Page MenuHomeFreeBSD

rrs (Randall Stewart)
User

Projects

User Details

User Since
Jan 22 2015, 5:22 AM (212 w, 6 d)

Recent Activity

Yesterday

rrs accepted D19254: Avoid dereferencing a NULL pointer when TCP_REASS_LOGGING is enabled.
Wed, Feb 20, 8:31 PM
rrs accepted D18974: Use exponential backoff also for SYN segments.
Wed, Feb 20, 12:08 PM

Tue, Feb 19

rrs accepted D19152: Various cleanups to the management of multiple TCP stacks..
Tue, Feb 19, 11:22 AM

Wed, Feb 13

rrs created D19179: Crashing in a race with SOCK_DGRAM and unix domain sockets.
Wed, Feb 13, 3:04 PM
rrs committed rS344099: This commit adds the missing release mechanism for the.
This commit adds the missing release mechanism for the
Wed, Feb 13, 2:58 PM
rrs closed D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.
Wed, Feb 13, 2:58 PM

Mon, Feb 11

rrs updated the diff for D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

get the update in and fix the pesky extra line.

Mon, Feb 11, 12:37 PM

Mon, Feb 4

rrs accepted D19071: Get TCP CDG CC working if net.inet.tcp.cc.cdg.smoothing_factor = 0 .
Mon, Feb 4, 4:25 PM
rrs added inline comments to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.
Mon, Feb 4, 12:11 PM

Fri, Feb 1

rrs added inline comments to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.
Fri, Feb 1, 4:05 PM
rrs updated the diff for D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

Update to hopefully get a better view and not have the stray line in the diff

Fri, Feb 1, 4:04 PM
rrs added a comment to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

Interesting my svn copy does not have that extra chunk.. let me re-update the diff and use -x -up

Fri, Feb 1, 4:03 PM
rrs added inline comments to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.
Fri, Feb 1, 4:01 PM
rrs updated the diff for D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

This updates the diff to Fix Hann's point on vlans, both the
vlan and lagg are now balanced.

Fri, Feb 1, 12:31 PM
rrs added a comment to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

Ahh good catch I did not know vlan had an allocator.. I will add it to the patch-set

Fri, Feb 1, 12:18 PM
rrs accepted D19033: Initialise snd_wnd on the server side always before processing the ack.
Fri, Feb 1, 11:26 AM

Thu, Jan 31

rrs added a comment to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

I don't follow your suggestion. Why would I add a if_snd_tag_free() in if_attach_internal() .
The driver below has to do that just like it has to add the if_snd_tag_alloc(). The
internal function has no way of knowing what is the right thing to do in
freeing a tag that the driver allocates. Having it be NULL as the default in
the internal just the same as if_snd_tag_alloc() is the right thing, only
if a driver sets these should they get used (when RATELIMIT is configured of course)

Thu, Jan 31, 8:07 PM

Wed, Jan 30

rrs added a comment to D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.

Hi,

This patch does not look correct. You will need to pass the flowid in order to correctly handle this inside lagg?

Why are you not using the tag->ifp->if_sndtag_free(tag) in the first place?

--HPS

Hann's

Wed, Jan 30, 4:31 PM
rrs created D19032: Balance the if_snd_tag_alloc with an appropriate if_snd_tag_free.
Wed, Jan 30, 1:43 PM

Dec 17 2018

rrs accepted D18580: Accept always TCP segment with SEG.SEQ==RCV.NXT.
Dec 17 2018, 9:43 PM
rrs added inline comments to D18580: Accept always TCP segment with SEG.SEQ==RCV.NXT.
Dec 17 2018, 4:59 PM
rrs requested changes to D18580: Accept always TCP segment with SEG.SEQ==RCV.NXT.
Dec 17 2018, 4:55 PM

Nov 29 2018

rrs accepted D18366: Limit option_len for the TCP_CCALGOOPT option.
Nov 29 2018, 5:19 PM
rrs accepted D18375: Improve receive window autoscaling for TCP.
Nov 29 2018, 5:19 PM

Nov 21 2018

rrs accepted D18021: Improve KASSERTs in rack.c.

Good catch Michael!!!

Nov 21 2018, 12:30 PM

Nov 20 2018

rrs accepted D18033: TCP RACK stack should honor net.inet.tcp.drop_synfin sysctl variable.

Looks good to me :)

Nov 20 2018, 11:01 PM
rrs accepted D18032: Support RTT measurement even if switching from the rack stack to the base stack.
Nov 20 2018, 9:36 PM
rrs accepted D18034: The TCP RACK stack need to verify SEG.ACK first when in SYS-SENT state.
Nov 20 2018, 1:07 PM
rrs added inline comments to D18032: Support RTT measurement even if switching from the rack stack to the base stack.
Nov 20 2018, 12:59 PM
rrs accepted D18021: Improve KASSERTs in rack.c.
Nov 20 2018, 12:57 PM
rrs accepted D17949: Advertise a zero window when sending TCP RST segments via tcp_output().

This looks fine Michael :)

Nov 20 2018, 12:53 PM

Oct 18 2018

rrs accepted D17595: Fix handling of RST segments in SYN-RCVD state via the syn cache code path.
Oct 18 2018, 6:04 PM

Oct 16 2018

rrs accepted D17588: Fix TCP module unloading code to work with epoch(9) changes.
Oct 16 2018, 6:10 PM

Aug 28 2018

rrs accepted D16912: Fix TCP Fast Open for TCP RACK.
Aug 28 2018, 5:00 AM

Aug 27 2018

rrs added inline comments to D16912: Fix TCP Fast Open for TCP RACK.
Aug 27 2018, 3:07 PM
rrs accepted D16891: Make tcp_hpts compile with option RSS.

Because in order to conserve space the fid in the in_pcb (IIRR) is a uint16_t.. I was thinking
64k cpu's is probably enough for now :)

Aug 27 2018, 2:55 PM

Aug 21 2018

rrs accepted D16791: Add support for SO_DOMAIN in getsockopt().
Aug 21 2018, 1:23 PM
rrs accepted D16792: Fix inheritance of IPv6 options.
Aug 21 2018, 1:22 PM
rrs accepted D16796: Fix IPV6_USE_MIN_MTU socket option for TCP sockets.
Aug 21 2018, 1:22 PM

Aug 20 2018

rrs committed rS338102: This change represents a substantial restructure of the way we.
This change represents a substantial restructure of the way we
Aug 20 2018, 12:43 PM
rrs closed D16626: Rewrite of TCP Reassembly code.
Aug 20 2018, 12:43 PM

Aug 18 2018

rrs added inline comments to D16626: Rewrite of TCP Reassembly code.
Aug 18 2018, 7:17 PM
rrs updated the diff for D16626: Rewrite of TCP Reassembly code.

Fix the silly cut and paste bug.

Aug 18 2018, 7:15 PM
rrs updated the diff for D16626: Rewrite of TCP Reassembly code.

This version incorporates all of Jonathans comments and suggested improvements. Thanks
Jonathan!!!

Aug 18 2018, 7:01 AM

Aug 10 2018

rrs updated the diff for D16626: Rewrite of TCP Reassembly code.

This version updates from the last to include all of Lawrence Stewarts comments
as well as those from Peter Lei. A significant change is altering the calculate
overhead function to also glean the mlast so that a second walk of the mbuf
chain is no longer needed.

Aug 10 2018, 11:59 AM

Aug 9 2018

rrs accepted D16636: Improve TCP timestamps.
Aug 9 2018, 10:11 AM
rrs accepted D16637: Remove ipsec includes from tcp_hpts.
Aug 9 2018, 10:10 AM
rrs updated the diff for D16626: Rewrite of TCP Reassembly code.

Fixes an accounting problem shown by testing

Aug 9 2018, 10:07 AM

Aug 8 2018

rrs updated the diff for D16626: Rewrite of TCP Reassembly code.

Testing with INVAR found a problem with my late morning last optimization. If the
seq is behind us but overlaps we can't do the optimization.

Aug 8 2018, 8:59 PM
rrs created D16626: Rewrite of TCP Reassembly code.
Aug 8 2018, 2:25 PM
rrs closed D16604: Rack sends extra FIN.
Aug 8 2018, 1:37 PM
rrs committed rS337455: Fix a small bug in rack where it will.
Fix a small bug in rack where it will
Aug 8 2018, 1:37 PM

Aug 6 2018

rrs created D16604: Rack sends extra FIN.
Aug 6 2018, 9:32 AM
rrs closed D16579: Rack delayed ack issue.
Aug 6 2018, 9:22 AM
rrs committed rS337375: This fixes a bug in Rack where we were.
This fixes a bug in Rack where we were
Aug 6 2018, 9:22 AM

Aug 5 2018

rrs updated the diff for D16579: Rack delayed ack issue.

Update my typo :)

Aug 5 2018, 11:47 AM
rrs added inline comments to D16579: Rack delayed ack issue.
Aug 5 2018, 11:45 AM

Aug 3 2018

rrs created D16579: Rack delayed ack issue.
Aug 3 2018, 11:48 AM

Jul 31 2018

rrs accepted D16377: Add dtrace network provider for UDP-Lite.
Jul 31 2018, 7:01 PM

Jul 30 2018

rrs accepted D16369: Add missing dtrace receive and send probes for TCP.
Jul 30 2018, 12:00 PM
rrs closed D16453: When data after close is on, you can mess up the reassm queue.
Jul 30 2018, 10:24 AM
rrs committed rS336893: This fixes a hole where rack could end up.
This fixes a hole where rack could end up
Jul 30 2018, 10:23 AM
rrs accepted D16503: Send consistent SEG.WIN when sending from the timewait code path.
Jul 30 2018, 10:19 AM
rrs accepted D16458: Allow implicit connection setup for TCP/IPv6.
Jul 30 2018, 10:17 AM
rrs accepted D16485: Fix some TCP fast open issues.
Jul 30 2018, 10:13 AM

Jul 26 2018

rrs created D16453: When data after close is on, you can mess up the reassm queue.
Jul 26 2018, 11:00 AM

Jul 24 2018

rrs committed rS336672: Delete the example tcp stack "fastpath" which.
Delete the example tcp stack "fastpath" which
Jul 24 2018, 2:56 PM
rrs closed D16420: Fastpath was a temp holder to give as an example and now needs to disappear.
Jul 24 2018, 2:56 PM
rrs updated the diff for D16420: Fastpath was a temp holder to give as an example and now needs to disappear.
Jul 24 2018, 1:27 PM
rrs created D16420: Fastpath was a temp holder to give as an example and now needs to disappear.
Jul 24 2018, 1:22 PM

Jul 19 2018

rrs accepted D16046: Add UDP receive probes.
Jul 19 2018, 3:36 PM

Jul 18 2018

rrs committed rS336465: Bump the ICMP echo limits to match the RFC.
Bump the ICMP echo limits to match the RFC
Jul 18 2018, 10:50 PM
rrs closed D16333: The ICMP echo is too small.
Jul 18 2018, 10:50 PM
rrs created D16333: The ICMP echo is too small.
Jul 18 2018, 10:46 PM

Jul 12 2018

rrs accepted D16201: acquire inp lock around ip6_pcbopt to fix IPV6_TCLASS panic.
Jul 12 2018, 4:26 PM

Jun 21 2018

rrs committed rS335502: This adds in an optimization so that we only walk one.
This adds in an optimization so that we only walk one
Jun 21 2018, 9:04 PM
rrs closed D15937: Optimize the TSO and copy paths to use the new tcp_m_copy routine.
Jun 21 2018, 9:04 PM

Jun 20 2018

rrs created D15937: Optimize the TSO and copy paths to use the new tcp_m_copy routine.
Jun 20 2018, 8:29 PM

Jun 19 2018

rrs committed rS335364: Make sure that the t_peakrate_thr is not compiled in.
Make sure that the t_peakrate_thr is not compiled in
Jun 19 2018, 11:20 AM
rrs committed rS335361: Move the tp set back to where it was before.
Move the tp set back to where it was before
Jun 19 2018, 5:28 AM

Jun 18 2018

rrs committed rS335317: Move to using the inp->vnet pointer has suggested by lstewart..
Move to using the inp->vnet pointer has suggested by lstewart.
Jun 18 2018, 2:10 PM

Jun 14 2018

rrs committed rS335106: This fixes several bugs that Larry Rosenman helped me find in.
This fixes several bugs that Larry Rosenman helped me find in
Jun 14 2018, 3:27 AM
rrs closed D15758: Rack crash that Larry Rosenman found...
Jun 14 2018, 3:27 AM

Jun 12 2018

rrs closed D15757: HPTS Is missing some support for Vnets.
Jun 12 2018, 11:54 PM
rrs committed rS335022: This fixes missing VNET sets in the hpts system. Basically.
This fixes missing VNET sets in the hpts system. Basically
Jun 12 2018, 11:54 PM
rrs updated the diff for D15758: Rack crash that Larry Rosenman found...

Turns out after looking at Larry's last core and contemplating it a bit if it
would have succeeded (which it did not) we would have crashed anyway.
This is because the SYN got included in the send_map and we would have
tried to trim 14 instead of the 13 that bytes in the socketbuffer. Fixing that
means the send_map cannot start at snd_una for the TFO case. This means
a bit more special code needs also to be added to the SYN_SENT case then
so you end up incrementing snd_una when you move to established and that
way when you call the ack processing code all will be well :)

Jun 12 2018, 11:59 AM
rrs updated the diff for D15758: Rack crash that Larry Rosenman found...

So this is weird we look like we sent a TFO and then hit a retransmit.. but somehow
the TFO flags was removed so we did not set len == 0 on having a SYN retransmitted.

Jun 12 2018, 5:58 AM

Jun 11 2018

rrs updated the diff for D15758: Rack crash that Larry Rosenman found...
Jun 11 2018, 11:24 PM
rrs updated the diff for D15757: HPTS Is missing some support for Vnets.

Get the vnet set much earlier and all the right bits around earlier too.

Jun 11 2018, 3:18 PM
rrs created D15758: Rack crash that Larry Rosenman found...
Jun 11 2018, 12:39 PM
rrs created D15757: HPTS Is missing some support for Vnets.
Jun 11 2018, 12:35 PM
rrs accepted D15756: Change RACK dependency on TCPHPTS from build-time to load-time.
Jun 11 2018, 10:43 AM

Jun 7 2018

rrs committed rS334815: Take out the stack alias inadvertantly added by my commit..
Take out the stack alias inadvertantly added by my commit.
Jun 7 2018, 8:57 PM
rrs committed rS334813: Fix build issue with const and volatile and the.
Fix build issue with const and volatile and the
Jun 7 2018, 7:58 PM
rrs committed rS334804: This commit brings in a new refactored TCP stack called Rack..
This commit brings in a new refactored TCP stack called Rack.
Jun 7 2018, 6:18 PM
rrs closed D15525: Rack from Netflix.
Jun 7 2018, 6:18 PM
rrs added a comment to D15373: Patch to dummynet to allow simulation of compressed ack's.

A couple of configs I use:

Jun 7 2018, 2:41 PM

May 23 2018

rrs created D15525: Rack from Netflix.
May 23 2018, 12:12 AM

May 22 2018

rrs added a comment to D15510: Defer inpcb deletion until after a grace period has elapsed.

Looking closely at this and thinking about tcp_hpts.c, if I grok this correctly
this should all integrate ok. Assuming that someone calls inp_pcbfree() on
something that is in hpts, even if the epoch_call happens, the inpcb won't
be freed until the hpts's main loop see the INP_FREED flag and does
its in_pcbrele_wlocked() call at which time the INP will finally get destroyed.

May 22 2018, 11:41 AM

May 9 2018

rrs created D15373: Patch to dummynet to allow simulation of compressed ack's.
May 9 2018, 9:14 PM