This is a discussion on Re: [PATCH] statetop hostname resolution and bug fixes within the IPFilter forums, part of the System Security and Security Related category; On Sun, Apr 03, 2005 at 04:07:47AM -0700, Phil Dibowitz wrote: > OK I made the changes I ...
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
|
|||
|
On Sun, Apr 03, 2005 at 04:07:47AM -0700, Phil Dibowitz wrote:
> OK I made the changes I mentioned before as well as updating the man > page for ipfstat. This should be a fully functional patch. > > It would be nice if someone would test this with IPv6, but short of > that, I think it's ready and done. > > As always, comments welcome. Here's the updated changelog: > Looks useful. I've some comments on the source, see below. IPv6 statetop still works fine for me, but resolving of IPv6 addresses does not. One suggestion: maybe implement port lookups as well (by using portname() from lib/portname.c) ? > > + sname = malloc(STSTRSIZE); > + dname = malloc(STSTRSIZE); > + malloc() can fail, better check for NULL, or use static allocation. > + hostinfo = gethostbyaddr((char *)&ips.is_src,sizeof(ips.is_src), AF_INET); I've two problems with this, the sizeof is wrong because is_src is actually much bigger than you want. Better replace this with sizeof(struct in_addr). Or use hostname() from lib/hostname.c. On NetBSD/sparc64, the resolving doesn't works unless you do one of the above. And secondly, this only works for AF_INET since you've hardcoded that. > + if (length < 2) { > + fprintf(stderr,"Error, screen too small\n"); > + exit(0); > + } This doesn't restore the terminal. > +SIGWINCH doesn't seem to get through to STATETOP. Resizing works for me (NetBSD-3.99.2/sparc64). Does handle_resize ever get set to 1 when you resize? Cheers, -- Peter Postma |