Bug in smtp retry timeout?

I've found something very odd with one particular machine I'm trying to
send mail to.

A customer complained about not being able to get mail to an address in
hong kong. I tried with exim in debug mode, and sure enough, exim
bounces it with an smtp retry timeout error, but, it does not try for
anywhere close to the 300 seconds it claims, it bounces immediately.
Here is the bottom of the trace.

returned hosts:
fw1.vtc.edu.hk 202.40.210.221 10
lookuphost router succeeded
routed by lookuphost router:
deliver to nwlem@???
transport: smtp
host fw1.vtc.edu.hk [202.40.210.221] MX=10
search_tidyup called
>>> Local deliveries >>>
search_tidyup called
>>> Remote deliveries >>>
smtp transport entered
fw1.vtc.edu.hk [202.40.210.221] status = usable
set_process_info: 3100 1.58 delivering 0vhGzW-0000nf-00 to
fw1.vtc.edu.hk [202.40.210.221
] (nwlem@???)
Connecting to fw1.vtc.edu.hk [202.40.210.221] ... connected
SMTP response timeout = 300
SMTP response timeout = 300

The last 2 lines were printed immediately, not after close to 300
seconds. Also, if I telnet to the smtp port of the mx host in question,
I can enter the mail by hand without problem.

This happens with exim on both SCO and Solaris. On linux, the full 300
seconds appears to pass. However, what is totally baffling to me is that
I can telnet to the smtp port in question and get a response very
quickly. Here's a session. There were no long pauses in it except those
cause by my slow typing. Is exim failing to recognise a valid response
here or something?


Trying 202.40.210.221...
Connected to fw1.vtc.edu.hk.
Escape character is '^]'.
220 fw0.vtc.edu.hk SMTPXD version 1.14 ready at Mon Jan 6 23:30:16 1997
ehlo bogon.com
500 Command unrecognized
helo bogon.com
250 Hello stdismas.bogon.com, why did you call yourself bogon.com
mail from: root@???
250 root@???... Sender ok
rcpt to: postmaster
250 postmaster... Recipient ok
data
354 Enter mail, end with "." on a line by itself
test. ignore