Debian Patches
Status for libreoffice/4:25.2.3-2+deb13u12
| Patch | Description | Author | Forwarded | Bugs | Origin | Last update |
|---|---|---|---|---|---|---|
| debian-opt.diff | diff --git a/configure.ac b/configure.ac index f06ef8e..91ce612 100644 |
no | ||||
| jurt-soffice-location.diff | commit b71107fb12e3c3125e0cb62c5a4f6636a80c6408 on debian-based systems, we know where our soffice binary is diff --git a/jurt/com/sun/star/lib/util/NativeLibraryLoader.java b/jurt/com/sun/star/lib/util/NativeLibraryLoader.jav index da22980..36664ca 100644 |
Bjoern Michaelsen <bjoern.michaelsen@canonical.com> | no | |||
| splits-adapt-registry.diff | diff --git a/scp2/source/gnome/file_gnome.scp b/scp2/source/gnome/file_gnome.scp index c7a13c8..a3f59a6 100644 |
no | ||||
| debian-debug.diff | diff --git a/solenv/gbuild/platform/com_GCC_defs.mk b/solenv/gbuild/platform/com_GCC_defs.mk index 9de88a2..9161a4e 100644 |
no | ||||
| build-against-shared-lpsolve.diff | diff --git a/configure.ac b/configure.ac index cdae8b5..49f3ba2 100644 |
no | ||||
| install-fixes.diff | diff --git a/bin/distro-install-file-lists b/bin/distro-install-file-lists index eaabf9c..c1b1ec6 100755 |
no | ||||
| mention-java-common-package.diff | diff --git a/jvmfwk/plugins/sunmajor/javaenvsetup/javaldx.cxx b/jvmfwk/plugins/sunmajor/javaenvsetup/javaldx.cxx index 718902caba6f..3c8499b5a42c 100644 |
no | ||||
| help-msg-add-package-info.diff | diff --git a/include/sfx2/strings.hrc b/include/sfx2/strings.hrc index 1b0ea6ccb2ff..630f4330552a 100644 |
no | ||||
| sensible-lomua.diff | =================================================================== | no | ||||
| reportdesign-mention-package.diff | diff --git a/dbaccess/inc/strings.hrc b/dbaccess/inc/strings.hrc index 47068f43d09e..1f8b9a2331b6 100644 |
no | ||||
| jdbc-driver-classpaths.diff | diff --git a/officecfg/registry/data/org/openoffice/Office/DataAccess.xcu b/officecfg/registry/data/org/openoffice/Office/DataAccess.xcu index 9be30a2..59c87cb 100644 |
no | ||||
| make-package-modules-not-suck.diff | diff --git a/scp2/InstallModule_draw.mk b/scp2/InstallModule_draw.mk index a7c02be..d48edf2 100644 |
no | ||||
| mediwiki-oor-replace.diff | diff --git a/swext/mediawiki/src/registry/data/org/openoffice/Office/OptionsDialog.xcu b/swext/mediawiki/src/registry/data/org/openoffice/Office/OptionsDialog.xcu index 2b35ced08f31..0c96070dd4f0 100644 |
no | ||||
| debian-hardened-buildflags-CPPFLAGS.diff | no | |||||
| debian-hardened-buildflags-no-LO-fstack-protector-strong.diff | don't hardcode -fstack-protector-strong in configure.ac/gbuild. We get the hardening flags from dpkg-buildflags anyway. diff --git a/solenv/gbuild/platform/com_GCC_defs.mk b/solenv/gbuild/platform/com_GCC_defs.mk index 712a61df544f..0d50f538ba7b 100644 |
no | ||||
| disable-flaky-tests.diff | 14:13 < mst__> _rene_, the toolkit unoapi tests are known to be flaky (in some system dependent way) e.g. on the Win@6 tinderbox it always crashes 14:14 < mst__> _rene_, sc.ScAccessible* tests also fail on some systems some of the time diff --git a/toolkit/Module_toolkit.mk b/toolkit/Module_toolkit.mk index 25db0b6..14e507c 100644 |
no | ||||
| fix-internal-hsqldb-build.diff | diff -urN hsqldb.old/patches/use-system-servlet-api.jar.diff hsqldb/patches/use-system-servlet-api.jar.diff | no | ||||
| javadoc-optional.diff | Gemeinsame Unterverzeichnisse: odk-old/config und odk/config. Gemeinsame Unterverzeichnisse: odk-old/docs und odk/docs. Gemeinsame Unterverzeichnisse: odk-old/examples und odk/examples. diff --git a/odk/Module_odk.mk b/odk/Module_odk.mk index 693885322115..5fcb39306a0b 100644 |
no | ||||
| no-packagekit-per-default.diff | diff --git a/officecfg/registry/data/org/openoffice/Office/Common.xcu b/officecfg/registry/data/org/openoffice/Office/Common.xcu index 3d138551b593..9cb9831f3236 100644 |
no | ||||
| system-officeotron-and-odfvalidator.diff | diff --git a/bin/odfvalidator.sh.in b/bin/odfvalidator.sh.in index 56e2f29..9415ef6 100644 |
no | ||||
| cppunit-optional.diff | diff --git a/configure.ac b/configure.ac index c12fe95a561c..66d327ae8fb8 100644 |
no | ||||
| apparmor-complain.diff | diff --git a/sysui/desktop/apparmor/program.oosplash b/sysui/desktop/apparmor/program.oosplash index fef54b7ee384..9dde31a63615 100644 |
no | ||||
| hide-math-desktop-file.patch | Hide startcenter and math from the shell | Olivier Tilloy <olivier.tilloy@canonical.com> | not-needed | |||
| appstream-ignore-startcenter.diff | no | |||||
| disable-java-in-odk-build-examples-on-zero-vm.diff | diff --git a/config_host.mk.in b/config_host.mk.in index 6cea6ccf795e..d08fac922239 100644 |
no | ||||
| do-not-hide-test-output.diff | diff --git a/odk/build-examples_common.mk b/odk/build-examples_common.mk index abcb3a3e2593..35d45ad23413 100644 |
no | ||||
| apparmor-allow-java.security.diff | diff --git a/sysui/desktop/apparmor/program.soffice.bin b/sysui/desktop/apparmor/program.soffice.bin index 2fc7fd6b5735..3fd82b08431e 100644 |
no | ||||
| apparmor-cleanups.diff | apparmor: use dri-enumerate abstraction Remove backported rule and use new dri-enumerate abstraction instead. dri-enumerate is available in AppArmor 2.13, which recently migrated into Debian Buster. |
Vincas Dargis <vindrg@gmail.com> | no | 2018-08-04 | ||
| apparmor-mesa.diff | no | |||||
| disableClassPathURLCheck.diff | no | |||||
| use-mariadb-java-instead-of-mysql-java.diff | mariadb | Markus Koschany <apo@debian.org> | no | 2018-11-09 | ||
| apparmor-opencl.diff | AppArmor in Debian Buster now has OpenCL abstractions. Include OpenCL abstractions to fix OpenCL usage in Calc. =================================================================== |
no | ||||
| fix-flaky-bridgetest.diff | Add safer float comparisons to bridgetest equals() | Marcus Tomlinson <marcus.tomlinson@canonical.com> | no | |||
| add-access2base-doc.diff | no | |||||
| disable-shortcuts_tab_navigation-uitest.diff | no | |||||
| no-opencl-per-default.diff | Resolves: rhbz#1432468 disable opencl by default | Caolán McNamara <caolanm@redhat.com> | no | 2017-03-27 | ||
| apparmor-updates.diff | diff --git a/sysui/desktop/apparmor/program.soffice.bin b/sysui/desktop/apparmor/program.soffice.bin index 42053db2abef..bf48bdb1e44d 100644 |
no | ||||
| apparmor-gnupg-tofu.diff | Support tofu+pgp trust model in GnuPG GnuPG supports a trust-on-first-use layer that sits on top of the standard PGP trust model. If this is enabled, 'gpg --list-keys' needs write and lock permissions on the TOFU database to return any useful data. Allow this access through AppArmor. |
Benjamin Barenblat <bbaren@google.com> | no | debian | ||
| arm-fp-mode.diff | Explicitly set fpu mode on armhf diff --git a/bridges/source/cpp_uno/gcc3_linux_arm/armhelper.S b/bridges/source/cpp_uno/gcc3_linux_arm/armhelper.S index 4eff3ff0a012..7996b9c5b3f0 100644 |
Rico Tzschichholz <ricotz@ubuntu.com> | no | |||
| moreIconsDialog-accesses-internet.diff | diff --git a/cui/source/dialogs/AdditionsDialog.cxx b/cui/source/dialogs/AdditionsDialog.cxx index a8eabf6d369c..676add18204e 100644 |
no | ||||
| do-not-abort-on-NON_APPLICATION_FONT_USE.diff | diff --git a/solenv/gbuild/CppunitTest.mk b/solenv/gbuild/CppunitTest.mk index b16e3d0e5417..33bdde04b072 100644 |
no | ||||
| we-dont-have-the-needed-fonts.diff | diff --git a/sw/qa/core/text/text.cxx b/sw/qa/core/text/text.cxx index a4d4540ab228..46d262f6f0b8 100644 |
no | ||||
| adapt-for-new-carlito.diff | diff --git a/sw/qa/extras/layout/layout3.cxx b/sw/qa/extras/layout/layout3.cxx index 7f62a3a9d6b8..ec1567a564dc 100644 |
no | ||||
| pdfium-ports.diff | no | |||||
| jdk-minimal-and-zero-paths.diff | sunjre: check all four possible OpenJDK variants for libjvm.so Since OpenJDK 21, the "minimal" and "zero" variants of JVM no longer reside in the "server" directory in $JAVA_HOME/lib, making the loader unable to find libjvm.so. This regresses LO builds on platforms with only those variants of JVM (i.e. without a HotSpot port). Example error message is like this: > exception occurred: Could not create Java implementation loader at > ./stoc/source/javaloader/javaloader.cxx:551 Fix it by probing all four possible variant directories for libjvm.so. This is tested on Debian unstable (loong64 port) to fix the packaging. |
WANG Xuerui <xen0n@gentoo.org> | no | 2024-08-22 | ||
| lo-xlate-lang-add-tl.diff | diff --git a/bin/lo-xlate-lang b/bin/lo-xlate-lang index 9b939012e99a..967a188c14e5 100755 |
no | ||||
| fix-32bits-test-build.diff | diff --git a/sd/qa/unit/tiledrendering/tiledrendering.cxx b/sd/qa/unit/tiledrendering/tiledrendering.cxx index 18b213e99abf..3a26f9493a7f 100644 |
no | ||||
| disable-uitest-xmlsecurity-gpg.diff | diff --git a/xmlsecurity/Module_xmlsecurity.mk b/xmlsecurity/Module_xmlsecurity.mk index afb1e251586a..12241f6b6520 100644 |
no | ||||
| system-colamd.diff | implement --with-system-colamd (for internal lpsolve) using suitesparse. Tested with both suitesparse 5.12.0 (Debian stable, without pc) and 7.9.0 (Debian unstable, with .pc) |
Rene Engelhard <rene@rene-engelhard.de> | no | 2025-03-03 | ||
| add-EUR-for-Bulgaria-Lew.diff | fix typo: BLN -> BGN After commit 37d1cc071ba0cc7f9922c4409d0a014a722ebbce add EUR for Bulgaria/Lew |
Rene Engelhard <rene@rene-engelhard.de> | no | 2025-08-09 | ||
| avmedia-qt-use-gstreamer-frame-grabber-by-default.diff | tdf#166055 avmedia qt: Use GStreamer frame grabber by default As described in tdf#166055, QtFrameGrabber currently still causes issues (freezes or crashes). The QtMultimedia API does not provide a simple way to synchronously retrieve a video frame, so the current approach is to connect to the QVideoSink::videoFrameChanged signal and start playing the video until the first frame arrives. There are various QtMultimedia plugins/backends (at least GStreamer and ffmpeg for Linux). Some use multiple threads internally. Some logic needs to be run in the main thread however, so it's not possible to move things to a separate thread/event loop to decouple it from the main thread. As a consequence, there is a need to trigger event processing while waiting for the frame, as QVideoSink::videoFrameChanged would otherwise never be called. Triggering event processing can have bad side-effects however, as seen in tdf#166055, e.g. leading to endless recursion or crashes when processing LO events. See also commit 697405b533a8ae5b6a8f5bd184b9344a96f71c69 Author: Michael Weghorn <m.weghorn@posteo.de> Date: Wed Apr 9 09:42:45 2025 +0200 tdf#166055 qt avmedia: Provide media/player size right away for more background and some thoughts. For now, avoid the problem by using the GStreamer based frame grabber even with QtPlayer by default. This means that the QtMultimedia framework is still used for video playback with the qt6 VCL plugin, where LO's GStreamer implementation doesn't work properly, see commit 441d8ed9be0e7f831b455a69b8688dcb79a8bc00 Author: Michael Weghorn <m.weghorn@posteo.de> Date: Mon May 20 16:25:09 2024 +0200 tdf#145735 avmedia qt: Use QtMultimedia for Qt 6 media playback . However, in order to preview images (that are e.g. shown in Impress when not in presentation mode) the LO GStreamer backend is used now. (That works fine, no UI interaction is needed for that.) Adjust QtPlayer::createFrameGrabber accordingly to return the LO default platform player/frame grabber implementation (which is the GStreamer one on Linux), see also MediaWindowImpl::createPlayer. Use the service to avoid having to link GStreamer libraries. However, allow to force the use of QtFrameGrabber by setting environment variable SAL_VCL_QT_USE_QT_FRAME_GRABBER. (cherry picked from commit 08533ca4e2526644b803c40c0c3d3c96f43762af) |
Michael Weghorn <m.weghorn@posteo.de> | no | 2025-05-02 | ||
| qt-Consolidate-to-one-toOUString-helper.diff | qt: Consolidate to one toOUString helper Instead of having one for vcl in vcl/inc/qt5/QtTools.hxx and one in avmedia/source/qt6/QtFrameGrabber.cxx, move the existing implementation to include/vcl/qt/QtUtils.hxx and use it everywhere. (cherry picked from commit d023035acf83ee1b61dfc03333bfc6e612bb58f6) |
Michael Weghorn <m.weghorn@posteo.de> | no | 2025-04-11 | ||
| default-to-Euro-for-Bulgaria.diff | default to EUR for Bulgaria Bulgaria will switch to EUR 2026-01-01. |
Rene Engelhard <rene@rene-engelhard.de> | no | 2025-11-09 | ||
| Conform-AlignEngine-parsing-to-spec.diff | Conform AlignEngine parsing to what section 2.3.4.10 of the spec has | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-03-16 | ||
| CVE-2026-6039.diff | stay within max Polygon points | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-09 | ||
| CVE-2026-6040.diff | process loext:blank-width-char better | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-09 | ||
| CVE-2026-6045.diff | check that the file can provide the claimed data and make sure we initialize these locals |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-09 | ||
| CVE-2026-8356.diff | SdrEscherImport::RecolorGraphic reads but doesn't use FillColors it never did, but at: commit 025bfa7e510bdab3ef93ad45a731fc25085ba3cc cppcheck: unreadVariable the wrong choice was made wrt the dead store, the dead loop should have been fixed rather than let the unused fill colors accumulate in the globalcolors buffer. This function doesn't need to consume the unused data, all callers don't expect the entire record to be read so just drop the loop. (cherry picked from commit 07faeadace3d8d93225f9dd72c4f911e331070e2) (cherry picked from commit 9ba100c0d2131862c9356a86906921abc80d8f98) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2015-04-23 | ||
| CVE-2026-8357.diff | A formula of length L, composed entirely of open tokens, needs L+1 slots (cherry picked from commit 0b7a9149d7c7b5e044bb4f02668297bdf41b64d6) (cherry picked from commit a28cead4b7796642d946173b479281a6272117f2) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-26 | ||
| CVE-2026-8358.diff | drop malformed duplicate-id calc change track actions (cherry picked from commit b5b56a2239369e45d7fac2ea965558a1f337b602) (cherry picked from commit a20ae7b5f80a45e49a47ce22f5e92749cd9816a0) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-26 | ||
| check-for-hb_shape_full-failure.diff | check for hb_shape_full failure (cherry picked from commit 22c19be8fae977a90f83a82d8acccd0b46c7cf20) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-06 | ||
| CVE-2026-63272.diff | ofz: ignore an advance array shorter than the string (cherry picked from commit 6ab602ca281f507f7e637183a49dd514ffb1376b) (cherry picked from commit 4406da79b0715411528c17f93ae7ebcb29b9a0d2) (cherry picked from commit acf64b67cceecf3d1acd26033668b6673208bcf2) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-04 | ||
| CVE-2026-63273.diff | ofz: pdf, check record againsts max allowed size (cherry picked from commit e230b10b7cd2d5d8623cdba0877ec68f67003690) (cherry picked from commit 21124741e0134f0f1222a42726c223bf6da41dca) (cherry picked from commit 3a92ea5af6d1bf3820ce0909fb9de003565ebf2b) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-04 | ||
| CVE-2026-63274.diff | ofz: pdf, clamp stream Length to the bytes actually read (cherry picked from commit 332e8831e0b0e580dae0d77a0ec27bd267af6b29) (cherry picked from commit 7906b5a8aac335a0155a90fc6f8310a92e558124) (cherry picked from commit b4b7fcc1f0551ccdba1ddc520c351eee9bfccd4c) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-04 | ||
| CVE-2026-63275.diff | ofz: wmf, check record against max allowed size (cherry picked from commit 9b412f7cda30a4e1b820e51249fa7b3c36a812d3) (cherry picked from commit 1db305f340787ab7e59d089e51ac9d6fc07c2715) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-04 | ||
| CVE-2026-63276.diff | ofz: cff, track the charstring output capacity (cherry picked from commit 0b090eff6ca18e3f2bacc7311b746d2a15ed1247) (cherry picked from commit 49360f34f5caff4eee29489e339def8cf13053cc) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-05 | ||
| CVE-2026-63278.diff | match the package content provider when checking a vnd.sun.star.pkg url IsExoticProtocol used GetHost() to recover the nested package url of a vnd.sun.star.pkg url, but the package content provider in ucb/source/ucp/package/pkguri.cxx uses the whole authority between "://" and the next '/', which GetHost() truncates at '@'. Check the same substring the provider does. The vnd.sun.star.zip branch below already did the equivalent extraction, so pull the shared parts into a helper that both branches call. (cherry picked from commit ce436efb34d2e0e90ee2cdb450da285530c562f7) (cherry picked from commit 4ba5567cc44be760066c0c45c99949f790cf181a) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-06-12 | ||
| CVE-2026-63279.diff | limit palette index use the same SanitizePaletteIndex guard ipsd and iras already have. (cherry picked from commit ada76821319b80671dba9908aeea8cabba5cedc0) (cherry picked from commit 6bc93227e7c603d3f314202b89b6c567680e6261) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-06-13 | ||
| CVE-2026-18566-6.diff | ofz: drawing, better calculation for how many points could be in the stream (cherry picked from commit d75f91c68970a4ef1397c881ab9f37002fe1e462) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-18566-7.diff | ofz: emfio, return early for bogus Polygon nextpos (cherry picked from commit 77127124cf8910574abf16ee7079c140086fc995) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-18566-8.diff | wmf: initialize each palette entry and stop reading on stream failure (cherry picked from commit 9d70e966bc709bd10f96f9ca699778b937c6dcdf) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-18566-9.diff | ofz: emfio, return early for bogus comment counts | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-18566-10.diff | ofz: emfio, remember where the first free object slot is | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-18566-11.diff | emf+: give the stream-read locals a starting value ReadUInt32 leaves the variable alone if the record ends early. |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-26 | ||
| CVE-2026-84234.diff | pdf import: check the sizes before building the png scan lines dismiss unreasonable sizes and ImageStream::getLine returns null when it could not allocate its line buffer. |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-01 | ||
| CVE-2026-98100.diff | hwp: take the picture link target from inside the follow block | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-16 | ||
| CVE-2026-101351.diff | start with RawBitmap pixels blanked so pixels that a reader never writes, for example when it stops a row early, are black rather than arbitrary noise |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-28 | ||
| CVE-2026-63277.diff | translate only file URLs (cherry picked from commit 478eb7de0d14a05e6501bd44851f73ec0df5c59a) (cherry picked from commit e663f8baed49976fd4d4d6d1c277ab036cd28736) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-06-12 | ||
| do-not-load-exotic-protocols-for-document-supplied-data.diff | don't bother loading exotic protocols for document-supplied data (cherry picked from commit 0274803ae273fa6bdbd1124cedf37395ce847148) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-11 | ||
| CVE-2026-101352-1.diff | MasterScriptProvider::getScript: match the location to the context exactly (cherry picked from commit ae5d7f7ce564b4cd4656366ee88afa762e457c8f) (cherry picked from commit 9919c425762c21dac30b7bfef9c11a5f9847aa95) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-63269-1.diff | limit the gstreamer backend to simple self-contained media The adaptive-streaming cases follow a more complicated indirect route of a manifest to further resources which is atypical for our use. (cherry picked from commit c7eada5bc2233369e196f2c93e7606b56e3acc62) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-13 | ||
| CVE-2026-63269-2.diff | put media files under link update control A media object whose content lives at an external URL, rather than inside the document, is registered with the link manager like a linked graphic. So it then shares the usual link update mechanism. Media copied into the document is extracted to a temp file and is not a link, so that's left alone. (cherry picked from commit 1ea4fc54036587c7eb4d4cea74e776de9e6baaf3) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-13 | ||
| CVE-2026-63269-3.diff | put slideshow media links under link control (cherry picked from commit f0768394c14f928d47547408fc1195368be81f86) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-14 | ||
| Show-infobar-for-calc-graphics-with-remote-content-too.diff | Show infobar for calc graphics with remote content too (cherry picked from commit 1f51bbee0e7661d5607efe2603c59d83f410dd3d) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-04-01 | ||
| CVE-2026-63267-1.diff | put calc external data mappings under link update control register each mapping as a LinkManager link, so its refresh goes through the same update control as sheet links and area links. Wait until after the infobar-controlled update mode is known to update. (cherry picked from commit b72d760465c857eb5dd10cdc45b107c49bb0caec) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-11 | ||
| CVE-2026-63267-2.diff | sc: check the host when fetching an external data range checked only the file path allowlist. reject hosts outside it too, like ScExternalRefManager and ScWebServiceLink. (cherry picked from commit 29b5cb45a18ca79b28ce3a02be22803d94a36368) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-11 | ||
| CVE-2026-63268-1.diff | sc: quote the table name in the sql data provider query (cherry picked from commit 1d24f1b1937995936011f61b6c1436d821f66faf) (cherry picked from commit d7712feea1d1992c62f72aeeec32c2053556898b) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-12 | ||
| CVE-2026-63268-2.diff | sc: only build the supported data providers when loading a document Build a data mapping only for the csv, html and xml providers. Any other provider name in the document is ignored. The sql provider is not included. It never worked and was dropped from the Data Provider dialog in tdf#169079. (cherry picked from commit 9b6d7ec607e68c79ce685f126139ef826776e783) (cherry picked from commit 104d2b4f5dae917661b20b18d5d2043fca4407e4) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-12 | ||
| CVE-2026-63266-2.diff | Reapply "firebird: keep each embedded database's files in one directory" This reverts commit 23a5d0dc305fc04bdd277b24db8d5c8e41aa9ccd. it also contains https://gerrit.collaboraoffice.com/c/core/+/7093 "firebird: upper-case the DatabaseAccess Restrict path on Windows" (cherry picked from commit e4dc6a3cb2bfbf9d4c1e30e777b0f46417cf4c92) |
Xisco Fauli <xiscofauli@libreoffice.org> | no | 2026-07-17 | ||
| CVE-2026-63266-4.diff | firebird: only write back an embedded database that was opened When construct throws before the attach succeeds, the connection is disposed with the temporary .fdb extracted but never opened, but storeDatabase ran the backup service against it anyway. Note whether the database was opened and skip the write back when it was not. (cherry picked from commit a73db8b145af1c74cb181e094c0b7ccc4d92d8f0) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-12 | ||
| CVE-2026-63266-1.diff | firebird: don't run an attached database's own event triggers for the embedded and file cases suppress the database's ON CONNECT and similar event triggers when attaching, they should not fire just because the document was opened (cherry picked from commit 36335ff89ea2309e99da02b449c9b35517cd6f9d) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-12 | ||
| CVE-2026-63266-3.diff | firebird: don't attach a database that is not in the normal backup state A database we extract from a .odb is always in the normal backup state, so refuse one that is not, we're not interested in the difference file mode case. firebird upstream attempt as: https://github.com/FirebirdSQL/firebird/pull/9089 And do it manually for the system-firebird case. (cherry picked from commit e7ea381e87cadaf31bf1ad2213a23f17a6284d1c) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-12 | ||
| firebird-db-connection.diff | commit 16cf06caff90c0e4d9fe2fc24eac378ed533c48d Resolves: tdf#172935 Connection from Base to external firebird database refused there is one firebird engine per process and it reads its configuration once from the directory we set up, so a .fdb the user picked somewhere else could not be attached. firebird resolves a name in databases.conf before it treats it as a path, so give the file a name there and pass firebird that name, dropped again when the connection is disposed. since: commit 520d8173cef39a4da85b3ba21bdaaa3150384c25 Date: Sun Jul 12 12:10:41 2026 +0000 firebird: keep each embedded database's files in one directory Change-Id: I4872a4e2165c4701bfb20be5dcd8c119b1723125 Signed-off-by: Caolán McNamara <caolan.mcnamara@collabora.com> Reviewed-on: https://gerrit.collaboraoffice.com/c/core/+/7990 Tested-by: Jenkins CPCI <releng@collaboraoffice.com> Reviewed-by: Michael Stahl <michael.stahl@collabora.com> (cherry picked from commit 6ab71a109bd8a8335208c9faa37a3cfa94523148) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/208773 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> (cherry picked from commit 99bf5222896886d062b6a6a67377e7671de1446a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/208855 diff --git a/connectivity/source/drivers/firebird/Connection.cxx b/connectivity/source/drivers/firebird/Connection.cxx index 3b918e696d87..c4b9126cd0c9 100644 |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-28 | ||
| firebird-module-path.diff | firebird: name the modules of a system firebird by absolute path Debian patches firebird to resolve its plugins, intl and conf directories under the FIREBIRD variable (instead of where it was built) and we want point that variable at a private directory. But then the modules aren't loaded and no database opens. So write a replacement plugins.conf that lists the absolute path of each plugin module (and the intl conf files) with their macros expanded. each lib<Name>.so (otherwise found by default) with no entry names gets an entry of its own. (cherry picked from commit b8407b18adf3c6d552d54d7b6657539e2769afd6) (cherry picked from commit 09dbbe16b04add973ab04fe5937c12478e52451c) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-28 | ||
| firebird-call-writeFile-outside-the-SAL_WARN_IF-cond.patch | firebird: call writeFile outside the SAL_WARN_IF condition a product build doesn't evaluate the condition, so none of the conf files were written there. since: commit b8407b18adf3c6d552d54d7b6657539e2769afd6 firebird: name the modules of a system firebird by absolute path and: commit 6ab71a109bd8a8335208c9faa37a3cfa94523148 Resolves: tdf#172935 Connection from Base to external firebird database refused |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-28 | ||
| CVE-2026-76181.diff | diff --git a/connectivity/source/drivers/dbase/DIndex.cxx b/connectivity/source/drivers/dbase/DIndex.cxx index 35a4087c0482..2c64b51b01e9 100644 |
no | ||||
| CVE-2026-18563-1.diff | ofz: check buffer bounds before reading multi-byte CFF dict operators (cherry picked from commit fa2ed413a2d1a6766d43ed1d094c2ce651c68ebb) (cherry picked from commit 1857f907a6f8a5ff8154bbbf0790475040e809b7) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-15 | ||
| CVE-2026-18563-2.diff | ofz#535775262 Out-of-memory (cherry picked from commit 7bbe72cd3bdced0c8617fc0246bea9ed5a09332f) (cherry picked from commit 771e094dd840e33955dc8263225a2ed12242547f) (cherry picked from commit c7ed860db049b62eed1e26c9eed79e47f42e3454) (cherry picked from commit 8a4330439c866fdd59f85880426ccad562c82f9e) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-19 | ||
| CVE-2026-18563-3.diff | ofz: cff, keep the operand index inside ValStack (cherry picked from commit c6bbeb56705e0212d9cfdf1cc9210e8f8bc833a9) (cherry picked from commit 9e04fdeb1ff80b0a66c3317504b5154f2c3709d3) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-29 | ||
| CVE-2026-18563-4.diff | ofz#533311278 stay within limits where the base passed in was just taken from getReadOfs() the reassignment of mpReadPtr in the helper is a no-op (cherry picked from commit 70ee40fd48b13bbedad12d096e96af6d898371a8) (cherry picked from commit 8db20269a8095a81bc0a8c0bfe36342a7fc6ff4a) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-14 | ||
| CVE-2026-18563-5.diff | ofz: check buffer bounds before reading offsets in CFF seekIndexData Same pattern as: commit 304208502164c07df8eb5a3677cb445ab0965f0e ofz#513301090 UNKNOWN READ (cherry picked from commit 7c237925e474326f3c8dea179264b60d835145d7) (cherry picked from commit 99f5d81441424ab097b06843c960102c90ea9894) (cherry picked from commit 70eccad0a34f419a88fafdc585f061d6ff495da3) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-15 | ||
| CVE-2026-18563-6.diff | ofz#513277199 Integer-overflow (cherry picked from commit 6eb214dbe24ac7c1cde3fe8a6ce242f2d7e24151) (cherry picked from commit 073913a2cc102f14aa54eaf9d1ecfa3bd44656fb) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-15 | ||
| CVE-2026-18563-7.diff | ofz#513301090 UNKNOWN READ (cherry picked from commit f685107dc8c93ee92761b0418a4c0e37b4f97174) (cherry picked from commit 760a5b7c6647788ef4608a927e451d5528b07774) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-15 | ||
| CVE-2026-18563-8.diff | ofz: initialCffRead fails during smoketest (cherry picked from commit 15d96dba8b96f1248840da1bafa3b564b2ad2a95) (cherry picked from commit 0bda4b8824202ff97b32179bd88df538e0a573a1) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-21 | ||
| CVE-2026-18563-9.diff | ofz#536319253 stay within buffer (cherry picked from commit 0d43de913089afdb4f8153fc74684dbc013db733) (cherry picked from commit b5086bd01e2d40cdbe3cb5116fa8a84c1d81ead4) (cherry picked from commit c45a9221b58c0093e24dd2fd084a9aca86db2621) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-19 | ||
| CVE-2026-18563-10.diff | ofz: check for failure here too (cherry picked from commit 51fa2756479e99d1fc8d804d50bfd969d582ebdd) (cherry picked from commit f62f3a2552e90b464c37e50a2cdbed0ffcc09917) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-05-15 | ||
| CVE-2026-18563-11.diff | ofz#532839163 Integer-overflow (cherry picked from commit fd90f5eb7e514a2975d034a1d9bc5cbd77cab820) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-09 | ||
| CVE-2026-18563-12.diff | vcl: skip the bytes reserved in the CFF DICT encoding they are neither operators nor operands, so step over them and carry on reading the rest of the dict taken from the cff.cxx part of commit f799626f98b227d997c4d149ef3e7b0eab888406 Author: Caolán McNamara <caolan.mcnamara@collabora.com> Date: Fri Apr 10 11:04:03 2026 +0100 revive sftfuzzer for a different code path (cherry picked from commit a0343839bdc63b5d09e543b52e23fd71ce87a20e) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-30 | ||
| CVE-2026-18563-13.diff | Assert less in CFF to Type 1 code The top-level function already returns a boolean, make many of the internal functions return false in case of failure and propagate that instead of asserting in debug builds and continuing with bad input in release builds. |
Khaled Hosny <khaled@aliftype.com> | no | 2026-03-08 | ||
| CVE-2026-18564.diff | oox,svl,vcl: use more appropriate random numbers (cherry picked from commit 52ba487c3d84a89a027c3130bc86afdcfe57a860) (cherry picked from commit 63ab1f905bddbac4871a246d61aa59d09f122e5c) (cherry picked from commit 396747a0e319943e7d411b0d6bc8d8abcc3e4f19) |
Michael Stahl <michael.stahl@collabora.com> | no | 2026-07-30 | ||
| CVE-2026-18565.diff | pass a referer through setSvxBrushItemAsFillAttributesToTargetSet SvxBrushItem::GetGraphic wants the name of the document the brush came from, so take it as an argument and hand in SwDoc::GetLinkReferer wherever a document is at hand, the Writer style, frame and background shell paths included. The dialog and example-window callers pass the name of the active document shell through a new svx helper getActiveDocumentLinkReferer. The hand-rolled copies of the GetPersist, HasName, GetMedium GetName sequence elsewhere in sw now call SwDoc::GetLinkReferer too. (cherry picked from commit 18f91ac71710027479dfe9f206ae2207bad8f2b7) (cherry picked from commit ecd9ee7aa1fe69399fbe0eb59d1b847e80851fd5) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-30 | ||
| CVE-2026-18566-1.diff | ofz: emf+, return early on short read (cherry picked from commit 6fa74f27a0717afed4fd8b8c45442f3dbea31a7e) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-24 | ||
| CVE-2026-18566-2.diff | ofz: emf+, check if enough data is available | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-25 | ||
| CVE-2026-18566-3.diff | ofz: emf+, null deref (cherry picked from commit 7c603236253501c5ceceece3c6b4dd5b1e580b5d) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-25 | ||
| CVE-2026-18566-4.diff | ofz: drawing, don't allow pops past starting depth since: commit bf8660708144ccb693063a20954f64a66a1c979d #i112117# changed the Metafile primitive decomposition to better handle sub-metafile contents which moved the pop out of the flags check that paired it with the push, so an empty PUSH, or more POPs than PUSHes, drains the stack and Current() hands back its fallback static (cherry picked from commit 9f1bf5b844295a532cc9e93f84acb44ed345294c) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2010-06-04 | ||
| CVE-2026-18566-5.diff | ofz: drawing, check if offset is correct (cherry picked from commit 67848b0ef12d7817719afa5eb597bc2c0a316f80) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-07-25 | ||
| CVE-2026-101352-2.diff | handle customized bindings of a document like its other content (cherry picked from commit 8c25f8c10100f3f2c93169303752faebced64f90) (cherry picked from commit 70ac0cabc5c94492cf81cd55cfc87f851bbc1e3d) (cherry picked from commit 9d375606275c5e71391d905e68d2e59a51d9833f) (cherry picked from commit 1e39ec81097975c16255ee4d138b732a00ab7660) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-101352-3.diff | reuse INetURLObject::IsExoticProtocol for other binding command schemes (cherry picked from commit 770fcb69cb232b9c9cbb17304aca9fc963621dc1) (cherry picked from commit ce19ab84625995f197887cf3559b1d1ce2a2f087) (cherry picked from commit e0ac8ad7e95e6cb1b1d3248950b352a59ca7a839) (cherry picked from commit 46a149c2d0ed40dee5bafa7b0c36f0c7d4ec04ea) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-101353-1.diff | hsqldb: check an existing lock file rather than taking it over backport of upstream HSQLDB svn r3155, from the 1.8.1 release |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-28 | ||
| CVE-2026-101353-2.diff | hsqldb: keep the files of an embedded database in a private directory hsqldb writes its own files next to the database path, and that path was the document's location. Put each embedded database in a subdirectory of a directory the driver owns and removes. |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-29 | ||
| CVE-2026-101353-3.diff | hsqldb: read textdb.allow_full_path only as a Java system property as hsqldb 2.x does since 2.1.0 |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-29 | ||
| CVE-2026-101353-4.diff | hsqldb: no text table files for an embedded database hsqldb already refuses text tables for a database that is not file based, so do the same for a database kept inside the document. An existing text table in such a database then opens disconnected. |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-29 | ||
| CVE-2026-104007-1.diff | dbaccess: take JavaDriverClassPath only from the driver configuration No UI ever wrote this setting into a database document. None of the (631) .odb files in the crashtesting test corpus have it set. (cherry picked from commit aae21cae8f686c0f4d43fcbee8b478aae8d60ff1) (cherry picked from commit de88188e2cd8cd78ca46ace27c0a16a48ee538bb) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-01 | ||
| CVE-2026-104007-2.diff | put calc pivot table database sources under link control like refreshing database range imports, so a pivot table whose source is a database document named by URL, rather than a registered data source, waits until link updates are allowed before it opens it, and counts as an external link so the update links infobar is offered (cherry picked from commit 0d27a2394257fe817160ed4092aecd4168b792e5) (cherry picked from commit 0eaa944aa5b0f743c9f58fb2f757eedc619773b9) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-01 | ||
| CVE-2026-104007-3.diff | dbaccess: take SystemProperties only from the driver configuration as with JavaDriverClassPath, a database document no longer supplies them (cherry picked from commit 1cb0c88c59759fad0956c6b1ab2383b2625faa05) (cherry picked from commit b8724f1ec896052986bb9ca7b71f4b0acfbbc6c5) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-09 | ||
| CVE-2026-104692-1.diff | hsqldb: treat an unset hsqldb.method_class_names like an empty one | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-01 | ||
| CVE-2026-104692-2.diff | start the JVM with an empty hsqldb.method_class_names so a system hsqldb gets the same default as the bundled one |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-01 | ||
| CVE-2026-101892.diff | only use the form import object when it is a form component | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-28 | ||
| CVE-2026-101929.diff | clear the old value before loading an old format SbxVariable as the newer format already does The old format path set the new type without clearing an existing value first, so the new type bogusly redefined whatever the union still held. (cherry picked from commit 35a9e5d20be854e8392795d06c168eca6d270a21) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-26 | ||
| CVE-2026-101930.diff | note a document event from HTML import like the other HTML events | Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-101933-1.diff | keep each string of a Basic image string pool inside the pool an image with string offsets outside its pool is not loaded (cherry picked from commit fc0c18c508cd34bc8a17d534193de8992acdade0) (cherry picked from commit 0e705d4a0b812d671814e38db3ca2fc97560c8c8) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-101933-2.diff | check the class of a nested user type found by name in SbiImage::Load (cherry picked from commit 48162866845dd588ec0a85a9e9e93ed4477d7771) (cherry picked from commit 0f2d89705944d97aee45a17c560825d0df6591e8) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-27 | ||
| CVE-2026-104920.diff | a Java class path entry has to be a plain local file URL with no host and an absolute path follow on from "translate only file URLs" (Change-Id I41076b8af670d673bf29b14c02db0c66744feb0b) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-22 | ||
| CVE-2026-84443-1.diff | move isScriptURLAllowed onto SfxObjectShell it was private to sfx2's eventsupplier.hxx, so nothing outside sfx2 could run the AllowedDocumentEventURLs test |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-01 | ||
| CVE-2026-84443-2.diff | dbaccess: use the same script checks as an office document a database document isn't an SfxObjectShell, so its bound event scripts run from DocumentEventExecutor rather than from SfxEvents_Impl::Execute (cherry picked from commit 7993a7a1b0cccba429fb82ca3723d2ef50a39277) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-01 | ||
| CVE-2026-84443-3.diff | dbaccess: decide the macro execution mode during load the mode was only decided when the first controller connected, which is later than the other document types do it |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-01 | ||
| e12c061dbf1330b00cdc73b7691171b6af883764.diff | #diff --git a/sw/qa/extras/ooxmlexport/ooxmlexport25.cxx b/sw/qa/extras/ooxmlexport/ooxmlexport25.cxx #index c160cb326996..fc1fe5637b56 100644 #--- a/sw/qa/extras/ooxmlexport/ooxmlexport25.cxx #+++ b/sw/qa/extras/ooxmlexport/ooxmlexport25.cxx #@@ -223,6 +223,10 @@ CPPUNIT_TEST_FIXTURE(Test, testTdf169101_datePicker) # // there is no valid date, so fullDate must not be provided (or MS Word says 'corrupt file') # assertXPathNoAttribute(pXmlDoc, "//w:sdt/w:sdtPr/w:date", "fullDate"); # assertXPathContent(pXmlDoc, "//w:sdt/w:sdtContent/w:r/w:t", u"Week #"); #+ #+ // if the relationship is missing, then MS Word shows placeholder text instead of customXml date #+ xmlDocUniquePtr pXmlCustomXmlRels = parseExport(u"customXml/_rels/item1.xml.rels"_ustr); #+ assertXPath(pXmlCustomXmlRels, "//rels:Relationship", "Target", u"itemProps1.xml"); # } # # CPPUNIT_TEST_FIXTURE(Test, testTdf170389_manyTabstops) diff --git a/sw/source/filter/ww8/docxexport.cxx b/sw/source/filter/ww8/docxexport.cxx index 9a6a9e159898..0f07bc346518 100644 |
no | ||||
| CVE-2026-87932-1.diff | docx export: set custom xml values through XPathAPI instead of XSLT each data binding built its own stylesheet and its own transformation pass over the item, because different bindings can use conflicting namespace prefixes The import side already resolves the same binding with XPathAPI and RegisterDataBindingNamespaces, so select the nodes that way here too and set their text on a copy of the custom xml. (cherry picked from commit fe2f99ef74ef02e803aaf01dcc2b7f6e1cea67f7) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-08-24 | ||
| CVE-2026-87932-2.diff | docx export: write each bound value only into its own customXml item every custom xml item was given the values of every data-bound control, so a document with two data stores came back with both controls showing the same value, and testTdf147724 failed in ci with: xmltesttools.cxx:214:Assertion Test name: testTdf147724::Load_Verify_Reload_Verify equality assertion failed - Expected: Placeholder -> *ABC* - Actual : Placeholder -> *HERUNTERLADEN* - In <>, XPath contents of child does not match the second control's data binding names no store, so the import picks a store by std::unordered_map iteration order. That order differs from one builder to the next, and the test accepts either value for that control, so only some machines saw the failure. since: commit b5c616d10bff3213840d4893d13b4493de71fa56 tdf#104823: support for sdt plain text fields so keep the storeItemID of the data binding along with the value and only update the item whose ds:itemID matches it (cherry picked from commit b522379deae4c9df243d0493a3d242e80822c171) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2021-11-24 | ||
| CVE-2026-87932-3.diff | move the data binding namespace registration to a shared header the docx export needs to register the same prefixes before it can evaluate an xpath from the data binding (cherry picked from commit b482d8a3f019ae725f8242cbd7f1e258ab2b6cbd) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-08-24 | ||
| CVE-2026-87932-4.diff | help: give every libxslt preference a value, not just one of them A preference left unset keeps whatever libxslt defaults to, so name each one instead of only XSLT_SECPREF_READ_FILE. Reads over the network are allowed as well now, as a help page pulls in another page through a vnd.sun.star.help URL and libxslt counts any scheme other than file as a network read. (cherry picked from commit d8923fe423af48cb02a60b01ec881aa12f6c14ca) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-08-24 | ||
| CVE-2026-87932-5.diff | xslt filter: set every libxslt transformation option explicitly libxslt leaves each option unset on a fresh transformation context and takes its own default for it. The service registers no libxml input handlers, and a stylesheet only has to produce the result tree, so set all of them and allow back local file reads alone, which is how a stylesheet picks up a second file beside the one it converts. (cherry picked from commit 02da3650b1bb71170112a480c7fa0220a2a34490) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-08-24 | ||
| CVE-2026-97892.diff | check the byte ranges of a pdf signature against the file The offsets and lengths come from the document, so a range that is not inside the file leaves the signature unchecked now. |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-22 | ||
| CVE-2026-97894.diff | ReadLine and ReadUniStringLine: stop at the limit in the loop the two have drifted apart, so check the limit as each character is read instead of correcting the count afterwards |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-09-15 | ||
| CVE-2026-97999.diff | firebird: size the setBytes grow check by sqllen 8000 was the default width the hsqldb migration gives a VARBINARY column, but sqldata is allocated from sqllen, which is less than that for a narrower column, so test against sqllen instead since: commit 54ddc4ff4c2ff7e8b2c502d6b475cfdc9b8e3cec tdf#117446: FB mig, (VAR)BINARY, fix memory management |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2018-05-26 | ||
| CVE-2026-104933.diff | gstreamer: only autoplug demuxers for common self-contained media formats rather than dropping just the Adaptive ones (cherry picked from commit a19df93839c759c1b18a4dc7738265349cb19326) (cherry picked from commit f9e9e278a26cf08b5c69dff2965f984c905da1a1) (cherry picked from commit 374e12122c3301ff4448a9c7a63518eddd98bcdc) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2026-10-02 | ||
| CVE-2026-108169.diff | Related: tdf#106359 only take an http or https url from a web query file since: commit 099f9406b6657a03a37fbe5cedf7cb7da765d4aa tdf#106359: Open Web Query (.iqy) files a web query names a web page, so leave the argument alone when the line found there is some other kind of url (cherry picked from commit e8e42c773264f0a70a39d164b34f440eefb05f16) (cherry picked from commit 8b1b10e802a36463683aca9c3e9aeef4b5199cbf) (cherry picked from commit c4d6d681499236935b3bd825efb6c6dcc7a5de8a) |
Caolán McNamara <caolan.mcnamara@collabora.com> | no | 2017-03-06 |
All known versions for source package 'libreoffice'
- 4:26.8.3.1-3 (sid)
- 4:26.8.2.2-2 (forky)
- 4:26.8.2.2-2~bpo13+1 (trixie-backports)
- 4:25.2.3-2+deb13u12 (trixie-security)
- 4:25.2.3-2+deb13u9 (trixie-proposed-updates)
- 4:25.2.3-2+deb13u6 (trixie)
- 4:25.2.3-2+deb13u5~bpo12+1 (bookworm-backports)
- 4:7.4.7-1+deb12u14 (bookworm)
- 4:7.4.7-1+deb12u13 (bookworm-security)
