From: Yonghong Song <yhs@fb.com>
To: KP Singh <kpsingh@kernel.org>, <bpf@vger.kernel.org>
Cc: Gilad Reti <gilad.reti@gmail.com>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
Andrii Nakryiko <andrii@kernel.org>,
Martin KaFai Lau <kafai@fb.com>
Subject: Re: [PATCH bpf v2 1/3] bpf: update local storage test to check handling of null ptrs
Date: Mon, 11 Jan 2021 23:24:41 -0800 [thread overview]
Message-ID: <317e2cb6-3774-b343-d93b-5b6f1d41b97e@fb.com> (raw)
In-Reply-To: <20210111212340.86393-2-kpsingh@kernel.org>
On 1/11/21 1:23 PM, KP Singh wrote:
> It was found in [1] that bpf_inode_storage_get helper did not check
> the nullness of the passed owner ptr which caused an oops when
> dereferenced. This change incorporates the example suggested in [1] into
> the local storage selftest.
>
> The test is updated to create a temporary directory instead of just
> using a tempfile. In order to replicate the issue this copied rm binary
> is renamed tiggering the inode_rename with a null pointer for the
> new_inode. The logic to verify the setting and deletion of the inode
> local storage of the old inode is also moved to this LSM hook.
>
> The change also removes the copy_rm function and simply shells out
> to copy files and recursively delete directories and consolidates the
> logic of setting the initial inode storage to the bprm_committed_creds
> hook and removes the file_open hook.
>
> [1]: https://lore.kernel.org/bpf/CANaYP3HWkH91SN=wTNO9FL_2ztHfqcXKX38SSE-JJ2voh+vssw@mail.gmail.com
>
> Suggested-by: Gilad Reti <gilad.reti@gmail.com>
> Signed-off-by: KP Singh <kpsingh@kernel.org>
Ack with one nit below.
Acked-by: Yonghong Song <yhs@fb.com>
> ---
> .../bpf/prog_tests/test_local_storage.c | 96 +++++--------------
> .../selftests/bpf/progs/local_storage.c | 62 ++++++------
> 2 files changed, 61 insertions(+), 97 deletions(-)
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/test_local_storage.c b/tools/testing/selftests/bpf/prog_tests/test_local_storage.c
> index c0fe73a17ed1..338475fe9ffb 100644
> --- a/tools/testing/selftests/bpf/prog_tests/test_local_storage.c
> +++ b/tools/testing/selftests/bpf/prog_tests/test_local_storage.c
> @@ -34,61 +34,6 @@ struct storage {
> struct bpf_spin_lock lock;
> };
>
> -/* Copies an rm binary to a temp file. dest is a mkstemp template */
> -static int copy_rm(char *dest)
> -{
> - int fd_in, fd_out = -1, ret = 0;
> - struct stat stat;
> - char *buf = NULL;
> -
> - fd_in = open("/bin/rm", O_RDONLY);
> - if (fd_in < 0)
> - return -errno;
> -
> - fd_out = mkstemp(dest);
> - if (fd_out < 0) {
> - ret = -errno;
> - goto out;
> - }
> -
> - ret = fstat(fd_in, &stat);
> - if (ret == -1) {
> - ret = -errno;
> - goto out;
> - }
> -
> - buf = malloc(stat.st_blksize);
> - if (!buf) {
> - ret = -errno;
> - goto out;
> - }
> -
> - while (ret = read(fd_in, buf, stat.st_blksize), ret > 0) {
> - ret = write(fd_out, buf, ret);
> - if (ret < 0) {
> - ret = -errno;
> - goto out;
> -
> - }
> - }
> - if (ret < 0) {
> - ret = -errno;
> - goto out;
> -
> - }
> -
> - /* Set executable permission on the copied file */
> - ret = chmod(dest, 0100);
> - if (ret == -1)
> - ret = -errno;
> -
> -out:
> - free(buf);
> - close(fd_in);
> - close(fd_out);
> - return ret;
> -}
> -
> /* Fork and exec the provided rm binary and return the exit code of the
> * forked process and its pid.
> */
> @@ -168,9 +113,11 @@ static bool check_syscall_operations(int map_fd, int obj_fd)
>
> void test_test_local_storage(void)
> {
> - char tmp_exec_path[PATH_MAX] = "/tmp/copy_of_rmXXXXXX";
> + char tmp_dir_path[64] = "/tmp/local_storageXXXXXX";
> int err, serv_sk = -1, task_fd = -1, rm_fd = -1;
> struct local_storage *skel = NULL;
> + char tmp_exec_path[64];
> + char cmd[256];
>
> skel = local_storage__open_and_load();
> if (CHECK(!skel, "skel_load", "lsm skeleton failed\n"))
> @@ -189,18 +136,24 @@ void test_test_local_storage(void)
> task_fd))
> goto close_prog;
>
> - err = copy_rm(tmp_exec_path);
> - if (CHECK(err < 0, "copy_rm", "err %d errno %d\n", err, errno))
> + mkdtemp(tmp_dir_path);
> + if (CHECK(errno < 0, "mkdtemp", "unable to create tmpdir: %d\n", errno))
I think checking mkdtemp return value is more reliable than checking
errno. It is possible mkdtemp returns 0 and errno is not 0 (inheritted
from previous syscall).
> goto close_prog;
>
> + snprintf(tmp_exec_path, sizeof(tmp_exec_path), "%s/copy_of_rm",
> + tmp_dir_path);
> + snprintf(cmd, sizeof(cmd), "cp /bin/rm %s", tmp_exec_path);
> + if (CHECK_FAIL(system(cmd)))
> + goto close_prog_rmdir;
> +
[...]
next prev parent reply other threads:[~2021-01-12 7:25 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-11 21:23 [PATCH bpf v2 0/3] Fix local storage helper OOPs KP Singh
2021-01-11 21:23 ` [PATCH bpf v2 1/3] bpf: update local storage test to check handling of null ptrs KP Singh
2021-01-12 7:24 ` Yonghong Song [this message]
2021-01-12 7:35 ` KP Singh
2021-01-11 21:23 ` [PATCH bpf v2 2/3] bpf: local storage helpers should check nullness of owner ptr passed KP Singh
2021-01-12 7:25 ` Yonghong Song
2021-01-11 21:23 ` [PATCH bpf v2 3/3] bpf: Fix typo in bpf_inode_storage.c KP Singh
2021-01-12 7:26 ` Yonghong Song
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=317e2cb6-3774-b343-d93b-5b6f1d41b97e@fb.com \
--to=yhs@fb.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=gilad.reti@gmail.com \
--cc=kafai@fb.com \
--cc=kpsingh@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).