On Tue, Jan 26, 2021 at 6:21 PM Daniel Borkmann <daniel@xxxxxxxxxxxxx> wrote:
>
> On 1/24/21 8:05 PM, Andrei Matei wrote:
> > This patch adds support to verifier tests to check for a succession of
> > verifier log messages on program load failure. This makes the
> > errstr field work uniformly across REJECT and VERBOSE_ACCEPT checks.
> >
> > This patch also increases the maximum size of an accepted series of
> > messages to test from 80 chars to 200 chars. This is in order to keep
> > existing tests working, which sometimes test for messages larger than 80
> > chars (which was accepted in the REJECT case, when testing for a single
> > message, but ironically not in the VERBOSE_ACCEPT case, when testing for
> > possibly multiple messages).
> > And example of such a long, checked message is in bounds.c:
> > "R1 has unknown scalar with mixed signed bounds, pointer arithmetic with
> > it prohibited for !root"
> >
> > Signed-off-by: Andrei Matei <andreimatei1@xxxxxxxxx>
> > ---
> > tools/testing/selftests/bpf/test_verifier.c | 15 ++++++++++++---
> > 1 file changed, 12 insertions(+), 3 deletions(-)
> >
> > diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c
> > index 59bfa6201d1d..69298bf8ee86 100644
> > --- a/tools/testing/selftests/bpf/test_verifier.c
> > +++ b/tools/testing/selftests/bpf/test_verifier.c
> > @@ -88,6 +88,9 @@ struct bpf_test {
> > int fixup_map_event_output[MAX_FIXUPS];
> > int fixup_map_reuseport_array[MAX_FIXUPS];
> > int fixup_map_ringbuf[MAX_FIXUPS];
> > + /* Expected verifier log output for result REJECT or VERBOSE_ACCEPT. Can be a
> > + * tab-separated sequence of expected strings.
> > + */
> > const char *errstr;
> > const char *errstr_unpriv;
> > uint32_t insn_processed;
> > @@ -995,9 +998,11 @@ static int do_prog_test_run(int fd_prog, bool unpriv, uint32_t expected_val,
> > return 0;
> > }
> >
> > +/* Returns true if every part of exp (tab-separated) appears in log, in order.
> > + */
> > static bool cmp_str_seq(const char *log, const char *exp)
> > {
> > - char needle[80];
> > + char needle[200];
> > const char *p, *q;
> > int len;
> >
> > @@ -1015,7 +1020,7 @@ static bool cmp_str_seq(const char *log, const char *exp)
> > needle[len] = 0;
> > q = strstr(log, needle);
> > if (!q) {
> > - printf("FAIL\nUnexpected verifier log in successful load!\n"
> > + printf("FAIL\nUnexpected verifier log!\n"
> > "EXP: %s\nRES:\n", needle);
> > return false;
> > }
> > @@ -1130,7 +1135,11 @@ static void do_test_single(struct bpf_test *test, bool unpriv,
> > printf("FAIL\nUnexpected success to load!\n");
> > goto fail_log;
> > }
> > - if (!expected_err || !strstr(bpf_vlog, expected_err)) {
> > + if (!expected_err) {
> > + printf("FAIL\nTestcase bug; missing expected_err\n");
> > + goto fail_log;
>
> Do we have an in-tree case like this?
You're asking if there are tests with expected_res == REJECT and
expected_err == NULL?
There are no such test cases, and the intention of this "testcase bug"
check was to keep it that way.
I can simply fold it into the test failure below, as you're suggesting.
> Given this would also be visible below with 'EXP:'
> being (null), I might simplify and just replace the strstr() with cmp_str_seq().
>
> Also, could you elaborate on which test cases need the cmp_str_seq() conversion?
There are VERBOSE_ACCEPT tests that you a tab-separated list of
expected messages; see precise.c.
There are no such REJECT tests yet. I was about to introduce one in
another patch that's inflight, but I ended
up not needing to. Still, I figured that unifying the capabilities of
.errstr between VERBOSE_ACCEPT and REJECT
is a good idea. If you don't think so, I'm happy to drop this patch.
>
> > + }
> > + if ((strlen(expected_err) > 0) && !cmp_str_seq(bpf_vlog, expected_err)) {
>
> (nit: no extra '()' needed, but either way I'd rather opt for '!expected_err ||
> !cmp_str_seq(bpf_vlog, expected_err)' anyway)
>
> > printf("FAIL\nUnexpected error message!\n\tEXP: %s\n\tRES: %s\n",
> > expected_err, bpf_vlog);
> > goto fail_log;
> >
>
