svn commit: r342181 - stable/12/sys/netinet/cc

Author: hiren Date: Mon Dec 17 21:46:42 2018 New Revision: 342181 URL: https://svnweb.freebsd.org/changeset/base/342181 Log: MFC r342127 Revert r331567 CC Cubic: fix underflow for cubic_cwnd() This change is causing TCP connections using cubic to hang. Need to dig more to find exact

svn commit: r342127 - head/sys/netinet/cc

Author: hiren Date: Sat Dec 15 17:01:16 2018 New Revision: 342127 URL: https://svnweb.freebsd.org/changeset/base/342127 Log: Revert r331567 CC Cubic: fix underflow for cubic_cwnd() This change is causing TCP connections using cubic to hang. Need to dig more to find exact cause and fix

Re: svn commit: r334804 - in head/sys: kern modules/tcp modules/tcp/rack netinet netinet/tcp_stacks sys

On 06/07/18 at 08:07P, Matthew Macy wrote: > > > > Okay. I believe there might be situations where we may want to still > > keep the 'default' stack alive. I know Windows doesn't yet use RACK when > > rtt is lesser than 10ms (or something like that), as an example. > > > > Is there any reason

Re: svn commit: r334804 - in head/sys: kern modules/tcp modules/tcp/rack netinet netinet/tcp_stacks sys

On 06/07/18 at 08:58P, Randall Stewart wrote: > > > > On Jun 7, 2018, at 6:01 PM, hiren panchasara > > wrote: > > > > On 06/07/18 at 06:18P, Randall Stewart wrote: > >> Author: rrs > >> Date: Thu Jun 7 18:18:13 2018 > >> New Revisi

Re: svn commit: r334804 - in head/sys: kern modules/tcp modules/tcp/rack netinet netinet/tcp_stacks sys

On 06/07/18 at 06:18P, Randall Stewart wrote: > Author: rrs > Date: Thu Jun 7 18:18:13 2018 > New Revision: 334804 > URL: https://svnweb.freebsd.org/changeset/base/334804 > > Log: > This commit brings in a new refactored TCP stack called Rack. > Rack includes the following features: >- A

Re: svn commit: r332770 - in head/sys: conf netinet netinet/tcp_stacks sys

On 05/01/18 at 05:11P, Warner Losh wrote: > On Tue, May 1, 2018 at 5:00 PM, Jonathan Looney <jonloo...@gmail.com> wrote: > > > On Mon, Apr 30, 2018 at 3:16 AM, hiren panchasara < > > hi...@strugglingcoder.info> wrote: > > > > > > In my u

Re: svn commit: r332770 - in head/sys: conf netinet netinet/tcp_stacks sys

On 04/19/18 at 01:38P, Randall Stewart wrote: > Author: rrs > Date: Thu Apr 19 13:37:59 2018 > New Revision: 332770 > URL: https://svnweb.freebsd.org/changeset/base/332770 > > Log: > This commit brings in the TCP high precision timer system (tcp_hpts). > It is the forerunner/foundational work

Re: svn commit: r327559 - in head: . sys/net

On 01/04/18 at 11:37P, Steven Hartland wrote: > > > On 04/01/2018 22:42, hiren panchasara wrote: > > On 01/04/18 at 09:52P, Steven Hartland wrote: > >> On 04/01/2018 20:50, Eugene Grosbein wrote: > >>> 05.01.2018 3:05, Steven Hartland wrote: > >>>

Re: svn commit: r327559 - in head: . sys/net

On 01/04/18 at 09:52P, Steven Hartland wrote: > On 04/01/2018 20:50, Eugene Grosbein wrote: > > 05.01.2018 3:05, Steven Hartland wrote: > > > >> Author: smh > >> Date: Thu Jan 4 20:05:47 2018 > >> New Revision: 327559 > >> URL: https://svnweb.freebsd.org/changeset/base/327559 > >> > >> Log: > >>

Re: svn commit: r324836 - in head/sys: kern sys

On 10/21/17 at 10:40P, Mateusz Guzik wrote: > Author: mjg > Date: Sat Oct 21 22:40:09 2017 > New Revision: 324836 > URL: https://svnweb.freebsd.org/changeset/base/324836 > > Log: > mtx: implement thread lock fastpath Hi, Can you please elaborate this commit-log message a little more? what is

Re: svn commit: r316699 - head/sys/net

On 04/11/17 at 08:56P, Andrey V. Elsukov wrote: > Author: ae > Date: Tue Apr 11 08:56:18 2017 > New Revision: 316699 > URL: https://svnweb.freebsd.org/changeset/base/316699 > > Log: > Do not adjust interface MTU automatically. Leave this task to the system > administrator. > > This

Re: svn commit: r316309 - head/sys/dev/qlxgbe

On 03/30/17 at 11:07P, Somayajulu, David wrote: > Hi Hiren, > > I know this is not a new topic but a little more descriptive commit-log > > would have been nicer. Also, you should update the manpage reflecting this > > change. i.e. now it also supports software LRO when h/w LRO is disabled. >

Re: svn commit: r316309 - head/sys/dev/qlxgbe

On 03/30/17 at 10:43P, David C Somayajulu wrote: > Author: davidcs > Date: Thu Mar 30 22:43:32 2017 > New Revision: 316309 > URL: https://svnweb.freebsd.org/changeset/base/316309 > > Log: > Add support for optional Soft LRO Hi David, I know this is not a new topic but a little more

Re: svn commit: r314813 - head/sys/kern

On 03/06/17 at 07:08P, Ngie Cooper (yaneurabeya) wrote: > > > On Mar 6, 2017, at 16:20, Hiren Panchasara <hi...@freebsd.org> wrote: > > > > Author: hiren > > Date: Tue Mar 7 00:20:01 2017 > > New Revision: 314813 > > URL: https://svnweb.freebsd.org/c

svn commit: r314838 - head/sys/kern

Author: hiren Date: Tue Mar 7 06:46:38 2017 New Revision: 314838 URL: https://svnweb.freebsd.org/changeset/base/314838 Log: Fix the KASSERT check from r314813. len being 0 is valid. Submitted by: ngie Reported by: ngie (via jenkins test run) Sponsored by: Limelight Networks

svn commit: r314813 - head/sys/kern

Author: hiren Date: Tue Mar 7 00:20:01 2017 New Revision: 314813 URL: https://svnweb.freebsd.org/changeset/base/314813 Log: We've found a recurring problem where some userland process would be stuck spinning at 100% cpu around sbcut_internal(). Inside sbflush_internal(), sb_ccc reached to

Re: svn commit: r314216 - head/sys/x86/x86

On 02/24/17 at 06:56P, Jonathan T. Looney wrote: > Author: jtl > Date: Fri Feb 24 18:56:00 2017 > New Revision: 314216 > URL: https://svnweb.freebsd.org/changeset/base/314216 > > Log: > We have seen several cases recently where we appear to get a double-fault: > We have an original panic.

Re: svn commit: r278729 - head/sys/sys

On 03/19/15 at 11:08P, hiren panchasara wrote: > On 03/16/15 at 06:06P, hiren panchasara wrote: > > On 03/16/15 at 03:39P, Gleb Smirnoff wrote: > > > On Wed, Mar 11, 2015 at 02:36:07PM -0700, hiren panchasara wrote: > > > h> On 02/13/15 at 11:19P, Simon J. Gerraty

Re: svn commit: r313043 - head/sys/kern

On 02/01/17 at 01:12P, Hartmut Brandt wrote: > Author: harti > Date: Wed Feb 1 13:12:07 2017 > New Revision: 313043 > URL: https://svnweb.freebsd.org/changeset/base/313043 > > Log: > Merge filt_soread and filt_solisten and decide what to do when checking > for EVFILT_READ at the point of the

Re: svn commit: r312905 - head/sys/net

On 01/27/17 at 11:19P, Cy Schubert wrote: > Really? Just a copyright notice? > > The reason I ask is because of this: > > ===> ae (all) > --- iflib.o --- > /export/home/cy/freebsd/svn/ip6-cksum/sys/net/iflib.c:1561:1: error: > function definition is not allowed here > { > Should these 2 lines

svn commit: r312907 - head/sys/netinet

Author: hiren Date: Fri Jan 27 23:10:46 2017 New Revision: 312907 URL: https://svnweb.freebsd.org/changeset/base/312907 Log: Add a knob to change default behavior of inheriting listen socket's tcp stack regardless of what the default stack for the system is set to. With current/default

Re: svn commit: r286700 - in head: sbin/ifconfig sys/net

e of fast (it was dead before > this commit, too). > > CID: 1305734 > CID: 1305692 > > -Alan > > On Wed, Aug 12, 2015 at 2:21 PM, Hiren Panchasara <hi...@freebsd.org> wrote: > > Author: hiren > > Date: Wed Aug 12 20:21:04 2015 > > New Revision:

Re: svn commit: r312277 - in head/sys: kern sys

On 01/16/17 at 12:39P, Sergey Kandaurov wrote: > On 16 January 2017 at 11:25, Hiren Panchasara <hi...@freebsd.org> wrote: > > > Author: hiren > > Date: Mon Jan 16 08:25:33 2017 > > New Revision: 312277 > > URL: https://svnweb.freebsd.org/changeset/base/

svn commit: r312277 - in head/sys: kern sys

Author: hiren Date: Mon Jan 16 08:25:33 2017 New Revision: 312277 URL: https://svnweb.freebsd.org/changeset/base/312277 Log: Add kevent EVFILT_EMPTY for notification when a client has received all data i.e. everything outstanding has been acked. Reviewed by: bz, gnn (previous version)

svn commit: r311955 - stable/11/sys/netinet

Author: hiren Date: Thu Jan 12 00:50:37 2017 New Revision: 311955 URL: https://svnweb.freebsd.org/changeset/base/311955 Log: MFC r311453 sysctl net.inet.tcp.hostcache.list in a jail can see connections from other jails and the host. This commit fixes it. PR: 200361

svn commit: r311453 - head/sys/netinet

Author: hiren Date: Thu Jan 5 17:22:09 2017 New Revision: 311453 URL: https://svnweb.freebsd.org/changeset/base/311453 Log: sysctl net.inet.tcp.hostcache.list in a jail can see connections from other jails and the host. This commit fixes it. PR: 200361 Submitted by: bz

svn commit: r311297 - stable/11/sys/netinet

Author: hiren Date: Wed Jan 4 18:03:27 2017 New Revision: 311297 URL: https://svnweb.freebsd.org/changeset/base/311297 Log: MFC r309858 We currently don't do TSO if ip options are present. In case of IPv6, we look at in6p_options to check that. That is incorrect as we carry ip options in

Re: svn commit: r310547 - head/sys/netinet

On 12/25/16 at 05:37P, Michael Tuexen wrote: > Author: tuexen > Date: Sun Dec 25 17:37:18 2016 > New Revision: 310547 > URL: https://svnweb.freebsd.org/changeset/base/310547 > > Log: > Remove a KASSERT which is not always true. > > In case of the empty queue tp->snd_holes and

Re: svn commit: r310433 - head/lib/libc/stdio

On 12/23/16 at 03:05P, Andrey Chernov wrote: > On 23.12.2016 1:39, Conrad Meyer wrote: > > This was unjustified and inappropriate. > > I don't think so. While being able to back it out or fix it by another > way by yourself, you just doing nothing for several days. Your personal > reasons of

Re: svn commit: r309745 - in head: share/man/man9 sys/kern sys/sys

On 12/22/16 at 10:33P, Gleb Smirnoff wrote: > Hi! > > On Wed, Dec 21, 2016 at 04:19:28PM -0800, hiren panchasara wrote: > h> On 12/09/16 at 05:58P, Gleb Smirnoff wrote: > h> > Author: glebius > h> > Date: Fri Dec 9 17:58:34 2016 > h> &

Re: svn commit: r309745 - in head: share/man/man9 sys/kern sys/sys

On 12/09/16 at 05:58P, Gleb Smirnoff wrote: > Author: glebius > Date: Fri Dec 9 17:58:34 2016 > New Revision: 309745 > URL: https://svnweb.freebsd.org/changeset/base/309745 > > Log: > Provide counter_ratecheck(), a MP-friendly substitution to ppsratecheck(). > When rated event happens at a

Re: svn commit: r307745 - head/sys/kern

On 12/10/16 at 09:35P, Oliver Pinter wrote: > On 10/21/16, Hiren Panchasara <hi...@freebsd.org> wrote: > > Author: hiren > > Date: Fri Oct 21 18:27:30 2016 > > New Revision: 307745 > > URL: https://svnweb.freebsd.org/changeset/base/307745 > > > > L

svn commit: r309859 - stable/11/sys/kern

Author: hiren Date: Sun Dec 11 23:36:11 2016 New Revision: 309859 URL: https://svnweb.freebsd.org/changeset/base/309859 Log: MFC r307745 In sendit(), if mp->msg_control is present, then in sockargs() we are allocating mbuf to store mp->msg_control. Later in kern_sendit(), call to

svn commit: r309858 - head/sys/netinet

Author: hiren Date: Sun Dec 11 23:14:47 2016 New Revision: 309858 URL: https://svnweb.freebsd.org/changeset/base/309858 Log: We currently don't do TSO if ip options are present. In case of IPv6, we look at in6p_options to check that. That is incorrect as we carry ip options in

svn commit: r309548 - stable/11/sys/netinet

Author: hiren Date: Mon Dec 5 06:20:27 2016 New Revision: 309548 URL: https://svnweb.freebsd.org/changeset/base/309548 Log: MFC r308943 For RTT calculations mid-session, we explicitly ignore ACKs with tsecr of 0 as many borken middle-boxes tend to do that. But during 3whs, in

Re: svn commit: r309351 - head/etc/defaults

On 12/01/16 at 08:48P, hiren panchasara wrote: > On 12/01/16 at 04:35P, Warner Losh wrote: > > Author: imp > > Date: Thu Dec 1 04:35:43 2016 > > New Revision: 309351 > > URL: https://svnweb.freebsd.org/changeset/base/309351 > > > > Log: > > Rever

Re: svn commit: r309351 - head/etc/defaults

On 12/01/16 at 04:35P, Warner Losh wrote: > Author: imp > Date: Thu Dec 1 04:35:43 2016 > New Revision: 309351 > URL: https://svnweb.freebsd.org/changeset/base/309351 > > Log: > Revert the 'performance' setting to 'NONE' from C2. Just a note that this commit reverts r282110. Cheers, Hiren

Re: svn commit: r308943 - head/sys/netinet

Forgot to tag for MFC. On 11/21/16 at 08:53P, Hiren Panchasara wrote: > Author: hiren > Date: Mon Nov 21 20:53:11 2016 > New Revision: 308943 > URL: https://svnweb.freebsd.org/changeset/base/308943 > > Log: > For RTT calculations mid-session, we explicitly ignore

svn commit: r308944 - head/sys/modules/bytgpio

Author: hiren Date: Mon Nov 21 21:07:43 2016 New Revision: 308944 URL: https://svnweb.freebsd.org/changeset/base/308944 Log: r308942 broke kernel build. Add acpi_if.h to module makefile to fix it. Submitted by: peter Modified: head/sys/modules/bytgpio/Makefile Modified:

svn commit: r308943 - head/sys/netinet

Author: hiren Date: Mon Nov 21 20:53:11 2016 New Revision: 308943 URL: https://svnweb.freebsd.org/changeset/base/308943 Log: For RTT calculations mid-session, we explicitly ignore ACKs with tsecr of 0 as many borken middle-boxes tend to do that. But during 3whs, in syncache_expand(), we

svn commit: r308701 - in stable/11: cddl/lib/libdtrace tests/sys/netinet

Author: hiren Date: Tue Nov 15 22:18:52 2016 New Revision: 308701 URL: https://svnweb.freebsd.org/changeset/base/308701 Log: MFC r302474 (By gnn) On FreeBSD there is a setsockopt option SO_USER_COOKIE which allows setting a 32 bit value on each socket. This can be used by applications

svn commit: r308184 - stable/11/sys/netinet

Author: hiren Date: Tue Nov 1 22:40:25 2016 New Revision: 308184 URL: https://svnweb.freebsd.org/changeset/base/308184 Log: MFC r307545 Make sure tcp_mss() has the same check as tcp_mss_update() to have t_maxseg set to at least 64. This is still just a coverup to avoid kernel panic

svn commit: r308180 - in head/sys/netinet: . cc

Author: hiren Date: Tue Nov 1 21:08:37 2016 New Revision: 308180 URL: https://svnweb.freebsd.org/changeset/base/308180 Log: Set slow start threshold more accurately on loss to be flightsize/2 instead of cwnd/2 as recommended by RFC5681. (spotted by mmacy at nextbsd dot org) Restore

svn commit: r307901 - in head/sys/netinet: . cc tcp_stacks

Author: hiren Date: Tue Oct 25 05:45:47 2016 New Revision: 307901 URL: https://svnweb.freebsd.org/changeset/base/307901 Log: FreeBSD tcp stack used to inform respective congestion control module about the loss event but not use or obay the recommendations i.e. values set by it in some

svn commit: r307900 - in head/sys/netinet: . cc

Author: hiren Date: Tue Oct 25 05:07:51 2016 New Revision: 307900 URL: https://svnweb.freebsd.org/changeset/base/307900 Log: Undo r307899. It needs a bit more work and proper commit log. Modified: head/sys/netinet/cc/cc_cdg.c head/sys/netinet/cc/cc_chd.c head/sys/netinet/cc/cc_cubic.c

Re: svn commit: r307899 - in head/sys/netinet: . cc

Sigh. I'll revert this and do it right. On 10/25/16 at 05:03P, Hiren Panchasara wrote: > Author: hiren > Date: Tue Oct 25 05:03:33 2016 > New Revision: 307899 > URL: https://svnweb.freebsd.org/changeset/base/307899 > > Log: > In Collaboration with: Ma

svn commit: r307899 - in head/sys/netinet: . cc

Author: hiren Date: Tue Oct 25 05:03:33 2016 New Revision: 307899 URL: https://svnweb.freebsd.org/changeset/base/307899 Log: In Collaboration with:Matt Macy Reviewed by: jtl Sponsored by: Limelight Networks Differential Revision:

svn commit: r307745 - head/sys/kern

Author: hiren Date: Fri Oct 21 18:27:30 2016 New Revision: 307745 URL: https://svnweb.freebsd.org/changeset/base/307745 Log: Rework r306337. In sendit(), if mp->msg_control is present, then in sockargs() we are allocating mbuf to store mp->msg_control. Later in kern_sendit(), call to

Re: svn commit: r307727 - head/share/man/man4

On 10/21/16 at 11:29P, Michael Tuexen wrote: > Author: tuexen > Date: Fri Oct 21 11:29:25 2016 > New Revision: 307727 > URL: https://svnweb.freebsd.org/changeset/base/307727 > > Log: > Document the TCP sysctl variables insecure_rst and insecure_syn. > > MFC after: 1 month > Sponsored by:

svn commit: r307545 - head/sys/netinet

Author: hiren Date: Tue Oct 18 02:40:25 2016 New Revision: 307545 URL: https://svnweb.freebsd.org/changeset/base/307545 Log: Make sure tcp_mss() has the same check as tcp_mss_update() to have t_maxseg set to at least 64. This is still just a coverup to avoid kernel panic and not an

svn commit: r307242 - stable/11/sys/netinet

Author: hiren Date: Fri Oct 14 01:38:29 2016 New Revision: 307242 URL: https://svnweb.freebsd.org/changeset/base/307242 Log: MFC r306464 This adds a sysctl which allows you to disable the TCP hostcache. This is handy during testing of network related changes where cached entries may

svn commit: r306517 - stable/11/sys/netinet

Author: hiren Date: Fri Sep 30 18:14:05 2016 New Revision: 306517 URL: https://svnweb.freebsd.org/changeset/base/306517 Log: MFC r304855 Update TCPS_HAVERCVDFIN() macro to correctly include all states a connection can be in after receiving a FIN. FWIW, NetBSD has this change for

svn commit: r306464 - head/sys/netinet

Author: hiren Date: Fri Sep 30 00:10:57 2016 New Revision: 306464 URL: https://svnweb.freebsd.org/changeset/base/306464 Log: This adds a sysctl which allows you to disable the TCP hostcache. This is handy during testing of network related changes where cached entries may pollute your

Re: svn commit: r306337 - head/sys/kern

On 09/26/16 at 10:18P, Bruce Evans wrote: > On Mon, 26 Sep 2016, Bruce Evans wrote: > Thanks Bruce for your inputs. I've reverted this change with r306348 for now as there was a panic reported with this. I do not have time to deal with it for at least a few days. I'll try to get back to it. If

svn commit: r306348 - head/sys/kern

Author: hiren Date: Mon Sep 26 15:45:30 2016 New Revision: 306348 URL: https://svnweb.freebsd.org/changeset/base/306348 Log: Revert r306337. dhw@ reproted a panic which seems related to this and bde@ has raised some issues. Modified: head/sys/kern/uipc_syscalls.c Modified:

svn commit: r306337 - head/sys/kern

Author: hiren Date: Mon Sep 26 10:13:58 2016 New Revision: 306337 URL: https://svnweb.freebsd.org/changeset/base/306337 Log: In sendit(), if mp->msg_control is present, then in sockargs() we are allocating mbuf to store mp->msg_control. Later in kern_sendit(), call to getsock_cap(), will

Re: svn commit: r306284 - head/sys/dev/bxe

On 09/23/16 at 11:18P, David C Somayajulu wrote: > Author: davidcs > Date: Fri Sep 23 23:18:54 2016 > New Revision: 306284 > URL: https://svnweb.freebsd.org/changeset/base/306284 > > Log: > Fixes for issues under high workloads David, It is really useful to have details about said issues in

svn commit: r305853 - in stable/10: share/man/man4 sys/kern

Author: hiren Date: Fri Sep 16 00:14:26 2016 New Revision: 305853 URL: https://svnweb.freebsd.org/changeset/base/305853 Log: MFC r301522 (by bz) Implement a `show panic` command to DDB which will helpfully print the panic string again if set, in case it scrolled out of the active

Re: svn commit: r304895 - head/usr.bin/netstat

On 08/29/16 at 05:14P, Bruce Evans wrote: > On Sun, 28 Aug 2016, hiren panchasara wrote: > > > On 08/27/16 at 11:06P, Bruce Evans wrote: > >> > >> Log: > >> Fix build without INET6 and with gcc. A function definition was ifdefed > >> for INE

svn commit: r305008 - in stable/11/sys: netinet netinet6

Author: hiren Date: Mon Aug 29 18:00:14 2016 New Revision: 305008 URL: https://svnweb.freebsd.org/changeset/base/305008 Log: MFC r303626 (by gallatin) Rework IPV6 TCP path MTU discovery to match IPv4. No objection: gallatin Modified: stable/11/sys/netinet/tcp_subr.c

Re: svn commit: r304895 - head/usr.bin/netstat

On 08/27/16 at 11:06P, Bruce Evans wrote: > Author: bde > Date: Sat Aug 27 11:06:06 2016 > New Revision: 304895 > URL: https://svnweb.freebsd.org/changeset/base/304895 > > Log: > Fix build without INET6 and with gcc. A function definition was ifdefed > for INET6, but its protototype was not,

Re: svn commit: r304857 - head/sys/netinet/tcp_stacks

On 08/27/16 at 01:47P, Lawrence Stewart wrote: > Pointy hat to: lstewart@ > > Apologies all for the breakage, thanks Hiren for fixing and apologies > also for missing your email. For some reason your reply to my commit did > not make it to my inbox and was filtered straight to my mailing list >

svn commit: r304857 - head/sys/netinet/tcp_stacks

Author: hiren Date: Fri Aug 26 19:23:17 2016 New Revision: 304857 URL: https://svnweb.freebsd.org/changeset/base/304857 Log: Adjust TCP module fastpath after r304803's cc_ack_received() changes. Reported by: hiren, bz, np Reviewed by: rrs Sponsored by:

svn commit: r304855 - head/sys/netinet

Author: hiren Date: Fri Aug 26 17:48:54 2016 New Revision: 304855 URL: https://svnweb.freebsd.org/changeset/base/304855 Log: Update TCPS_HAVERCVDFIN() macro to correctly include all states a connection can be in after receiving a FIN. FWIW, NetBSD has this change for quite some time.

Re: svn commit: r304835 - stable/11/sys/netinet

On 08/26/16 at 05:37P, Sepherosa Ziehau wrote: > Author: sephe > Date: Fri Aug 26 05:37:44 2016 > New Revision: 304835 > URL: https://svnweb.freebsd.org/changeset/base/304835 > > Log: > MFC 303766 Thanks! Cheers, Hiren pgpTQwTTkL1ou.pgp Description: PGP signature

Re: svn commit: r303656 - head/sys/netinet

On 08/26/16 at 09:23P, Sepherosa Ziehau wrote: > On Fri, Aug 26, 2016 at 1:54 AM, hiren panchasara > <hi...@strugglingcoder.info> wrote: > > On 08/02/16 at 06:36P, Sepherosa Ziehau wrote: > >> Author: sephe > >> Date: Tue Aug 2 06:36:47 2016 >

Re: svn commit: r304803 - in head/sys: netinet netinet/cc sys

On 08/25/16 at 01:33P, Lawrence Stewart wrote: > Author: lstewart > Date: Thu Aug 25 13:33:32 2016 > New Revision: 304803 > URL: https://svnweb.freebsd.org/changeset/base/304803 > > Log: > Pass the number of segments coalesced by LRO up the stack by repurposing the > tso_segsz pkthdr field

Re: svn commit: r304223 - in head: share/man/man4 share/man/man9 sys/netinet

On 08/16/16 at 03:11P, Randall Stewart wrote: > Author: rrs > Date: Tue Aug 16 15:11:46 2016 > New Revision: 304223 > URL: https://svnweb.freebsd.org/changeset/base/304223 > > Log: > Here we update the modular tcp to be able to switch to an > alternate TCP stack in other then the closed

Re: svn commit: r303626 - in head/sys: netinet netinet6

On 08/01/16 at 05:02P, Andrew Gallatin wrote: > Author: gallatin > Date: Mon Aug 1 17:02:21 2016 > New Revision: 303626 > URL: https://svnweb.freebsd.org/changeset/base/303626 > > Log: > Rework IPV6 TCP path MTU discovery to match IPv4 > > - Re-write tcp_ctlinput6() to closely mimic the

Re: svn commit: r303656 - head/sys/netinet

On 08/02/16 at 06:36P, Sepherosa Ziehau wrote: > Author: sephe > Date: Tue Aug 2 06:36:47 2016 > New Revision: 303656 > URL: https://svnweb.freebsd.org/changeset/base/303656 > > Log: > tcp/lro: Implement hash table for LRO entries. > > This significantly improves HTTP workload performance

Re: svn commit: r303766 - head/sys/netinet

On 08/05/16 at 09:08P, Sepherosa Ziehau wrote: > Author: sephe > Date: Fri Aug 5 09:08:00 2016 > New Revision: 303766 > URL: https://svnweb.freebsd.org/changeset/base/303766 > > Log: > tcp/lro: If timestamps mismatch or it's a FIN, force flush. > > This keeps the segments/ACK/FIN delivery

Re: svn commit: r302081 - head/sys/netinet6

On 06/23/16 at 11:48P, Andrey V. Elsukov wrote: > On 22.06.16 18:46, hiren panchasara wrote: > >> Fix the NULL pointer dereference for unresolved link layer entries in > >> the netinet6 code. Copy link layer address only when corresponding entry > >> has LLE

Re: svn commit: r302081 - head/sys/netinet6

On 06/22/16 at 11:29P, Andrey V. Elsukov wrote: > Author: ae > Date: Wed Jun 22 11:29:21 2016 > New Revision: 302081 > URL: https://svnweb.freebsd.org/changeset/base/302081 > > Log: > Fix the NULL pointer dereference for unresolved link layer entries in > the netinet6 code. Copy link layer

Re: svn commit: r300865 - in head/sys: sys vm

On 05/27/16 at 07:15P, Alan Cox wrote: > Author: alc > Date: Fri May 27 19:15:45 2016 > New Revision: 300865 > URL: https://svnweb.freebsd.org/changeset/base/300865 > > Log: > The flag "vm_pages_needed" has long served two distinct purposes: (1) to > indicate that threads are waiting for free

svn commit: r299280 - head/sys/netinet/cc

Author: hiren Date: Mon May 9 19:19:03 2016 New Revision: 299280 URL: https://svnweb.freebsd.org/changeset/base/299280 Log: Add an option to use rfc6675 based pipe/inflight bytes calculation in htcp. Submitted by: Kevin Bowling MFC after:1 week Sponsored

Re: svn commit: r298769 - head/sys/netinet

On 05/03/16 at 09:29P, Sepherosa Ziehau wrote: > On Fri, Apr 29, 2016 at 11:55 PM, hiren panchasara > <hi...@strugglingcoder.info> wrote: > > On 04/29/16 at 07:23P, Sepherosa Ziehau wrote: > >> Author: sephe > >> Date: Fri Apr 29 07:23:08 2016 >

Re: svn commit: r298769 - head/sys/netinet

On 04/29/16 at 07:23P, Sepherosa Ziehau wrote: > Author: sephe > Date: Fri Apr 29 07:23:08 2016 > New Revision: 298769 > URL: https://svnweb.freebsd.org/changeset/base/298769 > > Log: > tcp/syncache: Set flowid and hash type properly for SYN|ACK > > So the underlying drivers can use it to

Re: svn commit: r298696 - head/sys/netinet

On 04/27/16 at 09:40P, Sepherosa Ziehau wrote: > Author: sephe > Date: Wed Apr 27 09:40:55 2016 > New Revision: 298696 > URL: https://svnweb.freebsd.org/changeset/base/298696 > > Log: > tcp/lro: Fix typo. > > MFC after: 1 week > Sponsored by: Microsoft OSTC > > Modified: >

Re: svn commit: r298131 - head/sys/conf

On 04/17/16 at 12:20P, Alexander Leidinger wrote: > > Quoting Luiz Otavio O Souza (from Sat, 16 Apr 2016 > 20:54:55 + (UTC)): > > > Author: loos > > Date: Sat Apr 16 20:54:55 2016 > > New Revision: 298131 > > URL: https://svnweb.freebsd.org/changeset/base/298131 > > > >

svn commit: r298087 - in head: cddl/lib/libdtrace share/man/man4

Author: hiren Date: Fri Apr 15 20:27:36 2016 New Revision: 298087 URL: https://svnweb.freebsd.org/changeset/base/298087 Log: Fix the 'type' for a few variables from tcpcb. Reviewed by: markj Sponsored by: Limelight Networks Differential Revision:

Re: svn commit: r294840 - head/sys/netinet

Somehow I totally missed this email. On 01/26/16 at 02:53P, Gleb Smirnoff wrote: > Hiren, > > On Tue, Jan 26, 2016 at 04:33:38PM +0000, Hiren Panchasara wrote: > H> Author: hiren > H> Date: Tue Jan 26 16:33:38 2016 > H> New Revision: 294840 > H> URL: https://

svn commit: r295015 - stable/10/sys/netinet

Author: hiren Date: Thu Jan 28 21:30:49 2016 New Revision: 295015 URL: https://svnweb.freebsd.org/changeset/base/295015 Log: MFC r294840 Persist timers TCPTV_PERSMIN and TCPTV_PERSMAX are hardcoded with 5 seconds and 60 seconds, respectively. Turn them into sysctls that can be tuned

svn commit: r294840 - head/sys/netinet

Author: hiren Date: Tue Jan 26 16:33:38 2016 New Revision: 294840 URL: https://svnweb.freebsd.org/changeset/base/294840 Log: Persist timers TCPTV_PERSMIN and TCPTV_PERSMAX are hardcoded with 5 seconds and 60 seconds, respectively. Turn them into sysctls that can be tuned live. The default

Re: svn commit: r294535 - in head/sys/netinet: . cc tcp_stacks

On 01/22/16 at 04:06P, Bjoern A. Zeeb wrote: > > > On 22 Jan 2016, at 15:21 , George Neville-Neil > > wrote: > > > > > > > > On 22 Jan 2016, at 2:13, Lawrence Stewart wrote: > > > >> Hi Gleb, > >> > >> On 01/22/16 09:34, Gleb Smirnoff wrote: > >>> Author: glebius >

Re: svn commit: r294540 - in head: etc/mtree include share/man/man4 share/man/man9 sys/netinet

On 01/21/16 at 06:39P, Gleb Smirnoff wrote: > Hiren, > > On Thu, Jan 21, 2016 at 06:24:02PM -0800, hiren panchasara wrote: > h> > Log: > h> > Provide new socket option TCP_CCALGOOPT, which stands for TCP > congestion > h> > control algorithm op

Re: svn commit: r294540 - in head: etc/mtree include share/man/man4 share/man/man9 sys/netinet

On 01/22/16 at 02:07P, Gleb Smirnoff wrote: > Author: glebius > Date: Fri Jan 22 02:07:48 2016 > New Revision: 294540 > URL: https://svnweb.freebsd.org/changeset/base/294540 > > Log: > Provide new socket option TCP_CCALGOOPT, which stands for TCP congestion > control algorithm options. The

Re: svn commit: r294540 - in head: etc/mtree include share/man/man4 share/man/man9 sys/netinet

On 01/21/16 at 06:50P, Gleb Smirnoff wrote: > On Thu, Jan 21, 2016 at 06:43:10PM -0800, hiren panchasara wrote: > h> > h> > Provide new socket option TCP_CCALGOOPT, which stands for TCP > congestion > h> > h> > control algorithm options. The argument is

svn commit: r293710 - stable/10/sys/netinet

Author: hiren Date: Mon Jan 11 23:34:29 2016 New Revision: 293710 URL: https://svnweb.freebsd.org/changeset/base/293710 Log: MFC: r292003 Improve tcp duplicate ack processing when SACK is present. Modified: stable/10/sys/netinet/tcp_input.c stable/10/sys/netinet/tcp_sack.c

svn commit: r293709 - stable/10/sys/netinet

Author: hiren Date: Mon Jan 11 23:31:13 2016 New Revision: 293709 URL: https://svnweb.freebsd.org/changeset/base/293709 Log: MFC: r290122 Calculate the correct amount of bytes that are in-flight for a connection as suggested by RFC 6675. MFC: r292046 r290122 added 4 bytes and removed

svn commit: r293711 - stable/10/sys/netinet/cc

Author: hiren Date: Mon Jan 11 23:37:31 2016 New Revision: 293711 URL: https://svnweb.freebsd.org/changeset/base/293711 Log: MFC: r292011 MFC: r292012 Add an option to use rfc6675 based pipe/inflight bytes calculation in cubic and newreno. Modified:

Re: svn commit: r293439 - in head: lib/libc/sys sys/dev/ti sys/kern sys/sys usr.bin/netstat

I'd also stress on having important details in commit-log itself. On 01/08/16 at 01:32P, Gleb Smirnoff wrote: > I'm writing status report on this right now. Looking forward to it. > > Also there is slideshare: > http://www.slideshare.net/facepalmtarbz2/new-sendfile-in-english Thanks. Cheers,

svn commit: r292087 - head/sys/netinet

Author: hiren Date: Fri Dec 11 06:22:58 2015 New Revision: 292087 URL: https://svnweb.freebsd.org/changeset/base/292087 Log: Clean up unused bandwidth entry in the TCP hostcache. Submitted by: Jason Wolfe (j at nitrology dot com) Reviewed by: rrs, hiren Sponsored by:

svn commit: r292012 - head/sys/netinet/cc

Author: hiren Date: Wed Dec 9 08:53:41 2015 New Revision: 292012 URL: https://svnweb.freebsd.org/changeset/base/292012 Log: Add an option to use rfc6675 based pipe/inflight bytes calculation in newreno. MFC after:3 weeks Sponsored by: Limelight Networks Modified:

svn commit: r292046 - head/sys/netinet

Author: hiren Date: Thu Dec 10 03:20:10 2015 New Revision: 292046 URL: https://svnweb.freebsd.org/changeset/base/292046 Log: r290122 added 4 bytes and removed 8 in struct sackhint. Add a pad entry of 4 bytes to restore the size. Spotted by: rrs Reviewed by: rrs X-MFC with:

svn commit: r292011 - head/sys/netinet/cc

Author: hiren Date: Wed Dec 9 07:56:40 2015 New Revision: 292011 URL: https://svnweb.freebsd.org/changeset/base/292011 Log: Add an option to use rfc6675 based pipe/inflight bytes calculation in cubic. Reviewed by: gnn MFC after:3 weeks Sponsored by:

svn commit: r292003 - head/sys/netinet

Author: hiren Date: Tue Dec 8 21:21:48 2015 New Revision: 292003 URL: https://svnweb.freebsd.org/changeset/base/292003 Log: One of the ways to detect loss is to count duplicate acks coming back from the other end till it reaches predetermined threshold which is 3 for us right now. Once

svn commit: r290379 - head/sys/netinet

Author: hiren Date: Thu Nov 5 02:09:48 2015 New Revision: 290379 URL: https://svnweb.freebsd.org/changeset/base/290379 Log: Improve the sysctl node name. X-MFC with: r290122 Sponsored by: Limelight Networks Modified: head/sys/netinet/tcp_input.c Modified:

svn commit: r290122 - head/sys/netinet

Author: hiren Date: Wed Oct 28 22:57:51 2015 New Revision: 290122 URL: https://svnweb.freebsd.org/changeset/base/290122 Log: Calculate the correct amount of bytes that are in-flight for a connection as suggested by RFC 6675. Currently differnt places in the stack tries to guess this in

svn commit: r290125 - stable/10/sys/netinet

Author: hiren Date: Thu Oct 29 00:36:10 2015 New Revision: 290125 URL: https://svnweb.freebsd.org/changeset/base/290125 Log: MFC r289293 Fix an unnecessarily aggressive behavior where mtu clamping begins on first retransmission timeout (rto) when blackhole detection is enabled. Make

svn commit: r290043 - in head: share/man/man4 sys/netinet

Author: hiren Date: Tue Oct 27 09:43:05 2015 New Revision: 290043 URL: https://svnweb.freebsd.org/changeset/base/290043 Log: Add sysctl tunable net.inet.tcp.initcwnd_segments to specify initial congestion window in number of segments on fly. It is set to 10 segments by default. Remove

  1   2   3   >