It's definitely bug in Windows XP SP1, as it was supposed by Roland Postle <email@example.com> To reproduce it:
254 IN CNAME non.existant.name
Use test program attached
I did tests on Windows NT 4.0, Windows 2000 and Windows XP SP1. Results:
Windows NT 4.0:
c:\>test.exe 192.168.1.254 gethostbyaddr failed
C:\>test.exe 192.168.1.254 gethostbyaddr failed
Windows XP SP1:
C:\>test.exe 192.168.1.254 h_name: (null)
So, this problem is not specific to mIRC and it's possible to crash any application on Windows XP Sp1 where gethostbyaddr() or WSAAsyncGetHostByAddr() is used for reverse name resolution (IRC clients, Peer-to-Peer clients, personal firewalls, etc).
Can somebody test Windows 2003?
--Thursday, May 29, 2003, 11:52:59 AM, you wrote to firstname.lastname@example.org:
3> Dear Peter Pentchev,
3> A way to delegate reverse resolution for network less than /24 is 3> defined in RFC 2317. And it's different from one used.
3> But you're right: the problem is probably not in unresolvable PTR. The 3> problem is in unresolvable CNAME instead of PTR, so PTR is never found 3> at all. And yes: it may affect different applications where 3> gethostbyname() is used. I will test gethostbyname() behavior for this 3> case in Windows and Unix and report back.
3> --Thursday, May 29, 2003, 11:26:04 AM, you wrote to 3APA3A@SECURITY.NNOV.RU:
PP>> On Wed, May 28, 2003 at 02:45:25PM +0400, 3APA3A wrote: >>> Dear Davide Del Vecchio, >>> >>> Currently 126.96.36.199 doesn't resolve. But during original test it had >>> flowed PTR record: >>> >>> bash-2.03$ host 188.8.131.52 >>> 184.108.40.206.IN-ADDR.ARPA is a nickname for 220.127.116.11.210.IN-ADDR.ARPA >>> >>> (.16 is twice)
PP>> This is not necessarily a flawed record; I believe it might be as simple PP>> as the ultimate in classless reverse DNS delegation. Note that the PP>> 16.193.210.in-addr.arpa zone is delegated to ns.qala.com.sg, while PP>> this specific "alias" subdelegates the reverse DNS records for PP>> 18.104.22.168 to dns.lga.net.sg.
PP>> G'luck, PP>> Peter
-- ~/ZARAZA Машина оказалась способной к единственному действию, а именно умножению 2x2, да и то при этом ошибаясь. (Лем)