perf report: Show call graph from reference events

Introduce --show-ref-call-graph for perf report to print reference
callgraph for no callgraph event.

Here is an example.

 perf report --show-ref-call-graph --stdio

 # To display the perf.data header info, please use
 --header/--header-only options.
 #
 #
 # Total Lost Samples: 0
 #
 # Samples: 5  of event 'cpu/cpu-cycles,call-graph=fp/'
 # Event count (approx.): 144985
 #
 # Children      Self  Command  Shared Object     Symbol
 # ........  ........  .......  ................  ........................................
 #
    72.30%     0.00%  sleep    [kernel.vmlinux]  [k] entry_SYSCALL_64_fastpath
              |
              ---entry_SYSCALL_64_fastpath
                 |
                 |--22.62%-- __GI___libc_nanosleep
                  --77.38%-- [...]

......

 # Samples: 6  of event 'cpu/instructions,call-graph=no/', show reference callgraph
 # Event count (approx.): 172780
 #
 # Children      Self  Command  Shared Object     Symbol
 # ........  ........  .......  ................  ........................................
 #
    73.16%     0.00%  sleep    [kernel.vmlinux]  [k] entry_SYSCALL_64_fastpath
              |
              ---entry_SYSCALL_64_fastpath
                 |
                 |--31.44%-- __GI___libc_nanosleep
                  --68.56%-- [...]

Signed-off-by: Kan Liang <kan.liang@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lkml.kernel.org/r/1439289050-40510-3-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools/perf/Documentation/perf-report.txt

@@ -359,6 +359,17 @@ OPTIONS
 --full-source-path::
 	Show the full path for source files for srcline output.
 
+--show-ref-call-graph::
+	When multiple events are sampled, it may not be needed to collect
+	callgraphs for all of them. The sample sites are usually nearby,
+	and it's enough to collect the callgraphs on a reference event.
+	So user can use "call-graph=no" event modifier to disable callgraph
+	for other events to reduce the overhead.
+	However, perf report cannot show callgraphs for the event which
+	disable the callgraph.
+	This option extends the perf report to show reference callgraphs,
+	which collected by reference event, in no callgraph event.
+
 include::callchain-overhead-calculation.txt[]
 
 SEE ALSO

tools/perf/builtin-report.c

@@ -316,6 +316,11 @@ static size_t hists__fprintf_nr_sample_events(struct hists *hists, struct report
 	if (evname != NULL)
 		ret += fprintf(fp, " of event '%s'", evname);
 
+	if (symbol_conf.show_ref_callgraph &&
+	    strstr(evname, "call-graph=no")) {
+		ret += fprintf(fp, ", show reference callgraph");
+	}
+
 	if (rep->mem_mode) {
 		ret += fprintf(fp, "\n# Total weight : %" PRIu64, nr_events);
 		ret += fprintf(fp, "\n# Sort order   : %s", sort_order ? : default_mem_sort_order);
@@ -740,6 +745,8 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 			    itrace_parse_synth_opts),
 	OPT_BOOLEAN(0, "full-source-path", &srcline_full_filename,
 			"Show full source file name path for source lines"),
+	OPT_BOOLEAN(0, "show-ref-call-graph", &symbol_conf.show_ref_callgraph,
+		    "Show callgraph from reference event"),
 	OPT_END()
 	};
 	struct perf_data_file file = {

tools/perf/ui/browsers/hists.c

@@ -1267,6 +1267,8 @@ static int hists__browser_title(struct hists *hists,
 	const char *ev_name = perf_evsel__name(evsel);
 	char buf[512];
 	size_t buflen = sizeof(buf);
+	char ref[30] = " show reference callgraph, ";
+	bool enable_ref = false;
 
 	if (symbol_conf.filter_relative) {
 		nr_samples = hists->stats.nr_non_filtered_samples;
@@ -1292,10 +1294,13 @@ static int hists__browser_title(struct hists *hists,
 		}
 	}
 
+	if (symbol_conf.show_ref_callgraph &&
+	    strstr(ev_name, "call-graph=no"))
+		enable_ref = true;
 	nr_samples = convert_unit(nr_samples, &unit);
 	printed = scnprintf(bf, size,
-			   "Samples: %lu%c of event '%s', Event count (approx.): %" PRIu64,
+			   "Samples: %lu%c of event '%s',%sEvent count (approx.): %" PRIu64,
-			   nr_samples, unit, ev_name, nr_events);
+			   nr_samples, unit, ev_name, enable_ref ? ref : " ", nr_events);
 
 
 	if (hists->uid_filter_str)

tools/perf/util/hist.c

@@ -1141,7 +1141,12 @@ void hists__output_resort(struct hists *hists, struct ui_progress *prog)
 	struct hist_entry *n;
 	u64 min_callchain_hits;
 	struct perf_evsel *evsel = hists_to_evsel(hists);
-	bool use_callchain = evsel ? (evsel->attr.sample_type & PERF_SAMPLE_CALLCHAIN) : symbol_conf.use_callchain;
+	bool use_callchain;
+
+	if (evsel && !symbol_conf.show_ref_callgraph)
+		use_callchain = evsel->attr.sample_type & PERF_SAMPLE_CALLCHAIN;
+	else
+		use_callchain = symbol_conf.use_callchain;
 
 	min_callchain_hits = hists->stats.total_period * (callchain_param.min_percent / 100);
 

tools/perf/util/symbol.h

@@ -106,7 +106,8 @@ struct symbol_conf {
 			filter_relative,
 			show_hist_headers,
 			branch_callstack,
-			has_filter;
+			has_filter,
+			show_ref_callgraph;
 	const char	*vmlinux_name,
 			*kallsyms_name,
 			*source_prefix,