Phillip Hazel wrote:
>On Wed, 13 Feb 2002, Christian Balzer wrote:
>
>> Filter: end of processing
>> restored uid=0 gid=0 euid=8 egid=8
>> Segmentation fault
>> ---
>>
>> So was something changed in the filter specific or generic lsearch handling?
>> And if so, will it change back or do I need to deal with this differently
>> from now on? (partial lookups don't seem to be what I want either)
>
>Exim shouldn't segfault, whatever. I will see if I can reproduce this
>error. There was certainly no intention of any change between 3.33 and
>3.34. I'm winding down Exim 3 to bug fixes only.
>
Well, it turns out that this is not the lsearch bit I was thinking,
it must be something within the filter that breaks exim. Was lured
there by the -d9 output. The thing is, the filter matches and that
gets written to the filterlog. Alas the delivery never happens.
Since this happens for all kinds of transports and directors (and
since one would expect more debugging output otherwise) the filter
is my prime suspect. However parts of it fully work (it's broken
into several segments) and if you test it with -bF against the
tokens in the data files that cause it to segfault all works fine.
I'm quite puzzled at this point, if you would like the exim.conf
and filter in question, just let me know.
Regards,
Christian Balzer
--
Christian Balzer Network Engineer Engineering
chibi@??? Exodus Communications K.K.
Phone: +81 3 4354 0290 FAX: +81 3 ?? http://www.gol.com/
