Changelog for nbd-3.16.2-1.2.x86_64.rpm :

* Fri Jan 05 2018 mpluskalAATTsuse.com- Update to version 1.16.2:
* Make the test suite less chatty
* Various build system improvements
* Fixes to the systemd unit to make it work again with recent systemd
* Point to the nbd mailinglist, rather than to the maintainer\'s personal email address, for bug reports.
* Thu Nov 23 2017 rbrownAATTsuse.com- Replace references to /var/adm/fillup-templates with new %_fillupdir macro (boo#1069468)
* Fri Oct 13 2017 olafAATTaepfle.de- Disable restart on update to keep existing client connections alive (bsc#1063135)
* Thu Sep 07 2017 olafAATTaepfle.de- Add nbd-client.service which will provide a nbdAATT.service to allow other units to require a nbdAATTnbd0.service if /etc/nbdtab is configured accordingly
* Thu Jun 08 2017 mpluskalAATTsuse.com- Update to version 3.16.1:
* Only change is a minor bugfix from 3.16, so that - -enable-syslog is unbroken.- Changes for version 3.16.0:
* Implement the NBD_OPT_GO and NBD_OPT_INFO negotiation options in both the client and the server, and use them in preference over NBD_OPT_EXPORT_NAME (but fall back to the latter in case the serverdoesn\'t support NBD_OPT_GO).
* Make sure we still run the postrun command, even if the connection was closed improperly.
* Add a new \"waitfile\" parameter.
* Make sure that nbd compiles on non-Linux platforms again, in the face of all variants of the GnuTLS options.
* Various fixes for issues found by Coverity
* Fri Mar 24 2017 mpluskalAATTsuse.com- Update to version 3.15.2:
* Fix the copyonwrite situation with multiple threads. This has been broken since nbd 3.12, where multiple outstanding writes could corrupt the copyonwrite diff file.
* Build system improvements so that things work again if nbd-client isn\'t buildable while GnuTLS is enabled
* Documentation updates
* Fixes to the GnuTLS usage, and the ability to configure the priority string for that library- Changes for version 3.15.1:
* minor but crucial fix in the handling of the WRITE_ZEROES command, which would otherwise not work in cooperation with TLS.- Changes for version 3.15.0:
* Implement NBD_CMD_WRITE_ZEROES, for efficiently writing large numbers of zeroes (server side implementation only)
* (optional) support for NBD_OPT_STARTTLS, using GnuTLS.
* Support for multiplexing commands over multiple connections, for improving performance (requires Linux 4.10 or above).
* Support for using splice(), also for improving performance
* Protocol errors during negotiation now are accompanied by a human-readable error string.
* Fix the handling of maxconnections, so that it is not a global setting (over all exports) but rather a per-export setting.
* Fix handling of unknown options sent to the server. Previously we would desync from the protocol, causing another error and a disconnect.
* Produce a log message when the address family of the client socket does not match the expected address family in the authfile handling. A proper fix would convert IPv4 sockets to v6-mapped v4 sockets, but that\'s much more complicated.
* and various other minor fixes too small to mention here; see the changelog- Changes for version 3.14.0:
* Major review of doc/proto.md, with many clarifications.
* Some bug fixes for nbd-server that were discovered during the extensive updates to proto.md
* Some fixes for non-Linux platforms,
* Some minor incremental improvements to the new nbdtab functionality,
* Fix the way we tell the kernel about really small exports,
* Removal of references to the no longer supported oldstyle from all documentation
* Minor updates, clarifications, and syntax fixes for man pages and other documentation
* Addition of a systemd unit file for nbd-client. This requires a kernel which issues change events upon disconnect (commit 37091fdd83), and a udev which correctly sets SYSTEMD_READY- Drop init.nbd-server and support for non systemd distributions- Drop patches:
* nbd-2.9.25-close.diff (upstreamed)
* nbd-2.9.25-doxyfile.diff (no longer needed)- Do not build doc subpackage
* Mon Feb 08 2016 dmuellerAATTsuse.com- autoload nbd module
* Fri Feb 05 2016 dmuellerAATTsuse.com- nbd-server.service: implement reload
* Sat Jan 23 2016 mpluskalAATTsuse.com- Update to 3.13
* This is mainly a bugfix release, containing fixes found in 3.12.1 by compiling on different architectures, submitting the build to Coverity Scan, and some analysis of intermittent failures that found a race condition.
* Tue Dec 01 2015 p.drouandAATTgmail.com- Update to version 3.12.1
* ensures the source to the \"nbdtab\" man page is correctly shipped with the rest of the source.- Changes from version 3.12
* inetd mode should work again, now using the default export name in the newstyle protocol
* when specifying an export on the server command line or not specifying a name on a client command line, the newstyle protocol is used with the default export name (but a warning will be produced stating that oldstyle is no longer supported)
* All traces of oldstyle (except a few error messages) have been removed
* We now listen on IPv4 and IPv6 by default, regardless of what the system\'s configuration is like, by explicitly listening both on 0.0.0.0 and ::.
* Some bugs in the handling of the TRIM command have been removed
* nbd
*client now uses readit() in the negotiation, rather than plain non
*retrying read().
* It is now possible to listen on multiple addresses; the \"listenaddr\" parameter has been changed into a comma
*separated list of addresses to listen on
* nbd
*client now supports an nbdtab file, which allows connecting a device by just using the device name
* nbd
*server now uses a GThreadPool to implement asynchronous handling of requests
* Punching holes on files is now supported on cygwin, too.- Remove no-return-nonvoid.patch; fixed on upstream release
* Thu May 28 2015 mpluskalAATTsuse.com- Update to 3.11
* nbd-server: fix unsafe signal handling
* define error values as part of the protocol
* docs/proto: clarify NBD_CMD_FLUSH- Drop nbd_signaling_CVE-2015-0847.patch as fix is included in upstream
* Mon May 11 2015 mpluskalAATTsuse.com- Fix CVE-2015-0847
* nbd_signaling_CVE-2015-0847.patch
* Fri Apr 17 2015 mpluskalAATTsuse.com- Update to 3.10
* There really is only one change since 3.9: drop of the \"oldstyle\" support from nbd-server and nbd-client.
* Thu Mar 05 2015 mpluskalAATTsuse.com- Cleanup spec-file with spec-cleaner- Do not generate config.example from README
* add config.example as source- Add unit file (nbd-server.service)- Add sysconfig template (nbd-server.sysconfig)- Add no-return-nonvoid.patch- Update requirements- Update to 3.9
* AF_UNIX support
* Fixed test suite on non-GNU getopt() implementations
* New \"treefiles\" mode, which exports a gazillion of page-sized files; useful for exporting things which are stored on an SSHFS or amazon AWS (trough FUSE) or similar, where every write causes an upload to the backend storage
* New \"cowdir\" option, allowing to specify where copy-on-write files are written.
* Minor changes so that nbd-client can now also be compiled for the Android platform. This required removal of the -swap command line option, which requires the mlockall() system call, unavailable on Android.
* Protocol update: a reserved bit is used to avoid sending the 124 bytes of useless data at the beginning of the negotiation. The change is implemented so that things will still work with clients not supporting this option, however.
* gznbd is now built by the same build system, rather than a separate one. Note however that gznbd is still unmaintained; it should be considered a \"contrib\" feature.
* \"nbd-server -V\" will now output the nbd-server version number.
* Various fixes found through Coverity and the clang static analyzer, and lots of other minor things too small to mention here.
* Sun Nov 09 2014 Led - fix bashisms in pre script
* Thu Apr 18 2013 dmuellerAATTsuse.com- update to 3.3:
* and various bugfixes, upgrading is strongly recommended
* Fix various nbd-server misbehave bugs
* Tue Jul 10 2012 ajAATTsuse.de- Fix restart_on_update usage.
* Fri May 18 2012 tabrahamAATTnovell.com- Update to nbd-3.1
* minor bug fixes
* new feature: ability for the client to list the exports a server supports
* Wed Feb 29 2012 tabrahamAATTnovell.com- Update to nbd-3.0
* minor bug fixes- removed obsolete nbd-2.9.25-have_falloc_ph.diff- compile with -fno-strict-aliasing to prevent breakage with new gcc
* Thu Jan 05 2012 tabrahamAATTnovell.com- Update to nbd-2.9.25
* TRIM now uses FALLOC_FL_PUNCH_HOLE
* minor bug fixes- removed obsolete nbd-2.9.20-bigendian.diff- refreshed nbd-2.9.20-close.diff
* Fri Sep 16 2011 rfnoAATTfietze-home.de- Install nbd-server.5 man page
* Mon Apr 04 2011 msAATTsuse.de- socklen_t is always 32bit, size_t is 64bit. so a classical big endian 64bit error (bnc #674281)
* Fri Feb 25 2011 msAATTsuse.de- Update to nbd-2.9.20:
* Documentation update
* Fix buffer size checking (bnc #674281)
* Extend test suite and fix named exports
* Wed Aug 05 2009 msAATTsuse.de- fixed missing close to open (bnc #528272)
* Thu Jul 09 2009 msAATTsuse.de- Update to nbd-2.9.13:
* Documentation update
* Don\'t try working on partitions When a partition is given to the -d or -c command line options, reduce it to the master device instead. This assumes udev-style naming of device entries, though.
* Fix autodetection Everything else in nbd-server was already 64bit-clean, but this part, where the size of a block device is autodected, wasn\'t. Replace the ioctl() to BLKGETSIZE with its 64bit version, so that trying to detect the size of a multi-terabyte block device does not fail.
* Avoid unaligned access
* Actually go daemon just doing chdir(\"/\") leave file descriptors open, which is not ideal. Doing it this way closes everything, including any system-specific things we\'re not aware of.
* Remove support for timeout option
* Mon May 04 2009 msAATTsuse.de- Update to nbd-2.9.12:
* use atoll instead of atol ... so that multi-terabyte exports work, too
* Change meaning of -swap option: rather than trying to use a kernel patch which has been effectively dead since sometime around 2.1.something, do an mlockall() so that the kernel won\'t throw us out. The MCL_CURRENT | MCL_FUTURE bit might prove to be problematic, but if so, we\'ll deal with it when we get there
* Use right printf() format string