Merge branch 'master' into upstream

51 files changed, 1286 insertions, 350 deletions

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 43e3dd284b90..399751befeed 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -15,6 +15,23 @@ DESCRIPTION
 This command displays the symbolic event types which can be selected in the
 various perf commands with the -e option.
 
+EVENT MODIFIERS
+---------------
+
+Events can optionally have a modifer by appending a colon and one or
+more modifiers.  Modifiers allow the user to restrict when events are
+counted with 'u' for user-space, 'k' for kernel, 'h' for hypervisor.
+
+The 'p' modifier can be used for specifying how precise the instruction
+address should be. The 'p' modifier is currently only implemented for
+Intel PEBS and can be specified multiple times:
+  0 - SAMPLE_IP can have arbitrary skid
+  1 - SAMPLE_IP must have constant skid
+  2 - SAMPLE_IP requested to have 0 skid
+  3 - SAMPLE_IP must have 0 skid
+
+The PEBS implementation now supports up to 2.
+
 RAW HARDWARE EVENT DESCRIPTOR
 -----------------------------
 Even when an event is not available in a symbolic form within perf right now,
diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt
index 27d52dae5a43..62de1b7f4e76 100644
--- a/tools/perf/Documentation/perf-probe.txt
+++ b/tools/perf/Documentation/perf-probe.txt
@@ -16,7 +16,9 @@ or
 or
 'perf probe' --list
 or
-'perf probe' --line='FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]'
+'perf probe' [options] --line='FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]'
+or
+'perf probe' [options] --vars='PROBEPOINT'
 
 DESCRIPTION
 -----------
@@ -31,6 +33,11 @@ OPTIONS
 --vmlinux=PATH::
 	Specify vmlinux path which has debuginfo (Dwarf binary).
 
+-m::
+--module=MODNAME::
+	Specify module name in which perf-probe searches probe points
+	or lines.
+
 -s::
 --source=PATH::
 	Specify path to kernel source.
@@ -57,6 +64,15 @@ OPTIONS
 	Show source code lines which can be probed. This needs an argument
 	which specifies a range of the source code. (see LINE SYNTAX for detail)
 
+-V::
+--vars=::
+	Show available local variables at given probe point. The argument
+	syntax is same as PROBE SYNTAX, but NO ARGs.
+
+--externs::
+	(Only for --vars) Show external defined variables in addition to local
+	variables.
+
 -f::
 --force::
 	Forcibly add events with existing name.
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index 3ee27dccfde9..a91f9f9e6e5c 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -83,6 +83,10 @@ OPTIONS
 --call-graph::
 	Do call-graph (stack chain/backtrace) recording.
 
+-q::
+--quiet::
+	Don't print any message, useful for scripting.
+
 -v::
 --verbose::
 	Be more verbose (show counter open errors, etc).
diff --git a/tools/perf/Documentation/perf-trace.txt b/tools/perf/Documentation/perf-trace.txt
index 122ec9dc4853..26aff6bf9e50 100644
--- a/tools/perf/Documentation/perf-trace.txt
+++ b/tools/perf/Documentation/perf-trace.txt
@@ -8,7 +8,11 @@ perf-trace - Read perf.data (created by perf record) and display trace output
 SYNOPSIS
 --------
 [verse]
-'perf trace' {record <script> | report <script> [args] }
+'perf trace' [<options>]
+'perf trace' [<options>] record <script> [<record-options>] <command>
+'perf trace' [<options>] report <script> [script-args]
+'perf trace' [<options>] <script> <required-script-args> [<record-options>] <command>
+'perf trace' [<options>] <top-script> [script-args]
 
 DESCRIPTION
 -----------
@@ -24,23 +28,53 @@ There are several variants of perf trace:
   available via 'perf trace -l').  The following variants allow you to
   record and run those scripts:
 
-  'perf trace record <script>' to record the events required for 'perf
-  trace report'.  <script> is the name displayed in the output of
-  'perf trace --list' i.e. the actual script name minus any language
-  extension.
+  'perf trace record <script> <command>' to record the events required
+  for 'perf trace report'.  <script> is the name displayed in the
+  output of 'perf trace --list' i.e. the actual script name minus any
+  language extension.  If <command> is not specified, the events are
+  recorded using the -a (system-wide) 'perf record' option.
 
-  'perf trace report <script>' to run and display the results of
-  <script>.  <script> is the name displayed in the output of 'perf
+  'perf trace report <script> [args]' to run and display the results
+  of <script>.  <script> is the name displayed in the output of 'perf
   trace --list' i.e. the actual script name minus any language
   extension.  The perf.data output from a previous run of 'perf trace
   record <script>' is used and should be present for this command to
-  succeed.
+  succeed.  [args] refers to the (mainly optional) args expected by
+  the script.
+
+  'perf trace <script> <required-script-args> <command>' to both
+  record the events required for <script> and to run the <script>
+  using 'live-mode' i.e. without writing anything to disk.  <script>
+  is the name displayed in the output of 'perf trace --list' i.e. the
+  actual script name minus any language extension.  If <command> is
+  not specified, the events are recorded using the -a (system-wide)
+  'perf record' option.  If <script> has any required args, they
+  should be specified before <command>.  This mode doesn't allow for
+  optional script args to be specified; if optional script args are
+  desired, they can be specified using separate 'perf trace record'
+  and 'perf trace report' commands, with the stdout of the record step
+  piped to the stdin of the report script, using the '-o -' and '-i -'
+  options of the corresponding commands.
+
+  'perf trace <top-script>' to both record the events required for
+  <top-script> and to run the <top-script> using 'live-mode'
+  i.e. without writing anything to disk.  <top-script> is the name
+  displayed in the output of 'perf trace --list' i.e. the actual
+  script name minus any language extension; a <top-script> is defined
+  as any script name ending with the string 'top'.
+
+  [<record-options>] can be passed to the record steps of 'perf trace
+  record' and 'live-mode' variants; this isn't possible however for
+  <top-script> 'live-mode' or 'perf trace report' variants.
 
   See the 'SEE ALSO' section for links to language-specific
   information on how to write and run your own trace scripts.
 
 OPTIONS
 -------
+<command>...::
+	Any command you can specify in a shell.
+
 -D::
 --dump-raw-trace=::
         Display verbose dump of the trace data.
@@ -64,6 +98,13 @@ OPTIONS
         Generate perf-trace.[ext] starter script for given language,
         using current perf.data.
 
+-a::
+        Force system-wide collection.  Scripts run without a <command>
+        normally use -a by default, while scripts run with a <command>
+        normally don't - this option allows the latter to be run in
+        system-wide mode.
+
+
 SEE ALSO
 --------
 linkperf:perf-record[1], linkperf:perf-trace-perl[1],
diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c
index 199d5e19554f..2e000c068cc5 100644
--- a/tools/perf/builtin-probe.c
+++ b/tools/perf/builtin-probe.c
@@ -50,14 +50,17 @@ static struct {
 	bool list_events;
 	bool force_add;
 	bool show_lines;
+	bool show_vars;
+	bool show_ext_vars;
+	bool mod_events;
 	int nevents;
 	struct perf_probe_event events[MAX_PROBES];
 	struct strlist *dellist;
 	struct line_range line_range;
+	const char *target_module;
 	int max_probe_points;
 } params;
 
-
 /* Parse an event definition. Note that any error must die. */
 static int parse_probe_event(const char *str)
 {
@@ -92,6 +95,7 @@ static int parse_probe_event_argv(int argc, const char **argv)
 	len = 0;
 	for (i = 0; i < argc; i++)
 		len += sprintf(&buf[len], "%s ", argv[i]);
+	params.mod_events = true;
 	ret = parse_probe_event(buf);
 	free(buf);
 	return ret;
@@ -100,9 +104,10 @@ static int parse_probe_event_argv(int argc, const char **argv)
 static int opt_add_probe_event(const struct option *opt __used,
 			      const char *str, int unset __used)
 {
-	if (str)
+	if (str) {
+		params.mod_events = true;
 		return parse_probe_event(str);
-	else
+	} else
 		return 0;
 }
 
@@ -110,6 +115,7 @@ static int opt_del_probe_event(const struct option *opt __used,
 			       const char *str, int unset __used)
 {
 	if (str) {
+		params.mod_events = true;
 		if (!params.dellist)
 			params.dellist = strlist__new(true, NULL);
 		strlist__add(params.dellist, str);
@@ -130,6 +136,25 @@ static int opt_show_lines(const struct option *opt __used,
 
 	return ret;
 }
+
+static int opt_show_vars(const struct option *opt __used,
+			 const char *str, int unset __used)
+{
+	struct perf_probe_event *pev = &params.events[params.nevents];
+	int ret;
+
+	if (!str)
+		return 0;
+
+	ret = parse_probe_event(str);
+	if (!ret && pev->nargs != 0) {
+		pr_err("  Error: '--vars' doesn't accept arguments.\n");
+		return -EINVAL;
+	}
+	params.show_vars = true;
+
+	return ret;
+}
 #endif
 
 static const char * const probe_usage[] = {
@@ -138,7 +163,8 @@ static const char * const probe_usage[] = {
 	"perf probe [<options>] --del '[GROUP:]EVENT' ...",
 	"perf probe --list",
 #ifdef DWARF_SUPPORT
-	"perf probe --line 'LINEDESC'",
+	"perf probe [<options>] --line 'LINEDESC'",
+	"perf probe [<options>] --vars 'PROBEPOINT'",
 #endif
 	NULL
 };
@@ -180,10 +206,17 @@ static const struct option options[] = {
 	OPT_CALLBACK('L', "line", NULL,
 		     "FUNC[:RLN[+NUM|-RLN2]]|SRC:ALN[+NUM|-ALN2]",
 		     "Show source code lines.", opt_show_lines),
+	OPT_CALLBACK('V', "vars", NULL,
+		     "FUNC[@SRC][+OFF|%return|:RL|;PT]|SRC:AL|SRC;PT",
+		     "Show accessible variables on PROBEDEF", opt_show_vars),
+	OPT_BOOLEAN('\0', "externs", &params.show_ext_vars,
+		    "Show external variables too (with --vars only)"),
 	OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name,
 		   "file", "vmlinux pathname"),
 	OPT_STRING('s', "source", &symbol_conf.source_prefix,
 		   "directory", "path to kernel source"),
+	OPT_STRING('m', "module", &params.target_module,
+		   "modname", "target module name"),
 #endif
 	OPT__DRY_RUN(&probe_event_dry_run),
 	OPT_INTEGER('\0', "max-probes", &params.max_probe_points,
@@ -217,7 +250,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 		usage_with_options(probe_usage, options);
 
 	if (params.list_events) {
-		if (params.nevents != 0 || params.dellist) {
+		if (params.mod_events) {
 			pr_err("  Error: Don't use --list with --add/--del.\n");
 			usage_with_options(probe_usage, options);
 		}
@@ -225,6 +258,10 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 			pr_err("  Error: Don't use --list with --line.\n");
 			usage_with_options(probe_usage, options);
 		}
+		if (params.show_vars) {
+			pr_err(" Error: Don't use --list with --vars.\n");
+			usage_with_options(probe_usage, options);
+		}
 		ret = show_perf_probe_events();
 		if (ret < 0)
 			pr_err("  Error: Failed to show event list. (%d)\n",
@@ -234,17 +271,35 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 
 #ifdef DWARF_SUPPORT
 	if (params.show_lines) {
-		if (params.nevents != 0 || params.dellist) {
-			pr_warning("  Error: Don't use --line with"
-				   " --add/--del.\n");
+		if (params.mod_events) {
+			pr_err("  Error: Don't use --line with"
+			       " --add/--del.\n");
+			usage_with_options(probe_usage, options);
+		}
+		if (params.show_vars) {
+			pr_err(" Error: Don't use --line with --vars.\n");
 			usage_with_options(probe_usage, options);
 		}
 
-		ret = show_line_range(&params.line_range);
+		ret = show_line_range(&params.line_range, params.target_module);
 		if (ret < 0)
 			pr_err("  Error: Failed to show lines. (%d)\n", ret);
 		return ret;
 	}
+	if (params.show_vars) {
+		if (params.mod_events) {
+			pr_err("  Error: Don't use --vars with"
+			       " --add/--del.\n");
+			usage_with_options(probe_usage, options);
+		}
+		ret = show_available_vars(params.events, params.nevents,
+					  params.max_probe_points,
+					  params.target_module,
+					  params.show_ext_vars);
+		if (ret < 0)
+			pr_err("  Error: Failed to show vars. (%d)\n", ret);
+		return ret;
+	}
 #endif
 
 	if (params.dellist) {
@@ -258,8 +313,9 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 
 	if (params.nevents) {
 		ret = add_perf_probe_events(params.events, params.nevents,
-					    params.force_add,
-					    params.max_probe_points);
+					    params.max_probe_points,
+					    params.target_module,
+					    params.force_add);
 		if (ret < 0) {
 			pr_err("  Error: Failed to add events. (%d)\n", ret);
 			return ret;
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index ff77b805de71..564491fa18b2 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -197,7 +197,7 @@ static void sig_atexit(void)
 	if (child_pid > 0)
 		kill(child_pid, SIGTERM);
 
-	if (signr == -1)
+	if (signr == -1 || signr == SIGUSR1)
 		return;
 
 	signal(signr, SIG_DFL);
@@ -353,7 +353,7 @@ try_again:
 		}
 
 		if (read(fd[nr_cpu][counter][thread_index], &read_data, sizeof(read_data)) == -1) {
-			perror("Unable to read perf file descriptor\n");
+			perror("Unable to read perf file descriptor");
 			exit(-1);
 		}
 
@@ -515,6 +515,7 @@ static int __cmd_record(int argc, const char **argv)
 	atexit(sig_atexit);
 	signal(SIGCHLD, sig_handler);
 	signal(SIGINT, sig_handler);
+	signal(SIGUSR1, sig_handler);
 
 	if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) {
 		perror("failed to create pipes");
@@ -606,6 +607,7 @@ static int __cmd_record(int argc, const char **argv)
 			execvp(argv[0], (char **)argv);
 
 			perror(argv[0]);
+			kill(getppid(), SIGUSR1);
 			exit(-1);
 		}
 
@@ -626,7 +628,7 @@ static int __cmd_record(int argc, const char **argv)
 
 	nr_cpus = read_cpu_map(cpu_list);
 	if (nr_cpus < 1) {
-		perror("failed to collect number of CPUs\n");
+		perror("failed to collect number of CPUs");
 		return -1;
 	}
 
@@ -697,17 +699,18 @@ static int __cmd_record(int argc, const char **argv)
 	if (err < 0)
 		err = event__synthesize_kernel_mmap(process_synthesized_event,
 						    session, machine, "_stext");
-	if (err < 0) {
-		pr_err("Couldn't record kernel reference relocation symbol.\n");
-		return err;
-	}
+	if (err < 0)
+		pr_err("Couldn't record kernel reference relocation symbol\n"
+		       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
+		       "Check /proc/kallsyms permission or run as root.\n");
 
 	err = event__synthesize_modules(process_synthesized_event,
 					session, machine);
-	if (err < 0) {
-		pr_err("Couldn't record kernel reference relocation symbol.\n");
-		return err;
-	}
+	if (err < 0)
+		pr_err("Couldn't record kernel module information.\n"
+		       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
+		       "Check /proc/modules permission or run as root.\n");
+
 	if (perf_guest)
 		perf_session__process_machines(session, event__synthesize_guest_os);
 
@@ -761,6 +764,9 @@ static int __cmd_record(int argc, const char **argv)
 		}
 	}
 
+	if (quiet || signr == SIGUSR1)
+		return 0;
+
 	fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking);
 
 	/*
@@ -787,7 +793,7 @@ static const char * const record_usage[] = {
 
 static bool force, append_file;
 
-static const struct option options[] = {
+const struct option record_options[] = {
 	OPT_CALLBACK('e', "event", NULL, "event",
 		     "event selector. use 'perf list' to list available events",
 		     parse_events),
@@ -820,6 +826,7 @@ static const struct option options[] = {
 		    "do call-graph (stack chain/backtrace) recording"),
 	OPT_INCR('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
+	OPT_BOOLEAN('q', "quiet", &quiet, "don't print any message"),
 	OPT_BOOLEAN('s', "stat", &inherit_stat,
 		    "per thread counts"),
 	OPT_BOOLEAN('d', "data", &sample_address,
@@ -835,16 +842,16 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 {
 	int i, j, err = -ENOMEM;
 
-	argc = parse_options(argc, argv, options, record_usage,
+	argc = parse_options(argc, argv, record_options, record_usage,
 			    PARSE_OPT_STOP_AT_NON_OPTION);
 	if (!argc && target_pid == -1 && target_tid == -1 &&
 		!system_wide && !cpu_list)
-		usage_with_options(record_usage, options);
+		usage_with_options(record_usage, record_options);
 
 	if (force && append_file) {
 		fprintf(stderr, "Can't overwrite and append at the same time."
 				" You need to choose between -f and -A");
-		usage_with_options(record_usage, options);
+		usage_with_options(record_usage, record_options);
 	} else if (append_file) {
 		write_mode = WRITE_APPEND;
 	} else {
@@ -867,7 +874,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		if (thread_num <= 0) {
 			fprintf(stderr, "Can't find all threads of pid %d\n",
 					target_pid);
-			usage_with_options(record_usage, options);
+			usage_with_options(record_usage, record_options);
 		}
 	} else {
 		all_tids=malloc(sizeof(pid_t));
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index b513e40974f4..dd625808c2a5 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -69,7 +69,6 @@ static int			target_tid			=     -1;
 static pid_t			*all_tids			=      NULL;
 static int			thread_num			=      0;
 static bool			inherit				=  false;
-static int			profile_cpu			=     -1;
 static int			nr_cpus				=      0;
 static int			realtime_prio			=      0;
 static bool			group				=  false;
@@ -558,13 +557,13 @@ static void print_sym_table(void)
 	else
 		printf(" (all");
 
-	if (profile_cpu != -1)
-		printf(", cpu: %d)\n", profile_cpu);
+	if (cpu_list)
+		printf(", CPU%s: %s)\n", nr_cpus > 1 ? "s" : "", cpu_list);
 	else {
 		if (target_tid != -1)
 			printf(")\n");
 		else
-			printf(", %d CPUs)\n", nr_cpus);
+			printf(", %d CPU%s)\n", nr_cpus, nr_cpus > 1 ? "s" : "");
 	}
 
 	printf("%-*.*s\n", win_width, win_width, graph_dotted_line);
@@ -1187,11 +1186,10 @@ int group_fd;
 static void start_counter(int i, int counter)
 {
 	struct perf_event_attr *attr;
-	int cpu;
+	int cpu = -1;
 	int thread_index;
 
-	cpu = profile_cpu;
-	if (target_tid == -1 && profile_cpu == -1)
+	if (target_tid == -1)
 		cpu = cpumap[i];
 
 	attr = attrs + counter;
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 40a6a2992d15..86cfe3800e6b 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -10,6 +10,7 @@
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/trace-event.h"
+#include "util/parse-options.h"
 #include "util/util.h"
 
 static char const		*script_name;
@@ -17,6 +18,7 @@ static char const		*generate_script_lang;
 static bool			debug_mode;
 static u64			last_timestamp;
 static u64			nr_unordered;
+extern const struct option	record_options[];
 
 static int default_start_script(const char *script __unused,
 				int argc __unused,
@@ -46,9 +48,6 @@ static struct scripting_ops	*scripting_ops;
 
 static void setup_scripting(void)
 {
-	/* make sure PERF_EXEC_PATH is set for scripts */
-	perf_set_argv_exec_path(perf_exec_path());
-
 	setup_perl_scripting();
 	setup_python_scripting();
 
@@ -285,7 +284,7 @@ static int parse_scriptname(const struct option *opt __used,
 		script++;
 	} else {
 		script = str;
-		ext = strchr(script, '.');
+		ext = strrchr(script, '.');
 		if (!ext) {
 			fprintf(stderr, "invalid script extension");
 			return -1;
@@ -331,7 +330,7 @@ static struct script_desc *script_desc__new(const char *name)
 {
 	struct script_desc *s = zalloc(sizeof(*s));
 
-	if (s != NULL)
+	if (s != NULL && name)
 		s->name = strdup(name);
 
 	return s;
@@ -340,6 +339,8 @@ static struct script_desc *script_desc__new(const char *name)
 static void script_desc__delete(struct script_desc *s)
 {
 	free(s->name);
+	free(s->half_liner);
+	free(s->args);
 	free(s);
 }
 
@@ -540,8 +541,40 @@ static char *get_script_path(const char *script_root, const char *suffix)
 	return path;
 }
 
+static bool is_top_script(const char *script_path)
+{
+	return ends_with((char *)script_path, "top") == NULL ? false : true;
+}
+
+static int has_required_arg(char *script_path)
+{
+	struct script_desc *desc;
+	int n_args = 0;
+	char *p;
+
+	desc = script_desc__new(NULL);
+
+	if (read_script_info(desc, script_path))
+		goto out;
+
+	if (!desc->args)
+		goto out;
+
+	for (p = desc->args; *p; p++)
+		if (*p == '<')
+			n_args++;
+out:
+	script_desc__delete(desc);
+
+	return n_args;
+}
+
 static const char * const trace_usage[] = {
-	"perf trace [<options>] <command>",
+	"perf trace [<options>]",
+	"perf trace [<options>] record <script> [<record-options>] <command>",
+	"perf trace [<options>] report <script> [script-args]",
+	"perf trace [<options>] <script> [<record-options>] <command>",
+	"perf trace [<options>] <top-script> [script-args]",
 	NULL
 };
 
@@ -567,47 +600,81 @@ static const struct option options[] = {
 	OPT_END()
 };
 
+static bool have_cmd(int argc, const char **argv)
+{
+	char **__argv = malloc(sizeof(const char *) * argc);
+
+	if (!__argv)
+		die("malloc");
+	memcpy(__argv, argv, sizeof(const char *) * argc);
+	argc = parse_options(argc, (const char **)__argv, record_options,
+			     NULL, PARSE_OPT_STOP_AT_NON_OPTION);
+	free(__argv);
+
+	return argc != 0;
+}
+
 int cmd_trace(int argc, const char **argv, const char *prefix __used)
 {
+	char *rec_script_path = NULL;
+	char *rep_script_path = NULL;
 	struct perf_session *session;
-	const char *suffix = NULL;
+	char *script_path = NULL;
 	const char **__argv;
-	char *script_path;
-	int i, err;
+	bool system_wide;
+	int i, j, err;
 
-	if (argc >= 2 && strncmp(argv[1], "rec", strlen("rec")) == 0) {
-		if (argc < 3) {
-			fprintf(stderr,
-				"Please specify a record script\n");
-			return -1;
-		}
-		suffix = RECORD_SUFFIX;
+	setup_scripting();
+
+	argc = parse_options(argc, argv, options, trace_usage,
+			     PARSE_OPT_STOP_AT_NON_OPTION);
+
+	if (argc > 1 && !strncmp(argv[0], "rec", strlen("rec"))) {
+		rec_script_path = get_script_path(argv[1], RECORD_SUFFIX);
+		if (!rec_script_path)
+			return cmd_record(argc, argv, NULL);
 	}
 
-	if (argc >= 2 && strncmp(argv[1], "rep", strlen("rep")) == 0) {
-		if (argc < 3) {
+	if (argc > 1 && !strncmp(argv[0], "rep", strlen("rep"))) {
+		rep_script_path = get_script_path(argv[1], REPORT_SUFFIX);
+		if (!rep_script_path) {
 			fprintf(stderr,
-				"Please specify a report script\n");
+				"Please specify a valid report script"
+				"(see 'perf trace -l' for listing)\n");
 			return -1;
 		}
-		suffix = REPORT_SUFFIX;
 	}
 
-	if (!suffix && argc >= 2 && strncmp(argv[1], "-", strlen("-")) != 0) {
-		char *record_script_path, *report_script_path;
+	/* make sure PERF_EXEC_PATH is set for scripts */
+	perf_set_argv_exec_path(perf_exec_path());
+
+	if (argc && !script_name && !rec_script_path && !rep_script_path) {
 		int live_pipe[2];
+		int rep_args;
 		pid_t pid;
 
-		record_script_path = get_script_path(argv[1], RECORD_SUFFIX);
-		if (!record_script_path) {
-			fprintf(stderr, "record script not found\n");
-			return -1;
+		rec_script_path = get_script_path(argv[0], RECORD_SUFFIX);
+		rep_script_path = get_script_path(argv[0], REPORT_SUFFIX);
+
+		if (!rec_script_path && !rep_script_path) {
+			fprintf(stderr, " Couldn't find script %s\n\n See perf"
+				" trace -l for available scripts.\n", argv[0]);
+			usage_with_options(trace_usage, options);
 		}
 
-		report_script_path = get_script_path(argv[1], REPORT_SUFFIX);
-		if (!report_script_path) {
-			fprintf(stderr, "report script not found\n");
-			return -1;
+		if (is_top_script(argv[0])) {
+			rep_args = argc - 1;
+		} else {
+			int rec_args;
+
+			rep_args = has_required_arg(rep_script_path);
+			rec_args = (argc - 1) - rep_args;
+			if (rec_args < 0) {
+				fprintf(stderr, " %s script requires options."
+					"\n\n See perf trace -l for available "
+					"scripts and options.\n", argv[0]);
+				usage_with_options(trace_usage, options);
+			}
 		}
 
 		if (pipe(live_pipe) < 0) {
@@ -622,59 +689,84 @@ int cmd_trace(int argc, const char **argv, const char *prefix __used)
 		}
 
 		if (!pid) {
+			system_wide = true;
+			j = 0;
+
 			dup2(live_pipe[1], 1);
 			close(live_pipe[0]);
 
-			__argv = malloc(5 * sizeof(const char *));
-			__argv[0] = "/bin/sh";
-			__argv[1] = record_script_path;
-			__argv[2] = "-o";
-			__argv[3] = "-";
-			__argv[4] = NULL;
+			if (!is_top_script(argv[0]))
+				system_wide = !have_cmd(argc - rep_args,
+							&argv[rep_args]);
+
+			__argv = malloc((argc + 6) * sizeof(const char *));
+			if (!__argv)
+				die("malloc");
+
+			__argv[j++] = "/bin/sh";
+			__argv[j++] = rec_script_path;
+			if (system_wide)
+				__argv[j++] = "-a";
+			__argv[j++] = "-q";
+			__argv[j++] = "-o";
+			__argv[j++] = "-";
+			for (i = rep_args + 1; i < argc; i++)
+				__argv[j++] = argv[i];
+			__argv[j++] = NULL;
 
 			execvp("/bin/sh", (char **)__argv);
+			free(__argv);
 			exit(-1);
 		}
 
 		dup2(live_pipe[0], 0);
 		close(live_pipe[1]);
 
-		__argv = malloc((argc + 3) * sizeof(const char *));
-		__argv[0] = "/bin/sh";
-		__argv[1] = report_script_path;
-		for (i = 2; i < argc; i++)
-			__argv[i] = argv[i];
-		__argv[i++] = "-i";
-		__argv[i++] = "-";
-		__argv[i++] = NULL;
+		__argv = malloc((argc + 4) * sizeof(const char *));
+		if (!__argv)
+			die("malloc");
+		j = 0;
+		__argv[j++] = "/bin/sh";
+		__argv[j++] = rep_script_path;
+		for (i = 1; i < rep_args + 1; i++)
+			__argv[j++] = argv[i];
+		__argv[j++] = "-i";
+		__argv[j++] = "-";
+		__argv[j++] = NULL;
 
 		execvp("/bin/sh", (char **)__argv);
+		free(__argv);
 		exit(-1);
 	}
 
-	if (suffix) {
-		script_path = get_script_path(argv[2], suffix);
-		if (!script_path) {
-			fprintf(stderr, "script not found\n");
-			return -1;
-		}
-
-		__argv = malloc((argc + 1) * sizeof(const char *));
-		__argv[0] = "/bin/sh";
-		__argv[1] = script_path;
-		for (i = 3; i < argc; i++)
-			__argv[i - 1] = argv[i];
-		__argv[argc - 1] = NULL;
+	if (rec_script_path)
+		script_path = rec_script_path;
+	if (rep_script_path)
+		script_path = rep_script_path;
+
+	if (script_path) {
+		system_wide = false;
+		j = 0;
+
+		if (rec_script_path)
+			system_wide = !have_cmd(argc - 1, &argv[1]);
+
+		__argv = malloc((argc + 2) * sizeof(const char *));
+		if (!__argv)
+			die("malloc");
+		__argv[j++] = "/bin/sh";
+		__argv[j++] = script_path;
+		if (system_wide)
+			__argv[j++] = "-a";
+		for (i = 2; i < argc; i++)
+			__argv[j++] = argv[i];
+		__argv[j++] = NULL;
 
 		execvp("/bin/sh", (char **)__argv);
+		free(__argv);
 		exit(-1);
 	}
 
-	setup_scripting();
-
-	argc = parse_options(argc, argv, options, trace_usage,
-			     PARSE_OPT_STOP_AT_NON_OPTION);
-
 	if (symbol__init() < 0)
 		return -1;
 	if (!script_name)
diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-record b/tools/perf/scripts/perl/bin/failed-syscalls-record
index eb5846bcb565..8104895a7b67 100644
--- a/tools/perf/scripts/perl/bin/failed-syscalls-record
+++ b/tools/perf/scripts/perl/bin/failed-syscalls-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_exit $@
+perf record -e raw_syscalls:sys_exit $@
diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-report b/tools/perf/scripts/perl/bin/failed-syscalls-report
index e3a5e55d54ff..4028d92dc4ae 100644
--- a/tools/perf/scripts/perl/bin/failed-syscalls-report
+++ b/tools/perf/scripts/perl/bin/failed-syscalls-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
 	shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/failed-syscalls.pl $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/failed-syscalls.pl $comm
diff --git a/tools/perf/scripts/perl/bin/rw-by-file-record b/tools/perf/scripts/perl/bin/rw-by-file-record
index 5bfaae5a6cba..33efc8673aae 100644
--- a/tools/perf/scripts/perl/bin/rw-by-file-record
+++ b/tools/perf/scripts/perl/bin/rw-by-file-record
@@ -1,3 +1,3 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_enter_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_enter_write $@
 
diff --git a/tools/perf/scripts/perl/bin/rw-by-file-report b/tools/perf/scripts/perl/bin/rw-by-file-report
index d83070b7eeb5..ba25f4d41fb0 100644
--- a/tools/perf/scripts/perl/bin/rw-by-file-report
+++ b/tools/perf/scripts/perl/bin/rw-by-file-report
@@ -7,7 +7,7 @@ if [ $# -lt 1 ] ; then
 fi
 comm=$1
 shift
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rw-by-file.pl $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rw-by-file.pl $comm
 
 
 
diff --git a/tools/perf/scripts/perl/bin/rw-by-pid-record b/tools/perf/scripts/perl/bin/rw-by-pid-record
index 6e0b2f7755ac..7cb9db230448 100644
--- a/tools/perf/scripts/perl/bin/rw-by-pid-record
+++ b/tools/perf/scripts/perl/bin/rw-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
diff --git a/tools/perf/scripts/perl/bin/rw-by-pid-report b/tools/perf/scripts/perl/bin/rw-by-pid-report
index 7ef46983f62f..641a3f5d085c 100644
--- a/tools/perf/scripts/perl/bin/rw-by-pid-report
+++ b/tools/perf/scripts/perl/bin/rw-by-pid-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: system-wide r/w activity
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rw-by-pid.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rw-by-pid.pl
 
 
 
diff --git a/tools/perf/scripts/perl/bin/rwtop-record b/tools/perf/scripts/perl/bin/rwtop-record
index 6e0b2f7755ac..7cb9db230448 100644
--- a/tools/perf/scripts/perl/bin/rwtop-record
+++ b/tools/perf/scripts/perl/bin/rwtop-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
diff --git a/tools/perf/scripts/perl/bin/rwtop-report b/tools/perf/scripts/perl/bin/rwtop-report
index 93e698cd3f38..4918dba77021 100644
--- a/tools/perf/scripts/perl/bin/rwtop-report
+++ b/tools/perf/scripts/perl/bin/rwtop-report
@@ -17,7 +17,7 @@ if [ "$n_args" -gt 0 ] ; then
     interval=$1
     shift
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rwtop.pl $interval
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rwtop.pl $interval
 
 
 
diff --git a/tools/perf/scripts/perl/bin/wakeup-latency-record b/tools/perf/scripts/perl/bin/wakeup-latency-record
index 9f2acaaae9f0..464251a1bd7e 100644
--- a/tools/perf/scripts/perl/bin/wakeup-latency-record
+++ b/tools/perf/scripts/perl/bin/wakeup-latency-record
@@ -1,5 +1,5 @@
 #!/bin/bash
-perf record -a -e sched:sched_switch -e sched:sched_wakeup $@
+perf record -e sched:sched_switch -e sched:sched_wakeup $@
 
 
 
diff --git a/tools/perf/scripts/perl/bin/wakeup-latency-report b/tools/perf/scripts/perl/bin/wakeup-latency-report
index a0d898f9ca1d..49052ebcb632 100644
--- a/tools/perf/scripts/perl/bin/wakeup-latency-report
+++ b/tools/perf/scripts/perl/bin/wakeup-latency-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: system-wide min/max/avg wakeup latency
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/wakeup-latency.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/wakeup-latency.pl
 
 
 
diff --git a/tools/perf/scripts/perl/bin/workqueue-stats-record b/tools/perf/scripts/perl/bin/workqueue-stats-record
index 85301f2471ff..8edda9078d5d 100644
--- a/tools/perf/scripts/perl/bin/workqueue-stats-record
+++ b/tools/perf/scripts/perl/bin/workqueue-stats-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e workqueue:workqueue_creation -e workqueue:workqueue_destruction -e workqueue:workqueue_execution -e workqueue:workqueue_insertion $@
+perf record -e workqueue:workqueue_creation -e workqueue:workqueue_destruction -e workqueue:workqueue_execution -e workqueue:workqueue_insertion $@
diff --git a/tools/perf/scripts/perl/bin/workqueue-stats-report b/tools/perf/scripts/perl/bin/workqueue-stats-report
index 35081132ef97..df0c65f4ca93 100644
--- a/tools/perf/scripts/perl/bin/workqueue-stats-report
+++ b/tools/perf/scripts/perl/bin/workqueue-stats-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: workqueue stats (ins/exe/create/destroy)
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/workqueue-stats.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/workqueue-stats.pl
 
 
 
diff --git a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
index 9689bc0acd9f..13cc02b5893a 100644
--- a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
+++ b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
@@ -6,6 +6,14 @@
 # Public License ("GPL") version 2 as published by the Free Software
 # Foundation.
 
+import errno, os
+
+FUTEX_WAIT = 0
+FUTEX_WAKE = 1
+FUTEX_PRIVATE_FLAG = 128
+FUTEX_CLOCK_REALTIME = 256
+FUTEX_CMD_MASK = ~(FUTEX_PRIVATE_FLAG | FUTEX_CLOCK_REALTIME)
+
 NSECS_PER_SEC    = 1000000000
 
 def avg(total, n):
@@ -24,5 +32,55 @@ def nsecs_str(nsecs):
     str = "%5u.%09u" % (nsecs_secs(nsecs), nsecs_nsecs(nsecs)),
     return str
 
+def add_stats(dict, key, value):
+	if not dict.has_key(key):
+		dict[key] = (value, value, value, 1)
+	else:
+		min, max, avg, count = dict[key]
+		if value < min:
+			min = value
+		if value > max:
+			max = value
+		avg = (avg + value) / 2
+		dict[key] = (min, max, avg, count + 1)
+
 def clear_term():
     print("\x1b[H\x1b[2J")
+
+audit_package_warned = False
+
+try:
+	import audit
+	machine_to_id = {
+		'x86_64': audit.MACH_86_64,
+		'alpha'	: audit.MACH_ALPHA,
+		'ia64'	: audit.MACH_IA64,
+		'ppc'	: audit.MACH_PPC,
+		'ppc64'	: audit.MACH_PPC64,
+		's390'	: audit.MACH_S390,
+		's390x'	: audit.MACH_S390X,
+		'i386'	: audit.MACH_X86,
+		'i586'	: audit.MACH_X86,
+		'i686'	: audit.MACH_X86,
+	}
+	try:
+		machine_to_id['armeb'] = audit.MACH_ARMEB
+	except:
+		pass
+	machine_id = machine_to_id[os.uname()[4]]
+except:
+	if not audit_package_warned:
+		audit_package_warned = True
+		print "Install the audit-libs-python package to get syscall names"
+
+def syscall_name(id):
+	try:
+		return audit.audit_syscall_to_name(id, machine_id)
+	except:
+		return str(id)
+
+def strerror(nr):
+	try:
+		return errno.errorcode[abs(nr)]
+	except:
+		return "Unknown %d errno" % nr
diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
index eb5846bcb565..8104895a7b67 100644
--- a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
+++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_exit $@
+perf record -e raw_syscalls:sys_exit $@
diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
index 30293545fcc2..03587021463d 100644
--- a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
+++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
 	shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/failed-syscalls-by-pid.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/failed-syscalls-by-pid.py $comm
diff --git a/tools/perf/scripts/python/bin/futex-contention-record b/tools/perf/scripts/python/bin/futex-contention-record
new file mode 100644
index 000000000000..b1495c9a9b20
--- /dev/null
+++ b/tools/perf/scripts/python/bin/futex-contention-record
@@ -0,0 +1,2 @@
+#!/bin/bash
+perf record -e syscalls:sys_enter_futex -e syscalls:sys_exit_futex $@
diff --git a/tools/perf/scripts/python/bin/futex-contention-report b/tools/perf/scripts/python/bin/futex-contention-report
new file mode 100644
index 000000000000..c8268138fb7e
--- /dev/null
+++ b/tools/perf/scripts/python/bin/futex-contention-report
@@ -0,0 +1,4 @@
+#!/bin/bash
+# description: futext contention measurement
+
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/futex-contention.py
diff --git a/tools/perf/scripts/python/bin/netdev-times-record b/tools/perf/scripts/python/bin/netdev-times-record
index d931a828126b..558754b840a9 100644
--- a/tools/perf/scripts/python/bin/netdev-times-record
+++ b/tools/perf/scripts/python/bin/netdev-times-record
@@ -1,5 +1,5 @@
 #!/bin/bash
-perf record -a -e net:net_dev_xmit -e net:net_dev_queue		\
+perf record -e net:net_dev_xmit -e net:net_dev_queue		\
 		-e net:netif_receive_skb -e net:netif_rx		\
 		-e skb:consume_skb -e skb:kfree_skb			\
 		-e skb:skb_copy_datagram_iovec -e napi:napi_poll	\
diff --git a/tools/perf/scripts/python/bin/netdev-times-report b/tools/perf/scripts/python/bin/netdev-times-report
index c3d0a638123d..4ad361b31249 100644
--- a/tools/perf/scripts/python/bin/netdev-times-report
+++ b/tools/perf/scripts/python/bin/netdev-times-report
@@ -2,4 +2,4 @@
 # description: display a process of packet and processing time
 # args: [tx] [rx] [dev=] [debug]
 
-perf trace -s ~/libexec/perf-core/scripts/python/netdev-times.py $@
+perf trace -s "$PERF_EXEC_PATH"/scripts/python/netdev-times.py $@
diff --git a/tools/perf/scripts/python/bin/sched-migration-record b/tools/perf/scripts/python/bin/sched-migration-record
index 17a3e9bd9e8f..7493fddbe995 100644
--- a/tools/perf/scripts/python/bin/sched-migration-record
+++ b/tools/perf/scripts/python/bin/sched-migration-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -m 16384 -a -e sched:sched_wakeup -e sched:sched_wakeup_new -e sched:sched_switch -e sched:sched_migrate_task $@
+perf record -m 16384 -e sched:sched_wakeup -e sched:sched_wakeup_new -e sched:sched_switch -e sched:sched_migrate_task $@
diff --git a/tools/perf/scripts/python/bin/sched-migration-report b/tools/perf/scripts/python/bin/sched-migration-report
index 61d05f72e443..df1791f07c24 100644
--- a/tools/perf/scripts/python/bin/sched-migration-report
+++ b/tools/perf/scripts/python/bin/sched-migration-report
@@ -1,3 +1,3 @@
 #!/bin/bash
 # description: sched migration overview
-perf trace $@ -s ~/libexec/perf-core/scripts/python/sched-migration.py
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/sched-migration.py
diff --git a/tools/perf/scripts/python/bin/sctop-record b/tools/perf/scripts/python/bin/sctop-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/sctop-record
+++ b/tools/perf/scripts/python/bin/sctop-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/sctop-report b/tools/perf/scripts/python/bin/sctop-report
index b01c842ae7b4..36b409c05e50 100644
--- a/tools/perf/scripts/python/bin/sctop-report
+++ b/tools/perf/scripts/python/bin/sctop-report
@@ -21,4 +21,4 @@ elif [ "$n_args" -gt 0 ] ; then
     interval=$1
     shift
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/sctop.py $comm $interval
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/sctop.py $comm $interval
diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-record b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
+++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-report b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
index 9e9d8ddd72ce..4eb88c9fc83c 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
+++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
 	shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/syscall-counts-by-pid.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/syscall-counts-by-pid.py $comm
diff --git a/tools/perf/scripts/python/bin/syscall-counts-record b/tools/perf/scripts/python/bin/syscall-counts-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-record
+++ b/tools/perf/scripts/python/bin/syscall-counts-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/syscall-counts-report b/tools/perf/scripts/python/bin/syscall-counts-report
index dc076b618796..cb2f9c5cf17e 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-report
+++ b/tools/perf/scripts/python/bin/syscall-counts-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
 	shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/syscall-counts.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/syscall-counts.py $comm
diff --git a/tools/perf/scripts/python/failed-syscalls-by-pid.py b/tools/perf/scripts/python/failed-syscalls-by-pid.py
index 0ca02278fe69..acd7848717b3 100644
--- a/tools/perf/scripts/python/failed-syscalls-by-pid.py
+++ b/tools/perf/scripts/python/failed-syscalls-by-pid.py
@@ -13,21 +13,26 @@ sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 
 from perf_trace_context import *
 from Core import *
+from Util import *
 
-usage = "perf trace -s syscall-counts-by-pid.py [comm]\n";
+usage = "perf trace -s syscall-counts-by-pid.py [comm|pid]\n";
 
 for_comm = None
+for_pid = None
 
 if len(sys.argv) > 2:
 	sys.exit(usage)
 
 if len(sys.argv) > 1:
-	for_comm = sys.argv[1]
+	try:
+		for_pid = int(sys.argv[1])
+	except:
+		for_comm = sys.argv[1]
 
 syscalls = autodict()
 
 def trace_begin():
-	pass
+	print "Press control+C to stop and show the summary"
 
 def trace_end():
 	print_error_totals()
@@ -35,9 +40,9 @@ def trace_end():
 def raw_syscalls__sys_exit(event_name, context, common_cpu,
 	common_secs, common_nsecs, common_pid, common_comm,
 	id, ret):
-	if for_comm is not None:
-		if common_comm != for_comm:
-			return
+	if (for_comm and common_comm != for_comm) or \
+	   (for_pid  and common_pid  != for_pid ):
+		return
 
 	if ret < 0:
 		try:
@@ -62,7 +67,7 @@ def print_error_totals():
 		    print "\n%s [%d]\n" % (comm, pid),
 		    id_keys = syscalls[comm][pid].keys()
 		    for id in id_keys:
-			    print "  syscall: %-16d\n" % (id),
+			    print "  syscall: %-16s\n" % syscall_name(id),
 			    ret_keys = syscalls[comm][pid][id].keys()
 			    for ret, val in sorted(syscalls[comm][pid][id].iteritems(), key = lambda(k, v): (v, k),  reverse = True):
-				    print "    err = %-20d  %10d\n" % (ret, val),
+				    print "    err = %-20s  %10d\n" % (strerror(ret), val),
diff --git a/tools/perf/scripts/python/futex-contention.py b/tools/perf/scripts/python/futex-contention.py
new file mode 100644
index 000000000000..11e70a388d41
--- /dev/null
+++ b/tools/perf/scripts/python/futex-contention.py
@@ -0,0 +1,50 @@
+# futex contention
+# (c) 2010, Arnaldo Carvalho de Melo <acme@redhat.com>
+# Licensed under the terms of the GNU GPL License version 2
+#
+# Translation of:
+#
+# http://sourceware.org/systemtap/wiki/WSFutexContention
+#
+# to perf python scripting.
+#
+# Measures futex contention
+
+import os, sys
+sys.path.append(os.environ['PERF_EXEC_PATH'] + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
+from Util import *
+
+process_names = {}
+thread_thislock = {}
+thread_blocktime = {}
+
+lock_waits = {} # long-lived stats on (tid,lock) blockage elapsed time
+process_names = {} # long-lived pid-to-execname mapping
+
+def syscalls__sys_enter_futex(event, ctxt, cpu, s, ns, tid, comm,
+			      nr, uaddr, op, val, utime, uaddr2, val3):
+	cmd = op & FUTEX_CMD_MASK
+	if cmd != FUTEX_WAIT:
+		return # we don't care about originators of WAKE events
+
+	process_names[tid] = comm
+	thread_thislock[tid] = uaddr
+	thread_blocktime[tid] = nsecs(s, ns)
+
+def syscalls__sys_exit_futex(event, ctxt, cpu, s, ns, tid, comm,
+			     nr, ret):
+	if thread_blocktime.has_key(tid):
+		elapsed = nsecs(s, ns) - thread_blocktime[tid]
+		add_stats(lock_waits, (tid, thread_thislock[tid]), elapsed)
+		del thread_blocktime[tid]
+		del thread_thislock[tid]
+
+def trace_begin():
+	print "Press control+C to stop and show the summary"
+
+def trace_end():
+	for (tid, lock) in lock_waits:
+		min, max, avg, count = lock_waits[tid, lock]
+		print "%s[%d] lock %x contended %d times, %d avg ns" % \
+		      (process_names[tid], tid, lock, count, avg)
+
diff --git a/tools/perf/scripts/python/sctop.py b/tools/perf/scripts/python/sctop.py
index 6cafad40c296..7a6ec2c7d8ab 100644
--- a/tools/perf/scripts/python/sctop.py
+++ b/tools/perf/scripts/python/sctop.py
@@ -8,10 +8,7 @@
 # will be refreshed every [interval] seconds.  The default interval is
 # 3 seconds.
 
-import thread
-import time
-import os
-import sys
+import os, sys, thread, time
 
 sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 	'/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
@@ -20,7 +17,7 @@ from perf_trace_context import *
 from Core import *
 from Util import *
 
-usage = "perf trace -s syscall-counts.py [comm] [interval]\n";
+usage = "perf trace -s sctop.py [comm] [interval]\n";
 
 for_comm = None
 default_interval = 3
@@ -71,7 +68,7 @@ def print_syscall_totals(interval):
 		for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \
 					      reverse = True):
 			try:
-				print "%-40d  %10d\n" % (id, val),
+				print "%-40s  %10d\n" % (syscall_name(id), val),
 			except TypeError:
 				pass
 		syscalls.clear()
diff --git a/tools/perf/scripts/python/syscall-counts-by-pid.py b/tools/perf/scripts/python/syscall-counts-by-pid.py
index af722d6a4b3f..d1ee3ec10cf2 100644
--- a/tools/perf/scripts/python/syscall-counts-by-pid.py
+++ b/tools/perf/scripts/python/syscall-counts-by-pid.py
@@ -5,29 +5,33 @@
 # Displays system-wide system call totals, broken down by syscall.
 # If a [comm] arg is specified, only syscalls called by [comm] are displayed.
 
-import os
-import sys
+import os, sys
 
 sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 	'/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
 
 from perf_trace_context import *
 from Core import *
+from Util import syscall_name
 
 usage = "perf trace -s syscall-counts-by-pid.py [comm]\n";
 
 for_comm = None
+for_pid = None
 
 if len(sys.argv) > 2:
 	sys.exit(usage)
 
 if len(sys.argv) > 1:
-	for_comm = sys.argv[1]
+	try:
+		for_pid = int(sys.argv[1])
+	except:
+		for_comm = sys.argv[1]
 
 syscalls = autodict()
 
 def trace_begin():
-	pass
+	print "Press control+C to stop and show the summary"
 
 def trace_end():
 	print_syscall_totals()
@@ -35,9 +39,10 @@ def trace_end():
 def raw_syscalls__sys_enter(event_name, context, common_cpu,
 	common_secs, common_nsecs, common_pid, common_comm,
 	id, args):
-	if for_comm is not None:
-		if common_comm != for_comm:
-			return
+
+	if (for_comm and common_comm != for_comm) or \
+	   (for_pid  and common_pid  != for_pid ):
+		return
 	try:
 		syscalls[common_comm][common_pid][id] += 1
 	except TypeError:
@@ -61,4 +66,4 @@ def print_syscall_totals():
 		    id_keys = syscalls[comm][pid].keys()
 		    for id, val in sorted(syscalls[comm][pid].iteritems(), \
 				  key = lambda(k, v): (v, k),  reverse = True):
-			    print "  %-38d  %10d\n" % (id, val),
+			    print "  %-38s  %10d\n" % (syscall_name(id), val),
diff --git a/tools/perf/scripts/python/syscall-counts.py b/tools/perf/scripts/python/syscall-counts.py
index f977e85ff049..ea183dc82d29 100644
--- a/tools/perf/scripts/python/syscall-counts.py
+++ b/tools/perf/scripts/python/syscall-counts.py
@@ -13,6 +13,7 @@ sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 
 from perf_trace_context import *
 from Core import *
+from Util import syscall_name
 
 usage = "perf trace -s syscall-counts.py [comm]\n";
 
@@ -27,7 +28,7 @@ if len(sys.argv) > 1:
 syscalls = autodict()
 
 def trace_begin():
-	pass
+	print "Press control+C to stop and show the summary"
 
 def trace_end():
 	print_syscall_totals()
@@ -55,4 +56,4 @@ def print_syscall_totals():
 
     for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \
 				  reverse = True):
-	    print "%-40d  %10d\n" % (id, val),
+	    print "%-40s  %10d\n" % (syscall_name(id), val),
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index f9c7e3ad1aa7..c8d81b00089d 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -12,8 +12,8 @@
 #include "debug.h"
 #include "util.h"
 
-int verbose = 0;
-bool dump_trace = false;
+int verbose;
+bool dump_trace = false, quiet = false;
 
 int eprintf(int level, const char *fmt, ...)
 {
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index 7a17ee061bcb..7b514082bbaf 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -6,7 +6,7 @@
 #include "event.h"
 
 extern int verbose;
-extern bool dump_trace;
+extern bool quiet, dump_trace;
 
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
 void trace_event(event_t *event);
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index d7e67b167ea3..64a85bafde63 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -946,11 +946,16 @@ perf_header__find_attr(u64 id, struct perf_header *header)
 
 	/*
 	 * We set id to -1 if the data file doesn't contain sample
-	 * ids. Check for this and avoid walking through the entire
-	 * list of ids which may be large.
+	 * ids. This can happen when the data file contains one type
+	 * of event and in that case, the header can still store the
+	 * event attribute information. Check for this and avoid
+	 * walking through the entire list of ids which may be large.
 	 */
-	if (id == -1ULL)
+	if (id == -1ULL) {
+		if (header->attrs > 0)
+			return &header->attr[0]->attr;
 		return NULL;
+	}
 
 	for (i = 0; i < header->attrs; i++) {
 		struct perf_header_attr *attr = header->attr[i];
diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h
index 78575796d5f3..b397c0383728 100644
--- a/tools/perf/util/map.h
+++ b/tools/perf/util/map.h
@@ -215,6 +215,16 @@ struct symbol *map_groups__find_function_by_name(struct map_groups *self,
 	return map_groups__find_symbol_by_name(self, MAP__FUNCTION, name, mapp, filter);
 }
 
+static inline
+struct symbol *machine__find_kernel_function_by_name(struct machine *self,
+						     const char *name,
+						     struct map **mapp,
+						     symbol_filter_t filter)
+{
+	return map_groups__find_function_by_name(&self->kmaps, name, mapp,
+						 filter);
+}
+
 int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
 				   int verbose, FILE *fp);
 
diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
index fcc16e4349df..3b6a5297bf16 100644
--- a/tools/perf/util/probe-event.c
+++ b/tools/perf/util/probe-event.c
@@ -74,10 +74,9 @@ static int e_snprintf(char *str, size_t size, const char *format, ...)
 static char *synthesize_perf_probe_point(struct perf_probe_point *pp);
 static struct machine machine;
 
-/* Initialize symbol maps and path of vmlinux */
+/* Initialize symbol maps and path of vmlinux/modules */
 static int init_vmlinux(void)
 {
-	struct dso *kernel;
 	int ret;
 
 	symbol_conf.sort_by_name = true;
@@ -91,33 +90,61 @@ static int init_vmlinux(void)
 		goto out;
 	}
 
-	ret = machine__init(&machine, "/", 0);
+	ret = machine__init(&machine, "", HOST_KERNEL_ID);
 	if (ret < 0)
 		goto out;
 
-	kernel = dso__new_kernel(symbol_conf.vmlinux_name);
-	if (kernel == NULL)
-		die("Failed to create kernel dso.");
-
-	ret = __machine__create_kernel_maps(&machine, kernel);
-	if (ret < 0)
-		pr_debug("Failed to create kernel maps.\n");
-
+	if (machine__create_kernel_maps(&machine) < 0) {
+		pr_debug("machine__create_kernel_maps ");
+		goto out;
+	}
 out:
 	if (ret < 0)
 		pr_warning("Failed to init vmlinux path.\n");
 	return ret;
 }
 
+static struct symbol *__find_kernel_function_by_name(const char *name,
+						     struct map **mapp)
+{
+	return machine__find_kernel_function_by_name(&machine, name, mapp,
+						     NULL);
+}
+
+const char *kernel_get_module_path(const char *module)
+{
+	struct dso *dso;
+
+	if (module) {
+		list_for_each_entry(dso, &machine.kernel_dsos, node) {
+			if (strncmp(dso->short_name + 1, module,
+				    dso->short_name_len - 2) == 0)
+				goto found;
+		}
+		pr_debug("Failed to find module %s.\n", module);
+		return NULL;
+	} else {
+		dso = machine.vmlinux_maps[MAP__FUNCTION]->dso;
+		if (dso__load_vmlinux_path(dso,
+			 machine.vmlinux_maps[MAP__FUNCTION], NULL) < 0) {
+			pr_debug("Failed to load kernel map.\n");
+			return NULL;
+		}
+	}
+found:
+	return dso->long_name;
+}
+
 #ifdef DWARF_SUPPORT
-static int open_vmlinux(void)
+static int open_vmlinux(const char *module)
 {
-	if (map__load(machine.vmlinux_maps[MAP__FUNCTION], NULL) < 0) {
-		pr_debug("Failed to load kernel map.\n");
-		return -EINVAL;
+	const char *path = kernel_get_module_path(module);
+	if (!path) {
+		pr_err("Failed to find path of %s module", module ?: "kernel");
+		return -ENOENT;
 	}
-	pr_debug("Try to open %s\n", machine.vmlinux_maps[MAP__FUNCTION]->dso->long_name);
-	return open(machine.vmlinux_maps[MAP__FUNCTION]->dso->long_name, O_RDONLY);
+	pr_debug("Try to open %s\n", path);
+	return open(path, O_RDONLY);
 }
 
 /*
@@ -125,20 +152,19 @@ static int open_vmlinux(void)
  * Currently only handles kprobes.
  */
 static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
-				       struct perf_probe_point *pp)
+					struct perf_probe_point *pp)
 {
 	struct symbol *sym;
-	int fd, ret = -ENOENT;
+	struct map *map;
+	u64 addr;
+	int ret = -ENOENT;
 
-	sym = map__find_symbol_by_name(machine.vmlinux_maps[MAP__FUNCTION],
-				       tp->symbol, NULL);
+	sym = __find_kernel_function_by_name(tp->symbol, &map);
 	if (sym) {
-		fd = open_vmlinux();
-		if (fd >= 0) {
-			ret = find_perf_probe_point(fd,
-						 sym->start + tp->offset, pp);
-			close(fd);
-		}
+		addr = map->unmap_ip(map, sym->start + tp->offset);
+		pr_debug("try to find %s+%ld@%llx\n", tp->symbol,
+			 tp->offset, addr);
+		ret = find_perf_probe_point((unsigned long)addr, pp);
 	}
 	if (ret <= 0) {
 		pr_debug("Failed to find corresponding probes from "
@@ -156,12 +182,12 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
 /* Try to find perf_probe_event with debuginfo */
 static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
 					   struct probe_trace_event **tevs,
-					   int max_tevs)
+					   int max_tevs, const char *module)
 {
 	bool need_dwarf = perf_probe_event_need_dwarf(pev);
 	int fd, ntevs;
 
-	fd = open_vmlinux();
+	fd = open_vmlinux(module);
 	if (fd < 0) {
 		if (need_dwarf) {
 			pr_warning("Failed to open debuginfo file.\n");
@@ -300,7 +326,7 @@ error:
  * Show line-range always requires debuginfo to find source file and
  * line number.
  */
-int show_line_range(struct line_range *lr)
+int show_line_range(struct line_range *lr, const char *module)
 {
 	int l = 1;
 	struct line_node *ln;
@@ -313,7 +339,7 @@ int show_line_range(struct line_range *lr)
 	if (ret < 0)
 		return ret;
 
-	fd = open_vmlinux();
+	fd = open_vmlinux(module);
 	if (fd < 0) {
 		pr_warning("Failed to open debuginfo file.\n");
 		return fd;
@@ -378,11 +404,84 @@ end:
 	return ret;
 }
 
+static int show_available_vars_at(int fd, struct perf_probe_event *pev,
+				  int max_vls, bool externs)
+{
+	char *buf;
+	int ret, i;
+	struct str_node *node;
+	struct variable_list *vls = NULL, *vl;
+
+	buf = synthesize_perf_probe_point(&pev->point);
+	if (!buf)
+		return -EINVAL;
+	pr_debug("Searching variables at %s\n", buf);
+
+	ret = find_available_vars_at(fd, pev, &vls, max_vls, externs);
+	if (ret > 0) {
+		/* Some variables were found */
+		fprintf(stdout, "Available variables at %s\n", buf);
+		for (i = 0; i < ret; i++) {
+			vl = &vls[i];
+			/*
+			 * A probe point might be converted to
+			 * several trace points.
+			 */
+			fprintf(stdout, "\t@<%s+%lu>\n", vl->point.symbol,
+				vl->point.offset);
+			free(vl->point.symbol);
+			if (vl->vars) {
+				strlist__for_each(node, vl->vars)
+					fprintf(stdout, "\t\t%s\n", node->s);
+				strlist__delete(vl->vars);
+			} else
+				fprintf(stdout, "(No variables)\n");
+		}
+		free(vls);
+	} else
+		pr_err("Failed to find variables at %s (%d)\n", buf, ret);
+
+	free(buf);
+	return ret;
+}
+
+/* Show available variables on given probe point */
+int show_available_vars(struct perf_probe_event *pevs, int npevs,
+			int max_vls, const char *module, bool externs)
+{
+	int i, fd, ret = 0;
+
+	ret = init_vmlinux();
+	if (ret < 0)
+		return ret;
+
+	fd = open_vmlinux(module);
+	if (fd < 0) {
+		pr_warning("Failed to open debuginfo file.\n");
+		return fd;
+	}
+
+	setup_pager();
+
+	for (i = 0; i < npevs && ret >= 0; i++)
+		ret = show_available_vars_at(fd, &pevs[i], max_vls, externs);
+
+	close(fd);
+	return ret;
+}
+
 #else	/* !DWARF_SUPPORT */
 
 static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
-				       struct perf_probe_point *pp)
+					struct perf_probe_point *pp)
 {
+	struct symbol *sym;
+
+	sym = __find_kernel_function_by_name(tp->symbol, NULL);
+	if (!sym) {
+		pr_err("Failed to find symbol %s in kernel.\n", tp->symbol);
+		return -ENOENT;
+	}
 	pp->function = strdup(tp->symbol);
 	if (pp->function == NULL)
 		return -ENOMEM;
@@ -394,7 +493,7 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
 
 static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
 				struct probe_trace_event **tevs __unused,
-				int max_tevs __unused)
+				int max_tevs __unused, const char *mod __unused)
 {
 	if (perf_probe_event_need_dwarf(pev)) {
 		pr_warning("Debuginfo-analysis is not supported.\n");
@@ -403,12 +502,19 @@ static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
 	return 0;
 }
 
-int show_line_range(struct line_range *lr __unused)
+int show_line_range(struct line_range *lr __unused, const char *module __unused)
 {
 	pr_warning("Debuginfo-analysis is not supported.\n");
 	return -ENOSYS;
 }
 
+int show_available_vars(struct perf_probe_event *pevs __unused,
+			int npevs __unused, int max_vls __unused,
+			const char *module __unused, bool externs __unused)
+{
+	pr_warning("Debuginfo-analysis is not supported.\n");
+	return -ENOSYS;
+}
 #endif
 
 int parse_line_range_desc(const char *arg, struct line_range *lr)
@@ -1087,7 +1193,7 @@ error:
 }
 
 static int convert_to_perf_probe_event(struct probe_trace_event *tev,
-				struct perf_probe_event *pev)
+				       struct perf_probe_event *pev)
 {
 	char buf[64] = "";
 	int i, ret;
@@ -1516,14 +1622,14 @@ static int __add_probe_trace_events(struct perf_probe_event *pev,
 
 static int convert_to_probe_trace_events(struct perf_probe_event *pev,
 					  struct probe_trace_event **tevs,
-					  int max_tevs)
+					  int max_tevs, const char *module)
 {
 	struct symbol *sym;
 	int ret = 0, i;
 	struct probe_trace_event *tev;
 
 	/* Convert perf_probe_event with debuginfo */
-	ret = try_to_find_probe_trace_events(pev, tevs, max_tevs);
+	ret = try_to_find_probe_trace_events(pev, tevs, max_tevs, module);
 	if (ret != 0)
 		return ret;
 
@@ -1572,8 +1678,7 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev,
 	}
 
 	/* Currently just checking function name from symbol map */
-	sym = map__find_symbol_by_name(machine.vmlinux_maps[MAP__FUNCTION],
-				       tev->point.symbol, NULL);
+	sym = __find_kernel_function_by_name(tev->point.symbol, NULL);
 	if (!sym) {
 		pr_warning("Kernel symbol \'%s\' not found.\n",
 			   tev->point.symbol);
@@ -1596,7 +1701,7 @@ struct __event_package {
 };
 
 int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
-			  bool force_add, int max_tevs)
+			  int max_tevs, const char *module, bool force_add)
 {
 	int i, j, ret;
 	struct __event_package *pkgs;
@@ -1617,7 +1722,9 @@ int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
 		pkgs[i].pev = &pevs[i];
 		/* Convert with or without debuginfo */
 		ret  = convert_to_probe_trace_events(pkgs[i].pev,
-						      &pkgs[i].tevs, max_tevs);
+						     &pkgs[i].tevs,
+						     max_tevs,
+						     module);
 		if (ret < 0)
 			goto end;
 		pkgs[i].ntevs = ret;
diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h
index 5af39243a25b..5accbedfea37 100644
--- a/tools/perf/util/probe-event.h
+++ b/tools/perf/util/probe-event.h
@@ -90,6 +90,12 @@ struct line_range {
 	struct list_head	line_list;	/* Visible lines */
 };
 
+/* List of variables */
+struct variable_list {
+	struct probe_trace_point	point;	/* Actual probepoint */
+	struct strlist			*vars;	/* Available variables */
+};
+
 /* Command string to events */
 extern int parse_perf_probe_command(const char *cmd,
 				    struct perf_probe_event *pev);
@@ -109,12 +115,18 @@ extern void clear_perf_probe_event(struct perf_probe_event *pev);
 /* Command string to line-range */
 extern int parse_line_range_desc(const char *cmd, struct line_range *lr);
 
+/* Internal use: Return kernel/module path */
+extern const char *kernel_get_module_path(const char *module);
 
 extern int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
-				 bool force_add, int max_probe_points);
+				 int max_probe_points, const char *module,
+				 bool force_add);
 extern int del_perf_probe_events(struct strlist *dellist);
 extern int show_perf_probe_events(void);
-extern int show_line_range(struct line_range *lr);
+extern int show_line_range(struct line_range *lr, const char *module);
+extern int show_available_vars(struct perf_probe_event *pevs, int npevs,
+			       int max_probe_points, const char *module,
+			       bool externs);
 
 
 /* Maximum index number of event-name postfix */
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index 32b81f707ff5..3991d73d1cff 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -116,6 +116,101 @@ static void line_list__free(struct list_head *head)
 	}
 }
 
+/* Dwarf FL wrappers */
+
+static int __linux_kernel_find_elf(Dwfl_Module *mod,
+				   void **userdata,
+				   const char *module_name,
+				   Dwarf_Addr base,
+				   char **file_name, Elf **elfp)
+{
+	int fd;
+	const char *path = kernel_get_module_path(module_name);
+
+	if (path) {
+		fd = open(path, O_RDONLY);
+		if (fd >= 0) {
+			*file_name = strdup(path);
+			return fd;
+		}
+	}
+	/* If failed, try to call standard method */
+	return dwfl_linux_kernel_find_elf(mod, userdata, module_name, base,
+					  file_name, elfp);
+}
+
+static char *debuginfo_path;	/* Currently dummy */
+
+static const Dwfl_Callbacks offline_callbacks = {
+	.find_debuginfo = dwfl_standard_find_debuginfo,
+	.debuginfo_path = &debuginfo_path,
+
+	.section_address = dwfl_offline_section_address,
+
+	/* We use this table for core files too.  */
+	.find_elf = dwfl_build_id_find_elf,
+};
+
+static const Dwfl_Callbacks kernel_callbacks = {
+	.find_debuginfo = dwfl_standard_find_debuginfo,
+	.debuginfo_path = &debuginfo_path,
+
+	.find_elf = __linux_kernel_find_elf,
+	.section_address = dwfl_linux_kernel_module_section_address,
+};
+
+/* Get a Dwarf from offline image */
+static Dwarf *dwfl_init_offline_dwarf(int fd, Dwfl **dwflp, Dwarf_Addr *bias)
+{
+	Dwfl_Module *mod;
+	Dwarf *dbg = NULL;
+
+	if (!dwflp)
+		return NULL;
+
+	*dwflp = dwfl_begin(&offline_callbacks);
+	if (!*dwflp)
+		return NULL;
+
+	mod = dwfl_report_offline(*dwflp, "", "", fd);
+	if (!mod)
+		goto error;
+
+	dbg = dwfl_module_getdwarf(mod, bias);
+	if (!dbg) {
+error:
+		dwfl_end(*dwflp);
+		*dwflp = NULL;
+	}
+	return dbg;
+}
+
+/* Get a Dwarf from live kernel image */
+static Dwarf *dwfl_init_live_kernel_dwarf(Dwarf_Addr addr, Dwfl **dwflp,
+					  Dwarf_Addr *bias)
+{
+	Dwarf *dbg;
+
+	if (!dwflp)
+		return NULL;
+
+	*dwflp = dwfl_begin(&kernel_callbacks);
+	if (!*dwflp)
+		return NULL;
+
+	/* Load the kernel dwarves: Don't care the result here */
+	dwfl_linux_kernel_report_kernel(*dwflp);
+	dwfl_linux_kernel_report_modules(*dwflp);
+
+	dbg = dwfl_addrdwarf(*dwflp, addr, bias);
+	/* Here, check whether we could get a real dwarf */
+	if (!dbg) {
+		dwfl_end(*dwflp);
+		*dwflp = NULL;
+	}
+	return dbg;
+}
+
 /* Dwarf wrappers */
 
 /* Find the realpath of the target file. */
@@ -160,26 +255,44 @@ static bool die_compare_name(Dwarf_Die *dw_die, const char *tname)
 	return name ? (strcmp(tname, name) == 0) : false;
 }
 
-/* Get type die, but skip qualifiers and typedef */
-static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+/* Get type die */
+static Dwarf_Die *die_get_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
 {
 	Dwarf_Attribute attr;
+
+	if (dwarf_attr_integrate(vr_die, DW_AT_type, &attr) &&
+	    dwarf_formref_die(&attr, die_mem))
+		return die_mem;
+	else
+		return NULL;
+}
+
+/* Get a type die, but skip qualifiers */
+static Dwarf_Die *__die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+{
 	int tag;
 
 	do {
-		if (dwarf_attr(vr_die, DW_AT_type, &attr) == NULL ||
-		    dwarf_formref_die(&attr, die_mem) == NULL)
-			return NULL;
-
-		tag = dwarf_tag(die_mem);
-		vr_die = die_mem;
+		vr_die = die_get_type(vr_die, die_mem);
+		if (!vr_die)
+			break;
+		tag = dwarf_tag(vr_die);
 	} while (tag == DW_TAG_const_type ||
 		 tag == DW_TAG_restrict_type ||
 		 tag == DW_TAG_volatile_type ||
-		 tag == DW_TAG_shared_type ||
-		 tag == DW_TAG_typedef);
+		 tag == DW_TAG_shared_type);
+
+	return vr_die;
+}
 
-	return die_mem;
+/* Get a type die, but skip qualifiers and typedef */
+static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+{
+	do {
+		vr_die = __die_get_real_type(vr_die, die_mem);
+	} while (vr_die && dwarf_tag(vr_die) == DW_TAG_typedef);
+
+	return vr_die;
 }
 
 static bool die_is_signed_type(Dwarf_Die *tp_die)
@@ -320,25 +433,35 @@ static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr,
 	return die_find_child(sp_die, __die_find_inline_cb, &addr, die_mem);
 }
 
+struct __find_variable_param {
+	const char *name;
+	Dwarf_Addr addr;
+};
+
 static int __die_find_variable_cb(Dwarf_Die *die_mem, void *data)
 {
-	const char *name = data;
+	struct __find_variable_param *fvp = data;
 	int tag;
 
 	tag = dwarf_tag(die_mem);
 	if ((tag == DW_TAG_formal_parameter ||
 	     tag == DW_TAG_variable) &&
-	    die_compare_name(die_mem, name))
+	    die_compare_name(die_mem, fvp->name))
 		return DIE_FIND_CB_FOUND;
 
-	return DIE_FIND_CB_CONTINUE;
+	if (dwarf_haspc(die_mem, fvp->addr))
+		return DIE_FIND_CB_CONTINUE;
+	else
+		return DIE_FIND_CB_SIBLING;
 }
 
-/* Find a variable called 'name' */
-static Dwarf_Die *die_find_variable(Dwarf_Die *sp_die, const char *name,
-				    Dwarf_Die *die_mem)
+/* Find a variable called 'name' at given address */
+static Dwarf_Die *die_find_variable_at(Dwarf_Die *sp_die, const char *name,
+				       Dwarf_Addr addr, Dwarf_Die *die_mem)
 {
-	return die_find_child(sp_die, __die_find_variable_cb, (void *)name,
+	struct __find_variable_param fvp = { .name = name, .addr = addr};
+
+	return die_find_child(sp_die, __die_find_variable_cb, (void *)&fvp,
 			      die_mem);
 }
 
@@ -361,6 +484,60 @@ static Dwarf_Die *die_find_member(Dwarf_Die *st_die, const char *name,
 			      die_mem);
 }
 
+/* Get the name of given variable DIE */
+static int die_get_typename(Dwarf_Die *vr_die, char *buf, int len)
+{
+	Dwarf_Die type;
+	int tag, ret, ret2;
+	const char *tmp = "";
+
+	if (__die_get_real_type(vr_die, &type) == NULL)
+		return -ENOENT;
+
+	tag = dwarf_tag(&type);
+	if (tag == DW_TAG_array_type || tag == DW_TAG_pointer_type)
+		tmp = "*";
+	else if (tag == DW_TAG_subroutine_type) {
+		/* Function pointer */
+		ret = snprintf(buf, len, "(function_type)");
+		return (ret >= len) ? -E2BIG : ret;
+	} else {
+		if (!dwarf_diename(&type))
+			return -ENOENT;
+		if (tag == DW_TAG_union_type)
+			tmp = "union ";
+		else if (tag == DW_TAG_structure_type)
+			tmp = "struct ";
+		/* Write a base name */
+		ret = snprintf(buf, len, "%s%s", tmp, dwarf_diename(&type));
+		return (ret >= len) ? -E2BIG : ret;
+	}
+	ret = die_get_typename(&type, buf, len);
+	if (ret > 0) {
+		ret2 = snprintf(buf + ret, len - ret, "%s", tmp);
+		ret = (ret2 >= len - ret) ? -E2BIG : ret2 + ret;
+	}
+	return ret;
+}
+
+/* Get the name and type of given variable DIE, stored as "type\tname" */
+static int die_get_varname(Dwarf_Die *vr_die, char *buf, int len)
+{
+	int ret, ret2;
+
+	ret = die_get_typename(vr_die, buf, len);
+	if (ret < 0) {
+		pr_debug("Failed to get type, make it unknown.\n");
+		ret = snprintf(buf, len, "(unknown_type)");
+	}
+	if (ret > 0) {
+		ret2 = snprintf(buf + ret, len - ret, "\t%s",
+				dwarf_diename(vr_die));
+		ret = (ret2 >= len - ret) ? -E2BIG : ret2 + ret;
+	}
+	return ret;
+}
+
 /*
  * Probe finder related functions
  */
@@ -374,8 +551,13 @@ static struct probe_trace_arg_ref *alloc_trace_arg_ref(long offs)
 	return ref;
 }
 
-/* Show a location */
-static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
+/*
+ * Convert a location into trace_arg.
+ * If tvar == NULL, this just checks variable can be converted.
+ */
+static int convert_variable_location(Dwarf_Die *vr_die, Dwarf_Addr addr,
+				     Dwarf_Op *fb_ops,
+				     struct probe_trace_arg *tvar)
 {
 	Dwarf_Attribute attr;
 	Dwarf_Op *op;
@@ -384,20 +566,23 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
 	Dwarf_Word offs = 0;
 	bool ref = false;
 	const char *regs;
-	struct probe_trace_arg *tvar = pf->tvar;
 	int ret;
 
+	if (dwarf_attr(vr_die, DW_AT_external, &attr) != NULL)
+		goto static_var;
+
 	/* TODO: handle more than 1 exprs */
 	if (dwarf_attr(vr_die, DW_AT_location, &attr) == NULL ||
-	    dwarf_getlocation_addr(&attr, pf->addr, &op, &nops, 1) <= 0 ||
+	    dwarf_getlocation_addr(&attr, addr, &op, &nops, 1) <= 0 ||
 	    nops == 0) {
 		/* TODO: Support const_value */
-		pr_err("Failed to find the location of %s at this address.\n"
-		       " Perhaps, it has been optimized out.\n", pf->pvar->var);
 		return -ENOENT;
 	}
 
 	if (op->atom == DW_OP_addr) {
+static_var:
+		if (!tvar)
+			return 0;
 		/* Static variables on memory (not stack), make @varname */
 		ret = strlen(dwarf_diename(vr_die));
 		tvar->value = zalloc(ret + 2);
@@ -412,14 +597,11 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
 
 	/* If this is based on frame buffer, set the offset */
 	if (op->atom == DW_OP_fbreg) {
-		if (pf->fb_ops == NULL) {
-			pr_warning("The attribute of frame base is not "
-				   "supported.\n");
+		if (fb_ops == NULL)
 			return -ENOTSUP;
-		}
 		ref = true;
 		offs = op->number;
-		op = &pf->fb_ops[0];
+		op = &fb_ops[0];
 	}
 
 	if (op->atom >= DW_OP_breg0 && op->atom <= DW_OP_breg31) {
@@ -435,13 +617,18 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
 	} else if (op->atom == DW_OP_regx) {
 		regn = op->number;
 	} else {
-		pr_warning("DW_OP %x is not supported.\n", op->atom);
+		pr_debug("DW_OP %x is not supported.\n", op->atom);
 		return -ENOTSUP;
 	}
 
+	if (!tvar)
+		return 0;
+
 	regs = get_arch_regstr(regn);
 	if (!regs) {
-		pr_warning("Mapping for DWARF register number %u missing on this architecture.", regn);
+		/* This should be a bug in DWARF or this tool */
+		pr_warning("Mapping for DWARF register number %u "
+			   "missing on this architecture.", regn);
 		return -ERANGE;
 	}
 
@@ -666,8 +853,14 @@ static int convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf)
 	pr_debug("Converting variable %s into trace event.\n",
 		 dwarf_diename(vr_die));
 
-	ret = convert_variable_location(vr_die, pf);
-	if (ret == 0 && pf->pvar->field) {
+	ret = convert_variable_location(vr_die, pf->addr, pf->fb_ops,
+					pf->tvar);
+	if (ret == -ENOENT)
+		pr_err("Failed to find the location of %s at this address.\n"
+		       " Perhaps, it has been optimized out.\n", pf->pvar->var);
+	else if (ret == -ENOTSUP)
+		pr_err("Sorry, we don't support this variable location yet.\n");
+	else if (pf->pvar->field) {
 		ret = convert_variable_fields(vr_die, pf->pvar->var,
 					      pf->pvar->field, &pf->tvar->ref,
 					      &die_mem);
@@ -722,56 +915,39 @@ static int find_variable(Dwarf_Die *sp_die, struct probe_finder *pf)
 	pr_debug("Searching '%s' variable in context.\n",
 		 pf->pvar->var);
 	/* Search child die for local variables and parameters. */
-	if (die_find_variable(sp_die, pf->pvar->var, &vr_die))
+	if (die_find_variable_at(sp_die, pf->pvar->var, pf->addr, &vr_die))
 		ret = convert_variable(&vr_die, pf);
 	else {
 		/* Search upper class */
 		nscopes = dwarf_getscopes_die(sp_die, &scopes);
-		if (nscopes > 0) {
-			ret = dwarf_getscopevar(scopes, nscopes, pf->pvar->var,
-						0, NULL, 0, 0, &vr_die);
-			if (ret >= 0)
+		while (nscopes-- > 1) {
+			pr_debug("Searching variables in %s\n",
+				 dwarf_diename(&scopes[nscopes]));
+			/* We should check this scope, so give dummy address */
+			if (die_find_variable_at(&scopes[nscopes],
+						 pf->pvar->var, 0,
+						 &vr_die)) {
 				ret = convert_variable(&vr_die, pf);
-			else
-				ret = -ENOENT;
+				goto found;
+			}
+		}
+		if (scopes)
 			free(scopes);
-		} else
-			ret = -ENOENT;
+		ret = -ENOENT;
 	}
+found:
 	if (ret < 0)
 		pr_warning("Failed to find '%s' in this function.\n",
 			   pf->pvar->var);
 	return ret;
 }
 
-/* Show a probe point to output buffer */
-static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
+/* Convert subprogram DIE to trace point */
+static int convert_to_trace_point(Dwarf_Die *sp_die, Dwarf_Addr paddr,
+				  bool retprobe, struct probe_trace_point *tp)
 {
-	struct probe_trace_event *tev;
 	Dwarf_Addr eaddr;
-	Dwarf_Die die_mem;
 	const char *name;
-	int ret, i;
-	Dwarf_Attribute fb_attr;
-	size_t nops;
-
-	if (pf->ntevs == pf->max_tevs) {
-		pr_warning("Too many( > %d) probe point found.\n",
-			   pf->max_tevs);
-		return -ERANGE;
-	}
-	tev = &pf->tevs[pf->ntevs++];
-
-	/* If no real subprogram, find a real one */
-	if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) {
-		sp_die = die_find_real_subprogram(&pf->cu_die,
-						 pf->addr, &die_mem);
-		if (!sp_die) {
-			pr_warning("Failed to find probe point in any "
-				   "functions.\n");
-			return -ENOENT;
-		}
-	}
 
 	/* Copy the name of probe point */
 	name = dwarf_diename(sp_die);
@@ -781,26 +957,45 @@ static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
 				   dwarf_diename(sp_die));
 			return -ENOENT;
 		}
-		tev->point.symbol = strdup(name);
-		if (tev->point.symbol == NULL)
+		tp->symbol = strdup(name);
+		if (tp->symbol == NULL)
 			return -ENOMEM;
-		tev->point.offset = (unsigned long)(pf->addr - eaddr);
+		tp->offset = (unsigned long)(paddr - eaddr);
 	} else
 		/* This function has no name. */
-		tev->point.offset = (unsigned long)pf->addr;
+		tp->offset = (unsigned long)paddr;
 
 	/* Return probe must be on the head of a subprogram */
-	if (pf->pev->point.retprobe) {
-		if (tev->point.offset != 0) {
+	if (retprobe) {
+		if (eaddr != paddr) {
 			pr_warning("Return probe must be on the head of"
 				   " a real function\n");
 			return -EINVAL;
 		}
-		tev->point.retprobe = true;
+		tp->retprobe = true;
 	}
 
-	pr_debug("Probe point found: %s+%lu\n", tev->point.symbol,
-		 tev->point.offset);
+	return 0;
+}
+
+/* Call probe_finder callback with real subprogram DIE */
+static int call_probe_finder(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+	Dwarf_Die die_mem;
+	Dwarf_Attribute fb_attr;
+	size_t nops;
+	int ret;
+
+	/* If no real subprogram, find a real one */
+	if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) {
+		sp_die = die_find_real_subprogram(&pf->cu_die,
+						  pf->addr, &die_mem);
+		if (!sp_die) {
+			pr_warning("Failed to find probe point in any "
+				   "functions.\n");
+			return -ENOENT;
+		}
+	}
 
 	/* Get the frame base attribute/ops */
 	dwarf_attr(sp_die, DW_AT_frame_base, &fb_attr);
@@ -820,22 +1015,13 @@ static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
 #endif
 	}
 
-	/* Find each argument */
-	tev->nargs = pf->pev->nargs;
-	tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
-	if (tev->args == NULL)
-		return -ENOMEM;
-	for (i = 0; i < pf->pev->nargs; i++) {
-		pf->pvar = &pf->pev->args[i];
-		pf->tvar = &tev->args[i];
-		ret = find_variable(sp_die, pf);
-		if (ret != 0)
-			return ret;
-	}
+	/* Call finder's callback handler */
+	ret = pf->callback(sp_die, pf);
 
 	/* *pf->fb_ops will be cached in libdw. Don't free it. */
 	pf->fb_ops = NULL;
-	return 0;
+
+	return ret;
 }
 
 /* Find probe point from its line number */
@@ -871,7 +1057,7 @@ static int find_probe_point_by_line(struct probe_finder *pf)
 			 (int)i, lineno, (uintmax_t)addr);
 		pf->addr = addr;
 
-		ret = convert_probe_point(NULL, pf);
+		ret = call_probe_finder(NULL, pf);
 		/* Continuing, because target line might be inlined. */
 	}
 	return ret;
@@ -984,7 +1170,7 @@ static int find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf)
 			 (int)i, lineno, (unsigned long long)addr);
 		pf->addr = addr;
 
-		ret = convert_probe_point(sp_die, pf);
+		ret = call_probe_finder(sp_die, pf);
 		/* Continuing, because target line might be inlined. */
 	}
 	/* TODO: deallocate lines, but how? */
@@ -1019,7 +1205,7 @@ static int probe_point_inline_cb(Dwarf_Die *in_die, void *data)
 		pr_debug("found inline addr: 0x%jx\n",
 			 (uintmax_t)pf->addr);
 
-		param->retval = convert_probe_point(in_die, pf);
+		param->retval = call_probe_finder(in_die, pf);
 		if (param->retval < 0)
 			return DWARF_CB_ABORT;
 	}
@@ -1057,7 +1243,7 @@ static int probe_point_search_cb(Dwarf_Die *sp_die, void *data)
 			}
 			pf->addr += pp->offset;
 			/* TODO: Check the address in this function */
-			param->retval = convert_probe_point(sp_die, pf);
+			param->retval = call_probe_finder(sp_die, pf);
 		}
 	} else {
 		struct dwarf_callback_param _param = {.data = (void *)pf,
@@ -1079,90 +1265,276 @@ static int find_probe_point_by_func(struct probe_finder *pf)
 	return _param.retval;
 }
 
-/* Find probe_trace_events specified by perf_probe_event from debuginfo */
-int find_probe_trace_events(int fd, struct perf_probe_event *pev,
-			     struct probe_trace_event **tevs, int max_tevs)
+/* Find probe points from debuginfo */
+static int find_probes(int fd, struct probe_finder *pf)
 {
-	struct probe_finder pf = {.pev = pev, .max_tevs = max_tevs};
-	struct perf_probe_point *pp = &pev->point;
+	struct perf_probe_point *pp = &pf->pev->point;
 	Dwarf_Off off, noff;
 	size_t cuhl;
 	Dwarf_Die *diep;
-	Dwarf *dbg;
+	Dwarf *dbg = NULL;
+	Dwfl *dwfl;
+	Dwarf_Addr bias;	/* Currently ignored */
 	int ret = 0;
 
-	pf.tevs = zalloc(sizeof(struct probe_trace_event) * max_tevs);
-	if (pf.tevs == NULL)
-		return -ENOMEM;
-	*tevs = pf.tevs;
-	pf.ntevs = 0;
-
-	dbg = dwarf_begin(fd, DWARF_C_READ);
+	dbg = dwfl_init_offline_dwarf(fd, &dwfl, &bias);
 	if (!dbg) {
 		pr_warning("No dwarf info found in the vmlinux - "
 			"please rebuild with CONFIG_DEBUG_INFO=y.\n");
-		free(pf.tevs);
-		*tevs = NULL;
 		return -EBADF;
 	}
 
 #if _ELFUTILS_PREREQ(0, 142)
 	/* Get the call frame information from this dwarf */
-	pf.cfi = dwarf_getcfi(dbg);
+	pf->cfi = dwarf_getcfi(dbg);
 #endif
 
 	off = 0;
-	line_list__init(&pf.lcache);
+	line_list__init(&pf->lcache);
 	/* Loop on CUs (Compilation Unit) */
 	while (!dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL) &&
 	       ret >= 0) {
 		/* Get the DIE(Debugging Information Entry) of this CU */
-		diep = dwarf_offdie(dbg, off + cuhl, &pf.cu_die);
+		diep = dwarf_offdie(dbg, off + cuhl, &pf->cu_die);
 		if (!diep)
 			continue;
 
 		/* Check if target file is included. */
 		if (pp->file)
-			pf.fname = cu_find_realpath(&pf.cu_die, pp->file);
+			pf->fname = cu_find_realpath(&pf->cu_die, pp->file);
 		else
-			pf.fname = NULL;
+			pf->fname = NULL;
 
-		if (!pp->file || pf.fname) {
+		if (!pp->file || pf->fname) {
 			if (pp->function)
-				ret = find_probe_point_by_func(&pf);
+				ret = find_probe_point_by_func(pf);
 			else if (pp->lazy_line)
-				ret = find_probe_point_lazy(NULL, &pf);
+				ret = find_probe_point_lazy(NULL, pf);
 			else {
-				pf.lno = pp->line;
-				ret = find_probe_point_by_line(&pf);
+				pf->lno = pp->line;
+				ret = find_probe_point_by_line(pf);
 			}
 		}
 		off = noff;
 	}
-	line_list__free(&pf.lcache);
-	dwarf_end(dbg);
+	line_list__free(&pf->lcache);
+	if (dwfl)
+		dwfl_end(dwfl);
 
-	return (ret < 0) ? ret : pf.ntevs;
+	return ret;
+}
+
+/* Add a found probe point into trace event list */
+static int add_probe_trace_event(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+	struct trace_event_finder *tf =
+			container_of(pf, struct trace_event_finder, pf);
+	struct probe_trace_event *tev;
+	int ret, i;
+
+	/* Check number of tevs */
+	if (tf->ntevs == tf->max_tevs) {
+		pr_warning("Too many( > %d) probe point found.\n",
+			   tf->max_tevs);
+		return -ERANGE;
+	}
+	tev = &tf->tevs[tf->ntevs++];
+
+	ret = convert_to_trace_point(sp_die, pf->addr, pf->pev->point.retprobe,
+				     &tev->point);
+	if (ret < 0)
+		return ret;
+
+	pr_debug("Probe point found: %s+%lu\n", tev->point.symbol,
+		 tev->point.offset);
+
+	/* Find each argument */
+	tev->nargs = pf->pev->nargs;
+	tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
+	if (tev->args == NULL)
+		return -ENOMEM;
+	for (i = 0; i < pf->pev->nargs; i++) {
+		pf->pvar = &pf->pev->args[i];
+		pf->tvar = &tev->args[i];
+		ret = find_variable(sp_die, pf);
+		if (ret != 0)
+			return ret;
+	}
+
+	return 0;
+}
+
+/* Find probe_trace_events specified by perf_probe_event from debuginfo */
+int find_probe_trace_events(int fd, struct perf_probe_event *pev,
+			    struct probe_trace_event **tevs, int max_tevs)
+{
+	struct trace_event_finder tf = {
+			.pf = {.pev = pev, .callback = add_probe_trace_event},
+			.max_tevs = max_tevs};
+	int ret;
+
+	/* Allocate result tevs array */
+	*tevs = zalloc(sizeof(struct probe_trace_event) * max_tevs);
+	if (*tevs == NULL)
+		return -ENOMEM;
+
+	tf.tevs = *tevs;
+	tf.ntevs = 0;
+
+	ret = find_probes(fd, &tf.pf);
+	if (ret < 0) {
+		free(*tevs);
+		*tevs = NULL;
+		return ret;
+	}
+
+	return (ret < 0) ? ret : tf.ntevs;
+}
+
+#define MAX_VAR_LEN 64
+
+/* Collect available variables in this scope */
+static int collect_variables_cb(Dwarf_Die *die_mem, void *data)
+{
+	struct available_var_finder *af = data;
+	struct variable_list *vl;
+	char buf[MAX_VAR_LEN];
+	int tag, ret;
+
+	vl = &af->vls[af->nvls - 1];
+
+	tag = dwarf_tag(die_mem);
+	if (tag == DW_TAG_formal_parameter ||
+	    tag == DW_TAG_variable) {
+		ret = convert_variable_location(die_mem, af->pf.addr,
+						af->pf.fb_ops, NULL);
+		if (ret == 0) {
+			ret = die_get_varname(die_mem, buf, MAX_VAR_LEN);
+			pr_debug2("Add new var: %s\n", buf);
+			if (ret > 0)
+				strlist__add(vl->vars, buf);
+		}
+	}
+
+	if (af->child && dwarf_haspc(die_mem, af->pf.addr))
+		return DIE_FIND_CB_CONTINUE;
+	else
+		return DIE_FIND_CB_SIBLING;
+}
+
+/* Add a found vars into available variables list */
+static int add_available_vars(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+	struct available_var_finder *af =
+			container_of(pf, struct available_var_finder, pf);
+	struct variable_list *vl;
+	Dwarf_Die die_mem, *scopes = NULL;
+	int ret, nscopes;
+
+	/* Check number of tevs */
+	if (af->nvls == af->max_vls) {
+		pr_warning("Too many( > %d) probe point found.\n", af->max_vls);
+		return -ERANGE;
+	}
+	vl = &af->vls[af->nvls++];
+
+	ret = convert_to_trace_point(sp_die, pf->addr, pf->pev->point.retprobe,
+				     &vl->point);
+	if (ret < 0)
+		return ret;
+
+	pr_debug("Probe point found: %s+%lu\n", vl->point.symbol,
+		 vl->point.offset);
+
+	/* Find local variables */
+	vl->vars = strlist__new(true, NULL);
+	if (vl->vars == NULL)
+		return -ENOMEM;
+	af->child = true;
+	die_find_child(sp_die, collect_variables_cb, (void *)af, &die_mem);
+
+	/* Find external variables */
+	if (!af->externs)
+		goto out;
+	/* Don't need to search child DIE for externs. */
+	af->child = false;
+	nscopes = dwarf_getscopes_die(sp_die, &scopes);
+	while (nscopes-- > 1)
+		die_find_child(&scopes[nscopes], collect_variables_cb,
+			       (void *)af, &die_mem);
+	if (scopes)
+		free(scopes);
+
+out:
+	if (strlist__empty(vl->vars)) {
+		strlist__delete(vl->vars);
+		vl->vars = NULL;
+	}
+
+	return ret;
+}
+
+/* Find available variables at given probe point */
+int find_available_vars_at(int fd, struct perf_probe_event *pev,
+			   struct variable_list **vls, int max_vls,
+			   bool externs)
+{
+	struct available_var_finder af = {
+			.pf = {.pev = pev, .callback = add_available_vars},
+			.max_vls = max_vls, .externs = externs};
+	int ret;
+
+	/* Allocate result vls array */
+	*vls = zalloc(sizeof(struct variable_list) * max_vls);
+	if (*vls == NULL)
+		return -ENOMEM;
+
+	af.vls = *vls;
+	af.nvls = 0;
+
+	ret = find_probes(fd, &af.pf);
+	if (ret < 0) {
+		/* Free vlist for error */
+		while (af.nvls--) {
+			if (af.vls[af.nvls].point.symbol)
+				free(af.vls[af.nvls].point.symbol);
+			if (af.vls[af.nvls].vars)
+				strlist__delete(af.vls[af.nvls].vars);
+		}
+		free(af.vls);
+		*vls = NULL;
+		return ret;
+	}
+
+	return (ret < 0) ? ret : af.nvls;
 }
 
 /* Reverse search */
-int find_perf_probe_point(int fd, unsigned long addr,
-			  struct perf_probe_point *ppt)
+int find_perf_probe_point(unsigned long addr, struct perf_probe_point *ppt)
 {
 	Dwarf_Die cudie, spdie, indie;
-	Dwarf *dbg;
+	Dwarf *dbg = NULL;
+	Dwfl *dwfl = NULL;
 	Dwarf_Line *line;
-	Dwarf_Addr laddr, eaddr;
+	Dwarf_Addr laddr, eaddr, bias = 0;
 	const char *tmp;
 	int lineno, ret = 0;
 	bool found = false;
 
-	dbg = dwarf_begin(fd, DWARF_C_READ);
-	if (!dbg)
-		return -EBADF;
+	/* Open the live linux kernel */
+	dbg = dwfl_init_live_kernel_dwarf(addr, &dwfl, &bias);
+	if (!dbg) {
+		pr_warning("No dwarf info found in the vmlinux - "
+			"please rebuild with CONFIG_DEBUG_INFO=y.\n");
+		ret = -EINVAL;
+		goto end;
+	}
 
+	/* Adjust address with bias */
+	addr += bias;
 	/* Find cu die */
-	if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr, &cudie)) {
+	if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr - bias, &cudie)) {
+		pr_warning("No CU DIE is found at %lx\n", addr);
 		ret = -EINVAL;
 		goto end;
 	}
@@ -1225,7 +1597,8 @@ found:
 	}
 
 end:
-	dwarf_end(dbg);
+	if (dwfl)
+		dwfl_end(dwfl);
 	if (ret >= 0)
 		ret = found ? 1 : 0;
 	return ret;
@@ -1358,6 +1731,9 @@ static int line_range_search_cb(Dwarf_Die *sp_die, void *data)
 	struct line_finder *lf = param->data;
 	struct line_range *lr = lf->lr;
 
+	pr_debug("find (%llx) %s\n",
+		 (unsigned long long)dwarf_dieoffset(sp_die),
+		 dwarf_diename(sp_die));
 	if (dwarf_tag(sp_die) == DW_TAG_subprogram &&
 	    die_compare_name(sp_die, lr->function)) {
 		lf->fname = dwarf_decl_file(sp_die);
@@ -1401,10 +1777,12 @@ int find_line_range(int fd, struct line_range *lr)
 	Dwarf_Off off = 0, noff;
 	size_t cuhl;
 	Dwarf_Die *diep;
-	Dwarf *dbg;
+	Dwarf *dbg = NULL;
+	Dwfl *dwfl;
+	Dwarf_Addr bias;	/* Currently ignored */
 	const char *comp_dir;
 
-	dbg = dwarf_begin(fd, DWARF_C_READ);
+	dbg = dwfl_init_offline_dwarf(fd, &dwfl, &bias);
 	if (!dbg) {
 		pr_warning("No dwarf info found in the vmlinux - "
 			"please rebuild with CONFIG_DEBUG_INFO=y.\n");
@@ -1450,8 +1828,7 @@ int find_line_range(int fd, struct line_range *lr)
 	}
 
 	pr_debug("path: %s\n", lr->path);
-	dwarf_end(dbg);
-
+	dwfl_end(dwfl);
 	return (ret < 0) ? ret : lf.found;
 }
 
diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h
index 4507d519f183..bba69d455699 100644
--- a/tools/perf/util/probe-finder.h
+++ b/tools/perf/util/probe-finder.h
@@ -22,20 +22,27 @@ extern int find_probe_trace_events(int fd, struct perf_probe_event *pev,
 				    int max_tevs);
 
 /* Find a perf_probe_point from debuginfo */
-extern int find_perf_probe_point(int fd, unsigned long addr,
+extern int find_perf_probe_point(unsigned long addr,
 				 struct perf_probe_point *ppt);
 
+/* Find a line range */
 extern int find_line_range(int fd, struct line_range *lr);
 
+/* Find available variables */
+extern int find_available_vars_at(int fd, struct perf_probe_event *pev,
+				  struct variable_list **vls, int max_points,
+				  bool externs);
+
 #include <dwarf.h>
 #include <libdw.h>
+#include <libdwfl.h>
 #include <version.h>
 
 struct probe_finder {
 	struct perf_probe_event	*pev;		/* Target probe event */
-	struct probe_trace_event *tevs;		/* Result trace events */
-	int			ntevs;		/* Number of trace events */
-	int			max_tevs;	/* Max number of trace events */
+
+	/* Callback when a probe point is found */
+	int (*callback)(Dwarf_Die *sp_die, struct probe_finder *pf);
 
 	/* For function searching */
 	int			lno;		/* Line number */
@@ -53,6 +60,22 @@ struct probe_finder {
 	struct probe_trace_arg	*tvar;		/* Current result variable */
 };
 
+struct trace_event_finder {
+	struct probe_finder	pf;
+	struct probe_trace_event *tevs;		/* Found trace events */
+	int			ntevs;		/* Number of trace events */
+	int			max_tevs;	/* Max number of trace events */
+};
+
+struct available_var_finder {
+	struct probe_finder	pf;
+	struct variable_list	*vls;		/* Found variable lists */
+	int			nvls;		/* Number of variable lists */
+	int			max_vls;	/* Max no. of variable lists */
+	bool			externs;	/* Find external vars too */
+	bool			child;		/* Search child scopes */
+};
+
 struct line_finder {
 	struct line_range	*lr;		/* Target line range */
 
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index b39f499e575a..d628c8d1cf5e 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -295,7 +295,9 @@ static void symbols__insert_by_name(struct rb_root *self, struct symbol *sym)
 {
 	struct rb_node **p = &self->rb_node;
 	struct rb_node *parent = NULL;
-	struct symbol_name_rb_node *symn = ((void *)sym) - sizeof(*parent), *s;
+	struct symbol_name_rb_node *symn, *s;
+
+	symn = container_of(sym, struct symbol_name_rb_node, sym);
 
 	while (*p != NULL) {
 		parent = *p;
@@ -530,7 +532,7 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
 	struct machine *machine = kmaps->machine;
 	struct map *curr_map = map;
 	struct symbol *pos;
-	int count = 0;
+	int count = 0, moved = 0;	
 	struct rb_root *root = &self->symbols[map->type];
 	struct rb_node *next = rb_first(root);
 	int kernel_range = 0;
@@ -588,6 +590,11 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
 			char dso_name[PATH_MAX];
 			struct dso *dso;
 
+			if (count == 0) {
+				curr_map = map;
+				goto filter_symbol;
+			}
+
 			if (self->kernel == DSO_TYPE_GUEST_KERNEL)
 				snprintf(dso_name, sizeof(dso_name),
 					"[guest.kernel].%d",
@@ -613,7 +620,7 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
 			map_groups__insert(kmaps, curr_map);
 			++kernel_range;
 		}
-
+filter_symbol:
 		if (filter && filter(curr_map, pos)) {
 discard_symbol:		rb_erase(&pos->rb_node, root);
 			symbol__delete(pos);
@@ -621,8 +628,9 @@ discard_symbol:		rb_erase(&pos->rb_node, root);
 			if (curr_map != map) {
 				rb_erase(&pos->rb_node, root);
 				symbols__insert(&curr_map->dso->symbols[curr_map->type], pos);
-			}
-			count++;
+				++moved;
+			} else
+				++count;
 		}
 	}
 
@@ -632,7 +640,7 @@ discard_symbol:		rb_erase(&pos->rb_node, root);
 		dso__set_loaded(curr_map->dso, curr_map->type);
 	}
 
-	return count;
+	return count + moved;
 }
 
 int dso__load_kallsyms(struct dso *self, const char *filename,
@@ -2123,14 +2131,55 @@ static struct dso *machine__create_kernel(struct machine *self)
 	return kernel;
 }
 
+struct process_args {
+	u64 start;
+};
+
+static int symbol__in_kernel(void *arg, const char *name,
+			     char type __used, u64 start)
+{
+	struct process_args *args = arg;
+
+	if (strchr(name, '['))
+		return 0;
+
+	args->start = start;
+	return 1;
+}
+
+/* Figure out the start address of kernel map from /proc/kallsyms */
+static u64 machine__get_kernel_start_addr(struct machine *machine)
+{
+	const char *filename;
+	char path[PATH_MAX];
+	struct process_args args;
+
+	if (machine__is_host(machine)) {
+		filename = "/proc/kallsyms";
+	} else {
+		if (machine__is_default_guest(machine))
+			filename = (char *)symbol_conf.default_guest_kallsyms;
+		else {
+			sprintf(path, "%s/proc/kallsyms", machine->root_dir);
+			filename = path;
+		}
+	}
+
+	if (kallsyms__parse(filename, &args, symbol__in_kernel) <= 0)
+		return 0;
+
+	return args.start;
+}
+
 int __machine__create_kernel_maps(struct machine *self, struct dso *kernel)
 {
 	enum map_type type;
+	u64 start = machine__get_kernel_start_addr(self);
 
 	for (type = 0; type < MAP__NR_TYPES; ++type) {
 		struct kmap *kmap;
 
-		self->vmlinux_maps[type] = map__new2(0, kernel, type);
+		self->vmlinux_maps[type] = map__new2(start, kernel, type);
 		if (self->vmlinux_maps[type] == NULL)
 			return -1;
 
diff --git a/tools/perf/util/ui/browser.c b/tools/perf/util/ui/browser.c
index 6d0df809a2ed..8bc010edca25 100644
--- a/tools/perf/util/ui/browser.c
+++ b/tools/perf/util/ui/browser.c
@@ -1,4 +1,3 @@
-#include <slang.h>
 #include "libslang.h"
 #include <linux/compiler.h>
 #include <linux/list.h>
diff --git a/tools/perf/util/ui/util.c b/tools/perf/util/ui/util.c
index 9706d9d40279..056c69521a38 100644
--- a/tools/perf/util/ui/util.c
+++ b/tools/perf/util/ui/util.c
@@ -104,9 +104,10 @@ out_destroy_form:
 	return rc;
 }
 
+static const char yes[] = "Yes", no[] = "No";
+
 bool ui__dialog_yesno(const char *msg)
 {
 	/* newtWinChoice should really be accepting const char pointers... */
-	char yes[] = "Yes", no[] = "No";
-	return newtWinChoice(NULL, yes, no, (char *)msg) == 1;
+	return newtWinChoice(NULL, (char *)yes, (char *)no, (char *)msg) == 1;
 }