Bug 272754 - mail/exim unbroken with openssl31
Summary: mail/exim unbroken with openssl31
Status: Open
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: --- Affects Some People
Assignee: Dima Panov
URL: https://www.freshports.org/mail/exim/
Keywords:
Depends on:
Blocks:
 
Reported: 2023-07-27 13:22 UTC by Andrey Pevnev
Modified: 2023-11-10 07:21 UTC (History)
3 users (show)

See Also:
bugzilla: maintainer-feedback? (fluffy)


Attachments
build log (13.97 KB, text/plain)
2023-07-27 13:22 UTC, Andrey Pevnev
no flags Details
remove BROKEN lines (539 bytes, patch)
2023-07-27 13:23 UTC, Andrey Pevnev
no flags Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Andrey Pevnev 2023-07-27 13:22:08 UTC
Created attachment 243644 [details]
build log

I'm not sure why it's marked broken with openssl31, seems to build ok for me:
Comment 1 Andrey Pevnev 2023-07-27 13:23:48 UTC
Created attachment 243645 [details]
remove BROKEN lines
Comment 2 Igor Zabelin 2023-11-08 17:19:07 UTC
4.97 build ok
Comment 3 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 07:10:21 UTC
(In reply to Andrey Pevnev from comment #0)
the build log does not show that you build with openssl31. Can you provide a poudriere build log ?
Comment 4 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 07:11:12 UTC
(In reply to Igor Zabelin from comment #2)
On which versions of fbsd does it build ? I have a testbuild@work on current,
will test other versions if that one works.
Comment 5 Igor Zabelin 2023-11-09 08:40:45 UTC
(In reply to Kurt Jaeger from comment #4)

13.2-RELEASE-p4
Comment 6 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 09:05:05 UTC
(In reply to Igor Zabelin from comment #5)
Building with mysql=10.6m causes the build of mariadb106 to fail, and
therefore the build of exim to bail out.
Comment 7 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 09:06:01 UTC
(In reply to Kurt Jaeger from comment #6)
ups: on 13.2. It seems to build on 15.0...
Comment 8 Igor Zabelin 2023-11-09 09:53:15 UTC
(In reply to Kurt Jaeger from comment #6)

mariadb106-client-10.6.15 You are using OpenSSL from ports and have
selected GSSAPI from base, please select another GSSAPI value.

openssl31 or openssl32.

This is not about an exim build fail.
Comment 9 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 13:40:09 UTC
(In reply to Igor Zabelin from comment #8)
13.2, mariadb1011-client (builds), with openssl31, exim fails to build:

https://people.freebsd.org/~pi/exim-132-ossl31.txt
Comment 10 Andrey Pevnev 2023-11-09 14:27:52 UTC
(In reply to Kurt Jaeger from comment #3)
I don't run poudriere on that host (just cd /usr/ports/mail/exim && make install clean) but
===>   exim-4.96_1 depends on file: /usr/local/lib/libcrypto.so.13 - found

says it's using openssl3.1
Comment 11 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 15:39:01 UTC
(In reply to Andrey Pevnev from comment #10)
poudriere is normally the reproducable process to build ports. If it
fails, something is problematic.
Comment 12 Igor Zabelin 2023-11-09 18:11:09 UTC
(In reply to Kurt Jaeger from comment #11)

This fail is only with openssl31?
No problem with the openssl 3.0 and 3.2 versions?
Comment 13 Kurt Jaeger freebsd_committer freebsd_triage 2023-11-09 19:07:57 UTC
(In reply to Igor Zabelin from comment #12)
I have not tested every combination 8-} of openssl and fbsd version...

I'll test when I find the time.
Comment 14 Igor Zabelin 2023-11-09 19:28:56 UTC
I can't build mariadb with openssl31.
There is no way to use mariadb and openssl31 and default option GSSAPI_BASE.
Then I use the possible option GSSAPI support via security/krb5.
And there is no problem with exim. Build ok.

There is mess with mariadb and openssl 3.x and exim. 
The initial problem was with openssl31 and exim. And everything is fine there. And there is no problem with mysql 8 either.
Comment 15 Igor Zabelin 2023-11-10 07:21:05 UTC
And more:

BROKEN_SSL_REASON=      error: token is not a valid binary operator in a preprocessor subexpression

Where is this error in log?