selftests/bpf: Add tests for bpf_rcu_read_lock()

Add a few positive/negative tests to test bpf_rcu_read_lock()
and its corresponding verifier support. The new test will fail
on s390x and aarch64, so an entry is added to each of their
respective deny lists.

Acked-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/r/20221124053222.2374650-1-yhs@fb.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

tools/testing/selftests/bpf/DENYLIST.aarch64

@@ -45,6 +45,7 @@ modify_return                                    # modify_return__attach failed
 module_attach                                    # skel_attach skeleton attach failed: -524
 mptcp/base                                       # run_test mptcp unexpected error: -524 (errno 524)
 netcnt                                           # packets unexpected packets: actual 10001 != expected 10000
+rcu_read_lock                                    # failed to attach: ERROR: strerror_r(-524)=22
 recursion                                        # skel_attach unexpected error: -524 (errno 524)
 ringbuf                                          # skel_attach skeleton attachment failed: -1
 setget_sockopt                                   # attach_cgroup unexpected error: -524

tools/testing/selftests/bpf/DENYLIST.s390x

@@ -43,6 +43,7 @@ module_attach                            # skel_attach skeleton attach failed: -
 mptcp
 netcnt                                   # failed to load BPF skeleton 'netcnt_prog': -7                               (?)
 probe_user                               # check_kprobe_res wrong kprobe res from probe read                           (?)
+rcu_read_lock                            # failed to find kernel BTF type ID of '__x64_sys_getpgid': -3                (?)
 recursion                                # skel_attach unexpected error: -524                                          (trampoline)
 ringbuf                                  # skel_load skeleton load failed                                              (?)
 select_reuseport                         # intermittently fails on new s390x setup

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

@@ -0,0 +1,158 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022 Meta Platforms, Inc. and affiliates.*/
+
+#define _GNU_SOURCE
+#include <unistd.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <test_progs.h>
+#include <bpf/btf.h>
+#include "rcu_read_lock.skel.h"
+#include "cgroup_helpers.h"
+
+static unsigned long long cgroup_id;
+
+static void test_success(void)
+{
+	struct rcu_read_lock *skel;
+	int err;
+
+	skel = rcu_read_lock__open();
+	if (!ASSERT_OK_PTR(skel, "skel_open"))
+		return;
+
+	skel->bss->target_pid = syscall(SYS_gettid);
+
+	bpf_program__set_autoload(skel->progs.get_cgroup_id, true);
+	bpf_program__set_autoload(skel->progs.task_succ, true);
+	bpf_program__set_autoload(skel->progs.no_lock, true);
+	bpf_program__set_autoload(skel->progs.two_regions, true);
+	bpf_program__set_autoload(skel->progs.non_sleepable_1, true);
+	bpf_program__set_autoload(skel->progs.non_sleepable_2, true);
+	err = rcu_read_lock__load(skel);
+	if (!ASSERT_OK(err, "skel_load"))
+		goto out;
+
+	err = rcu_read_lock__attach(skel);
+	if (!ASSERT_OK(err, "skel_attach"))
+		goto out;
+
+	syscall(SYS_getpgid);
+
+	ASSERT_EQ(skel->bss->task_storage_val, 2, "task_storage_val");
+	ASSERT_EQ(skel->bss->cgroup_id, cgroup_id, "cgroup_id");
+out:
+	rcu_read_lock__destroy(skel);
+}
+
+static void test_rcuptr_acquire(void)
+{
+	struct rcu_read_lock *skel;
+	int err;
+
+	skel = rcu_read_lock__open();
+	if (!ASSERT_OK_PTR(skel, "skel_open"))
+		return;
+
+	skel->bss->target_pid = syscall(SYS_gettid);
+
+	bpf_program__set_autoload(skel->progs.task_acquire, true);
+	err = rcu_read_lock__load(skel);
+	if (!ASSERT_OK(err, "skel_load"))
+		goto out;
+
+	err = rcu_read_lock__attach(skel);
+	ASSERT_OK(err, "skel_attach");
+out:
+	rcu_read_lock__destroy(skel);
+}
+
+static const char * const inproper_region_tests[] = {
+	"miss_lock",
+	"miss_unlock",
+	"non_sleepable_rcu_mismatch",
+	"inproper_sleepable_helper",
+	"inproper_sleepable_kfunc",
+	"nested_rcu_region",
+};
+
+static void test_inproper_region(void)
+{
+	struct rcu_read_lock *skel;
+	struct bpf_program *prog;
+	int i, err;
+
+	for (i = 0; i < ARRAY_SIZE(inproper_region_tests); i++) {
+		skel = rcu_read_lock__open();
+		if (!ASSERT_OK_PTR(skel, "skel_open"))
+			return;
+
+		prog = bpf_object__find_program_by_name(skel->obj, inproper_region_tests[i]);
+		if (!ASSERT_OK_PTR(prog, "bpf_object__find_program_by_name"))
+			goto out;
+		bpf_program__set_autoload(prog, true);
+		err = rcu_read_lock__load(skel);
+		ASSERT_ERR(err, "skel_load");
+out:
+		rcu_read_lock__destroy(skel);
+	}
+}
+
+static const char * const rcuptr_misuse_tests[] = {
+	"task_untrusted_non_rcuptr",
+	"task_untrusted_rcuptr",
+	"cross_rcu_region",
+};
+
+static void test_rcuptr_misuse(void)
+{
+	struct rcu_read_lock *skel;
+	struct bpf_program *prog;
+	int i, err;
+
+	for (i = 0; i < ARRAY_SIZE(rcuptr_misuse_tests); i++) {
+		skel = rcu_read_lock__open();
+		if (!ASSERT_OK_PTR(skel, "skel_open"))
+			return;
+
+		prog = bpf_object__find_program_by_name(skel->obj, rcuptr_misuse_tests[i]);
+		if (!ASSERT_OK_PTR(prog, "bpf_object__find_program_by_name"))
+			goto out;
+		bpf_program__set_autoload(prog, true);
+		err = rcu_read_lock__load(skel);
+		ASSERT_ERR(err, "skel_load");
+out:
+		rcu_read_lock__destroy(skel);
+	}
+}
+
+void test_rcu_read_lock(void)
+{
+	struct btf *vmlinux_btf;
+	int cgroup_fd;
+
+	vmlinux_btf = btf__load_vmlinux_btf();
+	if (!ASSERT_OK_PTR(vmlinux_btf, "could not load vmlinux BTF"))
+		return;
+	if (btf__find_by_name_kind(vmlinux_btf, "rcu", BTF_KIND_TYPE_TAG) < 0) {
+		test__skip();
+		goto out;
+	}
+
+	cgroup_fd = test__join_cgroup("/rcu_read_lock");
+	if (!ASSERT_GE(cgroup_fd, 0, "join_cgroup /rcu_read_lock"))
+		goto out;
+
+	cgroup_id = get_cgroup_id("/rcu_read_lock");
+	if (test__start_subtest("success"))
+		test_success();
+	if (test__start_subtest("rcuptr_acquire"))
+		test_rcuptr_acquire();
+	if (test__start_subtest("negative_tests_inproper_region"))
+		test_inproper_region();
+	if (test__start_subtest("negative_tests_rcuptr_misuse"))
+		test_rcuptr_misuse();
+	close(cgroup_fd);
+out:
+	btf__free(vmlinux_btf);
+}

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

@@ -0,0 +1,290 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022 Meta Platforms, Inc. and affiliates. */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+#include "bpf_tracing_net.h"
+#include "bpf_misc.h"
+
+char _license[] SEC("license") = "GPL";
+
+struct {
+	__uint(type, BPF_MAP_TYPE_TASK_STORAGE);
+	__uint(map_flags, BPF_F_NO_PREALLOC);
+	__type(key, int);
+	__type(value, long);
+} map_a SEC(".maps");
+
+__u32 user_data, key_serial, target_pid;
+__u64 flags, task_storage_val, cgroup_id;
+
+struct bpf_key *bpf_lookup_user_key(__u32 serial, __u64 flags) __ksym;
+void bpf_key_put(struct bpf_key *key) __ksym;
+void bpf_rcu_read_lock(void) __ksym;
+void bpf_rcu_read_unlock(void) __ksym;
+struct task_struct *bpf_task_acquire(struct task_struct *p) __ksym;
+void bpf_task_release(struct task_struct *p) __ksym;
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int get_cgroup_id(void *ctx)
+{
+	struct task_struct *task;
+
+	task = bpf_get_current_task_btf();
+	if (task->pid != target_pid)
+		return 0;
+
+	/* simulate bpf_get_current_cgroup_id() helper */
+	bpf_rcu_read_lock();
+	cgroup_id = task->cgroups->dfl_cgrp->kn->id;
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int task_succ(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+	long init_val = 2;
+	long *ptr;
+
+	task = bpf_get_current_task_btf();
+	if (task->pid != target_pid)
+		return 0;
+
+	bpf_rcu_read_lock();
+	/* region including helper using rcu ptr real_parent */
+	real_parent = task->real_parent;
+	ptr = bpf_task_storage_get(&map_a, real_parent, &init_val,
+				   BPF_LOCAL_STORAGE_GET_F_CREATE);
+	if (!ptr)
+		goto out;
+	ptr = bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	if (!ptr)
+		goto out;
+	task_storage_val = *ptr;
+out:
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_nanosleep")
+int no_lock(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	/* no bpf_rcu_read_lock(), old code still works */
+	task = bpf_get_current_task_btf();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_nanosleep")
+int two_regions(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	/* two regions */
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	bpf_rcu_read_unlock();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry/" SYS_PREFIX "sys_getpgid")
+int non_sleepable_1(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry/" SYS_PREFIX "sys_getpgid")
+int non_sleepable_2(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	bpf_rcu_read_lock();
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_unlock();
+
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_nanosleep")
+int task_acquire(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	/* acquire a reference which can be used outside rcu read lock region */
+	real_parent = bpf_task_acquire(real_parent);
+	bpf_rcu_read_unlock();
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_task_release(real_parent);
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int miss_lock(void *ctx)
+{
+	struct task_struct *task;
+	struct css_set *cgroups;
+	struct cgroup *dfl_cgrp;
+
+	/* missing bpf_rcu_read_lock() */
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	(void)bpf_task_storage_get(&map_a, task, 0, 0);
+	bpf_rcu_read_unlock();
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int miss_unlock(void *ctx)
+{
+	struct task_struct *task;
+	struct css_set *cgroups;
+	struct cgroup *dfl_cgrp;
+
+	/* missing bpf_rcu_read_unlock() */
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	(void)bpf_task_storage_get(&map_a, task, 0, 0);
+	return 0;
+}
+
+SEC("?fentry/" SYS_PREFIX "sys_getpgid")
+int non_sleepable_rcu_mismatch(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	task = bpf_get_current_task_btf();
+	/* non-sleepable: missing bpf_rcu_read_unlock() in one path */
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	if (real_parent)
+		bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int inproper_sleepable_helper(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+	struct pt_regs *regs;
+	__u32 value = 0;
+	void *ptr;
+
+	task = bpf_get_current_task_btf();
+	/* sleepable helper in rcu read lock region */
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	regs = (struct pt_regs *)bpf_task_pt_regs(real_parent);
+	if (!regs) {
+		bpf_rcu_read_unlock();
+		return 0;
+	}
+
+	ptr = (void *)PT_REGS_IP(regs);
+	(void)bpf_copy_from_user_task(&value, sizeof(uint32_t), ptr, task, 0);
+	user_data = value;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?lsm.s/bpf")
+int BPF_PROG(inproper_sleepable_kfunc, int cmd, union bpf_attr *attr, unsigned int size)
+{
+	struct bpf_key *bkey;
+
+	/* sleepable kfunc in rcu read lock region */
+	bpf_rcu_read_lock();
+	bkey = bpf_lookup_user_key(key_serial, flags);
+	bpf_rcu_read_unlock();
+	if (!bkey)
+		return -1;
+	bpf_key_put(bkey);
+
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_nanosleep")
+int nested_rcu_region(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	/* nested rcu read lock regions */
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int task_untrusted_non_rcuptr(void *ctx)
+{
+	struct task_struct *task, *last_wakee;
+
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	/* the pointer last_wakee marked as untrusted */
+	last_wakee = task->real_parent->last_wakee;
+	(void)bpf_task_storage_get(&map_a, last_wakee, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+int task_untrusted_rcuptr(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	bpf_rcu_read_unlock();
+	/* helper use of rcu ptr outside the rcu read lock region */
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_nanosleep")
+int cross_rcu_region(void *ctx)
+{
+	struct task_struct *task, *real_parent;
+
+	/* rcu ptr define/use in different regions */
+	task = bpf_get_current_task_btf();
+	bpf_rcu_read_lock();
+	real_parent = task->real_parent;
+	bpf_rcu_read_unlock();
+	bpf_rcu_read_lock();
+	(void)bpf_task_storage_get(&map_a, real_parent, 0, 0);
+	bpf_rcu_read_unlock();
+	return 0;
+}