Hi, I received a bugfix for my IPv6 patch to the Windows TAP driver today, which affects IPv4 - and there is this open bug with small-size IPv4 packets in our trac, which this should fix...
The problem is a missing "break" statement at the end of the IPv6
block, which now leads to "fall into IPv6 block" and that one has a
larger minimum packet size...
ACK from me - David, could you please integrate the one-line patch,
and then we need James to compile + sign a new driver (and I think
this warrants a 2.2 bugfix release).
[I'm not on IRC these days as I'm technically on vacation, but I saw
this mail today and it warrants action]
gert
----- Forwarded message from Christian Niessner <bug-rep...@secadm.de> -----
@@ -1611,6 +1766,27 @@ AdapterTransmit (IN NDIS_HANDLE p_AdapterContext,
// Packet looks like IPv4, queue it.
l_PacketBuffer->m_SizeFlags |= TP_TUN;
+
+ case ETH_P_IPV6:
+ // make sure that packet is large
+ // enough to be IPv6
+ if (l_PacketLength
+ < ETHERNET_HEADER_SIZE + IPV6_HEADER_SIZE)
+ goto no_queue;
+
[..]
--- ../tap-win32-latest/tapdrvr.c 2011-11-09 17:12:19.000000000 +0100
+++ tapdrvr.c 2011-11-10 09:24:24.000000000 +0100
@@ -1766,6 +1766,7 @@
// Packet looks like IPv4, queue it.
l_PacketBuffer->m_SizeFlags |= TP_TUN;
+ break;
case ETH_P_IPV6:
// make sure that packet is large
----- End forwarded message -----
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany g...@greenie.muc.de
fax: +49-89-35655025 g...@net.informatik.tu-muenchen.de
pgpE9ila8_fWs.pgp
Description: PGP signature
