Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux…

…/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

User visible changes:

  - Fixes and improvements for supporting annotating ARM binaries, support ARM
    call and jump instructions, more work needed to have arch specific stuff
    separated into tools/perf/arch/*/annotate/ (Russell King)

  - Fix several 'perf test' entries broken by recent perf/core changes (Jiri Olsa)

Infrastructure changes:

  - Consolidate perf_ev{list,sel}__{enable,disable}() calls (Jiri Olsa)

  - Pass correct string to dso__adjust_kmod_long_name() (Wang Nan)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

tools/perf/builtin-stat.c

@@ -168,15 +168,25 @@ static int create_perf_stat_counter(struct perf_evsel *evsel)
 	attr->sample_period = 0;
 	attr->sample_type   = 0;
 
-	if (target__has_cpu(&target))
-		return perf_evsel__open_per_cpu(evsel, perf_evsel__cpus(evsel));
-
-	if (!target__has_task(&target) && perf_evsel__is_group_leader(evsel)) {
+	/*
+	 * Disabling all counters initially, they will be enabled
+	 * either manually by us or by kernel via enable_on_exec
+	 * set later.
+	 */
+	if (perf_evsel__is_group_leader(evsel)) {
 		attr->disabled = 1;
-		if (!initial_delay)
+
+		/*
+		 * In case of initial_delay we enable tracee
+		 * events manually.
+		 */
+		if (target__none(&target) && !initial_delay)
 			attr->enable_on_exec = 1;
 	}
 
+	if (target__has_cpu(&target))
+		return perf_evsel__open_per_cpu(evsel, perf_evsel__cpus(evsel));
+
 	return perf_evsel__open_per_thread(evsel, evsel_list->threads);
 }
 
@@ -251,18 +261,18 @@ static void process_interval(void)
 	print_counters(&rs, 0, NULL);
 }
 
-static void handle_initial_delay(void)
+static void enable_counters(void)
 {
-	struct perf_evsel *counter;
-
-	if (initial_delay) {
-		const int ncpus = cpu_map__nr(evsel_list->cpus),
-			nthreads = thread_map__nr(evsel_list->threads);
-
+	if (initial_delay)
 		usleep(initial_delay * 1000);
-		evlist__for_each(evsel_list, counter)
-			perf_evsel__enable(counter, ncpus, nthreads);
-	}
+
+	/*
+	 * We need to enable counters only if:
+	 * - we don't have tracee (attaching to task or cpu)
+	 * - we have initial delay configured
+	 */
+	if (!target__none(&target) || initial_delay)
+		perf_evlist__enable(evsel_list);
 }
 
 static volatile int workload_exec_errno;
@@ -359,7 +369,7 @@ static int __run_perf_stat(int argc, const char **argv)
 
 	if (forks) {
 		perf_evlist__start_workload(evsel_list);
-		handle_initial_delay();
+		enable_counters();
 
 		if (interval) {
 			while (!waitpid(child_pid, &status, WNOHANG)) {
@@ -378,7 +388,7 @@ static int __run_perf_stat(int argc, const char **argv)
 		if (WIFSIGNALED(status))
 			psignal(WTERMSIG(status), argv[0]);
 	} else {
-		handle_initial_delay();
+		enable_counters();
 		while (!done) {
 			nanosleep(&ts, NULL);
 			if (interval)

tools/perf/tests/code-reading.c

@@ -433,7 +433,6 @@ enum {
 
 static int do_test_code_reading(bool try_kcore)
 {
-	struct machines machines;
 	struct machine *machine;
 	struct thread *thread;
 	struct record_opts opts = {
@@ -459,8 +458,7 @@ static int do_test_code_reading(bool try_kcore)
 
 	pid = getpid();
 
-	machines__init(&machines);
-	machine = &machines.host;
+	machine = machine__new_host();
 
 	ret = machine__create_kernel_maps(machine);
 	if (ret < 0) {
@@ -549,6 +547,13 @@ static int do_test_code_reading(bool try_kcore)
 		if (ret < 0) {
 			if (!excl_kernel) {
 				excl_kernel = true;
+				/*
+				 * Both cpus and threads are now owned by evlist
+				 * and will be freed by following perf_evlist__set_maps
+				 * call. Getting refference to keep them alive.
+				 */
+				cpu_map__get(cpus);
+				thread_map__get(threads);
 				perf_evlist__set_maps(evlist, NULL, NULL);
 				perf_evlist__delete(evlist);
 				evlist = NULL;
@@ -594,9 +599,8 @@ static int do_test_code_reading(bool try_kcore)
 		cpu_map__put(cpus);
 		thread_map__put(threads);
 	}
-	machines__destroy_kernel_maps(&machines);
 	machine__delete_threads(machine);
-	machines__exit(&machines);
+	machine__delete(machine);
 
 	return err;
 }

tools/perf/tests/dwarf-unwind.c

@@ -160,14 +160,11 @@ static int krava_1(struct thread *thread)
 
 int test__dwarf_unwind(int subtest __maybe_unused)
 {
-	struct machines machines;
 	struct machine *machine;
 	struct thread *thread;
 	int err = -1;
 
-	machines__init(&machines);
-
-	machine = machines__find(&machines, HOST_KERNEL_ID);
+	machine = machine__new_host();
 	if (!machine) {
 		pr_err("Could not get machine\n");
 		return -1;
@@ -199,7 +196,6 @@ int test__dwarf_unwind(int subtest __maybe_unused)
 
  out:
 	machine__delete_threads(machine);
-	machine__exit(machine);
-	machines__exit(&machines);
+	machine__delete(machine);
 	return err;
 }

tools/perf/tests/evsel-roundtrip-name.c

@@ -103,7 +103,8 @@ int test__perf_evsel__roundtrip_name_test(int subtest __maybe_unused)
 	if (err)
 		ret = err;
 
-	err = perf_evsel__name_array_test(perf_evsel__sw_names);
+	err = __perf_evsel__name_array_test(perf_evsel__sw_names,
+					    PERF_COUNT_SW_DUMMY + 1);
 	if (err)
 		ret = err;
 

tools/perf/tests/hists_common.c

@@ -87,6 +87,11 @@ struct machine *setup_fake_machine(struct machines *machines)
 		return NULL;
 	}
 
+	if (machine__create_kernel_maps(machine)) {
+		pr_debug("Not enough memory for machine setup\n");
+		goto out;
+	}
+
 	for (i = 0; i < ARRAY_SIZE(fake_threads); i++) {
 		struct thread *thread;
 

tools/perf/tests/mmap-thread-lookup.c

@@ -149,7 +149,6 @@ static int synth_process(struct machine *machine)
 
 static int mmap_events(synth_cb synth)
 {
-	struct machines machines;
 	struct machine *machine;
 	int err, i;
 
@@ -162,8 +161,7 @@ static int mmap_events(synth_cb synth)
 	 */
 	TEST_ASSERT_VAL("failed to create threads", !threads_create());
 
-	machines__init(&machines);
-	machine = &machines.host;
+	machine = machine__new_host();
 
 	dump_trace = verbose > 1 ? 1 : 0;
 
@@ -203,7 +201,7 @@ static int mmap_events(synth_cb synth)
 	}
 
 	machine__delete_threads(machine);
-	machines__exit(&machines);
+	machine__delete(machine);
 	return err;
 }
 

tools/perf/util/annotate.c

@@ -65,6 +65,11 @@ static int call__parse(struct ins_operands *ops)
 
 	name++;
 
+#ifdef __arm__
+	if (strchr(name, '+'))
+		return -1;
+#endif
+
 	tok = strchr(name, '>');
 	if (tok == NULL)
 		return -1;
@@ -246,7 +251,11 @@ static int mov__parse(struct ins_operands *ops)
 		return -1;
 
 	target = ++s;
+#ifdef __arm__
+	comment = strchr(s, ';');
+#else
 	comment = strchr(s, '#');
+#endif
 
 	if (comment != NULL)
 		s = comment - 1;
@@ -354,6 +363,20 @@ static struct ins instructions[] = {
 	{ .name = "addq",  .ops  = &mov_ops, },
 	{ .name = "addw",  .ops  = &mov_ops, },
 	{ .name = "and",   .ops  = &mov_ops, },
+#ifdef __arm__
+	{ .name = "b",     .ops  = &jump_ops, }, // might also be a call
+	{ .name = "bcc",   .ops  = &jump_ops, },
+	{ .name = "bcs",   .ops  = &jump_ops, },
+	{ .name = "beq",   .ops  = &jump_ops, },
+	{ .name = "bge",   .ops  = &jump_ops, },
+	{ .name = "bgt",   .ops  = &jump_ops, },
+	{ .name = "bhi",   .ops  = &jump_ops, },
+	{ .name = "bl",    .ops  = &call_ops, },
+	{ .name = "blt",   .ops  = &jump_ops, },
+	{ .name = "bls",   .ops  = &jump_ops, },
+	{ .name = "blx",   .ops  = &call_ops, },
+	{ .name = "bne",   .ops  = &jump_ops, },
+#endif
 	{ .name = "bts",   .ops  = &mov_ops, },
 	{ .name = "call",  .ops  = &call_ops, },
 	{ .name = "callq", .ops  = &call_ops, },

tools/perf/util/evlist.c

@@ -336,41 +336,25 @@ static int perf_evlist__nr_threads(struct perf_evlist *evlist,
 
 void perf_evlist__disable(struct perf_evlist *evlist)
 {
-	int cpu, thread;
 	struct perf_evsel *pos;
-	int nr_cpus = cpu_map__nr(evlist->cpus);
-	int nr_threads;
 
-	for (cpu = 0; cpu < nr_cpus; cpu++) {
-		evlist__for_each(evlist, pos) {
-			if (!perf_evsel__is_group_leader(pos) || !pos->fd)
-				continue;
-			nr_threads = perf_evlist__nr_threads(evlist, pos);
-			for (thread = 0; thread < nr_threads; thread++)
-				ioctl(FD(pos, cpu, thread),
-				      PERF_EVENT_IOC_DISABLE, 0);
-		}
+	evlist__for_each(evlist, pos) {
+		if (!perf_evsel__is_group_leader(pos) || !pos->fd)
+			continue;
+		perf_evsel__disable(pos);
 	}
 
 	evlist->enabled = false;
 }
 
 void perf_evlist__enable(struct perf_evlist *evlist)
 {
-	int cpu, thread;
 	struct perf_evsel *pos;
-	int nr_cpus = cpu_map__nr(evlist->cpus);
-	int nr_threads;
 
-	for (cpu = 0; cpu < nr_cpus; cpu++) {
-		evlist__for_each(evlist, pos) {
-			if (!perf_evsel__is_group_leader(pos) || !pos->fd)
-				continue;
-			nr_threads = perf_evlist__nr_threads(evlist, pos);
-			for (thread = 0; thread < nr_threads; thread++)
-				ioctl(FD(pos, cpu, thread),
-				      PERF_EVENT_IOC_ENABLE, 0);
-		}
+	evlist__for_each(evlist, pos) {
+		if (!perf_evsel__is_group_leader(pos) || !pos->fd)
+			continue;
+		perf_evsel__enable(pos);
 	}
 
 	evlist->enabled = true;

tools/perf/util/evsel.c

@@ -981,13 +981,26 @@ int perf_evsel__append_filter(struct perf_evsel *evsel,
 	return -1;
 }
 
-int perf_evsel__enable(struct perf_evsel *evsel, int ncpus, int nthreads)
+int perf_evsel__enable(struct perf_evsel *evsel)
 {
+	int nthreads = thread_map__nr(evsel->threads);
+	int ncpus = cpu_map__nr(evsel->cpus);
+
 	return perf_evsel__run_ioctl(evsel, ncpus, nthreads,
 				     PERF_EVENT_IOC_ENABLE,
 				     0);
 }
 
+int perf_evsel__disable(struct perf_evsel *evsel)
+{
+	int nthreads = thread_map__nr(evsel->threads);
+	int ncpus = cpu_map__nr(evsel->cpus);
+
+	return perf_evsel__run_ioctl(evsel, ncpus, nthreads,
+				     PERF_EVENT_IOC_DISABLE,
+				     0);
+}
+
 int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads)
 {
 	if (ncpus == 0 || nthreads == 0)

tools/perf/util/evsel.h

@@ -227,7 +227,8 @@ int perf_evsel__append_filter(struct perf_evsel *evsel,
 			      const char *op, const char *filter);
 int perf_evsel__apply_filter(struct perf_evsel *evsel, int ncpus, int nthreads,
 			     const char *filter);
-int perf_evsel__enable(struct perf_evsel *evsel, int ncpus, int nthreads);
+int perf_evsel__enable(struct perf_evsel *evsel);
+int perf_evsel__disable(struct perf_evsel *evsel);
 
 int perf_evsel__open_per_cpu(struct perf_evsel *evsel,
 			     struct cpu_map *cpus);

tools/perf/util/machine.c

@@ -576,7 +576,7 @@ static void dso__adjust_kmod_long_name(struct dso *dso, const char *filename)
 	if (!dup_filename)
 		return;
 
-	dso__set_long_name(dso, filename, true);
+	dso__set_long_name(dso, dup_filename, true);
 }
 
 struct map *machine__findnew_module_map(struct machine *machine, u64 start,