1 files changed, 87 insertions, 27 deletions

diff --git a/tools/perf/util/bpf_ftrace.c b/tools/perf/util/bpf_ftrace.c
index 7a4297d8fd2c..c456d24efa30 100644
--- a/tools/perf/util/bpf_ftrace.c
+++ b/tools/perf/util/bpf_ftrace.c
@@ -1,8 +1,10 @@
-#include <stdio.h>
+#include <errno.h>
 #include <fcntl.h>
 #include <stdint.h>
+#include <stdio.h>
 #include <stdlib.h>
 
+#include <bpf/bpf.h>
 #include <linux/err.h>
 
 #include "util/ftrace.h"
@@ -11,6 +13,7 @@
 #include "util/debug.h"
 #include "util/evlist.h"
 #include "util/bpf_counter.h"
+#include "util/stat.h"
 
 #include "util/bpf_skel/func_latency.skel.h"
 
@@ -20,15 +23,26 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)
 {
 	int fd, err;
 	int i, ncpus = 1, ntasks = 1;
-	struct filter_entry *func;
+	struct filter_entry *func = NULL;
 
-	if (!list_is_singular(&ftrace->filters)) {
-		pr_err("ERROR: %s target function(s).\n",
-		       list_empty(&ftrace->filters) ? "No" : "Too many");
-		return -1;
-	}
+	if (!list_empty(&ftrace->filters)) {
+		if (!list_is_singular(&ftrace->filters)) {
+			pr_err("ERROR: Too many target functions.\n");
+			return -1;
+		}
+		func = list_first_entry(&ftrace->filters, struct filter_entry, list);
+	} else {
+		int count = 0;
+		struct list_head *pos;
 
-	func = list_first_entry(&ftrace->filters, struct filter_entry, list);
+		list_for_each(pos, &ftrace->event_pair)
+			count++;
+
+		if (count != 2) {
+			pr_err("ERROR: Needs two target events.\n");
+			return -1;
+		}
+	}
 
 	skel = func_latency_bpf__open();
 	if (!skel) {
@@ -36,17 +50,28 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)
 		return -1;
 	}
 
+	skel->rodata->bucket_range = ftrace->bucket_range;
+	skel->rodata->min_latency = ftrace->min_latency;
+	skel->rodata->bucket_num = ftrace->bucket_num;
+	if (ftrace->bucket_range && ftrace->bucket_num) {
+		bpf_map__set_max_entries(skel->maps.latency, ftrace->bucket_num);
+	}
+
 	/* don't need to set cpu filter for system-wide mode */
 	if (ftrace->target.cpu_list) {
 		ncpus = perf_cpu_map__nr(ftrace->evlist->core.user_requested_cpus);
 		bpf_map__set_max_entries(skel->maps.cpu_filter, ncpus);
+		skel->rodata->has_cpu = 1;
 	}
 
 	if (target__has_task(&ftrace->target) || target__none(&ftrace->target)) {
 		ntasks = perf_thread_map__nr(ftrace->evlist->core.threads);
 		bpf_map__set_max_entries(skel->maps.task_filter, ntasks);
+		skel->rodata->has_task = 1;
 	}
 
+	skel->rodata->use_nsec = ftrace->use_nsec;
+
 	set_max_rlimit();
 
 	err = func_latency_bpf__load(skel);
@@ -59,7 +84,6 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)
 		u32 cpu;
 		u8 val = 1;
 
-		skel->bss->has_cpu = 1;
 		fd = bpf_map__fd(skel->maps.cpu_filter);
 
 		for (i = 0; i < ncpus; i++) {
@@ -72,7 +96,6 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)
 		u32 pid;
 		u8 val = 1;
 
-		skel->bss->has_task = 1;
 		fd = bpf_map__fd(skel->maps.task_filter);
 
 		for (i = 0; i < ntasks; i++) {
@@ -81,22 +104,46 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)
 		}
 	}
 
-	skel->bss->use_nsec = ftrace->use_nsec;
+	skel->bss->min = INT64_MAX;
 
-	skel->links.func_begin = bpf_program__attach_kprobe(skel->progs.func_begin,
-							    false, func->name);
-	if (IS_ERR(skel->links.func_begin)) {
-		pr_err("Failed to attach fentry program\n");
-		err = PTR_ERR(skel->links.func_begin);
-		goto out;
-	}
+	if (func) {
+		skel->links.func_begin = bpf_program__attach_kprobe(skel->progs.func_begin,
+								    false, func->name);
+		if (IS_ERR(skel->links.func_begin)) {
+			pr_err("Failed to attach fentry program\n");
+			err = PTR_ERR(skel->links.func_begin);
+			goto out;
+		}
 
-	skel->links.func_end = bpf_program__attach_kprobe(skel->progs.func_end,
-							  true, func->name);
-	if (IS_ERR(skel->links.func_end)) {
-		pr_err("Failed to attach fexit program\n");
-		err = PTR_ERR(skel->links.func_end);
-		goto out;
+		skel->links.func_end = bpf_program__attach_kprobe(skel->progs.func_end,
+								  true, func->name);
+		if (IS_ERR(skel->links.func_end)) {
+			pr_err("Failed to attach fexit program\n");
+			err = PTR_ERR(skel->links.func_end);
+			goto out;
+		}
+	} else {
+		struct filter_entry *event;
+
+		event = list_first_entry(&ftrace->event_pair, struct filter_entry, list);
+
+		skel->links.event_begin = bpf_program__attach_raw_tracepoint(skel->progs.event_begin,
+									     event->name);
+		if (IS_ERR(skel->links.event_begin)) {
+			pr_err("Failed to attach first tracepoint program\n");
+			err = PTR_ERR(skel->links.event_begin);
+			goto out;
+		}
+
+		event = list_next_entry(event, list);
+
+		skel->links.event_end = bpf_program__attach_raw_tracepoint(skel->progs.event_end,
+									     event->name);
+		if (IS_ERR(skel->links.event_end)) {
+			pr_err("Failed to attach second tracepoint program\n");
+			err = PTR_ERR(skel->links.event_end);
+			goto out;
+		}
 	}
 
 	/* XXX: we don't actually use this fd - just for poll() */
@@ -118,8 +165,8 @@ int perf_ftrace__latency_stop_bpf(struct perf_ftrace *ftrace __maybe_unused)
 	return 0;
 }
 
-int perf_ftrace__latency_read_bpf(struct perf_ftrace *ftrace __maybe_unused,
-				  int buckets[])
+int perf_ftrace__latency_read_bpf(struct perf_ftrace *ftrace,
+				  int buckets[], struct stats *stats)
 {
 	int i, fd, err;
 	u32 idx;
@@ -132,7 +179,7 @@ int perf_ftrace__latency_read_bpf(struct perf_ftrace *ftrace __maybe_unused,
 	if (hist == NULL)
 		return -ENOMEM;
 
-	for (idx = 0; idx < NUM_BUCKET; idx++) {
+	for (idx = 0; idx < skel->rodata->bucket_num; idx++) {
 		err = bpf_map_lookup_elem(fd, &idx, hist);
 		if (err) {
 			buckets[idx] = 0;
@@ -143,6 +190,19 @@ int perf_ftrace__latency_read_bpf(struct perf_ftrace *ftrace __maybe_unused,
 			buckets[idx] += hist[i];
 	}
 
+	if (skel->bss->count) {
+		stats->mean = skel->bss->total / skel->bss->count;
+		stats->n = skel->bss->count;
+		stats->max = skel->bss->max;
+		stats->min = skel->bss->min;
+
+		if (!ftrace->use_nsec) {
+			stats->mean /= 1000;
+			stats->max /= 1000;
+			stats->min /= 1000;
+		}
+	}
+
 	free(hist);
 	return 0;
 }