Re: NEW: net/irssi-otr
On 2014/05/21 02:15, Vladimir Seleznev wrote: > Hi ports@, > > There is OTR plugin port for IRSSI. > > Homepage: https://github.com/cryptodotis/irssi-otr > > Tested on i386, -current #92. > > Please, comment. > > -- >With best regards, >Vladimir Seleznev There have already been a couple of different irssi-otr ports submitted, this one fixes a few porting issues present in your version: https://github.com/jasperla/openbsd-wip/tree/master/security/irssi-otr though adding your README to it would be useful.
NEW: net/irssi-otr
Hi ports@, There is OTR plugin port for IRSSI. Homepage: https://github.com/cryptodotis/irssi-otr Tested on i386, -current #92. Please, comment. -- With best regards, Vladimir Seleznev irssi-otr.tar.gz Description: Binary data
[new] mapcache
Hi, here's a port of mapcache (http://mapserver.org/mapcache/) , a WMS/WMTS/TMS tile cache server companion to geo/mapserver. I've taken the choice to only build it as a fcgi/cgi server (ie without the apache2 dependency) but that can be discussed if needed. Special care has been taken to ensure it properly runs chrooted within /var/www, and as a fastcgi server via spawn-fcgi, so please comment especially on those parts (ie the rc.d script, the PLIST and the README) - and if you want to try it for real, follow the instructions and let me know if you have it running in the end. Works fine here with the provided default config. ok to import? Landry mapcache-1.2.1.tgz Description: application/tar-gz
Re: UPDATE: py-django
On Tue, 20 May 2014 22:27:26 +0200 frantisek holop wrote: > here is an update for 1.6.5 > i have added py-sphinx, as i think > the html docs are really warrented > for a huge framework like this. > > > passes most of the tests with python2.7. > > """ > Ran 5980 tests in 1361.618s > > OK (skipped=446, expected failures=11) > """ > > please test and commit. Hi, could you base your diff with mine I sent recently (to update to 1.6.4) to have this port working with python 3 too). Thanks, Remi.
firefox "save as" dialog
i have noticed this before but forgot to ask about it. i have both comix and firefox installed. whenever i select a .tgz or .tar.bz2 archive to download, the 'save as' dialog is _always_ trying to shove comix down my throat (open with), and even when i select 'save to disk', the 'Do this automatically for files like this from now on' is greyed out and i cannot figure out why. comix is not even anywhere to be found in preferences/applications. -f -- unix should be used as an adjective. - at&t
Re: [NEW] beets
this program was just put into my queue for porting a couple of weeks ago, you beet me to it :] and the author beet me to writing it, i have just starting working a tag/filename normalizer. this is how people get lazy! others doing the heavy lifting... commit commit commit plz! -f -- program: n. used to turn valid data into error messages
[wip] Firefox 30.0b5 (+gtk3 FLAVOR) & Thunderbird 30.0b1
Hi,
another month, another round of betas, this time for 30.0. Note that the
next-next one will be 31.0 (obvious!) but also an ESR release, and
thunderbird will also see a release, so please test it too.
This time, and as a tech preview, some plumbing will be added to
mozilla.port.mk to allow building firefox against a gtk3 FLAVOR. While
this is not considered upstream ready for public use, lacks some
features wrt plugins, and still has some graphical bugs (see
https://bugzilla.mozilla.org/show_bug.cgi?id=983843 at least) basic
usage works fine so far here.
As usual, grab the port from here:
http://rhaalovely.net/cgit/mozilla-firefox/?h=beta
git clone -b beta http://rhaalovely.net/git/mozilla-firefox
http://rhaalovely.net/cgit/mozilla-thunderbird/?h=beta
git clone -b beta http://rhaalovely.net/git/mozilla-thunderbird
(or git pull if you already have a clone)
- ppc is still broken as in 29, but a patch is in the works (see
https://bugzilla.mozilla.org/show_bug.cgi?id=961488).
- sparc64 build is broken but will be fixed in beta6 (see
https://bugzilla.mozilla.org/show_bug.cgi?id=1005449).
- i386 is untested as i still dont have a builder.
- amd64 Packages for ffx 30.0b5 are provided for both flavors:
$PKG_PATH=http://rhaalovely.net/stuff/amd64 sudo pkg_add firefox
Ambiguous: choose package for firefox
a 0:
1: firefox-30.0beta5
2: firefox-30.0beta5-gtk3
If you want to build the gtk3 flavor yourself, apply the attached
mozilla.port.mk diff, and build with FLAVOR=gtk3. If you try this and
encounter issues, get involved and file bugs UPSTREAM.
Landry
Index: mozilla.port.mk
===
RCS file: /cvs/ports/www/mozilla/mozilla.port.mk,v
retrieving revision 1.69
diff -u -r1.69 mozilla.port.mk
--- mozilla.port.mk 29 Apr 2014 21:16:26 - 1.69
+++ mozilla.port.mk 18 May 2014 20:38:24 -
@@ -38,8 +38,7 @@
archivers/unzip \
archivers/zip>=2.3
-MODMOZ_LIB_DEPENDS = x11/gtk+2 \
- textproc/hunspell \
+MODMOZ_LIB_DEPENDS = textproc/hunspell \
devel/nspr>=4.10.3 \
security/nss>=3.16
@@ -51,10 +50,9 @@
MODMOZ_BUILD_DEPENDS +=devel/yasm
.endif
-MODMOZ_WANTLIB += X11 Xcomposite Xcursor Xdamage Xext Xfixes Xi \
- Xinerama Xrandr Xrender Xt atk-1.0 c cairo \
- fontconfig freetype gdk-x11-2.0 gdk_pixbuf-2.0 gio-2.0 glib-2.0
\
- gobject-2.0 gthread-2.0 gtk-x11-2.0 m \
+MODMOZ_WANTLIB += X11 Xext Xrender Xt atk-1.0 c cairo \
+ fontconfig freetype gdk_pixbuf-2.0 gio-2.0 glib-2.0 \
+ gobject-2.0 gthread-2.0 m \
nspr4 nss3 pango-1.0 pangocairo-1.0 pangoft2-1.0 \
plc4 plds4 pthread event kvm sqlite3>=27 \
smime3 sndio nssutil3 ssl3 stdc++ z hunspell-1.3
@@ -100,6 +98,9 @@
--disable-dbus
FLAVORS += debug
+.if ${PKGPATH} == "www/mozilla-firefox"
+FLAVORS += gtk3
+.endif
FLAVOR ?=
.if ${FLAVOR:Mdebug}
@@ -108,6 +109,18 @@
--enable-debug-symbols=yes \
--disable-install-strip
INSTALL_STRIP =
+.endif
+
+.if ${FLAVOR:Mgtk3}
+# https://bugzilla.mozilla.org/show_bug.cgi?id=983843
+CONFIGURE_ARGS += --with-system-cairo
+CONFIGURE_ARGS += --enable-default-toolkit=cairo-gtk3
+MODMOZ_LIB_DEPENDS += x11/gtk+3
+MODMOZ_WANTLIB += cairo-gobject gdk-3 gtk-3 pixman-1 pthread-stubs
+.else
+MODMOZ_LIB_DEPENDS += x11/gtk+2
+MODMOZ_WANTLIB += Xcomposite Xcursor Xdamage Xfixes Xi Xinerama \
+ Xrandr gdk-x11-2.0 gtk-x11-2.0
.endif
# from browser/config/mozconfig
UPDATE: py-jinja2
a straightforward update.
please test and commit.
-f
--
no sense being pessimistic. it wouldn't work anyway.
Index: Makefile
===
RCS file: /cvs/ports/www/py-jinja2/Makefile,v
retrieving revision 1.17
diff -u -p -r1.17 Makefile
--- Makefile28 Sep 2013 17:52:14 - 1.17
+++ Makefile20 May 2014 19:06:07 -
@@ -2,13 +2,13 @@
COMMENT = fast, optionally sandboxed, Python template engine
-MODPY_EGG_VERSION =2.7.1
+MODPY_EGG_VERSION =2.7.2
DISTNAME = Jinja2-${MODPY_EGG_VERSION}
PKGNAME = py-${DISTNAME:L}
CATEGORIES = www devel
-HOMEPAGE = http://jinja.pocoo.org/2/
+HOMEPAGE = http://jinja.pocoo.org/
MAINTAINER = frantisek holop
Index: distinfo
===
RCS file: /cvs/ports/www/py-jinja2/distinfo,v
retrieving revision 1.8
diff -u -p -r1.8 distinfo
--- distinfo28 Sep 2013 17:52:14 - 1.8
+++ distinfo20 May 2014 19:06:07 -
@@ -1,2 +1,2 @@
-SHA256 (Jinja2-2.7.1.tar.gz) = XMCgh6gdyhwINoSC+3qS/ivdjPuyK8D8z+bIWv+wTIs=
-SIZE (Jinja2-2.7.1.tar.gz) = 377813
+SHA256 (Jinja2-2.7.2.tar.gz) = MQo1+8ysOvE+v5Jyl/hxrGVrnaHSSLH+Z2Wv+nG1MjU=
+SIZE (Jinja2-2.7.2.tar.gz) = 378300
Re: ACPI Component Architecture
previously on this list People contributed: > I've updated the acpica port in -current but it won't help you with this. > On OpenBSD acpica is *only* used as a development tool > (disassembler/debugger). > we do not use Intel's acpi code in the kernel, we have our own implementation > - anything you read about doing this on FreeBSD is not directly relevant. > > For what you're looking for, I think you would need a kernel driver, > and I think it's going to be fairly machine-specific. > > I think that in OpenBSD there is currently no way to disable particular > > device, > > but as I understand PCI-Express has power-management capabilities and > > can turn off device. I guess you have checked for any bios settings and disabling the particular PCI Express port/device via boot -c won't work due to the single multiplexer if it would have worked anyway? What does the dmesg show out of interest? -- ___ 'Write programs that do one thing and do it well. Write programs to work together. Write programs to handle text streams, because that is a universal interface' (Doug McIlroy) In Other Words - Don't design like polkit or systemd ___ ___
Re: ACPI Component Architecture
On 2014/05/20 18:41, Lampshade wrote: > I have laptop with Optimus technology so it is without dedicated multiplexer. > Intel > GPU is proxy for Nvidia GPU. First Linux I installed on this laptop was Ubuntu > 12.10 and by default it disabled Nvidia GPU. It is good behaviour, under > Linux and > OpenBSD I want Nvidia GPU to be disabled. Nvidia GPU is useful for me in > Windows (example Autodesk's Inventor) so I don't want to change laptop. When I > installed OpenBSD 5.5-current Nvidia GPU is enabled and is consuming a lot of > power and heating my laptop to 64 Celsius degrees (°C) in idle when CPU was > underclocked to 800 Mhz... For two weeks I was searching for solution but I > haven't > finded it (I am not familiar with *BSD systems) so I decided to remove OpenBSD > from laptop :/ But I had searching for solution further and in FreeBSD forum > (in 9.2 they have version from 2013) I finded that it is a way to disable > Nvidia GPU > via acpica. Unfortunately on Liux forum there was an user who claimed that in > his > VPS acpica in OpenBSD 5.4 is old and can't send any signals to some devices. > Should I > question him about details? > > I am a regular user who just want some security in Internet, I don't how > programmers > are dealing with hardware, I can just write very, very simple program in C > and this is > all my hacking skill :/ I've updated the acpica port in -current but it won't help you with this. On OpenBSD acpica is *only* used as a development tool (disassembler/debugger). we do not use Intel's acpi code in the kernel, we have our own implementation - anything you read about doing this on FreeBSD is not directly relevant. For what you're looking for, I think you would need a kernel driver, and I think it's going to be fairly machine-specific. There's a general-purpose linux driver that allows calling specific acpi methods from userland via a /proc interface, plus some sample scripts to disable second gpu on certain known laptops (https://github.com/mkottman/acpi_call) and there's something similar for FreeBSD, but nothing directly suitable for porting to OpenBSD.
Re: ACPI Component Architecture
I have laptop with Optimus technology so it is without dedicated multiplexer. Intel GPU is proxy for Nvidia GPU. First Linux I installed on this laptop was Ubuntu 12.10 and by default it disabled Nvidia GPU. It is good behaviour, under Linux and OpenBSD I want Nvidia GPU to be disabled. Nvidia GPU is useful for me in Windows (example Autodesk's Inventor) so I don't want to change laptop. When I installed OpenBSD 5.5-current Nvidia GPU is enabled and is consuming a lot of power and heating my laptop to 64 Celsius degrees (°C) in idle when CPU was underclocked to 800 Mhz... For two weeks I was searching for solution but I haven't finded it (I am not familiar with *BSD systems) so I decided to remove OpenBSD from laptop :/ But I had searching for solution further and in FreeBSD forum (in 9.2 they have version from 2013) I finded that it is a way to disable Nvidia GPU via acpica. Unfortunately on Liux forum there was an user who claimed that in his VPS acpica in OpenBSD 5.4 is old and can't send any signals to some devices. Should I question him about details? I am a regular user who just want some security in Internet, I don't how programmers are dealing with hardware, I can just write very, very simple program in C and this is all my hacking skill :/ Od: "Daniel Dickman" Do: "Lampshade" ; Wysłane: 17:24 Wtorek 2014-05-20 Temat: Re: ACPI Component Architecture > > On May 20, 2014, at 11:08 AM, Lampshade wrote: > > > > > > Hello > > > > Is there any chance to include to current and in future to 5.6 newer > > version of acpica > > package? Currently in ports there is acpica-2023p0.tgz but ACPI > > Component > > Architecture is under active development and current release is 20140424. > > well someone has to do the work... > > in this case I've gotten it up to 2012-07-11 which works for me on i386. > patch is here if someone wants to commit it or give me an ok. > http://www.dickman.org/openbsd/patches/devel_acpica.diff > > > > I think it > > would be useful for lot of modern laptop users to disable Nvidia GPU and > > other > > interfaces. > > > > can you explain this a bit more? I'm not sure I'm fully understanding this > bit. > >
Re: opendnssec and softhsm revisited
On Mon, May 19, 2014 at 11:55:33PM +0200, Jérémie Courrèges-Anglas wrote: > > It is not about this particular occurrence. If you replace it by > a strlcpy call, you will probably see the same warning, coming from > another strcpy call elsewhere. > > Propose upstream strl* and snprintf as alternatives to the unchecked > functions. If upstream sees interest in this, then a careful audit of > the code can follow, and strl* functions make error checking easier. > Adding fallback code for OSes not providing those functions is easy too. > Upon mentioning these findings I was informed of the situation where an ongoing rewrite of the code for 2.0 will make sure strcpy/strcat is not used, so this will fix itself in time. It is available here: http://lists.opendnssec.org/pipermail/opendnssec-user/2014-May/002941.html > > > > The call in signer/src/wire/notify.c: > > random()%(extra-base); > > > > I guess #ifdef'ing this with HAVE_ARC4RANDOM like is done in ksm_policy.c > > would be nice. > > See arc4random_uniform(3), the configure seems to test for its presence. > The current random-related fixes on the upstream openbsd branch actually tests for both now. Thanks for the pointer! > > > Finally some unused variables, might as well be complete while I am at > > it: > > > > === > > daemon/cmdhandler.c:398: warning: unused variable 'task' > > > > shared/duration.c:442: warning: 'is_leap_year' defined but not used > > shared/duration.c:449: warning: 'leap_days' defined but not used > > > > shared/hsm.c:224: warning: unused variable 'retries' > > shared/hsm.c:220: warning: unused variable 'status' > > === > > Those could be harmless, or they could show a bug in the program logic. > If you want to audit them fine, but in the end it's rather upstream's > responsibility. > These was mostly intended to be seen by upstream, and some of them has been fixed on the branch. > We try to avoid having too much patches in the ports tree. The best way > is probably to work with upstream and get your patches integrated, after > proper reviewing. Some mistakes have been done while wrongly replacing > eg. strcpy by strlcpy. > I intend to ignore the strcat/strcpy stuff for now given the above status. The other stuff has been fixed. Thanks a lot for your input :). Regards, Patrik Lundin
Re: ACPI Component Architecture
On 2014/05/20 11:23, Daniel Dickman wrote: > > On May 20, 2014, at 11:08 AM, Lampshade wrote: > > > > > > Hello > > > > Is there any chance to include to current and in future to 5.6 newer > > version of acpica > > package? Currently in ports there is acpica-2023p0.tgz but ACPI > > Component > > Architecture is under active development and current release is 20140424. I tried updating to the latest version about a year ago, but decided it was too much of a pain to be worth spending any more time on (we basically just use this port for the disassembler). Note that current versions require a newer version of bison; newer bison doesn't work with OpenBSD's m4. > well someone has to do the work... > > in this case I've gotten it up to 2012-07-11 which works for me on i386. > patch is here if someone wants to commit it or give me an ok. > http://www.dickman.org/openbsd/patches/devel_acpica.diff patch: malformed patch at line 31: @@ -24,20 +23,19 @@ NO_TEST = Yes > > I think it > > would be useful for lot of modern laptop users to disable Nvidia GPU and > > other > > interfaces. > > > can you explain this a bit more? I'm not sure I'm fully understanding this > bit. > I'm guessing it's about decompiling/modifying/recompiling tables; but this won't help - OpenBSD only uses aml from the tables in the bios, it does not support reading custom aml from disk like some OS do.
Re: ACPI Component Architecture
> On May 20, 2014, at 11:08 AM, Lampshade wrote: > > > Hello > > Is there any chance to include to current and in future to 5.6 newer version > of acpica > package? Currently in ports there is acpica-2023p0.tgz but ACPI Component > Architecture is under active development and current release is 20140424. well someone has to do the work... in this case I've gotten it up to 2012-07-11 which works for me on i386. patch is here if someone wants to commit it or give me an ok. http://www.dickman.org/openbsd/patches/devel_acpica.diff > I think it > would be useful for lot of modern laptop users to disable Nvidia GPU and other > interfaces. > can you explain this a bit more? I'm not sure I'm fully understanding this bit.
ACPI Component Architecture
Hello Is there any chance to include to current and in future to 5.6 newer version of acpica package? Currently in ports there is acpica-2023p0.tgz but ACPI Component Architecture is under active development and current release is 20140424. I think it would be useful for lot of modern laptop users to disable Nvidia GPU and other interfaces.
Re: [UPDATE] yle-dl 2.2.1
On Tue, 20 May 2014 12:52:07 +0100 Stuart Henderson wrote: > Antti, OK? Sure. Thanks both! -- Antti Harri
Re: Reordering objects in linking command line for Samba4
On Tue, May 20, 2014 at 03:01:58PM +0400, Vadim Zhukov wrote:
> 2014-05-20 14:39 GMT+04:00, Stuart Henderson :
> > On 2014/05/19 17:06, Marc Espie wrote:
> >> > --- /dev/null1 Jan 1970 00:00:00 -
> >> > +++ patches/patch-buildtools_wafadmin_Tools_cxx_py 19 May 2014
> >> > 14:45:35
> >> > -
> >> > @@ -0,0 +1,14 @@
> >> > +$OpenBSD$
> >> > +Reorder object files to minimize library interrefs count.
> >> > +Speeds up loading of library files.
> >> > +--- buildtools/wafadmin/Tools/cxx.py.orig Mon May 19 18:30:35 2014
> >> > buildtools/wafadmin/Tools/cxx.pyMon May 19 18:31:28 2014
> >> > +@@ -97,7 +97,7 @@ cls = Task.simple_task_type('cxx', cxx_str,
> >> > color='GRE
> >> > + cls.scan = ccroot.scan
> >> > + cls.vars.append('CXXDEPS')
> >> > +
> >> > +-link_str = '${LINK_CXX} ${CXXLNK_SRC_F}${SRC}
> >> > ${CXXLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
> >> > ++link_str = '${LINK_CXX} ${CXXLNK_SRC_F} `set -- ${SRC}; test $# -gt 0
> >> > && lorder "$@" | tsort -q` ${CXXLNK_TGT_F}${TGT[0].abspath(env)}
> >> > ${LINKFLAGS}'
> >> > + cls = Task.simple_task_type('cxx_link', link_str, color='YELLOW',
> >> > ext_in='.o', ext_out='.bin', shell=False)
> >> > + cls.maxjobs = 1
> >> > + cls.install = Utils.nada
> >> This is definitely a good idea, this is what the base system does for
> >> linking
> >> libraries !
> >>
> >
> > Does src/usr.bin/libtool do the same?
>
> No. But it could; what do you think, Marc?
looks like a good idea. Of course, it's very likely to find other bugs,
so it should be exhaustively tested...
Re: [UPDATE] yle-dl 2.2.1
On 2014/05/20 14:40, vesa norrman wrote:
> Update for yle-dl. Old version is broken.
>
> - Vesa
tweaked: remove REVISION, use MODPY_ADJ_FILES rather than patching.
works for me with a random choice from areena.yle.fi.
Antti, OK?
Index: Makefile
===
RCS file: /cvs/ports/multimedia/yle-dl/Makefile,v
retrieving revision 1.5
diff -u -p -r1.5 Makefile
--- Makefile22 Apr 2014 08:54:57 - 1.5
+++ Makefile20 May 2014 11:50:57 -
@@ -1,44 +1,36 @@
# $OpenBSD: Makefile,v 1.5 2014/04/22 08:54:57 jca Exp $
-V =1.4.6
-REVISION = 0
+V =2.2.1
COMMENT = download videos from YLE Areena
-DISTNAME = rtmpdump-yle-$V
+DISTNAME = aajanki-yle-$V
PKGNAME = yle-dl-$V
CATEGORIES = multimedia
-HOMEPAGE = http://users.tkk.fi/~aajanki/rtmpdump-yle/
+HOMEPAGE = http://aajanki.github.io/yle-dl/
MAINTAINER = Antti Harri
# GPLv2
PERMIT_PACKAGE_CDROM = Yes
-WANTLIB += c crypto json-c ssl z
+MASTER_SITES = https://github.com/aajanki/yle-dl/tarball/${V}/
-MASTER_SITES = ${HOMEPAGE}
+MODULES = lang/python
+RUN_DEPENDS = net/rtmpdump \
+ security/py-crypto
-MODULES = converters/libiconv \
- lang/python
-LIB_DEPENDS = devel/json-c
-
-MAKE_FLAGS = OPT="" CC="${CC}"
-MAKE_ENV = XCFLAGS="${CFLAGS}" XLDFLAGS="${LDFLAGS}"
-
-USE_GMAKE =Yes
NO_TEST = Yes
-CFLAGS += `pkg-config --cflags json-c` -I${PREFIX}/include
-LDFLAGS += `pkg-config --libs json-c` -liconv
+WRKDIST = ${WRKDIR}/aajanki-yle-dl-b2b1869
+
+MODPY_ADJ_FILES = yle-dl
do-build:
cd ${WRKSRC} && env -i ${MAKE_ENV} ${MAKE_PROGRAM} ${MAKE_FLAGS} \
-f ${MAKE_FILE} ${ALL_TARGET}
do-install:
- ${SUBST_CMD} ${WRKSRC}/yle-dl.py
- ${INSTALL_SCRIPT} ${WRKSRC}/yle-dl.py ${PREFIX}/bin/yle-dl
- ${INSTALL_PROGRAM} ${WRKSRC}/rtmpdump-yle ${PREFIX}/bin/
+ ${INSTALL_SCRIPT} ${WRKSRC}/yle-dl ${PREFIX}/bin/yle-dl
${INSTALL_DATA_DIR} ${PREFIX}/share/doc/yle-dl
${INSTALL_DATA} ${WRKSRC}/COPYING ${WRKSRC}/README.fi \
${PREFIX}/share/doc/yle-dl
Index: distinfo
===
RCS file: /cvs/ports/multimedia/yle-dl/distinfo,v
retrieving revision 1.3
diff -u -p -r1.3 distinfo
--- distinfo22 Apr 2014 08:54:57 - 1.3
+++ distinfo20 May 2014 11:50:57 -
@@ -1,2 +1,2 @@
-SHA256 (rtmpdump-yle-1.4.6.tar.gz) =
BVlsnBGf2u7y8JuCxPCIK4HJhyb0p7twoGFWuQCLVnY=
-SIZE (rtmpdump-yle-1.4.6.tar.gz) = 167441
+SHA256 (aajanki-yle-2.2.1.tar.gz) =
/zERphVHkPjL/noVvFFohzdWtck3FYax+ob09suaRds=
+SIZE (aajanki-yle-2.2.1.tar.gz) = 25557
Index: patches/patch-Makefile
===
RCS file: patches/patch-Makefile
diff -N patches/patch-Makefile
--- patches/patch-Makefile 22 Apr 2014 08:54:57 - 1.1
+++ /dev/null 1 Jan 1970 00:00:00 -
@@ -1,12 +0,0 @@
-$OpenBSD: patch-Makefile,v 1.1 2014/04/22 08:54:57 jca Exp $
Makefile.orig Mon Apr 21 18:28:44 2014
-+++ Makefile Mon Apr 21 18:29:10 2014
-@@ -22,7 +22,7 @@ CRYPTO_DEF=$(DEF_$(CRYPTO))
-
- STATIC_JSON=
- #STATIC_JSON=/path/to/json-c
--JSON_LIB=$(if $(STATIC_JSON),$(STATIC_JSON)/.libs/libjson.a,-ljson)
-+JSON_LIB=$(if $(STATIC_JSON),$(STATIC_JSON)/.libs/libjson-c.a,-ljson-c)
-
- DEF=-DRTMPDUMP_VERSION=\"$(VERSION)\"
-DYLERTMPDUMP_VERSION=\"$(VERSION_YLE)\" $(CRYPTO_DEF) $(XDEF)
- OPT=-O2
Index: patches/patch-librtmp_Makefile
===
RCS file: patches/patch-librtmp_Makefile
diff -N patches/patch-librtmp_Makefile
--- patches/patch-librtmp_Makefile 22 Apr 2014 08:54:57 - 1.1
+++ /dev/null 1 Jan 1970 00:00:00 -
@@ -1,12 +0,0 @@
-$OpenBSD: patch-librtmp_Makefile,v 1.1 2014/04/22 08:54:57 jca Exp $
librtmp/Makefile.orig Mon Apr 21 18:28:22 2014
-+++ librtmp/Makefile Mon Apr 21 18:28:39 2014
-@@ -64,7 +64,7 @@ INSTALL_IMPLIB=$(INSTALL_IMPLIB_$(SYS))
-
- STATIC_JSON=
- JSON_DEF=$(if $(STATIC_JSON),-I$(STATIC_JSON),-I/usr/include/json)
--LIB_JSON=$(if $(STATIC_JSON),,-ljson)
-+LIB_JSON=$(if $(STATIC_JSON),,-ljson-c)
- REQ_JSON=json
-
- SHARED=
Index: patches/patch-yle-dl_py
===
RCS file: patches/patch-yle-dl_py
diff -N patches/patch-yle-dl_py
--- patches/patch-yle-dl_py 13 Dec 2011 13:10:47 - 1.2
+++ /dev/null 1 Jan 1970 00:00:00 -
@@ -1,20 +0,0 @@
-$OpenBSD: patch-yle-dl_py,v 1.2 2011/12/13 13:10:47 sthen Exp $
yle-dl.py.orig Tue Nov 29 21:41:58 2011
-+++ yle-dl.py Mon Dec 12 18:52:45 2011
-@@ -1,4 +1,4 @@
--#!/usr/bin/python
-+#!$
[UPDATE] yle-dl 2.2.1
Update for yle-dl. Old version is broken. - Vesa yle-dl-146to221 Description: Binary data
Re: Reordering objects in linking command line for Samba4
2014-05-20 14:39 GMT+04:00, Stuart Henderson :
> On 2014/05/19 17:06, Marc Espie wrote:
>> > --- /dev/null 1 Jan 1970 00:00:00 -
>> > +++ patches/patch-buildtools_wafadmin_Tools_cxx_py 19 May 2014 14:45:35
>> > -
>> > @@ -0,0 +1,14 @@
>> > +$OpenBSD$
>> > +Reorder object files to minimize library interrefs count.
>> > +Speeds up loading of library files.
>> > +--- buildtools/wafadmin/Tools/cxx.py.orig Mon May 19 18:30:35 2014
>> > buildtools/wafadmin/Tools/cxx.py Mon May 19 18:31:28 2014
>> > +@@ -97,7 +97,7 @@ cls = Task.simple_task_type('cxx', cxx_str,
>> > color='GRE
>> > + cls.scan = ccroot.scan
>> > + cls.vars.append('CXXDEPS')
>> > +
>> > +-link_str = '${LINK_CXX} ${CXXLNK_SRC_F}${SRC}
>> > ${CXXLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
>> > ++link_str = '${LINK_CXX} ${CXXLNK_SRC_F} `set -- ${SRC}; test $# -gt 0
>> > && lorder "$@" | tsort -q` ${CXXLNK_TGT_F}${TGT[0].abspath(env)}
>> > ${LINKFLAGS}'
>> > + cls = Task.simple_task_type('cxx_link', link_str, color='YELLOW',
>> > ext_in='.o', ext_out='.bin', shell=False)
>> > + cls.maxjobs = 1
>> > + cls.install = Utils.nada
>> This is definitely a good idea, this is what the base system does for
>> linking
>> libraries !
>>
>
> Does src/usr.bin/libtool do the same?
No. But it could; what do you think, Marc?
--
WBR,
Vadim Zhukov
Re: KiCAD update?
On 2014/05/20 07:42, Daniel Bolgheroni wrote: > Hi ports@, > > Anyone with a wip diff to update KiCAD? If not, as time permits, I'm > willing to put some effort on it. I've not heard of any.. > Anything I must be aware of? Yes, see the warning in the log message of the last commit: downloading during build is not permitted. - PatchSet 24 Date: 2013/12/12 20:21:45 Author: landry Branch: HEAD Tag: OPENBSD_5_5_BASE Log: Add boost to BDEP, kicad supposedly bundles the parts of boost it needs in the version we have in-tree, but experience has shown that if systemwide boost was installed at some point during build then junked by dpb build was failing. Boo. Note that in newer releases of kicad those bundled boost headers were removed, and now cmake downloads boost tarball during build and patch its headers on the fly. Yay. https://lists.launchpad.net/kicad-developers/msg10503.html Members: Makefile:1.19->1.20
KiCAD update?
Hi ports@, Anyone with a wip diff to update KiCAD? If not, as time permits, I'm willing to put some effort on it. Anything I must be aware of? Thank you.
Re: Reordering objects in linking command line for Samba4
On 2014/05/19 17:06, Marc Espie wrote:
> > --- /dev/null 1 Jan 1970 00:00:00 -
> > +++ patches/patch-buildtools_wafadmin_Tools_cxx_py 19 May 2014 14:45:35
> > -
> > @@ -0,0 +1,14 @@
> > +$OpenBSD$
> > +Reorder object files to minimize library interrefs count.
> > +Speeds up loading of library files.
> > +--- buildtools/wafadmin/Tools/cxx.py.orig Mon May 19 18:30:35 2014
> > buildtools/wafadmin/Tools/cxx.py Mon May 19 18:31:28 2014
> > +@@ -97,7 +97,7 @@ cls = Task.simple_task_type('cxx', cxx_str, color='GRE
> > + cls.scan = ccroot.scan
> > + cls.vars.append('CXXDEPS')
> > +
> > +-link_str = '${LINK_CXX} ${CXXLNK_SRC_F}${SRC}
> > ${CXXLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
> > ++link_str = '${LINK_CXX} ${CXXLNK_SRC_F} `set -- ${SRC}; test $# -gt 0 &&
> > lorder "$@" | tsort -q` ${CXXLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
> > + cls = Task.simple_task_type('cxx_link', link_str, color='YELLOW',
> > ext_in='.o', ext_out='.bin', shell=False)
> > + cls.maxjobs = 1
> > + cls.install = Utils.nada
> This is definitely a good idea, this is what the base system does for linking
> libraries !
>
Does src/usr.bin/libtool do the same?
Re: Reordering objects in linking command line for Samba4
On Tue, May 20th, 2014 at 1:06 AM, Marc Espie wrote:
> On Mon, May 19, 2014 at 06:56:48PM +0400, Vadim Zhukov wrote:
> > This patch improves loading time of Samba 4 binaries a bit. This is
> > a rude hack that involves lorder(1) to change order of object files
> > being linked. Unfortunately, it cannot be used to reorder libraries,
> > i.e., the "-lfoo" items.
> >
> > Bad thing is that we inject shell code into command line, which
> > forces us to switch shell=False to shell=True.
> >
> > I have at least samba-4.0.17 compiled with this patch, and results are
> > promising a bit: ~15 secs instead of 30-60 secs previously.
> >
> > If this is considered a good idea, I'll work on reordering of -lfoo.
> > --
> > WBR,
> > Vadim Zhukov
> >
> >
> > Index: patches/patch-buildtools_wafadmin_Tools_cc_py
> >
=
==
> > RCS file: patches/patch-buildtools_wafadmin_Tools_cc_py
> > diff -N patches/patch-buildtools_wafadmin_Tools_cc_py
> > --- /dev/null 1 Jan 1970 00:00:00 -
> > +++ patches/patch-buildtools_wafadmin_Tools_cc_py 19 May 2014 14:45:35
> -
> > @@ -0,0 +1,16 @@
> > +$OpenBSD$
> > +Reorder object files to minimize library interrefs count.
> > +Speeds up loading of library files.
> > +--- buildtools/wafadmin/Tools/cc.py.orig Mon May 19 17:29:02 2014
> > buildtools/wafadmin/Tools/cc.pyMon May 19 18:05:26 2014
> > +@@ -93,8 +93,8 @@ cls = Task.simple_task_type('cc', cc_str, 'GREEN',
> ext
> > + cls.scan = ccroot.scan
> > + cls.vars.append('CCDEPS')
> > +
> > +-link_str = '${LINK_CC} ${CCLNK_SRC_F}${SRC}
> ${CCLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
> > +-cls = Task.simple_task_type('cc_link', link_str, color='YELLOW',
> ext_in='.o', ext_out='.bin', shell=False)
> > ++link_str = '${LINK_CC} ${CCLNK_SRC_F} `set -- ${SRC}; test $# -gt 0 &&
> lorder "$@" | tsort -q` ${CCLNK_TGT_F}${TGT[0].abspath(env)}
> ${LINKFLAGS}'
> > ++cls = Task.simple_task_type('cc_link', link_str, color='YELLOW',
> ext_in='.o', ext_out='.bin', shell=True)
> > + cls.maxjobs = 1
> > + cls.install = Utils.nada
> > +
> > Index: patches/patch-buildtools_wafadmin_Tools_cxx_py
> >
=
==
> > RCS file: patches/patch-buildtools_wafadmin_Tools_cxx_py
> > diff -N patches/patch-buildtools_wafadmin_Tools_cxx_py
> > --- /dev/null 1 Jan 1970 00:00:00 -
> > +++ patches/patch-buildtools_wafadmin_Tools_cxx_py 19 May 2014 14:45:35
> -
> > @@ -0,0 +1,14 @@
> > +$OpenBSD$
> > +Reorder object files to minimize library interrefs count.
> > +Speeds up loading of library files.
> > +--- buildtools/wafadmin/Tools/cxx.py.orig Mon May 19 18:30:35 2014
> > buildtools/wafadmin/Tools/cxx.py Mon May 19 18:31:28 2014
> > +@@ -97,7 +97,7 @@ cls = Task.simple_task_type('cxx', cxx_str,
> color='GRE
> > + cls.scan = ccroot.scan
> > + cls.vars.append('CXXDEPS')
> > +
> > +-link_str = '${LINK_CXX} ${CXXLNK_SRC_F}${SRC}
> ${CXXLNK_TGT_F}${TGT[0].abspath(env)} ${LINKFLAGS}'
> > ++link_str = '${LINK_CXX} ${CXXLNK_SRC_F} `set -- ${SRC}; test $# -gt 0
> && lorder "$@" | tsort -q` ${CXXLNK_TGT_F}${TGT[0].abspath(env)}
> ${LINKFLAGS}'
> > + cls = Task.simple_task_type('cxx_link', link_str, color='YELLOW',
> ext_in='.o', ext_out='.bin', shell=False)
> > + cls.maxjobs = 1
> > + cls.install = Utils.nada
> This is definitely a good idea, this is what the base system does for
> linking
> libraries !
>
>
Sounds good to me. The whole waf build thingy i'm still coming to grips with so
if espie@ likes it then
run with it.
Ian McWilliam
