Page MenuHomeFreeBSD

tuexen (Michael Tuexen)
User

Projects

User Details

User Since
Feb 4 2016, 4:45 PM (210 w, 6 d)

Recent Activity

Yesterday

tuexen committed rS358083: Epochify SCTP..
Epochify SCTP.
Tue, Feb 18, 9:25 PM
tuexen committed rS358080: Remove unused function..
Remove unused function.
Tue, Feb 18, 7:42 PM

Mon, Feb 17

tuexen committed rS358028: Fix the non-default stream schedulers such that do not interleave.
Fix the non-default stream schedulers such that do not interleave
Mon, Feb 17, 6:05 PM
tuexen committed rS358023: Don't use uninitialised stack memory if the sysctl variable.
Don't use uninitialised stack memory if the sysctl variable
Mon, Feb 17, 2:54 PM
tuexen closed D23687: Fix sporadic odd MSS values when TCP HostCache is not in use.
Mon, Feb 17, 2:54 PM
tuexen accepted D23687: Fix sporadic odd MSS values when TCP HostCache is not in use.
Mon, Feb 17, 10:54 AM

Sun, Feb 16

tuexen added a comment to D23687: Fix sporadic odd MSS values when TCP HostCache is not in use.

OK, I double checked the code. I think the way to go is taking your change to tcp_hc_get() and not do the initialisations in cc_conn_init() and tcp_mss_update().
It is clear that tcp_hc_get() always returns the memory located at hc_metrics_lite in a correct state. So there is no need to initialise the memory before calling tcp_hc_get().
The code should be clear on this.

Sun, Feb 16, 5:24 PM

Sat, Feb 15

tuexen added a comment to D23687: Fix sporadic odd MSS values when TCP HostCache is not in use.

I think the proposed fix tcp_hc_get() is the correct thing to do. I don't think we need the initialisers in cc_conn_init() and tcp_mss_update. Which compiler warning are you referring to?

Sat, Feb 15, 12:17 AM

Thu, Feb 13

tuexen committed rS357858: sack_newdata and snd_recover hold the same value. Therefore, use only.
sack_newdata and snd_recover hold the same value. Therefore, use only
Thu, Feb 13, 3:15 PM
tuexen closed D18811: nuke sack_newdata.
Thu, Feb 13, 3:15 PM
tuexen accepted D18811: nuke sack_newdata.
Thu, Feb 13, 3:14 PM
tuexen committed rS357856: Whitespace cleanup. No functional change..
Whitespace cleanup. No functional change.
Thu, Feb 13, 1:58 PM

Wed, Feb 12

tuexen committed rS357830: Don't panic under INVARIANTS when we can't allocate memory for storing.
Don't panic under INVARIANTS when we can't allocate memory for storing
Wed, Feb 12, 5:05 PM
tuexen committed rS357829: Mark the socket as disconnected when freeing the association the first.
Mark the socket as disconnected when freeing the association the first
Wed, Feb 12, 5:02 PM
tuexen accepted D23645: ECN incorrectly set.
Wed, Feb 12, 12:47 PM

Tue, Feb 11

tuexen committed rS357775: Revert https://svnweb.freebsd.org/changeset/base/357761.
Revert https://svnweb.freebsd.org/changeset/base/357761
Tue, Feb 11, 8:03 PM
tuexen committed rS357768: Don't start an SCTP timer using a net, which has been removed..
Don't start an SCTP timer using a net, which has been removed.
Tue, Feb 11, 6:16 PM
tuexen committed rS357761: Use an int instead of a bool variable, since bool is not supported.
Use an int instead of a bool variable, since bool is not supported
Tue, Feb 11, 2:00 PM

Sun, Feb 9

tuexen committed rS357708: Stop the PMTU and HB timer when removing a net, not when freeing it..
Stop the PMTU and HB timer when removing a net, not when freeing it.
Sun, Feb 9, 10:40 PM
tuexen committed rS357705: Cleanup timer handling..
Cleanup timer handling.
Sun, Feb 9, 10:05 PM

Tue, Feb 4

tuexen committed rS357501: Remove unused timer..
Remove unused timer.
Tue, Feb 4, 2:01 PM
tuexen committed rS357500: Improve numbering of debug information..
Improve numbering of debug information.
Tue, Feb 4, 12:34 PM

Tue, Jan 28

tuexen committed rS357197: Fix build issues for the userland stack on 32-bit platforms..
Fix build issues for the userland stack on 32-bit platforms.
Tue, Jan 28, 10:09 AM
tuexen added inline comments to D23353: ECN and CUBIC do not play well together in very congested networks.
Tue, Jan 28, 8:17 AM

Sat, Jan 25

tuexen added a comment to D23353: ECN and CUBIC do not play well together in very congested networks.

Why are you enforcing a lower bond of two MSS instead of one MSS?

Sat, Jan 25, 3:24 PM
tuexen added a comment to D23281: Reduce Delayed ACK timeout to 40ms.

Would it make sense to make the default value for the delay ACK timer sysctl'able?

Sat, Jan 25, 1:47 PM
tuexen committed rS357116: Sending CWR after an RTO is according to RFC 3168 generally required.
Sending CWR after an RTO is according to RFC 3168 generally required
Sat, Jan 25, 1:45 PM
tuexen closed D23119: Send CWR whenever a ECN-enabled session runs into RTO.
Sat, Jan 25, 1:45 PM
tuexen accepted D23119: Send CWR whenever a ECN-enabled session runs into RTO.
Sat, Jan 25, 1:44 PM
tuexen closed D23118: Fix IP ECN codepoint for SACK retransmissions (RFC3168).
Sat, Jan 25, 1:35 PM
tuexen committed rS357115: Don't set the ECT codepoint on retransmitted packets during SACK loss.
Don't set the ECT codepoint on retransmitted packets during SACK loss
Sat, Jan 25, 1:35 PM
tuexen accepted D23118: Fix IP ECN codepoint for SACK retransmissions (RFC3168).
Sat, Jan 25, 1:28 PM
tuexen closed D23228: Fix passive ECN negotiation.
Sat, Jan 25, 1:12 PM
tuexen committed rS357114: As a TCP client only enable ECN when the corresponding sysctl variable.
As a TCP client only enable ECN when the corresponding sysctl variable
Sat, Jan 25, 1:11 PM
tuexen accepted D23228: Fix passive ECN negotiation.

The same change is needed for RACK. I will add it when committing.

Sat, Jan 25, 1:06 PM

Fri, Jan 24

tuexen closed D22670: Send immediate ACK on receipt of CWR.
Fri, Jan 24, 10:50 PM
tuexen committed rS357101: Don't delay the ACK for a TCP segment with the CWR flag set..
Don't delay the ACK for a TCP segment with the CWR flag set.
Fri, Jan 24, 10:50 PM
tuexen closed D23212: Fix server side TCP fast open.
Fri, Jan 24, 10:38 PM
tuexen committed rS357100: The server side of TCP fast open relies on the delayed ACK timer to allow.
The server side of TCP fast open relies on the delayed ACK timer to allow
Fri, Jan 24, 10:38 PM

Jan 16 2020

tuexen accepted D22670: Send immediate ACK on receipt of CWR.
Jan 16 2020, 9:40 PM
tuexen added a comment to D23212: Fix server side TCP fast open.

There is a special case handling in DSACK when start and end (2nd and 3rd param) are equal. Placing the todrop before the block will prevent DSACKs from generated if the last in-sequence packet was sent twice (is to be dsacked).

Please note that todrop is positive when we enter the block. It is only reduced when a SYN segment is received. So the case where todrop is zero and caught by the condition I added is only when todrop is initially 1 and reduced to 0 because we received a SYN segment. The condition todrop > 0 and tlen == 0 would be handled wrong. Can it happen?

I suggest to move the (todrop > 0) check just over the tf_ACKNOW statement, possibly with a check if we are not to send out a (D)SACK block (tp->rcv_numsacks >= 1).

Jan 16 2020, 8:24 PM
tuexen updated the diff for D23212: Fix server side TCP fast open.

Same change for RACK and BBR.

Jan 16 2020, 8:20 PM
tuexen created D23212: Fix server side TCP fast open.
Jan 16 2020, 7:34 PM
tuexen updated the summary of D23208: Remove DSACK debug logging.
Jan 16 2020, 5:16 PM
tuexen updated the summary of D23208: Remove DSACK debug logging.
Jan 16 2020, 5:16 PM
tuexen closed D23208: Remove DSACK debug logging.
Jan 16 2020, 5:15 PM
tuexen committed rS356796: Remove debug code not needed anymore..
Remove debug code not needed anymore.
Jan 16 2020, 5:15 PM
tuexen accepted D23208: Remove DSACK debug logging.
Jan 16 2020, 5:12 PM

Jan 12 2020

tuexen committed rS356663: Fix race when accepting TCP connections..
Fix race when accepting TCP connections.
Jan 12 2020, 5:52 PM
tuexen closed D22971: Fix race when accepting TCP connections.
Jan 12 2020, 5:52 PM
tuexen committed rS356660: Fix division by zero issue..
Fix division by zero issue.
Jan 12 2020, 3:45 PM

Jan 7 2020

tuexen accepted D23068: Do not define TCPOUTFLAGS in rack_bbr_common.c.
Jan 7 2020, 5:36 PM

Jan 6 2020

tuexen accepted D23052: Apply the same ECN updates that were in the default stack..
Jan 6 2020, 2:28 PM

Jan 5 2020

tuexen committed rS356378: Don't make the sendall iterator as being up if it could not be started..
Don't make the sendall iterator as being up if it could not be started.
Jan 5 2020, 2:08 PM
tuexen committed rS356377: Return -1 consistently if an error occurs..
Return -1 consistently if an error occurs.
Jan 5 2020, 2:07 PM
tuexen committed rS356376: Ensure that we don't miss a trigger for kicking off the SCTP iterator..
Ensure that we don't miss a trigger for kicking off the SCTP iterator.
Jan 5 2020, 1:56 PM

Jan 4 2020

tuexen committed rS356357: Make the message size limit used for SCTP_SENDALL configurable via.
Make the message size limit used for SCTP_SENDALL configurable via
Jan 4 2020, 8:33 PM

Jan 2 2020

tuexen committed rS356271: Remove empty line which was added in r356270 by accident..
Remove empty line which was added in r356270 by accident.
Jan 2 2020, 2:04 PM
tuexen committed rS356270: Improve input validation of the spp_pathmtu field in the.
Improve input validation of the spp_pathmtu field in the
Jan 2 2020, 1:55 PM

Dec 31 2019

tuexen committed rS356235: Fix delayed ACK generation for DCTCP..
Fix delayed ACK generation for DCTCP.
Dec 31 2019, 4:16 PM
tuexen closed D22644: Fix the CE reflector receiver state machine in DCTCP.
Dec 31 2019, 4:16 PM
tuexen accepted D22644: Fix the CE reflector receiver state machine in DCTCP.
Dec 31 2019, 4:12 PM
tuexen closed D22429: Addition of flag bit definitions in preparation of AccECN.
Dec 31 2019, 2:33 PM
tuexen committed rS356227: Add flags for upcoming patches related to improved ECN handling..
Add flags for upcoming patches related to improved ECN handling.
Dec 31 2019, 2:33 PM
tuexen accepted D22429: Addition of flag bit definitions in preparation of AccECN.
Dec 31 2019, 2:30 PM
tuexen committed rS356226: Clear the flag indicating that the last received packet was marked CE also.
Clear the flag indicating that the last received packet was marked CE also
Dec 31 2019, 2:24 PM
tuexen committed rS356224: Add curly braces missed in https://svnweb.freebsd.org/changeset/base/354773.
Add curly braces missed in https://svnweb.freebsd.org/changeset/base/354773
Dec 31 2019, 12:29 PM

Dec 30 2019

tuexen updated the test plan for D22971: Fix race when accepting TCP connections.
Dec 30 2019, 6:01 PM
tuexen updated the summary of D22971: Fix race when accepting TCP connections.
Dec 30 2019, 5:53 PM
tuexen created D22971: Fix race when accepting TCP connections.
Dec 30 2019, 5:51 PM

Dec 26 2019

tuexen accepted D22924: netinet/cc: make use of arc4random().

Unpredictability is not important here, it is more the distribution. Any reason why the change has to happen?

Dec 26 2019, 8:23 PM

Dec 20 2019

tuexen committed rS355931: Improve input validation for some parameters having a too small.
Improve input validation for some parameters having a too small
Dec 20 2019, 3:25 PM

Dec 6 2019

tuexen added a comment to D22704: Remove superfluous trailing whitespaces in TCP source files.

Why would cleaning up trailing whitespaces make MFCing harder? Other than possibly adding one adjacent line?

If you don't MFC the whitespace cleanup, you might get merge conflicts...

I made this huge diff, because sneaking these in one-by-one whenever I encounter them around the lines I was modifying was shot down so far ;)

As I said, doing this once is a nice thing in my opinion...

BTW - since lint was removed, what other tool could I use to validate the fbsd code style(9)? Any hint as to how i could integrate that into git or arcanist would be appreciated.

I"m not doing this automated... I just try to follow it (as much as I think is a good thing) when coding... But this is not perfect and others are also not perfect.

Dec 6 2019, 1:08 PM
tuexen added a comment to D22704: Remove superfluous trailing whitespaces in TCP source files.

When I did similar things in the past, one comment I received was that this does not improve the functionality, but
might make MFCing harder... However, only dealing with this in a single commit is definitely a good idea in my view.
The question is whether to do it at all. Maybe rrs@ or rgrimes@ have an opinion on this.

Dec 6 2019, 10:55 AM

Dec 1 2019

tuexen added a comment to D15818: Add a delayed VT attachment for ofwfb.

@jhibbits Thanks for the explanation. I'm using it on a blackbird. I have also a G5 Powermac in my lab, so if there is a need for testing, just let me know. I would like to run an unmodified version of FreeBSD just to make sure I don't commit anything by accident...

Dec 1 2019, 10:10 PM
tuexen added 1 blocking reviewer(s) for D22438: Improve NewReno after_idle ssthresh per RFC2861 Sec.2: transport.
Dec 1 2019, 9:08 PM
tuexen added a reviewer for D22438: Improve NewReno after_idle ssthresh per RFC2861 Sec.2: transport.
Dec 1 2019, 9:07 PM
tuexen requested changes to D22438: Improve NewReno after_idle ssthresh per RFC2861 Sec.2.
Dec 1 2019, 9:06 PM
tuexen closed D22497: Moving all ECN related flags to flags2.
Dec 1 2019, 9:01 PM
tuexen committed rS355273: Move all ECN related flags from the flags to the flags2 field..
Move all ECN related flags from the flags to the flags2 field.
Dec 1 2019, 9:01 PM
tuexen accepted D22497: Moving all ECN related flags to flags2.
Dec 1 2019, 8:51 PM
tuexen committed rS355272: Make the TF_* flags easier readable by humans by adding leading zeroes.
Make the TF_* flags easier readable by humans by adding leading zeroes
Dec 1 2019, 8:46 PM
tuexen closed D22428: Editorial change to enhance readability of the TF_* bit constants.
Dec 1 2019, 8:46 PM
tuexen accepted D22428: Editorial change to enhance readability of the TF_* bit constants.
Dec 1 2019, 8:36 PM
tuexen closed D21798: Restrict cwnd growth on app-limited flows.
Dec 1 2019, 8:36 PM
tuexen committed rS355269: This is an initial step in implementing the new congestion window.
This is an initial step in implementing the new congestion window
Dec 1 2019, 8:36 PM
tuexen accepted D21798: Restrict cwnd growth on app-limited flows.
Dec 1 2019, 7:22 PM
tuexen committed rS355268: Add a description for the TCP sysctl variable rfc6675_pipe..
Add a description for the TCP sysctl variable rfc6675_pipe.
Dec 1 2019, 7:20 PM
tuexen closed D22426: Add access to TOS(ECN bits) byte in TCP RACK.
Dec 1 2019, 6:48 PM
tuexen committed rS355267: Make the IPTOS value available to all substate handlers. This will allow.
Make the IPTOS value available to all substate handlers. This will allow
Dec 1 2019, 6:48 PM
tuexen accepted D22426: Add access to TOS(ECN bits) byte in TCP RACK.

Don't we need a similar change to BBR?

Dec 1 2019, 6:47 PM
tuexen committed rS355266: In order for the TCP Handshake to support ECN++, and further ECN-related.
In order for the TCP Handshake to support ECN++, and further ECN-related
Dec 1 2019, 6:05 PM
tuexen closed D22436: Add access to TOS(ECN bits) to the TCP syncache.
Dec 1 2019, 6:05 PM
tuexen accepted D22436: Add access to TOS(ECN bits) to the TCP syncache.
Dec 1 2019, 6:02 PM
tuexen closed D22547: schedule new RTT measurement after MTU update.
Dec 1 2019, 5:36 PM
tuexen committed rS355265: When changing the MTU of an SCTP path, not only cancel all ongoing.
When changing the MTU of an SCTP path, not only cancel all ongoing
Dec 1 2019, 5:35 PM
tuexen accepted D22547: schedule new RTT measurement after MTU update.
Dec 1 2019, 5:31 PM
tuexen closed D22605: SCTP over v6: Update host cache after MTU change.
Dec 1 2019, 4:15 PM
tuexen committed rS355264: Update the hostcache also for PTB messages received for SCTP/IPv6..
Update the hostcache also for PTB messages received for SCTP/IPv6.
Dec 1 2019, 4:15 PM
tuexen accepted D22605: SCTP over v6: Update host cache after MTU change.
Dec 1 2019, 4:11 PM