Date: Fri, 28 Apr 2017 19:58:02 -0700 From: Mark Millard <markmi@dsl-only.net> To: Mark Linimon <linimon@lonesome.com> Cc: svn-ports-head@freebsd.org, Baptiste Daroussin <bapt@FreeBSD.org>, Alexander Kabaev <kabaev@gmail.com> Subject: Re: svn commit: r439595 - in head/devel: aarch64-gcc aarch64-none-elf-gcc amd64-gcc arm-none-eabi-gcc arm-none-eabi-gcc492 mips-gcc mips64-gcc powerpc64-gcc riscv64-gcc sparc64-gcc Message-ID: <A6D8E222-75C6-4835-8847-938BE967AF9D@dsl-only.net> In-Reply-To: <A6D8B25D-27F6-4216-BD96-1C07E08E7847@dsl-only.net> References: <8E45FA57-8D2E-4159-8E02-6A5044000CC2@dsl-only.net> <20170429022831.GB15674@lonesome.com> <A6D8B25D-27F6-4216-BD96-1C07E08E7847@dsl-only.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2017-Apr-28, at 7:41 PM, Mark Millard <markmi at dsl-only.net> wrote:
> On 2017-Apr-28, at 7:28 PM, Mark Linimon <linimon at lonesome.com> wrote:
>
>> On Fri, Apr 28, 2017 at 03:27:31PM -0700, Mark Millard wrote:
>>> I do not have access to mips, mips64, riscv, sparc64
>>> so they could be targets but not hosts in my tests
>>
>> I don't think we have any native mips, mips64, or riscv package build
>> capability right now.
>>
>> I have sparc64 native. I have never gotten around to trying any, but
>> I can.
>>
>>> I have access to powerpc but currently am not well set up to use it
>>> without rebuilding it
>>
>> I have powerpc64 native and am currently attempting builds.
>
> I do have access to powerpc64 and tested it in my context.
> It is powerpc that I'm far from an appropriate configuration
> of for now.
>
> FYI: The D10537 patch for devel/[powerpc64-gcc looks like:
>
> # svnlite diff /usr/ports/devel/powerpc64-gcc
> Index: /usr/ports/devel/powerpc64-gcc/Makefile
> ===================================================================
> --- /usr/ports/devel/powerpc64-gcc/Makefile (revision 439263)
> +++ /usr/ports/devel/powerpc64-gcc/Makefile (working copy)
> @@ -74,11 +74,19 @@
> .endif
>
> post-install:
> - @if [ -f ${STAGEDIR}${PREFIX}/bin/cpp ] ; then \
> - mv ${STAGEDIR}${PREFIX}/bin/cpp \
> - ${STAGEDIR}${PREFIX}/bin/${GCC_TARGET}-cpp ; \
> +.for f in cpp gcov gcov-tool
> + @if [ -f ${STAGEDIR}${PREFIX}/bin/$f ] ; then \
> + ${MV} ${STAGEDIR}${PREFIX}/bin/$f \
> + ${STAGEDIR}${PREFIX}/bin/${GCC_TARGET}-$f ; \
> fi
> +.endfor
> .for f in c++ cpp g++ gcc gcc-ar gcc-nm gcc-ranlib gcov gcov-tool
> + @if [ ! -f ${STAGEDIR}${PREFIX}/man/man1/${GCC_TARGET}-$f.1 ] ; then \
> + if [ -f ${STAGEDIR}${PREFIX}/man/man1/$f.1 ] ; then \
> + ${CP} ${STAGEDIR}${PREFIX}/man/man1/$f.1 \
> + ${STAGEDIR}${PREFIX}/man/man1/${GCC_TARGET}-$f.1 ; \
> + fi \
> + fi
> @${RM} ${STAGEDIR}${PREFIX}/bin/$f
> @${RM} ${STAGEDIR}${PREFIX}/man/man1/$f.1
> @${RM} ${STAGEDIR}${PREFIX}/man/man1/$f.1.gz
>
> It contributes to all its slave ports.
>
> As for the other part of D10537's patch (the
> pkg-plist part of the below):
>
> # svnlite diff /usr/ports/devel/amd64-gcc
> Index: /usr/ports/devel/amd64-gcc/Makefile
> ===================================================================
> --- /usr/ports/devel/amd64-gcc/Makefile (revision 439263)
> +++ /usr/ports/devel/amd64-gcc/Makefile (working copy)
> @@ -8,6 +8,4 @@
> BU_PREFIX= x86_64-${OPSYS:tl}
> FREEBSD_EXTENSION= yes
>
> -BROKEN_powerpc64= Does not build
> -
> .include "${MASTERDIR}/Makefile"
> Index: /usr/ports/devel/amd64-gcc/pkg-plist
> ===================================================================
> --- /usr/ports/devel/amd64-gcc/pkg-plist (revision 439263)
> +++ /usr/ports/devel/amd64-gcc/pkg-plist (working copy)
> @@ -6,6 +6,8 @@
> bin/%%GCC_TARGET%%-gcc-ar
> bin/%%GCC_TARGET%%-gcc-nm
> bin/%%GCC_TARGET%%-gcc-ranlib
> +bin/%%GCC_TARGET%%-gcov
> +bin/%%GCC_TARGET%%-gcov-tool
> lib/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/include-fixed/README
> lib/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/include-fixed/limits.h
> lib/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/include-fixed/netinet/ip_fil.h
> @@ -489,3 +491,7 @@
> libexec/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/lto-wrapper
> libexec/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/lto1
> libexec/gcc/%%GCC_TARGET%%/%%GCC_VERSION%%/plugin/gengtype
> +man/man1/%%GCC_TARGET%%-cpp.1.gz
> +man/man1/%%GCC_TARGET%%-g++.1.gz
> +man/man1/%%GCC_TARGET%%-gcc.1.gz
> +man/man1/%%GCC_TARGET%%-gcov.1.gz
>
> devel/amd64-gcc was missing 6 files in pkg-plist
> that devel/aarch64-gcc and devel/powerpc64-gcc
> had in place. Plus I enabled attempting the build.
A likely powerpc64 difference in our contexts is
my use of a system-clang-based/libc++ based
head -r317015 vs. you not being likely to use
a system-clang based powerpc64 environment.
So we are likely using different compilers
to compile any devel/*-gcc when compiled
on powerpc64 (no qemu or the like involved):
clang vs. gcc (4.2.1?) and possibly libc++
vs. old(?) libstdc++ .
I do not expect that clang and gcc get the same
result for devel/amd64-gcc --but I've only tested
clang doing the compiles.
Also I use -r436731 for devel/*binutils (so
2.27) and you may be using 2.28. That may be
true of more than just powerpc64.
I also use:
# svnlite diff /usr/ports/Mk/
Index: /usr/ports/Mk/bsd.port.mk
===================================================================
--- /usr/ports/Mk/bsd.port.mk (revision 439263)
+++ /usr/ports/Mk/bsd.port.mk (working copy)
@@ -1644,7 +1644,11 @@
STRIP_CMD= ${TRUE}
.endif
DEBUG_FLAGS?= -g
+.if defined(ALLOW_OPTIMIZATIONS_FOR_WITH_DEBUG)
+CFLAGS:= ${CFLAGS} ${DEBUG_FLAGS}
+.else
CFLAGS:= ${CFLAGS:N-O*:N-fno-strict*} ${DEBUG_FLAGS}
+.endif
.if defined(INSTALL_TARGET)
INSTALL_TARGET:= ${INSTALL_TARGET:S/^install-strip$/install/g}
.endif
and. . .
# more /etc/make.conf
WANT_QT_VERBOSE_CONFIGURE=1
#
DEFAULT_VERSIONS+=perl5=5.24
WRKDIRPREFIX=/usr/obj/portswork
#
# From a local /usr/ports/Mk/bsd.port.mk extension:
ALLOW_OPTIMIZATIONS_FOR_WITH_DEBUG=
#
.if ${.CURDIR:M*/devel/*llvm*}
#WITH_DEBUG=
.elif ${.CURDIR:M*/www/webkit-qt5*}
#WITH_DEBUG=
.else
WITH_DEBUG=
.endif
WITH_DEBUG_FILES=
MALLOC_PRODUCTION=
which together mean having both optimization
and debug information for most things.
( devel/*llvm* and www/webkit-qt5* are massive
builds when WITH_DEBUG= is used and so I avoid
such.)
===
Mark Millard
markmi at dsl-only.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?A6D8E222-75C6-4835-8847-938BE967AF9D>