Lines Matching defs:reordering

605 	nr_segs = max_t(u32, nr_segs, tp->reordering + 1);
1216 	 * DSACKs that may have been due to reordering causing RACK to trigger
1218 	 * without having seen reordering, or that match TLP probes (TLP
1231 /* It's reordering when higher sequence was delivered (i.e. sacked) before
1232  * some lower never-retransmitted sequence ("low_seq"). The maximum reordering
1233  * distance is approximated in full-mss packet distance ("reordering").
1247 	if ((metric > tp->reordering * mss) && mss) {
1251 			 tp->reordering,
1256 		tp->reordering = min_t(u32, (metric + mss - 1) / mss,
1351  *    ever retransmitted -> reordering. Alas, we cannot use it
1354  *    for retransmitted and already SACKed segment -> reordering..
1401  * fragmentation and packet reordering past skb's retransmission. To consider
1577 				 * which was in hole. It is reordering.
2047 				/* Don't count olds caused by ACK reordering */
2206  * in assumption of absent reordering, interpret this as reordering.
2216 	tp->reordering = min_t(u32, tp->packets_out + addend,
2332 	u8 reordering;
2351 	 * suggests that the degree of reordering is over-estimated.
2353 	reordering = READ_ONCE(net->ipv4.sysctl_tcp_reordering);
2355 	    tp->sacked_out >= reordering)
2356 		tp->reordering = min_t(unsigned int, tp->reordering,
2357 				       reordering);
2476  * fast retransmit (reordering) and underestimated RTO, analyzing
3047 		 * packet, rather than with a retransmit. Check reordering.
3051 		/* We are getting evidence that the reordering degree is higher
3167 			 * starts a new recovery (e.g. reordering then loss);
3527 			/* Non-retransmitted hole got filled? That's reordering */
3613 	/* If reordering is high then always grow cwnd whenever data is
3619 	if (tcp_sk(sk)->reordering >
4519  * the biggest problem on large power networks even with minor reordering.