On Tue 2005.09.20 at 12:48 -0600, Chris Kuethe wrote: > On 9/20/05, Okan Demirmen <[EMAIL PROTECTED]> wrote: > > On Tue 2005.09.20 at 11:12 -0600, Chris Kuethe wrote: > > > On 9/20/05, Okan Demirmen <[EMAIL PROTECTED]> wrote: > > > > too many changes to list here, but one positive note - no more local > > > > patches needed...update from 3.81 -> 3.93 > > > > > > and on the up side, this makes os fingerprinting work again. > > > just tested on i386 with the sept 10 snapshot... > > > > it was broken before? not on any of my builds/archs. what was the > > error/problem/issue - on what archs? > > http://marc.theaimsgroup.com/?l=openbsd-misc&m=112685846427703&w=2 > > Much like the thread above, os detection scans would fail with "no > route to host". I was using snapshots and packages, though. This was > reproducible with both nmap 3.81 and 3.82.
yea, so turn off pf(4), or more specifically, state creation, especially when you are doing syn-scans (default when done as root). which archs did it break for you? on a side note, i recently found that state creation on amd64 and sparc64 seem to top out around 140, but on i386, i can create states up to the configured max. however, none of this is highly scientific. i still need to isolate more to figure it out.
