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

* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf sched: Fix build failure on sparc
  perf bench: Add "all" pseudo subsystem and "all" pseudo suite
  perf tools: Introduce perf_session class
  perf symbols: Ditch dso->find_symbol
  perf symbols: Allow lookups by symbol name too
  perf symbols: Add missing "Variables" entry to map_type__name
  perf symbols: Add support for 'variable' symtabs
  perf symbols: Introduce ELF counterparts to symbol_type__is_a
  perf symbols: Introduce symbol_type__is_a
  perf symbols: Rename kthreads to kmaps, using another abstraction for it
  perf tools: Allow building for ARM
  hw-breakpoints: Handle bad modify_user_hw_breakpoint off-case return value
  perf tools: Allow cross compiling
  tracing, slab: Fix no callsite ifndef CONFIG_KMEMTRACE
  tracing, slab: Define kmem_cache_alloc_notrace ifdef CONFIG_TRACING

Trivial conflict due to different fixes to modify_user_hw_breakpoint()
in include/linux/hw_breakpoint.h

26 files changed, 638 insertions, 340 deletions

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 23ec66098bd..406999668ca 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -237,8 +237,8 @@ lib = lib
 
 export prefix bindir sharedir sysconfdir
 
-CC = gcc
-AR = ar
+CC = $(CROSS_COMPILE)gcc
+AR = $(CROSS_COMPILE)ar
 RM = rm -f
 TAR = tar
 FIND = find
@@ -356,7 +356,9 @@ LIB_H += util/parse-options.h
 LIB_H += util/parse-events.h
 LIB_H += util/quote.h
 LIB_H += util/util.h
+LIB_H += util/header.h
 LIB_H += util/help.h
+LIB_H += util/session.h
 LIB_H += util/strbuf.h
 LIB_H += util/string.h
 LIB_H += util/strlist.h
@@ -405,6 +407,7 @@ LIB_OBJS += util/callchain.o
 LIB_OBJS += util/values.o
 LIB_OBJS += util/debug.o
 LIB_OBJS += util/map.o
+LIB_OBJS += util/session.o
 LIB_OBJS += util/thread.o
 LIB_OBJS += util/trace-event-parse.o
 LIB_OBJS += util/trace-event-read.o
@@ -492,8 +495,10 @@ else
 	LIB_OBJS += util/probe-finder.o
 endif
 
+ifndef NO_LIBPERL
 PERL_EMBED_LDOPTS = `perl -MExtUtils::Embed -e ldopts 2>/dev/null`
 PERL_EMBED_CCOPTS = `perl -MExtUtils::Embed -e ccopts 2>/dev/null`
+endif
 
 ifneq ($(shell sh -c "(echo '\#include <EXTERN.h>'; echo '\#include <perl.h>'; echo 'int main(void) { perl_alloc(); return 0; }') | $(CC) -x c - $(PERL_EMBED_CCOPTS) -o /dev/null $(PERL_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y)
 	BASIC_CFLAGS += -DNO_LIBPERL
diff --git a/tools/perf/bench/sched-messaging.c b/tools/perf/bench/sched-messaging.c
index 605a2a959aa..81cee78181f 100644
--- a/tools/perf/bench/sched-messaging.c
+++ b/tools/perf/bench/sched-messaging.c
@@ -1,6 +1,6 @@
 /*
  *
- * builtin-bench-messaging.c
+ * sched-messaging.c
  *
  * messaging: Benchmark for scheduler and IPC mechanisms
  *
@@ -320,10 +320,12 @@ int bench_sched_messaging(int argc, const char **argv,
 		       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);
+		       diff.tv_sec,
+		       (unsigned long) (diff.tv_usec/1000));
 		break;
 	case BENCH_FORMAT_SIMPLE:
-		printf("%lu.%03lu\n", diff.tv_sec, diff.tv_usec/1000);
+		printf("%lu.%03lu\n", diff.tv_sec,
+		       (unsigned long) (diff.tv_usec/1000));
 		break;
 	default:
 		/* reaching here is something disaster */
diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c
index 238185f9797..4f77c7c2764 100644
--- a/tools/perf/bench/sched-pipe.c
+++ b/tools/perf/bench/sched-pipe.c
@@ -1,6 +1,6 @@
 /*
  *
- * builtin-bench-pipe.c
+ * sched-pipe.c
  *
  * pipe: Benchmark for pipe()
  *
@@ -87,7 +87,8 @@ int bench_sched_pipe(int argc, const char **argv,
 	if (pid) {
 		retpid = waitpid(pid, &wait_stat, 0);
 		assert((retpid == pid) && WIFEXITED(wait_stat));
-		return 0;
+	} else {
+		exit(0);
 	}
 
 	switch (bench_format) {
@@ -99,7 +100,8 @@ int bench_sched_pipe(int argc, const char **argv,
 		result_usec += diff.tv_usec;
 
 		printf(" %14s: %lu.%03lu [sec]\n\n", "Total time",
-		       diff.tv_sec, diff.tv_usec/1000);
+		       diff.tv_sec,
+		       (unsigned long) (diff.tv_usec/1000));
 
 		printf(" %14lf usecs/op\n",
 		       (double)result_usec / (double)loops);
@@ -110,7 +112,8 @@ int bench_sched_pipe(int argc, const char **argv,
 
 	case BENCH_FORMAT_SIMPLE:
 		printf("%lu.%03lu\n",
-		       diff.tv_sec, diff.tv_usec / 1000);
+		       diff.tv_sec,
+		       (unsigned long) (diff.tv_usec / 1000));
 		break;
 
 	default:
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 0bf2e8f9af5..21a78d30d53 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -25,6 +25,7 @@
 #include "util/thread.h"
 #include "util/sort.h"
 #include "util/hist.h"
+#include "util/session.h"
 #include "util/data_map.h"
 
 static char		const *input_name = "perf.data";
@@ -462,21 +463,23 @@ static struct perf_file_handler file_handler = {
 
 static int __cmd_annotate(void)
 {
-	struct perf_header *header;
+	struct perf_session *session = perf_session__new(input_name, O_RDONLY, force);
 	struct thread *idle;
 	int ret;
 
+	if (session == NULL)
+		return -ENOMEM;
+
 	idle = register_idle_thread();
 	register_perf_file_handler(&file_handler);
 
-	ret = mmap_dispatch_perf_file(&header, input_name, 0, 0,
-				      &event__cwdlen, &event__cwd);
+	ret = perf_session__process_events(session, 0, &event__cwdlen, &event__cwd);
 	if (ret)
-		return ret;
+		goto out_delete;
 
 	if (dump_trace) {
 		event__print_totals();
-		return 0;
+		goto out_delete;
 	}
 
 	if (verbose > 3)
@@ -489,6 +492,8 @@ static int __cmd_annotate(void)
 	output__resort(event__total[0]);
 
 	find_annotations();
+out_delete:
+	perf_session__delete(session);
 
 	return ret;
 }
diff --git a/tools/perf/builtin-bench.c b/tools/perf/builtin-bench.c
index e043eb83092..46996774e55 100644
--- a/tools/perf/builtin-bench.c
+++ b/tools/perf/builtin-bench.c
@@ -31,6 +31,9 @@ struct bench_suite {
 	const char *summary;
 	int (*fn)(int, const char **, const char *);
 };
+						\
+/* sentinel: easy for help */
+#define suite_all { "all", "test all suite (pseudo suite)", NULL }
 
 static struct bench_suite sched_suites[] = {
 	{ "messaging",
@@ -39,6 +42,7 @@ static struct bench_suite sched_suites[] = {
 	{ "pipe",
 	  "Flood of communication over pipe() between two processes",
 	  bench_sched_pipe      },
+	suite_all,
 	{ NULL,
 	  NULL,
 	  NULL                  }
@@ -48,6 +52,7 @@ static struct bench_suite mem_suites[] = {
 	{ "memcpy",
 	  "Simple memory copy in various ways",
 	  bench_mem_memcpy },
+	suite_all,
 	{ NULL,
 	  NULL,
 	  NULL             }
@@ -66,6 +71,9 @@ static struct bench_subsys subsystems[] = {
 	{ "mem",
 	  "memory access performance",
 	  mem_suites },
+	{ "all",		/* sentinel: easy for help */
+	  "test all subsystem (pseudo subsystem)",
+	  NULL },
 	{ NULL,
 	  NULL,
 	  NULL       }
@@ -75,11 +83,11 @@ static void dump_suites(int subsys_index)
 {
 	int i;
 
-	printf("List of available suites for %s...\n\n",
+	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",
+		printf("%14s: %s\n",
 		       subsystems[subsys_index].suites[i].name,
 		       subsystems[subsys_index].suites[i].summary);
 
@@ -110,10 +118,10 @@ static void print_usage(void)
 		printf("\t%s\n", bench_usage[i]);
 	printf("\n");
 
-	printf("List of available subsystems...\n\n");
+	printf("# List of available subsystems...\n\n");
 
 	for (i = 0; subsystems[i].name; i++)
-		printf("\t%s: %s\n",
+		printf("%14s: %s\n",
 		       subsystems[i].name, subsystems[i].summary);
 	printf("\n");
 }
@@ -131,6 +139,37 @@ static int bench_str2int(char *str)
 	return BENCH_FORMAT_UNKNOWN;
 }
 
+static void all_suite(struct bench_subsys *subsys)	  /* FROM HERE */
+{
+	int i;
+	const char *argv[2];
+	struct bench_suite *suites = subsys->suites;
+
+	argv[1] = NULL;
+	/*
+	 * TODO:
+	 * preparing preset parameters for
+	 * embedded, ordinary PC, HPC, etc...
+	 * will be helpful
+	 */
+	for (i = 0; suites[i].fn; i++) {
+		printf("# Running %s/%s benchmark...\n",
+		       subsys->name,
+		       suites[i].name);
+
+		argv[1] = suites[i].name;
+		suites[i].fn(1, argv, NULL);
+		printf("\n");
+	}
+}
+
+static void all_subsystem(void)
+{
+	int i;
+	for (i = 0; subsystems[i].suites; i++)
+		all_suite(&subsystems[i]);
+}
+
 int cmd_bench(int argc, const char **argv, const char *prefix __used)
 {
 	int i, j, status = 0;
@@ -155,6 +194,11 @@ int cmd_bench(int argc, const char **argv, const char *prefix __used)
 		goto end;
 	}
 
+	if (!strcmp(argv[0], "all")) {
+		all_subsystem();
+		goto end;
+	}
+
 	for (i = 0; subsystems[i].name; i++) {
 		if (strcmp(subsystems[i].name, argv[0]))
 			continue;
@@ -165,6 +209,11 @@ int cmd_bench(int argc, const char **argv, const char *prefix __used)
 			goto end;
 		}
 
+		if (!strcmp(argv[1], "all")) {
+			all_suite(&subsystems[i]);
+			goto end;
+		}
+
 		for (j = 0; subsystems[i].suites[j].name; j++) {
 			if (strcmp(subsystems[i].suites[j].name, argv[1]))
 				continue;
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
index dcb6143a000..bfd16a1594e 100644
--- a/tools/perf/builtin-buildid-list.c
+++ b/tools/perf/builtin-buildid-list.c
@@ -11,8 +11,8 @@
 #include "util/cache.h"
 #include "util/data_map.h"
 #include "util/debug.h"
-#include "util/header.h"
 #include "util/parse-options.h"
+#include "util/session.h"
 #include "util/symbol.h"
 
 static char const *input_name = "perf.data";
@@ -55,56 +55,17 @@ static int perf_file_section__process_buildids(struct perf_file_section *self,
 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);
+	struct perf_session *session = perf_session__new(input_name, O_RDONLY, force);
 
-	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;
-	}
+	if (session == NULL)
+		return -1;
 
-	err = perf_header__process_sections(header, input,
+	err = perf_header__process_sections(&session->header, session->fd,
 				         perf_file_section__process_buildids);
+	if (err >= 0)
+		dsos__fprintf_buildid(stdout);
 
-	if (err < 0)
-		goto out_close;
-
-	dsos__fprintf_buildid(stdout);
-out_close:
-	close(input);
-out:
+	perf_session__delete(session);
 	return err;
 }
 
diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c
index 5f209514f65..2071d248591 100644
--- a/tools/perf/builtin-kmem.c
+++ b/tools/perf/builtin-kmem.c
@@ -6,6 +6,7 @@
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/header.h"
+#include "util/session.h"
 
 #include "util/parse-options.h"
 #include "util/trace-event.h"
@@ -20,7 +21,6 @@ typedef int (*sort_fn_t)(struct alloc_stat *, struct alloc_stat *);
 
 static char const		*input_name = "perf.data";
 
-static struct perf_header	*header;
 static u64			sample_type;
 
 static int			alloc_flag;
@@ -367,11 +367,18 @@ static struct perf_file_handler file_handler = {
 
 static int read_events(void)
 {
+	int err;
+	struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0);
+
+	if (session == NULL)
+		return -ENOMEM;
+
 	register_idle_thread();
 	register_perf_file_handler(&file_handler);
 
-	return mmap_dispatch_perf_file(&header, input_name, 0, 0,
-				       &event__cwdlen, &event__cwd);
+	err = perf_session__process_events(session, 0, &event__cwdlen, &event__cwd);
+	perf_session__delete(session);
+	return err;
 }
 
 static double fragmentation(unsigned long n_req, unsigned long n_alloc)
@@ -403,7 +410,7 @@ static void __print_result(struct rb_root *root, int n_lines, int is_caller)
 		if (is_caller) {
 			addr = data->call_site;
 			if (!raw_ip)
-				sym = thread__find_function(kthread, addr, NULL);
+				sym = map_groups__find_function(kmaps, addr, NULL);
 		} else
 			addr = data->ptr;
 
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 0e519c667e3..4decbd14eae 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/session.h"
 #include "util/symbol.h"
 
 #include <unistd.h>
@@ -62,7 +63,7 @@ static int			nr_cpu				=      0;
 
 static int			file_new			=      1;
 
-struct perf_header		*header				=   NULL;
+static struct perf_session	*session;
 
 struct mmap_data {
 	int			counter;
@@ -216,12 +217,12 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n
 {
 	struct perf_header_attr *h_attr;
 
-	if (nr < header->attrs) {
-		h_attr = header->attr[nr];
+	if (nr < session->header.attrs) {
+		h_attr = session->header.attr[nr];
 	} else {
 		h_attr = perf_header_attr__new(a);
 		if (h_attr != NULL)
-			if (perf_header__add_attr(header, h_attr) < 0) {
+			if (perf_header__add_attr(&session->header, h_attr) < 0) {
 				perf_header_attr__delete(h_attr);
 				h_attr = NULL;
 			}
@@ -395,9 +396,9 @@ static void open_counters(int cpu, pid_t pid)
 
 static void atexit_header(void)
 {
-	header->data_size += bytes_written;
+	session->header.data_size += bytes_written;
 
-	perf_header__write(header, output, true);
+	perf_header__write(&session->header, output, true);
 }
 
 static int __cmd_record(int argc, const char **argv)
@@ -440,24 +441,24 @@ static int __cmd_record(int argc, const char **argv)
 		exit(-1);
 	}
 
-	header = perf_header__new();
-	if (header == NULL) {
+	session = perf_session__new(output_name, O_WRONLY, force);
+	if (session == NULL) {
 		pr_err("Not enough memory for reading perf file header\n");
 		return -1;
 	}
 
 	if (!file_new) {
-		err = perf_header__read(header, output);
+		err = perf_header__read(&session->header, output);
 		if (err < 0)
 			return err;
 	}
 
 	if (raw_samples) {
-		perf_header__set_feat(header, HEADER_TRACE_INFO);
+		perf_header__set_feat(&session->header, HEADER_TRACE_INFO);
 	} else {
 		for (i = 0; i < nr_counters; i++) {
 			if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
-				perf_header__set_feat(header, HEADER_TRACE_INFO);
+				perf_header__set_feat(&session->header, HEADER_TRACE_INFO);
 				break;
 			}
 		}
@@ -481,7 +482,7 @@ static int __cmd_record(int argc, const char **argv)
 	}
 
 	if (file_new) {
-		err = perf_header__write(header, output, false);
+		err = perf_header__write(&session->header, output, false);
 		if (err < 0)
 			return err;
 	}
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 2b9eb3a553e..e2ec49a9b73 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -22,6 +22,7 @@
 #include "perf.h"
 #include "util/debug.h"
 #include "util/header.h"
+#include "util/session.h"
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
@@ -52,7 +53,7 @@ static int		exclude_other = 1;
 
 static char		callchain_default_opt[] = "fractal,0.5";
 
-static struct perf_header *header;
+static struct perf_session *session;
 
 static u64		sample_type;
 
@@ -701,7 +702,7 @@ static int process_read_event(event_t *event)
 {
 	struct perf_event_attr *attr;
 
-	attr = perf_header__find_attr(event->read.id, header);
+	attr = perf_header__find_attr(event->read.id, &session->header);
 
 	if (show_threads) {
 		const char *name = attr ? __event_name(attr->type, attr->config)
@@ -766,6 +767,10 @@ static int __cmd_report(void)
 	struct thread *idle;
 	int ret;
 
+	session = perf_session__new(input_name, O_RDONLY, force);
+	if (session == NULL)
+		return -ENOMEM;
+
 	idle = register_idle_thread();
 	thread__comm_adjust(idle);
 
@@ -774,14 +779,14 @@ static int __cmd_report(void)
 
 	register_perf_file_handler(&file_handler);
 
-	ret = mmap_dispatch_perf_file(&header, input_name, force,
-				      full_paths, &event__cwdlen, &event__cwd);
+	ret = perf_session__process_events(session, full_paths,
+					   &event__cwdlen, &event__cwd);
 	if (ret)
-		return ret;
+		goto out_delete;
 
 	if (dump_trace) {
 		event__print_totals();
-		return 0;
+		goto out_delete;
 	}
 
 	if (verbose > 3)
@@ -796,7 +801,8 @@ static int __cmd_report(void)
 
 	if (show_threads)
 		perf_read_values_destroy(&show_threads_values);
-
+out_delete:
+	perf_session__delete(session);
 	return ret;
 }
 
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 7cca7c15b40..65021fe1361 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -6,6 +6,7 @@
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/header.h"
+#include "util/session.h"
 
 #include "util/parse-options.h"
 #include "util/trace-event.h"
@@ -21,7 +22,6 @@
 
 static char			const *input_name = "perf.data";
 
-static struct perf_header	*header;
 static u64			sample_type;
 
 static char			default_sort_order[] = "avg, max, switch, runtime";
@@ -1663,11 +1663,18 @@ static struct perf_file_handler file_handler = {
 
 static int read_events(void)
 {
+	int err;
+	struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0);
+
+	if (session == NULL)
+		return -ENOMEM;
+
 	register_idle_thread();
 	register_perf_file_handler(&file_handler);
 
-	return mmap_dispatch_perf_file(&header, input_name, 0, 0,
-				       &event__cwdlen, &event__cwd);
+	err = perf_session__process_events(session, 0, &event__cwdlen, &event__cwd);
+	perf_session__delete(session);
+	return err;
 }
 
 static void print_bad_events(void)
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index f472df9561e..759dd2b35fd 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -1059,15 +1059,17 @@ static struct perf_file_handler file_handler = {
 
 static int __cmd_timechart(void)
 {
-	struct perf_header *header;
+	struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0);
 	int ret;
 
+	if (session == NULL)
+		return -ENOMEM;
+
 	register_perf_file_handler(&file_handler);
 
-	ret = mmap_dispatch_perf_file(&header, input_name, 0, 0,
-				      &event__cwdlen, &event__cwd);
+	ret = perf_session__process_events(session, 0, &event__cwdlen, &event__cwd);
 	if (ret)
-		return EXIT_FAILURE;
+		goto out_delete;
 
 	process_samples();
 
@@ -1079,8 +1081,9 @@ static int __cmd_timechart(void)
 
 	pr_info("Written %2.1f seconds of trace to %s.\n",
 		(last_time - first_time) / 1000000000.0, output_name);
-
-	return EXIT_SUCCESS;
+out_delete:
+	perf_session__delete(session);
+	return ret;
 }
 
 static const char * const timechart_usage[] = {
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index c2fcc34486f..0756664666f 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -7,6 +7,7 @@
 #include "util/header.h"
 #include "util/exec_cmd.h"
 #include "util/trace-event.h"
+#include "util/session.h"
 
 static char const		*script_name;
 static char const		*generate_script_lang;
@@ -61,7 +62,7 @@ static int cleanup_scripting(void)
 
 static char const		*input_name = "perf.data";
 
-static struct perf_header	*header;
+static struct perf_session 	*session;
 static u64			sample_type;
 
 static int process_sample_event(event_t *event)
@@ -126,11 +127,18 @@ static struct perf_file_handler file_handler = {
 
 static int __cmd_trace(void)
 {
+	int err;
+
+	session = perf_session__new(input_name, O_RDONLY, 0);
+	if (session == NULL)
+		return -ENOMEM;
+
 	register_idle_thread();
 	register_perf_file_handler(&file_handler);
 
-	return mmap_dispatch_perf_file(&header, input_name,
-				       0, 0, &event__cwdlen, &event__cwd);
+	err = perf_session__process_events(session, 0, &event__cwdlen, &event__cwd);
+	perf_session__delete(session);
+	return err;
 }
 
 struct script_spec {
@@ -348,11 +356,7 @@ int cmd_trace(int argc, const char **argv, const char *prefix __used)
 			return -1;
 		}
 
-		header = perf_header__new();
-		if (header == NULL)
-			return -1;
-
-		perf_header__read(header, input);
+		perf_header__read(&session->header, input);
 		err = scripting_ops->generate_script("perf-trace");
 		goto out;
 	}
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 454d5d55f32..75f941bfba9 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -59,6 +59,18 @@
 #define cpu_relax()	asm volatile ("hint @pause" ::: "memory")
 #endif
 
+#ifdef __arm__
+#include "../../arch/arm/include/asm/unistd.h"
+/*
+ * Use the __kuser_memory_barrier helper in the CPU helper page. See
+ * arch/arm/kernel/entry-armv.S in the kernel source for details.
+ */
+#define rmb()		asm volatile("mov r0, #0xffff0fff; mov lr, pc;" \
+				     "sub pc, r0, #95" ::: "r0", "lr", "cc", \
+				     "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 59b65d0bd7c..6d46dda53a2 100644
--- a/tools/perf/util/data_map.c
+++ b/tools/perf/util/data_map.c
@@ -129,23 +129,16 @@ out:
 	return err;
 }
 
-int mmap_dispatch_perf_file(struct perf_header **pheader,
-			    const char *input_name,
-			    int force,
-			    int full_paths,
-			    int *cwdlen,
-			    char **cwd)
+int perf_session__process_events(struct perf_session *self,
+				 int full_paths, int *cwdlen, char **cwd)
 {
 	int err;
-	struct perf_header *header;
 	unsigned long head, shift;
 	unsigned long offset = 0;
-	struct stat input_stat;
 	size_t	page_size;
 	u64 sample_type;
 	event_t *event;
 	uint32_t size;
-	int input;
 	char *buf;
 
 	if (curr_handler == NULL) {
@@ -155,56 +148,19 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
 
 	page_size = getpagesize();
 
-	input = open(input_name, O_RDONLY);
-	if (input < 0) {
-		pr_err("Failed to open file: %s", input_name);
-		if (!strcmp(input_name, "perf.data"))
-			pr_err("  (try 'perf record' first)");
-		pr_err("\n");
-		return -errno;
-	}
-
-	if (fstat(input, &input_stat) < 0) {
-		pr_err("failed to stat file");
-		err = -errno;
-		goto out_close;
-	}
-
-	err = -EACCES;
-	if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
-		pr_err("file: %s not owned by current user or root\n",
-			input_name);
-		goto out_close;
-	}
-
-	if (input_stat.st_size == 0) {
-		pr_info("zero-sized file, nothing to do!\n");
-		goto done;
-	}
-
-	err = -ENOMEM;
-	header = perf_header__new();
-	if (header == NULL)
-		goto out_close;
-
-	err = perf_header__read(header, input);
-	if (err < 0)
-		goto out_delete;
-	*pheader = header;
-	head = header->data_offset;
-
-	sample_type = perf_header__sample_type(header);
+	head = self->header.data_offset;
+	sample_type = perf_header__sample_type(&self->header);
 
 	err = -EINVAL;
 	if (curr_handler->sample_type_check &&
 	    curr_handler->sample_type_check(sample_type) < 0)
-		goto out_delete;
+		goto out_err;
 
 	if (!full_paths) {
 		if (getcwd(__cwd, sizeof(__cwd)) == NULL) {
 			pr_err("failed to get the current directory\n");
 			err = -errno;
-			goto out_delete;
+			goto out_err;
 		}
 		*cwd = __cwd;
 		*cwdlen = strlen(*cwd);
@@ -219,11 +175,11 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
 
 remap:
 	buf = mmap(NULL, page_size * mmap_window, PROT_READ,
-		   MAP_SHARED, input, offset);
+		   MAP_SHARED, self->fd, offset);
 	if (buf == MAP_FAILED) {
 		pr_err("failed to mmap file\n");
 		err = -errno;
-		goto out_delete;
+		goto out_err;
 	}
 
 more:
@@ -273,19 +229,14 @@ more:
 
 	head += size;
 
-	if (offset + head >= header->data_offset + header->data_size)
+	if (offset + head >= self->header.data_offset + self->header.data_size)
 		goto done;
 
-	if (offset + head < (unsigned long)input_stat.st_size)
+	if (offset + head < self->size)
 		goto more;
 
 done:
 	err = 0;
-out_close:
-	close(input);
-
+out_err:
 	return err;
-out_delete:
-	perf_header__delete(header);
-	goto out_close;
 }
diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h
index 258a87bcc4f..98c5b823388 100644
--- a/tools/perf/util/data_map.h
+++ b/tools/perf/util/data_map.h
@@ -3,6 +3,7 @@
 
 #include "event.h"
 #include "header.h"
+#include "session.h"
 
 typedef int (*event_type_handler_t)(event_t *);
 
@@ -21,12 +22,8 @@ struct perf_file_handler {
 };
 
 void register_perf_file_handler(struct perf_file_handler *handler);
-int mmap_dispatch_perf_file(struct perf_header **pheader,
-			    const char *input_name,
-			    int force,
-			    int full_paths,
-			    int *cwdlen,
-			    char **cwd);
+int perf_session__process_events(struct perf_session *self,
+				 int full_paths, int *cwdlen, char **cwd);
 int perf_header__read_build_ids(int input, u64 offset, u64 file_size);
 
 #endif
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index 4dcecafa85d..ba0de90cd3d 100644
--- a/tools/perf/util/event.c
+++ b/