Re: lsof, libgtop2: 'struct rb_entry' has no member named 'rbe_left' etc

On 2016-09-20, Stuart Henderson wrote: > Any suggestions what to do in the short term about these kvm-grovellers? I move that we delete lsof. It has been building fine on the amd64.ports machines and I was mystified how this could be until I realized that it uses

Re: FIX: www/chromium build

On 2016-08-18, Rafael Sadowski wrote: > the patch below fix my chromium build @amd64. What's the problem? -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: simple wxneeded setting from port Makefiles

On 2016-08-18, Stuart Henderson wrote: > This factors out the wrapper so it can be generated by ports > infrastructure instead (by setting USE_WXNEEDED=Yes), and adds > it to the sqlports database so we can spot them more easily. I guess ports that were changed to use

Re: xpdf crashes when going fullscreen

David Coppa: > Can you please try the x11/motif diff below and report back? The question isn't so much whether this fixes the xpdf problem, but whether is breaks anything else. -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: textlive_base amd64+dpb web2c == "No rule to make target 'all'

On 2016-08-21, Kenneth R Westerback wrote: > Trying to compile coccinelle on my amd64 dpb builder with -current > sources as of last night fails in the dependency textlive_base. The > last messages are FWIW, I have not seen this in the amd64 package builds. -- Christian

Fix *.googlecode.com master sites

AGS="${LDFLAGS}" Index: comms/wy60/Makefile === RCS file: /cvs/ports/comms/wy60/Makefile,v retrieving revision 1.12 diff -u -p -r1.12 Makefile --- comms/wy60/Makefile 11 Mar 2013 01:30:27 - 1.12 +++ comms/wy60/Makefile 30 Sep 2016 22:

Re: strange build failure with llvm

On 2016-09-30, patrick keshishian wrote: > I'm just curious if this type of failures have been seen by those > doing bulk builds? Nope (amd64). -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: lang/seed7

Stuart Henderson: > I'm seeing this failure on i386 and amd64: > > cmd_rtl.c:1449: error: 'NAN_COMPARISON_OKAY' undeclared (first use in this > function) > > Any clues? seed7 has its own autoconf replacement, chkccomp. The problem is SYSTEM_LIBS="-lm -lsqlite3". chkccomp tries to link with

Re: Update: devel/gmp 6.1.1

On 2016-11-02, Christian Weisgerber <na...@mips.inka.de> wrote: > This is a long overdue update of devel/gmp to the latest release > (6.1.1). Our old version of devel/mpfr does not build against this and also requires a straightforward update. (See below.) The lang/gcc/4.9 Ad

Re: krb5-config --libs

Stuart Henderson: > The downside is that then everybody gets asked which curl package to use > when it gets installed as a dependency.. and I bet many people will choose > "full" when they don't need it (and unnecessarily pull in a bunch of other > libraries to the address space, and maybe

x11/kde/base3 dependency error

x11/kde/base3 failed to build in my latest amd64 bulk build. >>> Building on amd64-2 under x11/kde/base3,-en_US BDEPENDS =

amd64: chronic breakage

I would like to remind everybody that there are a number of ports that keep breaking in my amd64 bulk builds: lang/mono This is almost always broken. It builds on an otherwise idle machine, but when there is other build activity going on in parallel, it almost always segfaults. This

Re: amd64: chronic breakage

Antoine Jacoutot: > > lang/guile2 > > I think the issue is that at some point it takes a huge amount of time and > dpb kills it because the stuck timeout it reached. That doesn't seem plausible. The last time guile2 successfully built, it took 21 minutes total. Meanwhile my stuck timeout is

graphics/imlib2: imlib2-config vs. pkg-config

imlib2-config and pkg-config imlib2 don't agree: $ imlib2-config --cflags -I/usr/local/include $ pkg-config --cflags imlib2 -I/usr/local/include -I/usr/X11R6/include/freetype2 -I/usr/X11R6/include $ imlib2-config --libs -L/usr/local/lib -lImlib2 $ pkg-config --libs imlib2

Re: Update: multimedia/mpv, feedback needed

On 2016-11-23, "Dmitrij D. Czarkoff" wrote: > I was asked off-list to add a "lite" flavor for multimedia/mpv. This > diff introduces such version. This iteration disables the fllowing in > "lite" FLAVOR: -snip- Well, if I was using mpv (as espie urges me to), I would want

Update: graphics/tiff 4.0.7

Here's an update to graphics/tiff 4.0.7. Changes: http://www.simplesystems.org/libtiff/v4.0.7.html TL;DR: Numerous security fixes in the library; some tools were removed. The changes to the port mostly consist of disentangling and removing various accumulated security patches. Unfortunately

Re: graphics/imlib2: imlib2-config vs. pkg-config

Jeremie Courreges-Anglas: > > imlib2-config and pkg-config imlib2 don't agree: > > I think that "-L/usr/local/lib -L/usr/X11R6/lib -lImlib2" and > "-I/usr/local/include -I/usr/X11R6/include" should be enough. Yes to "-I/usr/local/include -I/usr/X11R6/include". However, in "-L/usr/local/lib

Re: graphics/imlib2: imlib2-config vs. pkg-config

On 2016-11-27, Jeremie Courreges-Anglas wrote: >>> > imlib2-config and pkg-config imlib2 don't agree: >>> >>> I think that "-L/usr/local/lib -L/usr/X11R6/lib -lImlib2" and >>> "-I/usr/local/include -I/usr/X11R6/include" should be enough. >> >> Yes to "-I/usr/local/include

devel/libf2c: fix after bsd.obj.mk changes

After the recent bsd.obj.mk changes, devel/libf2c fails to build in a dpb setup where the build is run as the _pbuild user. ===> Configuring for libf2c-3.3.6p9 cd /usr/obj/ports/libf2c-3.3.6/libf2c && /usr/bin/make -f /usr/obj/ports/libf2c-3.3.6/libf2c/Makefile.bsd-wrapper

Update: devel/gmp 6.1.1

/ +HOMEPAGE= https://gmplib.org/ MAINTAINER=Christian Weisgerber <na...@openbsd.org> -EXTRACT_SUFX= .tar.bz2 +EXTRACT_SUFX= .tar.xz # LGPLv3+ PERMIT_PACKAGE_CDROM= Yes -WANTLIB += m stdc++ +WANTLIB= m stdc++ -MASTER_SITES= ftp://ftp.gmplib.org/pub/${DI

Re: Fix crashes in Aqualung

Jeremy Evans: > This fixes an occasional crash when loading files in aqualung. > This is a fix to an earlier patch, which was taken from Aqualung's > bug tracker. Why do we need that patch at all? > -Use glib character conversion function instead of MAC library function, > -since the function

Re: www/nginx: switch from flavors to subpackages using dynamic module support

Jeremy Evans: > The four current FLAVORs get changed to subpackages, and the following > subpackages are added: There should be some upgrade path for pkg_add if you currently have one of the flavors installed. I wonder whether it would make sense to declare the subpackage with the corresponding

New: devel/libidn2

Here's a port of libidn2 for comments. (I thought we might need this for curl 7.51.0, but we'll take a different direction.) * Not sure about the category. devel (like libidn)? net? * COMMENT and in particular DESCR are very sparse. * I have disabled iconv support. Supposedly this is only

Re: amd64: chronic breakage

On 2016-10-19, Christian Weisgerber <na...@mips.inka.de> wrote: > lang/mono > lang/guile2 > net/libaccounts-glib Some additional ports that keep failing _sometimes_: devel/ocaml-menhir Segfault. I haven't paid attention whether it's always in the same place. Here's the

Update: www/w3m (0.5.3+git20161120)

This is a security update and complete overhaul of the www/w3m port. It uses the Debian version as upstream, which fixes 23 (!) CVEs, and follows the Debian package in some other respects. People who use w3m need to test this! Details: The w3m upstream has been long dead again. New

Re: Update: graphics/tiff 4.0.7

On 2016-11-24, Christian Weisgerber <na...@mips.inka.de> wrote: > Here's an update to graphics/tiff 4.0.7. > > Changes: > http://www.simplesystems.org/libtiff/v4.0.7.html > TL;DR: Numerous security fixes in the library; some tools were removed. > > Unfortunatel

Re: Update: graphics/tiff 4.0.7

Jeremie Courreges-Anglas: > > The removed symbols are due to _static_ declarations in 4.0.7. We > > could revert that in -stable to maintain ABI compatibility. Does > > that sound sensible? > > It would be safer and more correct so +1 if it doesn't mean too many > patches. Here's the full

Re: Update: www/w3m (0.5.3+git20161120)

On 2016-12-17, Christian Weisgerber <na...@mips.inka.de> wrote: > diff -u -p -r1.20 PLIST > --- pkg/PLIST 12 Mar 2011 23:20:17 - 1.20 > +++ pkg/PLIST 17 Dec 2016 22:41:52 - > @@ -1,4 +1,6 @@ > @comment $OpenBSD: PLIST,v 1.20 2011/03/12 23:20:17 sthen Exp

Re: Update: www/w3m (0.5.3+git20161120)

On 2016-12-17, Christian Weisgerber <na...@mips.inka.de> wrote: > This is a security update and complete overhaul of the www/w3m port. > It uses the Debian version as upstream, which fixes 23 (!) CVEs, > and follows the Debian package in some other respects. I have committ

Re: [UPDATE] devel/jdk/1.8 u112

Kurt Miller: > Here is an update of 1.8 to u112. > > Tested on amd64. Needs to be thrown in bulk builds on both amd64 and i386. No problems in an amd64 bulk build. -- Christian "naddy" Weisgerber na...@mips.inka.de

Locking up the ports tree

We're locking up the ports tree for the 6.1 release. If you have FIXES or URGENT changes that still should go in, you need to convince sthen@ or me. -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: [PATCH] lang/brandy (add an SDL flavor)

On 2017-03-24, Frederic Cambus wrote: > Here is a diff to add an SDL flavor to lang/brandy. Why make this a flavor? Why not just add it unconditionally to the port? -- Christian "naddy" Weisgerber na...@mips.inka.de

news/nn overhaul

Does anybody still use news/nn? Here's an overhaul of the port: * Get rid of the MACHINE_ARCH dance and use a single m-openbsd.h file. We only need to define a few basic types that are the same on all our archs. NETWORK_BYTE_ORDER could be gotten from , but it is only used with

www/w3m: fix duplicate declaration

If you build www/w3m on amd64+clang you get this error: /usr/bin/ld: terms.o: relocation R_X86_64_PC32 against `COLS ' can not be used when making a shared object; recompile with -fPIC That's because w3m defines global variables COLS and LINES that clash with those exported by libcurses. The

Re: amd64-clang port build failures

On 2017-04-13, Christian Weisgerber <na...@mips.inka.de> wrote: > To get us going, I have started running test builds on amd64 with > clang as the compiler. Logs of the build failures are available > here: Update: http://build-failures.rhaalovely.net/amd64-clang/2017-04

USE_CXX11 (was: Re: UPDATE: graphics/slop)

On 2017-04-13, Stuart Henderson wrote: > If it's for c++11 support you could use > > MODULES= gcc4 > MODGCC4_ARCHS=${GCC3_ARCHS} ${GCC4_ARCHS} > MODGCC4_LANGS=c++ > > I've been trying to think of a nice abstraction so we don't need to > copy that

amd64-clang port build failures

As you may know, on the arm64 architecture we use clang 4.0 as the system compiler. Other architectures for which clang support is available, like amd64, will follow sooner or later. Currently, a lot of the ports tree fails to build with clang. This must be fixed. To get us going, I have

Re: amd64-clang port build failures

Stuart Henderson: > Interesting, boost succeeded here. Was this using libestdc++ or > libc++? No, boost was skipped completely due to ONLY_FOR_ARCHS=${GCC4_ARCHS}. -- Christian "naddy" Weisgerber na...@mips.inka.de

Remove math/libneural?

math/libneural will need C++ fixes for clang. Or should we just remove it? - Dates from the turn of the century. - Appears abandoned. Home page/master site is down. - Nothing in the ports tree uses it. -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: amd64-clang port build failures

On 2017-04-16, Marc Espie wrote: >> * In the clang module, also link /usr/local/bin/clang and clang++ >> into the leading PATH directory. > > Should that actually be default ? it will make sure the clang from ports > is used on clang architectures, if that's the intention.

Re: amd64-clang port build failures

The amd64 and i386 base snapshots now include clang. You don't need to worry how to install clang, just install a snapshot! -- Christian "naddy" Weisgerber na...@mips.inka.de

devel/boost vs. clang

I tried to get boost to build with clang, but failed pretty miserably because I don't know what I'm doing. If anybody wants to look into this, here's what I got: * Configure with --with-toolset=clang * Fix boost_has_nl_types_h.ipp, from upstream. * Some haphazard changes to clang-linux.jam,

Re: Ports with hardcoded "gcc", "g++"

Updated list, 2017-03-04: devel/boost devel/jdk/1.7 emulators/virtualjaguar graphics/ipe lang/nim misc/rocrail net/pidgin productivity/wyrd x11/p5-Wx -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: imake: honor CC, CFLAGS, CXX, CXXFLAGS

On 2017-03-06, Christian Weisgerber <na...@mips.inka.de> wrote: > Ports that use CONFIGURE_STYLE=imake don't actually honor the CC, > CFLAGS, CXX, CXXFLAGS settings. They just pick up whatever is > hardcoded in devel/imake-cf. This behavior has grown historically > and there

Re: Ports with hardcoded "gcc", "g++"

Updated list, 2017-03-08: lang/jruby lang/nim x11/p5-Wx -- Christian "naddy" Weisgerber na...@mips.inka.de

devel/boost: honor CC, CXX

Here's a patch to make devel/boost honor CC and CXX. Also CFLAGS and CXXFLAGS in some invisible places. There are three parts to this: (1) Pass CXX and CXXFLAGS to the configure script run in libs/config. (2) Separate out the build of the bjam tool from the bootstrap procedure. This is

imake: honor CC, CFLAGS, CXX, CXXFLAGS

Ports that use CONFIGURE_STYLE=imake don't actually honor the CC, CFLAGS, CXX, CXXFLAGS settings. They just pick up whatever is hardcoded in devel/imake-cf. This behavior has grown historically and there is really no good reason for it. Here's a patch to change it. ok? Index:

Re: Ports with hardcoded "gcc", "g++"

Updated list, 2017-03-05: devel/boost java/jna lang/jruby lang/nim net/pidgin x11/p5-Wx -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: Remove audio/gogo?

Frederic Cambus: > The version we have in ports is from 2000, gogo is dead upstream, and > this is an i386 only port. Nothing depends on it. > OK to remove it? ok But for courtesy's sake let's see what Bob says, who probably doesn't even remember that his name is on this piece of cruft... --

Re: Ports with hardcoded "gcc", "g++"

Juan Francisco Cantero Hurtado: > > ===> Building for nim-0.16.0p0 > > cd /usr/obj/nim-0.16.0/nim-0.16.0 && /usr/bin/env -i CC="gcc" LINKER="gcc" > > CFLA > > GS="-O2 -pipe " sh build.sh > > gcc -O2 -pipe -w -fno-strict-aliasing -Ic_code -c c_code/3_2/compiler_nim.c > > -o c > >

Re: Ports with hardcoded "gcc", "g++"

Juan Francisco Cantero Hurtado: > > I changed the variable to NIM_CC. OK? > > And now with "elif" -> "else" fixed. There is still confusion about CC and NIM_CC: ===> Building for nim-0.16.0p0 cd /usr/obj/nim-0.16.0/nim-0.16.0 && /usr/bin/env -i CC="gcc" LINKER="gcc" CFLA GS="-O2 -pipe " sh

Re: Ports with hardcoded "gcc", "g++"

Juan Francisco Cantero Hurtado: > They have a simple option to change the compilers but we need a variable > with the realname of the compiler, i.e. clang or gcc. > > You can use "nim c -cc:clang" or "nim c -cc:gcc" (the default) or "nim c > -cc:egcc". All of them are the name of the profile,

Re: dpb runs three parallel processes on 2 CPUs

On 2017-04-13, Andreas Kusalananda Kähäri wrote: > I have 2 CPUs, and dpb usually builds two ports at a time (or one port > in parallel). However, while rebuilding llvm and ghc I observed that > ghc was being built as one job while llvm clearly used a parallel build >

devel/boehm-gc: add aarch64 support

devel/boehm-gc compiles on arm64, but it doesn't actually work. The regression tests all fail. There is support for aarch64, but not on OpenBSD. The patch below copies the required pieces into place. With this, the regression tests all succeed. ok? Index: Makefile

Re: amd64-clang port build failures

Sebastien Marie: > I am also surprised that lang/rust succeed (and devel/cargo didn't). lang/rust builds with gcc 4.9, so it sidesteps clang. > Is it possible to have more information on the build configuration ? > Stuart already mentioned libestdc++ vs libc++ for example. I have the base

Re: NEW PORT: systemd

On 2017-07-31, Marc Espie wrote: > Port is attached. Needs more sparkle. -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: Port bulk with

On 2017-08-09, Martin Pieuchot wrote: > So here's a first step, introducing /usr/include/elf.h. Could some of > you run a bulk with it and report the possible breakages? This breaks devel/libelf, which indirectly takes out about a quarter of the tree.

clang/amd64 fallout (2017-08-10)

Here's the list of ports still broken on amd64 due to the clang switch: audio/festival/core devel/glog devel/mico devel/stp devel/ti-msp430gcc editors/TeXmacs emulators/emulationstation games/frozen-bubble games/vacuum graphics/simgear multimedia/avidemux security/encfs sysutils/memtest86+ Build

Update+new: audio/speex 1.2.0, audio/speexdsp

HARED_LIBS += speex8.0 # .6.0 -SHARED_LIBS += speexdsp 1.0 # .6.0 +HOMEPAGE= https://speex.org/ +SHARED_LIBS= speex 8.0 # 6.1 MAINTAINER= Christian Weisgerber <na...@openbsd.org> @@ -16,15 +14,15 @@ PERMIT_PACKAGE_CDROM= Yes MASTER_SITES= http://do

Re: Update+new: audio/speex 1.2.0, audio/speexdsp

On 2017-08-13, Christian Weisgerber <na...@mips.inka.de> wrote: > Since libspeex and libspeexdsp were already separate libraries > before, and speex depends on speexdsp, the dependent ports should > be fine. It turns out that ports which have a LIB_DEPENDS on audio/speex but only

libpng.pc

The recent graphics/png 1.6.31 update added a line Requires: zlib to libpng.pc. This makes sense as libpng requires libz. Consequently, the pkg-config output changes: pkg-config --cflags libpng old: -I/usr/local/include/libpng16 new: -I/usr/local/include/libpng16 -I/usr/include pkg-config

clang/amd64 fallout: sysutils/memtest86+

sysutils/memtest86+ fails to build with clang. That can be hacked around, but the resulting memtest crashes. (For the curious, I'll append the required changes below.) memtest86+ builds with gcc4.9, but again, the resulting executable crashes. We could use CC=/usr/bin/gcc, but that will likely

amd64-clang port build failures 2017-04-27

I am running test builds on amd64 with clang as the compiler. The latest logs of the build failures are available here: http://build-failures.rhaalovely.net/amd64-clang/2017-04-27/ -- Christian "naddy" Weisgerber na...@mips.inka.de

clang/amd64 fallout

Here's the corresponding list of bulk build failures on amd64. I've uploaded the logs to http://build-failures.rhaalovely.net/amd64/2017-07-27/ audio/festival/core devel/arm-none-eabi/gcc-linaro devel/glog devel/mico devel/mono-addins devel/p5-Alien-wxWidgets devel/stp devel/ti-msp430gcc

clang/amd64 fallout: net/argus

net/argus fails to build with clang on amd64: cc -O2 -pipe -I. -I./../include -DHAVE_CONFIG_H -DSYSCONFDIR=\"/etc\" -o ../bin/argus argus.o ArgusModeler.o ArgusSource.o ArgusUtil.o ArgusOutput.o ArgusUdp.o ArgusTcp.o ArgusIcmp.o ArgusIgmp.o ArgusEsp.o ArgusArp.o ArgusFrag.o ArgusUdt.o

Re: NEW: net/mcast-proxy

On 2017-07-27, Florian Obser wrote: >> mcast-proxy.c:662:30: warning: taking address of packed member 'ip6_dst' of >> class or structure 'ip6_hdr' may result in an unaligned pointer value >> [-Waddress-of-packed-member] > > We have similar warnings in base in rtadvd(8). In

clang/amd64 fallout (2017-08-03)

Here's the list of ports still broken on amd64 due to the clang switch: audio/festival/core devel/glog devel/mico devel/stp devel/ti-msp430gcc editors/TeXmacs emulators/emulationstation games/frozen-bubble games/vacuum graphics/simgear multimedia/avidemux net/isc-dhcp security/encfs

Re: Remove benchmarks/bytebench?

On 2017-05-01, Ryan Freeman wrote: > i came across 'unixbench'; a version of bytebench > 'updated and revised by many people over the years'. > > https://github.com/meteorfox/byte-unixbench/ > > If there is any sort of interest I can whip it into a port and submit. I don't

amd64-clang port build failures 2017-05-11

Here are the error logs from this week's amd64 bulk build with clang: http://build-failures.rhaalovely.net/amd64-clang/2017-05-11/ -- Christian "naddy" Weisgerber na...@mips.inka.de

Update: net/ncftp 3.2.6, drop IPv6 patch

-3.2.5 +DISTNAME= ncftp-3.2.6 CATEGORIES=net HOMEPAGE= http://www.ncftp.com/ncftp/ -MAINTAINER=Christian Weisgerber <na...@openbsd.org> - # Artistic PERMIT_PACKAGE_CDROM= Yes WANTLIB= c ncurses -MASTER_SITES= ftp://ftp.ncftp.com/ncftp/ \ -

Re: NEW: games/spaceinvaders

On 2017-05-17, Marcus Glocker wrote: > This is a display-based version of the Space Invaders game. What does "display-based" mean? -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: sysclean question

On 2017-05-07, Amit Kulkarni wrote: > When I try to upgrade a May 3 binaries to May 6 binaries on amd64, The May 6 snapshot is corrupt. It lacks lib{crypto,ssl,tls}.so.*. I'm not familiar with sysclean, but I'm fairly confident the rest follows from that. -- Christian

amd64-clang port build failures 2017-05-04

http://build-failures.rhaalovely.net/amd64-clang/2017-05-04/ The idea is that people start picking some low-hanging fruit, so that we can slowly shrink the list. -- Christian "naddy" Weisgerber na...@mips.inka.de

Remove benchmarks/bytebench?

benchmarks/bytebenchBYTE magazine benchmark suite The timestamps in the distfile are from February 1992. (BYTE magazine folded in 1998.) I was going to add some fixes to make this compile with clang, but there's no point. This long forgotten cruft can just be removed, ok? -- Christian

Re: UPDATE: devel/ninja

Giovanni Bechis: > > please find below a simple update diff for ninja. > > > > - add write_fake_manifests.py and measure.py to PLIST. > > - add NO_TEST otherwise there is an error if run regress. > > > added a do-test target to fix regression tests. There were no problems with this in an amd64

Re: Port bulk with

Karel Gardas: > I've update my elf.h transition patch and posted on tech@ here > https://marc.info/?l=openbsd-tech=150551268819592=2 -- I'm running This patch doesn't apply. There is no file src/lib/libc/gen/elf_hash.c in -CURRENT. -- Christian "naddy" Weisgerber

Re: Port bulk with

Karel Gardas: > I think it should be just enough if you remove this problematic hunk: Building a release fails. I should not be the one discovering this when you are asking for a ports build. cc -o rdsetroot /usr/src/distrib/amd64/ramdisk_cd/../../common/elfrdsetroot.c

Re: [UPDATE] lang/python/2.7

Remi Pointel: > this is the diff to update Python 2.7 to latest release: 2.7.14. > > Could it be tested in a bulk build please? There were no problems on amd64. -- Christian "naddy" Weisgerber na...@mips.inka.de

Re: UPDATE: graphics/png

Rafael Sadowski: > update png to 1.6.32. This release includes a security patch: > > Ok? Comments? ok naddy@ This should also be committed to 6.2-stable. The latest upstream version is in fact 1.6.34, but that isn't security relevant, so we can look at it later. -- Christian "naddy"

Re: Port bulk with

Martin Pieuchot: > > On amd64, two ports failed to build: devel/libdwarf and devel/valgrind. > > Thanks, here's an updated diff that should fix those. Nope, those two still fail. Full logs attached. ===> libdwarf c++ -I/usr/local/include/libelf -I/usr/local/include

Clang 5.0 fallout

Port build failures on amd64 -current after the clang 5.0 update: devel/valgrind games/qgo geo/gpstk math/freemat x11/kde/libs3 valgrind errors out in configure because it doesn't recognize the compiler version: configure: error: please use gcc >= 3.0 or clang >= 2.9 The other ports all suffer

Re: Port bulk with

Martin Pieuchot: > > So here's a first step, introducing /usr/include/elf.h. Could some of > > you run a bulk with it and report the possible breakages? > > Now that the offending function declaration has been remove from libc > libelf builds as before. > > Could you guys tell us if there's

Ports lock imminent

We're going to lock the ports tree for the release any day now. Please, no more imports or regular updates. Issues I'd like to see taken care of before the release: * Fix the remaining instances of nested functions in autoconf checks. * Pick up Firefox 56. * Mark as BROKEN the remaining ports

Re: ports broken on clang archs (was: Re: Ports lock imminent)

Jeremie Courreges-Anglas: > > * Mark as BROKEN the remaining ports that have failed to build since > > the clang switch on amd64/i386. > > Should probably use NOT_FOR_ARCHS = ${CLANG_ARCHS}. > > http://exopi.exo.bsdfrog.org/logs/old/amd64/2017-09-26T21:25:35+0200/paths/x11/kde4/cantor.log

Re: [security] libidn & libidn2 fixes for CVE-2017-14061 and CVE-2017-14062

On 2017-10-01, Jeremie Courreges-Anglas wrote: > --- /dev/null 1 Jan 1970 00:00:00 - > +++ devel/libidn/patches/patch-lib_punycode_c 1 Oct 2017 22:06:28 - > @@ -0,0 +1,27 @@ > +$OpenBSD$ > + > +commit e9e81b8063b095b02cf104bb992fa9bf9515b9d8 > +Author: Tim Rühsen

Re: aarch64 bulk build report

Juan Francisco Cantero Hurtado: > > bulk build on arm64.ports.openbsd.org > > build failures: 232 There is a lot to fix on aarch64, but it's too late for the release. > > http://build-failures.rhaalovely.net//aarch64/2017-09-08/lang/gambit.log > > MAXSSIZ is ((paddr_t)8*1024*1024) on arm64.

Re: lastpass-cli segfaulting

Stuart Henderson: > REVISION goes to 0 first. Add the upstream commit information to the > patch. > > I think this should probably go in if there's still time.. What do > you think naddy? Obvious fix, ok > Index: Makefile > === >

Re: Remaining clang fallout (amd64 2017-08-27)

Christian Weisgerber: > These ports still fail to build on amd64 after the clang switch: > > devel/ti-msp430gcc > editors/TeXmacs > emulators/emulationstation > games/frozen-bubble > games/vacuum > multimedia/avidemux PS: Logs at http://build-failures.rhaalov

Remaining clang fallout (amd64 2017-08-27)

These ports still fail to build on amd64 after the clang switch: devel/ti-msp430gcc editors/TeXmacs emulators/emulationstation games/frozen-bubble games/vacuum multimedia/avidemux -- Christian "naddy" Weisgerber na...@mips.inka.de

Ports using /dev/arandom

The /dev/arandom device was a detour that we are trying to get rid of. Nothing should use it any longer. Code should preferably use the arc4random() family of functions to get random numbers. If that isn't an option, e.g. in shell, use /dev/urandom. I extracted all packages from an amd64

Re: [UPDATE] lang/python/3.6

Remi Pointel: > this is the diff to update Python to latest release 3.6.4. > > Could someone please run a bulk build with this diff please? I have started an amd64 bulk build with this. -- Christian "naddy" Weisgerber na...@mips.inka.de

Gettext cleanup next step

Regarding the ongoing effort to get rid of the gettext module, the current state is that only "simple" ports still use it. No MULTI_PACKAGES, no FLAVORs. My next step will be to check all remaining ports whether they need gettext-tools and then move the BUILD_DEPENDS out of the module and into

Re: Upcoming firefox 57 release notes

On 2017-11-04, Landry Breuil wrote: > - and something which has been brought to my attention, the (somewhat > hidden imo) feature of 'pasting an url anywhere in the firefox window > would open it in a new tab' has been disabled by default, because it was (In the same tab.)

Re: Gettext cleanup next step

On 2017-11-20, Christian Weisgerber <na...@mips.inka.de> wrote: > Just a quick heads-up that I now have a plan for removing the last > uses of the gettext module. This will happen over the next few > days. Should be all done now. -- Christian "naddy" Weisgerber

Re: Gettext cleanup next step

On 2017-11-17, Christian Weisgerber <na...@mips.inka.de> wrote: > My next step will be to check all remaining ports whether they need > gettext-tools and then move the BUILD_DEPENDS out of the module and > into the ports that actually require it. Done. -- Christian &qu

Re: Gettext cleanup next step

Just a quick heads-up that I now have a plan for removing the last uses of the gettext module. This will happen over the next few days. -- Christian "naddy" Weisgerber na...@mips.inka.de

Remove x11/xdmchoose

When xdm is used to handle access to remote X11 servers, it can offer a chooser menu that allows selecting which host to connect to. x11/xdmchoose is a replacement for the default xdm chooser. Since the ability to manage remote X11 servers has been completely stripped from xenodm, there is no

Re: Gettext cleanup next step

Jeremie Courreges-Anglas: > I have an update (with an ok from maintainer) for devel/quilt, I can > handle the gettext removal too. Oh sure, I don't want to hold up work on individual ports. Just people who specifically want to remove the gettext module from a number of ports should talk to me

Remove mail/mixmaster?

mail/mixmaster (client for anonymous remailing) is fourteen years old. I don't think that security~crypto software that has been unmaintained for that long is still useful. Time to delete it? -- Christian "naddy" Weisgerber na...@mips.inka.de

Remaining /dev/srandom users fixed

After the removal of /dev/srandom, it turned out that this device wasn't as entirely unused as I had thought. I extracted all packages from an amd64 snapshot and ran "fgrep -a /dev/srandom" over the contents to identify all remaining /dev/srandom users. Today I went through the list and fixed

<    7   8   9   10   11   12   13   14   15   16   >