bpf, sockmap: Test shutdown() correctly exits epoll and recv()=0

When session gracefully shutdowns epoll needs to wake up and any recv()
readers should return 0 not the -EAGAIN they previously returned.

Note we use epoll instead of select to test the epoll wake on shutdown
event as well.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-12-john.fastabend@gmail.com

diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_basic.c b/tools/testing/selftests/bpf/prog_tests/sockmap_basic.c
index 0ce25a9..615a816 100644 (file)
--- a/tools/testing/selftests/bpf/prog_tests/sockmap_basic.c
+++ b/tools/testing/selftests/bpf/prog_tests/sockmap_basic.c
@@ -2,6 +2,7 @@
 // Copyright (c) 2020 Cloudflare
 #include <error.h>
 #include <netinet/tcp.h>
+#include <sys/epoll.h>
 
 #include "test_progs.h"
 #include "test_skmsg_load_helpers.skel.h"
@@ -9,8 +10,11 @@
 #include "test_sockmap_invalid_update.skel.h"
 #include "test_sockmap_skb_verdict_attach.skel.h"
 #include "test_sockmap_progs_query.skel.h"
+#include "test_sockmap_pass_prog.skel.h"
 #include "bpf_iter_sockmap.skel.h"
 
+#include "sockmap_helpers.h"
+
 #define TCP_REPAIR             19      /* TCP sock is under repair right now */
 
 #define TCP_REPAIR_ON          1
@@ -350,6 +354,62 @@ out:
        test_sockmap_progs_query__destroy(skel);
 }
 
+#define MAX_EVENTS 10
+static void test_sockmap_skb_verdict_shutdown(void)
+{
+       struct epoll_event ev, events[MAX_EVENTS];
+       int n, err, map, verdict, s, c1, p1;
+       struct test_sockmap_pass_prog *skel;
+       int epollfd;
+       int zero = 0;
+       char b;
+
+       skel = test_sockmap_pass_prog__open_and_load();
+       if (!ASSERT_OK_PTR(skel, "open_and_load"))
+               return;
+
+       verdict = bpf_program__fd(skel->progs.prog_skb_verdict);
+       map = bpf_map__fd(skel->maps.sock_map_rx);
+
+       err = bpf_prog_attach(verdict, map, BPF_SK_SKB_STREAM_VERDICT, 0);
+       if (!ASSERT_OK(err, "bpf_prog_attach"))
+               goto out;
+
+       s = socket_loopback(AF_INET, SOCK_STREAM);
+       if (s < 0)
+               goto out;
+       err = create_pair(s, AF_INET, SOCK_STREAM, &c1, &p1);
+       if (err < 0)
+               goto out;
+
+       err = bpf_map_update_elem(map, &zero, &c1, BPF_NOEXIST);
+       if (err < 0)
+               goto out_close;
+
+       shutdown(p1, SHUT_WR);
+
+       ev.events = EPOLLIN;
+       ev.data.fd = c1;
+
+       epollfd = epoll_create1(0);
+       if (!ASSERT_GT(epollfd, -1, "epoll_create(0)"))
+               goto out_close;
+       err = epoll_ctl(epollfd, EPOLL_CTL_ADD, c1, &ev);
+       if (!ASSERT_OK(err, "epoll_ctl(EPOLL_CTL_ADD)"))
+               goto out_close;
+       err = epoll_wait(epollfd, events, MAX_EVENTS, -1);
+       if (!ASSERT_EQ(err, 1, "epoll_wait(fd)"))
+               goto out_close;
+
+       n = recv(c1, &b, 1, SOCK_NONBLOCK);
+       ASSERT_EQ(n, 0, "recv_timeout(fin)");
+out_close:
+       close(c1);
+       close(p1);
+out:
+       test_sockmap_pass_prog__destroy(skel);
+}
+
 void test_sockmap_basic(void)
 {
        if (test__start_subtest("sockmap create_update_free"))
@@ -384,4 +444,6 @@ void test_sockmap_basic(void)
                test_sockmap_progs_query(BPF_SK_SKB_STREAM_VERDICT);
        if (test__start_subtest("sockmap skb_verdict progs query"))
                test_sockmap_progs_query(BPF_SK_SKB_VERDICT);
+       if (test__start_subtest("sockmap skb_verdict shutdown"))
+               test_sockmap_skb_verdict_shutdown();
 }

diff --git a/tools/testing/selftests/bpf/progs/test_sockmap_pass_prog.c b/tools/testing/selftests/bpf/progs/test_sockmap_pass_prog.c
new file mode 100644 (file)
index 0000000..1d86a71
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/test_sockmap_pass_prog.c
@@ -0,0 +1,32 @@
+#include <linux/bpf.h>
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_endian.h>
+
+struct {
+       __uint(type, BPF_MAP_TYPE_SOCKMAP);
+       __uint(max_entries, 20);
+       __type(key, int);
+       __type(value, int);
+} sock_map_rx SEC(".maps");
+
+struct {
+       __uint(type, BPF_MAP_TYPE_SOCKMAP);
+       __uint(max_entries, 20);
+       __type(key, int);
+       __type(value, int);
+} sock_map_tx SEC(".maps");
+
+struct {
+       __uint(type, BPF_MAP_TYPE_SOCKMAP);
+       __uint(max_entries, 20);
+       __type(key, int);
+       __type(value, int);
+} sock_map_msg SEC(".maps");
+
+SEC("sk_skb")
+int prog_skb_verdict(struct __sk_buff *skb)
+{
+       return SK_PASS;
+}
+
+char _license[] SEC("license") = "GPL";