>>On 6/24/16, 6:14 PM, "Sairam Venugopal" <vsai...@vmware.com> wrote:
>>
>>>Update the code to use tcp->flags. This keeps the kernel conntrack-tcp.c
>>>file in sync with userspace version.
>>>
>>>This patch also addresses an warning - 'Comparison of a boolean
>>>expression with an integer other than 0 or 1' - (tcp_flags &
>>>(TCP_ACK|TCP_RST)) == (TCP_ACK|TCP_RST))
>>>
>>>Signed-off-by: Sairam Venugopal <vsai...@vmware.com>
>
>>>@@ -232,31 +226,33 @@ OvsConntrackUpdateTcpEntry(OVS_CT_ENTRY* conn_,
>>> /* The peer that should receive 'pkt' */
>>> struct tcp_peer *dst = &conn->peer[reply ? 0 : 1];
>>> uint8_t sws = 0, dws = 0;
>>>+ UINT16 tcp_flags = tcp->flags;
>>> uint16_t win = ntohs(tcp->window);
>>> uint32_t ack, end, seq, orig_seq;
>>> uint32_t p_len = OvsGetTcpPayloadLength(nbl);
>>> int ackskew;
>>>
>>>- if (OvsConntrackValidateTcpFlags(tcp)) {
>>>+ if (OvsCtInvalidTcpFlags(tcp->flags)) {
>>> return CT_UPDATE_INVALID;
>>> }
>>>
>>>- if ((tcp->syn) && dst->state >= CT_DPIF_TCPS_FIN_WAIT_2 &&
>>>- src->state >= CT_DPIF_TCPS_FIN_WAIT_2) {
>>>+ if (((tcp_flags & (TCP_SYN|TCP_ACK)) == TCP_SYN)
>
>Why do we need to include TCP_ACK in the check?
We spoke offline. This is a bug fix.
Acked-by: Nithin Raju <nit...@vmware.com>
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
