AuxXxilium/linux_dsm_epyc7002
1
0
Fork 0
mirror of https://github.com/AuxXxilium/linux_dsm_epyc7002.git synced 2025-01-19 07:06:23 +07:00
Code Issues Actions Packages Projects Releases Wiki Activity
d41a69f1d3
linux_dsm_epyc7002/net/ipv4
History
Eric Dumazet d41a69f1d3 tcp: make tcp_sendmsg() aware of socket backlog 
Large sendmsg()/write() hold socket lock for the duration of the call,
unless sk->sk_sndbuf limit is hit. This is bad because incoming packets
are parked into socket backlog for a long time.
Critical decisions like fast retransmit might be delayed.
Receivers have to maintain a big out of order queue with additional cpu
overhead, and also possible stalls in TX once windows are full.

Bidirectional flows are particularly hurt since the backlog can become
quite big if the copy from user space triggers IO (page faults)

Some applications learnt to use sendmsg() (or sendmmsg()) with small
chunks to avoid this issue.

Kernel should know better, right ?

Add a generic sk_flush_backlog() helper and use it right
before a new skb is allocated. Typically we put 64KB of payload
per skb (unless MSG_EOR is requested) and checking socket backlog
every 64KB gives good results.

As a matter of fact, tests with TSO/GSO disabled give very nice
results, as we manage to keep a small write queue and smaller
perceived rtt.

Note that sk_flush_backlog() maintains socket ownership,
so is not equivalent to a {release_sock(sk); lock_sock(sk);},
to ensure implicit atomicity rules that sendmsg() was
giving to (possibly buggy) applications.

In this simple implementation, I chose to not call tcp_release_cb(),
but we might consider this later.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Alexei Starovoitov <ast@fb.com>
Cc: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-05-02 17:02:26 -04:00
..
netfilter Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next 2016-04-24 00:12:08 -04:00
af_inet.c GSO: Support partial segmentation offload 2016-04-14 16:23:41 -04:00
ah4.c ah4: Fix error return in ah_input(). 2015-08-25 13:38:50 -07:00
arp.c net: rename NET_{ADD|INC}_STATS_BH() 2016-04-27 22:48:24 -04:00
cipso_ipv4.c net: introduce lockdep_is_held and update various places to use it 2016-04-07 16:44:14 -04:00
datagram.c net: Set sk_txhash from a random number 2015-07-29 22:44:04 -07:00
devinet.c ipv4: Don't do expensive useless work during inetdev destroy. 2016-03-13 23:28:35 -04:00
esp4.c esp4: Switch to new AEAD interface 2015-05-28 11:23:20 +08:00
fib_frontend.c ipv4/fib: don't warn when primary address is missing if in_dev is dead 2016-04-24 23:26:29 -04:00
fib_lookup.h ipv4: consider TOS in fib_select_default 2015-07-24 22:46:11 -07:00
fib_rules.c net: ipv6: use common fib_default_rule_pref 2015-09-09 14:19:50 -07:00
fib_semantics.c net: ipv4: Consider failed nexthops in multipath routes 2016-04-11 15:16:13 -04:00
fib_trie.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2016-02-01 15:56:08 -08:00
fou.c ip_tunnel_core: iptunnel_handle_offloads returns int and doesn't free skb 2016-04-16 19:09:13 -04:00
gre_demux.c gre: Remove support for sharing GRE protocol hook. 2015-08-10 14:03:54 -07:00
gre_offload.c GRE: Add support for GRO/GSO of IPv6 GRE traffic 2016-04-16 19:09:13 -04:00
icmp.c net: icmp: rename ICMPMSGIN_INC_STATS_BH() 2016-04-27 22:48:23 -04:00
igmp.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2016-03-08 12:34:12 -05:00
inet_connection_sock.c net: rename IP_INC_STATS_BH() 2016-04-27 22:48:23 -04:00
inet_diag.c sock_diag: align nlattr properly when needed 2016-04-26 12:00:48 -04:00
inet_fragment.c inet: kill unused skb_free op 2016-01-05 22:25:57 -05:00
inet_hashtables.c net: rename NET_{ADD|INC}_STATS_BH() 2016-04-27 22:48:24 -04:00
inet_timewait_sock.c net: rename NET_{ADD|INC}_STATS_BH() 2016-04-27 22:48:24 -04:00
inetpeer.c net: Add helper function to compare inetpeer addresses 2015-08-28 13:32:36 -07:00
ip_forward.c net: rename IP_ADD_STATS_BH() 2016-04-27 22:48:24 -04:00
ip_fragment.c net: rename IP_INC_STATS_BH() 2016-04-27 22:48:23 -04:00
ip_gre.c ip_tunnel_core: iptunnel_handle_offloads returns int and doesn't free skb 2016-04-16 19:09:13 -04:00
ip_input.c net: rename NET_{ADD|INC}_STATS_BH() 2016-04-27 22:48:24 -04:00
ip_options.c net: ipv4: Convert IP network timestamps to be y2038 safe 2016-03-01 17:18:44 -05:00
ip_output.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2016-03-08 12:34:12 -05:00
ip_sockglue.c net: better drop monitoring in ip{6}_recv_error() 2016-04-25 15:48:10 -04:00
ip_tunnel_core.c libnl: nla_put_be64(): align on a 64-bit area 2016-04-23 20:13:24 -04:00
ip_tunnel.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2016-03-08 12:34:12 -05:00
ip_vti.c ip_tunnel: Move stats update to iptunnel_xmit() 2015-12-25 23:32:23 -05:00
ipcomp.c ipv4: coding style: comparison for equality with NULL 2015-04-03 12:11:15 -04:00
ipconfig.c ipv4: ipconfig: avoid unused ic_proto_used symbol 2016-01-29 19:39:09 -08:00
ipip.c ip_tunnel_core: iptunnel_handle_offloads returns int and doesn't free skb 2016-04-16 19:09:13 -04:00
ipmr.c ipmr: align RTA_MFC_STATS on 64-bit 2016-04-21 14:22:13 -04:00
Kconfig ipv4: Remove inet_lro library 2016-02-17 16:15:46 -05:00
Makefile ipv4: Remove inet_lro library 2016-02-17 16:15:46 -05:00
netfilter.c ipv4: Pass struct net into ip_route_me_harder 2015-09-29 20:21:32 +02:00
ping.c sock: enable timestamping using control messages 2016-04-04 15:50:30 -04:00
proc.c ipv4: Namespaceify ip_default_ttl sysctl knob 2016-02-16 20:42:54 -05:00
protocol.c net: Export inet_offloads and inet6_offloads 2014-09-19 17:15:31 -04:00
raw.c sock: enable timestamping using control messages 2016-04-04 15:50:30 -04:00
route.c net: rename IP_INC_STATS_BH() 2016-04-27 22:48:23 -04:00
syncookies.c net: rename NET_{ADD|INC}_STATS_BH() 2016-04-27 22:48:24 -04:00
sysctl_net_ipv4.c net: ipv4: Consider failed nexthops in multipath routes 2016-04-11 15:16:13 -04:00
tcp_bic.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_cdg.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_cong.c tcp: remove tcp_ecn_make_synack() socket argument 2015-09-25 13:00:38 -07:00
tcp_cubic.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_dctcp.c tcp: allow dctcp alpha to drop to zero 2015-10-23 02:46:52 -07:00
tcp_diag.c net: diag: Support destroying TCP sockets. 2015-12-15 23:26:52 -05:00
tcp_fastopen.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_highspeed.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_htcp.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_hybla.c tcp: do not slow start when cwnd equals ssthresh 2015-07-09 14:22:52 -07:00
tcp_illinois.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_input.c tcp: do not block bh during prequeue processing 2016-05-02 17:02:25 -04:00
tcp_ipv4.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_lp.c tcp: remove in_flight parameter from cong_avoid() methods 2014-05-03 19:23:07 -04:00
tcp_metrics.c libnl: nla_put_msecs(): align on a 64-bit area 2016-04-23 20:13:24 -04:00
tcp_minisocks.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_offload.c GSO: Support partial segmentation offload 2016-04-14 16:23:41 -04:00
tcp_output.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_probe.c net: ipv4: tcp_probe: Replace timespec with timespec64 2016-03-01 17:18:44 -05:00
tcp_recovery.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_scalable.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_timer.c tcp: do not assume TCP code is non preemptible 2016-05-02 17:02:25 -04:00
tcp_vegas.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_vegas.h tcp: prepare CC get_info() access from getsockopt() 2015-04-29 17:10:38 -04:00
tcp_veno.c tcp: add tcp_in_slow_start helper 2015-07-09 14:22:52 -07:00
tcp_westwood.c tcp_westwood: fix tcp_westwood_info() 2015-05-05 19:50:09 -04:00
tcp_yeah.c tcp_yeah: don't set ssthresh below 2 2016-01-11 17:25:16 -05:00
tcp.c tcp: make tcp_sendmsg() aware of socket backlog 2016-05-02 17:02:26 -04:00
tunnel4.c
udp_diag.c udp: no longer use SLAB_DESTROY_BY_RCU 2016-04-04 22:11:19 -04:00
udp_impl.h net: Remove iocb argument from sendmsg and recvmsg 2015-03-02 13:06:31 -05:00
udp_offload.c GSO: Support partial segmentation offload 2016-04-14 16:23:41 -04:00
udp_tunnel.c udp: Add socket based GRO and config 2016-04-07 16:53:29 -04:00
udp.c udp: prepare for non BH masking at backlog processing 2016-05-02 17:02:25 -04:00
udplite.c net: Eliminate no_check from protosw 2014-05-23 16:28:53 -04:00
xfrm4_input.c netfilter: Pass net into okfn 2015-09-17 17:18:37 -07:00
xfrm4_mode_beet.c ipv4: ERROR: code indent should use tabs where possible 2013-12-26 13:43:21 -05:00
xfrm4_mode_transport.c
xfrm4_mode_tunnel.c ipv4: hash net ptr into fragmentation bucket selection 2015-03-25 14:07:04 -04:00
xfrm4_output.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2015-10-24 06:54:12 -07:00
xfrm4_policy.c Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec 2015-12-22 16:26:31 -05:00
xfrm4_protocol.c xfrm4: Remove duplicate semicolon 2014-06-30 07:49:47 +02:00
xfrm4_state.c
xfrm4_tunnel.c sit: add IPv4 over IPv4 support 2013-05-31 17:19:05 -07:00