CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/18 05:06:41 Modified files: devel : Makefile Log message: +p5-CLI-Framework
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs
Module name:ports
Changes by: c...@cvs.openbsd.org2019/10/18 05:05:30
Log message:
Import p5-CLI-Framework-0.05
CLI::Framework ("CLIF") provides a framework and conceptual pattern
for building full-featured command line applications. It intends
to make this process simple and consistent. It assumes the
responsibility of implementing details that are common to all
command-line applications, making it possible for new applications
adhering to well-defined conventions to be built without the need
to repeatedly write the same command-line interface code.
From Peter Ezetta , with tweaks from afresh1@
OK afresh1@
Status:
Vendor Tag: cwen
Release Tags: cwen_20191018
N ports/devel/p5-CLI-Framework/Makefile
N ports/devel/p5-CLI-Framework/distinfo
N ports/devel/p5-CLI-Framework/pkg/PLIST
N ports/devel/p5-CLI-Framework/pkg/DESCR
No conflicts created by this import
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs
Module name:ports
Changes by: c...@cvs.openbsd.org2019/10/18 04:39:09
Modified files:
math/double-conversion: Makefile
Log message:
double-conversion: ${COMPILER_LIBCXX} in WANTLIB should come with a
COMPILER line including ports-gcc in it. This fixes the build on macppc.
Also tested on sparc64 by kmos@
OK rsadowski@ kmos@
Re: [macppc] Unbreak graphics/DevIL
On Thu, 17 Oct 2019 20:52:47 -0400
Kurt Mosiejczuk wrote:
> On Fri, Oct 18, 2019 at 01:51:35AM +0200, Charlene Wendling wrote:
>
> > (Pascal, only this port seems to have this issue, it's FYI)
>
> > > http://build-failures.rhaalovely.net/powerpc/2019-09-17/graphics/DevIL.log
>
> > __vector is an AltiVec keyword. AltiVec is not found by DevIL's
> > configure script, which is actually what we want (G3 machines don't
> > support AltiVec).
>
> > The "problem" comes from gcc-8's _mm_malloc(), that uses that
> > keyword. It seems to me that the old gcc-4.9 didn't have
> > _mm_malloc(), so that's why it's broken since gcc-8 is ports-gcc.
>
> > As such, i've disabled the use of _mm_malloc. While here i've
> > updated WANTLIB.
>
> > It builds fine on amd64, and macppc [0], where i have been able to
> > fix (diff incoming), build and run games/love, its only consumer on
> > macppc.
>
> > Comments/feedback are welcome!
>
> > Charl??ne.
>
> amd64 and sparc64 still build fine with this patch. Although I do get
> a complaint during packaging (and from portcheck) of:
> "LIB_DEPENDS graphics/openjpeg not needed for graphics/DevIL ?"
>
> You didn't touch LIB_DEPENDS though and it happens before your
> change... so OK kmos
It does not seem used at all indeed. Here is an alternate diff that
works fine as well:
Index: Makefile
===
RCS file: /cvs/ports/graphics/DevIL/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:46:56 - 1.26
+++ Makefile18 Oct 2019 01:25:18 -
@@ -3,7 +3,7 @@
COMMENT= library for powerful image loading capabilities
DISTNAME= DevIL-1.7.8
-REVISION= 14
+REVISION= 15
SHARED_LIBS+= IL 1.0 # 2.0
SHARED_LIBS+= ILU 1.0 # 2.0
SHARED_LIBS+= ILUT1.0 # 2.0
@@ -18,11 +18,12 @@ PERMIT_PACKAGE= Yes
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=openil/}
EXTRACT_SUFX= .zip
-WANTLIB += GL GLU Half ICE Iex IexMath IlmImf IlmThread Imath
-WANTLIB += SDL SM X11 X11-xcb Xdamage Xext Xfixes Xi
-WANTLIB += Xmu Xrandr Xrender Xt Xxf86vm c drm glapi glut iconv jasper
-WANTLIB += jpeg lcms m mng png pthread sndio ${COMPILER_LIBCXX}
-WANTLIB += tiff usbhid xcb xcb-dri2 xcb-glx z expat
+WANTLIB += ${COMPILER_LIBCXX} GL GLU Half ICE Iex IexMath IlmImf
+WANTLIB += IlmThread Imath SDL SM X11 X11-xcb Xdamage Xext Xfixes
+WANTLIB += Xi Xmu Xrandr Xrender Xt Xxf86vm c drm expat glapi
+WANTLIB += glut iconv jasper jpeg lcms m mng png sndio tiff usbhid
+WANTLIB += xcb xcb-dri2 xcb-dri3 xcb-glx xcb-present xcb-sync
+WANTLIB += xcb-xfixes xshmfence z
COMPILER = base-clang ports-gcc base-gcc
@@ -34,7 +35,6 @@ LIB_DEPENDS= converters/libiconv \
graphics/libmng \
graphics/lcms \
graphics/openexr \
- graphics/openjpeg \
graphics/tiff \
graphics/png
@@ -56,6 +56,12 @@ WRKDIST= ${WRKDIR}/${DISTNAME:L}
# Too many Linuxisms in testing tools
NO_TEST= Yes
+
+# gcc-8's _mm_malloc() uses the AltiVec keyword __vector, so we need
+# to disable it to make it run on macppc G3 machines
+.if ${MACHINE_ARCH:Mpowerpc}
+CONFIGURE_ENV += ac_cv_header_mm_malloc_h=
+.endif
post-extract:
chmod +x ${WRKSRC}/configure
[ports-gcc] Unbreak games/love
> http://build-failures.rhaalovely.net/sparc64/2019-10-11/games/love.log
(no macppc log because graphics/DevIL is broken, i've just posted a diff
for that)
This is a very typical case; gcc does not consider `NULL' as `false'.
I needed to dos2unix the source code as the "faulty" code had MS-DOS
line endings.
Once fixed it builds fine on amd64, and on macppc [0] where the runtime
has been tested successfully [1].
While here i've moved HOMEPAGE to https.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/love.log
[1] https://bsd.network/@julianaito/102980335050703097
Index: Makefile
===
RCS file: /cvs/ports/games/love/Makefile,v
retrieving revision 1.24
diff -u -p -u -p -r1.24 Makefile
--- Makefile12 Jul 2019 20:46:20 - 1.24
+++ Makefile17 Oct 2019 22:26:34 -
@@ -5,9 +5,9 @@ COMMENT=2D games framework for use wit
V= 0.8.0
DISTNAME= love-$V-linux-src
PKGNAME= love-$V
-REVISION= 9
+REVISION= 10
-HOMEPAGE= http://love2d.org/
+HOMEPAGE= https://love2d.org/
CATEGORIES=games lang/lua
@@ -55,5 +55,8 @@ post-extract:
for i in `find . -type f`; \
do [ "$(tail -c1 $i)" == "$(printf '\n')" ] || ls -l
$i; \
done
+# And some files have MS-DOS line endings we don't want for patches
+ cd ${WRKSRC} && perl -i -pe 's/\r$$//' \
+ ./src/libraries/Box2D/Collision/Shapes/b2ChainShape.h
.include
Index: patches/patch-src_libraries_Box2D_Collision_Shapes_b2ChainShape_h
===
RCS file: patches/patch-src_libraries_Box2D_Collision_Shapes_b2ChainShape_h
diff -N patches/patch-src_libraries_Box2D_Collision_Shapes_b2ChainShape_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_libraries_Box2D_Collision_Shapes_b2ChainShape_h 17 Oct
2019 22:26:34 -
@@ -0,0 +1,19 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: converting to 'bool' from 'std::nullptr_t' requires
direct-initialization
+
+Index: src/libraries/Box2D/Collision/Shapes/b2ChainShape.h
+--- src/libraries/Box2D/Collision/Shapes/b2ChainShape.h.orig
src/libraries/Box2D/Collision/Shapes/b2ChainShape.h
+@@ -95,8 +95,8 @@ inline b2ChainShape::b2ChainShape()
+ m_radius = b2_polygonRadius;
+ m_vertices = NULL;
+ m_count = 0;
+- m_hasPrevVertex = NULL;
+- m_hasNextVertex = NULL;
++ m_hasPrevVertex = false;
++ m_hasNextVertex = false;
+ }
+
+ #endif
[macppc] Unbreak graphics/DevIL
(Pascal, only this port seems to have this issue, it's FYI)
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/graphics/DevIL.log
__vector is an AltiVec keyword. AltiVec is not found by DevIL's
configure script, which is actually what we want (G3 machines don't
support AltiVec).
The "problem" comes from gcc-8's _mm_malloc(), that uses that keyword.
It seems to me that the old gcc-4.9 didn't have _mm_malloc(), so that's
why it's broken since gcc-8 is ports-gcc.
As such, i've disabled the use of _mm_malloc. While here i've updated
WANTLIB.
It builds fine on amd64, and macppc [0], where i have been able to
fix (diff incoming), build and run games/love, its only consumer on
macppc.
Comments/feedback are welcome!
Charlène.
[0] https://bin.charlenew.xyz/DevIL.log
Index: Makefile
===
RCS file: /cvs/ports/graphics/DevIL/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:46:56 - 1.26
+++ Makefile17 Oct 2019 21:53:10 -
@@ -3,7 +3,7 @@
COMMENT= library for powerful image loading capabilities
DISTNAME= DevIL-1.7.8
-REVISION= 14
+REVISION= 15
SHARED_LIBS+= IL 1.0 # 2.0
SHARED_LIBS+= ILU 1.0 # 2.0
SHARED_LIBS+= ILUT1.0 # 2.0
@@ -18,11 +18,12 @@ PERMIT_PACKAGE= Yes
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=openil/}
EXTRACT_SUFX= .zip
-WANTLIB += GL GLU Half ICE Iex IexMath IlmImf IlmThread Imath
-WANTLIB += SDL SM X11 X11-xcb Xdamage Xext Xfixes Xi
-WANTLIB += Xmu Xrandr Xrender Xt Xxf86vm c drm glapi glut iconv jasper
-WANTLIB += jpeg lcms m mng png pthread sndio ${COMPILER_LIBCXX}
-WANTLIB += tiff usbhid xcb xcb-dri2 xcb-glx z expat
+WANTLIB += ${COMPILER_LIBCXX} GL GLU Half ICE Iex IexMath IlmImf
+WANTLIB += IlmThread Imath SDL SM X11 X11-xcb Xdamage Xext Xfixes
+WANTLIB += Xi Xmu Xrandr Xrender Xt Xxf86vm c drm expat glapi
+WANTLIB += glut iconv jasper jpeg lcms m mng png sndio tiff usbhid
+WANTLIB += xcb xcb-dri2 xcb-dri3 xcb-glx xcb-present xcb-sync
+WANTLIB += xcb-xfixes xshmfence z
COMPILER = base-clang ports-gcc base-gcc
@@ -56,6 +57,12 @@ WRKDIST= ${WRKDIR}/${DISTNAME:L}
# Too many Linuxisms in testing tools
NO_TEST= Yes
+
+# gcc-8's _mm_malloc() uses the AltiVec keyword __vector, so we need
+# to disable it to make it run on macppc G3 machines
+.if ${MACHINE_ARCH:Mpowerpc}
+CONFIGURE_ENV += ac_cv_header_mm_malloc_h=
+.endif
post-extract:
chmod +x ${WRKSRC}/configure
[macppc, probably ports-gcc] Unbreak devel/double-conversion (was: Re: new: devel/double-conversion)
Hi,
On Fri, 27 Sep 2019 13:52:49 +0200
Rafael Sadowski wrote:
[...]
>
> Please find attached a re-write of your proposal.
>
> - move to math, devel to full and I see math is a better category.
> - Drop do-test, cmake will handle it for use.
> - Add WANTLIB
> - Drop Compiler, builds fine with all compilers. No C++11
Yes, but there are platform-specific quirks that ports-gcc handles
gracefully, unlike the old base one ;)
That's why portcheck(1) yells if COMPILER_LIBCXX is in WANTLIB and
no COMPILER line with at least ports-gcc is set [0].
The build fails on macppc:
>> https://bin.charlenew.xyz/double-conversion.fail.log
Once COMPILER is brought back, it works [1], and 'make test' as well.
OK?
Charlène.
[0] https://marc.info/?l=openbsd-ports=156561626006156=2
[1] https://bin.charlenew.xyz/double-conversion.working.log
Index: Makefile
===
RCS file: /cvs/ports/math/double-conversion/Makefile,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 Makefile
--- Makefile17 Oct 2019 19:53:58 - 1.1.1.1
+++ Makefile17 Oct 2019 23:01:37 -
@@ -15,6 +15,8 @@ PERMIT_PACKAGE= Yes
WANTLIB += ${COMPILER_LIBCXX} m
+COMPILER = base-clang ports-gcc
+
MODULES = devel/cmake
CONFIGURE_ARGS += -DBUILD_SHARED_LIBS=ON \
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/17 08:03:56 Modified files: lang/parrot: Makefile lang/parrot/pkg: DESCR Log message: parrot: _CALL_ELF is undefined on powerpc, force it as ELFv1. While here, refresh WANTLIB and fix a trailing whitespace. OK landry@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/17 07:32:31 Modified files: emulators/nestopia: Makefile Log message: nestopia,-libretro: fix packaging on base-gcc archs. Implicit MODGCC4_CPPLIBDEP in LIB_DEPENDS is not inherited by LIB_DEPENDS-*, despite COMPILER_LIBCXX being in WANTLIB-*. OK bentley@ (maintainer)
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/17 07:29:53 Modified files: emulators/gambatte: Makefile Log message: gambatte: fix packaging on base-gcc archs. Implicit MODGCC4_CPPLIBDEP in LIB_DEPENDS is not inherited by LIB_DEPENDS-*, despite COMPILER_LIBCXX being in WANTLIB-*. OK bentley@ (maintainer)
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/17 03:13:16 Modified files: devel/p5-Test-Spelling: Makefile distinfo Log message: p5-Test-Spelling: update to 0.25 Changelog: https://metacpan.org/changes/release/CAPOEIRAB/Test-Spelling-0.25 >From Wen Heping, thanks!
Re: [Update] devel/p5-System-Command : Update to 1.119
On Thu, 26 Sep 2019 11:51:20 +
wen heping wrote:
> Hi, ports@:
>
> Here is a patch for devel/p5-System-Command:
> i) Update to 1.119
> ii) Add AUTHOR_TESTING=yes to MKAE_ENV, then it could
> run pod tests.
> iii) Remove CONFIGURE_STYLE = modbuild
>
> It build well and passed all tests on amd64-head system. It
> was required by update of devel/p5-Git-Repository.
> One port depends on it :p5-Git-Repository.
>
> Comments? OK?
It requires devel/p5-IPC-Run as a RUN_DEPENDS. devel/p5-Test-Command
is also needed as a TEST_DEPENDS so all tests run, see the META.json
file [0].
Pod testing requires Pod::Coverage::TrustPod that we don't have in
ports, so we're better off disabling it.
Fixed diff below :)
Charlène.
[0] https://metacpan.org/source/BOOK/System-Command-1.119/META.json
Index: Makefile
===
RCS file: /cvs/ports/devel/p5-System-Command/Makefile,v
retrieving revision 1.6
diff -u -p -u -p -r1.6 Makefile
--- Makefile12 Jul 2019 20:45:07 - 1.6
+++ Makefile17 Oct 2019 08:36:22 -
@@ -4,18 +4,18 @@ COMMENT = object for running system com
MODULES = cpan
PKG_ARCH = *
-DISTNAME = System-Command-1.07
+DISTNAME = System-Command-1.119
CATEGORIES = devel
-REVISION = 0
-
# perl
PERMIT_PACKAGE = Yes
-CONFIGURE_STYLE = modbuild
+RUN_DEPENDS = devel/p5-IPC-Run
-MAKE_ENV +=TEST_POD=yes
+TEST_DEPENDS = devel/p5-Test-Command \
+ devel/p5-Test-Output
-TEST_DEPENDS = devel/p5-Test-Output
+# Requires Pod::Coverage::TrustPod that we don't have in ports
+MAKE_ENV +=TEST_POD= AUTHOR_TESTING=
.include
Index: distinfo
===
RCS file: /cvs/ports/devel/p5-System-Command/distinfo,v
retrieving revision 1.3
diff -u -p -u -p -r1.3 distinfo
--- distinfo1 May 2014 06:59:28 - 1.3
+++ distinfo17 Oct 2019 08:36:22 -
@@ -1,2 +1,2 @@
-SHA256 (System-Command-1.07.tar.gz) =
hDKsJCHQMOeU5c0IE3V54/VQ0sQw2Clu7s1v0E6uxWA=
-SIZE (System-Command-1.07.tar.gz) = 13261
+SHA256 (System-Command-1.119.tar.gz) =
yMn7HlJ8UkY8qxR2UA7+pwOWoLYr6mJdLW+uqZTcRuc=
+SIZE (System-Command-1.119.tar.gz) = 30479
Index: pkg/PLIST
===
RCS file: /cvs/ports/devel/p5-System-Command/pkg/PLIST,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 PLIST
--- pkg/PLIST 20 Jul 2011 18:04:48 - 1.1.1.1
+++ pkg/PLIST 17 Oct 2019 08:36:22 -
@@ -1,4 +1,7 @@
@comment $OpenBSD: PLIST,v 1.1.1.1 2011/07/20 18:04:48 jasper Exp $
${P5SITE}/System/
+${P5SITE}/System/Command/
${P5SITE}/System/Command.pm
+${P5SITE}/System/Command/Reaper.pm
@man man/man3p/System::Command.3p
+@man man/man3p/System::Command::Reaper.3p
Re: [NEW] devel/p5-USB-LibUSB
Hi! On Wed, 25 Sep 2019 07:52:45 + wen heping wrote: > Hi, ports@: > >Here is a patch to create new port devel/p5-USB-LibUSB. >It could replace devel p5-Device-USB in current portstree. >It build well and passed all tests on amd64-head system. > > Comments? OK? usb-1.0 is in WANTLIB, as such devel/libusb1 should be in LIB_DEPENDS. I'm attaching a tarball with that changed. It works fine on macppc and amd64. OK cwen@ if someone is willing to import it. p5-USB-LibUSB.tgz Description: Binary data
Re: [macppc] Unbreak lang/parrot
Ping :)
On Sun, 6 Oct 2019 12:02:51 +0200
Charlene Wendling wrote:
> Ping.
>
> On Tue, 17 Sep 2019 17:53:27 +0200
> Charlene Wendling wrote:
>
> > Ping.
> >
> > On Tue, 27 Aug 2019 22:50:18 +0200
> > Charlene Wendling wrote:
> >
> > >
> > > > http://build-failures.rhaalovely.net/powerpc/2019-08-09/lang/parrot.log
> > >
> > > ${LOCALBASE}/include/ffitarget.h is actually fine. The build is
> > > failing due to the (aggressive) usage of -Werror.
> > >
> > > I could have removed the annoying -Werror, but i preferred to play
> > > fair and defining it to (ELFv)1. While here, i've regen WANTLIB.
> > >
> > > It indeed builds fine [0], there is only one test out of 47767
> > > failing, as seen on amd64.
> > >
> > > Comments/feedback are welcome,
> > >
> > > Charlène.
> > >
> > >
> > > [0] https://bin.charlenew.xyz/parrot.log
Index: Makefile
===
RCS file: /cvs/ports/lang/parrot/Makefile,v
retrieving revision 1.47
diff -u -p -u -p -r1.47 Makefile
--- Makefile12 Jul 2019 20:47:21 - 1.47
+++ Makefile16 Oct 2019 19:37:15 -
@@ -4,7 +4,7 @@ COMMENT=virtual machine designed for in
V= 7.10.0
DISTNAME= parrot-$V
-REVISION= 7
+REVISION= 8
CATEGORIES=lang perl6
SHARED_LIBS= parrot 11.0
@@ -16,10 +16,10 @@ MASTER_SITES= http://ftp.parrot.org/rele
# Artistic2
PERMIT_PACKAGE=Yes
-WANTLIB += GL GLU c glut gmp iconv icudata icuuc intl m ncurses pthread
-WANTLIB += readline ${COMPILER_LIBCXX} util z ffi
+WANTLIB += ${COMPILER_LIBCXX} GL GLU c ffi glut gmp icudata icuuc
+WANTLIB += intl m readline z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER = base-clang ports-gcc
BUILD_DEPENDS= graphics/glew \
devel/pcre
@@ -55,6 +55,11 @@ INSTALL_TARGET = install install-doc
TEST_TARGET= fulltest
SUBST_VARS += V
+
+# fix for error: "_CALL_ELF" is not defined, evaluates to 0
+.if ${MACHINE_ARCH:Mpowerpc}
+CFLAGS += -D_CALL_ELF=1
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/config/init/hints/openbsd.pm
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/16 12:59:54 Modified files: geo/gpsbabel : Makefile Log message: gpsbabel: fix packaging on base-gcc archs. Implicit MODGCC4_CPPLIBDEP in LIB_DEPENDS is not inherited by LIB_DEPENDS-*, despite COMPILER_LIBCXX being in WANTLIB-*. With hints from kirby@, thanks! OK sthen@
[update] sysutils/screenfetch 3.9.0
Hi,
Upstream finally made a release! The most notable change for us is that
disk usage has been added.
Port-wise, i've removed the upstreamed fixes. Disk usage was
broken, also i've brought back architecture mention, so screenfetch
still needs a patch. All of this is upstreamed [0] already.
It has been tested successfully on my amd64 and macppc machines.
Comments/feedback are welcome,
Charlène.
[0] https://github.com/KittyKatt/screenFetch/pull/648
Index: Makefile
===
RCS file: /cvs/ports/sysutils/screenfetch/Makefile,v
retrieving revision 1.4
diff -u -p -u -p -r1.4 Makefile
--- Makefile12 Jul 2019 20:49:51 - 1.4
+++ Makefile13 Oct 2019 17:29:11 -
@@ -2,15 +2,14 @@
COMMENT = display system information in the terminal for screenshots
PKGNAME = ${DISTNAME:L}
-REVISION = 1
GH_ACCOUNT = KittyKatt
GH_PROJECT = screenFetch
-GH_TAGNAME = v3.8.0
+GH_TAGNAME = v3.9.0
CATEGORIES = sysutils misc
-MAINTAINER = Charlene Wendling
+MAINTAINER = Charlene Wendling
# GPLv3+
PERMIT_PACKAGE = Yes
Index: distinfo
===
RCS file: /cvs/ports/sysutils/screenfetch/distinfo,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 distinfo
--- distinfo24 Mar 2017 19:12:50 - 1.1.1.1
+++ distinfo13 Oct 2019 17:29:11 -
@@ -1,2 +1,2 @@
-SHA256 (screenFetch-3.8.0.tar.gz) =
JIKD7jwksNv/t57Whb3VGFVAcwkMHBZ9B60qcp2yZjM=
-SIZE (screenFetch-3.8.0.tar.gz) = 71970
+SHA256 (screenFetch-3.9.0.tar.gz) =
1t9O93Y/l2HYGMh4Rl1473AbcQAqUNTxUPZaMcwb6jc=
+SIZE (screenFetch-3.9.0.tar.gz) = 77670
Index: patches/patch-screenfetch-dev
===
RCS file: /cvs/ports/sysutils/screenfetch/patches/patch-screenfetch-dev,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 patch-screenfetch-dev
--- patches/patch-screenfetch-dev 28 Aug 2018 20:05:18 - 1.2
+++ patches/patch-screenfetch-dev 13 Oct 2019 17:29:11 -
@@ -1,37 +1,31 @@
$OpenBSD: patch-screenfetch-dev,v 1.2 2018/08/28 20:05:18 bcallah Exp $
-Fix "awk: cannot open /proc/fb"
-From upstream dc72b5932e86ba9c4e36110408690abeb2004070
-Fix "unary operator expected" when no GPU is detected
-From upstream 8346a75068323692243805fa702d02ec7538f3b9
-Fix RAM output
-From upstream 77613487a23fd5a12b70945cf8075c2773553643
+
+Improve kernel version and architecture reporting and fix disk total usage
+From https://github.com/KittyKatt/screenFetch/pull/648
+
Index: screenfetch-dev
--- screenfetch-dev.orig
+++ screenfetch-dev
-@@ -1332,13 +1332,15 @@ detectgpu () {
- # Detect Intel GPU #works in dash
- # Run it only on Intel Processors if GPU is unknown
- DetectIntelGPU() {
-- gpu=$(awk '{print $2}' /proc/fb)
-+ if [ -r /proc/fb ]; then
-+ gpu=$(awk '{print $2}' /proc/fb)
-+ fi
-
- case $gpu in
- *intel*) gpu=intel ;;
- esac
-
--if [ $gpu = intel ]; then
-+ if [ "$gpu" = "intel" ]; then
- #Detect CPU
- local CPU=$(uname -p | awk '{print $3}')
- CPU=${CPU#*'-'}; #Detect CPU number
-@@ -1421,7 +1423,7 @@ detectmem () {
- usedmem=$(($used_mem / 1024^2 ))
- elif [ "$distro" == "OpenBSD" ]; then
- totalmem=$(($(sysctl -n hw.physmem) / 1024 / 1024))
-- usedmem=$(($(vmstat | awk '!/[a-z]/{print $4}') / 1024))
-+ usedmem=$(vmstat | awk '!/[a-z]/{gsub("M",""); print $3}')
- elif [ "$distro" == "NetBSD" ]; then
- phys_mem=$(awk '/MemTotal/ { print $2 }' /proc/meminfo)
- totalmem=$((${phys_mem} / 1024))
+@@ -1271,7 +1271,7 @@ detecthost () {
+ # Kernel Version Detection - Begin
+ detectkernel () {
+ if [[ "$distro" == "OpenBSD" ]]; then
+- kernel=$(sysctl kern.version|awk -F'[ =:]' 'NR==1{print $3"
"$4" "$5}')
++ kernel=$(uname -a | cut -f 3- -d ' ')
+ else
+ # compatibility for older versions of OS X:
+ kernel=$(uname -m && uname -sr)
+@@ -1702,8 +1702,12 @@ DetectIntelGPU() {
+ detectdisk () {
+ diskusage="Unknown"
+ if type -p df >/dev/null 2>&1; then
+- if [[ "${distro}" =~ (Free|Net|Open|DragonFly)BSD ]]; then
++ if [[ "${distro}" =~ (Free|Net|DragonFly)BSD ]]; then
+ totaldisk=$(df -h -c 2>/dev/null | tail -1)
++ elif [[ "${distro}" == "OpenBSD" ]]; then
++ totaldisk=$(df -Pk 2> /dev/null | awk '
++ /^\// {total+=$2; used+=$3; avail+=$4}
++ END{printf("total %.1fG %.1fG %.1fG %d%%\n",
total/1048576, used/1048576, avail/1048576, used*100/total)}')
+ elif [[ "${distro}" == "Mac OS X" ]]; then
+ totaldisk=$(df -H / 2>/dev/null | tail -1)
+ else
[macppc] Unbreak games/pentobi
Hi!
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/games/pentobi.log
I've already been there in the past with libsass [0]. In that case the
COMPILER line has just no reason to be used anymore.
Pentobi needed clang in the past as gcc-4.9 did not support C++14, but
it's not the case anymore, so i've just let it honour x11/qt5 MODULE's
COMPILER line, that is "base-clang ports-gcc".
It builds [1] and works fine then.
I've bumped REVISION, as it was built with clang when Qt5 itself was
built with ports-clang.
Comments/feedback are welcome,
Charlène.
[0] https://marc.info/?l=openbsd-ports=155767262330313=2
[1] https://bin.charlenew.xyz/pentobi-14.1p4.log
Index: Makefile
===
RCS file: /cvs/ports/games/pentobi/Makefile,v
retrieving revision 1.13
diff -u -p -u -p -r1.13 Makefile
--- Makefile12 Jul 2019 20:46:22 - 1.13
+++ Makefile12 Oct 2019 23:43:47 -
@@ -4,7 +4,7 @@ COMMENT = free clone of Blokus board gam
DISTNAME = pentobi-14.1
EXTRACT_SUFX = .tar.xz
CATEGORIES = games x11
-REVISION = 3
+REVISION = 4
HOMEPAGE = http://pentobi.sourceforge.net/
MAINTAINER = Brian Callahan
@@ -16,9 +16,6 @@ WANTLIB += Qt5Concurrent Qt5Core Qt5Gui
WANTLIB += ${COMPILER_LIBCXX}
MASTER_SITES = ${MASTER_SITE_SOURCEFORGE:=pentobi/}
-
-# C++14
-COMPILER = base-clang ports-clang ports-gcc
MODULES = devel/cmake \
x11/qt5
[ld.bfd] Unbreak devel/py-unicorn
Hi,
> http://build-failures.rhaalovely.net/sparc64/2019-10-06/devel/py-unicorn%2Cpython3.log
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/devel/py-unicorn.log
It seems it is happening since the python3 flavor has been enabled.
'-Bsymbolic-functions' is not supported by our ld.bfd, so i've simply
split linking options according to which linker is used. I could have
totally removed that option, but i heard it allows better performances.
This builds fine on macppc [0] and amd64.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/py-unicorn.tgz
Index: Makefile
===
RCS file: /cvs/ports/devel/py-unicorn/Makefile,v
retrieving revision 1.6
diff -u -p -u -p -r1.6 Makefile
--- Makefile12 Jul 2019 20:45:55 - 1.6
+++ Makefile12 Oct 2019 23:17:27 -
@@ -5,7 +5,7 @@ COMMENT = Unicorn CPU emulator engine
MODPY_EGG_VERSION =1.0.1
DISTNAME = unicorn-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME}
-REVISION = 1
+REVISION = 2
CATEGORIES = devel
@@ -32,6 +32,14 @@ BUILD_DEPENDS = lang/python/${MODPY_DEF
USE_GMAKE =Yes
MAKE_ENV +=V=1
+
+# our ld.bfd has no '-Bsymbolic-functions' option support
+.include
+.if ${PROPERTIES:Mlld}
+MAKE_ENV +=LD_IS_LLD=1
+.else
+MAKE_ENV +=LD_IS_LLD=0
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/src/Makefile
Index: patches/patch-src_Makefile
===
RCS file: /cvs/ports/devel/py-unicorn/patches/patch-src_Makefile,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 patch-src_Makefile
--- patches/patch-src_Makefile 1 Jun 2019 20:00:01 - 1.2
+++ patches/patch-src_Makefile 12 Oct 2019 23:17:27 -
@@ -1,11 +1,25 @@
$OpenBSD: patch-src_Makefile,v 1.2 2019/06/01 20:00:01 kn Exp $
-ERROR: Cannot use '/usr/local/bin/python3.6', Python 2.4 or later is required.
+Hunk#1: -Bsymbolic-functions is not supported by ld.bfd
+Hunk#2: Fix error: Cannot use '/usr/local/bin/python3.6', Python 2.4 or
+later is required.
Index: src/Makefile
--- src/Makefile.orig
+++ src/Makefile
-@@ -212,7 +212,7 @@ all: unicorn
+@@ -141,7 +141,11 @@ else
+ EXT = so
+ VERSION_EXT = $(EXT).$(API_MAJOR)
+ AR_EXT = a
++ifeq ($(LD_IS_LLD),1)
+ $(LIBNAME)_LDFLAGS +=
-Wl,-Bsymbolic-functions,-soname,lib$(LIBNAME).$(VERSION_EXT)
++else
++$(LIBNAME)_LDFLAGS += -Wl,-soname,lib$(LIBNAME).$(VERSION_EXT)
++endif
+ UNICORN_CFLAGS += -fvisibility=hidden
+ endif
+
+@@ -212,7 +216,7 @@ all: unicorn
qemu/config-host.h-timestamp:
cd qemu && \
[ports-gcc] Unbreak emulators/gambatte
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/emulators/gambatte%2C-main.log
> http://build-failures.rhaalovely.net/sparc64/2019-10-06/emulators/gambatte%2C-main.log
Once again this is the same issue as gpsbabel.
Once MODGCC4_CPPLIBDEP is provided as a LIB_DEPENDS for gambatte,-main,
this just packages [0] and works [1] fine on macppc, without breaking
amd64.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/gambatte.log
[1] https://bsd.network/@julianaito/102946308754054753
Index: Makefile
===
RCS file: /cvs/ports/emulators/gambatte/Makefile,v
retrieving revision 1.22
diff -u -p -u -p -r1.22 Makefile
--- Makefile12 Jul 2019 20:46:08 - 1.22
+++ Makefile11 Oct 2019 23:38:01 -
@@ -38,7 +38,8 @@ MODQMAKE_PROJECTS = gambatte_qt
SEPARATE_BUILD = No
CONFIGURE_STYLE = qmake specific # avoid
qmake.port.mk add do-build
-LIB_DEPENDS-main = devel/sdl
+LIB_DEPENDS-main = ${MODGCC4_CPPLIBDEP} \
+ devel/sdl
TEST_DEPENDS = graphics/png
[ports-gcc] Unbreak emulators/nestopia,-libretro
Hi,
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/emulators/nestopia%2C-libretro.log
> http://build-failures.rhaalovely.net/sparc64/2019-10-06/emulators/nestopia%2C-libretro.log
This is the same issue as gpsbabel [0], and i applied the same fix.
It packages [1] fine and runs [2] fine on macppc and amd64.
Comments/feedback are welcome :)
Charlène.
[0] https://marc.info/?l=openbsd-ports=157066541411347=2
[1] https://bin.charlenew.xyz/nestopia.log
[2] https://bsd.network/@julianaito/102940668487274726
Index: Makefile
===
RCS file: /cvs/ports/emulators/nestopia/Makefile,v
retrieving revision 1.19
diff -u -p -u -p -r1.19 Makefile
--- Makefile12 Jul 2019 20:46:09 - 1.19
+++ Makefile10 Oct 2019 23:31:43 -
@@ -38,7 +38,7 @@ LIB_DEPENDS-main =archivers/libarchive
audio/libao \
devel/sdl2 \
x11/gtk+3
-LIB_DEPENDS-libretro =
+LIB_DEPENDS-libretro = ${MODGCC4_CPPLIBDEP}
MAKE_FLAGS = CC="${CC}" \
CXX="${CXX}" \
Re: [ports-gcc] Unbreak geo/gpsbabel
On Thu, 10 Oct 2019 09:17:33 +0300
Kirill Bychkov wrote:
> On Thu, October 10, 2019 02:55, Charlene Wendling wrote:
> > Hi,
> >
> > I was trying to build qt5^W geo/qlandkartegt with naddy's latest
> > fix on macppc, but i hit that:
> >
> >> http://build-failures.rhaalovely.net//sparc64/2019-10-06/geo/gpsbabel.log
> >> http://build-failures.rhaalovely.net//powerpc/2019-09-17/geo/gpsbabel.log
> >
> > I found out that landry met the issue with print/poppler, i quote
> > the commit [0] message:
> >
> >> Add ${MODGCC4_CPPLIBDEP} to LIB_DEPENDS-*, changes nothing for
> >> clang archs, but fixes the dreaded 'Missing library for estdc+
> >> +>=17.0' error message at pkg_create on other archs.
> >> COMPILER_LIBCXX is in WANTLIB-*, but nothing brought it into
> >> context, as ${MODGCC4_CPPLIBDEP} is only in LIB_DEPENDS, which
> >> isnt inherited by LIB_DEPENDS-* here.
> >
> > I bumped REVISION as this version was already built during a
> > previous macppc fix for qlandkartegt iirc.
> >
> > This builds fine on amd64 and macppc [1].
> >
> > Two ports depend on it: geo/qlandkartegt (it's still building) and
> > geo/viking
> >
> > naddy, sthen: i let you decide if it's critical enough to get
> > committed, otherwise i'll ping post-unlock :)
> >
> > Charl?ne.
> >
> >
> > [0]
> > https://github.com/openbsd/ports/commit/10eedd1e2cf2afea1bcbe41e374ce44fb7be4a8a
> > [1] https://bin.charlenew.xyz/gpsbabel.missinglib.log
> >
> >
[...]
>
> Hi,
> I was testing simillar diff. Looks like ${MODGCC4_CPPLIBDEP} is
> needed only for -main and -qt.
>
It makes sense indeed given the WANTLIB! I tested again on macppc and
amd64 - it builds fine.
Charlène.
Index: Makefile
===
RCS file: /cvs/ports/geo/gpsbabel/Makefile,v
retrieving revision 1.35
diff -u -p -u -p -r1.35 Makefile
--- Makefile3 Sep 2019 13:48:25 - 1.35
+++ Makefile10 Oct 2019 09:07:01 -
@@ -12,6 +12,9 @@ DISTNAME= gpsbabel-${VERSION}
PKGNAME-main= gpsbabel-${VERSION}
PKGNAME-tk=gpsbabel-tk-${VERSION}
PKGNAME-qt=gpsbabel-qt-${VERSION}
+REVISION-main= 0
+REVISION-tk= 0
+REVISION-qt= 0
CATEGORIES=geo
HOMEPAGE= https://www.gpsbabel.org/
@@ -34,7 +37,8 @@ MULTI_PACKAGES= -main -tk -qt
MODULES= devel/qmake x11/tk x11/qt5
MODQMAKE_PROJECTS =gui/app.pro
-LIB_DEPENDS-main= devel/libusb-compat \
+LIB_DEPENDS-main= ${MODGCC4_CPPLIBDEP} \
+ devel/libusb-compat \
devel/shapelib
cWANTLIB = c m pthread
@@ -44,7 +48,8 @@ WANTLIB-qt += GL Qt5Core Qt5Gui Qt5Netwo
WANTLIB-qt += Qt5Widgets Qt5Xml ${COMPILER_LIBCXX} ${cWANTLIB}
LIB_DEPENDS-tk=
-LIB_DEPENDS-qt=x11/qt5/qtwebkit
+LIB_DEPENDS-qt=${MODGCC4_CPPLIBDEP} \
+ x11/qt5/qtwebkit
PKG_ARCH-tk= *
RUN_DEPENDS-tk=geo/gpsbabel \
${MODTK_RUN_DEPENDS}
[ports-gcc] Unbreak geo/gpsbabel
Hi,
I was trying to build qt5^W geo/qlandkartegt with naddy's latest fix on
macppc, but i hit that:
> http://build-failures.rhaalovely.net//sparc64/2019-10-06/geo/gpsbabel.log
> http://build-failures.rhaalovely.net//powerpc/2019-09-17/geo/gpsbabel.log
I found out that landry met the issue with print/poppler, i quote the
commit [0] message:
> Add ${MODGCC4_CPPLIBDEP} to LIB_DEPENDS-*, changes nothing for clang
> archs, but fixes the dreaded 'Missing library for estdc++>=17.0' error
> message at pkg_create on other archs. COMPILER_LIBCXX is in WANTLIB-*,
> but nothing brought it into context, as ${MODGCC4_CPPLIBDEP} is only in
> LIB_DEPENDS, which isnt inherited by LIB_DEPENDS-* here.
I bumped REVISION as this version was already built during a previous
macppc fix for qlandkartegt iirc.
This builds fine on amd64 and macppc [1].
Two ports depend on it: geo/qlandkartegt (it's still building) and
geo/viking
naddy, sthen: i let you decide if it's critical enough to get
committed, otherwise i'll ping post-unlock :)
Charlène.
[0]
https://github.com/openbsd/ports/commit/10eedd1e2cf2afea1bcbe41e374ce44fb7be4a8a
[1] https://bin.charlenew.xyz/gpsbabel.missinglib.log
Index: Makefile
===
RCS file: /cvs/ports/geo/gpsbabel/Makefile,v
retrieving revision 1.35
diff -u -p -u -p -r1.35 Makefile
--- Makefile3 Sep 2019 13:48:25 - 1.35
+++ Makefile9 Oct 2019 23:41:05 -
@@ -12,6 +12,9 @@ DISTNAME= gpsbabel-${VERSION}
PKGNAME-main= gpsbabel-${VERSION}
PKGNAME-tk=gpsbabel-tk-${VERSION}
PKGNAME-qt=gpsbabel-qt-${VERSION}
+REVISION-main= 0
+REVISION-tk= 0
+REVISION-qt= 0
CATEGORIES=geo
HOMEPAGE= https://www.gpsbabel.org/
@@ -34,7 +37,8 @@ MULTI_PACKAGES= -main -tk -qt
MODULES= devel/qmake x11/tk x11/qt5
MODQMAKE_PROJECTS =gui/app.pro
-LIB_DEPENDS-main= devel/libusb-compat \
+LIB_DEPENDS-main= ${MODGCC4_CPPLIBDEP} \
+ devel/libusb-compat \
devel/shapelib
cWANTLIB = c m pthread
@@ -43,8 +47,9 @@ WANTLIB-tk =
WANTLIB-qt += GL Qt5Core Qt5Gui Qt5Network Qt5WebKit Qt5WebKitWidgets
WANTLIB-qt += Qt5Widgets Qt5Xml ${COMPILER_LIBCXX} ${cWANTLIB}
-LIB_DEPENDS-tk=
-LIB_DEPENDS-qt=x11/qt5/qtwebkit
+LIB_DEPENDS-tk=${MODGCC4_CPPLIBDEP}
+LIB_DEPENDS-qt=${MODGCC4_CPPLIBDEP} \
+ x11/qt5/qtwebkit
PKG_ARCH-tk= *
RUN_DEPENDS-tk=geo/gpsbabel \
${MODTK_RUN_DEPENDS}
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/10/09 17:12:39 Modified files: net/fastnetmon : Makefile Log message: fastnetmon: use __atomic* primitives instead of __sync* ones This fixes the build on macppc, and probably hppa. OK naddy@ jasper@ (maintainer)
Re: [macppc/hppa] Unbreak net/fastnetmon
On Tue, 8 Oct 2019 20:21:37 - (UTC)
Christian Weisgerber wrote:
[...]
> More generally, I assume if an arch supports (1) 64-bit __sync*
> primitives and (2) __atomic* primitives, then it will also have
> 64-bit __atomic*.
>
> > +# Use __atomic* function instead of __sync* ones, allowing to
> > build on +# the following archs.
> > +CONFIGURE_ARGS += -DUSE_NEW_ATOMIC_BUILTINS=On
> > +.if ${MACHINE_ARCH} == "powerpc" || ${MACHINE_ARCH} == "hppa"
> > +WANTLIB += atomic
> > +.endif
>
> I like this better.
> But I don't want to start backseat-driving atomic fixes now, after
> ignoring them all throughout the release cycle...
Do you mean we should wait post unlock?
> --
> Christian "naddy" Weisgerber
> na...@mips.inka.de
>
Re: [macppc/hppa] Unbreak net/fastnetmon
On Mon, 7 Oct 2019 15:50:14 - (UTC)
Christian Weisgerber wrote:
> On 2019-10-06, Charlene Wendling wrote:
>
> > Upstream introduced since 1.1.4 a cmake flag allowing the use of
> > __atomic* functions instead of unsupported __sync* ones.
> >
> > Once provided, it builds fine on macppc [0]. I have not bumped
> > REVISION, it never had a package built before that.
>
> I admit I haven't followed the subject of __atomic* at all.
> What's the reason to enable the use of these primitives only for
> powerpc/hppa and not on all archs?
GCC has no 64-bits __sync* operators support on macppc and hppa, the
newer __atomic* ones have it.
We sure can enable __atomic* for all archs, it works at least on amd64,
but may break somewhere else - upstream's intention was to provide a
workaround [0].
I'm joining an alternate diff that i tried on amd64.
Charlène.
> --
> Christian "naddy" Weisgerber
> na...@mips.inka.de
>
[0]
https://github.com/pavel-odintsov/fastnetmon/issues/703#issuecomment-362906382
Index: Makefile
===
RCS file: /cvs/ports/net/fastnetmon/Makefile,v
retrieving revision 1.6
diff -u -p -r1.6 Makefile
--- Makefile12 Jul 2019 20:48:26 - 1.6
+++ Makefile7 Oct 2019 23:32:06 -
@@ -5,6 +5,7 @@ COMMENT = DDoS detector with multiple p
GH_ACCOUNT = pavel-odintsov
GH_PROJECT = fastnetmon
GH_TAGNAME = v1.1.4
+REVISION = 0
HOMEPAGE = https://fastnetmon.com/guides/
@@ -35,6 +36,13 @@ CONFIGURE_ARGS +=-DENABLE_LUA_SUPPORT=O
-DLOG4CPP_INCLUDES_FOLDER="${LOCALBASE}/include/log4cpp" \
-DJSONC_INCLUDES_FOLDER="${LOCALBASE}/include/json-c" \
-DHIREDIS_INCLUDES_FOLDER="${LOCALBASE}/include/hiredis"
+
+# Use __atomic* function instead of __sync* ones, allowing to build on
+# the following archs.
+CONFIGURE_ARGS += -DUSE_NEW_ATOMIC_BUILTINS=On
+.if ${MACHINE_ARCH} == "powerpc" || ${MACHINE_ARCH} == "hppa"
+WANTLIB += atomic
+.endif
WRKSRC = ${WRKDIST}/src
[macppc/hppa] Unbreak net/fastnetmon
Hi,
> http://build-failures.rhaalovely.net/powerpc/2019-09-17/net/fastnetmon.log
Upstream introduced since 1.1.4 a cmake flag allowing the use of
__atomic* functions instead of unsupported __sync* ones.
Once provided, it builds fine on macppc [0]. I have not bumped
REVISION, it never had a package built before that.
OK?
Charlène.
[0] https://bin.charlenew.xyz/fastnetmon-1.1.4.log
Index: Makefile
===
RCS file: /cvs/ports/net/fastnetmon/Makefile,v
retrieving revision 1.6
diff -u -p -u -p -r1.6 Makefile
--- Makefile12 Jul 2019 20:48:26 - 1.6
+++ Makefile6 Oct 2019 13:01:44 -
@@ -40,6 +40,13 @@ WRKSRC = ${WRKDIST}/src
NO_TEST = Yes
+# Use __atomic* function instead of __sync* ones that are not fully supported
+# on these archs
+.if ${MACHINE_ARCH} == "powerpc" || ${MACHINE_ARCH} == "hppa"
+CONFIGURE_ARGS += -DUSE_NEW_ATOMIC_BUILTINS=On
+WANTLIB += atomic
+.endif
+
pre-configure:
${SUBST_CMD} ${WRKSRC}/CMakeLists.txt
sed -i -e 's,/var/log,&/fastnetmon,g;' \
Re: [macppc] Unbreak lang/parrot
Ping :)
On Tue, 17 Sep 2019 17:53:27 +0200
Charlene Wendling wrote:
> Ping.
>
> On Tue, 27 Aug 2019 22:50:18 +0200
> Charlene Wendling wrote:
>
> >
> > > http://build-failures.rhaalovely.net/powerpc/2019-08-09/lang/parrot.log
> >
> > ${LOCALBASE}/include/ffitarget.h is actually fine. The build is
> > failing due to the (aggressive) usage of -Werror.
> >
> > I could have removed the annoying -Werror, but i preferred to play
> > fair and defining it to (ELFv)1. While here, i've regen WANTLIB.
> >
> > It indeed builds fine [0], there is only one test out of 47767
> > failing, as seen on amd64.
> >
> > Comments/feedback are welcome,
> >
> > Charlène.
> >
> >
> > [0] https://bin.charlenew.xyz/parrot.log
>
>
> Index: Makefile
> ===
> RCS file: /cvs/ports/lang/parrot/Makefile,v
> retrieving revision 1.47
> diff -u -p -u -p -r1.47 Makefile
> --- Makefile 12 Jul 2019 20:47:21 - 1.47
> +++ Makefile 27 Aug 2019 20:27:09 -
> @@ -4,7 +4,7 @@ COMMENT= virtual machine designed for in
>
> V= 7.10.0
> DISTNAME=parrot-$V
> -REVISION=7
> +REVISION=8
> CATEGORIES= lang perl6
> SHARED_LIBS= parrot 11.0
>
> @@ -16,10 +16,10 @@ MASTER_SITES= http://ftp.parrot.org/rele
> # Artistic2
> PERMIT_PACKAGE= Yes
>
> -WANTLIB += GL GLU c glut gmp iconv icudata icuuc intl m ncurses
> pthread -WANTLIB += readline ${COMPILER_LIBCXX} util z ffi
> +WANTLIB += ${COMPILER_LIBCXX} GL GLU c ffi glut gmp icudata icuuc
> +WANTLIB += intl m readline z
>
> -COMPILER = base-clang ports-gcc base-gcc
> +COMPILER = base-clang ports-gcc
>
> BUILD_DEPENDS= graphics/glew \
> devel/pcre
> @@ -55,6 +55,11 @@ INSTALL_TARGET = install install-doc
> TEST_TARGET= fulltest
>
> SUBST_VARS +=V
> +
> +# fix for error: "_CALL_ELF" is not defined, evaluates to 0
> +.if ${MACHINE_ARCH:Mpowerpc}
> +CFLAGS +=-D_CALL_ELF=1
> +.endif
>
> pre-configure:
> ${SUBST_CMD} ${WRKSRC}/config/init/hints/openbsd.pm
>
Re: games/ace crashes
On Tue, 1 Oct 2019 18:04:49 +0200 Solene Rapenne wrote: > on amd64, few days old snap (I can't upgrade now for some reasons..) The build log complains about: imagelib.c:109:17: warning: implicit declaration of function 'malloc' is invalid in C99 [-Wimplicit-function-declaration] Clang uses C11 by default, that's probably why it works fine on macppc with the good^W old base-gcc, that uses gnu89 by default. So i added the header and... it works :] Feedback is welcome! Charlène. Index: Makefile === RCS file: /cvs/ports/games/ace/Makefile,v retrieving revision 1.4 diff -u -p -u -p -r1.4 Makefile --- Makefile12 Jul 2019 20:46:14 - 1.4 +++ Makefile2 Oct 2019 13:17:54 - @@ -3,7 +3,7 @@ COMMENT = solitaire games DISTNAME = ace-1.4 -REVISION = 0 +REVISION = 1 SHARED_LIBS += cards 0.0 # 1.0 Index: patches/patch-lib_imagelib_c === RCS file: patches/patch-lib_imagelib_c diff -N patches/patch-lib_imagelib_c --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-lib_imagelib_c2 Oct 2019 13:17:54 - @@ -0,0 +1,11 @@ +$OpenBSD$ + +Index: lib/imagelib.c +--- lib/imagelib.c.orig lib/imagelib.c +@@ -1,4 +1,5 @@ + #include ++#include + #include + #include "imagelib.h" + #include "cards.h"
Re: NEW: games/fallingtime
Hi, On Mon, 23 Sep 2019 18:29:03 -0400 Brian Callahan wrote: > Hi ports -- > > Attached is a new port, games/fallingtime. FallingTime is a game > where you are a constantly falling ball. > > --- > pkg/DESCR: > FallingTime is an arcade game inspired by Fall Down. Like the Fall > Down games, you are a ball and you must fall into the holes to avoid > being crushed by the top of the screen! > --- > > Simple port; works well on amd64. > > OK? It builds and runs fine on macppc. Colors aren't off as well. OK cwen@ > ~Brian >
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/09/26 04:36:53 Modified files: x11/waimea : Makefile Added files: x11/waimea/patches: patch-src_Resources_cc patch-src_Resources_hh patch-src_Window_cc Log message: waimea: unbreak with ports-gcc by fixing a namespace clash. Also move HOMEPAGE to https and refresh WANTLIB. Tested on sparc64 by kmos@ and macppc by me. OK kmos@
Re: [ports-gcc] Unbreak multimedia/mkvtoolnix (impacts all archs)
Hi, There have been some out of list discussion about it. COMPILER_VERSION is unused in ports and we have PROPERTIES for the same purpose. jca@ proposed to remove COMPILER_VERSION from MAKE_ENV and test bulk the change. The issue being i can't do a (required) full bulk myself. Meanwhile mkvtoolnix still doesn't build for a trivial reason, and if no one is willing to test bulk the MAKE_ENV change, i think we should at least unset COMPILER_VERSION for that port. Opinions? Charlène. On Wed, 21 Aug 2019 21:09:33 +0200 Charlene Wendling wrote: > Hi, > > > http://build-failures.rhaalovely.net/sparc64/2019-08-18/multimedia/mkvtoolnix%2Cno_x11.log > (no full macppc bulk report available for now) > > What happens behind the scene (--trace didn't help): > > - $WRKSRC/rake.d/config.rb loads the 'build-config' file, that says > that "COMPILER_VERSION = 8.3.0" (fine) > - $WRKSRC/Rakefile does various compiler version checks to determine > proper build flags to use > - Compiler versions are checked in $WRKSRC/rake.d/helpers.rb using > Ruby's check_version() function, that seems to not like version > numbers with letters at the start, as we can see in the log. > The compiler version itself comes from the c() function... > - ...that is defined in $WRKSRC/rake.d/config.rb, where the line 24 is > interesting: > > var = (ENV[idx_s].nil? ? $config[idx.to_sym] : ENV[idx_s]).to_s > > It appears that this 'gcc4' is coming from our default MAKE_ENV: > > $ make show=MAKE_ENV > [...] COMPILER_VERSION="gcc4" [...] > (on amd64, it returns COMPILER_VERSION="clang") > > It did not break on amd64, because with clang, upstream uses the same > compilation flags for all versions, so there is no version comparison > done. > > As such, i've overridden bsd.port.mk's COMPILER_VERSION, letting > upstream's build system doing its job. I decided to go for it for all > archs, as this may prevent possible future breakage on base-clang > archs. > > I did not bump REVISION because this version never built on ports-gcc > archs and it causes no change on base-clang ones. > > This builds, and tests pass with flying colors, on amd64 and macppc > [0]. > > Comments/feedback are welcome :) > > Charlène. > > > [0] https://bin.charlenew.xyz/mkvtoolnix.log.gz > > > Index: Makefile > === > RCS file: /cvs/ports/multimedia/mkvtoolnix/Makefile,v > retrieving revision 1.92 > diff -u -p -u -p -r1.92 Makefile > --- Makefile 12 Jul 2019 20:47:57 - 1.92 > +++ Makefile 21 Aug 2019 19:02:57 - > @@ -29,6 +29,11 @@ MODRUBY_RUNDEP=No > # C++11 / C++14 > COMPILER = base-clang ports-gcc > > +# Override bsd.port.mk's COMPILER_VERSION because 'gcc4' is incorrect > +# with ports-gcc and not recognised as a version number by Ruby's > +# check_version(), and is just 'clang' with base-clang. > +COMPILER_VERSION= > + > BUILD_DEPENDS= devel/gettext,-tools \ > devel/gtest \ > textproc/docbook \ >
Re: [ports-gcc] Unbreak x11/waimea
Ping :)
On Fri, 23 Aug 2019 15:54:18 +0200
Charlene Wendling wrote:
> ping.
>
> > > http://build-failures.rhaalovely.net/powerpc/2019-07-29/x11/waimea.log
> > > http://build-failures.rhaalovely.net/sparc64/2019-08-12/x11/waimea.log
> >
> > There is a clash between std::list and the "list" class member.
> >
> > I've just renamed that member and its instances. It builds [0] and
> > works fine on macppc, and amd64 is not broken.
> >
> > While here WANTLIB needed a refresh, and i moved HOMEPAGE to https.
> >
> > OK?
> >
> > Charlène.
> >
> >
> > [0] https://bin.charlenew.xyz/waimea.log
> >
Index: Makefile
===
RCS file: /cvs/ports/x11/waimea/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:51:23 - 1.26
+++ Makefile15 Aug 2019 09:56:07 -
@@ -4,21 +4,20 @@ COMMENT= virtual desktop manager for X1
VERSION= 0.4.0
DISTNAME= waimea-${VERSION}
-REVISION = 9
+REVISION= 10
CATEGORIES=x11
-HOMEPAGE= http://www.freedesktop.org/wiki/Software/waimea
+HOMEPAGE= https://www.freedesktop.org/wiki/Software/waimea
# GPLv2
PERMIT_PACKAGE=Yes
-MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
+WANTLIB += ${COMPILER_LIBCXX} ICE SM X11 Xext Xft Xinerama Xrandr
+WANTLIB += Xrender c m
-WANTLIB += ICE SM X11 Xau Xdmcp Xext Xft Xinerama Xrandr Xrender
-WANTLIB += c expat fontconfig freetype m ${COMPILER_LIBCXX}
-WANTLIB += xcb z
+MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
AUTOCONF_VERSION= 2.52
CONFIGURE_STYLE= autoconf
Index: patches/patch-src_Resources_cc
===
RCS file: patches/patch-src_Resources_cc
diff -N patches/patch-src_Resources_cc
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_cc 15 Aug 2019 09:56:07 -
@@ -0,0 +1,35 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.cc
+--- src/Resources.cc.orig
src/Resources.cc
+@@ -1568,7 +1568,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 'n' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1583,7 +1583,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 't' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1598,7 +1598,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+str + 2);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (! strncasecmp(str, "window", 6)) {
+ str = str + 6;
Index: patches/patch-src_Resources_hh
===
RCS file: patches/patch-src_Resources_hh
diff -N patches/patch-src_Resources_hh
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_hh 15 Aug 2019 09:56:07 -
@@ -0,0 +1,24 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.hh
+--- src/Resources.hh.orig
src/Resources.hh
+@@ -193,13 +193,13 @@ class WaActionExtList { (public)
+ delete name;
+ delete cl;
+ delete title;
+-ACTLISTCLEAR(list);
++ACTLISTCLEAR(alist);
+ }
+
+ Regex *name;
+ Regex *cl;
+ Regex *title;
+-list list;
++list alist;
+ };
+
+ class StrComp {
Index:
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/09/22 03:10:55 Modified files: x11/grantlee-qt5: Makefile x11/grantlee-qt5/patches: patch-CMakeLists_txt Log message: grantlee-qt5: unbreak with ports-gcc Remove '-ansi' from cxxflags as it implies c++98 with gcc, and grantlee wants c++11. OK rsadowski@
[ports-gcc] Unbreak x11/grantlee-qt5
Hi,
> http://build-failures.rhaalovely.net/powerpc/2019-08-30/x11/grantlee-qt5.log
> http://build-failures.rhaalovely.net/sparc64/2019-09-08/x11/grantlee-qt5.log
There is some funny stuff going on here. On base-gcc archs,
grantlee-qt5 is built with c++98, while it requires at least c++11.
eg++(1) manpage:
-ansi
In C mode, this is equivalent to -std=c90. In C++ mode, it is
equivalent to -std=c++98.
clang++(1) manpage:
-ansi Same as -std=c89.
Removing '-ansi' is solving the issue on macppc [0], while keeping
amd64 fine.
Comments/feedback are welcome!
Charlène.
[0] https://bin.charlenew.xyz/grantlee-qt5-5.1.0p2.log
Index: Makefile
===
RCS file: /cvs/ports/x11/grantlee-qt5/Makefile,v
retrieving revision 1.4
diff -u -p -u -p -r1.4 Makefile
--- Makefile12 Jul 2019 20:51:12 - 1.4
+++ Makefile19 Sep 2019 16:22:26 -
@@ -5,7 +5,7 @@ COMMENT = template and text document li
VERSION = 5.1.0
PKGNAME = grantlee-qt5-${VERSION}
DISTNAME = grantlee-${VERSION}
-REVISION = 1
+REVISION = 2
SHARED_LIBS += Grantlee_Templates0.0 # 5.1
SHARED_LIBS += Grantlee_TextDocument 0.0 # 5.1
Index: patches/patch-CMakeLists_txt
===
RCS file: /cvs/ports/x11/grantlee-qt5/patches/patch-CMakeLists_txt,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 patch-CMakeLists_txt
--- patches/patch-CMakeLists_txt14 Aug 2018 05:23:08 - 1.2
+++ patches/patch-CMakeLists_txt19 Sep 2019 16:22:26 -
@@ -1,4 +1,6 @@
$OpenBSD: patch-CMakeLists_txt,v 1.2 2018/08/14 05:23:08 rsadowski Exp $
+Hunk #2: ports-gcc fix because with it, -ansi means -std=c++98, and
+grantlee wants c++11
Index: CMakeLists.txt
--- CMakeLists.txt.orig
@@ -20,6 +22,15 @@ Index: CMakeLists.txt
set(CMAKE_CXX_FLAGS_COVERAGE "${CMAKE_CXX_FLAGS_RELEASE} -fprofile-arcs
-ftest-coverage")
set(CMAKE_C_FLAGS_COVERAGE "${CMAKE_C_FLAGS_RELEASE} -fprofile-arcs
-ftest-coverage")
set(CMAKE_SHARED_LINKER_FLAGS_COVERAGE
"${CMAKE_SHARED_LINKER_FLAGS_RELEASE} -fprofile-arcs -ftest-coverage")
+@@ -56,7 +56,7 @@ endif()
+
+ if (CMAKE_COMPILER_IS_GNUCXX)
+ set (CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wno-long-long -std=iso9899:1990
-Wundef -Wcast-align -Werror-implicit-function-declaration -Wchar-subscripts
-Wall -W -Wpointer-arith -Wwrite-strings -Wformat-security
-Wmissing-format-attribute -fno-common")
+- set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wnon-virtual-dtor -Wno-long-long
-ansi -Wundef -Wcast-align -Wchar-subscripts -Wall -W -Wpointer-arith
-Wformat-security -fno-check-new -fno-common ")
++ set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wnon-virtual-dtor -Wno-long-long
-Wundef -Wcast-align -Wchar-subscripts -Wall -W -Wpointer-arith
-Wformat-security -fno-check-new -fno-common ")
+
+ if (CMAKE_BUILD_TYPE MATCHES TestCocoon)
+ # Needed on 64 bit
@@ -100,7 +100,7 @@ set (LIB_SUFFIX "" CACHE STRING "Define suffix of libr
set( LIB_INSTALL_DIR lib${LIB_SUFFIX} )
set( PLUGIN_INSTALL_DIR
${LIB_INSTALL_DIR}/grantlee/${Grantlee5_MAJOR_MINOR_VERSION_STRING} )
Re: [macppc] Unbreak lang/parrot
Ping.
On Tue, 27 Aug 2019 22:50:18 +0200
Charlene Wendling wrote:
>
> > http://build-failures.rhaalovely.net/powerpc/2019-08-09/lang/parrot.log
>
> ${LOCALBASE}/include/ffitarget.h is actually fine. The build is
> failing due to the (aggressive) usage of -Werror.
>
> I could have removed the annoying -Werror, but i preferred to play
> fair and defining it to (ELFv)1. While here, i've regen WANTLIB.
>
> It indeed builds fine [0], there is only one test out of 47767
> failing, as seen on amd64.
>
> Comments/feedback are welcome,
>
> Charlène.
>
>
> [0] https://bin.charlenew.xyz/parrot.log
Index: Makefile
===
RCS file: /cvs/ports/lang/parrot/Makefile,v
retrieving revision 1.47
diff -u -p -u -p -r1.47 Makefile
--- Makefile12 Jul 2019 20:47:21 - 1.47
+++ Makefile27 Aug 2019 20:27:09 -
@@ -4,7 +4,7 @@ COMMENT=virtual machine designed for in
V= 7.10.0
DISTNAME= parrot-$V
-REVISION= 7
+REVISION= 8
CATEGORIES=lang perl6
SHARED_LIBS= parrot 11.0
@@ -16,10 +16,10 @@ MASTER_SITES= http://ftp.parrot.org/rele
# Artistic2
PERMIT_PACKAGE=Yes
-WANTLIB += GL GLU c glut gmp iconv icudata icuuc intl m ncurses pthread
-WANTLIB += readline ${COMPILER_LIBCXX} util z ffi
+WANTLIB += ${COMPILER_LIBCXX} GL GLU c ffi glut gmp icudata icuuc
+WANTLIB += intl m readline z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER = base-clang ports-gcc
BUILD_DEPENDS= graphics/glew \
devel/pcre
@@ -55,6 +55,11 @@ INSTALL_TARGET = install install-doc
TEST_TARGET= fulltest
SUBST_VARS += V
+
+# fix for error: "_CALL_ELF" is not defined, evaluates to 0
+.if ${MACHINE_ARCH:Mpowerpc}
+CFLAGS += -D_CALL_ELF=1
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/config/init/hints/openbsd.pm
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/09/16 14:23:48 Modified files: sysutils/neofetch: Makefile distinfo Removed files: sysutils/neofetch/patches: patch-neofetch Log message: neofetch: update to 6.1.0 Changelog: https://github.com/dylanaraps/neofetch/releases/tag/6.1.0 OK bcallah@
[update] sysutils/neofetch 6.1.0
Hi,
Here is an update to 6.1.0 [0] for neofetch. There is no
OpenBSD-specific changes for this release, but Puffy's tail is fixed
upstream :)
Port-wise it's a simple version bump with only the tail patch removed.
It works fine on my 2 amd64 machines and my PowerBook.
OK?
Charlène.
[0] https://github.com/dylanaraps/neofetch/releases/tag/6.1.0
Index: Makefile
===
RCS file: /cvs/ports/sysutils/neofetch/Makefile,v
retrieving revision 1.6
diff -u -p -u -p -r1.6 Makefile
--- Makefile12 Jul 2019 20:49:47 - 1.6
+++ Makefile6 Sep 2019 17:25:47 -
@@ -2,11 +2,10 @@
COMMENT = system information tool written in bash
CATEGORIES = sysutils misc
-REVISION = 0
GH_ACCOUNT = dylanaraps
GH_PROJECT = neofetch
-GH_TAGNAME = 6.0.0
+GH_TAGNAME = 6.1.0
MAINTAINER = Charlene Wendling
Index: distinfo
===
RCS file: /cvs/ports/sysutils/neofetch/distinfo,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 distinfo
--- distinfo9 Jan 2019 16:46:08 - 1.2
+++ distinfo6 Sep 2019 17:25:47 -
@@ -1,2 +1,2 @@
-SHA256 (neofetch-6.0.0.tar.gz) = Jkp2iVYbtJj5fxAjGVm92PfIc2cbrC/7Zg3ppYY7HHY=
-SIZE (neofetch-6.0.0.tar.gz) = 115538
+SHA256 (neofetch-6.1.0.tar.gz) = 7ONR41KGtk02IADUCbJ1l/y9z3fo5g+gra4fKdPCljc=
+SIZE (neofetch-6.1.0.tar.gz) = 118346
Index: patches/patch-neofetch
===
RCS file: patches/patch-neofetch
diff -N patches/patch-neofetch
--- patches/patch-neofetch 1 Feb 2019 21:23:39 - 1.4
+++ /dev/null 1 Jan 1970 00:00:00 -
@@ -1,15 +0,0 @@
-$OpenBSD: patch-neofetch,v 1.4 2019/02/01 21:23:39 cwen Exp $
-Fix Puffy's ascii logo
-From https://github.com/dylanaraps/neofetch/pull/1182
-Index: neofetch
neofetch.orig
-+++ neofetch
-@@ -7201,7 +7201,7 @@ ${c1} `-|.' /_. ${c4}\_| ${c1} F
-|/`. `-. `._)
- / .-.\\
- \\ ( `\\
-- `.\
-+ `.\\
- EOF
- ;;
-
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/30 13:54:47 Modified files: databases/p5-Mojo-Pg: Makefile distinfo Log message: p5-Mojo-Pg: update to 4.15 Changelog: https://metacpan.org/source/SRI/Mojo-Pg-4.15/Changes OK maintainer
Re: [Update] databases/p5-Mojo-Pg : Update to 4.15
Hi, On Fri, 30 Aug 2019 13:32:52 -0500 Abel Abraham Camarillo Ojeda wrote: > On Friday, August 30, 2019, wen heping > wrote: > > > Hi, > > > >Here is a patch for databases/p5-Mojo-Pg: > > i) Update to 4.15 > > ii) Fix LICENSE to artistic 2.0 > > > > Hi Wen, > > I think that perl implied artistic 2.0, is that on longer true? (Or > has it never been true? ) It's true, but artistic 2.0 per se does not imply Perl as a license marker (may it be for CPAN or for us). Perl is distributed under either the GPL or Artistic 2.0 [0], and only modules mentioning the dual licensing, or the famous "under the same terms as Perl itself" should have the Perl license marker. Mojo-Pg is released only under the term of the Artistic 2.0 license [1], so the fix is real. OK cwen@ [0] https://dev.perl.org/licenses/ [1] https://metacpan.org/pod/Mojo::Pg#COPYRIGHT-AND-LICENSE
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/30 07:02:19 Modified files: www/webkitgtk4 : Makefile Added files: www/webkitgtk4/patches: patch-Source_WebCore_Modules_indexeddb_server_SQLiteIDBBackingStore_cpp Log message: webkitgtk4: fix linking error when built with ports-clang OK jca@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/29 15:30:21 Modified files: graphics/makehuman: Makefile Added files: graphics/makehuman/patches: patch-src_FileTools_cpp patch-src_util_h Log message: makehuman: add missing headers to fix the build with ports-gcc OK sthen@
Re: [Update] www/p5-Dancer-Plugin-Database: Update to 2.17
On Thu, 29 Aug 2019 02:15:41 + wen heping wrote: > Hi, ports@: > >Here is a patch for www/p5-Dancer-Plugin-Database: > i) Update to 2.17 > ii) Add TEST_POD=yes to MAKE_ENV > iii) Update RUN_DEPENDS > > It build well and passed all tests on amd64-head system. > > No other port depends on it. > > Comments? OK? > wen > OK cwen@
Re: [Update] www/p5-Dancer-Plugin-Database-Core : Update to 0.20
On Thu, 29 Aug 2019 02:13:57 + wen heping wrote: > Hi, ports@: > >Here is a patch for www/p5-Dancer-Plugin-Database-Core: > i) Update to 0.20 > ii) Add TEST_POD=yes to MAKE_ENV > > It build well and passed all tests on amd64-head system. > > Only one port depends on it :www/p5-Dancer-Plugin-Database, > it build well and passed all tests with this patch too. > > Comments? OK? The license is Artistic 2.0 - not Perl. OK cwen@ with that changed. > wen
Re: [Update] www/p5-Dancer2-Plugin-Auth-Tiny : Update to 0.008
Hi, On Wed, 28 Aug 2019 14:11:57 + wen heping wrote: > Hi, ports@: > > Here is a patch for www/p5-Dancer2-Plugin-Auth-Tiny to update to > 0.008. It build well and passed all tests on amd64-head system > No other ports depends on it. > > Comments?OK? - License is "Apache 2.0" - There is a trailing whitespace line 18 in the Makefile (please use portcheck(1)) > wen With that fixed, OK cwen@
[ports-clang] Unbreak www/webkitgtk4 2.24.2
Hi,
As i feared it, the newer webkitgtk4 cannot be built on my macppc box, due
to some undefined references errors.
WebKit proper has a fix for this [0], allowing webkitgtk4 to build.
amd64 is still fine with that.
As far as runtime is concerned, the situation is still the same.
Excepted file://, all other protocols cause an internal webkit error.
Comments/feedback are welcome :)
Charlène.
[0] https://trac.webkit.org/changeset/246922/webkit
Index: Makefile
===
RCS file: /cvs/ports/www/webkitgtk4/Makefile,v
retrieving revision 1.107
diff -u -p -u -p -r1.107 Makefile
--- Makefile28 Aug 2019 12:24:21 - 1.107
+++ Makefile29 Aug 2019 14:14:46 -
@@ -13,6 +13,7 @@ PORTROACH = limitw:1,even
COMMENT = GTK+ port of the WebKit rendering engine
V =2.24.4
+REVISION = 0
DISTNAME = webkitgtk-${V}
PKGNAME = webkitgtk4-${V}
EXTRACT_SUFX = .tar.xz
Index:
patches/patch-Source_WebCore_Modules_indexeddb_server_SQLiteIDBBackingStore_cpp
===
RCS file:
patches/patch-Source_WebCore_Modules_indexeddb_server_SQLiteIDBBackingStore_cpp
diff -N
patches/patch-Source_WebCore_Modules_indexeddb_server_SQLiteIDBBackingStore_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++
patches/patch-Source_WebCore_Modules_indexeddb_server_SQLiteIDBBackingStore_cpp
29 Aug 2019 14:14:46 -
@@ -0,0 +1,22 @@
+$OpenBSD$
+
+ports-clang: fix undefined reference to `JSC::JSObject::didBecomePrototype()'
+from:
+https://trac.webkit.org/changeset/246922/webkit
+
+Index: Source/WebCore/Modules/indexeddb/server/SQLiteIDBBackingStore.cpp
+--- Source/WebCore/Modules/indexeddb/server/SQLiteIDBBackingStore.cpp.orig
Source/WebCore/Modules/indexeddb/server/SQLiteIDBBackingStore.cpp
+@@ -48,11 +48,8 @@
+ #include "SQLiteTransaction.h"
+ #include "ThreadSafeDataBuffer.h"
+ #include
+-#include
+-#include
+-#include
++#include
+ #include
+-#include
+ #include
+ #include
+ #include
[macppc] Unbreak lang/parrot
> http://build-failures.rhaalovely.net/powerpc/2019-08-09/lang/parrot.log
${LOCALBASE}/include/ffitarget.h is actually fine. The build is failing
due to the (aggressive) usage of -Werror.
According to what i've read, _CALL_ELF is used to separate ppc64be from
ppc64le.
I could have removed the annoying -Werror, but i preferred to play fair
and defining it to (ELFv)1. While here, i've regen WANTLIB.
It indeed builds fine [0], there is only one test out of 47767 failing,
as seen on amd64.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/parrot.log
Index: Makefile
===
RCS file: /cvs/ports/lang/parrot/Makefile,v
retrieving revision 1.47
diff -u -p -u -p -r1.47 Makefile
--- Makefile12 Jul 2019 20:47:21 - 1.47
+++ Makefile27 Aug 2019 20:27:09 -
@@ -4,7 +4,7 @@ COMMENT=virtual machine designed for in
V= 7.10.0
DISTNAME= parrot-$V
-REVISION= 7
+REVISION= 8
CATEGORIES=lang perl6
SHARED_LIBS= parrot 11.0
@@ -16,10 +16,10 @@ MASTER_SITES= http://ftp.parrot.org/rele
# Artistic2
PERMIT_PACKAGE=Yes
-WANTLIB += GL GLU c glut gmp iconv icudata icuuc intl m ncurses pthread
-WANTLIB += readline ${COMPILER_LIBCXX} util z ffi
+WANTLIB += ${COMPILER_LIBCXX} GL GLU c ffi glut gmp icudata icuuc
+WANTLIB += intl m readline z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER = base-clang ports-gcc
BUILD_DEPENDS= graphics/glew \
devel/pcre
@@ -55,6 +55,11 @@ INSTALL_TARGET = install install-doc
TEST_TARGET= fulltest
SUBST_VARS += V
+
+# fix for error: "_CALL_ELF" is not defined, evaluates to 0
+.if ${MACHINE_ARCH:Mpowerpc}
+CFLAGS += -D_CALL_ELF=1
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/config/init/hints/openbsd.pm
[ports-gcc] Unbreak graphics/makehuman
Hi! > http://build-failures.rhaalovely.net/powerpc/2019-08-09/graphics/makehuman.log > http://build-failures.rhaalovely.net/sparc64/2019-08-21/graphics/makehuman.log It's just some missing headers actually, and once provided it builds [0] and runs fine on macppc. amd64 is still good. While here i've removed base-gcc from COMPILER, because of COMPILER_LIBCXX being in WANTLIB. There is still no https version of HOMEPAGE. OK? Charlène. [0] https://bin.charlenew.xyz/makehuman.log Index: Makefile === RCS file: /cvs/ports/graphics/makehuman/Makefile,v retrieving revision 1.28 diff -u -p -u -p -r1.28 Makefile --- Makefile12 Jul 2019 20:47:03 - 1.28 +++ Makefile26 Aug 2019 16:16:31 - @@ -5,7 +5,7 @@ COMMENT=parametrical modeling of 3D hum DISTNAME= makehuman-0.9.1-rc1a PKGNAME= makehuman-0.9.1rc1 CATEGORIES=graphics -REVISION= 6 +REVISION= 7 # Newer releases available at https://github.com/makehumancommunity/makehuman HOMEPAGE= http://www.makehumancommunity.org/ @@ -18,7 +18,7 @@ WANTLIB += Xfixes Xi Xmu Xrandr Xrender WANTLIB += glapi glut m mhgui png pthread xcb xcb-dri2 xcb-dri3 xcb-glx WANTLIB += xcb-present xcb-sync xcb-xfixes xshmfence z -COMPILER = base-clang ports-gcc base-gcc +COMPILER = base-clang ports-gcc # gcp(1), see pre-configure BUILD_DEPENDS= sysutils/coreutils Index: patches/patch-src_FileTools_cpp === RCS file: patches/patch-src_FileTools_cpp diff -N patches/patch-src_FileTools_cpp --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_FileTools_cpp 26 Aug 2019 16:16:31 - @@ -0,0 +1,15 @@ +$OpenBSD$ + +Add missing header for ports-gcc + +Index: src/FileTools.cpp +--- src/FileTools.cpp.orig src/FileTools.cpp +@@ -38,6 +38,7 @@ + #endif // _FileTools_H_ + + #include ++#include + #include + #include + #include Index: patches/patch-src_util_h === RCS file: patches/patch-src_util_h diff -N patches/patch-src_util_h --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_util_h26 Aug 2019 16:16:31 - @@ -0,0 +1,15 @@ +$OpenBSD$ + +Add missing headers for ports-gcc + +Index: src/util.h +--- src/util.h.orig src/util.h +@@ -31,6 +31,7 @@ + #include + #endif + ++#include + #include + #include + #include
Re: UPDATE: x11/icewm 1.5.5 => 1.6.1
Hi, On Sun, 25 Aug 2019 20:14:51 -0400 Brian Callahan wrote: > Hi ports -- > > Attached is an update to IceWM. > Changelog: https://github.com/ice-wm/icewm/releases > > Works fine on amd64; big endian tests appreciated. It builds and works fine on macppc. > Note that I've removed myself as maintainer; I've moved on from > IceWM. Now would be a good time for a user to step up and maintain > this. I only use it on macppc, and clearly not enough to take it. > OK? OK cwen@ > ~Brian
Re: [ports-gcc] Unbreak x11/kde4/krfb
On Sun, 25 Aug 2019 15:20:01 -0600
Theo de Raadt wrote:
> I suspect the question is why is it getting the max() macro. is it
> getting from sys/param.h? Does some code have an #ifdef which pulls
> sys/param.h on some platforms or not others?
I used eg++ -M to see which headers are pulled, and sys/param.h does
not appear. I also used '-dD -E' to no avail - actually min and max are
undef'd by bits/c++config.h.
Even working against a way simpler port that has this issue
(games/xevil) didn't bring something interesting.
base-gcc and base-clang c++ STL use a similar definition of max() and
min(), so it works. ports-gcc uses an inlined version instead. That's
the only difference i've seen.
As such, i have no clue about what is going on.
> If this is from sys/param.h, I would suggest finding out why and
> stopping that practice. I think all systems still have it in
> sys/param.h which is why I suspect someone has an #ifdef to get at
> something else in the file, and therefore causing this problem
>
> Charlene Wendling wrote:
>
> >
> >
> > > http://build-failures.rhaalovely.net/sparc64/2019-08-21/x11/kde4/krfb.log
> > > http://build-failures.rhaalovely.net/powerpc/2019-08-09/x11/kde4/krfb.log
> >
> > We've 3 issues here:
> >
> > - a typical max() macro clash in the bundled libvncviewer. I did it
> > the harsh way, upstream has a different fix, but it has a bigger
> > footprint on the codebase [0]
> > - there was an occurrence where FALSE and TRUE were undefined, so i
> > redefined them. Looks like clang and gcc approaches are different
> > on that matter.
> > - headers must be placed as defined in style(9) in various places,
> > otherwise some variables/functions are undefined, heavily
> > inspired by how Debian did it [1].
> >
> > While here i've moved HOMEPAGE to https.
> >
> > It builds fine on macppc [2] and amd64. The runtime is also good on
> > these two platforms.
> >
> > Comments/feedback are welcome :)
> >
> > Charl??ne.
> >
> >
> > [0]
> > https://github.com/LibVNC/libvncserver/commit/785f0fa2d1fe18f08bf521ed4a0d34a5d0c87782
> > [1]
> > https://sources.debian.org/patches/krfb/4:17.08.3-1/build_against_gcc-6.patch/
> > [2] https://bin.charlenew.xyz/krfb.log
> >
> >
> > Index: Makefile
> > ===
> > RCS file: /cvs/ports/x11/kde4/krfb/Makefile,v
> > retrieving revision 1.9
> > diff -u -p -u -p -r1.9 Makefile
> > --- Makefile27 Jun 2018 21:04:17 - 1.9
> > +++ Makefile25 Aug 2019 20:35:56 -
> > @@ -2,10 +2,10 @@
> >
> > COMMENT = desktop sharing
> > DISTNAME = krfb-${MODKDE4_VERSION}
> > -REVISION = 4
> > +REVISION = 5
> >
> > CATEGORIES = net
> > -HOMEPAGE =
> > http://projects.kde.org/projects/kde/kdenetwork/krfb +HOMEPAGE
> > = https://userbase.kde.org/Krfb
> > SHARED_LIBS = krfbprivate 52.0
> >
> > Index: patches/patch-krfb_events_cpp
> > ===
> > RCS file: patches/patch-krfb_events_cpp
> > diff -N patches/patch-krfb_events_cpp
> > --- /dev/null 1 Jan 1970 00:00:00 -
> > +++ patches/patch-krfb_events_cpp 25 Aug 2019 20:35:56 -
> > @@ -0,0 +1,25 @@
> > +$OpenBSD$
> > +
> > +Fix for ports-gcc>=6
> > +
> > +Index: krfb/events.cpp
> > +--- krfb/events.cpp.orig
> > krfb/events.cpp
> > +@@ -22,8 +22,6 @@
> > +Boston, MA 02110-1301, USA.
> > + */
> > +
> > +-#include "events.h"
> > +-
> > + #include
> > + #include
> > + #include
> > +@@ -32,6 +30,8 @@
> > + #include
> > + #include
> > + #include
> > ++
> > ++#include "events.h"
> > +
> > + enum {
> > + LEFTSHIFT = 1,
> > Index: patches/patch-krfb_framebuffer_h
> > ===
> > RCS file: patches/patch-krfb_framebuffer_h
> > diff -N patches/patch-krfb_framebuffer_h
> > --- /dev/null 1 Jan 1970 00:00:00 -
> > +++ patches/patch-krfb_framebuffer_h25 Aug 2019 20:35:56
> > - @@ -0,0 +1,24 @@
> > +$OpenBSD$
> > +Fix for ports-gcc>=6
> > +
> > +Index: krfb/framebuffer.h
> > +--- krfb/framebuffer.h.orig
> > krfb/framebuffer.h
> > +@@ -10,8 +10,6 @@
> > + #ifndef FRAMEBUFFER_H
> > + #define FRAMEBUFFER_H
> > +
> > +-#include "rfb.h"
> > +
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/25 15:43:39 Modified files: x11/kde4/kopete: Makefile Added files: x11/kde4/kopete/patches: patch-protocols_jabber_libiris_src_xmpp_base64_base64_cpp Log message: kopete: fix the build on archs where char is unsigned by default (arm, ppc). While here, move HOMEPAGE to https. OK rsadowski@
[ports-gcc] Unbreak x11/kde4/krfb
> http://build-failures.rhaalovely.net/sparc64/2019-08-21/x11/kde4/krfb.log
> http://build-failures.rhaalovely.net/powerpc/2019-08-09/x11/kde4/krfb.log
We've 3 issues here:
- a typical max() macro clash in the bundled libvncviewer. I did it the
harsh way, upstream has a different fix, but it has a bigger footprint
on the codebase [0]
- there was an occurrence where FALSE and TRUE were undefined, so i
redefined them. Looks like clang and gcc approaches are different on
that matter.
- headers must be placed as defined in style(9) in various places,
otherwise some variables/functions are undefined, heavily inspired by
how Debian did it [1].
While here i've moved HOMEPAGE to https.
It builds fine on macppc [2] and amd64. The runtime is also good on
these two platforms.
Comments/feedback are welcome :)
Charlène.
[0]
https://github.com/LibVNC/libvncserver/commit/785f0fa2d1fe18f08bf521ed4a0d34a5d0c87782
[1]
https://sources.debian.org/patches/krfb/4:17.08.3-1/build_against_gcc-6.patch/
[2] https://bin.charlenew.xyz/krfb.log
Index: Makefile
===
RCS file: /cvs/ports/x11/kde4/krfb/Makefile,v
retrieving revision 1.9
diff -u -p -u -p -r1.9 Makefile
--- Makefile27 Jun 2018 21:04:17 - 1.9
+++ Makefile25 Aug 2019 20:35:56 -
@@ -2,10 +2,10 @@
COMMENT = desktop sharing
DISTNAME = krfb-${MODKDE4_VERSION}
-REVISION = 4
+REVISION = 5
CATEGORIES = net
-HOMEPAGE = http://projects.kde.org/projects/kde/kdenetwork/krfb
+HOMEPAGE = https://userbase.kde.org/Krfb
SHARED_LIBS = krfbprivate 52.0
Index: patches/patch-krfb_events_cpp
===
RCS file: patches/patch-krfb_events_cpp
diff -N patches/patch-krfb_events_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-krfb_events_cpp 25 Aug 2019 20:35:56 -
@@ -0,0 +1,25 @@
+$OpenBSD$
+
+Fix for ports-gcc>=6
+
+Index: krfb/events.cpp
+--- krfb/events.cpp.orig
krfb/events.cpp
+@@ -22,8 +22,6 @@
+Boston, MA 02110-1301, USA.
+ */
+
+-#include "events.h"
+-
+ #include
+ #include
+ #include
+@@ -32,6 +30,8 @@
+ #include
+ #include
+ #include
++
++#include "events.h"
+
+ enum {
+ LEFTSHIFT = 1,
Index: patches/patch-krfb_framebuffer_h
===
RCS file: patches/patch-krfb_framebuffer_h
diff -N patches/patch-krfb_framebuffer_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-krfb_framebuffer_h25 Aug 2019 20:35:56 -
@@ -0,0 +1,24 @@
+$OpenBSD$
+Fix for ports-gcc>=6
+
+Index: krfb/framebuffer.h
+--- krfb/framebuffer.h.orig
krfb/framebuffer.h
+@@ -10,8 +10,6 @@
+ #ifndef FRAMEBUFFER_H
+ #define FRAMEBUFFER_H
+
+-#include "rfb.h"
+-
+ #include
+
+ #include
+@@ -19,6 +17,8 @@
+ #include
+
+ #include
++
++#include "rfb.h"
+
+ class FrameBuffer;
+ /**
Index: patches/patch-krfb_invitationsrfbclient_cpp
===
RCS file: patches/patch-krfb_invitationsrfbclient_cpp
diff -N patches/patch-krfb_invitationsrfbclient_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-krfb_invitationsrfbclient_cpp 25 Aug 2019 20:35:56 -
@@ -0,0 +1,28 @@
+$OpenBSD$
+Fix for ports-gcc>=6
+
+Index: krfb/invitationsrfbclient.cpp
+--- krfb/invitationsrfbclient.cpp.orig
krfb/invitationsrfbclient.cpp
+@@ -18,16 +18,17 @@
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/25 14:10:38 Modified files: x11/qt5/qt3d : Makefile Added files: x11/qt5/qt3d/patches: patch-src_3rdparty_assimp_code_glTFAsset_inl Log message: qt3d: fix the build with ports-gcc. There was a namespace issue in the bundled assimp. OK jca@, "go ahead" rsadowski@
[arm*, macppc] Unbreak x11/kde4/kopete
Hi,
> http://build-failures.rhaalovely.net/aarch64/2019-08-16/x11/kde4/kopete.log
> http://build-failures.rhaalovely.net/powerpc/2019-08-09/x11/kde4/kopete.log
It's the same base64 code copied and pasted from projects to projects.
Once that array is declared as signed, it builds on macppc [0], and it's
still fine on amd64.
While here, i've updated HOMEPAGE, and moved to https in the same time.
OK?
Charlène.
Index: Makefile
===
RCS file: /cvs/ports/x11/kde4/kopete/Makefile,v
retrieving revision 1.22
diff -u -p -u -p -r1.22 Makefile
--- Makefile23 Jan 2019 18:45:21 - 1.22
+++ Makefile25 Aug 2019 19:58:14 -
@@ -6,11 +6,11 @@ DISTNAME =kopete-${MODKDE4_VERSION}
PKGNAME-main = ${DISTNAME}
PKGNAME-sounds = kde4-kopete-sounds-${MODKDE4_VERSION}
PKG_ARCH-sounds = *
-REVISION-main =9
-REVISION-sounds = 4
+REVISION-main =10
+REVISION-sounds = 5
CATEGORIES = net
-HOMEPAGE = http://kopete.kde.org/
+HOMEPAGE = https://userbase.kde.org/Kopete
SHARED_LIBS += kopete 52.0 # 4.0
SHARED_LIBS += kopete_videodevice 52.0 # 4.0
Index: patches/patch-protocols_jabber_libiris_src_xmpp_base64_base64_cpp
===
RCS file: patches/patch-protocols_jabber_libiris_src_xmpp_base64_base64_cpp
diff -N patches/patch-protocols_jabber_libiris_src_xmpp_base64_base64_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-protocols_jabber_libiris_src_xmpp_base64_base64_cpp 25 Aug
2019 19:58:14 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+fix the build on archs where char is unsigned by default (ppc, arm)
+
+Index: protocols/jabber/libiris/src/xmpp/base64/base64.cpp
+--- protocols/jabber/libiris/src/xmpp/base64/base64.cpp.orig
protocols/jabber/libiris/src/xmpp/base64/base64.cpp
+@@ -45,7 +45,7 @@ QByteArray Base64::decode(const QString& input)
+ // 64 specifies eof
+ // everything else specifies data
+
+- char tbl[] = {
++ signed char tbl[] = {
+ -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
+ -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
+ -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,62,-1,-1,-1,63,
Re: [macppc, hppa?] Unbreak games/godot
Ping.
On Sat, 17 Aug 2019 00:38:43 +0200
Charlene Wendling wrote:
>
> > https://marc.info/?l=openbsd-ports=15654325765=2
>
> As promised, here is the conversion from __sync_* to __atomic_*
> functions, allowing godot to build on macppc at least, and maybe
> hppa, on top of the ports-gcc/sparc64 fixes, that Thomas committed a
> few hours ago.
>
> The patch was originally written for godot-3.1.1, and has been
> upstreamed [0]. I backported it for godot-3.0.6, it builds
> successfully on amd64 and macppc.
>
> Due to relocations errors, i had to add address relaxing and
> long calls.
>
> Runtime cannot be tested with my video card (radeon 9700) with 3.0.6,
> because it requires OpenGL(ES) 3, even for opening the editor, unlike
> godot-3.1.1.
>
> Comments/feedback are welcome,
>
> Charlène.
>
>
> [0] https://github.com/godotengine/godot/pull/31321
Index: Makefile
===
RCS file: /cvs/ports/games/godot/Makefile,v
retrieving revision 1.8
diff -u -p -u -p -r1.8 Makefile
--- Makefile16 Aug 2019 15:38:15 - 1.8
+++ Makefile16 Aug 2019 22:02:03 -
@@ -8,7 +8,7 @@ PKGNAME = godot-${V}
CATEGORIES = games
HOMEPAGE = https://godotengine.org/
MAINTAINER = Thomas Frohwein
-REVISION = 2
+REVISION = 3
# MIT
PERMIT_PACKAGE = Yes
@@ -68,6 +68,18 @@ LIB_DEPENDS =archivers/zstd \
net/enet
NO_TEST = Yes
+
+.if ${MACHINE_ARCH:Mhppa} || ${MACHINE_ARCH:Mpowerpc}
+LDFLAGS += -latomic
+WANTLIB += atomic
+.endif
+
+# Fix relocation overflows
+.if ${MACHINE_ARCH:Mpowerpc}
+CFLAGS += -mlongcall
+CXXFLAGS +=-mlongcall
+LDFLAGS += -Wl,--relax
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/drivers/unix/os_unix.cpp
Index: patches/patch-core_safe_refcount_h
===
RCS file: patches/patch-core_safe_refcount_h
diff -N patches/patch-core_safe_refcount_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-core_safe_refcount_h 16 Aug 2019 22:02:03 -
@@ -0,0 +1,68 @@
+$OpenBSD$
+
+hppa, ppc: use __atomic functions as 64-bit __sync operators
+are not supported, from:
+https://github.com/godotengine/godot/pull/31321
+
+Index: core/safe_refcount.h
+--- core/safe_refcount.h.orig
core/safe_refcount.h
+@@ -99,8 +99,8 @@ static _ALWAYS_INLINE_ T atomic_exchange_if_greater(re
+
+ /* Implementation for GCC & Clang */
+
+-// GCC guarantees atomic intrinsics for sizes of 1, 2, 4 and 8 bytes.
+-// Clang states it supports GCC atomic builtins.
++#include
++#include
+
+ template
+ static _ALWAYS_INLINE_ T atomic_conditional_increment(register T *pw) {
+@@ -109,7 +109,7 @@ static _ALWAYS_INLINE_ T atomic_conditional_increment(
+ T tmp = static_cast(*pw);
+ if (tmp == 0)
+ return 0; // if zero, can't add to it anymore
+- if (__sync_val_compare_and_swap(pw, tmp, tmp + 1) == tmp)
++ if (__atomic_compare_exchange_n(pw, , tmp + 1, false,
__ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) == true)
+ return tmp + 1;
+ }
+ }
+@@ -117,25 +117,25 @@ static _ALWAYS_INLINE_ T atomic_conditional_increment(
+ template
+ static _ALWAYS_INLINE_ T atomic_decrement(register T *pw) {
+
+- return __sync_sub_and_fetch(pw, 1);
++ return __atomic_sub_fetch(pw, 1, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_increment(register T *pw) {
+
+- return __sync_add_and_fetch(pw, 1);
++ return __atomic_add_fetch(pw, 1, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_sub(register T *pw, register V val) {
+
+- return __sync_sub_and_fetch(pw, val);
++ return __atomic_sub_fetch(pw, val, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_add(register T *pw, register V val) {
+
+- return __sync_add_and_fetch(pw, val);
++ return __atomic_add_fetch(pw, val, __ATOMIC_SEQ_CST);
+ }
+
+ template
+@@ -145,7 +145,7 @@ static _ALWAYS_INLINE_ T atomic_exchange_if_greater(re
+ T tmp = static_cast(*pw);
+ if (tmp >= val)
+ return tmp; // already greater, or equal
+- if (__sync_val_compare_and_swap(pw, tmp, val) == tmp)
++ if (__atomic_compare_exchange_n(pw, , val, false,
__ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) == true)
+ return val;
+ }
+ }
[ports-gcc] Unbreak x11/qt5/qt3d
Hi,
> http://build-failures.rhaalovely.net/powerpc/2019-08-09/x11/qt5/qt3d.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-21/x11/qt5/qt3d.log
There used to have more issues with gcc-4.9, but with gcc-8.3, only the
bundled assimp is troublesome :)
Recent assimp versions have a fix for it [0], that allows qt3d to build
on macppc [1]. Amd64 is still fine and geo/qgis does not misbehave.
OK?
Charlène.
[0] https://github.com/assimp/assimp/pull/986
[1] https://bin.charlenew.xyz/qt3d.log.gz
Index: Makefile
===
RCS file: /cvs/ports/x11/qt5/qt3d/Makefile,v
retrieving revision 1.10
diff -u -p -u -p -r1.10 Makefile
--- Makefile24 Jun 2019 20:56:45 - 1.10
+++ Makefile24 Aug 2019 12:15:31 -
@@ -5,7 +5,7 @@
QT5NAME = Qt3D
COMMENT-main = Qt5 components for 3D graphics
-REVISION-main =1
+REVISION-main =2
MULTI_PACKAGES = -main -examples
Index: patches/patch-src_3rdparty_assimp_code_glTFAsset_inl
===
RCS file: patches/patch-src_3rdparty_assimp_code_glTFAsset_inl
diff -N patches/patch-src_3rdparty_assimp_code_glTFAsset_inl
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_3rdparty_assimp_code_glTFAsset_inl24 Aug 2019
12:15:31 -
@@ -0,0 +1,17 @@
+$OpenBSD$
+
+Fix the build on big endian archs. From:
+https://github.com/assimp/assimp/pull/986/commits/c2c12c1db5c8d7610ad46b1970aabf4afde0eef4
+
+Index: src/3rdparty/assimp/code/glTFAsset.inl
+--- src/3rdparty/assimp/code/glTFAsset.inl.orig
src/3rdparty/assimp/code/glTFAsset.inl
+@@ -40,6 +40,8 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY O
+
+ #include "StringUtils.h"
+
++using namespace Assimp;
++
+ namespace glTF {
+
+ namespace {
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/23 14:20:03 Modified files: audio/py-acoustid: Makefile audio/picard : Makefile Log message: Bump consumers REVISIONs following the audio/chromaprint update. Hints and OK sthen@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/23 14:18:04 Modified files: audio/chromaprint: Makefile distinfo audio/chromaprint/patches: patch-cmake_modules_FindGTest_cmake patch-tests_CMakeLists_txt audio/chromaprint/pkg: DESCR Added files: audio/chromaprint/patches: patch-tests_test_utils_cpp patch-tests_test_utils_h Removed files: audio/chromaprint/patches: patch-examples_fpcalc_c Log message: chromaprint: update to 1.4.3, enable all tests again Changelog: https://github.com/acoustid/chromaprint/releases/tag/v1.4.3 OK sthen@
Re: [ports-gcc] Unbreak x11/waimea
ping.
> > http://build-failures.rhaalovely.net/powerpc/2019-07-29/x11/waimea.log
> > http://build-failures.rhaalovely.net/sparc64/2019-08-12/x11/waimea.log
>
> There is a clash between std::list and the "list" class member.
>
> I've just renamed that member and its instances. It builds [0] and
> works fine on macppc, and amd64 is not broken.
>
> While here WANTLIB needed a refresh, and i moved HOMEPAGE to https.
>
> OK?
>
> Charlène.
>
>
> [0] https://bin.charlenew.xyz/waimea.log
>
Index: Makefile
===
RCS file: /cvs/ports/x11/waimea/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:51:23 - 1.26
+++ Makefile15 Aug 2019 09:56:07 -
@@ -4,21 +4,20 @@ COMMENT= virtual desktop manager for X1
VERSION= 0.4.0
DISTNAME= waimea-${VERSION}
-REVISION = 9
+REVISION= 10
CATEGORIES=x11
-HOMEPAGE= http://www.freedesktop.org/wiki/Software/waimea
+HOMEPAGE= https://www.freedesktop.org/wiki/Software/waimea
# GPLv2
PERMIT_PACKAGE=Yes
-MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
+WANTLIB += ${COMPILER_LIBCXX} ICE SM X11 Xext Xft Xinerama Xrandr
+WANTLIB += Xrender c m
-WANTLIB += ICE SM X11 Xau Xdmcp Xext Xft Xinerama Xrandr Xrender
-WANTLIB += c expat fontconfig freetype m ${COMPILER_LIBCXX}
-WANTLIB += xcb z
+MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
AUTOCONF_VERSION= 2.52
CONFIGURE_STYLE= autoconf
Index: patches/patch-src_Resources_cc
===
RCS file: patches/patch-src_Resources_cc
diff -N patches/patch-src_Resources_cc
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_cc 15 Aug 2019 09:56:07 -
@@ -0,0 +1,35 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.cc
+--- src/Resources.cc.orig
src/Resources.cc
+@@ -1568,7 +1568,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 'n' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1583,7 +1583,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 't' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1598,7 +1598,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+str + 2);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (! strncasecmp(str, "window", 6)) {
+ str = str + 6;
Index: patches/patch-src_Resources_hh
===
RCS file: patches/patch-src_Resources_hh
diff -N patches/patch-src_Resources_hh
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_hh 15 Aug 2019 09:56:07 -
@@ -0,0 +1,24 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.hh
+--- src/Resources.hh.orig
src/Resources.hh
+@@ -193,13 +193,13 @@ class WaActionExtList { (public)
+ delete name;
+ delete cl;
+ delete title;
+-ACTLISTCLEAR(list);
++ACTLISTCLEAR(alist);
+ }
+
+ Regex *name;
+ Regex *cl;
+ Regex *title;
+-list list;
++list alist;
+ };
+
+ class StrComp {
Index: patches/patch-src_Window_cc
===
RCS file: patches/patch-src_Window_cc
diff -N patches/patch-src_Window_cc
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Window_cc 15 Aug 2019 09:56:07 -
@@ -0,0 +1,25 @@
+$OpenBSD$
+
Re: math/plplot on powerpc
On Thu, 22 Aug 2019 20:42:59 +0200 Martin Reindl wrote: > Hello, > > plplot on powerpc works just fine now, this part did not make it with > the cmake/plplot update. > > -m It works fine on my powerbook as well, thanks :) OK cwen@
Re: Update: archivers/bzip2 1.0.8
On Wed, 21 Aug 2019 22:40:52 +0200 Christian Weisgerber wrote: > Update archivers/bzip2 to 1.0.8. The original author is involved > again. > > From CHANGES: > * Accept as many selectors as the file format allows. > This relaxes the fix for CVE-2019-12900 from 1.0.7 > so that bzip2 allows decompression of bz2 files that > use (too) many selectors again. > * Cleanup of bzdiff and bzgrep scripts so they don't use > any bash extensions and handle multiple archives correctly. > > There's also a HOMEPAGE again. > > I have not touched the additional CFLAGS we have been cargo-culting > forward since forever. > > Builds and passes make test on amd64 and arm64. > > OK? On macppc, it builds and tests pass as well. OK cwen@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/21 13:30:52 Modified files: x11/ctwm : Makefile distinfo Log message: ctwm: update to 4.0.3 Changelog: https://www.ctwm.org/CHANGES.html OK sthen@
[ports-gcc] Unbreak multimedia/mkvtoolnix (impacts all archs)
Hi, > http://build-failures.rhaalovely.net/sparc64/2019-08-18/multimedia/mkvtoolnix%2Cno_x11.log (no full macppc bulk report available for now) What happens behind the scene (--trace didn't help): - $WRKSRC/rake.d/config.rb loads the 'build-config' file, that says that "COMPILER_VERSION = 8.3.0" (fine) - $WRKSRC/Rakefile does various compiler version checks to determine proper build flags to use - Compiler versions are checked in $WRKSRC/rake.d/helpers.rb using Ruby's check_version() function, that seems to not like version numbers with letters at the start, as we can see in the log. The compiler version itself comes from the c() function... - ...that is defined in $WRKSRC/rake.d/config.rb, where the line 24 is interesting: var = (ENV[idx_s].nil? ? $config[idx.to_sym] : ENV[idx_s]).to_s It appears that this 'gcc4' is coming from our default MAKE_ENV: $ make show=MAKE_ENV [...] COMPILER_VERSION="gcc4" [...] (on amd64, it returns COMPILER_VERSION="clang") It did not break on amd64, because with clang, upstream uses the same compilation flags for all versions, so there is no version comparison done. As such, i've overridden bsd.port.mk's COMPILER_VERSION, letting upstream's build system doing its job. I decided to go for it for all archs, as this may prevent possible future breakage on base-clang archs. I did not bump REVISION because this version never built on ports-gcc archs and it causes no change on base-clang ones. This builds, and tests pass with flying colors, on amd64 and macppc [0]. Comments/feedback are welcome :) Charlène. [0] https://bin.charlenew.xyz/mkvtoolnix.log.gz Index: Makefile === RCS file: /cvs/ports/multimedia/mkvtoolnix/Makefile,v retrieving revision 1.92 diff -u -p -u -p -r1.92 Makefile --- Makefile12 Jul 2019 20:47:57 - 1.92 +++ Makefile21 Aug 2019 19:02:57 - @@ -29,6 +29,11 @@ MODRUBY_RUNDEP= No # C++11 / C++14 COMPILER = base-clang ports-gcc +# Override bsd.port.mk's COMPILER_VERSION because 'gcc4' is incorrect +# with ports-gcc and not recognised as a version number by Ruby's +# check_version(), and is just 'clang' with base-clang. +COMPILER_VERSION= + BUILD_DEPENDS= devel/gettext,-tools \ devel/gtest \ textproc/docbook \
Re: [update] x11/ctwm 4.0.2 -> 4.0.3
ping. On Wed, 14 Aug 2019 17:52:15 +0200 Charlene Wendling wrote: > > Here is a very simple update for ctwm. This only contains bugfixes > [0]. > > It builds and works fine on amd64 and macppc. > > Comments/feedback are welcome, > > Charlène. > > > [0] https://www.ctwm.org/CHANGES.html Index: Makefile === RCS file: /cvs/ports/x11/ctwm/Makefile,v retrieving revision 1.28 diff -u -p -u -p -r1.28 Makefile --- Makefile12 Jul 2019 20:51:09 - 1.28 +++ Makefile14 Aug 2019 15:45:38 - @@ -2,7 +2,7 @@ COMMENT= twm, plus support for multiple virtual screens, etc -DISTNAME= ctwm-4.0.2 +DISTNAME= ctwm-4.0.3 CATEGORIES=x11 Index: distinfo === RCS file: /cvs/ports/x11/ctwm/distinfo,v retrieving revision 1.7 diff -u -p -u -p -r1.7 distinfo --- distinfo29 Aug 2018 21:06:43 - 1.7 +++ distinfo14 Aug 2019 15:45:38 - @@ -1,2 +1,2 @@ -SHA256 (ctwm-4.0.2.tar.gz) = U3Slep45FuFOm4qrt+BsZYt9gu2FrozUZEqWJoXIZNU= -SIZE (ctwm-4.0.2.tar.gz) = 915254 +SHA256 (ctwm-4.0.3.tar.gz) = 3ezUUw/R76LRnVlNmAPm6fFXiDlviRwxldWtIyBS4Bc= +SIZE (ctwm-4.0.3.tar.gz) = 934415
Re: [update] audio/chromaprint 1.1 -> 1.4.3
ping.
On Wed, 14 Aug 2019 15:02:54 +0200
Charlene Wendling wrote:
> Hi,
>
> Here is an update for chromaprint that fixes build issues on ppc/arm.
>
> Upstream didn't pin a 1.4.4 release [1], but this is what you get with
> that diff.
>
> What's new upstream (see [0]):
>
> - move to github
> - C++11 is required, chromaprint doesn't use boost anymore
> - use newer ffmpeg api
>
> What's new in the port:
>
> - reorder the Makefile according to Makefile.template. It doesn't
> impact a lot the diff readability given the amount of changes.
> - major bump SHARED_LIBS. check_sym reports deletions.
> - s/sed/SUBST_CMD
> - remove hardcoded path when finding gtest, it also required to add
> linking instructions for cmake to allow the use of devel/gtest
> - add "1.4.4" fixes that make test endian neutral
>
> Testing (macppc/amd64):
>
> - 'make test' passes with all tests enabled
> - All consumers build and run without issue. Clementine auto-tagging
> doesn't work with chromaprint-1.1 as well.
>
>
> Comments/feedback are welcome :)
>
> Charlène.
>
>
> [0]
> https://github.com/acoustid/chromaprint/blob/master/NEWS.txt
> [1] https://github.com/acoustid/chromaprint/issues/59
Index: chromaprint/Makefile
===
RCS file: /cvs/ports/audio/chromaprint/Makefile,v
retrieving revision 1.12
diff -u -p -r1.12 Makefile
--- chromaprint/Makefile12 Jul 2019 20:43:32 - 1.12
+++ chromaprint/Makefile13 Aug 2019 21:18:50 -
@@ -2,37 +2,39 @@
COMMENT = audio fingerprint extraction library
-DISTNAME = chromaprint-1.1
+GH_ACCOUNT = acoustid
+GH_PROJECT = chromaprint
+GH_TAGNAME = v1.4.3
+
+SHARED_LIBS = chromaprint 2.0 # 1.4.3
+
CATEGORIES = audio devel
-HOMEPAGE = https://acoustid.org/chromaprint
-MASTER_SITES = https://bitbucket.org/acoustid/chromaprint/downloads/
-REVISION = 3
-SHARED_LIBS = chromaprint 1.0 # 0.1
+HOMEPAGE = https://acoustid.org/chromaprint
-# LGPL2.1+
+# MIT with LGPL2.1+ parts
PERMIT_PACKAGE = Yes
-WANTLIB += avcodec avformat avutil c m pthread ${COMPILER_LIBCXX} swresample
+WANTLIB += ${COMPILER_LIBCXX} avcodec avformat avutil c m swresample
-COMPILER = base-clang ports-gcc base-gcc
+# C++11
+COMPILER = base-clang ports-gcc
MODULES = devel/cmake
-BUILD_DEPENDS =audio/taglib \
- devel/boost
+BUILD_DEPENDS =audio/taglib
+# gtest presence is checked in configure stage, so this cannot be
+# in TEST_DEPENDS
+BUILD_DEPENDS += devel/gtest
LIB_DEPENDS = graphics/ffmpeg
-# gtest presence is checked in configure stage, so this cannot be in
TEST_DEPENDS
-BUILD_DEPENDS += devel/gtest
-CONFIGURE_ARGS = -DBUILD_TESTS:Bool=Yes -DBUILD_EXAMPLES=ON
+CONFIGURE_ARGS = -DBUILD_TOOLS=ON \
+ -DBUILD_TESTS=ON
-post-patch:
- sed -i 's,/usr/local,${LOCALBASE},' \
- ${WRKSRC}/cmake/modules/FindGTest.cmake
+TEST_TARGET = check
-do-test:
- cd ${WRKBUILD}/tests && ./all_tests
+pre-configure:
+ ${SUBST_CMD} ${WRKSRC}/cmake/modules/FindGTest.cmake
.include
Index: chromaprint/distinfo
===
RCS file: /cvs/ports/audio/chromaprint/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- chromaprint/distinfo9 May 2014 09:00:19 - 1.2
+++ chromaprint/distinfo13 Aug 2019 21:18:50 -
@@ -1,2 +1,2 @@
-SHA256 (chromaprint-1.1.tar.gz) = axTX6klkWBtzvT+AOMiFfAHkRkIcGumcu/ZN4mtHzRI=
-SIZE (chromaprint-1.1.tar.gz) = 542360
+SHA256 (chromaprint-1.4.3.tar.gz) =
1K5llig6rXoBWlsERQEgVMY0pLkynssjAAzTVLQKKDs=
+SIZE (chromaprint-1.4.3.tar.gz) = 613718
Index: chromaprint/patches/patch-cmake_modules_FindGTest_cmake
===
RCS file:
/cvs/ports/audio/chromaprint/patches/patch-cmake_modules_FindGTest_cmake,v
retrieving revision 1.1
diff -u -p -r1.1 patch-cmake_modules_FindGTest_cmake
--- chromaprint/patches/patch-cmake_modules_FindGTest_cmake 9 May 2014
09:00:19 - 1.1
+++ chromaprint/patches/patch-cmake_modules_FindGTest_cmake 13 Aug 2019
21:18:50 -
@@ -1,6 +1,10 @@
$OpenBSD: patch-cmake_modules_FindGTest_cmake,v 1.1 2014/05/09 09:00:19 sthen
Exp $
cmake/modules/FindGTest.cmake.orig Sat Nov 23 16:43:42 2013
-+++ cmake/modules/FindGTest.cmake Wed May 7 16:29:15 2014
+
+Find gtest from devel/gtest.
+
+Index: cmake/modules/FindGTest.cmake
+--- cmake/modules/FindGTest.cmake.orig
cmake/modules/FindGTest.cmake
@@ -71,12 +71,24 @@ find_path(GTEST_INCLUDE_D
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/21 05:20:25 Modified files: devel/woboq_codebrowser: Makefile Log message: woboq_codebrowser: requires ports-clang or ports-gcc on base-gcc archs. Also remove the clang module and use devel/llvm as a BUILD_DEPENDS (from jca@, thanks!). OK kmos@ jca@ (who tested on sparc64)
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/21 05:13:35 Modified files: graphics/dibuja: Makefile Log message: dibuja: fix the build on base-gcc archs. LIBCXX_COMPILER is defined in WANTLIB, as such ports-gcc is needed. OK jca@ bcallah@ (maintainer)
Re: [macppc/ports-gcc] Unbreak benchmarks/wrk
On Tue, 20 Aug 2019 21:00:15 +0200 Frederic Cambus wrote: > On Mon, Aug 12, 2019 at 10:05:30PM +0200, Charlene Wendling wrote: > > > > > http://build-failures.rhaalovely.net/powerpc/2019-07-29/benchmarks/wrk.log > > > > It requires a move from __sync to __atomic. Debian did that already > > [0], so i've taken their stuff. > > > > While wrk is written in C, the use of __atomic functions requires > > estdc++ and ports-gcc (we can't use COMPILER_LANGS=c). > > > > It builds [1] and runs fine on macppc and amd64 with multiple > > threads. > > > > Comments/feedback are welcome, > > The patches from Debian are for wrk 4.0.2, released in March 2016. > There has since been a 4.1.0 release in January 2018, without those > changes. I don't know if the Debian package maintainer tried to > upstream this diff or not, but it would be nice to find out, as the > diff changes code for all platforms. > It seems Debian didn't. I did a PR [0], but it seems pretty clear to me that upstream will refactor it. I'll resend a diff once things are settled. Charlène. [0] https://github.com/wg/wrk/pull/395
Re: [ports-clang] Unbreak devel/woboq_codebrowser
On Mon, 19 Aug 2019 22:39:34 +0200
Jeremie Courreges-Anglas wrote:
> On Mon, Aug 19 2019, Jeremie Courreges-Anglas wrote:
> > On Mon, Aug 19 2019, Charlene Wendling wrote:
> >>> http://build-failures.rhaalovely.net/sparc64/2019-08-15/devel/woboq_codebrowser.log
> >> (same thing for macppc, but latest full bulk report is archived)
> >>
[...]
> >>
> >> OK?
> >
> > yep, but obvious comment is obvious...
>
> Thinking more about it...
It's more explicit that my comment, and it works fine as well.
I attached the diff below.
[...]
> >> Index: Makefile
> >> ===
> >> RCS file: /cvs/ports/devel/woboq_codebrowser/Makefile,v
> >> retrieving revision 1.1.1.1
> >> diff -u -p -u -p -r1.1.1.1 Makefile
> >> --- Makefile 22 Jun 2019 12:16:56 - 1.1.1.1
> >> +++ Makefile 19 Aug 2019 18:25:27 -
> >> @@ -26,6 +26,9 @@ PERMIT_PACKAGE = Yes
> >> # uses pledge()
> >> WANTLIB = ${COMPILER_LIBCXX} c m
> >>
> >> +# required to use ports-clang as CXX on base-gcc arches
> >> +COMPILER =base-clang ports-clang
> >> +
> >> MODULES = devel/cmake \
> >>lang/clang
>
> COMPILER and MODULES = lang/clang look a bit redundant. I'd suggest
> using something more explicit:
>
> COMPILER = base-clang ports-clang
> BUILD_DEPENDS = devel/llvm
> RUN_DEPENDS = devel/llvm
>
> like done in devel/include-what-you-use.
>
> >> RUN_DEPENDS = devel/llvm
> --
> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE
> 1524 E7EE
>
Index: Makefile
===
RCS file: /cvs/ports/devel/woboq_codebrowser/Makefile,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 Makefile
--- Makefile22 Jun 2019 12:16:56 - 1.1.1.1
+++ Makefile19 Aug 2019 20:53:27 -
@@ -26,8 +26,10 @@ PERMIT_PACKAGE = Yes
# uses pledge()
WANTLIB = ${COMPILER_LIBCXX} c m
-MODULES = devel/cmake \
- lang/clang
+COMPILER = base-clang ports-clang
+
+MODULES = devel/cmake
+BUILD_DEPENDS =devel/llvm
RUN_DEPENDS = devel/llvm
NO_TEST = Yes
Re: [macppc/ports-gcc] Unbreak benchmarks/wrk
Ping.
On Mon, 12 Aug 2019 22:05:30 +0200
Charlene Wendling wrote:
>
> > http://build-failures.rhaalovely.net/powerpc/2019-07-29/benchmarks/wrk.log
>
> It requires a move from __sync to __atomic. Debian did that already
> [0], so i've taken their stuff.
>
> While wrk is written in C, the use of __atomic functions requires
> estdc++ and ports-gcc (we can't use COMPILER_LANGS=c).
>
> It builds [1] and runs fine on macppc and amd64 with multiple threads.
>
> Comments/feedback are welcome,
>
> Charlène.
>
>
> [0] https://sources.debian.org/patches/wrk/4.0.2-2/debian-changes/
> [1] https://bin.charlenew.xyz/wrk.log
Index: Makefile
===
RCS file: /cvs/ports/benchmarks/wrk/Makefile,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 Makefile
--- Makefile12 Jul 2019 20:43:43 - 1.2
+++ Makefile12 Aug 2019 19:40:01 -
@@ -5,6 +5,7 @@ COMMENT = modern HTTP benchmarking tool
GH_ACCOUNT = wg
GH_PROJECT = wrk
GH_TAGNAME = 4.1.0
+REVISION = 0
CATEGORIES = benchmarks
@@ -15,6 +16,9 @@ PERMIT_PACKAGE = Yes
WANTLIB += c crypto luajit-5.1 m pthread ssl
+# __atomic support (GCC>4.7.4)
+COMPILER = base-clang ports-gcc
+
LIB_DEPENDS = lang/luajit
USE_GMAKE =Yes
@@ -26,6 +30,12 @@ MAKE_ENV = WITH_LUAJIT=${LOCALBASE} WITH
MAKE_ENV +=LIBS=-lc++abi
WANTLIB += c++abi
.endif
+
+.if ${MACHINE_ARCH:Mhppa} || ${MACHINE_ARCH:Mpowerpc}
+MAKE_ENV +=LIBS=-latomic
+WANTLIB += atomic
+.endif
+
NO_TEST = Yes
Index: patches/patch-src_stats_c
===
RCS file: patches/patch-src_stats_c
diff -N patches/patch-src_stats_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_stats_c 12 Aug 2019 19:40:01 -
@@ -0,0 +1,33 @@
+$OpenBSD$
+
+ppc, hppa: we can't use 64-bit __sync operators there
+
+Index: src/stats.c
+--- src/stats.c.orig
src/stats.c
+@@ -21,12 +21,21 @@ void stats_free(stats *stats) {
+
+ int stats_record(stats *stats, uint64_t n) {
+ if (n >= stats->limit) return 0;
+-__sync_fetch_and_add(>data[n], 1);
+-__sync_fetch_and_add(>count, 1);
++__atomic_fetch_add(>data[n], 1, __ATOMIC_SEQ_CST);
++__atomic_fetch_add(>count, 1, __ATOMIC_SEQ_CST);
+ uint64_t min = stats->min;
+ uint64_t max = stats->max;
+-while (n < min) min = __sync_val_compare_and_swap(>min, min, n);
+-while (n > max) max = __sync_val_compare_and_swap(>max, max, n);
++while (n < min) {
++__atomic_compare_exchange(>min, , , false,
++ __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST);
++min = stats->min;
++}
++while (n > max) {
++__atomic_compare_exchange(>max, , , false,
++ __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST);
++max = stats->max;
++ }
++
+ return 1;
+ }
+
[ports-clang] Unbreak devel/woboq_codebrowser
> http://build-failures.rhaalovely.net/sparc64/2019-08-15/devel/woboq_codebrowser.log
(same thing for macppc, but latest full bulk report is archived)
This software requires clang as it uses libclang to analyse the code,
but on base-gcc archs:
===> Compiler link: gcc -> /usr/local/bin/clang
===> Compiler link: cc -> /usr/local/bin/clang
===> Compiler link: clang -> /usr/local/bin/clang
===> Compiler link: c++ -> /usr/bin/c++
[...]
-- The C compiler identification is Clang 8.0.0
-- The CXX compiler identification is GNU 4.2.1
It is lacking a COMPILER line to make it pick clang++. No REVISION
bump is needed, as it has never been built on base-gcc archs
since its import.
Once added it builds fine on macppc [0].
OK?
Charlène.
[0] https://bin.charlenew.xyz/woboq_codebrowser.log
Index: Makefile
===
RCS file: /cvs/ports/devel/woboq_codebrowser/Makefile,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 Makefile
--- Makefile22 Jun 2019 12:16:56 - 1.1.1.1
+++ Makefile19 Aug 2019 18:25:27 -
@@ -26,6 +26,9 @@ PERMIT_PACKAGE = Yes
# uses pledge()
WANTLIB = ${COMPILER_LIBCXX} c m
+# required to use ports-clang as CXX on base-gcc arches
+COMPILER = base-clang ports-clang
+
MODULES = devel/cmake \
lang/clang
RUN_DEPENDS = devel/llvm
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/19 12:12:31 Modified files: games/puzzles : Makefile Added files: games/puzzles/patches: patch-palisade_c Log message: puzzles: unbreak on archs where char is unsigned by default (ppc, arm) OK landry@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/18 13:23:41 Modified files: inputmethods/scim-tables: Makefile Added files: inputmethods/scim-tables/patches: patch-src_scim_generic_table_h Log message: scim-tables: add missing headers for ports-gcc. Regen WANTLIB and remove base-gcc from COMPILER while here. OK kmos@
[macppc,arm*] Unbreak games/puzzles
> http://build-failures.rhaalovely.net/aarch64/2019-08-16/games/puzzles.log
(no link to macppc build failure, landry has been a bit too much
aggressive while archiving failure logs)
Well, this is exactly what the report says, char being unsigned by
default on ppc and arm archs.
I decided to not remove -Werror while here - it has failed to build for
a legit reason this time. I will if asked to do so.
It builds and run fine on macppc [0] and amd64.
OK?
Charlène.
[0] https://bin.charlenew.xyz/puzzles.log
Index: Makefile
===
RCS file: /cvs/ports/games/puzzles/Makefile,v
retrieving revision 1.18
diff -u -p -u -p -r1.18 Makefile
--- Makefile12 Jul 2019 20:46:23 - 1.18
+++ Makefile17 Aug 2019 21:54:21 -
@@ -3,6 +3,7 @@
COMMENT = collection of puzzles games
DISTNAME = puzzles-20190415
+REVISION = 0
CATEGORIES = games
MASTER_SITES = https://rhaalovely.net/stuff/
Index: patches/patch-palisade_c
===
RCS file: patches/patch-palisade_c
diff -N patches/patch-palisade_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-palisade_c17 Aug 2019 21:54:21 -
@@ -0,0 +1,18 @@
+$OpenBSD$
+
+Fix for archs where char is unsigned by default (ppc, arm):
+error: result of comparison of constant -1 with expression of type
+'clue' (aka 'char')
+
+Index: palisade.c
+--- palisade.c.orig
palisade.c
+@@ -46,7 +46,7 @@ struct game_params {
+ int w, h, k;
+ };
+
+-typedef char clue;
++typedef signed char clue;
+ typedef unsigned char borderflag;
+
+ typedef struct shared_state {
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/17 16:17:15 Modified files: inputmethods/scim-pinyin: Makefile Added files: inputmethods/scim-pinyin/patches: patch-src_scim_phrase_cpp patch-src_scim_pinyin_cpp patch-src_scim_pinyin_phrase_cpp patch-src_scim_special_table_cpp Log message: scim-pinyin: add missing headers for ports-gcc. Regen WANTLIB and remove base-gcc from COMPILER while here. OK jca@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/17 16:15:03 ports/inputmethods/scim-pinyin/patches Update of /cvs/ports/inputmethods/scim-pinyin/patches In directory cvs.openbsd.org:/tmp/cvs-serv59882/patches Log Message: Directory /cvs/ports/inputmethods/scim-pinyin/patches added to the repository
[ports-gcc] Unbreak inputmethods/scim-tables
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/inputmethods/scim-tables.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/inputmethods/scim-tables.log
This is hopefully the last one.
With the below patch, it builds on macppc [0], and is still fine on
amd64.
OK?
Charlène.
[0] https://bin.charlenew.xyz/scim-tables.log
Index: Makefile
===
RCS file: /cvs/ports/inputmethods/scim-tables/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:47:14 - 1.26
+++ Makefile17 Aug 2019 20:16:31 -
@@ -3,7 +3,7 @@
COMMENT= input method data tables for scim
DISTNAME= scim-tables-0.5.8
-REVISION= 13
+REVISION= 14
CATEGORIES=inputmethods chinese japanese korean
@@ -12,16 +12,16 @@ HOMEPAGE= https://github.com/scim-im/sci
# GPLv2
PERMIT_PACKAGE=Yes
-WANTLIB+= GL X11 Xau Xcomposite Xcursor Xdamage Xdmcp Xext Xfixes Xi
-WANTLIB+= Xinerama Xrandr Xrender Xxf86vm atk-1.0 c cairo drm expat ffi
-WANTLIB+= fontconfig freetype gdk-x11-2.0 gdk_pixbuf-2.0 gio-2.0
-WANTLIB+= glib-2.0 gmodule-2.0 gobject-2.0 graphite2 gthread-2.0
-WANTLIB+= gtk-x11-2.0 harfbuzz iconv intl m pango-1.0 pangocairo-1.0
-WANTLIB+= pangoft2-1.0 pcre pixman-1 png pthread scim-1.0
-WANTLIB+= scim-gtkutils-1.0 scim-x11utils-1.0 ${COMPILER_LIBCXX} xcb xcb-render
-WANTLIB+= xcb-shm z fribidi
+WANTLIB += ${COMPILER_LIBCXX} X11 Xcomposite Xcursor Xdamage Xext
+WANTLIB += Xfixes Xi Xinerama Xrandr Xrender atk-1.0 c cairo expat
+WANTLIB += ffi fontconfig freetype fribidi gdk-x11-2.0 gdk_pixbuf-2.0
+WANTLIB += gio-2.0 glib-2.0 gmodule-2.0 gobject-2.0 graphite2
+WANTLIB += gthread-2.0 gtk-x11-2.0 harfbuzz iconv intl m pango-1.0
+WANTLIB += pangocairo-1.0 pangoft2-1.0 pcre pixman-1 png scim-1.0
+WANTLIB += scim-gtkutils-1.0 scim-x11utils-1.0 xcb xcb-render
+WANTLIB += xcb-shm z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER = base-clang ports-gcc
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=scim/}
Index: patches/patch-src_scim_generic_table_h
===
RCS file: patches/patch-src_scim_generic_table_h
diff -N patches/patch-src_scim_generic_table_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_generic_table_h 17 Aug 2019 20:16:31 -
@@ -0,0 +1,15 @@
+$OpenBSD$
+
+Add missing header for ports-gcc
+
+Index: src/scim_generic_table.h
+--- src/scim_generic_table.h.orig
src/scim_generic_table.h
+@@ -34,6 +34,7 @@
+ #include
+ #include
+ #include
++#include
+
+ #define SCIM_GT_MAX_KEY_LENGTH 63
+
[ports-gcc] Unbreak inputmethods/scim-pinyin
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/inputmethods/scim-pinyin.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/inputmethods/scim-pinyin.log
Once all headers are provided, it builds fine on macppc [0] and amd64 is
not broken.
OK?
Charlène.
[0] https://bin.charlenew.xyz/scim-pinyin.log
Index: Makefile
===
RCS file: /cvs/ports/inputmethods/scim-pinyin/Makefile,v
retrieving revision 1.24
diff -u -p -u -p -r1.24 Makefile
--- Makefile12 Jul 2019 20:47:14 - 1.24
+++ Makefile17 Aug 2019 19:45:00 -
@@ -3,7 +3,7 @@
COMMENT= scim input method module for pinyin
DISTNAME= scim-pinyin-0.5.91
-REVISION= 13
+REVISION= 14
CATEGORIES=inputmethods chinese
@@ -12,16 +12,16 @@ HOMEPAGE= https://github.com/scim-im/sci
# GPLv2
PERMIT_PACKAGE=Yes
-WANTLIB+= GL X11 Xau Xcomposite Xcursor Xdamage Xdmcp Xext Xfixes Xi
-WANTLIB+= Xinerama Xrandr Xrender Xxf86vm atk-1.0 cairo drm expat ffi
-WANTLIB+= fontconfig freetype gdk-x11-2.0 gdk-x11-2.0 gdk_pixbuf-2.0
-WANTLIB+= gdk_pixbuf-2.0 gio-2.0 glib-2.0 gmodule-2.0 gobject-2.0
-WANTLIB+= graphite2 gthread-2.0 gtk-x11-2.0 gtk-x11-2.0 harfbuzz iconv intl m
-WANTLIB+= pango-1.0 pangocairo-1.0 pangoft2-1.0 pcre pixman-1 png
-WANTLIB+= pthread scim-1.0 scim-gtkutils-1.0 fribidi
-WANTLIB+= scim-x11utils-1.0 ${COMPILER_LIBCXX} xcb xcb-render xcb-shm z
+WANTLIB += ${COMPILER_LIBCXX} X11 Xcomposite Xcursor Xdamage Xext
+WANTLIB += Xfixes Xi Xinerama Xrandr Xrender atk-1.0 cairo expat
+WANTLIB += ffi fontconfig freetype fribidi gdk-x11-2.0 gdk_pixbuf-2.0
+WANTLIB += gio-2.0 glib-2.0 gmodule-2.0 gobject-2.0 graphite2
+WANTLIB += gthread-2.0 gtk-x11-2.0 harfbuzz iconv intl m pango-1.0
+WANTLIB += pangocairo-1.0 pangoft2-1.0 pcre pixman-1 png scim-1.0
+WANTLIB += scim-gtkutils-1.0 scim-x11utils-1.0 xcb xcb-render
+WANTLIB += xcb-shm z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=scim/}
Index: patches/patch-src_scim_phrase_cpp
===
RCS file: patches/patch-src_scim_phrase_cpp
diff -N patches/patch-src_scim_phrase_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_phrase_cpp 17 Aug 2019 19:45:00 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Add missing header for ports-gcc
+
+Index: src/scim_phrase.cpp
+--- src/scim_phrase.cpp.orig
src/scim_phrase.cpp
+@@ -27,6 +27,8 @@
+
+ #define SCIM_PHRASE_MAX_RELATION 1000
+
++#include
++
+ #include
+ #include "scim_pinyin_private.h"
+ #include "scim_phrase.h"
Index: patches/patch-src_scim_pinyin_cpp
===
RCS file: patches/patch-src_scim_pinyin_cpp
diff -N patches/patch-src_scim_pinyin_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_pinyin_cpp 17 Aug 2019 19:45:00 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Add missing header for ports-gcc
+
+Index: src/scim_pinyin.cpp
+--- src/scim_pinyin.cpp.orig
src/scim_pinyin.cpp
+@@ -28,6 +28,8 @@
+ #define Uses_SCIM_CONFIG_PATH
+ #define Uses_SCIM_LOOKUP_TABLE
+
++#include
++
+ #include
+ #include "scim_pinyin.h"
+
Index: patches/patch-src_scim_pinyin_phrase_cpp
===
RCS file: patches/patch-src_scim_pinyin_phrase_cpp
diff -N patches/patch-src_scim_pinyin_phrase_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_pinyin_phrase_cpp17 Aug 2019 19:45:00 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Add missing header for ports-gcc
+
+Index: src/scim_pinyin_phrase.cpp
+--- src/scim_pinyin_phrase.cpp.orig
src/scim_pinyin_phrase.cpp
+@@ -26,6 +26,8 @@
+ #define Uses_SCIM_CONFIG_PATH
+ #define Uses_SCIM_LOOKUP_TABLE
+
++#include
++
+ #include
+ #include "scim_pinyin_private.h"
+ #include "scim_phrase.h"
Index: patches/patch-src_scim_special_table_cpp
===
RCS file: patches/patch-src_scim_special_table_cpp
diff -N patches/patch-src_scim_special_table_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_special_table_cpp17 Aug 2019 19:45:00 -
@@ -0,0 +1,15 @@
+$OpenBSD$
+
+Add missing header for ports-gcc
+
+Index: src/scim_special_table.cpp
+--- src/scim_special_table.cpp.orig
src/scim_special_table.cpp
+@@ -20,6 +20,7 @@
+ #define Uses_STL_IOMANIP
+ #define Uscs_C_STRING
+
++#include
+ #include
+ #include
+ #include "scim_pinyin_private.h"
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/17 13:08:38 Modified files: inputmethods/scim-hangul: Makefile inputmethods/scim-hangul/patches: patch-src_scim_hangul_imengine_cpp Log message: scim-hangul: add missing headers for ports-gcc. Regen WANTLIB and remove base-gcc from COMPILER while here. OK jca@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/17 13:07:21 Modified files: inputmethods/scim-fcitx: Makefile inputmethods/scim-fcitx/patches: patch-src_sp_cpp Added files: inputmethods/scim-fcitx/patches: patch-src_ime_cpp patch-src_scim_fcitx_imengine_cpp Log message: scim-fcitx: add missing headers for ports-gcc. Regen WANTLIB and remove base-gcc from COMPILER while here. OK jca@
[ports-gcc] Unbreak inputmethods/scim-hangul
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/inputmethods/scim-hangul.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/inputmethods/scim-hangul.log
Same thing as scim-fcitx.
Once all headers are provided, it builds fine on macppc [0] and amd64 is
not broken.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/scim-hangul.log
Index: Makefile
===
RCS file: /cvs/ports/inputmethods/scim-hangul/Makefile,v
retrieving revision 1.27
diff -u -p -u -p -r1.27 Makefile
--- Makefile12 Jul 2019 20:47:13 - 1.27
+++ Makefile17 Aug 2019 10:51:48 -
@@ -3,7 +3,7 @@
COMMENT= scim input method module for hangul
DISTNAME= scim-hangul-0.3.2
-REVISION= 14
+REVISION= 15
CATEGORIES=inputmethods korean
@@ -12,16 +12,16 @@ HOMEPAGE= https://sourceforge.net/projec
# GPLv2
PERMIT_PACKAGE=Yes
-WANTLIB+= GL X11 Xau Xcomposite Xcursor Xdamage Xdmcp Xext Xfixes Xi
-WANTLIB+= Xinerama Xrandr Xrender Xxf86vm atk-1.0 cairo drm expat ffi
-WANTLIB+= fontconfig freetype gdk-x11-2.0 gdk_pixbuf-2.0 gio-2.0
-WANTLIB+= glib-2.0 gmodule-2.0 gobject-2.0 graphite2 gthread-2.0
-WANTLIB+= gtk-x11-2.0 hangul harfbuzz iconv intl m pango-1.0 pangocairo-1.0
-WANTLIB+= pangoft2-1.0 pcre pixman-1 png pthread scim-1.0
-WANTLIB+= scim-gtkutils-1.0 scim-x11utils-1.0 ${COMPILER_LIBCXX} xcb xcb-render
-WANTLIB+= xcb-shm z fribidi
+WANTLIB += ${COMPILER_LIBCXX} X11 Xcomposite Xcursor Xdamage Xext
+WANTLIB += Xfixes Xi Xinerama Xrandr Xrender atk-1.0 cairo expat
+WANTLIB += ffi fontconfig freetype fribidi gdk-x11-2.0 gdk_pixbuf-2.0
+WANTLIB += gio-2.0 glib-2.0 gmodule-2.0 gobject-2.0 graphite2
+WANTLIB += gthread-2.0 gtk-x11-2.0 hangul harfbuzz iconv intl
+WANTLIB += m pango-1.0 pangocairo-1.0 pangoft2-1.0 pcre pixman-1
+WANTLIB += png scim-1.0 scim-gtkutils-1.0 scim-x11utils-1.0 xcb
+WANTLIB += xcb-render xcb-shm z
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=scim/}
Index: patches/patch-src_scim_hangul_imengine_cpp
===
RCS file:
/cvs/ports/inputmethods/scim-hangul/patches/patch-src_scim_hangul_imengine_cpp,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 patch-src_scim_hangul_imengine_cpp
--- patches/patch-src_scim_hangul_imengine_cpp 31 Oct 2008 01:42:00 -
1.1.1.1
+++ patches/patch-src_scim_hangul_imengine_cpp 17 Aug 2019 10:51:48 -
@@ -1,10 +1,15 @@
$OpenBSD: patch-src_scim_hangul_imengine_cpp,v 1.1.1.1 2008/10/31 01:42:00
kevlo Exp $
src/scim_hangul_imengine.cpp.orig Wed Oct 29 18:29:20 2008
-+++ src/scim_hangul_imengine.cpp Wed Oct 29 18:31:21 2008
-@@ -48,6 +48,8 @@
+
+Add missing headers fort ports-gcc
+
+Index: src/scim_hangul_imengine.cpp
+--- src/scim_hangul_imengine.cpp.orig
src/scim_hangul_imengine.cpp
+@@ -48,6 +48,9 @@
#define bind_textdomain_codeset(domain,codeset)
#endif
++#include
+#include
+
#define scim_module_init hangul_LTX_scim_module_init
[ports-gcc] Unbreak inputmethods/scim-fcitx
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/inputmethods/scim-fcitx.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/inputmethods/scim-fcitx.log
This is just a handful of missing headers.
Once provided, it builds on macppc [0] and amd64 is still fine indeed.
While here, i've refreshed WANTLIB and COMPILER.
Comments/feedback are welcome,
Charlène.
[0] https://bin.charlenew.xyz/scim-fcitx.log
Index: Makefile
===
RCS file: /cvs/ports/inputmethods/scim-fcitx/Makefile,v
retrieving revision 1.17
diff -u -p -u -p -r1.17 Makefile
--- Makefile12 Jul 2019 20:47:13 - 1.17
+++ Makefile17 Aug 2019 10:33:36 -
@@ -5,7 +5,7 @@ COMMENT=scim input method module for fc
V= 3.1.1
DISTNAME= scim-fcitx.$V
PKGNAME= scim-fcitx-$V
-REVISION= 7
+REVISION= 8
CATEGORIES=inputmethods chinese
@@ -14,9 +14,9 @@ HOMEPAGE= https://github.com/scim-im/sci
# GPLv2
PERMIT_PACKAGE=Yes
-WANTLIB= c iconv intl m ${COMPILER_LIBCXX} scim-1.0 pthread
+WANTLIB += ${COMPILER_LIBCXX} c iconv intl m scim-1.0
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=scim/}
EXTRACT_SUFX= .tar.bz2
Index: patches/patch-src_ime_cpp
===
RCS file: patches/patch-src_ime_cpp
diff -N patches/patch-src_ime_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_ime_cpp 17 Aug 2019 10:33:36 -
@@ -0,0 +1,15 @@
+$OpenBSD$
+
+ports-gcc: add missing header
+
+Index: src/ime.cpp
+--- src/ime.cpp.orig
src/ime.cpp
+@@ -3,6 +3,7 @@
+ #define Uses_SCIM_CONFIG_BASE
+ #define Uses_SCIM_CONFIG_PATH
+ #include
++#include
+
+ #include
+ #include "scim_fcitx_imengine.h"
Index: patches/patch-src_scim_fcitx_imengine_cpp
===
RCS file: patches/patch-src_scim_fcitx_imengine_cpp
diff -N patches/patch-src_scim_fcitx_imengine_cpp
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_scim_fcitx_imengine_cpp 17 Aug 2019 10:33:36 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+ports-gcc: add missing header
+
+Index: src/scim_fcitx_imengine.cpp
+--- src/scim_fcitx_imengine.cpp.orig
src/scim_fcitx_imengine.cpp
+@@ -35,6 +35,8 @@
+ #define Uses_SCIM_CONFIG_BASE
+ #define Uses_SCIM_CONFIG_PATH
+
++#include
++
+ #include
+ #include "scim_fcitx_imengine.h"
+ #include "main.h"
Index: patches/patch-src_sp_cpp
===
RCS file: /cvs/ports/inputmethods/scim-fcitx/patches/patch-src_sp_cpp,v
retrieving revision 1.1.1.1
diff -u -p -u -p -r1.1.1.1 patch-src_sp_cpp
--- patches/patch-src_sp_cpp29 Oct 2008 02:24:17 - 1.1.1.1
+++ patches/patch-src_sp_cpp17 Aug 2019 10:33:36 -
@@ -1,7 +1,15 @@
$OpenBSD: patch-src_sp_cpp,v 1.1.1.1 2008/10/29 02:24:17 kevlo Exp $
src/sp.cpp.origSat Oct 25 15:36:52 2008
-+++ src/sp.cpp Sat Oct 25 15:37:30 2008
-@@ -11,6 +11,7 @@ using namespace scim;
+
+ports-gcc: add missing headers
+
+Index: src/sp.cpp
+--- src/sp.cpp.orig
src/sp.cpp
+@@ -8,9 +8,11 @@
+ using namespace scim;
+
+ #include
++#include
#include
#include
#include
Re: portcheck: whine with default COMPILER and C++ libs
I was surprised it has not been committed already, it works great, and
may hopefully reduce the number of build failures on base-gcc archs in
the future.
OK cwen@
On Mon, 12 Aug 2019 11:13:35 -0400
Kurt Mosiejczuk wrote:
> On Mon, Aug 12, 2019 at 02:23:55PM +0100, Stuart Henderson wrote:
> > any comments? ok?
>
> Looks like a good check to add. OK kmos
>
> --Kurt
>
> > Index: mk/bsd.port.mk
> > ===
> > RCS file: /cvs/ports/infrastructure/mk/bsd.port.mk,v
> > retrieving revision 1.1475
> > diff -u -p -r1.1475 bsd.port.mk
> > --- mk/bsd.port.mk 14 Jul 2019 11:32:05 - 1.1475
> > +++ mk/bsd.port.mk 12 Aug 2019 13:23:06 -
> > @@ -304,7 +304,7 @@ _MODULES_DONE =
> > .endif
>
> > # this only happens if we exit modules without having ever gone
> > -# thru compiler.port.mk
> > +# thru compiler.port.mk - update portcheck if changing COMPILER
> > default COMPILER ?= base-clang base-gcc gcc3
> > COMPILER_LANGS ?= c c++
> > .if ${PROPERTIES:Mclang}
> > Index: bin/portcheck
> > ===
> > RCS file: /cvs/ports/infrastructure/bin/portcheck,v
> > retrieving revision 1.128
> > diff -u -p -r1.128 portcheck
> > --- bin/portcheck 23 Jun 2019 16:06:23 - 1.128
> > +++ bin/portcheck 12 Aug 2019 13:23:06 -
> > @@ -844,10 +844,12 @@ sub_checks() {
> > vars="$vars MODULES"
> > vars="$vars PKG_ARCH$subpkg $wantlib_var
> > WANTLIB-" vars="$vars PERMIT_PACKAGE${subpkg%-}"
> > + vars="$vars COMPILER"
> > make "${make_args[@]}" show="$vars" | {
> > local comment fullpkgname modules
> > pkg_arch local wantlib permit_package
> > local arch_independent=false
> > + local default_compiler=false
> > read -r comment
> > read -r fullpkgname
> > read -r modules
> > @@ -856,6 +858,7 @@ sub_checks() {
> > read -r wantlib
> > read -r wantlib_ss
> > read -r permit_package
> > + read -r compiler
> >
> > if [[ $comment == @(a|an|the)"
> > "* ]]; then err "${portref}no leading articles in" \
> > @@ -869,7 +872,12 @@ sub_checks() {
> > fi
> > fi
> >
> > - check_wantlib "$portref"
> > "$modules" $wantlib
> > + [[ "$compiler" == "base-clang
> > base-gcc gcc3" ]] \
> > + && default_compiler=true
> > +
> > + check_wantlib "$portref"
> > "$modules" \
> > + "$default_compiler" $wantlib
> > +
> > check_permit_subpkg "$portref"
> > "$subpkg" \ "$permit_package"
> >
> > @@ -1301,6 +1309,7 @@ check_plist() {
> > check_wantlib() {
> > local portref="$1"; shift
> > local modules="$1"; shift
> > + local default_compiler="$1"; shift
> >
> > local phonon_s_wantlib=false
> >
> > @@ -1345,6 +1354,15 @@ check_wantlib() {
> > "in MODULES (check other libs,
> > too!)" fi
> > ;;
> > +
> > + @(c++|stdc++)?(?('>')=+([0-9])))
> > + if $default_compiler; then
> > + err "C++ libraries in WANTLIB with
> > default COMPILER" \
> > + "(most ports need
> > 'COMPILER=base-clang ports-gcc'" \
> > + "or 'COMPILER=base-clang
> > ports-gcc base-gcc')"
> > + fi
> > + ;;
> > +
> > stdc++?(?('>')=+([0-9])))
> > if $gcc4_module; then
> > err "$portref$v in WANTLIB when
> > gcc4 is" \
> >
>
[ports-gcc] Unbreak graphics/dibuja
Hi,
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/graphics/dibuja.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/graphics/dibuja.log
Not much to say here. dibuja has:
WANTLIB += ${COMPILER_LIBCXX}
It has also been spotted by sthen@'s proposal [0] to add a COMPILER
check in portcheck(1) for ports having c++ libs but using default
compilers.
As such, ports-gcc should be used. This fixes the build on macppc [1],
the runtime is fine as well. I'm bumping REVISION, dibuja has been
already built on base-gcc archs.
OK?
Charlène.
[0] https://marc.info/?l=openbsd-ports=156561626006156=2
[1] https://bin.charlenew.xyz/dibuja.log
Index: Makefile
===
RCS file: /cvs/ports/graphics/dibuja/Makefile,v
retrieving revision 1.4
diff -u -p -u -p -r1.4 Makefile
--- Makefile12 Jul 2019 20:46:57 - 1.4
+++ Makefile17 Aug 2019 09:18:16 -
@@ -3,7 +3,7 @@
V =0.10.0
COMMENT = simple to use paint program
DISTNAME = dibuja-${V}
-REVISION = 0
+REVISION = 1
CATEGORIES = graphics x11
HOMEPAGE = https://launchpad.net/dibuja
@@ -22,6 +22,8 @@ WANTLIB += pangoft2-1.0 pcre pixman-1 pn
WANTLIB += z
MASTER_SITES = https://launchpad.net/dibuja/trunk/${V}/+download/
+
+COMPILER = base-clang ports-gcc
MODULES = textproc/intltool
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/16 17:27:09 Modified files: devel/angr/py-z3-solver: Makefile Added files: devel/angr/py-z3-solver/patches: patch-core_src_util_lp_permutation_matrix_h Log message: py-z3-solver: unbreak with ports-gcc, by fixing an incorrect conversion. Tested on sparc64 (by kmos@, thanks!) and macppc. OK kmos@ kn@ (maintainer)
[macppc, hppa?] Unbreak games/godot
> https://marc.info/?l=openbsd-ports=15654325765=2
As promised, here is the conversion from __sync_* to __atomic_*
functions, allowing godot to build on macppc at least, and maybe
hppa, on top of the ports-gcc/sparc64 fixes, that Thomas committed a
few hours ago.
The patch was originally written for godot-3.1.1, and has been
upstreamed [0]. I backported it for godot-3.0.6, it builds
successfully on amd64 and macppc.
Due to relocations errors, i had to add address relaxing and
long calls.
Runtime cannot be tested with my video card (radeon 9700) with 3.0.6,
because it requires OpenGL(ES) 3, even for opening the editor, unlike
godot-3.1.1.
Comments/feedback are welcome,
Charlène.
[0] https://github.com/godotengine/godot/pull/31321
Index: Makefile
===
RCS file: /cvs/ports/games/godot/Makefile,v
retrieving revision 1.8
diff -u -p -u -p -r1.8 Makefile
--- Makefile16 Aug 2019 15:38:15 - 1.8
+++ Makefile16 Aug 2019 22:02:03 -
@@ -8,7 +8,7 @@ PKGNAME = godot-${V}
CATEGORIES = games
HOMEPAGE = https://godotengine.org/
MAINTAINER = Thomas Frohwein
-REVISION = 2
+REVISION = 3
# MIT
PERMIT_PACKAGE = Yes
@@ -68,6 +68,18 @@ LIB_DEPENDS =archivers/zstd \
net/enet
NO_TEST = Yes
+
+.if ${MACHINE_ARCH:Mhppa} || ${MACHINE_ARCH:Mpowerpc}
+LDFLAGS += -latomic
+WANTLIB += atomic
+.endif
+
+# Fix relocation overflows
+.if ${MACHINE_ARCH:Mpowerpc}
+CFLAGS += -mlongcall
+CXXFLAGS +=-mlongcall
+LDFLAGS += -Wl,--relax
+.endif
pre-configure:
${SUBST_CMD} ${WRKSRC}/drivers/unix/os_unix.cpp
Index: patches/patch-core_safe_refcount_h
===
RCS file: patches/patch-core_safe_refcount_h
diff -N patches/patch-core_safe_refcount_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-core_safe_refcount_h 16 Aug 2019 22:02:03 -
@@ -0,0 +1,68 @@
+$OpenBSD$
+
+hppa, ppc: use __atomic functions as 64-bit __sync operators
+are not supported, from:
+https://github.com/godotengine/godot/pull/31321
+
+Index: core/safe_refcount.h
+--- core/safe_refcount.h.orig
core/safe_refcount.h
+@@ -99,8 +99,8 @@ static _ALWAYS_INLINE_ T atomic_exchange_if_greater(re
+
+ /* Implementation for GCC & Clang */
+
+-// GCC guarantees atomic intrinsics for sizes of 1, 2, 4 and 8 bytes.
+-// Clang states it supports GCC atomic builtins.
++#include
++#include
+
+ template
+ static _ALWAYS_INLINE_ T atomic_conditional_increment(register T *pw) {
+@@ -109,7 +109,7 @@ static _ALWAYS_INLINE_ T atomic_conditional_increment(
+ T tmp = static_cast(*pw);
+ if (tmp == 0)
+ return 0; // if zero, can't add to it anymore
+- if (__sync_val_compare_and_swap(pw, tmp, tmp + 1) == tmp)
++ if (__atomic_compare_exchange_n(pw, , tmp + 1, false,
__ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) == true)
+ return tmp + 1;
+ }
+ }
+@@ -117,25 +117,25 @@ static _ALWAYS_INLINE_ T atomic_conditional_increment(
+ template
+ static _ALWAYS_INLINE_ T atomic_decrement(register T *pw) {
+
+- return __sync_sub_and_fetch(pw, 1);
++ return __atomic_sub_fetch(pw, 1, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_increment(register T *pw) {
+
+- return __sync_add_and_fetch(pw, 1);
++ return __atomic_add_fetch(pw, 1, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_sub(register T *pw, register V val) {
+
+- return __sync_sub_and_fetch(pw, val);
++ return __atomic_sub_fetch(pw, val, __ATOMIC_SEQ_CST);
+ }
+
+ template
+ static _ALWAYS_INLINE_ T atomic_add(register T *pw, register V val) {
+
+- return __sync_add_and_fetch(pw, val);
++ return __atomic_add_fetch(pw, val, __ATOMIC_SEQ_CST);
+ }
+
+ template
+@@ -145,7 +145,7 @@ static _ALWAYS_INLINE_ T atomic_exchange_if_greater(re
+ T tmp = static_cast(*pw);
+ if (tmp >= val)
+ return tmp; // already greater, or equal
+- if (__sync_val_compare_and_swap(pw, tmp, val) == tmp)
++ if (__atomic_compare_exchange_n(pw, , val, false,
__ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) == true)
+ return val;
+ }
+ }
[ports-gcc] Unbreak devel/angr/py-z3-solver
Hi,
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/devel/angr/py-z3-solver.log
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/devel/angr/py-z3-solver.log
I've found a fix for this from the "standard" z3 repo [0].
As expected, it builds on macppc [1] and amd64.
OK?
Charlène.
[0] https://github.com/Z3Prover/z3/pull/1612
[1] https://bin.charlenew.xyz/py-z3-solver.log
Index: Makefile
===
RCS file: /cvs/ports/devel/angr/py-z3-solver/Makefile,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 Makefile
--- Makefile23 Jul 2019 19:50:03 - 1.2
+++ Makefile16 Aug 2019 16:04:13 -
@@ -5,6 +5,7 @@ COMMENT = efficient SMT solver library
ANGR_PYTHON_MODULE = z3-solver
# devel/angr/py-claripy requires this exact version, newer 4.8.5.0 breaks it.
MODPY_EGG_VERSION =4.5.1.0.post2
+REVISION = 0
CATEGORIES += math
Index: patches/patch-core_src_util_lp_permutation_matrix_h
===
RCS file: patches/patch-core_src_util_lp_permutation_matrix_h
diff -N patches/patch-core_src_util_lp_permutation_matrix_h
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-core_src_util_lp_permutation_matrix_h 16 Aug 2019 16:04:13
-
@@ -0,0 +1,17 @@
+$OpenBSD$
+
+Unbreak with ports-gcc, from:
+https://github.com/Z3Prover/z3/commit/2d5dd802386d78117d5ed9ddcbf8bc22ab3cb461
+
+Index: core/src/util/lp/permutation_matrix.h
+--- core/src/util/lp/permutation_matrix.h.orig
core/src/util/lp/permutation_matrix.h
+@@ -117,7 +117,7 @@ class permutation_matrix : public tail_matrix {
+
+ unsigned size() const { return static_cast(m_rev.size()); }
+
+-unsigned * values() const { return m_permutation; }
++unsigned * values() const { return m_permutation.c_ptr(); }
+
+ void resize(unsigned size) {
+ unsigned old_size = m_permutation.size();
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/15 12:22:34 Added files: audio/flac/patches: patch-src_libFLAC_cpu_c Log message: flac: unbreak on macppc, by removing some use of getauxval(3), that we don't implement. No revision bump because flac-1.3.3 never built on macppc. OK jca@ naddy@ (maintainer)
[ports-gcc] Unbreak x11/waimea
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/x11/waimea.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-12/x11/waimea.log
There is a clash between std::list and the "list" class member.
I've just renamed that member and its instances. It builds [0] and works
fine on macppc, and amd64 is not broken.
While here WANTLIB needed a refresh, and i moved HOMEPAGE to https.
OK?
Charlène.
[0] https://bin.charlenew.xyz/waimea.log
Index: Makefile
===
RCS file: /cvs/ports/x11/waimea/Makefile,v
retrieving revision 1.26
diff -u -p -u -p -r1.26 Makefile
--- Makefile12 Jul 2019 20:51:23 - 1.26
+++ Makefile15 Aug 2019 09:56:07 -
@@ -4,21 +4,20 @@ COMMENT= virtual desktop manager for X1
VERSION= 0.4.0
DISTNAME= waimea-${VERSION}
-REVISION = 9
+REVISION= 10
CATEGORIES=x11
-HOMEPAGE= http://www.freedesktop.org/wiki/Software/waimea
+HOMEPAGE= https://www.freedesktop.org/wiki/Software/waimea
# GPLv2
PERMIT_PACKAGE=Yes
-MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
+WANTLIB += ${COMPILER_LIBCXX} ICE SM X11 Xext Xft Xinerama Xrandr
+WANTLIB += Xrender c m
-WANTLIB += ICE SM X11 Xau Xdmcp Xext Xft Xinerama Xrandr Xrender
-WANTLIB += c expat fontconfig freetype m ${COMPILER_LIBCXX}
-WANTLIB += xcb z
+MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=waimea/}
-COMPILER = base-clang ports-gcc base-gcc
+COMPILER= base-clang ports-gcc
AUTOCONF_VERSION= 2.52
CONFIGURE_STYLE= autoconf
Index: patches/patch-src_Resources_cc
===
RCS file: patches/patch-src_Resources_cc
diff -N patches/patch-src_Resources_cc
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_cc 15 Aug 2019 09:56:07 -
@@ -0,0 +1,35 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.cc
+--- src/Resources.cc.orig
src/Resources.cc
+@@ -1568,7 +1568,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 'n' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1583,7 +1583,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+NULL);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (str[0] == 't' && str[1] == '/') {
+ for (i3 = 2; str[i3] != '\0' &&
+@@ -1598,7 +1598,7 @@ void ResourceHandler::LoadActions(WaScreen *wascreen)
+str + 2);
+ str = str + i3 + 1;
+ ReadActions((char *) buffer2, defs, ,
+-_list->list, wascreen);
++_list->alist, wascreen);
+ }
+ else if (! strncasecmp(str, "window", 6)) {
+ str = str + 6;
Index: patches/patch-src_Resources_hh
===
RCS file: patches/patch-src_Resources_hh
diff -N patches/patch-src_Resources_hh
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Resources_hh 15 Aug 2019 09:56:07 -
@@ -0,0 +1,24 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning of 'list' from 'class std::__cxx11::list<_WaAction*>'
+
+Index: src/Resources.hh
+--- src/Resources.hh.orig
src/Resources.hh
+@@ -193,13 +193,13 @@ class WaActionExtList { (public)
+ delete name;
+ delete cl;
+ delete title;
+-ACTLISTCLEAR(list);
++ACTLISTCLEAR(alist);
+ }
+
+ Regex *name;
+ Regex *cl;
+ Regex *title;
+-list list;
++list alist;
+ };
+
+ class StrComp {
Index: patches/patch-src_Window_cc
===
RCS file: patches/patch-src_Window_cc
diff -N patches/patch-src_Window_cc
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_Window_cc 15 Aug 2019 09:56:07 -
@@ -0,0 +1,25 @@
+$OpenBSD$
+
+ports-gcc fix for:
+error: changes meaning
[macppc] Unbreak audio/flac 1.3.3
Hi,
That newer version of flac makes use of getauxval(3) on powerpc, that
apparently we don't implement, so the builds fails while not finding
sys/auxv.h.
The below diff allows to build flac on macppc, where tests are passing
[1]. I've not bumped revision, the change impacts powerpc where it has
never been built.
Comments/feedback are welcome,
Charlène.
[0] http://man7.org/linux/man-pages/man3/getauxval.3.html
[1] https://bin.charlenew.xyz/flac.log
Index: patches/patch-src_libFLAC_cpu_c
===
RCS file: patches/patch-src_libFLAC_cpu_c
diff -N patches/patch-src_libFLAC_cpu_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_libFLAC_cpu_c 15 Aug 2019 09:43:18 -
@@ -0,0 +1,43 @@
+$OpenBSD$
+
+Remove getauxval(3) code for ppc because we don't implement this
+
+Index: src/libFLAC/cpu.c
+--- src/libFLAC/cpu.c.orig
src/libFLAC/cpu.c
+@@ -53,10 +53,6 @@
+ #define dfprintf(file, format, ...)
+ #endif
+
+-#if defined FLAC__CPU_PPC
+-#include
+-#endif
+-
+ #if (defined FLAC__CPU_IA32 || defined FLAC__CPU_X86_64) && (defined
FLAC__HAS_NASM || FLAC__HAS_X86INTRIN) && !defined FLAC__NO_ASM
+
+ /* these are flags in EDX of CPUID AX=0001 */
+@@ -236,24 +232,8 @@ x86_cpu_info (FLAC__CPUInfo *info)
+ static void
+ ppc_cpu_info (FLAC__CPUInfo *info)
+ {
+-#if defined FLAC__CPU_PPC
+-#ifndef PPC_FEATURE2_ARCH_3_00
+-#define PPC_FEATURE2_ARCH_3_000x0080
+-#endif
+-
+-#ifndef PPC_FEATURE2_ARCH_2_07
+-#define PPC_FEATURE2_ARCH_2_070x8000
+-#endif
+-
+- if (getauxval(AT_HWCAP2) & PPC_FEATURE2_ARCH_3_00) {
+- info->ppc.arch_3_00 = true;
+- } else if (getauxval(AT_HWCAP2) & PPC_FEATURE2_ARCH_2_07) {
+- info->ppc.arch_2_07 = true;
+- }
+-#else
+ info->ppc.arch_2_07 = false;
+ info->ppc.arch_3_00 = false;
+-#endif
+ }
+
+ void FLAC__cpu_info (FLAC__CPUInfo *info)
[update] x11/ctwm 4.0.2 -> 4.0.3
Here is a very simple update for ctwm. This only contains bugfixes [0]. It builds and works fine on amd64 and macppc. Comments/feedback are welcome, Charlène. [0] https://www.ctwm.org/CHANGES.html Index: Makefile === RCS file: /cvs/ports/x11/ctwm/Makefile,v retrieving revision 1.28 diff -u -p -u -p -r1.28 Makefile --- Makefile12 Jul 2019 20:51:09 - 1.28 +++ Makefile14 Aug 2019 15:45:38 - @@ -2,7 +2,7 @@ COMMENT= twm, plus support for multiple virtual screens, etc -DISTNAME= ctwm-4.0.2 +DISTNAME= ctwm-4.0.3 CATEGORIES=x11 Index: distinfo === RCS file: /cvs/ports/x11/ctwm/distinfo,v retrieving revision 1.7 diff -u -p -u -p -r1.7 distinfo --- distinfo29 Aug 2018 21:06:43 - 1.7 +++ distinfo14 Aug 2019 15:45:38 - @@ -1,2 +1,2 @@ -SHA256 (ctwm-4.0.2.tar.gz) = U3Slep45FuFOm4qrt+BsZYt9gu2FrozUZEqWJoXIZNU= -SIZE (ctwm-4.0.2.tar.gz) = 915254 +SHA256 (ctwm-4.0.3.tar.gz) = 3ezUUw/R76LRnVlNmAPm6fFXiDlviRwxldWtIyBS4Bc= +SIZE (ctwm-4.0.3.tar.gz) = 934415
[update] audio/chromaprint 1.1 -> 1.4.3
Hi,
Here is an update for chromaprint that fixes build issues on ppc/arm.
Upstream didn't pin a 1.4.4 release [1], but this is what you get with
that diff.
What's new upstream (see [0]):
- move to github
- C++11 is required, chromaprint doesn't use boost anymore
- use newer ffmpeg api
What's new in the port:
- reorder the Makefile according to Makefile.template. It doesn't
impact a lot the diff readability given the amount of changes.
- major bump SHARED_LIBS. check_sym reports deletions.
- s/sed/SUBST_CMD
- remove hardcoded path when finding gtest, it also required to add
linking instructions for cmake to allow the use of devel/gtest
- add "1.4.4" fixes that make test endian neutral
Testing (macppc/amd64):
- 'make test' passes with all tests enabled
- All consumers build and run without issue. Clementine auto-tagging
doesn't work with chromaprint-1.1 as well.
Comments/feedback are welcome :)
Charlène.
[0]
https://github.com/acoustid/chromaprint/blob/master/NEWS.txt
[1] https://github.com/acoustid/chromaprint/issues/59
Index: chromaprint/Makefile
===
RCS file: /cvs/ports/audio/chromaprint/Makefile,v
retrieving revision 1.12
diff -u -p -r1.12 Makefile
--- chromaprint/Makefile12 Jul 2019 20:43:32 - 1.12
+++ chromaprint/Makefile13 Aug 2019 21:18:50 -
@@ -2,37 +2,39 @@
COMMENT = audio fingerprint extraction library
-DISTNAME = chromaprint-1.1
+GH_ACCOUNT = acoustid
+GH_PROJECT = chromaprint
+GH_TAGNAME = v1.4.3
+
+SHARED_LIBS = chromaprint 2.0 # 1.4.3
+
CATEGORIES = audio devel
-HOMEPAGE = https://acoustid.org/chromaprint
-MASTER_SITES = https://bitbucket.org/acoustid/chromaprint/downloads/
-REVISION = 3
-SHARED_LIBS = chromaprint 1.0 # 0.1
+HOMEPAGE = https://acoustid.org/chromaprint
-# LGPL2.1+
+# MIT with LGPL2.1+ parts
PERMIT_PACKAGE = Yes
-WANTLIB += avcodec avformat avutil c m pthread ${COMPILER_LIBCXX} swresample
+WANTLIB += ${COMPILER_LIBCXX} avcodec avformat avutil c m swresample
-COMPILER = base-clang ports-gcc base-gcc
+# C++11
+COMPILER = base-clang ports-gcc
MODULES = devel/cmake
-BUILD_DEPENDS =audio/taglib \
- devel/boost
+BUILD_DEPENDS =audio/taglib
+# gtest presence is checked in configure stage, so this cannot be
+# in TEST_DEPENDS
+BUILD_DEPENDS += devel/gtest
LIB_DEPENDS = graphics/ffmpeg
-# gtest presence is checked in configure stage, so this cannot be in
TEST_DEPENDS
-BUILD_DEPENDS += devel/gtest
-CONFIGURE_ARGS = -DBUILD_TESTS:Bool=Yes -DBUILD_EXAMPLES=ON
+CONFIGURE_ARGS = -DBUILD_TOOLS=ON \
+ -DBUILD_TESTS=ON
-post-patch:
- sed -i 's,/usr/local,${LOCALBASE},' \
- ${WRKSRC}/cmake/modules/FindGTest.cmake
+TEST_TARGET = check
-do-test:
- cd ${WRKBUILD}/tests && ./all_tests
+pre-configure:
+ ${SUBST_CMD} ${WRKSRC}/cmake/modules/FindGTest.cmake
.include
Index: chromaprint/distinfo
===
RCS file: /cvs/ports/audio/chromaprint/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- chromaprint/distinfo9 May 2014 09:00:19 - 1.2
+++ chromaprint/distinfo13 Aug 2019 21:18:50 -
@@ -1,2 +1,2 @@
-SHA256 (chromaprint-1.1.tar.gz) = axTX6klkWBtzvT+AOMiFfAHkRkIcGumcu/ZN4mtHzRI=
-SIZE (chromaprint-1.1.tar.gz) = 542360
+SHA256 (chromaprint-1.4.3.tar.gz) =
1K5llig6rXoBWlsERQEgVMY0pLkynssjAAzTVLQKKDs=
+SIZE (chromaprint-1.4.3.tar.gz) = 613718
Index: chromaprint/patches/patch-cmake_modules_FindGTest_cmake
===
RCS file:
/cvs/ports/audio/chromaprint/patches/patch-cmake_modules_FindGTest_cmake,v
retrieving revision 1.1
diff -u -p -r1.1 patch-cmake_modules_FindGTest_cmake
--- chromaprint/patches/patch-cmake_modules_FindGTest_cmake 9 May 2014
09:00:19 - 1.1
+++ chromaprint/patches/patch-cmake_modules_FindGTest_cmake 13 Aug 2019
21:18:50 -
@@ -1,6 +1,10 @@
$OpenBSD: patch-cmake_modules_FindGTest_cmake,v 1.1 2014/05/09 09:00:19 sthen
Exp $
cmake/modules/FindGTest.cmake.orig Sat Nov 23 16:43:42 2013
-+++ cmake/modules/FindGTest.cmake Wed May 7 16:29:15 2014
+
+Find gtest from devel/gtest.
+
+Index: cmake/modules/FindGTest.cmake
+--- cmake/modules/FindGTest.cmake.orig
cmake/modules/FindGTest.cmake
@@ -71,12 +71,24 @@ find_path(GTEST_INCLUDE_DIR
)
mark_as_advanced(GTEST_INCLUDE_DIR)
@@ -10,7 +14,7 @@ $OpenBSD: patch-cmake_modules_FindGTest_
+FIND_LIBRARY(${varname}
+NAMES ${shortname}
+PATHS
-+/usr/local/lib
++ ${LOCALBASE}/lib
+ NO_DEFAULT_PATH
+DOC "Location
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/13 07:36:00 Modified files: cad/gnucap : Makefile Added files: cad/gnucap/patches: patch-src_plot_cc Log message: gnucap: fix the build with ports-gcc OK bentley@
Re: audio/chromaprint: fix compilation on unsigned char arches
Hi,
On Tue, 13 Aug 2019 08:14:24 +0200
Peter Hessler wrote:
> test_bit_string_* both use constructs like:
>
> char data[] = { -28 };
>
> which won't work on arches where char is unsigned, such as aarch64.
>
> This unblocks a lot of packages. No bump because none of the test
> files are ran, only compiled.
>
> OK?
So that's why chromaprint vanished from the aarch64 build failures (:
It works as well on macppc.
If people have objections, properly fixing the code was non trivial
(at least to me) and updating to 1.4.3 was simpler. I've an update
almost ready for submission [0], that fixes the build and can
run all the tests.
It took some time to get OKs for a needed audio/clementine fix, so
i prefer to test the update again on macppc (it's fine on amd64),
before submitting it here. That could take a few more days if it goes
bad.
Meanwhile, OK cwen@
[0]
https://github.com/jasperla/openbsd-wip/tree/master/audio/chromaprint
>
> Index: audio/chromaprint/patches/patch-tests_CMakeLists_txt
> ===
> RCS
> file: /cvs/ports/audio/chromaprint/patches/patch-tests_CMakeLists_txt,v
> retrieving revision 1.1 diff -u -p -u -p -r1.1
> patch-tests_CMakeLists_txt
> --- audio/chromaprint/patches/patch-tests_CMakeLists_txt 9 May
> 2014 09:00:19 - 1.1 +++
> audio/chromaprint/patches/patch-tests_CMakeLists_txt 12 Aug
> 2019 22:41:33 - @@ -1,6 +1,10 @@ $OpenBSD:
> patch-tests_CMakeLists_txt,v 1.1 2014/05/09 09:00:19 sthen Exp $
> tests/CMakeLists.txt.origSat Nov 23 16:43:42 2013
> -+++ tests/CMakeLists.txt Wed May 7 15:12:15 2014
> +
> +test-bit_string_*.cpp depends on signed char, failing on aarch64
> +
> +Index: tests/CMakeLists.txt
> +--- tests/CMakeLists.txt.orig
> tests/CMakeLists.txt
> @@ -3,7 +3,6 @@ include_directories(
> ${Boost_INCLUDE_DIRS}
> ${CMAKE_CURRENT_SOURCE_DIR}/../src
> @@ -9,3 +13,12 @@ $OpenBSD: patch-tests_CMakeLists_txt,v 1
>
> set(tests_SOURCES
> main.cpp
> +@@ -15,8 +14,6 @@ set(tests_SOURCES
> + test_integral_image.cpp
> + test_lloyds.cpp
> + test_audio_processor.cpp
> +-test_bit_string_reader.cpp
> +-test_bit_string_writer.cpp
> + test_chromaprint.cpp
> + test_chroma.cpp
> + test_chroma_filter.cpp
>
>
>
> --
> The probability of someone watching you is proportional to the
> stupidity of your action.
>
[macppc/ports-gcc] Unbreak benchmarks/wrk
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/benchmarks/wrk.log
It requires a move from __sync to __atomic. Debian did that already [0],
so i've taken their stuff.
While wrk is written in C, the use of __atomic functions requires
estdc++ and ports-gcc (we can't use COMPILER_LANGS=c).
It builds [1] and runs fine on macppc and amd64 with multiple threads.
Comments/feedback are welcome,
Charlène.
[0] https://sources.debian.org/patches/wrk/4.0.2-2/debian-changes/
[1] https://bin.charlenew.xyz/wrk.log
Index: Makefile
===
RCS file: /cvs/ports/benchmarks/wrk/Makefile,v
retrieving revision 1.2
diff -u -p -u -p -r1.2 Makefile
--- Makefile12 Jul 2019 20:43:43 - 1.2
+++ Makefile12 Aug 2019 19:40:01 -
@@ -5,6 +5,7 @@ COMMENT = modern HTTP benchmarking tool
GH_ACCOUNT = wg
GH_PROJECT = wrk
GH_TAGNAME = 4.1.0
+REVISION = 0
CATEGORIES = benchmarks
@@ -15,6 +16,9 @@ PERMIT_PACKAGE = Yes
WANTLIB += c crypto luajit-5.1 m pthread ssl
+# __atomic support (GCC>4.7.4)
+COMPILER = base-clang ports-gcc
+
LIB_DEPENDS = lang/luajit
USE_GMAKE =Yes
@@ -26,6 +30,12 @@ MAKE_ENV = WITH_LUAJIT=${LOCALBASE} WITH
MAKE_ENV +=LIBS=-lc++abi
WANTLIB += c++abi
.endif
+
+.if ${MACHINE_ARCH:Mhppa} || ${MACHINE_ARCH:Mpowerpc}
+MAKE_ENV +=LIBS=-latomic
+WANTLIB += atomic
+.endif
+
NO_TEST = Yes
Index: patches/patch-src_stats_c
===
RCS file: patches/patch-src_stats_c
diff -N patches/patch-src_stats_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-src_stats_c 12 Aug 2019 19:40:01 -
@@ -0,0 +1,33 @@
+$OpenBSD$
+
+ppc, hppa: we can't use 64-bit __sync operators there
+
+Index: src/stats.c
+--- src/stats.c.orig
src/stats.c
+@@ -21,12 +21,21 @@ void stats_free(stats *stats) {
+
+ int stats_record(stats *stats, uint64_t n) {
+ if (n >= stats->limit) return 0;
+-__sync_fetch_and_add(>data[n], 1);
+-__sync_fetch_and_add(>count, 1);
++__atomic_fetch_add(>data[n], 1, __ATOMIC_SEQ_CST);
++__atomic_fetch_add(>count, 1, __ATOMIC_SEQ_CST);
+ uint64_t min = stats->min;
+ uint64_t max = stats->max;
+-while (n < min) min = __sync_val_compare_and_swap(>min, min, n);
+-while (n > max) max = __sync_val_compare_and_swap(>max, max, n);
++while (n < min) {
++__atomic_compare_exchange(>min, , , false,
++ __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST);
++min = stats->min;
++}
++while (n > max) {
++__atomic_compare_exchange(>max, , , false,
++ __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST);
++max = stats->max;
++ }
++
+ return 1;
+ }
+
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/12 11:47:52 Modified files: emulators/xnp2 : Makefile Added files: emulators/xnp2/patches: patch-x11_compiler_h Log message: xnp2: fix the build with ports-gcc OK bentley@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/12 08:40:04 Modified files: security/aircrack-ng: Makefile security/aircrack-ng/pkg: PFRAG.ppc Log message: aircrack-ng: allow packaging on macppc POWER8 libaircrack-crypto isn't built on powerpc -- remove them from PFRAG. OK benoit@ (maintainer)
Re: UPDATE: lang/gawk 5.0.0 => 5.0.1
On Sun, 11 Aug 2019 21:57:07 -0400 Brian Callahan wrote: > > > On 7/29/19 9:12 PM, Brian Callahan wrote: > > Hi ports -- > > > > Attached is a simple update to lang/gawk. All tests pass on amd64 > > and macppc. > > There is something in FILESDIR. But removing it seems to have no > > consequence on building the port whatsoever. So let's get rid of it. > > > > Take MAINTAINER because sure why not. > > > > OK? OK cwen@ (i've tested on amd64 only) > > ~Brian > > > > Ping. Diff reattached for convenience. > > ~Brian >
[ports-gcc] Unbreak graphics/vulkan-{tools,validation-layers}
Hi!
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/graphics/vulkan-tools.log
> http://build-failures.rhaalovely.net/powerpc/2019-07-29/graphics/vulkan-validation-layers.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-06/graphics/vulkan-tools.log
> http://build-failures.rhaalovely.net/sparc64/2019-08-06/graphics/vulkan-validation-layers.log
This is C++11 code and as such require ports-gcc, it builds once
changed [0] [1].
Charlène.
[0] https://bin.charlenew.xyz/vulkan-tools.log
[1] https://bin.charlenew.xyz/vulkan-validation-layers.log
Index: graphics/vulkan-tools/Makefile
===
RCS file: /cvs/ports/graphics/vulkan-tools/Makefile,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile
--- graphics/vulkan-tools/Makefile 22 Jun 2019 16:51:06 - 1.2
+++ graphics/vulkan-tools/Makefile 12 Aug 2019 09:33:27 -
@@ -18,6 +18,9 @@ PERMIT_PACKAGE = Yes
WANTLIB += ${COMPILER_LIBCXX} ICE SM X11 Xext c m vulkan xcb
+# C++11
+COMPILER = base-clang ports-gcc
+
MODULES = devel/cmake \
lang/python
MODPY_VERSION =${MODPY_DEFAULT_VERSION_3}
Index: graphics/vulkan-validation-layers/Makefile
===
RCS file: /cvs/ports/graphics/vulkan-validation-layers/Makefile,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile
--- graphics/vulkan-validation-layers/Makefile 22 Jun 2019 16:51:06 -
1.2
+++ graphics/vulkan-validation-layers/Makefile 12 Aug 2019 09:33:27 -
@@ -18,6 +18,9 @@ PERMIT_PACKAGE = Yes
WANTLIB += ${COMPILER_LIBCXX} m
+# C++11
+COMPILER = base-clang ports-gcc
+
MODULES = devel/cmake \
lang/python
MODPY_VERSION =${MODPY_DEFAULT_VERSION_3}
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/11 14:53:07 Modified files: lang/gcc/8 : Makefile lang/gcc/8/patches: patch-libgcc_config_host Added files: lang/gcc/8/patches: patch-libgcc_config_rs6000_t-openbsd Log message: gcc/8: add crt* symbols on powerpc These symbols are needed at least for php-7.3. Tips (thanks a lot!) and OK pascal@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: c...@cvs.openbsd.org2019/08/11 14:42:44 Modified files: audio/clementine: Makefile Added files: audio/clementine/patches: patch-3rdparty_libechonest_CMakeLists_txt patch-src_devices_giolister_cpp patch-src_musicbrainz_chromaprinter_cpp Log message: clementine: add missing headers and remove -Werror on 3rd party libechonest so it builds with ports-gcc. While here, prepare for the future update to chromaprint-1.4x in a backward compatible way. OK bket@ kmos@
