[Swan-commit] Changes to ref refs/heads/main
Andrew Cagney
cagney at vault.libreswan.fi
Fri May 21 15:42:31 UTC 2021
New commits:
commit 0710654fdff55bb95744dcd0e2467bba8caf9520
Merge: 0c1004c 384d902
Author: Andrew Cagney <cagney at gnu.org>
Date: Fri May 21 11:39:13 2021 -0400
ikev2: make the IKE SA responsible for all retransmits
Code was trying to use the child sa for this.
Also, when retransmits fail, an there's no retry left,
just kill the entire IKE family. (I suspect this code path
needs more work; even when a connection has retry left,
the IKE family is dead and should be deleted).
Some, but not all tests are updated (having the IKE SA
initiator not switch to the child will fix more).
Merge commit '384d902fe41d7b8c20ef80b7d62008c0c0458dbf'
commit 384d902fe41d7b8c20ef80b7d62008c0c0458dbf
Author: Andrew Cagney <cagney at gnu.org>
Date: Fri May 21 11:32:30 2021 -0400
testing: switch IKEv2 retransmit logs to the IKE SA
The IKEv2 code's been switched to using the IKE SA when managing
retransmits (it was using the current SA which could be a child).
This updates all the logs that clearly need changing. However, it
leaves some more obscure cases (for instance a test expecting to
find a retransmitting CHILD SA using grep).
commit 5f4b291c78bac414f57c5a6ec93af7b44ddc67bf
Author: Andrew Cagney <cagney at gnu.org>
Date: Thu May 20 16:16:52 2021 -0400
ikev2: blame the IKE SA for retransmits
... and get retransmit parameters from the IKE SA's connection.
The code was getting the connection retransmit parameters using the
current SA (which can be a CHILD SA with a different connection) and
not the IKE SA.
commit e80920659c62eb615e14b3a7bd824c97565f36be
Author: Andrew Cagney <cagney at gnu.org>
Date: Thu May 20 21:46:27 2021 -0400
ikev2: when retransmits fail, delete the ike family
More information about the Swan-commit
mailing list