This problem is, indeed, very similar to the November 23rd bug fix; in fact, it was that bug fix that caused me to notice this problem. The difference is that the November 23rd bug fix covers the Greatest Acknowledgment Number Received; whereas, this bug fix covers the Greatest Sequence Number Received. On Tue, 2010-12-28 at 10:44 +0100, Gerrit Renker wrote: > Thanks a lot for looking into this. > > I think the problem below is similar to or even covered by the bug fix > submitted on 23th November (it is in net-2.6 but may not yet be in mainline): > > http://eden-feed.erg.abdn.ac.uk/cgi-bin/gitweb.cgi?p=dccp_exp.git;a=commitdiff;h=0ac78870220b6e0ac74dd9292bcfa7b18718babd > > I will have a look at the other patches, but it maybe "next year" until > able to get back. > > > Quoting Samuel Jero: > | Currently dccp_check_seqno allows any valid packet to update the > | Greatest Sequence Number Received, even if that packet's sequence number > | is less than the current GSR. This patch adds a check to make sure that > | the new packet's sequence number is greater than GSR. > | > | >>>>>>>>>>>>>>>>>>>>>>>>> Patch Follows <<<<<<<<<<<<<<<<<<<<<<<<<< > | diff --git a/net/dccp/input.c b/net/dccp/input.c > | --- a/net/dccp/input.c > | +++ b/net/dccp/input.c > | @@ -252,7 +253,10 @@ static int dccp_check_seqno(struct sock *sk, struct > | sk_buff *skb) > | if (between48(seqno, lswl, dp->dccps_swh) && > | (ackno == DCCP_PKT_WITHOUT_ACK_SEQ || > | between48(ackno, lawl, dp->dccps_awh))) { > | - dccp_update_gsr(sk, seqno); > | + > | + if(after48(seqno, dp->dccps_gsr)){ > | + dccp_update_gsr(sk, seqno); > | + } > | > | if (dh->dccph_type != DCCP_PKT_SYNC && > | ackno != DCCP_PKT_WITHOUT_ACK_SEQ && >
Attachment:
signature.asc
Description: This is a digitally signed message part
