On Tue, Mar 19 2019, Steven Mestdagh <[email protected]> wrote:
> Charlene Wendling [2019-03-18, 14:57:23]:
>> Hi ports, Steven,
>>
>> > http://build-failures.rhaalovely.net/sparc64/2019-02-03/graphics/pstoedit.log
>> > http://build-failures.rhaalovely.net/powerpc/last/graphics/pstoedit.log
>>
>> It requires -std=c++11 once again, it builds fine on macppc after it's
>> provided [1]. While here i've fixed a spacing inconsistency in COMPILER.
>>
>> I didn't bump REVISION, since pstoedit-3.73 has never been built on
>> ports-gcc archs.
>>
>> Any comment?
>>
>> Charlène.
>>
>> [1] http://0x0.st/z8SZ.txt
>>
>> Index: Makefile
>> ===================================================================
>> RCS file: /cvs/ports/graphics/pstoedit/Makefile,v
>> retrieving revision 1.27
>> diff -u -p -u -p -r1.27 Makefile
>> --- Makefile 5 Jan 2019 10:08:33 -0000 1.27
>> +++ Makefile 18 Mar 2019 13:36:47 -0000
>> @@ -26,9 +26,15 @@ CONFIGURE_ARGS= --without-libplot \
>> --without-magick
>> WANTLIB= c ${COMPILER_LIBCXX} m
>>
>> -COMPILER = base-clang ports-gcc base-gcc
>> +# c++11
>> +COMPILER = base-clang ports-gcc base-gcc
>>
>> post-install:
>> ${INSTALL_MAN} ${WRKSRC}/doc/pstoedit.1 ${PREFIX}/man/man1
>>
>> .include <bsd.port.mk>
>> +
>> +# XXX this should be retried once moving to ports-gcc>=8
I guess we could add a guard for this kind of workaround, as in:
&& ${MODGCC4_VERSION} == "4.9"
But this should probably be done for the whole tree, in one sweep, after
testing with gcc8.
>> +.if ${CHOSEN_COMPILER} == "ports-gcc"
>> +CONFIGURE_ENV += CXXFLAGS="${CXXFLAGS} -std=c++11"
>> +.endif
I would suggest -std=gnu++11 here, if only because -std=gnu... is what
gcc and clang default to. But I must admit that I don't know what it
actually changes in C++ land.
But I see that you have already used -std=c++11 in other parts of the
tree, maybe it's not a problem in practice?
Anyway, ok jca@
> sure, but -std=c++11 is also supported by clang, we could always apply it?
clang defaults to -std=gnu++14 so this would change the defaults on
clang archs, possibly warranting a REVISION bump.
--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
