TCP拥塞控制算法 — CUBIC的补丁（4）

TCP拥塞控制算法 — CUBIC的补丁（四）
描述

以下是提交者Stephen Hemminger对这个patch的描述：

enable high resolution ack time if needed

This is a refined version of an earlier patch by Lucas Nussbaum.

Cubic needs RTT values in miliseconds. If HZ < 1000 then the values will be too coarse.

--- a/net/ipv4/tcp_cubic.c+++ b/net/ipv4/tcp_cubic.c@@ -459,6 +459,10 @@ static int __init cubictcp_register(void) /* divide by bic_scale and by constant Srtt (100ms) */ do_div(cube_factor, bic_scale * 10); +/* hystart needs ms clock resolution */+if (hystart && HZ < 1000)+cubictcp.flags |= TCP_CONG_RTT_STAMP;+ return tcp_register_congestion_control(&cubictcp); }

@include/net/tcp.h：#define TCP_CONG_RTT_STAMP 0x02struct tcp_congestion_ops {    ...    unsigned long flags;    ...};/** * @tstamp: Time we arrived */struct sk_buff {    ...    ktime_t tstamp;    ...};static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, u32 prior_snd_una){    ktime_t last_ackt = net_invalid_timestamp();    ...        ca_seq_rtt = now - scb->when;        last_ackt = skb->tstamp;    ...    if (ca_ops->pkts_acked) {        s32 rtt_us = -1;        /* Is the ACK triggering packet unambiguous? */        if (! (flag & FLAG_RETRANS_DATA_ACKED)) {            /* High resolution needed and available? */            if (ca_ops->flags & TCP_CONG_RTT_STAMP &&                ! ktime_equal(last_ackt, net_invalid_timestamp()))                rtt_us = ktime_us_delta(ktime_get_real(), last_ackt); /* 精确到微秒*/            else if (ca_seq_rtt >= 0)                rtt_us = jiffies_to_usecs(ca_seq_rtt); /*精确度只有jiffies，一般是毫秒 */        }        ca_ops->pkts_acked(sk, pkts_acked, rtt_us);    }    ...}/** * ktime_equal - Compares two ktime_t variables to see if they are equal * @cmp1: comparable1 * @cmp2: comparable2 * Compare two ktime_t variables, returns 1 if equal */static inline int ktime_equal(const ktime_t cmp1, const ktime_t cmp2){    return cmp1.tv64 == cmp2.tv64;}static inline ktime_t net_invalid_timestamp(void){    return ktime_set(0, 0);}/* Set a ktime_t variable to a value in sec/nsec representation: */static inline ktime_t ktime_set (const long secs, const unsigned long nsecs){    return (ktime_t) { .tv = { .sec = secs, .nsec = nsecs} };}

我们知道当HZ < 1000时，计算得到的RTT的精确度就在1ms以上。这个时候可以通过设置

TCP_CONG_RTT_STAMP标志来获得更高精度的RTT，测量得到的RTT可以精确到微秒。

但是由于x86机器的HZ一般为1000，所以这个补丁的影响有限。

zhangskd @ csdn blog