synapse filepaths + rc script query
Hi Renaud,
Many thanks for your work on synapse. I'm upgrading one of my machines
to 6.7 and am pleased I no longer have to maintain my own
virtualenv-based synapse.
As I am preparing a port for mautrix-hangouts, I would like to check a
couple of decisions in the synapse port that differed from what I had
chosen in my local version. They are really very minor nitpicks, but I'd
like to ask your reasoning so I may be consistent:
- config files stored in /var/synapse; I had these all in /etc/synapse
- pid file stored in /var/synapse/homeserver.pid; I had this in
/var/run/synapse/homeserver.pid
Lastly, instead of relying on a custom python init system (synctl), I
found the builtin rc.subr setup worked well:
> #!/bin/ksh
>
> daemon="/usr/local/bin/python3"
> daemon_flags="-m synapse.app.homeserver -c /etc/synapse/homeserver.yaml
> --daemonize"
> daemon_user="_synapse"
>
> . /etc/rc.d/rc.subr
>
> rc_pre() {
> install -d -o _synapse -g _synapse /var/run/synapse/
> }
>
> rc_cmd $1
Cheers,
Tom
Re: UPDATE: VLC 3.0.11.1
On Sat, 8 Aug 2020 11:34:47 -0400 Brad Smith wrote: > Here is an update to VLC 3.0.11.1. This is good on macppc. I built it along with your updates to FFmpeg 4.3.1 and libvpx 1.9.0. (My macppc still has LLVM 8.) There are no regressions when, in vlc, I play a .webm file with VP8 video and Vorbis audio. I use XVideo (XCB) to play video in vlc, because the default (OpenGL?) freezes the kernel. The video looks and sounds correct, except that my macppc cpu is too slow: a few video frames are blocky, and the audio sometimes has short pauses.
Re: purritobin-0.1.2 - new package + dependencies
On 7/5/20 5:58 AM, Aisha Tammy wrote: > Bump, > >> On 2020/06/23 23:33, Aisha Tammy wrote: >>> On 6/22/20 6:56 AM, Stuart Henderson wrote: On 2020/06/21 21:32, Aisha Tammy wrote: >> I've added them again with the two from Brian and the uwebsockets >> updated to 18.3.0 > > Another bump. > Also updated uwebsockets to 18.4.0 OK sthen@ if someone wants to import uwebsockets/usockets with usockets moved from devel to net, tgz attached. For purritobin, could you install the README under share/doc and maybe also paste.html and POSIX_shell_client.sh (either in share/examples or somewhere else). Might be a good idea to add an rc script and a userid to run it under (@newuser in plist) too? >>> >>> I've attached purritobin with the README and an rc file. >>> Changes >>> - new user and group >>> >>> - rc file running as new user >>> - added paste.html and shell clients to /usr/local/share/purritobin/ >>> - added pkg-readme with examples for httpd, cron cleanup and pf rate >>> limiting >>> >>> Also needs that uwebsockets and usockets be updated to latest version as the >>> current ones in the tree are broken (upstream messed up some SSL >>> declarations). >>> I've added the diff for that as well. >> >> I've committed those, diff didn't work as it had extra blank lines >> so I did them by hand, SHARED_LIBS needed a bump too due to additional >> functions. >> >> purritobin looks good now, ok if someone would like to import (wonder if >> it might be better in www rather than net though) >> > Another bump. There's one OK, so I think it should be good. Thanks a lot, Aisha purritobin.tgz Description: application/compressed-tar
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/09 18:56:40 Modified files: lang/luajit: Makefile lang/luajit/pkg: PLIST Log message: Link libluajit.so with libc++abi to unbreak rspamd with newer llvm luajit uses symbols from libunwind which is embedded in libc++abi. Use the appropriate knob to link libluajit.so (not just the interpreter) with libc++abi and register the shared lib dep. ok sthen@ (rspamd maintainer)
NEW py-synapse-imap
Hi, I've attached a new port py-synapse-imap which is a IMAP authentication provider for synapse Its a lot more convenient to use IMAP when creating users for multiple different services so I'd like to get this port in the tree. Thanks, Aisha py-synapse-imap.tgz Description: application/compressed-tar
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/09 18:47:44 Modified files: sysutils/xxhash: Makefile Log message: Don't clobber CFLAGS to let this build on i386 with newer llvm Upstream's (GNU)Makefile properly amends CFLAGS on a per-target basis, but that can't happen if we override CFLAGS on gmake's command line. We already provide CFLAGS through MAKE_ENV, let the Makefile do its job and provide -fPIC where needed. ok bket@ (maintainer)
aarch64 bulk build report
bulk build on arm64.ports.openbsd.org started on Sat Aug 8 16:30:26 MDT 2020 finished at Sun Aug 9 17:58:55 MDT 2020 lasted 1D01h28m done with kern.version=OpenBSD 6.7-current (GENERIC.MP) #757: Sat Aug 8 03:51:17 MDT 2020 built packages:7766 Aug 8:3082 Aug 9:4683 critical path missing pkgs: http://build-failures.rhaalovely.net/aarch64/2020-08-08/summary.log build failures: 12 http://build-failures.rhaalovely.net/aarch64/2020-08-08/devel/py-llvmlite,python3.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/devel/rttr.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/editors/xwpe.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/lang/pfe.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/net/bro.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/net/libtorrent-rasterbar.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/security/libgcrypt.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/sysutils/nomad.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/sysutils/rclone.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/sysutils/telegraf.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/sysutils/terragrunt.log http://build-failures.rhaalovely.net/aarch64/2020-08-08/textproc/source-highlight.log recurrent failures failures/devel/py-llvmlite,python3.log failures/devel/rttr.log failures/editors/xwpe.log failures/lang/pfe.log failures/net/bro.log failures/net/libtorrent-rasterbar.log failures/security/libgcrypt.log failures/sysutils/nomad.log failures/sysutils/rclone.log failures/sysutils/telegraf.log failures/sysutils/terragrunt.log new failures +++ ls-failures Sun Aug 9 17:59:06 2020 +failures/textproc/source-highlight.log resolved failures --- ../old/aarch64/last//ls-failuresFri Aug 7 13:32:24 2020 -failures/devel/woboq_codebrowser.log -failures/net/rsync,iconv.log -failures/shells/nsh.log
gimp install - Can't find gcc-libs-8.3.0p5
amd64 recent snapshot install, getting this error. Checking bulk build report, last, there is no mention of this package as not being able to compile. Did I overlooked? Also another error, first time when i see this: https://ftp2.eu.openbsd.org/pub/OpenBSD/snapshots/packages/amd64/lame-3.100p1.tgz: signify: signature truncated Thanks.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/09 16:19:32 Modified files: www/urlwatch : Makefile distinfo www/urlwatch/pkg: PLIST Log message: update to urlwatch-2.21
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/09 16:17:48 Modified files: textproc/p5-PDF-API2: Makefile distinfo Log message: update to p5-PDF-API2-2.037
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/09 16:12:36 Modified files: devel/re2c : Makefile distinfo devel/re2c/pkg : PLIST Log message: update to re2c-2.0.2
Re: add notmuch flavor to mail/neomutt
Please either send directly to the maintainer, or at least CC the maintainer,
on ports changes.
On 2020/08/09 12:03, Olivier Taïbi wrote:
> diff -uNpr /usr/ports/mail/neomutt/Makefile
> /build/usr/ports/mail/neomutt/Makefile
Not sure what went wrong, but the emailed diff doesn't apply with patch(1).
> --- /usr/ports/mail/neomutt/Makefile Tue Jul 28 23:54:23 2020
> +++ /build/usr/ports/mail/neomutt/MakefileFri Aug 7 23:46:50 2020
> @@ -26,7 +26,7 @@ BUILD_DEPENDS+= textproc/docbook-xsl \
> DEBUG_PACKAGES= ${BUILD_PACKAGES}
>
> SEPARATE_BUILD= Yes
> -FLAVORS= gpgme sasl
> +FLAVORS= gpgme sasl notmuch
> FLAVOR?=
>
> CONFIGURE_STYLE= simple
> @@ -63,6 +63,12 @@ WANTLIB+= assuan gpg-error gpgme sqlite3
> CONFIGURE_ARGS+= --sasl
> LIB_DEPENDS+=security/cyrus-sasl2
> WANTLIB+=sasl2
> +.endif
> +
> +.if ${FLAVOR:Mnotmuch}
> +CONFIGURE_ARGS+= --notmuch
> +LIB_DEPENDS+=mail/notmuch/notmuch,-main
> +WANTLIB+=notmuch
> .endif
>
> post-install:
> diff -uNpr /usr/ports/mail/neomutt/patches/patch-notmuch_private_h
> /build/usr/ports/mail/neomutt/patches/patch-notmuch_private_h
> --- /usr/ports/mail/neomutt/patches/patch-notmuch_private_h Thu Jan 1
> 01:00:00 1970
> +++ /build/usr/ports/mail/neomutt/patches/patch-notmuch_private_h Fri Aug
> 7 23:40:05 2020
> @@ -0,0 +1,20 @@
> +$OpenBSD$
> +
> +Index: notmuch/private.h
> +--- notmuch/private.h.orig
> notmuch/private.h
> +@@ -34,11 +34,7 @@
> + #endif
> +
> + /* The definition in is broken */
Funny because it looks like notmuch.h has the same
LIBNOTMUCH_CHECK_VERSION macro, just with different whitespace..
> +-#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro)
> \
> +- (LIBNOTMUCH_MAJOR_VERSION > (major) ||
> \
> +- (LIBNOTMUCH_MAJOR_VERSION == (major) && LIBNOTMUCH_MINOR_VERSION >
> (minor)) || \
> +- (LIBNOTMUCH_MAJOR_VERSION == (major) &&
> \
> +-LIBNOTMUCH_MINOR_VERSION == (minor) && LIBNOTMUCH_MICRO_VERSION >=
> (micro)))
> ++#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro) (1>0)
This suggests that the way the so version was set in notmuch isn't
ideal, perhaps it would be better to do this instead so the check
should then work as expected?
Index: notmuch/Makefile
===
RCS file: /cvs/ports/mail/notmuch/notmuch/Makefile,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile
--- notmuch/Makefile19 May 2020 20:45:49 - 1.2
+++ notmuch/Makefile9 Aug 2020 22:03:09 -
@@ -6,6 +6,8 @@ COMMENT-emacs = Emacs bindings for notm
PKGNAME-main = notmuch-${V}
PKGNAME-emacs =notmuch-emacs-${V}
+REVISION-main =0
+
SHARED_LIBS += notmuch 0.0
MULTI_PACKAGES = -main -emacs
@@ -68,7 +70,7 @@ post-extract:
# Avoid conflict with C++20 , also see patches
pre-configure:
cd ${WRKSRC}; mv version version.txt; \
- ${SUBST_CMD} lib/notmuch.h
+ ${SUBST_CMD} configure
ln -s ${PREFIX}/bin/sphinx-build${MODPY_BIN_SUFFIX}
${WRKDIR}/bin/sphinx-build
post-install:
Index: notmuch/patches/patch-configure
===
RCS file: /cvs/ports/mail/notmuch/notmuch/patches/patch-configure,v
retrieving revision 1.2
diff -u -p -r1.2 patch-configure
--- notmuch/patches/patch-configure 16 May 2020 11:43:51 - 1.2
+++ notmuch/patches/patch-configure 9 Aug 2020 22:03:09 -
@@ -24,3 +24,25 @@ Index: configure
printf "Yes.\n"
have_zlib=1
zlib_cflags=$(pkg-config --cflags zlib)
+@@ -1076,18 +1076,18 @@ vpath %.rst \$(srcdir)
+ # the library interface, (such as the deletion of an API or a major
+ # semantic change that breaks formerly functioning code).
+ #
+-LIBNOTMUCH_VERSION_MAJOR = ${libnotmuch_version_major}
++LIBNOTMUCH_VERSION_MAJOR = ${LIBnotmuch_MAJOR}
+
+ # The minor version of the library interface. This should be incremented at
+ # the time of release for any additions to the library interface,
+ # (and when it is incremented, the release version of the library should
+ # be reset to 0).
+-LIBNOTMUCH_VERSION_MINOR = ${libnotmuch_version_minor}
++LIBNOTMUCH_VERSION_MINOR = ${LIBnotmuch_MINOR}
+
+ # The release version the library interface. This should be incremented at
+ # the time of release if there have been no changes to the interface, (but
+ # simply compatible changes to the implementation).
+-LIBNOTMUCH_VERSION_RELEASE = ${libnotmuch_version_release}
++LIBNOTMUCH_VERSION_RELEASE = 0
+
+ # These are derived from the VERSION macros in lib/notmuch.h so
+ # if you have to change them, something is wrong.
Index: notmuch/patches/patch-lib_notmuch_h
===
RCS file: notmuch/patches/patch-lib_notmuch_h
diff -N notmuch/patches/patch-lib_notmuch_h
---
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: sebas...@cvs.openbsd.org2020/08/09 15:20:52 Modified files: devel/cpp-hocon: Makefile distinfo Log message: simple update 0.2.0 > 0.3.0
NEW: games/mkxp
Hi ports@,
Attached is a new port for mkxp [1] which is an implementation of the Ruby
Game Scripting System used in "RPG Maker XP", "RPG Maker VX" and "RPG Maker
VX Ace" games.
This port needs a fork [2] of devel/sdl-sound that has SDL2 support and is
only intended to be used by mkxp. It is also attached as a new port
(devel/sdl-sound-mkxp).
Notes:
- Both ports pass port-lib-depends-check and portcheck.
- Built them on amd64/arm64 and tested mkxp on amd64 with some games.
- Both ports have no releases, so I used the GH_* variables.
- mkxp looks for ruby 2.1 but is now using the default ruby version in
ports (currently 2.6, mentioned in ruby-module(5)).
- mkxp has vim as a BDEP because it needs xxd(1).
- Some instructions about running RPG Maker {XP,VX,VX Ace} games are
included in mkxp's pkg README.
- The sdl-sound-mkxp port is based on sdl-sound and most of the changes
made by the patches of sdl-sound are already included. Only patch-configure_in
is still present on sdl-sound-mkxp and patch-playsound_physfsrwops_h is
not included because the playsound binary is not being built.
- Originally sdl-sound and its fork has conflicts with some filenames but
I took care of it in the Makefile by renaming SDL_sound to SDL_sound_mkxp
for those files and made a similar change in mkxp's Makefile to look for
SDL_sound_mkxp instead of SDL_sound.
Comments? Thoughts?
[1] https://github.com/Ancurio/mkxp
[2] https://github.com/Ancurio/SDL_sound
mkxp.tar.gz
Description: application/tar-gz
sdl-sound-mkxp.tar.gz
Description: application/tar-gz
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: sebas...@cvs.openbsd.org2020/08/09 14:46:06 Modified files: sysutils/ruby-puppet/5: Makefile distinfo Log message: update 5.5.20 -> 5.5.21
Re: mail/rspamd and luajit (was: Re: amd64 bulk build failures 2020-08-07)
> From: Jeremie Courreges-Anglas
> Date: Sun, 09 Aug 2020 12:23:05 +0200
>
> > mail/rspamd
>
> http://build-failures.rhaalovely.net/amd64/2020-08-07/mail/rspamd.log
>
> --8<--
> ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
> _Unwind_DeleteException
> ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
> _Unwind_GetCFA
> ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
> _Unwind_RaiseException
> ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
> _Unwind_SetGR
> ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
> _Unwind_SetIP
> -->8--
>
> This is fixed by linking libluajit.so with libc++abi (which ships the
> required the required unwind symbols). TARGET_LDFLAGS were only used to
> link the luajit interpreter, not the library. I suspect that a few other
> ports can benefit from this diff.
>
> I have no idea why ld.lld starts looking at undefined symbols in library
> dependencies (cc'ing kettenis so that he knows about this case).
Because it is the right thing to do. The old ld.lld already did this
so I don't fully understands why this suddenly breaks the build.
Maybe there is a corner case where the check wasn't performed properly
that is now fixed.
> ok?
>
>
> Index: Makefile
> ===
> RCS file: /cvs/ports/lang/luajit/Makefile,v
> retrieving revision 1.30
> diff -u -p -r1.30 Makefile
> --- Makefile 12 Jul 2019 20:47:19 - 1.30
> +++ Makefile 9 Aug 2020 10:17:32 -
> @@ -11,7 +11,7 @@ SHARED_LIBS += luajit-${MODLUA_VERSION}
>
> COMMENT =just-in-time compiler for Lua
> V = 2.0.5
> -REVISION = 1
> +REVISION = 2
> DISTNAME = LuaJIT-${V}
> PKGNAME =${DISTNAME:L}
>
> @@ -61,7 +61,7 @@ do-test:
> .include
>
> .if ${PROPERTIES:Mclang}
> -MAKE_FLAGS += TARGET_LDFLAGS='-lc++abi -lpthread'
> +MAKE_FLAGS += TARGET_LIBS='-lc++abi -lpthread'
> WANTLIB += c++abi pthread
> .endif
>
> Index: pkg/PLIST
> ===
> RCS file: /cvs/ports/lang/luajit/pkg/PLIST,v
> retrieving revision 1.2
> diff -u -p -r1.2 PLIST
> --- pkg/PLIST 3 May 2014 14:24:03 - 1.2
> +++ pkg/PLIST 9 Aug 2020 10:17:32 -
> @@ -8,12 +8,10 @@ include/luajit-2.0/luaconf.h
> include/luajit-2.0/luajit.h
> include/luajit-2.0/lualib.h
> @comment lib/libluajit-${MODLUA_VERSION}.so
> -lib/libluajit-${MODLUA_VERSION}.a
> +@static-lib lib/libluajit-${MODLUA_VERSION}.a
> @lib
> lib/libluajit-${MODLUA_VERSION}.so.${LIBluajit-${MODLUA_VERSION}_VERSION}
> -lib/lua/${MODLUA_VERSION}/
> lib/pkgconfig/luajit.pc
> @man man/man1/luajit.1
> -share/lua/${MODLUA_VERSION}/
> share/lua/${MODLUA_VERSION}/jit/
> share/lua/${MODLUA_VERSION}/jit/bc.lua
> share/lua/${MODLUA_VERSION}/jit/bcsave.lua
>
>
> --
> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
>
[sparc64/base-gcc] Fix build of devel/keystone/python
The build of devel/keystone/python fails with
-- Looking for __atomic_fetch_add_4 in atomic - not found
CMake Error at llvm/cmake/modules/CheckAtomic.cmake:36 (message):
Host compiler appears to require libatomic, but cannot find it.
Adding a COMPILER line to use ports-gcc on base-gcc arches fixes it.
ok?
--Kurt
Index: python/Makefile
===
RCS file: /cvs/ports/devel/keystone/python/Makefile,v
retrieving revision 1.6
diff -u -p -r1.6 Makefile
--- python/Makefile 8 Aug 2020 07:39:25 - 1.6
+++ python/Makefile 9 Aug 2020 19:52:30 -
@@ -10,6 +10,8 @@ MODULES = devel/cmake \
lang/python
MODPY_SETUPTOOLS =Yes
+COMPILER = base-clang ports-gcc ports-clang
+
WRKDIST = ${WRKDIR}/${DISTNAME}/bindings/python
FLAVORS = python3
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: juan...@cvs.openbsd.org 2020/08/09 13:44:28 Modified files: lang/racket-minimal: Makefile distinfo lang/racket-minimal/pkg: PLIST Removed files: lang/racket-minimal/patches: patch-src_ChezScheme_s_cpnanopass_ss patch-src_ChezScheme_s_x86_64_ss patch-src_cs_c_Makefile_in patch-src_cs_c_embed-boot_rkt patch-src_racket_src_optimize_c Log message: Update to racket 7.8.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: juan...@cvs.openbsd.org 2020/08/09 13:39:15 Modified files: emulators/xroar: Makefile distinfo Log message: Update to xroar 0.36.2.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: juan...@cvs.openbsd.org 2020/08/09 13:38:18 Modified files: devel/mercurial: Makefile distinfo devel/mercurial/pkg: PLIST-main Log message: Update to mercurial 5.5.
Re: rsync: update to 3.2.2
On 2020-08-04, Klemens Nanni wrote: > I also looked at iconv but rsync configures and builds with it even when > --disable-iconv gets passed, this smells like some configure* bug. 3.2.2p0 now does not pick up iconv. I quickly looked over configure.ac, and as far as I can tell (1) detection that iconv is available and how, and (2) the choice to enable --iconv are independent of each other. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: NEW: net/i2p
Ping On Sun, August 2, 2020 18:44, Dimitri Karamazov wrote: > I2P software bundle, includes the standard I2P implementation. > https://geti2p.net/ > > > About: > *I2P is an anonymous overlay network - a network within a network. > *It is intended to protect communication from dragnet surveillance > and monitoring by third parties such as ISPs. *Is decentralised completely > unlike tor. Each client routes > anonymous and encrypted traffic from other clients by default. No central > directory for routing. > *I2P incorporates applications in the form of webapps. > *Webapps plugins (*.su3) can be loaded easily through the console. > *Includes a ready mail(smtp pop3) client, i2psnark (bittorrent) client, > and Jetty (webserver). Or host with your own webserver. *Allows anonymous > chats through IRC tunnel service. > *Plugins for blogging, chatting, file sharing have already been written. > *Every service(plugin) works within the i2p network itself(Except mail). > *Not supposed to be used as an outproxy, still an outproxy is included. > *I2P source code includes a java-service-wrapper compiled for specific > targets but is not compatible with OpenBSD. So I've included a new port > tanukiwrapper, and configured I2P to use it. > *Very exhaustive(functionality) but easy to use router console. > > > Testing: > *Run the rc service, point browser to localhost:7657. > *Set HTTP proxy to localhost: to view eepsites. > https://geti2p.net/en/about/browser-config > *Been running the router for a week now, with no problems. > *Personally tested the default and external plugins(e.g muwire). > *Test dependencies not present in ports, will add to it. > *Mount separate partition for /var/i2p if you do heavy transfers. > > > Index: user.list > === > RCS file: /cvs/ports/infrastructure/db/user.list,v > retrieving revision 1.371 diff -u -p -r1.371 user.list --- user.list 12 Jun > 2020 11:27:08 - 1.371 > +++ user.list 2 Aug 2020 18:16:51 - > @@ -366,3 +366,4 @@ id user group port > 855 _repro_repro telephony/resiprocate > 856 _return _return telephony/resiprocate > 857 _web2ldap _web2ldap sysutils/web2ldap > +858 _i2p _i2pnet/i2p > > > Any comments?
math/py-numexpr: remove py2-flavor
Remove py2-flavor of math/py-numexpr. Consumers are already py3-only.
OK?
diff --git devel/quirks/Makefile devel/quirks/Makefile
index 3c7f46a4a3e..33f744dfa26 100644
--- devel/quirks/Makefile
+++ devel/quirks/Makefile
@@ -5,7 +5,7 @@ CATEGORIES =devel databases
DISTFILES =
# API.rev
-PKGNAME = quirks-3.396
+PKGNAME = quirks-3.397
PKG_ARCH = *
MAINTAINER = Marc Espie
diff --git devel/quirks/files/Quirks.pm devel/quirks/files/Quirks.pm
index 7f620c19826..90300d8e3bd 100644
--- devel/quirks/files/Quirks.pm
+++ devel/quirks/files/Quirks.pm
@@ -454,6 +454,7 @@ my $stem_extensions = {
'py-cftime' => 'py3-cftime',
'py-astral' => 'py3-astral',
'py-ropper' => 'py3-ropper',
+ 'py-numexpr' => 'py3-numexpr',
};
my $obsolete_reason = {
diff --git math/Makefile math/Makefile
index 90041544d47..c1dac648268 100644
--- math/Makefile
+++ math/Makefile
@@ -133,7 +133,6 @@
SUBDIR += py-mpmath,python3
SUBDIR += py-networkx
SUBDIR += py-netcdf4
- SUBDIR += py-numexpr
SUBDIR += py-numexpr,python3
SUBDIR += py-numpy
SUBDIR += py-numpy,python3
diff --git math/py-numexpr/Makefile math/py-numexpr/Makefile
index 6e6d53e33b0..f0687eb0b5c 100644
--- math/py-numexpr/Makefile
+++ math/py-numexpr/Makefile
@@ -6,7 +6,7 @@ MODPY_EGG_VERSION = 2.6.4
DISTNAME = numexpr-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME}
CATEGORIES = math
-REVISION = 3
+REVISION = 4
# MIT
PERMIT_PACKAGE = Yes
@@ -30,6 +30,6 @@ RUN_DEPENDS = math/py-numpy${MODPY_FLAVOR}
BUILD_DEPENDS =${RUN_DEPENDS}
FLAVORS = python3
-FLAVOR ?=
+FLAVOR ?= python3
.include
diff --git math/py-numexpr/pkg/PLIST math/py-numexpr/pkg/PLIST
index 562a4deac37..c9f63abe582 100644
--- math/py-numexpr/pkg/PLIST
+++ math/py-numexpr/pkg/PLIST
@@ -1,4 +1,6 @@
@comment $OpenBSD: PLIST,v 1.1.1.1 2017/04/23 17:45:20 shadchin Exp $
+@conflict py-numexpr-*
+@pkgpath math/py-numexpr
lib/python${MODPY_VERSION}/site-packages/numexpr/
lib/python${MODPY_VERSION}/site-packages/numexpr-${MODPY_EGG_VERSION}-py${MODPY_VERSION}.egg-info/
lib/python${MODPY_VERSION}/site-packages/numexpr-${MODPY_EGG_VERSION}-py${MODPY_VERSION}.egg-info/PKG-INFO
@@ -19,7 +21,7 @@
lib/python${MODPY_VERSION}/site-packages/numexpr/${MODPY_PYCACHE}utils.${MODPY_P
lib/python${MODPY_VERSION}/site-packages/numexpr/${MODPY_PYCACHE}version.${MODPY_PYC_MAGIC_TAG}pyc
lib/python${MODPY_VERSION}/site-packages/numexpr/cpuinfo.py
lib/python${MODPY_VERSION}/site-packages/numexpr/expressions.py
-lib/python${MODPY_VERSION}/site-packages/numexpr/interpreter.so
+@so lib/python${MODPY_VERSION}/site-packages/numexpr/interpreter.so
lib/python${MODPY_VERSION}/site-packages/numexpr/necompiler.py
lib/python${MODPY_VERSION}/site-packages/numexpr/tests/
lib/python${MODPY_VERSION}/site-packages/numexpr/tests/__init__.py
minor update: neovim to 0.4.4
Hey Edd and ports@
Looks like a new Neovim minor release came out a few days ago.
https://github.com/neovim/neovim/releases/tag/v0.4.4
It's a trivial version bump and it builds and works fine for me. Built on the
latest snapshot with llvm 10.
I also removed the comment about a single test failure because that is fixed
now. I get a clean 'make test'.
diff --git a/Makefile b/Makefile
index eb913e1..67ba720 100644
--- a/Makefile
+++ b/Makefile
@@ -4,7 +4,7 @@ COMMENT = continuation and extension of Vim
GH_ACCOUNT = neovim
GH_PROJECT = neovim
-GH_TAGNAME = v0.4.3
+GH_TAGNAME = v0.4.4
CATEGORIES = editors devel
HOMEPAGE = https://neovim.io
@@ -101,9 +101,6 @@ pre-configure:
# These are the "old tests". There is also a new suite, but we would need the
# "busted" test suite for Lua, which is not yet ported.
-#
-# There is currently one (minor) test failure:
-# https://github.com/neovim/neovim/issues/10420
do-test:
cd ${WRKSRC}/src/nvim/testdir && ${SETENV} LC_CTYPE=en_US.UTF-8 \
${MAKE_PROGRAM} NVIM_PRG=${WRKBUILD}/bin/nvim ${MAKE_FLAGS}
diff --git a/distinfo b/distinfo
index 1c1e58a..9375c6b 100644
--- a/distinfo
+++ b/distinfo
@@ -1,4 +1,4 @@
SHA256 (luv-1.30.1-1.tar.gz) = Tih77W9R/VDOA7p/qMwz2E4bnLhpEcSBK7H0eh4+0So=
-SHA256 (neovim-0.4.3.tar.gz) = kaC10yIEqCG/QUaQ5rSM9pIk0ZYdNxWMKzg/amz4VNI=
+SHA256 (neovim-0.4.4.tar.gz) = L3aqxZNjZ383WS6FOrLAYVHMqIMNSz/kZ1tKUtQfxCw=
SIZE (luv-1.30.1-1.tar.gz) = 1354232
-SIZE (neovim-0.4.3.tar.gz) = 9556199
+SIZE (neovim-0.4.4.tar.gz) = 9558246
UPDATE math/py-pandas-1.1.0
Pandas 1.1.0 is a minor release, which includes some new features, bug
fixes, and performance improvements. Changes are explained at
https://pandas.pydata.org/pandas-docs/dev/whatsnew/v1.1.0.html.
Run tested on amd64 (while running $DAYJOB related tools).
Comments/OK?
diff --git Makefile Makefile
index 55ca90e54b0..02503ab2195 100644
--- Makefile
+++ Makefile
@@ -2,7 +2,7 @@
COMMENT = data analysis and manipulation library
-MODPY_EGG_VERSION =1.0.5
+MODPY_EGG_VERSION =1.1.0
DISTNAME = pandas-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME}
@@ -26,13 +26,13 @@ WANTLIB += ${COMPILER_LIBCXX} m ${MODPY_WANTLIB}
COMPILER = base-clang ports-gcc base-gcc
-BUILD_DEPENDS += lang/cython${MODPY_FLAVOR} \
- math/py-numpy${MODPY_FLAVOR}>=1.13.3
+BUILD_DEPENDS += lang/cython${MODPY_FLAVOR}>=0.29.16 \
+ math/py-numpy${MODPY_FLAVOR}>=1.15.4
-RUN_DEPENDS += devel/py-dateutil${MODPY_FLAVOR}>=2.6.1 \
+RUN_DEPENDS += devel/py-dateutil${MODPY_FLAVOR}>=2.7.3 \
devel/py-tz${MODPY_FLAVOR}>=2017.2 \
- lang/cython${MODPY_FLAVOR} \
- math/py-numpy${MODPY_FLAVOR}>=1.13.3
+ lang/cython${MODPY_FLAVOR}>=0.29.16 \
+ math/py-numpy${MODPY_FLAVOR}>=1.15.4
TEST_DEPENDS +=databases/py-mysqlclient${MODPY_FLAVOR} \
databases/py-psycopg2${MODPY_FLAVOR} \
diff --git distinfo distinfo
index 0b99eba959e..1165021ef2b 100644
--- distinfo
+++ distinfo
@@ -1,2 +1,2 @@
-SHA256 (pandas-1.0.5.tar.gz) = acXZIKCyqYOOZ3949N3lBrleqOTTDaJYWdtkad7YT6g=
-SIZE (pandas-1.0.5.tar.gz) = 5007108
+SHA256 (pandas-1.1.0.tar.gz) = s5UIVirQuz84Sw2yTafWiiYIud3IWx2THMqqktXkUnM=
+SIZE (pandas-1.1.0.tar.gz) = 5211071
diff --git patches/patch-setup_cfg patches/patch-setup_cfg
index 62e8906b29a..cb716ac7da5 100644
--- patches/patch-setup_cfg
+++ patches/patch-setup_cfg
@@ -3,10 +3,10 @@ $OpenBSD: patch-setup_cfg,v 1.3 2020/02/27 18:23:39 bket Exp $
Index: setup.cfg
--- setup.cfg.orig
+++ setup.cfg
-@@ -52,7 +52,6 @@ markers =
- high_memory: mark a test as a high-memory only
- clipboard: mark a pd.read_clipboard test
- doctest_optionflags = NORMALIZE_WHITESPACE IGNORE_EXCEPTION_DETAIL
+@@ -46,7 +46,6 @@ exclude =
+ minversion = 4.0.2
+ testpaths = pandas
+ doctest_optionflags = NORMALIZE_WHITESPACE IGNORE_EXCEPTION_DETAIL ELLIPSIS
-addopts = --strict-data-files
xfail_strict = True
filterwarnings =
diff --git patches/patch-setup_py patches/patch-setup_py
new file mode 100644
index 000..701c42a9379
--- /dev/null
+++ patches/patch-setup_py
@@ -0,0 +1,17 @@
+$OpenBSD$
+
+Fix Py38 build issue. Adapted from
+https://github.com/pandas-dev/pandas/pull/33431
+
+Index: setup.py
+--- setup.py.orig
setup.py
+@@ -433,7 +433,7 @@ if is_platform_windows():
+ extra_compile_args.append("/Z7")
+ extra_link_args.append("/DEBUG")
+ else:
+-extra_compile_args = ["-Werror"]
++extra_compile_args = ["-Werror", "-Wno-error=deprecated-declarations"]
+ extra_link_args = []
+ if debugging_symbols_requested:
+ extra_compile_args.append("-g")
diff --git pkg/PLIST pkg/PLIST
index 612f65f0556..c650de0bfda 100644
--- pkg/PLIST
+++ pkg/PLIST
@@ -59,22 +59,22 @@
lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/
lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/__init__.py
${MODPY_COMMENT}lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/${MODPY_PYCACHE}/
lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/${MODPY_PYCACHE}__init__.${MODPY_PYC_MAGIC_TAG}pyc
-@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/c_timestamp.so
+@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/base.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/ccalendar.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/conversion.so
+@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/dtypes.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/fields.so
-@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/frequencies.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/nattype.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/np_datetime.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/offsets.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/parsing.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/period.so
-@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/resolution.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/strptime.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/timedeltas.so
@so lib/python${MODPY_VERSION}/site-packages/pandas/_libs/tslibs/timestamps.so
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: t...@cvs.openbsd.org2020/08/09 11:52:18 ports/security/py-tlsfuzzer/files Update of /cvs/ports/security/py-tlsfuzzer/files In directory cvs.openbsd.org:/tmp/cvs-serv54180/files Log Message: Directory /cvs/ports/security/py-tlsfuzzer/files added to the repository
UPDATE math/py-numpy-1.16.6
NumPy 1.16.6 fixes bugs reported against the 1.16.5 release, and also
backports several enhancements from master that seem appropriate for a
release series that is the last to support Python 2.7. Overview changes:
https://github.com/numpy/numpy/releases/tag/v1.16.6
Testing:
- Run tested on amd64 in combination with several consumers;
- Output 'make test':
1.16.6:
13 failed, 7300 passed, 92 skipped, 168 deselected, 12 xfailed, 1 xpassed, 8
warnings in 256.89 seconds
1.16.5:
13 failed, 7258 passed, 90 skipped, 167 deselected, 12 xfailed, 1 xpassed, 8
warnings in 331.49 seconds
Comments/OK?
diff --git Makefile Makefile
index 480522f06fe..94913b37bd0 100644
--- Makefile
+++ Makefile
@@ -2,11 +2,10 @@
COMMENT= fast array and numeric programming library for Python
-MODPY_EGG_VERSION= 1.16.5
+MODPY_EGG_VERSION= 1.16.6
DISTNAME= numpy-${MODPY_EGG_VERSION}
PKGNAME= py-${DISTNAME}
EXTRACT_SUFX= .zip
-REVISION= 0
CATEGORIES=math devel
diff --git distinfo distinfo
index f8c8f20fed7..ba59222d7ac 100644
--- distinfo
+++ distinfo
@@ -1,2 +1,2 @@
-SHA256 (numpy-1.16.5.zip) = i7RS2U6WSzEiBbDeEjjdcgnaRSNDZTqyFLXWgXgOegw=
-SIZE (numpy-1.16.5.zip) = 5138208
+SHA256 (numpy-1.16.6.zip) = 5c8/3xNAGIXo7qgXBiTsliJeIXTrDGEcbybdM7SJ4/8=
+SIZE (numpy-1.16.6.zip) = 5143340
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: t...@cvs.openbsd.org2020/08/09 08:46:23 Added files: emulators/dolphin/patches: patch-Externals_glslang_glslang_Include_PoolAlloc_h Log message: fix build with clang 10. Found by Ricardo Mestre for another port. Tested by me and naddy@ No revision bump as port didn't build
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/09 08:17:27 Modified files: devel/llvm/patches: patch-lib_CodeGen_AsmPrinter_AsmPrinter_cpp patch-lib_MC_MCELFStreamer_cpp Log message: Add a comment and fix a typo
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/09 08:14:42 Removed files: devel/llvm/patches: patch-test_Assembler_debug-variant-discriminator_ll Log message: Drop patch, merged upstream The patch in cvs duplicated the content of the original testcase.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/09 08:12:30 Modified files: devel/llvm/patches: patch-lib_Target_Mips_MipsAsmPrinter_cpp patch-lib_Target_Sparc_SparcISelLowering_cpp Log message: Drop two outdated comments from Brad
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: mes...@cvs.openbsd.org 2020/08/09 07:44:56 Modified files: emulators/mednafen: Makefile distinfo Log message: update to 1.24.3 OK jeremy@
Re: sysutils/xxhash
On Sun 09/08/2020 09:58, Jeremie Courreges-Anglas wrote:
> On Sun, Aug 09 2020, Bjorn Ketelaars wrote:
> > On Sun 09/08/2020 00:10, Jeremie Courreges-Anglas wrote:
> >> > ==> failures/sysutils/xxhash.log <==
> >> > Hunk #1 succeeded at 447.
> >> > done
> >> > ===> Compiler link: clang -> /usr/bin/clang
> >> > ===> Compiler link: clang++ -> /usr/bin/clang++
> >> > ===> Compiler link: cc -> /usr/bin/cc
> >> > ===> Compiler link: c++ -> /usr/bin/c++
> >> > ===> Generating configure for xxhash-0.8.0
> >> > ===> Configuring for xxhash-0.8.0
> >> > ===> Building for xxhash-0.8.0
> >> > cc -O2 -pipe -c -o xxhash.o xxhash.c
> >> > ar rcs libxxhash.a xxhash.o
> >> > cc -O2 -pipexxhash.c -shared -o libxxhash.so.0.2
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_64bits_withSeed; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH3_64bits_digest)
> >> >
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_64bits_withSecret; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH3_64bits_digest)
> >> >
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_128bits_withSeed; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH3_generateSecret)
> >> >
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_128bits_withSeed; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH128)
> >> >
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_128bits_withSeed; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH3_128bits_digest)
> >> >
> >> > ld: error: relocation R_386_PC32 cannot be used against symbol
> >> > XXH3_128bits_withSecret; recompile with -fPIC
> >> defined in /tmp/xxhash-53bf5d.o
> >> referenced by xxhash.c
> >> /tmp/xxhash-53bf5d.o:(XXH3_128bits_digest)
> >> > cc: error: linker command failed with exit code 1 (use -v to see
> >> > invocation)
> >> > gmake: *** [Makefile:129: libxxhash.so.0.2] Error 1
> >> > *** Error 2 in sysutils/xxhash
> >> > (/usr/ports/infrastructure/mk/bsd.port.mk:2925
> >> > '/pobj/xxhash-0.8.0/.build_done': @cd /pobj/xxhash-0.8.0/xxHas...)
> >> > *** Error 2 in sysutils/xxhash
> >> > (/usr/ports/infrastructure/mk/bsd.port.mk:2584 'build':
> >> > @lock=xxhash-0.8.0; export _LOCKS_HELD=" xxhash-0.8)
> >> > ===> Exiting sysutils/xxhash with an error
> >> > *** Error 1 in /usr/ports (infrastructure/mk/bsd.port.subdir.mk:137
> >> > 'build': @: ${echo_msg:=echo}; : ${target:=build}; for i in ; do
> >> > eval...)
> >> Ended at 1596797814.95
> >> > max_stuck=0.00/depends=1.29/show-prepare-results=0.32/build=4.64
> >> > Error: job failed with 512 on i386-1 at 1596797814
> >>
> >> Shared libraries should be built with -fpic/-fPIC, i386 is more
> >> sensitive than amd64 in this regard.
> >>
> >> In the diff below I strip the CXX bits since this is a C-only port.
> >> I can leave them alone if you prefer, though.
> >>
> >> ok?
> >
> > Thank you addressing this issue!
> >
> > Please note that the CXX bits are needed for the testing phase. As such,
> > I think it makes sense to leave them alone (or move them to TEST_FLAGS).
>
> Oh indeed. I've left them alone.
>
> Here's an alternative take regarding the CFLAGS vs -fPIC issue.
> Upstream's GNU Makefile does respect CFLAGS (we pass them via MAKE_ENV)
> and properly tweaks them when needed (eg when building the shared lib).
> By also passing CFLAGS via MAKE_FLAGS we prevent such tweaking. So the
> fix is just to do less.
>
> ok?
>
> Index: Makefile
> ===
> RCS file: /cvs/ports/sysutils/xxhash/Makefile,v
> retrieving revision 1.7
> diff -u -p -r1.7 Makefile
> --- Makefile 28 Jul 2020 04:41:18 - 1.7
> +++ Makefile 9 Aug 2020 07:34:49 -
> @@ -5,6 +5,7 @@ COMMENT = extremely fast non-cryptograph
> GH_ACCOUNT = Cyan4973
> GH_PROJECT = xxHash
> GH_TAGNAME = v0.8.0
> +REVISION = 0
> PKGNAME =${DISTNAME:L}
>
> SHARED_LIBS =xxhash 0.2 # 0.7.4
> @@ -22,7 +23,6 @@ WANTLIB = c
>
> MAKE_FLAGS = CC="${CC}" \
> CXX="${CXX}" \
> - CFLAGS="${CFLAGS}" \
> CXXFLAGS="${CXXFLAGS}" \
> SONAME_FLAGS= \
> SHARED_EXT_VER="so.${LIBxxhash_VERSION}" \
OK bket@
Re: add notmuch flavor to mail/neomutt
Seems plausible/fine to me, but please document the flavor in DESCR as well (warranting a REVISION bump). The flavor must also be hooked up in mail/Makefile. On Sun, Aug 09, 2020 at 12:03:17PM +0200, Olivier Taïbi wrote: > +$OpenBSD$ Please comment patches. > +Index: notmuch/private.h > +--- notmuch/private.h.orig > notmuch/private.h > +@@ -34,11 +34,7 @@ > + #endif > + > + /* The definition in is broken */ > +-#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro) > \ > +- (LIBNOTMUCH_MAJOR_VERSION > (major) || > \ > +- (LIBNOTMUCH_MAJOR_VERSION == (major) && LIBNOTMUCH_MINOR_VERSION > > (minor)) || \ > +- (LIBNOTMUCH_MAJOR_VERSION == (major) && > \ > +-LIBNOTMUCH_MINOR_VERSION == (minor) && LIBNOTMUCH_MICRO_VERSION >= > (micro))) > ++#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro) (1>0) Why is this needed?
Re: update editors/kakoune
Hi,
Thank you solene.
Built, package made, updating and running fine on amd64 -current.
OK (from maintainer)
Best regards.
fredg
Le Friday 07 Aug 2020 à 15:55 (+0200), Solene Rapenne a écrit :
> this updates editors/kakoune
>
> Changelog since last version
>
> * Introduce `User` hook support.
>
> * The `bold` and `italic` faces are no longer built-in. Highlighters
> are expected to use face attributes (`+b` and `+i`, respectively) to
> decorate text.
>
> * The `lint-enable` command no longer needs to be called to display
> linting errors. The `lint-disable` command was renamed into
> `lint-hide-diagnostics`.
>
> * The `+` part of a `range-specs` highlighter consistently
> refers to the length of the target range.
>
> * clients stdin is transferred to the server, making it possible
> to pipe into `kak -c `
>
> * Faces can have an alpha channel, specified using the
> `rgba:RRGGBBAA` format.
>
> * replace-ranges highlighter now support empty and multi-lines ranges
>
> * `%val{...}` now expands to list of strings, `$kak_quoted_...` now work
> as expected with these.
>
> * `*SetOption` hooks filter string will contain a value only for options
> of int/str/bool types to avoid performance issue with generating those
> on more complex option types.
>
>
>
> Index: Makefile
> ===
> RCS file: /home/reposync/ports/editors/kakoune/Makefile,v
> retrieving revision 1.1.1.1
> diff -u -p -r1.1.1.1 Makefile
> --- Makefile 14 Jun 2020 13:07:04 - 1.1.1.1
> +++ Makefile 7 Aug 2020 08:13:24 -
> @@ -2,7 +2,7 @@
>
> COMMENT =modal code editor with a focus on interactivity
>
> -V = 2020.01.16
> +V = 2020.08.04
> DISTNAME = kakoune-${V}
>
> CATEGORIES = editors
> Index: distinfo
> ===
> RCS file: /home/reposync/ports/editors/kakoune/distinfo,v
> retrieving revision 1.1.1.1
> diff -u -p -r1.1.1.1 distinfo
> --- distinfo 14 Jun 2020 13:07:04 - 1.1.1.1
> +++ distinfo 7 Aug 2020 08:13:31 -
> @@ -1,2 +1,2 @@
> -SHA256 (kakoune-2020.01.16.tar.bz2) =
> oJTxaJ8CKDCPYx43GzgrDAUiOR/ItsI6bLxx/0BKDa4=
> -SIZE (kakoune-2020.01.16.tar.bz2) = 478509
> +SHA256 (kakoune-2020.08.04.tar.bz2) =
> AkafrB/4MZEWVTb6UvAZB9tb4Kc0vJBXCSSz7buc8SE=
> +SIZE (kakoune-2020.08.04.tar.bz2) = 499765
> Index: pkg/PLIST
> ===
> RCS file: /home/reposync/ports/editors/kakoune/pkg/PLIST,v
> retrieving revision 1.1.1.1
> diff -u -p -r1.1.1.1 PLIST
> --- pkg/PLIST 14 Jun 2020 13:07:04 - 1.1.1.1
> +++ pkg/PLIST 7 Aug 2020 08:40:46 -
> @@ -59,9 +59,11 @@ share/kak/rc/filetype/cabal.kak
> share/kak/rc/filetype/clojure.kak
> share/kak/rc/filetype/cmake.kak
> share/kak/rc/filetype/coffee.kak
> +share/kak/rc/filetype/coq.kak
> share/kak/rc/filetype/crystal.kak
> share/kak/rc/filetype/css.kak
> share/kak/rc/filetype/cucumber.kak
> +share/kak/rc/filetype/cue.kak
> share/kak/rc/filetype/d.kak
> share/kak/rc/filetype/dart.kak
> share/kak/rc/filetype/dhall.kak
> @@ -69,6 +71,7 @@ share/kak/rc/filetype/diff.kak
> share/kak/rc/filetype/dockerfile.kak
> share/kak/rc/filetype/elixir.kak
> share/kak/rc/filetype/elm.kak
> +share/kak/rc/filetype/eruby.kak
> share/kak/rc/filetype/etc.kak
> share/kak/rc/filetype/exherbo.kak
> share/kak/rc/filetype/fish.kak
> @@ -99,6 +102,8 @@ share/kak/rc/filetype/mail.kak
> share/kak/rc/filetype/makefile.kak
> share/kak/rc/filetype/markdown.kak
> share/kak/rc/filetype/mercurial.kak
> +share/kak/rc/filetype/meson.kak
> +share/kak/rc/filetype/mlb.kak
> share/kak/rc/filetype/moon.kak
> share/kak/rc/filetype/nim.kak
> share/kak/rc/filetype/nix.kak
> @@ -119,6 +124,7 @@ share/kak/rc/filetype/scala.kak
> share/kak/rc/filetype/scheme.kak
> share/kak/rc/filetype/scss.kak
> share/kak/rc/filetype/sh.kak
> +share/kak/rc/filetype/sml.kak
> share/kak/rc/filetype/sql.kak
> share/kak/rc/filetype/swift.kak
> share/kak/rc/filetype/systemd.kak
> @@ -129,6 +135,7 @@ share/kak/rc/filetype/troff.kak
> share/kak/rc/filetype/tupfile.kak
> share/kak/rc/filetype/void-linux.kak
> share/kak/rc/filetype/yaml.kak
> +share/kak/rc/filetype/zig.kak
> share/kak/rc/tools/
> share/kak/rc/tools/autorestore.kak
> share/kak/rc/tools/autowrap.kak
> @@ -150,10 +157,12 @@ share/kak/rc/tools/rust/
> share/kak/rc/tools/rust/racer.kak
> share/kak/rc/tools/spell.kak
> share/kak/rc/windowing/
> +share/kak/rc/windowing/detection.kak
> share/kak/rc/windowing/iterm.kak
> share/kak/rc/windowing/kitty.kak
> share/kak/rc/windowing/new-client.kak
> share/kak/rc/windowing/repl/
> +share/kak/rc/windowing/repl/kitty.kak
> share/kak/rc/windowing/repl/tmux.kak
> share/kak/rc/windowing/repl/x11.kak
> share/kak/rc/windowing/screen.kak
>
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2020/08/09 05:10:07 Modified files: sysutils/salt/patches: patch-salt_utils_network_py Log message: submitted upstream
Re: update mail/notmuch to 0.30
ping, I also added bash as a build dependency (for bash-completion)
On Wed, Jul 15, 2020 at 11:33:45PM +0200, Olivier Taïbi wrote:
> Below is a diff to update the recently added mail/notmuch to 0.30.
>
> changes (from upstream):
>
> What's new in notmuch 0.30
> =
>
> S/MIME
> --
>
> Handle S/MIME (PKCS#7) messages -- one-part signed messages, encrypted
> messages, and multilayer messages. Treat them symmetrically to
> OpenPGP messages. This includes handling protected headers
> gracefully.
>
> If you're using Notmuch with S/MIME, you currently need to configure
> gpgsm appropriately.
>
> Mixed-up MIME Repair
>
>
> Detect and automatically repair a common form of message mangling
> created by Microsoft Exchange (see index.repaired=mixedup in
> notmuch-properties(7)).
>
> Protected Headers
> -
>
> Avoid indexing the legacy-display part of an encrypted message that
> has protected headers (see
> index.repaired=skip-protected-headers-legacy-display in
> notmuch-properties(7)).
>
> Python
> --
>
> Drop support for python2, focus on python3.
>
> Introduce new CFFI-based python bindings in the python module named
> "notmuch2". Officially deprecate (but still support) the older
> "notmuch" module.
>
> Dependencies
>
>
> Support for Xapian 1.2 is removed. The minimum supported version of
> Xapian is now 1.4.0.
>
>
> In particular I only kept the python3 flavor for the python bindings.
>
> result of /usr/src/lib/check_sym (although mail/notmuch is still a leaf
> port):
>
> oldlibnotmuch.so.0.0 --> lib/libnotmuch.so.0.0
> No dynamic export changes
> External reference changes:
> added:
> g_mime_application_pkcs7_mime_decrypt
> g_mime_application_pkcs7_mime_get_smime_type
> g_mime_application_pkcs7_mime_get_type
> g_mime_application_pkcs7_mime_verify
> g_mime_multipart_encrypted_new
> g_mime_multipart_insert
> g_mime_object_set_content_type_parameter
> g_mime_text_part_get_text
> g_mime_text_part_get_type
>
>
> All tests pass except for the "unicode yen sign" ones, as expected (see
> https://marc.info/?l=openbsd-ports=158706294808158). The way the
> --gzip option for "notmuch dump" is removed (see explanation in previous
> url) is now slightly simpler because macros around gzprintf/gzputs were
> added upstream.
diff -uNpr /usr/ports/mail/notmuch/Makefile
/build/usr/ports/mail/notmuch/Makefile
--- /usr/ports/mail/notmuch/MakefileFri May 15 10:58:14 2020
+++ /build/usr/ports/mail/notmuch/Makefile Wed Jul 29 00:02:56 2020
@@ -2,7 +2,6 @@
SUBDIR =
SUBDIR += notmuch
-SUBDIR += py-notmuch
SUBDIR += py-notmuch,python3
.include
diff -uNpr /usr/ports/mail/notmuch/Makefile.inc
/build/usr/ports/mail/notmuch/Makefile.inc
--- /usr/ports/mail/notmuch/Makefile.incFri May 15 10:58:14 2020
+++ /build/usr/ports/mail/notmuch/Makefile.inc Wed Jul 29 00:02:56 2020
@@ -1,6 +1,6 @@
# $OpenBSD: Makefile.inc,v 1.1.1.1 2020/05/15 08:58:14 sthen Exp $
-V =0.29.3
+V =0.30
DISTNAME = notmuch-$V
EXTRACT_SUFX = .tar.xz
diff -uNpr /usr/ports/mail/notmuch/notmuch/Makefile
/build/usr/ports/mail/notmuch/notmuch/Makefile
--- /usr/ports/mail/notmuch/notmuch/MakefileTue May 19 22:45:49 2020
+++ /build/usr/ports/mail/notmuch/notmuch/Makefile Wed Jul 29 08:28:55 2020
@@ -14,7 +14,7 @@ DEBUG_PACKAGES = -main
COMPILER= base-clang ports-gcc
WANTLIB-main +=${COMPILER_LIBCXX} c glib-2.0 gmime-3.0
-WANTLIB-main +=gobject-2.0 intl m talloc xapian z
+WANTLIB-main +=gobject-2.0 intl m pthread talloc xapian z
LIB_DEPENDS = databases/xapian-core \
devel/libtalloc \
@@ -29,6 +29,7 @@ BUILD_DEPENDS = devel/doxygen \
editors/emacs,gtk3 \
security/gnupg2 \
shells/bash \
+ shells/bash-completion \
textproc/py-sphinx${MODPY_FLAVOR}
TEST_DEPENDS = archivers/gtar \
diff -uNpr /usr/ports/mail/notmuch/notmuch/distinfo
/build/usr/ports/mail/notmuch/notmuch/distinfo
--- /usr/ports/mail/notmuch/notmuch/distinfoFri May 15 10:58:14 2020
+++ /build/usr/ports/mail/notmuch/notmuch/distinfo Wed Jul 29 00:02:56 2020
@@ -1,2 +1,2 @@
-SHA256 (notmuch-0.29.3.tar.xz) = 1fcEuacjleQzA96bH02OFN0nvzZG/bs3S7Pbt9FQ3DU=
-SIZE (notmuch-0.29.3.tar.xz) = 660536
+SHA256 (notmuch-0.30.tar.xz) = Xjuqb+EdZcZ+Jq5Ii+EbMguuBOM2rMnGRiH340SQlvo=
+SIZE (notmuch-0.30.tar.xz) = 711904
diff -uNpr
[NEW] x11/screenkey
screenkey echoes keypresses to screen (which is useful for screencasts / teaching and the like). There's a simple demo on the homepage: https://www.thregr.org/~wavexx/software/screenkey/ This is a port of screenkey-1.2. One obvious problem is that there are some localisation files included, but they cause the setup.py installation to fail. Maybe there's an obvious Python setup fix, but I couldn't find it -- the best I could manage is to delete the initialisation files before building... Suggestions welcome! Laurie screenkey.tgz Description: application/tar-gz
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2020/08/09 04:37:46 Modified files: sysutils/salt/patches: patch-salt_grains_core_py patch-salt_modules_vmctl_py patch-salt_returners_highstate_return_py patch-salt_returners_nagios_nrdp_return_py patch-salt_states_sysctl_py Log message: tweak patch comments and add references to upstream pull requests where applicable
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2020/08/09 04:33:44 Modified files: sysutils/salt : Makefile distinfo sysutils/salt/patches: patch-salt_grains_core_py sysutils/salt/pkg: PLIST Log message: update to salt-3001.1
add notmuch flavor to mail/neomutt
diff -uNpr /usr/ports/mail/neomutt/Makefile
/build/usr/ports/mail/neomutt/Makefile
--- /usr/ports/mail/neomutt/MakefileTue Jul 28 23:54:23 2020
+++ /build/usr/ports/mail/neomutt/Makefile Fri Aug 7 23:46:50 2020
@@ -26,7 +26,7 @@ BUILD_DEPENDS+= textproc/docbook-xsl \
DEBUG_PACKAGES=${BUILD_PACKAGES}
SEPARATE_BUILD=Yes
-FLAVORS= gpgme sasl
+FLAVORS= gpgme sasl notmuch
FLAVOR?=
CONFIGURE_STYLE= simple
@@ -63,6 +63,12 @@ WANTLIB+=assuan gpg-error gpgme sqlite3
CONFIGURE_ARGS+= --sasl
LIB_DEPENDS+= security/cyrus-sasl2
WANTLIB+= sasl2
+.endif
+
+.if ${FLAVOR:Mnotmuch}
+CONFIGURE_ARGS+= --notmuch
+LIB_DEPENDS+= mail/notmuch/notmuch,-main
+WANTLIB+= notmuch
.endif
post-install:
diff -uNpr /usr/ports/mail/neomutt/patches/patch-notmuch_private_h
/build/usr/ports/mail/neomutt/patches/patch-notmuch_private_h
--- /usr/ports/mail/neomutt/patches/patch-notmuch_private_h Thu Jan 1
01:00:00 1970
+++ /build/usr/ports/mail/neomutt/patches/patch-notmuch_private_h Fri Aug
7 23:40:05 2020
@@ -0,0 +1,20 @@
+$OpenBSD$
+
+Index: notmuch/private.h
+--- notmuch/private.h.orig
notmuch/private.h
+@@ -34,11 +34,7 @@
+ #endif
+
+ /* The definition in is broken */
+-#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro)
\
+- (LIBNOTMUCH_MAJOR_VERSION > (major) ||
\
+- (LIBNOTMUCH_MAJOR_VERSION == (major) && LIBNOTMUCH_MINOR_VERSION >
(minor)) || \
+- (LIBNOTMUCH_MAJOR_VERSION == (major) &&
\
+-LIBNOTMUCH_MINOR_VERSION == (minor) && LIBNOTMUCH_MICRO_VERSION >=
(micro)))
++#define LIBNOTMUCH_CHECK_VERSION(major, minor, micro) (1>0)
+
+ extern const int NmUrlProtocolLen;
+
diff -uNpr /usr/ports/mail/neomutt/pkg/PLIST
/build/usr/ports/mail/neomutt/pkg/PLIST
--- /usr/ports/mail/neomutt/pkg/PLIST Fri May 15 10:19:08 2020
+++ /build/usr/ports/mail/neomutt/pkg/PLIST Sat Aug 8 22:34:35 2020
@@ -88,8 +88,6 @@ share/locale/cs/LC_MESSAGES/neomutt.mo
share/locale/da/LC_MESSAGES/neomutt.mo
share/locale/de/LC_MESSAGES/neomutt.mo
share/locale/el/LC_MESSAGES/neomutt.mo
-share/locale/en_GB/
-share/locale/en_GB/LC_MESSAGES/
share/locale/en_GB/LC_MESSAGES/neomutt.mo
share/locale/eo/LC_MESSAGES/neomutt.mo
share/locale/es/LC_MESSAGES/neomutt.mo
@@ -104,8 +102,6 @@ share/locale/id/LC_MESSAGES/neomutt.mo
share/locale/it/LC_MESSAGES/neomutt.mo
share/locale/ja/LC_MESSAGES/neomutt.mo
share/locale/ko/LC_MESSAGES/neomutt.mo
-share/locale/lt/
-share/locale/lt/LC_MESSAGES/
share/locale/lt/LC_MESSAGES/neomutt.mo
share/locale/nl/LC_MESSAGES/neomutt.mo
share/locale/pl/LC_MESSAGES/neomutt.mo
mail/rspamd and luajit (was: Re: amd64 bulk build failures 2020-08-07)
> mail/rspamd
http://build-failures.rhaalovely.net/amd64/2020-08-07/mail/rspamd.log
--8<--
ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
_Unwind_DeleteException
ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
_Unwind_GetCFA
ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
_Unwind_RaiseException
ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
_Unwind_SetGR
ld: error: /usr/local/lib/libluajit-5.1.so.1.0: undefined reference to
_Unwind_SetIP
-->8--
This is fixed by linking libluajit.so with libc++abi (which ships the
required the required unwind symbols). TARGET_LDFLAGS were only used to
link the luajit interpreter, not the library. I suspect that a few other
ports can benefit from this diff.
I have no idea why ld.lld starts looking at undefined symbols in library
dependencies (cc'ing kettenis so that he knows about this case).
ok?
Index: Makefile
===
RCS file: /cvs/ports/lang/luajit/Makefile,v
retrieving revision 1.30
diff -u -p -r1.30 Makefile
--- Makefile12 Jul 2019 20:47:19 - 1.30
+++ Makefile9 Aug 2020 10:17:32 -
@@ -11,7 +11,7 @@ SHARED_LIBS +=luajit-${MODLUA_VERSION}
COMMENT = just-in-time compiler for Lua
V =2.0.5
-REVISION = 1
+REVISION = 2
DISTNAME = LuaJIT-${V}
PKGNAME = ${DISTNAME:L}
@@ -61,7 +61,7 @@ do-test:
.include
.if ${PROPERTIES:Mclang}
-MAKE_FLAGS += TARGET_LDFLAGS='-lc++abi -lpthread'
+MAKE_FLAGS += TARGET_LIBS='-lc++abi -lpthread'
WANTLIB += c++abi pthread
.endif
Index: pkg/PLIST
===
RCS file: /cvs/ports/lang/luajit/pkg/PLIST,v
retrieving revision 1.2
diff -u -p -r1.2 PLIST
--- pkg/PLIST 3 May 2014 14:24:03 - 1.2
+++ pkg/PLIST 9 Aug 2020 10:17:32 -
@@ -8,12 +8,10 @@ include/luajit-2.0/luaconf.h
include/luajit-2.0/luajit.h
include/luajit-2.0/lualib.h
@comment lib/libluajit-${MODLUA_VERSION}.so
-lib/libluajit-${MODLUA_VERSION}.a
+@static-lib lib/libluajit-${MODLUA_VERSION}.a
@lib lib/libluajit-${MODLUA_VERSION}.so.${LIBluajit-${MODLUA_VERSION}_VERSION}
-lib/lua/${MODLUA_VERSION}/
lib/pkgconfig/luajit.pc
@man man/man1/luajit.1
-share/lua/${MODLUA_VERSION}/
share/lua/${MODLUA_VERSION}/jit/
share/lua/${MODLUA_VERSION}/jit/bc.lua
share/lua/${MODLUA_VERSION}/jit/bcsave.lua
--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ki...@cvs.openbsd.org 2020/08/09 03:37:07 Modified files: games/freeorion: Makefile distinfo games/freeorion/patches: patch-CMakeLists_txt games/freeorion/pkg: PLIST Added files: games/freeorion/patches: patch-GG_CMakeLists_txt patch-cmake_FindBoost_cmake patch-cmake_FindPyCodestyle_cmake patch-cmake_make_versioncpp_py Log message: Update to freeorion-0.4.10 and switch to python3 >From Tom Murphy, maintainer. Minor tweaks by me.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ki...@cvs.openbsd.org 2020/08/09 03:29:13 Modified files: net/seafile/client: Makefile distinfo net/seafile/client/patches: patch-CMakeLists_txt net/seafile/seafile: Makefile distinfo Log message: Update seafile to 7.0.9
Re: sysutils/xxhash
On Sun, Aug 09 2020, Bjorn Ketelaars wrote:
> On Sun 09/08/2020 00:10, Jeremie Courreges-Anglas wrote:
>> > ==> failures/sysutils/xxhash.log <==
>> > Hunk #1 succeeded at 447.
>> > done
>> > ===> Compiler link: clang -> /usr/bin/clang
>> > ===> Compiler link: clang++ -> /usr/bin/clang++
>> > ===> Compiler link: cc -> /usr/bin/cc
>> > ===> Compiler link: c++ -> /usr/bin/c++
>> > ===> Generating configure for xxhash-0.8.0
>> > ===> Configuring for xxhash-0.8.0
>> > ===> Building for xxhash-0.8.0
>> > cc -O2 -pipe -c -o xxhash.o xxhash.c
>> > ar rcs libxxhash.a xxhash.o
>> > cc -O2 -pipexxhash.c -shared -o libxxhash.so.0.2
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_64bits_withSeed; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH3_64bits_digest)
>> >
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_64bits_withSecret; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH3_64bits_digest)
>> >
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_128bits_withSeed; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH3_generateSecret)
>> >
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_128bits_withSeed; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH128)
>> >
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_128bits_withSeed; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH3_128bits_digest)
>> >
>> > ld: error: relocation R_386_PC32 cannot be used against symbol
>> > XXH3_128bits_withSecret; recompile with -fPIC
>> defined in /tmp/xxhash-53bf5d.o
>> referenced by xxhash.c
>> /tmp/xxhash-53bf5d.o:(XXH3_128bits_digest)
>> > cc: error: linker command failed with exit code 1 (use -v to see
>> > invocation)
>> > gmake: *** [Makefile:129: libxxhash.so.0.2] Error 1
>> > *** Error 2 in sysutils/xxhash
>> > (/usr/ports/infrastructure/mk/bsd.port.mk:2925
>> > '/pobj/xxhash-0.8.0/.build_done': @cd /pobj/xxhash-0.8.0/xxHas...)
>> > *** Error 2 in sysutils/xxhash
>> > (/usr/ports/infrastructure/mk/bsd.port.mk:2584 'build':
>> > @lock=xxhash-0.8.0; export _LOCKS_HELD=" xxhash-0.8)
>> > ===> Exiting sysutils/xxhash 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 1596797814.95
>> > max_stuck=0.00/depends=1.29/show-prepare-results=0.32/build=4.64
>> > Error: job failed with 512 on i386-1 at 1596797814
>>
>> Shared libraries should be built with -fpic/-fPIC, i386 is more
>> sensitive than amd64 in this regard.
>>
>> In the diff below I strip the CXX bits since this is a C-only port.
>> I can leave them alone if you prefer, though.
>>
>> ok?
>
> Thank you addressing this issue!
>
> Please note that the CXX bits are needed for the testing phase. As such,
> I think it makes sense to leave them alone (or move them to TEST_FLAGS).
Oh indeed. I've left them alone.
Here's an alternative take regarding the CFLAGS vs -fPIC issue.
Upstream's GNU Makefile does respect CFLAGS (we pass them via MAKE_ENV)
and properly tweaks them when needed (eg when building the shared lib).
By also passing CFLAGS via MAKE_FLAGS we prevent such tweaking. So the
fix is just to do less.
ok?
Index: Makefile
===
RCS file: /cvs/ports/sysutils/xxhash/Makefile,v
retrieving revision 1.7
diff -u -p -r1.7 Makefile
--- Makefile28 Jul 2020 04:41:18 - 1.7
+++ Makefile9 Aug 2020 07:34:49 -
@@ -5,6 +5,7 @@ COMMENT = extremely fast non-cryptograph
GH_ACCOUNT = Cyan4973
GH_PROJECT = xxHash
GH_TAGNAME = v0.8.0
+REVISION = 0
PKGNAME = ${DISTNAME:L}
SHARED_LIBS = xxhash 0.2 # 0.7.4
@@ -22,7 +23,6 @@ WANTLIB = c
MAKE_FLAGS = CC="${CC}" \
CXX="${CXX}" \
- CFLAGS="${CFLAGS}" \
CXXFLAGS="${CXXFLAGS}" \
SONAME_FLAGS= \
SHARED_EXT_VER="so.${LIBxxhash_VERSION}" \
--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: i386/LLVM 10
On Sat, 08 Aug 2020 23:16:09 +0200
Jeremie Courreges-Anglas wrote:
>
> < sthen> ld.lld:
> error: /pobj/webkitgtk4-2.28.4/build-i386/lib/libjavascriptcoregtk-4.0.so:
> undefined reference to WTF::Optional
> JSC::parse(char const*)
>
> cwen did hit this too on powerpc when it was switched to clang,
> I propose to extend the workaround to all archs. Disclaimer: diff not
> tested on i386, no build regression on amd64.
I've been able to build webkitgtk4 without issues on powerpc with that
diff.
OK cwen@ if i386 survives to this :)
> Index: Makefile
> ===
> RCS file: /cvs/ports/www/webkitgtk4/Makefile,v
> retrieving revision 1.125
> diff -u -p -r1.125 Makefile
> --- Makefile 7 Aug 2020 14:30:10 - 1.125
> +++ Makefile 8 Aug 2020 21:09:58 -
> @@ -16,7 +16,7 @@ COMMENT = GTK+ port of the WebKit
> rende V = 2.28.4
> DISTNAME = webkitgtk-${V}
> PKGNAME =webkitgtk4-${V}
> -REVISION = 0
> +REVISION = 1
> EXTRACT_SUFX = .tar.xz
> API =4.0
> SUBST_VARS = API
> Index: patches/patch-Source_JavaScriptCore_runtime_Options_cpp
> ===
> RCS
> file:
> /cvs/ports/www/webkitgtk4/patches/patch-Source_JavaScriptCore_runtime_Options_cpp,v
> retrieving revision 1.3 diff -u -p -r1.3
> patch-Source_JavaScriptCore_runtime_Options_cpp
> --- patches/patch-Source_JavaScriptCore_runtime_Options_cpp 18
> Apr 2020 16:14:28 - 1.3 +++
> patches/patch-Source_JavaScriptCore_runtime_Options_cpp 8 Aug
> 2020 21:09:58 - @@ -1,6 +1,6 @@ $OpenBSD:
> patch-Source_JavaScriptCore_runtime_Options_cpp,v 1.3 2020/04/18
> 16:14:28 cwen Exp $ -powerpc fix for:
> +powerpc and i386 fix for:
> undefined reference to `WTF::Optional
> JSC::parse(char const*)'
>
> @@ -12,7 +12,7 @@ Index: Source/JavaScriptCore/runtime/Opt
> }
>
> -#if CPU(ADDRESS64) || OS(DARWIN)
> -+#if CPU(ADDRESS64) || OS(DARWIN) || ( defined(__OpenBSD__) &&
> defined(__powerpc__) ) ++#if CPU(ADDRESS64) || OS(DARWIN) ||
> OS(OPENBSD) template<>
> Optional parse(const char* string)
> {
> @@ -21,7 +21,7 @@ Index: Source/JavaScriptCore/runtime/Opt
> return WTF::nullopt;
> }
> -#endif // CPU(ADDRESS64) || OS(DARWIN)
> -+#endif // CPU(ADDRESS64) || OS(DARWIN) || ( defined(__OpenBSD__) &&
> defined(__powerpc__) ) ++#endif // CPU(ADDRESS64) || OS(DARWIN) ||
> OS(OPENBSD)
> template<>
> Optional parse(const char* string)
>
>
> --
> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE
> 1524 E7EE
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rob...@cvs.openbsd.org 2020/08/09 00:33:26 Modified files: editors/libreoffice: Makefile Added files: editors/libreoffice/patches: patch-connectivity_source_drivers_odbc_OConnection_cxx patch-external_skia_Library_skia_mk patch-vcl_skia_gdiimpl_cxx Log message: add some missing OpenBSD ifdefs to take the Linux codepath
