cpet/openbsd-ports
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
47,988 Commits 1 Branch 0 Tags 554 MiB
f2aab6a0da
Commit Graph

1115 Commits

Author SHA1 Message Date
espie
bf85c6df5a make it possible to lock/unlock manually a given directory. 
This can be used to work on a port update while building other stuff
without risking the global build from interfering until the update is
finished.
 2008-01-12 14:12:43 +00:00
espie
9ca2e064cf do the same thing with REGRESS_FLAGS as FAKE_FLAGS 2008-01-04 18:38:51 +00:00
espie
17d70806a3 tweak FAKE_FLAGS semantics to saner defaults. 2008-01-04 17:48:33 +00:00
espie
e0c15c5448 scaffold for running make in parallel, to be documented soon. 2007-12-28 12:49:12 +00:00
espie
4cdccf971e add a PKGDB_LOCK that can be used by pkg_add to -F nolock in src/ build 
mode.
 2007-12-28 12:46:03 +00:00
jolan
72e58f4c30 support .tbz2, ok espie@ 2007-12-05 06:55:41 +00:00
merdely
32c84fc9c3 Create $PLIST_DB directory if defined. 
Fix a spacing nit.

ok espie@
 2007-12-01 14:44:47 +00:00
sturm
6438da2ade back out, this breaks PLIST_DB 2007-10-29 21:50:40 +00:00
sturm
09b5470166 unbreak, the regular expression never matched and $' was empty there 
ok espie@
 2007-10-29 20:00:53 +00:00
steven
19811ccb50 protect FLAVOR with :Q when passed to the shell, since it may contain 
whitespace.
 2007-09-30 15:07:40 +00:00
steven
bad04c5480 pass FLAVOR to update-plist. 
ok espie@
 2007-09-21 08:04:06 +00:00
steven
9cb9ca7c2e use :Q for BROKEN 
ok sturm@
 2007-09-21 06:58:21 +00:00
naddy
1715498931 quote shell meta characters when generating readmes; ok espie@ 2007-09-16 21:38:34 +00:00
espie
7a9f1afb15 add message that shows we exited a directory after an error, stops log there 
during rebuilds.
 2007-08-25 07:56:04 +00:00
espie
fb338f26c3 make mirror-maker sturdier: 
- pass an argument for the file into which we want to save the result, to
avoid getting it polluted with error messages (defaults to /dev/stdout for
debug).
- create the makefile fragment as a temp file and only copy it when complete.
- copy it in one chunk, so that one can read a partial mirror-maker file
and have it be usable.

This does allow for people to start a make mirror-maker in one shell, and
start fetching stuff right away, before mirror-maker is finished.

This also produces usable mirror-maker Makefiles even if the ports tree
contains bogus entries.
 2007-07-28 12:58:34 +00:00
espie
a52a0f1494 make sure ARCH gets defined also for individual port. 
Noticed by Stuart Henderson
 2007-07-09 13:32:56 +00:00
espie
43c841c254 avoid recomputing ARCH all the time. 2007-07-08 17:57:56 +00:00
simon
641ac136d2 let MODAPACHE_FINAL point to the correct location 
noticed by & ok steven@
 2007-06-30 15:27:29 +00:00
espie
33473a3fbb better, make it fetch file under DISTDIR 2007-06-30 14:48:04 +00:00
espie
b96d64afbf debug target for mirror-maker: allows one to check quickly whether a given port 
generates correct mirror-maker fragments.
 2007-06-30 14:43:49 +00:00
espie
3c826c4a33 rework the way subdirs are skipped so that they're skipped a the right 
location.
Allows stuff like make ... STARTDIR=x11/openmotif SKIPDIR=x11/openmotif
to work.
 2007-06-30 14:31:00 +00:00
espie
07262849b8 add support for FTP_KEEPALIVE. 
You need a somewhat current ftp(1) for this to work !!!
 2007-06-29 10:24:23 +00:00
espie
9eb5a6bae5 tweak getpkgpath to display a diagnostic, but not exit 1, in case of 
error.

Do the real erroring out through a fatal error in pkgpath.mk (and set
PKGPATH to something less confusing than an empty variable). This
solves PR 5510 for real.
 2007-06-28 21:05:47 +00:00
steven
eb5a2e2d25 log results of regression tests into a file. 
ok simon@
 2007-06-23 09:39:18 +00:00
simon
ac9f495cdd set MODAPACHE_ENABLE and MODAPACHE_MODULE conditional 
for apache modules with mixed interal/external names

ok espie@
 2007-06-22 19:07:33 +00:00
ajacoutot
e6e9840691 - set SHARED_ONLY 
ok simon@ espie@
 2007-06-21 07:37:58 +00:00
simon
8055ea218d adjust the comment according to changed subst_vars 
ok espie@
 2007-06-21 06:11:11 +00:00
espie
7d0d0ae02a don't substitute MODAPACHE_NAME, this is often too short and causes 
too much trouble
 2007-06-20 14:43:17 +00:00
espie
5e5ec2f584 fix name of installed file 2007-06-20 13:55:41 +00:00
espie
b869babbd8 glue to simplify installation/updates of apache modules 2007-06-20 12:23:17 +00:00
jakemsr
28e134ba9c pass $PORTSDIR to ${PORTSDIR}/infrastructure/package/check-lib-depends 
"Sure" espie@
 2007-06-18 23:04:02 +00:00
espie
343aff0c83 tweak FETCH_PACKAGES: do not fetch all multi-packages to satisfy a 
dependency, but only the ones actually required.
If we end up needing to rebuild them from source, then change back
to _internal-package-only, so that we benefit from the whole build
and get all packages again.
 2007-06-16 09:57:03 +00:00
espie
b257e4d40f put all phony targets inside a variable, so I can check quickly that those 
targets do still exist.
 2007-06-04 12:15:09 +00:00
espie
479b2ddcdc since we no longer have to tsort the output of _print-package-args, 
we can compute it first, and only start pkg_create if it didn't error
out.

This gets rid of the very verbose and confusing error messages
pkg_create meets when the ports tree cannot solve some dependencies.
 2007-06-03 22:30:25 +00:00
espie
7f62033c1e zap duplicate lines 2007-06-03 22:25:01 +00:00
espie
1196a0d7f3 reorganize to avoid repeating lists of targets: 
classify stuff that is used in bsd.port.subdir.mk, name them in pkgpath.mk,
and use them as .PHONY targets in both bsd.port.mk and bsd.port.subdir.mk.
 2007-06-03 12:51:59 +00:00
espie
b4a526a3bb tell make-plist about .saved-libs 2007-06-03 11:06:41 +00:00
espie
8aed1af7f0 Let lock save the BUILD_PKGPATH. 
Do port-lib-depends-check in two steps: first generate a list of
binary: lib1,lib2,lib3
in WRKINST/.saved_libs
then run check-lib-depends on it.

Speeds up tweaks to WANTLIB quite a lot, as we do not rescan every
binary all the time...
 2007-06-03 11:03:06 +00:00
espie
4fb932ea5d finish renaming check-newlib-depends -> check-lib-depends 2007-06-01 13:15:21 +00:00
espie
3eae0fe94a kill REQ script support, flag INSTALL/DEINSTALL as warnings. 
Remove obsolete _PKG_PREREQ cruft.
 2007-05-31 10:52:16 +00:00
espie
1a7def9f37 use the same rules to represent comments in describe that are used to 
build packages.
 2007-05-27 11:53:39 +00:00
espie
f693d15fb3 use the new features of pkg_create. A few packages may no longer build 
without a packagename bump, or because they have too long comments.

steven@ and I cleaned up most of them.

As a result, there's no longer any WRKPKG directory with their temporary
files.

We also use the `sort -u' feature of pkg_create for dependencies, so that
the command lines to pkg_create get simpler, which will allow for easier
error-checking later on.
 2007-05-25 13:07:41 +00:00
espie
9fd3895989 fix infinite recursion for p5-version like we already did for modbuild 
proper... nothing new under the sun.
 2007-05-22 20:57:25 +00:00
steven
f1aafbc927 add clean=plist to clean a port's registered plists. 
fine with bernd@, ok espie@
 2007-05-21 11:18:10 +00:00
sturm
7c066d7098 fix the recursive case for FETCH_PACKAGES by appending the PACKAGE_COOKIE 
to _TRIED_FETCHING, this broke in 1.774

ok bernd, espie
 2007-05-01 17:17:54 +00:00
espie
77b04141fd pedantic fix: typo without actual consequences. 2007-04-30 12:44:46 +00:00
espie
e55de40a3f handle exact matches on STARTDIR. 
allows `partial' STARTDIR, such as make describe STARTDIR=x11/kde
to work as expected.
 2007-04-28 10:19:41 +00:00
espie
2e20da7e2c make it work with >2 directories in PORTSDIR_PATH. 
From Mikolaj Kujars
 2007-04-17 15:22:46 +00:00
espie
e4dcfe629b avoid Makemaker going interactive, okay naddy@ 2007-04-08 16:16:06 +00:00
espie
7c076f597b port-lib-depends-check should ignore errors, because it's much less useful 
that way.
 2007-04-08 11:26:39 +00:00
espie
04b97b2f22 STARTDIR functionality: allows one to skip directly to a given directory 
and start from there.
 2007-04-07 09:55:13 +00:00
espie
97e872a96d sha256 by default 2007-04-05 18:32:26 +00:00
espie
064be6c9e4 revert order of tests, if CHECK_LIB_DEPENDS is active, do not register plist 
until we're satisfied.
 2007-04-03 13:32:45 +00:00
espie
23c2d177a9 allow people to run check-lib-depends automatically, grabbing the files 
from WRKINST and the packing-list from the just built package.
Reuse the register_plist pattern.
 2007-04-03 10:14:14 +00:00
espie
1cca659066 configure may not exist. found oud by steven@ 2007-04-02 10:43:39 +00:00
espie
f020e9e32b new target: port-lib-depends-check, can verify libraries directly from the 
fake area, without needing to build bogus packages first.
 2007-03-31 15:36:43 +00:00
espie
5bc25e411a display warnings if autoconf version differs. 2007-03-30 13:55:07 +00:00
espie
949afb38a5 like xenocara: REORDER_DEPENDENCIES has been tested enough, silence it 
by default.
 2007-03-30 13:44:50 +00:00
espie
c6897dc077 CHECKSUM_FILE never changes -> can be used directly 2007-03-30 13:41:44 +00:00
espie
634644d8e4 tweak the way mirror-maker emits stuff, so that the Makefile contains lines 
suitable for cksum -c.
 2007-03-28 15:45:03 +00:00
espie
fb9fcd5fbc use cksum -c so we can get `magic' base64/hex matching. 2007-03-28 13:21:43 +00:00
espie
bc0a71b1bb remove stupid hack that makes _print-packagename display README on 
subdirs. Confuse the hell out of broken dependencies, and is not
even needed.

fix typo.
 2007-03-19 21:32:35 +00:00
naddy
2c2c250df3 makesum: don't try to create distinfo if there are no distfiles; ok espie@ 2007-03-15 18:11:54 +00:00
espie
b5a3fca2a6 use _CKSUMFILES in clean=dist, and recurse if we notice there are 
SUPDISTFILES, so we get them as well.

This misbehavior noticed by Mikolaj Kucharski.

(the intention is obviously to regrab everything to verify whether
anything changed, and that includes SUPDISTFILES)
 2007-02-16 19:08:54 +00:00
espie
3b1db2b2c1 use cksum -a $$cipher instead of $$cipher to compute crypto hashes. 
Only functional change:
Add sha256 at the end of the list of supported ciphers, so that new
makesums will add the new cipher.

The rest is business as usual.
 2007-02-11 11:44:17 +00:00
espie
2c0953c14e minor tidbit: define FULLPKGPATH based on FULLPKGPATH${SUBPACKAGE} in 
the multi-packages case, as should be.

introduce _ALLPKGPATHS, and use it in show-required-by, so that
show-required-by works for multi-packages.

problem noticed by bernd@
 2007-02-06 20:04:01 +00:00
espie
50ea26a11c rely on new make feature, avoid old work-around (need a current make to 
work)
 2007-02-03 17:09:39 +00:00
espie
fa6fc8a2f7 sometimes (not very often), config.h.in files live in any directory 2007-02-03 17:08:59 +00:00
espie
a7cedd9db7 add the ability to cleanly delimit embedded plist info. 
will help unconfuse lib-depends-check and make-plist wrt error messages.
 2007-01-15 13:43:52 +00:00
bernd
c5e1d48b7a Fix typo. This unbreaks 'make update'. 2007-01-04 11:34:14 +00:00
espie
9f019509e8 prefer subpackage-dependent version of variables for substitutions. 
this lets the correct PREFIX be substituted in packages, among other things.

Problem discovered in php5-extensions
 2006-12-31 13:12:35 +00:00
espie
8b31dd60d8 use a simple perl script to compute initial pkgpaths... Cwd can replace 
pwd -P, and we select the longest prefix from PORTSDIR_PATH.

This lets mystuff-like dirs get stripped from PKGPATH correctly, as long
as PORTSDIR_PATH is set correctly.
 2006-12-26 15:00:24 +00:00
espie
5efd6b68e4 duplicate the print-plist targets, add a version -with-depends, which is 
slower, but shows dependency lines as well...

Also fix target evaluation to work like it does elsewhere
 2006-12-18 12:52:34 +00:00
espie
25c626b21e activate our pkg-config. 2006-12-16 11:53:46 +00:00
espie
b3ba10ebac repair pkg_mklocatedb from src 2006-12-11 15:52:14 +00:00
steven
e3820f5171 DEPPATHS per subpackage, ok espie 2006-12-11 14:01:03 +00:00
espie
c89eab44fc pass all subpackage information to make-plist. 
Adapt make-plist to use the same code as pkg_create to deduce fragment
names, remove some old special cases that should no longer matter.

Zap directories from dependencies in a way depending on the dependency.
Far from perfect yet, as directories should be registered multiple times
in packages that do not depend on each other, and could also be used to
figure out where to put new files...
 2006-12-11 13:36:06 +00:00
espie
9d1d088f4c missed a few = signs. 
Avoid writing anything when no libspecs.
 2006-12-11 11:05:43 +00:00
espie
5792254153 Add proper "Returning to build of..." to unconfuse the logger in the case 
of FORCE_UPDATE.   Also break out in case update fails (otherwise the old
installed package would be enough to continue building).

Sprinkle REPORT_PROBLEM to know more precisely what stuff is breaking.

replace some
if eval cmd; then
with
if (eval exec cmd); then

because if eval... seems to break early, even with the test guard, thus
giving a chance to REPORT_PROBLEM to do stuff instead of having to wait.

(Note: a trap would probably be simpler ?)
 2006-12-09 14:56:41 +00:00
espie
1afd447f95 wrap calls for _print-package-args to print reliable errors (better than: 
`some Makefile somewhere is broken ! find it')
 2006-12-08 10:19:08 +00:00
espie
a9bf5ac052 ending up with an empty SUBPACKAGE is an explicit error. 2006-12-05 19:23:42 +00:00
espie
26075e9de6 move TMPDIR to pkgpath.mk to allow its use from both bsd.port.mk and 
bsd.port.subdir.mk

Use absolute paths to build readmes files, based on READMES_TOP, which
can be (or not) PORTSDIR.

Generate category readmes in tmp directory as well.

minor clean-ups
 2006-12-02 11:27:46 +00:00
espie
b219b82d19 move the generated readme stuff to a temporary directory, much cleaner that way. 2006-12-02 11:08:49 +00:00
espie
6f8c254c9e fix READMES again 2006-12-02 10:27:40 +00:00
espie
8fb5e71cef clean-up spacing 2006-12-02 00:10:24 +00:00
espie
1259dc14cb let license-check recurse correctly, thus we are checking ALL licences, 
finally.

Remove work-around for undefined SUBPACKAGE stuff, since we're no longer
hardcoding it at the Makefile level, and thus the ERROR check works.
 2006-12-01 17:56:38 +00:00
espie
5f5c120220 use the pkg-config in /usr/local preferably, until we get a decent 
pkg-config in /usr/bin...
 2006-12-01 17:37:15 +00:00
espie
197d308a2f remove the last redirectors, regenerate readmes on the fly... 2006-12-01 17:33:16 +00:00
espie
0e7d3606d9 make it possible to override a FULLPKGPATH if you REALLY know what you're 
doing... DO NOT USE, unless you really know.
 2006-12-01 11:34:04 +00:00
espie
dfa3a77ddb fix describe 2006-11-30 23:08:07 +00:00
espie
a8356d9ae2 un-recursive dump-vars, remove vars that don't make a lot of sense now. 2006-11-29 09:42:08 +00:00
espie
c72f6e29fd hidden SUBPACKAGE in _register_plist: put it back into ${_PACKAGE_COOKIE${_S}} 2006-11-28 23:13:30 +00:00
espie
6d481f3f77 apply the same transformation to install/install-all 2006-11-28 20:40:55 +00:00
espie
cd97d2252c move a few things around inside one single MULTI_PACKAGES loop 2006-11-28 20:31:25 +00:00
espie
46484d5a90 fix a typo 
mv some PKG_ARGS where they belong, and s/SUBPACKAGE/_S
 2006-11-28 20:26:38 +00:00
espie
fe36feb551 remove a lot of redirectors, add a lot more loops, build stuff based 
on internal variables mostly.
 2006-11-28 20:20:25 +00:00
espie
fda8ee6238 let PKG_ARGS be SUBPACKAGE dependent, define them all 2006-11-28 19:59:15 +00:00
espie
4d572376b6 fix for xview 2006-11-28 19:13:47 +00:00
espie
0f300c1d5c big internal changes: if MULTI_PACKAGES is not defined, define it to '-', 
so that the `normal' cases is MULTI_PACKAGES, (with possibly one
special '-' subpackage).

Adjust a few tests accordingly so that people don't notice the '-'
SUBPACKAGE: mostly do not print some directory changes, adjust FULLPKGPATH,
and set up PLIST- templates correctly.
 2006-11-28 18:25:42 +00:00
espie
2e52dd7f15 display something when switching directories during updates 2006-11-28 18:17:57 +00:00