Page MenuHomeFreeBSD

tuexen (Michael Tuexen)
User

Projects

User Details

User Since
Feb 4 2016, 4:45 PM (382 w, 6 d)

Recent Activity

Tue, Jun 6

tuexen accepted D40436: tcp: Update cubic CC module variable names to align with RFC8312bis.

I think you are not only renaming variables, but also adding new ones. Please reflect that in the commit message.

Tue, Jun 6, 8:21 PM
tuexen accepted D40424: tcp: make the maximum number of retransmissions tunable per VNET.
Tue, Jun 6, 8:17 PM

Mon, Jun 5

tuexen closed D40423: tcp: improve sending of TTL/hoplimit and DSCP.
Mon, Jun 5, 9:02 PM
tuexen committed rGd66540e829ec: tcp: improve sending of TTL/hoplimit and DSCP (authored by tuexen).
tcp: improve sending of TTL/hoplimit and DSCP
Mon, Jun 5, 9:02 PM
tuexen added inline comments to D40424: tcp: make the maximum number of retransmissions tunable per VNET.
Mon, Jun 5, 7:14 PM
tuexen added inline comments to D40423: tcp: improve sending of TTL/hoplimit and DSCP.
Mon, Jun 5, 5:28 PM
tuexen updated the diff for D40423: tcp: improve sending of TTL/hoplimit and DSCP.

Be consistent.

Mon, Jun 5, 5:26 PM
tuexen added inline comments to D40423: tcp: improve sending of TTL/hoplimit and DSCP.
Mon, Jun 5, 2:29 PM
tuexen updated the diff for D40423: tcp: improve sending of TTL/hoplimit and DSCP.

Address most of cc's comments.

Mon, Jun 5, 2:27 PM
tuexen added inline comments to D40423: tcp: improve sending of TTL/hoplimit and DSCP.
Mon, Jun 5, 11:56 AM
tuexen requested review of D40423: tcp: improve sending of TTL/hoplimit and DSCP.
Mon, Jun 5, 10:46 AM

Fri, Jun 2

tuexen accepted D40391: tcp: Rack fixes and misc updates.
Fri, Jun 2, 2:16 PM

Tue, May 30

tuexen accepted D40336: siftr: bring back the siftr_pkts_per_log feature.
Tue, May 30, 3:14 PM
tuexen accepted D40323: siftr: convert this tval.tv_sec to type intmax_t to print across platforms.
Tue, May 30, 5:12 AM

Mon, May 29

tuexen accepted D40322: siftr: sync-up man page with recent code changes, and cleanup code.
Mon, May 29, 7:30 PM
tuexen accepted D40320: siftr: fix a build error for powerpc and arm platforms.
Mon, May 29, 7:28 PM
tuexen accepted D40302: siftr: three changes that improve performance.
Mon, May 29, 12:43 PM
tuexen added inline comments to D40302: siftr: three changes that improve performance.
Mon, May 29, 9:35 AM
tuexen added inline comments to D40302: siftr: three changes that improve performance.
Mon, May 29, 8:53 AM

Sat, May 27

tuexen accepted D40304: Refactor tcp_get_srtt().

There is the plan to move the base stack away from ticks and use usecs everywhere. But we wanted not to do this before stable/14 is branched.

Sat, May 27, 7:01 PM

Fri, May 26

tuexen accepted D40185: cc_cubic: Use units of micro seconds (usecs) instead of ticks in rtt..

You might want to wait also for Richards' approval, since he is much more familiar with the code than I am.

Fri, May 26, 5:48 PM
tuexen accepted D40267: tcp: fix a bug where unshifting should be put last in tcp_get_srtt().
Fri, May 26, 7:42 AM

Thu, May 25

tuexen added inline comments to D40267: tcp: fix a bug where unshifting should be put last in tcp_get_srtt().
Thu, May 25, 8:31 PM

Tue, May 23

tuexen accepted D40229: tcp: request tracking is not http specific..
Tue, May 23, 9:22 PM

Sat, May 20

tuexen added inline comments to D40185: cc_cubic: Use units of micro seconds (usecs) instead of ticks in rtt..
Sat, May 20, 8:25 PM

Thu, May 18

tuexen accepted D40146: There are congestion control algorithms will that pull in srtt, and this can cause issues with rack..
Thu, May 18, 6:50 PM
tuexen accepted D40061: Under RSS, assign a TCP flow's inp_flowid anyway..
Thu, May 18, 6:48 PM
tuexen added inline comments to D40061: Under RSS, assign a TCP flow's inp_flowid anyway..
Thu, May 18, 3:39 PM

Thu, May 11

tuexen added inline comments to D40061: Under RSS, assign a TCP flow's inp_flowid anyway..
Thu, May 11, 2:49 PM

May 7 2023

tuexen added a comment to D39848: siftr: use inp_flowid or a new packet hash for a flow identification.

Changing the algorithm is one thing I agree with. But why change the output in the summary from IP addresses / port numbers to flow IDs? For me the IP addresses and port numbers are easier to read and more helpful.

May 7 2023, 9:00 PM

May 6 2023

tuexen committed rGe044a0bce495: bblog: inherit TCP_LOG option from listener (authored by tuexen).
bblog: inherit TCP_LOG option from listener
May 6 2023, 9:24 AM
tuexen closed D38436: bblog: inherit TCP_LOG option.
May 6 2023, 9:24 AM
tuexen committed rGc2399dd2e227: tcp: improve BBLoging for PRUs (authored by tuexen).
tcp: improve BBLoging for PRUs
May 6 2023, 9:16 AM
tuexen closed D39591: tcp: improve logging for PRUs .
May 6 2023, 9:15 AM
tuexen committed rG113f56ba4930: tcp_rack: allow the module to be loaded without TCP_BLACKBOX (authored by tuexen).
tcp_rack: allow the module to be loaded without TCP_BLACKBOX
May 6 2023, 9:05 AM
tuexen closed D39860: tcp_rack: allow the module to be loaded without TCP_BLACKBOX.
May 6 2023, 9:05 AM

May 3 2023

tuexen committed rG04ede3675e44: sctp: only start shutdown guard timer when sending SHUTDOWN chunk (authored by tuexen).
sctp: only start shutdown guard timer when sending SHUTDOWN chunk
May 3 2023, 6:32 PM
tuexen committed rGd9ae4adff271: sctp: improve shutdown(..., SHUT_WR) handling (authored by tuexen).
sctp: improve shutdown(..., SHUT_WR) handling
May 3 2023, 3:34 PM

Apr 30 2023

tuexen committed rG1f0e13449bee: sctp: improve handling of stale cookie error causes (authored by tuexen).
sctp: improve handling of stale cookie error causes
Apr 30 2023, 9:46 AM

Apr 28 2023

tuexen added a comment to D39860: tcp_rack: allow the module to be loaded without TCP_BLACKBOX.

Should the man-page tcp_rack(4) be changed at the same time to mention TCP_BLACKBOX as an optional dependency (and what feature it unlocks in rack)?

Apr 28 2023, 10:55 AM

Apr 27 2023

tuexen requested review of D39860: tcp_rack: allow the module to be loaded without TCP_BLACKBOX.
Apr 27 2023, 9:15 PM

Apr 26 2023

tuexen accepted D39835: siftr: remove barely used hash generation per record.
In D39835#906912, @cc wrote:
In D39835#906901, @cc wrote:
In D39835#906795, @cc wrote:

What do we gain from removing this feature? My understanding is that if it is not enabled, it does not have a substantial CPU overhead. I have no idea who uses siftr and if they use this feature...

This feature is not used AFAIK. We have less carbon print (at least 10 chars of "0x00000000") per record after removing this feature.

OK, so I understand that it makes sense not to print it, if the feature is not used. But I have (in general) a hard time to tell, which features are used and which are not... So how do you know that you are not removing a feature considered important by some people?
An argument I see in favor of this change is that I'm not aware of any tool reading a .pcap file and computing the hashes for the packets.

Let's assume this is a dead feature, or the original authors do not need it any longer. Also, a hash value is hard to use for a packet identification. I will consider tcp.seq, tcp.ack, and tcp.len in the tcphdr for such identification. Besides, the inconsistency between IPv4 and IPv6 for this feature does not make sense.

On a second thought, if there is really someone who uses this feature. That also means this user knows how/where to add this missing code to study TCP, under the BSD license.

Apr 26 2023, 10:37 PM
tuexen added a comment to D39835: siftr: remove barely used hash generation per record.
In D39835#906795, @cc wrote:

What do we gain from removing this feature? My understanding is that if it is not enabled, it does not have a substantial CPU overhead. I have no idea who uses siftr and if they use this feature...

This feature is not used AFAIK. We have less carbon print (at least 10 chars of "0x00000000") per record after removing this feature.

Apr 26 2023, 6:59 PM
tuexen added a comment to D39835: siftr: remove barely used hash generation per record.

What do we gain from removing this feature? My understanding is that if it is not enabled, it does not have a substantial CPU overhead. I have no idea who uses siftr and if they use this feature...

Apr 26 2023, 6:25 PM

Apr 25 2023

tuexen accepted D39803: Change the unit of srtt and rto to usec, inspired by these in struct "tcp_info". Therefore, no need hz and tcp_rtt_scale in the headline of the log. Update the man page as well..

Isn't that another backwards incompatible change?

Apr 25 2023, 7:22 PM

Apr 24 2023

tuexen accepted D39776: Remove unused fields in siftr_stats. Thus, update the man page as well..
Apr 24 2023, 6:53 PM
tuexen added inline comments to D39776: Remove unused fields in siftr_stats. Thus, update the man page as well..
Apr 24 2023, 2:24 PM
tuexen added inline comments to D39776: Remove unused fields in siftr_stats. Thus, update the man page as well..
Apr 24 2023, 12:03 PM
tuexen accepted D39716: Correct the value of macro TF2_TCP_ACCOUNTING..
Apr 24 2023, 11:22 AM

Apr 23 2023

tuexen committed rG66d6fd532282: sctp: use constants from RFC 8260 to improve compliance (authored by tuexen).
sctp: use constants from RFC 8260 to improve compliance
Apr 23 2023, 3:50 PM

Apr 21 2023

tuexen updated the test plan for D38436: bblog: inherit TCP_LOG option.
Apr 21 2023, 8:16 AM

Apr 20 2023

tuexen updated the diff for D38436: bblog: inherit TCP_LOG option.

Update to latest version of the main tree.

Apr 20 2023, 10:46 PM
tuexen committed rGd7339c42673a: trpt: fix several bugs and add deprecation notice (authored by tuexen).
trpt: fix several bugs and add deprecation notice
Apr 20 2023, 10:16 PM
tuexen closed D39436: trpt: fix logging.
Apr 20 2023, 10:16 PM
tuexen committed rGddf24a507c1a: netstat: cleanup (authored by tuexen).
netstat: cleanup
Apr 20 2023, 10:06 PM
tuexen closed D39688: netstat: cleanup.
Apr 20 2023, 10:06 PM
tuexen accepted D39738: tcp: hpts needs to still call output even after input..
Apr 20 2023, 7:53 PM
tuexen accepted D39736: netstat: fix printing of TCP pcbs with -A.
Apr 20 2023, 7:16 PM
tuexen abandoned D39735: tcp: fix netstat -A.
Apr 20 2023, 7:14 PM
tuexen requested review of D39735: tcp: fix netstat -A.
Apr 20 2023, 6:53 PM

Apr 19 2023

tuexen accepted D39699: tcp: rack the request level logging is a bit too noisy when doing point logging..
Apr 19 2023, 5:58 PM
tuexen accepted D39690: tcp: Rack can crash with the new non-TSO fix...
Apr 19 2023, 5:01 PM
tuexen requested review of D39688: netstat: cleanup.
Apr 19 2023, 10:38 AM
tuexen committed rG373b95976bce: netstat: document that PCB information can't be read from corefiles (authored by tuexen).
netstat: document that PCB information can't be read from corefiles
Apr 19 2023, 10:17 AM
tuexen closed D39610: netstat: document that PCB information can't be read from corefiles.
Apr 19 2023, 10:17 AM

Apr 18 2023

tuexen accepted D39672: tcp: bbr.c is non-capable of doing ECN and sets an INP flag to fend off ECN however our syncache is not aware of that flag..
Apr 18 2023, 10:11 PM
tuexen accepted D39653: tcp: Inconsistent use of hpts_calling flag.
Apr 18 2023, 2:34 PM

Apr 17 2023

tuexen accepted D39622: tcp: Blackbox logging and tcp accounting together can cause a crash..
Apr 17 2023, 8:33 PM
tuexen added a comment to D39591: tcp: improve logging for PRUs .

I like the combing of all these functions prologue.

I have one note, though. The "PRU" in the commit message made me notice that tcp_bblog_pru() again started to use this PRU_XXX namespace. If you started using it again, it needs to get out of in_kdtrace.h. But I'd rather suggest to create some own namespace for BB logging and forget this legacy. The long story of PRU_XXX namespace can be found in a comment in in_kdtrace.h.

Apr 17 2023, 7:57 PM

Apr 16 2023

tuexen updated the diff for D39610: netstat: document that PCB information can't be read from corefiles.

Provide context in patch file.

Apr 16 2023, 9:39 PM
tuexen requested review of D39610: netstat: document that PCB information can't be read from corefiles.
Apr 16 2023, 9:32 PM

Apr 15 2023

tuexen requested review of D39591: tcp: improve logging for PRUs .
Apr 15 2023, 10:59 PM

Apr 14 2023

tuexen accepted D39576: tcp: stack unloading crash in rack and bbr.
Apr 14 2023, 7:28 PM
tuexen added inline comments to D39574: tcp_hpts: use queue(9) STAILQ for the input queue.
Apr 14 2023, 4:50 PM

Apr 13 2023

tuexen accepted D39565: tcp: rack pacing has some caveats that need to be obeyed when LRO is missing..
Apr 13 2023, 8:38 PM
tuexen updated the summary of D39436: trpt: fix logging.
Apr 13 2023, 2:55 PM
tuexen updated the summary of D39436: trpt: fix logging.
Apr 13 2023, 2:55 PM
tuexen updated the diff for D39436: trpt: fix logging.

Add changes to the man page.

Apr 13 2023, 2:49 PM

Apr 12 2023

tuexen accepted D39501: TCP: Misc cleanups of tcp_subr.c.

Would be nice to know what the problem is you are experiencing with the DSCP for IPv6.

Apr 12 2023, 6:37 PM
tuexen added inline comments to D39501: TCP: Misc cleanups of tcp_subr.c.
Apr 12 2023, 6:31 PM
tuexen added inline comments to D39516: tcp: make net.inet.tcp.functions_default vnet specific.
Apr 12 2023, 4:14 PM
tuexen committed rG2ba2849c8275: tcp: fix typo in comment (authored by tuexen).
tcp: fix typo in comment
Apr 12 2023, 4:11 PM
tuexen committed rGc687f21add77: tcp: make net.inet.tcp.functions_default vnet specific (authored by tuexen).
tcp: make net.inet.tcp.functions_default vnet specific
Apr 12 2023, 4:07 PM
tuexen closed D39516: tcp: make net.inet.tcp.functions_default vnet specific.
Apr 12 2023, 4:07 PM
tuexen added a comment to D39516: tcp: make net.inet.tcp.functions_default vnet specific.

Thanks for taking care of the problem. Will it grow vnet memory requirements? Can the change be applied safely to stable/13? Are any plans to MFC when gets committed to HEAD ?

Apr 12 2023, 10:09 AM
tuexen requested review of D39516: tcp: make net.inet.tcp.functions_default vnet specific.
Apr 12 2023, 9:08 AM
tuexen committed rG73c48d9d8f89: tcp: fix deregistering stacks when vnets are used (authored by tuexen).
tcp: fix deregistering stacks when vnets are used
Apr 12 2023, 8:56 AM
tuexen closed D39514: tcp: fix stack switching when vnets are used.
Apr 12 2023, 8:56 AM

Apr 11 2023

tuexen requested review of D39514: tcp: fix stack switching when vnets are used.
Apr 11 2023, 10:40 PM
tuexen added inline comments to D39501: TCP: Misc cleanups of tcp_subr.c.
Apr 11 2023, 6:40 PM
tuexen accepted D39502: tcp_lro: When processing compressed acks lets support the new early wake feature for rack..
Apr 11 2023, 4:42 PM
tuexen accepted D39503: use IPV6_FLOWLABEL_LEN.
Apr 11 2023, 4:42 PM

Apr 10 2023

tuexen accepted D39494: tcp: Rack - in the absence of LRO fixed rate pacing (loopback or interfaces with no LRO) does not work correctly..
Apr 10 2023, 8:18 PM

Apr 5 2023

tuexen requested review of D39436: trpt: fix logging.
Apr 5 2023, 4:54 PM
tuexen added inline comments to D39427: tcp: misc cleanup of options for rack as well as socket option logging..
Apr 5 2023, 4:43 PM
tuexen accepted D39427: tcp: misc cleanup of options for rack as well as socket option logging..
Apr 5 2023, 3:24 PM

Apr 4 2023

tuexen accepted D39402: Update rack to the latest code used at NF..
Apr 4 2023, 6:01 PM

Apr 3 2023

tuexen accepted D39210: Fixes in the tcp infrastructure with respect to stack changes as well as other infrastructure updates for incoming rack features..
Apr 3 2023, 9:03 PM

Mar 31 2023

tuexen accepted D39368: Add myself (cc) as a src commiter..
Mar 31 2023, 9:41 PM
tuexen accepted D39363: Revert "Add myself (cc) as a src commiter.".

The diff looks good.

Mar 31 2023, 11:45 AM