Changes in t_ipsec_policy.c http://nxr.netbsd.org/rss/src/tests/net/ipsec/t_ipsec_policy.c en Copyright 2005 Java t_ipsec_policy: Skip if net.inet.ipsec.enabled doesn't exist too.<br/><br/>Presumably this means the kernel was built without IPsec support.<br/><br/>PR kern/60669: netipsec key_sp2msg buffer overrun /src/tests/net/ipsec/t_ipsec_policy.c - 1.6 Fri Oct 02 16:14:37 UTC 2026 riastradh t_ipsec_policy: Skip if IPsec is disabled in the kernel.<br/><br/>Currently failing on the sparc testbed with ENOPROTOOPT, which likely<br/>indicates that IPsec is disabled on that host -- although other IPsec<br/>tests are passing just fine? Maybe they all use rump?<br/><br/>XXX Should maybe do this as a rumpy test instead so we can always<br/>ensure IPsec is enabled.<br/><br/>PR kern/60669: netipsec key_sp2msg buffer overrun /src/tests/net/ipsec/t_ipsec_policy.c - 1.5 Tue Sep 22 15:11:43 UTC 2026 riastradh t_ipsec_policy: Send all output to stderr.<br/><br/>This way it doesn't get separated in atf output, so you can see which<br/>subcases correspond to which check failures.<br/><br/>PR kern/60669: netipsec key_sp2msg buffer overrun /src/tests/net/ipsec/t_ipsec_policy.c - 1.4 Tue Sep 22 15:04:22 UTC 2026 riastradh tests/ipsec/t_ipsec_policy.c -- appease clang<br/><br/>Cast to (void) not (void *) to explicitly ignore a function return<br/>value, even though that value is a pointer.<br/><br/>Might fix the clang builds (should go part way at least). /src/tests/net/ipsec/t_ipsec_policy.c - 1.3 Mon Sep 21 10:52:42 UTC 2026 kre ipsec: Require matching src/dst addr family in IPsec policies.<br/><br/>PR kern/60669: netipsec key_sp2msg buffer overrun /src/tests/net/ipsec/t_ipsec_policy.c - 1.2 Fri Sep 18 13:33:46 UTC 2026 riastradh ipsec: Test edge cases of setting and getting IPsec policies.<br/><br/>PR kern/60669: netipsec key_sp2msg buffer overrun /src/tests/net/ipsec/t_ipsec_policy.c - 1.1 Fri Sep 18 13:32:55 UTC 2026 riastradh