November 2017 - openSUSE Factory - openSUSE Mailing Lists

[opensuse-factory] OBS: nothing provides /usr/bin/cdrdao needed by k3b
by Linux Kamarada 02 Dec '17

02 Dec '17

Hi everyone! First of all, congratulations to the OBS team, the SUSE Studio team, the KIWI team and any other teams involved in doing Studio Express. Keep up the good work! The only thing I'm really missing is testdrive, but considering that now we can use all the KIWI power on OBS, downloading the built image and testing locally on VirtualBox is just a little price worth paying! As a matter of fact, the heavy work is done on the cloud by SUSE servers free of charge... I'm trying to build a LiveDVD as close as possible from a clean openSUSE install on disk. That should be enough for it: config.kiwi (https://build.opensuse.org/package/view_file/home:livesuse:dev/LiveSUSE-Lea…) <packages type="image" patternType="plusRecommended">   <package name="patterns-openSUSE-base" /> <package name="patterns-openSUSE-x11" /> <package name="patterns-openSUSE-kde" /> <package name="sddm" /> <package name="branding-openSUSE"/> (observe the plusRecommended setting) (actually, some more packages are needed, but that is the idea, anyone interested can follow the link to the complete file) OBS succeeds. The image is built. I download it. Test it on VirtualBox. Then I realize it does not have K3b. I install K3b using zypper. Works like a charm. On purpose, I add the k3b package to the package list in config.kiwi: <package name="k3b" /> Then OBS starts building the image again, but fails. The build log shows: [ 143s] Resolving package dependencies... [ 143s] [ 143s] Problem: nothing provides /usr/bin/cdrdao needed by k3b-17.04.2-4.1.x86_64 [ 143s] Solution 1: do not install k3b-17.04.2-4.1.x86_64 [ 143s] Solution 2: break k3b-17.04.2-4.1.x86_64 by ignoring some of its dependencies [ 143s] [ 143s] Choose from above solutions by number or cancel [1/2/c] (c): c https://build.opensuse.org/package/live_build_log/home:livesuse:dev/LiveSUS… I do not understand, because the OSS repo has both k3b and cdrdao. Any ideas on what is going wrong? Thank you very much! Antonio The Linux Kamarada Project http://kamarada.github.io/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

12 30

[opensuse-factory] scons is going to use python3 as interpreter
by Dominique Leuenberger / DimStar 01 Dec '17

01 Dec '17

Dear Tumbleweed hackers and packagers, As you all know, python3 is 'the new thing' and pytho2 is being moved to the background. Various packages have already been switched to be using python3 - probably 'the easy ones' are all done. so we enter the jounery of more complex things. One such package is scons, which has been switched to be using python3 in request https://build.opensuse.org/request/show/542087 Now, scons is a bit 'special', as it is written in python and is basiclaly compatible to python2 AND python3, BUT it parses SConstruct/SConscript of different packages. And there the results are a bit mixed when it comes to python3 compliance Martin setup a test branch where he monitors all packages in openSUSE:Factory using scons as their build script, see https://build.opensuse.org/project/monitor/home:pluskalm:python3 Out of those only two packages are tracked as part of the rings (ffado and gpsd), which means from a Factory process perspective, once those two are fixed, the scons-switch to python3 could happen. As this will impact some other packages, I'd like to ask for your help to fix the other failing packages there. Simply branch them using "osc branch openSUSE:Factory $pkg" and submit them back to their original devel project. Martin's project only serves as monitor and should not receive submissions. In many cases, the issue in SConstruct are the usual python3 offenders: print "FOO" needs to be written as print ("FOO") now; so patches can turn out to be easy (in some cases they are more complex though) This mail serves as an info so you won't be surprised when scons is being merged and some packagaes stop building (of course some people will still be surprised, but such is life) Cheers Dominique

3 3

[opensuse-factory] apparmor, kernel 4.14 and libvirtd
by Michael Ströder 01 Dec '17

01 Dec '17

HI! It seems the kernel upgrade needs another modification to apparmor profile(s) for libvirtd: type=VIRT_RESOURCE msg=audit(1511353655.324:343): pid=1528 uid=0 auid=4294967295 ses=4294967295 msg='virt=kvm resrc=cgroup reason=deny vm="ae-dir-deb-p1" uuid=35bee50f-d977-48d4-88d1-9af4bfd1b6c7 cgroup="/sys/fs/cgroup/devices/machine.slice/machine-qemu\x2d2\x2dae\x2ddir\x2ddeb\x2dp1.scope/" class=all exe="/usr/sbin/libvirtd" hostname=? addr=? terminal=? res=success' Ciao, Michael.

3 8

[opensuse-factory] Tumbleweed will change suse_version to 1550
by Dominique Leuenberger / DimStar 01 Dec '17

01 Dec '17

Dear Tumbleweed hackers and packagers, It is about time that openSUSE:Factory aka Tumbleweed moves its suse_version definition up again. There were quite some build hacks in place that were explicitly pinned on 1330, as the idea was to be able to get rid of them by the time we move the version. As this is dragging on for too long though, we changed the hacks to also apply on Tumbleweed with a newer suse_version and all those tweaks are now in place. Which means we can move forward. The new suse_version will be 1550 - starting with December 1st 2017 What's the rationale behind 1550: * SLE15 / Leap 15 are using 1500 (for their entire lifecycle). * Tumbleweed is supposed to be 'ahead of the game', so we plant it between SLe/Leap15 and SLE/Leap16 (don't ask about timeframes for those, nobody knows) While packaging, you should never reference version 1550 though. your thing are compatible to suse_version >= 1500 (meaning all Leap 15.x versions) OR, if something comes into play on newer .x levels, you'd do: %if %{suse_version} > 1500 || %{sle_version} >= 150100 # Feature is available since Leap 15.1 BuildRequires: new-dep %endif This will indicate that starting with 15.1 things are present, and for TW as well. If you don't know yet about the Leap version it might become available, simply skip the sle_version check (but keep the check for suse_version > 1500) The change of suse_version has no visible impact to the end user, this is merely of interest to developers/packagers. Cheers, Dominique

1 1

[opensuse-factory] libevtlog
by Peter Czanik 01 Dec '17

01 Dec '17

Hi, The libevtlog package is a dependency of syslog-ng. As far as I know no other package is using it. With syslog-ng version 3.11.1 libeventlog was merged into syslog-ng as having it as a separate package cause more confusion than good. I plan to update syslog-ng to 3.12.1 to resolve compile issues on Tumbleweed. How can I check if any other package uses libevtlog? If nothing, it is safe to drop it after syslog-ng 3.12.1 arrived to Tumbleweed. Bye, CzP -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

2 12

[opensuse-factory] New Tumbleweed snapshot 20171129 released!
by Dominique Leuenberger 01 Dec '17

01 Dec '17

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&versio… When you reply to report some issues, make sure to change the subject. It is not helpful to keep the release announcement subject in a thread while discussing a specific problem. Packages changed: amarok at atftp clamav claws-mail create-resources cronie goffice (0.10.35 -> 0.10.36) gthumb (3.5.3 -> 3.6.0) hddtemp icecream icinga kernel-source (4.14.1 -> 4.14.2) ldns (1.6.17 -> 1.7.0) libchewing (0.4.0+git20150820.285b47d -> 0.5.1+git20171114.3df07c9) lomoco mailman mcelog openssl (1.0.2m -> 1.1.0g) osc (0.161.1 -> 0.162.0) postgresql (9.6 -> 10) postgresql10 (9.6.4 -> 10.1) ppp radvd rpcbind rsync scons (3.0.0 -> 3.0.1) sensors smartmontools susepaste syslogd tftp tlp wondershaper words === Details === ==== amarok ==== Subpackages: amarok-lang - Fix build with libmariadb by disabling using mysql_config/mariadb_config (DMYSQLCONFIG_EXECUTABLE:BOOL=OFF) and let find_library() find respective libraries in predefined paths [bsc#1067898] - Add amarok-2.8.0-find_mysql.patch to fix obvious typo in paths definition in find_library() - BuildIgnore openssl 1.0 headers to not conflict with 1.1 headers required by some deps ==== at ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== atftp ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== clamav ==== Subpackages: libclamav7 - clamav-0.99.2-openssl-1.1.patch: fixed build against openssl 1.1 ==== claws-mail ==== Subpackages: claws-mail-lang - Disable fancy (HTML messages) plugin and drop build requires on libwebkit, this can be re-enabled when code is updated to later releases of libwebkit, see claws#3763. ==== create-resources ==== - Add create-scons-on-py3.patch: Fix build with scons using python3 as interpreter. ==== cronie ==== Subpackages: cron - Ensure that /etc/cron.{hourly,daily,weekly,monthly} have proper permissions and owner. This is racy but prevents some LPE vectors - Requires smtp_daemon (not just Recommends) as it's needed by run-crons script [bsc#1064834] - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== goffice ==== Version update (0.10.35 -> 0.10.36) Subpackages: goffice-lang libgoffice-0_10-10 - Update to version 0.10.36: + Bug Fixes: - Engineering format (bgo#785669). - Crash in 3D plots when X or Y values are missing (bgo#788437). - Documentation build with recent gtk-doc (bgo#788710). - Don't use invalid styles in legend from contour or surface plots imported from Excel (bgo#788861). - Add gio-2.0, gmodule-2.0 and gobject-2.0 pkgconfig(*) BuildRequires: - Add pkgconfig(libxml-2.0) BuildRequires: required dependency checked by configure. Note that it was already being used but not declared. - Update typelib-1_0-GOffice-0_10 summary for something shorter and correct GOffice spelling in devel package summary. ==== gthumb ==== Version update (3.5.3 -> 3.6.0) Subpackages: gthumb-lang - Update to version 3.6.0: + Bugs fixed: Numeric keypad nav keys don't work (bgo#667390). - Readd bison, flex and itstool BuildRequires since configures actually do check for them. - Update to version 3.5.4: + Bugs fixed: - No application icon on Wayland. - Tags dialog: improved usability. Do not select the entry content after opening the dialog; add a comma at the end of the tags entry to simplify the addition of a new tag; support keyboard navigation (bgo#777234). - Hi-DPI: "High" quality zoom is pixelated on hi-dpi monitors (bgo#779340). - Web services: fixed errors and removed features no longer supported by the APIs: . Facebook: fixed authorization process; removed photo importer. . PicasaWeb: removed album creation; remember application authorization (bgo#784142). . Photobucket: removed uploader. - Image viewer: use a white background for transparency (bgo#780207). - Fixed saving of uris in settings (bgo#783498). - Color manager: fixed crash when no device is found (bgo#789347). - Do not save selections. This broke workflows where selections were involved. - Removed obsolete file gthumb.convert (bgo#767034). + Updated manual translations. + Updated translations. - Update Url to https://wiki.gnome.org/Apps/gthumb: current gThumb's web page. - Point fdupes to the data directory instead of the build root, which is a practice that must be avoided. - Drop pkgconfig(sm) and its --with-smclient=xsmp configure parameter: not needed/used anymore since its parameter no longer exist. - Drop bison, flex, pkgconfig(ice) and itstool BuildRequires: they are not needed/used anymore. - Replace pkgconfig(libsoup-gnome-2.4) with pkgconfig(libsoup-2.4): libsoup is the required dependency and was already in use. They are different files but live in the same package-devel. ==== hddtemp ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== icecream ==== Subpackages: icecream-clang-wrappers - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== icinga ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== kernel-source ==== Version update (4.14.1 -> 4.14.2) Subpackages: kernel-default kernel-default-devel kernel-devel kernel-docs kernel-macros kernel-syms - ipmi_si: fix memory leak on new_smi (git-fixes). - commit 4ca3b35 - Linux 4.14.2 (bnc#1012628). - af_netlink: ensure that NLMSG_DONE never fails in dumps (bnc#1012628). - vxlan: fix the issue that neigh proxy blocks all icmpv6 packets (bnc#1012628). - net: cdc_ncm: GetNtbFormat endian fix (bnc#1012628). - fealnx: Fix building error on MIPS (bnc#1012628). - net/sctp: Always set scope_id in sctp_inet6_skb_msgname (bnc#1012628). - ima: do not update security.ima if appraisal status is not INTEGRITY_PASS (bnc#1012628). - serial: omap: Fix EFR write on RTS deassertion (bnc#1012628). - serial: 8250_fintek: Fix finding base_port with activated SuperIO (bnc#1012628). - tpm-dev-common: Reject too short writes (bnc#1012628). - rcu: Fix up pending cbs check in rcu_prepare_for_idle (bnc#1012628). - mm/pagewalk.c: report holes in hugetlb ranges (bnc#1012628). - ocfs2: fix cluster hang after a node dies (bnc#1012628). - ocfs2: should wait dio before inode lock in ocfs2_setattr() (bnc#1012628). - ipmi: fix unsigned long underflow (bnc#1012628). - mm/page_alloc.c: broken deferred calculation (bnc#1012628). - mm/page_ext.c: check if page_ext is not prepared (bnc#1012628). - coda: fix 'kernel memory exposure attempt' in fsync (bnc#1012628). - ipmi: Prefer ACPI system interfaces over SMBIOS ones (bnc#1012628). - commit 295c90a - apparmor: fix oops in audit_signal_cb hook (bnc#1069562). - Refresh patches.suse/0001-AppArmor-basic-networking-rules.patch. - commit d091ad8 - bio: ensure __bio_clone_fast copies bi_partno (bnc#1069605). - commit 59c6ade ==== ldns ==== Version update (1.6.17 -> 1.7.0) - disable DANE verification when building with openssl < 1.1 to fix build on distributions that have openssl 1.0.x - Update descriptions. - Update to version 1.7.0 * Ldns built with openssl-1.1.0 [bsc#1042653] * Fix #551 change Regent to Copyright holder in BSD license in some of the headings of the file, to match the opensource.org BSD license. * -e option makes ldns-compare-zones exit with status code 2 on difference * Filter out specified RR types with ldns-read-zone -e and -E options * bugfix #563: Correct DNSKEY from DSA private key. * bugfix #562: ldns-keygen match DSA key maximum size with library. And check keysizes with all algorithms. * ldns-verify-zone accepts only one single zonefile as argument. * bugfix #573: ldns-keygen write private keys with mode 0600. * Fix configure to make ldns compile with LibreSSL 2.0 * drill now also accepts dig style -y option (-y <[algo:]name:key> i.s.o. -y <name:key[:algo]>) * OPENPGPKEY draft rr types. Enable with: --enable-rrtype-openpgpkey * bugfix #608: Correct comment about escaped characters * CDS and CDNSKEY rr type from RFC 7344. --enable-rrtype-cds configure option removed * fix: Memory leak in ldns_pkt_rr_list_by_name() * fix: Memory leak in ldns_dname2buffer_wire_compress() * bugfix #613: Allow tab as whitespace too in last rdata field of types of variable length. * bugfix: strip trailing whitespace from $ORIGIN lines in zone files * Let ldns-keygen output .ds files only for KSK keys * Parse RFC7218 TLSA mnemonics, but do not output them * Let ldns-dane use SPKI as the default selector i.s.o. Cert * bugfix: Fit left over NSEC3s once more before adding empty non terminals * bugfix #605: Determine default trust anchor location at compile time * bugfix #697: Double free with ldns-dane create * bugfix #623: Do not redefine bool type and boolean values * bugfix #570: Add TLSA, CDS, CDNSKEY and OPENPGPKEY RR types to ldnsx * bugfix #575: ldns_pkt_clone() does not copy timestamp field * bugfix #584: ldns-update fixes. Send update to port 53, bring manpage in sync with the usage text, and don't alter the ldns_resolver passed to ldns_update_soa_zone_mname(). Created a ldns_resolver_clone() function in the process. * bugfix #633: ldns_pkt_clone() parameter isn't const. * bugfix: ldns-dane manpage correction * RFC7553 RR Type URI is supported by default. * Fix ECDSA signature generation, do not omit leading zeroes. * bugfix: Get rid of superfluous newline in ldns-keyfetcher * bugfix: -U option to ldns-signzone to sign with every algorithm * const function parameters whenever possible. * bugfix #725: allow RR-types on the type bitmap window border * Add type CSYNC support, RFC 7477. * Prepare for ED25519, ED448 support: todo convert* routines in dnssec.h, once openssl has support for signing with these algorithms. The dns algorithm number is not yet allocated. These features are not fully implemented yet, openssl (1.1) does not support the algorithms enough to generate keys and sign and verify with them. * Fix drill axfr ipv4/ipv6 queries. * Fix for openssl 1.1.0 API changes. * bugfix #825: Module import breaks with newer SWIG versions. * bugfix #769: Add support for :: in an IPv6 address * bugfix #708: warnings and errors with xcode 6.1/7.0 * bugfix #754: Memory leak in ldns_str2rdf_ipseckey * bugfix #661: Fail NSEC3 signing when NSEC domainname length would overflow. * bugfix #771: hmac-sha224, hmac-sha384 and hmac-sha512 keys. * bugfix #680: ldns fails to reject invalidly formatted RFC 7553 URI RRs. * bugfix #678: Use poll i.s.o. select to support > 1024 fds * Use OpenSSL DANE functions for verification (unless explicitly disabled with --disable-dane-ta-usage). * Bumb .so version * Include OPENPGPKEY RR type by default * rdata processing for SMIMEA RR type * Fix crash in displaying TLSA RR's. * Update ldns-key2ds man page to mention GOST and SHA384 hash functions. * Add sha384 and sha512 tsig algorithm. * Clarify data ownership with consts for tsig parameters. * bugfix: Fix detection of DSA support with OpenSSL >= 1.1.0 * bugfix #1160: Provide sha256 for release tarballs * --enable-gost-anyway compiles GOST support with OpenSSL >= 1.1.0 even when the GOST engine is not available. - Dropped patch ldns-perl-5.22.patch ==== libchewing ==== Version update (0.4.0+git20150820.285b47d -> 0.5.1+git20171114.3df07c9) Subpackages: chewing-data chewing-utils libchewing3 - Update to version 0.4.0+git20171114.3df07c9: * CMakeLists.txt: Add stress * sort data/tsi.sr * Add Carpalx keyboard layout support * Fix warning in GCC * Let CHEWING_DATA_PREFIX and TEST_DATA_DIR be absolute * Fix wrong CHEWING_DATADIR definition in CMake build * Fix jk_selection * an API to convert phone to bopomofo * Test chewing_phone_to_bopomofo with invalid phone * Implement has_userphrase with chewing_userphrase_lookup * libchewing.texi: Fix format tweaks * Adjust behavior of chewing_phone_to_bopomofo * Fix array index out of bound error in stress.c * Make stress.c less verbose * Prefer https for web links * stress.c: Don't force set data path * Fix #235: Unable to input some Bopomofos with ET26 layout * prevent memory leak in error case of GetDefaultUserPhrasePath * stress: clean_userphrase before each test * stress: clean up temp ~userphrase after test * stress: fix -extra argument parsing * refactoring key tables in chewing test * simplify simulate.c * testchewing: support read input from file * stress: support read input from file * Fix PinYinData keySeq buffer overflow in PinYinInput() * Fix trying to allocate buffer of negative size * Fix access violation due to negative cursor position * stress.c: more verbose output * Fix heap-buffer-overflow when read wordSeq * Use strncmp to avoid buffer overflow * tree: avoid undefined allocation of 0 bytes (CERT MEM04-C; CWE-131) * pinyin: avoid null-pointer dereference * chewingutil: fix dead assignment * Revert "libchewing.texi: Add chewing_bopomofo_String" * Fix memory leak in Discard2 * Fix memory leak in UserUpdatePhrase(). * code clean up, use new DestroyUserPhraseData() * userphrase-hash: avoid double free in TerminateUserphrase * test-userphrase: add test_userphrase_double_free * hash: fix hash item offset calculating in uhash.dat * hash: fix memory leak in HashFileOffsetWithUserPhrase * hash: fix memory access alignment * Fix ChoiceEndChoice expression always true * hash: follow consistent coding style * Update jvc-convert of KB layout Hsu * Auto learn phrases only after committing * Implement chewing_set_autoLearn() and chewing_get_autoLearn() * Fix indent in test/stress.c * Fix build with internal sqlite3 * Fix: chewing_handle_Space() cannot input space when buffer is empty * hash: remove goto statement * Remove API changes/compatibility report link * hash: follow consistent coding style ==== lomoco ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== mailman ==== - remove distributable flag (which is always true): drops SuSEconfig.mailman-SuSE, mailman-SuSE.patch, mailman-SuSE2.patch - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== mcelog ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== openssl ==== Version update (1.0.2m -> 1.1.0g) Subpackages: libopenssl-devel - Update to 1.1.0g release ==== osc ==== Version update (0.161.1 -> 0.162.0) - 0.162.0 - Additional file content check with SHA256 when using OBS 2.9 - support deb format with control.tar.xz ==== postgresql ==== Version update (9.6 -> 10) Subpackages: postgresql-contrib postgresql-devel postgresql-docs postgresql-server - instead of hardcoding version 10, just check if the prio is smaller than 80. - simplify the postgresql-install-alternatives to only require the pg version and go from there. supported calls of the script are: /usr/share/postgresql/install-alternatives postgresqlXY /usr/share/postgresql/install-alternatives XY where XY can currently be: 92 93 94 95 96 10 - CVE-2017-14798, bsc#1062722: Fix LPE via postgres init script. This only affected SysV-init based systems (i.e. SLE11). - postgresql-test was missing a %files section. - Move the start/stop script to /usr/share/postgresql - Add a script for handling update-alternatives as subpackages get installed/uninstalled, so that the implementation packages don't have to repeat the logic (postgresql-install-alternatives). - Add the postgresql-test subpackage. - Fix some dependencies. - Save the enabled and running state when upgrading from the obsolete postgresql-init package. - properly guard the systemd macro in %pre server so it wont be used on sle11 - add Requires(pre) so we have groupadd/useradd available - Refine the dependencies between the "dummy" packages and the actual implementation of PostgreSQL. The dummy packages now require an arbitrary implementation and recommend the default version. - Move the contents of postgresql-init to the server package. - bump to 10 - Bump version and defaultpackage to 9.6 in SLE12 (bsc#1046324). ==== postgresql10 ==== Version update (9.6.4 -> 10.1) Subpackages: libecpg6 libpq5 - port /usr/share/postgresql/install-alternatives usage to the new scheme - update to 10.1 * Security Issues - CVE-2017-15098: Memory disclosure in JSON functions - CVE-2017-15099: INSERT ... ON CONFLICT DO UPDATE fails to enforce SELECT privileges Prior to this release, the "INSERT ... ON CONFLICT DO UPDATE" would not check to see if the executing user had permission to perform a "SELECT" on the index performing the conflicting check. Additionally, in a table with row-level security enabled, the "INSERT ... ON CONFLICT DO UPDATE" would not check the SELECT policies for that table before performing the update. This fix ensures that "INSERT ... ON CONFLICT DO UPDATE" checks against table permissions and RLS policies before executing. * Bug Fixes and Improvements This update also fixes a number of bugs reported in the last few months. Some of these issues affect only version 10, but many affect all supported versions: - Fix a race condition in BRIN indexing that could cause some rows to not be included in the indexing. - Fix crash when logical decoding is invoked from a PL language function. - Several fixes for logical replication. - Restored behavior for CTEs attached to INSERT/UPDATE/DELETE statements to pre-version 10. - Prevent low-probability crash in processing of nested trigger firings. - Do not evaluate an aggregate function's argument expressions when the conditions in the FILTER clause evaluate to FALSE. This complies with SQL-standard behavior. - Fix incorrect query results when multiple GROUPING SETS columns contain the same simple variable. - Fix memory leak over the lifespan of a query when evaluating a set-returning function from the target list in a SELECT. - Several fixes for parallel query execution, including fixing a crash in the parallel execution of certain queries that contain a certain type of bitmap scan. - Fix json_build_array(), json_build_object(), jsonb_build_array(), and jsonb_build_object() to handle explicit VARIADIC arguments correctly. - Prevent infinite float values from being casted to the numeric type. - Fix autovacuum's ?work item? logic to prevent possible crashes and silent loss of work items. - Several fixes for VIEWs around adding columns to the end of a view. - Fix for hashability detection of range data types that are created by a user. - Improvements on using extended statistics on columns for the purposes of query planning. - Prevent idle_in_transaction_session_timeout from being ignored when a statement_timeout occurred earlier. - Fix low-probability loss of NOTIFY messages due more than 2 billion transactions processing before any queries are executed in the session. - Several file system interaction fixes. - Correctly restore the umask setting when file creation fails in COPY or lo_export(). - Fix pg_dump to ensure that it emits GRANT commands in a valid order. - Fix pg_basebackup's matching of tablespace paths to canonicalize both paths before comparing to help improve Windows compatibility. - Fix libpq to not require user's home directory to exist when trying to read the "~/.pgpass" file. - Several fixes for ecpg. - This update also contains tzdata release 2017c, with updates for Fiji, Namibia, Northern Cyprus, Sudan, Tonga, and Turks & Caicos Islands, plus historical corrections for Alaska, Apia, Burma, Calcutta, Detroit, Ireland, Namibia, and Pago Pago. For more details see: https://www.postgresql.org/docs/current/static/release-10-1.html - Use /usr/share/postgresql/install-alternatives in the respective scriptlets. - Fix the logic around restart on upgrade and stop on removal. We bring the binaries, but we don't own the unit file. - Update to the final 10.0 release. Major enhancements include: * Logical replication using publish/subscribe * Declarative table partitioning * Improved query parallelism * Significant general performance improvements * Stronger password authentication based on SCRAM-SHA-256 * Improved monitoring and control A dump/restore using pg_dumpall, or use of pg_upgrade, is required for those wishing to migrate data from any previous release. Version 10 contains a number of changes that may affect compatibility with previous releases. See the release notes for details: https://www.postgresql.org/docs/10/static/release-10.html - Reduce the number of version-related macros in the spec file. - Break a build dependency loop in the server packages that led to bootstrap problems. - Update to 10rc1 PostgreSQL 10 RC 1 requires an upgrade from beta 4, or earlier either using pg_dump / pg_restore or pg_upgrade. Changes since beta4 include: * Add psql variables showing server version and psql version. * Several fixes for partitioning * Several fixes for logical replication * Several fixes for transition tables * Fix for query that could end up in an uninterruptible state - Adjust dependencies to the new packaging schema. - update to 10beta4 PostgreSQL 10 beta 4 requires an upgrade from beta 3, or earlier either using pg_dump / pg_restore or pg_upgrade. Any bugfixes applied to 9.6 or earlier that also affected 10 are included in beta 4. Our users and contributors also reported bugs against 10 beta 3, and many of them have been fixed in this release. We urge our community to re-test to ensure that these bugs are actually fixed, including: - Show foreign tables in information_schema.table_privileges view. This fix applies to new databases, see the release notes for the procedure to apply the fix to an existing database. - Clean up handling of a fatal exit (e.g., due to receipt of SIGTERM) that occurs while trying to execute a ROLLBACK of a failed transaction - Remove assertion that could trigger during a fatal exit - Correctly identify columns that are of a range type or domain type over a composite type or domain type being searched for - Prevent crash when passing fixed-length pass-by-reference data types to parallel worker processes - Fix crash in pg_restore when using parallel mode and using a list file to select a subset of items to restore - Change ecpg?s parser to allow RETURNING clauses without attached C variables - Change ecpg?s parser to recognize backslash continuation of C preprocessor command lines - Improve selection of compiler flags for PL/Perl on Windows - Fix make check to behave correctly when invoked via a non-GNU make program Note that some known issues remain unfixed. Before reporting a bug in the beta, please check the Open Items page. https://wiki.postgresql.org/wiki/PostgreSQL_10_Open_Items - update to 10beta3 - hash: Fix write-ahead logging bugs related to init forks - Fix oddity in error handling of constraint violation in ExecConstraints for partitioned tables - Use a real RT index when setting up partition tuple routing - Fix serious performance problems in json(b) to_tsvector() - Fix problems defining multi-column range partition bounds - Fix partitioning crashes during error reporting - Fix race conditions in replication slot operations - Fix very minor memory leaks in psql's command.c - PL/Perl portability fix: avoid including XSUB.h in plperl.c - Fix inadequate stack depth checking in the wake of expression execution changes - Allow creation of C/POSIX collations without depending on libc behavior - Fix OBJECT_TYPE/OBJECT_DOMAIN confusion - Remove duplicate setting of SSL_OP_SINGLE_DH_USE option - Fix crash with logical replication on a function index - Teach map_partition_varattnos to handle whole-row expressions - Fix lock upgrade hazard in ATExecAttachPartition - Apply ALTER ... SET NOT NULL recursively in ALTER ... ADD PRIMARY KEY - hash: Increase the number of possible overflow bitmaps by 8x - Only kill sync workers at commit time in subscription DDL - Fix bug in deciding whether to scan newly-attached partition - Make pg_stop_backup's wait_for_archive flag work on standbys - Fix handling of dropped columns in logical replication - Fix local/remote attribute mix-up in logical replication - bump version in update-alternatives call - use multibuild - initial package ==== ppp ==== - Add Requires group(dialout) (boo#1067511). ==== radvd ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== rpcbind ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== rsync ==== - Security fix [bsc#1066644, CVE-2017-16548] * Missing trailing '\0' character check could lead to remote DoS * Added rsync-CVE-2017-16548.patch ==== scons ==== Version update (3.0.0 -> 3.0.1) - Explicitly require python3 for python3 version of scons as dependency does not get generated automatically - SCons 3.0.1: * Fix return value handling in to_String_for_subst() * Fixe Variables.GenerateHelpText() to now use the sort parameter * Fix Tool loading logic from exploding sys.path with many site_scons/site_tools prepended on py3. * Add additional output with time to process each SConscript file when using --debug=time. * Fix broken subst logic with "$$([...])" * Java/Jar building improvements and fixes - Packaging changes: * drop scons-3.0.0-support-python-2-prints.patch, now upstream * Restore python2 support for SLE 12 and Leap 42.x - Switch to python3 ==== sensors ==== Subpackages: libsensors4 libsensors4-devel - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== smartmontools ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== susepaste ==== Subpackages: susepaste-screenshot - Cleanup spec file with spec-cleaner ==== syslogd ==== Subpackages: klogd syslog-service - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== tftp ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== tlp ==== Subpackages: tlp-rdw - Add tlp-1.0-run-mount.patch: Move runtime data from /var/run/tlp to /run/tlp (boo#1058983) ==== wondershaper ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) ==== words ==== - Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

6 6

[opensuse-factory] Error during upgrade to Tumbleweed 20171129
by Freek de Kruijf 30 Nov '17

30 Nov '17

During the upgrade systemctl status saned.socket gives error message: ● saned.socket Loaded: error (Reason: Is a directory) Active: inactive (dead) Freek de Kruijf member openSUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

1 0

[opensuse-factory] Tumbleweed: Manpage formatting gone wrong
by Achim Gratz 30 Nov '17

30 Nov '17

Some of the manpages seem to be incorrectly formatted, for example the systemd manpages: --8<---------------cut here---------------start------------->8--- TIMESYNCD.CONF(5) timesyncd.conf TIMESYNCD.CONF(5) .SH "NAME" timesyncd.conf, timesyncd.conf.d - Network Time Synchronization configuration files --8<---------------cut here---------------end--------------->8--- Is this a configuration problem on my side or something that I should report as a bug? Regards, Achim. -- +<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+ DIY Stuff: http://Synth.Stromeko.net/DIY.html -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

5 6

[opensuse-factory] Bluetooth keeps failing in Tumbleweed, including the current version 20171125, Linux 4.14.1-1-default #1 SMP PREEMPT
by Foolish Ewe 30 Nov '17

30 Nov '17

Hello All: I am using opensuse tumbleweed for the last 6 weeks or so and keep having my bluetooth support fail while I'm using my laptop. I'm currently up to date (as of a few hours ago): $uname -a Linux linux 4.14.1-1-default #1 SMP PREEMPT Tue Nov 21 18:26:02 UTC 2017 (a5bca71) x86_64 x86_64 x86_64 GNU/Linux $ cat /etc/os-release NAME="openSUSE Tumbleweed" # VERSION="20171125" ID=opensuse ID_LIKE="suse" VERSION_ID="20171125" PRETTY_NAME="openSUSE Tumbleweed" ANSI_COLOR="0;32" CPE_NAME="cpe:/o:opensuse:tumbleweed:20171125" BUG_REPORT_URL="https://bugs.opensuse.org" HOME_URL="https://www.opensuse.org/" Consider the following log, I think bluetooth failed around 21:53:54, what could be going on here? Is this an usb error or bluetooth specific? What information do I need to track this down? $journalctl -all [Stuff removed for brevity] Nov 28 21:39:08 linux dbus[1126]: [system] Successfully activated service 'org.kde.powerdevil.backlighthelper' Nov 28 21:39:18 linux systemd[1]: Starting Cleanup of Temporary Directories... Nov 28 21:39:18 linux systemd[1]: Started Cleanup of Temporary Directories. Nov 28 21:41:51 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:43:39 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:44:20 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:47:20 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:47:28 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:52:04 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:53:54 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 21:53:54 linux kernel: xhci_hcd 0000:00:14.0: xHCI host controller not responding, assume dead Nov 28 21:53:54 linux kernel: xhci_hcd 0000:00:14.0: HC died; cleaning up Nov 28 21:53:54 linux kernel: usb 1-5: USB disconnect, device number 2 Nov 28 21:53:54 linux kernel: usb 1-7: USB disconnect, device number 3 Nov 28 21:53:54 linux systemd[1]: Starting Load/Save RF Kill Switch Status... Nov 28 21:53:54 linux systemd[1]: bluetooth.target: Unit not needed anymore. Stopping. Nov 28 21:53:54 linux systemd[1]: Stopped target Bluetooth. Nov 28 21:53:54 linux systemd[1]: Started Load/Save RF Kill Switch Status. Nov 28 21:53:54 linux bluetoothd[1117]: Endpoint unregistered: sender=:1.39 path=/MediaEndpoint/A2DPSource Nov 28 21:53:54 linux bluetoothd[1117]: Endpoint unregistered: sender=:1.39 path=/MediaEndpoint/A2DPSink Nov 28 21:53:54 linux dbus[1126]: [system] Rejected send message, 1 matched rules; type="error", sender=":1.39" (uid=1000 pid=2094 comm="/usr/bin/pulseaudio --start --log-target=syslog ") interface="(unset)" member="(unset)" error name="org.bluez.MediaEndpoint1.Error.NotImplemented" requested_reply="0" destination=":1.1" (uid=0 pid=1117 comm="/usr/lib/bluetooth/bluetoothd ") Nov 28 21:53:54 linux dbus[1126]: [system] Rejected send message, 1 matched rules; type="error", sender=":1.39" (uid=1000 pid=2094 comm="/usr/bin/pulseaudio --start --log-target=syslog ") interface="(unset)" member="(unset)" error name="org.bluez.MediaEndpoint1.Error.NotImplemented" requested_reply="0" destination=":1.1" (uid=0 pid=1117 comm="/usr/lib/bluetooth/bluetoothd ") Nov 28 21:53:54 linux dbus[1126]: [system] Rejected send message, 1 matched rules; type="error", sender=":1.39" (uid=1000 pid=2094 comm="/usr/bin/pulseaudio --start --log-target=syslog ") interface="(unset)" member="(unset)" error name="org.bluez.MediaEndpoint1.Error.NotImplemented" requested_reply="0" destination=":1.1" (uid=0 pid=1117 comm="/usr/lib/bluetooth/bluetoothd ") Nov 28 21:53:54 linux dbus[1126]: [system] Rejected send message, 1 matched rules; type="error", sender=":1.39" (uid=1000 pid=2094 comm="/usr/bin/pulseaudio --start --log-target=syslog ") interface="(unset)" member="(unset)" error name="org.bluez.MediaEndpoint1.Error.NotImplemented" requested_reply="0" destination=":1.1" (uid=0 pid=1117 comm="/usr/lib/bluetooth/bluetoothd ") After a reboot, I bluetooth failed pretty quickly, I see: ov 28 22:02:44 linux nm-dispatcher[2752]: req:2 'connectivity-change': start running ordered scripts... Nov 28 22:02:51 linux PackageKit[2580]: daemon quit Nov 28 22:02:51 linux packagekitd[2580]: Source ID 11 was not found when attempting to remove it Nov 28 22:02:57 linux kernel: xhci_hcd 0000:00:14.0: xHC is not running. Nov 28 22:02:57 linux kernel: xhci_hcd 0000:00:14.0: xHCI host controller not responding, assume dead Nov 28 22:02:57 linux kernel: xhci_hcd 0000:00:14.0: HC died; cleaning up Nov 28 22:02:57 linux kernel: usb 1-5: USB disconnect, device number 2 Nov 28 22:02:57 linux kernel: usb 1-7: USB disconnect, device number 3 Nov 28 22:02:57 linux systemd[1]: Starting Load/Save RF Kill Switch Status... Nov 28 22:02:57 linux systemd[1]: bluetooth.target: Unit not needed anymore. Stopping. Nov 28 22:02:57 linux systemd[1]: Stopped target Bluetooth. With best regards: Bill -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

5 11

[opensuse-factory] ffmpeg multithreads support
by Daniele 29 Nov '17

29 Nov '17

HI, while rendering with kdenlive, 1 core at 100% and 3 doing nothing. It' not clear to me if is a problem withn kdenlive, mlt or ffmpeg. I'm not the only one with this problem.. I've found some old post about ffmpeg that needs to be compiled with the option "--enable-pthreads". It seem that both version (suse and packman) are compiled without this option. So, is still valid ? If so, why is missing ? What else could be the problem ? Thanks, Daniele -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

7 18