On 2026/10/20 17:54, Kirill A. Korinsky wrote:
> On Tue, 20 Oct 2026 17:25:19 +0200,
> Stuart Henderson <[email protected]> wrote:
> >
> > On 2026/10/20 14:39, Kirill A. Korinsky wrote:
> > > ports@,
> > >
> > > I think GCC49_ARCHS is very misslead and using PORTSGCC_ARCHS instead
> > > makes sense.
> >
> > > @@ -40,10 +40,10 @@ LLD_ARCHS = aarch64 amd64 arm i386 power
> > > # as well as available for PROPERTIES checks.
> > > LLVM_ARCHS = aarch64 amd64 arm i386 mips64 mips64el powerpc powerpc64
> > > riscv64 sparc64
> > > # arches where ports-gcc >4.9 exists. To be used again for modules
> > > -GCC49_ARCHS = aarch64 alpha amd64 arm hppa i386 mips64 mips64el powerpc
> > > powerpc64 riscv64 sparc64
> > > +PORTSGCC_ARCHS = aarch64 alpha amd64 arm hppa i386 mips64 mips64el
> > > powerpc powerpc64 riscv64 sparc64
> > >
> > > # arches where there is a C++11 compiler, either clang in base or
> > > ports-gcc
> > > -CXX11_ARCHS = ${CLANG_ARCHS} ${GCC49_ARCHS}
> > > +CXX11_ARCHS = ${CLANG_ARCHS} ${PORTSGCC_ARCHS}
> > > DEBUGINFO_ARCHS = aarch64 amd64
> > >
> > > .for PROP in ALL APM BE LE LP64 CLANG GCC4 GCC49 MONO LLVM LUAJIT \
> > ^^^^^
> > this needs changing too
> >
>
> As this step?
I think it should be - the above xx_ARCHS are used to set 'xx' in PROPERTIES:
.for PROP in ALL APM BE LE LP64 CLANG GCC4 GCC49 MONO LLVM LUAJIT \
CXX11 OCAML_NATIVE OCAML_NATIVE_DYNLINK GO \
LLD RUST DEBUGINFO
. for A B in ${MACHINE_ARCH} ${ARCH}
. if !empty(${PROP}_ARCHS:M$A) || !empty(${PROP}_ARCHS:M$B)
PROPERTIES += ${PROP:L}
. endif
. endfor
.endfor
I just fixed math/hdf5 and telephony/asterisk because neither of them
need to set ONLY_FOR_ARCHS themselves, so no need to touch those any
more.
> I had the next step with rename gcc4.port.mk to portsgcc.port.mk where I
> removed that GCC49 completley.
