[exim-dev] [Bug 1671] segfault after delivery

Top Page
Delete this message
Reply to this message
Author: admin
Date:  
To: exim-dev
Subject: [exim-dev] [Bug 1671] segfault after delivery
https://bugs.exim.org/show_bug.cgi?id=1671

Phil Pennock <pdp@???> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pdp@???


--- Comment #3 from Phil Pennock <pdp@???> ---
This appears to be new logic introduced with commit `a39bd74d` 2015-03-15:
http://git.exim.org/exim.git/commit/a39bd74d3e942db1b465cae643e1a4766ffb304e

This was introduced to fix bug 1141, "remote_smtp reuse when should not be".

I'm not familiar with this new logic, but at a glance, it looks like the usage
of `deliver_get_sender_address()` is assuming that the previous message from
the same connection is still on disk, but when it's delivered before the next
message from that source, the `-H` file is missing,
`deliver_get_sender_address()` returns `NULL` and things go wrong.

But that doesn't explain why the backtrace shows `tblock=0x0` here, because it
should have done the same on lines 1278/1279 as it does on 1282 and it should
have crashed there.

--
You are receiving this mail because:
You are on the CC list for the bug.