CVS log for src/tools/Makefile.gnuhost

Up to [cvs.NetBSD.org] / src / tools

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Pull up following revision(s) (requested by lukem in ticket #177):

	tools/Makefile.gnuhost: revision 1.56
	external/gpl3/gcc/usr.bin/host-libcpp/Makefile: revision 1.6

Fix passing -j NNN to gmake

Use a more restrictive pattern to extract -j NNN from MAKEFLAGS
into GMAKE_J_ARGS, to avoid false positives when the source directory
has "-j" in the path (e.g "amd64-job-000012" or "src-j9999").

Previously this could pass either -"-j" or "-j BIGNUM" to gmake
and result in "vfork: Resource temporarily unavailable" failures.

PR misc/54456

Fix passing -j NNN to gmake

Use a more restrictive pattern to extract -j NNN from MAKEFLAGS
into GMAKE_J_ARGS, to avoid false positives when the source directory
has "-j" in the path (e.g "amd64-job-000012" or "src-j9999").
Previously this could pass either -"-j" or "-j BIGNUM" to gmake
and result in "vfork: Resource temporarily unavailable" failures.

PR misc/54456

tools: if MAKEVERBOSE >=2, noisy automake builds

Explicitly build with V=1 for MAKEVERBOSE >= 2
so that Makefile.gnuhost (automake) consumers
are more consistent with native builds.

(If the upstream is using AM_SILENT_RULES(yes)
then the default is V=0).

tools: in non-verbose mode, build the tools silently

tools: configure --silent if MAKEVERBOSE == 0

tools: if MAKEVERBOSE < 2, quieten automake builds

Sync with HEAD

Sync with HEAD, resolve a couple of conflicts

Apply patch, requested by maya in ticket #1070:

force build of gcc and friends with -std=gnu++03 when the host compiler
is clang. New clang versions default to newer c++ standards and the
old gcc in this branch has not been updated to those.

Adjust whitespace to look more like assignment than comparison. NFC

detect clang instead of bundling it with Darwin

Sync with head

Add FLEX to the configure env, from Lloyd Parkes on tech-toolchain.
Some configure scripts seem to need both LEX and FLEX.

Sync with HEAD

Pull up following revision(s) (requested by maya in ticket #775):

	tools/Makefile.gnuhost: revision 1.46-1.48
	external/gpl3/gcc/dist/gcc/genattrtab.c: revision 1.2

do the bracket nesting only for clang for now.

Use the __clang__ preprocessor symbol to check for clang, since --version
might barf. From joerg@

Apply upstream commit:
From: ppalka <ppalka@138bc75d-0d04-0410-961f-82ee72b054a4>
Date: Wed, 27 Apr 2016 21:18:05 +0000
Subject: [PATCH] Reduce nesting of parentheses in conditionals generated by
 genattrtab

gcc/ChangeLog:
        * genattrtab.c (write_test_expr): New parameter EMIT_PARENS
        which defaults to true.  Emit an outer pair of parentheses only if
        EMIT_PARENS.  When continuing a chain of && or || (or & or |),
        don't emit parentheses for the right-hand operand.
git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@235536

gcc/arm generates so many parens it hits a bracket depth limited which is
enforced by clang. This reduces the number of parens generated and avoids the
need to increase bracket depth.

Fixes PR toolchain/53178 properly.

Remove hack previously needed to build gcc/arm with clang.
genattrtab.c:1.2 makes this unnecessary.

Tested by thorpej.

Remove hack previously needed to build gcc/arm with clang.
genattrtab.c:1.2 makes this unnecessary.

Tested by thorpej.

Use the __clang__ preprocessor symbol to check for clang, since --version
might barf. From joerg@

Sync with HEAD, resolve some conflicts

do the bracket nesting only for clang for now.

Pull up following revision(s) (requested by maya in ticket #754):

	tools/Makefile.gnuhost: revision 1.45

GCC build exceeds the macOS clang default bracket nesting level of 256.

Work around with -fbracket-depth=512.

From potr in PR toolchain/53178.

GCC build exceeds the macOS clang default bracket nesting level of 256.
Work around with -fbracket-depth=512.

From potr in PR toolchain/53178.

look in gpl3 for stuff too (autoconf)

look in external first.

Rebase to HEAD as of a few days ago.

sync with head.

for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.

this commit was splitted into small chunks to avoid
a limitation of cvs.  ("Protocol error: too many arguments")

pass HOST_SH to the build, to keep it from picking up a "bash"
from $PATH

Revert previous change to tools/Makefile.gnuhost.  When getenv("M4")
contains multiple words, flex fails.

Adjust tools/autoconf/Makefile to pass the necessary extra arguments via
M4=... in CONFIGURE_ENV.  This is similar to what used to be in
revision 1.4 of tools/autoconf/Makefile, but it's now after instead of
before the inclusion of Makefile.gnuhost, because it has to override the
value of M4=...  that Makefile.gnuhost stores in CONFIGURE_ENV.

Add the "-g -D__gnu__" flags when invoking M4 via CONFIGURE_ARGS
in tools/Makefile.gnuhost.  This makes the tools version of m4
behave like GNU m4.

Remove similar code from tools/autoconf/Makefile, where it didn't
work because Makefile.gnuhost would previously have overridden it.

This fixes a problem in which the tools version of autoconf complained
"M4sugar requires GNU M4.
Install it before installing M4sugar or set the M4 environment variable to its path name."

sync with head

Don't use -print0/-0 in find | xargs, it is non-standard and not needed at all.

sync with head.

Pull up following revision(s) (requested by tsutsui in ticket #234):
	tools/Makefile.gnuhost: revision 1.38
Replace not only space but also '(' and ')' on generating BUILD_PLATFORM
strings to appease shell's echo.
Cygwin's uname -r returns "1.7.9(0.237/5/3)" etc.

Replace not only space but also '(' and ')' on generating BUILD_PLATFORM
strings to appease shell's echo.
Cygwin's uname -r returns "1.7.9(0.237/5/3)" etc.

sync with head

remove dup target.

Merge the two Makefiles that were mostly copies of each others.
The reason to do this was to add support for cleaning up the config.cache
files when the build environment (as determined by uname -srm) changes.

Pass M4=${TOOL_M4:Q} to CONFIGURE_ENV for tools/gmp.
Fixes configure error on cygwin.

Actually use timestamp file's timestamps so that things rebuild correctly.

Use ${TOOL_AWK} instead of ${AWK} or plain "awk" in make commands.
Pass AWK=${TOOL_AWK:Q} to shell scripts that use awk.

Put back @ on configure line. Was off for debugging and shouldn't have been
committed.

Make a BUILD_ENV that by defaults matches CONFIGURE_ENV and pass it along on
the build step. Catches things like binutils which do a bunch of configures
on the build step and lose possibly. Fixes issues from PR#29197 for lex
not being picked up here.

Pass AR and RANLIB also.

OK'ed by mrg.

Pull up revision 1.29 (requested by jmc in ticket #531):
Ignore errors on some rm -rf's for platforms that aren't quite
POSIX compliant
and return errors on r/o source (FreeBSD) for -rf. Fixes PR#25022

Pull up revision 1.28 (requested by jmc in ticket #529):
Add $MAKE to env also. From PR#25952

Ignore errors on some rm -rf's for platforms that aren't quite
POSIX compliant
and return errors on r/o source (FreeBSD) for -rf. Fixes PR#25022

Add $MAKE to env also. From PR#25952

Pull up revision 1.27 (requested by jmc in ticket #473):
MKUPDATE cannot possibly do the right thing with GNU wrapped tools installs as
no dependencies are known in advance. So a simple 'build.sh -r -u' will
often lose and end up with a TOOLDIR without a toolchain, groff, etc. Fix
by forcing .install_done to always run.

MKUPDATE cannot possibly do the right thing with GNU wrapped tools installs as
no dependencies are known in advance. So a simple 'build.sh -r -u' will
often lose and end up with a TOOLDIR without a toolchain, groff, etc. Fix
by forcing .install_done to always run.

Backout recent changes made to gnu/lib/libgcc3 for a forced MKPIC=no build.

Fix the behaviour of native and tools gcc when MKPIC=no is specified for
platforms that mknative has determined support shared libraries.

XXX distrib/sets/sets.subr doesn't support MKPIC=no

Be more gcc like in the way we deal with libgcc so that sh3/hppa millicode
and exception handling have a chance of working properly.

        - creates libgcc, libgcc_eh and libgcc_s

        - updates LIBGCC_SPEC to use them appropriately.

          There's a hack in here at the moment with respect to libgcc_so in that it
          is preferable to link against libgcc_so will only when -shared-libgcc is
          specified (the c++ frontend does this automatically.) Configurations where
          LINK_EH_SPEC is defined already do this. The gcc configuration for
          NetBSD/alpha and another NetBSD platform (I forget which) actually define
          LINK_EH_SPEC probably by accident rather than design.

        - updates share/mk to use the compiler's knowledge of what needs linking into
          libraries and executables. This removes an hppa hack.

        - updates the sets for the newly created libgcc* files.

	- support for linking against the _pg version of libgcc has been removed.

Use ${HOST_SH} instead of `sh'.

If necessary, pull in <bsd.sys.mk> to get the definition of HOST_SH;
Makefiles that pull in one of (most of) <bsd.*.mk> will get this anyway.

Rework how MAKEVERBOSE operates:

    *	Don't bother prefixing commands with a line of  ${_MKCMD}\
	and instead rely upon "make -s".  This is less intrusive on
	all the Makefiles than the former.  Idea from David Laight.

    *	Rename the variables use to print messages.  The scheme now is:
	    _MKMSG_FOO		Run  _MKMSG 'foo'
	    _MKTARGET_FOO	Run  _MKMSG_FOO ${.TARGET}
	From discussion with Alistair Crooks.

MAKEVERBOSE support

add -O2 to HOST_CFLAGS on Darwin since adding "-no-cpp-precomp" there blows
it away otherwise

set HOST_CFLAGS+= -no-cpp-precomp when configuring gnu software on Darwin

Add MKUNPRIVED; if not no has the same semantics as if UNPRIVED was defined.
Replace defined(UNPRIVED) tests with ${MKUNPRIVED} != "no"

Add MKUPDATE; if not no has the same semantics as if UPDATE was defined.
Replace defined(UPDATE) tests with ${MKUPDATE} != "no"

Improve documentation for these and other make flags.

Put -X into MAKEFLAGS on FreeBSD and Mac OS X; they have a small ARG_MAX.

replace DIST with GNUHOSTDIST, and don't unconditionally define it
in Makefile.gnuhost (use ?= instead of =, to allow overriding)

Remove the "-j1" hack.  This needs to be fixed better.  For now, -j is
disabled at the top level anyway, making this extra hack unnecessary.

Roll in fixes to permit cross-compiling from non-NetBSD hosts.  This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
  sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

Make sure LEX and YACC are properly defined (mainly for groff); also use
a more portable grep construct.

Remove stray diagnostic "echo".

* Honor BUILD on "make install" (will do "make all" first).

* Make "make all" update the .build_done timestamp more intelligently;
  if no files changed in the build tree, don't touch the timestamp.
  (This allows UPDATE and BUILD to function correctly.)

Avoid building and installing every time, so that we don't have toolchain
dependencies cause things like constantly regenerating catpages....

Make sure and run sh configure as it may not have execute permissions.

replace instances of INSTALL_FILE with HOST_INSTALL_FILE

Workaround some make(1) braindamage wrt when it tries to "auto-chdir"
back to ${.CURDIR}.  How losing....

Recheck the GNU configuration if ${.CURDIR}/Makefile changes.

Use ${INSTALL_FILE:N-U}.

Change where the -j1 is set to disable parallelism.  XXX This should not
be necessary; .NOTPARALLEL needs to be told how to work properly for
specific recursion targets.

Remove extraneous @echo from local changes.

Move around the MAKE_ARGS assignments to put the := as early as possible.

Make the wrapper work correctly for groff too, which uses a rather freaky
Makefile layout.

Major overhaul of src/tools and host toolchain.  Summary of changes:

* Rewrite src/tools Make logic to work like the rest of the tree wrt
  "dependall" and "install".  The old "make build" hack is gone.

* Remove the MKTOOLS logic.  This was linked to the "make build" hack,
  and was only needed because TOOLDIR originally had no writable default.

* Redo the GNU configure/make logic to make it fit reasonably in a
  BSD make wrapper.  Use new ${.ALLTARGETS} variable to scan for
  targets in $(srcdir), and mark them with .MADE: to prevent rebuilding.

* Only build cross tools in src/tools; remove some messy logic in
  src/usr.* and src/gnu/usr.* that would do target filename rewriting
  (improves consistency and readability).

* Add the ability to build cross gdb at tool build time by setting
  MKCROSSGDB (default no) to "yes" in mk.conf.

* Add src/tools/groff and set up paths to work with this cross groff.