I have seen this bug happening a few times in test now, and I do not exactly
know how to reproduce, but I have seen references to it before as well, what
I see is the following:
1) exim connects to remote mx to transport a message
2) When first message is sent, exim apparently searches its db for other mails
with same destination
3) exim tries to open a queued file it finds in its database, but
apparently cant
find on the filesystem, and segfaults.
4) the message that actually got delivered is still in the db, but I
think not on the
filesystem.
My setup is the follwing:
1) use localhost_number
2) use split_spool_directory
3) gnutls
I think the problem here is first and foremost that exim segfaults when it cant
find a message that exists in its db. I cant be the only one who is seeing this?
