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

Pull perf updates from Ingo Molnar.

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (21 commits)
  perf ui browser: Stop using 'self'
  perf annotate browser: Read perf config file for settings
  perf config: Allow '_' in config file variable names
  perf annotate browser: Make feature toggles global
  perf annotate browser: The idx_asm field should be used in asm only view
  perf tools: Convert critical messages to ui__error()
  perf ui: Make --stdio default when TUI is not supported
  tools lib traceevent: Silence compiler warning on 32bit build
  perf record: Fix branch_stack type in perf_record_opts
  perf tools: Reconstruct event with modifiers from perf_event_attr
  perf top: Fix counter name fixup when fallbacking to cpu-clock
  perf tools: fix thread_map__new_by_pid_str() memory leak in error path
  perf tools: Do not use _FORTIFY_SOURCE when DEBUG=1 is specified
  tools lib traceevent: Fix signature of create_arg_item()
  tools lib traceevent: Use proper function parameter type
  tools lib traceevent: Fix freeing arg on process_dynamic_array()
  tools lib traceevent: Fix a possibly wrong memory dereference
  tools lib traceevent: Fix a possible memory leak
  tools lib traceevent: Allow expressions in __print_symbolic() fields
  perf evlist: Explicititely initialize input_name
  ...

kernel/trace/ring_buffer.c

@@ -1486,6 +1486,11 @@ int ring_buffer_resize(struct ring_buffer *buffer, unsigned long size,
 	if (!buffer)
 		return size;
 
+	/* Make sure the requested buffer exists */
+	if (cpu_id != RING_BUFFER_ALL_CPUS &&
+	    !cpumask_test_cpu(cpu_id, buffer->cpumask))
+		return size;
+
 	size = DIV_ROUND_UP(size, BUF_PAGE_SIZE);
 	size *= BUF_PAGE_SIZE;
 

tools/lib/traceevent/event-parse.c

@@ -1434,8 +1434,11 @@ static int event_read_fields(struct event_format *event, struct format_field **f
 fail:
 	free_token(token);
 fail_expect:
-	if (field)
+	if (field) {
+		free(field->type);
+		free(field->name);
 		free(field);
+	}
 	return -1;
 }
 
@@ -1712,6 +1715,8 @@ process_op(struct event_format *event, struct print_arg *arg, char **tok)
 
 		if (set_op_prio(arg) == -1) {
 			event->flags |= EVENT_FL_FAILED;
+			/* arg->op.op (= token) will be freed at out_free */
+			arg->op.op = NULL;
 			goto out_free;
 		}
 
@@ -2124,6 +2129,13 @@ process_fields(struct event_format *event, struct print_flag_sym **list, char **
 
 		free_token(token);
 		type = process_arg(event, arg, &token);
+
+		if (type == EVENT_OP)
+			type = process_op(event, arg, &token);
+
+		if (type == EVENT_ERROR)
+			goto out_free;
+
 		if (test_type_token(type, token, EVENT_DELIM, ","))
 			goto out_free;
 
@@ -2288,17 +2300,18 @@ process_dynamic_array(struct event_format *event, struct print_arg *arg, char **
 	arg = alloc_arg();
 	type = process_arg(event, arg, &token);
 	if (type == EVENT_ERROR)
-		goto out_free;
+		goto out_free_arg;
 
 	if (!test_type_token(type, token, EVENT_OP, "]"))
-		goto out_free;
+		goto out_free_arg;
 
 	free_token(token);
 	type = read_token_item(tok);
 	return type;
 
+ out_free_arg:
+	free_arg(arg);
  out_free:
-	free(arg);
 	free_token(token);
 	*tok = NULL;
 	return EVENT_ERROR;
@@ -3362,6 +3375,7 @@ process_defined_func(struct trace_seq *s, void *data, int size,
 			break;
 		}
 		farg = farg->next;
+		param = param->next;
 	}
 
 	ret = (*func_handle->func)(s, args);

tools/lib/traceevent/parse-filter.c

@@ -325,9 +325,8 @@ static void free_events(struct event_list *events)
 }
 
 static struct filter_arg *
-create_arg_item(struct event_format *event,
-		const char *token, enum filter_arg_type type,
-		char **error_str)
+create_arg_item(struct event_format *event, const char *token,
+		enum event_type type, char **error_str)
 {
 	struct format_field *field;
 	struct filter_arg *arg;
@@ -1585,7 +1584,7 @@ get_value(struct event_format *event,
 		const char *name;
 
 		name = get_comm(event, record);
-		return (unsigned long long)name;
+		return (unsigned long)name;
 	}
 
 	pevent_read_number_field(field, record->data, &val);

tools/perf/Documentation/perfconfig.example

@@ -19,3 +19,11 @@
 
 	# Default, disable using /dev/null
 	dir = /root/.debug
+
+[annotate]
+
+	# Defaults
+	hide_src_code = false
+	use_offset = true
+	jump_arrows = true
+	show_nr_jumps = false

tools/perf/Makefile

@@ -80,7 +80,7 @@ ifeq ("$(origin DEBUG)", "command line")
   PERF_DEBUG = $(DEBUG)
 endif
 ifndef PERF_DEBUG
-  CFLAGS_OPTIMIZE = -O6
+  CFLAGS_OPTIMIZE = -O6 -D_FORTIFY_SOURCE=2
 endif
 
 ifdef PARSER_DEBUG
@@ -89,7 +89,7 @@ ifdef PARSER_DEBUG
 	PARSER_DEBUG_CFLAGS := -DPARSER_DEBUG
 endif
 
-CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_WERROR) $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) $(PARSER_DEBUG_CFLAGS)
+CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_WERROR) $(CFLAGS_OPTIMIZE) $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) $(PARSER_DEBUG_CFLAGS)
 EXTLIBS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE
 ALL_LDFLAGS = $(LDFLAGS)

tools/perf/builtin-annotate.c

@@ -215,7 +215,7 @@ static int __cmd_annotate(struct perf_annotate *ann)
 	}
 
 	if (total_nr_samples == 0) {
-		ui__warning("The %s file has no samples!\n", session->filename);
+		ui__error("The %s file has no samples!\n", session->filename);
 		goto out_delete;
 	}
 out_delete:

tools/perf/builtin-evlist.c

@@ -116,7 +116,7 @@ static const char * const evlist_usage[] = {
 int cmd_evlist(int argc, const char **argv, const char *prefix __used)
 {
 	struct perf_attr_details details = { .verbose = false, };
-	const char *input_name;
+	const char *input_name = NULL;
 	const struct option options[] = {
 		OPT_STRING('i', "input", &input_name, "file",
 			    "Input file name"),

tools/perf/builtin-record.c

@@ -264,7 +264,7 @@ static void perf_record__open(struct perf_record *rec)
 			}
 
 			if (err == ENOENT) {
-				ui__warning("The %s event is not supported.\n",
+				ui__error("The %s event is not supported.\n",
 					    event_name(pos));
 				exit(EXIT_FAILURE);
 			}
@@ -858,8 +858,8 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		usage_with_options(record_usage, record_options);
 
 	if (rec->force && rec->append_file) {
-		fprintf(stderr, "Can't overwrite and append at the same time."
-				" You need to choose between -f and -A");
+		ui__error("Can't overwrite and append at the same time."
+			  " You need to choose between -f and -A");
 		usage_with_options(record_usage, record_options);
 	} else if (rec->append_file) {
 		rec->write_mode = WRITE_APPEND;
@@ -868,8 +868,8 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 	}
 
 	if (nr_cgroups && !rec->opts.target.system_wide) {
-		fprintf(stderr, "cgroup monitoring only available in"
-			" system-wide mode\n");
+		ui__error("cgroup monitoring only available in"
+			  " system-wide mode\n");
 		usage_with_options(record_usage, record_options);
 	}
 
@@ -905,7 +905,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		int saved_errno = errno;
 
 		perf_target__strerror(&rec->opts.target, err, errbuf, BUFSIZ);
-		ui__warning("%s", errbuf);
+		ui__error("%s", errbuf);
 
 		err = -saved_errno;
 		goto out_free_fd;
@@ -933,7 +933,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 	else if (rec->opts.freq) {
 		rec->opts.default_interval = rec->opts.freq;
 	} else {
-		fprintf(stderr, "frequency and count are zero, aborting\n");
+		ui__error("frequency and count are zero, aborting\n");
 		err = -EINVAL;
 		goto out_free_fd;
 	}

tools/perf/builtin-report.c

@@ -251,13 +251,13 @@ static int perf_report__setup_sample_type(struct perf_report *rep)
 
 	if (!(self->sample_type & PERF_SAMPLE_CALLCHAIN)) {
 		if (sort__has_parent) {
-			ui__warning("Selected --sort parent, but no "
+			ui__error("Selected --sort parent, but no "
 				    "callchain data. Did you call "
 				    "'perf record' without -g?\n");
 			return -EINVAL;
 		}
 		if (symbol_conf.use_callchain) {
-			ui__warning("Selected -g but no callchain data. Did "
+			ui__error("Selected -g but no callchain data. Did "
 				    "you call 'perf record' without -g?\n");
 			return -1;
 		}
@@ -266,17 +266,15 @@ static int perf_report__setup_sample_type(struct perf_report *rep)
 		   !symbol_conf.use_callchain) {
 			symbol_conf.use_callchain = true;
 			if (callchain_register_param(&callchain_param) < 0) {
-				ui__warning("Can't register callchain "
-					    "params.\n");
+				ui__error("Can't register callchain params.\n");
 				return -EINVAL;
 			}
 	}
 
 	if (sort__branch_mode == 1) {
 		if (!(self->sample_type & PERF_SAMPLE_BRANCH_STACK)) {
-			fprintf(stderr, "selected -b but no branch data."
-					" Did you call perf record without"
-					" -b?\n");
+			ui__error("Selected -b but no branch data. "
+				  "Did you call perf record without -b?\n");
 			return -1;
 		}
 	}
@@ -420,7 +418,7 @@ static int __cmd_report(struct perf_report *rep)
 	}
 
 	if (nr_samples == 0) {
-		ui__warning("The %s file has no samples!\n", session->filename);
+		ui__error("The %s file has no samples!\n", session->filename);
 		goto out_delete;
 	}
 

tools/perf/builtin-top.c

@@ -953,22 +953,22 @@ static void perf_top__start_counters(struct perf_top *top)
 				attr->config = PERF_COUNT_SW_CPU_CLOCK;
 				if (counter->name) {
 					free(counter->name);
-					counter->name = strdup(event_name(counter));
+					counter->name = NULL;
 				}
 				goto try_again;
 			}
 
 			if (err == ENOENT) {
-				ui__warning("The %s event is not supported.\n",
+				ui__error("The %s event is not supported.\n",
 					    event_name(counter));
 				goto out_err;
 			} else if (err == EMFILE) {
-				ui__warning("Too many events are opened.\n"
+				ui__error("Too many events are opened.\n"
 					    "Try again after reducing the number of events\n");
 				goto out_err;
 			}
 
-			ui__warning("The sys_perf_event_open() syscall "
+			ui__error("The sys_perf_event_open() syscall "
 				    "returned with %d (%s).  /bin/dmesg "
 				    "may provide additional information.\n"
 				    "No CONFIG_PERF_EVENTS=y kernel support "
@@ -978,7 +978,7 @@ static void perf_top__start_counters(struct perf_top *top)
 	}
 
 	if (perf_evlist__mmap(evlist, top->mmap_pages, false) < 0) {
-		ui__warning("Failed to mmap with %d (%s)\n",
+		ui__error("Failed to mmap with %d (%s)\n",
 			    errno, strerror(errno));
 		goto out_err;
 	}
@@ -994,12 +994,12 @@ static int perf_top__setup_sample_type(struct perf_top *top)
 {
 	if (!top->sort_has_symbols) {
 		if (symbol_conf.use_callchain) {
-			ui__warning("Selected -g but \"sym\" not present in --sort/-s.");
+			ui__error("Selected -g but \"sym\" not present in --sort/-s.");
 			return -EINVAL;
 		}
 	} else if (!top->dont_use_callchains && callchain_param.mode != CHAIN_NONE) {
 		if (callchain_register_param(&callchain_param) < 0) {
-			ui__warning("Can't register callchain params.\n");
+			ui__error("Can't register callchain params.\n");
 			return -EINVAL;
 		}
 	}
@@ -1041,7 +1041,7 @@ static int __cmd_top(struct perf_top *top)
 
 	if (pthread_create(&thread, NULL, (use_browser > 0 ? display_thread_tui :
 							    display_thread), top)) {
-		printf("Could not create display thread.\n");
+		ui__error("Could not create display thread.\n");
 		exit(-1);
 	}
 
@@ -1050,7 +1050,7 @@ static int __cmd_top(struct perf_top *top)
 
 		param.sched_priority = top->realtime_prio;
 		if (sched_setscheduler(0, SCHED_FIFO, &param)) {
-			printf("Could not set realtime priority.\n");
+			ui__error("Could not set realtime priority.\n");
 			exit(-1);
 		}
 	}
@@ -1274,7 +1274,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 		int saved_errno = errno;
 
 		perf_target__strerror(&top.target, status, errbuf, BUFSIZ);
-		ui__warning("%s", errbuf);
+		ui__error("%s", errbuf);
 
 		status = -saved_errno;
 		goto out_delete_evlist;
@@ -1288,7 +1288,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 
 	if (!top.evlist->nr_entries &&
 	    perf_evlist__add_default(top.evlist) < 0) {
-		pr_err("Not enough memory for event selector list\n");
+		ui__error("Not enough memory for event selector list\n");
 		return -ENOMEM;
 	}
 
@@ -1305,7 +1305,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 	else if (top.freq) {
 		top.default_interval = top.freq;
 	} else {
-		fprintf(stderr, "frequency and count are zero, aborting\n");
+		ui__error("frequency and count are zero, aborting\n");
 		exit(EXIT_FAILURE);
 	}
 

tools/perf/perf.h

@@ -227,7 +227,7 @@ struct perf_record_opts {
 	unsigned int freq;
 	unsigned int mmap_pages;
 	unsigned int user_freq;
-	int	     branch_stack;
+	u64          branch_stack;
 	u64	     default_interval;
 	u64	     user_interval;
 };

tools/perf/ui/browser.h

@@ -69,4 +69,5 @@ void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whenc
 unsigned int ui_browser__list_head_refresh(struct ui_browser *self);
 
 void ui_browser__init(void);
+void annotate_browser__init(void);
 #endif /* _PERF_UI_BROWSER_H_ */

tools/perf/ui/setup.c

@@ -22,6 +22,7 @@ void setup_browser(bool fallback_to_pager)
 			break;
 		/* fall through */
 	default:
+		use_browser = 0;
 		if (fallback_to_pager)
 			setup_pager();
 		break;

tools/perf/util/config.c

@@ -120,7 +120,7 @@ static char *parse_value(void)
 
 static inline int iskeychar(int c)
 {
-	return isalnum(c) || c == '-';
+	return isalnum(c) || c == '-' || c == '_';
 }
 
 static int get_value(config_fn_t fn, void *data, char *name, unsigned int len)

tools/perf/util/evsel.c

@@ -15,6 +15,7 @@
 #include "cpumap.h"
 #include "thread_map.h"
 #include "target.h"
+#include "../../include/linux/perf_event.h"
 
 #define FD(e, x, y) (*(int *)xyarray__entry(e->fd, x, y))
 #define GROUP_FD(group_fd, cpu) (*(int *)xyarray__entry(group_fd, cpu, 0))
@@ -64,6 +65,95 @@ struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx)
 	return evsel;
 }
 
+static const char *perf_evsel__hw_names[PERF_COUNT_HW_MAX] = {
+	"cycles",
+	"instructions",
+	"cache-references",
+	"cache-misses",
+	"branches",
+	"branch-misses",
+	"bus-cycles",
+	"stalled-cycles-frontend",
+	"stalled-cycles-backend",
+	"ref-cycles",
+};
+
+const char *__perf_evsel__hw_name(u64 config)
+{
+	if (config < PERF_COUNT_HW_MAX && perf_evsel__hw_names[config])
+		return perf_evsel__hw_names[config];
+
+	return "unknown-hardware";
+}
+
+static int perf_evsel__hw_name(struct perf_evsel *evsel, char *bf, size_t size)
+{
+	int colon = 0;
+	struct perf_event_attr *attr = &evsel->attr;
+	int r = scnprintf(bf, size, "%s", __perf_evsel__hw_name(attr->config));
+	bool exclude_guest_default = false;
+
+#define MOD_PRINT(context, mod)	do {					\
+		if (!attr->exclude_##context) {				\
+			if (!colon) colon = r++;			\
+			r += scnprintf(bf + r, size - r, "%c", mod);	\
+		} } while(0)
+
+	if (attr->exclude_kernel || attr->exclude_user || attr->exclude_hv) {
+		MOD_PRINT(kernel, 'k');
+		MOD_PRINT(user, 'u');
+		MOD_PRINT(hv, 'h');
+		exclude_guest_default = true;
+	}
+
+	if (attr->precise_ip) {
+		if (!colon)
+			colon = r++;
+		r += scnprintf(bf + r, size - r, "%.*s", attr->precise_ip, "ppp");
+		exclude_guest_default = true;
+	}
+
+	if (attr->exclude_host || attr->exclude_guest == exclude_guest_default) {
+		MOD_PRINT(host, 'H');
+		MOD_PRINT(guest, 'G');
+	}
+#undef MOD_PRINT
+	if (colon)
+		bf[colon] = ':';
+	return r;
+}
+
+int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size)
+{
+	int ret;
+
+	switch (evsel->attr.type) {
+	case PERF_TYPE_RAW:
+		ret = scnprintf(bf, size, "raw 0x%" PRIx64, evsel->attr.config);
+		break;
+
+	case PERF_TYPE_HARDWARE:
+		ret = perf_evsel__hw_name(evsel, bf, size);
+		break;
+	default:
+		/*
+		 * FIXME
+ 		 *
+		 * This is the minimal perf_evsel__name so that we can
+		 * reconstruct event names taking into account event modifiers.
+		 *
+		 * The old event_name uses it now for raw anr hw events, so that
+		 * we don't drag all the parsing stuff into the python binding.
+		 *
+		 * On the next devel cycle the rest of the event naming will be
+		 * brought here.
+ 		 */
+		return 0;
+	}
+
+	return ret;
+}
+
 void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts,
 			struct perf_evsel *first)
 {

tools/perf/util/evsel.h

@@ -83,6 +83,9 @@ void perf_evsel__config(struct perf_evsel *evsel,
 			struct perf_record_opts *opts,
 			struct perf_evsel *first);
 
+const char* __perf_evsel__hw_name(u64 config);
+int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size);
+
 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);

tools/perf/util/parse-events.c

@@ -62,19 +62,6 @@ static struct event_symbol event_symbols[] = {
 #define PERF_EVENT_TYPE(config)		__PERF_EVENT_FIELD(config, TYPE)
 #define PERF_EVENT_ID(config)		__PERF_EVENT_FIELD(config, EVENT)
 
-static const char *hw_event_names[PERF_COUNT_HW_MAX] = {
-	"cycles",
-	"instructions",
-	"cache-references",
-	"cache-misses",
-	"branches",
-	"branch-misses",
-	"bus-cycles",
-	"stalled-cycles-frontend",
-	"stalled-cycles-backend",
-	"ref-cycles",
-};
-
 static const char *sw_event_names[PERF_COUNT_SW_MAX] = {
 	"cpu-clock",
 	"task-clock",
@@ -300,6 +287,16 @@ const char *event_name(struct perf_evsel *evsel)
 	u64 config = evsel->attr.config;
 	int type = evsel->attr.type;
 
+	if (type == PERF_TYPE_RAW || type == PERF_TYPE_HARDWARE) {
+		/*
+ 		 * XXX minimal fix, see comment on perf_evsen__name, this static buffer
+ 		 * will go away together with event_name in the next devel cycle.
+ 		 */
+		static char bf[128];
+		perf_evsel__name(evsel, bf, sizeof(bf));
+		return bf;
+	}
+
 	if (evsel->name)
 		return evsel->name;
 
@@ -317,9 +314,7 @@ const char *__event_name(int type, u64 config)
 
 	switch (type) {
 	case PERF_TYPE_HARDWARE:
-		if (config < PERF_COUNT_HW_MAX && hw_event_names[config])
-			return hw_event_names[config];
-		return "unknown-hardware";
+		return __perf_evsel__hw_name(config);
 
 	case PERF_TYPE_HW_CACHE: {
 		u8 cache_type, cache_op, cache_result;

tools/perf/util/thread_map.c

@@ -188,28 +188,27 @@ static struct thread_map *thread_map__new_by_pid_str(const char *pid_str)
 		nt = realloc(threads, (sizeof(*threads) +
 				       sizeof(pid_t) * total_tasks));
 		if (nt == NULL)
-			goto out_free_threads;
+			goto out_free_namelist;
 
 		threads = nt;
 
-		if (threads) {
-			for (i = 0; i < items; i++)
-				threads->map[j++] = atoi(namelist[i]->d_name);
-			threads->nr = total_tasks;
-		}
-
-		for (i = 0; i < items; i++)
+		for (i = 0; i < items; i++) {
+			threads->map[j++] = atoi(namelist[i]->d_name);
 			free(namelist[i]);
+		}
+		threads->nr = total_tasks;
 		free(namelist);
-
-		if (!threads)
-			break;
 	}
 
 out:
 	strlist__delete(slist);
 	return threads;
 
+out_free_namelist:
+	for (i = 0; i < items; i++)
+		free(namelist[i]);
+	free(namelist);
+
 out_free_threads:
 	free(threads);
 	threads = NULL;