Please note that this mail was generated by a script. The described changes are computed based on the x86_64 DVD. The full online repo contains too many changes to be listed here. Please check the known defects of this snapshot before upgrading: https://openqa.opensuse.org/tests/overview?distri=microos&groupid=1&version=Tumbleweed&build=20221124 https://bugzilla.opensuse.org/buglist.cgi?product=openSUSE%20Tumbleweed&component=MicroOS&query_format=advanced&resolution=--- Please do not reply to this email to report issues, rather file a bug on bugzilla.opensuse.org. For more information on filing bugs please see https://en.opensuse.org/openSUSE:Submitting_bug_reports Packages changed: apparmor (3.0.7 -> 3.1.2) busybox conmon (2.1.3 -> 2.1.5) hicolor-icon-theme inxi (3.3.21 -> 3.3.23) libapparmor (3.0.7 -> 3.1.2) libdrm memcached nano (6.4 -> 7.0) nghttp2 (1.50.0 -> 1.51.0) openvpn (2.5.7 -> 2.5.8) perl-IO-Socket-SSL (2.075 -> 2.077) python-httpcore python-jsonschema (4.17.0 -> 4.17.1) python-pylibmc (1.6.1 -> 1.6.3) schily vim (9.0.0814 -> 9.0.0924) === Details === ==== apparmor ==== Version update (3.0.7 -> 3.1.2) Subpackages: apparmor-abstractions apparmor-parser apparmor-parser-lang apparmor-profiles apparmor-utils apparmor-utils-lang python3-apparmor - update to AppArmor 3.1.2 - lots of cleanups, improvements and bugfixes in all areas - rework internal profile storage and handling in the aa-* tools - support boolean variable definitions in the aa-* tools - see https://gitlab.com/apparmor/apparmor/-/wikis/Release_Notes_3.1.1 and https://gitlab.com/apparmor/apparmor/-/wikis/Release_Notes_3.1.2 for the detailed upstream changelog - remove upstream(ed) patches: - apparmor-3.0.7-egrep.patch - dnsmasq.diff - profiles-permit-php-fpm-pid-files-directly-under-run.patch - zgrep-profile-mr870.diff - no longer ship precompiled profile cache for Tumbleweed (boo#1205659) - BuildRequire iproute2 (needed for aa-unconfined tests) ==== busybox ==== Subpackages: busybox-static - Add e63d7cdf.patch: awk: fix use after free (CVE-2022-30065, boo#1199744). ==== conmon ==== Version update (2.1.3 -> 2.1.5) - New upstream release 2.1.5: Bug Fixes don't leak syslog_identifier logging: do not read more that the buf size logging: fix error handling Misc Makefile: Fix install for FreeBSD signal: Track changes to get_signal_descriptor in the FreeBSD version Packit: initial enablement - New upstream release 2.1.4: Bug fixes Fix a bug where conmon crashed when it got a SIGCHLD ==== hicolor-icon-theme ==== - Add 1024x1024 directory for package ownership ==== inxi ==== Version update (3.3.21 -> 3.3.23) - Update to version 3.3.23: + /usr/share/doc/packages/inxi/inxi.changelog. ==== libapparmor ==== Version update (3.0.7 -> 3.1.2) - update to AppArmor 3.1.2 - lots of cleanups, improvements and bugfixes in all areas - rework internal profile storage and handling in the aa-* tools - support boolean variable definitions in the aa-* tools - see https://gitlab.com/apparmor/apparmor/-/wikis/Release_Notes_3.1.1 and https://gitlab.com/apparmor/apparmor/-/wikis/Release_Notes_3.1.2 for the detailed upstream changelog - remove upstream(ed) patches: - apparmor-3.0.7-egrep.patch - dnsmasq.diff - profiles-permit-php-fpm-pid-files-directly-under-run.patch - zgrep-profile-mr870.diff - no longer ship precompiled profile cache for Tumbleweed (boo#1205659) - BuildRequire iproute2 (needed for aa-unconfined tests) ==== libdrm ==== Subpackages: libdrm2 libdrm_amdgpu1 libdrm_intel1 libdrm_nouveau2 libdrm_radeon1 - Apply n_libdrm-drop-valgrind-dep-generic.patch and n_libdrm-drop-valgrind-dep-intel.patch only when the build uses meson < 0.64. With meson 0.64, we don't get the dependency on valgraind added. - split n_libdrm-drop-valgrind-dep.patch into n_libdrm-drop-valgrind-dep-generic.patch and n_libdrm-drop-valgrind-dep-intel.patch to fix build on s390 and armv7l - Only apply libdrm-drop-valgrind-dep.patch if valgrnid_support is enabled (fix build on e.g. aarch64). - renamed libdrm-drop-valgrind-dep.patch to n_libdrm-drop-valgrind-dep.patch in order to mark it as 'never to be upstreamed' - Add libdrm-drop-valgrind-dep.patch (as source): drop dependency on valgrind on generated pkgconfig files. The .pc files are auto-generated by meson and are 'technically' correct, but we do not want to inject valgrind here (we can get away with this hack as it's only relevant when using pkg-config --static, and we do not provide static libs anyway). ==== memcached ==== - remove sysv init case - move memcache binary to %{_bindir} like on Debian and RHEL like distributions ==== nano ==== Version update (6.4 -> 7.0) Subpackages: nano-lang - update to 7.0: * String binds may contain bindable function names between braces * Unicode codes can be entered (via M-V) without leading zeroes, by finishing short codes with <Space> or <Enter> * Word completion (^]) looks for candidates in all open buffers * No regular expression matches the final empty line any more ==== nghttp2 ==== Version update (1.50.0 -> 1.51.0) - update to 1.51.0: * https://nghttp2.org/blog/2022/11/13/nghttp2-v1-51-0/ This release fixes affinity-cookie-stickiness parameter handling. ==== openvpn ==== Version update (2.5.7 -> 2.5.8) Subpackages: openvpn-auth-pam-plugin - update to 2.5.8: * allow running a default configuration with TLS libraries without BF-CBC (even if TLS cipher negotiation would not actually use BF-CBC, the long-term compatibility "default cipher BF-CBC" would trigger an error on such TLS libraries) * ``--auth-nocache'' was not always correctly clearing username+password after a renegotiation * ensure that auth-token received from server is cleared if requested by the management interface ("forget password" or automatically via ``--management-forget-disconnect'') * in a setup without username+password, but with auth-token and auth-token-username pushed by the server, OpenVPN would start asking for username+password on token expiry. Fix. * using ``--auth-token`` together with ``--management-client-auth`` (on the server) would lead to TLS keys getting out of sync and client being disconnected. Fix. * management interface would sometimes get stuck if client and server try to write something simultaneously. Fix by allowing a limited level of recursion in virtual_output_callback() * fix management interface not returning ERROR:/SUCCESS: response on "signal SIGxxx" commands when in HOLD state * tls-crypt-v2: abort connection if client-key is too short * make man page agree with actual code on replay-window backtrag log message * remove useless empty line from CR_RESPONSE message ==== perl-IO-Socket-SSL ==== Version update (2.075 -> 2.077) - updated to 2.077 see /usr/share/doc/packages/perl-IO-Socket-SSL/Changes 2.077 2022/11/21 - fix memory leak in session cache, thanks to genuaboro https://github.com/noxxi/p5-io-socket-ssl/pull/118 - more race conditions in tests fixed thanks to jddurand https://github.com/noxxi/p5-io-socket-ssl/issues/97 2.076 2022/11/12 - added curl like tracing based on contribution from jddurand https://github.com/noxxi/p5-io-socket-ssl/pull/117 - fixed race condition in t/sni_verify.t based on analysis from jddurand https://github.com/noxxi/p5-io-socket-ssl/issues/97 ==== python-httpcore ==== - Skip failing test_request_with_content test (gh#encode/httpcore#622). ==== python-jsonschema ==== Version update (4.17.0 -> 4.17.1) - update to 4.17.1: * No nice changelog provided upstream Full Changelog: https://github.com/python-jsonschema/jsonschema/compare/v4.17.0...v4.17.1 ==== python-pylibmc ==== Version update (1.6.1 -> 1.6.3) - Update to 1.6.3 * Fix logger.warn() deprecation warnings * Modernize tests and test harness * Python 3.10 support - Remove python-pylibmc-remove-nose.patch - Cope with memcached moving from sbindir to bindir and safeguard against non-existing memcached binary: it if can't be started, there is no reason to even attempt to run the test suite. ==== schily ==== Subpackages: libcdrdeflt1_0 libdeflt1_0 libfile1_0 libfind4_0 librmt1_0 librscg1_0 libscg1_0 libscgcmd1_0 libschily2_0 mkisofs spax star - Update to release 2022.10.16 * bsh: Implement ^L support * libmdigest: Add support for BLAKE2 digests * mkisofs: No longer warn about Joliet being nonstandard. With the 2019 revision of ECMA 119, Joliet has been added as an annex to the standard. ==== vim ==== Version update (9.0.0814 -> 9.0.0924) Subpackages: vim-data vim-data-common vim-small - Updated to version 9.0.0924, fixes the following problems * Aws config files are not recognized. * ":!" does not switch to the alternate screen. * CTRL-Z at end of file is always dropped. * Build error. * "!ls" does not work. * Still a build error, tests are failing. * Memory leak with empty shell command. * Crash when using win_move_statusline() in another tab page. * Crash when dragging the statusline with a mapping. * Mouse drag test fails. * Crash when using win_move_separator() in other tab page. * If 'endofline' is set the CTRL-Z may be written in the wrong place. * The <Home> key in tmux doesn't work when 'term' is set to "xterm". (Dominique Pellé) * Various typos. * Wrong counts in macro comment. * Compiler warning for redefining HAVE_DUP. * Deprecation warning causes build failure. * Warning for missing return type. * The window title is not redrawn when 'endoffile' changes. * Wrong error when using extend() with funcref. * append() reports failure when not appending anything. * Compiler warnings for unused variables. * Test may fail depending on sequence of events. * Cannot change a slice of a const list. (Takumi KAGIYAMA) * deletebufline() does not always return 1 on failure. * Unicode range for Apple SF symbols is outdated. * VHS tape files are not recognized. * Handling 'statusline' errors is spread out. * Shell command with just space gives strange error. * Using assert_fails() may cause hit-enter prompt. * CI: not totally clear what MS-Windows version is used. * Help item for --log argument is not aligned nicely. * Terminal mouse test is a bit flaky. * Terminal mouse test is still flaky. * Crypt test is skipped if xxd is not found. * No proper test for what 9.0.0846 fixes. * Comment not located above the code it refers to. * "!!sort" in a closed fold sorts too many lines. * Compiler warning for unused variable. * Solution for "!!sort" in closed fold is not optimal. * Default value of 'endoffile' is wrong. * col() and charcol() only work for the current window. * Crash when using "!!" without a previous shell command. * Duplicate arguments are not always detected. * No test for what patch 8.2.2207 fixes. * Wildmenu redrawing code is spread out. * Bogus error when string used after :elseif. * Get E967 when using text property in quickfix window. (Sergey Vlasov) * Using freed memory when clearing augroup at more prompt. * Code is indented more than needed. * Using freed memory when executing mapclear at the more prompt. * Using freed memory when executing unmenu at the more prompt. * Using freed memory when executing delfunc at the more prompt. * Code is indented more than needed. * Using freed memory with :comclear while listing commands. * Coverity warns for dead code. * Unnecessary nesting in makefile. * Preprocessor indenting is off. * Cannot get the currently showing mouse shape. * Using freed memory after SpellFileMissing autocmd uses bwipe. * A silent mapping may cause dots on the command line. * Mouse shape remains in op-pending mode after failed change. * Informational message has an error message number. * Horizontal mouse scroll only works in the GUI. * Cannot easily try out what codes various keys produce. * Keycode check script has a few flaws. * No test for what patch 9.0.0827 fixes. * Virtual text below after match has wrong highlight. * May redraw when not needed, causing slow scrolling. * 'smoothscroll' cursor calculations wrong when 'number' is set. * Virtual text property highlight ignores window background. * File renamed twice in test; missing feature check. * Test for home key fails when 'term' is "tmux". * Clinical Quality Language files are not recognized. * With 'smoothscroll' cursor is one screen line too far down. (Ernie Rael) * The builtin terminals are in one long list. * Cursor moves too far with 'smoothscroll'. * Setting w_leftcol and handling side effects is confusing. * Some mouse scroll code is not in a good place. * Key code checker doesn't check modifyOtherKeys resource. * Various comment and indent flaws. * Virtual text after the line wraps when 'wrap' is off. * Mouse scroll code is not optimal. * Restoring window after WinScrolled may fail. * With 'smoothscroll' cursor may end up in wrong position. * Error message for layout change does not match action. * Setting lines in another buffer may not work well. * With 'smoothscroll' set mouse click position may be wrong. * libvterm with modifyOtherKeys level 2 does not match xterm. * Only a change in the current window triggers the WinScrolled event. * deletebufline() may move marks in the wrong window. * WinScrolled may trigger immediately when defined. * getbufline() is inefficient for getting a single line. * The WinScrolled autocommand event is not enough. * MS-Windows: modifier keys do not work with mouse scroll events. * Build failure with tiny features. * Cannot find an import prefixed with "s:". (Doug Kearns) * Missing defined(PROTO) in #ifdef. * Mermaid files are not recognized. * Second SIGWINCH signal may be ignored.