From: "selvanair (Code Review)" Date: Mon, 3 Jun 2024 03:04:01 +0000 In-Reply-To: References: Reply-To: selva.nair@...277..., arne-openvpn@...1227..., ssbssa@...360..., openvpn-devel@lists.sourceforge.net, frank@...2641... Message-ID: <2bd78e8fe192563a135fa0fedcf0ef62a2509ade-EmailReviewComments-HTML@...2715...> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Disposition: inline Content-Type: multipart/alternative; boundary="oK+EI5JDJqI="; charset=UTF-8 Subject: [Openvpn-devel] [M] Change in openvpn[master]: Implement Windows CA template match for Crypto-API selector List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: ssbssa Cc: plaisthos , flichtenheld , openvpn-devel --oK+EI5JDJqI= Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Attention is currently required from: flichtenheld, plaisthos, ssbssa. selvanair has posted comments on this change. ( http://gerrit.openvpn.net/c= /openvpn/+/621?usp=3Demail ) Change subject: Implement Windows CA template match for Crypto-API selector ...................................................................... Patch Set 1: (6 comments) Patchset: PS1: I can't comment on the usefulness of this, but the changes look easy to mai= ntain and clean, so a conditional feature-ACK from me. At least one function call involves network search in the AD, a test report= of what happens when no DC's are reachable would help. If that could cause= a slow timeout, or error, we should document it. In typical setups DCs are= often unreachable before the tunnel is up. File doc/man-sections/windows-options.rst: http://gerrit.openvpn.net/c/openvpn/+/621/comment/98846a64_c997a4b3 : PS1, Line 62: cryptoapicert "TMPL:1.3.6.1.4..." Add an example where the template name is used instead of OID? File src/openvpn/cryptoapi.c: http://gerrit.openvpn.net/c/openvpn/+/621/comment/41d62b25_9fc645be : PS1, Line 217: } I think I said this in response to the original in 2021 as well: why separa= te into two calls like this? The use of the flag is to avoid network calls,= but if we are anyway going to allow network access, its more efficient to = allow it in the first place. I'm assuming the call will return quickly if a= local search succeeds even if the flag is not there. http://gerrit.openvpn.net/c/openvpn/+/621/comment/aa376a5e_4ae3dc25 : PS1, Line 220: if (!info && fallback && groupid) I see the only place this is called is with fallback=3Dtrue, so we do not n= eed that argument and can always fallback if groupid !=3D 0. http://gerrit.openvpn.net/c/openvpn/+/621/comment/9ee83cff_cbebf4f4 : PS1, Line 277: * TMPL: