- provide cpp wrapper using cc -E
- provide c++_r wrapper working like c++
- provide -rpath handling for cc
- fix some debug messages to provide the correct script name
as a modifier on another variable and would need to be escaped
as :M${_dep_\:C/\\\:.*$//} in order for the modifiers to be parsed correctly.
This all used to work because _dep_ is a .for loop control variable which
was substituted in an earlier pass through the line.
Recent changes leave .for loop control variables to be substituted during
normal expansion.
Adding the \ would leave the script non-working with theolder make.
OK'd by Beefy
- Add cc_r/xlc_r wrapper using the same rules as xlc itself.
It is used for example by Perl.
- Improve the RPATH emulation:
- Always set -blibpath, use /usr/lib:/lib as default
- If -blibpath is exlicitly given, add to the default
- Additionally append any -Wl,-rpath given.
The PKG_FAIL_REASON isn't executed by default, no idea why, but this
makes it a bit cleaner why the compiler wrapper will fail and fixes the
more important AIX case.
1.) Create a user and group "mediatomb" for running the server to restrict
the files which can be accessed.
2.) Add a startup script to start the server automatically.
Bump package revision because of these fixes. Changes approved by
Alistair Crooks.
- Use PGKSRC-WARNING to note it's coming from pkgsrc and not always fatal.
- Describe more precisely what's happening when you get this warning.
Hopefuly this will stop the misundersanding of the error message I could
see quite often amongst users.
Don't call pkg_info to get the installed Emacs version; always use the
version matching EMACS_TYPE set by users. Be DEPENDS to it. This should
address pkg/37146 by Aleksey Cheusov.
While here convert some emacs lisp packages to user-destdir.
Ideally, X11_TYPE modular would set up such packages so that
PREFER.<BUILTIN_PKG> is set to "pkgsrc", but it's a more complex change.
Reported by Adam Hoka.
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.
support for creating empty files as CONF_FILES.
The usual way is to add
CONF_FILES= /dev/null /some/file
However, some parts of the infrastructure check if the "source" is a
file -- this fails for /dev/null obviously (other parts accept
character devices already).
Fix this. Will follow up with PKGREVISION bumps for affected packages.
Ok during freeze: agc@
Instead, let curses.buildlink3.mk find out a usable one.
This fixes the error reported in PR 39531. However, the PR will remain
open for now because it raises another issue whether tnftp should depend
on curses or not.
- add missing modules to DFLT_APACHE_MODULES
- bump PKGREVISION for new modules
- replace APACHE_MODULES=all-shared with apache-shared-modules option
- '--enable-mods-shared=all' is not sufficient, that's why DFLT_APACHE_MODULES
lists all those others; use DFLT_APACHE_MODULES instead of 'all'
- add apache-shared-modules to options.description
approved by tron
It was uncompleted and now service suspended, noticed by Zafer Aydogan.
While here, update MSTER_SITE_MOZILLA taken from mirrors list:
http://www.mozilla.org/mirrors.html
as we use pdksh on this platform.
Also use /usr/bin/ginstall when it's present (it is on OpenSolaris)
for the same reason.
This problem was reported by Andras Barna.
the full requirement rather than just the package name. This message
should never be seen (after all, the package we need is supposed to
*get* installed) but sometimes if things are screwed up in one way or
another it does show up. Since often what's wrong is that the package
that's installed is the wrong version, not that it's missing entirely,
this way the error message makes a lot more sense.
E.g. http://mail-index.netbsd.org/tech-pkg/2008/06/12/msg001126.html et seq.
ok dillo@
build and kill bmake. This isn't really a great solution as it means it
won't get built on any platform, maybe we should just make pdksh a bmake
dependancy on every OS.
by default. Previous behaviour would try to echo /<pkgpath> which leads to
interesting behaviour when /net is an autofs map...
Reduces runtime of mk/bulk/printdepends by 5.25 days on my test system.
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.
You must *not* use "test ... && ..." when you use "set -e". Because if the
first expression fails your shell script will abort.
This should fix problem with NetBSD-current's (correctly behaving)
"/bin/sh" report on current-users by David Holland and Kurt Schreiner.
patches to add it). Drop pax from the default USE_TOOLS list.
Make bsdtar the default for those places that wanted gtar to extract
long links etc, as bsdtar can be built of the tree.
version of tnftp, otherwise it will use net/tnftp. Require ftp as
bootstrap dependency when the package has files to fetch.
The dependency is currently unconditional and FETCH_CMD is set to
TOOLS_PATH.ftp by default.
Make _ALL_FILES and in turn _CKSUMFILES available earlier.
Use it to only depend on digest if _CKSUMFILES is non-empty.
NO_CHECKSUM will now only skip the checksum and checksum-phase
targets. FAILOVER_FETCH will independently check the sums on
distfiles and depend on digest, fixing the remaining issues
originally raised in PR 34914.
Make sure _PKGSRC_BARRIER is set when using just-in-time su.
I've been running this for a long time in my tree without ill effects.
For more information see PR pkg/34470 and PR pkg/38505.
environment.
Otherwise BSD style packages will use the install(1) which
bootstrap-mk-files' sys.mk picked, rather than the one from
TOOLS_PLATFORM.install (which may be specified by the user in mk.conf).
targets. Values are now case-insensitive, and the expanded set of values
that can be used are:
Added => add*
Updated => up*
Removed => rem*, rm
Renamed, Moved => ren*, mov*, mv
This changes is purely for usability reasons -- I have trouble remembering
the exact values, and I don't like to type with capitalization unless.
other libraries (specified with -l).
Fixes a linking problem in net-snmp, where the Perl DynaLoader.a library
becomes out-of-order when all -l arguments have been moved to the end of
the command line.
"pkg_admin audit-history" command to print the full list of
vulnerabilities and use further pkg_admin pmatch logic to mark them as
fixed or still open. For pkg_install versions before 20080415, skip the
vulnerability processing.
Discussed with: dmcmahill
anyway, simplify logic a bit:
Add Python 2.5 to the default list, but also mark it as incompatible if
wip/python25 is not present. Move the Darwin handling after setting a
default value.
Provide a new variable _PYTHON_VERSIONS_ACCEPTED that is filtered by
PYTHON_VERSIONS_INCOMPATIBLE. This helps to avoid providing broken
dependencies when a version is not supported as PYPKGPREFIX wouldn't be
set in that case.
PLIST_VARS is the list of names corresponding to automatic variables
generated by plist.mk to simplify having conditionally-present entries
in the PLIST. If "var" is listed in PLIST_VARS, then the automatic
variable is named PLIST.var. If PLIST.var is defined, then in the PLIST
generation, the ${PLIST.var} symbol is replaced with the empty string,
or "@comment " otherwise.
phase pkg_install-depends before bootstrap-depends that just tries to
install a new pkg_install if the current version is too old. Still
keep it as bootstrap dependency for the bulk build code.
For NetBSD, PKG_TOOLS_BIN has to be computed in shell code due to a make
bug.
OK: jlam@
Instead of just looking at the libraries, we check for the headers as
well, and if multiple implementations exist (usually because of symlinks
to shared libraries), then we order the search as:
tinfo, curses, termcap, termlib, c
in bsd.buildlink3.mk was broken with pkg_install-20080309 was it
returned a relative path. It would have failed before e.g. with symbolic
links in the path. pkg_info -E is simpler and was added exactly for this
purpose. Fixes PR 38213 and PR 38211.
code was moved to the builtin.mk file. The buildlink3.mk file should
contain the bits that always apply to all packages that include it.
The builtin.mk file should include the bits that only apply if
"termcap" is listed in ${BUILDLINK_PACKAGES} (this isn't the case
if we use curses to replace termcap). As such, redistribute the
code as follows:
+ Move the parts that remove -l options for terminal libraries we
don't support, as well as for transforming "-ltermcap" into the
appropriate libraries, from the builtin.mk back into the
buildlink3.mk.
+ Leave the parts the remove -lcurses and -lncurses in the
builtin.mk. We can remove the ${TERMCAP_TYPE} == "curses" check
since that part of the file is protected by CHECK_BUILTIN.termcap,
so it should only be triggered if "termcap" is in BUILDLINK_PACKAGES,
which only happens if ${TERMCAP_TYPE} isn't "curses".
package does not use either curses or ncurses. We determine this
by inspecting BUILDLINK_PACKAGES and looking for "curses" and
"ncurses".
+ Because the above logic uses BUILDLINK_PACKAGES, the code must be
moved from termcap.buildlink3.mk into termcap.builtin.mk where it
is safe to inspect BUILDLINK_PACKAGES.
the command line so we don't find any system ncurses library.
XXX This currently causes problems with packages using both termcap
XXX (usually via readline) and curses. This will be fixed in time.
