Commit graph

107 commits

Author SHA1 Message Date
jlam
3ef633718c (1) Publicly export the value of _OPSYS_RPATH_NAME as RPATH_FLAG;
Makefiles simply need to use this value often, for better or for
    worse.

(2) Create a new variable FIX_RPATH that lists variables that should
    be cleansed of -R or -rpath values if ${_USE_RPATH} is "no".  By
    default, FIX_RPATH contains LIBS, X11_LDFLAGS, and LDFLAGS, and
    additional variables may be appended from package Makefiles.
2003-03-14 19:37:30 +00:00
jschauma
ce9144b328 Add some LOWER_VENDOR definitions on Linux to better customize packages
in the future (on Debian, for example, there is no libtermcap, and we need to
depend on ncurses etc.)
2003-01-25 22:39:56 +00:00
jlam
63029a767b Make BUILD_DIR always point to a physical path by invoking /bin/pwd instead
of relying on the shell's builtin pwd.  This makes BUILD_DIR consistently
point to the correct directory regardless of the definition of SHELL.  This
fixes elusive some buildlink2 errors due to the fact that BUILDLINK_DIR is
derived from BUILD_DIR and some paths were incorrectly being translated.
2003-01-15 20:55:38 +00:00
jlam
bae57c8408 Revert previous change. Defining BSD_PREFS_MK instead of BSD_PKG_MK
breaks too many users' /etc/mk.conf files that have a

	.ifdef BSD_PKG_MK
	# /usr/pkgsrc stuff
	.else
	# /usr/src stuff
	.endif

structure.  We'll think of another way to protect against multiple
inclusion of bsd.pkg.mk and bsd.prefs.mk.
2003-01-15 10:06:48 +00:00
wiz
d2bb2ec016 Define and check for BSD_PREFS_MK, not BSD_PKG_MK. Proposed by Jan Schaumann. 2003-01-14 16:25:41 +00:00
jschauma
d07838581f On Linux, LOWER_ARCH is generally not yet defined. Move assignment. 2003-01-14 03:01:37 +00:00
jschauma
207359eaa9 According to seb, make on 1.5 does not know about ${VAR:tl}, so let's go back to
echo VAR | tr A-Z a-z
2003-01-12 22:36:14 +00:00
jschauma
a2843dc815 efficiency++;
Make can take care of substitutions itself, so we don't need to use
echo VAR | sed
or
echo VAR | tr
Suggested by seb.
2003-01-12 22:30:08 +00:00
jschauma
638f4888bd On Linux, s/i.86/i386/ for LOWER_ARCH and MACHINE_ARCH as suggested by
Jeremy C. Reed on tech-pkg.
Even though bootstrap's bmake will do this at build-time, this ensures
that even older bmake's don't fail if a package checks it for == i386.
2003-01-12 20:36:58 +00:00
jlam
60415acdfc Remove a unused hack (it was only used by pthread.buildlink.mk, which is
now gone).
2003-01-05 22:48:12 +00:00
schmonz
e1c77ef496 Introduce _OPSYS_HAS_INET6, and use it to determine the default
value of USE_INET6.
2002-12-18 06:18:14 +00:00
taca
6e1eb8c978 Previous commit breaks pkgsrc on NetBSD current since OS_VERSION isn't
pure numeric value "1.6K".

So, sperate ".if" directive comparing OS_VERSION into two ".if" directives.
2002-12-15 01:15:24 +00:00
schmonz
a582dc31ff Default to USE_INET6 = NO on Darwin < 6.0, whose IPv6 bits are a red herring. 2002-12-15 00:40:29 +00:00
jschauma
ea31bf10a2 Make sure that on IRIX{,64} LOWER_OPSYS matches what most configure
scripts seem to expect.  While we're at it, add a vendor.
2002-12-12 03:02:41 +00:00
wiz
6e77852ef9 Remove trailing whitespace. 2002-12-10 12:50:48 +00:00
grant
a352f694c2 turn on USE_INET6 in a KAME IPv6 environment other than NetBSD. 2002-11-30 16:49:00 +00:00
grant
9731c03451 LOWER_VENDOR is 'pc' on FreeBSD only if running on i386. 2002-11-17 22:35:23 +00:00
grant
cf547a3787 Add initial pkgsrc FreeBSD support :-)
Tested on FreeBSD 4.7-RELEASE/i386.
2002-11-17 15:11:40 +00:00
jlam
3ea1f6abe7 When "make install" su's to root, su is invoked with the "-l" option, which
causes the shell environment to be discarded.  This also discards
OBJMACHINE or OBJHOSTNAME, which causes WRKDIR_BASENAME to be different
between the "non-root" and "root" make targets and leads to the breakage
seen in pkg/18879 by Simon Burge.  Fix this by saving the OBJHOSTNAME or
OBJMACHINE setting in MAKEFLAGS so that it is seen even after we "su -l" to
root.
2002-11-01 06:29:23 +00:00
jlam
6bbf080d53 Whitespace and minor cosmetic fixes. 2002-11-01 06:19:47 +00:00
jlam
4d0aee253b Move X11BASE setting for Solaris to a more appropriate place now that we
don't need to worry about USE_XPM.
2002-10-21 21:46:00 +00:00
wiz
555b0a21c7 Purge unused USE_XPM (use graphics/xpm/buildlink2.mk instead). 2002-10-21 13:58:14 +00:00
wiz
144d00183a Remove traces of now unused USE_XAW. 2002-10-21 01:40:56 +00:00
wiz
16407c1798 Remove USE_MESA -- now unused. 2002-10-21 01:17:11 +00:00
wiz
accb9cb69d Remove obsolete USE_FREETYPE2 support; all users have been converted to use
freetype2's buildlink2.mk.
2002-10-20 11:47:04 +00:00
jlam
0905c3678b Move some variable definitions from bsd.pkg.mk to bsd.prefs.mk so that
they may be used in conditional statements in *.mk files before bsd.pkg.mk
is included.
2002-10-19 20:33:59 +00:00
bouyer
90ee03f829 Assign a default value to SPARC_TARGET_ARCH for the if MACHINE_ARCH == sun4
too.
2002-10-12 20:25:49 +00:00
agc
acd94fb634 On Solaris, MACHINE_ARCH on Suns can be sun4c or i86pc with
bootstrap-pkgsrc, so adjust accordingly.

Also indent a little better.
2002-10-12 17:17:22 +00:00
jlam
4a7c8310c5 Move pkgtools-related definitions into bsd.prefs.mk so that they may be
used by buildink2.mk files.
2002-09-24 21:54:49 +00:00
wiz
6c40f123f7 Some whitespace cleanup. 2002-09-24 13:59:20 +00:00
rh
5d9dea26a6 Move xpkgwedge check from bsd.pkg.mk to bsd.prefs.mk so that X11PREFIX
location check within package Makefiles actually works as expected.
2002-09-15 11:58:22 +00:00
tron
df0b4367aa The "Xpm" library is distributed with Solaris 9. Handle this case like
XFree86 4.x and don't install the package.
2002-09-01 15:13:41 +00:00
grant
a4a0e792bf set USE_INET6 if we are on an IPv6 capable Solaris host. 2002-08-21 08:40:40 +00:00
jlam
554bc08289 We no longer expect PTHREAD_TYPE to be defined after including
bsd.prefs.mk.  It's value is now supposed to be checked after including
pthread.buildlink.mk.
2002-08-01 05:36:33 +00:00
jlam
753e8e8e87 Move the definition of OBJECT_FMT for NetBSD-1.3.x back _below_ the
inclusion of bsd.own.mk.  This is to allow OBJECT_FMT to be set explicitly
in ${MAKECONF} (/etc/mk.conf) to override the setting in bsd.own.mk on the
older NetBSD systems.
2002-06-21 21:06:48 +00:00
jlam
d52785d9bf Move the definition of OBJECT_FMT for NetBSD-1.3.x systems ahead of the
inclusion of defs.*.mk so that they can conditionalize on OBJECT_FMT.
2002-06-21 17:49:47 +00:00
schmonz
4a19db1481 On Darwin, before including <bsd.own.mk>, set OBJECT_FMT to "Mach-O".
(<bsd.own.mk> tries to set it to "ELF" otherwise.) The problem was
unmasked by -r1.983 of bsd.pkg.mk.

Approved by agc.
2002-05-30 22:15:40 +00:00
fredb
d9607f2757 Append "elf" to ${MACHINE_GNU_PLATFORM}, for NetBSD archs that were once
a.out but are now ELF, in a way that's consistent with bsd.own.mk in
NetBSD-current. This, incidently, makes the gcc package build again on
NetBSD-1.5.3_ALPHA/i386.
2002-03-18 05:46:42 +00:00
jlam
e023ba1995 Define a symbol BSD_PREFS_MK that may be checked by Makefile fragments to
see whether they are being included from within bsd.prefs.mk or from
without.
2001-12-26 19:16:59 +00:00
jlam
d0bd34b556 Initial implementation of pthread.buildlink.mk to transparently allow for
using either the native pthread implementation or using a package pthread
implementation instead, e.g. pth, ptl2, mit-pthreads, etc.  The only
currently supported package pthread implementation is pth as it's the only
one with a buildlink.mk file.  An example usage is:

	USE_PTHREAD=	native pth
	.include "../../mk/pthread.buildlink.mk"

or a fancier example is:

	USE_PTHREAD=	native
	.include "../../mk/bsd.prefs.mk"
	.if defined(PTHREAD_TYPE) && (${PTHREAD_TYPE} == "none")
	CONFIGURE_ARGS+= --without-pthreads
	.endif
	.include "../../mk/pthread.buildlink.mk"
2001-12-24 19:29:32 +00:00
wiz
9d774a36a3 Retire USE_CURSES, which was superseded by devel/ncurses/buildlink.mk, and
has now been purged from pkgsrc.
2001-12-12 16:30:04 +00:00
abs
e40b0cc41e Add LOWER_OS_VERSION to match LOWER_OPSYS, LOWER_VENDOR, and LOWER_ARCH 2001-12-12 12:31:14 +00:00
tv
92d724ccfe In the odd possible case of someone using "pkgsrc" underneath "src"
(even though that's still a bad idea), set USETOOLS=no explicitly before
including <bsd.own.mk>, and add this also to MAKE_ENV to make BSD-Makefile
based packages build properly.
2001-11-28 19:58:13 +00:00
agc
6dde5fcc79 Move some operating-system specific definitions to the individual
defs.${OPSYS}.mk files.
2001-11-21 14:10:06 +00:00
agc
55b1c236e7 My thanks to Thorsten - this part of my change for Darwin support was
lost.

However, modify previous change to let Darwin know about SHAREOWN,
SHAREGRP and SHAREMODE in a different way - no point in checking for
OPSYS here, since every OPSYS has to know about them.
2001-11-04 19:49:56 +00:00
frueauf
8e847c8588 "Darwin" needs to know about SHAREOWN, SHAREGRP and SHAREMODE too. 2001-11-04 13:42:01 +00:00
agc
7167cd3880 Add initial support for Darwin (1.4) to pkgsrc. 2001-10-30 13:53:15 +00:00
tron
85eb51b084 Include "bsd.own.mk" before "bsd.pkg.defaults.mk" because otherwise "?="
assignments in "/etc/mk.conf" won't work anymore.
2001-09-21 15:28:35 +00:00
agc
91d1845f89 Add a new bsd.pkg.defaults.mk file, derived from the old mk.conf.example.
This file is "included" automatically before <bsd.own.mk> includes
/etc/mk.conf, so that pkgsrc-wide default values are set.

It is now possible just to set values in mk.conf only where they differ
from the default, thereby easing the problems of updating mk.conf when
new values get added.
2001-09-18 21:30:05 +00:00
abs
7cb5dda995 shuffle some whitespace to make it more obvious what is going on 2001-08-24 09:00:44 +00:00