Fatal if field[0].start is null #2612
Conversation
7eb3a0a to
e485de7
Compare
|
This looks reasonable, but shouldn't we make the same change in |
|
Here is actually more to do - next_field_w() should get the same length protection as next_field(), otherwise you may end in out of bounds reads or an infinite loop. So archive_acl_from_text_w shoud wrap in a new archive_acl_from_text_wl(). Does an malformed ACL entry mean we completely bail out or do we just skip it? |
We should not get here, but given that the check exists, we should not let it happen if this is NULL because otherwise we just dereference it later on.
I don't know what to do here in this case. I just think we should protect from this. Was wondering if I had to do anything else but yeah this seems fine for now. Wondering what else we should do in this area. |
We should not get here, but given that the check exists, we should not let it happen if this is NULL because otherwise we just dereference it later on.