[PATCH 12/31] perf tools: Allow filter option to be applied to bof object

[Wang Nan <wangnan0@huawei.com>]

Before this patch, --filter options can't be applied to BPF object
'events'. For example, the following command:

 # perf record -e cycles -e test_bpf.o --exclude-perf -a sleep 1

doesn't apply '--exclude-perf' to events in test_bpf.o. Instead, the
filter will be applied to 'cycles' event. This is caused by the delay
manner of adding real BPF events. Because all BPF probing points are
probed by one call, we can't add real events until all BPF objects
are collected. In previous patch (perf tools: Enable passing bpf object
file to --event), nothing is appended to evlist.

This patch fixes this by utilizing the dummy event linked during
parse_events(). Filter settings goes to dummy event, and be synced with
real events in add_bpf_event().

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David Ahern <dsahern@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kaixu Xia <xiakaixu@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Link: http://lkml.kernel.org/r/1440742821-44548-5-git-send-email-wangnan0@huawei.com
---
 tools/perf/builtin-record.c  |  6 ++++-
 tools/perf/util/bpf-loader.c |  8 ++++++-
 tools/perf/util/bpf-loader.h |  2 ++
 tools/perf/util/evlist.c     | 53 +++++++++++++++++++++++++++++++++++++++++---
 4 files changed, 64 insertions(+), 5 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 5051d3b..fd56a5b 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -1113,7 +1113,6 @@ int cmd_record(int argc, const char **argv, const char *prefix __maybe_unused)
 
 	argc = parse_options(argc, argv, record_options, record_usage,
 			    PARSE_OPT_STOP_AT_NON_OPTION);
-	perf_evlist__purge_dummy(rec->evlist);
 
 	if (!argc && target__none(&rec->opts.target))
 		usage_with_options(record_usage, record_options);
@@ -1178,6 +1177,11 @@ int cmd_record(int argc, const char **argv, const char *prefix __maybe_unused)
 		pr_err("Failed to add events from BPF object(s)\n");
 		goto out_symbol_exit;
 	}
+	/*
+	 * Until now let's purge dummy event. Filter options should
+	 * have been attached to real events by perf_evlist__add_bpf().
+	 */
+	perf_evlist__purge_dummy(rec->evlist);
 
 	symbol__init(NULL);
 
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 126aa71..c3bc0a8 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -293,6 +293,12 @@ int bpf__foreach_tev(bpf_prog_iter_callback_t func, void *arg)
 	int err;
 
 	bpf_object__for_each_safe(obj, tmp) {
+		const char *obj_name;
+
+		obj_name = bpf_object__get_name(obj);
+		if (!obj_name)
+			obj_name = "[unknown]";
+
 		bpf_object__for_each_program(prog, obj) {
 			struct probe_trace_event *tev;
 			struct perf_probe_event *pev;
@@ -316,7 +322,7 @@ int bpf__foreach_tev(bpf_prog_iter_callback_t func, void *arg)
 					return fd;
 				}
 
-				err = func(tev, fd, arg);
+				err = func(tev, obj_name, fd, arg);
 				if (err) {
 					pr_debug("bpf: call back failed, stop iterate\n");
 					return err;
diff --git a/tools/perf/util/bpf-loader.h b/tools/perf/util/bpf-loader.h
index 34656f8..323e664 100644
--- a/tools/perf/util/bpf-loader.h
+++ b/tools/perf/util/bpf-loader.h
@@ -6,6 +6,7 @@
 #define __BPF_LOADER_H
 
 #include <linux/compiler.h>
+#include <linux/perf_event.h>
 #include <string.h>
 #include "probe-event.h"
 #include "debug.h"
@@ -13,6 +14,7 @@
 #define PERF_BPF_PROBE_GROUP "perf_bpf_probe"
 
 typedef int (*bpf_prog_iter_callback_t)(struct probe_trace_event *tev,
+					const char *obj_name,
 					int fd, void *arg);
 
 #ifdef HAVE_LIBBPF_SUPPORT
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index f79bbf8..c00e939 100644
--- a/tools/perf/util/evlist.c
+++ b/tools/perf/util/evlist.c
@@ -197,7 +197,45 @@ error:
 	return -ENOMEM;
 }
 
-static int add_bpf_event(struct probe_trace_event *tev, int fd,
+static void
+sync_with_dummy(struct perf_evlist *evlist, const char *obj_name,
+		struct list_head *list)
+{
+	struct perf_evsel *dummy_evsel, *pos;
+	const char *filter;
+	bool found = false;
+	int err;
+
+	evlist__for_each(evlist, dummy_evsel) {
+		if (!perf_evsel__is_dummy(dummy_evsel))
+			continue;
+
+		if (strcmp(dummy_evsel->name, obj_name) == 0) {
+			found = true;
+			break;
+		}
+	}
+
+	if (!found) {
+		pr_debug("Failed to find dummy event of '%s'\n",
+			 obj_name);
+		return;
+	}
+
+	filter = dummy_evsel->filter;
+	if (!filter)
+		return;
+
+	list_for_each_entry(pos, list, node) {
+		err = perf_evsel__set_filter(pos, filter);
+		if (err)
+			pr_debug("Failed to set filter '%s' to evsel %s\n",
+				 filter, pos->name);
+	}
+}
+
+static int add_bpf_event(struct probe_trace_event *tev,
+			 const char *obj_name, int fd,
 			 void *arg)
 {
 	struct perf_evlist *evlist = arg;
@@ -205,8 +243,8 @@ static int add_bpf_event(struct probe_trace_event *tev, int fd,
 	struct list_head list;
 	int err, idx, entries;
 
-	pr_debug("add bpf event %s:%s and attach bpf program %d\n",
-			tev->group, tev->event, fd);
+	pr_debug("add bpf event %s:%s and attach bpf program %d (from %s)\n",
+			tev->group, tev->event, fd, obj_name);
 	INIT_LIST_HEAD(&list);
 	idx = evlist->nr_entries;
 
@@ -228,6 +266,15 @@ static int add_bpf_event(struct probe_trace_event *tev, int fd,
 	list_for_each_entry(pos, &list, node)
 		pos->bpf_fd = fd;
 	entries = idx - evlist->nr_entries;
+
+	sync_with_dummy(evlist, obj_name, &list);
+
+	/*
+	 * Currectly we don't need to link those new events at the
+	 * same place where dummy node reside because order of
+	 * events in cmdline won't be used after
+	 * 'perf_evlist__add_bpf'.
+	 */
 	perf_evlist__splice_list_tail(evlist, &list, entries);
 	return 0;
 }
-- 
2.1.0