On Thu, 1 Dec 2005, John Jetmore wrote:
> Report from a Darwin machine
Fixed some things. Comments on others:
> Basic/0142 open file descriptor test
<snip>
> I have no idea what those extra FDs are. If you have any ideas about next
> steps for this I'd love to hear them =).
I am mystified too. This is during a transport filter. What the filter
process is supposed to do is to transfer the pipe fds onto 0 and 1
(stdin and stdout), and close the original fds. Maybe something odd is
happening about that. OTOH, your extra fds seem to be read only. Hmm.
> 0222 test output <<END OF 0222 OUTPUT
> Basic/0222 autoreply and once and once_file_size
> I honestly have no idea what the deal is with the above output. Submitted
> for your opinion.
I have made a change that might or might not make a difference. Let's
see what happens next time round.
> Basic/0226 redirect: file and pipe addresses
> 1999-03-02 09:44:33 10HmaZ-0005vi-00 Process failed (-256) when writing
> error me
> ssage to CALLER@??? (frozen)
> I got the above exactly once. Every time I tried to reproduce it the
> script ended normally. Submitted here as a curiosity
That -256 indicates a timeout while the first Exim was waiting for the
second Exim (which is receiving a bounce message) to return. However, in
this particular situation, there is not supposed to be any timeout. It
suggests that somehow, somewhere, an alarm() didn't get cancelled when
it should have, but tracking down something like that is difficult.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
Get the Exim 4 book: http://www.uit.co.uk/exim-book
