On Fri, Mar 22, 2024 at 08:48:29AM +0100, Mickaël Salaün wrote:
> It might be interesting to create a layout with one file of each type
> and use that for the IOCTL tests.
We have already written these tests and we can keep them, but I think that we
only gain little additional confidence from testing non-device files. The
implementation is saying pretty directly that IOCTLs are permitted if the file
is not a character or block device, at the top of the file_ioctl hook. I don't
see much value in testing this even more exhaustively and would like to keep it
as it is for now.
> On Sat, Mar 09, 2024 at 07:53:17AM +0000, Günther Noack wrote:
> > Named pipes should behave like pipes created with pipe(2),
> > so we don't want to restrict IOCTLs on them.
> >
> > Suggested-by: Mickaël Salaün <mic@xxxxxxxxxxx>
> > Signed-off-by: Günther Noack <gnoack@xxxxxxxxxx>
> > ---
> > tools/testing/selftests/landlock/fs_test.c | 61 ++++++++++++++++++----
> > 1 file changed, 52 insertions(+), 9 deletions(-)
> >
> > diff --git a/tools/testing/selftests/landlock/fs_test.c b/tools/testing/selftests/landlock/fs_test.c
> > index 5c47231a722e..d991f44875bc 100644
> > --- a/tools/testing/selftests/landlock/fs_test.c
> > +++ b/tools/testing/selftests/landlock/fs_test.c
> > @@ -3924,6 +3924,58 @@ TEST_F_FORK(layout1, o_path_ftruncate_and_ioctl)
> > ASSERT_EQ(0, close(fd));
> > }
> >
> > +static int test_fionread_ioctl(int fd)
> > +{
> > + size_t sz = 0;
> > +
> > + if (ioctl(fd, FIONREAD, &sz) < 0 && errno == EACCES)
> > + return errno;
> > + return 0;
> > +}
> > +
> > +/*
> > + * Named pipes are not governed by the LANDLOCK_ACCESS_FS_IOCTL_DEV right,
> > + * because they are not character or block devices.
> > + */
> > +TEST_F_FORK(layout1, named_pipe_ioctl)
> > +{
> > + pid_t child_pid;
> > + int fd, ruleset_fd;
> > + const char *const path = file1_s1d1;
> > + const struct landlock_ruleset_attr attr = {
> > + .handled_access_fs = LANDLOCK_ACCESS_FS_IOCTL_DEV,
> > + };
> > +
> > + ASSERT_EQ(0, unlink(path));
> > + ASSERT_EQ(0, mkfifo(path, 0600));
> > +
> > + /* Enables Landlock. */
> > + ruleset_fd = landlock_create_ruleset(&attr, sizeof(attr), 0);
> > + ASSERT_LE(0, ruleset_fd);
> > + enforce_ruleset(_metadata, ruleset_fd);
> > + ASSERT_EQ(0, close(ruleset_fd));
> > +
> > + /* The child process opens the pipe for writing. */
> > + child_pid = fork();
> > + ASSERT_NE(-1, child_pid);
> > + if (child_pid == 0) {
>
> What is the purpose of this child's code?
>From fifo(7):
Opening the FIFO blocks until the other end is opened also.
So the child and parent process both wait for the other open to happen.
I suspect I could technically also use O_RDWR here, but that is undefined
behaviour in POSIX and less conventional code. (This is described further down,
also in fifo(7).)
>
> > + fd = open(path, O_WRONLY);
> > + close(fd);
> > + exit(0);
> > + }
> > +
> > + fd = open(path, O_RDONLY);
> > + ASSERT_LE(0, fd);
> > +
> > + /* FIONREAD is implemented by pipefifo_fops. */
> > + EXPECT_EQ(0, test_fionread_ioctl(fd));
> > +
> > + ASSERT_EQ(0, close(fd));
> > + ASSERT_EQ(0, unlink(path));
> > +
> > + ASSERT_EQ(child_pid, waitpid(child_pid, NULL, 0));
> > +}
> > +
> > /* clang-format off */
> > FIXTURE(ioctl) {};
> >
> > @@ -3997,15 +4049,6 @@ static int test_tcgets_ioctl(int fd)
> > return 0;
> > }
> >
> > -static int test_fionread_ioctl(int fd)
> > -{
> > - size_t sz = 0;
> > -
> > - if (ioctl(fd, FIONREAD, &sz) < 0 && errno == EACCES)
> > - return errno;
> > - return 0;
> > -}
> > -
>
> You should add test_fionread_ioctl() at the right place from the start.
Fair enough, done.
> > TEST_F_FORK(ioctl, handle_dir_access_file)
> > {
> > const int flag = 0;
> > --
> > 2.44.0.278.ge034bb2e1d-goog
> >
> >
—Günther
