that location in the new find-headers infrastructure when they have not
been installed into /usr/include.
This allows us to remove the hardcoded builtins, as they can now be
correctly determined automatically.
are instead moved to SDK-specific locations. This breaks many builtin checks
which assume headers are in a fixed place.
Rather than trying to dig around finding where Xcode.app might be installed,
just override IS_BUILTIN for libraries that we know exist.
This file is from libtool-1.x, which is not in pkgsrc any longer
(using libtool-2.x since 2009). Also, it was only used for packages
using autoconf-2.13 AND libtool; I found three packages in pkgsrc with
that combination, and all of them still build on NetBSD-6.99.24/amd64
after this change.
catman pages are installed with a suffix which matches their section
instead of the default '.0'.
Enable it by default on SunOS, which requires that particular layout.
The nroff tool is primarily used to generate catman pages, however there
were circular dependency issues with using groff as the provider, as it
has a large number of dependencies, some of which depend on nroff.
mdocml is much smaller, has fewer dependencies, is BSD licensed, and is
just as capable in nroff mode.
per-platform default. Previously PREFER.<pkg> was used, and as that
has the highest precedence it meant the defaults could not be overridden
with the PREFER_PKGSRC and PREFER_NATIVE user variables.
While here, set the openssl default for SunOS back to pkgsrc, now that
users who wish to use the builtin can do so via PREFER_NATIVE=openssl.
* Correct detection of following cases.
non-editline/readline, editline/non-readline, and editline/readline.
* If builtin editline has header files in include/editline, create
include/readline/* symlinks.
* Fix PR pkg/48062 with above fixes. Confirmed on Ubuntu Linux/amd64 13.04.
Many thanks for obache@.
* READLINE_DEFAULT is depends on builtin editline/readline type if possible.
* _READLINE_ACCEPTED is always "editline readline", both are provided.
Tested on OmniOS (builtin readline-6.2; with some modifications) and NetBSD.
XXX If buitin readline is incompatible, READLINE_DEFAULT is set as readline.
According to devel/readline/builtin.mk, SunOS, Darwin, and Interix's
readline is incompatible with GNU readline.
This behavior should be fixed.
Our make(1) now sets $MAKELEVEL. While this should cause no harm, gmake
detects a non-zero $MAKELEVEL and automatically sets "w" in $MAKEFLAGS
for subordinate makes, in order to print the entry and exit directories.
Our make, does not understand -w, so it prints an error message and exits.
In order to catch this everywhere (since cmake for example can invoke
either our make or gmake depending on how it feels), we reset the variable
for any top level command. This effectively reverts to the behavior of
our make not setting $MAKELEVEL.
satisfy the BISON_REQD check, it does not function correctly in the tools
environment when not called as /usr/bin/bison, as it is unable to find its
m4sugar.m4 without BISON_PKGDATADIR being set.
Whilst we could work around that in bison.mk I feel that's something of a
hack, and it is simpler and cleaner to just use the pkgsrc tool instead.
With this change, .include "../../devel/readline/buildlink3.mk" with
USE_GNU_READLINE=yes should be replaced with
.include "../../devel/readline/buildlink3.mk",
and .include "../../devel/readline/buildlink3.mk" without USE_GNU_READLINE
should be replaced .include "../../mk/readline.buildlink3.mk".
USE_GNU_READLINE is removed.
so that they are correctly calculated as independent.
This avoids issues in bulk builds where the package version was taking
precedence and causing the wrong user package to be depended upon.
from 2006 and the OSX bison has been upgraded long since then. In any
case, if the bison is too old, the BISON_REQD check will ensure that a
working version is pulled in if necessary.
Only one compiler is used when "ada" is listed in LANGUAGES, and that
is the one built by the lang/gcc-aux source package. When PKGSRC_COMPILER
is defined as anything else other than "gcc", the Ada packages fail to
build. This can be seen when clang is used with CLANGBASE=${LOCALBASE}.
This straight-forward fix is to override the user specification of
PKGSRC_COMPILER when Ada is specified and define it as "gcc" in all cases.
Tested on NetBSD 6.1 amd64 with CLANGBASE=${LOCALBASE}
Whilst it works for the most part, the mk/extract/extract script expects
an -O flag which it does not support, and adding conditionals to that
script would be messy.
Fixes 5 direct packages.
the package. For f2c, all Fortran 95+ programs are broken and it is
generally not possible to mix output from different Fortran compilers.
Default to g95 for now as fallback compiler.
This is a provisional kludge to work around PR pkg/47838. Sorry for
taking far too long to find a workaround that doesn't break various
other stuff too -- this duration of time was ridiculous, and it was
entirely my fault.
We can get rid of this kludge when we start using `env -i' in the
build phase or when we replace TARGET_ARCH by TARGET_MACHINE_PLATFORM
(and replace the make-internal variable MACHINE_ARCH by
MACHINE_PLATFORM -- that is part of what makes the logic in
pkgformat/pkg/depends.mk and bsd.prefs.mk so fragile). However,
although I intend to do both of these things, they were deemed too
likely to cause too much fallout just before the freeze, so they'll
wait until after the freeze.
and force USE_BUILTIN.pkg=no for packages that depend on packages where
USE_BUILTIN.pkg is no. The names of any such packages are accumulated in
the variable FORCED_PKGSRC for reference; this is currently undocumented
and could be dropped in the future.
This makes it a lot safer to install pkgsrc versions of selected X
libraries without switching wholesale to pkgsrc X; however, other
issues may still exist and caution is still advisable.
As seen on tech-pkg.
Also note: this may affect the builds of packages we don't realize are
affected and that haven't been revbumped. If you find one, let us know
so we can bump its version (or do that yourself) -- most likely this
change will produce in working, properly-linked packages that were
previously broken, but if problems arise please speak up.
1.) It breaks the build of "www/firefox" which gets upset if "SHELL" is
not defined in the environment. There are probably more packages
which similar problems.
2.) It breaks established use case like this one:
export ALLOW_VULNERABLE_PACKAGES=yes
cd pkgsrc/multimedia/ffmpeg2theora
bmake install
In this case the value of "ALLOW_VULNERABLE_PACKAGES" will not be
passed to the build of "pkgsrc/multimedia/ffmpeg". And the build of
this package will fail due to known vulnerabilities.
`install' with USE_DESTDIR=yes.
This changes prevent to unwanted overwite of existing binary packages with
test installation (`stage-install', `replace' & `undo-replace', and so on).
To do both `install' and `package', you can still use `package-install' target,
same as USE_DESTDIR=no.
substitutions will not be correct.
Fixes issue with ABI=64 where /usr/lib/amd64 was being exposed, but
packages will need to be rebuilt for the change to take effect.
however that is not the case. To get that behaviour use ':S/c/ /g'.
Fixes a number of issues on various OPSYS introduced with the recent
COMPILER_* and SYSTEM_DEFAULT_RPATH abstractions.
It is used in limited case, and does not exist by default on some platforms.
proposed at over 30 months ago, and no negative feedback (only one request).
hierarchy.
This values should be generated from output of some commands,
but I cannot find the rule.
Tested on armel and x86_64 Debian GNU/Linux environment.
The values are shown in http://wiki.debian.org/Multiarch/Tuples .
Build depends are target packages that are needed at build-time for,
e.g., static libraries to link against, header files to include, &c.
Tool depends are native packages that are needed at build-time for,
e.g., compilers/linkers/&c. to run.
ok agc
The NATIVE_xyz versions are for packages that build tools that they
run natively but don't end up in the final product.
This is a provisional scheme -- it should be replaced eventually by
something more principled.
ok agc
