only there to support this option. Maybe I will reinvent the advanced
autofix code someday, hopefully in a programming language that provides
more error checking than Perl.
"!~ qr" with "!~ m", to work around a memory leak in Perl 5.10.0.
(See Perl bug #59994 or just run perl -e 'while(1){qr""}'.)
This change also speeds up pkglint by around 15 percent, although I
noticed that updating from Perl 5.8.8. to 5.10.0 slowed down pkglint by
about 2 percent.
to trigger/signal a rebuild for the transition 5.8.8 -> 5.10.0.
The list of packages is computed by finding all packages which end
up having either of PERL5_USE_PACKLIST, BUILDLINK_API_DEPENDS.perl,
or PERL5_PACKLIST defined in their make setup (tested via
"make show-vars VARNAMES=...").
o Documentation fix: the blbump script does not print the names of
the package directories it succeeded in bumping, only the ones
which it failed to bump
o Code fix: remove a pointless assignment / substitution in blbump
- Perl 5.10 complains when closedir() is called on an undefined
directory handle, and is correct to do so. Since directory handles are
global variables, they must be used carefully in recursive functions.
- The PLIST.* variables still had some warning suspecting a "spelling
mistake". This has been fixed by marking all PLIST_VARS as being used.
- DESTDIR support
- Add check for inttypes.h and sys/socket.h
- Provide compatibility fallback for netdb.h and the RFC2553 API
(getaddrinfo), supoprting IPv4 only. Code from OpenSSH.
- Use normal autoconf checks for fixed size bit types. Stop using the
older BSD types (u_intXX_t) in libnbcompat.
- Add implementation of shquote (from NetBSD).
- Fix a buglet in the getenv/setenv/unsetenv detection.
- Fix a buglet in the mkdtemp/mkstemp detection.
X.Org found in NetBSD-current.
Thanks a lot to all who helped, especially Matthias Scheler who did
repeated tests on Mac OS X and older versions of NetBSD to make sure the
support for those platforms wouldn't be broken (or at least, not fatally,
as I would still expect a few hiccups here and there, because there is
only so much one can test in such limited time).
On the infrastructure side, this branch brings pkgconfig-builtin.mk, in
order to write very easily new builtin.mk files. It can actually handle
more than just pkgconfig files, but it will provide a version if it finds
such a file. x11.builtin.mk has also been made more useful and now all
existing (and future!) native-X11-related builtin.mk files should include
it.
Sync Dewey with pkg_install(-renovation) and add a fast check if a
pattern could ever match. This reduces the time for pbulk-resolve on a
full tree on my laptop from 12.9s to 5.3s.
Merge from changes for read_plist from pkg_install-renovation to always
initialize the plist and add append_plist for the one case where this is
not desired. Fixes PR 39276.
- version configuration file, object if the version doesn't match
- add an option for pbulk-resolve to ignore unresolvable dependencies
similiar to the incremental mode
- use this option to run pbulk-resolve again if ignore_missing_depencies
is set to yes. Report this explicitly to un-confuse readers and point
them to the logfile.
Merge a number of bugfixes from the pkg_install-renovation branch:
- explicit include of nbcompat/md5.h
- use errx when dealing with libfetch as it doesn't set errno
- avoid optind = 0 as GNUish getopt will reset itself otherwise
Isolate rules to compute the restricted subset in the upload script.
Make the upload script independent from the report file to allow
uploading partial builds.
Only depend on NO_BIN_ON_FTP to decide what to upload and what not.
provided by the system and if one is lacking use the file. Don't short
cut the list as before and do a second iteration, it is just confusing.
This breaks the regression introduced by the last commit on OS X.
- declare endpwent and endgrent if system headers are missing them
- check for existance of statfs.f_flags before using it
- drop a bunch of redundant function checks
Based on the work of Timothy Larson.
Add newer library version as provided by Mac OS 10.5.4. Based on a patch
submitted by Brian de Alwis in PR pkg/39207.
Tested under NetBSD/i386 4.0_STABLE and Mac OS PPC 10.5.4.
Make bulkbuild-rebuild over 200 times faster when running against a full
tree scan by using awk to filter out the data instead of the while read
shell builtin.
- For direct build failures, make the package location in the first
column a hyper link to the last (supposedly failing) phase.
- For indirect failures, list the first 10 failing dependencies.
doing an old-style bulk build on 4.0 or older systems:
- in mk/bulk/build, do "make update" instead of "make install + clean"
when installing a new pkg_install
- in pkgtools/pkg_install/Makefile, don't try to use our own
executables (${WORKDIR}/pkg_add/pkg_add etc.) if they don't exist
Discussed with joerg, and even though he's not entirely happy
with the latter change, he didn't appear to have a better suggestion.
This, and putting /usr/pkg/sbin earlier than /usr/sbin in the $PATH
appears to be required to get an old-style bulk build going.
Fix a bug exposed by the "set -e" improvements in NetBSD-current's
"/bin/sh" which would prevent the script from cleaning up its
temporary directory.
Change approved by David Brownlee.
This package attempts to resolve deficiencies in system headers by
providing missing types, definitions and prototypes in shadow headers.
New in this version:
- Rewrote all of the tests to use autoconf.
- Implemented tests and support for more standard integer types.
nor intends to provide full posix(r) compatibility, I've decided to rename
it.
An improved version of the package will be re-imported later, probably
under the name "compat_headers".
- Explicitly clear automatic flag for packages user wants installed
that got pulled in previously.
- Only ignore pkg_summary.gz in local repository if newer files
with name matching *.t[bg]z exist. (Otherwise, pkg_summary.gz
created by pbulk would be ignored because pkg_summary.bz2 is
newer.)
Okayed by abs@.
Makefiles. As a first pass make MAINTAINER optional and add an optional
OWNER directly after it in the same section. This will pass valid
OWNER/MAINTAINER combinations but not reject some invalid ones.
Full fix (hopefully) coming soon...
Many of the hash function headers are incompatible with the OpenSSL
headers, so don't include them by default and require the caller to
include them on demand.
* Look for $MAKECONF in @MAKECONF@, @PREFIX@/etc/mk.conf,
and /etc/mk.conf, in that order.
* Look for $PKGSRCDIR in the Makefile referred to by $MAKECONF,
".", "..", "../..", and "/usr/pkgsrc", in that order.
* Convert the Makefil to use the SUBST framework.
Proposed in tech-pkg; OK (in principle) David Brownlee
* Look for $MAKECONF in @MAKECONF@, @PREFIX@/etc/mk.conf,
and /etc/mk.conf, in that order.
* Look for $PKGSRCDIR in the Makefile referred to by $MAKECONF,
".", "..", "../..", and "/usr/pkgsrc", in that order.
Proposed in tech-pkg; OK Greg Troxel.
Known regressions:
- "pkg_add -" (aka reading from stdin) is currently not supported
- "pkg_add -s" is not supported either
- no progress reports for the downloads
- binary packages with hardlinks created by pkg_create before
pkg_install-20080422 will not extract correctly (libarchive issue)
- no adhoc check for potential conflicts between dependencies and
already installed packages
Features:
- Twice as fast for the typical case of /var/tmp and /usr/pkg on
different filesystems
- Standalone
- implicit conflict detection before actual installation.
Make use of the linkresolver from libarchive and thereby fix the
permissions of hardlinks in archives. This wasn't a problem so far
as nbpax is (incorrectly) ignoring the permissions of the other entries.
