CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rsadow...@cvs.openbsd.org 2020/08/26 23:21:57 Modified files: audio/musique : Makefile distinfo Added files: audio/musique/patches: patch-src_mainwindow_cpp Log message: Update musique to 1.9 Remove the gzip hack and take maintainer.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rsadow...@cvs.openbsd.org 2020/08/26 23:19:17 Modified files: devel/qjson: Makefile devel/qjson/pkg: PLIST Log message: Switch qjson from qt4 to qt5
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ke...@cvs.openbsd.org 2020/08/26 19:00:46 Modified files: inputmethods/fcitx-qt5: Makefile distinfo inputmethods/fcitx-qt5/pkg: PLIST Log message: Update fcitx-qt5 to 1.2.5. "patch is good with me" bcallah@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ke...@cvs.openbsd.org 2020/08/26 18:59:48 Modified files: inputmethods/fcitx: Makefile distinfo inputmethods/fcitx/patches: patch-src_frontend_qt_fcitxwatcher_cpp inputmethods/fcitx/pkg: PLIST-main Log message: Update to fcitx-4.2.9.8. Input by bcallah@, "patch is good with me" bcallah@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/26 18:09:09 Removed files: devel/llvm/patches: patch-lib_CodeGen_SelectionDAG_DAGCombiner_cpp Log message: Zap local comment that wasn't picked up upstream >From Brad Smith.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: bcal...@cvs.openbsd.org 2020/08/26 16:44:48 Modified files: x11/worker : Makefile distinfo Removed files: x11/worker/patches: patch-src_aguix_aguix_cc patch-src_aguix_aguix_h patch-src_aguix_awindow_cc patch-src_worker_net_wm_icon_h Log message: Update to worker-4.5.1 This is a bugfix release that fixes the 32-bit build error I reported, so all the patches go away. Tested on amd64 and i386.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: fcam...@cvs.openbsd.org 2020/08/26 15:06:42 Modified files: www/slowhttptest: Makefile distinfo Removed files: www/slowhttptest/patches: patch-src_slowhttptest_cc Log message: Update slowhttptest to 1.8.2, and remove patch which was merged upstream.
Re: cvs diff patch with new directories
That's right. Thanks! A new diff coming soon then. ;) On Wed, Aug 26, 2020 at 4:54 PM Stuart Henderson wrote: > > patch -Ep0 > > On 26 August 2020 20:32:11 "Elias M. Mariani" wrote: > >> Hi ports@, >> Sorry for this silly question. >> I'm trying to generate a diff file for an upcoming x11/lxqt update. >> But if I use "cvs diff > file.diff" and then try to apply it with >> "patch < file.diff" the new directories (*/patches) are not generated >> and the files just... drop in the working directory. (why?) >> The problem goes away if I create the "patches" directories manually. >> Is there a secret flag on cvs diff or patch to generate those directories ? >> I tried looking in man cvs, rcsdiff and patch but I didn't find an >> answer. And my almost zero experience with cvs is not making this >> easier. :) >> >> Cheers. >> Elias mariani@ > >
Re: cvs diff patch with new directories
patch -Ep0 On 26 August 2020 20:32:11 "Elias M. Mariani" wrote: Hi ports@, Sorry for this silly question. I'm trying to generate a diff file for an upcoming x11/lxqt update. But if I use "cvs diff > file.diff" and then try to apply it with "patch < file.diff" the new directories (*/patches) are not generated and the files just... drop in the working directory. (why?) The problem goes away if I create the "patches" directories manually. Is there a secret flag on cvs diff or patch to generate those directories ? I tried looking in man cvs, rcsdiff and patch but I didn't find an answer. And my almost zero experience with cvs is not making this easier. :) Cheers. Elias mariani@
Re: cvs diff patch with new directories
On Wed, Aug 26, 2020 at 04:31:32PM -0300, Elias M. Mariani wrote: > But if I use "cvs diff > file.diff" and then try to apply it with > "patch < file.diff" the new directories (*/patches) are not generated > and the files just... drop in the working directory. (why?) See the documentation of patch's -p flag (in fact, patch defaults to -p 957). Try 'patch -p 0 < file.diff'.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: e...@cvs.openbsd.org2020/08/26 13:34:01 Modified files: sysutils/fzf : Makefile Log message: Add myself as MAINTAINER of fzf.
cvs diff patch with new directories
Hi ports@, Sorry for this silly question. I'm trying to generate a diff file for an upcoming x11/lxqt update. But if I use "cvs diff > file.diff" and then try to apply it with "patch < file.diff" the new directories (*/patches) are not generated and the files just... drop in the working directory. (why?) The problem goes away if I create the "patches" directories manually. Is there a secret flag on cvs diff or patch to generate those directories ? I tried looking in man cvs, rcsdiff and patch but I didn't find an answer. And my almost zero experience with cvs is not making this easier. :) Cheers. Elias mariani@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: e...@cvs.openbsd.org2020/08/26 13:29:30 Modified files: sysutils/fzf : Makefile distinfo Log message: Update fzf to version 0.22.0. OK abieber@ and paco@. Many thanks!
Re: rsync 3.2.3
On 2020-08-26, "Theo de Raadt" wrote: > I've encountered a few situations with _mostly offline machines_ where > the rsync package is need, and thus, 1 file can be copied over for > pkg_add. I don't think I'm the only person to encounter this situation. We could link rsync statically with some dependencies, or provide a static flavor that does this. The flip side is that somebody needs to remember to bump rsync every time one of those dependencies is updated, otherwise the installed rsync package will not be updated. -- Christian "naddy" Weisgerber na...@mips.inka.de
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: lan...@cvs.openbsd.org 2020/08/26 12:00:24 Modified files: mail/mozilla-thunderbird: Makefile distinfo mail/thunderbird-i18n: Makefile.inc distinfo Log message: Update to thunderbird 68.12.0. See https://www.thunderbird.net/en-US/thunderbird/68.12.0/releasenotes/ Same fixes as corresponding firefox esr: https://www.mozilla.org/en-US/security/advisories/mfsa2020-37/ Note that the switch to 78.x branch is imminent, pending openpgp being enabled by default upstream, hopefully for 78.2.1 per #1659536
Re: [Maintainer Update] www/zola 0.10.0 -> 0.11.0
Hi Sebastien, I removed Japanese support by modifying the Cargo.toml to add a language list, and regenerated the Cargo.lock via a `cargo update`. The issues I had were around the build hanging while lindera-ipaddic builds, similar to what happens in this issue https://github.com/getzola/zola/issues/1062. This solution matches the upstream fix for the problem in their "next" branch. Here's the commit: https://github.com/getzola/zola/commit/351c9b7ff34ed2872e0b915990219b6b4783642f It appears from the above issue thread that lindera may be modifying their build so that the dictionary is pre-calculated, which would take care of the memory-heavy build. Thanks, Peter On Wed, Aug 26, 2020 at 10:27 AM Sebastien Marie wrote: > Hi, > > First, sorry for the delay. > > Some comments inlined. > > > >>> On Mon, Jul 27, 2020 at 8:00 AM Peter Ezetta > > >>> wrote: > > >>> > > Hello ports@, > > > > Attached is an update for www/zola to version 0.11.0. The changelog > can > > be viewed here: > https://github.com/getzola/zola/releases/tag/v0.11.0 > > > > I was able to drop my patches to rustls-native-certs as this was > fixed > > upstream. > > > > I ran into issues with building lindera-ipadic, and wound up > picking up > > new patches to disable the dependency on it. Upstream is aware of > this > > patch, and has already disabled the dependency in their "next" > branch, so I > > should be able to drop these patches next release. > > How do you manage the disabling of lindera-ipadic ? It seems that the whole > Cargo.lock has been regenerated with new versions for crates, and not just > lindera-ipadic removed. > > Can you explain the issues you had ? Here the build of plain 0.11.0 (so > with > lindera-ipadic) is fine. > > > Should there be a note > > in the pkg readme mentioning that Japanese language searching has > been > > disabled? > > I think it is fine to not mention it in pkg readme. > > > Does anyone know if Ring is still broken on sparc64/aarch64? > > I think it is still broken. It needs patches to support openbsd. > > > All of the regression tests pass, and zola seems to be functioning > well > > for me. > > -- > Sebastien Marie > >
Re: Hacking cabal v2-build stuff together
Hi, On Thu, Aug 20, 2020 at 01:10:53PM -0700, Greg Steuck wrote: > Hope your summer is going well. I got devel/happy to build with cabal > v2-build but the lack of datadir is a show stopper (asking Gleb for > pointers). Still if you wanna take a look I pushed the current state of my > ports tree to: > https://github.com/blackgnezdo/ports/commit/74a8e81582027dbcaefd91f40ddbd45ce633d870 Sorry for slacking so much ;-) I hope to get some time next week to look at your recent work. Ciao, Kili
Re: [Maintainer Update] www/zola 0.10.0 -> 0.11.0
Hi, First, sorry for the delay. Some comments inlined. > >>> On Mon, Jul 27, 2020 at 8:00 AM Peter Ezetta > >>> wrote: > >>> > Hello ports@, > > Attached is an update for www/zola to version 0.11.0. The changelog can > be viewed here: https://github.com/getzola/zola/releases/tag/v0.11.0 > > I was able to drop my patches to rustls-native-certs as this was fixed > upstream. > > I ran into issues with building lindera-ipadic, and wound up picking up > new patches to disable the dependency on it. Upstream is aware of this > patch, and has already disabled the dependency in their "next" branch, > so I > should be able to drop these patches next release. How do you manage the disabling of lindera-ipadic ? It seems that the whole Cargo.lock has been regenerated with new versions for crates, and not just lindera-ipadic removed. Can you explain the issues you had ? Here the build of plain 0.11.0 (so with lindera-ipadic) is fine. > Should there be a note > in the pkg readme mentioning that Japanese language searching has been > disabled? I think it is fine to not mention it in pkg readme. > Does anyone know if Ring is still broken on sparc64/aarch64? I think it is still broken. It needs patches to support openbsd. > All of the regression tests pass, and zola seems to be functioning well > for me. -- Sebastien Marie
Re: UPDATE: fzf-0.22.0
On Tue, 25 Aug 2020, Edd Barrett wrote: > Hi, > > Here's an update to fzf-0.22.0. Seems to work fine here. Builds and works for me on amd64. > Removed NO_TEST since although there are no tests, the target does run > with success. So if tests appear later, we won't miss them. > > OK? ok paco. > > P.S. I'm semi-considering taking MAINTAINER on this. Would anyone mind? You already host the distfile so ... :-P > > Index: Makefile > === > RCS file: /cvs/ports/sysutils/fzf/Makefile,v > retrieving revision 1.6 > diff -u -p -r1.6 Makefile > --- Makefile 18 Jul 2020 11:32:27 - 1.6 > +++ Makefile 25 Aug 2020 18:34:41 - > @@ -2,8 +2,8 @@ > > COMMENT =command-line fuzzy finder > > -DISTNAME = fzf-0.21.1 > -REVISION = 0 > +# Note: we make our own vendored distfile. > +DISTNAME = fzf-0.22.0 > > CATEGORIES = sysutils > > @@ -15,15 +15,12 @@ PERMIT_PACKAGE = Yes > # uses pledge() > WANTLIB += c pthread > > -MASTER_SITES = https://deftly.net/dist/ > +MASTER_SITES = https://theunixzoo.co.uk/distfiles/ > > MODULES =lang/go > MODGO_TYPE = bin > > -NO_TEST =Yes > - > ALL_TARGET = github.com/junegunn/fzf > - > > # Note that unlike zsh and fish, bash has no well-defined site functions > # directory from which to autoload stuff. > Index: distinfo > === > RCS file: /cvs/ports/sysutils/fzf/distinfo,v > retrieving revision 1.4 > diff -u -p -r1.4 distinfo > --- distinfo 7 Apr 2020 08:32:50 - 1.4 > +++ distinfo 25 Aug 2020 18:25:19 - > @@ -1,2 +1,2 @@ > -SHA256 (fzf-0.21.1.tar.gz) = fiUVWVnr+d4pMs6YD8oCGExrujlnVTLkrPBBG2O7n9I= > -SIZE (fzf-0.21.1.tar.gz) = 2512770 > +SHA256 (fzf-0.22.0.tar.gz) = TTGoiXiIpNJMFRLNqkTmHPkwUCCiJ+elc9suTT4hq4E= > +SIZE (fzf-0.22.0.tar.gz) = 2517641 > > -- > Best Regards > Edd Barrett > > http://www.theunixzoo.co.uk > -- Paco Esteban. 0x5818130B8A6DBC03
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: tra...@cvs.openbsd.org 2020/08/26 10:59:56 Modified files: net/thingsd: Makefile distinfo Log message: Update to 3.0.3. ok jca@
Re: [Maintainer Update] www/zola 0.10.0 -> 0.11.0
Ping. On Thu, Aug 13, 2020 at 7:03 AM Peter Ezetta wrote: > Ping. Diff reattached for convenience. > > Thanks, > Peter > > On Thu, Aug 6, 2020 at 7:15 AM Peter Ezetta wrote: > >> Hi ports@, >> >> Is anyone able to provide an OK or a commit for this? I've had two users >> ping me off-list hoping to get this one in due to some config syntax >> changes included in 0.11.0. Regression tests pass, the application is >> working fine for me during functional testing on amd64. >> >> Thanks, >> Peter >> >> On Wed, Jul 29, 2020 at 6:20 PM Peter Ezetta >> wrote: >> >>> Hi ports@, >>> >>> After a bit more testing, I found another bug. Updated patch to www/zola >>> is attached. A patch for the bugfix has been submitted upstream. >>> >>> Comments? OK? >>> >>> Thanks, >>> Peter >>> >>> On Mon, Jul 27, 2020 at 8:00 AM Peter Ezetta >>> wrote: >>> Hello ports@, Attached is an update for www/zola to version 0.11.0. The changelog can be viewed here: https://github.com/getzola/zola/releases/tag/v0.11.0 I was able to drop my patches to rustls-native-certs as this was fixed upstream. I ran into issues with building lindera-ipadic, and wound up picking up new patches to disable the dependency on it. Upstream is aware of this patch, and has already disabled the dependency in their "next" branch, so I should be able to drop these patches next release. Should there be a note in the pkg readme mentioning that Japanese language searching has been disabled? Does anyone know if Ring is still broken on sparc64/aarch64? All of the regression tests pass, and zola seems to be functioning well for me. Comments? OK? Thanks, Peter >>>
[MAINTAINER UPDATE] net/thingsd 3.0.3
Hello, Here is a bump for net/thingsd to 3.0.3. Changelog: * fix for root required on kill clients, show packets, and set log levels * uncomment notification of root requirement in thingsctl * fix bad multi-subscription design and grammar related to subscribing to things * use the correct event for pausing accept-reserve when we're out of fds * fix inflight count * cleanup some dereference, free, and other tiny logic problems Ok? -- Tracey Emery Index: Makefile === RCS file: /cvs/ports/net/thingsd/Makefile,v retrieving revision 1.5 diff -u -p -u -p -r1.5 Makefile --- Makefile10 Aug 2020 16:16:15 - 1.5 +++ Makefile26 Aug 2020 15:11:53 - @@ -5,7 +5,7 @@ CATEGORIES =net GH_ACCOUNT = basepr1me GH_PROJECT = thingsd -GH_TAGNAME = 3.0.2 +GH_TAGNAME = 3.0.3 MAINTAINER = Tracey Emery Index: distinfo === RCS file: /cvs/ports/net/thingsd/distinfo,v retrieving revision 1.4 diff -u -p -u -p -r1.4 distinfo --- distinfo10 Aug 2020 16:16:15 - 1.4 +++ distinfo26 Aug 2020 15:11:53 - @@ -1,2 +1,2 @@ -SHA256 (thingsd-3.0.2.tar.gz) = 9OdlSxuCC8c3PnlXFHBWjzMPPrBkGTzIG8kfb0dX9Ms= -SIZE (thingsd-3.0.2.tar.gz) = 47892 +SHA256 (thingsd-3.0.3.tar.gz) = fnFCY77nbK/vLhkRaWAVxYqrtrvp/PJOsO/hyQztRfY= +SIZE (thingsd-3.0.3.tar.gz) = 48078
Re: rsync 3.2.3
Stuart Henderson wrote: > > I don't think we'll be able to make it default, Theo will object. > > If you want it, you have to convince *him*, not me. I've encountered a few situations with _mostly offline machines_ where the rsync package is need, and thus, 1 file can be copied over for pkg_add. I don't think I'm the only person to encounter this situation. My objection was that rsync suddenly needed a pile of dependencies, and it would take multiple steps to satisfy them all. A simple procedure becomes onerous. > > The problem is not onerousness, but rather Theo being scared of more > > compression algorithms, which tend to be rife with overflows... No, my concern was about the flood of hard to discern dependencies, and a semi-common installation workaround. > Oh, I guess the commit message was wrong then > > "Disable lz4, zstd and xxhash again > The dependencies are too big" > > lz4 has a reasonably good track record (not spotless but not rife > with overflows) and (unlike rsync's internal zlib, which is 1.2.8 plus > some rsync patch) is now extensively fuzzed > https://github.com/lz4/lz4/releases Maybe capabilities those make sense. It would be even better if they were not dependencies, but included as internal. I don't know how common that method is. > I am far more concerned about rsync itself which is already a labyrinth > of so many options that it's nigh on impossible to get wide test > coverage, but *shrug* there's no usable alternative for what it does, > I jail it, what more can I do? BTW, it would be great if openrsync could be used in such situations instead. The first 90% of the work is done in it, but the next 10% towards making it fully capable turns out to be another 90% task, and I guess we are waiting for a wave of developers to step in there improve those issues. openrsync would satisfy a majority of use cases. History shows us we often have bad outcomes when there is only one implementation of a tool.
Re: NEW: security/libmodsecurity
On 24.08.2020 - 16:19:06, Stuart Henderson wrote: > On 2020/08/24 14:28, Matthias Pitzl wrote: > > Hello, > > > > This new port is needed for adding ModSecurity support to nginx. > > > > Excerpt from github: > > ModSecurity is an open source, cross platform web application firewall > > (WAF) > > engine for Apache, IIS and Nginx that is developed by Trustwave's > > SpiderLabs. > > It has a robust event-based programming language which provides protection > > from > > a range of attacks against web applications and allows for HTTP traffic > > monitoring, logging and real-time analys... > > > > I will send an updated diff for adding the libmodsecurity connector to nginx > > too. > > > > I've sent this port a few weeks ago but haven't got an OK or further > > feedback > > except from Stuart. > > > > Greetings, > > Matthias > > Diff below with the following: > > - omit SEPARATE_BUILD=no, it's the default > - -std=c++11 is used in build so base-gcc doesn't work > - some things are compiled with -O3 [...] $CFLAGS, remove -O3 > - regen plist > > ... also, this (not addressed in my diff) - there's "checking for PCRE > JIT... yes" - I think this it should not be enabling this, IIRC pcre's > jit isn't W^X friendly - looks like pcre maybe giving the wrong answer > when the autoconf check is testing for this. Hi, The attached version contains the diff of Stuart. In addition I made a patch for configure which disables the PCRE JIT by disabling the whole check for it. A make test did not find any problems. Greetings, Matthias security_libmodsecurity.tgz Description: application/gzip
archiver/zstd: don't rely on COMPILER
I have the diff below sitting in my tree for a couple of weeks:
As sthen@ explained [0]: "COMPILER=base-clang ports-gcc" is needed by
archiver/zstd because of -Wvla in a couple of Makefiles. Fortunate this
is easily patched out. While here:
- switch ALL_TARGET to 'default' as 'allmost' builds stuff that we do
not package. This saves a couple of seconds while building;
- No need for WANTLIB+=atomic/LDFLAGS+=-latomic (hppa only). There is
nothing related to atomics in the stuff that is build (related to
ALL_TARGET);
- No need to set CXX* in MAKE_ENV as the build is C-only;
- No need to set CFLAGS. Upstream's (GNU)Makefile properly amends CFLAGS
on a per-target basis. We already provide CFLAGS through MAKE_ENV, let
the Makefile do its job;
- Update CPPFLAGS.
Builds successfully using either clang or gcc4. All tests are OK.
This diff needs to be tested on hppa. Unfortunately I have no access to
this arch. Anyone willing to test?
Comments/OK?
[0] https://marc.info/?l=openbsd-ports=159844258927938=2
diff --git Makefile Makefile
index 231f0a93d6c..782ee6b9655 100644
--- Makefile
+++ Makefile
@@ -3,6 +3,7 @@
COMMENT = zstandard fast real-time compression algorithm
V =1.4.5
+REVISION = 0
DISTNAME = zstd-${V}
SHARED_LIBS = zstd3.2 # 1.4.5
@@ -24,29 +25,19 @@ LIB_DEPENDS = archivers/lz4 \
archivers/xz
BUILD_DEPENDS =sysutils/ggrep
-COMPILER = base-clang ports-gcc
-
MAKE_ENV = CC="${CC}" \
- CFLAGS="${CFLAGS}" \
- CPPFLAGS="-I${WRKSRC}/lib -I${WRKSRC}/lib/common
-I${LOCALBASE}/include" \
- CXX="${CXX}" \
- CXXFLAGS="${CXXFLAGS}" \
+ CPPFLAGS="-I${LOCALBASE}/include" \
LDFLAGS="${LDFLAGS} -L${LOCALBASE}/lib"
MAKE_FLAGS = SHARED_EXT_VER="so.$(LIBzstd_VERSION)" \
SONAME_FLAGS= \
V=1
FAKE_FLAGS = PREFIX="${PREFIX}"
-.if ${MACHINE_ARCH:Mhppa}
-WANTLIB += atomic
-LDFLAGS += -latomic
-.endif
-
USE_GMAKE =Yes
DEBUG_PACKAGES = ${BUILD_PACKAGES}
-ALL_TARGET = allmost
+ALL_TARGET = default
TEST_TARGET = check
pre-build:
diff --git patches/patch-lib_Makefile patches/patch-lib_Makefile
index 48a5c2d802a..9d9f7f90115 100644
--- patches/patch-lib_Makefile
+++ patches/patch-lib_Makefile
@@ -1,8 +1,21 @@
$OpenBSD: patch-lib_Makefile,v 1.8 2020/05/28 19:31:56 bket Exp $
+- Some compilers don't have -Wvla
+- Don't set compiler optimisation level
+- Don't create symbolic links to SHARED_LIB
+
Index: lib/Makefile
--- lib/Makefile.orig
+++ lib/Makefile
+@@ -38,7 +38,7 @@ endif
+ DEBUGFLAGS= -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \
+ -Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement \
+ -Wstrict-prototypes -Wundef -Wpointer-arith \
+--Wvla -Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \
++-Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \
+ -Wredundant-decls -Wmissing-prototypes -Wc++-compat
+ CFLAGS += $(DEBUGFLAGS) $(MOREFLAGS)
+ FLAGS= $(CPPFLAGS) $(CFLAGS)
@@ -81,7 +81,7 @@ ifneq ($(ZSTD_LIB_MINIFY), 0)
CFLAGS += -fno-stack-protector -fomit-frame-pointer -fno-ident \
-DDYNAMIC_BMI2=0 -DNDEBUG
diff --git patches/patch-programs_Makefile patches/patch-programs_Makefile
new file mode 100644
index 000..d0dda7f7233
--- /dev/null
+++ patches/patch-programs_Makefile
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Some compilers don't have -Wvla
+
+Index: programs/Makefile
+--- programs/Makefile.orig
programs/Makefile
+@@ -51,7 +51,7 @@ CFLAGS ?= -O3
+ DEBUGFLAGS+=-Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \
+ -Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement \
+ -Wstrict-prototypes -Wundef -Wpointer-arith \
+--Wvla -Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \
++-Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \
+ -Wredundant-decls -Wmissing-prototypes -Wc++-compat
+ CFLAGS += $(DEBUGFLAGS) $(MOREFLAGS)
+ FLAGS= $(CPPFLAGS) $(CFLAGS) $(LDFLAGS)
diff --git patches/patch-tests_Makefile patches/patch-tests_Makefile
new file mode 100644
index 000..fe60f9fb452
--- /dev/null
+++ patches/patch-tests_Makefile
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+Some compilers don't have -Wvla
+
+Index: tests/Makefile
+--- tests/Makefile.orig
tests/Makefile
+@@ -35,7 +35,7 @@ CFLAGS ?= -O3
+ CFLAGS += -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow
\
+ -Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement
\
+ -Wstrict-prototypes -Wundef
\
+- -Wvla -Wformat=2
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/26 06:59:29 Modified files: graphics/darktable: Makefile Log message: darktable: update ONLY_FOR_ARCHS following version update; upstream only supports amd64, aarch64, powerpc64
Re: rsync 3.2.3
On 2020/08/26 14:03, Marc Espie wrote:
> On Wed, Aug 26, 2020 at 12:48:27PM +0100, Stuart Henderson wrote:
> > This updates rsync to 3.2.3, which amongst other things includes
> > https://github.com/WayneD/rsync/commit/9375a8c4c2b79a9d44ceae00de7284b77d79033c
> > which greatly reduces maxrss (https://github.com/WayneD/rsync/issues/81).
> >
> > Also adds a patch for the --rsh problem I ran into (issue open upstream).
> >
> > OK?
>
> Okay.
>
>
> > Not included in the diff, but also I would like to reopen the discussion
> > about dependencies. In order to lower resource use for CVS repository
> > mirrors (7.3G of data across 450K files) I would like to have xxhash
> > and lz4 enabled in the default ports build.
> >
> > xxhash and lz4 are nearly self-contained ports, they use the base
> > compiler, they use gmake but otherwise add no build dependencies above
> > what's already needed for rsync, and are available on all archs we
> > build packages for.
> >
> > xxhash is a single C file, takes ~10 seconds to build, the package
> > weighs in at 200K.
> >
> > lz4 is a little larger, 4 C files in the library, and there are tools
> > too so it takes more like 45 seconds to build, the package is 300K.
> >
> > These are not onerous requirements.
> >
> > zstd is a bit different, it uses "COMPILER=base-clang ports-gcc".
> > Seems this is needed for -Wvla which is used in the Makefile (easily
> > patched), and maybe for something related to atomics on certain archs
> > (hppa and powerpc are the usual suspects for this). It would be nice
> > to have zstd as well (it's an excellent compression tool and nicely
> > licensed) but that wants more work first so I am not proposing that at
> > this point.
>
> I don't think we'll be able to make it default, Theo will object.
> If you want it, you have to convince *him*, not me.
>
> The problem is not onerousness, but rather Theo being scared of more
> compression algorithms, which tend to be rife with overflows...
Oh, I guess the commit message was wrong then
"Disable lz4, zstd and xxhash again
The dependencies are too big"
lz4 has a reasonably good track record (not spotless but not rife
with overflows) and (unlike rsync's internal zlib, which is 1.2.8 plus
some rsync patch) is now extensively fuzzed https://github.com/lz4/lz4/releases
I am far more concerned about rsync itself which is already a labyrinth
of so many options that it's nigh on impossible to get wide test
coverage, but *shrug* there's no usable alternative for what it does,
I jail it, what more can I do?
> rsync is cheap enough that more flavors is not really an issue.
>
It is an issue, after being presented with an overwhelming list of
various flavours people will tend to pick the default option, so mirror
operators won't benefit from clients having access to cheaper algorithms.
FWIW this is the diff I have locally,
Index: Makefile
===
RCS file: /cvs/ports/net/rsync/Makefile,v
retrieving revision 1.89
diff -u -p -r1.89 Makefile
--- Makefile26 Aug 2020 12:40:39 - 1.89
+++ Makefile26 Aug 2020 12:42:56 -
@@ -5,6 +5,7 @@ COMMENT = mirroring/synchronization over
DISTNAME = rsync-3.2.3
CATEGORIES = net
HOMEPAGE = https://rsync.samba.org/
+REVISION = 0
MAINTAINER = Marc Espie
@@ -14,27 +15,32 @@ FLAVOR ?=
# GPLv3
PERMIT_PACKAGE = Yes
-WANTLIB = c crypto
+WANTLIB = c crypto lz4 xxhash
MASTER_SITES = https://rsync.samba.org/ftp/rsync/src/ \
https://ftp.funet.fi/pub/mirrors/samba.org/pub/rsync/src/
SEPARATE_BUILD =Yes
CONFIGURE_STYLE =gnu
-CONFIGURE_ARGS =--disable-lz4 \
- --disable-xxhash \
- --disable-zstd \
+CONFIGURE_ARGS = --disable-zstd \
--with-included-popt \
--with-included-zlib \
--with-rsyncd-conf="${SYSCONFDIR}/rsyncd.conf" \
--with-rsh=/usr/bin/ssh \
--with-nobody-group=_rsync
-.if ${FLAVOR:Miconv}
-CONFIGURE_ENV +=CPPFLAGS='-I${LOCALBASE}/include' \
+CONFIGURE_ENV += CPPFLAGS='-I${LOCALBASE}/include' \
LDFLAGS='-L${LOCALBASE}/lib'
+
+LIB_DEPENDS += archivers/lz4 \
+ sysutils/xxhash
+
+.if ${FLAVOR:Miconv}
LIB_DEPENDS += converters/libiconv
WANTLIB += iconv
+.else
+# --disable-iconv-open only covers iconv_open not libiconv_open
+CONFIGURE_ENV += ac_cv_func_libiconv_open=false
.endif
DOCDIR = ${PREFIX}/share/doc/rsync
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/26 06:40:39 Modified files: net/rsync : Makefile Log message: https for MASTER_SITES
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2020/08/26 06:15:56 Modified files: net/rsync : Makefile distinfo net/rsync/patches: patch-authenticate_c patch-configure_sh patch-rsync_1 patch-rsyncd_conf_5 patch-support_rrsync Added files: net/rsync/patches: patch-compat_c Log message: update to rsync-3.2.3, patch for overflow in read_varint() ok espie
Re: rsync 3.2.3
On Wed, Aug 26, 2020 at 12:48:27PM +0100, Stuart Henderson wrote: > This updates rsync to 3.2.3, which amongst other things includes > https://github.com/WayneD/rsync/commit/9375a8c4c2b79a9d44ceae00de7284b77d79033c > which greatly reduces maxrss (https://github.com/WayneD/rsync/issues/81). > > Also adds a patch for the --rsh problem I ran into (issue open upstream). > > OK? Okay. > Not included in the diff, but also I would like to reopen the discussion > about dependencies. In order to lower resource use for CVS repository > mirrors (7.3G of data across 450K files) I would like to have xxhash > and lz4 enabled in the default ports build. > > xxhash and lz4 are nearly self-contained ports, they use the base > compiler, they use gmake but otherwise add no build dependencies above > what's already needed for rsync, and are available on all archs we > build packages for. > > xxhash is a single C file, takes ~10 seconds to build, the package > weighs in at 200K. > > lz4 is a little larger, 4 C files in the library, and there are tools > too so it takes more like 45 seconds to build, the package is 300K. > > These are not onerous requirements. > > zstd is a bit different, it uses "COMPILER=base-clang ports-gcc". > Seems this is needed for -Wvla which is used in the Makefile (easily > patched), and maybe for something related to atomics on certain archs > (hppa and powerpc are the usual suspects for this). It would be nice > to have zstd as well (it's an excellent compression tool and nicely > licensed) but that wants more work first so I am not proposing that at > this point. I don't think we'll be able to make it default, Theo will object. If you want it, you have to convince *him*, not me. The problem is not onerousness, but rather Theo being scared of more compression algorithms, which tend to be rife with overflows... rsync is cheap enough that more flavors is not really an issue.
rsync 3.2.3
This updates rsync to 3.2.3, which amongst other things includes
https://github.com/WayneD/rsync/commit/9375a8c4c2b79a9d44ceae00de7284b77d79033c
which greatly reduces maxrss (https://github.com/WayneD/rsync/issues/81).
Also adds a patch for the --rsh problem I ran into (issue open upstream).
OK?
Not included in the diff, but also I would like to reopen the discussion
about dependencies. In order to lower resource use for CVS repository
mirrors (7.3G of data across 450K files) I would like to have xxhash
and lz4 enabled in the default ports build.
xxhash and lz4 are nearly self-contained ports, they use the base
compiler, they use gmake but otherwise add no build dependencies above
what's already needed for rsync, and are available on all archs we
build packages for.
xxhash is a single C file, takes ~10 seconds to build, the package
weighs in at 200K.
lz4 is a little larger, 4 C files in the library, and there are tools
too so it takes more like 45 seconds to build, the package is 300K.
These are not onerous requirements.
zstd is a bit different, it uses "COMPILER=base-clang ports-gcc".
Seems this is needed for -Wvla which is used in the Makefile (easily
patched), and maybe for something related to atomics on certain archs
(hppa and powerpc are the usual suspects for this). It would be nice
to have zstd as well (it's an excellent compression tool and nicely
licensed) but that wants more work first so I am not proposing that at
this point.
Index: Makefile
===
RCS file: /cvs/ports/net/rsync/Makefile,v
retrieving revision 1.87
diff -u -p -r1.87 Makefile
--- Makefile8 Aug 2020 21:53:46 - 1.87
+++ Makefile26 Aug 2020 11:24:23 -
@@ -2,8 +2,7 @@
COMMENT = mirroring/synchronization over low bandwidth links
-DISTNAME = rsync-3.2.2
-REVISION = 0
+DISTNAME = rsync-3.2.3
CATEGORIES = net
HOMEPAGE = https://rsync.samba.org/
Index: distinfo
===
RCS file: /cvs/ports/net/rsync/distinfo,v
retrieving revision 1.30
diff -u -p -r1.30 distinfo
--- distinfo3 Aug 2020 21:35:36 - 1.30
+++ distinfo26 Aug 2020 11:24:23 -
@@ -1,2 +1,2 @@
-SHA256 (rsync-3.2.2.tar.gz) = ZEvThBd5UHZlIR/X24NZyKEGcMV+MFtKq2G05AA3r6g=
-SIZE (rsync-3.2.2.tar.gz) = 1057001
+SHA256 (rsync-3.2.3.tar.gz) = vsw8UEzupJn0FnomAEDM9Nny75SZrVaDwXmmlxRs5Q4=
+SIZE (rsync-3.2.3.tar.gz) = 1069784
Index: patches/patch-authenticate_c
===
RCS file: /cvs/ports/net/rsync/patches/patch-authenticate_c,v
retrieving revision 1.7
diff -u -p -r1.7 patch-authenticate_c
--- patches/patch-authenticate_c3 Aug 2020 21:35:36 - 1.7
+++ patches/patch-authenticate_c26 Aug 2020 11:24:23 -
@@ -1,4 +1,5 @@
-$OpenBSD: patch-authenticate_c,v 1.7 2020/08/03 21:35:36 kn Exp $
+$OpenBSD$
+
Index: authenticate.c
--- authenticate.c.orig
+++ authenticate.c
Index: patches/patch-compat_c
===
RCS file: patches/patch-compat_c
diff -N patches/patch-compat_c
--- /dev/null 1 Jan 1970 00:00:00 -
+++ patches/patch-compat_c 26 Aug 2020 11:24:23 -
@@ -0,0 +1,16 @@
+$OpenBSD$
+
+https://github.com/WayneD/rsync/issues/84
+
+Index: compat.c
+--- compat.c.orig
compat.c
+@@ -705,7 +705,7 @@ void setup_protocol(int f_out,int f_in)
+ do_negotiated_strings = 1;
+ compat_flags |= CF_VARINT_FLIST_FLAGS;
+ }
+- if (strchr(client_info, 'V') != NULL) { /* Support a
pre-release 'V' that got superseded */
++ if (!local_server && strchr(client_info, 'V') != NULL)
{ /* Support a pre-release 'V' that got superseded */
+ if (!write_batch)
+ compat_flags |= CF_VARINT_FLIST_FLAGS;
+ write_byte(f_out, compat_flags);
Index: patches/patch-configure_sh
===
RCS file: /cvs/ports/net/rsync/patches/patch-configure_sh,v
retrieving revision 1.5
diff -u -p -r1.5 patch-configure_sh
--- patches/patch-configure_sh 3 Aug 2020 21:35:36 - 1.5
+++ patches/patch-configure_sh 26 Aug 2020 11:24:23 -
@@ -1,8 +1,9 @@
-$OpenBSD: patch-configure_sh,v 1.5 2020/08/03 21:35:36 kn Exp $
+$OpenBSD$
+
Index: configure.sh
--- configure.sh.orig
+++ configure.sh
-@@ -4901,7 +4901,7 @@ fi
+@@ -4908,7 +4908,7 @@ fi
cat >>confdefs.h <<_ACEOF
Index: patches/patch-rsync_1
===
RCS file: /cvs/ports/net/rsync/patches/patch-rsync_1,v
retrieving revision 1.7
diff -u -p -r1.7 patch-rsync_1
--- patches/patch-rsync_1 3 Aug 2020 21:35:36 - 1.7
+++ patches/patch-rsync_1 26 Aug 2020 11:24:23 -
Re: NEW: dnspy - .NET/C# disassembler
On Sun, Aug 23, 2020 at 11:49:45PM -0600, Thomas Frohwein wrote: > Hi, > > Attached is a port of dnSpy, one of the better known .NET/C# > decompilers/assembly editors. As we don't have one available so far, > this would be a useful addition. (mono ships with monodis and ikdasm, > but they only create CIL (common intermediate language)). Sorry, new tarball attached. The previous one accidentally created distinfo from a prior version that was in my /usr/ports/distfiles. (unfortunately upstream doesn't version its distfiles) Thanks bentley@ for catching this. > > The port ships only the console version. I have tested it with a > few .NET assemblies with good results. The GUI binaries don't run as > PresentationCore is missing (for now). > > Similar to nuget and depotdownloader, this port also uses the compiled > form rather than building directly from source. This is because xbuild > and msbuild fail; it seems some Visual Studio Code components are > missing for a build. > > ok to import? dnspy.tgz Description: Binary data
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: j...@cvs.openbsd.org2020/08/26 04:33:47 Modified files: devel/llvm : Makefile devel/llvm/patches: patch-lib_Target_PowerPC_PPCISelLowering_cpp Removed files: devel/llvm/patches: patch-lib_Target_PowerPC_PPCSubtarget_h Log message: Sync with base Zap outdated powerpc alignement fix, fixed upstream in a different way. >From Brad Smith.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: lan...@cvs.openbsd.org 2020/08/26 01:37:39 Modified files: meta/tor-browser: Makefile www/tor-browser: Makefile.inc www/tor-browser/browser: Makefile distinfo www/tor-browser/browser/patches: patch-config_makefiles_rust_mk www/tor-browser/https-everywhere: Makefile distinfo www/tor-browser/noscript: Makefile distinfo Log message: Update to tor-browser 9.5.4 See https://blog.torproject.org/new-release-tor-browser-954 from MAINTAINER Caspar Schutijser
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 01:04:35 Modified files: net/py-boto3 : Makefile distinfo Log message: Update to py3-boto3-1.14.48.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 01:04:21 Modified files: net/py-botocore: Makefile distinfo Log message: Update to py3-botocore-1.17.48.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 01:04:55 Modified files: sysutils/awscli: Makefile distinfo sysutils/awscli/pkg: PLIST Log message: Update to awscli-1.18.125.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 01:02:34 Modified files: x11/gnome/tracker-miners: Makefile distinfo x11/gnome/tracker-miners/pkg: PLIST Log message: Update to tracker-miners-2.3.4.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ki...@cvs.openbsd.org 2020/08/26 00:57:33 Modified files: www/linkchecker: Makefile distinfo www/linkchecker/patches: patch-setup_py Added files: www/linkchecker/pkg: DESCR PLIST Removed files: www/linkchecker/patches: patch-linkcheck___init___py www/linkchecker/pkg: DESCR-main DESCR-x11 PLIST-main PLIST-x11 Log message: Update to linkchecker-9.4.0. GUI is not maintained upstream and was removed from the port. One less Qt4 consumer. Requested, tweaks and OK rsadowski@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 00:50:14 Modified files: x11/gnome/tracker: Makefile distinfo x11/gnome/tracker/pkg: PLIST Log message: Update to (meta-)tracker-2.3.5.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2020/08/26 00:46:53 Modified files: print/cups-filters: Makefile distinfo print/cups-filters/pkg: PLIST Added files: print/cups-filters/patches: patch-utils_driverless-fax Log message: Update to cups-filters-1.28.0.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rsadow...@cvs.openbsd.org 2020/08/26 00:11:52 Modified files: devel/jenkins/devel: Makefile distinfo Log message: Update jenkins-devel to 2.254
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: k...@cvs.openbsd.org2020/08/26 00:09:21 Modified files: lang/python/3.7/pkg: Tag: OPENBSD_6_7 PLIST-idle PLIST-main Log message: Hopefully last bit of fallout doing the -stable commit. Most of this caused because we didn't do 3.7.8 for -stable while we did for -current so the patch missed some things. Pointed out again by Ian McWilliam. Used his diff with some slight PLIST tweaks.
Re: Spam: Re: OpenBSD 6.7-stable Python 3.7 update
On Wed, Aug 26, 2020 at 02:00:42AM +, Ian McWilliam wrote: > Package errors. Patch attached. Alright. I checked out a ports tree on a 6.7 machine and checked it this time. Your patch would work, but this is the actual result of regenerating the PLIST. --Kurt Index: pkg/PLIST-idle === RCS file: /cvs/ports/lang/python/3.7/pkg/PLIST-idle,v retrieving revision 1.8.2.1 diff -u -p -r1.8.2.1 PLIST-idle --- pkg/PLIST-idle 24 Aug 2020 15:01:20 - 1.8.2.1 +++ pkg/PLIST-idle 26 Aug 2020 06:06:02 - @@ -17,7 +17,6 @@ lib/python3.7/idlelib/ChangeLog lib/python3.7/idlelib/HISTORY.txt lib/python3.7/idlelib/Icons/ lib/python3.7/idlelib/Icons/folder.gif -lib/python3.7/idlelib/Icons/idle.icns lib/python3.7/idlelib/Icons/idle.ico lib/python3.7/idlelib/Icons/idle_16.gif lib/python3.7/idlelib/Icons/idle_16.png Index: pkg/PLIST-main === RCS file: /cvs/ports/lang/python/3.7/pkg/PLIST-main,v retrieving revision 1.12.2.1 diff -u -p -r1.12.2.1 PLIST-main --- pkg/PLIST-main 24 Aug 2020 15:01:20 - 1.12.2.1 +++ pkg/PLIST-main 26 Aug 2020 06:06:03 - @@ -1798,8 +1798,8 @@ lib/python3.7/ensurepip/__pycache__/_uni lib/python3.7/ensurepip/__pycache__/_uninstall.cpython-37.opt-2.pyc lib/python3.7/ensurepip/__pycache__/_uninstall.cpython-37.pyc lib/python3.7/ensurepip/_bundled/ -lib/python3.7/ensurepip/_bundled/pip-19.2.3-py2.py3-none-any.whl -lib/python3.7/ensurepip/_bundled/setuptools-41.2.0-py2.py3-none-any.whl +lib/python3.7/ensurepip/_bundled/pip-20.1.1-py2.py3-none-any.whl +lib/python3.7/ensurepip/_bundled/setuptools-47.1.0-py3-none-any.whl lib/python3.7/ensurepip/_uninstall.py lib/python3.7/enum.py lib/python3.7/filecmp.py @@ -1854,6 +1854,7 @@ lib/python3.7/http/client.py lib/python3.7/http/cookiejar.py lib/python3.7/http/cookies.py lib/python3.7/http/server.py +lib/python3.7/idlelib/Icons/README.txt lib/python3.7/imaplib.py lib/python3.7/imghdr.py lib/python3.7/imp.py
