Thread (2 messages) 2 messages, 2 authors, 2021-01-04

Re: [PATCH v11 bpf-next] bpf/selftests: fold test_current_pid_tgid_new_ns into test_progs.

From: Andrii Nakryiko <hidden>
Date: 2021-01-04 23:52:58
Also in: netdev

On Mon, Dec 28, 2020 at 7:48 AM Carlos Neira [off-list ref] wrote:
Currently tests for bpf_get_ns_current_pid_tgid() are outside test_progs.
This change folds test cases into test_progs.

Changes from v10:

 - Code style fixes.
 - Remove redundant code.

Signed-off-by: Carlos Neira <redacted>
---
 tools/testing/selftests/bpf/.gitignore        |   1 -
 tools/testing/selftests/bpf/Makefile          |   3 +-
 .../bpf/prog_tests/ns_current_pid_tgid.c      | 118 ++++++-------
 .../bpf/progs/test_ns_current_pid_tgid.c      |  28 +--
 .../bpf/test_current_pid_tgid_new_ns.c        | 160 ------------------
 5 files changed, 69 insertions(+), 241 deletions(-)
 delete mode 100644 tools/testing/selftests/bpf/test_current_pid_tgid_new_ns.c
[...]
-void test_ns_current_pid_tgid(void)
+static int test_current_pid_tgid(void *args)
 {
-       const char *probe_name = "raw_tracepoint/sys_enter";
-       const char *file = "test_ns_current_pid_tgid.o";
-       int err, key = 0, duration = 0;
-       struct bpf_link *link = NULL;
-       struct bpf_program *prog;
-       struct bpf_map *bss_map;
-       struct bpf_object *obj;
-       struct bss bss;
+       struct test_ns_current_pid_tgid__bss  *bss;
+       struct test_ns_current_pid_tgid *skel;
+       int err = 0, duration = 0;
+       pid_t tgid, pid;
        struct stat st;
-       __u64 id;
-
-       obj = bpf_object__open_file(file, NULL);
-       if (CHECK(IS_ERR(obj), "obj_open", "err %ld\n", PTR_ERR(obj)))
-               return;

-       err = bpf_object__load(obj);
-       if (CHECK(err, "obj_load", "err %d errno %d\n", err, errno))
+       skel = test_ns_current_pid_tgid__open_and_load();
+       if (CHECK(!skel, "skel_open_load", "failed to load skeleton\n"))
                goto cleanup;
err will be 0 here, and test failure won't be detected (if it happens
in a child process), right? please check all the other cases. It might
be better to assign err = -1 at the beginning and just assign err = 0
after all tests succeeded, right before the clean up.
-       bss_map = bpf_object__find_map_by_name(obj, "test_ns_.bss");
-       if (CHECK(!bss_map, "find_bss_map", "failed\n"))
-               goto cleanup;
+       pid = syscall(SYS_gettid);
+       tgid = getpid();
[...]
-cleanup:
-       bpf_link__destroy(link);
-       bpf_object__close(obj);
+       if (CHECK(waitpid(cpid, &wstatus, 0) == -1, "waitpid", strerror(errno)))
+               exit(EXIT_FAILURE);
+
+       if (CHECK(WEXITSTATUS(wstatus) != 0, "newns_pidtgid", "failed"))
+               exit(EXIT_FAILURE);
this exit will happen in the parent process, which is test_progs test
runner itself, right? Which will prematurely stop all of the remaining
tests to run. I.e., you can't just exit(), you need to just return and
let test_progs keep running.
+}
+
+void test_ns_current_pid_tgid(void)
+{
+       if (test__start_subtest("ns_current_pid_tgid_root_ns"))
+               test_current_pid_tgid(NULL);
+       if (test__start_subtest("ns_current_pid_tgid_new_ns"))
+               test_ns_current_pid_tgid_new_ns();
 }
[...]
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help