selftests/bpf: Add test relying only on CO-RE and no recent kernel fe…

…atures

Add a test that relies on CO-RE, but doesn't expect any of the recent
features, not available on old kernels. This is useful for Travis CI tests
running against very old kernels (e.g., libbpf has 4.9 kernel testing now), to
verify that CO-RE still works, even if kernel itself doesn't support BTF yet,
as long as there is .BTF embedded into vmlinux image by pahole. Given most of
CO-RE doesn't require any kernel awareness of BTF, it is a useful test to
validate that libbpf's BTF sanitization is working well even with ancient
kernels.

Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20200708015318.3827358-5-andriin@fb.com

tools/testing/selftests/bpf/prog_tests/core_retro.c

@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (c) 2020 Facebook
+#define _GNU_SOURCE
+#include <test_progs.h>
+#include "test_core_retro.skel.h"
+
+void test_core_retro(void)
+{
+	int err, zero = 0, res, duration = 0;
+	struct test_core_retro *skel;
+
+	/* load program */
+	skel = test_core_retro__open_and_load();
+	if (CHECK(!skel, "skel_load", "skeleton open/load failed\n"))
+		goto out_close;
+
+	/* attach probe */
+	err = test_core_retro__attach(skel);
+	if (CHECK(err, "attach_kprobe", "err %d\n", err))
+		goto out_close;
+
+	/* trigger */
+	usleep(1);
+
+	err = bpf_map_lookup_elem(bpf_map__fd(skel->maps.results), &zero, &res);
+	if (CHECK(err, "map_lookup", "failed to lookup result: %d\n", errno))
+		goto out_close;
+
+	CHECK(res != getpid(), "pid_check", "got %d != exp %d\n", res, getpid());
+
+out_close:
+	test_core_retro__destroy(skel);
+}

tools/testing/selftests/bpf/progs/test_core_retro.c

@@ -0,0 +1,30 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (c) 2020 Facebook
+#include <linux/bpf.h>
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_core_read.h>
+
+struct task_struct {
+	int tgid;
+} __attribute__((preserve_access_index));
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__uint(max_entries, 1);
+	__type(key, int);
+	__type(value, int);
+} results SEC(".maps");
+
+SEC("tp/raw_syscalls/sys_enter")
+int handle_sys_enter(void *ctx)
+{
+	struct task_struct *task = (void *)bpf_get_current_task();
+	int tgid = BPF_CORE_READ(task, tgid);
+	int zero = 0;
+
+	bpf_map_update_elem(&results, &zero, &tgid, 0);
+
+	return 0;
+}
+
+char _license[] SEC("license") = "GPL";