Specifically, newer autoconf (> 2.13) has different semantic of the
configure target. In short, one should use --build=CONFIGURE_TARGET
instead of CONFIGURE_TARGET directly. Otherwise, you will get a warning
and the old semantic may be removed in later autoconf releases.
To workaround this issue, many ports hack the CONFIGURE_TARGET variable
so that it contains the ``--build='' prefix.
To solve this issue, under the fact that some ports still have
configure script generated by the old autoconf, we use runtime detection
in the do-configure target so that the proper argument can be used.
Changes to Mk/*:
- Add runtime detection magic in bsd.port.mk
- Remove CONFIGURE_TARGET hack in various bsd.*.mk
- USE_GNOME=gnometarget is now an no-op
Changes to individual ports, other than removing the CONFIGURE_TARGET hack:
= pkg-plist changed (due to the ugly CONFIGURE_TARGET prefix in * executables)
- comms/gnuradio
- science/abinit
- science/elmer-fem
- science/elmer-matc
- science/elmer-meshgen2d
- science/elmerfront
- science/elmerpost
= use x86_64 as ARCH
- devel/g-wrap
= other changes
- print/magicfilter
GNU_CONFIGURE -> HAS_CONFIGURE since it's not generated by autoconf
Total # of ports modified: 1,027
Total # of ports affected: ~7,000 (set GNU_CONFIGURE to yes)
PR: 126524 (obsoletes 52917)
Submitted by: rafan
Tested on: two pointyhat 7-amd64 exp runs (by pav)
Approved by: portmgr (pav)
allow removing either .tgz or .tbz temporary pkgs.
- Refactor the existing patches so that we have one patch per file to ease
maintenance.
- Pass maintainership to submitter.
- Bump PORTREVISION
PR: ports/106652
Submitted by: Robert C. Noland III <rnoland@2hip.net>
if portmanager has been installed for the first time it can't copy
pm-020.conf-SAMPLE to pm-020.conf.
PR: ports/90957
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
make config have their video scrambled. Fixed by adding TERM= back to environment
during make.
PR: ports/90633
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
update sysutils/portmanager to 0.3.9_7
1. Fixes portmanager not honoring STOP/START statements
from its own config file, reported by Gerard Seibert
<gerard@seibercom.net>, Thanks!
PR: ports/89913
Submitted by: Michael C. Shultz <ringworm01@gmail.com>
1, Fixes corruption of +CONTENTS files when portmanager is compiled in
BUILD_DEPENDS_ARE_LEAVES mode Bug identified by Jiawei Ye
<leafy7382@gmail.com>
2. Fixes parsing pkgtools.conf when arrays are used, patch
provided by dan@slightlystrange.org, thank you!
PR: ports/89903
Submitted by: Michael C. Shultz <ringworm01 -at- gmail.com> (maintainer)
- Fixes when adding a new port with portmanager, +CONTENTS files being
corrupted.
- Improved logging of conflicting ports
Submitted by: "Michael C. Shultz" <ringworm01@gmail.com> (maintainer)
PR: ports/89716
- Fixes bug where if run with -f switch when portmanager updates itsself goes into endless loop.
Thanks to Gerard Seibert for helping to track this one down!
PR: 89609
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
- Using ruby script provided by Jan Grant <Jan.Grant@bristol.ac.uk> if portupgrade is installed
portmanager also uses settings from pkgtools.conf in addition to those from its own conf file.
- Wild cards "*" may now be used in pm-020.conf, see man portmanager for details
- Automatic timeouts are re-enabled, stdin problem is solved.
- Fix build on 4.x (0.3.9)
PR: 89533
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Had to disable automatic timeouts until I learn how to properly close then open stdin.
0.3.8_1 will seg fault upon a second timeout and 0.3.8 leaves these random errors:
awk: i/o error occurred while closing /dev/stdin input record number
PR: 89424
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Added --pristine mode that guarentees no inconsistancies in +CONTENTS files
* Fixed (I hope) stdin not being reopened properly after a [Y/N/A] timeout
* Fixed a segfault when portmanager reads a very old pre 4.11 +CONTENTS file,
identified by Andrew Gordon <arg@arg.me.uk>
PR: 89397
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Craig Boston <craig@yekse.gank.org> brought to my attention a problem
with how dependency information is gathered in +CONTENT files.
portmanager -s (status report) uses port +CONTENT files to generate its
report and now that is proven as not 100% accurate. This patch modifies
the data source for status reports from +CONTENT files to the results
obtained when running make -V *_DEPENDS in each port directory.
PR: 89084
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Fixes bug where on initial installation if portmanager is run in single port update
mode first it is missing a database
* Added WITH_BUILD_DEPENDS_ARE_LEAVES information to man page
PR: 88953
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Fixed seg fault in -slid on 6.0 machines
* Fixed segfault when cvsup not set to delete and an installed port
is removed from the ports tree
PR: 88888
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* Adds ability to stop programs like postfix
from running during upgrade then restarts them
afterwards. See pm-020.conf-example or man page
for instructions
* Fixes reported bug where -slid option removes
the wrong port
* May fix a reported segfault at:
MGPMrCreateAllUpdateStatusDb.c:162
since user didn't include portmanager version or
uname -a output can't be certain fix is good
PR: 88800
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
* At user's request portmanager now respects a defined BATCH
* -sl works again, -sl and -slid options no longer consider
build dependencies as "leaves"
PR: 88744
Submitted by: Michael C. Shultz <ringworm01@gmail.com> (maintainer)
