Arkadiusz Mi�kiewicz <[EMAIL PROTECTED]> writes: > This is bad thing ;-) > > Also you should use getaddrinfo() && getnameinfo() instead gethostbyaddr() > etc., > avoid sockaddr_in and sockaddr_in6; use sockaddr_storage etc....
I'm not the author of the patches, I'm just helping him test them since most of his stuff is fbsd with the kame stack, and his email address is on the web site I posted in my mail to the list. I do know he's in the process of reworking major portions of the patches, it'll probably be a few days before an updated patch is available as I think he's still out of town. > Another thing: > Jul 19 21:18:39 linstar postfix/smtp[1212]: 31CA426107: to=<[EMAIL > PROTECTED]>, relay=none, delay=0, status=bounced (Name service error for > domain linstar.ipv6.zsz2.starachowice.pl: Host not found) > while > [EMAIL PROTECTED] /root# host -t MX ipv6.zsz2.starachowice.pl > ipv6.zsz2.starachowice.pl mail is handled (pri=0) by > linstar.ipv6.zsz2.starachowice.pl > > and linstar.ipv6.zsz2.starachowice.pl has _only_ IPv6 address. hmmm, I'll hafta do some more testing on my end and see if I can track that down.. > ps. small patch for your patch ;) > -+#ifdef INET6 > ++#if defined(INET6) && !defined(linux) this was basically the way I patched this first, but to avoid tampering with the nrl stack I changed this test to detect a glibc2.1 system, since just testing for linux wasn't clear enough. The testing packages I made for this are already patched as such and compile cleanly. Marc
