Changelog for
memcached-1.4.25-1.3.x86_64.rpm :
Tue Dec 1 13:00:00 2015 p.drouandAATTgmail.com
- Update to version 1.4.25:
* Please read the news at
https://github.com/memcached/memcached/wiki/ReleaseNotes1425
- Update memcached-autofoo.patch
Mon Jun 15 14:00:00 2015 mrueckertAATTsuse.de
- fix comment in the sysconfig file
Sun Feb 22 13:00:00 2015 mpluskalAATTsuse.com
- Cleanup spec file
* using spec-cleaner
* remove unnecessary %defines
- Create new package (devel)
- Install either init script or unit file
- Refresh dependencies
- Update to 1.4.22
* gatkq: return key in response
* Handle SIGTERM the same as SIGINT
* Fix off-by-one causing segfault in lru_crawler
* Fix potential corruption for incr/decr of 0b items
* Fix issue #369 - uninitialized stats_lock
* issue#370 : slab re-balance is not thread-safe in function
do_item_get
* Fix potential corruption in hash table expansion
* use item lock instead of global lock when hash expanding
* fix hang when lru crawler started from commandline
* rename thread_init to avoid runtime failure on AIX
* Support -V (version option)
- Changes for 1.4.21
* makefile cleanups
* Avoid OOM errors when locked items stuck in tail
Sun Nov 9 13:00:00 2014 Led
- fix bashisms in pre script
Mon Jun 23 14:00:00 2014 andreas.stiegerAATTgmx.de
- fix source URL
Mon Jun 16 14:00:00 2014 mrueckertAATTsuse.de
- disable testsuite. passes with chroot build but fails on server
side build
Mon Jun 16 14:00:00 2014 mrueckertAATTsuse.de
- Bumping to latest version to include all fixes for:
bnc#858677 CVE-2013-7290
bnc#858676 CVE-2013-7291
bnc#857188 CVE-2013-7239
bnc#817781 CVE-2011-4971
bnc#798458 CVE-2013-0179
- update to 1.4.20
- Fix a race condition causing new connections to appear closed,
causing an inifinte loop.
- additional changes from 1.4.19
- Fix endianness detection during configure.
- Fixes a performance regression with binary protocol (up to
20%)
- Fix rare segfault in incr/decr.
- disable tail_repair_time by default.
- Likely not needed anymore, and can rarely cause bugs.
- use the right hashpower for the item_locks table. Small perf
improvement.
- Fix crash for LRU crawler while using lock elision (haswell+
processors)
- additional changes from 1.4.18
- Fixes
- fix LRU contention for first minute of uptime
- This made some synthetic benchmarks look awful.
- Make hash table algorithm selectable
- Don\'t lose item_size_max units in command line
- Add a \"stats conns\" command to show the states of open
connections.
- Allow caller-specific error text in binary protocol
- Stop returning ASCII error messages to binary clients
- Fix reference leak in binary protocol \"get\" and \"touch\"
handlers
- Fix reference leak in process_get_command()
- New Features
- New \"stats conns\" command, which will show you what currently
open connections are up to, how idle they\'ve been, etc.
- The jenkins hash was getting a little long in the tooth, and
we might want to add specific hash algorithms for different
platforms in the future. This makes it selectable in some
sense. We\'ve initially added murmur3 hash to the lineup and
that seems to run a tiny bit faster in some tests. -o
hash_algorithm=murmur3
- A new background thread emerges! Currently experimental, so
the syntax might change. If you run into bugs please let us
know (though it\'s been testing fine in torture tests so far).
- additional changes from 1.4.17
- Fixes
- Fix potential segfault in incr/decr routine.
- Fix potential unbounded key prints (leading to crashes in
logging code)
- Fix bug which allowed invalid SASL credentials to
authenticate.
- Fix udp mode when listening on ipv6 addresses.
- Fix for incorrect length of initial value set via binary
increment protocol.
- New Features
- Add linux accept4() support. Removes one syscall for each new
tcp connection
- scripts/memcached-tool gets \"settings\" and \"sizes\" commands.
- Add parameter (-F) to disable flush_all. Useful if you never
want to be able to run a full cache flush on production
instances.
- additional changes from 1.4.16
- Fixes
- Builds on OS X Mavericks (with clang)
- Add statistics for allocation failures
- Issue 294 : Check for allocation failure
- Make tail leak expiry time configurable (-o
tail_repair_time=60)
- Fix segfault on specially crafted packet.
- Close connection on update_event error while parsing new
commands
- Don\'t truncate maxbytes stat from \'stats settings\'
- Add the \"shutdown\" command to the server. This allows for
better
- automation
- fix enable-sasl-pwdb
- New Features
Adjusting tail repair time: -o tail_repair_time=60 (in seconds)
- dropped memcached-1.4.x_delete_verbose_mode_dos.patch:
included upstream
- freshed memcached-autofoo.patch
Tue Jan 15 13:00:00 2013 mrueckertAATTsuse.de
- added memcached-1.4.x_delete_verbose_mode_dos.patch (bnc#798458)
DoS when printing out keys to be deleted in verbose mode
Upstream bug 306 (CVE-2013-0179)
Tue Nov 20 13:00:00 2012 dimstarAATTopensuse.org
- Fix useradd invocation: -o is useless without -u and newer
versions of pwdutils/shadowutils fail on this now.
Wed Nov 7 13:00:00 2012 chrisAATTcomputersalat.de
- update to version 1.4.15
* Add some mild thread documentation
* README.md was missing from dist tarball
* Issue 286 : --disable-coverage drops \"-pthread\" option
* Reduce odds of getting OOM errors in some odd cases
- rebase use-endian_h, autofoo patch and 1.4.5.dif
- fix build <= 1140
* export LIBEVENT_CFLAGS and LIBEVENT_LIBS so we dont need
pkgconfig check for libevent on <= 1140
Wed Nov 7 13:00:00 2012 mrueckertAATTsuse.de
- fix build on older distros
- memcached-autofoo.patch: removed no-dist-gzip dist-xz
- added new conditional to guard all the systemd stuff and
guarded the general bcond_without with an suse_version > 12.2
- export LIBEVENT_CFLAGS and LIBEVENT_LIBS so we dont need
pkgconfig check for libevent on sles11
- use makeinstall instead of make_install
Tue Aug 14 14:00:00 2012 crrodriguezAATTopensuse.org
- Update to version 1.4.14
* Avoid race condition in test during pid creation by blind retrying
* Fixed issue with invalid binary protocol touch command expiration time
- If the test suite fails, package must fail build.
- Use byteswapping macros from endian.h and not some ad-hoc/slow
function.
- Add systemd units.
Tue Apr 3 14:00:00 2012 tabrahamAATTnovell.com
- removed fix-strict-aliasing.patch deprecated by this release
- Update to version 1.4.13
* Fix inline issue with older compilers (gcc 4.2.2)
* Better detection of sasl_callback_ft
- Changes from version 1.4.12
Fixes:
* fix glitch with flush_all (exptime)
* Skip SASL tests unless RUN_SASL_TESTS is defined.
* Look around for saslpasswd2 (typically not in the user\'s path).
* build fix: Define sasl_callback_ft on older versions of sasl.
* fix segfault when sending a zero byte command
* fix warning in UDP test
* properly detect GCC atomics
* tests: loop on short binary packet reads
* fix slabs_reassign tests on 32bit hosts
- Changes from version 1.4.11
Fixes:
* bug237 : Don\'t compute incorrect argc for timedrun
* fix \'age\' stat for stats items
* binary deletes were not ticking stats counters
* Fix a race condition from 1.4.10 on item_remove
* close some idiotic race conditions
* initial slab automover
* slab reassignment
* clean do_item_get logic a bit. fix race.
* clean up the do_item_alloc logic
* shorten lock for item allocation more
* Fix to build with cyrus sasl 2.1.25
New features:
* Slab page reassignment and bug fixes over 1.4.10.
- Changes from version 1.4.10
Fixes:
* Disable issue 140 \'s test.
* Push cache_lock deeper into item_alloc
* Use item partitioned lock for as much as possible
* Remove the depth search from item_alloc
* Move hash calls outside of cache_lock
* Use spinlocks for main cache lock
* Remove uncommon branch from asciiprot hot path
* Allow all tests to run as root
New features:
* tested improvements in speed between 3 and 6 worker
threads (-t 3 to -t 6) More than 6 reduced speed
- Changes from version 1.4.9
* Add a systemd service file
* Fix some minor typos in the protocol doc
* Issue 224 - check retval of main event loop
* Fix -c so maxconns can be raised above default.
- Changes from version 1.4.8
Fixes:
* Fix to write correct pid from start-memcached
* Fix to enable LRU when using binary protocol
* Upgrade stats items counters to 64bit
* Add new stats expired_unfetched, evicted_unfetched
* Allow setting initial size of the hash table
* Expose stats for the internal hash table
* bug220 : incr would sometimes return the previous item\'s CAS
* Fixed bug on multi get processing
* Experimental maxconns_fast option
* Add an ASCII touch command
* Add binary GATK/GATKQ
* Backport binary TOUCH/GAT/GATQ commands
* Issue 221 : Increment treats leading spaces as 0
* Fix compile error on OS X
New features:
* touch commands
* fast connection limit handling
* internal hash table
* expored_unfetched, evicted_unfetched
- Changes from version 1.4.7
Fixes:
* Use a monotonically increasing timer
* Immediately expire items when given a negative expiration time
* fix memcached-tool to print about all slabs
* Properly daemonize memcached for debian
* Don\'t permanently close UDP listeners on error
* Allow memcached-init to start multiple instances (not recommended)
* Issue 214 : Search for network libraries before searching for libevent
* Issue 213 : Search for clock_gettime in librt
* Issue 115 : accont for CAS in item_size_ok
* Fix incredibly slim race for maxconns handler. Should no longer hang ever
* Issue 183 - Reclaim items dead by flush_all
* Issue 200 : Don\'t fire dtrace probe as the last thing in a function
New features:
* monotonic clock
- Changes from version 1.4.6
* Gcc on Solaris sparc wants -R and not -rpath
* Issue 121 : Set the runtime path when --with-libevent is used
* Fix autogen failure when unable to find supported command.
* fix race crash for accepting new connections
* fix incr/decr race conditions for binary prot
* fix incr/decr race conditions for ASCII prot
* Compile fix (-Werror=unused-but-set-variable warnings)
* Bind each UDP socket to an a single worker thread in multiport env
* Add support for using multiple ports
* Issue 154 : pid file out of sync (created before socket binding)
* Issue 163 : Buggy mem_requested values
* Fix cross compilation issues in configure
* Issue 140 - Fix age for items stats
* Issue 131 - ChangeLog is outdated
* Issue 155 : bind to multiple interface
* Issue 161 incorrect allocation in cache_create
* Fix type-punning issues exposed with GCC 4.5.1
* Simplify stats aggregation code
* Reverse backward expected/actual params in test
* Issue 152 : Fix error message from mget
* Refuse to start if we detect libevent 1.12
* Fix compilation issue on Solaris 9 wrt isspace() macro - Resolves
issue 111
New features:
* multiple port binding
Fri May 27 14:00:00 2011 cooloAATTnovell.com
- follow fedora\'s lead and remove Werror
Thu May 6 14:00:00 2010 freitagAATTnovell.com
- update to version 1.4.5:
This is a maintenance release with some build fixes, doc fixes, and one new stat.
Fixes:
* Properly detect CPU alignment on ARM. bug100
* Remove 1MB assertion. bug 119
* More automake versions supported.
* Compiler warning fixes for OpenBSD.
* potential buffer overflow in vperror
* Report errors opening pidfiles using vperror
New Features::
* New stat: reclaimed
* sasl_pwdb for more simple auth deployments
Fri Jun 5 14:00:00 2009 mrueckertAATTsuse.de
- update to version 1.2.8:
- make -b command actually work
-
*critical bugfix
*. In 1.2.7 under multithreaded mode, memcached
would never restart accepting connections after hitting the
maximum connection limit.
- remove \'stats maps\' command, as it is a potential information
leak, usable if versions prior to 1.2.8 ever have buffer
overflows discovered. (bnc#501656) CVE-2009-1494
- additional changes from version 1.2.7
- reset new stats with \'stats reset\'
- Slew of new tests. (misc, mostly Dustin Sallings)
- Minor bug fixes. (misc, mostly Dustin Sallings, some Dormando)
(see git history for full list)
- -b command for setting the tcp listen backlog (Chris Goffinet)
- Workaround for a more major bug that very rarely makes
memcached stop allowing new data to be set. (Dormando)
- Print why a key was expired in -vv mode (Dormando)
- cmd_flush stat (Dormando)
- listen_disabled_num stat for determining if you\'ve hit the
maxconns limit (Dormando)
- Display error status on listen failures (Dormando)
- Remove managed instance code. Incomplete/etc. (Dormando)
- Handle broken IPV6 stacks better (Brian Aker)
- Generate warnings on setsockopt() failures (Brian Aker)
- Fix some indentation issues (Brian Aker)
- UDP/TCP can be disabled by setting their port to zero (Brian
Aker)
- Zero out libevent thread structures before use (Ricky Zhou)
- New stat: Last accessed time for last evicted item per slab
class. (Dormando)
- Use a dedicated socket accept thread (Facebook)
- Add -R option. Limit the number of requests processed by a
connection at once. Prevents starving other threads if bulk
loading. (Facebook)
