From: Peter Maydell <[email protected]> In qemu_receive_packet() we check to see if we should pad a short packet. This is doing the wrong test: because this function is used when the device adds a packet to its own incoming queue (i.e. for loopback), we should be checking the NetClientState's own do_not_pad flag, not that for its peer.
We didn't notice this earlier, because at the moment all the real peers of a network device (i.e. the network backends) do not set do_not_pad, so net_peer_needs_padding() always returns true except in the corner case where the network device has no peer at all. The effect of this is that if a network device has no peer (e.g. because QEMU was started with -net none or with -nodefaults) then we can still let through the kind of "guest misprograms the network device to loopback-transmit a short packet and then we mishandle it in the receive path" bug like #3043 which commit a01344d9d78 was trying to fix. Since the distinction between "we should check nc->do_not_pad" and "we should check nc->peer->do_not_pad" is a bit subtle, add enough documentation commentary to make it more obvious. Cc: [email protected] Fixes: a01344d9d78 ("net: pad packets to minimum length in qemu_receive_packet()") Suggested-by: Bin Meng <[email protected]> Signed-off-by: Peter Maydell <[email protected]> Reviewed-by: Bin Meng <[email protected]> Message-ID: <[email protected]> Signed-off-by: Philippe Mathieu-Daudé <[email protected]> (cherry picked from commit b1e73b2ddbb2850e394b9c47cfcb14dba8048a71) Signed-off-by: Michael Tokarev <[email protected]> diff --git a/include/net/net.h b/include/net/net.h index cdd5b109b0d..e42e8263e8a 100644 --- a/include/net/net.h +++ b/include/net/net.h @@ -329,9 +329,32 @@ uint32_t net_crc32_le(const uint8_t *p, int len); .offset = vmstate_offset_macaddr(_state, _field), \ } +/** + * net_peer_needs_padding: Should we pad as we send out packets? + * @nc: NetClientState + * + * Return true if the peer of this NetClientState (i.e. the + * destination that qemu_send_packet() etc send to) requires us to pad + * out packets that are shorter than the minimum ethernet frame + * length. + */ static inline bool net_peer_needs_padding(NetClientState *nc) { return nc->peer && !nc->peer->do_not_pad; } +/** + * net_client_needs_padding: Should we pad as we queue packets to ourselves? + * @nc: NetClientState + * + * Return true if this NetClientState requires us to pad out packets + * that are shorter than the minimum ethernet frame length. This is + * the check to make in qemu_receive_packet() when we are queuing a + * packet back into ourselves (i.e. loopback). + */ +static inline bool net_client_needs_padding(NetClientState *nc) +{ + return !nc->do_not_pad; +} + #endif diff --git a/net/net.c b/net/net.c index c2d0d83bc96..4a5ceaf7865 100644 --- a/net/net.c +++ b/net/net.c @@ -764,7 +764,7 @@ ssize_t qemu_receive_packet(NetClientState *nc, const uint8_t *buf, int size) return 0; } - if (net_peer_needs_padding(nc)) { + if (net_client_needs_padding(nc)) { if (eth_pad_short_frame(min_pkt, &min_pktsz, buf, size)) { buf = min_pkt; size = min_pktsz; -- 2.47.3
