d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
91a790ee6e
freebsd-skq/sys/netinet
History
lstewart 91a790ee6e Retire the system-wide, per-reassembly queue segment limit. The mechanism is far 
too coarse grained to be useful and the default value significantly degrades TCP
performance on moderate to high bandwidth-delay product paths with non-zero loss
(e.g. 5+Mbps connections across the public Internet often suffer).

Replace the outgoing mechanism with an individual per-queue limit based on the
number of MSS segments that fit into the socket's receive buffer. This should
strike a good balance between performance and the potential for resource
exhaustion when FreeBSD is acting as a TCP receiver. With socket buffer
autotuning (which is enabled by default), the reassembly queue tracks the
socket buffer and benefits too.

As the XXX comment suggests, my testing uncovered some unexpected behaviour
which requires further investigation. By using so->so_rcv.sb_hiwat
instead of sbspace(&so->so_rcv), we allow more segments to be held across both
the socket receive buffer and reassembly queue than we probably should. The
tradeoff is better performance in at least one common scenario, versus a devious
sender's ability to consume more resources on a FreeBSD receiver.

Sponsored by:	FreeBSD Foundation
Reviewed by:	andre, gnn, rpaulo
MFC after:	2 weeks
2010-10-16 07:12:39 +00:00
..
ipfw put back the assigment to sched_time. It was correct, and 2010-10-01 15:38:35 +00:00
libalias Don't use struct timezone. 2010-08-08 02:51:32 +00:00
accf_data.c Rework socket upcalls to close some races with setup/teardown of upcalls. 2009-06-01 21:17:03 +00:00
accf_dns.c Rework socket upcalls to close some races with setup/teardown of upcalls. 2009-06-01 21:17:03 +00:00
accf_http.c Rework socket upcalls to close some races with setup/teardown of upcalls. 2009-06-01 21:17:03 +00:00
icmp6.h MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
icmp_var.h Many network stack subsystems use a single global data structure to hold 2009-08-02 19:43:32 +00:00
if_atm.c
if_atm.h
if_ether.c Unbreak LINT by moving all carp hooks to net/if.c / netinet/ip_carp.h, with 2010-08-11 20:18:19 +00:00
if_ether.h Add arp_update_event. This replaces route_arp_update_event, which 2009-09-08 21:17:17 +00:00
igmp_var.h Update stats in struct igmpstat using two new macros: 2009-04-12 13:41:13 +00:00
igmp.c Proper bracketing. 2010-10-01 11:48:14 +00:00
igmp.h These are no longer referenced in the tree, so can be safely removed. 2009-06-10 18:12:15 +00:00
in_cksum.c
in_gif.c MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
in_gif.h
in_mcast.c Fix a few issues related to the legacy 4.4 BSD multicast APIs. 2010-04-10 12:05:31 +00:00
in_pcb.c This patch fixes the problem where proxy ARP entries cannot be added 2010-05-25 20:42:35 +00:00
in_pcb.h Abstract out initialization of most aspects of struct inpcbinfo from 2010-03-14 18:59:11 +00:00
in_proto.c Allow carp(4) to be loaded as a kernel module. Follow precedent set by 2010-08-11 00:51:50 +00:00
in_rmx.c Add a bandaid for a long-standing race condition during route entry 2010-09-27 19:26:56 +00:00
in_systm.h
in_var.h Remove unused VNET_SET() and related macros; only VNET_GET() is 2009-07-16 21:13:04 +00:00
in.c In case of RADIX_MPATH do not leak the IN_IFADDR read lock on 2010-09-04 16:06:01 +00:00
in.h Make the RPC specific __rpc_inet_ntop() and __rpc_inet_pton() general 2010-09-24 15:01:45 +00:00
ip6.h
ip_carp.c Do not convert some meaningful error value to EINVAL. 2010-09-20 12:23:10 +00:00
ip_carp.h Unbreak LINT by moving all carp hooks to net/if.c / netinet/ip_carp.h, with 2010-08-11 20:18:19 +00:00
ip_divert.c Ensure a minimum "slop" of 10 extra pcb structures when providing a 2010-08-17 16:41:16 +00:00
ip_divert.h Various cleanup done in ipfw3-head branch including: 2010-01-04 19:01:22 +00:00
ip_dummynet.h whitespace fixes (trailing whitespace, bad indentation 2010-04-19 16:17:30 +00:00
ip_ecn.c
ip_ecn.h
ip_encap.c (S)LIST_HEAD_INITIALIZER takes a (S)LIST_HEAD as an argument. 2009-12-28 22:56:30 +00:00
ip_encap.h
ip_fastfwd.c Virtualize the pfil hooks so that different jails may chose different 2009-10-11 05:59:43 +00:00
ip_fw.h + implement (two lines) the kernel side of 'lookup dscp N' to use the 2010-03-15 17:14:27 +00:00
ip_gre.c The NetBSD Foundation has granted permission to remove clause 3 and 4 from 2010-03-01 17:05:46 +00:00
ip_gre.h The NetBSD Foundation has granted permission to remove clause 3 and 4 from 2010-03-01 17:05:46 +00:00
ip_icmp.c Change the messages of the ICMP bad port bandwidth limiter from 2010-08-14 21:04:27 +00:00
ip_icmp.h MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
ip_id.c
ip_input.c MFp4 CH=183052 183053 183258: 2010-09-02 17:43:44 +00:00
ip_ipsec.c The proper fix for the delayed SCTP checksum is to 2010-03-12 22:58:52 +00:00
ip_ipsec.h Remove ifdefed out part of code, which seems to have originated a decade ago 2009-11-09 19:53:34 +00:00
ip_mroute.c Virtualize the IPv4 multicast routing code. 2010-06-02 15:44:43 +00:00
ip_mroute.h Virtualize the IPv4 multicast routing code. 2010-06-02 15:44:43 +00:00
ip_options.c Use ifa_ifwithaddr_check() rather than ifa_ifwithaddr() as we are not 2010-10-14 12:32:49 +00:00
ip_options.h
ip_output.c IP_BINDANY is not correctly handled in getsockopt() case. 2010-09-24 14:38:54 +00:00
ip_var.h MFp4 CH=183052 183053 183258: 2010-09-02 17:43:44 +00:00
ip.h use u_char instead of u_int for short bitfields. 2010-02-01 14:13:44 +00:00
pim_var.h Virtualize the IPv4 multicast routing code. 2010-06-02 15:44:43 +00:00
pim.h
raw_ip.c Adding an address on an interface also requires the loopback route to 2010-09-12 18:04:47 +00:00
sctp_asconf.c Delay the assignment of a path for DATA chunk until they hit 2010-09-15 23:10:45 +00:00
sctp_asconf.h Delay the assignment of a path for DATA chunk until they hit 2010-09-15 23:10:45 +00:00
sctp_auth.c Fix a locking issue which shows up when the code is used 2010-09-19 11:42:16 +00:00
sctp_auth.h Use [] instead of [0] for flexible arrays. 2010-01-22 07:53:41 +00:00
sctp_bsd_addr.c Remove unused variable/assignment. 2010-09-15 21:19:54 +00:00
sctp_bsd_addr.h This adds back the Iterator to the sctp 2010-05-16 17:03:56 +00:00
sctp_cc_functions.c Make the initial congestion window configurable via sysctl. 2010-09-17 18:53:07 +00:00
sctp_cc_functions.h
sctp_constants.h * Implement initial version of send buffer splitting. 2010-09-17 16:20:29 +00:00
sctp_crc32.c Fix the the SCTP_WITH_NO_CSUM option when used in combination with 2010-08-29 18:50:30 +00:00
sctp_crc32.h Fix the the SCTP_WITH_NO_CSUM option when used in combination with 2010-08-29 18:50:30 +00:00
sctp_header.h Add missing SCTP_PACKED. Spotted by Irene Ruengeler. 2010-02-13 21:38:15 +00:00
sctp_indata.c Fix a locking issue which resulted in aborted associations 2010-09-20 12:19:11 +00:00
sctp_indata.h * Fix some race condition in SACK/NR-SACK processing. 2010-04-03 15:40:14 +00:00
sctp_input.c Delay the assignment of a path for DATA chunk until they hit 2010-09-15 23:10:45 +00:00
sctp_input.h
sctp_lock_bsd.h 1) Optimize the cleanup and don't always depend on 2010-06-06 20:34:17 +00:00
sctp_os_bsd.h Fix the the SCTP_WITH_NO_CSUM option when used in combination with 2010-08-29 18:50:30 +00:00
sctp_os.h
sctp_output.c Fix a bug where the wrong PR-SCTP policy was considered. 2010-09-17 19:20:39 +00:00
sctp_output.h Delay the assignment of a path for DATA chunk until they hit 2010-09-15 23:10:45 +00:00
sctp_pcb.c Fix a locking issue which shows up when the code is used 2010-09-19 11:42:16 +00:00
sctp_pcb.h Fix the switching on/off of CMT using sysctl and socket option. 2010-08-28 17:59:51 +00:00
sctp_peeloff.c Fix the switching on/off of CMT using sysctl and socket option. 2010-08-28 17:59:51 +00:00
sctp_peeloff.h
sctp_structs.h PR SCTP Bugs. Basically a full sized frame of 2010-07-29 11:37:04 +00:00
sctp_sysctl.c Make the initial congestion window configurable via sysctl. 2010-09-17 18:53:07 +00:00
sctp_sysctl.h Allow the initial congestion window to be configure 2010-09-19 11:57:21 +00:00
sctp_timer.c Fix a bug where the wrong PR-SCTP policy was considered. 2010-09-17 19:20:39 +00:00
sctp_timer.h
sctp_uio.h Implement correct handling of address parameter and 2010-09-05 20:13:07 +00:00
sctp_usrreq.c Remove old debug code. 2010-09-15 23:56:25 +00:00
sctp_var.h * Fix some race condition in SACK/NR-SACK processing. 2010-04-03 15:40:14 +00:00
sctp.h Fix the switching on/off of CMT using sysctl and socket option. 2010-08-28 17:59:51 +00:00
sctputil.c Remove old debug code. 2010-09-15 23:56:25 +00:00
sctputil.h * Fix some race condition in SACK/NR-SACK processing. 2010-04-03 15:40:14 +00:00
siftr.c Log the number of segments currently in the reassembly queue. 2010-09-25 09:16:46 +00:00
tcp_debug.c Remove the "The option TCPDEBUG requires option INET." requirement. 2009-06-10 10:39:41 +00:00
tcp_debug.h
tcp_fsm.h
tcp_hostcache.c MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
tcp_hostcache.h
tcp_input.c Remove the TCP inflight bandwidth limiter as announced in r211315 2010-09-16 21:06:45 +00:00
tcp_lro.c
tcp_lro.h
tcp_offload.c Merge the remainder of kern_vimage.c and vimage.h into vnet.c and 2009-08-01 19:26:27 +00:00
tcp_offload.h Replace struct tcpopt with a proxy toeopt struct in the TOE driver interface to 2009-07-13 11:51:02 +00:00
tcp_output.c Rearrange the TSO code to make it more readable and to clearly 2010-09-17 22:05:27 +00:00
tcp_reass.c Retire the system-wide, per-reassembly queue segment limit. The mechanism is far 2010-10-16 07:12:39 +00:00
tcp_sack.c MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
tcp_seq.h
tcp_subr.c Internalise reassembly queue related functionality and variables which should 2010-09-25 04:58:46 +00:00
tcp_syncache.c Initializing the new error variable to zero in syncache_socket() 2010-08-15 13:07:08 +00:00
tcp_syncache.h Replace struct tcpopt with a proxy toeopt struct in the TOE driver interface to 2009-07-13 11:51:02 +00:00
tcp_timer.c - spread tcp timer callout load evenly across cpus if net.inet.tcp.per_cpu_timers is set to 1 2010-03-20 19:47:30 +00:00
tcp_timer.h Remove the TCP inflight bandwidth limiter as announced in r211315 2010-09-16 21:06:45 +00:00
tcp_timewait.c MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
tcp_usrreq.c Remove the TCP inflight bandwidth limiter as announced in r211315 2010-09-16 21:06:45 +00:00
tcp_var.h Internalise reassembly queue related functionality and variables which should 2010-09-25 04:58:46 +00:00
tcp.h Remove the TCP inflight bandwidth limiter as announced in r211315 2010-09-16 21:06:45 +00:00
tcpip.h
toedev.h
udp_usrreq.c Ensure a minimum "slop" of 10 extra pcb structures when providing a 2010-08-17 16:41:16 +00:00
udp_var.h MFP4: @176978-176982, 176984, 176990-176994, 177441 2010-04-29 11:52:42 +00:00
udp.h Added support for NAT-Traversal (RFC 3948) in IPsec stack. 2009-06-12 15:44:35 +00:00