from the LLVM 3.4 branch starting.
Apply the clang portion of r255321 from base switching the default C++
library to libc++ on 10+.
Correct clang's notion of the underlying type of (u)intmax_t on MIPS64.
2013-11-01 sysutils/mktool
2013-11-01 databases/ruby-memcache: Development has been abandoned upstream
2013-11-01 deskutils/blogtk: This is dead upstream and is no longer being maintained
While here:
* remove indefinite article from COMMENT
* make USE_DOS2UNIX unconditional (for all patches)
* make use of OSVERSION for FreeBSD only
* convert USE_GMAKE to USES format
Approved by: portmgr (bapt, implicit)
with new versions of texinfo. [1]
Adopt the new LIB_DEPENDS standard and replace USE_GMAKE by USES=gmake.
Remove an obsolete dependency on bison. [2]
STAGEify.
PR: 183342 [1]
Reported by: Christoph Moench-Tegeder <cmt@burggraben.net> [1], marino [2]
With the 2.6.9 release, and five years after its first release, the
Python 2.6 series is now officially retired. All official maintenance
for Python 2.6, including security patches, has ended. [1]
Please migrate to lang/python27 or later as soon as possible.
lang/python26 will expire and be removed in January 2014.
[1] http://www.python.org/download/releases/2.6.9/
With hat on: python@
Reviewed by: koobs@
Changes:
* Compiler and runtime library have been improved to support
big-endian systems. Thanks go to Brian Callahan, for his support
and for providing access to his PowerPC Macintosh under OpenBSD.
* The list of source files and the chapter about the defines used in
version.h have been improved in src/read_me.txt.
* The library tls.s7i has been improved to avoid raising an
exception, when the remote side closes the socket.
* The program chkccomp.c has been improved to write defines for
CHECK_INT_DIV_BY_ZERO, INT_DIV_BY_ZERO_SIGNALS and
DO_SIGFPE_WITH_DIV_BY_ZERO to version.h.
* Documentation comments have been added or improved in fillib.c,
int_rtl.c, soclib.c, tim_dos.c, tim_unx.c, tim_win.c, ut8lib.c,
ut8_rtl.c, tls.s7i and bytedata.s7i.
* The generic compare functions bigCmpGeneric, bstCmpGeneric,
chrCmpGeneric, intCmpGeneric, rflCmpGeneric, setCmpGeneric,
strCmpGeneric and typCmpGeneric have been improved to work
endianness neutral. Now they work correct on little- and
big-endian systems.
* The generic create functions bigCreateGeneric, bstCreateGeneric,
drwCreateGeneric, rflCreateGeneric, setCreateGeneric,
strCreateGeneric and itfCreateGeneric have been improved to work
endianness neutral. Now they work correct on little- and
big-endian systems.
* The new generic copy functions bigCpyGeneric, bstCpyGeneric,
drwCpyGeneric, fltCpyGeneric, rflCpyGeneric, setCpyGeneric and
strCpyGeneric have been added. This functions work endianness
neutral.
* The new generic destroy functions bigDestrGeneric,
drwDestrGeneric, rflDestrGeneric, setDestrGeneric and
strDestrGeneric have been added. This functions work endianness
neutral.
* The functions uintCmpGeneric, intCreateGeneric and intDestrGeneric
have been renamed to genericCmp, genericCreate respectively
genericDestr.
* The functions genericCpy, ptrCmp, ptrCmpGeneric, ptrCpyGeneric and
ptrCreateGeneric have been added to int_rtl.c.
* The function cmdConfigValue in cmd_rtl.c has been improved to
support the config value DO_SIGFPE_WITH_DIV_BY_ZERO.
* The handling of the option -l in processOptions (file s7.c) has
been improved to work endianness neutral.
* Several functions have been optimized:
In infile.c: get_file_name and get_file_name_ustri
In prg_comp.c: prgSyobject
In rfllib.c: rfl_elemcpy and rfl_idx
In rfl_data.c: rflElemcpy, rflIdx and rflTail
* The function get_ident and the macro IDENT_TABLE have been changed
to have an additional progtype argument.
* The function typNum in typ_data.c has been improved to use a
hashtable.
* In typlib.c the functions typ_num and typ_str have been improved
to call typNum respectively typStr.
* Support for the actions ACT_ICONV and ACT_ORD has been added to
interpreter, compiler and runtime library.
* The function id_string2 has been added to datautl.c, and used in
ref_data.c and typ_data.c.
- Switched to automake 1.11.6, see CVE-2012-3386.
- #14669: Fixed extraction of CC from gmp.h.
- Fixed case of intermediate zero real or imaginary part in mpc_fma,
found by hydra with GMP_CHECK_RANDOMIZE=1346362345.
This is on top of the following changes from version 1.0
- Licence change towards LGPLv3+ for the code and GFDLv1.3+ (with no
invariant sections) for the documentation.
- 100% of all lines are covered by tests
- Renamed functions
. mpc_mul_2exp to mpc_mul_2ui
. mpc_div_2exp to mpc_div_2ui
- 0^0, which returned (NaN,NaN) previously, now returns (1,+0).
- Removed compatibility with K&R compilers, which was untestable due
to lack of such compilers.
- New functions
. mpc_log10
. mpc_mul_2si, mpc_div_2si
- Speed-ups
. mpc_fma
- Bug fixes
. mpc_div and mpc_norm now return a value indicating the effective
rounding direction, as the other functions.
. mpc_mul, mpc_sqr and mpc_norm now return correct results even if
there are over- or underflows during the computation.
. mpc_asin, mpc_proj, mpc_sqr: Wrong result when input variable has
infinite part and equals output variable is corrected.
. mpc_fr_sub: Wrong return value for imaginary part is corrected.
Convert to the new LIB_DEPENDS standard and remove hard-coded
.so versions from a couple of dependent ports.
Bump PORTREVISIONS of all dependent ports.
PR: 183141
Approved by: portmgr (bdrewery)
- Add LDFLAGS to unbreak the ODBC build.
- Remove unnecessary USE_CSTD (see r331232). [1]
- Remove build_plt target, which is overkill for normal usage anyway.
PR: ports/180017 [1]
Submitted by: Stefan Grundmann <sg2342@googlemail.com> [1]
- MKDIR is silent
- ECHO is silent
- INSTALL_* are not silent
- CP/FIND/... are not silent
I fixed a few PORTDOCS misusage, I'll do a second pass.
With Hat: perl@
2013-10-22 games/magicmaze: Does not work with Ruby 1.9
2013-10-10 lang/ruby18: Please use lang/ruby19
2013-10-10 www/rubygem-mongrel_cluster: Does not work with Ruby 1.9
2013-10-10 www/rubygem-mongrel: Does not work with Ruby 1.9
2013-10-10 www/rubygem-rails-app-installer: Does not work with Ruby 1.9
2013-10-22 www/typo: Does not work with Ruby 1.9
is enabled. It was added in r300328 to fix compilation with clang but
testing has shown it is no longer necessary.
PR: ports/180017
Tested by: Stefan Grundmann <sg2342@googlemail.com>
These ports install binaries with "32" suffixes and most data lives in
${PREFIX}/llvm32 so there are no conflicts.
NOTE: devel/llvm and lang/clang will shortly be removed in favor of
these ports.
Switch to USES=gmake.
STAGIFY.
Don't try to build when the default compiler uses libc++.
Initiate deorbit burn sequence.
Submitted by: John Marino <freebsd@marino.st> [1]
Changes:
20131013:
- Libraries for TLS (Transport Layer Security), SSL (Secure Sockets
Layer) and HTTPS (HyperText Transfer Protocol Secure) support have
been added. The new libraries are: arc4.s7i, asn1.s7i, cipher.s7i,
gethttps.s7i, hmac.s7i, pkcs1.s7i, tls.s7i and x509cert.s7i.
- Comparison operators for < (proper subset) and > (proper superset)
have been added to hashsetof.s7i.
- The type digestAlgorithm and the functions msgDigest, blockSize
and digestSize have been added to msgdigest.s7i.
- The function hex has been added to bytedata.s7i.
- The example program castle.sd7 has been improved to avoid, that
an item can be dropped on an illegal place.
- The library make.s7i has been improved to accept the option -p for
the cp command.
- Documentation comments have been added to sockbase.s7i,
socket.s7i, atom.c, entutl.c, soclib.c, soc_rtl.c and striutl.c.
- The padding in the message digest functions of msgdigest.s7i has
been corrected. Now the functions pad nothing instead of 64 bytes,
when the message length is 55 (mod 64).
- The function getHttp in gethttp.s7i has been improved to use
a case-insensitive comparison for HTTP header field names.
- In exec.c the functions eval_arg_list and dump_arg_list have been
improved to avoid reading from freed memory. This occured in very
rare cases.
- The functions find_action (in actutl.c), fltParse (in flt_rtl.c),
cmdConfigValue (in cmd_rtl.c) and find_sysvar (in syvarutl.c)
have been improved to use the macros MAX_STRI_EXPORT_LEN and
max_utf8_size.
- Interpreter and runtime library have been improved to ignore the
signal SIGPIPE. Writing to a broken pipe or socket raises the
exception FILE_ERROR instead.
- The functions bln_and and bln_or in blnlib.c have been improved
to return either SYS_TRUE_OBJECT or SYS_FALSE_OBJECT.
- The program chkccomp.c has been improved to determine the
alignment used by malloc and whether an access to unaligned memory
succeeds.
- Usages of ALLOC_CHUNK in flistutl.h have been improved to allocate
aligned memory the same way as malloc returns aligned memory.
- The program chkccomp.c has been improved to define the macro
'inline', when the C compiler does not support the keyword inline.
- Occurances of 'INLINE' have been replaced with 'inline' in the
files analyze.c, atom.c, big_gmp.c, big_rtl.c, chr_rtl.c, exec.c,
findid.c, infile.c, int_rtl.c, numlit.c, parser.c, rfl_data.c,
scanner.c, striutl.c, strlib.c, str_rtl.c, syntax.c and ut8_rtl.c.
20130929:
- The functions bStriLe, bytesLe, bStriBe and bytesBe have been
added to bytedata.s7i. This functions convert an integer to a
sequence of bytes.
- The compiler has been improved to generate inline code for the
action STR_PUSH. This action is used when a character is appended
to a string (with the operator &:= ).
- The function sendGet has been added to gethttp.s7i.
- The functions str_yy_mm_dd and str_hh_mm_ss have been added to
time.s7i.
- Checks for +:= and -:= have been added to chkbig.sd7.
- Checks for bytesBe and bytesLe have been added to chkint.sd7.
- Checks for & and &:= have been added to chkstr.sd7.
- The function doLocalDeclsOfStatement has been added to the
compiler. This function is used to simplify the processing
of various statements.
- The functions that process the statement actions BIG_GROW,
BIG_LSHIFT_ASSIGN, BIG_MULT_ASSIGN, BIG_RSHIFT_ASSIGN,
BIG_SHRINK, BLN_CPY, CHR_CPY, CHR_GROW, CHR_SHRINK, ENU_CPY,
FIL_CPY, FLT_CPY, FLT_DIV_ASSIGN, FLT_GROW, FLT_MULT_ASSIGN,
FLT_SHRINK, INT_AND_ASSIGN, INT_CPY, INT_GROW, INT_LSHIFT_ASSIGN,
INT_MULT_ASSIGN, INT_OR_ASSIGN, INT_RSHIFT_ASSIGN, INT_SHRINK,
INT_URSHIFT_ASSIGN, REF_CPY, RFL_INCL, SET_EXCL, SET_INCL,
SOC_CPY, STR_PUSH and TYP_CPY in the compiler have been
simplified.
- The functions int_toBStriBe and int_toBStriLe have been added to
intlib.c.
- The functions intToBStriBe and intToBStriLe have been added to
int_rtl.c
- In big_rtl.c and big_gmp.c the functions bigFromBytesBe and
bigFromBytesLe have been renamed to bigFromByteBufferBe
respectively bigFromByteBufferLe.
- The performance of the functions bst_cmp (in bstlib.c), bstCmp
(in bst_rtl.c) and strMult (in str_rtl.c) has been improved.
- The macro 'unlikely' has been added to functions in bst_rtl.c.
STAGIFY
Mark dragonegg IGNORE on systems where the default compiler defaults to
libc++ on the grounds the the clang/llvm ports are likely linked with it
and thus won't work with gcc and libstdc++.
2013-10-10 www/ruby-nora: Does not work with Ruby 1.9
www/aswiki
2013-10-10 textproc/ruby-amrita: Does not work with Ruby 1.9
2013-10-15 security/flowtag: Does not work with Ruby 1.9
2013-10-10 lang/ruby-lua4: Does not work with Ruby 1.9
2013-10-15 dns/dnsdoctor: Does not work with Ruby 1.9
2013-10-10 audio/ruby-xmms: Does not work with Ruby 1.9
2013-10-10 databases/ruby-interbase: Does not work with Ruby 1.9
2013-10-10 databases/ruby-bdb1: Does not work with Ruby 1.9
2013-10-10 devel/ruby-fam: Does not work with Ruby 1.9
2013-10-10 devel/ruby-rcov: Does not work with Ruby 1.9
2013-10-10 devel/ruby-rudl: Does not work with Ruby 1.9
2013-10-10 devel/ruby-gemfinder: Does not work with Ruby 1.9
2013-10-10 devel/ruby-sysvipc: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-sdl: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-ncurses: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-parsetree: Does not work with Ruby 1.9
2013-10-10 devel/ruby-sexp: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-getopt-declare: Does not work with Ruby 1.9
2013-10-10 devel/ruby-slang: Does not work with Ruby 1.9
2013-10-10 japanese/ruby-slang: Does not work with Ruby 1.9
2013-10-10 devel/ruby-gettext: Does not work with Ruby 1.9
2013-10-10 devel/ruby-jttui: Does not work with Ruby 1.9
2013-10-10 devel/ruby-mmap: Does not work with Ruby 1.9
2013-10-10 devel/ruby-racc: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-rparsec: Does not work with Ruby 1.9
2013-10-10 devel/rubygem-zoom: Does not work with Ruby 1.9
2013-10-10 games/ruby-exmars: Does not work with Ruby 1.9
2013-10-10 graphics/ruby-pgplot: Does not work with Ruby 1.9
2013-10-10 graphics/ruby-opengl: Does not work with Ruby 1.9
2013-10-10 graphics/rubygem-turing: Does not work with Ruby 1.9
2013-10-10 japanese/ruby-chasen: Does not work with Ruby 1.9
2013-10-10 japanese/ruby-gyokuro: Does not work with Ruby 1.9
2013-10-10 japanese/ruby-rdic: Does not work with Ruby 1.9
2013-10-10 lang/ruby-mode.el: Does not work with Ruby 1.9
2013-10-10 misc/rubygem-ohcount: Does not work with Ruby 1.9
2013-10-10 net-im/rubygem-xmpp4r: Does not work with Ruby 1.9
2013-10-10 security/ruby-acl: Does not work with Ruby 1.9
2013-10-10 textproc/ruby-xslt: Does not work with Ruby 1.9
2013-10-10 textproc/ruby-erbscan: Does not work with Ruby 1.9
2013-10-10 textproc/ruby-mwdom: Does not work with Ruby 1.9
2013-10-10 textproc/ruby-xmlscan-old: Does not work with Ruby 1.9
2013-10-10 textproc/rubygem-formosa: Does not work with Ruby 1.9
2013-10-10 textproc/rubygem-htmltools: Does not work with Ruby 1.9
2013-10-10 www/ruby-cruisecontrolrb: Does not work with Ruby 1.9
2013-10-10 www/rubygem-rubyfulsoup: Does not work with Ruby 1.9
2013-10-10 x11/ruby-gtktrayicon: Does not work with Ruby 1.9
Yap-6.3.3 is the development version of lang/yap (6.2.2).
YAP is a high-performance Prolog compiler developed at
LIACC/Universidade do Porto and at COPPE Sistemas/UFRJ. Its Prolog
engine is based in the WAM (Warren Abstract Machine), with several
optimizations for better performance. YAP follows the Edinburgh
tradition, and is largely compatible with the ISO-Prolog standard and
with Quintus and SICStus Prolog.
PR: ports/181334
Submitted by: Danilo Egea Gondolfo <danilogondolfo@gmail.com>
Approved by: wg/culot (mentors)
- Enable stage support
- Add work-around to fix OS detection, enables TCP/IP support
- Add work-around to fix dlopen detection, enables compilation support
The following ports will not build with a binutils 2.22+ linker built
with standard options. This has been obvious with DPorts, but difficult
to see on FreeBSD. However, setting the ports compiler as a recent gcc
(e.g lang/gcc48) is an excellent way to detect the unspecified but needed
libraries as these recent GCC compilers use the latest binutils linkers.
These patches were tested on FreeBSD 8.4 and DragonFly 3.5
Pending issues:
- Support for long doubles was disabled (does not build on 9.x.)
- Using readline/rep causes a core dump on SIGWINCH.
- Calls to (find-seconds 0 0 0 1 5 1970 #f) fail, so a bit of
documentation was cut out as it broke the build by using this
specific call.
install absolute symlinks pointing to the stagedir! Poudriere has
been improved to check for this issue now.
With hat: portmgr
Reported by: madpilot, jpaetzel
- old USE_ZOPE knob support was removed from bsd.python.mk
- update CHANGES and bsd.sanity.mk accordingly
- add ZOPE options knob and use it in lang/py-mx-base
The work is done by Marcus von Appen, but any problems are mine.
Submitted by: mva (python ML)
You must switch to DEFAULT_VERSIONS (see ports/UPDATING entry 20130920 by bapt@)
if you are still using PERL_VERSION to define a specific version requirement.
- bump portrevision
Reviewed by: bapt@
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
binary in $PATH. It can be set in case the user wants to use a
specific python2 version as a default. Its behaviour is similar
to the existing PYTHON3_DEFAULT_VERSION and fixes a small issue with
package builds for different python versions.
- Outline that PYTHON_DEFAULT_VERSION always takes precedence for a
specific python major version.
- Update lang/python2 to use PYTHON2_DEFAULT_VERSION and bump the
PORTREVISION to let the installed port catch up with the change.
- Allow PYTHON3_DEFAULT_VERSION to overriden by a user choice, if
PYTHON_DEFAULT_VERSION is not set to a python3 port.
Reported by: David Demelier <demelier.david@gmail.com>
Reviewed by: koobs@, sbz@
With hat on: python@
The simplexml module uses spl which uses pcre (php53,54,55). For php52,
the fix is applied directly to spl. In any case, the bundled PCRE is not
used, so it needs to be able to find the external pcre library and related
headers. Unfortunately, there is no include for ${LOCALBASE}/include
directory. This commit adds to fix simplexml (or spl) module build.
Approved by: portmgr (bapt)
7.6.3-based bootstrap compilers without any external dependencies [1]
- Update bootstrap hscolour to 1.20.3, so both 7.4 and 7.6 bootstraps can
work with it
- Remove the line for the 7.x bootstrap as no such one exists any more
Thanks to: cognet [1]
Obtained from: FreeBSD Haskell
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
Changes:
* The compiler has been improved to use charvalue instead of
intvalue, when characters are initialized. Thanks go to Brian
Callahan, for reporting the bug. Btw.: Brian Callahan created an
openBSD port for Seed7.
* The makefile mk_freebsd.mk has been added to support compilation
under freeBSD. Thanks go to Pietro Cerutti, for providing it.
* The makefiles mk_clang.mak and mk_freebsd.mk have been improved
to use the options -ffunction-sections, -fdata-sections and
*Wl,--strip-debug.
* The functions bStriLe2BigInt, bytesLe2BigInt, bStriBe2BigInt,
bytesBe2BigInt, bStriLe, bytesLe, bStriBe and bytesBe have been
added to bytedata.s7i. This functions convert between sequences
of bytes and bigInteger numbers.
* In bytedata.s7i the functions bytesLeAsInt and bytesBeAsInt have
been renamed to bytesLe2Int respectively bytesBe2Int.
* Parameter names have been added to seed7_05.s7i, graph_file.s7i,
graph.s7i, reference.s7i, pac.sd7, tetg.sd7, tet.sd7 and wiz.sd7.
* Checks for the Karatsuba multiplication and the functions bytesBe,
bytesLe, bytesBe2BigInt and bytesLe2BigInt have been added to
chkbig.sd7.
* Checks for set membership in character sets have been added to
chkset.sd7.
* The readability of the bitset initialisation in the compiler (s7c)
has been improved.
* Interpreter and compiler have been improved to support the new
actions BIG_FROM_BSTRI_BE, BIG_FROM_BSTRI_LE, BIG_TO_BSTRI_BE and
BIG_TO_BSTRI_LE.
* The functions bigFromBytesBe, bigFromBytesLe, bigFromBStriBe,
bigFromBStriLe, bigToBStriBe and bigToBStriLe have been added to
both, big_gmp.c and big_rtl.c.
* In big_gmp.c and big_rtl.c the function bigImport has been
rewritten to use bigFromBytesBe.
* The readability of the function uint2_mult in int_rtl.c has been
improved.
* The program chkccomp.c has been improved to check for a ones
complement inttype and if casting can be used to get the element
of a union.
* Several makefiles have been adjusted to use 'cp -p' instead of
'cp -a'. The cp command openBSD does not support -a.
* The function bigRand in big_rtl.c has been improved to avoid that
a random number is outside the required range.
- Replace NO_LATEST_LINK with PKGNAMESUFFIX to accomodate pkg.
- Update to 3.3.1.2013.09.07 which brings experimental support for
STMicroelectronics STM8.
if present (even if the libiconv port is present and requested) 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
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
if present instead of relying on the port we actually depend on.
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
if present instead of relying on the port we actually depend on.
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.
Interestingly enough, lang/gcc46 and lang/gcc48 did not fail under the
same test scenarios and 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)
after r254273
- Fix a bunch of ports to properly work after this
- Mark converters/libiconv as IGNORE for systems with iconv in libc
Reviewed by: bapt
Approved by: portmgr (bapt)
Discussed with: bapt, bsam (who both contributed ideas and code)
Work around a bug in libcpp that pulls in the optional system iconv
if present instead of relying on the port we actually depend on. [1]
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.
Interestingly enough, lang/gcc46 and lang/gcc48 did not fail under the
same test scenarios and libstdc++ does not seem to require adjustments,
so PR 161417 may not be relevant in full.
PR: 161417 [1]