Hello, Well, I didn't have enough sleep lately, so probably this is an obvious and expected result.
But connecting to a closed TCP port (that replies with RST) from OpenBSD doesn't immediately return. Extremely trivial demo (assuming nothing listens on 4242): $ nc 127.0.0.1 4242 It waits a while before throwing Connection refused. Same thing for a remote host: $ nc openbsd.org 4242 The same command with any another OS immediately returns ECONNREFUSED. PF isn't enabled. The behavior is the same with -current, OpenBSD 4.1 and OpenBSD 3.9, so this is not new. Is it intentional? What is the point? Another thing, with -current : $ sudo tcpdump -nvi vge0 dst host 199.185.137.3 & $ nc 199.185.137.3 4242 tcpdump: listening on vge0, link-type EN10MB 01:24:53.473955 88.191.38.240.48850 > 199.185.137.3.4242: SWE [bad tcp cksum d965!] 1361218769:1361218769(0) win 16384 <mss 1460,nop,nop,sackOK,nop,wscale 0,nop,nop,timestamp 138557634 0> (DF) (ttl 64, id 18485, len 64, bad cksum 14!) 01:24:59.462216 88.191.38.240.48850 > 199.185.137.3.4242: S [bad tcp cksum 8d66!] 1361218769:1361218769(0) win 16384 <mss 1460,nop,nop,sackOK,nop,wscale 0,nop,nop,timestamp 138557646 0> (DF) (ttl 64, id 52715, len 64, bad cksum 14!) This is with a vge NIC on OpenBSD/i386. I tried on completely different hardware, on OpenBSD/amd64 with a re NIC and also got those bad checksums everytime. -- Frank Denis - j [at] pureftpd.org - http://00f.net - http://www.cotery.com
