https://issues.apache.org/bugzilla/show_bug.cgi?id=56452
--- Comment #4 from Christopher Schultz <ch...@christopherschultz.net> --- (In reply to Konstantin Kolinko from comment #3) > (In reply to Christopher Schultz from comment #2) > > Created attachment 31557 [details] > > Proposed patch (against mod_jk/trunk) > > > > I have raised the size of buf from 64 to 100 characters. The message looks > > like it will need 96 bytes for a complete IPv6 message (e.g. > > 2001:0db8:0000:0000:0000:ff00:0042:8329:65556 -> > > 2001:0db8:0000:0000:0000:ff00:0042:8329:65535 + \0), so I just rounded up to > > 100 bytes. > > > > I've also added proper length-tracking to the buffer and more protection > > against overruns. > > The following discussion says that IPv6 address needs maximum 45 characters, > not 39. So you would need 12 bytes more than 96 = 108. Bah, I forgot about tunnelled-IPv4. However: "0000:0000:0000:0000:0000:0000:192.168.0.1:65535 -> 0000:0000:0000:0000:0000:0000:192.168.0.1:65535" That's only 97 bytes plus a NULL-terminator. Where did you get the 39 from? > http://stackoverflow.com/questions/166132/maximum-length-of-the-textual- > representation-of-an-ipv6-address > > Also see tmp buffer size here: > http://svn.apache.org/viewvc/apr/apr/trunk/network_io/unix/inet_ntop. > c?revision=573491&view=markup#l149 I did see that code in APR while researching. I suppose I could use that, but I figured that "buf" might be used for other things so I didn't consider it. I think it turns out that "buf" is really just for this purpose, so we could do something like this if you'd prefer: char buf[sizeof "0000:0000:0000:0000:0000:0000:192.168.0.1:65535 -> 0000:0000:0000:0000:0000:0000:192.168.0.1:65535" + 1]; -- You are receiving this mail because: You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
