git: d174085f9f89 - main - tcp: improve sender side SWS avoidance
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Mon, 16 Nov 2026 17:29:40 UTC
The branch main has been updated by tuexen:
URL: https://cgit.FreeBSD.org/src/commit/?id=d174085f9f895ed444bd95120da3eff481a58f1a
commit d174085f9f895ed444bd95120da3eff481a58f1a
Author: Michael Tuexen <tuexen@FreeBSD.org>
AuthorDate: 2026-11-16 17:25:32 +0000
Commit: Michael Tuexen <tuexen@FreeBSD.org>
CommitDate: 2026-11-16 17:25:32 +0000
tcp: improve sender side SWS avoidance
When checking for a full sized segment, take all TCP options that will
be included into account. This fixes a bug which can trigger the
"sent too much" KASSERT related to limited transmit.
Reported by: olivier
Reviewed by: rscheff, Peter Lei
MFC after: 3 days
MFC to: stable/15
MFC to: stable/14
Sponsored by: Netflix, Inc.
Differential Revision: https://reviews.freebsd.org/D61605
---
sys/netinet/tcp_output.c | 14 +-------------
1 file changed, 1 insertion(+), 13 deletions(-)
diff --git a/sys/netinet/tcp_output.c b/sys/netinet/tcp_output.c
index 2ed2a7319aba..b37edfa35248 100644
--- a/sys/netinet/tcp_output.c
+++ b/sys/netinet/tcp_output.c
@@ -587,19 +587,7 @@ after_sack_rexmit:
if (len) {
if (len >= tp->t_maxseg)
goto send;
- /*
- * As the TCP header options are now
- * considered when setting up the initial
- * window, we would not send the last segment
- * if we skip considering the option length here.
- * Note: this may not work when tcp headers change
- * very dynamically in the future.
- */
- if ((((tp->t_flags & TF_SIGNATURE) ?
- PADTCPOLEN(TCPOLEN_SIGNATURE) : 0) +
- ((tp->t_flags & TF_RCVD_TSTMP) ?
- PADTCPOLEN(TCPOLEN_TIMESTAMP) : 0) +
- len) >= tp->t_maxseg)
+ if (len >= tcp_maxseg(tp))
goto send;
/*
* NOTE! on localhost connections an 'ack' from the remote