on FreeBSD 10, and amd64 on earlier versions.
SSP_UNSAFE is added to disable in a port if it fails to build, but
this should only be used in rare circumstances such as kernel modules.
Otherwise, the port may just be failing due to lack of respecting
LDFLAGS.
On FreeBSD 10, this uses an ldscript in /usr/lib/libc.so to pull in
libssp_nonshared.a to address issues linking on i386 [1].
On earlier FreeBSD versions the WITH_SSP knob will add -lssp_nonshared
to LDFLAGS on i386. This is not needed on amd64. However, several hundred
ports do not currently respect LDFLAGS, so this support is disabled currently
as it causes build failures if a dependency is looking for the stack_chk
symbols.
Many thanks to jlh@ for this as he had many years of patience in getting
all of the necessary pieces [1][2] in.
[1] http://svnweb.freebsd.org/base/head/lib/libc/libc.ldscript?revision=251668&view=markup
PR: ports/138228 [2]
Submitted by: jlh (bsd.ssp.mk based on)
Reviewed by: bapt
With hat: portmgr
exp-runs done: 37 over a month on 91i386,91amd64,10i386,10amd64
Instead of hardcoding the compiler target as FreeBSD, use the OPSYS
variable to it. This makes no practical difference for FreeBSD, but
it helps DragonFly get properly configured.
Approved by: gerald
if present (even if the libiconv port is present) and fails using that.
The issue is that /usr/include/iconv.h has #include <stdbool.h> which
in turn, since both are included very late in the game, conflicts with
similar definintions by libcpp itself.
libstdc++ does not seem to require adjustments, so PR 161417 may not
be relevant in full.
PR: 161417
It's value is "--with-libiconv-prefix=/usr/local" for systems
before 100043 with ports libiconv and to use at systems post
100043 with base iconv it's value is "" (NULL).
Co-authors: bapt, madpilot and bsam (me)
It brings bison as a build dependency in case it is set the following way:
USES= bison or USES= bison:build
it brings bison as a run dependency in case it is set the following way:
USES= bison:run
it brings bison both as a run and build dependency in case it the set the following way:
USES= bison:both
While here trim some headers
Convert some USE_GNOME= gnomehack to USES= pathfix
rely on gcc. The patch uses the new USE_GCC=any code in Mk/bsd.gcc.mk to
accomplish this.
The ports chosen were ports that blocked 2 or more ports from building with
clang. (There are several hundred other ports that still fail to build with
clang, even with this patch. This is merely one step along the way.)
Those interested in fixing these ports with clang, and have clang as their
default compiler, can simply set FORCE_BASE_CC_FOR_TESTING=yes.
For those who have gcc as their default compiler, this change is believed
to cause no change.
Hat: portmgr
Tested with: multiple runs on amd64-8-exp-bcm and 9-exp-clang, with various
combinations of patch/no-patch and flag settings.
This speeds up a full port/package test cycle by 50% on a 4-core
system; a simple build/install will benefit even more.
On the way, have he compiler report itself as "FreeBSD Ports Collection".
the tree any more and has been obsoleted by several newer versions of
GCC -- ones that actually build on FreeBSD 10.x out of the box and
provide better support for newer CPUs, better optimizations overall,
and more advanced language support.
GCC 4.6 aka lang/gcc or lang/gcc46 is the new default version of GCC
in ports.
The system compiler version of GCC 4.2 is still available up to and
including FreeBSD 9.x and users of USE_GCC=4.2+ will be transparently
redirected to lang/gcc if, or rather when, this changes.
PR: 163103
Feature safe: yes
port consistent across installations and avoids confusing error modes
for users who may have others flavors of these tools earlier in PATH.
On the way, sort CONFIGURE_ARGS.
Triggered by: glarkin
contrib/download_ecj script to the fetch phase using extra MASTER_SITES
and DISTFILES. Pointyhat does not allow fetching outside of the
fetch phase, and the size and checksums of ecj.jar need to be recorded
in distinfo.
Reported by: pointyhat (pav)
Approved by: gerald (via private email)
- Added support for gcj on amd64
- Added call to contrib/download_ecj so gcj executable doesn't die with
a "file not found" error after it's built
Submitted by: glarkin
Approved by: gerald (via email)
served using GCC 4.4 or later. This makes this port a whole lot cheaper
and also allows us to remove the (somewhat bogus) check for lang/gcc295
as well as WANT_GNOME.
No longer download Objective-C related files which we don't actually use.
Set MAINTAINER unconditionally.
to allow for PORTREVISION bumps.
Simplify pre-everything and adjust PLIST_SUB accordingly.
Extract SUFFIX from PORTVERSION. Use SUFFIX for TARGLIB (and thus for
the library path used by this port). Also use SUFFIX for the libexec
directory instead of the full port version and flatten the directory
structure and simplify the logic along the way.
Move VERSIONSTRING down to pacify portlint after all those years and
extract it from PORTVERSION, too.
Make the determination of CONFIGURE_SCRIPT more robust.
Address a long term FIXME related to older versions of FreeBSD by
removing NOMANCOMPRESS=yes.
Adjust Makefile header; nothing really left from the original.
Remove obsolete code to create (empty) man pages in case regular man
pages cannot be created. This catches four dummy man pages we are now
able to remove: grepjar42, jar42, rmic42, and rmiregistry42.
using ?=.
Add powerpc to NOT_FOR_ARCHS. [1]
No longer include the Fortran frontend. This version has been super-
seded by later ports, and the main purpose of lang/gcc42 is being on
par with the GCC 4.2-based system compiler recent versions of FreeBSD
feature.
Remove gcc-testsuite from DISTFILES, along with the post-build and
check targets. This reduces disk and bandwidth consumptions for a
feature (apparently) never used.
PR: 130964 [1]
