. kibble . wrote:
>> On Wed, 19 Jan 2005 20:57:26 Philip Hazel wrote:
>>
>> On Wed, 19 Jan 2005, Fred Viles wrote:
>>
>> > | That there is no core is probably because Exim is a setuid binary.
>> >
>> > Um, in the case of local_scan code, isn't it more likely because of
>> > this? (from Ch 39):
>> >
>> > Exim applies a timeout to calls of the local scan function, and
>> there is an
>> > option called "local_scan_timeout" for setting it. The default is 5
>> minutes.
>> > Zero means 'no timeout'. Exim also sets up signal handlers for SIGSEGV,
>> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Of course it is. Silly me. That's how it logged the error...
>>
>
> But of course:
> "The incident is logged on the main and reject logs."
>
> I' am getting the logs [in my main exim log that is] which only reports:
> "local_scan() function crashed with signal 11 - message temporarily
> rejected "
>
> This however doesn't uncover the exact reason why it crashed... Not had
> time to look into my code yet, going away to the Czech Rep tonight and
> won't be back till end of the month. Hopefully I will be able to fill my
> code with tonnes of debug_printf commands and find out where it goes wrong.
>
> Best regards, kibble.
>
when I was creating my local_scan function, I too had signal 11
crashes.
I had some pointer problems in my code which I found by sucking the
routine into it's own executable and thrashing it with the
header&bodies of the precise messages which caused the sig 11
problems.
Jim
