Failed to emerge net-mail/courier-imap-4.0.6-r3 make[3]: Entering directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/numlib' Compiling atotimet.c ./libtool: line 454: CDPATH: command not found ./libtool: line 1139: func_opt_split: command not found libtool: Version mismatch error. This is libtool 2.2.6, but the libtool: definition of this LT_INIT comes from an older release. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.6 libtool: and run autoconf again. make[3]: *** [atotimet.lo] Ошибка 1 make[3]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/numlib' make[2]: *** [all] Ошибка 2 make[2]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/numlib' make[1]: *** [all-recursive] Ошибка 1 make[1]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6' Reproducible: Always Steps to Reproduce: 1. emerge sys-devel/libtool 2.2.6a 2. emerge courier-imap
Full build.log, please. Anyway, courier-imap has a really messy build system and it's a package, that nobody really cares about (or so it seems). See perhaps if 4.5.0 handles it better.
Created attachment 204301 [details] build.log
I think adding this block: cd "${S}/numlib" ebegin "Recreating numlib/configure" eautoreconf || die "eautoreconf on numlib failed" eend $? may help. Not sure though, it should fix the problem of old libtool macro definition, but new problem may surface.
(In reply to comment #3) I cannot understand how I can use your advice. I must place this block in ebuild file?
Yes, in the ebuild, right after other such blocks in src_unpack.
I have the same Problem. In the courier ebuild you mean ? Can you tell us on what position we should add this Block?
Created attachment 204512 [details, diff] Ebuild-Patch for #c3 (fixing libtool issue) for all thos who don't know where to put the block Rafa? Mu?y?o posted in #c3
it's not only for "numlib", the patch fixes that one but you'll end up with the next libtool error in the "md5"-directory (and further i'd guess)
For those just trying to get courier-imap (r)emerged, an easy work-around is temporarily downgrading libtool: emerge -av1 =sys-devel/libtool-1.5.26 courier-imap But this is obviously not a long-term fix (er, unless libtool can be slotted).
Seems I'm having the same problem with courier-authlib-0.62.2: http://bugs.gentoo.org/show_bug.cgi?id=286811
There's a little catch about this "workaround": as it seems those courier packages use libltdl, emerging them after emerging old libtool gets them linked to the wrong libltdl (as major lib version has changed). Perhaps somebody could test, if now a fully recursive eautoreconf works (instead of doing just a few dirs).
i have these files: /usr/include/ltdl.h /usr/share/aclocal/ltdl.m4 /usr/share/libtool/libltdl/ltdl.c /usr/share/libtool/libltdl/ltdl.h all of these files belongs to sys-devel/libtool there's no a package named libltdl or *tdl*
(In reply to comment #12) > i have these files: > .... > all of these files belongs to sys-devel/libtool > there's no a package named libltdl or *tdl* > Of course there's none - libltdl is provided by libtool.
and how to fix this bug? i`m don`t understand what mean: Ebuild-Patch for #c3 (fixing libtool issue) at #7 comment how to use this patch? p.s. sorry for my english )
(In reply to comment #14) > and how to fix this bug? i`m don`t understand what mean: > Ebuild-Patch for #c3 (fixing libtool issue) at #7 comment > how to use this patch? > > p.s. sorry for my english ) > I believe this is what You should do (someone correct me if I'm wrong)...: Download courier-4.0.6-r3_libtoolfix.patch to /usr/portage/net-mail/courier-imap and cd (cd /usr/portage/net-mail/courier-imap) to the directory Run the following: patch courier-imap-4.0.6-r3.ebuild < courier-4.0.6-r3_libtoolfix.patch ebuild courier-imap-4.0.6-r3.ebuild digest emerge courier-imap
Never mind... Proposed patch doesn't work: [...] make[3]: Entering directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/md5' Compiling md5.c Compiling md5_hash.c Compiling redhat-crypt-md5.c Compiling hmac.c Compiling md5test.c ./libtool: line 454: CDPATH: command not found ./libtool: line 454: CDPATH: command not found ./libtool: line 454: CDPATH: command not found ./libtool: line 454: CDPATH: command not found ./libtool: line 1139: func_opt_split: command not found ./libtool: line 1139: func_opt_split: command not found ./libtool: line 1139: func_opt_split: command not found libtool: Version mismatch error. This is libtool 2.2.6, but the libtool: definition of this LT_INIT comes from an older release. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.6 libtool: and run autoconf again. libtool: Version mismatch error. This is libtool 2.2.6, but the libtool: definition of this LT_INIT comes from an older release. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.6 libtool: and run autoconf again. libtool: Version mismatch error. This is libtool 2.2.6, but the libtool: definition of this LT_INIT comes from an older release. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.6 libtool: and run autoconf again. make[3]: *** [md5_hash.lo] Error 1 make[3]: *** Waiting for unfinished jobs.... make[3]: *** [md5.lo] Error 1 ./libtool: line 1139: func_opt_split: command not found make[3]: *** [redhat-crypt-md5.lo] Error 1 libtool: Version mismatch error. This is libtool 2.2.6, but the libtool: definition of this LT_INIT comes from an older release. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.6 libtool: and run autoconf again. make[3]: *** [hmac.lo] Error 1 make[3]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/md5' make[2]: *** [all] Error 2 make[2]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6/md5' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/work/courier-imap-4.0.6' make: *** [all] Error 2 * * ERROR: net-mail/courier-imap-4.0.6-r3 failed. * Call stack: * ebuild.sh, line 49: Called src_compile * environment, line 2797: Called die * The specific snippet of code: * emake || die "emake failed" * The die message: * emake failed * * If you need support, post the topmost build error, and the call stack if relevant. * A complete build log is located at '/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/temp/environment'. * >>> Failed to emerge net-mail/courier-imap-4.0.6-r3, Log file: >>> '/var/tmp/portage/net-mail/courier-imap-4.0.6-r3/temp/build.log'
Just had a look at the ChangeLog for courier-imap: http://www.courier-mta.org/imap/changelog.html It seems that courier-imap-4.0.6 was released in 2005...?? Newest version is 4.6.0 released 2009-09-05...
may be forget change version in gentoo ports...?
i've got it, too. very annoying. maybe we shouldn't care about fixing such old versions and just try to stabilize a newer one?
Sounds like a good idea... ;) Another possible "solution" would be to use another IMAP server (like Dovecot), but I really don't want to migrate since it might break my current setup...
There are a lot of unstable versions in the portage: http://gentoo-portage.com/net-mail/courier-imap Now trying to upgrade to version 4.5.0 (~x86), I hope it will work (and that it is going to be marked as stable soon)
Now using courier-imap 4.5.0, so far it CAN be emerged without problems, and seems to do it's job.
I'm emerging as I write... :)
I've also upgraded to 4.5.0 and haven't seen any problems so far... ;) Maybe someone "forgot" to mark it stable...?? ;)
plz, tel me, how can i unmask courier-imap-4.5.0 or 4.6.0??
(In reply to comment #25) > plz, tel me, how can i unmask courier-imap-4.5.0 or 4.6.0?? > echo "=net-mail/courier-imap-4.5.0" >> /etc/portage/package.keywords This info is also available in the Gentoo Handbook :) http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=3&chap=3
(In reply to comment #26) > (In reply to comment #25) > > plz, tel me, how can i unmask courier-imap-4.5.0 or 4.6.0?? > > > > echo "=net-mail/courier-imap-4.5.0" >> /etc/portage/package.keywords > > This info is also available in the Gentoo Handbook :) > http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=3&chap=3 > when i compile courier authlib without postgres useflag it compile fine
I never compiled courier-authlib with the postgres use flag, and still, the stable version throws an error when compiling. (still running unstable on several servers. Still no problems)
(In reply to comment #28) > I never compiled courier-authlib with the postgres use flag, and still, the > stable version throws an error when compiling. (still running unstable on > several servers. Still no problems) > Same here... :)
(In reply to comment #29) > (In reply to comment #28) > > I never compiled courier-authlib with the postgres use flag, and still, the > > stable version throws an error when compiling. (still running unstable on > > several servers. Still no problems) > > > > Same here... :) > and here (4.5.0)... if there is not a good reason to keep 4.5.0 Masked... how about it be unmasked... this (along with the xorg and gnome upgrades has made this an ugly weekend of emerges trying to get current...
Created attachment 207859 [details, diff] Expanded version of patch that makes courier-imap actually compile. I'm attaching a patch (courier-imap-4.0.6-r3.ebuild.patch) that makes this compile for me. Basically, I just followed the previous instructions (adding in the "eautoreconf") for each thing that failed, until everything compiled. I've briefly tested this, and it appears that the compiled code actually functions correctly, too.
I've upgraded to 4.5.0, too. Error-free operation with about 45 users on a daily basis.
sucessfully compiled unmasking 2.5.0 version =)
(In reply to comment #33) > sucessfully compiled unmasking 2.5.0 version =) > I mean 4.5.0 version, sorry
courier-imap-4.5.0 is stable on most arches now, and fixes this bug Thanks for the report!