Skip to content

K
Kernel

openeuler / Kernel
大约 1 年 前同步成功

通知 6
Star 0
Fork 0
K
Kernel

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 16 10月, 2018 2 次提交
    • E
      net: extend sk_pacing_rate to unsigned long · 76a9ebe8
      由 Eric Dumazet 提交于 
      sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.

We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.

This patch adds no cost for 32bit kernels.

The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.

Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.

State      Recv-Q Send-Q Local Address:Port             Peer Address:Port
ESTAB      0      1787144  10.246.9.76:49992             10.246.9.77:36741
                 timer:(on,003ms,0) ino:91863 sk:2 <->
 skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
 ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
 rcvmss:536 advmss:1448
 cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
 segs_in:3916318 data_segs_out:177279175
 bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
 send 28045.5Mbps lastrcv:73333
 pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
 busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
 notsent:2085120 minrtt:0.013
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      76a9ebe8
    • E
      tcp: do not change tcp_wstamp_ns in tcp_mstamp_refresh · 5f6188a8
      由 Eric Dumazet 提交于 
      In EDT design, I made the mistake of using tcp_wstamp_ns
to store the last tcp_clock_ns() sample and to store the
pacing virtual timer.

This causes major regressions at high speed flows.

Introduce tcp_clock_cache to store last tcp_clock_ns().
This is needed because some arches have slow high-resolution
kernel time service.

tcp_wstamp_ns is only updated when a packet is sent.

Note that we can remove tcp_mstamp in the future since
tcp_mstamp is essentially tcp_clock_cache/1000, so the
apparent socket size increase is temporary.

Fixes: 9799ccb0 ("tcp: add tcp_wstamp_ns socket field")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NSoheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      5f6188a8
  3. 13 10月, 2018 1 次提交
    • D
      net: Evict neighbor entries on carrier down · 859bd2ef
      由 David Ahern 提交于 
      When a link's carrier goes down it could be a sign of the port changing
networks. If the new network has overlapping addresses with the old one,
then the kernel will continue trying to use neighbor entries established
based on the old network until the entries finally age out - meaning a
potentially long delay with communications not working.

This patch evicts neighbor entries on carrier down with the exception of
those marked permanent. Permanent entries are managed by userspace (either
an admin or a routing daemon such as FRR).
Signed-off-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      859bd2ef
  5. 11 10月, 2018 3 次提交
  7. 09 10月, 2018 5 次提交
  9. 08 10月, 2018 1 次提交
    • J
      udp: Unbreak modules that rely on external __skb_recv_udp() availability · 7e823644
      由 Jiri Kosina 提交于 
      Commit 2276f58a ("udp: use a separate rx queue for packet reception")
turned static inline __skb_recv_udp() from being a trivial helper around
__skb_recv_datagram() into a UDP specific implementaion, making it
EXPORT_SYMBOL_GPL() at the same time.

There are external modules that got broken by __skb_recv_udp() not being
visible to them. Let's unbreak them by making __skb_recv_udp EXPORT_SYMBOL().

Rationale (one of those) why this is actually "technically correct" thing
to do: __skb_recv_udp() used to be an inline wrapper around
__skb_recv_datagram(), which itself (still, and correctly so, I believe)
is EXPORT_SYMBOL().

Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Eric Dumazet <edumazet@google.com>
Fixes: 2276f58a ("udp: use a separate rx queue for packet reception")
Signed-off-by: NJiri Kosina <jkosina@suse.cz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      7e823644
  11. 06 10月, 2018 1 次提交
  13. 05 10月, 2018 4 次提交
  15. 03 10月, 2018 6 次提交
  17. 02 10月, 2018 5 次提交
  19. 30 9月, 2018 1 次提交
    • Y
      tcp: up initial rmem to 128KB and SYN rwin to around 64KB · a337531b
      由 Yuchung Cheng 提交于 
      Previously TCP initial receive buffer is ~87KB by default and
the initial receive window is ~29KB (20 MSS). This patch changes
the two numbers to 128KB and ~64KB (rounding down to the multiples
of MSS) respectively. The patch also simplifies the calculations s.t.
the two numbers are directly controlled by sysctl tcp_rmem[1]:

  1) Initial receiver buffer budget (sk_rcvbuf): while this should
     be configured via sysctl tcp_rmem[1], previously tcp_fixup_rcvbuf()
     always override and set a larger size when a new connection
     establishes.

  2) Initial receive window in SYN: previously it is set to 20
     packets if MSS <= 1460. The number 20 was based on the initial
     congestion window of 10: the receiver needs twice amount to
     avoid being limited by the receive window upon out-of-order
     delivery in the first window burst. But since this only
     applies if the receiving MSS <= 1460, connection using large MTU
     (e.g. to utilize receiver zero-copy) may be limited by the
     receive window.

With this patch TCP memory configuration is more straight-forward and
more properly sized to modern high-speed networks by default. Several
popular stacks have been announcing 64KB rwin in SYNs as well.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NWei Wang <weiwan@google.com>
Signed-off-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reviewed-by: NSoheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      a337531b
  21. 28 9月, 2018 1 次提交
    • T
      netfilter: masquerade: don't flush all conntracks if only one address deleted on device · 097f95d3
      由 Tan Hu 提交于 
      We configured iptables as below, which only allowed incoming data on
established connections:

iptables -t mangle -A PREROUTING -m state --state ESTABLISHED -j ACCEPT
iptables -t mangle -P PREROUTING DROP

When deleting a secondary address, current masquerade implements would
flush all conntracks on this device. All the established connections on
primary address also be deleted, then subsequent incoming data on the
connections would be dropped wrongly because it was identified as NEW
connection.

So when an address was delete, it should only flush connections related
with the address.
Signed-off-by: NTan Hu <tan.hu@zte.com.cn>
Signed-off-by: NPablo Neira Ayuso <pablo@netfilter.org>
      097f95d3
  23. 27 9月, 2018 3 次提交
  25. 25 9月, 2018 1 次提交
  27. 22 9月, 2018 6 次提交
    • P
      net/ipfrag: let ip[6]frag_high_thresh in ns be higher than in init_net · 83619623
      由 Peter Oskolkov 提交于 
      Currently, ip[6]frag_high_thresh sysctl values in new namespaces are
hard-limited to those of the root/init ns.

There are at least two use cases when it would be desirable to
set the high_thresh values higher in a child namespace vs the global hard
limit:

- a security/ddos protection policy may lower the thresholds in the
  root/init ns but allow for a special exception in a child namespace
- testing: a test running in a namespace may want to set these
  thresholds higher in its namespace than what is in the root/init ns

The new behavior:

 # ip netns add testns
 # ip netns exec testns bash

 # sysctl -w net.ipv4.ipfrag_high_thresh=9000000
 net.ipv4.ipfrag_high_thresh = 9000000

 # sysctl net.ipv4.ipfrag_high_thresh
 net.ipv4.ipfrag_high_thresh = 9000000

 # sysctl -w net.ipv6.ip6frag_high_thresh=9000000
 net.ipv6.ip6frag_high_thresh = 9000000

 # sysctl net.ipv6.ip6frag_high_thresh
 net.ipv6.ip6frag_high_thresh = 9000000

The old behavior:

 # ip netns add testns
 # ip netns exec testns bash

 # sysctl -w net.ipv4.ipfrag_high_thresh=9000000
 net.ipv4.ipfrag_high_thresh = 9000000

 # sysctl net.ipv4.ipfrag_high_thresh
 net.ipv4.ipfrag_high_thresh = 4194304

 # sysctl -w net.ipv6.ip6frag_high_thresh=9000000
 net.ipv6.ip6frag_high_thresh = 9000000

 # sysctl net.ipv6.ip6frag_high_thresh
 net.ipv6.ip6frag_high_thresh = 4194304
Signed-off-by: NPeter Oskolkov <posk@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      83619623
    • E
      net/ipv4: avoid compile error in fib_info_nh_uses_dev · 075e264f
      由 Eric Dumazet 提交于 
      net/ipv4/fib_frontend.c: In function 'fib_info_nh_uses_dev':
net/ipv4/fib_frontend.c:322:6: error: unused variable 'ret' [-Werror=unused-variable]
cc1: all warnings being treated as errors

Fixes: 78f2756c ("net/ipv4: Move device validation to helper")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: David Ahern <dsahern@gmail.com>
Reviewed-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      075e264f
    • E
      tcp: switch tcp_internal_pacing() to tcp_wstamp_ns · c092dd5f
      由 Eric Dumazet 提交于 
      Now TCP keeps track of tcp_wstamp_ns, recording the earliest
departure time of next packet, we can remove duplicate code
from tcp_internal_pacing()

This removes one ktime_get_tai_ns() call, and a divide.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      c092dd5f
    • E
      tcp: switch tcp and sch_fq to new earliest departure time model · ab408b6d
      由 Eric Dumazet 提交于 
      TCP keeps track of tcp_wstamp_ns by itself, meaning sch_fq
no longer has to do it.

Thanks to this model, TCP can get more accurate RTT samples,
since pacing no longer inflates them.

This has the nice effect of removing some delays caused by FQ
quantum mechanism, causing inflated max/P99 latencies.

Also we might relax TCP Small Queue tight limits in the future,
since this new model allow TCP to build bigger batches, since
sch_fq (or a device with earliest departure time offload) ensure
these packets will be delivered on time.

Note that other protocols are not converted (they will probably
never be) so sch_fq has still support for SO_MAX_PACING_RATE

Tested:

Test showing FQ pacing quantum artifact for low-rate flows,
adding unexpected throttles for RPC flows, inflating max and P99 latencies.

The parameters chosen here are to show what happens typically when
a TCP flow has a reduced pacing rate (this can be caused by a reduced
cwin after few losses, or/and rtt above few ms)

MIBS="MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY,P99_LATENCY,STDDEV_LATENCY"
Before :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
 Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
19,82.78,5279,3825,482.02

After :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
20,49.94,128,63,3.18
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      ab408b6d
    • E
      tcp: switch internal pacing timer to CLOCK_TAI · fd2bca2a
      由 Eric Dumazet 提交于 
      Next patch will use tcp_wstamp_ns to feed internal
TCP pacing timer, so switch to CLOCK_TAI to share same base.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      fd2bca2a
    • E
      tcp: provide earliest departure time in skb->tstamp · d3edd06e
      由 Eric Dumazet 提交于 
      Switch internal TCP skb->skb_mstamp to skb->skb_mstamp_ns,
from usec units to nsec units.

Do not clear skb->tstamp before entering IP stacks in TX,
so that qdisc or devices can implement pacing based on the
earliest departure time instead of socket sk->sk_pacing_rate

Packets are fed with tcp_wstamp_ns, and following patch
will update tcp_wstamp_ns when both TCP and sch_fq switch to
the earliest departure time mechanism.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      d3edd06e