From: job Date: Fri, 7 Jun 2024 11:48:05 +0000 (+0000) Subject: Fine-tune the TA tiebreaker logic X-Git-Url: http://artulab.com/gitweb/?a=commitdiff_plain;h=93ab8db6af8b51343d56596a7cd9ae7b56dff6e5;p=openbsd Fine-tune the TA tiebreaker logic Additional tiebreaker: prefer TA certificates with the narrower validity window OK tb@ --- diff --git a/usr.sbin/rpki-client/parser.c b/usr.sbin/rpki-client/parser.c index fec811e472d..50f0ee72465 100644 --- a/usr.sbin/rpki-client/parser.c +++ b/usr.sbin/rpki-client/parser.c @@ -1,4 +1,4 @@ -/* $OpenBSD: parser.c,v 1.137 2024/06/07 08:33:12 tb Exp $ */ +/* $OpenBSD: parser.c,v 1.138 2024/06/07 11:48:05 job Exp $ */ /* * Copyright (c) 2019 Claudio Jeker * Copyright (c) 2019 Kristaps Dzonsons @@ -581,12 +581,15 @@ proc_parser_ta_cmp(const struct cert *cert1, const struct cert *cert2) return 1; /* - * Good tie breakers are not obvious. While RFC 6487 and other sources - * advise against backdating, it's explicitly allowed and some TAs do. - * Some TAs have also re-issued with new dates and old serialNumber. - * Ignore the notAfter as hopefully future best practices will encourage - * significant reduction of some egregiously long TA validity periods - * and we want to pick such certs up. + * The standards don't specify tiebreakers. While RFC 6487 and other + * sources advise against backdating, it's explicitly allowed and some + * TAs do. Some TAs have also re-issued with new dates and old + * serialNumber. + * Our tiebreaker logic: a more recent notBefore is taken to mean a + * more recent issuance, and thus preferable. Given equal notBefore + * values, prefer the TA cert with the narrower validity window. This + * hopefully encourages TA operators to reduce egregiously long TA + * validity periods. */ if (cert1->notbefore < cert2->notbefore) @@ -594,9 +597,15 @@ proc_parser_ta_cmp(const struct cert *cert1, const struct cert *cert2) if (cert1->notbefore > cert2->notbefore) return 1; + if (cert1->notafter > cert2->notafter) + return -1; + if (cert1->notafter < cert2->notafter) + return 1; + /* * The serialNumber isn't monotonic and some TAs use semi-random ones. - * If the number changed prefer the freshly-fetched cert. + * If the freshly-fetched cert's serial number is different from the + * cached one, prefer the freshly-fetched cert. */ serial1 = X509_get0_serialNumber(cert1->x509);