_DEFAULT:
+		printf("# %d sender and receiver %s per group\n",
+		       num_fds, thread_mode ? "threads" : "processes");
+		printf("# %d groups == %d %s run\n\n",
+		       num_groups, num_groups * 2 * num_fds,
+		       thread_mode ? "threads" : "processes");
+		printf(" %14s: %lu.%03lu [sec]\n", "Total time",
+		       diff.tv_sec, diff.tv_usec/1000);
+		break;
+	case BENCH_FORMAT_SIMPLE:
+		printf("%lu.%03lu\n", diff.tv_sec, diff.tv_usec/1000);
+		break;
+	default:
+		/* reaching here is something disaster */
+		fprintf(stderr, "Unknown format:%d\n", bench_format);
+		exit(1);
+		break;
+	}
+
+	return 0;
+}
diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c
new file mode 100644
index 0000000..238185f
--- /dev/null
+++ b/tools/perf/bench/sched-pipe.c
@@ -0,0 +1,124 @@
+/*
+ *
+ * builtin-bench-pipe.c
+ *
+ * pipe: Benchmark for pipe()
+ *
+ * Based on pipe-test-1m.c by Ingo Molnar <mingo@redhat.com>
+ *  http://people.redhat.com/mingo/cfs-scheduler/tools/pipe-test-1m.c
+ * Ported to perf by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/parse-options.h"
+#include "../builtin.h"
+#include "bench.h"
+
+#include <unistd.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <sys/wait.h>
+#include <linux/unistd.h>
+#include <string.h>
+#include <errno.h>
+#include <assert.h>
+#include <sys/time.h>
+#include <sys/types.h>
+
+#define LOOPS_DEFAULT 1000000
+static int loops = LOOPS_DEFAULT;
+
+static const struct option options[] = {
+	OPT_INTEGER('l', "loop", &loops,
+		    "Specify number of loops"),
+	OPT_END()
+};
+
+static const char * const bench_sched_pipe_usage[] = {
+	"perf bench sched pipe <options>",
+	NULL
+};
+
+int bench_sched_pipe(int argc, const char **argv,
+		     const char *prefix __used)
+{
+	int pipe_1[2], pipe_2[2];
+	int m = 0, i;
+	struct timeval start, stop, diff;
+	unsigned long long result_usec = 0;
+
+	/*
+	 * why does "ret" exist?
+	 * discarding returned value of read(), write()
+	 * causes error in building environment for perf
+	 */
+	int ret, wait_stat;
+	pid_t pid, retpid;
+
+	argc = parse_options(argc, argv, options,
+			     bench_sched_pipe_usage, 0);
+
+	assert(!pipe(pipe_1));
+	assert(!pipe(pipe_2));
+
+	pid = fork();
+	assert(pid >= 0);
+
+	gettimeofday(&start, NULL);
+
+	if (!pid) {
+		for (i = 0; i < loops; i++) {
+			ret = read(pipe_1[0], &m, sizeof(int));
+			ret = write(pipe_2[1], &m, sizeof(int));
+		}
+	} else {
+		for (i = 0; i < loops; i++) {
+			ret = write(pipe_1[1], &m, sizeof(int));
+			ret = read(pipe_2[0], &m, sizeof(int));
+		}
+	}
+
+	gettimeofday(&stop, NULL);
+	timersub(&stop, &start, &diff);
+
+	if (pid) {
+		retpid = waitpid(pid, &wait_stat, 0);
+		assert((retpid == pid) && WIFEXITED(wait_stat));
+		return 0;
+	}
+
+	switch (bench_format) {
+	case BENCH_FORMAT_DEFAULT:
+		printf("# Extecuted %d pipe operations between two tasks\n\n",
+			loops);
+
+		result_usec = diff.tv_sec * 1000000;
+		result_usec += diff.tv_usec;
+
+		printf(" %14s: %lu.%03lu [sec]\n\n", "Total time",
+		       diff.tv_sec, diff.tv_usec/1000);
+
+		printf(" %14lf usecs/op\n",
+		       (double)result_usec / (double)loops);
+		printf(" %14d ops/sec\n",
+		       (int)((double)loops /
+			     ((double)result_usec / (double)1000000)));
+		break;
+
+	case BENCH_FORMAT_SIMPLE:
+		printf("%lu.%03lu\n",
+		       diff.tv_sec, diff.tv_usec / 1000);
+		break;
+
+	default:
+		/* reaching here is something disaster */
+		fprintf(stderr, "Unknown format:%d\n", bench_format);
+		exit(1);
+		break;
+	}
+
+	return 0;
+}
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 6d63c2e..77d50a6 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -55,11 +55,11 @@ struct sym_priv {
 
 static const char *sym_hist_filter;
 
-static int symbol_filter(struct map *map, struct symbol *sym)
+static int symbol_filter(struct map *map __used, struct symbol *sym)
 {
 	if (sym_hist_filter == NULL ||
 	    strcmp(sym->name, sym_hist_filter) == 0) {
-		struct sym_priv *priv = dso__sym_priv(map->dso, sym);
+		struct sym_priv *priv = symbol__priv(sym);
 		const int size = (sizeof(*priv->hist) +
 				  (sym->end - sym->start) * sizeof(u64));
 
@@ -92,7 +92,7 @@ static void hist_hit(struct hist_entry *he, u64 ip)
 	if (!sym || !he->map)
 		return;
 
-	priv = dso__sym_priv(he->map->dso, sym);
+	priv = symbol__priv(sym);
 	if (!priv->hist)
 		return;
 
@@ -165,7 +165,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		if (map != NULL) {
 got_map:
 			ip = map->map_ip(map, ip);
-			sym = map->dso->find_symbol(map->dso, ip);
+			sym = map__find_symbol(map, ip, symbol_filter);
 		} else {
 			/*
 			 * If this is outside of all known maps,
@@ -202,8 +202,7 @@ got_map:
 static int
 process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct map *map = map__new(&event->mmap, NULL, 0,
-				   sizeof(struct sym_priv), symbol_filter);
+	struct map *map = map__new(&event->mmap, NULL, 0);
 	struct thread *thread = threads__findnew(event->mmap.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_MMAP %d: [%p(%p) @ %p]: %s\n",
@@ -355,7 +354,7 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len)
 		unsigned int hits = 0;
 		double percent = 0.0;
 		const char *color;
-		struct sym_priv *priv = dso__sym_priv(he->map->dso, sym);
+		struct sym_priv *priv = symbol__priv(sym);
 		struct sym_ext *sym_ext = priv->ext;
 		struct sym_hist *h = priv->hist;
 
@@ -422,7 +421,7 @@ static void insert_source_line(struct sym_ext *sym_ext)
 
 static void free_source_line(struct hist_entry *he, int len)
 {
-	struct sym_priv *priv = dso__sym_priv(he->map->dso, he->sym);
+	struct sym_priv *priv = symbol__priv(he->sym);
 	struct sym_ext *sym_ext = priv->ext;
 	int i;
 
@@ -446,7 +445,7 @@ get_source_line(struct hist_entry *he, int len, const char *filename)
 	int i;
 	char cmd[PATH_MAX * 2];
 	struct sym_ext *sym_ext;
-	struct sym_priv *priv = dso__sym_priv(he->map->dso, sym);
+	struct sym_priv *priv = symbol__priv(sym);
 	struct sym_hist *h = priv->hist;
 
 	if (!h->sum)
@@ -589,7 +588,7 @@ static void find_annotations(void)
 		if (he->sym == NULL)
 			continue;
 
-		priv = dso__sym_priv(he->map->dso, he->sym);
+		priv = symbol__priv(he->sym);
 		if (priv->hist == NULL)
 			continue;
 
@@ -637,7 +636,7 @@ static int __cmd_annotate(void)
 		exit(0);
 	}
 
-	if (load_kernel(sizeof(struct sym_priv), symbol_filter) < 0) {
+	if (load_kernel(symbol_filter) < 0) {
 		perror("failed to load kernel symbols");
 		return EXIT_FAILURE;
 	}
@@ -769,7 +768,7 @@ static void setup_sorting(void)
 
 int cmd_annotate(int argc, const char **argv, const char *prefix __used)
 {
-	symbol__init();
+	symbol__init(sizeof(struct sym_priv));
 
 	page_size = getpagesize();
 
diff --git a/tools/perf/builtin-bench.c b/tools/perf/builtin-bench.c
new file mode 100644
index 0000000..90c39ba
--- /dev/null
+++ b/tools/perf/builtin-bench.c
@@ -0,0 +1,183 @@
+/*
+ *
+ * builtin-bench.c
+ *
+ * General benchmarking subsystem provided by perf
+ *
+ * Copyright (C) 2009, Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+/*
+ *
+ * Available subsystem list:
+ *  sched ... scheduler and IPC mechanism
+ *
+ */
+
+#include "perf.h"
+#include "util/util.h"
+#include "util/parse-options.h"
+#include "builtin.h"
+#include "bench/bench.h"
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+struct bench_suite {
+	const char *name;
+	const char *summary;
+	int (*fn)(int, const char **, const char *);
+};
+
+static struct bench_suite sched_suites[] = {
+	{ "messaging",
+	  "Benchmark for scheduler and IPC mechanisms",
+	  bench_sched_messaging },
+	{ "pipe",
+	  "Flood of communication over pipe() between two processes",
+	  bench_sched_pipe      },
+	{ NULL,
+	  NULL,
+	  NULL                  }
+};
+
+struct bench_subsys {
+	const char *name;
+	const char *summary;
+	struct bench_suite *suites;
+};
+
+static struct bench_subsys subsystems[] = {
+	{ "sched",
+	  "scheduler and IPC mechanism",
+	  sched_suites },
+	{ NULL,
+	  NULL,
+	  NULL         }
+};
+
+static void dump_suites(int subsys_index)
+{
+	int i;
+
+	printf("List of available suites for %s...\n\n",
+	       subsystems[subsys_index].name);
+
+	for (i = 0; subsystems[subsys_index].suites[i].name; i++)
+		printf("\t%s: %s\n",
+		       subsystems[subsys_index].suites[i].name,
+		       subsystems[subsys_index].suites[i].summary);
+
+	printf("\n");
+	return;
+}
+
+static char *bench_format_str;
+int bench_format = BENCH_FORMAT_DEFAULT;
+
+static const struct option bench_options[] = {
+	OPT_STRING('f', "format", &bench_format_str, "default",
+		    "Specify format style"),
+	OPT_END()
+};
+
+static const char * const bench_usage[] = {
+	"perf bench [<common options>] <subsystem> <suite> [<options>]",
+	NULL
+};
+
+static void print_usage(void)
+{
+	int i;
+
+	printf("Usage: \n");
+	for (i = 0; bench_usage[i]; i++)
+		printf("\t%s\n", bench_usage[i]);
+	printf("\n");
+
+	printf("List of available subsystems...\n\n");
+
+	for (i = 0; subsystems[i].name; i++)
+		printf("\t%s: %s\n",
+		       subsystems[i].name, subsystems[i].summary);
+	printf("\n");
+}
+
+static int bench_str2int(char *str)
+{
+	if (!str)
+		return BENCH_FORMAT_DEFAULT;
+
+	if (!strcmp(str, BENCH_FORMAT_DEFAULT_STR))
+		return BENCH_FORMAT_DEFAULT;
+	else if (!strcmp(str, BENCH_FORMAT_SIMPLE_STR))
+		return BENCH_FORMAT_SIMPLE;
+
+	return BENCH_FORMAT_UNKNOWN;
+}
+
+int cmd_bench(int argc, const char **argv, const char *prefix __used)
+{
+	int i, j, status = 0;
+
+	if (argc < 2) {
+		/* No subsystem specified. */
+		print_usage();
+		goto end;
+	}
+
+	argc = parse_options(argc, argv, bench_options, bench_usage,
+			     PARSE_OPT_STOP_AT_NON_OPTION);
+
+	bench_format = bench_str2int(bench_format_str);
+	if (bench_format == BENCH_FORMAT_UNKNOWN) {
+		printf("Unknown format descriptor:%s\n", bench_format_str);
+		goto end;
+	}
+
+	if (argc < 1) {
+		print_usage();
+		goto end;
+	}
+
+	for (i = 0; subsystems[i].name; i++) {
+		if (strcmp(subsystems[i].name, argv[0]))
+			continue;
+
+		if (argc < 2) {
+			/* No suite specified. */
+			dump_suites(i);
+			goto end;
+		}
+
+		for (j = 0; subsystems[i].suites[j].name; j++) {
+			if (strcmp(subsystems[i].suites[j].name, argv[1]))
+				continue;
+
+			if (bench_format == BENCH_FORMAT_DEFAULT)
+				printf("# Running %s/%s benchmark...\n",
+				       subsystems[i].name,
+				       subsystems[i].suites[j].name);
+			status = subsystems[i].suites[j].fn(argc - 1,
+							    argv + 1, prefix);
+			goto end;
+		}
+
+		if (!strcmp(argv[1], "-h") || !strcmp(argv[1], "--help")) {
+			dump_suites(i);
+			goto end;
+		}
+
+		printf("Unknown suite:%s for %s\n", argv[1], argv[0]);
+		status = 1;
+		goto end;
+	}
+
+	printf("Unknown subsystem:%s\n", argv[0]);
+	status = 1;
+
+end:
+	return status;
+}
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
new file mode 100644
index 0000000..7dee9d19
--- /dev/null
+++ b/tools/perf/builtin-buildid-list.c
@@ -0,0 +1,116 @@
+/*
+ * builtin-buildid-list.c
+ *
+ * Builtin buildid-list command: list buildids in perf.data
+ *
+ * Copyright (C) 2009, Red Hat Inc.
+ * Copyright (C) 2009, Arnaldo Carvalho de Melo <acme@redhat.com>
+ */
+#include "builtin.h"
+#include "perf.h"
+#include "util/cache.h"
+#include "util/data_map.h"
+#include "util/debug.h"
+#include "util/header.h"
+#include "util/parse-options.h"
+#include "util/symbol.h"
+
+static char const *input_name = "perf.data";
+static int force;
+
+static const char *const buildid_list_usage[] = {
+	"perf report [<options>]",
+	NULL
+};
+
+static const struct option options[] = {
+	OPT_STRING('i', "input", &input_name, "file",
+		    "input file name"),
+	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
+	OPT_BOOLEAN('v', "verbose", &verbose,
+		    "be more verbose"),
+	OPT_END()
+};
+
+static int perf_file_section__process_buildids(struct perf_file_section *self,
+					       int feat, int fd)
+{
+	if (feat != HEADER_BUILD_ID)
+		return 0;
+
+	if (lseek(fd, self->offset, SEEK_SET) < 0) {
+		pr_warning("Failed to lseek to %Ld offset for buildids!\n",
+			   self->offset);
+		return -1;
+	}
+
+	if (perf_header__read_build_ids(fd, self->offset, self->size)) {
+		pr_warning("Failed to read buildids!\n");
+		return -1;
+	}
+
+	return 0;
+}
+
+static int __cmd_buildid_list(void)
+{
+	int err = -1;
+	struct perf_header *header;
+	struct perf_file_header f_header;
+	struct stat input_stat;
+	int 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");
+		goto out;
+	}
+
+	err = fstat(input, &input_stat);
+	if (err < 0) {
+		perror("failed to stat file");
+		goto out_close;
+	}
+
+	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) {
+		pr_info("zero-sized file, nothing to do!\n");
+		goto out_close;
+	}
+
+	err = -1;
+	header = perf_header__new();
+	if (header == NULL)
+		goto out_close;
+
+	if (perf_file_header__read(&f_header, header, input) < 0) {
+		pr_warning("incompatible file format");
+		goto out_close;
+	}
+
+	err = perf_header__process_sections(header, input,
+				         perf_file_section__process_buildids);
+
+	if (err < 0)
+		goto out_close;
+
+	dsos__fprintf_buildid(stdout);
+out_close:
+	close(input);
+out:
+	return err;
+}
+
+int cmd_buildid_list(int argc, const char **argv, const char *prefix __used)
+{
+	argc = parse_options(argc, argv, options, buildid_list_usage, 0);
+	setup_pager();
+	return __cmd_buildid_list();
+}
diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
index 4fb8734..768f9c8 100644
--- a/tools/perf/builtin-help.c
+++ b/tools/perf/builtin-help.c
@@ -61,8 +61,7 @@ static const char *get_man_viewer_info(const char *name)
 {
 	struct man_viewer_info_list *viewer;
 
-	for (viewer = man_viewer_info_list; viewer; viewer = viewer->next)
-	{
+	for (viewer = man_viewer_info_list; viewer; viewer = viewer->next) {
 		if (!strcasecmp(name, viewer->name))
 			return viewer->info;
 	}
@@ -115,7 +114,7 @@ static int check_emacsclient_version(void)
 	return 0;
 }
 
-static void exec_woman_emacs(const char* path, const char *page)
+static void exec_woman_emacs(const char *path, const char *page)
 {
 	if (!check_emacsclient_version()) {
 		/* This works only with emacsclient version >= 22. */
@@ -129,7 +128,7 @@ static void exec_woman_emacs(const char* path, const char *page)
 	}
 }
 
-static void exec_man_konqueror(const char* path, const char *page)
+static void exec_man_konqueror(const char *path, const char *page)
 {
 	const char *display = getenv("DISPLAY");
 	if (display && *display) {
@@ -157,7 +156,7 @@ static void exec_man_konqueror(const char* path, const char *page)
 	}
 }
 
-static void exec_man_man(const char* path, const char *page)
+static void exec_man_man(const char *path, const char *page)
 {
 	if (!path)
 		path = "man";
@@ -364,9 +363,8 @@ static void show_man_page(const char *perf_cmd)
 
 	setup_man_path();
 	for (viewer = man_viewer_list; viewer; viewer = viewer->next)
-	{
 		exec_viewer(viewer->name, page); /* will return when unable */
-	}
+
 	if (fallback)
 		exec_viewer(fallback, page);
 	exec_viewer("man", page);
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index ac5ddff..82260c5 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -17,6 +17,7 @@
 #include "util/header.h"
 #include "util/event.h"
 #include "util/debug.h"
+#include "util/symbol.h"
 
 #include <unistd.h>
 #include <sched.h>
@@ -109,6 +110,24 @@ static void write_output(void *buf, size_t size)
 	}
 }
 
+static void write_event(event_t *buf, size_t size)
+{
+	/*
+	* Add it to the list of DSOs, so that when we finish this
+	 * record session we can pick the available build-ids.
+	 */
+	if (buf->header.type == PERF_RECORD_MMAP)
+		dsos__findnew(buf->mmap.filename);
+
+	write_output(buf, size);
+}
+
+static int process_synthesized_event(event_t *event)
+{
+	write_event(event, event->header.size);
+	return 0;
+}
+
 static void mmap_read(struct mmap_data *md)
 {
 	unsigned int head = mmap_read_head(md);
@@ -157,14 +176,14 @@ static void mmap_read(struct mmap_data *md)
 		size = md->mask + 1 - (old & md->mask);
 		old += size;
 
-		write_output(buf, size);
+		write_event(buf, size);
 	}
 
 	buf = &data[old & md->mask];
 	size = head - old;
 	old += size;
 
-	write_output(buf, size);
+	write_event(buf, size);
 
 	md->prev = old;
 	mmap_write_tail(md, old);
@@ -191,168 +210,6 @@ static void sig_atexit(void)
 	kill(getpid(), signr);
 }
 
-static pid_t pid_synthesize_comm_event(pid_t pid, int full)
-{
-	struct comm_event comm_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) {
-		/*
-		 * We raced with a task exiting - just return:
-		 */
-		if (verbose)
-			fprintf(stderr, "couldn't open %s\n", filename);
-		return 0;
-	}
-
-	memset(&comm_ev, 0, sizeof(comm_ev));
-	while (!comm_ev.comm[0] || !comm_ev.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(comm_ev.comm, name, size++);
-		} else if (memcmp(bf, "Tgid:", 5) == 0) {
-			char *tgids = bf + 5;
-			while (*tgids && isspace(*tgids))
-				++tgids;
-			tgid = comm_ev.pid = atoi(tgids);
-		}
-	}
-
-	comm_ev.header.type = PERF_RECORD_COMM;
-	size = ALIGN(size, sizeof(u64));
-	comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size);
-
-	if (!full) {
-		comm_ev.tid = pid;
-
-		write_output(&comm_ev, comm_ev.header.size);
-		goto out_fclose;
-	}
-
-	snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
-
-	tasks = opendir(filename);
-	while (!readdir_r(tasks, &dirent, &next) && next) {
-		char *end;
-		pid = strtol(dirent.d_name, &end, 10);
-		if (*end)
-			continue;
-
-		comm_ev.tid = pid;
-
-		write_output(&comm_ev, comm_ev.header.size);
-	}
-	closedir(tasks);
-
-out_fclose:
-	fclose(fp);
-	return tgid;
-
-out_failure:
-	fprintf(stderr, "couldn't get COMM and pgid, malformed %s\n",
-		filename);
-	exit(EXIT_FAILURE);
-}
-
-static void pid_synthesize_mmap_samples(pid_t pid, pid_t tgid)
-{
-	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:
-		 */
-		if (verbose)
-			fprintf(stderr, "couldn't open %s\n", filename);
-		return;
-	}
-	while (1) {
-		char bf[BUFSIZ], *pbf = bf;
-		struct mmap_event mmap_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, &mmap_ev.start);
-		if (n < 0)
-			continue;
-		pbf += n + 1;
-		n = hex2u64(pbf, &mmap_ev.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(mmap_ev.filename, execname, size);
-			size = ALIGN(size, sizeof(u64));
-			mmap_ev.len -= mmap_ev.start;
-			mmap_ev.header.size = (sizeof(mmap_ev) -
-					       (sizeof(mmap_ev.filename) - size));
-			mmap_ev.pid = tgid;
-			mmap_ev.tid = pid;
-
-			write_output(&mmap_ev, mmap_ev.header.size);
-		}
-	}
-
-	fclose(fp);
-}
-
-static void synthesize_all(void)
-{
-	DIR *proc;
-	struct dirent dirent, *next;
-
-	proc = opendir("/proc");
-
-	while (!readdir_r(proc, &dirent, &next) && next) {
-		char *end;
-		pid_t pid, tgid;
-
-		pid = strtol(dirent.d_name, &end, 10);
-		if (*end) /* only interested in proper numerical dirents */
-			continue;
-
-		tgid = pid_synthesize_comm_event(pid, 1);
-		pid_synthesize_mmap_samples(pid, tgid);
-	}
-
-	closedir(proc);
-}
-
 static int group_fd;
 
 static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int nr)
@@ -363,7 +220,11 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n
 		h_attr = header->attr[nr];
 	} else {
 		h_attr = perf_header_attr__new(a);
-		perf_header__add_attr(header, h_attr);
+		if (h_attr != NULL)
+			if (perf_header__add_attr(header, h_attr) < 0) {
+				perf_header_attr__delete(h_attr);
+				h_attr = NULL;
+			}
 	}
 
 	return h_attr;
@@ -424,7 +285,7 @@ try_again:
 	if (fd[nr_cpu][counter] < 0) {
 		int err = errno;
 
-		if (err == EPERM)
+		if (err == EPERM || err == EACCES)
 			die("Permission error - are you root?\n");
 		else if (err ==  ENODEV && profile_cpu != -1)
 			die("No such device - did you specify an out-of-range profile CPU?\n");
@@ -451,6 +312,8 @@ try_again:
 	}
 
 	h_attr = get_header_attr(attr, counter);
+	if (h_attr == NULL)
+		die("nomem\n");
 
 	if (!file_new) {
 		if (memcmp(&h_attr->attr, attr, sizeof(*attr))) {
@@ -464,7 +327,10 @@ try_again:
 		exit(-1);
 	}
 
-	perf_header_attr__add_id(h_attr, read_data.id);
+	if (perf_header_attr__add_id(h_attr, read_data.id) < 0) {
+		pr_warning("Not enough memory to add id\n");
+		exit(-1);
+	}
 
 	assert(fd[nr_cpu][counter] >= 0);
 	fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK);
@@ -525,7 +391,7 @@ static void atexit_header(void)
 {
 	header->data_size += bytes_written;
 
-	perf_header__write(header, output);
+	perf_header__write(header, output, true);
 }
 
 static int __cmd_record(int argc, const char **argv)
@@ -573,12 +439,17 @@ static int __cmd_record(int argc, const char **argv)
 	else
 		header = perf_header__new();
 
+	if (header == NULL) {
+		pr_err("Not enough memory for reading perf file header\n");
+		return -1;
+	}
+
 	if (raw_samples) {
-		perf_header__feat_trace_info(header);
+		perf_header__set_feat(header, HEADER_TRACE_INFO);
 	} else {
 		for (i = 0; i < nr_counters; i++) {
 			if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
-				perf_header__feat_trace_info(header);
+				perf_header__set_feat(header, HEADER_TRACE_INFO);
 				break;
 			}
 		}
@@ -602,24 +473,32 @@ static int __cmd_record(int argc, const char **argv)
 	}
 
 	if (file_new)
-		perf_header__write(header, output);
+		perf_header__write(header, output, false);
 
-	if (!system_wide) {
-		pid_t tgid = pid_synthesize_comm_event(pid, 0);
-		pid_synthesize_mmap_samples(pid, tgid);
-	} else
-		synthesize_all();
+	if (!system_wide)
+		event__synthesize_thread(pid, process_synthesized_event);
+	else
+		event__synthesize_threads(process_synthesized_event);
 
 	if (target_pid == -1 && argc) {
 		pid = fork();
 		if (pid < 0)
-			perror("failed to fork");
+			die("failed to fork");
 
 		if (!pid) {
 			if (execvp(argv[0], (char **)argv)) {
 				perror(argv[0]);
 				exit(-1);
 			}
+		} else {
+			/*
+			 * Wait a bit for the execv'ed child to appear
+			 * and be updated in /proc
+			 * FIXME: Do you know a less heuristical solution?
+			 */
+			usleep(1000);
+			event__synthesize_thread(pid,
+						 process_synthesized_event);
 		}
 
 		child_pid = pid;
@@ -729,6 +608,8 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 {
 	int counter;
 
+	symbol__init(0);
+
 	argc = parse_options(argc, argv, options, record_usage,
 		PARSE_OPT_STOP_AT_NON_OPTION);
 	if (!argc && target_pid == -1 && !system_wide)
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index b3d814b..1a806d5 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -455,7 +455,7 @@ got_map:
 	dump_printf(" ...... map: %Lx -> %Lx\n", *ipp, ip);
 	*ipp  = ip;
 
-	return map ? map->dso->find_symbol(map->dso, ip) : NULL;
+	return map ? map__find_symbol(map, ip, NULL) : NULL;
 }
 
 static int call__match(struct symbol *sym)
@@ -751,7 +751,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct map *map = map__new(&event->mmap, cwd, cwdlen, 0, NULL);
+	struct map *map = map__new(&event->mmap, cwd, cwdlen);
 	struct thread *thread = threads__findnew(event->mmap.pid);
 
 	dump_printf("%p [%p]: PERF_RECORD_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
@@ -1093,7 +1093,7 @@ static void setup_list(struct strlist **list, const char *list_str,
 
 int cmd_report(int argc, const char **argv, const char *prefix __used)
 {
-	symbol__init();
+	symbol__init(0);
 
 	argc = parse_options(argc, argv, options, report_usage, 0);
 
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 9a48d96..df44b75 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1937,7 +1937,7 @@ static int __cmd_record(int argc, const char **argv)
 
 int cmd_sched(int argc, const char **argv, const char *prefix __used)
 {
-	symbol__init();
+	symbol__init(0);
 
 	argc = parse_options(argc, argv, sched_options, sched_usage,
 			     PARSE_OPT_STOP_AT_NON_OPTION);
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index c6df377..c70d720 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -357,7 +357,8 @@ static void abs_printout(int counter, double avg)
 			ratio = avg / total;
 
 		fprintf(stderr, " # %10.3f IPC  ", ratio);
-	} else if (MATCH_EVENT(HARDWARE, HW_BRANCH_MISSES, counter)) {
+	} else if (MATCH_EVENT(HARDWARE, HW_BRANCH_MISSES, counter) &&
+			runtime_branches_stats.n != 0) {
 		total = avg_stats(&runtime_branches_stats);
 
 		if (total)
@@ -365,7 +366,7 @@ static void abs_printout(int counter, double avg)
 
 		fprintf(stderr, " # %10.3f %%    ", ratio);
 
-	} else {
+	} else if (runtime_nsecs_stats.n != 0) {
 		total = avg_stats(&runtime_nsecs_stats);
 
 		if (total)
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 0a2f222..665877e 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -1266,7 +1266,7 @@ static const struct option options[] = {
 
 int cmd_timechart(int argc, const char **argv, const char *prefix __used)
 {
-	symbol__init();
+	symbol__init(0);
 
 	page_size = getpagesize();
 
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index 4a9fe22..89b7f68 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -60,7 +60,7 @@ static int			system_wide			=      0;
 static int			default_interval		=      0;
 
 static int			count_filter			=      5;
-static int			print_entries			=     15;
+static int			print_entries;
 
 static int			target_pid			=     -1;
 static int			inherit				=      0;
@@ -76,6 +76,9 @@ static int			delay_secs			=      2;
 static int			zero                            =      0;
 static int			dump_symtab                     =      0;
 
+static bool			hide_kernel_symbols		=  false;
+static bool			hide_user_symbols		=  false;
+
 /*
  * Source
  */
@@ -104,6 +107,7 @@ struct sym_entry {
 	unsigned long		snap_count;
 	double			weight;
 	int			skip;
+	u8			origin;
 	struct map		*map;
 	struct source_line	*source;
 	struct source_line	*lines;
@@ -115,6 +119,36 @@ struct sym_entry {
  * Source functions
  */
 
+/* most GUI terminals set LINES (although some don't export it) */
+static int term_rows(void)
+{
+	char *lines_string = getenv("LINES");
+	int n_lines;
+
+	if (lines_string && (n_lines = atoi(lines_string)) > 0)
+		return n_lines;
+#ifdef TIOCGWINSZ
+	else {
+		struct winsize ws;
+		if (!ioctl(1, TIOCGWINSZ, &ws) && ws.ws_row)
+			return ws.ws_row;
+	}
+#endif
+	return 25;
+}
+
+static void update_print_entries(void)
+{
+	print_entries = term_rows();
+	if (print_entries > 9)
+		print_entries -= 9;
+}
+
+static void sig_winch_handler(int sig __used)
+{
+	update_print_entries();
+}
+
 static void parse_source(struct sym_entry *syme)
 {
 	struct symbol *sym;
@@ -318,7 +352,7 @@ static void show_details(struct sym_entry *syme)
 }
 
 /*
- * Symbols will be added here in record_ip and will get out
+ * Symbols will be added here in event__process_sample and will get out
  * after decayed.
  */
 static LIST_HEAD(active_symbols);
@@ -400,6 +434,13 @@ static void print_sym_table(void)
 	list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
 		syme->snap_count = syme->count[snap];
 		if (syme->snap_count != 0) {
+			if ((hide_user_symbols &&
+			     syme->origin == PERF_RECORD_MISC_USER) ||
+			    (hide_kernel_symbols &&
+			     syme->origin == PERF_RECORD_MISC_KERNEL)) {
+				list_remove_active_sym(syme);
+				continue;
+			}
 			syme->weight = sym_weight(syme);
 			rb_insert_active_sym(&tmp, syme);
 			sum_ksamples += syme->snap_count;
@@ -459,18 +500,18 @@ static void print_sym_table(void)
 	}
 
 	if (nr_counters == 1)
-		printf("             samples    pcnt");
+		printf("             samples  pcnt");
 	else
-		printf("   weight    samples    pcnt");
+		printf("   weight    samples  pcnt");
 
 	if (verbose)
 		printf("         RIP       ");
-	printf("   kernel function\n");
-	printf("   %s    _______   _____",
+	printf(" function                                 DSO\n");
+	printf("   %s    _______ _____",
 	       nr_counters == 1 ? "      " : "______");
 	if (verbose)
-		printf("   ________________");
-	printf("   _______________\n\n");
+		printf(" ________________");
+	printf(" ________________________________ ________________\n\n");
 
 	for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) {
 		struct symbol *sym;
@@ -486,16 +527,15 @@ static void print_sym_table(void)
 					 sum_ksamples));
 
 		if (nr_counters == 1 || !display_weighted)
-			printf("%20.2f - ", syme->weight);
+			printf("%20.2f ", syme->weight);
 		else
-			printf("%9.1f %10ld - ", syme->weight, syme->snap_count);
+			printf("%9.1f %10ld ", syme->weight, syme->snap_count);
 
 		percent_color_fprintf(stdout, "%4.1f%%", pcnt);
 		if (verbose)
-			printf(" - %016llx", sym->start);
-		printf(" : %s", sym->name);
-		if (syme->map->dso->name[0] == '[')
-			printf(" \t%s", syme->map->dso->name);
+			printf(" %016llx", sym->start);
+		printf(" %-32s", sym->name);
+		printf(" %s", syme->map->dso->short_name);
 		printf("\n");
 	}
 }
@@ -608,6 +648,12 @@ static void print_mapped_keys(void)
 	if (nr_counters > 1)
 		fprintf(stdout, "\t[w]     toggle display weighted/count[E]r. \t(%d)\n", display_weighted ? 1 : 0);
 
+	fprintf(stdout,
+		"\t[K]     hide kernel_symbols symbols.             \t(%s)\n",
+		hide_kernel_symbols ? "yes" : "no");
+	fprintf(stdout,
+		"\t[U]     hide user symbols.               \t(%s)\n",
+		hide_user_symbols ? "yes" : "no");
 	fprintf(stdout, "\t[z]     toggle sample zeroing.             \t(%d)\n", zero ? 1 : 0);
 	fprintf(stdout, "\t[qQ]    quit.\n");
 }
@@ -621,6 +667,8 @@ static int key_mapped(int c)
 		case 'z':
 		case 'q':
 		case 'Q':
+		case 'K':
+		case 'U':
 			return 1;
 		case 'E':
 		case 'w':
@@ -669,6 +717,11 @@ static void handle_keypress(int c)
 			break;
 		case 'e':
 			prompt_integer(&print_entries, "Enter display entries (lines)");
+			if (print_entries == 0) {
+				update_print_entries();
+				signal(SIGWINCH, sig_winch_handler);
+			} else
+				signal(SIGWINCH, SIG_DFL);
 			break;
 		case 'E':
 			if (nr_counters > 1) {
@@ -693,6 +746,9 @@ static void handle_keypress(int c)
 		case 'F':
 			prompt_percent(&sym_pcnt_filter, "Enter details display event filter (percent)");
 			break;
+		case 'K':
+			hide_kernel_symbols = !hide_kernel_symbols;
+			break;
 		case 'q':
 		case 'Q':
 			printf("exiting.\n");
@@ -712,6 +768,9 @@ static void handle_keypress(int c)
 				pthread_mutex_unlock(&syme->source_lock);
 			}
 			break;
+		case 'U':
+			hide_user_symbols = !hide_user_symbols;
+			break;
 		case 'w':
 			display_weighted = ~display_weighted;
 			break;
@@ -790,7 +849,7 @@ static int symbol_filter(struct map *map, struct symbol *sym)
 	    strstr(name, "_text_end"))
 		return 1;
 
-	syme = dso__sym_priv(map->dso, sym);
+	syme = symbol__priv(sym);
 	syme->map = map;
 	pthread_mutex_init(&syme->source_lock, NULL);
 	if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter))
@@ -808,8 +867,7 @@ static int symbol_filter(struct map *map, struct symbol *sym)
 
 static int parse_symbols(void)
 {
-	if (dsos__load_kernel(vmlinux_name, sizeof(struct sym_entry),
-			      symbol_filter, 1) <= 0)
+	if (dsos__load_kernel(vmlinux_name, symbol_filter, 1) <= 0)
 		return -1;
 
 	if (dump_symtab)
@@ -818,41 +876,104 @@ static int parse_symbols(void)
 	return 0;
 }
 
-/*
- * Binary search in the histogram table and record the hit:
- */
-static void record_ip(u64 ip, int counter)
+static void event__process_sample(const event_t *self, int counter)
 {
+	u64 ip = self->ip.ip;
 	struct map *map;
-	struct symbol *sym = kernel_maps__find_symbol(ip, &map);
-
-	if (sym != NULL) {
-		struct sym_entry *syme = dso__sym_priv(map->dso, sym);
-
-		if (!syme->skip) {
-			syme->count[counter]++;
-			record_precise_ip(syme, counter, ip);
-			pthread_mutex_lock(&active_symbols_lock);
-			if (list_empty(&syme->node) || !syme->node.next)
-				__list_insert_active_sym(syme);
-			pthread_mutex_unlock(&active_symbols_lock);
+	struct sym_entry *syme;
+	struct symbol *sym;
+	u8 origin = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+
+	switch (origin) {
+	case PERF_RECORD_MISC_USER: {
+		struct thread *thread;
+
+		if (hide_user_symbols)
+			return;
+
+		thread = threads__findnew(self->ip.pid);
+		if (thread == NULL)
 			return;
+
+		map = thread__find_map(thread, ip);
+		if (map != NULL) {
+			ip = map->map_ip(map, ip);
+			sym = map__find_symbol(map, ip, symbol_filter);
+			if (sym == NULL)
+				return;
+			userspace_samples++;
+			break;
 		}
 	}
+		/*
+		 * 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).
+		 */
+		if ((long long)ip >= 0)
+			return;
+		/* Fall thru */
+	case PERF_RECORD_MISC_KERNEL:
+		if (hide_kernel_symbols)
+			return;
+
+		sym = kernel_maps__find_symbol(ip, &map);
+		if (sym == NULL)
+			return;
+		break;
+	default:
+		return;
+	}
 
-	samples--;
+	syme = symbol__priv(sym);
+
+	if (!syme->skip) {
+		syme->count[counter]++;
+		syme->origin = origin;
+		record_precise_ip(syme, counter, ip);
+		pthread_mutex_lock(&active_symbols_lock);
+		if (list_empty(&syme->node) || !syme->node.next)
+			__list_insert_active_sym(syme);
+		pthread_mutex_unlock(&active_symbols_lock);
+		++samples;
+		return;
+	}
 }
 
-static void process_event(u64 ip, int counter, int user)
+static void event__process_mmap(event_t *self)
 {
-	samples++;
+	struct thread *thread = threads__findnew(self->mmap.pid);
 
-	if (user) {
-		userspace_samples++;
-		return;
+	if (thread != NULL) {
+		struct map *map = map__new(&self->mmap, NULL, 0);
+		if (map != NULL)
+			thread__insert_map(thread, map);
 	}
+}
 
-	record_ip(ip, counter);
+static void event__process_comm(event_t *self)
+{
+	struct thread *thread = threads__findnew(self->comm.pid);
+
+	if (thread != NULL)
+		thread__set_comm(thread, self->comm.comm);
+}
+
+static int event__process(event_t *event)
+{
+	switch (event->header.type) {
+	case PERF_RECORD_COMM:
+		event__process_comm(event);
+		break;
+	case PERF_RECORD_MMAP:
+		event__process_mmap(event);
+		break;
+	default:
+		break;
+	}
+
+	return 0;
 }
 
 struct mmap_data {
@@ -925,13 +1046,11 @@ static void mmap_read_counter(struct mmap_data *md)
 			event = &event_copy;
 		}
 
+		if (event->header.type == PERF_RECORD_SAMPLE)
+			event__process_sample(event, md->counter);
+		else
+			event__process(event);
 		old += size;
-
-		if (event->header.type == PERF_RECORD_SAMPLE) {
-			int user =
-	(event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK) == PERF_RECORD_MISC_USER;
-			process_event(event->ip.ip, md->counter, user);
-		}
 	}
 
 	md->prev = old;
@@ -973,6 +1092,7 @@ static void start_counter(int i, int counter)
 	}
 
 	attr->inherit		= (cpu < 0) && inherit;
+	attr->mmap		= 1;
 
 try_again:
 	fd[i][counter] = sys_perf_event_open(attr, target_pid, cpu, group_fd, 0);
@@ -980,7 +1100,7 @@ try_again:
 	if (fd[i][counter] < 0) {
 		int err = errno;
 
-		if (err == EPERM)
+		if (err == EPERM || err == EACCES)
 			die("No permission - are you root?\n");
 		/*
 		 * If it's cycles then fall back to hrtimer
@@ -1031,6 +1151,11 @@ static int __cmd_top(void)
 	int i, counter;
 	int ret;
 
+	if (target_pid != -1)
+		event__synthesize_thread(target_pid, event__process);
+	else
+		event__synthesize_threads(event__process);
+
 	for (i = 0; i < nr_cpus; i++) {
 		group_fd = -1;
 		for (counter = 0; counter < nr_counters; counter++)
@@ -1087,6 +1212,8 @@ static const struct option options[] = {
 	OPT_INTEGER('C', "CPU", &profile_cpu,
 		    "CPU to profile on"),
 	OPT_STRING('k', "vmlinux", &vmlinux_name, "file", "vmlinux pathname"),
+	OPT_BOOLEAN('K', "hide_kernel_symbols", &hide_kernel_symbols,
+		    "hide kernel symbols"),
 	OPT_INTEGER('m', "mmap-pages", &mmap_pages,
 		    "number of mmap data pages"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
@@ -1109,6 +1236,8 @@ static const struct option options[] = {
 		    "profile at this frequency"),
 	OPT_INTEGER('E', "entries", &print_entries,
 		    "display this many functions"),
+	OPT_BOOLEAN('U', "hide_user_symbols", &hide_user_symbols,
+		    "hide user symbols"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
 	OPT_END()
@@ -1118,7 +1247,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 {
 	int counter;
 
-	symbol__init();
+	symbol__init(sizeof(struct sym_entry));
 
 	page_size = sysconf(_SC_PAGE_SIZE);
 
@@ -1172,5 +1301,10 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 	if (target_pid != -1 || profile_cpu != -1)
 		nr_cpus = 1;
 
+	if (print_entries == 0) {
+		update_print_entries();
+		signal(SIGWINCH, sig_winch_handler);
+	}
+
 	return __cmd_top();
 }
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index e566bbe..d042d65 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -151,7 +151,7 @@ static const struct option options[] = {
 
 int cmd_trace(int argc, const char **argv, const char *prefix __used)
 {
-	symbol__init();
+	symbol__init(0);
 
 	argc = parse_options(argc, argv, options, annotate_usage, 0);
 	if (argc) {
diff --git a/tools/perf/builtin.h b/tools/perf/builtin.h
index ad5f0f4..9b02d85 100644
--- a/tools/perf/builtin.h
+++ b/tools/perf/builtin.h
@@ -15,6 +15,8 @@ extern int read_line_with_nul(char *buf, int size, FILE *file);
 extern int check_pager_config(const char *cmd);
 
 extern int cmd_annotate(int argc, const char **argv, const char *prefix);
+extern int cmd_bench(int argc, const char **argv, const char *prefix);
+extern int cmd_buildid_list(int argc, const char **argv, const char *prefix);
 extern int cmd_help(int argc, const char **argv, const char *prefix);
 extern int cmd_sched(int argc, const char **argv, const char *prefix);
 extern int cmd_list(int argc, const char **argv, const char *prefix);
diff --git a/tools/perf/command-list.txt b/tools/perf/command-list.txt
index 6475db4..d3a6e18 100644
--- a/tools/perf/command-list.txt
+++ b/tools/perf/command-list.txt
@@ -3,6 +3,8 @@
 # command name			category [deprecated] [common]
 #
 perf-annotate			mainporcelain common
+perf-bench			mainporcelain common
+perf-buildid-list		mainporcelain common
 perf-list			mainporcelain common
 perf-sched			mainporcelain common
 perf-record			mainporcelain common
diff --git a/tools/perf/design.txt b/tools/perf/design.txt
index fdd42a8..f000c30 100644
--- a/tools/perf/design.txt
+++ b/tools/perf/design.txt
@@ -137,6 +137,8 @@ enum sw_event_ids {
 	PERF_COUNT_SW_CPU_MIGRATIONS	= 4,
 	PERF_COUNT_SW_PAGE_FAULTS_MIN	= 5,
 	PERF_COUNT_SW_PAGE_FAULTS_MAJ	= 6,
+	PERF_COUNT_SW_ALIGNMENT_FAULTS	= 7,
+	PERF_COUNT_SW_EMULATION_FAULTS	= 8,
 };
 
 Counters of the type PERF_TYPE_TRACEPOINT are available when the ftrace event
diff --git a/tools/perf/perf.c b/tools/perf/perf.c
index 9cafe54..89b82ac 100644
--- a/tools/perf/perf.c
+++ b/tools/perf/perf.c
@@ -14,6 +14,7 @@
 #include "util/run-command.h"
 #include "util/parse-events.h"
 #include "util/string.h"
+#include "util/debugfs.h"
 
 const char perf_usage_string[] =
 	"perf [--version] [--help] COMMAND [ARGS]";
@@ -286,8 +287,10 @@ static void handle_internal_command(int argc, const char **argv)
 	static struct cmd_struct commands[] = {
 		{ "help", cmd_help, 0 },
 		{ "list", cmd_list, 0 },
+		{ "buildid-list", cmd_buildid_list, 0 },
 		{ "record", cmd_record, 0 },
 		{ "report", cmd_report, 0 },
+		{ "bench", cmd_bench, 0 },
 		{ "stat", cmd_stat, 0 },
 		{ "timechart", cmd_timechart, 0 },
 		{ "top", cmd_top, 0 },
@@ -383,45 +386,12 @@ static int run_argv(int *argcp, const char ***argv)
 /* mini /proc/mounts parser: searching for "^blah /mount/point debugfs" */
 static void get_debugfs_mntpt(void)
 {
-	FILE *file;
-	char fs_type[100];
-	char debugfs[MAXPATHLEN];
+	const char *path = debugfs_find_mountpoint();
 
-	/*
-	 * try the standard location
-	 */
-	if (valid_debugfs_mount("/sys/kernel/debug/") == 0) {
-		strcpy(debugfs_mntpt, "/sys/kernel/debug/");
-		return;
-	}
-
-	/*
-	 * try the sane location
-	 */
-	if (valid_debugfs_mount("/debug/") == 0) {
-		strcpy(debugfs_mntpt, "/debug/");
-		return;
-	}
-
-	/*
-	 * give up and parse /proc/mounts
-	 */
-	file = fopen("/proc/mounts", "r");
-	if (file == NULL)
-		return;
-
-	while (fscanf(file, "%*s %"
-		      STR(MAXPATHLEN)
-		      "s %99s %*s %*d %*d\n",
-		      debugfs, fs_type) == 2) {
-		if (strcmp(fs_type, "debugfs") == 0)
-			break;
-	}
-	fclose(file);
-	if (strcmp(fs_type, "debugfs") == 0) {
-		strncpy(debugfs_mntpt, debugfs, MAXPATHLEN);
-		debugfs_mntpt[MAXPATHLEN - 1] = '\0';
-	}
+	if (path)
+		strncpy(debugfs_mntpt, path, sizeof(debugfs_mntpt));
+	else
+		debugfs_mntpt[0] = '\0';
 }
 
 int main(int argc, const char **argv)
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 8cc4623..216bdb2 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -47,6 +47,12 @@
 #define cpu_relax()	asm volatile("":::"memory")
 #endif
 
+#ifdef __alpha__
+#include "../../arch/alpha/include/asm/unistd.h"
+#define rmb()		asm volatile("mb" ::: "memory")
+#define cpu_relax()	asm volatile("" ::: "memory")
+#endif
+
 #include <time.h>
 #include <unistd.h>
 #include <sys/types.h>
diff --git a/tools/perf/util/data_map.c b/tools/perf/util/data_map.c
index 18accb8..14cb846 100644
--- a/tools/perf/util/data_map.c
+++ b/tools/perf/util/data_map.c
@@ -70,6 +70,35 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 	}
 }
 
+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,
@@ -130,7 +159,7 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
 		if (curr_handler->sample_type_check(sample_type) < 0)
 			exit(-1);
 
-	if (load_kernel(0, NULL) < 0) {
+	if (load_kernel(NULL) < 0) {
 		perror("failed to load kernel symbols");
 		return EXIT_FAILURE;
 	}
diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h
index 716d105..ae036ec 100644
--- a/tools/perf/util/data_map.h
+++ b/tools/perf/util/data_map.h
@@ -27,5 +27,6 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
 			    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.h b/tools/perf/util/debug.h
index e8b18a1..c6c24c5 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -2,6 +2,8 @@
 #ifndef __PERF_DEBUG_H
 #define __PERF_DEBUG_H
 
+#include "event.h"
+
 extern int verbose;
 extern int dump_trace;
 
diff --git a/tools/perf/util/debugfs.c b/tools/perf/util/debugfs.c
new file mode 100644
index 0000000..06b73ee
--- /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 0000000..3cd14f9
--- /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 0000000..1dae7e3
--- /dev/null
+++ b/tools/perf/util/event.c
@@ -0,0 +1,177 @@
+#include <linux/types.h>
+#include "event.h"
+#include "debug.h"
+#include "string.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);
+}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index d972b4b..1f771ce 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -61,6 +61,20 @@ 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[];
+};
+
+struct build_id_list {
+	struct build_id_event	event;
+	struct list_head	list;
+	const char		*dso_name;
+	int			len;
+};
 
 typedef union event_union {
 	struct perf_event_header	header;
@@ -105,10 +119,15 @@ struct symbol;
 
 typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym);
 
-struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen,
-		     unsigned int sym_priv_size, symbol_filter_t filter);
+void map__init(struct map *self, u64 start, u64 end, u64 pgoff,
+	       struct dso *dso);
+struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen);
 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 ip, symbol_filter_t filter);
+
+int event__synthesize_thread(pid_t pid, int (*process)(event_t *event));
+void event__synthesize_threads(int (*process)(event_t *event));
 
 #endif /* __PERF_RECORD_H */
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 7d26659..b01a953 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -2,11 +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:
@@ -15,32 +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;
 }
 
 /*
@@ -50,34 +65,41 @@ struct perf_header *perf_header__new(void)
 {
 	struct perf_header *self = calloc(sizeof(*self), 1);
 
-	if (!self)
-		die("nomem");
-
-	self->size = 1;
-	self->attr = malloc(sizeof(void *));
+	if (self != NULL) {
+		self->size = 1;
+		self->attr = malloc(sizeof(void *));
 
-	if (!self->attr)
-		die("nomem");
+		if (self->attr == NULL) {
+			free(self);
+			self = NULL;
+		}
+	}
 
 	return 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");
+		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;
+	return 0;
 }
 
 #define MAX_EVENT_NAME 64
@@ -124,71 +146,110 @@ 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;
-	DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
-};
+void perf_header__set_feat(struct perf_header *self, int feat)
+{
+	set_bit(feat, self->adds_features);
+}
 
-void perf_header__feat_trace_info(struct perf_header *header)
+bool perf_header__has_feat(const struct perf_header *self, int feat)
 {
-	set_bit(HEADER_TRACE_INFO, header->adds_features);
+	return test_bit(feat, self->adds_features);
 }
 
-static void do_write(int fd, void *buf, size_t size)
+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 -1;
 
 		size -= ret;
 		buf += ret;
 	}
+
+	return 0;
+}
+
+static int write_buildid_table(int fd, struct list_head *id_head)
+{
+	struct build_id_list *iter, *next;
+
+	list_for_each_entry_safe(iter, next, id_head, list) {
+		struct build_id_event *b = &iter->event;
+
+		if (do_write(fd, b, sizeof(*b)) < 0 ||
+		    do_write(fd, iter->dso_name, iter->len) < 0)
+			return -1;
+		list_del(&iter->list);
+		free(iter);
+	}
+
+	return 0;
 }
 
-static void perf_header__adds_write(struct perf_header *self, int fd)
+static void
+perf_header__adds_write(struct perf_header *self, int fd)
 {
-	struct perf_file_section trace_sec;
-	u64 cur_offset = lseek(fd, 0, SEEK_CUR);
-	unsigned long *feat_mask = self->adds_features;
+	LIST_HEAD(id_list);
+	int nr_sections;
+	struct perf_file_section *feat_sec;
+	int sec_size;
+	u64 sec_start;
+	int idx = 0;
+
+	if (fetch_build_id_table(&id_list))
+		perf_header__set_feat(self, HEADER_BUILD_ID);
+
+	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	if (!nr_sections)
+		return;
+
+	feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+	if (!feat_sec)
+		die("No memory");
+
+	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++];
 
-	if (test_bit(HEADER_TRACE_INFO, feat_mask)) {
 		/* Write trace info */
-		trace_sec.offset = lseek(fd, sizeof(trace_sec), SEEK_CUR);
+		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;
-
-		/* Write trace info headers */
-		lseek(fd, cur_offset, SEEK_SET);
-		do_write(fd, &trace_sec, sizeof(trace_sec));
-
-		/*
-		 * Update cur_offset. So that other (future)
-		 * features can set their own infos in this place. But if we are
-		 * the only feature, at least that seeks to the place the data
-		 * should begin.
-		 */
-		cur_offset = lseek(fd, trace_sec.offset + trace_sec.size, SEEK_SET);
+		trace_sec->size = lseek(fd, 0, SEEK_CUR) - trace_sec->offset;
 	}
-};
 
-void perf_header__write(struct perf_header *self, int fd)
+
+	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);
+		if (write_buildid_table(fd, &id_list) < 0)
+			die("failed to write buildid table");
+		buildid_sec->size = lseek(fd, 0, SEEK_CUR) - buildid_sec->offset;
+	}
+
+	lseek(fd, sec_start, SEEK_SET);
+	if (do_write(fd, feat_sec, sec_size) < 0)
+		die("failed to write feature section");
+	free(feat_sec);
+}
+
+void perf_header__write(struct perf_header *self, int fd, bool at_exit)
 {
 	struct perf_file_header f_header;
 	struct perf_file_attr   f_attr;
@@ -202,7 +263,8 @@ 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));
+		if (do_write(fd, attr->id, attr->ids * sizeof(u64)) < 0)
+			die("failed to write perf header");
 	}
 
 
@@ -218,18 +280,21 @@ void perf_header__write(struct perf_header *self, int fd)
 				.size   = attr->ids * sizeof(u64),
 			}
 		};
-		do_write(fd, &f_attr, sizeof(f_attr));
+		if (do_write(fd, &f_attr, sizeof(f_attr)) < 0)
+			die("failed to write perf header attribute");
 	}
 
 	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);
-
-	perf_header__adds_write(self, fd);
+		if (do_write(fd, events, self->event_size) < 0)
+			die("failed to write perf header events");
 
 	self->data_offset = lseek(fd, 0, SEEK_CUR);
 
+	if (at_exit)
+		perf_header__adds_write(self, fd);
+
 	f_header = (struct perf_file_header){
 		.magic	   = PERF_MAGIC,
 		.size	   = sizeof(f_header),
@@ -251,7 +316,8 @@ void perf_header__write(struct perf_header *self, int fd)
 	memcpy(&f_header.adds_features, &self->adds_features, sizeof(self->adds_features));
 
 	lseek(fd, 0, SEEK_SET);
-	do_write(fd, &f_header, sizeof(f_header));
+	if (do_write(fd, &f_header, sizeof(f_header)) < 0)
+		die("failed to write perf header");
 	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
 
 	self->frozen = 1;
@@ -272,43 +338,112 @@ static void do_read(int fd, void *buf, size_t size)
 	}
 }
 
-static void perf_header__adds_read(struct perf_header *self, int fd)
+int perf_header__process_sections(struct perf_header *self, int fd,
+				  int (*process)(struct perf_file_section *self,
+						 int feat, int fd))
 {
-	const unsigned long *feat_mask = self->adds_features;
+	struct perf_file_section *feat_sec;
+	int nr_sections;
+	int sec_size;
+	int idx = 0;
+	int err = 0, feat = 1;
 
-	if (test_bit(HEADER_TRACE_INFO, feat_mask)) {
-		struct perf_file_section trace_sec;
+	nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+	if (!nr_sections)
+		return 0;
 
-		do_read(fd, &trace_sec, sizeof(trace_sec));
-		lseek(fd, trace_sec.offset, SEEK_SET);
-		trace_report(fd);
-		lseek(fd, trace_sec.offset + trace_sec.size, SEEK_SET);
+	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;
+}
+
 struct perf_header *perf_header__read(int fd)
 {
 	struct perf_header	*self = perf_header__new();
 	struct perf_file_header f_header;
 	struct perf_file_attr	f_attr;
 	u64			f_id;
-
 	int nr_attrs, nr_ids, i, j;
 
-	lseek(fd, 0, SEEK_SET);
-	do_read(fd, &f_header, sizeof(f_header));
+	if (self == NULL)
+		die("nomem");
 
-	if (f_header.magic	!= PERF_MAGIC		||
-	    f_header.attr_size	!= sizeof(f_attr))
+	if (perf_file_header__read(&f_header, self, fd) < 0)
 		die("incompatible file format");
 
-	if (f_header.size != sizeof(f_header)) {
-		/* Support the previous format */
-		if (f_header.size == offsetof(typeof(f_header), adds_features))
-			bitmap_zero(f_header.adds_features, HEADER_FEAT_BITS);
-		else
-			die("incompatible file format");
-	}
 	nr_attrs = f_header.attrs.size / sizeof(f_attr);
 	lseek(fd, f_header.attrs.offset, SEEK_SET);
 
@@ -320,6 +455,8 @@ 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)
+			 die("nomem");
 
 		nr_ids = f_attr.ids.size / sizeof(u64);
 		lseek(fd, f_attr.ids.offset, SEEK_SET);
@@ -327,9 +464,12 @@ 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)
+				die("nomem");
 		}
-		perf_header__add_attr(self, attr);
+		if (perf_header__add_attr(self, attr) < 0)
+			 die("nomem");
+
 		lseek(fd, tmp, SEEK_SET);
 	}
 
@@ -342,15 +482,7 @@ struct perf_header *perf_header__read(int fd)
 		event_count =  f_header.event_types.size / sizeof(struct perf_trace_event_type);
 	}
 
-	memcpy(&self->adds_features, &f_header.adds_features, sizeof(f_header.adds_features));
-
-	perf_header__adds_read(self, fd);
-
-	self->event_offset = f_header.event_types.offset;
-	self->event_size   = f_header.event_types.size;
-
-	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);
 
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
index 2ea9dfb..f46a94e 100644
--- a/tools/perf/util/header.h
+++ b/tools/perf/util/header.h
@@ -3,6 +3,7 @@
 
 #include "../../../include/linux/perf_event.h"
 #include <sys/types.h>
+#include <stdbool.h>
 #include "types.h"
 
 #include <linux/bitmap.h>
@@ -14,10 +15,34 @@ struct perf_header_attr {
 	off_t id_offset;
 };
 
-#define HEADER_TRACE_INFO			1
+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;
@@ -31,24 +56,29 @@ struct perf_header {
 };
 
 struct perf_header *perf_header__read(int fd);
-void perf_header__write(struct perf_header *self, int fd);
+void perf_header__write(struct perf_header *self, int fd, bool at_exit);
 
-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);
 
 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__feat_trace_info(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);
 
 struct perf_header *perf_header__new(void);
 
+int perf_header__process_sections(struct perf_header *self, int fd,
+				  int (*process)(struct perf_file_section *self,
+						 int feat, int fd));
+
 #endif /* __PERF_HEADER_H */
diff --git a/tools/perf/util/include/linux/bitmap.h b/tools/perf/util/include/linux/bitmap.h
index 821c103..9450763 100644
--- a/tools/perf/util/include/linux/bitmap.h
+++ b/tools/perf/util/include/linux/bitmap.h
@@ -1,2 +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/ctype.h b/tools/perf/util/include/linux/ctype.h
index bae5783..a53d4ee 100644
--- a/tools/perf/util/include/linux/ctype.h
+++ b/tools/perf/util/include/linux/ctype.h
@@ -1 +1 @@
-#include "../../../../include/linux/ctype.h"
+#include "../util.h"
diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index c1c5568..94ca950 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -20,16 +20,27 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
 	return n;
 }
 
-struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen,
-		     unsigned int sym_priv_size, symbol_filter_t filter)
+void map__init(struct map *self, u64 start, u64 end, u64 pgoff,
+	       struct dso *dso)
+{
+	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, 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;
-		bool new_dso;
 
 		if (cwd) {
 			int n = strcommon(filename, cwd, cwdlen);
@@ -48,33 +59,15 @@ struct map *map__new(struct mmap_event *event, 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, sym_priv_size, &new_dso);
-		if (self->dso == NULL)
+		dso = dsos__findnew(filename);
+		if (dso == NULL)
 			goto out_delete;
 
-		if (new_dso) {
-			int nr = dso__load(self->dso, self, filter);
-
-			if (nr < 0)
-				pr_warning("Failed to open %s, continuing "
-					   "without symbols\n",
-					   self->dso->long_name);
-			else if (nr == 0)
-				pr_warning("No symbols found in %s, maybe "
-					   "install a debug package?\n",
-					   self->dso->long_name);
-		}
+		map__init(self, event->start, event->start + event->len,
+			  event->pgoff, dso);
 
 		if (self->dso == vdso || anon)
 			self->map_ip = self->unmap_ip = identity__map_ip;
-		else {
-			self->map_ip = map__map_ip;
-			self->unmap_ip = map__unmap_ip;
-		}
 	}
 	return self;
 out_delete:
@@ -82,6 +75,47 @@ out_delete:
 	return NULL;
 }
 
+#define DSO__DELETED "(deleted)"
+
+struct symbol *
+map__find_symbol(struct map *self, u64 ip, symbol_filter_t filter)
+{
+	if (!self->dso->loaded) {
+		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, ip);
+}
+
 struct map *map__clone(struct map *self)
 {
 	struct map *map = malloc(sizeof(*self));
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index b097570..0faf4f2 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -7,6 +7,7 @@
 #include "string.h"
 #include "cache.h"
 #include "header.h"
+#include "debugfs.h"
 
 int				nr_counters;
 
@@ -47,6 +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) \
@@ -75,6 +78,8 @@ static const char *sw_event_names[] = {
 	"CPU-migrations",
 	"minor-faults",
 	"major-faults",
+	"alignment-faults",
+	"emulation-faults",
 };
 
 #define MAX_ALIASES 8
@@ -149,16 +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)
 {
@@ -171,7 +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);
@@ -510,7 +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, ':');
@@ -678,6 +673,8 @@ parse_event_symbols(const char **str, struct perf_event_attr *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:
@@ -786,7 +783,7 @@ 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);
@@ -804,7 +801,7 @@ 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);
@@ -821,8 +818,8 @@ 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;
@@ -830,19 +827,19 @@ 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 */
@@ -850,17 +847,17 @@ 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");
 
 	print_tracepoint_events();
 
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index 04743d3..2270435 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -1,5 +1,7 @@
 #include <string.h>
+#include <stdlib.h>
 #include "string.h"
+#include "util.h"
 
 static int hex(char ch)
 {
@@ -43,3 +45,85 @@ char *strxfrchar(char *s, char from, char 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;
+}
diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h
index 2c84bf6..e50b07f 100644
--- a/tools/perf/util/string.h
+++ b/tools/perf/util/string.h
@@ -5,6 +5,7 @@
 
 int hex2u64(const char *ptr, u64 *val);
 char *strxfrchar(char *s, char from, char to);
+s64 perf_atoll(const char *str);
 
 #define _STR(x) #x
 #define STR(x) _STR(x)
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 8f0208c..1b77e81 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -26,6 +26,7 @@ static void dsos__add(struct dso *dso);
 static struct dso *dsos__find(const char *name);
 static struct map *map__new2(u64 start, struct dso *dso);
 static void kernel_maps__insert(struct map *map);
+unsigned int symbol__priv_size;
 
 static struct rb_root kernel_maps;
 
@@ -75,18 +76,17 @@ static void kernel_maps__fixup_end(void)
 	}
 }
 
-static struct symbol *symbol__new(u64 start, u64 len, const char *name,
-				  unsigned int priv_size)
+static struct symbol *symbol__new(u64 start, u64 len, const char *name)
 {
 	size_t namelen = strlen(name) + 1;
-	struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen);
-
+	struct symbol *self = calloc(1, (symbol__priv_size +
+					 sizeof(*self) + namelen));
 	if (!self)
 		return NULL;
 
-	if (priv_size) {
-		memset(self, 0, priv_size);
-		self = ((void *)self) + priv_size;
+	if (symbol__priv_size) {
+		memset(self, 0, symbol__priv_size);
+		self = ((void *)self) + symbol__priv_size;
 	}
 	self->start = start;
 	self->end   = len ? start + len - 1 : start;
@@ -98,9 +98,9 @@ static struct symbol *symbol__new(u64 start, u64 len, const char *name,
 	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)
@@ -109,7 +109,7 @@ static size_t symbol__fprintf(struct symbol *self, FILE *fp)
 		       self->start, self->end, self->name);
 }
 
-struct dso *dso__new(const char *name, unsigned int sym_priv_size)
+struct dso *dso__new(const char *name)
 {
 	struct dso *self = malloc(sizeof(*self) + strlen(name) + 1);
 
@@ -118,10 +118,11 @@ struct dso *dso__new(const char *name, unsigned int sym_priv_size)
 		self->long_name = self->name;
 		self->short_name = self->name;
 		self->syms = RB_ROOT;
-		self->sym_priv_size = sym_priv_size;
 		self->find_symbol = dso__find_symbol;
 		self->slen_calculated = 0;
 		self->origin = DSO__ORIG_NOT_FOUND;
+		self->loaded = 0;
+		self->has_build_id = 0;
 	}
 
 	return self;
@@ -136,7 +137,7 @@ static void dso__delete_symbols(struct dso *self)
 		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);
+		symbol__delete(pos);
 	}
 }
 
@@ -148,6 +149,12 @@ void dso__delete(struct dso *self)
 	free(self);
 }
 
+void dso__set_build_id(struct dso *self, void *build_id)
+{
+	memcpy(self->build_id, build_id, sizeof(self->build_id));
+	self->has_build_id = 1;
+}
+
 static void dso__insert_symbol(struct dso *self, struct symbol *sym)
 {
 	struct rb_node **p = &self->syms.rb_node;
@@ -190,11 +197,37 @@ struct symbol *dso__find_symbol(struct dso *self, u64 ip)
 	return NULL;
 }
 
-size_t dso__fprintf(struct dso *self, FILE *fp)
+int build_id__sprintf(u8 *self, int len, char *bf)
 {
-	size_t ret = fprintf(fp, "dso: %s\n", self->short_name);
+	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, FILE *fp)
+{
 	struct rb_node *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->syms); nd; nd = rb_next(nd)) {
 		struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
 		ret += symbol__fprintf(pos, fp);
@@ -250,12 +283,16 @@ static int kernel_maps__load_all_kallsyms(void)
 		/*
 		 * Will fix up the end later, when we have all symbols sorted.
 		 */
-		sym = symbol__new(start, 0, symbol_name,
-				  kernel_map->dso->sym_priv_size);
+		sym = symbol__new(start, 0, symbol_name);
 
 		if (sym == NULL)
 			goto out_delete_line;
 
+		/*
+		 * We will pass the symbols to the filter later, in
+		 * kernel_maps__split_kallsyms, when we have split the
+		 * maps per module
+		 */
 		dso__insert_symbol(kernel_map->dso, sym);
 	}
 
@@ -317,8 +354,7 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
 			snprintf(dso_name, sizeof(dso_name), "[kernel].%d",
 				 kernel_range++);
 
-			dso = dso__new(dso_name,
-				       kernel_map->dso->sym_priv_size);
+			dso = dso__new(dso_name);
 			if (dso == NULL)
 				return -1;
 
@@ -336,7 +372,7 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
 		if (filter && filter(map, pos)) {
 delete_symbol:
 			rb_erase(&pos->rb_node, &kernel_map->dso->syms);
-			symbol__delete(pos, kernel_map->dso->sym_priv_size);
+			symbol__delete(pos);
 		} else {
 			if (map != kernel_map) {
 				rb_erase(&pos->rb_node, &kernel_map->dso->syms);
@@ -417,14 +453,13 @@ static int dso__load_perf_map(struct dso *self, struct map *map,
 		if (len + 2 >= line_len)
 			continue;
 
-		sym = symbol__new(start, size, line + len,
-				  self->sym_priv_size);
+		sym = symbol__new(start, size, line + len);
 
 		if (sym == NULL)
 			goto out_delete_line;
 
 		if (filter && filter(map, sym))
-			symbol__delete(sym, self->sym_priv_size);
+			symbol__delete(sym);
 		else {
 			dso__insert_symbol(self, sym);
 			nr_syms++;
@@ -532,7 +567,8 @@ 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)
+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;
@@ -552,7 +588,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self)
 	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;
 
@@ -616,12 +652,16 @@ static int dso__synthesize_plt_symbols(struct  dso *self)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size);
+					sympltname);
 			if (!f)
 				goto out_elf_end;
 
-			dso__insert_symbol(self, f);
-			++nr;
+			if (filter && filter(map, f))
+				symbol__delete(f);
+			else {
+				dso__insert_symbol(self, f);
+				++nr;
+			}
 		}
 	} else if (shdr_rel_plt.sh_type == SHT_REL) {
 		GElf_Rel pos_mem, *pos;
@@ -634,12 +674,16 @@ static int dso__synthesize_plt_symbols(struct  dso *self)
 				 "%s@plt", elf_sym__name(&sym, symstrs));
 
 			f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-					sympltname, self->sym_priv_size);
+					sympltname);
 			if (!f)
 				goto out_elf_end;
 
-			dso__insert_symbol(self, f);
-			++nr;
+			if (filter && filter(map, f))
+				symbol__delete(f);
+			else {
+				dso__insert_symbol(self, f);
+				++nr;
+			}
 		}
 	}
 
@@ -676,7 +720,7 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
 	Elf *elf;
 	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) {
 		pr_err("%s: cannot read %s ELF file.\n", __func__, name);
 		goto out_close;
@@ -769,7 +813,7 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
 				if (kmodule)
 					start += map->start + shdr.sh_offset;
 
-				curr_dso = dso__new(dso_name, self->sym_priv_size);
+				curr_dso = dso__new(dso_name);
 				if (curr_dso == NULL)
 					goto out_elf_end;
 				curr_map = map__new2(start, curr_dso);
@@ -803,14 +847,13 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
 		if (demangled != NULL)
 			elf_name = demangled;
 new_symbol:
-		f = symbol__new(sym.st_value, sym.st_size, elf_name,
-				curr_dso->sym_priv_size);
+		f = symbol__new(sym.st_value, sym.st_size, elf_name);
 		free(demangled);
 		if (!f)
 			goto out_elf_end;
 
 		if (filter && filter(curr_map, f))
-			symbol__delete(f, curr_dso->sym_priv_size);
+			symbol__delete(f);
 		else {
 			dso__insert_symbol(curr_dso, f);
 			nr++;
@@ -829,27 +872,59 @@ out_close:
 	return err;
 }
 
-#define BUILD_ID_SIZE 128
+bool fetch_build_id_table(struct list_head *head)
+{
+	bool have_buildid = false;
+	struct dso *pos;
 
-static char *dso__read_build_id(struct dso *self)
+	list_for_each_entry(pos, &dsos, node) {
+		struct build_id_list *new;
+		struct build_id_event b;
+		size_t len;
+
+		if (filename__read_build_id(pos->long_name,
+					    &b.build_id,
+					    sizeof(b.build_id)) < 0)
+			continue;
+		have_buildid = true;
+		memset(&b.header, 0, sizeof(b.header));
+		len = strlen(pos->long_name) + 1;
+		len = ALIGN(len, 64);
+		b.header.size = sizeof(b) + len;
+
+		new = malloc(sizeof(*new));
+		if (!new)
+			die("No memory\n");
+
+		memcpy(&new->event, &b, sizeof(b));
+		new->dso_name = pos->long_name;
+		new->len = len;
+
+		list_add_tail(&new->list, head);
+	}
+
+	return have_buildid;
+}
+
+int filename__read_build_id(const char *filename, void *bf, size_t size)
 {
-	int i;
+	int fd, err = -1;
 	GElf_Ehdr ehdr;
 	GElf_Shdr shdr;
 	Elf_Data *build_id_data;
 	Elf_Scn *sec;
-	char *build_id = NULL, *bid;
-	unsigned char *raw;
 	Elf *elf;
-	int fd = open(self->long_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) {
-		pr_err("%s: cannot read %s ELF file.\n", __func__,
-		       self->long_name);
+		pr_debug2("%s: cannot read %s ELF file.\n", __func__, filename);
 		goto out_close;
 	}
 
@@ -858,30 +933,40 @@ static char *dso__read_build_id(struct dso *self)
 		goto out_elf_end;
 	}
 
-	sec = elf_section_by_name(elf, &ehdr, &shdr, ".note.gnu.build-id", NULL);
+	sec = elf_section_by_name(elf, &ehdr, &shdr,
+				  ".note.gnu.build-id", 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)
-		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;
-	}
-	pr_debug2("%s(%s): %s\n", __func__, self->long_name, build_id);
+	memcpy(bf, build_id_data->d_buf + 16, BUILD_ID_SIZE);
+	err = BUILD_ID_SIZE;
 out_elf_end:
 	elf_end(elf);
 out_close:
 	close(fd);
 out:
+	return err;
+}
+
+static char *dso__read_build_id(struct dso *self)
+{
+	int len;
+	char *build_id = NULL;
+	unsigned char rawbf[BUILD_ID_SIZE];
+
+	len = filename__read_build_id(self->long_name, rawbf, sizeof(rawbf));
+	if (len < 0)
+		goto out;
+
+	build_id = malloc(len * 2 + 1);
+	if (build_id == NULL)
+		goto out;
+
+	build_id__sprintf(rawbf, len, build_id);
+out:
 	return build_id;
 }
 
@@ -909,6 +994,8 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 	int ret = -1;
 	int fd;
 
+	self->loaded = 1;
+
 	if (!name)
 		return -1;
 
@@ -925,6 +1012,8 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 
 more:
 	do {
+		int berr = 0;
+
 		self->origin++;
 		switch (self->origin) {
 		case DSO__ORIG_FEDORA:
@@ -941,8 +1030,7 @@ more:
 				snprintf(name, size,
 					 "/usr/lib/debug/.build-id/%.2s/%s.debug",
 					build_id, build_id + 2);
-				free(build_id);
-				break;
+				goto compare_build_id;
 			}
 			self->origin++;
 			/* Fall thru */
@@ -954,6 +1042,22 @@ more:
 			goto out;
 		}
 
+		if (self->has_build_id) {
+			bool match;
+			build_id = malloc(BUILD_ID_SIZE);
+			if (build_id == NULL)
+				goto more;
+			berr = filename__read_build_id(name, build_id,
+						       BUILD_ID_SIZE);
+compare_build_id:
+			match = berr > 0 && memcmp(build_id, self->build_id,
+						   sizeof(self->build_id)) == 0;
+			free(build_id);
+			build_id = NULL;
+			if (!match)
+				goto more;
+		}
+
 		fd = open(name, O_RDONLY);
 	} while (fd < 0);
 
@@ -967,7 +1071,7 @@ more:
 		goto more;
 
 	if (ret > 0) {
-		int nr_plt = dso__synthesize_plt_symbols(self);
+		int nr_plt = dso__synthesize_plt_symbols(self, map, filter);
 		if (nr_plt > 0)
 			ret += nr_plt;
 	}
@@ -1019,6 +1123,8 @@ static int dso__load_module_sym(struct dso *self, struct map *map,
 {
 	int err = 0, fd = open(self->long_name, O_RDONLY);
 
+	self->loaded = 1;
+
 	if (fd < 0) {
 		pr_err("%s: cannot open %s\n", __func__, self->long_name);
 		return err;
@@ -1128,22 +1234,16 @@ static struct map *map__new2(u64 start, struct dso *dso)
 	struct map *self = malloc(sizeof(*self));
 
 	if (self != NULL) {
-		self->start = start;
 		/*
-		 * Will be filled after we load all the symbols
+		 * ->end will be filled after we load all the symbols
 		 */
-		self->end = 0;
-
-		self->pgoff = 0;
-		self->dso = dso;
-		self->map_ip = map__map_ip;
-		self->unmap_ip = map__unmap_ip;
-		RB_CLEAR_NODE(&self->rb_node);
+		map__init(self, start, 0, 0, dso);
 	}
+
 	return self;
 }
 
-static int dsos__load_modules(unsigned int sym_priv_size)
+static int dsos__load_modules(void)
 {
 	char *line = NULL;
 	size_t n;
@@ -1182,7 +1282,7 @@ static int dsos__load_modules(unsigned int sym_priv_size)
 		*sep = '\0';
 
 		snprintf(name, sizeof(name), "[%s]", line);
-		dso = dso__new(name, sym_priv_size);
+		dso = dso__new(name);
 
 		if (dso == NULL)
 			goto out_delete_line;
@@ -1214,6 +1314,8 @@ static int dso__load_vmlinux(struct dso *self, struct map *map,
 {
 	int err, fd = open(vmlinux, O_RDONLY);
 
+	self->loaded = 1;
+
 	if (fd < 0)
 		return -1;
 
@@ -1224,11 +1326,11 @@ static int dso__load_vmlinux(struct dso *self, struct map *map,
 	return err;
 }
 
-int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
-		      symbol_filter_t filter, int use_modules)
+int dsos__load_kernel(const char *vmlinux, symbol_filter_t filter,
+		      int use_modules)
 {
 	int err = -1;
-	struct dso *dso = dso__new(vmlinux, sym_priv_size);
+	struct dso *dso = dso__new(vmlinux);
 
 	if (dso == NULL)
 		return -1;
@@ -1240,7 +1342,7 @@ int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
 
 	kernel_map->map_ip = kernel_map->unmap_ip = identity__map_ip;
 
-	if (use_modules && dsos__load_modules(sym_priv_size) < 0) {
+	if (use_modules && dsos__load_modules() < 0) {
 		pr_warning("Failed to load list of modules in use! "
 			   "Continuing...\n");
 		use_modules = 0;
@@ -1312,19 +1414,15 @@ static struct dso *dsos__find(const char *name)
 	return NULL;
 }
 
-struct dso *dsos__findnew(const char *name, unsigned int sym_priv_size,
-			  bool *is_new)
+struct dso *dsos__findnew(const char *name)
 {
 	struct dso *dso = dsos__find(name);
 
 	if (!dso) {
-		dso = dso__new(name, sym_priv_size);
-		if (dso) {
+		dso = dso__new(name);
+		if (dso != NULL)
 			dsos__add(dso);
-			*is_new = true;
-		}
-	} else
-		*is_new = false;
+	}
 
 	return dso;
 }
@@ -1337,13 +1435,24 @@ void dsos__fprintf(FILE *fp)
 		dso__fprintf(pos, fp);
 }
 
-int load_kernel(unsigned int sym_priv_size, symbol_filter_t filter)
+size_t dsos__fprintf_buildid(FILE *fp)
+{
+	struct dso *pos;
+	size_t ret = 0;
+
+	list_for_each_entry(pos, &dsos, node) {
+		ret += dso__fprintf_buildid(pos, fp);
+		ret += fprintf(fp, " %s\n", pos->long_name);
+	}
+	return ret;
+}
+
+int load_kernel(symbol_filter_t filter)
 {
-	if (dsos__load_kernel(vmlinux_name, sym_priv_size, filter,
-			      modules) <= 0)
+	if (dsos__load_kernel(vmlinux_name, filter, modules) <= 0)
 		return -1;
 
-	vdso = dso__new("[vdso]", 0);
+	vdso = dso__new("[vdso]");
 	if (!vdso)
 		return -1;
 
@@ -1352,7 +1461,8 @@ int load_kernel(unsigned int sym_priv_size, symbol_filter_t filter)
 	return 0;
 }
 
-void symbol__init(void)
+void symbol__init(unsigned int priv_size)
 {
 	elf_version(EV_CURRENT);
+	symbol__priv_size = priv_size;
 }
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 77b7b3e..51c5a4a 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -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 */
@@ -39,42 +49,51 @@ struct symbol {
 	char		name[0];
 };
 
+extern unsigned int symbol__priv_size;
+
+static inline void *symbol__priv(struct symbol *self)
+{
+	return ((void *)self) - symbol__priv_size;
+}
+
 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;
+	u8		 adjust_symbols:1;
+	u8		 slen_calculated:1;
+	u8		 loaded:1;
+	u8		 has_build_id:1;
 	unsigned char	 origin;
+	u8		 build_id[BUILD_ID_SIZE];
 	const char	 *short_name;
 	char	 	 *long_name;
 	char		 name[0];
 };
 
-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;
-}
-
 struct symbol *dso__find_symbol(struct dso *self, u64 ip);
 
-int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
-		      symbol_filter_t filter, int modules);
-struct dso *dsos__findnew(const char *name, unsigned int sym_priv_size,
-			  bool *is_new);
+int dsos__load_kernel(const char *vmlinux, symbol_filter_t filter, int modules);
+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_buildid(struct dso *self, FILE *fp);
 size_t dso__fprintf(struct dso *self, 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);
+bool fetch_build_id_table(struct list_head *head);
+int build_id__sprintf(u8 *self, int len, char *bf);
 
-int load_kernel(unsigned int sym_priv_size, symbol_filter_t filter);
+int load_kernel(symbol_filter_t filter);
 
-void symbol__init(void);
+void symbol__init(unsigned int priv_size);
 
 extern struct list_head dsos;
 extern struct map *kernel_map;
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 0daa341..f2203a0 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -315,6 +315,7 @@ 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
@@ -332,6 +333,8 @@ 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)