Age | Commit message (Collapse) | Author |
|
different versions of git-bash come with different msys library that behave
slightly differently when it comes to the path mangling that occurs. When msys
encounters something that looks like a path, but the path doesn't exist, then
it prefixes that path using the home of the git-bash installation. That path is
using the long form and since git-bash usually is installed to Program Files
the build then breaks because the path arguments aren't quoted.
The path-mangling can be prevented by starting the path with a double-slash, so
--prefix=//usr/local will take care of those instances where git-bash from
Visual Studio works, but git-bash installed via winget fails.
Luckily that seems to be the only case where the differences actually show
themselves.
Change-Id: I702da05b34dc153267f7251810a1c93babf66022
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/181061
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
Change-Id: Ibffac03be5f22852e647340ddc4f5211d0159c4f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180997
Tested-by: Jenkins
Reviewed-by: Andras Timar <andras.timar@collabora.com>
|
|
There is no point to bundle it, because
1) it is an old version
2) since LibreOffice 7.5 there are built-in number formats in
Calc with the same functionality (using libnumbertext FWIW)
Change-Id: I8cfb24f7e6f045bd51fea620aca6a61269dba77a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180996
Tested-by: Jenkins
Reviewed-by: Andras Timar <andras.timar@collabora.com>
|
|
This reverts commit 3500c44bccafd442e50da604287342708506bb5c.
Reason for revert: At least for Linux --without-system-cairo builds, this causes
workdir/UnpackedTarball/cairo/src/.libs/libcairo.so (which is the lib that other
libs like Library_drawinglayer and Library_vcl will link against) to have an
SONAME of libcairo.so.2 (instead of libcairo-lo.so.2), so at runtime libs like
instdir/program/libdrawinglayerlo.so and instdir/program/libvcllo.so will have a
NEEDED of libcairo.so.2 (instead of libcairo-lo.so.2), so will pick up whatever
system libcairo.so.2 (/usr/lib64/libcairo.so.2 from cairo-1.18.2-2.fc41.x86_64,
in my case) instead of instdir/program/libcairo-lo.so.2, which caused e.g.
CppunitTest_vcl_bitmap_render_test
CPPUNIT_TEST_NAME=BitmapRenderTest::testTdf113918 to happen to fail for me with
> [_RUN_____] BitmapRenderTest::testTdf113918
> warn:vcl.gdi:2234659:2234659:vcl/headless/CairoCommon.cxx:1619: unsupported SvpSalGraphics::drawAlphaBitmap case
> vcl/qa/cppunit/bitmaprender/BitmapRenderTest.cxx:100:BitmapRenderTest::testTdf113918
> equality assertion failed
> - Expected: rgba[ffffffff]
> - Actual : rgba[008000ff]
Change-Id: I9a2a64197185b7c057352c3fb7f3e3f6dbcf70db
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180884
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
This was introduced by
commit 2def5485aa57d7c407a84ecc73b22083579e9a98
Author: Tor Lillqvist <tml@collabora.com>
Date: Tue Sep 20 16:07:14 2022 +0300
Enable opening of downloaded fonts only in ForKit in Online
but we don't need to this, since we rename the .so file when we copy it
in external/cairo/ExternalPackage_cairo.mk
This change makes a later patch less noisy/
Change-Id: I06f78d7ff068df14e95542391e20b669487ebb5b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180794
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
In order to fix
ERROR: Could not install packages due to an OSError: Could not find a suitable TLS CA certificate bundle, invalid path:
/home/xisco/libreoffice/instdir/program/python-core-3.11.11/lib/pip/_vendor/certifi/cacert.pem
when using pip install
Change-Id: I84d6c23021523c11803bc2b0858d69a706da0dd0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180788
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Otherwise libraries installed through pip do not work
Change-Id: Ib916964ce7d309faffaad1763884b1600b139c18
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180781
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
on branches 25-2 and 24-8 the python tkinter dylib is back.
The corresponding python module is ignored already, and the tkinter
cpython dylib already was removed in the past, but apparently at some
point that file wasn't generated anymore and the corresponding line to
delete it was removed in 6e62c9df969fdf3a599e4f7b0e46a70e8f0780e5
That same commit also changed the filenames to use a * wildcard after
the python version, apparently to account for a "d" suffix in the
--enable-dbgutil case. In the spirit of the old commit to avoid rm -f
and to be explicit and catch potential other issues in future remove
that wildcard again/replace the suffix with a macro that contains the
appropriate suffix already as used in ExternalPackage_python3.mk.
Change-Id: I22b49af7dbf9a0bed5c124d9f137b8a38fee7471
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180705
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
For that, just unzip the wheel files to the Lib directory
Change-Id: I3998e427aeadb840b626ebbb6ce91938de47d2b2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180271
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
* external/fontconfig/ubsan.patch is fixed upstream
Downloaded from https://www.freedesktop.org/software/fontconfig/release/fontconfig-2.16.0.tar.xz
Change-Id: Id1acbc473cc03e35c8a7a6bd2b2c080cbf653d1f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180441
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Updates the Skia shared context patch to avoid surface/swapchain
construction on invalid X11 windows.
Change-Id: Ia591fb3755ed5adeec2b6246cf9a4a2fb9dcc74d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180420
Reviewed-by: Jonathan Clark <jonathan@libreoffice.org>
Tested-by: Jenkins
|
|
It's the only one the unordf component invokes.
CVE-2024-57823 CVE-2024-57822 affect the "ntriples" and "turtle"
parsers.
However it appears that the function raptor_uri_normalize_path() could
also be called from raptor_libxml_* functions? Somewhat unclear, let's
add the patch just in case.
Change-Id: Idd7ebbc29c63e84ca2434b06c26f7aca34bdcaa5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180272
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
...to avoid
> [build CHK] CustomTarget/postprocess/check_dynamic_objects/check.done
> instdir/program/python-core-3.11.11/lib/lib-dynload/_lzma.cpython-3.11d.so has suspicious NEEDED: liblzma.so.5
as seen at least when building on a Fedora 41 machine that happens to have
xz-devel-5.6.2-2.fc41.x86_64 installed (so that Python's configure detects
have_liblzma=yes).
The _lzma library had been added to external/python3/ExternalPackage_python3.mk
in 892c0d2ef0ee43e079aff51dcc09572d76689ec1 "python3: update
python3_EXTENSION_MODULES", but without giving any rationale. (It had
previously been removed there with c2e4e74138c0a816f6aace4d336c7d8cb611d4fa
"Python3.5: Remove external dependencies: readline and lzma".)
Change-Id: I4afe9fb08891dda116f115dac7b2f948aac84e06
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180194
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
|
|
Cloph reports it is not supported by macOS baseline XCode 14.2
Change-Id: I70a2e4c3ffd91c1a41123d1fa4d371406b300cba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180026
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Tested-by: Jenkins
|
|
* external/boost/0001-Add-a-comma-before-ellipsis-in-constexpr_swap.patch.2
has been fixed upstream
* Add external/boost/armv7.patch.0 based on following patches:
- https://github.com/boostorg/interprocess/commit/f58e968f13e693ffaa78fb283972e8598dc48c9f
- https://github.com/boostorg/unordered/commit/07f6463c1c302c5b1d28aa253e0b768e71c90235
- https://github.com/boostorg/json/commit/57d42f07158b8ea4a713378c90d22b28ca4787f8
- https://github.com/boostorg/outcome/commit/83e5127dd20b15e3f59059e5862e9926283fee80
to fix the problem described in 4f60246a1de68f7b47713ff926902791abe7a9ca
"Revert "boost: upgrade to 1_87_0" for now to unbreak android arm32
build"
<https://dev-www.libreoffice.org/src/boost_1_87_0.tar.xz> has been generated (on
Debian 12) with
> $ wget https://boostorg.jfrog.io/artifactory/main/release/1.87.0/source/boost_1_87_0.tar.bz2
> $ printf 'af57be25cb4c4f4b413ed692fe378affb4352ea50fbe294a11ef548f4d527d89 boost_1_87_0.tar.bz2' | sha256sum -c # cf. <https://www.boost.org/users/history/version_1_87_0.html>
> boost_1_87_0.tar.bz2: OK
> $ external/boost/repack_tarball.sh boost_1_87_0.tar.bz2
> Unpacking boost_1_87_0.tar.bz2 ...
> Removing unnecessary files ...
> Creating boost_1_87_0.tar.xz ...
> Cleaning up ...
> efd6d4ce7e8571ba87f77a30bee2d3dd8dccd306721351464fc6998dd00b0c8c boost_1_87_0.tar.xz
> Done.
Change-Id: I4bb14ba60f0b19c4908a0ffbdc66cc23c1b2581e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180169
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
|
|
Instead of downloading nuget.exe and then python.exe
as explained in https://lists.freedesktop.org/archives/libreoffice/2025-January/092881.html,
Add python.3.12.8.nupkg to extern and download it as a tarball
to unzip it in external/pythonx86.
find_python.bat will look for python.exe in external/pythonx86/tools
Downloaded from
https://globalcdn.nuget.org/packages/python.3.12.8.nupkg?packageVersion=3.12.8
Change-Id: Ib1638fa66b3983048c2f9a472d00ce365602fcbb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180078
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
git-bash/msys tar fails when trying to extract the zxing tarball because
it contains dangling symlinks and MSYS doesn't create symlinks by
default.
Change-Id: I51f3bb9e2a8cf5395b3fa314f3ac1d10b8873da7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179948
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
removed by mistake in
commit 50524481b30d904ee4a12ef478eeae05647a465d
Author: Xisco Fauli <xiscofauli@libreoffice.org>
Date: Tue Dec 17 17:25:59 2024 +0100
Python: upgrade to 3.11.11 (master only)
Change-Id: Ib1f79713e81238fd5c26cbbaf87e260a0a7e4f41
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180032
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
...when generating Python/deepfreeze/deepfreeze.c during the build of
ExternalProject_python3 after 50524481b30d904ee4a12ef478eeae05647a465d "Python:
upgrade to 3.11.11 (master only)",
> Objects/dictobject.c:1500:40: runtime error: load of misaligned address 0x555853494d81 for type 'PyDictUnicodeEntry *', which requires 8 byte alignment
> 0x555853494d81: note: pointer points here
> 00 00 00 ff ff ff ff ff ff ff ff 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> ^
> #0 in dictresize at workdir/UnpackedTarball/python3/Objects/dictobject.c:1500:21
> #1 in dict_merge at workdir/UnpackedTarball/python3/Objects/dictobject.c:2880:17
> #2 in _PyDict_MergeEx at workdir/UnpackedTarball/python3/Objects/dictobject.c:3003:12
> #3 in _PyEval_EvalFrameDefault at workdir/UnpackedTarball/python3/Python/ceval.c:3432:17
> #4 in _PyEval_EvalFrame at workdir/UnpackedTarball/python3/./Include/internal/pycore_ceval.h:73:16
> #5 in _PyEval_Vector at workdir/UnpackedTarball/python3/Python/ceval.c:6434:24
> #6 in _PyFunction_Vectorcall at workdir/UnpackedTarball/python3/Objects/call.c:393:16
> #7 in _PyObject_FastCallDictTstate at workdir/UnpackedTarball/python3/Objects/call.c:141:15
> #8 in _PyObject_Call_Prepend at workdir/UnpackedTarball/python3/Objects/call.c:482:24
> #9 in slot_tp_init at workdir/UnpackedTarball/python3/Objects/typeobject.c:7855:15
> #10 in type_call at workdir/UnpackedTarball/python3/Objects/typeobject.c:1103:19
> #11 in _PyObject_MakeTpCall at workdir/UnpackedTarball/python3/Objects/call.c:214:18
> #12 in _PyObject_VectorcallTstate at workdir/UnpackedTarball/python3/./Include/internal/pycore_call.h:90:16
> #13 in PyObject_Vectorcall at workdir/UnpackedTarball/python3/Objects/call.c:299:12
> #14 in _PyEval_EvalFrameDefault at workdir/UnpackedTarball/python3/Python/ceval.c:4769:23
> #15 in _PyEval_EvalFrame at workdir/UnpackedTarball/python3/./Include/internal/pycore_ceval.h:73:16
> #16 in _PyEval_Vector at workdir/UnpackedTarball/python3/Python/ceval.c:6434:24
> #17 in PyEval_EvalCode at workdir/UnpackedTarball/python3/Python/ceval.c:1148:21
> #18 in run_eval_code_obj at workdir/UnpackedTarball/python3/Python/pythonrun.c:1741:9
> #19 in run_mod at workdir/UnpackedTarball/python3/Python/pythonrun.c:1762:19
> #20 in pyrun_file at workdir/UnpackedTarball/python3/Python/pythonrun.c:1657:15
> #21 in _PyRun_SimpleFileObject at workdir/UnpackedTarball/python3/Python/pythonrun.c:440:13
> #22 in _PyRun_AnyFileObject at workdir/UnpackedTarball/python3/Python/pythonrun.c:79:15
> #23 in pymain_run_file_obj at workdir/UnpackedTarball/python3/Modules/main.c:360:15
> #24 in pymain_run_file at workdir/UnpackedTarball/python3/Modules/main.c:379:15
> #25 in pymain_run_python at workdir/UnpackedTarball/python3/Modules/main.c:605:21
> #26 in Py_RunMain at workdir/UnpackedTarball/python3/Modules/main.c:684:5
> #27 in main at workdir/UnpackedTarball/python3/Programs/_bootstrap_python.c:109:12
Change-Id: I0bacef11b5874bca6104516a206076a0505528ed
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/180037
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
|
|
Downloaded from https://www.python.org/ftp/python/3.11.11/Python-3.11.11.tar.xz
* Add external/python3/replace-powershell-with-wget.patch.1
to avoid
FindPythonForBuild:
Invoke-WebRequest : The request was aborted: Could not create SSL/TLS secure
channel.
At line:1 char:1
+ Invoke-WebRequest https://aka.ms/nugetclidl -OutFile
'C:\cygwin64\home\tdf\jenki ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~
+ CategoryInfo : InvalidOperation: (System.Net.HttpWebRequest:Htt
pWebRequest) [Invoke-WebRequest], WebException
+ FullyQualifiedErrorId : WebCmdletWebResponseException,Microsoft.PowerShe
ll.Commands.InvokeWebRequestCommand
'py' is not recognized as an internal or external command,
operable program or batch file.
when downloading nuget.exe from https://aka.ms/nugetclidl
using powershell. By default powershell uses TLS 1.0 and the site security requires TLS 1.2,
so it fails with 'The request was aborted: Could not create SSL/TLS
secure channel.'
Kudos to Caolán for helping with a compiling issue.
See https://lists.freedesktop.org/archives/libreoffice/2024-December/092830.html
Change-Id: If66a88dd8f8c7774c49f8c33e9e40d2d5418e7c6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178681
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Fixes CVE-2024-56378
- remove disable-nss-and-gpgmepp.patch.1 and instead build
CryptoSignBackend.cc which appears to have working ifdefs now
- run the update_pch script
- add clang-std-rotate.patch.1 because jenkins failed
Change-Id: I7fcdd10766287e83dc666041644f129378584d91
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179903
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
All three patches are upstreamed.
Downloaded from https://kumisystems.dl.sourceforge.net/project/libpng/libpng16/1.6.45/libpng-1.6.45.tar.xz?viasf=1
Change-Id: I61d336704b96e2a25d92b4512507cc168648ddec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179930
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Change-Id: Ia636690f0dd9497e62ac9fd7a1ed4d00e71cc299
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179896
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
Using an autogen.input somewhat different from what I usually use for ASan/UBSan
builds, that caused
> libtool: link: /home/sberg/llvm/inst/bin/clang -fdiagnostics-color=always -fsanitize=address -fsanitize-address-globals-dead-stripping -fsanitize-address-use-after-return=always -fsanitize=undefined -fsanitize=float-divide-by-zero -fsanitize=local-bounds -fsanitize=nullability -fno-sanitize=function -O0 -Wp,-U_FORTIFY_SOURCE -fstrict-aliasing -fstrict-overflow -ggdb2 -gsplit-dwarf -Xclang -debug-info-kind=constructor -ggnu-pubnames -Wl,-z -Wl,origin -Wl,-rpath -Wl,\$ORIGIN -Wl,-rpath-link -Wl,/home/sberg/lo-asanubsan/core/instdir/program -o .libs/rapper rapper.o ../src/.libs/libraptor2.so -Wl,-rpath -Wl,/usr/local/lib
> /usr/bin/ld: /home/sberg/lo-asanubsan/core/instdir/program/libicuuc.so.75: undefined reference to `__ubsan_vptr_type_cache'
> /usr/bin/ld: /home/sberg/lo-asanubsan/core/instdir/program/libicuuc.so.75: undefined reference to `__ubsan_handle_dynamic_type_cache_miss'
> clang: error: linker command failed with exit code 1 (use -v to see invocation)
> make[3]: *** [Makefile:540: rapper] Error 1
> make[3]: *** Waiting for unfinished jobs....
> libtool: link: /home/sberg/llvm/inst/bin/clang -fdiagnostics-color=always -fsanitize=address -fsanitize-address-globals-dead-stripping -fsanitize-address-use-after-return=always -fsanitize=undefined -fsanitize=float-divide-by-zero -fsanitize=local-bounds -fsanitize=nullability -fno-sanitize=function -O0 -Wp,-U_FORTIFY_SOURCE -fstrict-aliasing -fstrict-overflow -ggdb2 -gsplit-dwarf -Xclang -debug-info-kind=constructor -ggnu-pubnames -Wl,-z -Wl,origin -Wl,-rpath -Wl,\$ORIGIN -Wl,-rpath-link -Wl,/home/sberg/lo-asanubsan/core/instdir/program -o .libs/rdfdiff rdfdiff.o ../src/.libs/libraptor2.so -Wl,-rpath -Wl,/usr/local/lib
> /usr/bin/ld: /home/sberg/lo-asanubsan/core/instdir/program/libicuuc.so.75: undefined reference to `__ubsan_vptr_type_cache'
> /usr/bin/ld: /home/sberg/lo-asanubsan/core/instdir/program/libicuuc.so.75: undefined reference to `__ubsan_handle_dynamic_type_cache_miss'
> clang: error: linker command failed with exit code 1 (use -v to see invocation)
It appears that different configurations cause different libraries to be created
in ExternalProject_raptor, so that's why I never ran into this before, but the
solution is the same as in b3300387f2e14ea59c70db513a8f6eda6fe17afa "Fix UBSan
build of ExternalProject_redland".
Change-Id: I8bf955632af8c6490d9bd66a1d7dcd19b623c305
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179898
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
|
|
Downloaded from https://github.com/zxing-cpp/zxing-cpp/archive/refs/tags/v2.3.0.tar.gz
Change-Id: I79a9d63cfafb44d733231b78bc19a45b43baa6b7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179604
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
Change-Id: Ica4cacc702f338e9e0ebcf6e4b6a48d1d4f3bef8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/179605
Reviewed-by: Taichi Haradaguchi <vdnznnwii@mozmail.com>
Tested-by: Jenkins
|
|
Add missing _lasx files, because otherwise we get
/usr/bin/ld: /<<PKGBUILDDIR>>/workdir/GenCxxObject/UnpackedTarball/skia/src/core/SkBitmapProcState_opts.o: in function `SkOpts::Init_BitmapProcState()':
./././workdir/UnpackedTarball/skia/src/core/SkBitmapProcState_opts.cpp:36:(.text+0x58): undefined reference to `SkOpts::Init_BitmapProcState_lasx()'
/usr/bin/ld: /<<PKGBUILDDIR>>/workdir/GenCxxObject/UnpackedTarball/skia/src/core/SkBlitRow_opts.o: in function `SkOpts::Init_BlitRow()':
./././workdir/UnpackedTarball/skia/src/core/SkBlitRow_opts.cpp:36:(.text+0x58): undefined reference to `SkOpts::Init_BlitRow_lasx()'
/usr/bin/ld: /<<PKGBUILDDIR>>/workdir/GenCxxObject/UnpackedTarball/skia/src/core/SkOpts.o: in function `SkOpts::Init()':
./././workdir/UnpackedTarball/skia/src/core/SkOpts.cpp:65:(.text+0x237d8): undefined reference to `SkOpts::Init_lasx()'
/usr/bin/ld: /<<PKGBUILDDIR>>/workdir/GenCxxObject/UnpackedTarball/skia/src/core/SkSwizzler_opts.o: in function `SkOpts::Init_Swizzler()':
./././workdir/UnpackedTarball/skia/src/core/SkSwizzler_opts.cpp:51:(.text+0x58): undefined reference to `SkOpts::Init_Swizzler_lasx()'
cf. https://buildd.debian.org/status/fetch.php?pkg=libreoffice&arch=loong64&ver=4%3A25.2.0~beta1-1&stamp=1733592509&raw=1
Change-Id: I232b4656a87a4c3d87d669769c6483e01425c0e6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178051
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: wu jiahuan <wujiahuan@loongson.cn>
Tested-by: René Engelhard <rene@debian.org>
|
|
This change adds the --enable-skia-vulkan-validation build option. When
this option is set, Skia will be built with Vulkan validation layers
enabled. Vulkan validation layers must also be installed on the
development machine (e.g. via the Windows Vulkan SDK, or the Ubuntu
vulkan-validationlayers package).
Vulkan validation layers are an important debugging tool for Vulkan
application development. Validation layers externally verify that Vulkan
API calls are correct and comply with the standard. This is particularly
important as Vulkan does not require drivers to reject invalid uses;
some hardware or drivers may tolerate certain developer mistakes, while
other drivers reject them. Without using validation layers to confirm,
this may create a false impression that certain hardware/drivers are
faulty, when it is our own code at fault.
Change-Id: If39a09e3d1043d37b465dd3cc5fcd8e5c364a1a2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178736
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Change-Id: Ia38a81cbcd14776b7fa1398856bd3f8997a869ff
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178741
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
Change-Id: Ifc24cb42df76c50f0ca40e3c793f36c355e483d8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178730
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
...where "will be needed" is approximated by "install location is under
C:\Program Files" (i.e., where admin privileges will be necessary to do an
update)
Change-Id: I00c93df8c0c2c5132fffa4056b88af098cb4a644
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178679
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
Tested-by: Jenkins
|
|
For some (still unknown) reason the upgrade from boost 1.86 to 1.87
caused the 32-bit ARM Android builds to fail
with errors like
<inline asm>:1:41: error: expected '%<type>' or "<type>"
.pushsection ".debug_gdb_scripts", "MS",@progbits,1
^
<inline asm>:318:12: error: .popsection without corresponding .pushsection
.popsection
^
2 errors generated.
make[1]: *** [/home/tdf/lode/jenkins/workspace/android_arm/solenv/gbuild/LinkTarget.mk:339: /home/tdf/lode/jenkins/workspace/android_arm/workdir/CxxObject/configmgr/source/groupnode.o] Error 1
make[1]: *** Waiting for unfinished jobs....
[build CXX] configmgr/source/node.cxx
<inline asm>:1:41: error: expected '%<type>' or "<type>"
.pushsection ".debug_gdb_scripts", "MS",@progbits,1
^
, see e.g. [1] and the list of all Android ARM 32 bit builds [2].
A local bisect points to this commit, and reverting it
fixes the build.
Do this for now to work around the build failure and give
time for finding a better solution.
This reverts commit 25e7e1677b13db818b5458cd0c3fd3d9dfc88557.
[1] https://ci.libreoffice.org/job/gerrit_android_arm/43354/console
[2] https://ci.libreoffice.org/job/gerrit_android_arm/
Change-Id: Ifad1412b691adce74bb16805975e3439898e3b74
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178697
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
* external/boost/0001-Add-a-comma-before-ellipsis-in-constexpr_swap.patch.2
has been fixed upstream
<https://dev-www.libreoffice.org/src/boost_1_87_0.tar.xz> has been generated (on
Debian 12) with
> $ wget https://boostorg.jfrog.io/artifactory/main/release/1.87.0/source/boost_1_87_0.tar.bz2
> $ printf 'af57be25cb4c4f4b413ed692fe378affb4352ea50fbe294a11ef548f4d527d89 boost_1_87_0.tar.bz2' | sha256sum -c # cf. <https://www.boost.org/users/history/version_1_87_0.html>
> boost_1_87_0.tar.bz2: OK
> $ external/boost/repack_tarball.sh boost_1_87_0.tar.bz2
> Unpacking boost_1_87_0.tar.bz2 ...
> Removing unnecessary files ...
> Creating boost_1_87_0.tar.xz ...
> Cleaning up ...
> efd6d4ce7e8571ba87f77a30bee2d3dd8dccd306721351464fc6998dd00b0c8c boost_1_87_0.tar.xz
> Done.
Change-Id: I9aaa9fc435b5b1890f9038ccfbfa26707a57a8da
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178574
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
CPWL_Dash-switch-to-universal-init.patch.1 is upstreamed, remove it.
Change-Id: I4ff68f15d58b853896626b0458d3817f64fbc32f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178297
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
|
|
So it'll be easier to drop it the next time pdfium is updated.
Change-Id: Iebeb892f62ee8002460142adf28f75666c92c2e4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178237
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
|
|
Change-Id: I161f3cfc334d6be0263536800cb028b8e7310823
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178080
Tested-by: Jenkins
Reviewed-by: Ilmari Lauhakangas <ilmari.lauhakangas@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
Change-Id: Idb458e3e65bc22d148cc68e496aa0dda32fec2a2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178079
Tested-by: Jenkins
Reviewed-by: Ilmari Lauhakangas <ilmari.lauhakangas@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
There is no simple "Windows is 10 or newer" check in MSI, with the
Microsoft policy of "you don't need to know real Windows version".
See also https://developercommunity.visualstudio.com/t/detect-a-windows-version-from-msi/412198
The values that MSI sees (e.g. on my Win11) are:
Property(C): VersionNT = 603
Property(C): VersionNT64 = 603
Property(C): WindowsBuild = 9600
So those can't be used for the check. I use the WINMAJORVER that is
read from registry (Software\Microsoft\Windows NT\CurrentVersion).
There is no more need to check and install UCRT; it was dropped.
Change-Id: Ic02d6aeb965e9684bafd4357f969980f77000e73
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/178049
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
When live resizing a window, replacing the CAMetalLayer with each
resize event repaints the window's background which causes a
noticeable flicker. So reuse any existing CAMetalLayer already
assigned to the native view.
Change-Id: I03bda5f0d40b84606b6602961e5f0d3b0dfcc6ba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177921
Tested-by: Jenkins
Reviewed-by: Patrick Luby <guibomacdev@gmail.com>
|
|
* external/gpgmepp/gettid.patch issue fixed upstream
Downloaded from https://gnupg.org/ftp/gcrypt/gpgme/gpgme-1.24.1.tar.bz2
Change-Id: Id2d984f5e68b8681bf5e635a1afe5d90605711ec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177821
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
jom (https://wiki.qt.io/Jom) is a parallel version of nmake that
especially helps with building openssl since everything is compiled
individually/the compiler cannot use parallelism either.
Also in a LibreOffice build there's not much else that can be done in
parallel/other stuff requires openssl to be built to continue.
Change-Id: Ib16b0f947991a1d3255aeae7dc5d9bdee0b157a0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177743
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Tested-by: Jenkins
|
|
...new with GCC 15 trunk -std=c++26 since
<https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=cc67d95dc100706ea665e8cce581d59466aba62e>
"c++: Implement C++26 P3176R1 - The Oxford variadic comma",
> In file included from workdir/UnpackedTarball/boost/boost/rational.hpp:82,
> from source/external/boost/include/boost/rational.hpp:30,
> from source/tools/source/generic/fract.cxx:32:
> workdir/UnpackedTarball/boost/boost/integer/common_factor_rt.hpp:66:56: error: omission of ‘,’ before varargs ‘...’ is deprecated in C++26 [-Werror=deprecated-variadic-comma-omission]
> 66 | inline constexpr void constexpr_swap(T&a, U& b...) BOOST_GCD_NOEXCEPT(T)
> | ^~~
> | ,
and
> In file included from workdir/UnpackedTarball/boost/boost/move/unique_ptr.hpp:24,
> from workdir/UnpackedTarball/boost/boost/spirit/home/classic/core/non_terminal/impl/grammar.ipp:18,
> from workdir/UnpackedTarball/boost/boost/spirit/home/classic/core/non_terminal/grammar.hpp:21,
> from workdir/UnpackedTarball/boost/boost/spirit/home/classic/core.hpp:42,
> from workdir/UnpackedTarball/boost/boost/spirit/home/classic.hpp:24,
> from workdir/UnpackedTarball/boost/boost/spirit/include/classic.hpp:11,
> from source/external/boost/include/boost/spirit/include/classic.hpp:30,
> from sdext/source/pdfimport/pdfparse/pdfparse.cxx:23:
> workdir/UnpackedTarball/boost/boost/move/detail/unique_ptr_meta_utils.hpp:500:39: error: omission of ‘,’ before varargs ‘...’ is deprecated in C++26 [-Werror=deprecated-variadic-comma-omission]
> 500 | struct is_unary_function_impl<R (*)(T0...)>
> | ^~~
> | ,
Change-Id: I56856e274a7c577c53d96343992bb79d4c3d602e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177730
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
Tested-by: Jenkins
|
|
For some unkown reason (maybe due to the macOS two-level namespace?)
including the $(ICU_LIBS) link flags causes a crash on startup in
the macOS dynamic library loader on some versions of macOS running
on Intel machines. LibreOffice 24.8.3 and earlier did not have this
bug since libraptor2-lo was not linked against libicuuc. So, revert
back to the LibreOffice 24.8.3 macOS build behavior.
On macOS, libraptor2-lo still links successfully even without
libicuuc and the bundled JavaScript HelloWorld macro still runs. So
it appears that libraptor2-lo does not need any symbols in libicuuc
or, if it does, libicuuc has already been loaded by the time
libraptor2-lo is loaded.
Change-Id: I2e09ce57b5f7aea631a522f5d33f3a4b1ebfd419
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177594
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Reviewed-by: Patrick Luby <guibomacdev@gmail.com>
|
|
Because the mutex has already been freed because raptor calls
xmlCleanupParser
I have logged a feature request against raptor for a flag to turn off
this behaviour, but until that lands, this will have to do.
-----
make CppunitTest_filter_pdf results in
WARNING: ThreadSanitizer: use of an invalid mutex (e.g. uninitialized or
destroyed) (pid=330602)
0 pthread_mutex_lock
/home/noel/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_posix.cpp:1371
(discriminator 8) (cppunittester+0x7e1be)
1 xmlDictFree ??:? (libxml2.so.2+0x3404b) (BuildId:
14c9410dbe8278dfa121dabc48c9ae25eac169d1)
2 lt_xml_unref
/home/noel/libo-tsan/workdir/UnpackedTarball/liblangtag/liblangtag/lt-xml.c:351
(liblangtag-lo.so.1+0x1fc6a)
3 lt_db_finalize
/home/noel/libo-tsan/workdir/UnpackedTarball/liblangtag/liblangtag/lt-database.c:144
(liblangtag-lo.so.1+0xac22)
4 (anonymous namespace)::LiblangtagDataRef::teardown()
/home/noel/libo-tsan/i18nlangtag/source/languagetag/languagetag.cxx:219
(libi18nlangtag.so+0x20951)
5 (anonymous namespace)::LiblangtagDataRef::~LiblangtagDataRef()
/home/noel/libo-tsan/i18nlangtag/source/languagetag/languagetag.cxx:204
(libi18nlangtag.so+0x20951)
6 cxa_at_exit_callback_installed_at(void*)
/home/noel/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_posix.cpp:445
(cppunittester+0x6e3af)
7 LanguageTagImpl::canonicalize()
/home/noel/libo-tsan/i18nlangtag/source/languagetag/languagetag.cxx:190
(libi18nlangtag.so+0xe370)
8 main /home/noel/libo-tsan/sal/cppunittester/cppunittester.cxx:482
(cppunittester+0x10cf6c)
Change-Id: I5d9ff1c5667f8e19e5858cd5a408e8f95d2a98e4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177320
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Generated with:
./autogen.sh
./configure
make dist-xz
Change-Id: I46b921f3c2ef41ab52e272df5afd82a1d9331f42
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177332
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
Since
commit 05aedcba6672979d317b540bbfa74f5c9b409402
Author: Stephan Bergmann <sbergman@redhat.com>
Date: Thu Jan 21 23:19:20 2021 +0100
Fix use of -fvisibility=hidden with Clang in external/libcdr, external/libqxp
Change-Id: Idb97eaea30f5a2b15379b18981455532de0308b6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177330
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Generated with:
./autogen.sh
./configure
make dist-xz
Change-Id: I74ac661c5f8db9abf7644212d98b29c3ec5b422f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/176036
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
...by backporting upstream
<https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=bc5a4dfcf7390145dc3ba3c1c18c5ce561b778cd>
"Assume that <stdbool.h> conforms to the C standard." to our REL_14_15 version.
GCC 15 trunk defaults to C23 now, so started to fail with
> In file included from fls.c:49:
> ../../src/include/c.h:419:23: error: two or more data types in declaration specifiers
> 419 | typedef unsigned char bool;
> | ^~~~
> ../../src/include/c.h:419:1: warning: useless type name in empty declaration
> 419 | typedef unsigned char bool;
> | ^~~~~~~
> make: *** [<builtin>: fls.o] Error 1
> make: Leaving directory 'workdir/UnpackedTarball/postgresql/src/port'
(Removing HAVE__BOOL and HAVE_STDBOOL_H also from src/tools/msvc/Solution.pm is
necessary to avoid failures like
> "C:/cygwin64/home/tdf/jenkins/workspace/gerrit_windows/workdir/UnpackedTarball/openssl\apps\openssl.exe" version 2>&1unused defines: HAVE_STDBOOL_H HAVE__BOOL at /home/tdf/jenkins/workspace/gerrit_windows/workdir/UnpackedTarball/postgresql/src/tools/msvc/Mkvcbuild.pm line 870.
> make[1]: *** [C:/cygwin64/home/tdf/jenkins/workspace/gerrit_windows/external/postgresql/ExternalProject_postgresql.mk:27: C:/cygwin64/home/tdf/jenkins/workspace/gerrit_windows/workdir/ExternalProject/postgresql/build] Error 1
in MSVC builds.)
Change-Id: I1f6bd0d613ae4d6ce70feaaf67aaf95496eeff87
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177280
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins
|
|
...so the existing C code started to fail with
> In file included from workdir/UnpackedTarball/mariadb-connector-c/libmariadb/ma_alloc.c:20:
> workdir/UnpackedTarball/mariadb-connector-c/include/ma_global.h:687:25: error: two or more data types in declaration specifiers
> 687 | typedef char bool; /* Ordinary boolean values 0 1 */
> | ^~~~
(And the !defined(HAVE_BOOL) there, always being true, apparently started to get
in the way now and thus needed to be removed, whatever its original purpose.)
Change-Id: I781458d643e01a7199e19a178da3a32520d16b34
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177260
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
Tested-by: Jenkins
|
|
This reverts commit 32eb29377c5bbf6a965357756d2813add2789053.
Reason for revert: Original commit did not fix tdf#163945
Change-Id: Ie66edde40ed44f89bb28043db074fab7e6e06a49
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177025
Tested-by: Jenkins
Reviewed-by: Patrick Luby <guibomacdev@gmail.com>
|