Created attachment 143821 [details] plop latest sendmail (8.14.9) fails to build when ssl from ports is depended on, due to -rpath syntax with clang How-To-Repeat: cd /usr/ports/mail/sendmail make (set option TLS on and make sure WITH_OPENSSL_BASE is undefined.) Fix: apply patch to /usr/ports/mail/sendmail/files/site.config.m4.ssl
Over to maintainer.
I can not reproduce your problem Fresh build: # uname -a FreeBSD bamd10.dinoex.sub.de 10.0-RELEASE-p4 FreeBSD 10.0-RELEASE-p4 #0: Tue Jun 3 13:14:57 UTC 2014 root@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC amd64 # ldd /usr/local/sbin/sendmail /usr/local/sbin/sendmail: libwrap.so.6 => /usr/lib/libwrap.so.6 (0x8008c1000) libsasl2.so.3 => /usr/local/lib/libsasl2.so.3 (0x800aca000) libssl.so.8 => /usr/local/lib/libssl.so.8 (0x800ce5000) libcrypto.so.8 => /usr/local/lib/libcrypto.so.8 (0x800f4b000) libutil.so.9 => /lib/libutil.so.9 (0x801347000) libc.so.7 => /lib/libc.so.7 (0x801559000) libthr.so.3 => /lib/libthr.so.3 (0x8018f2000)
Created attachment 143979 [details] build log
(In reply to Dirk Meyer from comment #2) > I can not reproduce your problem Closed bug? Nice.... As I originally wrote: " due to -rpath syntax with clang". If you took 2 seconds to look at the patch, you'd see that the *non-clang compatible* use of '-rpath' is used instead of '-Wl,rpath' clang 3.3 ignores this, clang 3.4 (merged into 10-STABLE) a few months ago aborts. See http://docs.freebsd.org/cgi/getmsg.cgi?fetch=40565+0+/usr/local/www/db/text/2014/freebsd-python/20140223.freebsd-python Even on working 10-RELEASE this option should either be corrected, or removed. See the attached build.log I hope this is now sufficient, or do I need to wait until 10.1-RELEASE before reporting it? There is no issue for me - I've solved the problem, but when bug reports with fixes are simply closed without any meaningful being done, I'm sure I'm not alone in sometimes wonder why I bother raising them.
Sorry if you feel personal offended. Closing an issue is normal work flow in Bugzilla, there is no "feedback" state. a submitter can reopen the Ticket as you did.
Attached patch "plop" is of wrong type. Supposed patch rejected. Better report in 191290. *** This bug has been marked as a duplicate of bug 191290 ***
> --- Comment #5 from Dirk Meyer <dinoex@FreeBSD.org> --- > > Sorry if you feel personal offended. > > Closing an issue is normal work flow in Bugzilla, > there is no "feedback" state. > > a submitter can reopen the Ticket as you did. OK. *blush* My fault for not RTFM. I'd just had one of those days from hell, and took it out on you. Still, that's no excuse. Sorry, I'm not usually a grump old git. (And I'll use more appropriate attachment names in future!) Cheers, Jamie