Re: i386 build failures: eduke32 netsurf nblood openclonk compcert dino

2024-04-26 Thread Theo Buehler 
On Fri, Apr 26, 2024 at 06:41:18PM +0100, Stuart Henderson wrote:
> > eduke32, nblood:
> 
> ld: error: undefined symbol: startbanner_pixdata
> 
> > openclonk, netsurf/browser:
> 
> [ 89%] Generating mape-resource.c
> failed to load 
> "/pobj/openclonk-8.1/openclonk-release-8.1-src/src/res/ocd.ico": Couldn?t 
> recognize the image file format for file ?/pobj/openclonk-8.1/openclonk
> -release-8.1-src/src/res/ocd.ico?
> /pobj/openclonk-8.1/openclonk-release-8.1-src/src/res/mape.xml: Child process 
> exited with code 1.
> 
> glib-compile-resources --generate-source --sourcedir frontends/gtk/res 
> --target=build/OpenBSD-gtk3/netsurf_gresource.c 
> frontends/gtk/res/netsurf.gresource.xml
> failed to load "frontends/gtk/res/netsurf.xpm": Couldn?t recognize the image 
> file format for file ?frontends/gtk/res/netsurf.xpm?
> frontends/gtk/res/netsurf.gresource.xml: Child process exited with code 1.
> 

All these were fallout from graphics/gdk-pixbuf. aja's last two commits
to that port fixed them on amd64.

> 
> > dino:
> 
> /pobj/dino-0.4.3/build-i386/exports/gpgme.vapi:59:25: error: no member named 
> 'subkeys_length1' in 'struct _gpgme_key'
> - neither gpgme nor gnupg have been updated recently enough to be implicated

Same on amd64 and aarch64. vapi smells like vala might be involved.

i386 build failures: eduke32 netsurf nblood openclonk compcert dino

2024-04-26 Thread Stuart Henderson 
> eduke32, nblood:

ld: error: undefined symbol: startbanner_pixdata

> openclonk, netsurf/browser:

[ 89%] Generating mape-resource.c
failed to load "/pobj/openclonk-8.1/openclonk-release-8.1-src/src/res/ocd.ico": 
Couldn?t recognize the image file format for file ?/pobj/openclonk-8.1/openclonk
-release-8.1-src/src/res/ocd.ico?
/pobj/openclonk-8.1/openclonk-release-8.1-src/src/res/mape.xml: Child process 
exited with code 1.

glib-compile-resources --generate-source --sourcedir frontends/gtk/res 
--target=build/OpenBSD-gtk3/netsurf_gresource.c 
frontends/gtk/res/netsurf.gresource.xml
failed to load "frontends/gtk/res/netsurf.xpm": Couldn?t recognize the image 
file format for file ?frontends/gtk/res/netsurf.xpm?
frontends/gtk/res/netsurf.gresource.xml: Child process exited with code 1.

> compcert:

various linker relocation/preempted symbol errors following ocaml update

> dino:

/pobj/dino-0.4.3/build-i386/exports/gpgme.vapi:59:25: error: no member named 
'subkeys_length1' in 'struct _gpgme_key'
- neither gpgme nor gnupg have been updated recently enough to be implicated

Re: i386 build failures: graphics/birdfont, x11/qt5/docs,-html

2021-05-17 Thread Matthias Kilian 
Hi,

On Mon, May 17, 2021 at 08:59:36AM +0100, Stuart Henderson wrote:
> gnome/maps is already fixed, mozillas I mentioned yesterday, here
> are two others:
[...]
> Done
> Writing Config.vala
> >>> Running build in graphics/birdfont at 1621205656.53
> ===> graphics/birdfont
> ===>  Building for birdfont-2.29.4
> cd /pobj/birdfont-2.29.4/birdfont-2.29.4 && /usr/local/bin/doit
> build/libbirdfont/load_font.c:20:10: fatal error: 'birdfont.h' file not found
> #include 
>  ^~~~
> 1 error generated.

FWIW: I've seen this on amd64, too. Two times today while running
dpb -- I've removed the build dir of birdfont and the lock file.
The third built the succeeded.

No idea why it fails in such a non-deterministic way.

Ciao,
Kili

Re: i386 build failures: graphics/birdfont, x11/qt5/docs,-html

2021-05-17 Thread Rafael Sadowski 
On Mon May 17, 2021 at 08:59:36AM +0100, Stuart Henderson wrote:
> gnome/maps is already fixed, mozillas I mentioned yesterday, here
> are two others:
> 

No idea what changed in the ports tree...

Index: Makefile
===
RCS file: /cvs/ports/x11/qt5/docs/Makefile,v
retrieving revision 1.13
diff -u -p -u -p -r1.13 Makefile
--- Makefile9 Jun 2020 22:41:43 -   1.13
+++ Makefile17 May 2021 19:40:51 -
@@ -6,6 +6,7 @@ COMMENT-qch =   qdoc-compiled documentati
 PKGNAME =  qt5-docs-${QT5_VERSION}
 PKGNAME-html = qt5-html-${QT5_VERSION}
 PKGNAME-qch =  qt5-qch-${QT5_VERSION}
+REVISION-html =0
 
 MULTI_PACKAGES =   -html -qch
 SUBPACKAGE ?=  -html
Index: pkg/PLIST-html
===
RCS file: /cvs/ports/x11/qt5/docs/pkg/PLIST-html,v
retrieving revision 1.11
diff -u -p -u -p -r1.11 PLIST-html
--- pkg/PLIST-html  19 Jan 2021 06:16:29 -  1.11
+++ pkg/PLIST-html  17 May 2021 19:40:51 -
@@ -553,24 +553,11 @@ share/doc/qt5/qt3d/qt3danimation.html
 share/doc/qt5/qt3d/qt3dcore-module.html
 share/doc/qt5/qt3d/qt3dcore-qabstractaspect-proxy.html
 share/doc/qt5/qt3d/qt3dcore-qaspectengine-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qcomponentaddedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qcomponentremovedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qdynamicpropertyupdatedchange-proxy.html
 share/doc/qt5/qt3d/qt3dcore-qentity-proxy.html
 share/doc/qt5/qt3d/qt3dcore-qnode-proxy.html
 share/doc/qt5/qt3d/qt3dcore-qnodecommand-proxy.html
 share/doc/qt5/qt3d/qt3dcore-qnodecreatedchange.html
 share/doc/qt5/qt3d/qt3dcore-qnodecreatedchangebase-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qnodedestroyedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertynodeaddedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertynoderemovedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyupdatedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyupdatedchangebase-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyvalueaddedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyvalueaddedchangebase-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyvalueremovedchange-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qpropertyvalueremovedchangebase-proxy.html
-share/doc/qt5/qt3d/qt3dcore-qstaticpropertyupdatedchangebase-proxy.html
 share/doc/qt5/qt3d/qt3dcore.html
 share/doc/qt5/qt3d/qt3dextras-module.html
 share/doc/qt5/qt3d/qt3dinput-module.html

i386 build failures: graphics/birdfont, x11/qt5/docs,-html

2021-05-17 Thread Stuart Henderson 
gnome/maps is already fixed, mozillas I mentioned yesterday, here
are two others:

- Forwarded message from Stuart Henderson  
-

From: Stuart Henderson 
Date: Sun, 16 May 2021 23:02:40 -0600 (MDT)
To: st...@i386.ports.openbsd.org
Subject: i386.ports.openbsd.org build done

bulk build on i386.ports.openbsd.org
started at 1621095835 Sat May 15 10:23:55 MDT 2021
finished at 1621227733 Sun May 16 23:02:13 MDT 2021
report generated at 1621227759 Sun May 16 23:02:39 MDT 2021
total   36h38m
done with kern.version=OpenBSD 6.9-current (GENERIC.MP) #0: Sat May 15 10:04:05 
MDT 2021

built packages10288
3904
6383

build failures: 7
graphics/birdfont
mail/mozilla-thunderbird
www/firefox-esr
www/mozilla-firefox
www/tor-browser/browser
x11/gnome/maps
x11/qt5/docs,-html

new failures
resolved failures

packages newly built
packages not built this time
=
==> failures/graphics/birdfont.log <==
Looking for library: sqlite3= 9
Looking for library: xmlbird= 9
Looking for library: gee-0.8= 9

Done
Writing Config.vala
>>> Running build in graphics/birdfont at 1621205656.53
===> graphics/birdfont
===>  Building for birdfont-2.29.4
cd /pobj/birdfont-2.29.4/birdfont-2.29.4 && /usr/local/bin/doit
build/libbirdfont/load_font.c:20:10: fatal error: 'birdfont.h' file not found
#include 
 ^~~~
1 error generated.
.  build
.  compile_translations
.  man:birdfont.1
.  man:birdfont-export.1
.  man:birdfont-import.1
.  man:birdfont-autotrace.1
.  mkdir build/libbirdfont
.  copy libbirdfont/OpenFontFormat/loadfont.h
.  copy libbirdfont/OpenFontFormat/load_font.c
.  copy libbirdfont/Renderer/fontconfig.c
.  copy libbirdfont/overview_glyph.c
.  mkdir build/libbirdgems
.  copy libbirdgems/GraphicsGems.h
.  valac libbirdgems
.  copy libbirdgems/GGVecLib.c
.  copy libbirdgems/fit_cubic.c
.  compile build/libbirdgems/birdgems.c
.  compile build/libbirdgems/GGVecLib.c
.  compile build/libbirdgems/fit_cubic.c
.  libbirdgems
.  Create link libbirdgems.so.17.0 libbirdgems.so
.  compile build/libbirdfont/load_font.c
TaskFailed - taskid:compile build/libbirdfont/load_font.c
Command failed: 'cc  -c build/libbirdfont/load_font.c 
-fPIC -D 'GETTEXT_PACKAGE="birdfont"' -I 
./build/libbirdfont -I ./build/libbirdgems $(pkg-config 
--cflags sqlite3) $(pkg-config --cflags fontconfig) 
$(pkg-config --cflags gee-0.8) $(pkg-config --cflags gio-2.0)   
  $(pkg-config --cflags cairo) $(pkg-config --cflags glib-2.0)  
   $(pkg-config --cflags xmlbird) -o 
build/libbirdfont/load_font.o' returned 1


compile build/libbirdfont/load_font.c :

*** Error 1 in graphics/birdfont (Makefile:58 'do-build')
*** Error 2 in graphics/birdfont (/usr/ports/infrastructure/mk/bsd.port.mk:2952 
'/pobj/birdfont-2.29.4/.build_done': @cd /usr/ports/graphics...)
*** Error 2 in graphics/birdfont (/usr/ports/infrastructure/mk/bsd.port.mk:2602 
'build': @lock=birdfont-2.29.4;  export _LOCKS_HELD=" birdfo...)
===> Exiting graphics/birdfont with an error
*** Error 1 in /usr/ports (infrastructure/mk/bsd.port.subdir.mk:137 'build': @: 
${echo_msg:=echo};  : ${target:=build};  for i in ; do  eval...)
>>> Ended at 1621205658.82
max_stuck=7.15/depends=7.47/show-prepare-results=1.13/patch=0.66/configure=0.99/build=2.29
Error: job failed with 512 on i386-2 at 1621205658

==> failures/mail/mozilla-thunderbird.log <==
[glslopt 0.1.2] running: "/pobj/thunderbird-78.10.1/bin/c++" "-std=gnu++17" 
"-O0" "-ffunction-sections" "-fdata-sections" "-fPIC" "-m32" "-I" 
"glsl-optimizer/include" "-I" "glsl-optimizer/src/mesa" "-I" 
"glsl-optimizer/src/mapi" "-I" "glsl-optimizer/src/compiler" "-I" 
"glsl-optimizer/src/compiler/glsl" "-I" "glsl-optimizer/src/gallium/auxiliary" 
"-I" "glsl-optimizer/src/gallium/include" "-I" "glsl-optimizer/src" "-I" 
"glsl-optimizer/src/util" "-D__STDC_FORMAT_MACROS" "-DHAVE_PTHREAD" 
"-DHAVE_TIMESPEC_GET" "-o" 
"/pobj/thunderbird-78.10.1/build-i386/release/build/glslopt-4c728f9446d0d023/out/glsl-optimizer/src/compiler/glsl/opt_structure_splitting.o"
 "-c" "glsl-optimizer/src/compiler/glsl/opt_structure_splitting.cpp"
[glslopt 0.1.2] exit code: 0
[glslopt 0.1.2] running: "/pobj/thunderbird-78.10.1/bin/c++" "-std=gnu++17" 
"-O0" "-ffunction-sections" "-fdata-sections" "-fPIC" "-m32" "-I" 
"glsl-optimizer/include" "-I" "glsl-optimizer/src/mesa" "-I" 
"glsl-optimizer/src/mapi" "-I" "glsl-optimizer/src/compiler" "-I" 
"glsl-optimizer/src/compiler/glsl" "-I" "glsl-optimizer/src/gallium/auxiliary" 
"-I" "glsl-optimizer/src/gallium/include" "-I" "glsl-optimizer/src" "-I" 
"glsl-optimizer/src/util" "-D__STDC_FORMAT_MACROS" "-DHAVE_PTHREAD" 
"-DHAVE_TIMESPEC_GET" "-o" 
"/pobj/thunderbird-78.10.1/build-i386/release/build/glslopt-4c728f9446d0d023/out/glsl-optimizer/src/compiler/glsl/opt_swizzle.o"

Re: i386 build failures: graphics/makehuman, graphics/vulkan-loader

2020-07-16 Thread Thomas Frohwein 
On Thu, Jul 16, 2020 at 11:52:26AM -0600, Thomas Frohwein wrote:
[...]
> +# Integrated clang assembler doesn't work with x86 - Bug #698164
> +# remove if build can be updated to clang 9.0.0 (see
> +# https://github.com/KhronosGroup/Vulkan-Loader/issues/252)
> +#.if ${MACHINE_ARCH}==i386 && ${PROPERTIES:Mclang}
> +#CFLAGS +=   -fno-integrated-as
> +#.endif
[...]

Oops, this wasn't meant to be commented out.

Here's the proper diff:

Index: Makefile
===
RCS file: /cvs/ports/graphics/vulkan-loader/Makefile,v
retrieving revision 1.7
diff -u -p -r1.7 Makefile
--- Makefile7 Jul 2020 08:08:06 -   1.7
+++ Makefile16 Jul 2020 18:17:21 -
@@ -7,6 +7,7 @@ PKGNAME =   vulkan-loader-${V}
 GH_TAGNAME =   sdk-${V}
 GH_ACCOUNT =   KhronosGroup
 GH_PROJECT =   Vulkan-Loader
+REVISION = 0
 
 SHARED_LIBS += vulkan  0.1 # 1.2
 
@@ -34,6 +35,15 @@ MODPY_RUNDEP =   No
 
 # Tests only build if Google Test framework is in directory external/
 NO_TEST =  Yes
+
+.include 
+
+# Integrated clang assembler doesn't work with x86 - Bug #698164
+# remove if build can be updated to clang 9.0.0 (see
+# https://github.com/KhronosGroup/Vulkan-Loader/issues/252)
+.if ${MACHINE_ARCH}==i386 && ${PROPERTIES:Mclang}
+CFLAGS +=  -fno-integrated-as
+.endif
 
 pre-configure:
${SUBST_CMD} ${WRKSRC}/CMakeLists.txt

Re: i386 build failures: graphics/makehuman, graphics/vulkan-loader

2020-07-16 Thread Thomas Frohwein 
On Wed, Jul 15, 2020 at 10:51:21AM +0100, Stuart Henderson wrote:
[...]
> ==> failures/graphics/vulkan-loader.log <==
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:628:5:
>  note: while in macro instantiation
> PhysDevExtTermin 243
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:629:5:
>  note: while in macro instantiation
> PhysDevExtTermin 244
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:630:5:
>  note: while in macro instantiation
> PhysDevExtTermin 245
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:631:5:
>  note: while in macro instantiation
> PhysDevExtTermin 246
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:632:5:
>  note: while in macro instantiation
> PhysDevExtTermin 247
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:633:5:
>  note: while in macro instantiation
> PhysDevExtTermin 248
> ^
> :14:17: error: cannot use more than one symbol in memory 
> operand
> push offset termin_error_string@GOT # Push the error string (fourth arg)
> ^
> /pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:634:5:
>  note: while in macro instantiation
> PhysDevExtTermin 249
> ^
> ninja: build stopped: subcommand failed.
> ===> Exiting graphics/vulkan-loader with an error
> *** Error 1 in graphics/vulkan-loader 
> (/usr/ports/devel/cmake/cmake.port.mk:34 'do-build': @cd 
> /pobj/vulkan-loader-1.2.141.0/build-i386 && e...)
> *** Error 2 in graphics/vulkan-loader 
> (/usr/ports/infrastructure/mk/bsd.port.mk:2921 
> '/pobj/vulkan-loader-1.2.141.0/build-i386/.build_done')
> *** Error 2 in graphics/vulkan-loader 
> (/usr/ports/infrastructure/mk/bsd.port.mk:2584 'build': 
> @lock=vulkan-loader-1.2.141.0;  export _LOCKS_...)
> *** Error 1 in /usr/ports (infrastructure/mk/bsd.port.subdir.mk:137 'build': 
> @: ${echo_msg:=echo};  : ${target:=build};  for i in ; do  eval...)
> >>> Ended at 1594798947.24
> max_stuck=0.00/depends=2.85/show-prepare-results=0.66/build=10.58
> Error: job failed with 512 on i386-3 at 1594798947
> 

Issue 252 on GitHub discusses this [1], and it seems that adding
'-fno-integrated-as' should fix it. I don't have an i386 build system
set up. Stuart, could you test if the following diff fixes the build?

[1] https://github.com/KhronosGroup/Vulkan-Loader/issues/252

Index: Makefile
===
RCS file: /cvs/ports/graphics/vulkan-loader/Makefile,v
retrieving revision 1.7
diff -u -p -r1.7 Makefile
--- Makefile7 Jul 2020 08:08:06 -   1.7
+++ Makefile16 Jul 2020 17:48:43 -
@@ -7,6 +7,7 @@ PKGNAME =   vulkan-loader-${V}
 GH_TAGNAME =   sdk-${V}
 GH_ACCOUNT =   KhronosGroup
 GH_PROJECT =   Vulkan-Loader
+REVISION = 0
 
 SHARED_LIBS += vulkan  0.1 # 1.2
 
@@ -34,6 +35,15 @@ MODPY_RUNDEP =   No
 
 # Tests only build if Google Test framework is in directory external/
 NO_TEST =  Yes
+
+.include 
+
+# Integrated clang assembler doesn't work with x86 - Bug #698164
+# remove if build can be updated to clang 9.0.0 (see
+# https://github.com/KhronosGroup/Vulkan-Loader/issues/252)
+#.if ${MACHINE_ARCH}==i386 && ${PROPERTIES:Mclang}
+#CFLAGS += -fno-integrated-as
+#.endif
 
 pre-configure:
${SUBST_CMD} ${WRKSRC}/CMakeLists.txt

makehuman (was: i386 build failures: graphics/makehuman, graphics/vulkan-loader)

2020-07-15 Thread Matthias Kilian 
Hi,

On Wed, Jul 15, 2020 at 10:51:21AM +0100, Stuart Henderson wrote:
> graphics/makehuman
[...]

> cd 
> /pobj/makehuman-1.2.0beta2/makehuman-1c6a4147eeec0f1f07f73904f33ebd4e8ef0065a/makehuman
>  && /usr/local/bin/python3.8 compile_models.py
> Traceback (most recent call last):
>   File "compile_models.py", line 44, in 
> import module3d
>   File "./core/module3d.py", line 41, in 
> import numpy as np
> ModuleNotFoundError: No module named 'numpy'

makehuman needs (at least) two of its run dependencies at build time:

Index: Makefile
===
RCS file: /cvs/ports/graphics/makehuman/Makefile,v
retrieving revision 1.31
diff -u -p -r1.31 Makefile
--- Makefile13 Jul 2020 20:09:27 -  1.31
+++ Makefile15 Jul 2020 22:46:02 -
@@ -27,6 +27,9 @@ RUN_DEPENDS=  devel/desktop-file-utils \
math/py-numpy${MODPY_FLAVOR} \
x11/py-qt5${MODPY_FLAVOR}
 
+BUILD_DEPENDS= math/py-numpy${MODPY_FLAVOR} \
+   x11/py-qt5${MODPY_FLAVOR}
+
 do-build:
cd ${WRKSRC}/makehuman && ${MODPY_BIN} ${MODPY_LIBDIR}/compileall.py .
 .for TARGET in models proxies targets

i386 build failures: graphics/makehuman, graphics/vulkan-loader

2020-07-15 Thread Stuart Henderson 
started at 1594676294 Mon Jul 13 15:38:14 MDT 2020
finished at 1594805629 Wed Jul 15 03:33:49 MDT 2020
report generated at 1594805659 Wed Jul 15 03:34:19 MDT 2020
total   35h55m
done with kern.version=OpenBSD 6.7-current (GENERIC.MP) #0: Mon Jul 13 15:19:13 
MDT 2020

built packages
build failures: 2
graphics/makehuman
graphics/vulkan-loader

new failures
resolved failures

packages newly built
packages not built this time
=
==> failures/graphics/makehuman.log <==
Compiling './plugins/9_export_fbx/fbx_skeleton.py'...
Compiling './plugins/9_export_fbx/fbx_utils.py'...
Compiling './plugins/9_export_fbx/fbx_utils_bin.py'...
Compiling './plugins/9_export_fbx/mh2fbx.py'...
Compiling './plugins/9_export_light.py'...
Listing './plugins/9_export_obj'...
Compiling './plugins/9_export_obj/__init__.py'...
Compiling './plugins/9_export_obj/mh2obj.py'...
Listing './plugins/9_export_ogre'...
Compiling './plugins/9_export_ogre/__init__.py'...
Compiling './plugins/9_export_ogre/mh2ogre.py'...
Listing './plugins/9_export_stl'...
Compiling './plugins/9_export_stl/__init__.py'...
Compiling './plugins/9_export_stl/mh2stl.py'...
Compiling './plugins/9_export_uvmap.py'...
Listing './shared'...
Compiling './shared/animation.py'...
Compiling './shared/bvh.py'...
Compiling './shared/cache.py'...
Compiling './shared/image_operations.py'...
Compiling './shared/material.py'...
Compiling './shared/mesh_operations.py'...
Compiling './shared/mh2renderman.py'...
Compiling './shared/progress.py'...
Compiling './shared/projection.py'...
Compiling './shared/proxy.py'...
Compiling './shared/scene.py'...
Compiling './shared/skeleton.py'...
Compiling './shared/skeleton_drawing.py'...
Compiling './shared/subtexture.py'...
Compiling './shared/wavefront.py'...
Listing './testsuite'...
Compiling './testsuite/__init__.py'...
Compiling './testsuite/blender_initTest.py'...
Compiling './testsuite/test_blender.py'...
cd 
/pobj/makehuman-1.2.0beta2/makehuman-1c6a4147eeec0f1f07f73904f33ebd4e8ef0065a/makehuman
 && /usr/local/bin/python3.8 compile_models.py
Traceback (most recent call last):
  File "compile_models.py", line 44, in 
import module3d
  File "./core/module3d.py", line 41, in 
import numpy as np
ModuleNotFoundError: No module named 'numpy'
===> Exiting graphics/makehuman with an error
*** Error 1 in graphics/makehuman (Makefile:33 'do-build')
*** Error 2 in graphics/makehuman 
(/usr/ports/infrastructure/mk/bsd.port.mk:2921 
'/pobj/makehuman-1.2.0beta2/.build_done': @cd /usr/ports/gr...)
*** Error 2 in graphics/makehuman 
(/usr/ports/infrastructure/mk/bsd.port.mk:2584 'build': 
@lock=makehuman-1.2.0beta2;  export _LOCKS_HELD=" ...)
*** Error 1 in /usr/ports (infrastructure/mk/bsd.port.subdir.mk:137 'build': @: 
${echo_msg:=echo};  : ${target:=build};  for i in ; do  eval...)
>>> Ended at 1594776991.41
max_stuck=0.00/depends=1.59/show-prepare-results=0.33/build=3.25
Error: job failed with 512 on i386-3 at 1594776991

==> failures/graphics/vulkan-loader.log <==
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:628:5:
 note: while in macro instantiation
PhysDevExtTermin 243
^
:14:17: error: cannot use more than one symbol in memory operand
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:629:5:
 note: while in macro instantiation
PhysDevExtTermin 244
^
:14:17: error: cannot use more than one symbol in memory operand
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:630:5:
 note: while in macro instantiation
PhysDevExtTermin 245
^
:14:17: error: cannot use more than one symbol in memory operand
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:631:5:
 note: while in macro instantiation
PhysDevExtTermin 246
^
:14:17: error: cannot use more than one symbol in memory operand
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:632:5:
 note: while in macro instantiation
PhysDevExtTermin 247
^
:14:17: error: cannot use more than one symbol in memory operand
push offset termin_error_string@GOT # Push the error string (fourth arg)
^
/pobj/vulkan-loader-1.2.141.0/Vulkan-Loader-sdk-1.2.141.0/loader/unknown_ext_chain_gas.S:633:5:
 note: while in macro instantiation
PhysDevExtTermin 248
^
:14:17: error: cannot use more than one symbol in memory operand
push offset

Re: i386 build failures: devel/kf5/kdelibs4support, devel/pygame

2020-05-25 Thread Stuart Henderson 
On 2020/05/24 23:42, Rafael Sadowski wrote:
> On Sun May 24, 2020 at 03:28:41PM +0100, Stuart Henderson wrote:
> > -- Configuring done
> > -- Generating done
> > -- Build files have been written to: /pobj/kdelibs4support-5.68.0/build-i386
> > ninja: error: manifest 'build.ninja' still dirty after 100 tries
> > 
> Time out of sync? It was always this problem.
> 

oh, this thing where it cares about the timestamps of files *outside* the build
directory again. Yes, looks like that's it. The machines swap a lot and don't
always keep great time.

Re: i386 build failures: devel/kf5/kdelibs4support, devel/pygame

2020-05-24 Thread Rafael Sadowski 
On Sun May 24, 2020 at 03:28:41PM +0100, Stuart Henderson wrote:
> -- Configuring done
> -- Generating done
> -- Build files have been written to: /pobj/kdelibs4support-5.68.0/build-i386
> ninja: error: manifest 'build.ninja' still dirty after 100 tries
> 
Time out of sync? It was always this problem.

i386 build failures: devel/kf5/kdelibs4support, devel/pygame

2020-05-24 Thread Stuart Henderson 
==> failures/devel/kf5/kdelibs4support.log <==
-- Configuring done
-- Generating done
-- Build files have been written to: /pobj/kdelibs4support-5.68.0/build-i386
[0/1] /usr/local/bin/cmake 
-S/pobj/kdelibs4support-5.68.0/kdelibs4support-5.68.0 
-B/pobj/kdelibs4support-5.68.0/build-i386
Re-run cmake no build system arguments
-- 

Installing in /usr/local. Run /pobj/kdelibs4support-5.68.0/build-i386/prefix.sh 
to set the environment for KDELibs4Support.
-- Could not set up the appstream test. appstreamcli is missing.
-- QNtrack not found.  Can be used as a data input to Solid network status.  
See https://launchpad.net/ntrack
-- The following features have been enabled:

 * X11 Session Management (libSM), Support for session management in 
KApplication

-- The following OPTIONAL packages have been found:

 * X11
 * Libintl
 * DocBookXML4 (required version >= 4.2), DocBook XML 4, 


-- The following RECOMMENDED packages have been found:

 * OpenSSL, Support for secure network communications (SSL and TLS), 

   KDE uses OpenSSL for the bulk of secure communications, including secure web 
browsing via HTTPS

-- The following REQUIRED packages have been found:

 * ECM (required version >= 5.68.0), Extra CMake Modules., 

 * Qt5Test
 * Qt5Svg
 * Qt5PrintSupport
 * Qt5UiPlugin (required version >= 5.13.2)
 * Qt5Designer
 * KF5Crash (required version >= 5.68.0)
 * KF5DesignerPlugin (required version >= 5.68.0)
 * KF5GlobalAccel (required version >= 5.68.0)
 * KF5DocTools (required version >= 5.68.0)
 * KF5Emoticons (required version >= 5.68.0)
 * Qt5Gui (required version >= 5.12.0)
 * KF5GuiAddons (required version >= 5.68.0)
 * KF5IconThemes (required version >= 5.68.0)
 * KF5Notifications (required version >= 5.68.0)
 * KF5Bookmarks (required version >= 5.68.0)
 * KF5Completion (required version >= 5.68.0)
 * KF5ItemViews (required version >= 5.68.0)
 * KF5JobWidgets (required version >= 5.68.0)
 * KF5Solid (required version >= 5.68.0)
 * Qt5Network (required version >= 5.12.0)
 * KF5KIO (required version >= 5.68.0)
 * KF5Parts (required version >= 5.68.0)
 * KF5Service (required version >= 5.68.0)
 * KF5Sonnet (required version >= 5.68.0)
 * Gettext
 * KF5I18n (required version >= 5.68.0)
 * KF5TextWidgets (required version >= 5.68.0)
 * KF5UnitConversion (required version >= 5.68.0)
 * KF5WindowSystem (required version >= 5.68.0)
 * KF5CoreAddons (required version >= 5.68.0)
 * KF5Auth (required version >= 5.68.0)
 * Qt5Core (required version >= 5.12.0)
 * KF5Codecs (required version >= 5.68.0)
 * Qt5Xml (required version >= 5.12.0)
 * KF5Config (required version >= 5.68.0)
 * Qt5Widgets (required version >= 5.12.0)
 * KF5WidgetsAddons (required version >= 5.68.0)
 * KF5ConfigWidgets (required version >= 5.68.0)
 * KF5XmlGui (required version >= 5.68.0)
 * Qt5DBus (required version >= 5.12.0)
 * KF5DBusAddons (required version >= 5.68.0)
 * KDED (required version >= 5.68.0)
 * XCB, X protocol C-language Binding, 
 * Perl
 * PerlModules, 
 * Qt5Concurrent
 * Qt5 (required version >= 5.12.0)
 * Qt5X11Extras (required version >= 5.12.0)

-- Configuring done
-- Generating done
-- Build files have been written to: /pobj/kdelibs4support-5.68.0/build-i386
ninja: error: manifest 'build.ninja' still dirty after 100 tries

*** Error 1 in devel/kf5/kdelibs4support 
(/usr/ports/devel/cmake/cmake.port.mk:34 'do-build': @cd 
/pobj/kdelibs4support-5.68.0/build-i386 &&...)
*** Error 2 in devel/kf5/kdelibs4support 
(/usr/ports/infrastructure/mk/bsd.port.mk:2907 
'/pobj/kdelibs4support-5.68.0/build-i386/.build_done')
*** Error 2 in devel/kf5/kdelibs4support 
(/usr/ports/infrastructure/mk/bsd.port.mk:2577 'build': 
@lock=kdelibs4support-5.68.0p0;  export _LO...)
===> Exiting devel/kf5/kdelibs4support with an error
*** Error 1 in /usr/ports (infrastructure/mk/bsd.port.subdir.mk:137 'build': @: 
${echo_msg:=echo};  : ${target:=build};  for i in ; do  eval...)
>>> Ended at 1590300538.72
Error: job failed with 512 on localhost at 1590300538

==> failures/devel/pygame,python3.log <==
>>> Building on localhost under devel/pygame,python3
 BDEPENDS = 
[devel/sdl-image;devel/sdl-mixer;devel/sdl-ttf;math/py-Numeric;lang/python/3.7;devel/py-setuptools,python3]
 DIST = [devel/pygame:pygame-1.9.3.tar.gz]
 FULLPKGNAME = py3-game-1.9.3p1
 RDEPENDS = 
[lang/python/3.7;devel/py-setuptools,python3;devel/sdl-image;devel/sdl-ttf;devel/sdl-mixer]
(Junk lock obtained for localhost at 1590317732.65)
>>> Running depends in devel/pygame,python3 at 1590317732.69
   last junk was in net/py-netaddr
/usr/sbin/pkg_add -aI -Drepair py-Numeric-23.1p9 py3-setuptools-41.6.0v0 
python-3.7.7 sdl-image-1.2.12p4 sdl-mixer-1.2.12p5 sdl-ttf-2.0.11p3
was: /usr/sbin/pkg_add -aI -Drepair py-Numeric-23.1p9 py3-setuptools-41.6.0v0 
python-3.7.7 sdl-image-1.2.12p4 sdl-mixer-1.2.12p5

i386 build failures: poppler, gnucash, terraform-provider-azurerm, chromium, firefox-esr

2019-09-21 Thread Stuart Henderson 
> print/poppler

-- Set runtime path of "/pobj/poppler-0.80.0/fake-i386/usr/local/bin/pdfunite" 
to ""
-- Installing: /pobj/poppler-0.80.0/fake-i386/usr/local/man/man1/pdfunite.1
CMake Error at glib/cmake_install.cmake:52 (file):
  file INSTALL cannot find
  "/pobj/poppler-0.80.0/build-i386/glib/libpoppler-glib.so.8.14.0".
Call Stack (most recent call first):
  cmake_install.cmake:237 (include)


FAILED: CMakeFiles/install.util 
cd /pobj/poppler-0.80.0/build-i386 && /usr/local/bin/cmake -P 
cmake_install.cmake
ninja: build stopped: subcommand failed.

> sysutils/terraform/provider-azurerm

# github.com/terraform-providers/terraform-provider-azurerm
runtime: pointer 0xbb1e9c77 to unallocated span span.base()=0xbabbe000 
span.limit=0x0 span.state=3
runtime: found in object at *(0x9540e000+0x19f74)
object=0x9540e000 s.base()=0x9540e000 s.limit=0x954285e0 s.spanclass=0 
s.elemsize=114688 s.state=mSpanInUse
 *(object+0) = 0x953fc540
 *(object+4) = 0x5d
.
 *(object+106412) = 0xdd37bb5
 *(object+106416) = 0x0
 ...
fatal error: found bad pointer in Go heap (incorrect use of unsafe or cgo?)

> www/chromium

In file included from 
../../chrome/browser/media/router/mojo/media_router_mojo_impl.cc:18:
../../chrome/browser/media/cast_mirroring_service_host.h:13:10: fatal error: 
'components/mirroring/mojom/mirroring_service.mojom.h' file not found
#include "components/mirroring/mojom/mirroring_service.mojom.h"

> www/firefox-esr

error: Resource temporarily unavailable (os error 35)
gmake[3]: Leaving directory 
'/pobj/firefox-esr-68.1.0/build-i386/toolkit/library/rust'

> productivity/gnucash

wrote 
`/pobj/gnucash-3.7/build-i386/lib/gnucash/scm/ccache/2.2/gnucash/gnc-modules/test/test-gnc-module-scm-dynload.go'
[373/868] cd /pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test && 
/usr/local/bin/cmake -E env 
LD_LIBRARY_PATH=/pobj/gnucash-3.7/build-i386/lib:/pobj/gnucash-3.7/build-i386/lib/gnucash
 GNC_UNINSTALLED=YES GNC_BUILDDIR="/pobj/gnucash-3.7/build-i386" 
GUILE_LOAD_PATH="/pobj/gnucash-3.7/gnucash-3.7/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/libgnucash/scm:/pobj/gnucash-3.7/build-i386/share/gnucash/scm"
 
GUILE_LOAD_COMPILED_PATH=3D"/pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/lib/gnucash/scm/ccache/2.2"
 
GNC_MODULE_PATH="/pobj/gnucash-3.7/build-i386/lib:/pobj/gnucash-3.7/build-i386/lib/gnucash:"
 /usr/local/bin/guile2.2 -e '(@@ (guild) main)' -s /usr/local/bin/guild2.2 
compile -o 
/pobj/gnucash-3.7/build-i386/lib/gnucash/scm/ccache/2.2/gnucash/gnc-modules/test/test-gnc-module-swigged-c.go
 
/pobj/gnucash-3.7/gnucash-3.7/libgnucash/gnc-module/test/test-gnc-module-swigged-c.scm
FAILED: 
lib/gnucash/scm/ccache/2.2/gnucash/gnc-modules/test/test-gnc-module-swigged-c.go
 
cd /pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test && 
/usr/local/bin/cmake -E env 
LD_LIBRARY_PATH=/pobj/gnucash-3.7/build-i386/lib:/pobj/gnucash-3.7/build-i386/lib/gnucash
 GNC_UNINSTALLED=YES GNC_BUILDDIR="/pobj/gnucash-3.7/build-i386" 
GUILE_LOAD_PATH="/pobj/gnucash-3.7/gnucash-3.7/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/libgnucash/scm:/pobj/gnucash-3.7/build-i386/share/gnucash/scm"
 
GUILE_LOAD_COMPILED_PATH="/pobj/gnucash-3.7/build-i386/libgnucash/gnc-module/test:/pobj/gnucash-3.7/build-i386/lib/gnucash/scm/ccache/2.2"
 
GNC_MODULE_PATH="/pobj/gnucash-3.7/build-i386/lib:/pobj/gnucash-3.7/build-i386/lib/gnucash:"
 /usr/local/bin/guile2.2 -e '(@@ (guild) main)' -s /usr/local/bin/guild2.2 
compile -o 
/pobj/gnucash-3.7/build-i386/lib/gnucash/scm/ccache/2.2/gnucash/gnc-modules/test/test-gnc-module-swigged-c.go
 
/pobj/gnucash-3.7/gnucash-3.7/libgnucash/gnc-module/test/test-gnc-module-swigged-c.scm
Backtrace:
In scripts/compile.scm:
   264:26 19 (_ _)
In system/base/target.scm:
 57:6 18 (with-target _ _)
In system/base/compile.scm:
152:6 17 (compile-file _ #:output-file _ #:from _ #:to _ #:env _ ?)
 43:4 16 (call-once _)
In ice-9/boot-9.scm:
841:4 15 (with-throw-handler _ _ _)
In system/base/compile.scm:
59:11 14 (_)
   155:11 13 (_ #)
   235:18 12 (read-and-compile # ?)
   183:32 11 (compile-fold (#) ?)
In ice-9/boot-9.scm:
   2312:4 10 (save-module-excursion #)
In language/scheme/compile-tree-il.scm:
31:15  9 (_)
In ice-9/psyntax.scm:
  1262:36  8 (expand-top-sequence ((use-modules (gnucash #))) _ _ #f ?)
  1209:24  7 (parse _ (("placeholder" placeholder)) ((top) #(# # ?)) ?)
   285:10  6 (parse _ (("placeholder" placeholder)) (()) _ c (# #) #)
In ice-9/boot-9.scm:
  3377:20  5 (process-use-modules _)
   222:17  4 (map1 (((gnucash gnc-module
  3378:31  3 (_ ((gnucash gnc-module)))
   2803:6  2 (resolve-interface _ #:select _ #:hide _ #:prefix _ # _ ?)
In unknown file:
   1 (scm-error misc-error #f "~A ~S" ("no code for modu?" ?) ?)
In ice-9/boot-9.scm:
   752:25  0 (dispatch-exception _ _ _)

Re: i386 build failures: games/dxx-rebirth

2019-06-27 Thread Stuart Henderson 
On 2019/06/27 09:24, Thomas Frohwein wrote:
> On Thu, Jun 27, 2019 at 12:28:46PM +0100, Stuart Henderson wrote:
> > Nothing changed in dxx-rebirth itself recently, possibly due to
> > libc++/llvm updates.
> > 
> > ld: error: undefined symbol: dxx_error_object_type_mismatch()
> > >>> referenced by collide.cpp
> > >>>   
> > >>> similar/main/.d1x-rebirth.collide.o:(bump_this_object(valptridx::ptridx > object>::vm>, valptridx::ptridx::vm>, 
> > dcx::vms_vector const&, int))
> > c++: error: linker command failed with exit code 1 (use -v to see 
> > invocation)
> > scons: *** [d1x-rebirth/d1x-rebirth] Error 1
> > scons: building terminated because of errors.
> > Failed target count: total=1; targets with enable_build_failure_summary=1: 1
> > Failed node list:
> > 
> 
> This looks like thes same error as on amd64 after clang/llvm update. On
> amd64 this is resolved with an updated version of the port that first
> needs an update to physfs. I'm working on addressing kn's questions
> about the physfs update, after which an update of dxx-rebirth should
> resolve the build failure.
> 

Thanks, I've added a BROKEN marker on clang arches for now.

Re: i386 build failures: games/dxx-rebirth

2019-06-27 Thread Thomas Frohwein 
On Thu, Jun 27, 2019 at 12:28:46PM +0100, Stuart Henderson wrote:
> Nothing changed in dxx-rebirth itself recently, possibly due to
> libc++/llvm updates.
> 
> ld: error: undefined symbol: dxx_error_object_type_mismatch()
> >>> referenced by collide.cpp
> >>>   
> >>> similar/main/.d1x-rebirth.collide.o:(bump_this_object(valptridx::ptridx object>::vm>, valptridx::ptridx::vm>, 
> dcx::vms_vector const&, int))
> c++: error: linker command failed with exit code 1 (use -v to see invocation)
> scons: *** [d1x-rebirth/d1x-rebirth] Error 1
> scons: building terminated because of errors.
> Failed target count: total=1; targets with enable_build_failure_summary=1: 1
> Failed node list:
> 

This looks like thes same error as on amd64 after clang/llvm update. On
amd64 this is resolved with an updated version of the port that first
needs an update to physfs. I'm working on addressing kn's questions
about the physfs update, after which an update of dxx-rebirth should
resolve the build failure.

i386 build failures: games/dxx-rebirth

2019-06-27 Thread Stuart Henderson 
Nothing changed in dxx-rebirth itself recently, possibly due to
libc++/llvm updates.

ld: error: undefined symbol: dxx_error_object_type_mismatch()
>>> referenced by collide.cpp
>>>   
>>> similar/main/.d1x-rebirth.collide.o:(bump_this_object(valptridx::ptridx::vm>, valptridx::ptridx::vm>, 
dcx::vms_vector const&, int))
c++: error: linker command failed with exit code 1 (use -v to see invocation)
scons: *** [d1x-rebirth/d1x-rebirth] Error 1
scons: building terminated because of errors.
Failed target count: total=1; targets with enable_build_failure_summary=1: 1
Failed node list:

i386 build failures

2018-04-24 Thread Stuart Henderson 
various kde4 things that probably also fail on amd64:

./x11/kde4/bovo
./x11/kde4/kopete
./x11/kde4/kross-interpreters
./x11/kde4/kuser
./x11/kde4/p5-qt
./x11/kde4/pim-runtime
./x11/kde4/py-kde
./x11/kde4/rocs
./x11/kde4/superkaramba

others are more likely ILP32-specific:

./cad/kicad: likely to be fixed in newer versions, but from previous
ports@ mails it seems that they may need things in boost that we don't
have (ucontext-related perhaps?)

/usr/obj/ports/kicad-20100505/bin/c++  -D_FILE_OFFSET_BITS=64 -D_IODBC_ 
-D_LARGE_FILES -D__WXGTK__ -I/usr/local/lib/wx/include/gtk2-unicode-release-2.8 
-I/usr/local/inc
lude -I/usr/X11R6/include 
-I/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/include 
-I/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/shar
e -I. -I/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common 
-I/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/../3d-viewer
 -I/us
r/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/../pcbnew 
-I/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/../polygon
 -O2 -pip
e -pthread -DNDEBUG -MD -MT common/CMakeFiles/pcbcommon.dir/pcbcommon.cpp.o -MF 
common/CMakeFiles/pcbcommon.dir/pcbcommon.cpp.o.d -o 
common/CMakeFiles/pcbcommon.dir/pcb
common.cpp.o -c 
/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/pcbcommon.cpp
/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/pcbcommon.cpp:48:41:
 error: constant expression evaluates to 2147483648 which cannot be narrowed t
o type 'int' [-Wc++11-narrowing]
0x1000, 0x2000, 0x4000, 0x8000
^~
/usr/obj/ports/kicad-20100505/kicad-sources-2010-05-05-BZR2356/common/pcbcommon.cpp:48:41:
 note: insert an explicit cast to silence this issue
0x1000, 0x2000, 0x4000, 0x8000
^~
static_cast( )
1 error generated.
ninja: build stopped: subcommand failed.


./graphics/aqsis: linking fails with undefined references linking to boost

./lang/sbcl: fails randomly, I think.

debugger invoked on a SB-INT:BUG:
failed AVER: (EQ SB-VM::START SB-VM::END)
  This is probably a bug in SBCL itself. (Alternatively, SBCL might have been
  corrupted by bad user code, e.g. by an undefined Lisp operation like
  (FMAKUNBOUND 'COMPILE), or by stray pointers from alien code or from unsafe
  Lisp code; or there might be a bug in the OS or hardware that SBCL is running
  on.) If it seems to be a bug in SBCL itself, the maintainers would like to
  know about it. Bug reports are welcome on the SBCL mailing lists, which you
  can find at .
  2: [ABORT   ] Abort loading file 
"/usr/obj/ports/sbcl-1.4.5/sbcl-1.4.5/make-target-2-load.lisp".
  3:Abort building SBCL.
  4:Ignore runtime option --eval "(sb-fasl::!warm-load 
\"make-target-2-load.lisp\")".
  5:Skip rest of --eval and --load options.
  6:Skip to toplevel READ/EVAL/PRINT loop.
  7: [EXIT] Exit SBCL (calling #'EXIT, killing the process).


./x11/kde/libs3:

In file included from libkjs_html_la.all_cpp.cpp:7:
/usr/obj/ports/kdelibs-3.5.10/kdelibs-3.5.10/khtml/ecma/kjs_proxy.cpp:391:22: 
error: non-constant-expression cannot be narrowed from type 'unsigned int' to 
'suseconds_t' (aka 'long') in initializer list [-Wc++11-narrowing]
  { i_ms / 1000, (i_ms % 1000) * 1000 },
 ^~~~
/usr/obj/ports/kdelibs-3.5.10/kdelibs-3.5.10/khtml/ecma/kjs_proxy.cpp:391:22: 
note: insert an explicit cast to silence this issue
  { i_ms / 1000, (i_ms % 1000) * 1000 },
 ^~~~
 static_cast( )
/usr/obj/ports/kdelibs-3.5.10/kdelibs-3.5.10/khtml/ecma/kjs_proxy.cpp:392:20: 
error: non-constant-expression cannot be narrowed from type 'unsigned int' to 
'suseconds
_t' (aka 'long') in initializer list [-Wc++11-narrowing]
  { ms / 1000, (ms % 1000) * 1000 }
   ^~
/usr/obj/ports/kdelibs-3.5.10/kdelibs-3.5.10/khtml/ecma/kjs_proxy.cpp:392:20: 
note: insert an explicit cast to silence this issue
  { ms / 1000, (ms % 1000) * 1000 }
   ^~
   static_cast( )

i386 build failures

2017-08-28 Thread Stuart Henderson 
https://junkpile.org/p/failures/i386-20170828/devel/ti-msp430gcc.log
/usr/include/c++/v1/exception:267:5: error: no member named 'fancy_abort' in 
namespace 'std::__1'; did you mean simply 'fancy_abort'?

https://junkpile.org/p/failures/i386-20170828/editors/xemacs21/stable.log
"normal" intermittent hang, not clang

https://junkpile.org/p/failures/i386-20170828/editors/TeXmacs.log
undefined references

https://junkpile.org/p/failures/i386-20170828/emulators/emulationstation.log
/usr/include/c++/v1/utility:308:15: error: no matching constructor for 
initialization of 'std::__1::unique_ptr'

https://junkpile.org/p/failures/i386-20170828/lang/jamvm.log
clang ICE

https://junkpile.org/p/failures/i386-20170828/lang/STk.log
invalid use of a cast in a inline asm context requiring an l-value: remove the 
cast or build with -fheinous-gnu-extensions

https://junkpile.org/p/failures/i386-20170828/lang/oo2c.log
./lib/obj/ADT/ArrayList.c:518:3: error: illegal storage class on function

https://junkpile.org/p/failures/i386-20170828/lang/smlnj.log
undefined references

https://junkpile.org/p/failures/i386-20170828/multimedia/avidemux.log
i386/fft_sse.c:172:43: error: unknown token in expression
"movaps  %1, %%xmm3 \n\t"   // xmm3 = X  i1 X  i0: in1

https://junkpile.org/p/failures/i386-20170828/multimedia/transcode,lzo,mjpegtools.log
/usr/obj/ports/transcode-1.1.7-lzo-mjpegtools/transcode-1.1.7/filter/yuvdenoise/motion.c:353:4:
 error: no matching '.endr' in definition
  " pxor %%mm0 , %%mm0;  /* clear mm0   
   */"
  ^
:1:466: note: instantiated into assembly here

https://junkpile.org/p/failures/i386-20170828/sysutils/grub.log
checking for C compiler default output file name... configure: error: C 
compiler cannot create executables

https://junkpile.org/p/failures/i386-20170828/www/mongrel2.log
sh ./tests/runtests.sh
Running unit tests:
Segmentation fault (core dumped)
ERROR in test tests/cli_tests:

(plus the ones which have already been marked BROKEN).

i386 build failures

2016-08-18 Thread Stuart Henderson 
not W^X

devel/lazarus   random build failure on i386

W^X enforcement

lang/hugs   ERROR - mallocBytesRWX: failed to protect 0x0x7becea20
lang/sbcl   mmap: Not supported
lang/obcmprotect failed: Not supported

editors/libreoffice (with USE_WXNEEDED) - unsure if this is W^X enforcement
or another problem:

[...]
S=/usr/obj/ports/libreoffice-5.1.4.2/libreoffice-5.1.4.2 && I=$S/instdir && 
W=$S/workdir &&  mkdir -p $W/JavaClassSet/Jar/report/ &
&  rm -rf $W/JavaClassSet/Jar/report/* && RESPONSEFILE=/tmp/gbuild.ybliBP &&  
/usr/local/jdk-1.7.0/bin/javac -J-Xmx128M -source 1.5
 -target 1.5 -Xlint:-options  -classpath 
"$I/program/classes/ridl.jar:$I/program/classes/unoil.jar:$I/program/classes/jurt.jar:$I/p
rogram/classes/juh.jar:$I/program/classes/java_uno.jar:$I/program/classes/commonwizards.jar:$W/JavaClassSet/Jar/report"
 -d $W/JavaC
lassSet/Jar/report @$RESPONSEFILE && rm -f $RESPONSEFILE && touch 
$W/JavaClassSet/Jar/report/done
[build JCS] Jar/table
S=/usr/obj/ports/libreoffice-5.1.4.2/libreoffice-5.1.4.2 && I=$S/instdir && 
W=$S/workdir &&  mkdir -p $W/JavaClassSet/Jar/table/ &&
  rm -rf $W/JavaClassSet/Jar/table/* && RESPONSEFILE=/tmp/gbuild.77y9Tv &&  
/usr/local/jdk-1.7.0/bin/javac -J-Xmx128M -source 1.5 -
target 1.5 -Xlint:-options  -classpath 
"$I/program/classes/ridl.jar:$I/program/classes/unoil.jar:$I/program/classes/jurt.jar:$I/pro
gram/classes/juh.jar:$I/program/classes/java_uno.jar:$I/program/classes/commonwizards.jar:$W/JavaClassSet/Jar/table"
 -d $W/JavaClas
sSet/Jar/table @$RESPONSEFILE && rm -f $RESPONSEFILE && touch 
$W/JavaClassSet/Jar/table/done
[build JCS] Jar/XMergeBridge
S=/usr/obj/ports/libreoffice-5.1.4.2/libreoffice-5.1.4.2 && I=$S/instdir && 
W=$S/workdir &&  mkdir -p $W/JavaClassSet/Jar/XMergeBri
dge/ &&  rm -rf $W/JavaClassSet/Jar/XMergeBridge/* && 
RESPONSEFILE=/tmp/gbuild.ECsFpl &&  /usr/local/jdk-1.7.0/bin/javac -J-Xmx128M
 -source 1.5 -target 1.5 -Xlint:-options  -classpath 
"$I/program/classes/xmerge.jar:$I/program/classes/unoil.jar:$I/program/classes
/ridl.jar:$I/program/classes/jurt.jar:$I/program/classes/juh.jar:$W/JavaClassSet/Jar/XMergeBridge"
 -d $W/JavaClassSet/Jar/XMergeBri
dge @$RESPONSEFILE && rm -f $RESPONSEFILE && touch 
$W/JavaClassSet/Jar/XMergeBridge/done
S=/usr/obj/ports/libreoffice-5.1.4.2/libreoffice-5.1.4.2 && I=$S/instdir && 
W=$S/workdir &&  LD_LIBRARY_PATH=${LD_LIBRARY_PATH:+$LD
_LIBRARY_PATH:}"$I/program:$I/program"   $I/program/uno.bin -s 
com.sun.star.test.bridge.BridgeTest -- com.sun.star.test.bridge.CppT
estObject -env:LO_BUILD_LIB_DIR=file://$W/LinkTarget/Library 
-env:URE_MORE_SERVICES=file://$W/Rdb/uno_services.rdb -env:URE_MORE_TY
PES=file://$W/UnoApiTarget/bridgetest.rdb
[build LNK] CppunitTest/libtest_tools_test.so
mkstemp("/nonexistent/.execooolO58Ku") failed: No such file or directory
terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc
Abort trap (core dumped)

Re: i386 build failures: claws-mail [encrypt()], emacs21, mongodb, firefox/seamonkey

2014-12-10 Thread Stuart Henderson 
From the last build:

mail/claws-mail

Building with GnuTLS
checking for GNUTLS... yes
checking for library containing encrypt... no
configure: error: 'encrypt'-function not found.

editors/emacs21

LC_ALL=C ./temacs -batch -l loadup dump
Segmentation fault (core dumped) 

databases/mongodb

scons: *** 
[build/openbsd5/use-system-boost/use-system-pcre/use-system-v8/usev8/mongo/util/options_parser/options_parser.o]
 /usr/local/include/yaml-cpp/emitter.h: No such file or directory
scons: building terminated because of errors.

www/mozilla-firefox
www/seamonkey

SSE4 skia patches lost again

Re: i386 build failures: claws-mail [encrypt()], emacs21, mongodb, firefox/seamonkey

2014-12-10 Thread Christian Weisgerber 
On 2014-12-10, Stuart Henderson st...@openbsd.org wrote:

mail/claws-mail

 configure: error: 'encrypt'-function not found.

encrypt() has been removed from libc.

databases/mongodb

 scons: *** 
 [build/openbsd5/use-system-boost/use-system-pcre/use-system-v8/usev8/mongo/util/options_parser/options_parser.o]
  /usr/local/include/yaml-cpp/emitter.h: No such file or directory

That happens all the time.  Presumably some interaction with
devel/yaml-cpp.

-- 
Christian naddy Weisgerber  na...@mips.inka.de

Re: i386 build failures: claws-mail [encrypt()], emacs21, mongodb, firefox/seamonkey

2014-12-10 Thread David Coppa 

 From: Christian Weisgerber na...@mips.inka.de
 Date: Wed, Dec 10, 2014 at 4:18 PM
 Subject: Re: i386 build failures: claws-mail [encrypt()], emacs21,
 mongodb, firefox/seamonkey
 To: ports@openbsd.org
 
 
 On 2014-12-10, Stuart Henderson st...@openbsd.org wrote:
 
 mail/claws-mail
 
  configure: error: 'encrypt'-function not found.
 
 encrypt() has been removed from libc.
 
 databases/mongodb
 
  scons: *** 
  [build/openbsd5/use-system-boost/use-system-pcre/use-system-v8/usev8/mongo/util/options_parser/options_parser.o]
   /usr/local/include/yaml-cpp/emitter.h: No such file or directory
 
 That happens all the time.  Presumably some interaction with
 devel/yaml-cpp.

What about the following diff?

Index: Makefile
===
RCS file: /cvs/ports/databases/mongodb/Makefile,v
retrieving revision 1.16
diff -u -p -u -p -r1.16 Makefile
--- Makefile22 Oct 2014 20:11:27 -  1.16
+++ Makefile10 Dec 2014 16:03:28 -
@@ -9,7 +9,7 @@ COMMENT =   scalable, high-performance doc
 
 DISTNAME = mongodb-src-r2.6.4
 PKGNAME =  ${DISTNAME:S/src-r//}
-
+REVISION = 0
 CATEGORIES =   databases
 
 HOMEPAGE = http://www.mongodb.org/
@@ -19,7 +19,7 @@ PERMIT_PACKAGE_CDROM =Yes
 
 WANTLIB =  boost_filesystem-mt boost_program_options-mt boost_system-mt \
boost_thread-mt c execinfo m pcap pcre pcrecpp pthread \
-   stdc++ v8
+   stdc++ v8 yaml-cpp
 
 MASTER_SITES = http://downloads.mongodb.org/src/
 
@@ -32,7 +32,8 @@ MODSCONS_FLAGS = --prefix=${PREFIX} \
 --full \
 --use-system-pcre \
 --use-system-boost \
---use-system-v8
+--use-system-v8 \
+--use-system-yaml
 
 MAKE_ENV +=HOME=${HOME} \
TERM=${TERM} \
@@ -41,6 +42,7 @@ MAKE_ENV +=   HOME=${HOME} \
 LIB_DEPENDS =  devel/boost \
devel/libexecinfo \
devel/pcre \
+   devel/yaml-cpp \
lang/libv8
 
 pre-configure:
Index: patches/patch-SConstruct
===
RCS file: /cvs/ports/databases/mongodb/patches/patch-SConstruct,v
retrieving revision 1.4
diff -u -p -u -p -r1.4 patch-SConstruct
--- patches/patch-SConstruct16 Sep 2014 17:28:07 -  1.4
+++ patches/patch-SConstruct10 Dec 2014 16:03:28 -
@@ -1,6 +1,10 @@
 $OpenBSD: patch-SConstruct,v 1.4 2014/09/16 17:28:07 jasper Exp $
 SConstruct.origFri Aug  8 22:13:51 2014
-+++ SConstruct Tue Sep 16 16:08:02 2014
+
+Fixup detection of yaml-cpp system library
+https://github.com/mongodb/mongo/commit/44b0e3591839525a0c951319540488814e5c25fd
+
+--- SConstruct.origFri Aug  8 15:13:51 2014
 SConstruct Wed Dec 10 09:28:50 2014
 @@ -814,7 +814,6 @@ if nix:
  # -Winvalid-pch Warn if a precompiled header (see Precompiled Headers) is 
found in the search path but can't be used.
  env.Append( CCFLAGS=[-fPIC,
@@ -33,3 +37,12 @@ $OpenBSD: patch-SConstruct,v 1.4 2014/09
  env.Append( LINKFLAGS=[-fstack-protector] )
  env.Append( SHLINKFLAGS=[-fstack-protector] )
  env['ENV']['GLIBCXX_FORCE_NEW'] = 1; # play nice with valgrind
+@@ -1352,7 +1347,7 @@ def doConfigure(myenv):
+ conf.FindSysLibDep(stemmer, [stemmer])
+ 
+ if use_system_version_of_library(yaml):
+-conf.FindSysLibDep(yaml, [yaml])
++conf.FindSysLibDep(yaml, [yaml-cpp])
+ 
+ if use_system_version_of_library(boost):
+ if not conf.CheckCXXHeader( boost/filesystem/operations.hpp ):

Re: i386 build failures: claws-mail [encrypt()], emacs21, mongodb, firefox/seamonkey

2014-12-10 Thread Landry Breuil 
On Wed, Dec 10, 2014 at 01:26:41PM +, Stuart Henderson wrote:
 From the last build:
 
 mail/claws-mail
 
 Building with GnuTLS
 checking for GNUTLS... yes
 checking for library containing encrypt... no
 configure: error: 'encrypt'-function not found.
 
 editors/emacs21
 
 LC_ALL=C ./temacs -batch -l loadup dump
 Segmentation fault (core dumped) 
 
 databases/mongodb
 
 scons: *** 
 [build/openbsd5/use-system-boost/use-system-pcre/use-system-v8/usev8/mongo/util/options_parser/options_parser.o]
  /usr/local/include/yaml-cpp/emitter.h: No such file or directory
 scons: building terminated because of errors.
 
 www/mozilla-firefox
 www/seamonkey
 
 SSE4 skia patches lost again

Already fixed, or it still fails ? built for me with what i commited
yesterday.

Landry

Re: i386 build failures: claws-mail [encrypt()], emacs21, mongodb, firefox/seamonkey

2014-12-10 Thread Stuart Henderson 
On 2014/12/10 18:55, Landry Breuil wrote:
  www/mozilla-firefox
  www/seamonkey
  
  SSE4 skia patches lost again
 
 Already fixed, or it still fails ? built for me with what i commited
 yesterday.

I'll know after my next build :)

Re: i386 build failures

2014-11-27 Thread Christian Weisgerber 
On 2014-11-27, Stuart Henderson st...@openbsd.org wrote:

 I've seen these on multiple builds, though cleaning and rebuilding the
 affected ports (including sdl-ttf in the last case) has eventually got
 them to package.

 x11/kde/webdev(i386) ... seems to be some conflict between kde4 and kde3.

This affects various KDE3 ports.  Recently, it's been x11/kde/edu3
in my bulk builds.

-- 
Christian naddy Weisgerber  na...@mips.inka.de

i386 build failures

2014-11-26 Thread Stuart Henderson 
I've seen these on multiple builds, though cleaning and rebuilding the
affected ports (including sdl-ttf in the last case) has eventually got
them to package.


x11/kde/webdev(i386) ... seems to be some conflict between kde4 and kde3.

/usr/bin/ld: warning: libkdecore.so.7.0, needed by 
/usr/local/lib/libkwalletclient.so.2.0, may conflict with libkdecore.so.50.2
/usr/X11R6/lib/libGL.so.15.0: warning: random() isn't random; consider using 
arc4random()
kfilereplace.all_cpp.o(.text+0x77): In function `KFileReplace::qt_emit(int, 
QUObject*)':
: undefined reference to `KParts::MainWindow::qt_emit(int, QUObject*)'
kfilereplace.all_cpp.o(.text+0xbe): In function `KFileReplace::qt_property(int, 
int, QVariant*)':
: undefined reference to `KParts::MainWindow::qt_property(int, int, QVariant*)'
  * lots


graphics/gwenview(i386) ... ???
./.libs/libkdeinit_gwenview.so: undefined reference to `KMultiTabBar::tabs()'
./.libs/libkdeinit_gwenview.so: undefined reference to 
`KMultiTabBar::KMultiTabBar(KMultiTabBar::KMultiTabBarMode, QWidget*, char 
const*)'


games/pushover(i386) ... libSDL_ttf.la somehow ended up containing this line:
-
dependency_libs='-R/usr/X11R6/lib -pthread -L/usr/local/lib -L/usr/X11R6/lib 
-lfreetype -lz /usr/obj/ports/sdl-1.2.15/fake-i386/usr/local/lib/libSDL.la 
/usr/local/lib/libiconv.la -lm -lsndio -lusbhid'
-
Unsure how sdl-ttf got built like that; the build command lines in the
broken version had an extra
-rpath,/usr/obj/ports/sdl-1.2.15/fake-i386/usr/local/lib which wasn't
present in the working build. In any event this resulted in pushover build
dying after failure to find libSDL.la:

/usr/bin/libtool --tag=CXX--mode=link c++ -Wall 
-DPKGDATADIR=\/usr/local/share/pushover\ 
-DLOCALEDIR=\/usr/local/share/locale\  -O2 -pipe  -L/usr/local/lib -o 
pushover pushover-ant.o  pushover-graphics.o pushover-graphicsn.o  
pushover-leveldata.o pushover-leveldisplay.o  pushover-levelplayer.o 
pushover-levelset.o  pushover-luaclass.o pushover-main.o  pushover-pngloader.o 
pushover-recorder.o  pushover-screen.o pushover-sha1.o  pushover-solvedmap.o 
pushover-soundsys.o  pushover-textsections.o pushover-tools.o  
pushover-window.o  -lintl -L/usr/lib -L/usr/local/lib -pthread -L/usr/X11R6/lib 
-R/usr/X11R6/lib -lz -lpng -lSDL_mixer -lSDL_ttf -lSDL -llua5.1 -lm
Cannot read /usr/obj/ports/sdl-1.2.15/fake-i386/usr/local/lib/libSDL.la: No 
such file or directory
 at /usr/libdata/perl5/LT/LaLoFile.pm line 50.
[..]

lang/ghc again (was: i386 build failures)

2014-09-16 Thread David Coppa 

 From: Matthias Kilian k...@outback.escape.de
 Date: Thu, Jan 30, 2014 at 10:18 PM
 Subject: lang/ghc (was: i386 build failures)
 To: ports ports@openbsd.org
 
 
 On Tue, Jan 28, 2014 at 08:36:30PM +0100, Matthias Kilian wrote:
   Though OTOH, the port is built with -fno-pie -nopie, so should it even
   be necessary to support RELOC_GOTPC? Is it building the internal copy of 
   gmp
   incorrectly or something?
 
  After looking at the build logs: yes, there's no trace of no-?pie
  durinng the build of the bundled libgmp.
 
  Also, objdump -r libHSinteger-gmp-0.5.0.0.a shows entries like
 
R_386_GOTPC   _GLOBAL_OFFSET_TABLE_
 
  for all the object files from gmp.
 
  I'll try to fix that part, first (when my current test bulk build
  finished, may be today or tomorrow).
 
 new trouble happened:
 
 $ ghci
 GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
 Loading package ghc-prim ... linking ... done.
 Loading package integer-gmp ... linking ... done.
 Loading package base ...
 
 GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
 Loading package ghc-prim ... linking ... done.
 Loading package integer-gmp ... linking ... done.
 Loading package base ...
 
 GHCi runtime linker: fatal error: I found a duplicate definition for symbol
__i686.get_pc_thunk.bx
 whilst processing object file
/usr/local/lib/libiconv.a
 This could be caused by:
* Loading two different object files which export the same symbol
* Specifying the same object file twice on the GHCi command line
* An incorrect `package.conf' entry, causing some object to be
  loaded twice.
 GHCi cannot safely continue in this situation.  Exiting now.  Sorry.
 
 So I've to spend a little bit more work on this :-(
 
 (or touch that fucking rts/Linker.c, but that one is going to die in
 future releases of ghc, anyway, so any work on it is a waste of time)

Please, can commit the diff below?

---8---

Index: Makefile
===
RCS file: /cvs/ports/lang/ghc/Makefile,v
retrieving revision 1.107
diff -u -p -r1.107 Makefile
--- Makefile12 Dec 2013 22:13:37 -  1.107
+++ Makefile16 Sep 2014 07:19:03 -
@@ -11,7 +11,7 @@ COMMENT-doc = documentation for GHC
 
 DISTNAME = ghc-${MODGHC_VER}
 PKGNAME-main = ghc-${MODGHC_VER}
-REVISION-main =1
+REVISION-main =2
 PKGNAME-doc =  ghc-doc-${MODGHC_VER}
 REVISION-doc = 0
 CATEGORIES =   lang devel
@@ -64,6 +64,10 @@ SUPDISTFILES +=  ${BINDISTFILE-$m}
 
 USE_GMAKE =Yes
 USE_GROFF =Yes
+
+.if ${MACHINE_ARCH} == i386
+CFLAGS +=  -fno-pie
+.endif
 
 CONFIGURE_STYLE =  gnu
 CONFIGURE_ARGS +=  --with-iconv-includes=${LOCALBASE}/include \

---8---

With this in, I can have a working ghci to do some Haskell
tests/exercises on my i386-only netbook, using a quick trick.

It does not cause any harm, after all...

The quick dirty trick is:

---8---

Index: Makefile
===
RCS file: /cvs/ports/converters/libiconv/Makefile,v
retrieving revision 1.47
diff -u -p -r1.47 Makefile
--- Makefile9 Nov 2013 23:19:02 -   1.47
+++ Makefile16 Sep 2014 07:19:54 -
@@ -3,7 +3,7 @@
 COMMENT=   character set conversion library
 
 DISTNAME=  libiconv-1.14
-REVISION=  1
+REVISION=  2
 CATEGORIES=converters devel
 MASTER_SITES=  ${MASTER_SITE_GNU:=libiconv/}
 
@@ -21,6 +21,11 @@ WANTLIB= c
 BUILD_DEPENDS= devel/gperf
 
 MAKE_FLAGS=INSTALL_DATA=cp
+
+# XXX: unbreak lang/ghc on i386
+.if ${MACHINE_ARCH} == i386
+CFLAGS +=  -fno-pie
+.endif
 
 SEPARATE_BUILD=Yes
 CONFIGURE_STYLE= gnu

---8---

Rebuilding libiconv is faster than rebuilding ghc for sure ;)

Ciao!
David

Re: lang/ghc again (was: i386 build failures)

2014-09-16 Thread David Coppa 
On Tue, Sep 16, 2014 at 10:14 AM, Matthias Kilian k...@gaertner.de wrote:
 On Tue, Sep 16, 2014 at 01:37:44AM -0600, David Coppa wrote:
  GHCi cannot safely continue in this situation.  Exiting now.  Sorry.
 
  So I've to spend a little bit more work on this :-(
 
  (or touch that fucking rts/Linker.c, but that one is going to die in
  future releases of ghc, anyway, so any work on it is a waste of time)

 Please, can commit the diff below?

 The one for ghc? No objections from me.

Yeah, that one.

 Ciao,
 Kili

Ciao,
David

 Index: Makefile
 ===
 RCS file: /cvs/ports/lang/ghc/Makefile,v
 retrieving revision 1.107
 diff -u -p -r1.107 Makefile
 --- Makefile  12 Dec 2013 22:13:37 -  1.107
 +++ Makefile  16 Sep 2014 07:19:03 -
 @@ -11,7 +11,7 @@ COMMENT-doc =   documentation for GHC

  DISTNAME =   ghc-${MODGHC_VER}
  PKGNAME-main =   ghc-${MODGHC_VER}
 -REVISION-main =  1
 +REVISION-main =  2
  PKGNAME-doc =ghc-doc-${MODGHC_VER}
  REVISION-doc =   0
  CATEGORIES = lang devel
 @@ -64,6 +64,10 @@ SUPDISTFILES +=${BINDISTFILE-$m}

  USE_GMAKE =  Yes
  USE_GROFF =  Yes
 +
 +.if ${MACHINE_ARCH} == i386
 +CFLAGS +=-fno-pie
 +.endif

  CONFIGURE_STYLE =gnu
  CONFIGURE_ARGS +=--with-iconv-includes=${LOCALBASE}/include \

Re: lang/ghc again (was: i386 build failures)

2014-09-16 Thread Matthias Kilian 
On Tue, Sep 16, 2014 at 01:37:44AM -0600, David Coppa wrote:
  GHCi cannot safely continue in this situation.  Exiting now.  Sorry.
  
  So I've to spend a little bit more work on this :-(
  
  (or touch that fucking rts/Linker.c, but that one is going to die in
  future releases of ghc, anyway, so any work on it is a waste of time)
 
 Please, can commit the diff below?

The one for ghc? No objections from me.

Ciao,
Kili

 ---8---
 
 Index: Makefile
 ===
 RCS file: /cvs/ports/lang/ghc/Makefile,v
 retrieving revision 1.107
 diff -u -p -r1.107 Makefile
 --- Makefile  12 Dec 2013 22:13:37 -  1.107
 +++ Makefile  16 Sep 2014 07:19:03 -
 @@ -11,7 +11,7 @@ COMMENT-doc =   documentation for GHC
  
  DISTNAME =   ghc-${MODGHC_VER}
  PKGNAME-main =   ghc-${MODGHC_VER}
 -REVISION-main =  1
 +REVISION-main =  2
  PKGNAME-doc =ghc-doc-${MODGHC_VER}
  REVISION-doc =   0
  CATEGORIES = lang devel
 @@ -64,6 +64,10 @@ SUPDISTFILES +=${BINDISTFILE-$m}
  
  USE_GMAKE =  Yes
  USE_GROFF =  Yes
 +
 +.if ${MACHINE_ARCH} == i386
 +CFLAGS +=-fno-pie
 +.endif
  
  CONFIGURE_STYLE =gnu
  CONFIGURE_ARGS +=--with-iconv-includes=${LOCALBASE}/include \
 
 ---8---

Re: lang/ghc again (was: i386 build failures)

2014-09-16 Thread David Coppa 
On Tue, Sep 16, 2014 at 10:14 AM, Matthias Kilian k...@gaertner.de wrote:
 On Tue, Sep 16, 2014 at 01:37:44AM -0600, David Coppa wrote:
  GHCi cannot safely continue in this situation.  Exiting now.  Sorry.
 
  So I've to spend a little bit more work on this :-(
 
  (or touch that fucking rts/Linker.c, but that one is going to die in
  future releases of ghc, anyway, so any work on it is a waste of time)

 Please, can commit the diff below?

 The one for ghc? No objections from me.

Since we're here,
what would be needed to have this[1] working for our beloved OS?

[1] https://ghc.haskell.org/trac/ghc/wiki/DynamicByDefault

Cheers!
David

Re: lang/ghc again (was: i386 build failures)

2014-09-16 Thread Matthias Kilian 
On Tue, Sep 16, 2014 at 03:53:45PM +0200, David Coppa wrote:
 Since we're here,
 what would be needed to have this[1] working for our beloved OS?
 
 [1] https://ghc.haskell.org/trac/ghc/wiki/DynamicByDefault

First: me unslacking and updating ghc ;-)

Ciao,
Kili

lang/ghc (was: i386 build failures)

2014-01-30 Thread Matthias Kilian 
On Tue, Jan 28, 2014 at 08:36:30PM +0100, Matthias Kilian wrote:
  Though OTOH, the port is built with -fno-pie -nopie, so should it even
  be necessary to support RELOC_GOTPC? Is it building the internal copy of gmp
  incorrectly or something?
 
 After looking at the build logs: yes, there's no trace of no-?pie
 durinng the build of the bundled libgmp.
 
 Also, objdump -r libHSinteger-gmp-0.5.0.0.a shows entries like
 
   R_386_GOTPC   _GLOBAL_OFFSET_TABLE_
 
 for all the object files from gmp.
 
 I'll try to fix that part, first (when my current test bulk build
 finished, may be today or tomorrow).

Bah!

with this diff (to build libgmp and maybe other bundled libs
with -fno-pie -nopie):

Index: Makefile
===
RCS file: /cvs/ports/lang/ghc/Makefile,v
retrieving revision 1.107
diff -u -p -r1.107 Makefile
--- Makefile12 Dec 2013 22:13:37 -  1.107
+++ Makefile30 Jan 2014 15:54:06 -
@@ -65,6 +65,8 @@ SUPDISTFILES +=   ${BINDISTFILE-$m}
 USE_GMAKE =Yes
 USE_GROFF =Yes
 
+CFLAGS +=  -fno-pie -nopie
+
 CONFIGURE_STYLE =  gnu
 CONFIGURE_ARGS +=  --with-iconv-includes=${LOCALBASE}/include \
--with-iconv-libraries=${LOCALBASE}/lib

new trouble happened:

$ ghci
GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ...

GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ...

GHCi runtime linker: fatal error: I found a duplicate definition for symbol
   __i686.get_pc_thunk.bx
whilst processing object file
   /usr/local/lib/libiconv.a
This could be caused by:
   * Loading two different object files which export the same symbol
   * Specifying the same object file twice on the GHCi command line
   * An incorrect `package.conf' entry, causing some object to be
 loaded twice.
GHCi cannot safely continue in this situation.  Exiting now.  Sorry.

So I've to spend a little bit more work on this :-(

(or touch that fucking rts/Linker.c, but that one is going to die in
future releases of ghc, anyway, so any work on it is a waste of time)

Ciao,
Kili

Re: i386 build failures

2014-01-28 Thread Stuart Henderson 
On 2014/01/27 23:02, Matthias Kilian wrote:
 On Mon, Jan 27, 2014 at 12:51:07PM +, Stuart Henderson wrote:
  PIE?:
  devel/hs-type-level, devel/hs-vector
 
 Not sure about this. but why does it work on amd64? The patches in
 lang/ghc for PIE don't look arch-dependent.

i386 uses REL relocations (do_Elf_Rel_relocations in ghc's rts/Linker.c)
whereas amd64 does RELA (do_Elf_Rela_relocations).

do_Elf_Rel_relocations currently has:

  switch (ELF_R_TYPE(info)) {
#ifdef i386_HOST_ARCH
 case R_386_32:   *pP = value; break;
 case R_386_PC32: *pP = value - P; break;
#endif

R_386_32 and R_386_PC32 are #defined to RELOC_32 and RELOC_PC32
respectively.

from i386/include/reloc.h:

#define RELOC_321   /* Direct 32 bit  */
#define RELOC_PC32  2   /* PC relative 32 bit */

and type 10 is what the build is complaining about:

Loading package integer-gmp ... linking ... ghc: 
/usr/local/lib/ghc/integer-gmp-0.5.0.0/HSinteger-gmp-0.5.0.0.o: unhandled ELF 
relocation(Rel) type 10

#define RELOC_GOTPC 10  /* 32 bit PC relative offset to GOT */

So I think ghc's Linker.c also needs to support RELOC_GOTPC. There's
some code in do_Elf_Rela_relocations like this which I think covers
the same sort of case so it might be possible to adapt:

  case R_X86_64_GOTPCREL:
  {
  StgInt64 gotAddress = (StgInt64) makeSymbolExtra(oc, 
ELF_R_SYM(info), S)-addr;
  StgInt64 off = gotAddress + A - P;
  *(Elf64_Word *)P = (Elf64_Word)off;
  break;
  }

Though OTOH, the port is built with -fno-pie -nopie, so should it even
be necessary to support RELOC_GOTPC? Is it building the internal copy of gmp
incorrectly or something?

Re: i386 build failures

2014-01-28 Thread Stuart Henderson 
On 2014/01/27 11:53, David Coppa wrote:
 
  Da: Stuart Henderson st...@openbsd.org
  Data: 27/gen/2014 13.51
  Oggetto: i386 build failures
  A: ports ports@openbsd.org
 
  www/chromium: intermittent (dependency issue?)
  In file included from ../../chrome/browser/ui/webui/policy_ui.cc:20:
  ../../chrome/browser/policy/cloud/cloud_policy_client.h:18:10: fatal error:
  'chrome/browser/policy/proto/cloud/device_management_backend.pb.h' file not
  found
 
 Not 100% sure, but this should fix the dependency issue:

Thanks, I'll throw this into my next bulk, but I don't see a downside so
I think just commit it unless Robert disagrees.


 Index: patches/patch-chrome_chrome_browser_ui_gypi
 ===
 RCS file: patches/patch-chrome_chrome_browser_ui_gypi
 diff -N patches/patch-chrome_chrome_browser_ui_gypi
 --- /dev/null 1 Jan 1970 00:00:00 -
 +++ patches/patch-chrome_chrome_browser_ui_gypi   27 Jan 2014 18:52:44 
 -
 @@ -0,0 +1,11 @@
 +$OpenBSD$
 +--- chrome/chrome_browser_ui.gypi.orig.port  Mon Jan 27 15:59:23 2014
  chrome/chrome_browser_ui.gypiMon Jan 27 16:23:52 2014
 +@@ -2647,6 +2647,7 @@
 + ['configuration_policy==1', {
 +   'dependencies': [
 + 'app/policy/cloud_policy_codegen.gyp:policy',
 ++'cloud_policy_proto',
 +   ],
 + }, {  # configuration_policy==0
 +   'sources/': [

Re: i386 build failures

2014-01-28 Thread Matthias Kilian 
On Tue, Jan 28, 2014 at 12:04:45PM +, Stuart Henderson wrote:
  Not sure about this. but why does it work on amd64? The patches in
  lang/ghc for PIE don't look arch-dependent.
 
 i386 uses REL relocations (do_Elf_Rel_relocations in ghc's rts/Linker.c)
 whereas amd64 does RELA (do_Elf_Rela_relocations).
 
 do_Elf_Rel_relocations currently has:
 
   switch (ELF_R_TYPE(info)) {
 #ifdef i386_HOST_ARCH
  case R_386_32:   *pP = value; break;
  case R_386_PC32: *pP = value - P; break;
 #endif
 
 R_386_32 and R_386_PC32 are #defined to RELOC_32 and RELOC_PC32
 respectively.
 
 from i386/include/reloc.h:
 
 #define RELOC_321   /* Direct 32 bit  */
 #define RELOC_PC32  2   /* PC relative 32 bit */
 
 and type 10 is what the build is complaining about:
 
 Loading package integer-gmp ... linking ... ghc: 
 /usr/local/lib/ghc/integer-gmp-0.5.0.0/HSinteger-gmp-0.5.0.0.o: unhandled ELF 
 relocation(Rel) type 10
 
 #define RELOC_GOTPC 10  /* 32 bit PC relative offset to GOT */
 
 So I think ghc's Linker.c also needs to support RELOC_GOTPC. There's
 some code in do_Elf_Rela_relocations like this which I think covers
 the same sort of case so it might be possible to adapt:
 
   case R_X86_64_GOTPCREL:
   {
   StgInt64 gotAddress = (StgInt64) makeSymbolExtra(oc, 
 ELF_R_SYM(info), S)-addr;
   StgInt64 off = gotAddress + A - P;
   *(Elf64_Word *)P = (Elf64_Word)off;
   break;
   }

Thanks for hitting me with the cluestick ;-)

 Though OTOH, the port is built with -fno-pie -nopie, so should it even
 be necessary to support RELOC_GOTPC? Is it building the internal copy of gmp
 incorrectly or something?

After looking at the build logs: yes, there's no trace of no-?pie
durinng the build of the bundled libgmp.

Also, objdump -r libHSinteger-gmp-0.5.0.0.a shows entries like

R_386_GOTPC   _GLOBAL_OFFSET_TABLE_

for all the object files from gmp.

I'll try to fix that part, first (when my current test bulk build
finished, may be today or tomorrow).

Fixing the ghc linker (and getting rid of all the no-?pie patches)
is the next step, but may take a little bit longer.

Ciao,
Kili

Re: i386 build failures

2014-01-28 Thread Jérémie Courrèges-Anglas 
Stuart Henderson st...@openbsd.org writes:

[...]

 PIE:
 multimedia/avidemux, games/openarena, games/megaglest/base

[...]

Here are possible fixes for those.
- avidemux: I tried to enable the #ifdef'd PIC code but it seems that
  register pressure is too high anyway, so disable PIE.
- openarena: disable PIE.
- megaglest: quickfix to use %esi instead of %ebx, and preserve the
  value of the latter across the cpuid call.

All those changes built-tested only, but I think they're reasonable.
Please prove me wrong!

Index: games/megaglest/base/Makefile
===
RCS file: /cvs/ports/games/megaglest/base/Makefile,v
retrieving revision 1.3
diff -u -p -r1.3 Makefile
--- games/megaglest/base/Makefile   21 May 2013 10:34:14 -  1.3
+++ games/megaglest/base/Makefile   29 Jan 2014 06:30:16 -
@@ -3,7 +3,7 @@
 COMMENT =  3D real-time strategy game
 DISTNAME = megaglest-source-${V}
 PKGNAME =  megaglest-${V}
-REVISION = 1
+REVISION = 2
 
 WANTLIB += GL GLU GLEW ICE SDL SM X11 Xext c crypto curl fontconfig ftgl
 WANTLIB += freetype idn ircclient jpeg ${MODLUA_WANTLIB}
Index: 
games/megaglest/base/patches/patch-source_shared_lib_include_platform_sdl_platform_main_h
===
RCS file: 
games/megaglest/base/patches/patch-source_shared_lib_include_platform_sdl_platform_main_h
diff -N 
games/megaglest/base/patches/patch-source_shared_lib_include_platform_sdl_platform_main_h
--- /dev/null   1 Jan 1970 00:00:00 -
+++ 
games/megaglest/base/patches/patch-source_shared_lib_include_platform_sdl_platform_main_h
   29 Jan 2014 06:29:24 -
@@ -0,0 +1,21 @@
+$OpenBSD$
+Fix for i386 PIE (ebx is reserved).
+--- source/shared_lib/include/platform/sdl/platform_main.h.origWed Nov 
14 20:20:08 2012
 source/shared_lib/include/platform/sdl/platform_main.h Wed Jan 29 
07:29:15 2014
+@@ -582,9 +582,16 @@ int mainSetup(int argc, char **argv) {
+ 
+ #define CHECK_BIT(var,pos) ((var)  (1(pos)))
+ 
++#if defined(__i386__)  defined(__PIE__)
+ #define cpuid(func,ax,bx,cx,dx)\
++  __asm__ __volatile__ (\
++  xchg %%ebx, %%esi; cpuid; xchg %%ebx, %%esi:\
++  =a (ax), =S (bx), =c (cx), =d (dx) : a (func));
++#else
++#define cpuid(func,ax,bx,cx,dx)\
+   __asm__ __volatile__ (cpuid:\
+   =a (ax), =b (bx), =c (cx), =d (dx) : a (func));
++#endif
+ 
+   int ax=0,bx=0,cx=0,dx=0;
+   cpuid(0x001,ax,bx,cx,dx)
Index: games/openarena/Makefile
===
RCS file: /cvs/ports/games/openarena/Makefile,v
retrieving revision 1.18
diff -u -p -r1.18 Makefile
--- games/openarena/Makefile1 Jun 2013 12:05:14 -   1.18
+++ games/openarena/Makefile29 Jan 2014 06:55:12 -
@@ -57,6 +57,10 @@ MAKE_FLAGS +=ARCH=ppc \
COMPILE_ARCH=ppc
 .endif
 
+.if ${MACHINE_ARCH} == i386
+MAKE_FLAGS +=  PIE_CFLAGS=-fno-pie PIE_LDFLAGS=-nopie
+.endif
+
 NO_TEST =  Yes
 WRKDIST =  ${WRKDIR}/openarena-engine-source-${E_V}
 ALL_TARGET =   release
Index: games/openarena/patches/patch-Makefile
===
RCS file: /cvs/ports/games/openarena/patches/patch-Makefile,v
retrieving revision 1.4
diff -u -p -r1.4 patch-Makefile
--- games/openarena/patches/patch-Makefile  18 Mar 2012 15:47:28 -  
1.4
+++ games/openarena/patches/patch-Makefile  29 Jan 2014 06:55:09 -
@@ -1,6 +1,6 @@
 $OpenBSD: patch-Makefile,v 1.4 2012/03/18 15:47:28 kirby Exp $
 Makefile.orig  Tue Dec 13 00:48:49 2011
-+++ Makefile   Wed Mar  7 11:21:26 2012
+--- Makefile.orig  Mon Dec 12 23:48:49 2011
 Makefile   Wed Jan 29 02:12:09 2014
 @@ -164,13 +164,14 @@ endif
  
  #
@@ -27,8 +27,9 @@ $OpenBSD: patch-Makefile,v 1.4 2012/03/1
 +  # Get the machine type
 +  ARCH=$(shell uname -m)
  
-   BASE_CFLAGS = -Wall -fno-strict-aliasing -Wimplicit -Wstrict-prototypes \
+-  BASE_CFLAGS = -Wall -fno-strict-aliasing -Wimplicit -Wstrict-prototypes \
 --DUSE_ICON
++  BASE_CFLAGS = $(PIE_CFLAGS) -Wall -fno-strict-aliasing -Wimplicit 
-Wstrict-prototypes \
 +-DUSE_ICON -DARCH_STRING=\\\${ARCH}\\\ -DMAP_ANONYMOUS=MAP_ANON
CLIENT_CFLAGS = $(SDL_CFLAGS)
SERVER_CFLAGS = 
@@ -43,7 +44,12 @@ $OpenBSD: patch-Makefile,v 1.4 2012/03/1
  
SHLIBEXT=so
SHLIBNAME=.$(SHLIBEXT)
-@@ -675,11 +675,12 @@ ifeq ($(PLATFORM),openbsd)
+@@ -671,15 +671,16 @@ ifeq ($(PLATFORM),openbsd)
+   SHLIBLDFLAGS=-shared $(LDFLAGS)
+ 
+   THREAD_LIBS=-pthread
+-  LIBS=-lm
++  LIBS= $(PIE_LDFLAGS) -lm
  
CLIENT_LIBS =
  
Index: multimedia/avidemux/Makefile
===
RCS file: /cvs/ports/multimedia/avidemux/Makefile,v
retrieving revision 1.43

i386 build failures

2014-01-27 Thread Stuart Henderson 
Here's a quick list of failures in the last i386 bulk build.

network header changes:
shells/nsh, devel/py-openbsd

modes:
emulators/fedora/base, emulators/fedora/cups

PIE:
multimedia/avidemux, games/openarena, games/megaglest/base

PIE?:
devel/hs-type-level, devel/hs-vector

www/chromium: intermittent (dependency issue?)
In file included from ../../chrome/browser/ui/webui/policy_ui.cc:20:
../../chrome/browser/policy/cloud/cloud_policy_client.h:18:10: fatal error: 
'chrome/browser/policy/proto/cloud/device_management_backend.pb.h' file not 
found

Re: i386 build failures

2014-01-27 Thread Antoine Jacoutot 
 modes:
 emulators/fedora/base, emulators/fedora/cups

I'll take care of these.

-- 
Antoine

Re: i386 build failures

2014-01-27 Thread David Coppa 
On Mon, Jan 27, 2014 at 1:51 PM, Stuart Henderson st...@openbsd.org wrote:

 www/chromium: intermittent (dependency issue?)
 In file included from ../../chrome/browser/ui/webui/policy_ui.cc:20:
 ../../chrome/browser/policy/cloud/cloud_policy_client.h:18:10: fatal error: 
 'chrome/browser/policy/proto/cloud/device_management_backend.pb.h' file not 
 found

I'm going to have a look at this...

Re: i386 build failures

2014-01-27 Thread David Coppa 

 Da: Stuart Henderson st...@openbsd.org
 Data: 27/gen/2014 13.51
 Oggetto: i386 build failures
 A: ports ports@openbsd.org

 www/chromium: intermittent (dependency issue?)
 In file included from ../../chrome/browser/ui/webui/policy_ui.cc:20:
 ../../chrome/browser/policy/cloud/cloud_policy_client.h:18:10: fatal error:
 'chrome/browser/policy/proto/cloud/device_management_backend.pb.h' file not
 found

Not 100% sure, but this should fix the dependency issue:

Index: patches/patch-chrome_chrome_browser_ui_gypi
===
RCS file: patches/patch-chrome_chrome_browser_ui_gypi
diff -N patches/patch-chrome_chrome_browser_ui_gypi
--- /dev/null   1 Jan 1970 00:00:00 -
+++ patches/patch-chrome_chrome_browser_ui_gypi 27 Jan 2014 18:52:44 -
@@ -0,0 +1,11 @@
+$OpenBSD$
+--- chrome/chrome_browser_ui.gypi.orig.portMon Jan 27 15:59:23 2014
 chrome/chrome_browser_ui.gypi  Mon Jan 27 16:23:52 2014
+@@ -2647,6 +2647,7 @@
+ ['configuration_policy==1', {
+   'dependencies': [
+ 'app/policy/cloud_policy_codegen.gyp:policy',
++'cloud_policy_proto',
+   ],
+ }, {  # configuration_policy==0
+   'sources/': [

Re: i386 build failures: mapnik, gcc/4.8

2013-07-16 Thread Christian Weisgerber 
Stuart Henderson st...@openbsd.org wrote:

 gcc/4.8 intermittent:
 http://junkpile.org/4.8,,-estdc.log

I also see this sometimes on amd64, and recently it happened with
4.6 as well.

-- 
Christian naddy Weisgerber  na...@mips.inka.de

Re: i386 build failures: mapnik, gcc/4.8

2013-07-13 Thread Stuart Henderson 
On 2013/07/13 12:02, Stuart Henderson wrote:
 mapnik is permanent:
 http://rhaalovely.net/build-failures/i386/2013-07-10/graphics/mapnik.log
 
 tests/cpp_tests/conversions_test.cpp:169: error: integer constant is too 
 large for 'long' type

oh, also spotted some -O3 crept in.

i386 build failures: pdflib, pwm, xdtm, xforms

2013-05-24 Thread Stuart Henderson 
some of these are probably related to X library updates..
not sure about pdflib though..

http://rhaalovely.net/build-failures/i386/2013-05-23/print/pdflib.log
don't know how to make ../../pdflib/libpdf.la (prerequisite of: hello)

http://rhaalovely.net/build-failures/i386/2013-05-23/x11/pwm.log
ld: cannot find -ltu

http://rhaalovely.net/build-failures/i386/2013-05-23/x11/xdtm.log
ld: cannot find -lExt

http://rhaalovely.net/build-failures/i386/2013-05-23/x11/xforms.log
don't know how to make /usr/include/machine/ansi.h (prerequisite of: align.o)

i386 build failures: audio/lmms math/yorick lang/jruby

2012-12-19 Thread Stuart Henderson 
Here are some build failures from my last i386 bulk build.



lang/jruby - SIGSEGV in jdk-1.7.0/bin/apt.
hs_err log, gdb backtrace (all threads) and build log at
http://junkpile.org/jruby-build.txt




audio/lmms fails intermittently due to dpb junking; somehow lmms cmake
picks up AppKit headers, which aren't a listed dependency, so of course
dpb removes them partway through the build.  I can't see where it would
be picking those headers up though, anyone have an idea?

Otherwise I suppose we could make it build_depends on gnustep,
but that's a bit of a sledgehammer approach.

 Running fake in audio/lmms
=== audio/lmms
===  Faking installation for lmms-0.4.8p3
/usr/local/bin/cmake -H/usr/obj/lmms-0.4.8/lmms-0.4.8 
-B/usr/obj/lmms-0.4.8/build-i386 --check-build-system CMakeFiles/Makefile.cmake 0
/usr/bin/make -f CMakeFiles/Makefile2 preinstall
/usr/bin/make -f CMakeFiles/lmms.dir/build.make CMakeFiles/lmms.dir/preinstall
make: don't know how to make /usr/local/include/AppKit/AppKit.h (prerequisite 
of: CMakeFiles/lmms.dir/src/core/Controller.o)
Stop in .



math/yorick consistently fails:

yinit.o(.text+0x2e): In function `on_launch':
: undefined reference to `y_launch'
yinit.o(.data+0x0): undefined reference to `yk_yor'
../drat/libdrat.a(ydrat.o)(.text+0xc8): In function `PrintDM':
: undefined reference to `ForceNewline'
../drat/libdrat.a(ydrat.o)(.text+0xf5): In function `PrintDM':
: undefined reference to `PrintFunc'
../drat/libdrat.a(ydrat.o)(.text+0xfa): In function `PrintDM':
: undefined reference to `ForceNewline'
../drat/libdrat.a(ydrat.o)(.text+0x12d): In function `YGetDMesh':
: undefined reference to `referenceSym'
../drat/libdrat.a(ydrat.o)(.text+0x134): In function `YGetDMesh':
: undefined reference to `dataBlockSym'
(etc etc etc).

Re: i386 build failures: audio/lmms math/yorick lang/jruby

2012-12-19 Thread David Coppa 
On Wed, Dec 19, 2012 at 11:04 AM, Stuart Henderson s...@spacehopper.org wrote:

 Running fake in audio/lmms
 === audio/lmms
 ===  Faking installation for lmms-0.4.8p3
 /usr/local/bin/cmake -H/usr/obj/lmms-0.4.8/lmms-0.4.8 
 -B/usr/obj/lmms-0.4.8/build-i386 --check-build-system 
 CMakeFiles/Makefile.cmake 0
 /usr/bin/make -f CMakeFiles/Makefile2 preinstall
 /usr/bin/make -f CMakeFiles/lmms.dir/build.make CMakeFiles/lmms.dir/preinstall
 make: don't know how to make /usr/local/include/AppKit/AppKit.h (prerequisite 
 of: CMakeFiles/lmms.dir/src/core/Controller.o)

Can you check for AppKit.h into the *.includecache files?
To understand who is bringing it into the game...

Re: i386 build failures: audio/lmms math/yorick lang/jruby

2012-12-19 Thread Stuart Henderson 
On 2012/12/19 15:31, David Coppa wrote:
 On Wed, Dec 19, 2012 at 11:04 AM, Stuart Henderson s...@spacehopper.org 
 wrote:
 
  Running fake in audio/lmms
  === audio/lmms
  ===  Faking installation for lmms-0.4.8p3
  /usr/local/bin/cmake -H/usr/obj/lmms-0.4.8/lmms-0.4.8 
  -B/usr/obj/lmms-0.4.8/build-i386 --check-build-system 
  CMakeFiles/Makefile.cmake 0
  /usr/bin/make -f CMakeFiles/Makefile2 preinstall
  /usr/bin/make -f CMakeFiles/lmms.dir/build.make 
  CMakeFiles/lmms.dir/preinstall
  make: don't know how to make /usr/local/include/AppKit/AppKit.h 
  (prerequisite of: CMakeFiles/lmms.dir/src/core/Controller.o)
 
 Can you check for AppKit.h into the *.includecache files?
 To understand who is bringing it into the game...
 

build-i386/CMakeFiles/lmms.dir/CXX.includecache starts like this:-

-- -- -- -- --
#IncludeRegexLine: ^[   ]*#[]*(include|import)[ ]*[]([^]+)([])

#IncludeRegexScan: ^.*$

#IncludeRegexComplain: ^$

#IncludeRegexTransform:

/usr/local/include/AppKit/AppKit.h
GNUstepBase/GSVersionMacros.h
-
GNUstepGUI/GSVersion.h
-
Foundation/Foundation.h
-
AppKit/NSGraphics.h
-
AppKit/NSActionCell.h
-
AppKit/NSApplication.h
-
followed by lots more AppKit/*.h references
-- -- -- -- --

... and there's another mention of AppKit.h like this,

-- -- -- -- --
/usr/local/include/AppKit/PSOperators.h
AppKit/DPSOperators.h
-

/usr/local/include/Cocoa/Cocoa.h
Foundation/Foundation.h
-
AppKit/AppKit.h
-
-- -- -- -- --


sthen@i386[/usr/obj/lmms-0.4.8] grep -R AppKit .|cut -d: -f1 | uniq -c
18189 ./build-i386/CMakeFiles/lmms.dir/depend.make
 453 ./build-i386/CMakeFiles/lmms.dir/CXX.includecache
18189 ./build-i386/CMakeFiles/lmms.dir/depend.internal

^^ no references to AppKit in any files other than these in all of lmms'
source/build directories.

I would send a tarred build directory but it's close to ~50MB even
after compressing with xz..

i386 build failures: cil, jruby, smlnj, yorick

2012-11-19 Thread Stuart Henderson 
http://junkpile.org/i386-20121117/devel/cil.log
- mono-related?
- last built 2012/08/04 (5.2)

http://junkpile.org/i386-20121117/lang/jruby.log
- JRE segfault
- last built 2012/11/01

http://junkpile.org/i386-20121117/lang/smlnj.log
- various failures on different builds, sometimes seems sysconf-related
- last built 2012/05/30
[Loading posix-procenv.sml.bin]
/usr/obj/smlnj-110.0.7/bin/.run/run.x86-openbsd: Fatal error -- Uncaught excepti

http://junkpile.org/i386-20121117/math/yorick.log
- intermittent; on the same machine without touching anything
else between runs, I can build OK, then clean and rebuild and fail,
then clean and rebuild again OK.
- haven't seen failures before 2012/11/14
cc -O2-o yorick main.o yinit.o -L. -L../drat -L../hex  -L../drat -ldrat -L..
./libyor.a(pstrcpy.o)(.text+0x33): In function `p_strcpy':
: warning: strcpy() is almost always misused, please use strlcpy()
/usr/X11R6/lib/libX11.so.15.1: warning: strcat() is almost always misused, pleas
../drat/libdrat.a(ydrat.o)(.text+0xed): In function `PrintDM':
: warning: sprintf() is often misused, please use snprintf()
yinit.o(.text+0x2e): In function `on_launch':
: undefined reference to `y_launch'
yinit.o(.data+0x0): undefined reference to `yk_yor'
../drat/libdrat.a(ydrat.o)(.text+0xc8): In function `PrintDM':
: undefined reference to `ForceNewline'
../drat/libdrat.a(ydrat.o)(.text+0xf5): In function `PrintDM':
: undefined reference to `PrintFunc'
../drat/libdrat.a(ydrat.o)(.text+0xfa): In function `PrintDM':
: undefined reference to `ForceNewline'
../drat/libdrat.a(ydrat.o)(.text+0x12d): In function `YGetDMesh':
: undefined reference to `referenceSym'
[.]

start of diff between good and failing builds:-

--- 20121114/paths/math/yorick.log  Tue Nov 13 22:08:08 2012
+++ 20121117/paths/math/yorick.log  Sat Nov 17 10:35:08 2012
@@ -224,11 +224,8 @@ cc -O2 -pipe   -I. -I. -I../play -I../gist  -c funcdef
 cc -O2 -pipe   -I. -I. -I../play -I../gist  -c spawn.c
 cc  -O2 -pipe   -I. -I. -I../play -I../gist -Df_linkage_ -o fortrn.o -c 
fortrn.c
 cc  -O2 -pipe   -I. -I. -I../play -I../gist -Df_linkage_ -o codger codger.c
-/tmp//cc8QEter.o(.text+0x21b3): In function `main':
-: warning: strcpy() is almost always misused, please use strlcpy()
-/tmp//cc8QEter.o(.text+0x21df): In function `main':
-: warning: strcat() is almost always misused, please use strlcat()
-./codger w yor -I../i0 std.i matrix.i fft.i graph.i
+/tmp//ccwZM1RI.o(.text+0x21b3): In function `main':
+: warning: strcpy() is almost always misused,./codger w yor -I../i0 std.i 
matrix.i fft.i graph.i
 found std.i in ../i0
 found matrix.i in ../i0
 found fft.i in ../i0

Re: i386 build failures: cil, jruby, smlnj, yorick

2012-11-19 Thread Nigel Taylor 
On 11/19/12 11:05, Stuart Henderson wrote:
 http://junkpile.org/i386-20121117/devel/cil.log
 - mono-related?
 - last built 2012/08/04 (5.2)
To make this work you need to change this in Makefile...
-   ${WRKBUILD}/obj/x86_LINUX/machdep-ml*.exe \
+   ${WRKBUILD}/obj/x86_LINUX/machdep-ml.exe \

machdep-ml64.exe is not created on i386 never was
machdep-ml64.exe is link to machdep-ml.exe on amd64
machdep-ml32.exe is link to machdep-ml.exe on i386

Could copy machdep-ml32 on both i386/amd64 in Makefile and adjust PLIST
to include. Attached is the diff I use to build... Never use cil so not
tested beyond building.

 
 http://junkpile.org/i386-20121117/lang/jruby.log
 - JRE segfault
 - last built 2012/11/01
 
 http://junkpile.org/i386-20121117/lang/smlnj.log
 - various failures on different builds, sometimes seems sysconf-related
 - last built 2012/05/30
 [Loading posix-procenv.sml.bin]
 /usr/obj/smlnj-110.0.7/bin/.run/run.x86-openbsd: Fatal error -- Uncaught 
 excepti
 
 http://junkpile.org/i386-20121117/math/yorick.log
 - intermittent; on the same machine without touching anything
 else between runs, I can build OK, then clean and rebuild and fail,
 then clean and rebuild again OK.
 - haven't seen failures before 2012/11/14
 cc -O2-o yorick main.o yinit.o -L. -L../drat -L../hex  -L../drat -ldrat 
 -L..
 ./libyor.a(pstrcpy.o)(.text+0x33): In function `p_strcpy':
 : warning: strcpy() is almost always misused, please use strlcpy()
 /usr/X11R6/lib/libX11.so.15.1: warning: strcat() is almost always misused, 
 pleas
 ../drat/libdrat.a(ydrat.o)(.text+0xed): In function `PrintDM':
 : warning: sprintf() is often misused, please use snprintf()
 yinit.o(.text+0x2e): In function `on_launch':
 : undefined reference to `y_launch'
 yinit.o(.data+0x0): undefined reference to `yk_yor'
 ../drat/libdrat.a(ydrat.o)(.text+0xc8): In function `PrintDM':
 : undefined reference to `ForceNewline'
 ../drat/libdrat.a(ydrat.o)(.text+0xf5): In function `PrintDM':
 : undefined reference to `PrintFunc'
 ../drat/libdrat.a(ydrat.o)(.text+0xfa): In function `PrintDM':
 : undefined reference to `ForceNewline'
 ../drat/libdrat.a(ydrat.o)(.text+0x12d): In function `YGetDMesh':
 : undefined reference to `referenceSym'
 [.]
 
 start of diff between good and failing builds:-
 
 --- 20121114/paths/math/yorick.log  Tue Nov 13 22:08:08 2012
 +++ 20121117/paths/math/yorick.log  Sat Nov 17 10:35:08 2012
 @@ -224,11 +224,8 @@ cc -O2 -pipe   -I. -I. -I../play -I../gist  -c funcdef
  cc -O2 -pipe   -I. -I. -I../play -I../gist  -c spawn.c
  cc  -O2 -pipe   -I. -I. -I../play -I../gist -Df_linkage_ -o fortrn.o -c 
 fortrn.c
  cc  -O2 -pipe   -I. -I. -I../play -I../gist -Df_linkage_ -o codger codger.c
 -/tmp//cc8QEter.o(.text+0x21b3): In function `main':
 -: warning: strcpy() is almost always misused, please use strlcpy()
 -/tmp//cc8QEter.o(.text+0x21df): In function `main':
 -: warning: strcat() is almost always misused, please use strlcat()
 -./codger w yor -I../i0 std.i matrix.i fft.i graph.i
 +/tmp//ccwZM1RI.o(.text+0x21b3): In function `main':
 +: warning: strcpy() is almost always misused,./codger w yor -I../i0 std.i 
 matrix.i fft.i graph.i
  found std.i in ../i0
  found matrix.i in ../i0
  found fft.i in ../i0
 
 
Seen yorick fail recently. Also seen these fail...
mail/mozilla-thunderbird
www/mozilla-firefox
www/seamonkey
x11/mono-gtk2
x11/qt4

One mono related, others down to parallel build issue.
Index: Makefile
===
RCS file: /home/cvs/ports/devel/cil/Makefile,v
retrieving revision 1.23
diff -u -p -r1.23 Makefile
--- Makefile30 Sep 2012 09:16:36 -  1.23
+++ Makefile30 Sep 2012 10:36:02 -
@@ -7,7 +7,7 @@ COMMENT=framework for analysis and tran
 CATEGORIES =   devel lang
 
 DISTNAME = cil-1.5.1
-REVISION =  1
+REVISION =  2
 EXTRACT_SUFX = .tar.gz
 MASTER_SITES = ${MASTER_SITE_SOURCEFORGE:=cil/}
 
@@ -44,7 +44,7 @@ post-install:
${INSTALL_SCRIPT} ${WRKBUILD}/bin/patcher ${PREFIX}/bin
${INSTALL_DATA_DIR} ${PREFIX}/libexec/cil
${INSTALL_PROGRAM} \
-   ${WRKBUILD}/obj/x86_LINUX/machdep-ml*.exe \
+   ${WRKBUILD}/obj/x86_LINUX/machdep-ml.exe \
${PREFIX}/libexec/cil
${INSTALL_DATA_DIR} ${P5INST}/CIL
mv ${PREFIX}/lib/*.pm ${P5INST}/CIL/
Index: pkg/PLIST
===
RCS file: /home/cvs/ports/devel/cil/pkg/PLIST,v
retrieving revision 1.6
diff -u -p -r1.6 PLIST
--- pkg/PLIST   18 Aug 2012 22:00:56 -  1.6
+++ pkg/PLIST   6 Sep 2012 10:56:22 -
@@ -1,6 +1,7 @@
 @comment $OpenBSD: PLIST,v 1.6 2012/08/18 22:00:56 avsm Exp $
 bin/cilly
 bin/patcher
+lib/ocaml/
 lib/ocaml/cil/
 lib/ocaml/cil/META
 lib/ocaml/cil/alpha.cmi
@@ -128,6 +129,7 @@ lib/ocaml/cil/whitetrack.mli
 lib/ocaml/cil/zrapp.mli
 %%native%%
 @comment ${P5SITE}/CIL/CilConfig.pm
+lib/ocaml/stublibs/
 ${P5SITE}/CIL/

Re: i386 build failures: cil, jruby, smlnj, yorick

2012-11-19 Thread Stuart Henderson 
On 2012/11/19 11:50, Nigel Taylor wrote:
 On 11/19/12 11:05, Stuart Henderson wrote:
  http://junkpile.org/i386-20121117/devel/cil.log
  - mono-related?
  - last built 2012/08/04 (5.2)
 To make this work you need to change this in Makefile...
 - ${WRKBUILD}/obj/x86_LINUX/machdep-ml*.exe \
 + ${WRKBUILD}/obj/x86_LINUX/machdep-ml.exe \
 
 machdep-ml64.exe is not created on i386 never was
 machdep-ml64.exe is link to machdep-ml.exe on amd64
 machdep-ml32.exe is link to machdep-ml.exe on i386
 
 Could copy machdep-ml32 on both i386/amd64 in Makefile and adjust PLIST
 to include. Attached is the diff I use to build... Never use cil so not
 tested beyond building.

Oh it's ocaml? Weird file extensions...

Your diff is a good start but it's still broken at runtime (easy
to see; regress fails immediately too) - chrisz could you take
a look please?

$ make regress
===  Regression check for cil-1.5.1p1
bin/cilly -c test/small1/func.c -otest/small1/func.o
cc -D_GNUCC -E -DCIL=1 test/small1/func.c -o /tmp/cil-0HD1JWn2.i
/x86_LINUX/cilly.asm.exe --out /tmp/cil-V0oX4vjo.cil.c /tmp/cil-0HD1JWn2.i
The command
/x86_LINUX/cilly.asm.exe --out /tmp/cil-V0oX4vjo.cil.c /tmp/cil-0HD1JWn2.i
received signal %d
127gmake: *** [test/func] Error 255
*** Error 2 in . (/usr/ports/infrastructure/mk/bsd.port.mk:2599 
'/usr/obj/ports/cil-1.5.1/.regress_done')
*** Error 1 in /usr/ports/devel/cil 
(/usr/ports/infrastructure/mk/bsd.port.mk:2287 'regress')

$ wrksrc
$ bin/cilly -c test/small1/func.c -otest/small1/func.o
cc -D_GNUCC -E -DCIL=1 test/small1/func.c -o /tmp/cil-j5NnwJxL.i
/x86_LINUX/cilly.asm.exe --out /tmp/cil-NZebDpqs.cil.c /tmp/cil-j5NnwJxL.i
The command
/x86_LINUX/cilly.asm.exe --out /tmp/cil-NZebDpqs.cil.c /tmp/cil-j5NnwJxL.i
received signal %d

 and if I try running it from the installed package:

$ cilly -c test/small1/func.c -otest/small1/func.o
/usr/bin/gcc -D_GNUCC -E -DCIL=1 test/small1/func.c -o /tmp/cil-hzT3Piqg.i
/usr/local/libexec/cil/cilly.asm.exe --out /tmp/cil-qpayf6IQ.cil.c 
/tmp/cil-hzT3Piqg.i
The command /usr/local/libexec/cil/cilly.asm.exe --out /tmp/cil-qpayf6IQ.cil.c 
/tmp/cil-hzT3Piqg.i received signal %d

cilly.asm.exe / cilly.byte.exe are installed to
/usr/local/libexec/cil/obj/x86_LINUX but the /usr/local/bin/cilly
perl script looks for them in /usr/local/libexec/cil.

 Seen yorick fail recently. Also seen these fail...
 mail/mozilla-thunderbird
 www/mozilla-firefox
 www/seamonkey
 x11/mono-gtk2
 x11/qt4
 
 One mono related, others down to parallel build issue.

I haven't seen those parallel build issues here, it might be worth
giving them a try with PARALLEL_INSTALL=No.

i386 build failures, two python things

2012-10-02 Thread Stuart Henderson 
Anyone have ideas about these?

- www/py-paste-script, ImportError: No module named peak

- audio/py-tagpy, src/wrapper/id3.cpp:215: error: no matching function for 
call to 'boost::python::class_TagLib::ID3v2::Tag, 
boost::noncopyable_::noncopyable, boost::python::basesTagLib::Tag, 
mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, 
mpl_::void_, mpl_::void_, mpl_::void_, 
boost::python::detail::not_specified::def(const char [7], unresolved 
overloaded function type)' - boost is installed.


 Building under www/py-paste-script
 BDEPENDS = 
[www/py-paste;www/py-flup;devel/py-cheetah;devel/py-setuptools;www/py-paste-deploy;devel/py-wsgiutils;lang/python/2.7]
 DIST = [www/py-paste-script:PasteScript-1.7.5.tar.gz]
 FULLPKGNAME = py-paste-script-1.7.5
 RDEPENDS = 
[www/py-paste;www/py-flup;devel/py-cheetah;devel/py-setuptools;devel/py-wsgiutils;www/py-paste-deploy;lang/python/2.7]
/usr/sbin/pkg_add -aI py-cheetah-2.4.4p2 py-flup-1.0.2p1 py-paste-1.7.5.1 
py-paste-deploy-1.5.0 py-setuptools-0.6.11p4v0 py-wsgiutils-0.7p4 python-2.7.3p1
 Running prepare in www/py-paste-script
Use of uninitialized value $v in concatenation (.) or string at 
/usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
=== www/py-paste-script
=== py-paste-script-1.7.5 depends on: py-cheetah-* - py-cheetah-2.4.4p2
=== py-paste-script-1.7.5 depends on: py-wsgiutils-* - py-wsgiutils-0.7p4
=== py-paste-script-1.7.5 depends on: py-flup-* - py-flup-1.0.2p1
=== py-paste-script-1.7.5 depends on: py-paste-* - py-paste-1.7.5.1
=== py-paste-script-1.7.5 depends on: py-paste-deploy-* - 
py-paste-deploy-1.5.0
=== py-paste-script-1.7.5 depends on: python-=2.7,2.8 - python-2.7.3p1
=== py-paste-script-1.7.5 depends on: py-setuptools-* - 
py-setuptools-0.6.11p4v0
 Running patch in www/py-paste-script
Use of uninitialized value $v in concatenation (.) or string at 
/usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
=== www/py-paste-script
===  Checking files for py-paste-script-1.7.5
`/usr/ports/distfiles/PasteScript-1.7.5.tar.gz' is up to date.
===  Extracting for py-paste-script-1.7.5
===  Patching for py-paste-script-1.7.5
 Running configure in www/py-paste-script
Use of uninitialized value $v in concatenation (.) or string at 
/usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
=== www/py-paste-script
===  Configuring for py-paste-script-1.7.5
 Running build in www/py-paste-script
Use of uninitialized value $v in concatenation (.) or string at 
/usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
=== www/py-paste-script
===  Building for py-paste-script-1.7.5
Traceback (most recent call last):
  File ./setup.py, line 1, in module
from setuptools import setup, find_packages
  File /usr/local/lib/python2.7/site-packages/setuptools/__init__.py, line 2, 
in module
from setuptools.extension import Extension, Library
  File /usr/local/lib/python2.7/site-packages/setuptools/extension.py, line 
2, in module
from dist import _get_unpatched
  File /usr/local/lib/python2.7/site-packages/setuptools/dist.py, line 6, in 
module
from setuptools.command.sdist import sdist
  File /usr/local/lib/python2.7/site-packages/setuptools/command/sdist.py, 
line 5, in module
import os, re, sys, pkg_resources
  File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2623, in 
module
add_activation_listener(lambda dist: dist.activate())
  File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 679, in 
subscribe
callback(dist)
  File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2623, in 
lambda
add_activation_listener(lambda dist: dist.activate())
  File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2172, in 
activate
map(declare_namespace, self._get_metadata('namespace_packages.txt'))
  File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 1776, in 
declare_namespace
__import__(parent)
ImportError: No module named peak

Stop in /usr/ports/www/py-paste-script:
 Exit status 1 (line 176 of /usr/ports/lang/python/python.port.mk,
 target do-build: @cd /usr/obj/py-paste-script-1.7.5/PasteScript-1.7.5  
/usr/bin/env -i CC=cc PYTHONUSERBASE= 
PATH='/usr/obj/py-paste-script-1.7.5/bin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/local/bin:/usr/X11R6/bin'
 PREFIX='/usr/local'  LOCALBASE='/usr/local' DEPBASE='/usr/local' 
X11BASE='/usr/X11R6'  CFLAGS='-O2 -pipe'  TRUEPREFIX='/usr/local' DESTDIR=''  
HOME='/py-paste-script-1.7.5_writes_to_HOME' ELF_TOOLCHAIN=yes 
COMPILER_VERSION=gcc4  PICFLAG=-fpic ASPICFLAG=  BINGRP=bin BINOWN=root 
BINMODE=555 NONBINMODE=444 DIRMODE=755  INSTALL_COPY=-c INSTALL_STRIP=-s  
MANGRP=bin MANOWN=root MANMODE=444 BSD_INSTALL_PROGRAM=install -c -s -o root 
-g bin -m 555  BSD_INSTALL_SCRIPT=install -c -o root -g bin -m 555  
BSD_INSTALL_DATA=install -c -o root -g bin -m 444  BSD_INSTALL_MAN=install 
-c -o root -g bin -m 444  BSD_INSTALL_PROGRAM_DIR=install -d -o root

Re: i386 build failures, two python things

2012-10-02 Thread Nigel Taylor 
On 10/02/12 21:59, Stuart Henderson wrote:
 Anyone have ideas about these?
 
 - www/py-paste-script, ImportError: No module named peak
 
 - audio/py-tagpy, src/wrapper/id3.cpp:215: error: no matching function for 
 call to 'boost::python::class_TagLib::ID3v2::Tag, 
 boost::noncopyable_::noncopyable, boost::python::basesTagLib::Tag, 
 mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, mpl_::void_, 
 mpl_::void_, mpl_::void_, mpl_::void_, 
 boost::python::detail::not_specified::def(const char [7], unresolved 
 overloaded function type)' - boost is installed.
 
 
 Building under www/py-paste-script
BDEPENDS = 
 [www/py-paste;www/py-flup;devel/py-cheetah;devel/py-setuptools;www/py-paste-deploy;devel/py-wsgiutils;lang/python/2.7]
DIST = [www/py-paste-script:PasteScript-1.7.5.tar.gz]
FULLPKGNAME = py-paste-script-1.7.5
RDEPENDS = 
 [www/py-paste;www/py-flup;devel/py-cheetah;devel/py-setuptools;devel/py-wsgiutils;www/py-paste-deploy;lang/python/2.7]
 /usr/sbin/pkg_add -aI py-cheetah-2.4.4p2 py-flup-1.0.2p1 py-paste-1.7.5.1 
 py-paste-deploy-1.5.0 py-setuptools-0.6.11p4v0 py-wsgiutils-0.7p4 
 python-2.7.3p1
 Running prepare in www/py-paste-script
 Use of uninitialized value $v in concatenation (.) or string at 
 /usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
 === www/py-paste-script
 === py-paste-script-1.7.5 depends on: py-cheetah-* - py-cheetah-2.4.4p2
 === py-paste-script-1.7.5 depends on: py-wsgiutils-* - py-wsgiutils-0.7p4
 === py-paste-script-1.7.5 depends on: py-flup-* - py-flup-1.0.2p1
 === py-paste-script-1.7.5 depends on: py-paste-* - py-paste-1.7.5.1
 === py-paste-script-1.7.5 depends on: py-paste-deploy-* - 
 py-paste-deploy-1.5.0
 === py-paste-script-1.7.5 depends on: python-=2.7,2.8 - python-2.7.3p1
 === py-paste-script-1.7.5 depends on: py-setuptools-* - 
 py-setuptools-0.6.11p4v0
 Running patch in www/py-paste-script
 Use of uninitialized value $v in concatenation (.) or string at 
 /usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
 === www/py-paste-script
 ===  Checking files for py-paste-script-1.7.5
 `/usr/ports/distfiles/PasteScript-1.7.5.tar.gz' is up to date.
 ===  Extracting for py-paste-script-1.7.5
 ===  Patching for py-paste-script-1.7.5
 Running configure in www/py-paste-script
 Use of uninitialized value $v in concatenation (.) or string at 
 /usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
 === www/py-paste-script
 ===  Configuring for py-paste-script-1.7.5
 Running build in www/py-paste-script
 Use of uninitialized value $v in concatenation (.) or string at 
 /usr/ports/infrastructure/lib/DPB/Core/Distant.pm line 79.
 === www/py-paste-script
 ===  Building for py-paste-script-1.7.5
 Traceback (most recent call last):
   File ./setup.py, line 1, in module
 from setuptools import setup, find_packages
   File /usr/local/lib/python2.7/site-packages/setuptools/__init__.py, line 
 2, in module
 from setuptools.extension import Extension, Library
   File /usr/local/lib/python2.7/site-packages/setuptools/extension.py, line 
 2, in module
 from dist import _get_unpatched
   File /usr/local/lib/python2.7/site-packages/setuptools/dist.py, line 6, 
 in module
 from setuptools.command.sdist import sdist
   File /usr/local/lib/python2.7/site-packages/setuptools/command/sdist.py, 
 line 5, in module
 import os, re, sys, pkg_resources
   File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2623, 
 in module
 add_activation_listener(lambda dist: dist.activate())
   File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 679, 
 in subscribe
 callback(dist)
   File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2623, 
 in lambda
 add_activation_listener(lambda dist: dist.activate())
   File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 2172, 
 in activate
 map(declare_namespace, self._get_metadata('namespace_packages.txt'))
   File /usr/local/lib/python2.7/site-packages/pkg_resources.py, line 1776, 
 in declare_namespace
 __import__(parent)
 ImportError: No module named peak
 python peak module
 Stop in /usr/ports/www/py-paste-script:
  Exit status 1 (line 176 of /usr/ports/lang/python/python.port.mk,
  target do-build: @cd /usr/obj/py-paste-script-1.7.5/PasteScript-1.7.5  
 /usr/bin/env -i CC=cc PYTHONUSERBASE= 
 PATH='/usr/obj/py-paste-script-1.7.5/bin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/local/bin:/usr/X11R6/bin'
  PREFIX='/usr/local'  LOCALBASE='/usr/local' DEPBASE='/usr/local' 
 X11BASE='/usr/X11R6'  CFLAGS='-O2 -pipe'  TRUEPREFIX='/usr/local' DESTDIR=''  
 HOME='/py-paste-script-1.7.5_writes_to_HOME' ELF_TOOLCHAIN=yes 
 COMPILER_VERSION=gcc4  PICFLAG=-fpic ASPICFLAG=  BINGRP=bin BINOWN=root 
 BINMODE=555 NONBINMODE=444 DIRMODE=755  INSTALL_COPY=-c INSTALL_STRIP=-s  
 MANGRP=bin MANOWN=root MANMODE=444 BSD_INSTALL_PROGRAM=install -c -s -o root 
 -g bin -m 555  BSD_INSTALL_SCRIPT=install -c -o root -g bin -m 555

Re: i386 build failures, two python things

2012-10-02 Thread Stuart Henderson 
On 2012/10/02 23:48, Nigel Taylor wrote:
 www/py-paste-script built for me on both amd64 and i386... release built
 today / port tree updated about 15:00 for both amd64/i386. Deleting all
 the packages and rebuilding on amd64 I could not repeat your problem.
 i386 tried the same, and py-paste-script still built.

I suspect it maybe something that *is* installed which breaks it, rather than
something which isn't, maybe I'll try removing them one by one..

 audio/py-tagpy I have seen this, occurs on amd64 also

Had a diff from dcoppa for this, not checked it yet though.

 In the end I reverted audio/taglib in ports tree to 29/09/2012 so
 previous version was used, 1.7.2p0 then py-tagpy built.
 
 Could be fixed in taglib 1.9, could need py-tagpy updating to match
 later versions of taglib.

i386 build failures

2012-08-21 Thread Stuart Henderson 
E=net/nfsen,-web(i386) devel/p5-SVN-Web(i386) devel/p5-SVN-Simple 
productivity/wyrd lang/smlnj(i386) devel/cil games/frozen-bubble,-server 
math/maple math/maple-share math/maple

nfsen, the p5-SVN things, frozen-bubble are pthread fallout, possibly best
fixed by threaded perl.

wyrd, cil are ocaml fallout.

smlnj is still broken following sysconf changes a while ago.

(and maple is commercial, no distfiles).