Commit Graph

1196 Commits

Author SHA1 Message Date
espie
dd6c1654d5 fix stupid typo.
introduce WANTLIB.
2004-11-21 11:06:38 +00:00
espie
5982ed59bb new support routine _print-package-args, that prints new-style package
dependencies.
_syslib_resolve_fragment that can resolve system libs.
2004-11-21 10:43:48 +00:00
espie
b8dfc30a1a don't report false positives for libraries with similar names 2004-11-21 10:39:24 +00:00
sturm
ec0870206a use TRUEPREFIX locally
sort SYSTRACE_SUBST_VARS while here
2004-11-19 22:12:28 +00:00
espie
723c37630f can't depend on WRKDIR unconditionally, throws update off when
UPDATE_COOKIES_DIR is set.
2004-11-17 11:09:44 +00:00
mbalmer
bce2473cd5 Added uid/gid _vilter for mail/smtp-vilter.
ok naddy@, alek@
2004-11-17 10:45:25 +00:00
espie
3979adeb86 add missing stuff around update so that BULK works for it too. 2004-11-17 10:38:37 +00:00
espie
65134e9f73 only pass FULLPKGPATH, PERMIT_PACKAGE_CDROM, PERMIT_PACKAGE_FTP to
pkg_create, and quote the later properly...
2004-11-16 02:08:59 +00:00
espie
1b8f6e0f0c Pass PERMIT_PACKAGE_CDROM, PERMIT_PACKAGE_FTP, FULLPKGPATH as parameters
to pkg_create... This should allow the automatic generation of ExtraInfo
without needing to do it manually.
2004-11-15 16:31:28 +00:00
espie
e39dc1ae33 mark update as .PHONY target, necessary with UPDATE_COOKIES_DIR ;-) 2004-11-15 13:53:43 +00:00
espie
147d3cf94a industrial-strength updates:
- separate UPDATE_COOKIES_DIR, unless set to empty.
Makes for easier management.
- FORCE_UPDATE variable, that makes certain packages are updated when
a dependency is run, and that forces the replacement of packages that are
already there.  Still uses the update cookie, so that this update occurs
just once unless you wipe out yur UPDATE_COOKIES_DIR (which justifies putting
update cookies in a separate directory).
2004-11-15 13:52:03 +00:00
espie
54a74ef157 use pkg_info -q -e instead of pkg dependencies check now that it does
what it should.

Use the exact same rules for checking LIB_DEPENDS that will later be used
to build the package itself, e.g., add the pkg dependency check.
2004-11-15 11:49:05 +00:00
espie
b6797baf63 don't try to register anything that isn't at least a FileObject...
Found out by alek@
2004-11-14 19:42:04 +00:00
sturm
78074d531a add _maradns 2004-11-14 11:59:31 +00:00
sturm
7a4d2974e7 sparc64.p -> localhost 2004-11-14 11:56:07 +00:00
sturm
0148b33be9 improved SMP handling
- automatically detect number of cpus per host
- differentiate between hosts and nodes
  => each cpu is its own node
  => ports are build on nodes now
  => multiple host entries are no longer necessary for smp machines
2004-11-14 11:46:15 +00:00
sturm
9dca1f7475 a little cleanup in preparation for improved SMP handling
print() -> warn() where useful
2004-11-14 11:25:53 +00:00
espie
f4978ac9af synch with changes 2004-11-13 11:49:48 +00:00
espie
e7dd701872 synch with packing-list changes 2004-11-13 11:49:03 +00:00
espie
8af46cc105 fix cwd issues.
try harder to resolve links
2004-11-13 11:25:31 +00:00
alek
0caade6762 User and group _xcept are now created automagically.
looks good pvalchev@
2004-11-13 01:52:25 +00:00
espie
5dc2fdab3a tell all big analyzers they can forget about the packing-list. 2004-11-11 15:38:02 +00:00
espie
d9536a7ac4 rename _UPDATE_COOKIE to always be defined (wilfried) and to have a name
changing when the FULLPKGNAME changes.
2004-11-10 23:40:42 +00:00
espie
f2ae2fd4b7 allows it to recurse. 2004-11-10 10:23:42 +00:00
espie
eb3edb001b Add a simple update target, that relies on pkg_info -e pkgpath to know
if there is something to update...
2004-11-10 10:16:41 +00:00
espie
1c062c3c2e allows for updates 2004-11-10 10:15:42 +00:00
pvalchev
0a88d9c9c2 amd64.ports is now SMP 2004-11-07 18:32:01 +00:00
espie
e94445485e provide a simple locking command in perl. 2004-11-07 10:09:35 +00:00
pvalchev
68c791bb78 add mips64 to LP64_ARCHS 2004-10-26 06:16:53 +00:00
jolan
1ae09c843d sync sourceforge mirrors with the list on their site, omitting two out of
sync mirrors
2004-10-24 18:21:09 +00:00
sturm
bc39f7be50 new option -c for "clean builds", i.e. remove all installed packages
before building a new one; don't use this when building multiple packages
on one host
2004-10-24 14:16:38 +00:00
espie
8c4d857d3d allow us to specialize DEPENDS_TARGET depending on whether this is
a LIB/RUN/BUILD/REGRESS_DEPENDS.
2004-10-23 10:02:51 +00:00
espie
97804eddfe switch FAKE's test from == "yes" to != "no" 2004-10-23 09:58:13 +00:00
espie
0d09be0752 shared fragment in need of recording.
found by wilfried@
2004-10-23 06:51:54 +00:00
espie
63f4447086 avoid more get*_r functions until we decide upon an API... 2004-10-21 20:47:38 +00:00
robert
cba5c3b7b6 Missing support for a module should be a fatal error. So don't continue
the build if you miss a module.

ok espie@
2004-10-18 15:17:17 +00:00
espie
b5445b3596 forgot to copy @newgroup and @newuser items over.
noticed by nikolay.
2004-10-17 09:20:06 +00:00
espie
4ddb942744 getservbyname_r doesn't work/ doesn't have a prototype/ will vanish soon
from libc.

Make sure we don't pick it up in the interim.
2004-10-13 17:23:38 +00:00
jolan
c20baa5bed +_nostromo 2004-10-13 06:29:22 +00:00
espie
1bbdaa1cbd UNMESSAGE 2004-10-11 10:32:04 +00:00
sturm
8511b00703 add TRUEPREFIX to SYSTRACE_SUBST_VARS 2004-10-08 19:58:30 +00:00
espie
b71495c6f2 use base classes more 2004-10-05 20:40:11 +00:00
espie
3b550a0182 config.site support 2004-10-03 09:51:21 +00:00
espie
dfe26ec92e prime autoconf scripts with useful stuff.
okay naddy@
2004-10-03 09:50:44 +00:00
jolan
df8ad03ef6 scanlogd is auto now 2004-09-28 01:57:51 +00:00
sturm
82dfd19593 add 'patch' to handled targets 2004-09-26 09:59:14 +00:00
naddy
1db4c1063b Changes for symon 2.65 - 2.67:
- removed netiso depends
- _symon uid 535, 525 never made it into ports/infrastructure/db/user.list
- ntp clockadjusts no longer trigger extra measurements
- daemon parts ported to freebsd/netbsd/linux

From: Willem Dijkstra <wpd@xs4all.nl>
2004-09-21 16:44:04 +00:00
espie
e4b643ad88 don't create patches for Oops.rej 2004-09-21 09:27:15 +00:00
sturm
ea2b00d80a $OpenBSD$ 2004-09-20 20:17:33 +00:00
sturm
1f2b8ae4df - use sh on localhost* and don't check_host() on these
=> you have an SMP machine, use localhost0, localhost1, ...
- for remote sessions, use ssh multiplexing, this speeds up a dpb
  restart by about 1/4 - 1/3
- new option '-s', which starts building in cwd
  => use 'dpb -s' instead of 'make package'
  => mostly useful for SMP machines or if you have your cluster powered
     on all the time
- '-b' now forces creation of the dependency file, if it doesn't exist,
  it is created automatically
- check signals on returning ssh sessions to notice which ones really
  succeeded and which ones where killed
2004-09-20 18:57:53 +00:00
pvalchev
be6617f5f7 change CDROM_SITE to default to empty; ok espie 2004-09-19 09:00:15 +00:00
espie
784afb2649 kill old check-libs script, after checking the new one handles a.out
just fine.

Kill old subst that's not used anymore as well.
2004-09-18 13:48:43 +00:00
espie
0046b33eec Strip spaces from CFLAGS everywhere.
Pass off -a to pkg_add if used to solve dependencies.
2004-09-18 13:45:23 +00:00
espie
292285c3fe clone new-style conflicts too 2004-09-18 11:03:45 +00:00
espie
fb42eebbe1 new check libs that should work with a.out, and uses the current framework
for accessing plists.
okay naddy@
2004-09-18 08:12:11 +00:00
espie
b14b123e17 WRKCONF?=${WRKBUILD} 2004-09-15 18:58:49 +00:00
espie
613aa9a6c4 remove all SED_PLIST work, since pkg_create handles it now.
defines appropriate PKG_ARG arguments instead.
Move message to -M ${MESSAGE}.
2004-09-15 18:57:31 +00:00
pvalchev
def59b4c36 update list of arches for 3.6 2004-09-15 18:57:16 +00:00
pvalchev
c862b93395 ignore "index.txt" file found in ftp directories 2004-09-15 18:53:24 +00:00
pvalchev
2a5411ff4c had to get changed 2004-09-15 18:47:47 +00:00
espie
d67be4b0b9 remove trailing spaces from CFLAGS/CXXFLAGS, to please autoconf... 2004-09-14 23:07:20 +00:00
espie
907ba20320 explicitly forbid flavors starting with [0-9]. 2004-09-14 23:06:02 +00:00
espie
eb04306537 handle -> check_common_dirs
use visit
accurate list of existing directories with Mtree.
build better results.
2004-09-14 22:31:48 +00:00
espie
67877397d3 visit in the base system. 2004-09-14 22:30:36 +00:00
espie
91d926e81f - lots of comments
- Mtree module, shared with check-common-dirs
- update list of `known' directories to match actual OpenBSD, so
that stuff outside of /usr/local will work.
- drop dead if destdir does not exist (e.g., fake got a problem)
- if a file is found in a fragment, and as a comment in the original
file, then that was a limitation of the old make-plist: don't look at
the comment.
- copy modes for @sample
- do fragments for multi-packages too.
- pass existing files around for more accurate @exec tests.
- don't copy install-info/mkdir if files/dir have been found.
- register dirrm dir as dir/, but using original object, so that it will
need to be found once to be accounted for.
- don't register extra as `files', otherwise we end up with bogus
${SYSCONFDIR}/file entries.
- classify directories and files in a unified way.
- more thorough tests for info files, more names, look at two lines context.
- ibid dir file.
- ibid subinfo.
- provide a way to erase a directory that only holds one file.
- use for .packlist, perllocal.pod.
- use perl Config to look for those in the right location.
- don't look at file dates, we always use fake.
- special case /etc/X11/app-defaults becomes /usr/local/lib/X11/app-defaults
- use plist state info to avoid extraneous mode/owner/group entries.
- sort comments that are not accounted for.
- warn about .orig files.
- man pages for linux.
- don't backsubst SYSCONFDIR except at start of line.
- handle @comment @dir like @comment @dirrm.
- respect fragment locations
- put %%SHARED%% at top by default.
- handle recursive fragments correctly (simpler code, needs pkg_add support)
- mark genuine fragments, and don't try to write empty ones, assuming
the chosen flavor doesn't hold the required information.
2004-09-14 22:24:39 +00:00
espie
438553ff5e check lib dependencies on binary packages. 2004-09-14 22:22:17 +00:00
pvalchev
5b5b01e1ba grow slightly 2004-09-02 16:58:02 +00:00
pvalchev
7930bb72d2 more tweak 2004-08-24 22:35:13 +00:00
pvalchev
15a2294e9f more tweaks, can grow this 2004-08-24 21:40:30 +00:00
sturm
1f1c3ee467 wrong order when redirecting STDOUT/STDERR
ok pval
2004-08-24 20:30:07 +00:00
pvalchev
2104c574bf no longer interactive at pkg_add * 2004-08-24 19:17:32 +00:00
sturm
0c0f93c97d raise default timeout to 1 minute
requested by pval@
2004-08-23 20:46:42 +00:00
sturm
91eb345927 improve reliability:
- don't invoke intermediate shells on exec(), allowing to actually kill children
- when we are going down, only kill remaining children, not the whole pgrp
- don't use a SIGCHLD handler, but call waitpid explicitly at the right places
- exit(1) in case of an error

others:
- a little cleanup
- new option -t <timeout>

ok pval@
2004-08-23 19:54:11 +00:00
pvalchev
dcd774811e grow 2004-08-23 07:12:34 +00:00
pvalchev
1eb61e6684 expand 2004-08-23 06:46:21 +00:00
pvalchev
fb1323e187 adjust 2004-08-17 21:03:44 +00:00
pvalchev
c508632507 sync 2004-08-14 01:15:50 +00:00
pvalchev
8c30f51213 shrink by a lot 2004-08-14 00:23:35 +00:00
pvalchev
b06c4bcaef ignore files generated by check-plist 2004-08-14 00:23:00 +00:00
pvalchev
74efe665a3 sync 2004-08-13 23:48:48 +00:00
brad
2545045f21 Side-effect of BULK=Yes: WRKDIR gets wiped out before the depend cookie
is written.

So, we simply recreate WRKDIR manually.
This is a kluge, but it should do the trick.

From espie@

ok pvalchev@ espie@
2004-08-13 23:28:40 +00:00
pvalchev
7fade6547c bash2->bash; naddy 2004-08-13 21:00:29 +00:00
espie
2d48535d1e don't output pkgname if no problem to output.
okay pvalchev@
2004-08-13 07:45:22 +00:00
espie
9560823362 set _MASTER_LOCK in the correct location.
fix postgresql locking *for good*.
okay pvalchev@
2004-08-12 19:02:45 +00:00
espie
b2fffcb71e script to check that all common dirs are properly tagged in every package
that should tag them (e.g., any package without a parent `dependency'
should have the dir marked).

okay pvalchev@
2004-08-12 18:58:34 +00:00
espie
a8174b7541 obvious mistake, FULLPKGPATH for recursive clean, duh!
problem found by brad@
2004-08-11 22:42:47 +00:00
espie
fdb200935c fix postgresql lock issue: in recursive locking, don't relock the same
port.

okay pvalchev@
2004-08-11 22:25:23 +00:00
brad
e12fe83b57 mips stuff 2004-08-11 20:45:38 +00:00
espie
e99d16b029 Use original item to build new one: so we can clone things we do not yet
detect, like @mandir and @fontdir.

more precise logic for changes: do not check for orig files and die if
nothing changed.
2004-08-11 09:40:17 +00:00
espie
7413f2d2da better diagnostic logic:
first output all PLISTs in a tmp dir, then compare with
original set, and report all differences.
Then check that original can be moved to orig, and do the move.
That way, make plist on a set of unchanged lists will be silent,
and can be used irregardless of the presence of P*.orig files.
2004-08-10 21:01:01 +00:00
pvalchev
aa821a5387 sparc cluster too 2004-08-10 20:57:03 +00:00
sturm
02a925a3c7 Distributed Package Build is a helper script to permit several hosts
working together in doing a bulk package build.

ok pvalchev@, espie@
2004-08-10 20:54:38 +00:00
xsa
a3237a6832 prepare for new-style MODULES. ok sturm@. 2004-08-10 20:39:37 +00:00
espie
9ab9eeaa46 prefer pkg_create -B 2004-08-10 13:48:21 +00:00
espie
a6ddcef28f register sample dirs. 2004-08-10 08:45:21 +00:00
xsa
9d761fa2f0 prepare for new-style MODULES; ok dons@. 2004-08-10 08:14:34 +00:00
xsa
70c60882e0 prepare for new-style MODULES; ok couderc@. 2004-08-09 19:54:20 +00:00
espie
9503cf0454 more accurate info test. 2004-08-09 12:48:56 +00:00
espie
0b2cdd25b2 next -> return 2004-08-09 12:41:03 +00:00
espie
2cce78f090 next->return 2004-08-09 12:35:45 +00:00