devel/*{-xtoolchain,}-gcc: Retire the old GCC external toolchain ports.

These have been supplanted by the devel/freebsd-gcc6 port which uses
separate flavors for each target architecture and bundles the
CROSS_TOOLCHAIN makefile.

Feedback from:	jrm
Reviewed by:	lwhsu, bapt
Differential Revision:	https://reviews.freebsd.org/D30765

devel/i386-gcc: fix build (by patching devel/powerpc64-gcc)

One more small cleanup, forgotten yesterday.
Reported by:	lwhsu

Remove # $FreeBSD$ from Makefiles.

Update to MPC version 1.2.0 with the following changes:
 - New functions:
   . mpc_sum
   . mpc_dot
 - Several functions are more robust with a reduced exponent range
   (for example corresponding to IEEE 754 binary formats).
 - New mpcheck tool for comparison with the native C library (which
   is not installed by default).

Bump all directly dependent ports. Do not bump those indirectly
dependent via the lang/gcc* family since their run-time dependencies
and code generated should not be affected.

PR:		249950
Submitted by:	wen

devel/powerpc64-gcc: fix build on powerpc64le

Use gcc patch that other gcc ports use.

devel/powerpc64-gcc: fix build on powerpc64 elfv2

Use the same patch that is already widely used in the ports tree.

devel/powerpc64-gcc: add binutils dep on objdump

/usr/bin/objdump will be removed from FreeBSD before FreeBSD 13.0.

PR:		241159
Approved by:	bapt
Sponsored by:	The FreeBSD Foundation

Make the old xtoolchain GCC ports conflict with devel/freebsd-gcc6.

Reviewed by:	bapt
Differential Revision:	https://reviews.freebsd.org/D22861

Update binutils to 2.33.1

While here, convert binutils into flavors, that ease a lot maintenance
and helps cleaning out the code.

This is inspired by the same work on going on the xtoolchained gcc by jhb@

PR:		241756
Exp-run:	antoine (portmgr)
Discussed with:	jhb
Differential Revision:	https://reviews.freebsd.org/D22258

Divorce the ARM bare-metal GCC ports from the FreeBSD GCC xtoolchain ports.

In the past, changes to the FreeBSD GCC xtoolchain ports have broken
the ARM bare-metal GCC ports by accident.  In addition, we'd like to
move to having multiple FreeBSD GCC xtoolchain ports for different GCC
versions.  It seems best if we make the two groups of ports
independent so that changes to one do not affect the other.

Reviewed by:	manu
Differential Revision:	https://reviews.freebsd.org/D22064

Always set the endian flag for the linker emulation for MIPS.

Fix a regression in r472011 where gcc would pass a linker emulation
without an endian flag if neither -EB nor -EL were provided on the gcc
command line.

Bump PORTREVISION.

Reviewed by:	imp
Differential Revision:	https://reviews.freebsd.org/D22004

Bump PORTREVISION for ports depending on the canonical version of GCC
as defined in Mk/bsd.default-versions.mk which has moved from GCC 8.3
to GCC 9.1 under most circumstances now after revision 507371.

This includes ports
 - with USE_GCC=yes or USE_GCC=any,
 - with USES=fortran,
 - using Mk/bsd.octave.mk which in turn features USES=fortran, and
 - with USES=compiler specifying openmp, nestedfct, c11, c++0x, c++11-lang,
   c++11-lib, c++14-lang, c++17-lang, or gcc-c++11-lib
plus, everything INDEX-11 shows with a dependency on lang/gcc9 now.

PR:		238330

Return kan@'s ports to the pool after his commit bit was safekept.

Fix a regression for devel/amd64-gcc introduced in r488642.

Restore the value of TARGETARCH to be the FreeBSD architecture
name.  This fixes the block to remove float.h in the amd64
compiler.  This showed up as a packing list failure on amd64-gcc,
but it also undid r475290 for amd64-gcc.

Reviewed by:	manu
Differential Revision:	https://reviews.freebsd.org/D19126

Override the library search directories for external GCC toolchains.

The default library search directories when building a native GCC end
up falling back to /usr/local/lib (for devel/*-gcc) and /usr/lib (for
base/gcc) even when --sysroot is used.  For devel/*-gcc, I previously
forced CROSS_DIRECTORY_STRUCTURE on in r466699 to workaround this.
However, this solution could not be used for base/gcc.  Instead, patch
the sources to remove several of the library search directories and
override STARTFILE_PREFIX_SPEC to only look in /usr/lib.  GCC still
adds tuple-specific subdirs to /usr/lib, but this is ok.

Bump PORTREVISION for both the external GCC toolchains and base/gcc.

Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D18677

Simplify some of the logic for external GCC toolchain packages.

- Always define BU_PREFIX as GCC_TARGET if needed.  This was already
  effectively true, but BU_PREFIX was just being set to the same value
  in the case that GCC_TARGET wasn't defined.  This is simpler and
  avoids duplicating some logic.
- Simplify amd64-gcc by not defining GCC_TARGET and BU_PREFIX in that
  port but instead replacing "amd64" with "x86_64" in the
  devel/powerpc64-gcc Makefile.  This matches what base/gcc does.
- Remove FREEBSD_EXTENSION now that it is no longer needed.

Approved by:	zeising
Differential Revision:	https://reviews.freebsd.org/D18349

Bump PORTREVISION for ports depending on the canonical version of GCC
defined via Mk/bsd.default-versions.mk which has moved from GCC 7.4 t
GCC 8.2 under most circumstances.

This includes ports
 - with USE_GCC=yes or USE_GCC=any,
 - with USES=fortran,
 - using Mk/bsd.octave.mk which in turn features USES=fortran, and
 - with USES=compiler specifying openmp, nestedfct, c11, c++0x, c++11-lang,
   c++11-lib, c++14-lang, c++17-lang, or gcc-c++11-lib
plus, as a double check, everything INDEX-11 showed depending on lang/gcc7.

PR:		231590

Remove orphans after INFO_PATH move

USES=compiler:c++11-lang is necessary to build arm-none-eabi-gcc on
powerpc64.  Since it should be similar with other slave ports, add
USES to the masterport.

PR:		232521
Submitted by:	Piotr Kubaj
Approved by:	portmgr (tier-2 blanket)

devel/aarch64-none-elf-gcc: Unbreak on arm64

The plugins ae not built on arm64.
While we might want them in the future, we also need gcc on this platform
as it's the compiler used for u-boot.

Reviewed by:	jhb, kan
Approved by:	kan (maintainer)

Remove a spurious $ introduced in r475290.

Reported by:	Mark Millard
Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D16551

Cleanup packing lists for xtoolchain gcc packages.

- Add missing entries for gcov-dump and the gcov-tool manpage to the
  various xtoolchain gcc ports.
- Add gcov-dump to the post-install loops that handle optional renaming
  of tools and manpages.

Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D16465

Add ports for an i386 external GCC toolchain.

- devel/binutils: Remove i386 a.out linker scripts when building
  i386-binutils or a base/binutils that targets i386.
- devel/powerpc64-gcc: Remove float.h on i386 since it conflicts
  with sys/x86/include/float.h.
- devel/i386-{binutils,gcc,xtoolchain}: New ports.

Reviewed by:	bapt (previous version)
Differential Revision:	https://reviews.freebsd.org/D16228

Properly set CONFIGURE_TARGET for when amd64 is either the host or target.

CONFIGURE_TARGET is used for --build which is the host performing the
build.  TARGETARCH is used in powerpc64-gcc/Makefile as the target
architecture of the binaries generated by the compiler being built.
When building a non-amd64 target compiler on an amd64 host,
CONFIGURE_TARGET was set to amd64-unknown-freebsd which confused GCC
(it breaks compilation of an i386 target compiler on an amd64 host).
It would also result in setting CONFIGURE_TARGET improperly when
building an amd64 compiler on a non-amd64 host.  Fix by remapping
amd64 to x86-64 if the ARCH is amd64 rather than the TARGETARCH.

Reviewed by:	bapt
Differential Revision:	https://reviews.freebsd.org/D16227

Drop builtin float.h for amd64-gcc.

GCC's builtin <float.h> header is not compatible with
sys/x86/include/float.h.  Drop the builtin header for now.  If at
some point GCC's notion of floating point constants for i386 can
converge with sys/x86/include/float.h this can be restored.

Reviewed by:	bapt
Differential Revision:	https://reviews.freebsd.org/D16073

Enable .init_array and .fini_array with xtoolchain GCC.

For base/gcc, only enable .init_array and .fini_array for 12.0 and later
to match the policy used by clang and llvm.  For the xtoolchain ports,
enable these unconditionally since the only OS version handy is the OS
version of the build host, not the target OS to be built.  Unlike llvm,
the use of .init_array and .fini_array is set at compile-time for GCC
and can't be changed at runtime.

Bump PORTREVISION for base/gcc and the xtoolchain GCC ports.

Reviewed by:	bapt (earlier version)
Differential Revision:	https://reviews.freebsd.org/D16014

Use mips3 as the default MIPS ISA for 32-bit FreeBSD/mips.

The in-tree GCC 4.2.1 defaults to a MIPS ISA of MIPS3 (which includes ll
and sc instructions needed for SMP-aware atomics).  However, out of the
box GCC defaults to MIPS1 for 32-bit MIPS (N32 and N64 both require a
minimum of MIPS3).  Change both devel/mips-gcc and base/gcc to default to
MIPS3 for 32-bit MIPS on FreeBSD.
- Fix the default target cpu to include MASK_ABICALLS in configure.tgt.
  The gcc/config/mips/freebsd.h header already overrides this anway, but
  it is more correct to fix this here.  We could perhaps remove the
  hack from freebsd.h now but I haven't done that.
- Fix the case that checks for 32-bit mips tuples to match on 'mips*'
  instead of 'mips32*' in configure.tgt.  We don't use mips32* in our
  tuples for O32 MIPS.
- Set MIPS_ISA_DEFAULT to 3 (MIPS3) rather than 33 (MIPS32R6 or some such)

Make mips64 compiler use /lib:/usr/lib instead if lib64 variants

This allows mips64 compiler to be used bo build base/ packages,
while it was failing to find crt1.o and like before.

Upgrade ext-toolchain GCC ports to GCC 6.4.0

This is the latest supported release in GCC 6.x series.
Drop aarch64 support patch - it has been committed upstream.

Enable ifunc suppport by default.

PR:	228161
Reviewed by: bapt (update part)
Differential Revision:	https://reviews.freebsd.org/D13602

Update mpfr to 4.0.1 release and bump PORTVERSION of dependent ports.

Fix arm-none-eabi-gcc/aarch64-none-elf-gcc plist after r466699

jhb fixed these ports in r466699, but include-fixed headers has changed
since the last update, perhaps due to --sysroot and these ports being built
differently since then.

Add the extra headers to the plist and bump PORTREVISION due to package
differences. This fixes some sanity checking in the plist, since these files
are installed to the stage dir.

Reported by:	Phillip R. Jaenke <prj@rootwyrm.com>
Approved by:	ler (ports)
MFH:		2018Q2

Fix two more issues with r465416.

- Force build of a cross-compiler by defining CROSS_DIRECTORY_STRUCTURE
  in CFLAGS even if the build host matches the build target.  This
  fixes such a cross compiler to not include /usr/local/lib in its default
  library path (e.g. amd64-gcc when built on amd64).
- Don't remove the include-fixed headers for the aarch64-none-elf-gcc
  and arm-none-eabi-gcc packages.
- Bump PORTREVISION.

Reported by:	kevans (2)
Reviewed by:	bdrewery, kevans
Differential Revision:	https://reviews.freebsd.org/D14925

Remove CONFIGURE_ENV to set C++ include path.

The previous change to fix the name of the argument passed to configure
to set this path means that the setting is no longer required in the
environment.

Did not bump PORTREVISION since this should be a no-op.

Reviewed by:	bdrewery
Differential Revision:	https://reviews.freebsd.org/D14882

Fix --sysroot for cross-toolchain GCC packages

By default, GCC assumes that a cross compiler should not have any valid
initial include or library paths aside from /usr/local/lib/gcc/<mumble>/.
This means that one cannot use --sysroot to point to a system root for
another architecture and have GCC automatically look for headers in
${sysroot}/usr/include, etc.  Currently we workaround this in FreeBSD's
build system with explicit -isystem, -B, and -L directives.  However,
this is cumbersome compared to clang (where a bare --sysroot DTRT)
especially when using the compiler to build other software (such as test
programs, etc.).

One can override GCC's assumption and force it to assume that it should
honor --sysroot by setting the '--with-sysroot' option to force GCC to
assume a specific system root.  By setting this to '/', this means that

Bump PORTREVISIONs of all users of math/mpc that we just updated to
version 1.1.0 (via revision 464079).

Fix cross build of base/ ports

This is a follow up to r461057 and fixes base/binutils and base/gcc in my
testing.

PR:		224217
Submitted by:	nwhitehorn (partially, I made additional changes)
Reviewed by:	bapt

Fix build with clang 6

/wrkdirs/usr/ports/devel/powerpc64-gcc/work/gcc-6.3.0/gcc/ubsan.c:1474:20:
error: comparison between pointer and integer ('const char *' and 'int')
      || xloc.file == '\0' || xloc.file[0] == '\xff'
         ~~~~~~~~~ ^  ~~~~

PR:		224669
Reported by:	pkg-fallout

Make sure <arch-gcc> come with consistent content

Expand hackery in post-install target to
make sure that we always end up with same set of
$GCC_TARGET-prefixed files no matter if port is
built for cross target or for native target to be
used as external toolchain.

Take over the maintainership.

Reviewed by:	emaste, imp, bapt
Approved by:	bapt
Differential Revision:	https://reviews.freebsd.org/D10537

Mark some ports failing on power64.  In cases where the error message
was a stub, provide a real one.

While here, pet portlint.

Reported by:	swills
Approved by:	portmgr (tier-2 blanket)

Mark the gcc crosscompiler ports as broken on various tier-2 archs.

Unfortunately neither the maintainer or I have time at the moment
to chase down all of these failures.  The "fails to package" ones
ought to be fairly easy to fix.

Discussed with:		maintainer

Add external toolchain ports for aarch64

This involves copying patch file from lang/gcc6
and minor plist tweaks.

Reviewed by:	emaste, andreast, brooks
Approved by:	brooks
Differential Revision:	https://reviews.freebsd.org/D10292

The output of tools like awk, date, sort, tr,... depends on the current
locale set by the user.  Add LANG=C and LC_ALL=C at the beginning of
bsd.port.mk and export them so all commands are executed with the C locale.
LC_ALL=C overrides all other LC_* variables.  LANG is used by setlocale(3)
as default value for LC_* variables, so normally it isn't used when LC_ALL
is set, but there's code out there that looks at LANG directly so it's safer
to set it as well.  The only commands not captured by this are !=
assignments before any inclusion of bsd.port.*mk.

Introduce USE_LOCALE=<locale> that adds LANG=<locale> and LC_ALL=<locale> to
CONFIGURE_ENV and MAKE_ENV so upstream build systems can be executed with a
different locale (e.g. USE_LOCALE=en_US.UTF-8).

PR:		215882
Exp-run by:	antoine
Approved by:	portmgr (antoine)

Remove BROKEN_FreeBSD_9

Approved by:	portmgr (blanket)

Update to 6.3.0

Make amd64-gcc respect the 'unknown' second component of the triplets
as other cross toolchain gcc ports

Update cross toolchain gcc to 6.2.0
Remove the powerpc64 patch which has been upstreamed

Spell BROKEN_FreeBSD_9 properly.

devel/powerpc64-gcc: use libc++

Approved by:	bapt (maintainer)

devel/*gcc*: convert to CONFIGURE_OUTSOURCE in trivial cases

PR:		208294
Exp-run by:	antoine
Approved by:	portmgr blanket
Differential Revision:	https://reviews.freebsd.org/D4157

Remove ${PORTSDIR}/ from dependencies, categories d, e, f, and g.

With hat:	portmgr
Sponsored by:	Absolight

Update cross gcc to 5.3.0
While here:
- fix packaging on amd64
- fix building with utf8 locales en recent head by enforcing the locale to be C

Bring the fixes mentioned below from upstream to our gcc-5.2 powerpc64-gcc
port.

 [1]: https://gcc.gnu.org/ml/gcc-patches/2015-12/msg02057.html
 [2]: https://gcc.gnu.org/ml/gcc-patches/2015-12/msg00685.html

PR:		205394, 205440
Approved by:	bapt (maintainer)

Add ELFv2 support for FreeBSD PowerPC.

Submitted by:	nwhitehorn@
Approved by	bapt@:

Update the cross gcc to 5.2.0
Simplify a bit the plist
Drop upstreamed arm support patches not needed anymore

Add a patch to allow building u-boot on arm

Submitted by:	kientzle

MASTER_SITES cleanup.

- Replace ${MASTER_SITE_FOO} with FOO.
- Merge MASTER_SITE_SUBDIR into MASTER_SITES when possible. (This means 99.9%
  of the time.)
- Remove occurrences of MASTER_SITE_LOCAL when no subdirectory was present and
  no hint of what it should be was present.
- Fix some logic.
- And generally, make things more simple and easy to understand.

While there, add magic values to the FESTIVAL, GENTOO, GIMP, GNUPG, QT and
SAMBA macros.

Also, replace some EXTRACT_SUFX occurences with USES=tar:*.

Checked by:	make fetch-urlall-list
With hat:	portmgr
Sponsored by:	Absolight

gcc: Bump port to 4.9.2; add MIPS patches (from kan, thanks!).

Tested:

* compiled mips, mips64, arm-none-eabi, sparc64 cross compile
  packages.

Reviewed by:	bapt

Conditionally assign variables so that slave ports can override them.

Approved by:	imp

turns out -fbracket-depth is clang specific, only use it with clang.

Add the needed changes for amd64-gcc to build

Reported by:	antoine
Pointyhat:	bapt

Unbreak arm-none-eabi-gcc

Reported by:	pkg-fallout

Enable shared library

New port devel/arm-none-eabi-gcc for ARM bare metal programming.

Differential Revision:	https://reviews.freebsd.org/D974
Approved by:	bapt

Allow devel/powerpc64-gcc to be used to target non-FreeBSD ports.

 * The format extension patch should only be used on FreeBSD.
 * The name of the target may not follow the portbld scheme.

Differential Revision:	https://reviews.freebsd.org/D972
Approved by:	bapt

Add support for -fformat-extensions meaning we can now build the kernel

Submitted by:	kan

Cleanup garbage

Make the plist a bit more generic
Remove garbage from distinfo

Add a cross version of gcc minimalistic and suitable enough to cross build world
Current targets: powerpc64 sparc64

Note that gcc 4.9.1 never heard of mips on freebsd neither of arm on freebsd
Note that this does not yet includes the freebsd extensions meaning it cannot
cross build the kernel yet