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:

New features:

 - Add 'L' hotkey to dynamicly set the percent threshold for histogram
   entries and callchains, i.e. dynamicly do what the --percent-limit
   command line option to 'top' and 'report' does. (Namhyung Kim)

Infrastructure changes:

 - Per hists field and sort lists, that will be used, for instance,
   in the c2c tool (Jiri Olsa)

Documentation changes:

 - Update documentation of --sort and --perf-limit options
   for 'perf report' (Namhyung Kim)

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

tools/perf/Documentation/perf-report.txt

@@ -117,6 +117,22 @@ OPTIONS
 	And default sort keys are changed to comm, dso_from, symbol_from, dso_to
 	and symbol_to, see '--branch-stack'.
 
+	If the --mem-mode option is used, the following sort keys are also available
+	(incompatible with --branch-stack):
+	symbol_daddr, dso_daddr, locked, tlb, mem, snoop, dcacheline.
+
+	- symbol_daddr: name of data symbol being executed on at the time of sample
+	- dso_daddr: name of library or module containing the data being executed
+	on at the time of the sample
+	- locked: whether the bus was locked at the time of the sample
+	- tlb: type of tlb access for the data at the time of the sample
+	- mem: type of memory access for the data at the time of the sample
+	- snoop: type of snoop (if any) for the data at the time of the sample
+	- dcacheline: the cacheline the data address is on at the time of the sample
+
+	And the default sort keys are changed to local_weight, mem, sym, dso,
+	symbol_daddr, dso_daddr, snoop, tlb, locked, see '--mem-mode'.
+
 	If the data file has tracepoint event(s), following (dynamic) sort keys
 	are also available:
 	trace, trace_fields, [<event>.]<field>[/raw]
@@ -151,22 +167,6 @@ OPTIONS
 	By default, every sort keys not specified in -F will be appended
 	automatically.
 
-	If --mem-mode option is used, following sort keys are also available
-	(incompatible with --branch-stack):
-	symbol_daddr, dso_daddr, locked, tlb, mem, snoop, dcacheline.
-
-	- symbol_daddr: name of data symbol being executed on at the time of sample
-	- dso_daddr: name of library or module containing the data being executed
-	on at the time of sample
-	- locked: whether the bus was locked at the time of sample
-	- tlb: type of tlb access for the data at the time of sample
-	- mem: type of memory access for the data at the time of sample
-	- snoop: type of snoop (if any) for the data at the time of sample
-	- dcacheline: the cacheline the data address is on at the time of sample
-
-	And default sort keys are changed to local_weight, mem, sym, dso,
-	symbol_daddr, dso_daddr, snoop, tlb, locked, see '--mem-mode'.
-
 -p::
 --parent=<regex>::
         A regex filter to identify parent. The parent is a caller of this
@@ -351,7 +351,10 @@ OPTIONS
 
 --percent-limit::
 	Do not show entries which have an overhead under that percent.
-	(Default: 0).
+	(Default: 0).  Note that this option also sets the percent limit (threshold)
+	of callchains.  However the default value of callchain threshold is
+	different than the default value of hist entries.  Please see the
+	--call-graph option for details.
 
 --percentage::
 	Determine how to display the overhead percentage of filtered entries.

tools/perf/builtin-annotate.c

@@ -245,7 +245,7 @@ static int __cmd_annotate(struct perf_annotate *ann)
 			hists__collapse_resort(hists, NULL);
 			/* Don't sort callchain */
 			perf_evsel__reset_sample_bit(pos, CALLCHAIN);
-			hists__output_resort(hists, NULL);
+			perf_evsel__output_resort(pos, NULL);
 
 			if (symbol_conf.event_group &&
 			    !perf_evsel__is_group_leader(pos))

tools/perf/builtin-report.c

@@ -507,7 +507,7 @@ static void report__output_resort(struct report *rep)
 	ui_progress__init(&prog, rep->nr_entries, "Sorting events for output...");
 
 	evlist__for_each(rep->session->evlist, pos)
-		hists__output_resort(evsel__hists(pos), &prog);
+		perf_evsel__output_resort(pos, &prog);
 
 	ui_progress__finish();
 }
@@ -912,15 +912,6 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 		symbol_conf.cumulate_callchain = false;
 	}
 
-	if (setup_sorting(session->evlist) < 0) {
-		if (sort_order)
-			parse_options_usage(report_usage, options, "s", 1);
-		if (field_order)
-			parse_options_usage(sort_order ? NULL : report_usage,
-					    options, "F", 1);
-		goto error;
-	}
-
 	/* Force tty output for header output and per-thread stat. */
 	if (report.header || report.header_only || report.show_threads)
 		use_browser = 0;
@@ -930,6 +921,15 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 	else
 		use_browser = 0;
 
+	if (setup_sorting(session->evlist) < 0) {
+		if (sort_order)
+			parse_options_usage(report_usage, options, "s", 1);
+		if (field_order)
+			parse_options_usage(sort_order ? NULL : report_usage,
+					    options, "F", 1);
+		goto error;
+	}
+
 	if (report.header || report.header_only) {
 		perf_session__fprintf_info(session, stdout,
 					   report.show_full_info);

tools/perf/builtin-top.c

@@ -252,7 +252,8 @@ static void perf_top__print_sym_table(struct perf_top *top)
 	char bf[160];
 	int printed = 0;
 	const int win_width = top->winsize.ws_col - 1;
-	struct hists *hists = evsel__hists(top->sym_evsel);
+	struct perf_evsel *evsel = top->sym_evsel;
+	struct hists *hists = evsel__hists(evsel);
 
 	puts(CONSOLE_CLEAR);
 
@@ -288,7 +289,7 @@ static void perf_top__print_sym_table(struct perf_top *top)
 	}
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	hists__output_recalc_col_len(hists, top->print_entries - printed);
 	putchar('\n');
@@ -540,14 +541,15 @@ static bool perf_top__handle_keypress(struct perf_top *top, int c)
 static void perf_top__sort_new_samples(void *arg)
 {
 	struct perf_top *t = arg;
+	struct perf_evsel *evsel = t->sym_evsel;
 	struct hists *hists;
 
 	perf_top__reset_sample_counters(t);
 
 	if (t->evlist->selected != NULL)
 		t->sym_evsel = t->evlist->selected;
 
-	hists = evsel__hists(t->sym_evsel);
+	hists = evsel__hists(evsel);
 
 	if (t->evlist->enabled) {
 		if (t->zero) {
@@ -559,7 +561,7 @@ static void perf_top__sort_new_samples(void *arg)
 	}
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 }
 
 static void *display_thread_tui(void *arg)
@@ -1243,6 +1245,13 @@ int cmd_top(int argc, const char **argv, const char *prefix __maybe_unused)
 	/* display thread wants entries to be collapsed in a different tree */
 	sort__need_collapse = 1;
 
+	if (top.use_stdio)
+		use_browser = 0;
+	else if (top.use_tui)
+		use_browser = 1;
+
+	setup_browser(false);
+
 	if (setup_sorting(top.evlist) < 0) {
 		if (sort_order)
 			parse_options_usage(top_usage, options, "s", 1);
@@ -1252,13 +1261,6 @@ int cmd_top(int argc, const char **argv, const char *prefix __maybe_unused)
 		goto out_delete_evlist;
 	}
 
-	if (top.use_stdio)
-		use_browser = 0;
-	else if (top.use_tui)
-		use_browser = 1;
-
-	setup_browser(false);
-
 	status = target__validate(target);
 	if (status) {
 		target__strerror(target, status, errbuf, BUFSIZ);

tools/perf/tests/hists_cumulate.c

@@ -191,7 +191,7 @@ static int do_test(struct hists *hists, struct result *expected, size_t nr_expec
 	 * function since TEST_ASSERT_VAL() returns in case of failure.
 	 */
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(hists_to_evsel(hists), NULL);
 
 	if (verbose > 2) {
 		pr_info("use callchain: %d, cumulate callchain: %d\n",

tools/perf/tests/hists_filter.c

@@ -145,7 +145,7 @@ int test__hists_filter(int subtest __maybe_unused)
 		struct hists *hists = evsel__hists(evsel);
 
 		hists__collapse_resort(hists, NULL);
-		hists__output_resort(hists, NULL);
+		perf_evsel__output_resort(evsel, NULL);
 
 		if (verbose > 2) {
 			pr_info("Normal histogram\n");

tools/perf/tests/hists_output.c

@@ -156,7 +156,7 @@ static int test1(struct perf_evsel *evsel, struct machine *machine)
 		goto out;
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	if (verbose > 2) {
 		pr_info("[fields = %s, sort = %s]\n", field_order, sort_order);
@@ -256,7 +256,7 @@ static int test2(struct perf_evsel *evsel, struct machine *machine)
 		goto out;
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	if (verbose > 2) {
 		pr_info("[fields = %s, sort = %s]\n", field_order, sort_order);
@@ -310,7 +310,7 @@ static int test3(struct perf_evsel *evsel, struct machine *machine)
 		goto out;
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	if (verbose > 2) {
 		pr_info("[fields = %s, sort = %s]\n", field_order, sort_order);
@@ -388,7 +388,7 @@ static int test4(struct perf_evsel *evsel, struct machine *machine)
 		goto out;
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	if (verbose > 2) {
 		pr_info("[fields = %s, sort = %s]\n", field_order, sort_order);
@@ -491,7 +491,7 @@ static int test5(struct perf_evsel *evsel, struct machine *machine)
 		goto out;
 
 	hists__collapse_resort(hists, NULL);
-	hists__output_resort(hists, NULL);
+	perf_evsel__output_resort(evsel, NULL);
 
 	if (verbose > 2) {
 		pr_info("[fields = %s, sort = %s]\n", field_order, sort_order);

tools/perf/ui/browsers/hists.c

@@ -1095,7 +1095,7 @@ static int hist_browser__show_entry(struct hist_browser *browser,
 
 		hist_browser__gotorc(browser, row, 0);
 
-		perf_hpp__for_each_format(fmt) {
+		hists__for_each_format(browser->hists, fmt) {
 			if (perf_hpp__should_skip(fmt, entry->hists) ||
 			    column++ < browser->b.horiz_scroll)
 				continue;
@@ -1175,7 +1175,7 @@ static int hists_browser__scnprintf_headers(struct hist_browser *browser, char *
 			return ret;
 	}
 
-	perf_hpp__for_each_format(fmt) {
+	hists__for_each_format(browser->hists, fmt) {
 		if (perf_hpp__should_skip(fmt, hists)  || column++ < browser->b.horiz_scroll)
 			continue;
 
@@ -1441,7 +1441,7 @@ static int hist_browser__fprintf_entry(struct hist_browser *browser,
 	if (symbol_conf.use_callchain)
 		printed += fprintf(fp, "%c ", folded_sign);
 
-	perf_hpp__for_each_format(fmt) {
+	hists__for_each_format(browser->hists, fmt) {
 		if (perf_hpp__should_skip(fmt, he->hists))
 			continue;
 
@@ -2029,6 +2029,42 @@ static void hist_browser__update_nr_entries(struct hist_browser *hb)
 	hb->nr_non_filtered_entries = nr_entries;
 }
 
+static void hist_browser__update_percent_limit(struct hist_browser *hb,
+					       double percent)
+{
+	struct hist_entry *he;
+	struct rb_node *nd = rb_first(&hb->hists->entries);
+	u64 total = hists__total_period(hb->hists);
+	u64 min_callchain_hits = total * (percent / 100);
+
+	hb->min_pcnt = callchain_param.min_percent = percent;
+
+	if (!symbol_conf.use_callchain)
+		return;
+
+	while ((nd = hists__filter_entries(nd, hb->min_pcnt)) != NULL) {
+		he = rb_entry(nd, struct hist_entry, rb_node);
+
+		if (callchain_param.mode == CHAIN_GRAPH_REL) {
+			total = he->stat.period;
+
+			if (symbol_conf.cumulate_callchain)
+				total = he->stat_acc->period;
+
+			min_callchain_hits = total * (percent / 100);
+		}
+
+		callchain_param.sort(&he->sorted_chain, he->callchain,
+				     min_callchain_hits, &callchain_param);
+
+		/* force to re-evaluate folding state of callchains */
+		he->init_have_children = false;
+		hist_entry__set_folding(he, false);
+
+		nd = rb_next(nd);
+	}
+}
+
 static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events,
 				    const char *helpline,
 				    bool left_exits,
@@ -2064,6 +2100,7 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events,
 	"E             Expand all callchains\n"				\
 	"F             Toggle percentage of filtered entries\n"		\
 	"H             Display column headers\n"			\
+	"L             Change percent limit\n"				\
 	"m             Display context menu\n"				\
 	"S             Zoom into current Processor Socket\n"		\
 
@@ -2104,7 +2141,7 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events,
 	memset(options, 0, sizeof(options));
 	memset(actions, 0, sizeof(actions));
 
-	perf_hpp__for_each_format(fmt) {
+	hists__for_each_format(browser->hists, fmt) {
 		perf_hpp__reset_width(fmt, hists);
 		/*
 		 * This is done just once, and activates the horizontal scrolling
@@ -2219,6 +2256,24 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events,
 				top->zero = !top->zero;
 			}
 			continue;
+		case 'L':
+			if (ui_browser__input_window("Percent Limit",
+					"Please enter the value you want to hide entries under that percent.",
+					buf, "ENTER: OK, ESC: Cancel",
+					delay_secs * 2) == K_ENTER) {
+				char *end;
+				double new_percent = strtod(buf, &end);
+
+				if (new_percent < 0 || new_percent > 100) {
+					ui_browser__warning(&browser->b, delay_secs * 2,
+						"Invalid percent: %.2f", new_percent);
+					continue;
+				}
+
+				hist_browser__update_percent_limit(browser, new_percent);
+				hist_browser__reset(browser);
+			}
+			continue;
 		case K_F1:
 		case 'h':
 		case '?':

tools/perf/ui/gtk/hists.c

@@ -306,7 +306,7 @@ static void perf_gtk__show_hists(GtkWidget *window, struct hists *hists,
 
 	nr_cols = 0;
 
-	perf_hpp__for_each_format(fmt)
+	hists__for_each_format(hists, fmt)
 		col_types[nr_cols++] = G_TYPE_STRING;
 
 	store = gtk_tree_store_newv(nr_cols, col_types);
@@ -317,7 +317,7 @@ static void perf_gtk__show_hists(GtkWidget *window, struct hists *hists,
 
 	col_idx = 0;
 
-	perf_hpp__for_each_format(fmt) {
+	hists__for_each_format(hists, fmt) {
 		if (perf_hpp__should_skip(fmt, hists))
 			continue;
 
@@ -367,7 +367,7 @@ static void perf_gtk__show_hists(GtkWidget *window, struct hists *hists,
 
 		col_idx = 0;
 
-		perf_hpp__for_each_format(fmt) {
+		hists__for_each_format(hists, fmt) {
 			if (perf_hpp__should_skip(fmt, h->hists))
 				continue;