when ever maildrop starts processing my .mailfilter it crashes with a 0x06 error code. -Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient------- 1813A5782F 6662 Fri Sep 24 16:31:14 xxxxx@xxxxxx.com (temporary failure. Command output: maildrop: signal 0x06) moxum@localhost.xxxxx.xxxxx.xxxxxx -- 6 Kbytes in 1 Request. Reproducible: Always Steps to Reproduce: 1. 2. 3. I have tried to compile maildrop without the -O2 flags, that changed the error code to 0xB (segfault?) For now I just keep my .mailfilter emtpy, that way maildrop delivers perfectly.
I resolved the issue by editting the ebuild definition. I disabled mysql and ldap recompiled it and it works for me now.
What version of maildrop, and what did you do to the ebuild?
I also see maildrop 0x0B (Signal 11) errors when maildrop is compiled with -Os. This is ONLY a problem when it attempts to filter an email [it processes fine, but if a filter match occurs, so does a 0x0B].
maildrop-1.8.0 is in portage right now. Still fails for you ? If it is fixed maybe is time to stabilize it. Cheers, Ferdy
maildrop-1.8.0-r1 is on CVS. Need to know if still fails for you. Cheers, Ferdy
just wanted to add, that i have got the same problem. couldn't find a solution for that. as mentioned above, an error only occurs if maildrop starts to filter something. without any rules in ~/.mailfilter, it delivers into DEFAULT without problems.
I do not use Gentoo at the moment, so my ability to verify is non-existant. However, it seems it has been seen by somebody else. I do recall the issue being partly based upon optimization level and only -O will work. Any greater optimization (such as -O2, -O3 or -Os) causes this failure. I think there is a memory bug in the software that is triggered by the optimization level, but has nothing to do with the optimization level in and of itself.
