tcp: implement TSQ for retransmits

We saw sch_fq drops caused by the per flow limit of 100 packets and TCP
when dealing with large cwnd and bursts of retransmits.

Even after increasing the limit to 1000, and even after commit
10d3be569243 ("tcp-tso: do not split TSO packets at retransmit time"),
we can still have these drops.

Under certain conditions, TCP can spend a considerable amount of
time queuing thousands of skbs in a single tcp_xmit_retransmit_queue()
invocation, incurring latency spikes and stalls of other softirq
handlers.

This patch implements TSQ for retransmits, limiting number of packets
and giving more chance for scheduling packets in both ways.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

path: root/proc.h

#ifndef PROC_H
#define PROC_H

#include <stdlib.h>

extern void cpu_affinity(int cpu);
extern int set_proc_prio(int prio);
extern int set_sched_status(int policy, int priority);
extern ssize_t proc_get_cmdline(unsigned int pid, char *cmdline, size_t len);
extern int proc_exec(const char *proc, char *const argv[]);

#endif /* PROC_H */