Re: [exim] Failed to read delivery status

On 2013-02-09 at 09:29 -0500, Pavel Sedlacek wrote:
> 2013-02-09 07:33:08 1U49cB-0000A6-6a failed to read delivery status for xxxx@??? from delivery subprocess
> 2013-02-09 07:33:09 1U49cB-0000A6-6a appendfile transport process returned non-zero status 0x000b: terminated by signal 11

That's bad, that's a segfault, and I don't think there's any external
library code being called here.

So either you're exposing a previously unknown bug in Exim, or you have
bad RAM. For now, we'd better assume the former. It might be related
to check_local_user being missing.

> And my exim.conf setting for local delivery:
>
> local_delivery:
> driver = appendfile
> directory = $home/Maildir
> maildir_format
> maildir_use_size_file
> delivery_date_add
> envelope_to_add
> return_path_add
> group = mail
> mode = 0660

You're missing check_local_user, so $home will be unset. So you're
trying to write into /Maildir and likely failing. If this is causing a
segfault, then we're missing an error check somewhere in our code. :(

Add check_local_user, to ensure that the recipient is a local system
user. If some recipients are not local system users, you'll need to
rethink the use of $home here.

> Any idea what may cause this? I think I am missing something in my configuration but cannot find out what it may be.

If the above doesn't fix it, then thaw it and run a debug delivery:

# exim -Mt 1U49cB-0000A6-6a && exim -d+transport -M 1U49cB-0000A6-6a

That should provide more evidence as to where things are going wrong.

-Phil