Re: [new] productivity/khal
On Fri, 5 May 2017 22:22:54 +0200, Joerg Jung wrote: > Hi, > > please find attached a port for khal. While sometimes a bit buggy it > mostly works fine for my basic usage. > > $ cat pkg/DESCR > Khal is a standards based CLI (console) calendar program, able to > synchronize with CalDAV servers through vdirsyncer. > > > Comments? OK to import? ok danj@ with the following diff --- Makefile.orig Tue May 23 20:54:25 2017 +++ MakefileTue May 23 20:55:23 2017 @@ -24,12 +24,12 @@ RUN_DEPENDS = devel/py-atomicwrites${MODPY_FLAVOR} \ devel/py-click${MODPY_FLAVOR} \ devel/py-configobj${MODPY_FLAVOR} \ devel/py-dateutil${MODPY_FLAVOR} \ + devel/py-icalendar${MODPY_FLAVOR} \ devel/py-setproctitle${MODPY_FLAVOR} \ devel/py-tz${MODPY_FLAVOR} \ devel/py-tzlocal${MODPY_FLAVOR} \ devel/py-urwid${MODPY_FLAVOR} \ devel/py-xdg${MODPY_FLAVOR} \ - devel/py-icalendar${MODPY_FLAVOR} \ productivity/vdirsyncer TEST_DEPENDS = ${RUN_DEPENDS} \ devel/py-test${MODPY_FLAVOR} \ @@ -37,9 +37,7 @@ TEST_DEPENDS =${RUN_DEPENDS} \ devel/py-test-capturelog${MODPY_FLAVOR} \ devel/py-freezegun${MODPY_FLAVOR} -# todo: fix tests -NO_TEST = yes -#do-test: -# cd ${WRKSRC} && ${LOCALBASE}/bin/py.test-3 --cov khal +do-test: + cd ${WRKSRC} && ${MODPY_BIN} -m pytest .include
Re: [new] devel/py-icalendar
On Fri, 5 May 2017 22:20:07 +0200, Joerg Jung wrote: > Hi, > > please find attached a port for py-icalendar, a dependency for the > upcoming khal port. > > $ cat pkg/DESCR > This is an attempt at making an iCalendar package for Python, to > parse and generate iCalendar files. The inspiration has come from > the email package in the standard lib, yet efficient and powerful. > > > Comments? OK to import? ok danj@
Re: [new] devel/py-freezegun
On Fri, 5 May 2017 22:21:07 +0200, Joerg Jung wrote: > Hi, > > please find attached a port for py-freezegun, a test dependency for > the upcoming khal port. > > $ cat pkg/DESCR > FreezeGun is a library that allows your python tests to travel > through time by mocking the datetime module. > > > Comments? OK to import? > > Thanks, > Regards, > Joerg Here's a diff to remove HOMEPAGE as it seems nonexistent and MODPY_PI sets up a HOMEPAGE, and make tests work (there's one failing but shrug). With that, ok danj@ --- Makefile.orig Tue May 23 20:37:18 2017 +++ MakefileTue May 23 20:39:54 2017 @@ -8,8 +8,6 @@ PKGNAME = py-freezegun-${MODPY_EGG_VERSION} CATEGORIES = devel -HOMEPAGE = https://freezegun.readthedocs.org - MAINTAINER = Joerg Jung # Apache 2.0 @@ -30,9 +28,8 @@ TEST_DEPENDS =${RUN_DEPENDS} \ FLAVORS = python3 FLAVOR ?= -# todo: fix tests -NO_TEST = yes -#do-test: -# cd ${WRKSRC} && ${LOCALBASE}/bin/nosetests-3 +do-test: + cd ${WRKSRC} && ${LOCALBASE}/bin/nosetests${MODPY_BIN_SUFFIX} tests/*.py + .include
Re: www/webkit is old/insecure, news at 11
On Sun, Feb 12, 2017 at 06:51:06PM +0100, Joerg Jung wrote: > > > On 31 Jan 2017, at 10:43, Landry Breuil wrote: > > > > Hi, > > > > so the following ports use www/webkit or www/webkit,gtk3: > > > > cad/xtrkcad > > graphics/ebook-viewer > > mail/claws-mail > > mail/geary > > productivity/gnucash > > www/luakit > > www/midori > > www/surf > > www/vimb > > > > www/webkit is the 2.4 branch, which is dead/abandoned/unmaintained > > upstream, cf > > https://blogs.gnome.org/mcatanzaro/2016/02/01/on-webkit-security-updates/ > > and https://lists.webkit.org/pipermail/webkit-gtk/2016-September/002779.html > > > > The security circus being what it is (OMG YOUR OS SHIPS AN INSECURE > > LIBRARY STOP THE PRESS IM GOING TO SHAME YOU ON TWITTER), it would be > > nice to get rid of this, so can the ones maintaining those ports try to > > reach the respective upstreams and see if there's a way to make those > > ports use www/webkitgtk4 (which is maintained, until upstream webkit > > becomes crazy again..) ? gmpc-wikipedia and py-webkitgtk have already > > been removed. > > > > I'll have a look at what can be done for geary and midori. As for surf, > > there's surf2… ? > > While surf is still under active development and repo is seeing commits, > there seem to be no interest from the upstream maintainers to move > to a newer webkit. > AFAIK, surf2 is just a hand-rolled tarball from “experimental” branch > and has it’s own problems. Meanwhile, suckless upstream switched main surf repo to formerly surf2 branch and released a 2.0 tarball. Please find below a diff which brings surf to the latest release. OK? I think, once this is in, www/surf2 can be removed. Thanks, Regards, Joerg Index: Makefile === RCS file: /cvs/ports/www/surf/Makefile,v retrieving revision 1.22 diff -u -p -r1.22 Makefile --- Makefile23 Dec 2015 20:40:03 - 1.22 +++ Makefile23 May 2017 21:38:56 - @@ -2,7 +2,7 @@ COMMENT = simple webbrowser based on webkit/gtk+ -DISTNAME = surf-0.7 +DISTNAME = surf-2.0 CATEGORIES = www HOMEPAGE = http://surf.suckless.org/ @@ -11,18 +11,16 @@ MAINTAINER= Joerg Jung http://dl.suckless.org/surf/ -MODULES = devel/gettext - -LIB_DEPENDS = www/webkit>=1.6.1v0 -RUN_DEPENDS = x11/dmenu +LIB_DEPENDS = devel/gettext \ + www/webkitgtk4 +RUN_DEPENDS = x11/dmenu \ + net/curl # webkit browsing RUN_DEPENDS += multimedia/gstreamer1/plugins-libav Index: distinfo === RCS file: /cvs/ports/www/surf/distinfo,v retrieving revision 1.6 diff -u -p -r1.6 distinfo --- distinfo23 Dec 2015 20:40:03 - 1.6 +++ distinfo23 May 2017 21:38:56 - @@ -1,2 +1,2 @@ -SHA256 (surf-0.7.tar.gz) = lWCFRvtk0Bx6gVPDVr4OKE6+Egw8WWqU6z861H4cSUo= -SIZE (surf-0.7.tar.gz) = 18792 +SHA256 (surf-2.0.tar.gz) = +u5MemLDj8l5Hv8a0GeHw8myt58ziAaCf1FSp7xUlR0= +SIZE (surf-2.0.tar.gz) = 19056 Index: patches/patch-config_def_h === RCS file: /cvs/ports/www/surf/patches/patch-config_def_h,v retrieving revision 1.4 diff -u -p -r1.4 patch-config_def_h --- patches/patch-config_def_h 23 Dec 2015 20:40:03 - 1.4 +++ patches/patch-config_def_h 23 May 2017 21:38:56 - @@ -1,43 +1,22 @@ $OpenBSD: patch-config_def_h,v 1.4 2015/12/23 20:40:03 jung Exp $ config.def.h.orig Sat Dec 19 15:59:30 2015 -+++ config.def.h Sat Dec 19 20:45:50 2015 -@@ -16,9 +16,9 @@ static gfloat zoomlevel = 1.0; /* Default zoom l - - /* Soup default features */ - static char *cookiefile = "~/.surf/cookies.txt"; --static char *cookiepolicies = "Aa@"; /* A: accept all; a: accept nothing, -+static char *cookiepolicies = "@Aa"; /* A: accept all; a: accept nothing, - * @: accept all except third party */ --static char *cafile = "/etc/ssl/certs/ca-certificates.crt"; -+static char *cafile = "/etc/ssl/cert.pem"; - static Bool strictssl = FALSE; /* Refuse untrusted SSL connections */ - static time_t sessiontime = 3600; - -@@ -29,11 +29,11 @@ static Bool enablediskcache = TRUE; - static int diskcachebytes = 5 * 1024 * 1024; - static Bool enableplugins = TRUE; - static Bool enablescripts = TRUE; --static Bool enableinspector = TRUE; -+static Bool enableinspector = FALSE; - static Bool enablestyle = TRUE; - static Bool loadimages= TRUE; - static Bool hidebackground= FALSE; --static Bool allowgeolocation = TRUE; -+static Bool allowgeolocation = FALSE; - - #define SETPROP(p, q) { \ - .v = (char *[]){ "/bin/sh", "-c", \ -@@ -48,10 +48,9 @@ static Bool allowgeolocation = TRUE; +Index: config.def.h +--- config.def.h.orig config.def.h +@@ -30,7 +30,7 @@ static Parameter defconfig[ParameterLast] = { + SETB(SiteQuirks,
[UPDATE] databases/py-mongo to 3.4.0
Hello, This patch updates the databases/py-mongo port to 3.4.0, and adds a python3 flavor. It removes building the Sphinx documentation; it seems unnecessary, given that all of the docs are available at [1]. That said, I'm happy to re-add them. All tests pass on -current/amd64. [1]: http://api.mongodb.com/python/current/ Thank you, -- Andrew AldridgeIndex: Makefile === RCS file: /home/andrew/Documents/openbsd/cvs/ports/databases/py-mongo/Makefile,v retrieving revision 1.7 diff -u -p -r1.7 Makefile --- Makefile 15 Apr 2016 08:53:49 - 1.7 +++ Makefile 23 May 2017 20:15:01 - @@ -2,7 +2,7 @@ COMMENT = Python driver for mongodb -MODPY_EGG_VERSION = 2.7.2 +MODPY_EGG_VERSION = 3.4.0 DISTNAME = pymongo-${MODPY_EGG_VERSION} PKGNAME = py-${DISTNAME:S/py//} REVISION = 1 @@ -22,22 +22,12 @@ MODULES = lang/python MODPY_SETUPTOOLS = Yes -BUILD_DEPENDS = textproc/py-sphinx -TEST_DEPENDS = devel/py-nose +TEST_DEPENDS = devel/py-nose${MODPY_FLAVOR} -# needs a running mongodb instance in 127.0.0.1:27017 +# needs a running mongodb instance in 127.0.0.1:27017, +# and increased kern.somaxconn TEST_IS_INTERACTIVE = Yes -post-build: - @cd ${WRKSRC} && \ - ${SETENV} ${MAKE_ENV} ${MODPY_BIN} ${MODPY_SETUP} doc - -post-install: - ${INSTALL_DATA_DIR} ${PREFIX}/share/doc/${MODPY_PY_PREFIX}mongo - cd ${WRKSRC}/doc/_build/${MODPY_EGG_VERSION} && \ - pax -rw * ${PREFIX}/share/doc/${MODPY_PY_PREFIX}mongo - rm -rf ${PREFIX}/share/doc/${MODPY_PY_PREFIX}mongo/_sources - .include .if ${PROPERTIES:Mle} @@ -45,5 +35,8 @@ PKG_ARGS += -Dno_be=1 .else PKG_ARGS += -Dno_be=0 .endif + +FLAVORS = python3 +FLAVOR ?= .include Index: distinfo === RCS file: /home/andrew/Documents/openbsd/cvs/ports/databases/py-mongo/distinfo,v retrieving revision 1.2 diff -u -p -r1.2 distinfo --- distinfo 29 Sep 2014 19:39:53 - 1.2 +++ distinfo 23 May 2017 18:54:55 - @@ -1,2 +1,2 @@ -SHA256 (pymongo-2.7.2.tar.gz) = j7RePt7ABvSxFYD8V5vsd4GosyAc4KvzFZPwhgcOUfY= -SIZE (pymongo-2.7.2.tar.gz) = 381455 +SHA256 (pymongo-3.4.0.tar.gz) = 01k0nGyf+fSCgF+J5m5HaEYxfcex7ql519qcCFfuJyE= +SIZE (pymongo-3.4.0.tar.gz) = 583303 Index: pkg/PLIST === RCS file: /home/andrew/Documents/openbsd/cvs/ports/databases/py-mongo/pkg/PLIST,v retrieving revision 1.4 diff -u -p -r1.4 PLIST --- pkg/PLIST 15 Apr 2016 08:53:49 - 1.4 +++ pkg/PLIST 23 May 2017 19:04:37 - @@ -2,177 +2,141 @@ %%no_be%% lib/python${MODPY_VERSION}/site-packages/bson/ lib/python${MODPY_VERSION}/site-packages/bson/__init__.py -lib/python${MODPY_VERSION}/site-packages/bson/__init__.pyc +${MODPY_COMMENT}lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}/ +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}__init__.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}binary.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}code.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}codec_options.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}dbref.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}decimal128.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}errors.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}int64.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}json_util.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}max_key.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}min_key.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}objectid.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}py3compat.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}raw_bson.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}regex.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}son.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}timestamp.${MODPY_PYC_MAGIC_TAG}pyc +lib/python${MODPY_VERSION}/site-packages/bson/${MODPY_PYCACHE}tz_util.${MODPY_PYC_MAGIC_TAG}pyc lib/python${MODPY_VERSION}/site-packages/bson/binary.py -lib/python${MODPY_VERSION}/site-packages/bson/binary.pyc lib/python${MODPY_VERSION}/site-packages/bson/code.py -lib/python${MODPY_VERSION}/site-packages/bson/code.pyc +lib/python${MODPY_VERSION}/site-packages/bson/codec_options.py lib/python${MODPY_VERSION}/site-packages/bson/dbref.py -lib/python${MODPY_VERSION}/site-packages/bson/dbref
Re: mpd does not work on i386 (patch from FreeBSD)
On Tue, May 23, 2017 at 03:47:56PM -0400, Andre Smagin wrote: > On Tue, 23 May 2017 20:53:12 +0200 > Landry Breuil wrote: > > > Better with an update that actually compiles. works here, but doesnt > > directly fix my own issue... strange. Worked after forcing an update of > > the problematic dir via gmpc, and a rescan. > > > > Either way, we should go for the update if there are no regressions. > > David ? > > Does the update fix the database issue for you? It does not seem to > have any effect here, not without the #define gcc_pure bit. > (0.20.8 was the first thing I checked yesterday, hoping that a simple > update would be enough :) > > terminate called after throwing an instance of 'std::system_error' > what(): Failed to access /storage/music/test: No such file or directory > > May 23 15:18 : exception: Failed to access /storage/music/dts/test: No such > file or directory I'm not sure we were seeing the same issue.. in the end, going into the database in gmpc and updating the problematic dir fixed the issue. And i wasnt seeing a program termination. > When I add the Compiler.h patch to 0.20.8, it starts working, > so perhaps that commit you mentioned was not enough. Then that should be discussed with upstream.. > Also, you have > + static > +-#if GCC_OLDER_THAN(5,0) > ++#if GCC_OLDER_THAN(5,0) || defined(__clang__) > + /* gcc 4.x has no "constexpr" for std::max() */ > + const > > but with > > #define GCC_OLDER_THAN(major, minor) \ > (GCC_VERSION > 0 && CLANG_VERSION == 0 && \ >GCC_VERSION < GCC_MAKE_VERSION(major, minor, 0)) > > we end up with "if not clang or if clang", which works for us, > but does not feel right. I got a bit lost in all the compiler checks > defined in Compiler.h, not sure what the right way is though. > (The logic seems reversed to begin with and clang 4 is supposed > to support constexpr for std::max() I think.) > > But this compiler/library/c++ maze is way above my abilities, sorry. Ditto - now it works, that's all i was looking for. The patch might of course be wrong, that was just to make it build so that i could test it.
Re: mpd does not work on i386 (patch from FreeBSD)
(mailed landry@, but forgot to cc ports) On Tue, 23 May 2017 20:53:12 +0200 Landry Breuil wrote: > Better with an update that actually compiles. works here, but doesnt > directly fix my own issue... strange. Worked after forcing an update of > the problematic dir via gmpc, and a rescan. > > Either way, we should go for the update if there are no regressions. > David ? Does the update fix the database issue for you? It does not seem to have any effect here, not without the #define gcc_pure bit. (0.20.8 was the first thing I checked yesterday, hoping that a simple update would be enough :) terminate called after throwing an instance of 'std::system_error' what(): Failed to access /storage/music/test: No such file or directory May 23 15:18 : exception: Failed to access /storage/music/dts/test: No such file or directory When I add the Compiler.h patch to 0.20.8, it starts working, so perhaps that commit you mentioned was not enough. Also, you have + static +-#if GCC_OLDER_THAN(5,0) ++#if GCC_OLDER_THAN(5,0) || defined(__clang__) + /* gcc 4.x has no "constexpr" for std::max() */ + const but with #define GCC_OLDER_THAN(major, minor) \ (GCC_VERSION > 0 && CLANG_VERSION == 0 && \ GCC_VERSION < GCC_MAKE_VERSION(major, minor, 0)) we end up with "if not clang or if clang", which works for us, but does not feel right. I got a bit lost in all the compiler checks defined in Compiler.h, not sure what the right way is though. (The logic seems reversed to begin with and clang 4 is supposed to support constexpr for std::max() I think.) But this compiler/library/c++ maze is way above my abilities, sorry.
Re: [new] productivity/khal
Ping. On Fri, May 05, 2017 at 10:22:54PM +0200, Joerg Jung wrote: > Hi, > > please find attached a port for khal. While sometimes a bit buggy it > mostly works fine for my basic usage. > > $ cat pkg/DESCR > Khal is a standards based CLI (console) calendar program, able to synchronize > with CalDAV servers through vdirsyncer. > > > Comments? OK to import? > > Thanks, > Regards, > Joerg
Re: [new] devel/py-icalendar
Ping. Anynone? On Fri, May 05, 2017 at 10:20:07PM +0200, Joerg Jung wrote: > Hi, > > please find attached a port for py-icalendar, a dependency for the > upcoming khal port. > > $ cat pkg/DESCR > This is an attempt at making an iCalendar package for Python, to parse and > generate iCalendar files. The inspiration has come from the email package in > the standard lib, yet efficient and powerful. > > > Comments? OK to import? > > Thanks, > Regards, > Joerg
Re: [new] devel/py-freezegun
Ping. Anynone? On Fri, May 05, 2017 at 10:21:07PM +0200, Joerg Jung wrote: > Hi, > > please find attached a port for py-freezegun, a test dependency for the > upcoming khal port. > > $ cat pkg/DESCR > FreezeGun is a library that allows your python tests to travel through time by > mocking the datetime module. > > > Comments? OK to import? > > Thanks, > Regards, > Joerg
cannot change php 7 values through httpd.conf
Hello everybody! I noticed that i can't change php values through apache configuration. Here is example virtualhost: ServerName example.com DocumentRoot /var/www/htdocs/phpmy Options FollowSymLinks DirectoryIndex index.php Require ip 192.168.0 php_value upload_max_filesize 40M php_value post_max_size 50M It doesn't work! But if i create .htaccess with php_value upload_max_filesize 40M php_value post_max_size 50M It works! What am I doing wrong? This config I used earlier with php5, I only changed "mod_php5.c" to "mod_php7.c"
Re: mpd does not work on i386 (patch from FreeBSD)
OK aja On Tue, May 23, 2017 at 08:16:47PM +0200, Landry Breuil wrote: > On Tue, May 23, 2017 at 08:01:51PM +0200, Landry Breuil wrote: > > On Tue, May 23, 2017 at 01:18:17PM -0400, Andre Smagin wrote: > > > I think this is the relevant discussion about the same issue FreeBSD had: > > > > > > https://github.com/MusicPlayerDaemon/MPD/issues/41 > > > > > > This is the FreeBSD commit to fix it: > > > https://svnweb.freebsd.org/ports/head/audio/musicpd/files/patch-src_Compiler.h?revision=440334&view=markup > > > > > > "Fix library updating > > > Details: > > > Apparently, clang optimises exceptions with __attribute__((pure)) > > > more aggressively than gcc, causing mpd to not notice files removed > > > from the library. This patch causes mpd to be built without the pure > > > attribute." > > > > > > > > > I am testing with this diff (from FreeBSD) now and at a first glance > > > it seems to fix the database update problem: > > > > Good catch, testing it here too. > > Actually, we might aswell update to 0.20.8, which contains the fix > commmited upstream (ie removing the gcc_pure usage) in > https://github.com/MusicPlayerDaemon/MPD/commit/788e3b31e1ab8243386339eaf136ede18f015d55 > > > Index: Makefile > === > RCS file: /cvs/ports/audio/mpd/Makefile,v > retrieving revision 1.103 > diff -u -r1.103 Makefile > --- Makefile12 Mar 2017 10:13:49 - 1.103 > +++ Makefile23 May 2017 18:14:59 - > @@ -2,7 +2,7 @@ > > BROKEN-hppa = no atomic ops > COMMENT = Music Player Daemon > -VER = 0.20.6 > +VER = 0.20.8 > DISTNAME = mpd-${VER} > CATEGORIES = audio > HOMEPAGE = http://www.musicpd.org/ > Index: distinfo > === > RCS file: /cvs/ports/audio/mpd/distinfo,v > retrieving revision 1.44 > diff -u -r1.44 distinfo > --- distinfo12 Mar 2017 10:13:49 - 1.44 > +++ distinfo23 May 2017 18:14:59 - > @@ -1,2 +1,2 @@ > -SHA256 (mpd-0.20.6.tar.gz) = HRTyzRcGcIpRm7yI8wqsgF+HkSOMoIN0r18IJYcfVQI= > -SIZE (mpd-0.20.6.tar.gz) = 1091201 > +SHA256 (mpd-0.20.8.tar.gz) = RRlzEbbDAlUoFQTjamrvtHktAqGmM4xVaq1jN0vL73g= > +SIZE (mpd-0.20.8.tar.gz) = 1093412 > -- Antoine
Re: mpd does not work on i386 (patch from FreeBSD)
On Tue, May 23, 2017 at 08:16:47PM +0200, Landry Breuil wrote: > On Tue, May 23, 2017 at 08:01:51PM +0200, Landry Breuil wrote: > > On Tue, May 23, 2017 at 01:18:17PM -0400, Andre Smagin wrote: > > > I think this is the relevant discussion about the same issue FreeBSD had: > > > > > > https://github.com/MusicPlayerDaemon/MPD/issues/41 > > > > > > This is the FreeBSD commit to fix it: > > > https://svnweb.freebsd.org/ports/head/audio/musicpd/files/patch-src_Compiler.h?revision=440334&view=markup > > > > > > "Fix library updating > > > Details: > > > Apparently, clang optimises exceptions with __attribute__((pure)) > > > more aggressively than gcc, causing mpd to not notice files removed > > > from the library. This patch causes mpd to be built without the pure > > > attribute." > > > > > > > > > I am testing with this diff (from FreeBSD) now and at a first glance > > > it seems to fix the database update problem: > > > > Good catch, testing it here too. > > Actually, we might aswell update to 0.20.8, which contains the fix > commmited upstream (ie removing the gcc_pure usage) in > https://github.com/MusicPlayerDaemon/MPD/commit/788e3b31e1ab8243386339eaf136ede18f015d55 Better with an update that actually compiles. works here, but doesnt directly fix my own issue... strange. Worked after forcing an update of the problematic dir via gmpc, and a rescan. Either way, we should go for the update if there are no regressions. David ? ? mpd-0.18.10_rev2.diff ? patch-src_Compiler_h Index: Makefile === RCS file: /cvs/ports/audio/mpd/Makefile,v retrieving revision 1.103 diff -u -r1.103 Makefile --- Makefile12 Mar 2017 10:13:49 - 1.103 +++ Makefile23 May 2017 18:47:19 - @@ -2,7 +2,7 @@ BROKEN-hppa = no atomic ops COMMENT = Music Player Daemon -VER = 0.20.6 +VER = 0.20.8 DISTNAME = mpd-${VER} CATEGORIES = audio HOMEPAGE = http://www.musicpd.org/ Index: distinfo === RCS file: /cvs/ports/audio/mpd/distinfo,v retrieving revision 1.44 diff -u -r1.44 distinfo --- distinfo12 Mar 2017 10:13:49 - 1.44 +++ distinfo23 May 2017 18:47:19 - @@ -1,2 +1,2 @@ -SHA256 (mpd-0.20.6.tar.gz) = HRTyzRcGcIpRm7yI8wqsgF+HkSOMoIN0r18IJYcfVQI= -SIZE (mpd-0.20.6.tar.gz) = 1091201 +SHA256 (mpd-0.20.8.tar.gz) = RRlzEbbDAlUoFQTjamrvtHktAqGmM4xVaq1jN0vL73g= +SIZE (mpd-0.20.8.tar.gz) = 1093412 Index: patches/patch-src_Main_cxx === RCS file: patches/patch-src_Main_cxx diff -N patches/patch-src_Main_cxx --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_Main_cxx 23 May 2017 18:47:19 - @@ -0,0 +1,16 @@ +$OpenBSD$ + +Fix logic + +Index: src/Main.cxx +--- src/Main.cxx.orig src/Main.cxx +@@ -123,7 +123,7 @@ static constexpr size_t MEGABYTE = 1024 * KILOBYTE; + static constexpr size_t DEFAULT_BUFFER_SIZE = 4 * MEGABYTE; + + static +-#if GCC_OLDER_THAN(5,0) ++#if GCC_OLDER_THAN(5,0) || defined(__clang__) + /* gcc 4.x has no "constexpr" for std::max() */ + const + #else
[UPDATE] databases/mongodb 3.2.13
Hello, This patch updates MongoDB to 3.2.13 which, among a lot of other things[1], updates the bundled SpiderMonkey release. It also makes two changes to the sets of patches: * It REMOVES a patch that worked around siginfo->si_addr not being POSIX-compliant. This has been fixed in OpenBSD-current for about six weeks now. * It ADDS a patch implementing the fix for SERVER-29083[2], which unfortunately will not be backported to the 3.2 branch. This fixes a use-after-free, meaning that the core integration test suite now passes in addition to the unit tests. As a note: it seems that you have to reboot between building and running the tests, because building MongoDB wires all the pages (systat reports negative wired pages; integer overflow?) preventing mlock(..., 4096) from succeeding. This may be a regression since OpenBSD 6.1. [1]: https://docs.mongodb.com/manual/release-notes/3.2-changelog/#id1 [2]: https://jira.mongodb.org/browse/SERVER-29083 Thank you, -- Andrew AldridgeIndex: Makefile === RCS file: /home/andrew/Documents/openbsd/cvs/ports/databases/mongodb/Makefile,v retrieving revision 1.25 diff -u -p -r1.25 Makefile --- Makefile 22 May 2017 14:00:21 - 1.25 +++ Makefile 23 May 2017 18:03:25 - @@ -9,7 +9,7 @@ DPB_PROPERTIES = parallel COMMENT = scalable, high-performance document-oriented database -DISTNAME = mongodb-src-r3.2.12 +DISTNAME = mongodb-src-r3.2.13 PKGNAME = ${DISTNAME:S/src-r//} CATEGORIES = databases Index: distinfo === RCS file: /home/andrew/Documents/openbsd/cvs/ports/databases/mongodb/distinfo,v retrieving revision 1.11 diff -u -p -r1.11 distinfo --- distinfo 6 Apr 2017 22:11:23 - 1.11 +++ distinfo 23 May 2017 15:45:41 - @@ -1,2 +1,2 @@ -SHA256 (mongodb-src-r3.2.12.tar.gz) = QGzthzqavoo/D5KTm/U3RmO6yGxyGByAvsFErt9s0wY= -SIZE (mongodb-src-r3.2.12.tar.gz) = 29178424 +SHA256 (mongodb-src-r3.2.13.tar.gz) = MUkrmdEtY2Om68u6Miac6el9re3ajEcMv/jBryJ+l1M= +SIZE (mongodb-src-r3.2.13.tar.gz) = 29273616 Index: patches/patch-src_mongo_util_processinfo_openbsd_cpp === RCS file: patches/patch-src_mongo_util_processinfo_openbsd_cpp diff -N patches/patch-src_mongo_util_processinfo_openbsd_cpp --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_mongo_util_processinfo_openbsd_cpp 23 May 2017 15:50:57 - @@ -0,0 +1,29 @@ +$OpenBSD$ + +Index: src/mongo/util/processinfo_openbsd.cpp +--- src/mongo/util/processinfo_openbsd.cpp.orig src/mongo/util/processinfo_openbsd.cpp +@@ -110,9 +110,10 @@ int ProcessInfo::getVirtualMemorySize() { + } + + kinfo_proc* task = kvm_getprocs(kd, KERN_PROC_PID, _pid.toNative(), sizeof(kinfo_proc), &cnt); +-kvm_close(kd); +-return ((task->p_vm_dsize + task->p_vm_ssize + task->p_vm_tsize) * sysconf(_SC_PAGESIZE)) / ++int rss = ((task->p_vm_dsize + task->p_vm_ssize + task->p_vm_tsize) * sysconf(_SC_PAGESIZE)) / + 1048576; ++kvm_close(kd); ++return rss; + } + + int ProcessInfo::getResidentSize() { +@@ -124,8 +125,9 @@ int ProcessInfo::getResidentSize() { + return -1; + } + kinfo_proc* task = kvm_getprocs(kd, KERN_PROC_PID, _pid.toNative(), sizeof(kinfo_proc), &cnt); ++int rss = (task->p_vm_rssize * sysconf(_SC_PAGESIZE)) / 1048576; // convert from pages to MB + kvm_close(kd); +-return (task->p_vm_rssize * sysconf(_SC_PAGESIZE)) / 1048576; // convert from pages to MB ++return rss; + } + + double ProcessInfo::getSystemMemoryPressurePercentage() { Index: patches/patch-src_mongo_util_signal_handlers_synchronous_cpp === RCS file: patches/patch-src_mongo_util_signal_handlers_synchronous_cpp diff -N patches/patch-src_mongo_util_signal_handlers_synchronous_cpp --- patches/patch-src_mongo_util_signal_handlers_synchronous_cpp 6 May 2016 15:33:37 - 1.1 +++ /dev/null 1 Jan 1970 00:00:00 - @@ -1,12 +0,0 @@ -$OpenBSD: patch-src_mongo_util_signal_handlers_synchronous_cpp,v 1.1 2016/05/06 15:33:37 sthen Exp $ src/mongo/util/signal_handlers_synchronous.cpp.orig Thu Apr 21 16:21:39 2016 -+++ src/mongo/util/signal_handlers_synchronous.cpp Sun May 1 23:59:28 2016 -@@ -266,7 +266,7 @@ void abruptQuitWithAddrSignal(int signalNum, siginfo_t - MallocFreeOStreamGuard lk{}; - - const char* action = (signalNum == SIGSEGV || signalNum == SIGBUS) ? "access" : "operation"; --mallocFreeOStream << "Invalid " << action << " at address: " << siginfo->si_addr; -+mallocFreeOStream << "Invalid " << action << " at address: " << reinterpret_cast(siginfo->si_addr); - - // Writing out message to log separate from the stack trace so at least that much gets - // logged. This is important because we may get here by jumping to an invalid address which
Re: mpd does not work on i386 (patch from FreeBSD)
On Tue, May 23, 2017 at 08:01:51PM +0200, Landry Breuil wrote: > On Tue, May 23, 2017 at 01:18:17PM -0400, Andre Smagin wrote: > > I think this is the relevant discussion about the same issue FreeBSD had: > > > > https://github.com/MusicPlayerDaemon/MPD/issues/41 > > > > This is the FreeBSD commit to fix it: > > https://svnweb.freebsd.org/ports/head/audio/musicpd/files/patch-src_Compiler.h?revision=440334&view=markup > > > > "Fix library updating > > Details: > > Apparently, clang optimises exceptions with __attribute__((pure)) > > more aggressively than gcc, causing mpd to not notice files removed > > from the library. This patch causes mpd to be built without the pure > > attribute." > > > > > > I am testing with this diff (from FreeBSD) now and at a first glance > > it seems to fix the database update problem: > > Good catch, testing it here too. Actually, we might aswell update to 0.20.8, which contains the fix commmited upstream (ie removing the gcc_pure usage) in https://github.com/MusicPlayerDaemon/MPD/commit/788e3b31e1ab8243386339eaf136ede18f015d55 Index: Makefile === RCS file: /cvs/ports/audio/mpd/Makefile,v retrieving revision 1.103 diff -u -r1.103 Makefile --- Makefile12 Mar 2017 10:13:49 - 1.103 +++ Makefile23 May 2017 18:14:59 - @@ -2,7 +2,7 @@ BROKEN-hppa = no atomic ops COMMENT = Music Player Daemon -VER = 0.20.6 +VER = 0.20.8 DISTNAME = mpd-${VER} CATEGORIES = audio HOMEPAGE = http://www.musicpd.org/ Index: distinfo === RCS file: /cvs/ports/audio/mpd/distinfo,v retrieving revision 1.44 diff -u -r1.44 distinfo --- distinfo12 Mar 2017 10:13:49 - 1.44 +++ distinfo23 May 2017 18:14:59 - @@ -1,2 +1,2 @@ -SHA256 (mpd-0.20.6.tar.gz) = HRTyzRcGcIpRm7yI8wqsgF+HkSOMoIN0r18IJYcfVQI= -SIZE (mpd-0.20.6.tar.gz) = 1091201 +SHA256 (mpd-0.20.8.tar.gz) = RRlzEbbDAlUoFQTjamrvtHktAqGmM4xVaq1jN0vL73g= +SIZE (mpd-0.20.8.tar.gz) = 1093412
Re: mpd does not work on i386 (patch from FreeBSD)
On Tue, May 23, 2017 at 01:18:17PM -0400, Andre Smagin wrote: > I think this is the relevant discussion about the same issue FreeBSD had: > > https://github.com/MusicPlayerDaemon/MPD/issues/41 > > This is the FreeBSD commit to fix it: > https://svnweb.freebsd.org/ports/head/audio/musicpd/files/patch-src_Compiler.h?revision=440334&view=markup > > "Fix library updating > Details: > Apparently, clang optimises exceptions with __attribute__((pure)) > more aggressively than gcc, causing mpd to not notice files removed > from the library. This patch causes mpd to be built without the pure > attribute." > > > I am testing with this diff (from FreeBSD) now and at a first glance > it seems to fix the database update problem: Good catch, testing it here too.
Re: mpd does not work on i386 (patch from FreeBSD)
I think this is the relevant discussion about the same issue FreeBSD had: https://github.com/MusicPlayerDaemon/MPD/issues/41 This is the FreeBSD commit to fix it: https://svnweb.freebsd.org/ports/head/audio/musicpd/files/patch-src_Compiler.h?revision=440334&view=markup "Fix library updating Details: Apparently, clang optimises exceptions with __attribute__((pure)) more aggressively than gcc, causing mpd to not notice files removed from the library. This patch causes mpd to be built without the pure attribute." I am testing with this diff (from FreeBSD) now and at a first glance it seems to fix the database update problem: Index: patches/patch-src_Compiler_h === RCS file: patches/patch-src_Compiler_h diff -N patches/patch-src_Compiler_h --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_Compiler_h23 May 2017 17:00:24 - @@ -0,0 +1,14 @@ +$OpenBSD$ + +Index: src/Compiler.h +--- src/Compiler.h.orig src/Compiler.h +@@ -86,7 +86,7 @@ + #define gcc_noreturn __attribute__((noreturn)) + #define gcc_packed __attribute__((packed)) + #define gcc_printf(a,b) __attribute__((format(printf, a, b))) +-#define gcc_pure __attribute__((pure)) ++#define gcc_pure + #define gcc_sentinel __attribute__((sentinel)) + #define gcc_unused __attribute__((unused)) + #define gcc_warn_unused_result __attribute__((warn_unused_result))
Re: Update to py-argcomplete-1.8.2
On Tue, May 23, 2017 at 12:48:42PM +0100, Stuart Henderson wrote: > On 2017/03/07 20:15, Daniel Jakots wrote: > > On Tue, 21 Feb 2017 16:13:50 +, Stuart Henderson > > wrote: > > > > Sorry for the delay, I slacked a bit. > > > > > Tests are a bit messed up though, some missing deps - py-coverage, > > > py-pexpect, bash, tcsh we have - there's also devel/flake8 of which > > > we only have a py2 version (and this was imported as just "flake8" > > > rather than "py-flake8" so adding a py3 flavour is going to be more > > > fiddly than usual). It also uses "wheel" which isn't in ports as far > > > as I can see. > > > > I talked with flake8 maintainer about splitting it and he said that it > > was useless. Users only need the 'binary' and the fact that > > it is made from py2 or py3 changes nothing. Sadly he never had the time > > to write it on ports@. > > > > His point sounds legit so I checked for argcomplete and it seems it's > > not even used. I just added a patch to zap it from setup.py so we > > can drop the need to split flake8. > > I see where he's coming from, but if you're just working on things > written using the current version of python, it's annoying to have to > install a legacy version just to run flake8 (and vice-versa, I suppose..) > > > And probably wheel isn't needed neither. ??\_(???)_/?? > > > > > The do-test override isn't needed in this version, the default test > > > target gives much more information about failures. But it does need > > > "python" to be in the path, it can't cope with "python2.7". > > > > > > Also I see it runs out of PTYs during tests unless I create more > > > device nodes. > > > > > > So here's a better-but-not-perfect diff.. > > > > Updated diff, ok danj@ if someone wants to commit > > Looks good to me. Though I would also support switching flake8 to py3 > and removing the patch.. > I think the same as sthen@. > > Index: Makefile > > === > > RCS file: /cvs/ports/devel/py-argcomplete/Makefile,v > > retrieving revision 1.7 > > diff -u -p -r1.7 Makefile > > --- Makefile3 Jan 2017 19:19:48 - 1.7 > > +++ Makefile8 Mar 2017 01:13:23 - > > @@ -2,11 +2,10 @@ > > > > COMMENT = bash tab completion for argparse > > > > -MODPY_EGG_VERSION =1.4.1 > > +MODPY_EGG_VERSION =1.8.2 > > DISTNAME = argcomplete-${MODPY_EGG_VERSION} > > PKGNAME = py-${DISTNAME} > > CATEGORIES = devel > > -REVISION = 0 > > > > # GPLv3+ > > PERMIT_PACKAGE_CDROM= Yes > > @@ -19,12 +18,20 @@ MODPY_SETUPTOOLS = Yes > > FLAVORS = python3 > > FLAVOR ?= > > > > +TEST_DEPENDS = devel/py-coverage${MODPY_FLAVOR} \ > > + devel/py-pexpect${MODPY_FLAVOR} \ > > + devel/py-wheel${MODPY_FLAVOR} \ > > + shells/bash \ > > + shells/tcsh > > + > > post-install: > > for i in ${PREFIX}/bin/*; do \ > > mv $${i} $${i}${MODPY_BIN_SUFFIX} ;\ > > done > > > > -do-test: > > - cd ${WRKSRC}/test && ${MODPY_BIN} test.py > > +pre-test: > > + @ln -fs ${MODPY_BIN} ${WRKDIR}/bin/python > > + @[[ -r /dev/ptyz0 ]] || (printf "\\nWARNING: tests are likely to fail\ > > + unless additional ptys are available.\\n\\n" >&2; sleep 2) > > > > .include > > Index: distinfo > > === > > RCS file: /cvs/ports/devel/py-argcomplete/distinfo,v > > retrieving revision 1.3 > > diff -u -p -r1.3 distinfo > > --- distinfo28 Aug 2016 10:55:38 - 1.3 > > +++ distinfo8 Mar 2017 01:13:23 - > > @@ -1,2 +1,2 @@ > > -SHA256 (argcomplete-1.4.1.tar.gz) = > > p7X6jRrLaeSbPGs/YiW8cJCS4Mf2ITEbrFB6T27+YJ0= > > -SIZE (argcomplete-1.4.1.tar.gz) = 37219 > > +SHA256 (argcomplete-1.8.2.tar.gz) = > > eizN4HTrZjEjKy+R+NjVmF1b6/DjawkREwluOgOFVGs= > > +SIZE (argcomplete-1.8.2.tar.gz) = 53626 > > Index: patches/patch-setup_py > > === > > RCS file: patches/patch-setup_py > > diff -N patches/patch-setup_py > > --- /dev/null 1 Jan 1970 00:00:00 - > > +++ patches/patch-setup_py 8 Mar 2017 01:13:23 - > > @@ -0,0 +1,12 @@ > > +$OpenBSD$ > > +--- setup.py.orig Tue Mar 7 19:55:42 2017 > > setup.py Tue Mar 7 19:55:53 2017 > > +@@ -4,7 +4,7 @@ import glob > > + from setuptools import setup, find_packages > > + > > + install_requires = [] > > +-tests_require = ["coverage", "flake8", "pexpect", "wheel"] > > ++tests_require = ["coverage", "pexpect", "wheel"] > > + > > + try: > > + import argparse # noqa > > Index: pkg/PLIST > > === > > RCS file: /cvs/ports/devel/py-argcomplete/pkg/PLIST,v > > retrieving revision 1.2 > > diff -u -p -r1.2 PLIST > > --- pkg/PLIST 28 Aug 2016 10:55:38 - 1.2 > > +++ pkg/PLIST 8 Mar 2017 01:13:23 - > > @@ -1,6 +1,7 @@ > > @comment $OpenBS
Re: mpd does not work on i386
On Tue, May 23, 2017 at 05:05:38PM +0200, Jeremie Courreges-Anglas wrote: > Andre Smagin writes: > > > On Sat, 13 May 2017 18:35:41 +0200 > > Antoine Jacoutot wrote: > > > >> Hi. > >> > >> audio/mpd is pretty useless on 6.1 i386, it cannot update it's database. > > > > Hello. > > > > mpd has issues updating the database on amd64 as well. > > > > If a directory in the root of music_directory (/storage/music in my case) > > is deleted and database update requested, mpd dies with this in the log: > > > > terminate called after throwing an instance of 'std::system_error' > > what(): Failed to access /storage/music/test: No such file or directory > > > > If a nested directory is deleted, mpd does not die on db update, > > but fails to remove the deleted directory from the database: > > > > May 22 09:48 : exception: Failed to access /storage/music/dts/test: No such > > file or directory > > > > Requesting an update specifically on the deleted directory works fine: > > mpc update test > > May 22 09:50 : update: removing test/The Four Seasons (Blu-ray, DTS, > > 2011)/Antonio Vialdi - The Four Seasons.flac > > > > This issue appears to exist in mpd-0.20.5, mpd-0.20.6, and mpd-0.20.8. > > > > Reverting back to mpd-0.19.21 appears to fix the problem. > > cvs up -D "Feb 21"; make install > > > > An addition, a thing I have not noticed before: mpd-0.19.21 is less > > forgiving to the filenames with, what I assume, are illegal UTF-8 > > characters than the newer releases. Upon downgrading and requesting > > a database update, I had a few entries in the log like this: > > > > May 22 09:54 : update: removing flac/Classical/Edvard Grieg/The Grieg > > Edition/cd20 - Music for Strings/04 - ÅSe's Death, Op.46 No. 2.flac > > May 22 09:54 : update: removing flac/Classical/Gustav > > Mahler/Symphonies/Symphony No. 9/02 - Symphony No. 9 In D Major: II. Im > > Tempo Eines GemäChlichen LäNders.flac > > May 22 09:54 : update: removing mp3/_Korean/Park Hwa Yobi/5집/14 - 36.5°.mp3 > > > > I suspect that the filesystem walking code had changed since mpd-0.19.21. > > I just hit this on amd64, update with Ctrl+u in ncmpc. A backtrace with > egdb gives more details: > > Thread 2 (thread 401205): > #0 0x0f5a1a67463a in thrkill () at {standard input}:5 > #1 0x0f5a1a69b379 in *_libc_abort () at > /usr/src/lib/libc/stdlib/abort.c:52 > #2 0x0f59f7765ef5 in __gnu_cxx::__verbose_terminate_handler () at > /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/vterminate.cc:95 > #3 0x0f59f7763a88 in __cxxabiv1::__terminate (handler=) > at > /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_terminate.cc:47 > #4 0x0f59f7763af1 in std::terminate () at > /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_terminate.cc:57 > #5 0x0f59f7763d08 in __cxxabiv1::__cxa_throw (obj=0xf599b607980, > tinfo=0xf59f7a6ec60 , dest=0xf59f77c3ff0 > ) > at > /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_throw.cc:87 > #6 0x0f575372ea85 in FileInfo::FileInfo (this=, path=..., > follow_symlinks=) at ./src/fs/FileInfo.hxx:72 > #7 0x0f575373cbed in LocalStorage::GetInfo(char const*, bool) () at > /usr/local/include/c++/4.9.4/bits/basic_string.h:249 > #8 0x0f575372fb72 in GetDirectoryChildInfo (storage=..., directory=..., > name_utf8=0xf59fd3e7838 "audio.mp2") at src/db/update/UpdateIO.cxx:77 > #9 directory_child_is_regular (storage=..., directory=..., > name_utf8=0xf59fd3e7838 "audio.mp2") at src/db/update/UpdateIO.cxx:84 > #10 0x0f57537228d3 in > UpdateWalk::PurgeDeletedFromDirectory(Directory&)::{lambda(Song&)#1}::operator()(Song&) > const (song=..., this=) at src/db/update/Walk.cxx:116 > #11 > Directory::ForEachSongSafe(UpdateWalk::PurgeDeletedFromDirectory(Directory&)::{lambda(Song&)#1}&&) > (this=, t=) > at ./src/db/plugins/simple/Directory.hxx:222 > #12 UpdateWalk::PurgeDeletedFromDirectory (this=0xf5a30e59d40, directory=...) > at src/db/update/Walk.cxx:115 > #13 0x0f575372233f in UpdateWalk::UpdateDirectory (this=0xf5a30e59d40, > directory=..., exclude_list=..., info=...) at src/db/update/Walk.cxx:358 > #14 0x0f5753722a72 in UpdateWalk::Walk (this=0xf5a30e59d40, root=..., > path=, discard=) at src/db/update/Walk.cxx:499 > #15 0x0f575372145b in UpdateService::Task() () at > /usr/local/include/c++/4.9.4/bits/basic_string.h:249 > #16 0x0f57537776fe in Thread::ThreadProc (ctx=0x0) at > src/thread/Thread.cxx:105 > #17 0x0f59b24768ce in _rthread_start (v=0x0) at > /usr/src/lib/librthread/rthread.c:114 > #18 0x0f5a1a6919bb in __tfork_thread () at > /usr/src/lib/libc/arch/amd64/sys/tfork_thread.S:75 > #19 0x in ?? () > > The exception comes from an audio file that I have removed since the > last db update. That kinda makes your music collection append-only... > I just removed the existing db file for now. I've also seen this a while ago but here it didnt crash, it refused to update the db further. 0.20.
NEW: tdbc-mysql
Provides a database interface that conforms to Tcl DataBase Connectivity (TDBC) and allows a Tcl script to connect to a MariaDB database. One test failure: tdbc::mysql-19.11 $connection configure - -isolation FAILED Contents of test case: list [::db configure -isolation readuncommitted] [::db configure -isolation] [::db configure -isolation readcommitte$ Result was: {} repeatableread {} repeatableread {} repeatableread {} repeatableread Result should have been (exact matching): {} readuncommitted {} readcommitted {} serializable {} repeatableread tdbc::mysql-19.11 FAILED I suspect that the problem is with mariadb on obsd. $ uname -mrsv Linux 2.6.18-416.el5 #1 SMP Fri Oct 28 11:52:49 UTC 2016 x86_64 $ mysql --version mysql Ver 14.12 Distrib 5.0.95, for redhat-linux-gnu (x86_64) using readline 5.1 mysql> select @@global.tx_isolation, @@tx_isolation;set transaction isolation level serializable;select @@global.tx_isolation, @@tx_isolation; +---+-+ | @@global.tx_isolation | @@tx_isolation | +---+-+ | REPEATABLE-READ | REPEATABLE-READ | +---+-+ 1 row in set (0.00 sec) Query OK, 0 rows affected (0.00 sec) +---++ | @@global.tx_isolation | @@tx_isolation | +---++ | REPEATABLE-READ | SERIALIZABLE | +---++ 1 row in set (0.00 sec) $ uname -mrsv OpenBSD 6.1 GENERIC.MP#59 i386 $ mysql --version mysql Ver 15.1 Distrib 10.0.30-MariaDB, for OpenBSD (i386) using readline 4.3 MariaDB [(none)]> select @@global.tx_isolation, @@tx_isolation;set transaction isolation level serializable;select @@global.tx_isolation, @@tx_isolation; +---+-+ | @@global.tx_isolation | @@tx_isolation | +---+-+ | REPEATABLE-READ | REPEATABLE-READ | +---+-+ 1 row in set (0.00 sec) Query OK, 0 rows affected (0.00 sec) +---+-+ | @@global.tx_isolation | @@tx_isolation | +---+-+ | REPEATABLE-READ | REPEATABLE-READ | +---+-+ 1 row in set (0.00 sec) Stu tdbc-mysql-1.0.4-port.tar.gz Description: application/gzip
Re: mpd does not work on i386
Andre Smagin writes: > On Sat, 13 May 2017 18:35:41 +0200 > Antoine Jacoutot wrote: > >> Hi. >> >> audio/mpd is pretty useless on 6.1 i386, it cannot update it's database. > > Hello. > > mpd has issues updating the database on amd64 as well. > > If a directory in the root of music_directory (/storage/music in my case) > is deleted and database update requested, mpd dies with this in the log: > > terminate called after throwing an instance of 'std::system_error' > what(): Failed to access /storage/music/test: No such file or directory > > If a nested directory is deleted, mpd does not die on db update, > but fails to remove the deleted directory from the database: > > May 22 09:48 : exception: Failed to access /storage/music/dts/test: No such > file or directory > > Requesting an update specifically on the deleted directory works fine: > mpc update test > May 22 09:50 : update: removing test/The Four Seasons (Blu-ray, DTS, > 2011)/Antonio Vialdi - The Four Seasons.flac > > This issue appears to exist in mpd-0.20.5, mpd-0.20.6, and mpd-0.20.8. > > Reverting back to mpd-0.19.21 appears to fix the problem. > cvs up -D "Feb 21"; make install > > An addition, a thing I have not noticed before: mpd-0.19.21 is less > forgiving to the filenames with, what I assume, are illegal UTF-8 > characters than the newer releases. Upon downgrading and requesting > a database update, I had a few entries in the log like this: > > May 22 09:54 : update: removing flac/Classical/Edvard Grieg/The Grieg > Edition/cd20 - Music for Strings/04 - ÅSe's Death, Op.46 No. 2.flac > May 22 09:54 : update: removing flac/Classical/Gustav > Mahler/Symphonies/Symphony No. 9/02 - Symphony No. 9 In D Major: II. Im Tempo > Eines GemäChlichen LäNders.flac > May 22 09:54 : update: removing mp3/_Korean/Park Hwa Yobi/5집/14 - 36.5°.mp3 > > I suspect that the filesystem walking code had changed since mpd-0.19.21. I just hit this on amd64, update with Ctrl+u in ncmpc. A backtrace with egdb gives more details: Thread 2 (thread 401205): #0 0x0f5a1a67463a in thrkill () at {standard input}:5 #1 0x0f5a1a69b379 in *_libc_abort () at /usr/src/lib/libc/stdlib/abort.c:52 #2 0x0f59f7765ef5 in __gnu_cxx::__verbose_terminate_handler () at /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/vterminate.cc:95 #3 0x0f59f7763a88 in __cxxabiv1::__terminate (handler=) at /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_terminate.cc:47 #4 0x0f59f7763af1 in std::terminate () at /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_terminate.cc:57 #5 0x0f59f7763d08 in __cxxabiv1::__cxa_throw (obj=0xf599b607980, tinfo=0xf59f7a6ec60 , dest=0xf59f77c3ff0 ) at /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/libstdc++-v3/libsupc++/eh_throw.cc:87 #6 0x0f575372ea85 in FileInfo::FileInfo (this=, path=..., follow_symlinks=) at ./src/fs/FileInfo.hxx:72 #7 0x0f575373cbed in LocalStorage::GetInfo(char const*, bool) () at /usr/local/include/c++/4.9.4/bits/basic_string.h:249 #8 0x0f575372fb72 in GetDirectoryChildInfo (storage=..., directory=..., name_utf8=0xf59fd3e7838 "audio.mp2") at src/db/update/UpdateIO.cxx:77 #9 directory_child_is_regular (storage=..., directory=..., name_utf8=0xf59fd3e7838 "audio.mp2") at src/db/update/UpdateIO.cxx:84 #10 0x0f57537228d3 in UpdateWalk::PurgeDeletedFromDirectory(Directory&)::{lambda(Song&)#1}::operator()(Song&) const (song=..., this=) at src/db/update/Walk.cxx:116 #11 Directory::ForEachSongSafe(UpdateWalk::PurgeDeletedFromDirectory(Directory&)::{lambda(Song&)#1}&&) (this=, t=) at ./src/db/plugins/simple/Directory.hxx:222 #12 UpdateWalk::PurgeDeletedFromDirectory (this=0xf5a30e59d40, directory=...) at src/db/update/Walk.cxx:115 #13 0x0f575372233f in UpdateWalk::UpdateDirectory (this=0xf5a30e59d40, directory=..., exclude_list=..., info=...) at src/db/update/Walk.cxx:358 #14 0x0f5753722a72 in UpdateWalk::Walk (this=0xf5a30e59d40, root=..., path=, discard=) at src/db/update/Walk.cxx:499 #15 0x0f575372145b in UpdateService::Task() () at /usr/local/include/c++/4.9.4/bits/basic_string.h:249 #16 0x0f57537776fe in Thread::ThreadProc (ctx=0x0) at src/thread/Thread.cxx:105 #17 0x0f59b24768ce in _rthread_start (v=0x0) at /usr/src/lib/librthread/rthread.c:114 #18 0x0f5a1a6919bb in __tfork_thread () at /usr/src/lib/libc/arch/amd64/sys/tfork_thread.S:75 #19 0x in ?? () The exception comes from an audio file that I have removed since the last db update. That kinda makes your music collection append-only... I just removed the existing db file for now. -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: [UPDATE] latexmk
> Edd Barrett hat am 19. März 2017 um 16:13 geschrieben: > > > On Fri, Feb 24, 2017 at 01:56:34PM +0100, Nils Reuße wrote: > > Here is a small bump to the latest version of latexmk (4.52c), the version > > in ports being nearly two years old now. The new version fixed some small > > bugs regarding lualatex and works well for me. This version is also > > available in CTAN: http://www.ctan.org/pkg/latexmk/. > > > > Any comments? > > This looks good to me. Tested lightly. > > Can I get an OK someone? > > In a separate commit I'd like to remove the trailing whitespace in the > Makefile, and I suppose `perl` should be `Perl` in the pkg/DESCR. > > Thanks > > -- > Best Regards > Edd Barrett > > http://www.theunixzoo.co.uk > Now that texlive 2016 is in, i thought i could give this another try. Any comments?Index: Makefile === RCS file: /cvs/ports/textproc/latexmk/Makefile,v retrieving revision 1.5 diff -u -p -r1.5 Makefile --- Makefile 21 Mar 2015 13:54:17 - 1.5 +++ Makefile 24 Feb 2017 12:49:03 - @@ -2,9 +2,8 @@ COMMENT= LaTeX documents output generation automation tool -DISTNAME= latexmk-443a +DISTNAME= latexmk-452c EXTRACT_SUFX= .zip -REVISION= 0 CATEGORIES= textproc print Index: distinfo === RCS file: /cvs/ports/textproc/latexmk/distinfo,v retrieving revision 1.3 diff -u -p -r1.3 distinfo --- distinfo 21 Mar 2015 13:52:13 - 1.3 +++ distinfo 24 Feb 2017 12:49:03 - @@ -1,2 +1,2 @@ -SHA256 (latexmk-443a.zip) = 5BDSlcCkcye5U+zltYLClDWb34kTjvmQ1WIbAg/yu+U= -SIZE (latexmk-443a.zip) = 434070 +SHA256 (latexmk-452c.zip) = eo/VDgvUbIcFqk4Py8NtAdUozJKpbXHhwA4fsuZITbI= +SIZE (latexmk-452c.zip) = 391327 Index: patches/patch-latexmk_pl === RCS file: /cvs/ports/textproc/latexmk/patches/patch-latexmk_pl,v retrieving revision 1.4 diff -u -p -r1.4 patch-latexmk_pl --- patches/patch-latexmk_pl 21 Mar 2015 13:54:17 - 1.4 +++ patches/patch-latexmk_pl 24 Feb 2017 12:49:03 - @@ -1,7 +1,7 @@ $OpenBSD: patch-latexmk_pl,v 1.4 2015/03/21 13:54:17 ajacoutot Exp $ latexmk.pl.orig Thu Feb 5 23:00:49 2015 -+++ latexmk.pl Sat Mar 21 14:53:34 2015 -@@ -1053,6 +1053,10 @@ else { +--- latexmk.pl.orig Thu Jan 19 15:52:57 2017 latexmk.pl Fri Feb 24 10:36:01 2017 +@@ -978,6 +978,10 @@ else { $lpr_pdf = 'lpr %O %S'; $pscmd = "ps -ww -u $ENV{USER}"; }
Re: Update to py-argcomplete-1.8.2
On 2017/03/07 20:15, Daniel Jakots wrote: > On Tue, 21 Feb 2017 16:13:50 +, Stuart Henderson > wrote: > > Sorry for the delay, I slacked a bit. > > > Tests are a bit messed up though, some missing deps - py-coverage, > > py-pexpect, bash, tcsh we have - there's also devel/flake8 of which > > we only have a py2 version (and this was imported as just "flake8" > > rather than "py-flake8" so adding a py3 flavour is going to be more > > fiddly than usual). It also uses "wheel" which isn't in ports as far > > as I can see. > > I talked with flake8 maintainer about splitting it and he said that it > was useless. Users only need the 'binary' and the fact that > it is made from py2 or py3 changes nothing. Sadly he never had the time > to write it on ports@. > > His point sounds legit so I checked for argcomplete and it seems it's > not even used. I just added a patch to zap it from setup.py so we > can drop the need to split flake8. I see where he's coming from, but if you're just working on things written using the current version of python, it's annoying to have to install a legacy version just to run flake8 (and vice-versa, I suppose..) > And probably wheel isn't needed neither. ¯\_(ツ)_/¯ > > > The do-test override isn't needed in this version, the default test > > target gives much more information about failures. But it does need > > "python" to be in the path, it can't cope with "python2.7". > > > > Also I see it runs out of PTYs during tests unless I create more > > device nodes. > > > > So here's a better-but-not-perfect diff.. > > Updated diff, ok danj@ if someone wants to commit Looks good to me. Though I would also support switching flake8 to py3 and removing the patch.. > Index: Makefile > === > RCS file: /cvs/ports/devel/py-argcomplete/Makefile,v > retrieving revision 1.7 > diff -u -p -r1.7 Makefile > --- Makefile 3 Jan 2017 19:19:48 - 1.7 > +++ Makefile 8 Mar 2017 01:13:23 - > @@ -2,11 +2,10 @@ > > COMMENT =bash tab completion for argparse > > -MODPY_EGG_VERSION = 1.4.1 > +MODPY_EGG_VERSION = 1.8.2 > DISTNAME = argcomplete-${MODPY_EGG_VERSION} > PKGNAME =py-${DISTNAME} > CATEGORIES = devel > -REVISION = 0 > > # GPLv3+ > PERMIT_PACKAGE_CDROM=Yes > @@ -19,12 +18,20 @@ MODPY_SETUPTOOLS =Yes > FLAVORS =python3 > FLAVOR ?= > > +TEST_DEPENDS = devel/py-coverage${MODPY_FLAVOR} \ > + devel/py-pexpect${MODPY_FLAVOR} \ > + devel/py-wheel${MODPY_FLAVOR} \ > + shells/bash \ > + shells/tcsh > + > post-install: > for i in ${PREFIX}/bin/*; do \ > mv $${i} $${i}${MODPY_BIN_SUFFIX} ;\ > done > > -do-test: > - cd ${WRKSRC}/test && ${MODPY_BIN} test.py > +pre-test: > + @ln -fs ${MODPY_BIN} ${WRKDIR}/bin/python > + @[[ -r /dev/ptyz0 ]] || (printf "\\nWARNING: tests are likely to fail\ > + unless additional ptys are available.\\n\\n" >&2; sleep 2) > > .include > Index: distinfo > === > RCS file: /cvs/ports/devel/py-argcomplete/distinfo,v > retrieving revision 1.3 > diff -u -p -r1.3 distinfo > --- distinfo 28 Aug 2016 10:55:38 - 1.3 > +++ distinfo 8 Mar 2017 01:13:23 - > @@ -1,2 +1,2 @@ > -SHA256 (argcomplete-1.4.1.tar.gz) = > p7X6jRrLaeSbPGs/YiW8cJCS4Mf2ITEbrFB6T27+YJ0= > -SIZE (argcomplete-1.4.1.tar.gz) = 37219 > +SHA256 (argcomplete-1.8.2.tar.gz) = > eizN4HTrZjEjKy+R+NjVmF1b6/DjawkREwluOgOFVGs= > +SIZE (argcomplete-1.8.2.tar.gz) = 53626 > Index: patches/patch-setup_py > === > RCS file: patches/patch-setup_py > diff -N patches/patch-setup_py > --- /dev/null 1 Jan 1970 00:00:00 - > +++ patches/patch-setup_py8 Mar 2017 01:13:23 - > @@ -0,0 +1,12 @@ > +$OpenBSD$ > +--- setup.py.origTue Mar 7 19:55:42 2017 > setup.py Tue Mar 7 19:55:53 2017 > +@@ -4,7 +4,7 @@ import glob > + from setuptools import setup, find_packages > + > + install_requires = [] > +-tests_require = ["coverage", "flake8", "pexpect", "wheel"] > ++tests_require = ["coverage", "pexpect", "wheel"] > + > + try: > + import argparse # noqa > Index: pkg/PLIST > === > RCS file: /cvs/ports/devel/py-argcomplete/pkg/PLIST,v > retrieving revision 1.2 > diff -u -p -r1.2 PLIST > --- pkg/PLIST 28 Aug 2016 10:55:38 - 1.2 > +++ pkg/PLIST 8 Mar 2017 01:13:23 - > @@ -1,6 +1,7 @@ > @comment $OpenBSD: PLIST,v 1.2 2016/08/28 10:55:38 danj Exp $ > bin/activate-global-python-argcomplete${MODPY_BIN_SUFFIX} > bin/python-argcomplete-check-easy-install-script${MODPY_BIN_SUFFIX} > +bin/python-argcomplete-tcsh${MODPY_BIN_SUFFIX} > bin/register-python-argcomplete${MODPY_BIN_SUFFIX} > lib/python${MODPY_VERSION}/site-packages/argcomplete/ > > lib/python${MODPY_VERSION}/site-p
Re: Update to py-argcomplete-1.8.2
On Tue, Mar 07, 2017 at 08:15:07PM -0500, Daniel Jakots wrote: > On Tue, 21 Feb 2017 16:13:50 +, Stuart Henderson > wrote: > > Sorry for the delay, I slacked a bit. > > > Tests are a bit messed up though, some missing deps - py-coverage, > > py-pexpect, bash, tcsh we have - there's also devel/flake8 of which > > we only have a py2 version (and this was imported as just "flake8" > > rather than "py-flake8" so adding a py3 flavour is going to be more > > fiddly than usual). It also uses "wheel" which isn't in ports as far > > as I can see. > > I talked with flake8 maintainer about splitting it and he said that it > was useless. Users only need the 'binary' and the fact that > it is made from py2 or py3 changes nothing. Sadly he never had the time > to write it on ports@. > > His point sounds legit so I checked for argcomplete and it seems it's > not even used. I just added a patch to zap it from setup.py so we > can drop the need to split flake8. > > And probably wheel isn't needed neither. ??\_(???)_/?? > > > The do-test override isn't needed in this version, the default test > > target gives much more information about failures. But it does need > > "python" to be in the path, it can't cope with "python2.7". > > > > Also I see it runs out of PTYs during tests unless I create more > > device nodes. > > > > So here's a better-but-not-perfect diff.. > > Updated diff, ok danj@ if someone wants to commit Can one of these diffs go in? > > Index: Makefile > === > RCS file: /cvs/ports/devel/py-argcomplete/Makefile,v > retrieving revision 1.7 > diff -u -p -r1.7 Makefile > --- Makefile 3 Jan 2017 19:19:48 - 1.7 > +++ Makefile 8 Mar 2017 01:13:23 - > @@ -2,11 +2,10 @@ > > COMMENT =bash tab completion for argparse > > -MODPY_EGG_VERSION = 1.4.1 > +MODPY_EGG_VERSION = 1.8.2 > DISTNAME = argcomplete-${MODPY_EGG_VERSION} > PKGNAME =py-${DISTNAME} > CATEGORIES = devel > -REVISION = 0 > > # GPLv3+ > PERMIT_PACKAGE_CDROM=Yes > @@ -19,12 +18,20 @@ MODPY_SETUPTOOLS =Yes > FLAVORS =python3 > FLAVOR ?= > > +TEST_DEPENDS = devel/py-coverage${MODPY_FLAVOR} \ > + devel/py-pexpect${MODPY_FLAVOR} \ > + devel/py-wheel${MODPY_FLAVOR} \ > + shells/bash \ > + shells/tcsh > + > post-install: > for i in ${PREFIX}/bin/*; do \ > mv $${i} $${i}${MODPY_BIN_SUFFIX} ;\ > done > > -do-test: > - cd ${WRKSRC}/test && ${MODPY_BIN} test.py > +pre-test: > + @ln -fs ${MODPY_BIN} ${WRKDIR}/bin/python > + @[[ -r /dev/ptyz0 ]] || (printf "\\nWARNING: tests are likely to fail\ > + unless additional ptys are available.\\n\\n" >&2; sleep 2) > > .include > Index: distinfo > === > RCS file: /cvs/ports/devel/py-argcomplete/distinfo,v > retrieving revision 1.3 > diff -u -p -r1.3 distinfo > --- distinfo 28 Aug 2016 10:55:38 - 1.3 > +++ distinfo 8 Mar 2017 01:13:23 - > @@ -1,2 +1,2 @@ > -SHA256 (argcomplete-1.4.1.tar.gz) = > p7X6jRrLaeSbPGs/YiW8cJCS4Mf2ITEbrFB6T27+YJ0= > -SIZE (argcomplete-1.4.1.tar.gz) = 37219 > +SHA256 (argcomplete-1.8.2.tar.gz) = > eizN4HTrZjEjKy+R+NjVmF1b6/DjawkREwluOgOFVGs= > +SIZE (argcomplete-1.8.2.tar.gz) = 53626 > Index: patches/patch-setup_py > === > RCS file: patches/patch-setup_py > diff -N patches/patch-setup_py > --- /dev/null 1 Jan 1970 00:00:00 - > +++ patches/patch-setup_py8 Mar 2017 01:13:23 - > @@ -0,0 +1,12 @@ > +$OpenBSD$ > +--- setup.py.origTue Mar 7 19:55:42 2017 > setup.py Tue Mar 7 19:55:53 2017 > +@@ -4,7 +4,7 @@ import glob > + from setuptools import setup, find_packages > + > + install_requires = [] > +-tests_require = ["coverage", "flake8", "pexpect", "wheel"] > ++tests_require = ["coverage", "pexpect", "wheel"] > + > + try: > + import argparse # noqa > Index: pkg/PLIST > === > RCS file: /cvs/ports/devel/py-argcomplete/pkg/PLIST,v > retrieving revision 1.2 > diff -u -p -r1.2 PLIST > --- pkg/PLIST 28 Aug 2016 10:55:38 - 1.2 > +++ pkg/PLIST 8 Mar 2017 01:13:23 - > @@ -1,6 +1,7 @@ > @comment $OpenBSD: PLIST,v 1.2 2016/08/28 10:55:38 danj Exp $ > bin/activate-global-python-argcomplete${MODPY_BIN_SUFFIX} > bin/python-argcomplete-check-easy-install-script${MODPY_BIN_SUFFIX} > +bin/python-argcomplete-tcsh${MODPY_BIN_SUFFIX} > bin/register-python-argcomplete${MODPY_BIN_SUFFIX} > lib/python${MODPY_VERSION}/site-packages/argcomplete/ > > lib/python${MODPY_VERSION}/site-packages/argcomplete-${MODPY_EGG_VERSION}-py${MODPY_VERSION}.egg-info/ > @@ -8,6 +9,7 @@ lib/python${MODPY_VERSION}/site-packages > > lib/python${MODPY_VERSION}/site-packages/argcomplete-${MODPY_EGG_VERSION}-py${MODPY_VERSION}.egg-info/SOURCES.tx
e: LibreOffice Crash if started other than opening a document
> From: Ed Ahlsen-Girard > Date: 2017-01-12 12:27:55 > Message-ID: 20170112062755.15f180c5 () foop ! my ! domain > [Download message RAW] > > On Wed, 11 Jan 2017 12:51:12 +0100 > David Coppa wrote: > > > On Wed, Jan 11, 2017 at 11:55 AM, Robert Nagy > > wrote: [...] > > > > Or you can try with wmname. > > > > Install wmname with 'pkg_add wmname' > > > > And then put: > > > > /usr/local/bin/wmname LG3D > > > > into your ~/.xinitrc right before the line that exec your window > > manager... > > > > Ciao! > > David > > > > [...] > > [...] > > [...] > > [...] > > [...] > > [...] > > > > > > > > That did the trick. Thank you. > > -- > > Edward Ahlsen-Girard > Ft Walton Beach, FL Well, it doesn't do the trick anymore. Upstream I got a recommendation to install 32 bit Java. Even if I had a clear idea of how to do that, I'm not sure I'd want to. -- Edward Ahlsen-Girard Ft Walton Beach, FL
Re: [UPDATE] net/irssi 1.0.2
On 2017-05-20 22:15:27, viq wrote: > Long overdue and very simple patch bringing irssi to latest 1.0.2 > Lightly tested on amd64. > Works for me on amd64 -CURRENT. No regressions seen with any of the various scripts I use from the irssi website either. -- Bryan
Re: akpop3d crashed with core-dump
On Tue, May 23, 2017 at 09:23:12AM +0200, Giovanni Bechis wrote: > On 05/23/17 08:24, I. Schuchardt wrote: > > hi folks > > > > i testing akpop3d on openbsd 6.1 (amd64 and i386). > > > > with "akpop3d -d" starts the deamon und opening the port 110 on my machine. > > so i can check > > the pop3 with telnet on port 110: > > > fixed yesterday in -current, ok for the same diff for 6.1 ? sure, ok Thanks for taking care of this.
Re: akpop3d crashed with core-dump
On 05/23/17 08:24, I. Schuchardt wrote: > hi folks > > i testing akpop3d on openbsd 6.1 (amd64 and i386). > > with "akpop3d -d" starts the deamon und opening the port 110 on my machine. > so i can check > the pop3 with telnet on port 110: > fixed yesterday in -current, ok for the same diff for 6.1 ? Cheers Giovanni > ---snipp--- > ~% telnet my_machine 110 > Trying 1.2.3.4... > Connected to my_machine. > Escape character is '^]'. > +OK > user foo > +OK > pass bar > Connection closed by foreign host. > ---snipp--- > > > after this.. there is a core-dump-file "akpop3d.core" in "/" - and no other > logging-information. > > > thanks, > ingolf > Index: Makefile === RCS file: /var/cvs/ports/mail/akpop3d/Makefile,v retrieving revision 1.11 diff -u -p -r1.11 Makefile --- Makefile 10 Sep 2016 13:03:42 - 1.11 +++ Makefile 23 May 2017 07:22:13 - @@ -3,7 +3,7 @@ COMMENT= small and secure POP3 daemon DISTNAME= akpop3d-0.7.7 -REVISION = 2 +REVISION = 3 CATEGORIES= mail HOMEPAGE= http://www.synflood.at/akpop3d.html @@ -16,6 +16,7 @@ MASTER_SITES= http://www.synflood.at/akp EXTRACT_SUFX= .tar.bz2 CONFIGURE_STYLE=gnu +CONFIGURE_ARGS=--with-mysql=no NO_TEST= Yes Index: patches/patch-authenticate_c === RCS file: /var/cvs/ports/mail/akpop3d/patches/patch-authenticate_c,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 patch-authenticate_c --- patches/patch-authenticate_c 8 Nov 2004 21:57:05 - 1.1.1.1 +++ patches/patch-authenticate_c 23 May 2017 07:22:13 - @@ -1,6 +1,7 @@ $OpenBSD: patch-authenticate_c,v 1.1.1.1 2004/11/08 21:57:05 naddy Exp $ authenticate.c.orig Sun Aug 17 19:44:55 2003 -+++ authenticate.c Mon Nov 8 22:49:53 2004 +Index: authenticate.c +--- authenticate.c.orig authenticate.c @@ -25,6 +25,9 @@ #include "strlcpy.h" #include "mysql.h" @@ -11,7 +12,7 @@ $OpenBSD: patch-authenticate_c,v 1.1.1.1 extern const char * authfile; char real_username[MAXLINE+1]; char real_maildrop[MAXLINE+1]; -@@ -104,7 +107,7 @@ static int user_in_file(char * user, cha +@@ -104,7 +107,7 @@ static int user_in_file(char * user, char * path) { static int is_user_allowed(char * user) { int allow, deny; @@ -20,7 +21,7 @@ $OpenBSD: patch-authenticate_c,v 1.1.1.1 case 0: allow = 0; break; -@@ -115,7 +118,7 @@ static int is_user_allowed(char * user) +@@ -115,7 +118,7 @@ static int is_user_allowed(char * user) { return 0; } @@ -29,7 +30,7 @@ $OpenBSD: patch-authenticate_c,v 1.1.1.1 case 0: deny = 0; break; -@@ -322,7 +325,7 @@ int authenticate(char * username, char * +@@ -322,7 +325,7 @@ int authenticate(char * username, char * password) { memset(pass,0,sizeof(pass)); strlcpy(pass,password,len+1); @@ -38,3 +39,12 @@ $OpenBSD: patch-authenticate_c,v 1.1.1.1 if (0!=use_pop3_allow_deny && 0==is_user_allowed(user)) { return 0; } +@@ -335,7 +338,7 @@ int authenticate(char * username, char * password) { + } + + #ifndef HAVE_LIBMYSQLCLIENT +- u = getpwnam(user); ++ u = getpwnam_shadow(user); + #else + u = getMpwnam( user ); /* getMpwnam first checks getpwnam() */ + #endif /* HAVE_LIBMYSQLCLIENT */ Index: pkg/PLIST === RCS file: /var/cvs/ports/mail/akpop3d/pkg/PLIST,v retrieving revision 1.2 diff -u -p -r1.2 PLIST --- pkg/PLIST 14 Dec 2005 06:02:09 - 1.2 +++ pkg/PLIST 23 May 2017 07:22:13 - @@ -1,4 +1,4 @@ @comment $OpenBSD: PLIST,v 1.2 2005/12/14 06:02:09 jolan Exp $ @newgroup _akpop3d:555 @man man/man8/akpop3d.8 -sbin/akpop3d +@bin sbin/akpop3d