Sunday, November 22, 2026

Re: wayland/hyprland: new tiling WM

On Sun, 22 Nov 2026 14:38:46 +0100, Nazarenko Mykyta <mn@superpc.suicide.solutions> wrote: > > [1 <text/plain; UTF-8 (quoted-printable)>] > 21 nov 2026 20:44:12 Klemens Nanni <kn@openbsd.org>: > > > 21.11.2026 22:30, Nazarenko Mykyta пишет: > >> Hola ports@ , kirill@, kn@. > >> > >> I'd like to suggest adding Hyprland, a beautiful and highly customizable tiling > >> window manager. > >> > >> Any feedback is very important. > > > > I have no clue about wayland and can't test such ports. > > > >> > >> You can find all tarballs here : https://nazarenko.sbs/ports > > > > All as in more than one for one new port? > > > > In any case, you should explain what was ported, how to test, > > what problems you may have had, where to pay attention, etc. > > > > The less information you provide in mails like these, > > the more porters have to find out themselves, which tends to > > simply leave such requests unanswered. > > > > Please attach tarballs and inline diffs, as usual. > > > >> > >> With best regards, > >> Nazarenko Mykyta > > Yeah, I should have explained a little better what to do with all this. > > What was ported ( including deps ) : > - wayland/aquamarine > - wayland/hyprcursor > - wayland/hyprgraphics > - wayland/hyprland-protocols > - wayland/hyprutils > - wayland/hyprwayland-scanner > - wayland/hyprwire > - wayland/hyprland > > and textproc/glaze : > https://marc.info/?l=openbsd-ports&m=179370723776203&w=2 > > This version was tested on amdgpu and i915, and everything works > flawlessly. However, this does not rule out the possibility that > problems may arise with a particular hardware configuration. > > So, it would be great if those who are interested could test this > on different hardware configurations to make sure there aren't > any issues. For example: problems with suspend, crashes under > heavy CPU load, freezes, etc. > > In the previous thread, we agreed that we should only consider > importing Hyprland once hyprlang has been fully removed from it. > In commit 6170b35b326ff133da74a0ece6ef214098326bcb - it was finally > removed. That is exactly why this hash is hardcoded in GH_COMMIT. > After the next release, we'll be able to put a stable version tag. > I'm not wayland user nor I do have setup to test, the best that I can do it builds your ports in clean chroot environment. -- wbr, Kirill

[update] devel/snare

A simple update for devel/snare to 0.4.14. Mostly minor bug fixes. Laurie diff --git devel/snare/Makefile devel/snare/Makefile index 4cdf5e8d033..d79cf9e0211 100644 --- devel/snare/Makefile +++ devel/snare/Makefile @@ -1,6 +1,6 @@ COMMENT = GitHub webhooks runner daemon -V = 0.4.13 +V = 0.4.14 DISTNAME = snare-$V CATEGORIES = devel diff --git devel/snare/distinfo devel/snare/distinfo index e44a536f421..1c10d9c1e4b 100644 --- devel/snare/distinfo +++ devel/snare/distinfo @@ -108,7 +108,7 @@ SHA256 (cargo/windows_x86_64_gnu-0.52.6.tar.gz) = FHpcgKq/vwx9kBy1iV0d4w7ykH6yH7 SHA256 (cargo/windows_x86_64_gnullvm-0.52.6.tar.gz) = JNWyPcQXQSZ5aBOW8rSfPejBRz3rUWvTRBCHLv9R7Q0= SHA256 (cargo/windows_x86_64_msvc-0.52.6.tar.gz) = WJ9tqExkYgR0fRJwoqVmHqZu0cztJjHVRv37FVlZ+ew= SHA256 (cargo/wit-bindgen-rt-0.39.0.tar.gz) = b0IyDmH+LP00NU7LWX+G9BNISnmLpEqMoRZcWNQtpsE= -SHA256 (snare-0.4.13.tgz) = cnlBsxaQCjv6dUBsMjSBMjNEQxbTgELj1F5XNQQhj4M= +SHA256 (snare-0.4.14.tgz) = FQic1XCkwKqaqn8bwj7GAttyKeG28lVpq+PTDY851fQ= SIZE (cargo/aho-corasick-1.1.3.tar.gz) = 183311 SIZE (cargo/anyhow-1.0.98.tar.gz) = 53334 SIZE (cargo/autocfg-1.4.0.tar.gz) = 17712 @@ -219,4 +219,4 @@ SIZE (cargo/windows_x86_64_gnu-0.52.6.tar.gz) = 836363 SIZE (cargo/windows_x86_64_gnullvm-0.52.6.tar.gz) = 435707 SIZE (cargo/windows_x86_64_msvc-0.52.6.tar.gz) = 832564 SIZE (cargo/wit-bindgen-rt-0.39.0.tar.gz) = 12241 -SIZE (snare-0.4.13.tgz) = 45100 +SIZE (snare-0.4.14.tgz) = 48686

Saturday, November 21, 2026

Re: UPDATE: FFmpeg 9.1

On Sat, 21 Nov 2026 20:54:46 +0100 Kirill A. Korinsky <kirill@korins.ky> wrote: > > I run my amd64 bulk with my diff which is like this. > > Does it fix issue on your side? > Sure, builds fine. With best regards, Nazarenko Mykyta

Re: UPDATE: FFmpeg 9.1

21 nov 2026 8:02:22 Brad Smith <brad@comstyle.com>: > Here is an update to FFmpeg 9.1. > > Tested on aarch64 with MPlayer, MPV, VLC, Xine-lib. Needs testing > with an amd64 BTI system. > /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_0_16_12_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_0_16_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_0_16_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_2_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_2_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_4_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_4_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_6_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_6_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_8_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_8_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_12_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_12_16_avx512icl' originally defined here src/libswscale/x86/ops_common.asm:270: error: label `ff_u8_rw_shuffle_x_ff_16_16_avx512icl' inconsistently redefined libswscale/x86/uops_macros.gen.asm:3: ... from macro `MULTILINE' defined here /home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/libswscale/x86/ops_include.asm:205: ... from macro `DECL_OP' defined here src/libswscale/x86/ops_common.asm:172: ... from macro `RW_SHUFFLE' defined here src//libavutil/x86/x86inc.asm:827: ... from macro `cglobal' defined here src//libavutil/x86/x86inc.asm:857: ... from macro `cglobal_internal' defined here src/libswscale/x86/ops_common.asm:270: info: label `ff_u8_rw_shuffle_x_ff_16_16_avx512icl' originally defined here gmake: *** [/home/mykyta/Dev/ports/pobj/ffmpeg-9.1/ffmpeg-9.1/ffbuild/common.mak:104: libswscale/x86/ops_common.o] Error 1 *** Error 2 in . (/usr/ports/infrastructure/mk/bsd.port.mk:3094 '/home/mykyta/Dev/ports/pobj/ffmpeg-9.1/build-amd64/.build_done': @cd /home/...) *** Error 2 in /usr/ports/graphics/ffmpeg (/usr/ports/infrastructure/mk/bsd.port.mk:2729 'all': @lock=ffmpeg-9.1v1; export _LOCKS_HELD=" ff...) That's what I get. It's amd64. With best regards, Nazarenko Mykyta

wayland/hyprland: new tiling WM

Hola ports@ , kirill@, kn@. I'd like to suggest adding Hyprland, a beautiful and highly customizable tiling window manager. Any feedback is very important. You can find all tarballs here : https://nazarenko.sbs/ports With best regards, Nazarenko Mykyta

Re: [UPDATE] net/lyrebird 0.8.1 -> 0.10.0

Hi, On Tue, Nov 17, 2026 at 08:26:16PM +0100, Igor Zornik wrote: > Two new versions for lyrebird. Some bug fixes, but mostly Go modules > update. Build and tested on today's amd64 snap. Also needs custom > TEST_TARGET for tests to run. QP-encoded diff below. Thanks for your diff. Looks good to me. Nice that "make test" now works too. I recently saw this email from MAINTAINER Douglas (I didn't see it at the time Douglas sent it): https://marc.info/?l=openbsd-ports&m=177964621553007&w=2 So I guess it makes sense to remove Douglas as MAINTAINER? But I'll still give Douglas the chance to respond, both to the diff and this question. I'll wait a couple of days for an answer and otherwise commit your diff. Caspar

Re: productivity/calcurse update patch src/day.c

I committed an update to that earlier today. The existing patches are based on much older code, much of which was no longer relevant even before the update (i.e. they already fixed a bunch of things for 32-bit machines with 64-bit time_t, though not quite all; reading times from files wasn't working properly which resulted in about half the tests failing). what I committed passes all tests on 32-bit and 64-bit archs but if you run into other problems at runtime let me know (with a reproducer if possible please). On 2026/11/21 10:06, Heppler, J. Scott wrote: > Upstream released 4.9.0 last week. It's was fairly extensive but missed > a few items which were subsequently addressed in 4.9.1. > > Updating the port mainly concerns applying patches for 64bit time. > With the exception of the following, they look fairly easy with shifts > of a few lines. > > src/day.c > Index: src/day.c > --- src/day.c.orig > +++ src/day.c > @@ -856,6 +856,7 @@ int day_paste_item(struct day_item *p, time_t date) > * Careful: p->start is not yet set. > */ > time_t until; > + long until_l; > switch (p->type) { > case EVNT: > > The new 4.9.1 code now assigns time_t until=0; > > Should the old patch src/day.c > > long until_l; > > be updated to > > long until_l=0; > ? > > Thanks, > -- > J. Scott Heppler >

productivity/calcurse update patch src/day.c

Upstream released 4.9.0 last week. It's was fairly extensive but missed a few items which were subsequently addressed in 4.9.1. Updating the port mainly concerns applying patches for 64bit time. With the exception of the following, they look fairly easy with shifts of a few lines. src/day.c Index: src/day.c --- src/day.c.orig +++ src/day.c @@ -856,6 +856,7 @@ int day_paste_item(struct day_item *p, time_t date) * Careful: p->start is not yet set. */ time_t until; + long until_l; switch (p->type) { case EVNT: The new 4.9.1 code now assigns time_t until=0; Should the old patch src/day.c long until_l; be updated to long until_l=0; ? Thanks, -- J. Scott Heppler

Update: Aqualung 2.1

This updates to the latest version of Aqualung. Release announcement at: https://github.com/jeremyevans/aqualung/releases/tag/2.1 Our only local patch was merged upstream. Tested on amd64. Will be committing in a couple days unless I hear objections. Best, Jeremy Index: Makefile =================================================================== RCS file: /cvs/ports/audio/aqualung/Makefile,v retrieving revision 1.65 diff -u -p -u -p -r1.65 Makefile --- Makefile 28 Dec 2025 20:07:45 -0000 1.65 +++ Makefile 21 Nov 2026 17:43:03 -0000 @@ -1,10 +1,9 @@ COMMENT= advanced music player -VERSION= 2.0 +VERSION= 2.1 DISTNAME= aqualung-${VERSION} EPOCH= 0 CATEGORIES= audio -REVISION= 1 HOMEPAGE= https://github.com/jeremyevans/aqualung @@ -86,9 +85,9 @@ CONFIGURE_ARGS= --with-cddb \ --with-vorbisenc \ --with-wavpack \ --without-alsa \ - --without-ifp \ --without-jack \ --without-oss \ + --without-pipewire \ --without-pulse \ --without-winmm Index: distinfo =================================================================== RCS file: /cvs/ports/audio/aqualung/distinfo,v retrieving revision 1.11 diff -u -p -u -p -r1.11 distinfo --- distinfo 21 Feb 2025 02:10:32 -0000 1.11 +++ distinfo 21 Nov 2026 17:43:03 -0000 @@ -1,2 +1,2 @@ -SHA256 (aqualung-2.0.tar.gz) = RP/p2Ai7QxQHaIT5H47OKEpvkkUA5jHRIiWQa9v4CBo= -SIZE (aqualung-2.0.tar.gz) = 1657433 +SHA256 (aqualung-2.1.tar.gz) = sVeW8Us2SeRwrlTpMq3in+zPOEHwePGiuKHPX+DG7bM= +SIZE (aqualung-2.1.tar.gz) = 1686657 Index: patches/patch-src_decoder_dec_lavc_c =================================================================== RCS file: patches/patch-src_decoder_dec_lavc_c diff -N patches/patch-src_decoder_dec_lavc_c --- patches/patch-src_decoder_dec_lavc_c 21 Oct 2025 12:47:24 -0000 1.8 +++ /dev/null 1 Jan 1970 00:00:00 -0000 @@ -1,15 +0,0 @@ -Index: src/decoder/dec_lavc.c ---- src/decoder/dec_lavc.c.orig -+++ src/decoder/dec_lavc.c -@@ -360,7 +360,11 @@ lavc_decoder_close(decoder_t * dec) { - - lavc_pdata_t * pd = (lavc_pdata_t *)dec->pdata; - -+#if LIBAVCODEC_VERSION_INT < AV_VERSION_INT(61, 3, 100) - avcodec_close(pd->avCodecCtx); -+#else -+ avcodec_free_context(&pd->avCodecCtx); -+

Re: UPDATE: Suricata-8.0.6

Hi gonzalo,

With this diff Suricata no longer drops privileges: it keeps running
as root even with --user/--group or the run-as: _suricata that the
port puts in suricata.yaml.

The diff drops the patch-src_suricata_c hunk that called
SCSetUserID() before SCPledge() (8.0 split SuricataMain(), so the old
context no longer applies). The rest is still there: the "#if 0"
hunks make --user/--group accepted, and patch-src_util-privs_c still
defines SCSetUserID(), but nothing calls it any more. Upstream has
no privilege drop without libcap-ng (SCDropMainThreadCaps() is an
empty macro), so the port's call is the only one.

How to reproduce (8.0/amd64, the diff as posted, make package,
pkg_add):

# install -d -o _suricata -g _suricata /var/run/suricata /var/log/suricata
# suricata -c /etc/suricata/suricata.yaml -i lo0 \
--user _suricata --group _suricata -D \
--pidfile /var/run/suricata/suricata.pid
# ps -o user,uid,ruid,svuid,rgid,svgid -p $(cat /var/run/suricata/suricata.pid)
USER UID RUID SVUID RGID SVGID
root 0 0 0 0 0
# ps -o user,pledge -p $(cat /var/run/suricata/suricata.pid)
USER PLEDGE
root stdio,rpath,wpath,cpath,fattr,unix,dns,bpf

With the patch below (on top of your diff) it is 800 everywhere
(_suricata), pledged as before.

Index: patches/patch-src_suricata_c
--- patches/patch-src_suricata_c
+++ patches/patch-src_suricata_c
@@ -22,3 +22,11 @@
SCLogError("libcap-ng is required to"
" drop privileges, but it was not compiled into Suricata.");
return TM_ECODE_FAILED;
+@@ -3192,5 +3192,7 @@ void SuricataPostInit(void)
+ SystemHugepageSnapshotDestroy(prerun_snap);
+ SystemHugepageSnapshotDestroy(postrun_snap);
+ }
++ if (suricata.do_setuid || suricata.do_setgid)
++ SCSetUserID(suricata.userid, suricata.groupid);
+ SCPledge();
+ }

(The do_setuid/do_setgid guard keeps suricata usable by a non-root
user without run-as, e.g. -r on a pcap file; the 7.0.7 call was
unconditional.)

8.0.8 is out. I tested it the same way: your diff +
the fix, SURICATA_V 8.0.8, SUPDATE_V 1.3.8 (the bundled one), new
distinfo. In 8.0.8 SCPledge() moved earlier in SuricataPostInit(),
before the threads are un-paused (upstream ticket #8300), so the
hunk above no longer applies there and moves with it:

+@@ -3184,6 +3184,8 @@ void SuricataPostInit(void)
+
+ /* pledge before allowing threads to continue to avoid an issue with pcap file directory mode,
+ * see ticket #8300. */
++ if (suricata.do_setuid || suricata.do_setgid)
++ SCSetUserID(suricata.userid, suricata.groupid);
+ SCPledge();
+
+ /* Un-pause all the paused threads */

Nothing else needed patching: the other patches apply with offsets,
and PLIST only gains share/suricata/rules/sip-events.rules. I also
built it with Hyperscan by dropping the "prevent textproc/hyperscan"
CONFIGURE_ENV block and adding:

WANTLIB += hs
LIB_DEPENDS += textproc/hyperscan,avx2
CONFIGURE_ARGS += --with-libhs-includes=${LOCALBASE}/include/hs \
--with-libhs-libraries=${LOCALBASE}/lib

Test report, OpenBSD 8.0/amd64 (KVM guest, 4 vCPU), 8.0.8 + fix +
Hyperscan: make fake 509 s, make package OK, the package installs
with pkg_add, build-info says "Hyperscan support: yes", the binary
links libhs, suricata -T passes with the default suricata.yaml and
with mpm-algo/spm-algo set to hs, live pcap on lo0 starts in 1 s,
and the engine runs as _suricata (uid/ruid/svuid/rgid/svgid 800),
pledged. The same check on 8.0.6: root with your diff as posted,
_suricata with the fix.

Pierre Warnier

Friday, November 20, 2026

UPDATE: FFmpeg 9.1

Here is an update to FFmpeg 9.1. Tested on aarch64 with MPlayer, MPV, VLC, Xine-lib. Needs testing with an amd64 BTI system. Index: graphics/ffmpeg/Makefile =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/Makefile,v retrieving revision 1.263 diff -u -p -u -p -r1.263 Makefile --- graphics/ffmpeg/Makefile 19 Sep 2026 19:36:37 -0000 1.263 +++ graphics/ffmpeg/Makefile 21 Nov 2026 03:05:39 -0000 @@ -1,7 +1,7 @@ COMMENT= audio/video converter and streamer # keep it synced with x11/mplayer -V= 9.0.2 +V= 9.1 DISTNAME= ffmpeg-${V} CATEGORIES= graphics multimedia SITES= https://ffmpeg.org/releases/ @@ -22,12 +22,12 @@ USE_NOBTCFI-amd64= Yes EPOCH= 1 -SHARED_LIBS= avcodec 28.0 \ +SHARED_LIBS= avcodec 28.1 \ avdevice 16.0 \ - avfilter 14.0 \ - avformat 25.0 \ - avutil 18.0 \ - swresample 7.0 \ + avfilter 14.1 \ + avformat 25.1 \ + avutil 18.1 \ + swresample 7.1 \ swscale 10.0 HOMEPAGE= https://ffmpeg.org/ @@ -120,6 +120,7 @@ CONFIGURE_ARGS+=--enable-shared \ --cc=${CC} \ --cxx=${CXX} \ --enable-debug \ + --disable-libcurl \ --disable-stripping \ --disable-indev=jack \ --disable-vulkan \ Index: graphics/ffmpeg/distinfo =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/distinfo,v retrieving revision 1.76 diff -u -p -u -p -r1.76 distinfo --- graphics/ffmpeg/distinfo 19 Sep 2026 19:36:37 -0000 1.76 +++ graphics/ffmpeg/distinfo 21 Nov 2026 03:05:39 -0000 @@ -1,2 +1,2 @@ -SHA256 (ffmpeg-9.0.2.tar.xz) = jDhQKD6yX6AmSCB4oEBR4L4XNHsJ74GghJvsFaluAC4= -SIZE (ffmpeg-9.0.2.tar.xz) = 12040788 +SHA256 (ffmpeg-9.1.tar.xz) = Gk05SuY9GQpSbNBS60QVwVa1xeLfN++rQafeC4/eJDM= +SIZE (ffmpeg-9.1.tar.xz) = 12557532 Index: graphics/ffmpeg/patches/patch-configure =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-configure,v retrieving revision 1.81 diff -u -p -u -p -r1.81 patch-configure --- graphics/ffmpeg/patches/patch-configure 25 Aug 2026 07:12:12 -0000 1.81 +++ graphics/ffmpeg/patches/patch-configure 21 Nov 2026 03:05:39 -0000 @@ -3,7 +3,7 @@ Index: configure --- configure.orig +++ configure -@@ -6159,7 +6159,6 @@ case $target_os in +@@ -6083,7 +6083,6 @@ case $target_os in enable section_data_rel_ro striptype="" SHFLAGS='-shared' @@ -11,7 +11,7 @@ Index: configure SLIB_INSTALL_LINKS= oss_indev_extralibs="-lossaudio" oss_outdev_extralibs="-lossaudio" -@@ -6534,7 +6533,7 @@ set_default libdir +@@ -6419,7 +6418,7 @@ set_default libdir set_default $PATHS_LIST set_default nm @@ -20,28 +20,7 @@ Index: configure enable_weak_pic() { disabled pic && return -@@ -6830,11 +6829,19 @@ elif enabled x86; then - # As 'i' is stored on the stack, this program will crash - # if the base pointer is used to access it because the - # base pointer is cleared in the inline assembly code. -- check_exec_crash <<EOF && enable ebp_available -+ # Since https://github.com/llvm/llvm-project/commit/0d471b3f64d3116bd57c79d872f7384fff80daa5, -+ # Clang can save/restore EBP around clobber-only asm, so the old -+ # crash probe can be a false positive. -+ check_exec_crash <<EOF && test_cc <<EOF_EBP_REGS && enable ebp_available - volatile int i=0; - __asm__ volatile ("xorl %%ebp, %%ebp" ::: "%ebp"); - return i; - EOF -+void foo(void *a, void *b, void *c, void *d, int e, void *f) -+{ -+ __asm__ volatile ("" :: "r"(a), "r"(b), "r"(c), "r"(d), "g"(e), "r"(f) : "%eax", "%edx"); -+} -+EOF_EBP_REGS - - # check whether EBX is available on x86 - check_inline_asm ebx_available '""::"b"(0)' && -@@ -7090,7 +7097,7 @@ check_headers asm/types.h +@@ -6987,7 +6986,7 @@ check_headers asm/types.h # # some configurations also require linking to libatomic, so try # both with -latomic and without @@ -50,7 +29,7 @@ Index: configure check_builtin stdatomic stdatomic.h \ "atomic_int foo, bar = -1; atomic_store(&foo, 0); foo += bar" \ $LATOMIC && eval stdatomic_extralibs="\$LATOMIC" && break -@@ -7551,7 +7558,7 @@ enabled opengl && { check_lib opengl GL/glx +@@ -7460,7 +7459,7 @@ enabled opencl && { check_pkg_config opencl enabled openssl && { { check_pkg_config openssl "openssl >= 3.0.0" openssl/ssl.h DTLS_get_data_mtu && { enabled gplv3 || ! enabled gpl || enabled nonfree || die "ERROR: OpenSSL >=3.0.0 requires --enable-version3"; }; } || { enabled gpl && ! enabled nonfree && die "ERROR: OpenSSL <3.0.0 is incompatible with the gpl"; } || Index: graphics/ffmpeg/patches/patch-ffbuild_library_mak =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-ffbuild_library_mak,v retrieving revision 1.6 diff -u -p -u -p -r1.6 patch-ffbuild_library_mak --- graphics/ffmpeg/patches/patch-ffbuild_library_mak 25 Aug 2026 07:12:12 -0000 1.6 +++ graphics/ffmpeg/patches/patch-ffbuild_library_mak 21 Nov 2026 03:05:39 -0000 @@ -1,12 +1,11 @@ Index: ffbuild/library.mak --- ffbuild/library.mak.orig +++ ffbuild/library.mak -@@ -113,8 +113,6 @@ clean:: - install-lib$(NAME)-shared: $(SUBDIR)$(SLIBNAME) - $(Q)mkdir -p "$(SHLIBDIR)" - $$(INSTALL) -m 755 $$< "$(SHLIBDIR)/$(SLIB_INSTALL_NAME)" -- $$(STRIP) "$(SHLIBDIR)/$(SLIB_INSTALL_NAME)" +@@ -116,7 +116,6 @@ install-lib$(NAME)-shared: $(SUBDIR)$(SLIBNAME) + ifneq ($(STRIPTYPE),nostrip) + $$(STRIP) "$(SHLIBDIR)/$(SLIB_INSTALL_NAME)" + endif - $(Q)$(foreach F,$(SLIB_INSTALL_LINKS),(cd "$(SHLIBDIR)" && $(LN_S) $(SLIB_INSTALL_NAME) $(F));) - $(if $(SLIB_INSTALL_EXTRA_SHLIB),$$(INSTALL) -m 644 $(SLIB_INSTALL_EXTRA_SHLIB:%=$(SUBDIR)%) "$(SHLIBDIR)") + $(if $(SLIB_INSTALL_EXTRA_SHLIB),$$(call INSTALL_FILES,-m 644,$(SLIB_INSTALL_EXTRA_SHLIB:%=$(SUBDIR)%),$(SHLIBDIR))) $(if $(SLIB_INSTALL_EXTRA_LIB),$(Q)mkdir -p "$(LIBDIR)") - $(if $(SLIB_INSTALL_EXTRA_LIB),$$(INSTALL) -m 644 $(SLIB_INSTALL_EXTRA_LIB:%=$(SUBDIR)%) "$(LIBDIR)") + $(if $(SLIB_INSTALL_EXTRA_LIB),$$(call INSTALL_FILES,-m 644,$(SLIB_INSTALL_EXTRA_LIB:%=$(SUBDIR)%),$(LIBDIR))) Index: graphics/ffmpeg/patches/patch-libavcodec_aarch64_h26x_sao_neon_S =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-libavcodec_aarch64_h26x_sao_neon_S,v retrieving revision 1.1 diff -u -p -u -p -r1.1 patch-libavcodec_aarch64_h26x_sao_neon_S --- graphics/ffmpeg/patches/patch-libavcodec_aarch64_h26x_sao_neon_S 25 Aug 2026 07:23:52 -0000 1.1 +++ graphics/ffmpeg/patches/patch-libavcodec_aarch64_h26x_sao_neon_S 21 Nov 2026 03:05:39 -0000 @@ -7,12 +7,12 @@ Index: libavcodec/aarch64/h26x/sao_neon. +.section .rodata + - .Lhevc_sao_edge_pos: + .Lhevc_sao_edge_pos_8: .word 1 // horizontal .word HEVC_SAO_STRIDE // vertical -@@ -112,8 +114,11 @@ endfunc - .word VVC_SAO_STRIDE + 1 // 45 degree - .word VVC_SAO_STRIDE - 1 // 135 degree +@@ -124,8 +126,11 @@ endfunc + .word VVC_SAO_STRIDE + 2 // 135 degree + .word VVC_SAO_STRIDE - 2 // 45 degree +.previous + @@ -23,17 +23,37 @@ Index: libavcodec/aarch64/h26x/sao_neon. mov x15, #VVC_SAO_STRIDE b 1f endfunc -@@ -121,7 +126,8 @@ endfunc +@@ -133,7 +138,8 @@ endfunc // ff_hevc_sao_edge_filter_16x16_8_neon(char *dst, char *src, ptrdiff stride_dst, // int16 *sao_offset_val, int eo, int width, int height) function ff_hevc_sao_edge_filter_16x16_8_neon, export=1 -- adr x7, .Lhevc_sao_edge_pos -+ adrp x7, .Lhevc_sao_edge_pos -+ add x7, x7, :lo12:.Lhevc_sao_edge_pos +- adr x7, .Lhevc_sao_edge_pos_8 ++ adrp x7, .Lhevc_sao_edge_pos_8 ++ add x7, x7, :lo12:.Lhevc_sao_edge_pos_8 mov x15, #HEVC_SAO_STRIDE 1: ld1 {v3.8h}, [x3] // load sao_offset_val -@@ -180,7 +186,8 @@ function ff_hevc_sao_edge_filter_16x16_8_neon, export= +@@ -241,7 +247,8 @@ endfunc + function ff_vvc_sao_edge_filter_16x16_10_neon, export=1 + mvni v31.8h, #0xfc, lsl #8 // 1023 + .Lvvc_sao_edge_16x16_16bit: +- adr x7, .Lvvc_sao_edge_pos_16bit ++ adrp x7, .Lvvc_sao_edge_pos_16bit ++ add x7, x7, :lo12:.Lvvc_sao_edge_pos_16bit + mov x15, #VVC_SAO_STRIDE + b .Lh26x_sao_edge_16x16_16bit + endfunc +@@ -258,7 +265,8 @@ endfunc + function ff_hevc_sao_edge_filter_16x16_10_neon, export=1 + mvni v31.8h, #0xfc, lsl #8 // 1023 + .Lhevc_sao_edge_16bit: +- adr x7, .Lhevc_sao_edge_pos_16bit ++ adrp x7, .Lhevc_sao_edge_pos_16bit ++ add x7, x7, :lo12:.Lhevc_sao_edge_pos_16bit + mov x15, #HEVC_SAO_STRIDE + .Lh26x_sao_edge_16x16_16bit: + add w5, w5, #7 +@@ -288,7 +296,8 @@ function ff_hevc_sao_edge_filter_16x16_10_neon, export endfunc function ff_vvc_sao_edge_filter_8x8_8_neon, export=1 @@ -43,13 +63,33 @@ Index: libavcodec/aarch64/h26x/sao_neon. mov x15, #VVC_SAO_STRIDE b 1f endfunc -@@ -188,7 +195,8 @@ endfunc +@@ -296,7 +305,8 @@ endfunc // ff_hevc_sao_edge_filter_8x8_8_neon(char *dst, char *src, ptrdiff stride_dst, // int16 *sao_offset_val, int eo, int width, int height) function ff_hevc_sao_edge_filter_8x8_8_neon, export=1 -- adr x7, .Lhevc_sao_edge_pos -+ adrp x7, .Lhevc_sao_edge_pos -+ add x7, x7, :lo12:.Lhevc_sao_edge_pos +- adr x7, .Lhevc_sao_edge_pos_8 ++ adrp x7, .Lhevc_sao_edge_pos_8 ++ add x7, x7, :lo12:.Lhevc_sao_edge_pos_8 mov x15, #HEVC_SAO_STRIDE 1: ldr w4, [x7, w4, uxtw #2] +@@ -356,7 +366,8 @@ endfunc + function ff_vvc_sao_edge_filter_8x8_10_neon, export=1 + mvni v31.8h, #0xfc, lsl #8 // 1023 + .Lvvc_sao_edge_8x8_16bit: +- adr x7, .Lvvc_sao_edge_pos_16bit ++ adrp x7, .Lvvc_sao_edge_pos_16bit ++ add x7, x7, :lo12:.Lvvc_sao_edge_pos_16bit + mov x15, #VVC_SAO_STRIDE + b .Lh26x_sao_edge_8x8_16bit + endfunc +@@ -373,7 +384,8 @@ endfunc + function ff_hevc_sao_edge_filter_8x8_10_neon, export=1 + mvni v31.8h, #0xfc, lsl #8 // 1023 + .Lhevc_sao_edge_8x8_16bit: +- adr x7, .Lhevc_sao_edge_pos_16bit ++ adrp x7, .Lhevc_sao_edge_pos_16bit ++ add x7, x7, :lo12:.Lhevc_sao_edge_pos_16bit + mov x15, #HEVC_SAO_STRIDE + .Lh26x_sao_edge_8x8_16bit: + ldr w4, [x7, w4, uxtw #2] // a/b offsets in bytes Index: graphics/ffmpeg/patches/patch-libavcodec_aarch64_vc1dsp_neon_S =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-libavcodec_aarch64_vc1dsp_neon_S,v retrieving revision 1.1 diff -u -p -u -p -r1.1 patch-libavcodec_aarch64_vc1dsp_neon_S --- graphics/ffmpeg/patches/patch-libavcodec_aarch64_vc1dsp_neon_S 25 Aug 2026 07:23:52 -0000 1.1 +++ graphics/ffmpeg/patches/patch-libavcodec_aarch64_vc1dsp_neon_S 21 Nov 2026 03:05:39 -0000 @@ -47,7 +47,7 @@ Index: libavcodec/aarch64/vc1dsp_neon.S +.section .rodata + - .align 5 + .align 4 .Lcoeffs_it8: .quad 0x000F00090003 @@ -704,6 +710,8 @@ endfunc Index: graphics/ffmpeg/patches/patch-libavformat_avformat_h =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-libavformat_avformat_h,v retrieving revision 1.8 diff -u -p -u -p -r1.8 patch-libavformat_avformat_h --- graphics/ffmpeg/patches/patch-libavformat_avformat_h 25 Aug 2026 07:12:12 -0000 1.8 +++ graphics/ffmpeg/patches/patch-libavformat_avformat_h 21 Nov 2026 03:05:39 -0000 @@ -3,7 +3,7 @@ Add av_stream_get_first_dts for Chromium Index: libavformat/avformat.h --- libavformat/avformat.h.orig +++ libavformat/avformat.h -@@ -1229,6 +1229,10 @@ typedef struct AVStreamGroup { +@@ -1316,6 +1316,10 @@ typedef struct AVStreamGroup { struct AVCodecParserContext *av_stream_get_parser(const AVStream *s); Index: graphics/ffmpeg/patches/patch-libavformat_tls_openssl_c =================================================================== RCS file: /cvs/ports/graphics/ffmpeg/patches/patch-libavformat_tls_openssl_c,v retrieving revision 1.5 diff -u -p -u -p -r1.5 patch-libavformat_tls_openssl_c --- graphics/ffmpeg/patches/patch-libavformat_tls_openssl_c 25 Aug 2026 07:12:12 -0000 1.5 +++ graphics/ffmpeg/patches/patch-libavformat_tls_openssl_c 21 Nov 2026 03:05:39 -0000 @@ -58,7 +58,7 @@ Index: libavformat/tls_openssl.c done: BIO_free(mem); if (out && out_sz) -@@ -847,7 +861,9 @@ static int tls_open(URLContext *h, const char *uri, in +@@ -858,7 +872,9 @@ static int tls_open(URLContext *h, const char *uri, in s->mtu = 1096; SSL_set_options(c->ssl, SSL_OP_NO_QUERY_MTU); SSL_set_mtu(c->ssl, s->mtu); @@ -68,7 +68,7 @@ Index: libavformat/tls_openssl.c } init_bio_method(h); -@@ -960,10 +976,12 @@ static int tls_write(URLContext *h, const uint8_t *buf +@@ -974,10 +990,12 @@ static int tls_write(URLContext *h, const uint8_t *buf uc->flags &= ~AVIO_FLAG_NONBLOCK; uc->flags |= h->flags & AVIO_FLAG_NONBLOCK; @@ -79,5 +79,5 @@ Index: libavformat/tls_openssl.c } +

Re: [UPDATE] libinput-openbsd 1.31.3

På tirsdag 3. november 2026 kl. 20:11, skrev Landry Breuil: > Le Sun, Nov 01, 2026 at 02:32:45PM +0100, Matthieu Herrb a écrit : > > Hi, > > > > I've updated the current upstream repository > > (github.com/sizeofvoid/libopeninput) to libinput 1.31.3 and merged the > > 2 open merge requests that were mergeable. > > > > An update to the port to use this version is below. > > Comments? Ok? > > seems to work in my limited testing with labwc on a t495s, ok Current/amd64. No tests. Checks fine. No nits I can find. Ran with sway for the last few weeks. No obvious issues. -- yaydn Hope this helps.

remove unused legacy setuid devel/kf5/kdelibs4support ?

COMMENT = legacy support for kdelibs-4.x compatibility @mode 4555 libexec/kf5/fileshareset No consumers, port last touched in 2025, can this go? 5 is "outdated and/or no longer required by other ports". Index: devel/kf5/Makefile =================================================================== RCS file: /cvs/ports/devel/kf5/Makefile,v diff -u -p -r1.30 Makefile --- devel/kf5/Makefile 27 Jun 2026 08:35:47 -0000 1.30 +++ devel/kf5/Makefile 20 Nov 2026 22:30:48 -0000 @@ -86,7 +86,6 @@ SUBDIR += khtml SUBDIR += kmediaplayer SUBDIR += kdesignerplugin SUBDIR += kross -SUBDIR += kdelibs4support # tier 4 SUBDIR += frameworkintegration Index: devel/quirks/Makefile =================================================================== RCS file: /cvs/ports/devel/quirks/Makefile,v diff -u -p -r1.1857 Makefile --- devel/quirks/Makefile 16 Nov 2026 15:15:52 -0000 1.1857 +++ devel/quirks/Makefile 20 Nov 2026 22:31:00 -0000 @@ -3,7 +3,7 @@ CATEGORIES = devel databases DISTFILES = # API.rev -PKGNAME = quirks-7.264 +PKGNAME = quirks-7.265 PKG_ARCH = * MAINTAINER = Marc Espie <espie@openbsd.org> Index: devel/quirks/files/Quirks.pm =================================================================== RCS file: /cvs/ports/devel/quirks/files/Quirks.pm,v diff -u -p -r1.1871 Quirks.pm --- devel/quirks/files/Quirks.pm 16 Nov 2026 15:15:52 -0000 1.1871 +++ devel/quirks/files/Quirks.pm 20 Nov 2026 22:32:47 -0000 @@ -688,6 +688,7 @@ setup_obsolete_reason( 6 => 'ctftools', 3 => 'postgresql-pllua', 3 => 'postgresql_autodoc', + 5 => 'kdelibs4support', ); # beware, if ports with an RDEP on Python 3 are removed in a release where # the Python major version has changed (e.g. 3.13->3.14 between 7.9 and 8.0)

NEW: www/frankenphp, lang/php-zts

Hi, Attached are two new ports: - www/frankenphp: FrankenPHP, a PHP app server built on Caddy. Supports classic mode and a worker mode that keeps the app booted between requests. https://frankenphp.dev - lang/php-zts: PHP built as a thread-safe (ZTS) embed library, which FrankenPHP links against. lang/php has no ZTS or embed build. When php -embed was dropped (dd3553de0c) the note said that if it came back it should "probably be a separate port", and now this is one. This port takes the distfile checksum and patches from lang/php/8.5 (CHECKSUM_FILE, PATCHDIR), builds only the embed SAPI, and installs under its own names to not conflict with php-8.5: lib/libphp-zts-8.5.so.0.0 (versioned via SHARED_LIBS, like libpython3.x), share/php-zts-8.5/include, php-config-zts-8.5 and /etc/php-zts-8.5.ini. Since it shares the distfile, it has to be updated together with lang/php/8.5. Extensions are built in, because existing php-* packages are not ZTS and can't be loaded into it. On top of the lang/php defaults it has curl, intl, mysqli/pdo_mysql, pgsql/pdo_pgsql, sqlite3/pdo_sqlite, zip, bz2, gmp, soap and xsl. I left out gd, which would need the X sets. If you'd rather prefer extensions to be subpackages like lang/php, let me know and I can rework it that way. Please find these two small diffs to go with it, inlined below: - lang/php/8.5: two new patches, used by php-zts. patch-TSRM_TSRM_h: OpenBSD uses emulated TLS, so there is no _tsrm_ls_cache symbol and the initial-exec asm in TSRM.c fails to link on aarch64. https://github.com/php/php-src/pull/25557 patch-Zend_zend_call_stack_c: pthread_stackseg_np() returns the top of the stack, but PHP took it as the bottom, so on any thread but the main one the stack limit was never reached. Deep recursion in a FrankenPHP request killed the whole server with SIGILL instead of throwing an Error. https://github.com/php/php-src/pull/25635 Neither changes anything for the php packages in practice (the TSRM.h part only applies to ZTS, and php runs PHP on the main thread), so I didn't bump REVISION. - infrastructure/db/user.list: 906 _frankenphp FrankenPHP's caddy/go.mod has "replace github.com/dunglas/frankenphp => ../", which lang/go can't resolve, therefore this port drops replace and gets the root module from the proxy like any other dependency. make modgo-gen-modules trips over the replace too, so I generated modules.inc with a small wrapper around modgo-gen-modules-helper that drops it first. I can send it, or a diff to the helper, if useful. Tested on -current/arm64: both build and package, portcheck and port-lib-depends-check are clean, and frankenphp runs from the rc.d script (start, reload, stop), serving PHP 8.5 over HTTPS, and in worker mode under 400 concurrent requests. A shared extension built with phpize-zts-8.5 loads from lib/php-zts-8.5/modules via an .ini in /etc/php-zts-8.5, so a later split into subpackages is possible. I have no amd64 machine, so tests there would be very welcome. I'm happy to maintain both. sthen, php-zts is tied closely to lang/php, so if you'd rather maintain it yourself or put it somewhere else, that's fine with me. Comments and OKs welcome. Thanks, Ostap Index: lang/php/8.5/patches/patch-TSRM_TSRM_h --- lang/php/8.5/patches/patch-TSRM_TSRM_h 1 Jan 1970 00:00:00 -0000 +++ lang/php/8.5/patches/patch-TSRM_TSRM_h 18 Nov 2026 00:00:00 -0000 @@ -0,0 +1,16 @@ +Use the default TLS model: OpenBSD has emulated TLS, making the +initial-exec asm fail to link on aarch64 (undefined _tsrm_ls_cache). +https://github.com/php/php-src/pull/25557 + +Index: TSRM/TSRM.h +--- TSRM/TSRM.h.orig ++++ TSRM/TSRM.h +@@ -152,7 +152,7 @@ TSRM_API bool tsrm_is_managed_thread(void); + # define __has_attribute(x) 0 +

Re: archivers/blosc2 3.7.0

On 20.11.2026 16:49, Martin Reindl wrote: > On Fri, Nov 20, 2026 at 03:41:13PM +0100, Mark Patruck wrote: > > On 20.11.2026 15:34, Martin Reindl wrote: > > > Am 2026-11-19 20:36, schrieb Mark Patruck: > > > > On 19.11.2026 19:35, Martijn van Duren wrote: > > > > > Did you look at my diff here[0] to make it build on alpha? > > > > > If so, I can see if it builds there. > > > > > > > > > > martijn@ > > > > > > > > > > [0] https://marc.info/?l=openbsd-ports&m=179416258678246&w=2 > > > > > > > > New diff attached for 3.7.0 which always sets -D_XOPEN_SOURCE=600 > > > > like also written in the comment upstream (for BSDs). I also made > > > > a change in the mincore() patch. > > > > > > > > Building fine on amd64. > > > > > > Hi Mark, > > > > > > is devel/zfp local? > > > > it's a new port attached to the first post to ports@ > > > > https://marc.info/?l=openbsd-ports&m=179511561328585&w=2 > > > > If you don't want to add devel/zfp for the test, simply change > > > > CONFIGURE_ARGS = -DBLOSC_DEPENDENCY_MODE=EXTERNAL \ > > -DDEACTIVATE_ZFP=ON > > > > Now I recall why blosc did not get an update, py-tables needs this diff on top: Builds fine, but needs RUNDEP math/py-numexpr, otherwise you get F.e. $ /bin/pt2to3 -h Traceback (most recent call last): File "/usr/local/bin/pt2to3", line 5, in <module> from tables.scripts.pt2to3 import main File "/usr/local/lib/python3.14/site-packages/tables/__init__.py", line 76, in <module> from .file import File, open_file, copy_file File "/usr/local/lib/python3.14/site-packages/tables/file.py", line 24, in <module> import numexpr as ne ModuleNotFoundError: No module named 'numexpr' I also wanted to run the test, but was a bit deterred by the 4h run warning. -- Mark Patruck ( mark at wrapped.cx ) GPG key 0xF2865E51 / 187F F6D3 EE04 1DCE 1C74 F644 0D3C F66F F286 5E51 https://www.wrapped.cx

Re: [new] sysutils/susmb

On Fri, Nov 20, 2026 at 04:24:24PM +0100, Benjamin Stürz wrote: > Hi Hiltjo, > > sorry I was very busy and then got sick. > > I was mounting my Hetzner Storage Box [1] using: > > # susmb -f smb://uxxx@uxxx.your-storagebox.de/backup /mnt/storage/ > > Here's the list of all files it tries to access (I stubbed out unveil): > > stat('/etc/gnutls/config'); > unveil('/', ''); > unveil('/dev/null', 'rw'); > unveil('/dev/fuse0', 'rw'); > unveil('/mnt/storage/', 'rwxc'); > unveil('(null)', '(null)'); > stat('/root/.smb/smb.conf'); > fopen('/root/.smb/smb.conf', 'r'); > stat('/etc/samba/smb.conf'); > fopen('/etc/samba/smb.conf', 'r'); > stat('/root/.smb/smb.conf.append'); > fopen('/root/.smb/smb.conf.append', 'r'); > mkdir('/var/run/samba', 0755); > open('/var/run/samba/gencache.tdb', 0201002, 0640); > open('/root/.rnd', 0200000, 0); > open('/dev/urandom', 04, 0); > fopen('/root/.krb5/config', 'r'); > fopen('/etc/krb5.conf', 'r'); > fopen('/root/.hx509/config', 'r'); > fopen('/etc/hx509.conf', 'r'); > fopen('/etc/samba/krb5.conf', 'r'); > fopen('/root/.krb5/config', 'r'); > fopen('/etc/krb5.conf', 'r'); > lstat('/tmp/krb5cc_0'); > open('/dev/fuse0', 0200002, 0); > Hi, Not a fix, but to get an idea what it could be, what if you unveil the following paths: unveil("/var/run/samba", "r"); unveil("/var/run/samba/gencache.tdb", "rw") or even: unveil("/var/run/samba", "rw"); I haven't looked too deeply into the samba source-code yet which combination could trigger it so it requires to read "/var/run/samba/gencache.tdb". Maybe (shot in the dark): in libsmb_context.c it mentions: smbc_setOptionUseCCache It could be changed by setting the environment variable LIBSMBCLIENT_NO_CCACHE="1" without changing the source-code. > Also I noticed that the performance using susmb is > much worse than with sshfs. > Its very possible because it is a different protocol etc. Under the hood it uses just samba / smbclient + the FUSE layer. Some options to smbclient could be improved maybe. Changing some of the buffer sizes already helped a lot compared to usmb. It should not be slower than usmb. Suggestions are welcome. > [1] https://www.hetzner.com/de/storage/storage-box/ > > Thanks for your time, > Benjamin Stürz > > On 11/19/26 6:25 PM, Hiltjo Posthuma wrote: > > Hi, > > > > I've privately e-mailed to ask for more details, but did not get a response to: > > > > I cannot reproduce the issue on OpenBSD 8.0. > > > > On my simple network share it does not read /var/run/samba I think (I checked > > the ktrace log). > > Running a ktrace I see it tries to read files from /etc/samba/. > > However for susmb it should be used stand-alone (smb:// as the applications > > arguments) and not use share names from /etc/samba/ or so. > > > > I'm willing to loosen unveil() a bit, maybe an unveil("/var/run/samba", "r") if > > it is really neccesary. > > > > Could you provide more information of your config? Maybe a ktrace or specific > > added Samba modules? I need more information and details. > > > > Thank you, > > > > On Mon, Nov 09, 2026 at 05:47:57PM +0100, Benjamin Stürz wrote: > > > Hi, > > > > > > thanks for the port, but the unveil() rules are too strict. > > > It needs at least /var/run/samba to work correctly. > > > > > > After writing an LD_PRELOAD shim that bypasses the unveil() calls, > > > mounting works. Otherwise susmb fails on OpenBSD 8.0 with: > > > > > > directory_create_or_exist: mkdir failed on directory /var/run/samba: No such > > > file or directory > > > directory_create_or_exist: mkdir failed on directory /var/run/samba: No such > > > file or directory > > > Connection failed: Invalid argument > > > > > > > > > Kind Regards, > > > Benjamin Stürz > > > > > > On 2/21/26 5:28 PM, Pascal Stumpf wrote: > > > > susmb lets you mount SMB/CIFS shares via FUSE, in the vein of the > > > > Map Network Drive functionality in Windows. > > > > > > > > > > > > This is a fork of usmb, most notably adding privdrop and unveil(2), and > > > > getting rid of the XML config file. Patches from sysutils/usmb have > > > > also been incorporated. > > > > > > -- Kind regards, Hiltjo

Re: archivers/blosc2 3.7.0

On Fri, Nov 20, 2026 at 03:41:13PM +0100, Mark Patruck wrote: > On 20.11.2026 15:34, Martin Reindl wrote: > > Am 2026-11-19 20:36, schrieb Mark Patruck: > > > On 19.11.2026 19:35, Martijn van Duren wrote: > > > > Did you look at my diff here[0] to make it build on alpha? > > > > If so, I can see if it builds there. > > > > > > > > martijn@ > > > > > > > > [0] https://marc.info/?l=openbsd-ports&m=179416258678246&w=2 > > > > > > New diff attached for 3.7.0 which always sets -D_XOPEN_SOURCE=600 > > > like also written in the comment upstream (for BSDs). I also made > > > a change in the mincore() patch. > > > > > > Building fine on amd64. > > > > Hi Mark, > > > > is devel/zfp local? > > it's a new port attached to the first post to ports@ > > https://marc.info/?l=openbsd-ports&m=179511561328585&w=2 > > If you don't want to add devel/zfp for the test, simply change > > CONFIGURE_ARGS = -DBLOSC_DEPENDENCY_MODE=EXTERNAL \ > -DDEACTIVATE_ZFP=ON > Now I recall why blosc did not get an update, py-tables needs this diff on top: Index: Makefile =================================================================== RCS file: /cvs/ports/math/py-tables/Makefile,v diff -u -p -r1.26 Makefile --- Makefile 24 Jul 2026 17:11:06 -0000 1.26 +++ Makefile 20 Nov 2026 15:48:08 -0000 @@ -1,7 +1,6 @@ COMMENT= Python package to manage extremely large amounts of data -MODPY_DISTV= 3.10.2 -REVISION= 2 +MODPY_DISTV= 3.11.1 DISTNAME= tables-${MODPY_DISTV} PKGNAME= py-${DISTNAME} CATEGORIES= math Index: distinfo =================================================================== RCS file: /cvs/ports/math/py-tables/distinfo,v diff -u -p -r1.5 distinfo --- distinfo 23 Feb 2025 20:04:38 -0000 1.5 +++ distinfo 20 Nov 2026 15:48:08 -0000 @@ -1,2 +1,2 @@ -SHA256 (tables-3.10.2.tar.gz) = JUSBKnGG+tuoMdbdNOtJzNeI1qg/TkwrQxuDW2eWyRA= -SIZE (tables-3.10.2.tar.gz) = 4779722 +SHA256 (tables-3.11.1.tar.gz) = eKvPQTCRvHweTowQ+7tDjRrAtah0NsW5csPoJThxtvs= +SIZE (tables-3.11.1.tar.gz) = 4790533 Index: patches/patch-hdf5-blosc2_src_blosc2_filter_c =================================================================== RCS file: patches/patch-hdf5-blosc2_src_blosc2_filter_c diff -N patches/patch-hdf5-blosc2_src_blosc2_filter_c --- /dev/null 1 Jan 1970 00:00:00 -0000 +++ patches/patch-hdf5-blosc2_src_blosc2_filter_c 20 Nov 2026 15:48:08 -0000 @@ -0,0 +1,15 @@ +Index: hdf5-blosc2/src/blosc2_filter.c +--- hdf5-blosc2/src/blosc2_filter.c.orig ++++ hdf5-blosc2/src/blosc2_filter.c +@@ -17,6 +17,11 @@ + #include "blosc2_filter.h" + #include "b2nd.h" + ++/* c-blosc2 >= 3.0 renamed BLOSC2_MAX_DIM to B2ND_MAX_DIM */ ++#ifndef BLOSC2_MAX_DIM ++#define BLOSC2_MAX_DIM B2ND_MAX_DIM ++

Re: [new] sysutils/susmb

Hi Hiltjo, sorry I was very busy and then got sick. I was mounting my Hetzner Storage Box [1] using: # susmb -f smb://uxxx@uxxx.your-storagebox.de/backup /mnt/storage/ Here's the list of all files it tries to access (I stubbed out unveil): stat('/etc/gnutls/config'); unveil('/', ''); unveil('/dev/null', 'rw'); unveil('/dev/fuse0', 'rw'); unveil('/mnt/storage/', 'rwxc'); unveil('(null)', '(null)'); stat('/root/.smb/smb.conf'); fopen('/root/.smb/smb.conf', 'r'); stat('/etc/samba/smb.conf'); fopen('/etc/samba/smb.conf', 'r'); stat('/root/.smb/smb.conf.append'); fopen('/root/.smb/smb.conf.append', 'r'); mkdir('/var/run/samba', 0755); open('/var/run/samba/gencache.tdb', 0201002, 0640); open('/root/.rnd', 0200000, 0); open('/dev/urandom', 04, 0); fopen('/root/.krb5/config', 'r'); fopen('/etc/krb5.conf', 'r'); fopen('/root/.hx509/config', 'r'); fopen('/etc/hx509.conf', 'r'); fopen('/etc/samba/krb5.conf', 'r'); fopen('/root/.krb5/config', 'r'); fopen('/etc/krb5.conf', 'r'); lstat('/tmp/krb5cc_0'); open('/dev/fuse0', 0200002, 0); Also I noticed that the performance using susmb is much worse than with sshfs. [1] https://www.hetzner.com/de/storage/storage-box/ Thanks for your time, Benjamin Stürz On 11/19/26 6:25 PM, Hiltjo Posthuma wrote: > Hi, > > I've privately e-mailed to ask for more details, but did not get a response to: > > I cannot reproduce the issue on OpenBSD 8.0. > > On my simple network share it does not read /var/run/samba I think (I checked > the ktrace log). > Running a ktrace I see it tries to read files from /etc/samba/. > However for susmb it should be used stand-alone (smb:// as the applications > arguments) and not use share names from /etc/samba/ or so. > > I'm willing to loosen unveil() a bit, maybe an unveil("/var/run/samba", "r") if > it is really neccesary. > > Could you provide more information of your config? Maybe a ktrace or specific > added Samba modules? I need more information and details. > > Thank you, > > On Mon, Nov 09, 2026 at 05:47:57PM +0100, Benjamin Stürz wrote: >> Hi, >> >> thanks for the port, but the unveil() rules are too strict. >> It needs at least /var/run/samba to work correctly. >> >> After writing an LD_PRELOAD shim that bypasses the unveil() calls, >> mounting works. Otherwise susmb fails on OpenBSD 8.0 with: >> >> directory_create_or_exist: mkdir failed on directory /var/run/samba: No such >> file or directory >> directory_create_or_exist: mkdir failed on directory /var/run/samba: No such >> file or directory >> Connection failed: Invalid argument >> >> >> Kind Regards, >> Benjamin Stürz >> >> On 2/21/26 5:28 PM, Pascal Stumpf wrote: >>> susmb lets you mount SMB/CIFS shares via FUSE, in the vein of the >>> Map Network Drive functionality in Windows. >>> >>> >>> This is a fork of usmb, most notably adding privdrop and unveil(2), and >>> getting rid of the XML config file. Patches from sysutils/usmb have >>> also been incorporated. >> >

Re: archivers/blosc2 3.7.0

On 20.11.2026 15:34, Martin Reindl wrote: > Am 2026-11-19 20:36, schrieb Mark Patruck: > > On 19.11.2026 19:35, Martijn van Duren wrote: > > > Did you look at my diff here[0] to make it build on alpha? > > > If so, I can see if it builds there. > > > > > > martijn@ > > > > > > [0] https://marc.info/?l=openbsd-ports&m=179416258678246&w=2 > > > > New diff attached for 3.7.0 which always sets -D_XOPEN_SOURCE=600 > > like also written in the comment upstream (for BSDs). I also made > > a change in the mincore() patch. > > > > Building fine on amd64. > > Hi Mark, > > is devel/zfp local? it's a new port attached to the first post to ports@ https://marc.info/?l=openbsd-ports&m=179511561328585&w=2 If you don't want to add devel/zfp for the test, simply change CONFIGURE_ARGS = -DBLOSC_DEPENDENCY_MODE=EXTERNAL \ -DDEACTIVATE_ZFP=ON in archivers/blosc2/Makefile > ===> Building from scratch blosc2-3.7.0 > ===> blosc2-3.7.0 depends on: cmake-core->=4 -> cmake-core-4.3.5 > ===> blosc2-3.7.0 depends on: ninja-* -> ninja-1.11.1p1v1 > ===> blosc2-3.7.0 depends on: lz4-* -> lz4-1.11.1 > ===> blosc2-3.7.0 depends on: zstd-* -> zstd-1.5.7p0 > > > Broken dependency: devel/zfp non existent (DEPENDS was devel/zfp) in > > > archivers/blosc2 > *** Error 1 in . (/home/martin/ports/infrastructure/mk/bsd.port.mk:2394 > '/home/martin/ports/pobj/blosc2-3.7.0/.dep-devel-zfp': @unset _DEPEN...) > *** Error 2 in /home/martin/ports/archivers/blosc2 > (/home/martin/ports/infrastructure/mk/bsd.port.mk:2732 'all': > @lock=blosc2-3.7.0; export...) > > py-tables was not happy with newer blosc2 last time I checked, I need to > revisit that. -- Mark Patruck ( mark at wrapped.cx ) GPG key 0xF2865E51 / 187F F6D3 EE04 1DCE 1C74 F644 0D3C F66F F286 5E51 https://www.wrapped.cx

Re: archivers/blosc2 3.7.0

Am 2026-11-19 20:36, schrieb Mark Patruck: > On 19.11.2026 19:35, Martijn van Duren wrote: >> Did you look at my diff here[0] to make it build on alpha? >> If so, I can see if it builds there. >> >> martijn@ >> >> [0] https://marc.info/?l=openbsd-ports&m=179416258678246&w=2 > > New diff attached for 3.7.0 which always sets -D_XOPEN_SOURCE=600 > like also written in the comment upstream (for BSDs). I also made > a change in the mincore() patch. > > Building fine on amd64. Hi Mark, is devel/zfp local? ===> Building from scratch blosc2-3.7.0 ===> blosc2-3.7.0 depends on: cmake-core->=4 -> cmake-core-4.3.5 ===> blosc2-3.7.0 depends on: ninja-* -> ninja-1.11.1p1v1 ===> blosc2-3.7.0 depends on: lz4-* -> lz4-1.11.1 ===> blosc2-3.7.0 depends on: zstd-* -> zstd-1.5.7p0 >> Broken dependency: devel/zfp non existent (DEPENDS was devel/zfp) in >> archivers/blosc2 *** Error 1 in . (/home/martin/ports/infrastructure/mk/bsd.port.mk:2394 '/home/martin/ports/pobj/blosc2-3.7.0/.dep-devel-zfp': @unset _DEPEN...) *** Error 2 in /home/martin/ports/archivers/blosc2 (/home/martin/ports/infrastructure/mk/bsd.port.mk:2732 'all': @lock=blosc2-3.7.0; export...) py-tables was not happy with newer blosc2 last time I checked, I need to revisit that.

Re: UPDATE: games/classicube -> 1.4.0

På fredag 20. november 2026 kl. 06:53, skrev izzy Meyer: > Hello ports@, > > Here's a relatively trivial diff updating classicube to 1.4.0. > > Notably, upstream changed their build infra so ALL_TARGET was adjusted > as well as some patches were moved around/adjusted accordingly. > > Also, the LibreSSL incompatibility patch that was there previously is > no longer needed. Upstream has done the work to properly support > LibreSSL. > > A bit smaller, I broke up one of the lines in the do-install target as > previously it ran on for a tad to long and made portcheck upset. > > I also did set the JOBS env var in MAKE_FLAGS (used by upstream's > GNU Make infra) to match MAKE_JOBS (as in ports). Unsure if this is > "correct" or even needed, however. > > There *is* a small security change with this update that protects > against blind accepting of some types of invalid PNG data when the > resources are loaded. Not sure about the impact of this, though. > Thought it'd be worth mentioning. Since this is security related, I am > open to providing a diff for this update against -stable as well if > someone would be interested in that. > > Thanks, > > -- > iz (she/her) > Both {current,stable-8.0}/amd64. I'll let someone else knowledgeable comment on your compile flag changes and whether it needs backporting to stable-8.0. Runs fine on sway/xwayland on both current and stable. Lgtm otherwise. -- yaydn Hope this helps. Happy hacking the planet. > > I say mundane things > > so the uninteresting > > just might get noticed. > > izder456 (dot) neocities (dot) org >

Thursday, November 19, 2026

Re: archivers/blosc2 3.7.0

On 19.11.2026 19:35, Martijn van Duren wrote: > Did you look at my diff here[0] to make it build on alpha? > If so, I can see if it builds there. > > martijn@ > > [0] https://marc.info/?l=openbsd-ports&m=179416258678246&w=2 New diff attached for 3.7.0 which always sets -D_XOPEN_SOURCE=600 like also written in the comment upstream (for BSDs). I also made a change in the mincore() patch. Building fine on amd64. Index: Makefile =================================================================== RCS file: /cvs/ports/archivers/blosc2/Makefile,v retrieving revision 1.11 diff -u -p -r1.11 Makefile --- Makefile 20 Sep 2026 15:35:58 -0000 1.11 +++ Makefile 19 Nov 2026 19:29:07 -0000 @@ -1,9 +1,11 @@ -V = 2.23.1 COMMENT = fast, compressed, persistent binary data store library for C + +V = 3.7.0 PKGNAME = blosc2-${V} + CATEGORIES = archivers -SHARED_LIBS += blosc2 6.0 # 0.0 +SHARED_LIBS += blosc2 7.0 # 0.0 GH_ACCOUNT = Blosc GH_PROJECT = c-blosc2 @@ -15,7 +17,7 @@ MAINTAINER = Martin Reindl <martin@catai # BSD PERMIT_PACKAGE = Yes -WANTLIB += m lz4 pthread z zstd +WANTLIB += m lz4 pthread z zfp zstd COMPILER = base-clang ports-gcc COMPILER_LANGS = c @@ -23,10 +25,9 @@ COMPILER_LANGS = c MODULES = devel/cmake LIB_DEPENDS = archivers/lz4 \ - archivers/zstd + archivers/zstd \ + devel/zfp -CONFIGURE_ARGS = -DPREFER_EXTERNAL_LZ4=ON \ - -DPREFER_EXTERNAL_ZLIB=ON \ - -DPREFER_EXTERNAL_ZSTD=ON +CONFIGURE_ARGS = -DBLOSC_DEPENDENCY_MODE=EXTERNAL .include <bsd.port.mk> Index: distinfo =================================================================== RCS file: /cvs/ports/archivers/blosc2/distinfo,v retrieving revision 1.8 diff -u -p -r1.8 distinfo --- distinfo 14 Mar 2026 09:41:24 -0000 1.8 +++ distinfo 19 Nov 2026 19:29:07 -0000 @@ -1,2 +1,2 @@ -SHA256 (c-blosc2-2.23.1.tar.gz) = OhpV0eN5T7K1GhLnItYRs+V3RDq7f/mVFmZRH1duo9o= -SIZE (c-blosc2-2.23.1.tar.gz) = 3338640 +SHA256 (c-blosc2-3.7.0.tar.gz) = xospCsmDYZQdqI1lgawnZomgbgi22l6QvIQDqbCIsJI= +SIZE (c-blosc2-3.7.0.tar.gz) = 2200899 Index: patches/patch-CMakeLists_txt =================================================================== RCS file: patches/patch-CMakeLists_txt diff -N patches/patch-CMakeLists_txt --- /dev/null 1 Jan 1970 00:00:00 -0000 +++ patches/patch-CMakeLists_txt 19 Nov 2026 19:29:07 -0000 @@ -0,0 +1,19 @@ +Index: CMakeLists.txt +--- CMakeLists.txt.orig ++++ CMakeLists.txt +@@ -661,14 +661,7 @@ if(CMAKE_C_COMPILER_ID MATCHES GNU|Clang|Intel|IntelLL + endif() + endif() + +-if(CMAKE_C_COMPILER_ID MATCHES Intel|IntelLLVM|Clang OR HAIKU) +- # We need to tell Intel and Clang compilers about what level of POSIX they support. +- # Note this cannot be dropped on the BSDs even though it hides their BSD extensions +- # (see the _XOPEN_SOURCE handling at the top of frame.c): my_malloc() in blosc2.c +- # keys the posix_memalign() path off it, and there is nothing else defining +- # _POSIX_C_SOURCE there, so dropping it would silently fall back to plain malloc(). +- SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -D_XOPEN_SOURCE=600") +-endif() ++SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -D_XOPEN_SOURCE=600") + + if(MSVC) + if(NOT CMAKE_C_FLAGS) Index: patches/patch-bench_CMakeLists_txt =================================================================== RCS file: /cvs/ports/archivers/blosc2/patches/patch-bench_CMakeLists_txt,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 patch-bench_CMakeLists_txt --- patches/patch-bench_CMakeLists_txt 5 Apr 2024 08:53:54 -0000 1.1.1.1 +++ patches/patch-bench_CMakeLists_txt 19 Nov 2026 19:29:07 -0000 @@ -1,12 +1,21 @@ Index: bench/CMakeLists.txt --- bench/CMakeLists.txt.orig +++ bench/CMakeLists.txt -@@ -26,7 +26,7 @@ add_executable(sum_openmp ${SOURCES_SUM_OPENMP}) - add_executable(zero_runlen ${SOURCES_ZERO_RUNLEN}) - add_executable(create_frame ${SOURCES_CFRAME}) +@@ -30,7 +30,7 @@ add_executable(create_frame ${SOURCES_CFRAME}) add_executable(sframe_bench ${SOURCES_SFRAME}) + add_executable(get_sparse ${SOURCES_GET_SPARSE}) + add_executable(frame_lock_bench ${SOURCES_FRAME_LOCK}) -if(UNIX AND NOT APPLE) +if(UNIX AND NOT APPLE AND NOT CMAKE_SYSTEM_NAME STREQUAL "OpenBSD") # cmake is complaining about LINK_PRIVATE in original PR # and removing it does not seem to hurt, so be it. # target_link_libraries(b2bench LINK_PRIVATE rt) +@@ -62,7 +62,7 @@ target_link_libraries(frame_lock_bench blosc_testing) + + add_executable(bench_compact_tail bench_compact_tail.c) + target_link_libraries(bench_compact_tail blosc_testing) +-if(UNIX AND NOT APPLE) ++if(UNIX AND NOT APPLE AND NOT CMAKE_SYSTEM_NAME STREQUAL "OpenBSD") + target_link_libraries(bench_compact_tail rt) + endif() + Index: patches/patch-blosc_CMakeLists_txt =================================================================== RCS file: /cvs/ports/archivers/blosc2/patches/patch-blosc_CMakeLists_txt,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 patch-blosc_CMakeLists_txt --- patches/patch-blosc_CMakeLists_txt 5 Apr 2024 08:53:54 -0000 1.1.1.1 +++ patches/patch-blosc_CMakeLists_txt 19 Nov 2026 19:29:07 -0000 @@ -1,7 +1,7 @@ Index: blosc/CMakeLists.txt --- blosc/CMakeLists.txt.orig +++ blosc/CMakeLists.txt -@@ -234,7 +234,9 @@ if(HAVE_IPP) +@@ -172,7 +172,9 @@ if(BUILD_PLUGINS AND TARGET zfp) endif() if(UNIX AND NOT APPLE) Index: patches/patch-blosc_prefetch_c =================================================================== RCS file: patches/patch-blosc_prefetch_c diff -N patches/patch-blosc_prefetch_c --- /dev/null 1 Jan 1970 00:00:00 -0000 +++ patches/patch-blosc_prefetch_c 19 Nov 2026 19:29:07 -0000 @@ -0,0 +1,22 @@ +Index: blosc/prefetch.c +--- blosc/prefetch.c.orig ++++ blosc/prefetch.c +@@ -560,6 +560,7 @@ static void cache_query_destroy(cache_query *q) { + cross a multiple of CHECK_NBYTES. Only the pages to be read are asked about: the kernel looks at each of them, and + the chunks of one slice can lie far apart. */ + static bool is_cached(const cache_query *q, int64_t start, int64_t stop) { ++#if defined(__linux__) + int64_t first = (start - q->lo) / q->page * q->page; // start of the first page, relative to lo + int64_t npages = (stop - q->lo - first + q->page - 1) / q->page; + #if defined(SYS_cachestat) +@@ -579,7 +580,9 @@ static bool is_cached(const cache_query *q, int64_t st + return false; + } + } +- return true; ++

Re: policy about ports updates on oldstable

On Thu, Nov 19, 2026 at 01:21:18PM +0100, Kirill A. Korinsky wrote: > On Thu, 19 Nov 2026 12:49:24 +0100, > Leah Rowe <info@minifree.org> wrote: > > > > i understand openbsd does 1 year of support for each release, which > > means that each release gets support and the previous release also > > > > i maintain librewolf on -current and yaydn backports to stable for me, > > but what about oldstable? is it common for maintainers to port package > > updates in ports over to oldstable? or is that generally a no-go? > > > > just a passing though, and i would like to know if this is a starter or > > not. probably not worth bothering with as i guess most people tend to > > just update anyway, openbsd makes that easy, but i thought i'd ask > > > > As far as I know we're backports some changes to -stable only in ports. > > All development happens in -current. Not really. There's the whole syspatch infrastructure for the base system. I don't know if the people responsible (robert@ ?) bother to commit the corresponding patch to the release branch though.

Re: [new] sysutils/susmb

Hi, I've privately e-mailed to ask for more details, but did not get a response to: I cannot reproduce the issue on OpenBSD 8.0. On my simple network share it does not read /var/run/samba I think (I checked the ktrace log). Running a ktrace I see it tries to read files from /etc/samba/. However for susmb it should be used stand-alone (smb:// as the applications arguments) and not use share names from /etc/samba/ or so. I'm willing to loosen unveil() a bit, maybe an unveil("/var/run/samba", "r") if it is really neccesary. Could you provide more information of your config? Maybe a ktrace or specific added Samba modules? I need more information and details. Thank you, On Mon, Nov 09, 2026 at 05:47:57PM +0100, Benjamin Stürz wrote: > Hi, > > thanks for the port, but the unveil() rules are too strict. > It needs at least /var/run/samba to work correctly. > > After writing an LD_PRELOAD shim that bypasses the unveil() calls, > mounting works. Otherwise susmb fails on OpenBSD 8.0 with: > > directory_create_or_exist: mkdir failed on directory /var/run/samba: No such > file or directory > directory_create_or_exist: mkdir failed on directory /var/run/samba: No such > file or directory > Connection failed: Invalid argument > > > Kind Regards, > Benjamin Stürz > > On 2/21/26 5:28 PM, Pascal Stumpf wrote: > > susmb lets you mount SMB/CIFS shares via FUSE, in the vein of the > > Map Network Drive functionality in Windows. > > > > > > This is a fork of usmb, most notably adding privdrop and unveil(2), and > > getting rid of the XML config file. Patches from sysutils/usmb have > > also been incorporated. > -- Kind regards, Hiltjo

Re: sysutils/obsd.sh: new port

I've given kirill a few feedback points that will make this easier for developers to use. I suspect he'll send a new one in a few days.

net/bgpq4: update to 1.17

I put a bugfix release out, OK? Index: Makefile =================================================================== RCS file: /cvs/ports/net/bgpq4/Makefile,v diff -u -p -r1.19 Makefile --- Makefile 28 Jan 2025 11:22:03 -0000 1.19 +++ Makefile 19 Nov 2026 16:43:59 -0000 @@ -1,6 +1,6 @@ COMMENT= generate access/prefix lists for BGP config -DIST_TUPLE= github bgp bgpq4 1.15 . +DIST_TUPLE= github bgp bgpq4 1.17 . REVISION= 0 CATEGORIES= net Index: distinfo =================================================================== RCS file: /cvs/ports/net/bgpq4/distinfo,v diff -u -p -r1.15 distinfo --- distinfo 15 May 2024 11:30:54 -0000 1.15 +++ distinfo 19 Nov 2026 16:43:59 -0000 @@ -1,2 +1,2 @@ -SHA256 (bgp-bgpq4-1.15.tar.gz) = MP6o1idK9uP7owSXuXepJMed4yauIhHkkDYtxygkA9Y= -SIZE (bgp-bgpq4-1.15.tar.gz) = 61084 +SHA256 (bgp-bgpq4-1.17.tar.gz) = 9CDoLaELmROZID4B6i6jNpQIrDZ0Jb9IouCvFFuWu/Q= +SIZE (bgp-bgpq4-1.17.tar.gz) = 61527

Re: policy about ports updates on oldstable

> > > There is no "oldstable" for ports. that's all i wanted to know, thank you. yes, because openbsd releases get one year of support, yet releases occur on a 6-month schedule. thus, the current and previous release are supported, and then there is -current. so there are 3 branches. noted. then i will disregard oldstable (i call it that. e.g. openbsd 7.9 is currently the oldstable release. i'm infected with the debian mindvirus) -- Company director, Minifree Ltd Registered in England, No. 9361826 | VAT No. GB202190462 Registered Office: 19 Hilton Road, Canvey Island, Essex SS8 9QA, UK

sysutils/obsd.sh: new port

ports@, I'm working for months to make using qemu-baes VMs with OpenBSD simple. Now I'm at the point when I'd like to add my work to ports because I think it can be quite usefull for others. Here my script which allows me to run my VMs as disposable on demand, or as long running services. Disposable usuall runs as: obsd.sh -s -i octeon to instal one and after that just obsd.sh octeon and I run it as services, under _vmd user by this was. To install: doas su -c obsd_octeon -s /bin/sh _vmd /usr/local/bin/obsd.sh -d /var/vm -s -i octeon make a services doas ln -s /etc/rc.d/obsd /etc/rc.d/obsd_octeon doas cp /etc/login.conf.d/obsd /etc/login.conf.d/obsd_octeon doas sed -i 's/^obsd:/obsd_octeon:/' /etc/login.conf.d/obsd_octeon doas rcctl enable obsd_octeon doas rcctl set obsd_octeon flags -p 22222 octeon after that I can connect to that VM via ssh with following conifg: Host obsd-octeon.some-host HostKeyAlias obsd-octeon.some-host Port 22222 Host obsd-*.some-host ProxyJump some-host HostName 127.0.0.1 User obsd IdentityFile /usr/local/share/examples/obsd.sh/obsd_ed25519.pem and at host to connect to its console: doas -u _vmd tmux -N -S /var/vm/obsd_octeon.sock attach-session Ok to import? -- wbr, Kirill

Re: [update] www/librewolf 160.0.1-1

Thank you for your kind words. As long as it is co-maintainer, yes please. Also appreciate the reviews and commits Landry. Look forward to being each other's fail safe maintainer for this. -- yaydn To good builds. Cheers. Happy hacking the planet. På torsdag 19. november 2026 kl. 13:08, skrev Landry Breuil: > CVSROOT: /cvs > Module name: ports > Changes by: landry@cvs.openbsd.org 2026/11/19 06:07:48 > > Modified files: > www/librewolf : Makefile distinfo > > Log message: > www/librewolf: update to 160.0.1-1, from MAINTAINER Leah Rowe > > while here add yaydn as co-MAINTAINER, agreed by current MAINTAINER. På torsdag 19. november 2026 kl. 13:09, skrev Landry Breuil: > CVSROOT: /cvs > Module name: ports > Changes by: landry@cvs.openbsd.org 2026/11/19 06:09:10 > > Modified files: > www/librewolf : Tag: OPENBSD_8_0 distinfo Makefile > > Log message: > www/librewolf: MFC update to 160.0.1-1, from yaydn På torsdag 19. november 2026 kl. 12:58, skrev Leah Rowe: > > seems fine to me > > yaydn has proven their worth > > > Am 19.11.26 um 11:55 schrieb landry@openbsd.org: > > Le Thu, Nov 19, 2026 at 11:47:13AM +0000, Leah Rowe a écrit : > >> i have a theory that you're probably closer to whatever timezone librewolf's > >> devs operate in than i am, because it seems i'm often a bit later than you > >> to do a diff :) > >> > >> it's good to know that there's a backup in case one of us goes awol. thank > >> you for being my human mirror, in updating this port. it's good that there > >> are two of us doing this. > > in that case, should we put both of you as MAINTAINERs ? this way i dont > > need to wait for feedback and just commit whichever's diff is first, > > given that they're mostly trivial... > > > > Landry > > -- > Company director, Minifree Ltd > Registered in England, No. 9361826 | VAT No. GB202190462 > Registered Office: 19 Hilton Road, Canvey Island, Essex SS8 9QA, UK > >

Wednesday, November 18, 2026

Re: move xpdf3 to xpopple?

Hi, On Thu, Nov 12, 2026 at 10:24:35PM +0100, Matthias Kilian wrote: > On Thu, Nov 12, 2026 at 06:55:08PM +0000, Stuart Henderson wrote: > > Debian and Gentoo use a version of xpdf 3 that has been modified to > > use Poppler. The fork is currently maintained by Adam Samson at > > https://offog.org/code/xpopple/ and regularly gets updates for new > > Poppler versions. > > > > I wonder about moving our version across; what does anyone think? > > I didn't test it yet, but given that the original xpdf3 is dead > upstream, I think it's a good approach. Now also tested. Works ok for me (except for a warning "Fontconfig warning: using without calling FcInit()", but i think that's harmless for now). I also tested printing via lpr(1) up to the point that a print job can be queued correctly (i don't have a printer here at home). For me, it's perfectly ok to move it to xpopple. Ciao, Kili

[maintainer update] benchmarks/librespeed-cli 1.0.13 -> 1.0.15

Good evening ports@ this patch bumps librespeed-cli. What can I say. It still does its job. portcheck and make port-lib-depends-check are happy. ok?

diff --git a/benchmarks/librespeed-cli/Makefile b/benchmarks/librespeed-cli/Makefile
index c05ad4d75c9..c8514e91278 100644
--- a/benchmarks/librespeed-cli/Makefile
+++ b/benchmarks/librespeed-cli/Makefile
@@ -3,8 +3,7 @@ ONLY_FOR_ARCHS =	${GO_ARCHS}
 COMMENT =		command line client for LibreSpeed
 
 MODGO_MODNAME =		github.com/librespeed/speedtest-cli
-MODGO_VERSION =		v1.0.13
-REVISION =		0
+MODGO_VERSION =		v1.0.15
 
 DISTNAME =		librespeed-cli-${MODGO_VERSION}
 
diff --git a/benchmarks/librespeed-cli/distinfo b/benchmarks/librespeed-cli/distinfo
index 112f54e7562..47671d81b78 100644
--- a/benchmarks/librespeed-cli/distinfo
+++ b/benchmarks/librespeed-cli/distinfo
@@ -1,9 +1,9 @@
-SHA256 (go_modules/github.com/!burnt!sushi/toml/@v/v1.3.2.mod) = JnfIL+dPIDdyiJpagBr7Mp2VF1UId92ssXIZFpurPt0=
-SHA256 (go_modules/github.com/!burnt!sushi/toml/@v/v1.3.2.zip) = XeJGoMtMJW8/1dDbigihFPWK8MLhk7vwrZASEErbtrI=
-SHA256 (go_modules/github.com/briandowns/spinner/@v/v1.23.1.mod) = 5kl+NIWPfHiQPOJBFVzd2gDGgl25k07QM5nDMxPiDZk=
-SHA256 (go_modules/github.com/briandowns/spinner/@v/v1.23.1.zip) = fz/VaPuSJcsvnt1L7v2duTP08XO5utCYMBx//gOEvfs=
-SHA256 (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.4.mod) = FMOvou9iNVXBtINNH4ixTfd1Sa/uheV1DBkQJGW3iro=
-SHA256 (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.4.zip) = mtuqF/F9vjauRk2AYqy7+5z4U7/r7UJENag0orC+2RA=
+SHA256 (go_modules/github.com/!burnt!sushi/toml/@v/v1.5.0.mod) = RdeLgFcZmbaiWNoPIWc9o55rUBNJwDOkrNFpQ3QR1zM=
+SHA256 (go_modules/github.com/!burnt!sushi/toml/@v/v1.5.0.zip) = oQyNPWxKm3PciFRkJF7sayfWT0MNaXk4nNnFit3hWFU=
+SHA256 (go_modules/github.com/briandowns/spinner/@v/v1.23.2.mod) = 5kl+NIWPfHiQPOJBFVzd2gDGgl25k07QM5nDMxPiDZk=
+SHA256 (go_modules/github.com/briandowns/spinner/@v/v1.23.2.zip) = 1OR9aAf3/DWEr+jmLgu9Tp4Ly5I3Fbqntr6rXIJx/LE=
+SHA256 (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.7.mod) = u4Gp4tFNmlhvBZGFO9qtIw2M3w2Nb/px3driSa1XPMk=
+SHA256 (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.7.zip) = mtXyy8ZEwLtkipsYLWtw+S56xN+rXPDvdGu+75RP2lk=
 SHA256 (go_modules/github.com/fatih/color/@v/v1.17.0.mod) = EVCWwLAxTBSaQR67bfCX187wr8cx2gHTTexr2pLfgQ4=
 SHA256 (go_modules/github.com/fatih/color/@v/v1.17.0.zip) = iltulKtJ3m+GnT79TgQv/UbCoSfyUTn267YUbkyescQ=
 SHA256 (go_modules/github.com/fatih/color/@v/v1.7.0.mod) = TDlJ5lwSok4Luy1SEYvASOaYDCm8SUAsWWWK3gNRmKM=
@@ -18,40 +18,43 @@ SHA256 (go_modules/github.com/mattn/go-isatty/@v/v0.0.16.mod) = cUdm7ILWfbiVreiv
 SHA256 (go_modules/github.com/mattn/go-isatty/@v/v0.0.20.mod) = Huq+8bTZjbc8bCBzKJFjZJvJoERVKsqGudWS/oTqocg=
 SHA256 (go_modules/github.com/mattn/go-isatty/@v/v0.0.20.zip) = 8tX4nKRRV34XRkubtZbcDQ7OzsteqmNiLEG1fNC3uMw=
 SHA256 (go_modules/github.com/mattn/go-isatty/@v/v0.0.8.mod) = T3daib6agBiagm7yLcMD44lVfFWPFL/UC3W17yhF5W8=
-SHA256 (go_modules/github.com/prometheus-community/pro-bing/@v/v0.8.0.mod) = fzgJNOxmJ8jW2ya1wm0v7mdhLSlbYm2UBMfJ5qCPmFA=
-SHA256 (go_modules/github.com/prometheus-community/pro-bing/@v/v0.8.0.zip) = ej6K+NZQMAq/tN+e7sQ11uXuJsxn6Hg9wk6Hw/gV9gI=
+SHA256 (go_modules/github.com/prometheus-community/pro-bing/@v/v0.9.1.mod) = TGaN5R/9mEdv5yQnIxojRDEXU3bn9mi+pzqafEYRsU0=
+SHA256 (go_modules/github.com/prometheus-community/pro-bing/@v/v0.9.1.zip) = tn7HVoxXU3YjF/XUDyPF2sgQKzVoR7PLR0QDMpSBDko=
 SHA256 (go_modules/github.com/russross/blackfriday/v2/@v/v2.1.0.mod) = w5VHzOboZujPo5XSk3vOMIAc0H8LkBwhMySPH7oS9Y0=
 SHA256 (go_modules/github.com/russross/blackfriday/v2/@v/v2.1.0.zip) = eFJ1DVigU844sB8sIDIIgXVk9VLr83GytjAIHXAExq4=
-SHA256 (go_modules/github.com/urfave/cli/v2/@v/v2.27.4.mod) = UmfHv9x/17zhZR7Sd9Vy1xcB3nmjKE1SPU9wztX+TeI=
-SHA256 (go_modules/github.com/urfave/cli/v2/@v/v2.27.4.zip) = HPSpifOKjsKcOZQAxAxY6NWeoa+cZpIcALK1HLacBTU=
+SHA256 (go_modules/github.com/urfave/cli/v2/@v/v2.27.7.mod) = rprCPdIb3cUE6IElYKUtPUq6KIoAZXnkkCV5sqCeHPE=
+SHA256 (go_modules/github.com/urfave/cli/v2/@v/v2.27.7.zip) = qaQXWrXVUPCT9xbolxpchTmqc4Z1ohD2Mhz7WQXwSIo=
 SHA256 (go_modules/github.com/xrash/smetrics/@v/v0.0.0-20240521201337-686a1a2994c1.mod) = 20TcnREk0ciM/YvNMyL5h+eslNKTQE/SouqOivhFOKo=
 SHA256 (go_modules/github.com/xrash/smetrics/@v/v0.0.0-20240521201337-686a1a2994c1.zip) = YGHZph6GPJuKTtPjSwDWeE2uTKfsZKDm8OXEu4V0Mo0=
-SHA256 (go_modules/golang.org/x/crypto/@v/v0.47.0.mod) = /7gT2Z5zjMawzCnpCjxY7w3P5kkNNsIVO5c9QcX7fZk=
-SHA256 (go_modules/golang.org/x/crypto/@v/v0.47.0.zip) = h0V0aoDB6/+R4ddztsKMSqTwdNaqk4t6UTMNUu0ioMA=
-SHA256 (go_modules/golang.org/x/net/@v/v0.49.0.mod) = 9juXIMbYfvBWYtRkHnhTGvmOahO+CGEb+6cSjpst6XA=
-SHA256 (go_modules/golang.org/x/net/@v/v0.49.0.zip) = wHpNVsPbUtwrKcZbE0+WwssHHopQ5PSIrH+stGAFzIU=
-SHA256 (go_modules/golang.org/x/sync/@v/v0.19.0.mod) = 0zPFS3SviguOx0jTfFly0nudCIueRci/XDq1INIRMJA=
-SHA256 (go_modules/golang.org/x/sync/@v/v0.19.0.zip) = JSEf4s/9gCC7QFua23qQ9eBnYPKBi4+y50qqohpm7Z4=
+SHA256 (go_modules/golang.org/x/crypto/@v/v0.54.0.mod) = 2FK5O5P5zZEXCP/UVAlDmVbU6chJfFuH80j8g9CtxXc=
+SHA256 (go_modules/golang.org/x/crypto/@v/v0.54.0.zip) = 7PQDDBxrVbiXgmuV8B10Fs8nhbRnWw+awhjd91SgOrs=
+SHA256 (go_modules/golang.org/x/net/@v/v0.56.0.mod) = EHKRUjOmYXpgOjaZaJLKiq/MfYuNwPQ+cUbYPH6knEM=
+SHA256 (go_modules/golang.org/x/net/@v/v0.57.0.mod) = v8TDS5+l1w5IN1Qld1YOkvdD4k9YSjmHeFBlDBmkRKE=
+SHA256 (go_modules/golang.org/x/net/@v/v0.57.0.zip) = +NZSxik3S0OEbvTy79QZ6NHIF3nrcoMG0J8N4cwbq4A=
+SHA256 (go_modules/golang.org/x/sync/@v/v0.21.0.mod) = o+KedgYL1WEGBFSx+ivc1mZ09gycqTgzuBBjVeNMYDw=
+SHA256 (go_modules/golang.org/x/sync/@v/v0.21.0.zip) = 7mVFkCPefySDb24hIxRLUym9Ck0FqHw8RIUJN44ua+c=
 SHA256 (go_modules/golang.org/x/sys/@v/v0.0.0-20220412211240-33da011f77ad.mod) = 8DMzMJb+GY8xUd7tk/LeunTlC7/nc5E0BFvDt85KUCQ=
 SHA256 (go_modules/golang.org/x/sys/@v/v0.0.0-20220811171246-fbc7d0a398ab.mod) = 8DMzMJb+GY8xUd7tk/LeunTlC7/nc5E0BFvDt85KUCQ=
 SHA256 (go_modules/golang.org/x/sys/@v/v0.18.0.mod) = 0iezJfYh9OvijTm6dz6pm4cPOTt8CcNFksNlsW3VYN4=
-SHA256 (go_modules/golang.org/x/sys/@v/v0.40.0.mod) = 9BGBTYOpboZ4Gx3uQcEluVBNpUItujfMHWOwFq45z+I=
-SHA256 (go_modules/golang.org/x/sys/@v/v0.40.0.zip) = 1/n+RDAPyNR7MNBs0AHZxixUxhGYGKGaLjOpNTiXFmk=
+SHA256 (go_modules/golang.org/x/sys/@v/v0.46.0.mod) = V/Q5PqGNVEahI2OzXCOmFthD+hZpxxIacKK8OpZ31mU=
+SHA256 (go_modules/golang.org/x/sys/@v/v0.47.0.mod) = V/Q5PqGNVEahI2OzXCOmFthD+hZpxxIacKK8OpZ31mU=
+SHA256 (go_modules/golang.org/x/sys/@v/v0.50.0.mod) = lSL0yYBtHCZYbqgPpWKSkpvSlYZecZgZJ/UplsaxyDE=
+SHA256 (go_modules/golang.org/x/sys/@v/v0.50.0.zip) = sRXooUSnydHfP/uK8gsoPhYoRLW2JKl+40KVXMppueY=
 SHA256 (go_modules/golang.org/x/sys/@v/v0.6.0.mod) = 8DMzMJb+GY8xUd7tk/LeunTlC7/nc5E0BFvDt85KUCQ=
 SHA256 (go_modules/golang.org/x/term/@v/v0.1.0.mod) = dAeU4pFPXMh+JWVsWEMiKyxlZlghIF9FYyoEd8jT95s=
-SHA256 (go_modules/golang.org/x/term/@v/v0.39.0.mod) = wDklZj+X1JDHRXWE2tqVS2z748R2K6AufPbw6694oss=
-SHA256 (go_modules/golang.org/x/term/@v/v0.39.0.zip) = RDux+HrRGn0tqP6k5YJy6G89cIlfOOE6mSU23EMfxS0=
-SHA256 (go_modules/golang.org/x/text/@v/v0.33.0.mod) = Ef7SIeJ/A1RzGaflrnBC6CNKhaPk5ZzFGx984iLJnkA=
-SHA256 (go_modules/golang.org/x/text/@v/v0.33.0.zip) = 8Bzfhfall6pFJtaLXC5aNHCm8WK5a7mGevN/BZuW21o=
+SHA256 (go_modules/golang.org/x/term/@v/v0.45.0.mod) = YsP50aWuEiZKtI4siswIyfIBvalhneFxVE61cUt0rsU=
+SHA256 (go_modules/golang.org/x/term/@v/v0.45.0.zip) = s6OvaJcotDASrhshSCVxo1wDPFRemOAhYGvU/PH3kQg=
+SHA256 (go_modules/golang.org/x/text/@v/v0.40.0.mod) = 5p+wmxRjfKtiRaElPCLYASG3nZ/j3dGbVYYp4Vuq2vM=
+SHA256 (go_modules/golang.org/x/text/@v/v0.40.0.zip) = B3VzhHKND1LdhcHz8WNAXmOWsrD7frfWthDypxHWnuM=
 SHA256 (go_modules/gopkg.in/yaml.v3/@v/v3.0.1.mod) = IVeYYKIDBvz0OxvSNNH7oxlJnHdhG3HAX5vzupDauTk=
 SHA256 (go_modules/gopkg.in/yaml.v3/@v/v3.0.1.zip) = qrj7xOYwDqCOav4crqGKIckMefSJ9SxT4vIEMfGpoBU=
-SHA256 (librespeed-cli-v1.0.13.zip) = 48jTiQdHZQgbWsxbqC0+OE/ONU/qk60t0RryCCbjfDs=
-SIZE (go_modules/github.com/!burnt!sushi/toml/@v/v1.3.2.mod) = 43
-SIZE (go_modules/github.com/!burnt!sushi/toml/@v/v1.3.2.zip) = 303020
-SIZE (go_modules/github.com/briandowns/spinner/@v/v1.23.1.mod) = 291
-SIZE (go_modules/github.com/briandowns/spinner/@v/v1.23.1.zip) = 226880
-SIZE (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.4.mod) = 100
-SIZE (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.4.zip) = 13136
+SHA256 (librespeed-cli-v1.0.15.zip) = /dRzOHUh2Gp2ZGeASEX3W11/NKXLi63rCtOcZn4Erpc=
+SIZE (go_modules/github.com/!burnt!sushi/toml/@v/v1.5.0.mod) = 43
+SIZE (go_modules/github.com/!burnt!sushi/toml/@v/v1.5.0.zip) = 352548
+SIZE (go_modules/github.com/briandowns/spinner/@v/v1.23.2.mod) = 291
+SIZE (go_modules/github.com/briandowns/spinner/@v/v1.23.2.zip) = 226874
+SIZE (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.7.mod) = 100
+SIZE (go_modules/github.com/cpuguy83/go-md2man/v2/@v/v2.0.7.zip) = 15041
 SIZE (go_modules/github.com/fatih/color/@v/v1.17.0.mod) = 161
 SIZE (go_modules/github.com/fatih/color/@v/v1.17.0.zip) = 14693
 SIZE (go_modules/github.com/fatih/color/@v/v1.7.0.mod) = 30
@@ -66,31 +69,34 @@ SIZE (go_modules/github.com/mattn/go-isatty/@v/v0.0.16.mod) = 104
 SIZE (go_modules/github.com/mattn/go-isatty/@v/v0.0.20.mod) = 76
 SIZE (go_modules/github.com/mattn/go-isatty/@v/v0.0.20.zip) = 8936
 SIZE (go_modules/github.com/mattn/go-isatty/@v/v0.0.8.mod) = 95
-SIZE (go_modules/github.com/prometheus-community/pro-bing/@v/v0.8.0.mod) = 202
-SIZE (go_modules/github.com/prometheus-community/pro-bing/@v/v0.8.0.zip) = 36233
+SIZE (go_modules/github.com/prometheus-community/pro-bing/@v/v0.9.1.mod) = 202
+SIZE (go_modules/github.com/prometheus-community/pro-bing/@v/v0.9.1.zip) = 45751
 SIZE (go_modules/github.com/russross/blackfriday/v2/@v/v2.1.0.mod) = 42
 SIZE (go_modules/github.com/russross/blackfriday/v2/@v/v2.1.0.zip) = 122358
-SIZE (go_modules/github.com/urfave/cli/v2/@v/v2.27.4.mod) = 280
-SIZE (go_modules/github.com/urfave/cli/v2/@v/v2.27.4.zip) = 3552751
+SIZE (go_modules/github.com/urfave/cli/v2/@v/v2.27.7.mod) = 280
+SIZE (go_modules/github.com/urfave/cli/v2/@v/v2.27.7.zip) = 3553436
 SIZE (go_modules/github.com/xrash/smetrics/@v/v0.0.0-20240521201337-686a1a2994c1.mod) = 42
 SIZE (go_modules/github.com/xrash/smetrics/@v/v0.0.0-20240521201337-686a1a2994c1.zip) = 1826468
-SIZE (go_modules/golang.org/x/crypto/@v/v0.47.0.mod) = 192
-SIZE (go_modules/golang.org/x/crypto/@v/v0.47.0.zip) = 2224706
-SIZE (go_modules/golang.org/x/net/@v/v0.49.0.mod) = 157
-SIZE (go_modules/golang.org/x/net/@v/v0.49.0.zip) = 1934922
-SIZE (go_modules/golang.org/x/sync/@v/v0.19.0.mod) = 36
-SIZE (go_modules/golang.org/x/sync/@v/v0.19.0.zip) = 25714
+SIZE (go_modules/golang.org/x/crypto/@v/v0.54.0.mod) = 192
+SIZE (go_modules/golang.org/x/crypto/@v/v0.54.0.zip) = 2245311
+SIZE (go_modules/golang.org/x/net/@v/v0.56.0.mod) = 157
+SIZE (go_modules/golang.org/x/net/@v/v0.57.0.mod) = 157
+SIZE (go_modules/golang.org/x/net/@v/v0.57.0.zip) = 1870963
+SIZE (go_modules/golang.org/x/sync/@v/v0.21.0.mod) = 36
+SIZE (go_modules/golang.org/x/sync/@v/v0.21.0.zip) = 25686
 SIZE (go_modules/golang.org/x/sys/@v/v0.0.0-20220412211240-33da011f77ad.mod) = 33
 SIZE (go_modules/golang.org/x/sys/@v/v0.0.0-20220811171246-fbc7d0a398ab.mod) = 33
 SIZE (go_modules/golang.org/x/sys/@v/v0.18.0.mod) = 33
-SIZE (go_modules/golang.org/x/sys/@v/v0.40.0.mod) = 35
-SIZE (go_modules/golang.org/x/sys/@v/v0.40.0.zip) = 2003724
+SIZE (go_modules/golang.org/x/sys/@v/v0.46.0.mod) = 35
+SIZE (go_modules/golang.org/x/sys/@v/v0.47.0.mod) = 35
+SIZE (go_modules/golang.org/x/sys/@v/v0.50.0.mod) = 35
+SIZE (go_modules/golang.org/x/sys/@v/v0.50.0.zip) = 2120584
 SIZE (go_modules/golang.org/x/sys/@v/v0.6.0.mod) = 33
 SIZE (go_modules/golang.org/x/term/@v/v0.1.0.mod) = 95
-SIZE (go_modules/golang.org/x/term/@v/v0.39.0.mod) = 70
-SIZE (go_modules/golang.org/x/term/@v/v0.39.0.zip) = 21694
-SIZE (go_modules/golang.org/x/text/@v/v0.33.0.mod) = 190
-SIZE (go_modules/golang.org/x/text/@v/v0.33.0.zip) = 9236387
+SIZE (go_modules/golang.org/x/term/@v/v0.45.0.mod) = 70
+SIZE (go_modules/golang.org/x/term/@v/v0.45.0.zip) = 21694
+SIZE (go_modules/golang.org/x/text/@v/v0.40.0.mod) = 190
+SIZE (go_modules/golang.org/x/text/@v/v0.40.0.zip) = 7015236
 SIZE (go_modules/gopkg.in/yaml.v3/@v/v3.0.1.mod) = 95
 SIZE (go_modules/gopkg.in/yaml.v3/@v/v3.0.1.zip) = 104623
-SIZE (librespeed-cli-v1.0.13.zip) = 34032
+SIZE (librespeed-cli-v1.0.15.zip) = 60051
diff --git a/benchmarks/librespeed-cli/modules.inc b/benchmarks/librespeed-cli/modules.inc
index fded9be8820..58148543cc0 100644
--- a/benchmarks/librespeed-cli/modules.inc
+++ b/benchmarks/librespeed-cli/modules.inc
@@ -1,22 +1,22 @@
 MODGO_MODULES =	\
-	github.com/!burnt!sushi/toml			 v1.3.2 \
-	github.com/briandowns/spinner			 v1.23.1 \
-	github.com/cpuguy83/go-md2man/v2		 v2.0.4 \
+	github.com/!burnt!sushi/toml			 v1.5.0 \
+	github.com/briandowns/spinner			 v1.23.2 \
+	github.com/cpuguy83/go-md2man/v2		 v2.0.7 \
 	github.com/fatih/color				 v1.17.0 \
 	github.com/gocarina/gocsv			 v0.0.0-20240520201108-78e41c74b4b1 \
 	github.com/google/uuid				 v1.6.0 \
 	github.com/mattn/go-colorable			 v0.1.13 \
 	github.com/mattn/go-isatty			 v0.0.20 \
-	github.com/prometheus-community/pro-bing	 v0.8.0 \
+	github.com/prometheus-community/pro-bing	 v0.9.1 \
 	github.com/russross/blackfriday/v2		 v2.1.0 \
-	github.com/urfave/cli/v2			 v2.27.4 \
+	github.com/urfave/cli/v2			 v2.27.7 \
 	github.com/xrash/smetrics			 v0.0.0-20240521201337-686a1a2994c1 \
-	golang.org/x/crypto				 v0.47.0 \
-	golang.org/x/net				 v0.49.0 \
-	golang.org/x/sync				 v0.19.0 \
-	golang.org/x/sys				 v0.40.0 \
-	golang.org/x/term				 v0.39.0 \
-	golang.org/x/text				 v0.33.0 \
+	golang.org/x/crypto				 v0.54.0 \
+	golang.org/x/net				 v0.57.0 \
+	golang.org/x/sync				 v0.21.0 \
+	golang.org/x/sys				 v0.50.0 \
+	golang.org/x/term				 v0.45.0 \
+	golang.org/x/text				 v0.40.0 \
 	gopkg.in/yaml.v3				 v3.0.1
 
 MODGO_MODFILES =	\
@@ -24,9 +24,12 @@ MODGO_MODFILES =	\
 	github.com/mattn/go-colorable	 v0.1.2 \
 	github.com/mattn/go-isatty	 v0.0.16 \
 	github.com/mattn/go-isatty	 v0.0.8 \
+	golang.org/x/net		 v0.56.0 \
 	golang.org/x/sys		 v0.0.0-20220412211240-33da011f77ad \
 	golang.org/x/sys		 v0.0.0-20220811171246-fbc7d0a398ab \
 	golang.org/x/sys		 v0.18.0 \
+	golang.org/x/sys		 v0.46.0 \
+	golang.org/x/sys		 v0.47.0 \
 	golang.org/x/sys		 v0.6.0 \
 	golang.org/x/term		 v0.1.0
 
diff --git a/benchmarks/librespeed-cli/patches/patch-main_go b/benchmarks/librespeed-cli/patches/patch-main_go
index ff57fee0733..a836613e180 100644
--- a/benchmarks/librespeed-cli/patches/patch-main_go
+++ b/benchmarks/librespeed-cli/patches/patch-main_go
@@ -3,7 +3,7 @@ binding to specific interface is non-functional
 Index: main.go
 --- main.go.orig
 +++ main.go
-@@ -124,10 +124,6 @@ func main() {
+@@ -133,10 +133,6 @@ func main() {
  				Name:  defs.OptionSource,
  				Usage: "`SOURCE` IP address to bind to",
  			},
diff --git a/benchmarks/librespeed-cli/patches/patch-speedtest_speedtest_go b/benchmarks/librespeed-cli/patches/patch-speedtest_speedtest_go
index 2d2972e26ae..9a9c885ced6 100644
--- a/benchmarks/librespeed-cli/patches/patch-speedtest_speedtest_go
+++ b/benchmarks/librespeed-cli/patches/patch-speedtest_speedtest_go
@@ -1,7 +1,7 @@
 Index: speedtest/speedtest.go
 --- speedtest/speedtest.go.orig
 +++ speedtest/speedtest.go
-@@ -15,6 +15,8 @@ import (
+@@ -16,6 +16,8 @@ import (
  	"sync"
  	"time"
  
@@ -10,18 +10,7 @@ Index: speedtest/speedtest.go
  	"github.com/gocarina/gocsv"
  	"github.com/urfave/cli/v2"
  
-@@ -33,8 +35,8 @@ const (
- 	defaultTelemetryShare  = "/results/"
- 
- 	forceNothing = 0
--	forceHttps = 1
--	forceHttp = 2
-+	forceHttps   = 1
-+	forceHttp    = 2
- )
- 
- type PingJob struct {
-@@ -169,7 +171,24 @@ func SpeedTest(c *cli.Context) error {
+@@ -189,7 +191,24 @@ func SpeedTest(c *cli.Context) error {
  	transport.MaxIdleConnsPerHost = concurrent + 2
  	transport.MaxConnsPerHost = concurrent + 2
  

Re: [BUG-FIX] www/nextcloud 34

On 2026/09/30 10:38, Stuart Henderson wrote: > On 2026/09/30 09:49, Stuart Henderson wrote: > > moved to ports@; ports-bugs@openbsd.org is a dead mailing list > > > > doesn't nextcloud have integrity checks that get in the way of patching? > > I had an off-list reply (please, keep it on ports@): > > "On the port, these checks are disabled by default" Seems they are not actually, I had an email warning about this, the things it didn't like were: <cal:~>$ doas -u www /var/www/nextcloud/occ integrity:check-core - INVALID_HASH: - occ: - expected: 4570f9ef62d6c33442bcc96cafb75adb6e7046a5d0ddfc93e81726ac4e7bd76729cf6382c0bd22c5dfdc1bbcd46530c643a763fa7218cda33a6c938d1b8624b7 - current: 674177132e1b05b1836b1ac013237e988080a619aa4034167226b756549dd7fbcb5edb8109dfa21e36c840faa8cf898c31cf112e580aa0f5d8b80414f3fabf4e - EXTRA_FILE: - .htaccess.dist: - expected: - current: a2d827cbc4eadda3ccfdcfa9391985ee63808925b3cf2a43683296de37bcbcab4075ab0c49c3898ada1dd5f71100a673e166ce5f87c002bf18a3960385f7823c We can avoid these. The default PHP version installs /usr/local/bin/php so there's no need to change the #! in occ, and we can move the @sample source file out of the subdir. any objections to doing this? if it's ok, I'll backport to -stable too. Index: Makefile.inc =================================================================== RCS file: /cvs/ports/www/nextcloud/Makefile.inc,v diff -u -p -r1.6 Makefile.inc --- Makefile.inc 4 Jul 2026 10:05:10 -0000 1.6 +++ Makefile.inc 18 Nov 2026 17:50:11 -0000 @@ -63,14 +63,14 @@ RUN_DEPENDS += databases/pecl-redis${MO # image processing RUN_DEPENDS += graphics/pecl-imagick${MODPHP_FLAVOR} +# nextcloud has its own integrity checks; modifying installed files +# or creating additional files in the core/app-owned directories +# triggers warnings. try to avoid making any changes which trigger +# this. (occ integrity:check-core, occ integrity:check-apps --all). do-install: cp -Rp ${WRKSRC} ${INSTDIR} - mv ${INSTDIR}/.htaccess ${INSTDIR}/.htaccess.dist - find ${INSTDIR} -name '*.orig' -delete + mv ${INSTDIR}/.htaccess ${INSTDIR}.htaccess.dist chown -R ${BINOWN}:${BINGRP} ${INSTDIR} - - perl -pi -e '$$. == 1 && s|^.*env php.*$$|\#!${MODPHP_BIN}|;' \ - -e 'close ARGV if eof;' ${INSTDIR}/occ chmod +x ${INSTDIR}/occ ${INSTALL_DATA_DIR} ${PREFIX}/conf/modules.sample Index: 34/Makefile =================================================================== RCS file: /cvs/ports/www/nextcloud/34/Makefile,v diff -u -p -r1.11 Makefile --- 34/Makefile 15 Nov 2026 11:34:23 -0000 1.11 +++ 34/Makefile 18 Nov 2026 17:50:11 -0000 @@ -1,6 +1,6 @@ NC_VERSION= 34.0.6 # keep above 8.0-stable -REVISION= 0 +REVISION= 1 .include <bsd.port.mk> Index: 34/pkg/PLIST =================================================================== RCS file: /cvs/ports/www/nextcloud/34/pkg/PLIST,v diff -u -p -r1.8 PLIST --- 34/pkg/PLIST 15 Nov 2026 11:34:23 -0000 1.8 +++ 34/pkg/PLIST 18 Nov 2026 17:50:11 -0000 @@ -5,7 +5,7 @@ conf/modules.sample/ conf/modules.sample/apache-nextcloud.conf.dist @sample conf/modules.sample/apache-nextcloud.conf nextcloud/ -nextcloud/.htaccess.dist +nextcloud.htaccess.dist @sample nextcloud/.htaccess nextcloud/.user.ini nextcloud/3rdparty/ Index: 35/Makefile =================================================================== RCS file: /cvs/ports/www/nextcloud/35/Makefile,v diff -u -p -r1.5 Makefile --- 35/Makefile 15 Nov 2026 11:32:25 -0000 1.5 +++ 35/Makefile 18 Nov 2026 17:50:11 -0000 @@ -1,7 +1,7 @@ NC_VERSION= 35.0.3 # keep above 8.0-stable -REVISION= 0 +REVISION= 1 # override PORTROACH from ../Makefile.inc for the highest release branch only PORTROACH= Index: 35/pkg/PLIST =================================================================== RCS file: /cvs/ports/www/nextcloud/35/pkg/PLIST,v diff -u -p -r1.3 PLIST --- 35/pkg/PLIST 15 Nov 2026 11:32:25 -0000 1.3 +++ 35/pkg/PLIST 18 Nov 2026 17:50:11 -0000 @@ -4,7 +4,7 @@ conf/modules.sample/ conf/modules.sample/apache-nextcloud.conf.dist @sample conf/modules.sample/apache-nextcloud.conf nextcloud/ -nextcloud/.htaccess.dist +nextcloud.htaccess.dist @sample nextcloud/.htaccess nextcloud/.user.ini nextcloud/3rdparty/

net/haproxy 3.4.8 - >10 MAJOR, >100 MEDIUM/MINOR fixes

Update net/haproxy to 3.4.8 released today fixing - >10 MAJOR - lots of MEDIUM/MINOR bugs Changelog: https://www.haproxy.org/download/3.4/src/CHANGELOG Running fine on multiple machines. Index: Makefile =================================================================== RCS file: /cvs/ports/net/haproxy/Makefile,v retrieving revision 1.144 diff -u -p -r1.144 Makefile --- Makefile 8 Nov 2026 15:47:04 -0000 1.144 +++ Makefile 18 Nov 2026 17:20:22 -0000 @@ -1,10 +1,9 @@ COMMENT = reliable, high performance TCP/HTTP load balancer -DISTNAME = haproxy-3.4.7 +DISTNAME = haproxy-3.4.8 CATEGORIES = net www HOMEPAGE = https://www.haproxy.org MAINTAINER = Lucas Gabriel Vuotto <lucas@sexy.is> -REVISION = 0 # GPLv2 PERMIT_PACKAGE = Yes Index: distinfo =================================================================== RCS file: /cvs/ports/net/haproxy/distinfo,v retrieving revision 1.102 diff -u -p -r1.102 distinfo --- distinfo 15 Oct 2026 14:24:40 -0000 1.102 +++ distinfo 18 Nov 2026 17:20:22 -0000 @@ -1,2 +1,2 @@ -SHA256 (haproxy-3.4.7.tar.gz) = zP00Lzckws1llXQETXiPROILISRnxr1e27/FF6J3Xco= -SIZE (haproxy-3.4.7.tar.gz) = 5531506 +SHA256 (haproxy-3.4.8.tar.gz) = AkvVIabLvjK0JoUxR+a71qibEHxaNBRIRhk/7lt8btE= +SIZE (haproxy-3.4.8.tar.gz) = 5552944 -- Mark Patruck ( mark at wrapped.cx ) GPG key 0xF2865E51 / 187F F6D3 EE04 1DCE 1C74 F644 0D3C F66F F286 5E51 https://www.wrapped.cx