Commit Graph

170 Commits

Author SHA1 Message Date
apb 97c727e8ec Fix some problems with the change from revision 1.167:
* Allow "" or "." as the last component of PATH, but do not
  allow arbitrary single-character components.
* Make a comment match the code.
2007-04-14 10:50:48 +00:00
scw 5764a76889 Remove support for NetBSD/{,evb}sh5. 2007-04-08 09:35:21 +00:00
apb 4b4de5fa3f Try not to rebuild nbmake so often:
* Add a try_set_TOOLDIR function which tries to be clever about setting
  TOOLDIR if possible;
* Call try_set_TOOLDIR from rebuildmake;
* Add nobomb_getmakevar, which works like raw_getmakevar except
  never bombs;
* Reimplement raw_getmakevar in terms of nobomb_getmakevar;
* Add a second arg to find_in_PATH to control the result when the
  program cannot be found.
2007-04-07 14:49:40 +00:00
dsl af519d4729 Don't abort the build if teh last component of $PATH is . 2007-04-05 10:02:10 +00:00
apb 2347a3fcd8 The "${progname} started: ${build_start}" message will appear in the
"Summary of results", so don't print it again just before that.
2007-04-02 11:00:10 +00:00
apb daaa947e26 * Report the NetBSD version being built.
* Report the host platform.
* Print more of the initial messages in the same part of the code.
2007-04-02 10:57:36 +00:00
apb b6c6d298a0 s/PATH/PATH environment variable" in an error message. 2007-03-25 12:36:01 +00:00
apb cabe8d739b If the PATH contains any components that are not absolute
directory names, print a message and abort.  In expert mode,
downgrade this error to a warning.

Fixes PR 36048
2007-03-25 07:56:38 +00:00
briggs d66d113907 Don't treat -m evbarm like the other arm-based platforms. Set
MACHINE_ARCH only if it's not already set.  This prevents
	build.sh -m evbarm-eb
from ending up with MACHINE_ARCH=arm instead of =armeb (since MACHINE is
stripped of the -eb suffix just above).
2007-02-19 14:20:11 +00:00
apb f7004b0cbc Whenever we try to parse the output from make, run make with the -B
flag.  This ensures that -j<N> isn't accidentally inherited from the
environment, because the "--- foo ---" headers printed with -j<N> would
interfere with parsing the output.
2007-02-05 18:26:01 +00:00
matt 4e9d3760b0 Remove duplicate evbarm-e[bl] cases. Add evbppc64 and macppc64. 2007-01-29 00:08:13 +00:00
matt d9adacfe08 Add evbarm-e[bl] machine aliases for evbarm/armeb and evbarm/arm,
respectively.
2007-01-28 20:15:44 +00:00
apb 90b76996ba Add aliases evbarm-eb and evbarm-el, for consistency with several
existing *-e[bl] aliases.  build.sh -m MACHINE_ALIAS is (almost)
equivalent to build.sh -m MACHINE -a MACHINE_ARCH.
2007-01-27 11:27:33 +00:00
rillig ddda16fb21 Added a friendly error message in case the user forgot to "build.sh tools". 2007-01-17 03:43:18 +00:00
dsl 6875838454 Only cd "$(dirname $0)" when usr.bin/make doesn't exist. This caused grief
when I make my wrapper script source build.sh.  I can also think of other
times when it would be a PITA.
Abort the build if make fails in raw_getmakevar() - as tends to happen when
make itself is badly broken.
2007-01-14 17:02:56 +00:00
ober 02f3407a26 Updated to include the new Zaurus port for ARM.
OK'd by gimpy
2006-12-16 02:59:33 +00:00
dyoung f3fbc3bb0f Change the name of the host disklabel to nbdisklabel-${MAKEWRAPPERMACHINE},
where ${MAKEWRAPPERMACHINE} is the suffix XXX on the target's make
wrapper, nbmake-XXX.

Fixes toolchain/30673, "single arch host disklabel isn't enough".
2006-12-09 20:13:13 +00:00
apb 1889c0e175 Use ${HOST_SH} instead of the host system's /bin/sh wherever possible:
* Try to guess a suitable value for HOST_SH, if it was not set in
  the environment.  First try host-specific heuristics (Solaris has
  /usr/xpg4/bin/sh); then try to find the name of the shell that was used
  to run build.sh itself (by parsing the output from ps -p $$ -o comm);
  then fall back to "sh".

* Having found a value for HOST_SH, copy it to BSHELL and CONFIG_SHELL.

* Use ${HOST_SH} instead of /bin/sh when creating the make wrapper.

* Use ${HOST_SH} instead of unqualified sh when running shell scripts.
2006-09-29 19:53:54 +00:00
uwe 2e25d643b7 Recognize MACHINE=landisk (sh3el). 2006-09-01 21:52:55 +00:00
matt da01e9eb57 When going from -m {evb,sb}mips64-e[bl] to MACHINE,
make sure to remove 64-e[bl] so we wind up with just evbmips or sbmips.
2006-08-27 05:05:05 +00:00
matt 9e051a8002 Add enough configury glue so that src/tools can build a N64 mips toolchain. 2006-08-25 23:35:04 +00:00
jnemeth 252669904e PR/30842 -- compatibility issue with mksh from Thorsten Glaser 2006-08-09 22:13:12 +00:00
mrg d4ce694d06 preliminary support for powerpc64 toolchain:
- allow "-m macppc/evbppc" to also take "-a powerpc64" to build a
  powerpc64 toolchain.

XXX this may change, as an evbppc64 port may be required.
XXX deal with that when we have some code...
2006-06-22 20:00:18 +00:00
dogcow f76552d855 Huzzah. /bin/pwd -P, instead of silently failing, now returns an error code
on many linuces I've tried. If so, bail and invoke /bin/pwd sans argument.
2006-06-19 18:12:30 +00:00
apb fd0db8a416 Fix the iso-image action. These changes were accidentally omitted
from the previous commit.
2006-02-03 12:29:41 +00:00
apb 21078fc390 * Add an "iso-image" target in the top level Makefile,
which invokes make iso-image in the etc directory.
* Add an "iso-image" action in build.sh, which
  invokes make iso-image.
* Document the above in doc/BUILDING.mdoc.
* Re-generate BUILDING.

Approved by christos
2006-01-28 21:34:07 +00:00
dsl 93ab8a734f Use separate invocations of make for cleandir dependall and install of
tools so that a parallel make doesn't start running dependall before
cleandir has finished.
Unfortunately make doesn't support .WAIT on the command line :-(
2006-01-08 11:03:40 +00:00
dsl 20683c5a8c Use parallel makes more often. 2006-01-07 18:45:15 +00:00
apb 026a2bfc50 Add a new "syspkgs" target, which simply chains to the "syspkgs" target
in the top level Makefile.

Reviewed by agc
2006-01-04 15:31:40 +00:00
tsutsui 32af391479 Add ews4800mips to mipseb ports.
Build for ews4800mips might fail (yet),
but please don't complain for now.
2005-12-29 15:38:54 +00:00
jmc ad14f15c1d Nuke LESSCHARSET as well since it affects the output character set of groff 2005-12-12 04:51:55 +00:00
isaki 9265690c77 It needs the braceexpand option for PD KSH, regardless of
whether the host system is OpenBSD.
from Wim Lewis. reviewed by sjg@.
2005-09-25 05:34:21 +00:00
skrll 784c9344e8 Initial pass at an in-tree ia64 toolchain. From PR 29141 with some fixes
from me.
2005-08-17 07:13:32 +00:00
bouyer 44d1677f84 Remove support for build.sh -m xen, this has been merged in the i386
build. Pointed out by Chuck Silvers.
2005-05-29 10:54:40 +00:00
lukem 6ec400133b Unset INFODIR in the makewrapper.
Should fix a build problem observed when installing .info files
when the build user's environment has INFODIR set.
2005-05-19 01:37:30 +00:00
isaki 5067941219 Avoid the brace expansion on OpenBSD.
approved by sjg@.
2005-03-26 06:02:13 +00:00
gavan 0560d8b947 Initial import of iyonix port.
The Iyonix is a desktop machine from Castle Technology, based on a 600MHz
XScale[tm] 80321 processor.

* Uses the bootloader from NetBSD/acorn32, which is now 32-bit compatible.
* Currently boots multiuser with a serial console.
* Device support is not yet complete.

With help from abs.
2004-10-13 23:28:34 +00:00
martin 17cfe6c93e Backout previous on request from mrg - it breaks cross compilation for
sparc64. Matt will recommit a fixed version later.
2004-10-11 06:18:38 +00:00
mrg 57d6df6697 add new ${uname_p} (as `uname -p`).
set ${arches} to "sparc64 sparc" for MACHINE=sparc64.
for 32 bit sparc64 host, default to building 32 bit world (same as native).
2004-10-09 20:38:01 +00:00
junyoung 0220a8417b Remove trailing / in pathname.
From Alan Barrett. See tech-toolchain@ for related discussion.
2004-08-17 14:00:30 +00:00
jmc bd789f7eba Add a change requested in PR#15548 and append the args passed to build.sh
into the comments in the make wrapper.
2004-07-02 04:25:24 +00:00
wiz 40726f5f0d Document -h. From Karsten Kruse in PR 25974. 2004-06-25 15:03:39 +00:00
cl 885c0b6fc9 ``build.sh -m xen-i386 release'' now builds a release for NetBSD/xen
for i386.  The resulting release consists of:
- NetBSD/xen for i386 kernel, loader and docuemntation
- NetBSD/i386 userland sets
2004-05-12 16:56:41 +00:00
lukem d0ca5ded84 Add
-x		set MKX11=yes
	-X x11src	set X11SRCDIR to x11src
2004-02-06 23:19:30 +00:00
lukem c77dab9416 highlight that the noisyless level is the MAKEVERBOSE level 2004-02-04 11:23:40 +00:00
lukem 1710dac7b6 tweak usage message.
when performing "sets", highlight that DESTDIR should be
pre-populated.  suggested by glen mccready.
2003-11-14 12:38:12 +00:00
lukem b60bb5a7dd Now that MKUPDATE and MKUNPRIVED are "normalized" with getmakevar(),
test with  "${MKxxx}" = "no"  instead of  -z "${MKxxx}"

Ignore errors when running pwd -P.  (GNU coreutils incorrectly complains)


Both problems pointed out in private email from Christian Limpach.
2003-11-12 15:51:45 +00:00
lukem f87ba55ff9 Always use `/bin/pwd -P` to get the "TOP" of the source tree, so that
MAKEOBJDIR (and other parameters) work correctly with symlinks, amd(8), etc.
Should fix PR [bin/23313], and possibly others.
2003-10-31 01:46:39 +00:00
lukem 301dd07e78 Explain why make is being rebuilt.
Patch from Hubert Feyrer in [toolchain/23156].
2003-10-26 03:12:21 +00:00
lukem 5daa33014e Improve how build.sh -N and MAKEVERBOSE interoperate;
rather than adding '-s' to make's command line in the makewrapper,
use the .SILENT target based on MAKEVERBOSE's value.

This means that you can do:
	./build.sh -N 1 makewrapper

	$TOOLDIR/bin/nbmake-$MACHINE
	#  runs at MAKEVERBOSE==1, with command lines suppressed

	$TOOLDIR/bin/nbmake-$MACHINE MAKEVERBOSE=2
	#  runs at MAKEVERBOSE==2 for this invocation, with command lines shown
2003-10-26 02:17:46 +00:00