bpf: Notify user if we ever hit a bpf_snprintf verifier bug

In check_bpf_snprintf_call(), a map_direct_value_addr() of the fmt map
should never fail because it has already been checked by
ARG_PTR_TO_CONST_STR. But if it ever fails, it's better to error out
with an explicit debug message rather than silently fail.

Reported-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Florent Revest <revest@chromium.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210422235543.4007694-2-revest@chromium.org

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 58730872f7e5e842c4d16cc42b2f5ccf8ae9f754..59799a9b014aa69a69f298a20d4b30b39241e9a2 100644 (file)
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -5940,8 +5940,10 @@ static int check_bpf_snprintf_call(struct bpf_verifier_env *env,
        fmt_map_off = fmt_reg->off + fmt_reg->var_off.value;
        err = fmt_map->ops->map_direct_value_addr(fmt_map, &fmt_addr,
                                                  fmt_map_off);
-       if (err)
-               return err;
+       if (err) {
+               verbose(env, "verifier bug\n");
+               return -EFAULT;
+       }
        fmt = (char *)(long)fmt_addr + fmt_map_off;
 
        /* We are also guaranteed that fmt+fmt_map_off is NULL terminated, we