New Tumbleweed snapshot 20210325 released!
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=opensuse&groupid=1&version=Tumbleweed&build=20210325 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: PackageKit checkpolicy (3.1 -> 3.2) cppcheck (2.3 -> 2.4.1) cups e2fsprogs (1.46.1 -> 1.46.2) elfutils elfutils-debuginfod emacs-w3m (1.4.632 -> 1.4.632+396+g051dba16) enchant (2.2.8 -> 2.2.15) evolution-data-server evolution-ews ffmpeg-4 (4.3.1 -> 4.3.2) gnome-shell java-11-openjdk kernel-firmware (20210208 -> 20210315) libnettle (3.7.1 -> 3.7.2) libnice (0.1.17 -> 0.1.18) libselinux (3.1 -> 3.2) libselinux-bindings (3.1 -> 3.2) libsemanage (3.1 -> 3.2) libtpms (0.7.4 -> 0.7.7) libvirt links (2.21 -> 2.22) man-pages (5.10 -> 5.11) mc (4.8.25 -> 4.8.26) openblas_pthreads (0.3.13 -> 0.3.14) p7zip perl-Net-HTTP (6.20 -> 6.21) policycoreutils (3.1 -> 3.2) python-semanage (3.1 -> 3.2) redis rng-tools (6.11 -> 6.12) rubygem-bootsnap (1.7.2 -> 1.7.3) rubygem-rspec-rails (5.0.0 -> 5.0.1) setools (4.3.0 -> 4.4.0) snapper (0.8.15 -> 0.8.16) spirv-tools taglib (1.11.2~git20190725.79bc9ccf -> 1.12) thunar (4.16.5 -> 4.16.6) === Details === ==== PackageKit ==== Subpackages: PackageKit-backend-zypp PackageKit-gstreamer-plugin PackageKit-gtk3-module PackageKit-lang libpackagekit-glib2-18 typelib-1_0-PackageKitGlib-1_0 - Add PackageKit-remove-transaction-size-limit.patch: Remove large transaction size sanity check (gh#hughsie/PackageKit/commit#ac5c8660) ==== checkpolicy ==== Version update (3.1 -> 3.2) - Update to version 3.2 * Fix a memleak and an integer overflow ==== cppcheck ==== Version update (2.3 -> 2.4.1) - update to 2.4.1: * fix for windows installer, no other changes - update to 2.4: * Detect one definition rule violations * MISRA improvements * ImportProject fixes * Various bug hunting improvements * Fixes when importing AST from clang ==== cups ==== Subpackages: cups-client cups-config libcups2 libcups2-32bit libcupsimage2 - fix-negotiate-authentication-between-CGIs-and-scheduler.patch fixes web UI Kerberos authentication (bsc#1175960) - Remove code comments from expanded scriptlets to reduce size ==== e2fsprogs ==== Version update (1.46.1 -> 1.46.2) Subpackages: e2fsprogs-scrub libcom_err2 libcom_err2-32bit libext2fs2 - e2fsprogs 1.46.2: * tune2fs -c now takes "random" argument * Add support for the FS_NOCOMP_FL flag to chattr and lsattr * Fix warnings when resizing small file systems to a super-large * Fix the debugfs rdump and ls commands so they will work correctly for uid's and gid's => 65536 * Fix the debugfs write and symlink commands so they support targets which contain a pathname * Fix Direct I/O support on block devices where the logical block size is greater 1k * Fix debugfs's logdump so it works on file systems whose block size is greater than 8k * Fix a crash when there is error while e2fsck is trying to open the file system, and e2fsck calls ext2fs_mmp_stop() before MMP has been initialized * Improved error checking in the fast commit replay code in e2fsck * Fix various compiler and Coverity warnings * Update the Spanish translation from the translation project ==== elfutils ==== Subpackages: elfutils-lang libasm1 libdw1 libelf1 - Add disable-run-readelf-self-test.patch in order to disable a failing test-case with GCC 11 (PR27367). ==== elfutils-debuginfod ==== - Enable https://debuginfod.opensuse.org/ debuginfod server by default now. - Add disable-run-readelf-self-test.patch in order to disable a failing test-case with GCC 11 (PR27367). ==== emacs-w3m ==== Version update (1.4.632 -> 1.4.632+396+g051dba16) - Update to version 1.4.632+396+g051dba16: * Don't show a backstage while performing `text-scale-adjust' * Make text-scale-adjust work * Fix it again * shimbun.el (shimbun-xml-parse-buffer): Fix typo * ChangeLog: fix date entries * * shimbun.el (shimbun-xml-parse-buffer): Work around malformed xml * Update copyright years * Docfix so not to be wider than 80 chars * Bugfix 'w3m-previous-image doesn't work' ([emacs-w3m:13747]) * Fix typo in ChangeLog ==== enchant ==== Version update (2.2.8 -> 2.2.15) Subpackages: enchant-2-backend-hunspell enchant-data libenchant-2-2 - Update to version 2.2.15: + Specify that nuspell >= 4.1.0 is required. + Fix some space leaks in the tests. + The nuspell backend is updated for newer versions. + Make the enchant program output its version to standard output, not standard error. This may help some programs that use this output. + Fix a bug in the Voikko and Zemberek back-ends that could cause spell checking and suggestion to fail. + Make enchant silently ignore -B flag, for better Emacs compatibility. + Make enchant_broker_list_dicts sort the tags, so that enchant-lsmod?s output is sorted. + Minor build system improvement: don?t use -D_FORTIFY_SOURCE, which can cause problems on Windows, and should be configured by the compiler vendor if desired. + Fix Hunspell backend to treat apostrophes as Hunspell does: if either straight or curly apostrophe is a word character, allow both. + Fix a couple of space leaks in the Nuspell back end. - Drop Fix_back-ends_that_want_a_NUL-terminated_string.patch: fixed upstream. ==== evolution-data-server ==== Subpackages: evolution-data-server-lang libcamel-1_2-62 libebackend-1_2-10 libebook-1_2-20 libebook-contacts-1_2-3 libecal-2_0-1 libedata-book-1_2-26 libedata-cal-2_0-1 libedataserver-1_2-25 libedataserverui-1_2-2 - Add evolution-data-server-boo1182882.patch: fix buffer overrun when parsing base64 data (boo#1182882). ==== evolution-ews ==== Subpackages: evolution-ews-lang - Add evolution-ews-boo1182882.patch: fix buffer overrun when parsing base64 data (boo#1182882). ==== ffmpeg-4 ==== Version update (4.3.1 -> 4.3.2) Subpackages: libavcodec58_91 libavdevice58_10 libavfilter7_85 libavformat58_45 libavresample4_0 libavutil56_51 libpostproc55_7 libswresample3_7 libswscale5_7 - update to 4.3.2: * lots of oss-fuzz reported overflow fixes, see included ChangeLog - drop ffmpeg.git-ba3e771a42c29ee02c34e7769cfc1b2dbc5c760a.patch 0001-lavf-srt-fix-build-fail-when-used-the-libsrt-1.4.1.patch 0001-avformat-vividas-improve-extradata-packing-checks-in.patch: upstream ==== gnome-shell ==== Subpackages: gnome-shell-calendar gnome-shell-lang - Disable gnome-shell-jsc#SLE-16051-Input-method-recommendation.patch: Clears existing keyboard on gnome-shell restart (boo#1183823). - Update gnome-shell-jsc#SLE-16051-Input-method-recommendation.patch: Modify the Japanese input engine load order which will more fit for our community(bnc#1183475); ==== java-11-openjdk ==== Subpackages: java-11-openjdk-headless - Added patches: * system-crypto-policy.patch + Let OpenJDK use system crypto policies unless explicitely told not to * nss-security-provider.patch + Add the NSS security provider with configuration in generated nss.cfg file * keytool-default-rsa.patch + Make keytool generate RSA keys by default, since only the LEGACY system crypto policy allows DSA ==== kernel-firmware ==== Version update (20210208 -> 20210315) Subpackages: kernel-firmware-all kernel-firmware-amdgpu kernel-firmware-ath10k kernel-firmware-ath11k kernel-firmware-atheros kernel-firmware-bluetooth kernel-firmware-bnx2 kernel-firmware-brcm kernel-firmware-chelsio kernel-firmware-dpaa2 kernel-firmware-i915 kernel-firmware-intel kernel-firmware-iwlwifi kernel-firmware-liquidio kernel-firmware-marvell kernel-firmware-media kernel-firmware-mediatek kernel-firmware-mellanox kernel-firmware-mwifiex kernel-firmware-network kernel-firmware-nfp kernel-firmware-nvidia kernel-firmware-platform kernel-firmware-prestera kernel-firmware-qlogic kernel-firmware-radeon kernel-firmware-realtek kernel-firmware-serial kernel-firmware-sound kernel-firmware-ti kernel-firmware-ueagle kernel-firmware-usb-network ucode-amd - Update to version 20210315 (git commit 3568f962908c): * linux-firmware: Update firmware file for Intel Bluetooth AX210 * linux-firmware: Update firmware file for Intel Bluetooth AX200 * linux-firmware: Update firmware file for Intel Bluetooth AX201 * rtw88: 8822c: Update normal firmware to v9.9.6 * iwlwifi: add new FWs from core59-66 release * iwlwifi: update 9000-family firmwares * iwlwifi: update 7265D firmware * Mellanox: Add new mlxsw_spectrum firmware xx.2008.2406 * linux-firmware: add frimware for mediatek bluetooth chip (MT7921) * rtw89: 8852a: add firmware v0.9.12.2 * WHENCE: add missing symlink for BananaPi M3 * Add symlink for BananaPi M2 to brcmfmac43430-sdio config * brcm: Fix Raspberry Pi 4B NVRAM file * silabs: add new firmware for WF200 * amdgpu: add initial firmware for green sardine * rtw88: RTL8822C: Update normal firmware to v9.9.5 - Drop obsoleted patch: Revert-brcm-rpi4-boardflags3-bit.patch - Update topics and aliases ==== libnettle ==== Version update (3.7.1 -> 3.7.2) Subpackages: libhogweed6 libhogweed6-32bit libnettle8 libnettle8-32bit - GNU Nettle 3.7.2: * fix a bug in ECDSA signature verification that could lead to a denial of service attack (via an assertion failure) or possibly incorrect results (boo#1183835) * fix a few related problems where scalars are required to be canonically reduced modulo the ECC group order, but in fact may be slightly larger ==== libnice ==== Version update (0.1.17 -> 0.1.18) Subpackages: gstreamer-libnice libnice10 - Update descriptions to be closer to the other GNOME packages. - Update to 0.1.18: * Remove the autotools build system, now only meson is available * Accept receiving messages in multiple steps over TCP * Accept duplicated ports as last option instead of spinning forever * Use sendmmsg if possible to send multiple packets in one call * Fail gathering if no port is available * Hide the implementation of NiceCandidate, this hides some parts that were previously visible * Enable TURN server connects where both TCP and UDP use the same port number * Don't count rejected STUN messages as keepalive packets * On Windows, the improvements and fixes - use crypto library instead of CryptGenRandom() which is deprecated - use GetBestInterfaceEx() for UWP compatibility - fix the listing of interfaces to use the correct APIs - implement ignoring interfaces * Add buildrquires: gobject-introspection-devel and meson * The move from autoFOO to meson ==== libselinux ==== Version update (3.1 -> 3.2) Subpackages: libselinux1 libselinux1-32bit selinux-tools - Switch to pcre2: + Replace pcre-devel BuildRequires with pkgconfig(libpcre2-8) + Pass USE_PCRE2=y to make. + Replace pkgconfig(libpcre) Requires in -devel static with pkgconfig(libpcre2-8). - Update to version 3.2: * Use mmap()'ed kernel status page instead of netlink by default. See "KERNEL STATUS PAGE" section in avc_init(3) for more details. * New log callback levels for enforcing and policy load notices - SELINUX_POLICYLOAD, SELINUX_SETENFORCE * Changed userspace AVC setenforce and policy load messages to audit format. ==== libselinux-bindings ==== Version update (3.1 -> 3.2) - Switch to pcre2: + Replace pcre-devel BuildRequires with pkgconfig(libpcre2-8) + Pass USE_PCRE2=y to make. - Update to version 3.2: * Use mmap()'ed kernel status page instead of netlink by default. See "KERNEL STATUS PAGE" section in avc_init(3) for more details. * New log callback levels for enforcing and policy load notices - SELINUX_POLICYLOAD, SELINUX_SETENFORCE * Changed userspace AVC setenforce and policy load messages to audit format. ==== libsemanage ==== Version update (3.1 -> 3.2) - Link to correct so version - Minor spec file cleanups - Move configuration file to separate libsemanage-conf package to allow for parallel installation in future versions - Update to version 3.2 * dropped old and deprecated symbols and functions libsemanage version was bumped to libsemanage.so.2 * libsemanage tries to sync data to prevent empty files in SELinux module store ==== libtpms ==== Version update (0.7.4 -> 0.7.7) - Update to version 0.7.7 * CryptSym: fix AES output IV (bsc#1183729, CVE-2021-3446) * tpm2: Fix public key context save due to ANY_OBJECT_Marshal usage * tpm2: Address some Coverity issues (false positives) * tpm1.2: Backported ASAN/UBSAN related fixes * tpm2: Return properly sized array for b parameter for NIST P521 (HLK) * tpm2: Addressed issues detected by UBSAN * tpm2: Addressed issues detected by cppcheck (false positives) ==== libvirt ==== Subpackages: libvirt-bash-completion libvirt-client libvirt-daemon libvirt-daemon-driver-interface libvirt-daemon-driver-libxl libvirt-daemon-driver-lxc libvirt-daemon-driver-network libvirt-daemon-driver-nodedev libvirt-daemon-driver-nwfilter libvirt-daemon-driver-qemu libvirt-daemon-driver-secret libvirt-daemon-driver-storage libvirt-daemon-driver-storage-core libvirt-daemon-driver-storage-disk libvirt-daemon-driver-storage-iscsi libvirt-daemon-driver-storage-iscsi-direct libvirt-daemon-driver-storage-logical libvirt-daemon-driver-storage-mpath libvirt-daemon-driver-storage-rbd libvirt-daemon-driver-storage-scsi libvirt-daemon-lxc libvirt-daemon-qemu libvirt-daemon-xen libvirt-libs - spec: Fix exec-restart of virtlockd and virtlogd on package upgrade bsc#1183411 - spec: Move netcat-openbsd requirement from the libs to the daemon subpackage. It is only needed by the daemon and introduces an unneeded dependency for users of libvirt-libs. ==== links ==== Version update (2.21 -> 2.22) - update to 2.22: * Save and restore the terminal using xterm escape codes * Save and restore the console using "cons.saver" from Midnight Commander * Support UTF-8 frames * Fixed a bug in displaying non-printable characters ==== man-pages ==== Version update (5.10 -> 5.11) - version update to 5.11 http://linux-man-pages.blogspot.com/2021/03/man-pages-511-is-released.html - modified patches % man-pages-tty_ioctl.patch (refreshed) ==== mc ==== Version update (4.8.25 -> 4.8.26) Subpackages: mc-lang - Midnight Commander 4.8.26: * Support file names of any length * Implement persistent command line buffer for subshell (bash >= 4, zsh and fish are supported) * Implement shadows of dialog windows and menus * Allow running clipboard commands if DISPLAY is not set * Add support of "alacritty", "tmux", and "tmux-256color" terminals * VFS: Support wim archive format (using wimtools) * VFS: Support pak archive format (using unar) * Editor: Add Swift syntax highlighting * Various bug fixes ==== openblas_pthreads ==== Version update (0.3.13 -> 0.3.14) - Update openblas-ppc64be_up2_p8.patch trimed by previous sr (still need changes in Makefile.system) - Update to version 0.3.14 common: * Fixed a race condition on thread shutdown in non-OpenMP builds * Fixed custom BUFFERSIZE option getting ignored in gmake builds * Fixed CMAKE compilation of the TRMM kernels for GENERIC platforms * Added CBLAS interfaces for CROTG, ZROTG, CSROT and ZDROT * Improved performance of OMATCOPY_RT across all platforms * Changed perl scripts to use env instead of a hardcoded /usr/bin/perl * Fixed potential misreading of the GCC compiler version in the build scripts * Fixed convergence problems in LAPACK complex GGEV/GGES (Reference-LAPACK #477) * Reduced the stacksize requirements for running the LAPACK testsuite (Reference-LAPACK #335) RISC V: * Fixed compilation on RISCV (missing entry in getarch) POWER: * Fixed compilation for DYNAMIC_ARCH with clang and with older gcc versions * Added support for compilation on FreeBSD/ppc64le * Added optimized POWER10 kernels for SSCAL, DSCAL, CSCAL, ZSCAL * Added optimized POWER10 kernels for SROT, DROT, CDOT, SASUM, DASUM * Improved SSWAP, DSWAP, CSWAP, ZSWAP performance on POWER10 * Improved SCOPY and CCOPY performance on POWER10 * Improved SGEMM and DGEMM performance on POWER10 * Added support for compilation with the NVIDIA HPC compiler x86_64: * Added an optimized bfloat16 GEMM kernel for Cooperlake * Added CPUID autodetection for Intel Rocket Lake and Tiger Lake cpus * Improved the performance of SASUM,DASUM,SROT,DROT on AMD Ryzen cpus * Added support for compilation with the NAG Fortran compiler * Fixed recognition of the AMD AOCC compiler * Fixed compilation for DYNAMIC_ARCH with clang on Windows * Added support for running the BLAS/CBLAS tests on Windows * Fixed signatures of the tls callback functions for Windows x64 * Fixed various issues with fma intrinsics support handling ARM: * Support compilation for embedded Cortex M4 targets via a new option EMBEDDED ARM64: * Fixed the THUNDERX2T99 and NEOVERSEN1 DNRM2/ZNRM2 kernels for inputs with Inf * Added support for the DYNAMIC_LIST option * Added support for compilation with the NVIDIA HPC compiler * Added support for compiling with the NAG Fortran compiler - Remove 0001-Require-gcc-11-for-builtin_cpu_is-power10.patch 0002-patch-to-support-power10-in-builtin_cpu_is-was-backp.patch Upstream fixed in a different way. ==== p7zip ==== Subpackages: p7zip-full - Add patch from Debian to fix build with GCC 11 (boo#1181880) * 0001-Fix-g-warning.patch ==== perl-Net-HTTP ==== Version update (6.20 -> 6.21) - updated to 6.21 see /usr/share/doc/packages/perl-Net-HTTP/Changes 6.21 2021-03-18 21:56:42Z - Accept PeerAddr of 0. (GH#72) (trwyant) ==== policycoreutils ==== Version update (3.1 -> 3.2) Subpackages: policycoreutils-lang python3-policycoreutils - Update to version 3.2 * Tools using sepolgen, e.g. audit2allow, print extended permissions in hexadecimal * sepolgen sorts extended rules like normal ones * `setfiles` doesn't abort on labeling errors - Refreshed get_os_version.patch ==== python-semanage ==== Version update (3.1 -> 3.2) - Minor spec file cleanups - Update to version 3.2 * dropped old and deprecated symbols and functions libsemanage version was bumped to libsemanage.so.2 * libsemanage tries to sync data to prevent empty files in SELinux module store ==== redis ==== - replaced /var/run with /run for all PID file paths ==== rng-tools ==== Version update (6.11 -> 6.12) - update to 6.12: * Fix compiler warning over log message format * Fix some typos in force-reseed documentation in rngd man page * Improve --list option so that we properly capture entropy sources that are available and configured on at build time, but failed initalization at run time (due to lack of hw, or some other error, etc) * Drop the use of libsysfs - we only used it to access a single file, and we can do so with a simple open/read/close. Given the lack of maintenance of libsysfs, we can save lots of effort by dropping this lib ==== rubygem-bootsnap ==== Version update (1.7.2 -> 1.7.3) Subpackages: ruby2.7-rubygem-bootsnap ruby3.0-rubygem-bootsnap - updated to version 1.7.3 * Disable YAML precompilation when encountering YAML tags. (#351) ==== rubygem-rspec-rails ==== Version update (5.0.0 -> 5.0.1) - updated to version 5.0.1 [Full Changelog](https://github.com/rspec/rspec-rails/compare/v5.0.0...v5.0.1) Bug Fixes: * Limit multibyte example descriptions when used in system tests for #method_name which ends up as screenshot names etc. (@y-yagi, #2405, #2487) ==== setools ==== Version update (4.3.0 -> 4.4.0) - Update to the version 4.4.0: * Added support for old Boolean name substitution in seinfo and sesearch. * Added sechecker tool which is a configuration file driven analysis tool. ==== snapper ==== Version update (0.8.15 -> 0.8.16) Subpackages: libsnapper5 snapper-zypp-plugin - fixed creating root config (root prefix handling) (gh#openSUSE/snapper#627) ==== spirv-tools ==== - Add -Wno-error ==== taglib ==== Version update (1.11.2~git20190725.79bc9ccf -> 1.12) Subpackages: libtag1 libtag_c0 - Add missing zlib dependency in devel package - reference download url of tarball - Update to version 1.12: * Added support for WinRT. * Added support for Linux on POWER. * Added support for classical music tags of iTunes 12.5. * Added support for file descriptor to FileStream. * Added support for 'cmID', 'purl', 'egid' MP4 atoms. * Added support for 'GRP1' ID3v2 frame. * Added support for extensible WAV subformat. * Enabled FileRef to detect file types based on the stream content. * Dropped support for Windows 9x and NT 4.0 or older. * Check for mandatory header objects in ASF files. * More tolerant handling of RIFF padding, WAV files, broken MPEG streams. * Improved calculation of Ogg, Opus, Speex, WAV, MP4 bitrates. * Improved Windows compatibility by storing FLAC picture after comments. * Fixed numerical genres in ID3v2.3.0 'TCON' frames. * Fixed consistency of API removing MP4 items when empty values are set. * Fixed consistency of API preferring COMM frames with no description. * Fixed OOB read on invalid Ogg FLAC files (CVE-2018-11439). * Fixed handling of empty MPEG files. * Fixed parsing MP4 mdhd timescale. * Fixed reading MP4 atoms with zero length. * Fixed reading FLAC files with zero-sized seektables. * Fixed handling of lowercase field names in Vorbis Comments. * Fixed handling of 'rate' atoms in MP4 files. * Fixed handling of invalid UTF-8 sequences. * Fixed possible file corruptions when saving Ogg files. * Fixed handling of non-audio blocks, sampling rates, DSD audio in WavPack files. * TableOfContentsFrame::toString() improved. * UserTextIdentificationFrame::toString() improved. * Marked FileRef::create() deprecated. * Marked MPEG::File::save() with boolean parameters deprecated, provide overloads with enum parameters. * Several smaller bug fixes and performance improvements. - Remove obsolete patches: * taglib-versionbump.patch * 0001-Changed-libdir-includedir-variables-to-change-based-.patch ==== thunar ==== Version update (4.16.5 -> 4.16.6) Subpackages: libthunarx-3-0 thunar-lang - Update to version 4.16.6 * Reload current directory before selecting new files (gxo#xfce/thunar#524) * tree-view: Hide menu-item "properties" for unmounted devices * Removed 'the root folder has no parent' dialog box * Revamp documentation to modernize/uniformize accross components * Remove watches on shortcuts (gxo#xfce/thunar#513, gxo#xfce/thunar#47) * Translation Updates
20210325 was the first large snapshot to trigger my new logic for a longer publish delay (6h instead of 2h). The publisher became active when the repo age was 22203 s and the repo rsync log had 32114 lines and 2091370 byte. This time, download.o.o seems to have had a much better time handling the load than with previous full rebuilds. Though this seems to have been only a partial rebuild because there are 62802 entries in factory/repo Medium-term we can do more fine-tuning to balance publish delay vs load of download.o.o Ciao Bernhard M.
On 2021-03-28 15:09, Bernhard M. Wiedemann wrote:
20210325 was the first large snapshot to trigger my new logic for a longer publish delay (6h instead of 2h).
The publisher became active when the repo age was 22203 s and the repo rsync log had 32114 lines and 2091370 byte.
This time, download.o.o seems to have had a much better time handling the load than with previous full rebuilds. Though this seems to have been only a partial rebuild because there are 62802 entries in factory/repo
Medium-term we can do more fine-tuning to balance publish delay vs load of download.o.o
Bernhard, A personal thanks from me. I'm looking after Factory while Dominique is on vacation and having a major rebuild as an unexpected side effect of the libselinux.so version bump was not how I wanted to start my two weeks holding the reigns. I'm very grateful for what you've done here, but I'll be doing my best to avoid similar big rebuilds over the next few weeks ;) Cheers, -- Richard Brown Linux Distribution Engineer - Future Technology Team SUSE Software Solutions Germany GmbH Maxfeldstr. 5 90409 Nuremberg Germany (HRB 36809, AG Nürnberg) Managing Director: Felix Imendörffer
On Sun 2021-03-28, Bernhard M. Wiedemann wrote:
20210325 was the first large snapshot to trigger my new logic for a longer publish delay (6h instead of 2h).
The publisher became active when the repo age was 22203 s and the repo rsync log had 32114 lines and 2091370 byte.
This is really cool. Thank you, Bernhard! (There was a bit of discussion on project@ earlier this month about mirrors and larger snapshots; let me share this there, too.) Gerald
Hello. Starting with this snapshot, debuginfod is enabled by default (via /etc/profile.d/debuginfod.sh ENV variable). Right now, our server instance indexes openSUSE:Factory packages for i586 and x86_64 targets. One small example: $ gdb `which make` -ex 'set height unlimited' -ex 'set confirm off' -ex start -ex bt -ex 'p argv[0]' -ex list -ex q ... Reading symbols from /usr/bin/make... Downloading separate debug info for /usr/bin/make... Reading symbols from /home/marxin/.cache/debuginfod_client/7daf1620b47e92a1f4452405a617418b214f0f0b/debuginfo... Downloading source file /usr/src/debug/make-4.3-2.22.x86_64/src/main.c... Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6... Temporary breakpoint 1, main (argc=1, argv=0x7fffffffde78, envp=0x7fffffffde88) at src/main.c:1056 1056 { #0 main (argc=1, argv=0x7fffffffde78, envp=0x7fffffffde88) at src/main.c:1056 $1 = 0x7fffffffe2a0 "/usr/bin/make" 1051 main (int argc, char **argv) 1052 #else 1053 int 1054 main (int argc, char **argv, char **envp) 1055 #endif 1056 { 1057 static char *stdin_nm = 0; 1058 int makefile_status = MAKE_SUCCESS; 1059 struct goaldep *read_files; 1060 PATH_VAR (current_directory); Martin
On Tuesday 2021-03-30 10:37, Martin Liška wrote:
Starting with this snapshot, debuginfod is enabled by default (via /etc/profile.d/debuginfod.sh ENV variable).
Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6...
I noticed. Does it store them anywhere, or will it naively re-download them everytime? Is it installing a proper RPM, or just drawing untracked files?
On 3/30/21 10:47 AM, Jan Engelhardt wrote:
On Tuesday 2021-03-30 10:37, Martin Liška wrote:
Starting with this snapshot, debuginfod is enabled by default (via /etc/profile.d/debuginfod.sh ENV variable).
Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6...
I noticed. Does it store them anywhere, or will it naively re-download them everytime?
Yes, the files are cached here ~/.cache/debuginfod_client/ and wiped after `cat ~/.cache/debuginfod_client/max_unused_age_s` == 604800 seconds.
Is it installing a proper RPM, or just drawing untracked files?
No, it's not installing any RPM files. It downloads debuginfo ELF files and source files which are needed by a client. Cheers, Martin
On Tuesday 2021-03-30 10:37, Martin Liška wrote:
Right now, our server instance indexes openSUSE:Factory packages for i586 and x86_64 targets.
One small example:
$ gdb `which make` -ex 'set height unlimited' -ex 'set confirm off' -ex start -ex bt -ex 'p argv[0]' -ex list -ex q ... Reading symbols from /usr/bin/make... Downloading separate debug info for /usr/bin/make... Reading symbols from /home/marxin/.cache/debuginfod_client/7daf1620b47e92a1f4452405a617418b214f0f0b/debuginfo... Downloading source file /usr/src/debug/make-4.3-2.22.x86_64/src/main.c... Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6...
What's really annoying is that signals like ^C, ^Z and ^\ are all ignored while it downloads something. What is the way to turn off this downloading altogether?
On 4/2/21 1:02 PM, Jan Engelhardt wrote:
On Tuesday 2021-03-30 10:37, Martin Liška wrote:
Right now, our server instance indexes openSUSE:Factory packages for i586 and x86_64 targets.
One small example:
$ gdb `which make` -ex 'set height unlimited' -ex 'set confirm off' -ex start -ex bt -ex 'p argv[0]' -ex list -ex q ... Reading symbols from /usr/bin/make... Downloading separate debug info for /usr/bin/make... Reading symbols from /home/marxin/.cache/debuginfod_client/7daf1620b47e92a1f4452405a617418b214f0f0b/debuginfo... Downloading source file /usr/src/debug/make-4.3-2.22.x86_64/src/main.c... Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6...
What's really annoying is that signals like ^C, ^Z and ^\ are all ignored while it downloads something.
Sorry about that, yes, it's a known problem https://sourceware.org/bugzilla/show_bug.cgi?id=27026 and there's a tentative patch for it.
What is the way to turn off this downloading altogether?
You can comment the corresponding line in /etc/profile.d/debuginfod.sh. Cheers, Martin
On Freitag, 2. April 2021 13:08:23 CET Martin Liška wrote:
On 4/2/21 1:02 PM, Jan Engelhardt wrote:
On Tuesday 2021-03-30 10:37, Martin Liška wrote:
Right now, our server instance indexes openSUSE:Factory packages for i586 and x86_64 targets.
One small example:
$ gdb `which make` -ex 'set height unlimited' -ex 'set confirm off' -ex start -ex bt -ex 'p argv[0]' -ex list -ex q ... Reading symbols from /usr/bin/make... Downloading separate debug info for /usr/bin/make... Reading symbols from /home/marxin/.cache/debuginfod_client/7daf1620b47e92a1f4452405a617418b214 f0f0b/debuginfo... Downloading source file /usr/src/debug/make-4.3-2.22.x86_64/src/main.c... Temporary breakpoint 1 at 0xa840: file src/main.c, line 1056. Starting program: /usr/bin/make Downloading separate debug info for /lib64/ld-linux-x86-64.so.2... Downloading separate debug info for /lib64/libdl.so.2... Downloading separate debug info for /lib64/libc.so.6...
What's really annoying is that signals like ^C, ^Z and ^\ are all ignored while it downloads something.
Sorry about that, yes, it's a known problem https://sourceware.org/bugzilla/show_bug.cgi?id=27026
and there's a tentative patch for it.
Apparently, this is still not fixed ...
What is the way to turn off this downloading altogether?
You can comment the corresponding line in /etc/profile.d/debuginfod.sh.
Can this please be moved to a separate package, or the default reverted? Has anyone not living on the same network as the debuginfod server actually tested this? I just had to wait 40 minutes for gdb to start up when debugging a graphical application. It downloaded 2.5 GBytes (libLLVM13 alone is 700MB, QtWebengineCore another 700), and that does not include the application itself. Most of this debuginfo is totally irrelevant, as it also includes tons of stuff not remotely relevant for debugging a specific issue, as it is not part of the current backtrace. Regards, Stefan -- Stefan Brüns / Bergstraße 21 / 52062 Aachen home: +49 241 53809034 mobile: +49 151 50412019
participants (7)
-
Bernhard M. Wiedemann
-
Dominique Leuenberger
-
Gerald Pfeifer
-
Jan Engelhardt
-
Martin Liška
-
Richard Brown
-
Stefan Brüns