Merge branch 'perf/core' of...

Merge branch 'perf/core' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 into perf/core

tools/perf/Documentation/perf-probe.txt

@@ -73,6 +73,10 @@ OPTIONS
 	(Only for --vars) Show external defined variables in addition to local
 	variables.
 
+-F::
+--funcs::
+	Show available functions in given module or kernel.
+
 -f::
 --force::
 	Forcibly add events with existing name.

tools/perf/Makefile

@@ -402,6 +402,7 @@ LIB_H += util/debug.h
 LIB_H += util/debugfs.h
 LIB_H += util/event.h
 LIB_H += util/evsel.h
+LIB_H += util/evlist.h
 LIB_H += util/exec_cmd.h
 LIB_H += util/types.h
 LIB_H += util/levenshtein.h
@@ -425,6 +426,7 @@ LIB_H += util/values.h
 LIB_H += util/sort.h
 LIB_H += util/hist.h
 LIB_H += util/thread.h
+LIB_H += util/thread_map.h
 LIB_H += util/trace-event.h
 LIB_H += util/probe-finder.h
 LIB_H += util/probe-event.h
@@ -440,6 +442,7 @@ LIB_OBJS += $(OUTPUT)util/ctype.o
 LIB_OBJS += $(OUTPUT)util/debugfs.o
 LIB_OBJS += $(OUTPUT)util/environment.o
 LIB_OBJS += $(OUTPUT)util/event.o
+LIB_OBJS += $(OUTPUT)util/evlist.o
 LIB_OBJS += $(OUTPUT)util/evsel.o
 LIB_OBJS += $(OUTPUT)util/exec_cmd.o
 LIB_OBJS += $(OUTPUT)util/help.o
@@ -469,6 +472,7 @@ LIB_OBJS += $(OUTPUT)util/map.o
 LIB_OBJS += $(OUTPUT)util/pstack.o
 LIB_OBJS += $(OUTPUT)util/session.o
 LIB_OBJS += $(OUTPUT)util/thread.o
+LIB_OBJS += $(OUTPUT)util/thread_map.o
 LIB_OBJS += $(OUTPUT)util/trace-event-parse.o
 LIB_OBJS += $(OUTPUT)util/trace-event-read.o
 LIB_OBJS += $(OUTPUT)util/trace-event-info.o

tools/perf/builtin-probe.c

@@ -52,6 +52,7 @@ static struct {
 	bool show_lines;
 	bool show_vars;
 	bool show_ext_vars;
+	bool show_funcs;
 	bool mod_events;
 	int nevents;
 	struct perf_probe_event events[MAX_PROBES];
@@ -221,6 +222,8 @@ static const struct option options[] = {
 	OPT__DRY_RUN(&probe_event_dry_run),
 	OPT_INTEGER('\0', "max-probes", &params.max_probe_points,
 		 "Set how many probe points can be found for a probe."),
+	OPT_BOOLEAN('F', "funcs", &params.show_funcs,
+		    "Show potential probe-able functions."),
 	OPT_END()
 };
 
@@ -246,7 +249,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 		params.max_probe_points = MAX_PROBES;
 
 	if ((!params.nevents && !params.dellist && !params.list_events &&
-	     !params.show_lines))
+	     !params.show_lines && !params.show_funcs))
 		usage_with_options(probe_usage, options);
 
 	/*
@@ -267,12 +270,36 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 			pr_err(" Error: Don't use --list with --vars.\n");
 			usage_with_options(probe_usage, options);
 		}
+		if (params.show_funcs) {
+			pr_err("  Error: Don't use --list with --funcs.\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",
 			       ret);
 		return ret;
 	}
+	if (params.show_funcs) {
+		if (params.nevents != 0 || params.dellist) {
+			pr_err("  Error: Don't use --funcs with"
+			       " --add/--del.\n");
+			usage_with_options(probe_usage, options);
+		}
+		if (params.show_lines) {
+			pr_err("  Error: Don't use --funcs with --line.\n");
+			usage_with_options(probe_usage, options);
+		}
+		if (params.show_vars) {
+			pr_err("  Error: Don't use --funcs with --vars.\n");
+			usage_with_options(probe_usage, options);
+		}
+		ret = show_available_funcs(params.target_module);
+		if (ret < 0)
+			pr_err("  Error: Failed to show functions."
+			       " (%d)\n", ret);
+		return ret;
+	}
 
 #ifdef DWARF_SUPPORT
 	if (params.show_lines) {

tools/perf/builtin-report.c

@@ -81,18 +81,17 @@ static int perf_session__add_hist_entry(struct perf_session *self,
 					struct addr_location *al,
 					struct sample_data *data)
 {
-	struct map_symbol *syms = NULL;
 	struct symbol *parent = NULL;
-	int err = -ENOMEM;
+	int err = 0;
 	struct hist_entry *he;
 	struct hists *hists;
 	struct perf_event_attr *attr;
 
 	if ((sort__has_parent || symbol_conf.use_callchain) && data->callchain) {
-		syms = perf_session__resolve_callchain(self, al->thread,
-						       data->callchain, &parent);
-		if (syms == NULL)
-			return -ENOMEM;
+		err = perf_session__resolve_callchain(self, al->thread,
+						      data->callchain, &parent);
+		if (err)
+			return err;
 	}
 
 	attr = perf_header__find_attr(data->id, &self->header);
@@ -101,16 +100,17 @@ static int perf_session__add_hist_entry(struct perf_session *self,
 	else
 		hists = perf_session__hists_findnew(self, data->id, 0, 0);
 	if (hists == NULL)
-		goto out_free_syms;
+		return -ENOMEM;
+
 	he = __hists__add_entry(hists, al, parent, data->period);
 	if (he == NULL)
-		goto out_free_syms;
-	err = 0;
+		return -ENOMEM;
+
 	if (symbol_conf.use_callchain) {
-		err = callchain_append(he->callchain, data->callchain, syms,
+		err = callchain_append(he->callchain, &self->callchain_cursor,
 				       data->period);
 		if (err)
-			goto out_free_syms;
+			return err;
 	}
 	/*
 	 * Only in the newt browser we are doing integrated annotation,
@@ -119,8 +119,7 @@ static int perf_session__add_hist_entry(struct perf_session *self,
 	 */
 	if (use_browser > 0)
 		err = hist_entry__inc_addr_samples(he, al->addr);
-out_free_syms:
-	free(syms);
+
 	return err;
 }
 
@@ -222,7 +221,7 @@ static int perf_session__setup_sample_type(struct perf_session *self)
 	} else if (!dont_use_callchains && callchain_param.mode != CHAIN_NONE &&
 		   !symbol_conf.use_callchain) {
 			symbol_conf.use_callchain = true;
-			if (register_callchain_param(&callchain_param) < 0) {
+			if (callchain_register_param(&callchain_param) < 0) {
 				fprintf(stderr, "Can't register callchain"
 						" params\n");
 				return -EINVAL;
@@ -424,7 +423,7 @@ parse_callchain_opt(const struct option *opt __used, const char *arg,
 	if (tok2)
 		callchain_param.print_limit = strtod(tok2, &endptr);
 setup:
-	if (register_callchain_param(&callchain_param) < 0) {
+	if (callchain_register_param(&callchain_param) < 0) {
 		fprintf(stderr, "Can't register callchain params\n");
 		return -1;
 	}

tools/perf/builtin-stat.c

@@ -43,11 +43,13 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 #include "util/event.h"
+#include "util/evlist.h"
 #include "util/evsel.h"
 #include "util/debug.h"
 #include "util/header.h"
 #include "util/cpumap.h"
 #include "util/thread.h"
+#include "util/thread_map.h"
 
 #include <sys/prctl.h>
 #include <math.h>
@@ -71,6 +73,8 @@ static struct perf_event_attr default_attrs[] = {
 
 };
 
+struct perf_evlist		*evsel_list;
+
 static bool			system_wide			=  false;
 static struct cpu_map		*cpus;
 static int			run_idx				=  0;
@@ -166,7 +170,7 @@ static int create_perf_stat_counter(struct perf_evsel *evsel)
 				    PERF_FORMAT_TOTAL_TIME_RUNNING;
 
 	if (system_wide)
-		return perf_evsel__open_per_cpu(evsel, cpus);
+		return perf_evsel__open_per_cpu(evsel, cpus, false, false);
 
 	attr->inherit = !no_inherit;
 	if (target_pid == -1 && target_tid == -1) {
@@ -174,7 +178,7 @@ static int create_perf_stat_counter(struct perf_evsel *evsel)
 		attr->enable_on_exec = 1;
 	}
 
-	return perf_evsel__open_per_thread(evsel, threads);
+	return perf_evsel__open_per_thread(evsel, threads, false, false);
 }
 
 /*
@@ -309,7 +313,7 @@ static int run_perf_stat(int argc __used, const char **argv)
 		close(child_ready_pipe[0]);
 	}
 
-	list_for_each_entry(counter, &evsel_list, node) {
+	list_for_each_entry(counter, &evsel_list->entries, node) {
 		if (create_perf_stat_counter(counter) < 0) {
 			if (errno == -EPERM || errno == -EACCES) {
 				error("You may not have permission to collect %sstats.\n"
@@ -347,12 +351,12 @@ static int run_perf_stat(int argc __used, const char **argv)
 	update_stats(&walltime_nsecs_stats, t1 - t0);
 
 	if (no_aggr) {
-		list_for_each_entry(counter, &evsel_list, node) {
+		list_for_each_entry(counter, &evsel_list->entries, node) {
 			read_counter(counter);
 			perf_evsel__close_fd(counter, cpus->nr, 1);
 		}
 	} else {
-		list_for_each_entry(counter, &evsel_list, node) {
+		list_for_each_entry(counter, &evsel_list->entries, node) {
 			read_counter_aggr(counter);
 			perf_evsel__close_fd(counter, cpus->nr, threads->nr);
 		}
@@ -555,10 +559,10 @@ static void print_stat(int argc, const char **argv)
 	}
 
 	if (no_aggr) {
-		list_for_each_entry(counter, &evsel_list, node)
+		list_for_each_entry(counter, &evsel_list->entries, node)
 			print_counter(counter);
 	} else {
-		list_for_each_entry(counter, &evsel_list, node)
+		list_for_each_entry(counter, &evsel_list->entries, node)
 			print_counter_aggr(counter);
 	}
 
@@ -610,7 +614,7 @@ static int stat__set_big_num(const struct option *opt __used,
 }
 
 static const struct option options[] = {
-	OPT_CALLBACK('e', "event", NULL, "event",
+	OPT_CALLBACK('e', "event", &evsel_list, "event",
 		     "event selector. use 'perf list' to list available events",
 		     parse_events),
 	OPT_BOOLEAN('i', "no-inherit", &no_inherit,
@@ -648,6 +652,10 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 
 	setlocale(LC_ALL, "");
 
+	evsel_list = perf_evlist__new();
+	if (evsel_list == NULL)
+		return -ENOMEM;
+
 	argc = parse_options(argc, argv, options, stat_usage,
 		PARSE_OPT_STOP_AT_NON_OPTION);
 
@@ -679,17 +687,14 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 		usage_with_options(stat_usage, options);
 
 	/* Set attrs and nr_counters if no event is selected and !null_run */
-	if (!null_run && !nr_counters) {
+	if (!null_run && !evsel_list->nr_entries) {
 		size_t c;
 
-		nr_counters = ARRAY_SIZE(default_attrs);
-
 		for (c = 0; c < ARRAY_SIZE(default_attrs); ++c) {
-			pos = perf_evsel__new(&default_attrs[c],
-					      nr_counters);
+			pos = perf_evsel__new(&default_attrs[c], c);
 			if (pos == NULL)
 				goto out;
-			list_add(&pos->node, &evsel_list);
+			perf_evlist__add(evsel_list, pos);
 		}
 	}
 
@@ -713,7 +718,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 		return -1;
 	}
 
-	list_for_each_entry(pos, &evsel_list, node) {
+	list_for_each_entry(pos, &evsel_list->entries, node) {
 		if (perf_evsel__alloc_stat_priv(pos) < 0 ||
 		    perf_evsel__alloc_counts(pos, cpus->nr) < 0 ||
 		    perf_evsel__alloc_fd(pos, cpus->nr, threads->nr) < 0)
@@ -741,9 +746,9 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 	if (status != -1)
 		print_stat(argc, argv);
 out_free_fd:
-	list_for_each_entry(pos, &evsel_list, node)
+	list_for_each_entry(pos, &evsel_list->entries, node)
 		perf_evsel__free_stat_priv(pos);
-	perf_evsel_list__delete();
+	perf_evlist__delete(evsel_list);
 out:
 	thread_map__delete(threads);
 	threads = NULL;

tools/perf/builtin-test.c

@@ -7,10 +7,11 @@
 
 #include "util/cache.h"
 #include "util/debug.h"
+#include "util/evlist.h"
 #include "util/parse-options.h"
-#include "util/session.h"
+#include "util/parse-events.h"
 #include "util/symbol.h"
-#include "util/thread.h"
+#include "util/thread_map.h"
 
 static long page_size;
 
@@ -238,14 +239,14 @@ static int test__vmlinux_matches_kallsyms(void)
 #include "util/evsel.h"
 #include <sys/types.h>
 
-static int trace_event__id(const char *event_name)
+static int trace_event__id(const char *evname)
 {
 	char *filename;
 	int err = -1, fd;
 
 	if (asprintf(&filename,
 		     "/sys/kernel/debug/tracing/events/syscalls/%s/id",
-		     event_name) < 0)
+		     evname) < 0)
 		return -1;
 
 	fd = open(filename, O_RDONLY);
@@ -289,7 +290,7 @@ static int test__open_syscall_event(void)
 		goto out_thread_map_delete;
 	}
 
-	if (perf_evsel__open_per_thread(evsel, threads) < 0) {
+	if (perf_evsel__open_per_thread(evsel, threads, false, false) < 0) {
 		pr_debug("failed to open counter: %s, "
 			 "tweak /proc/sys/kernel/perf_event_paranoid?\n",
 			 strerror(errno));
@@ -347,9 +348,9 @@ static int test__open_syscall_event_on_all_cpus(void)
 	}
 
 	cpus = cpu_map__new(NULL);
-	if (threads == NULL) {
-		pr_debug("thread_map__new\n");
-		return -1;
+	if (cpus == NULL) {
+		pr_debug("cpu_map__new\n");
+		goto out_thread_map_delete;
 	}
 
 
@@ -364,7 +365,7 @@ static int test__open_syscall_event_on_all_cpus(void)
 		goto out_thread_map_delete;
 	}
 
-	if (perf_evsel__open(evsel, cpus, threads) < 0) {
+	if (perf_evsel__open(evsel, cpus, threads, false, false) < 0) {
 		pr_debug("failed to open counter: %s, "
 			 "tweak /proc/sys/kernel/perf_event_paranoid?\n",
 			 strerror(errno));
@@ -408,6 +409,8 @@ static int test__open_syscall_event_on_all_cpus(void)
 		goto out_close_fd;
 	}
 
+	err = 0;
+
 	for (cpu = 0; cpu < cpus->nr; ++cpu) {
 		unsigned int expected;
 
@@ -416,18 +419,18 @@ static int test__open_syscall_event_on_all_cpus(void)
 
 		if (perf_evsel__read_on_cpu(evsel, cpu, 0) < 0) {
 			pr_debug("perf_evsel__open_read_on_cpu\n");
-			goto out_close_fd;
+			err = -1;
+			break;
 		}
 
 		expected = nr_open_calls + cpu;
 		if (evsel->counts->cpu[cpu].val != expected) {
 			pr_debug("perf_evsel__read_on_cpu: expected to intercept %d calls on cpu %d, got %" PRIu64 "\n",
 				 expected, cpus->map[cpu], evsel->counts->cpu[cpu].val);
-			goto out_close_fd;
+			err = -1;
 		}
 	}
 
-	err = 0;
 out_close_fd:
 	perf_evsel__close_fd(evsel, 1, threads->nr);
 out_evsel_delete:
@@ -437,6 +440,159 @@ static int test__open_syscall_event_on_all_cpus(void)
 	return err;
 }
 
+/*
+ * This test will generate random numbers of calls to some getpid syscalls,
+ * then establish an mmap for a group of events that are created to monitor
+ * the syscalls.
+ *
+ * It will receive the events, using mmap, use its PERF_SAMPLE_ID generated
+ * sample.id field to map back to its respective perf_evsel instance.
+ *
+ * Then it checks if the number of syscalls reported as perf events by
+ * the kernel corresponds to the number of syscalls made.
+ */
+static int test__basic_mmap(void)
+{
+	int err = -1;
+	event_t *event;
+	struct thread_map *threads;
+	struct cpu_map *cpus;
+	struct perf_evlist *evlist;
+	struct perf_event_attr attr = {
+		.type		= PERF_TYPE_TRACEPOINT,
+		.read_format	= PERF_FORMAT_ID,
+		.sample_type	= PERF_SAMPLE_ID,
+		.watermark	= 0,
+	};
+	cpu_set_t cpu_set;
+	const char *syscall_names[] = { "getsid", "getppid", "getpgrp",
+					"getpgid", };
+	pid_t (*syscalls[])(void) = { (void *)getsid, getppid, getpgrp,
+				      (void*)getpgid };
+#define nsyscalls ARRAY_SIZE(syscall_names)
+	int ids[nsyscalls];
+	unsigned int nr_events[nsyscalls],
+		     expected_nr_events[nsyscalls], i, j;
+	struct perf_evsel *evsels[nsyscalls], *evsel;
+
+	for (i = 0; i < nsyscalls; ++i) {
+		char name[64];
+
+		snprintf(name, sizeof(name), "sys_enter_%s", syscall_names[i]);
+		ids[i] = trace_event__id(name);
+		if (ids[i] < 0) {
+			pr_debug("Is debugfs mounted on /sys/kernel/debug?\n");
+			return -1;
+		}
+		nr_events[i] = 0;
+		expected_nr_events[i] = random() % 257;
+	}
+
+	threads = thread_map__new(-1, getpid());
+	if (threads == NULL) {
+		pr_debug("thread_map__new\n");
+		return -1;
+	}
+
+	cpus = cpu_map__new(NULL);
+	if (threads == NULL) {
+		pr_debug("thread_map__new\n");
+		goto out_free_threads;
+	}
+
+	CPU_ZERO(&cpu_set);
+	CPU_SET(cpus->map[0], &cpu_set);
+	sched_setaffinity(0, sizeof(cpu_set), &cpu_set);
+	if (sched_setaffinity(0, sizeof(cpu_set), &cpu_set) < 0) {
+		pr_debug("sched_setaffinity() failed on CPU %d: %s ",
+			 cpus->map[0], strerror(errno));
+		goto out_free_cpus;
+	}
+
+	evlist = perf_evlist__new();
+	if (threads == NULL) {
+		pr_debug("perf_evlist__new\n");
+		goto out_free_cpus;
+	}
+
+	/* anonymous union fields, can't be initialized above */
+	attr.wakeup_events = 1;
+	attr.sample_period = 1;
+
+	for (i = 0; i < nsyscalls; ++i) {
+		attr.config = ids[i];
+		evsels[i] = perf_evsel__new(&attr, i);
+		if (evsels[i] == NULL) {
+			pr_debug("perf_evsel__new\n");
+			goto out_free_evlist;
+		}
+
+		perf_evlist__add(evlist, evsels[i]);
+
+		if (perf_evsel__open(evsels[i], cpus, threads, false, false) < 0) {
+			pr_debug("failed to open counter: %s, "
+				 "tweak /proc/sys/kernel/perf_event_paranoid?\n",
+				 strerror(errno));
+			goto out_close_fd;
+		}
+	}
+
+	if (perf_evlist__mmap(evlist, cpus, threads, 128, true) < 0) {
+		pr_debug("failed to mmap events: %d (%s)\n", errno,
+			 strerror(errno));
+		goto out_close_fd;
+	}
+
+	for (i = 0; i < nsyscalls; ++i)
+		for (j = 0; j < expected_nr_events[i]; ++j) {
+			int foo = syscalls[i]();
+			++foo;
+		}
+
+	while ((event = perf_evlist__read_on_cpu(evlist, 0)) != NULL) {
+		struct sample_data sample;
+
+		if (event->header.type != PERF_RECORD_SAMPLE) {
+			pr_debug("unexpected %s event\n",
+				 event__get_event_name(event->header.type));
+			goto out_munmap;
+		}
+
+		event__parse_sample(event, attr.sample_type, false, &sample);
+		evsel = perf_evlist__id2evsel(evlist, sample.id);
+		if (evsel == NULL) {
+			pr_debug("event with id %" PRIu64
+				 " doesn't map to an evsel\n", sample.id);
+			goto out_munmap;
+		}
+		nr_events[evsel->idx]++;
+	}
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		if (nr_events[evsel->idx] != expected_nr_events[evsel->idx]) {
+			pr_debug("expected %d %s events, got %d\n",
+				 expected_nr_events[evsel->idx],
+				 event_name(evsel), nr_events[evsel->idx]);
+			goto out_munmap;
+		}
+	}
+
+	err = 0;
+out_munmap:
+	perf_evlist__munmap(evlist, 1);
+out_close_fd:
+	for (i = 0; i < nsyscalls; ++i)
+		perf_evsel__close_fd(evsels[i], 1, threads->nr);
+out_free_evlist:
+	perf_evlist__delete(evlist);
+out_free_cpus:
+	cpu_map__delete(cpus);
+out_free_threads:
+	thread_map__delete(threads);
+	return err;
+#undef nsyscalls
+}
+
 static struct test {
 	const char *desc;
 	int (*func)(void);
@@ -453,6 +609,10 @@ static struct test {
 		.desc = "detect open syscall event on all cpus",
 		.func = test__open_syscall_event_on_all_cpus,
 	},
+	{
+		.desc = "read samples using the mmap interface",
+		.func = test__basic_mmap,
+	},
 	{
 		.func = NULL,
 	},

tools/perf/perf.h

@@ -94,6 +94,32 @@ void get_term_dimensions(struct winsize *ws);
 #include "util/types.h"
 #include <stdbool.h>
 
+struct perf_mmap {
+	void			*base;
+	int			mask;
+	unsigned int		prev;
+};
+
+static inline unsigned int perf_mmap__read_head(struct perf_mmap *mm)
+{
+	struct perf_event_mmap_page *pc = mm->base;
+	int head = pc->data_head;
+	rmb();
+	return head;
+}
+
+static inline void perf_mmap__write_tail(struct perf_mmap *md,
+					 unsigned long tail)
+{
+	struct perf_event_mmap_page *pc = md->base;
+
+	/*
+	 * ensure all reads are done before we write the tail out.
+	 */
+	/* mb(); */
+	pc->data_tail = tail;
+}
+
 /*
  * prctl(PR_TASK_PERF_EVENTS_DISABLE) will (cheaply) disable all
  * counters in the current task.

tools/perf/util/callchain.h

@@ -16,7 +16,7 @@ enum chain_mode {
 
 struct callchain_node {
 	struct callchain_node	*parent;
-	struct list_head	brothers;
+	struct list_head	siblings;
 	struct list_head	children;
 	struct list_head	val;
 	struct rb_node		rb_node; /* to sort nodes in an rbtree */
@@ -49,9 +49,30 @@ struct callchain_list {
 	struct list_head	list;
 };
 
+/*
+ * A callchain cursor is a single linked list that
+ * let one feed a callchain progressively.
+ * It keeps persitent allocated entries to minimize
+ * allocations.
+ */
+struct callchain_cursor_node {
+	u64				ip;
+	struct map			*map;
+	struct symbol			*sym;
+	struct callchain_cursor_node	*next;
+};
+
+struct callchain_cursor {
+	u64				nr;
+	struct callchain_cursor_node	*first;
+	struct callchain_cursor_node	**last;
+	u64				pos;
+	struct callchain_cursor_node	*curr;
+};
+
 static inline void callchain_init(struct callchain_root *root)
 {
-	INIT_LIST_HEAD(&root->node.brothers);
+	INIT_LIST_HEAD(&root->node.siblings);
 	INIT_LIST_HEAD(&root->node.children);
 	INIT_LIST_HEAD(&root->node.val);
 
@@ -61,15 +82,54 @@ static inline void callchain_init(struct callchain_root *root)
 	root->max_depth = 0;
 }
 
-static inline u64 cumul_hits(struct callchain_node *node)
+static inline u64 callchain_cumul_hits(struct callchain_node *node)
 {
 	return node->hit + node->children_hit;
 }
 
-int register_callchain_param(struct callchain_param *param);
-int callchain_append(struct callchain_root *root, struct ip_callchain *chain,
-		     struct map_symbol *syms, u64 period);
-int callchain_merge(struct callchain_root *dst, struct callchain_root *src);
+int callchain_register_param(struct callchain_param *param);
+int callchain_append(struct callchain_root *root,
+		     struct callchain_cursor *cursor,
+		     u64 period);
+
+int callchain_merge(struct callchain_cursor *cursor,
+		    struct callchain_root *dst, struct callchain_root *src);
 
 bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event);
+
+/*
+ * Initialize a cursor before adding entries inside, but keep
+ * the previously allocated entries as a cache.
+ */
+static inline void callchain_cursor_reset(struct callchain_cursor *cursor)
+{
+	cursor->nr = 0;
+	cursor->last = &cursor->first;
+}
+
+int callchain_cursor_append(struct callchain_cursor *cursor, u64 ip,
+			    struct map *map, struct symbol *sym);
+
+/* Close a cursor writing session. Initialize for the reader */
+static inline void callchain_cursor_commit(struct callchain_cursor *cursor)
+{
+	cursor->curr = cursor->first;
+	cursor->pos = 0;
+}
+
+/* Cursor reading iteration helpers */
+static inline struct callchain_cursor_node *
+callchain_cursor_current(struct callchain_cursor *cursor)
+{
+	if (cursor->pos == cursor->nr)
+		return NULL;
+
+	return cursor->curr;
+}
+
+static inline void callchain_cursor_advance(struct callchain_cursor *cursor)
+{
+	cursor->curr = cursor->curr->next;
+	cursor->pos++;
+}
 #endif	/* __PERF_CALLCHAIN_H */

tools/perf/util/cpumap.c

@@ -177,3 +177,8 @@ struct cpu_map *cpu_map__dummy_new(void)
 
 	return cpus;
 }
+
+void cpu_map__delete(struct cpu_map *map)
+{
+	free(map);
+}

tools/perf/util/cpumap.h

@@ -8,6 +8,6 @@ struct cpu_map {
 
 struct cpu_map *cpu_map__new(const char *cpu_list);
 struct cpu_map *cpu_map__dummy_new(void);
-void *cpu_map__delete(struct cpu_map *map);
+void cpu_map__delete(struct cpu_map *map);
 
 #endif /* __PERF_CPUMAP_H */

tools/perf/util/event.c

@@ -826,128 +826,3 @@ int event__preprocess_sample(const event_t *self, struct perf_session *session,
 	al->filtered = true;
 	return 0;
 }
-
-static int event__parse_id_sample(const event_t *event,
-				  struct perf_session *session,
-				  struct sample_data *sample)
-{
-	const u64 *array;
-	u64 type;
-
-	sample->cpu = sample->pid = sample->tid = -1;
-	sample->stream_id = sample->id = sample->time = -1ULL;
-
-	if (!session->sample_id_all)
-		return 0;
-
-	array = event->sample.array;
-	array += ((event->header.size -
-		   sizeof(event->header)) / sizeof(u64)) - 1;
-	type = session->sample_type;
-
-	if (type & PERF_SAMPLE_CPU) {
-		u32 *p = (u32 *)array;
-		sample->cpu = *p;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_STREAM_ID) {
-		sample->stream_id = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_ID) {
-		sample->id = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_TIME) {
-		sample->time = *array;
-		array--;
-	}
-
-	if (type & PERF_SAMPLE_TID) {
-		u32 *p = (u32 *)array;
-		sample->pid = p[0];
-		sample->tid = p[1];
-	}
-
-	return 0;
-}
-
-int event__parse_sample(const event_t *event, struct perf_session *session,
-			struct sample_data *data)
-{
-	const u64 *array;
-	u64 type;
-
-	if (event->header.type != PERF_RECORD_SAMPLE)
-		return event__parse_id_sample(event, session, data);
-
-	array = event->sample.array;
-	type = session->sample_type;
-
-	if (type & PERF_SAMPLE_IP) {
-		data->ip = event->ip.ip;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_TID) {
-		u32 *p = (u32 *)array;
-		data->pid = p[0];
-		data->tid = p[1];
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_TIME) {
-		data->time = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_ADDR) {
-		data->addr = *array;
-		array++;
-	}
-
-	data->id = -1ULL;
-	if (type & PERF_SAMPLE_ID) {
-		data->id = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_STREAM_ID) {
-		data->stream_id = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_CPU) {
-		u32 *p = (u32 *)array;
-		data->cpu = *p;
-		array++;
-	} else
-		data->cpu = -1;
-
-	if (type & PERF_SAMPLE_PERIOD) {
-		data->period = *array;
-		array++;
-	}
-
-	if (type & PERF_SAMPLE_READ) {
-		pr_debug("PERF_SAMPLE_READ is unsuported for now\n");
-		return -1;
-	}
-
-	if (type & PERF_SAMPLE_CALLCHAIN) {
-		data->callchain = (struct ip_callchain *)array;
-		array += 1 + data->callchain->nr;
-	}
-
-	if (type & PERF_SAMPLE_RAW) {
-		u32 *p = (u32 *)array;
-		data->raw_size = *p;
-		p++;
-		data->raw_data = p;
-	}
-
-	return 0;
-}

tools/perf/util/event.h

@@ -169,9 +169,10 @@ struct addr_location;
 int event__preprocess_sample(const event_t *self, struct perf_session *session,
 			     struct addr_location *al, struct sample_data *data,
 			     symbol_filter_t filter);
-int event__parse_sample(const event_t *event, struct perf_session *session,
-			struct sample_data *sample);
 
 const char *event__get_event_name(unsigned int id);
 
+int event__parse_sample(const event_t *event, u64 type, bool sample_id_all,
+			struct sample_data *sample);
+
 #endif /* __PERF_RECORD_H */

tools/perf/util/evlist.c

+#include <poll.h>
+#include "evlist.h"
+#include "evsel.h"
+#include "util.h"
+
+#include <linux/bitops.h>
+#include <linux/hash.h>
+
+void perf_evlist__init(struct perf_evlist *evlist)
+{
+	int i;
+
+	for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i)
+		INIT_HLIST_HEAD(&evlist->heads[i]);
+	INIT_LIST_HEAD(&evlist->entries);
+}
+
+struct perf_evlist *perf_evlist__new(void)
+{
+	struct perf_evlist *evlist = zalloc(sizeof(*evlist));
+
+	if (evlist != NULL)
+		perf_evlist__init(evlist);
+
+	return evlist;
+}
+
+static void perf_evlist__purge(struct perf_evlist *evlist)
+{
+	struct perf_evsel *pos, *n;
+
+	list_for_each_entry_safe(pos, n, &evlist->entries, node) {
+		list_del_init(&pos->node);
+		perf_evsel__delete(pos);
+	}
+
+	evlist->nr_entries = 0;
+}
+
+void perf_evlist__exit(struct perf_evlist *evlist)
+{
+	free(evlist->mmap);
+	free(evlist->pollfd);
+	evlist->mmap = NULL;
+	evlist->pollfd = NULL;
+}
+
+void perf_evlist__delete(struct perf_evlist *evlist)
+{
+	perf_evlist__purge(evlist);
+	perf_evlist__exit(evlist);
+	free(evlist);
+}
+
+void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry)
+{
+	list_add_tail(&entry->node, &evlist->entries);
+	++evlist->nr_entries;
+}
+
+int perf_evlist__add_default(struct perf_evlist *evlist)
+{
+	struct perf_event_attr attr = {
+		.type = PERF_TYPE_HARDWARE,
+		.config = PERF_COUNT_HW_CPU_CYCLES,
+	};
+	struct perf_evsel *evsel = perf_evsel__new(&attr, 0);
+
+	if (evsel == NULL)
+		return -ENOMEM;
+
+	perf_evlist__add(evlist, evsel);
+	return 0;
+}
+
+int perf_evlist__alloc_pollfd(struct perf_evlist *evlist, int ncpus, int nthreads)
+{
+	int nfds = ncpus * nthreads * evlist->nr_entries;
+	evlist->pollfd = malloc(sizeof(struct pollfd) * nfds);
+	return evlist->pollfd != NULL ? 0 : -ENOMEM;
+}
+
+void perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd)
+{
+	fcntl(fd, F_SETFL, O_NONBLOCK);
+	evlist->pollfd[evlist->nr_fds].fd = fd;
+	evlist->pollfd[evlist->nr_fds].events = POLLIN;
+	evlist->nr_fds++;
+}
+
+struct perf_evsel *perf_evlist__id2evsel(struct perf_evlist *evlist, u64 id)
+{
+	struct hlist_head *head;
+	struct hlist_node *pos;
+	struct perf_sample_id *sid;
+	int hash;
+
+	if (evlist->nr_entries == 1)
+		return list_entry(evlist->entries.next, struct perf_evsel, node);
+
+	hash = hash_64(id, PERF_EVLIST__HLIST_BITS);
+	head = &evlist->heads[hash];
+
+	hlist_for_each_entry(sid, pos, head, node)
+		if (sid->id == id)
+			return sid->evsel;
+	return NULL;
+}
+
+event_t *perf_evlist__read_on_cpu(struct perf_evlist *evlist, int cpu)
+{
+	/* XXX Move this to perf.c, making it generally available */
+	unsigned int page_size = sysconf(_SC_PAGE_SIZE);
+	struct perf_mmap *md = &evlist->mmap[cpu];
+	unsigned int head = perf_mmap__read_head(md);
+	unsigned int old = md->prev;
+	unsigned char *data = md->base + page_size;
+	event_t *event = NULL;
+	int diff;
+
+	/*
+	 * If we're further behind than half the buffer, there's a chance
+	 * the writer will bite our tail and mess up the samples under us.
+	 *
+	 * If we somehow ended up ahead of the head, we got messed up.
+	 *
+	 * In either case, truncate and restart at head.
+	 */
+	diff = head - old;
+	if (diff > md->mask / 2 || diff < 0) {
+		fprintf(stderr, "WARNING: failed to keep up with mmap data.\n");
+
+		/*
+		 * head points to a known good entry, start there.
+		 */
+		old = head;
+	}
+
+	if (old != head) {
+		size_t size;
+
+		event = (event_t *)&data[old & md->mask];
+		size = event->header.size;
+
+		/*
+		 * Event straddles the mmap boundary -- header should always
+		 * be inside due to u64 alignment of output.
+		 */
+		if ((old & md->mask) + size != ((old + size) & md->mask)) {
+			unsigned int offset = old;
+			unsigned int len = min(sizeof(*event), size), cpy;
+			void *dst = &evlist->event_copy;
+
+			do {
+				cpy = min(md->mask + 1 - (offset & md->mask), len);
+				memcpy(dst, &data[offset & md->mask], cpy);
+				offset += cpy;
+				dst += cpy;
+				len -= cpy;
+			} while (len);
+
+			event = &evlist->event_copy;
+		}
+
+		old += size;
+	}
+
+	md->prev = old;
+	return event;
+}

tools/perf/util/evlist.h

+#ifndef __PERF_EVLIST_H
+#define __PERF_EVLIST_H 1
+
+#include <linux/list.h>
+#include "../perf.h"
+#include "event.h"
+
+struct pollfd;
+
+#define PERF_EVLIST__HLIST_BITS 8
+#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS)
+
+struct perf_evlist {
+	struct list_head entries;
+	struct hlist_head heads[PERF_EVLIST__HLIST_SIZE];
+	int		 nr_entries;
+	int		 nr_fds;
+	int		 mmap_len;
+	event_t		 event_copy;
+	struct perf_mmap *mmap;
+	struct pollfd	 *pollfd;
+};
+
+struct perf_evsel;
+
+struct perf_evlist *perf_evlist__new(void);
+void perf_evlist__init(struct perf_evlist *evlist);
+void perf_evlist__exit(struct perf_evlist *evlist);
+void perf_evlist__delete(struct perf_evlist *evlist);
+
+void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry);
+int perf_evlist__add_default(struct perf_evlist *evlist);
+
+int perf_evlist__alloc_pollfd(struct perf_evlist *evlist, int ncpus, int nthreads);
+void perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd);
+
+struct perf_evsel *perf_evlist__id2evsel(struct perf_evlist *evlist, u64 id);
+
+event_t *perf_evlist__read_on_cpu(struct perf_evlist *self, int cpu);
+
+#endif /* __PERF_EVLIST_H */

tools/perf/util/evsel.h

@@ -24,11 +24,24 @@ struct perf_counts {
 	struct perf_counts_values cpu[];
 };
 
+struct perf_evsel;
+
+/*
+ * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are
+ * more than one entry in the evlist.
+ */
+struct perf_sample_id {
+	struct hlist_node 	node;
+	u64		 	id;
+	struct perf_evsel	*evsel;
+};
+
 struct perf_evsel {
 	struct list_head	node;
 	struct perf_event_attr	attr;
 	char			*filter;
 	struct xyarray		*fd;
+	struct xyarray		*id;
 	struct perf_counts	*counts;
 	int			idx;
 	void			*priv;
@@ -36,19 +49,31 @@ struct perf_evsel {
 
 struct cpu_map;
 struct thread_map;
+struct perf_evlist;
 
 struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx);
+void perf_evsel__init(struct perf_evsel *evsel,
+		      struct perf_event_attr *attr, int idx);
+void perf_evsel__exit(struct perf_evsel *evsel);
 void perf_evsel__delete(struct perf_evsel *evsel);
 
 int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
+int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads);
 int perf_evsel__alloc_counts(struct perf_evsel *evsel, int ncpus);
+int perf_evlist__alloc_mmap(struct perf_evlist *evlist, int ncpus);
 void perf_evsel__free_fd(struct perf_evsel *evsel);
+void perf_evsel__free_id(struct perf_evsel *evsel);
 void perf_evsel__close_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
 
-int perf_evsel__open_per_cpu(struct perf_evsel *evsel, struct cpu_map *cpus);
-int perf_evsel__open_per_thread(struct perf_evsel *evsel, struct thread_map *threads);
-int perf_evsel__open(struct perf_evsel *evsel, 
-		     struct cpu_map *cpus, struct thread_map *threads);
+int perf_evsel__open_per_cpu(struct perf_evsel *evsel,
+			     struct cpu_map *cpus, bool group, bool inherit);
+int perf_evsel__open_per_thread(struct perf_evsel *evsel,
+				struct thread_map *threads, bool group, bool inherit);
+int perf_evsel__open(struct perf_evsel *evsel, struct cpu_map *cpus,
+		     struct thread_map *threads, bool group, bool inherit);
+int perf_evlist__mmap(struct perf_evlist *evlist, struct cpu_map *cpus,
+		      struct thread_map *threads, int pages, bool overwrite);
+void perf_evlist__munmap(struct perf_evlist *evlist, int ncpus);
 
 #define perf_evsel__match(evsel, t, c)		\
 	(evsel->attr.type == PERF_TYPE_##t &&	\

tools/perf/util/header.c

@@ -8,6 +8,7 @@
 #include <linux/list.h>
 #include <linux/kernel.h>
 
+#include "evlist.h"
 #include "util.h"
 #include "header.h"
 #include "../perf.h"
@@ -428,7 +429,8 @@ static bool perf_session__read_build_ids(struct perf_session *self, bool with_hi
 	return ret;
 }
 
-static int perf_header__adds_write(struct perf_header *self, int fd)
+static int perf_header__adds_write(struct perf_header *self,
+				   struct perf_evlist *evlist, int fd)
 {
 	int nr_sections;
 	struct perf_session *session;
@@ -463,7 +465,7 @@ static int perf_header__adds_write(struct perf_header *self, int fd)
 
 		/* Write trace info */
 		trace_sec->offset = lseek(fd, 0, SEEK_CUR);
-		read_tracing_data(fd, &evsel_list);
+		read_tracing_data(fd, &evlist->entries);
 		trace_sec->size = lseek(fd, 0, SEEK_CUR) - trace_sec->offset;
 	}
 
@@ -513,7 +515,8 @@ int perf_header__write_pipe(int fd)
 	return 0;
 }
 
-int perf_header__write(struct perf_header *self, int fd, bool at_exit)
+int perf_header__write(struct perf_header *self, struct perf_evlist *evlist,
+		       int fd, bool at_exit)
 {
 	struct perf_file_header f_header;
 	struct perf_file_attr   f_attr;
@@ -566,7 +569,7 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit)
 	self->data_offset = lseek(fd, 0, SEEK_CUR);
 
 	if (at_exit) {
-		err = perf_header__adds_write(self, fd);
+		err = perf_header__adds_write(self, evlist, fd);
 		if (err < 0)
 			return err;
 	}
@@ -1133,7 +1136,7 @@ int event__process_event_type(event_t *self,
 	return 0;
 }
 
-int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
+int event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
 				   event__handler_t process,
 				   struct perf_session *session __unused)
 {
@@ -1144,7 +1147,7 @@ int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
 	memset(&ev, 0, sizeof(ev));
 
 	ev.tracing_data.header.type = PERF_RECORD_HEADER_TRACING_DATA;
-	size = read_tracing_data_size(fd, pattrs);
+	size = read_tracing_data_size(fd, &evlist->entries);
 	if (size <= 0)
 		return size;
 	aligned_size = ALIGN(size, sizeof(u64));
@@ -1154,7 +1157,7 @@ int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
 
 	process(&ev, NULL, session);
 
-	err = read_tracing_data(fd, pattrs);
+	err = read_tracing_data(fd, &evlist->entries);
 	write_padded(fd, NULL, 0, padding);
 
 	return aligned_size;

tools/perf/util/header.h

@@ -65,8 +65,11 @@ struct perf_header {
 int perf_header__init(struct perf_header *self);
 void perf_header__exit(struct perf_header *self);
 
+struct perf_evlist;
+
 int perf_header__read(struct perf_session *session, int fd);
-int perf_header__write(struct perf_header *self, int fd, bool at_exit);
+int perf_header__write(struct perf_header *self, struct perf_evlist *evlist,
+		       int fd, bool at_exit);
 int perf_header__write_pipe(int fd);
 
 int perf_header__add_attr(struct perf_header *self,
@@ -113,7 +116,7 @@ int event__synthesize_event_types(event__handler_t process,
 int event__process_event_type(event_t *self,
 			      struct perf_session *session);
 
-int event__synthesize_tracing_data(int fd, struct list_head *pattrs,
+int event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
 				   event__handler_t process,
 				   struct perf_session *session);
 int event__process_tracing_data(event_t *self,

tools/perf/util/hist.c

@@ -211,7 +211,9 @@ void hist_entry__free(struct hist_entry *he)
  * collapse the histogram
  */
 
-static bool collapse__insert_entry(struct rb_root *root, struct hist_entry *he)
+static bool hists__collapse_insert_entry(struct hists *self,
+					 struct rb_root *root,
+					 struct hist_entry *he)
 {
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
@@ -226,8 +228,11 @@ static bool collapse__insert_entry(struct rb_root *root, struct hist_entry *he)
 
 		if (!cmp) {
 			iter->period += he->period;
-			if (symbol_conf.use_callchain)
-				callchain_merge(iter->callchain, he->callchain);
+			if (symbol_conf.use_callchain) {
+				callchain_cursor_reset(&self->callchain_cursor);
+				callchain_merge(&self->callchain_cursor, iter->callchain,
+						he->callchain);
+			}
 			hist_entry__free(he);
 			return false;
 		}
@@ -262,7 +267,7 @@ void hists__collapse_resort(struct hists *self)
 		next = rb_next(&n->rb_node);
 
 		rb_erase(&n->rb_node, &self->entries);
-		if (collapse__insert_entry(&tmp, n))
+		if (hists__collapse_insert_entry(self, &tmp, n))
 			hists__inc_nr_entries(self, n);
 	}
 
@@ -425,7 +430,7 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 		u64 cumul;
 
 		child = rb_entry(node, struct callchain_node, rb_node);
-		cumul = cumul_hits(child);
+		cumul = callchain_cumul_hits(child);
 		remaining -= cumul;
 
 		/*

tools/perf/util/hist.h

@@ -77,6 +77,8 @@ struct hists {
 	u64			event_stream;
 	u32			type;
 	u16			col_len[HISTC_NR_COLS];
+	/* Best would be to reuse the session callchain cursor */
+	struct callchain_cursor	callchain_cursor;
 };
 
 struct hist_entry *__hists__add_entry(struct hists *self,

tools/perf/util/include/linux/list.h

+#include <linux/kernel.h>
 #include "../../../../include/linux/list.h"
 
 #ifndef PERF_LIST_H

tools/perf/util/parse-events.c

 #include "../../../include/linux/hw_breakpoint.h"
 #include "util.h"
 #include "../perf.h"
+#include "evlist.h"
 #include "evsel.h"
 #include "parse-options.h"
 #include "parse-events.h"
@@ -11,10 +12,6 @@
 #include "header.h"
 #include "debugfs.h"
 
-int				nr_counters;
-
-LIST_HEAD(evsel_list);
-
 struct event_symbol {
 	u8		type;
 	u64		config;
@@ -449,8 +446,8 @@ parse_single_tracepoint_event(char *sys_name,
 /* sys + ':' + event + ':' + flags*/
 #define MAX_EVOPT_LEN	(MAX_EVENT_LENGTH * 2 + 2 + 128)
 static enum event_result
-parse_multiple_tracepoint_event(char *sys_name, const char *evt_exp,
-				char *flags)
+parse_multiple_tracepoint_event(const struct option *opt, char *sys_name,
+				const char *evt_exp, char *flags)
 {
 	char evt_path[MAXPATHLEN];
 	struct dirent *evt_ent;
@@ -483,15 +480,16 @@ parse_multiple_tracepoint_event(char *sys_name, const char *evt_exp,
 		if (len < 0)
 			return EVT_FAILED;
 
-		if (parse_events(NULL, event_opt, 0))
+		if (parse_events(opt, event_opt, 0))
 			return EVT_FAILED;
 	}
 
 	return EVT_HANDLED_ALL;
 }
 
-static enum event_result parse_tracepoint_event(const char **strp,
-				    struct perf_event_attr *attr)
+static enum event_result
+parse_tracepoint_event(const struct option *opt, const char **strp,
+		       struct perf_event_attr *attr)
 {
 	const char *evt_name;
 	char *flags = NULL, *comma_loc;
@@ -530,7 +528,7 @@ static enum event_result parse_tracepoint_event(const char **strp,
 		return EVT_FAILED;
 	if (strpbrk(evt_name, "*?")) {
 		*strp += strlen(sys_name) + evt_length + 1; /* 1 == the ':' */
-		return parse_multiple_tracepoint_event(sys_name, evt_name,
+		return parse_multiple_tracepoint_event(opt, sys_name, evt_name,
 						       flags);
 	} else {
 		return parse_single_tracepoint_event(sys_name, evt_name,
@@ -740,11 +738,12 @@ parse_event_modifier(const char **strp, struct perf_event_attr *attr)
  * Symbolic names are (almost) exactly matched.
  */
 static enum event_result
-parse_event_symbols(const char **str, struct perf_event_attr *attr)
+parse_event_symbols(const struct option *opt, const char **str,
+		    struct perf_event_attr *attr)
 {
 	enum event_result ret;
 
-	ret = parse_tracepoint_event(str, attr);
+	ret = parse_tracepoint_event(opt, str, attr);
 	if (ret != EVT_FAILED)
 		goto modifier;
 
@@ -778,14 +777,15 @@ parse_event_symbols(const char **str, struct perf_event_attr *attr)
 	return ret;
 }
 
-int parse_events(const struct option *opt __used, const char *str, int unset __used)
+int parse_events(const struct option *opt, const char *str, int unset __used)
 {
+	struct perf_evlist *evlist = *(struct perf_evlist **)opt->value;
 	struct perf_event_attr attr;
 	enum event_result ret;
 
 	for (;;) {
 		memset(&attr, 0, sizeof(attr));
-		ret = parse_event_symbols(&str, &attr);
+		ret = parse_event_symbols(opt, &str, &attr);
 		if (ret == EVT_FAILED)
 			return -1;
 
@@ -794,12 +794,10 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u
 
 		if (ret != EVT_HANDLED_ALL) {
 			struct perf_evsel *evsel;
-			evsel = perf_evsel__new(&attr,
-						nr_counters);
+			evsel = perf_evsel__new(&attr, evlist->nr_entries);
 			if (evsel == NULL)
 				return -1;
-			list_add_tail(&evsel->node, &evsel_list);
-			++nr_counters;
+			perf_evlist__add(evlist, evsel);
 		}
 
 		if (*str == 0)
@@ -813,13 +811,14 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u
 	return 0;
 }
 
-int parse_filter(const struct option *opt __used, const char *str,
+int parse_filter(const struct option *opt, const char *str,
 		 int unset __used)
 {
+	struct perf_evlist *evlist = *(struct perf_evlist **)opt->value;
 	struct perf_evsel *last = NULL;
 
-	if (!list_empty(&evsel_list))
-		last = list_entry(evsel_list.prev, struct perf_evsel, node);
+	if (evlist->nr_entries > 0)
+		last = list_entry(evlist->entries.prev, struct perf_evsel, node);
 
 	if (last == NULL || last->attr.type != PERF_TYPE_TRACEPOINT) {
 		fprintf(stderr,
@@ -981,33 +980,3 @@ void print_events(void)
 
 	exit(129);
 }
-
-int perf_evsel_list__create_default(void)
-{
-	struct perf_evsel *evsel;
-	struct perf_event_attr attr;
-
-	memset(&attr, 0, sizeof(attr));
-	attr.type = PERF_TYPE_HARDWARE;
-	attr.config = PERF_COUNT_HW_CPU_CYCLES;
-
-	evsel = perf_evsel__new(&attr, 0);
-
-	if (evsel == NULL)
-		return -ENOMEM;
-
-	list_add(&evsel->node, &evsel_list);
-	++nr_counters;
-	return 0;
-}
-
-void perf_evsel_list__delete(void)
-{
-	struct perf_evsel *pos, *n;
-
-	list_for_each_entry_safe(pos, n, &evsel_list, node) {
-		list_del_init(&pos->node);
-		perf_evsel__delete(pos);
-	}
-	nr_counters = 0;
-}

tools/perf/util/parse-events.h

@@ -9,11 +9,6 @@
 struct list_head;
 struct perf_evsel;
 
-extern struct list_head evsel_list;
-
-int perf_evsel_list__create_default(void);
-void perf_evsel_list__delete(void);
-
 struct option;
 
 struct tracepoint_path {
@@ -25,8 +20,6 @@ struct tracepoint_path {
 extern struct tracepoint_path *tracepoint_id_to_path(u64 config);
 extern bool have_tracepoints(struct list_head *evlist);
 
-extern int			nr_counters;
-
 const char *event_name(struct perf_evsel *event);
 extern const char *__event_name(int type, u64 config);
 

tools/perf/util/probe-event.c

@@ -31,6 +31,7 @@
 #include <string.h>
 #include <stdarg.h>
 #include <limits.h>
+#include <elf.h>
 
 #undef _GNU_SOURCE
 #include "util.h"
@@ -111,7 +112,25 @@ static struct symbol *__find_kernel_function_by_name(const char *name,
 						     NULL);
 }
 
-const char *kernel_get_module_path(const char *module)
+static struct map *kernel_get_module_map(const char *module)
+{
+	struct rb_node *nd;
+	struct map_groups *grp = &machine.kmaps;
+
+	if (!module)
+		module = "kernel";
+
+	for (nd = rb_first(&grp->maps[MAP__FUNCTION]); nd; nd = rb_next(nd)) {
+		struct map *pos = rb_entry(nd, struct map, rb_node);
+		if (strncmp(pos->dso->short_name + 1, module,
+			    pos->dso->short_name_len - 2) == 0) {
+			return pos;
+		}
+	}
+	return NULL;
+}
+
+static struct dso *kernel_get_module_dso(const char *module)
 {
 	struct dso *dso;
 	struct map *map;
@@ -141,7 +160,13 @@ const char *kernel_get_module_path(const char *module)
 		}
 	}
 found:
-	return dso->long_name;
+	return dso;
+}
+
+const char *kernel_get_module_path(const char *module)
+{
+	struct dso *dso = kernel_get_module_dso(module);
+	return (dso) ? dso->long_name : NULL;
 }
 
 #ifdef DWARF_SUPPORT
@@ -1913,3 +1938,42 @@ int del_perf_probe_events(struct strlist *dellist)
 	return ret;
 }
 
+/*
+ * If a symbol corresponds to a function with global binding return 0.
+ * For all others return 1.
+ */
+static int filter_non_global_functions(struct map *map __unused,
+					struct symbol *sym)
+{
+	if (sym->binding != STB_GLOBAL)
+		return 1;
+
+	return 0;
+}
+
+int show_available_funcs(const char *module)
+{
+	struct map *map;
+	int ret;
+
+	setup_pager();
+
+	ret = init_vmlinux();
+	if (ret < 0)
+		return ret;
+
+	map = kernel_get_module_map(module);
+	if (!map) {
+		pr_err("Failed to find %s map.\n", (module) ? : "kernel");
+		return -EINVAL;
+	}
+	if (map__load(map, filter_non_global_functions)) {
+		pr_err("Failed to load map.\n");
+		return -EINVAL;
+	}
+	if (!dso__sorted_by_name(map->dso, map->type))
+		dso__sort_by_name(map->dso, map->type);
+
+	dso__fprintf_symbols_by_name(map->dso, map->type, stdout);
+	return 0;
+}

tools/perf/util/probe-event.h

@@ -127,6 +127,7 @@ 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);
+extern int show_available_funcs(const char *module);
 
 
 /* Maximum index number of event-name postfix */

tools/perf/util/session.c

@@ -242,17 +242,16 @@ static bool symbol__match_parent_regex(struct symbol *sym)
 	return 0;
 }
 
-struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
-						   struct thread *thread,
-						   struct ip_callchain *chain,
-						   struct symbol **parent)
+int perf_session__resolve_callchain(struct perf_session *self,
+				    struct thread *thread,
+				    struct ip_callchain *chain,
+				    struct symbol **parent)
 {
 	u8 cpumode = PERF_RECORD_MISC_USER;
 	unsigned int i;
-	struct map_symbol *syms = calloc(chain->nr, sizeof(*syms));
+	int err;
 
-	if (!syms)
-		return NULL;
+	callchain_cursor_reset(&self->callchain_cursor);
 
 	for (i = 0; i < chain->nr; i++) {
 		u64 ip = chain->ips[i];
@@ -281,12 +280,15 @@ struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
 				*parent = al.sym;
 			if (!symbol_conf.use_callchain)
 				break;
-			syms[i].map = al.map;
-			syms[i].sym = al.sym;
 		}
+
+		err = callchain_cursor_append(&self->callchain_cursor,
+					      ip, al.map, al.sym);
+		if (err)
+			return err;
 	}
 
-	return syms;
+	return 0;
 }
 
 static int process_event_synth_stub(event_t *event __used,
@@ -494,7 +496,7 @@ static void flush_sample_queue(struct perf_session *s,
 		if (iter->timestamp > limit)
 			break;
 
-		event__parse_sample(iter->event, s, &sample);
+		perf_session__parse_sample(s, iter->event, &sample);
 		perf_session_deliver_event(s, iter->event, &sample, ops,
 					   iter->file_offset);
 
@@ -804,7 +806,7 @@ static int perf_session__process_event(struct perf_session *session,
 	/*
 	 * For all kernel events we get the sample data
 	 */
-	event__parse_sample(event, session, &sample);
+	perf_session__parse_sample(session, event, &sample);
 
 	/* Preprocess sample records - precheck callchains */
 	if (perf_session__preprocess_sample(session, event, &sample))

tools/perf/util/session.h

@@ -51,7 +51,8 @@ struct perf_session {
 	int			cwdlen;
 	char			*cwd;
 	struct ordered_samples	ordered_samples;
-	char filename[0];
+	struct callchain_cursor	callchain_cursor;
+	char			filename[0];
 };
 
 struct perf_event_ops;
@@ -94,10 +95,10 @@ int __perf_session__process_events(struct perf_session *self,
 int perf_session__process_events(struct perf_session *self,
 				 struct perf_event_ops *event_ops);
 
-struct map_symbol *perf_session__resolve_callchain(struct perf_session *self,
-						   struct thread *thread,
-						   struct ip_callchain *chain,
-						   struct symbol **parent);
+int perf_session__resolve_callchain(struct perf_session *self,
+				    struct thread *thread,
+				    struct ip_callchain *chain,
+				    struct symbol **parent);
 
 bool perf_session__has_traces(struct perf_session *self, const char *msg);
 
@@ -154,4 +155,13 @@ size_t perf_session__fprintf_nr_events(struct perf_session *self, FILE *fp)
 {
 	return hists__fprintf_nr_events(&self->hists, fp);
 }
+
+static inline int perf_session__parse_sample(struct perf_session *session,
+					     const event_t *event,
+					     struct sample_data *sample)
+{
+	return event__parse_sample(event, session->sample_type,
+				   session->sample_id_all, sample);
+}
+
 #endif /* __PERF_SESSION_H */

tools/perf/util/thread.c

@@ -7,61 +7,6 @@
 #include "util.h"
 #include "debug.h"
 
-/* Skip "." and ".." directories */
-static int filter(const struct dirent *dir)
-{
-	if (dir->d_name[0] == '.')
-		return 0;
-	else
-		return 1;
-}
-
-struct thread_map *thread_map__new_by_pid(pid_t pid)
-{
-	struct thread_map *threads;
-	char name[256];
-	int items;
-	struct dirent **namelist = NULL;
-	int i;
-
-	sprintf(name, "/proc/%d/task", pid);
-	items = scandir(name, &namelist, filter, NULL);
-	if (items <= 0)
-                return NULL;
-
-	threads = malloc(sizeof(*threads) + sizeof(pid_t) * items);
-	if (threads != NULL) {
-		for (i = 0; i < items; i++)
-			threads->map[i] = atoi(namelist[i]->d_name);
-		threads->nr = items;
-	}
-
-	for (i=0; i<items; i++)
-		free(namelist[i]);
-	free(namelist);
-
-	return threads;
-}
-
-struct thread_map *thread_map__new_by_tid(pid_t tid)
-{
-	struct thread_map *threads = malloc(sizeof(*threads) + sizeof(pid_t));
-
-	if (threads != NULL) {
-		threads->map[0] = tid;
-		threads->nr	= 1;
-	}
-
-	return threads;
-}
-
-struct thread_map *thread_map__new(pid_t pid, pid_t tid)
-{
-	if (pid != -1)
-		return thread_map__new_by_pid(pid);
-	return thread_map__new_by_tid(tid);
-}
-
 static struct thread *thread__new(pid_t pid)
 {
 	struct thread *self = zalloc(sizeof(*self));

tools/perf/util/thread.h

@@ -18,24 +18,10 @@ struct thread {
 	int			comm_len;
 };
 
-struct thread_map {
-	int nr;
-	int map[];
-};
-
 struct perf_session;
 
 void thread__delete(struct thread *self);
 
-struct thread_map *thread_map__new_by_pid(pid_t pid);
-struct thread_map *thread_map__new_by_tid(pid_t tid);
-struct thread_map *thread_map__new(pid_t pid, pid_t tid);
-
-static inline void thread_map__delete(struct thread_map *threads)
-{
-	free(threads);
-}
-
 int thread__set_comm(struct thread *self, const char *comm);
 int thread__comm_len(struct thread *self);
 struct thread *perf_session__findnew(struct perf_session *self, pid_t pid);

tools/perf/util/thread_map.c

+#include <dirent.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include "thread_map.h"
+
+/* Skip "." and ".." directories */
+static int filter(const struct dirent *dir)
+{
+	if (dir->d_name[0] == '.')
+		return 0;
+	else
+		return 1;
+}
+
+struct thread_map *thread_map__new_by_pid(pid_t pid)
+{
+	struct thread_map *threads;
+	char name[256];
+	int items;
+	struct dirent **namelist = NULL;
+	int i;
+
+	sprintf(name, "/proc/%d/task", pid);
+	items = scandir(name, &namelist, filter, NULL);
+	if (items <= 0)
+                return NULL;
+
+	threads = malloc(sizeof(*threads) + sizeof(pid_t) * items);
+	if (threads != NULL) {
+		for (i = 0; i < items; i++)
+			threads->map[i] = atoi(namelist[i]->d_name);
+		threads->nr = items;
+	}
+
+	for (i=0; i<items; i++)
+		free(namelist[i]);
+	free(namelist);
+
+	return threads;
+}
+
+struct thread_map *thread_map__new_by_tid(pid_t tid)
+{
+	struct thread_map *threads = malloc(sizeof(*threads) + sizeof(pid_t));
+
+	if (threads != NULL) {
+		threads->map[0] = tid;
+		threads->nr	= 1;
+	}
+
+	return threads;
+}
+
+struct thread_map *thread_map__new(pid_t pid, pid_t tid)
+{
+	if (pid != -1)
+		return thread_map__new_by_pid(pid);
+	return thread_map__new_by_tid(tid);
+}
+
+void thread_map__delete(struct thread_map *threads)
+{
+	free(threads);
+}

tools/perf/util/thread_map.h

+#ifndef __PERF_THREAD_MAP_H
+#define __PERF_THREAD_MAP_H
+
+#include <sys/types.h>
+
+struct thread_map {
+	int nr;
+	int map[];
+};
+
+struct thread_map *thread_map__new_by_pid(pid_t pid);
+struct thread_map *thread_map__new_by_tid(pid_t tid);
+struct thread_map *thread_map__new(pid_t pid, pid_t tid);
+void thread_map__delete(struct thread_map *threads);
+#endif	/* __PERF_THREAD_MAP_H */

tools/perf/util/ui/browsers/hists.c

@@ -377,7 +377,7 @@ static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self,
 	while (node) {
 		struct callchain_node *child = rb_entry(node, struct callchain_node, rb_node);
 		struct rb_node *next = rb_next(node);
-		u64 cumul = cumul_hits(child);
+		u64 cumul = callchain_cumul_hits(child);
 		struct callchain_list *chain;
 		char folded_sign = ' ';
 		int first = true;