1 files changed, 235 insertions, 198 deletions

diff --git a/tools/testing/selftests/bpf/benchs/bench_trigger.c b/tools/testing/selftests/bpf/benchs/bench_trigger.c
index ace0d1011a..4b05539f16 100644
--- a/tools/testing/selftests/bpf/benchs/bench_trigger.c
+++ b/tools/testing/selftests/bpf/benchs/bench_trigger.c
@@ -1,15 +1,95 @@
 // SPDX-License-Identifier: GPL-2.0
 /* Copyright (c) 2020 Facebook */
+#define _GNU_SOURCE
+#include <argp.h>
+#include <unistd.h>
+#include <stdint.h>
 #include "bench.h"
 #include "trigger_bench.skel.h"
 #include "trace_helpers.h"
 
+#define MAX_TRIG_BATCH_ITERS 1000
+
+static struct {
+	__u32 batch_iters;
+} args = {
+	.batch_iters = 100,
+};
+
+enum {
+	ARG_TRIG_BATCH_ITERS = 7000,
+};
+
+static const struct argp_option opts[] = {
+	{ "trig-batch-iters", ARG_TRIG_BATCH_ITERS, "BATCH_ITER_CNT", 0,
+		"Number of in-kernel iterations per one driver test run"},
+	{},
+};
+
+static error_t parse_arg(int key, char *arg, struct argp_state *state)
+{
+	long ret;
+
+	switch (key) {
+	case ARG_TRIG_BATCH_ITERS:
+		ret = strtol(arg, NULL, 10);
+		if (ret < 1 || ret > MAX_TRIG_BATCH_ITERS) {
+			fprintf(stderr, "invalid --trig-batch-iters value (should be between %d and %d)\n",
+				1, MAX_TRIG_BATCH_ITERS);
+			argp_usage(state);
+		}
+		args.batch_iters = ret;
+		break;
+	default:
+		return ARGP_ERR_UNKNOWN;
+	}
+
+	return 0;
+}
+
+const struct argp bench_trigger_batch_argp = {
+	.options = opts,
+	.parser = parse_arg,
+};
+
+/* adjust slot shift in inc_hits() if changing */
+#define MAX_BUCKETS 256
+
+#pragma GCC diagnostic ignored "-Wattributes"
+
 /* BPF triggering benchmarks */
 static struct trigger_ctx {
 	struct trigger_bench *skel;
+	bool usermode_counters;
+	int driver_prog_fd;
 } ctx;
 
-static struct counter base_hits;
+static struct counter base_hits[MAX_BUCKETS];
+
+static __always_inline void inc_counter(struct counter *counters)
+{
+	static __thread int tid = 0;
+	unsigned slot;
+
+	if (unlikely(tid == 0))
+		tid = syscall(SYS_gettid);
+
+	/* multiplicative hashing, it's fast */
+	slot = 2654435769U * tid;
+	slot >>= 24;
+
+	atomic_inc(&base_hits[slot].value); /* use highest byte as an index */
+}
+
+static long sum_and_reset_counters(struct counter *counters)
+{
+	int i;
+	long sum = 0;
+
+	for (i = 0; i < MAX_BUCKETS; i++)
+		sum += atomic_swap(&counters[i].value, 0);
+	return sum;
+}
 
 static void trigger_validate(void)
 {
@@ -19,41 +99,63 @@ static void trigger_validate(void)
 	}
 }
 
-static void *trigger_base_producer(void *input)
+static void *trigger_producer(void *input)
 {
-	while (true) {
-		(void)syscall(__NR_getpgid);
-		atomic_inc(&base_hits.value);
+	if (ctx.usermode_counters) {
+		while (true) {
+			(void)syscall(__NR_getpgid);
+			inc_counter(base_hits);
+		}
+	} else {
+		while (true)
+			(void)syscall(__NR_getpgid);
 	}
 	return NULL;
 }
 
-static void trigger_base_measure(struct bench_res *res)
+static void *trigger_producer_batch(void *input)
 {
-	res->hits = atomic_swap(&base_hits.value, 0);
-}
+	int fd = ctx.driver_prog_fd ?: bpf_program__fd(ctx.skel->progs.trigger_driver);
 
-static void *trigger_producer(void *input)
-{
 	while (true)
-		(void)syscall(__NR_getpgid);
+		bpf_prog_test_run_opts(fd, NULL);
+
 	return NULL;
 }
 
 static void trigger_measure(struct bench_res *res)
 {
-	res->hits = atomic_swap(&ctx.skel->bss->hits, 0);
+	if (ctx.usermode_counters)
+		res->hits = sum_and_reset_counters(base_hits);
+	else
+		res->hits = sum_and_reset_counters(ctx.skel->bss->hits);
 }
 
 static void setup_ctx(void)
 {
 	setup_libbpf();
 
-	ctx.skel = trigger_bench__open_and_load();
+	ctx.skel = trigger_bench__open();
 	if (!ctx.skel) {
 		fprintf(stderr, "failed to open skeleton\n");
 		exit(1);
 	}
+
+	/* default "driver" BPF program */
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver, true);
+
+	ctx.skel->rodata->batch_iters = args.batch_iters;
+}
+
+static void load_ctx(void)
+{
+	int err;
+
+	err = trigger_bench__load(ctx.skel);
+	if (err) {
+		fprintf(stderr, "failed to open skeleton\n");
+		exit(1);
+	}
 }
 
 static void attach_bpf(struct bpf_program *prog)
@@ -67,64 +169,104 @@ static void attach_bpf(struct bpf_program *prog)
 	}
 }
 
-static void trigger_tp_setup(void)
+static void trigger_syscall_count_setup(void)
 {
-	setup_ctx();
-	attach_bpf(ctx.skel->progs.bench_trigger_tp);
+	ctx.usermode_counters = true;
 }
 
-static void trigger_rawtp_setup(void)
+/* Batched, staying mostly in-kernel triggering setups */
+static void trigger_kernel_count_setup(void)
 {
 	setup_ctx();
-	attach_bpf(ctx.skel->progs.bench_trigger_raw_tp);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver, false);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_count, true);
+	load_ctx();
+	/* override driver program */
+	ctx.driver_prog_fd = bpf_program__fd(ctx.skel->progs.trigger_count);
 }
 
 static void trigger_kprobe_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_kprobe, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_kprobe);
 }
 
 static void trigger_kretprobe_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_kretprobe, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_kretprobe);
 }
 
 static void trigger_kprobe_multi_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_kprobe_multi, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_kprobe_multi);
 }
 
 static void trigger_kretprobe_multi_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_kretprobe_multi, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_kretprobe_multi);
 }
 
 static void trigger_fentry_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_fentry, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_fentry);
 }
 
 static void trigger_fexit_setup(void)
 {
 	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_fexit, true);
+	load_ctx();
 	attach_bpf(ctx.skel->progs.bench_trigger_fexit);
 }
 
-static void trigger_fentry_sleep_setup(void)
+static void trigger_fmodret_setup(void)
 {
 	setup_ctx();
-	attach_bpf(ctx.skel->progs.bench_trigger_fentry_sleep);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver, false);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver_kfunc, true);
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_fmodret, true);
+	load_ctx();
+	/* override driver program */
+	ctx.driver_prog_fd = bpf_program__fd(ctx.skel->progs.trigger_driver_kfunc);
+	attach_bpf(ctx.skel->progs.bench_trigger_fmodret);
 }
 
-static void trigger_fmodret_setup(void)
+static void trigger_tp_setup(void)
 {
 	setup_ctx();
-	attach_bpf(ctx.skel->progs.bench_trigger_fmodret);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver, false);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver_kfunc, true);
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_tp, true);
+	load_ctx();
+	/* override driver program */
+	ctx.driver_prog_fd = bpf_program__fd(ctx.skel->progs.trigger_driver_kfunc);
+	attach_bpf(ctx.skel->progs.bench_trigger_tp);
+}
+
+static void trigger_rawtp_setup(void)
+{
+	setup_ctx();
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver, false);
+	bpf_program__set_autoload(ctx.skel->progs.trigger_driver_kfunc, true);
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_rawtp, true);
+	load_ctx();
+	/* override driver program */
+	ctx.driver_prog_fd = bpf_program__fd(ctx.skel->progs.trigger_driver_kfunc);
+	attach_bpf(ctx.skel->progs.bench_trigger_rawtp);
 }
 
 /* make sure call is not inlined and not avoided by compiler, so __weak and
@@ -137,7 +279,7 @@ static void trigger_fmodret_setup(void)
  * GCC doesn't generate stack setup preample for these functions due to them
  * having no input arguments and doing nothing in the body.
  */
-__weak void uprobe_target_nop(void)
+__nocf_check __weak void uprobe_target_nop(void)
 {
 	asm volatile ("nop");
 }
@@ -146,7 +288,7 @@ __weak void opaque_noop_func(void)
 {
 }
 
-__weak int uprobe_target_push(void)
+__nocf_check __weak int uprobe_target_push(void)
 {
 	/* overhead of function call is negligible compared to uprobe
 	 * triggering, so this shouldn't affect benchmark results much
@@ -155,16 +297,16 @@ __weak int uprobe_target_push(void)
 	return 1;
 }
 
-__weak void uprobe_target_ret(void)
+__nocf_check __weak void uprobe_target_ret(void)
 {
 	asm volatile ("");
 }
 
-static void *uprobe_base_producer(void *input)
+static void *uprobe_producer_count(void *input)
 {
 	while (true) {
 		uprobe_target_nop();
-		atomic_inc(&base_hits.value);
+		inc_counter(base_hits);
 	}
 	return NULL;
 }
@@ -194,15 +336,24 @@ static void usetup(bool use_retprobe, void *target_addr)
 {
 	size_t uprobe_offset;
 	struct bpf_link *link;
+	int err;
 
 	setup_libbpf();
 
-	ctx.skel = trigger_bench__open_and_load();
+	ctx.skel = trigger_bench__open();
 	if (!ctx.skel) {
 		fprintf(stderr, "failed to open skeleton\n");
 		exit(1);
 	}
 
+	bpf_program__set_autoload(ctx.skel->progs.bench_trigger_uprobe, true);
+
+	err = trigger_bench__load(ctx.skel);
+	if (err) {
+		fprintf(stderr, "failed to load skeleton\n");
+		exit(1);
+	}
+
 	uprobe_offset = get_uprobe_offset(target_addr);
 	link = bpf_program__attach_uprobe(ctx.skel->progs.bench_trigger_uprobe,
 					  use_retprobe,
@@ -216,204 +367,90 @@ static void usetup(bool use_retprobe, void *target_addr)
 	ctx.skel->links.bench_trigger_uprobe = link;
 }
 
-static void uprobe_setup_nop(void)
+static void usermode_count_setup(void)
+{
+	ctx.usermode_counters = true;
+}
+
+static void uprobe_nop_setup(void)
 {
 	usetup(false, &uprobe_target_nop);
 }
 
-static void uretprobe_setup_nop(void)
+static void uretprobe_nop_setup(void)
 {
 	usetup(true, &uprobe_target_nop);
 }
 
-static void uprobe_setup_push(void)
+static void uprobe_push_setup(void)
 {
 	usetup(false, &uprobe_target_push);
 }
 
-static void uretprobe_setup_push(void)
+static void uretprobe_push_setup(void)
 {
 	usetup(true, &uprobe_target_push);
 }
 
-static void uprobe_setup_ret(void)
+static void uprobe_ret_setup(void)
 {
 	usetup(false, &uprobe_target_ret);
 }
 
-static void uretprobe_setup_ret(void)
+static void uretprobe_ret_setup(void)
 {
 	usetup(true, &uprobe_target_ret);
 }
 
-const struct bench bench_trig_base = {
-	.name = "trig-base",
+const struct bench bench_trig_syscall_count = {
+	.name = "trig-syscall-count",
 	.validate = trigger_validate,
-	.producer_thread = trigger_base_producer,
-	.measure = trigger_base_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_tp = {
-	.name = "trig-tp",
-	.validate = trigger_validate,
-	.setup = trigger_tp_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_rawtp = {
-	.name = "trig-rawtp",
-	.validate = trigger_validate,
-	.setup = trigger_rawtp_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_kprobe = {
-	.name = "trig-kprobe",
-	.validate = trigger_validate,
-	.setup = trigger_kprobe_setup,
+	.setup = trigger_syscall_count_setup,
 	.producer_thread = trigger_producer,
 	.measure = trigger_measure,
 	.report_progress = hits_drops_report_progress,
 	.report_final = hits_drops_report_final,
 };
 
-const struct bench bench_trig_kretprobe = {
-	.name = "trig-kretprobe",
-	.validate = trigger_validate,
-	.setup = trigger_kretprobe_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_kprobe_multi = {
-	.name = "trig-kprobe-multi",
-	.validate = trigger_validate,
-	.setup = trigger_kprobe_multi_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_kretprobe_multi = {
-	.name = "trig-kretprobe-multi",
-	.validate = trigger_validate,
-	.setup = trigger_kretprobe_multi_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_fentry = {
-	.name = "trig-fentry",
-	.validate = trigger_validate,
-	.setup = trigger_fentry_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_fexit = {
-	.name = "trig-fexit",
-	.validate = trigger_validate,
-	.setup = trigger_fexit_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_fentry_sleep = {
-	.name = "trig-fentry-sleep",
-	.validate = trigger_validate,
-	.setup = trigger_fentry_sleep_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_fmodret = {
-	.name = "trig-fmodret",
-	.validate = trigger_validate,
-	.setup = trigger_fmodret_setup,
-	.producer_thread = trigger_producer,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uprobe_base = {
-	.name = "trig-uprobe-base",
-	.setup = NULL, /* no uprobe/uretprobe is attached */
-	.producer_thread = uprobe_base_producer,
-	.measure = trigger_base_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uprobe_nop = {
-	.name = "trig-uprobe-nop",
-	.setup = uprobe_setup_nop,
-	.producer_thread = uprobe_producer_nop,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uretprobe_nop = {
-	.name = "trig-uretprobe-nop",
-	.setup = uretprobe_setup_nop,
-	.producer_thread = uprobe_producer_nop,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uprobe_push = {
-	.name = "trig-uprobe-push",
-	.setup = uprobe_setup_push,
-	.producer_thread = uprobe_producer_push,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uretprobe_push = {
-	.name = "trig-uretprobe-push",
-	.setup = uretprobe_setup_push,
-	.producer_thread = uprobe_producer_push,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uprobe_ret = {
-	.name = "trig-uprobe-ret",
-	.setup = uprobe_setup_ret,
-	.producer_thread = uprobe_producer_ret,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
-
-const struct bench bench_trig_uretprobe_ret = {
-	.name = "trig-uretprobe-ret",
-	.setup = uretprobe_setup_ret,
-	.producer_thread = uprobe_producer_ret,
-	.measure = trigger_measure,
-	.report_progress = hits_drops_report_progress,
-	.report_final = hits_drops_report_final,
-};
+/* batched (staying mostly in kernel) kprobe/fentry benchmarks */
+#define BENCH_TRIG_KERNEL(KIND, NAME)					\
+const struct bench bench_trig_##KIND = {				\
+	.name = "trig-" NAME,						\
+	.setup = trigger_##KIND##_setup,				\
+	.producer_thread = trigger_producer_batch,			\
+	.measure = trigger_measure,					\
+	.report_progress = hits_drops_report_progress,			\
+	.report_final = hits_drops_report_final,			\
+	.argp = &bench_trigger_batch_argp,				\
+}
+
+BENCH_TRIG_KERNEL(kernel_count, "kernel-count");
+BENCH_TRIG_KERNEL(kprobe, "kprobe");
+BENCH_TRIG_KERNEL(kretprobe, "kretprobe");
+BENCH_TRIG_KERNEL(kprobe_multi, "kprobe-multi");
+BENCH_TRIG_KERNEL(kretprobe_multi, "kretprobe-multi");
+BENCH_TRIG_KERNEL(fentry, "fentry");
+BENCH_TRIG_KERNEL(fexit, "fexit");
+BENCH_TRIG_KERNEL(fmodret, "fmodret");
+BENCH_TRIG_KERNEL(tp, "tp");
+BENCH_TRIG_KERNEL(rawtp, "rawtp");
+
+/* uprobe benchmarks */
+#define BENCH_TRIG_USERMODE(KIND, PRODUCER, NAME)			\
+const struct bench bench_trig_##KIND = {				\
+	.name = "trig-" NAME,						\
+	.validate = trigger_validate,					\
+	.setup = KIND##_setup,						\
+	.producer_thread = uprobe_producer_##PRODUCER,			\
+	.measure = trigger_measure,					\
+	.report_progress = hits_drops_report_progress,			\
+	.report_final = hits_drops_report_final,			\
+}
+
+BENCH_TRIG_USERMODE(usermode_count, count, "usermode-count");
+BENCH_TRIG_USERMODE(uprobe_nop, nop, "uprobe-nop");
+BENCH_TRIG_USERMODE(uprobe_push, push, "uprobe-push");
+BENCH_TRIG_USERMODE(uprobe_ret, ret, "uprobe-ret");
+BENCH_TRIG_USERMODE(uretprobe_nop, nop, "uretprobe-nop");
+BENCH_TRIG_USERMODE(uretprobe_push, push, "uretprobe-push");
+BENCH_TRIG_USERMODE(uretprobe_ret, ret, "uretprobe-ret");