]>
git.ipfire.org Git - thirdparty/kea.git/log
Naoki Kambe [Mon, 24 Sep 2012 05:03:44 +0000 (14:03 +0900)]
[2222] corrected the comment according to what the code does
Naoki Kambe [Mon, 24 Sep 2012 02:46:27 +0000 (11:46 +0900)]
[2222] expanded the lambda statement to the normal method definition
Naoki Kambe [Mon, 24 Sep 2012 02:43:12 +0000 (11:43 +0900)]
[2222] simplified complicated assignment of `incrementer`
Mukund Sivaraman [Mon, 24 Sep 2012 05:41:47 +0000 (11:11 +0530)]
[2218] Remove obsolete comment
Mukund Sivaraman [Mon, 24 Sep 2012 05:41:26 +0000 (11:11 +0530)]
[2218] Rename InMemoryClient::findZone2() to ::findZoneData()
Also make it non-const and non-virtual.
Mukund Sivaraman [Mon, 24 Sep 2012 05:24:54 +0000 (10:54 +0530)]
[2218] Change node and level count types from size_t to uint32_t
Ideally, size_t should be used for most kinds of memory sizes, but in
this case, we are trying to keep down the size of a DomainTree object.
Mukund Sivaraman [Mon, 24 Sep 2012 05:15:09 +0000 (10:45 +0530)]
[2218] Make some DomainTreeNode methods private again
isSubTreeRoot(), getSubTreeRoot() and getLargestInSubTree() were made
private again.
Mukund Sivaraman [Mon, 24 Sep 2012 05:06:28 +0000 (10:36 +0530)]
[2218] Use alternate DomainTree methods to get the covering node
This avoids the use of predecessor() and successor() and
getLargestInSubTree(), and uses previousNode() and largestNode()
instead.
Mukund Sivaraman [Mon, 24 Sep 2012 05:05:09 +0000 (10:35 +0530)]
[2218] Add DomainTree::largestNode() to return the largest node in tree-of-trees
There was an old version of this method, but that was buggy.
Mukund Sivaraman [Mon, 24 Sep 2012 05:00:33 +0000 (10:30 +0530)]
[2218] Rearrange method in domaintree.h
JINMEI Tatuya [Sat, 22 Sep 2012 05:53:35 +0000 (22:53 -0700)]
[2219] cover a minor case: zone finder's getOrigin() with non absolute label.
JINMEI Tatuya [Fri, 21 Sep 2012 23:18:03 +0000 (16:18 -0700)]
[2219] moved in-memory unit tests under datasrc/tests to resolve dependency.
the tests now depend on libdatasrc, while libdatasrc needs memory.la.
so we need to build any tests after building all libraries.
JINMEI Tatuya [Fri, 21 Sep 2012 22:39:35 +0000 (15:39 -0700)]
[2219] removed memory_datasrc.cc from b10-auth's Makefile.am's.
we don't need it any more.
JINMEI Tatuya [Fri, 21 Sep 2012 22:36:35 +0000 (15:36 -0700)]
[2219] workaround: avoid redundant initial configuration for data sources.
it can take unnecessarily long if in-memory cache contains a large zone.
JINMEI Tatuya [Fri, 21 Sep 2012 22:19:12 +0000 (15:19 -0700)]
[2219] cleanup: removed a redundant toText() in log arg.
JINMEI Tatuya [Fri, 21 Sep 2012 22:11:13 +0000 (15:11 -0700)]
[2219] bug fix: corrected too-string validity check in lblseq constructor.
offset values can actually be larger than max_labellen (63). we should have
check label length field of the name data. added tests that would trigger
this bug.
JINMEI Tatuya [Fri, 21 Sep 2012 19:47:11 +0000 (12:47 -0700)]
[2218] some more trivial cleanups
JINMEI Tatuya [Fri, 21 Sep 2012 18:40:16 +0000 (11:40 -0700)]
[2218] avoid initializing nsec3_hash_map namespace scope to avoid fiasco.
Marcin Siodelski [Fri, 21 Sep 2012 18:04:17 +0000 (20:04 +0200)]
[1960] Added comments why testing CommandOptions::isUseFirst().
Marcin Siodelski [Fri, 21 Sep 2012 17:50:13 +0000 (19:50 +0200)]
[1960] Fixed some comments in test_control.h.
Marcin Siodelski [Fri, 21 Sep 2012 17:42:05 +0000 (19:42 +0200)]
[1960] Add EXPECT_NO_THROW when parsing perfdhcp command line.
Marcin Siodelski [Fri, 21 Sep 2012 17:32:08 +0000 (19:32 +0200)]
[1960] Remove unused CLOCK_GETTIME_LDFLAGS and HAVE_GETIFADDRS.
The refactored code neither uses clock_gettime nor HAVE_GETIFADDRS.
Marcin Siodelski [Fri, 21 Sep 2012 17:19:04 +0000 (19:19 +0200)]
[1960] Fixed doxygen documentation.
Tomek Mrugalski [Fri, 21 Sep 2012 17:11:58 +0000 (19:11 +0200)]
Merge branch 'trac2238' into trac2269
Tomek Mrugalski [Fri, 21 Sep 2012 17:01:14 +0000 (19:01 +0200)]
[2238] Pool, Subnet and Triplet moved to separate files.
Tomek Mrugalski [Fri, 21 Sep 2012 16:45:21 +0000 (18:45 +0200)]
[2238] Changes after review.
Marcin Siodelski [Fri, 21 Sep 2012 16:22:45 +0000 (18:22 +0200)]
[2272] Added changes based on the code review.
Jelte Jansen [Fri, 21 Sep 2012 15:00:24 +0000 (17:00 +0200)]
[2254] fix 'config remove Boss' crash
now shows a (admittedly rather cryptic) error
Jelte Jansen [Fri, 21 Sep 2012 14:57:35 +0000 (16:57 +0200)]
[2254] Fix tab-completion in lists
Marcin Siodelski [Fri, 21 Sep 2012 14:46:44 +0000 (16:46 +0200)]
[1960] Implemented changes suggested in code review.
Jelte Jansen [Fri, 21 Sep 2012 13:32:50 +0000 (15:32 +0200)]
[2254] remove 'argument' and 'identifier' from tab hints
These are needed to work with the way bindctl originally handled commands, but should not appear in normal tab-completion
Tomek Mrugalski [Fri, 21 Sep 2012 13:06:06 +0000 (15:06 +0200)]
Merge branch 'master' into trac2238
Conflicts:
ChangeLog
src/lib/dhcp/tests/Makefile.am
Tomek Mrugalski [Fri, 21 Sep 2012 12:21:58 +0000 (14:21 +0200)]
[2269] dhcp6.spec update, documentation written
Michal 'vorner' Vaner [Fri, 21 Sep 2012 10:57:24 +0000 (12:57 +0200)]
Changelog for #1375
Michal 'vorner' Vaner [Fri, 21 Sep 2012 10:39:48 +0000 (12:39 +0200)]
Merge #1357
Mukund Sivaraman [Fri, 21 Sep 2012 09:41:33 +0000 (15:11 +0530)]
[2218] Combine blocks
Mukund Sivaraman [Fri, 21 Sep 2012 08:01:26 +0000 (13:31 +0530)]
[2218] Add recursive findNSEC3() tests
Mukund Sivaraman [Fri, 21 Sep 2012 07:33:00 +0000 (13:03 +0530)]
[2218] Remove duplicate copies of NSEC3 data
Mukund Sivaraman [Fri, 21 Sep 2012 07:10:48 +0000 (12:40 +0530)]
[2218] Add trace when walking the tree
Michal 'vorner' Vaner [Fri, 21 Sep 2012 06:57:52 +0000 (08:57 +0200)]
[1357] Test nothing is committed on error
Mukund Sivaraman [Fri, 21 Sep 2012 06:41:31 +0000 (12:11 +0530)]
[2218] Move NSEC3 hash calculation to NSEC3Hash class
Mukund Sivaraman [Fri, 21 Sep 2012 06:08:06 +0000 (11:38 +0530)]
[2218] Test findNSEC3() by walking the tree
Mukund Sivaraman [Fri, 21 Sep 2012 04:11:03 +0000 (09:41 +0530)]
[2218] Change DomainTreeNodeChain::getLevelCount()'s return type to size_t
... so that it's unsigned. Also rename node_count_ to level_count_
to avoid confusion with DomainTree::node_count_.
Mukund Sivaraman [Fri, 21 Sep 2012 04:09:25 +0000 (09:39 +0530)]
[2218] Change DomainTree::getNodeCount()'s return type to size_t
... so that it's unsigned.
Mukund Sivaraman [Fri, 21 Sep 2012 03:59:55 +0000 (09:29 +0530)]
[2218] Fix comment
Mukund Sivaraman [Fri, 21 Sep 2012 03:58:21 +0000 (09:28 +0530)]
[2218] Check that the NSEC3 tree has nodes before performing findNSEC3()
Mukund Sivaraman [Fri, 21 Sep 2012 03:45:16 +0000 (09:15 +0530)]
[2218] Assert that ZoneNodes in NSEC3 tree contain NSEC3 RRType
Mukund Sivaraman [Fri, 21 Sep 2012 03:32:21 +0000 (09:02 +0530)]
[2218] Don't upcase the first label when inserting into NSEC3 tree
This is no longer required as tree.find() works without it.
Mukund Sivaraman [Fri, 21 Sep 2012 03:22:44 +0000 (08:52 +0530)]
[2218] Simplify code, also shortcutting RRset construction
Mukund Sivaraman [Fri, 21 Sep 2012 03:17:58 +0000 (08:47 +0530)]
[2218] Rename set variable to rdataset
Mukund Sivaraman [Fri, 21 Sep 2012 03:15:50 +0000 (08:45 +0530)]
[2218] Rewrite Name generation to be more efficient
Mukund Sivaraman [Fri, 21 Sep 2012 03:12:42 +0000 (08:42 +0530)]
[2218] Move some statements out of loop
Mukund Sivaraman [Fri, 21 Sep 2012 03:09:51 +0000 (08:39 +0530)]
[2218] Move expression to simplify code
JINMEI Tatuya [Fri, 21 Sep 2012 01:27:32 +0000 (18:27 -0700)]
[2219] workaround for use-after-destroy cases in python client-list tests.
by making sure cache client/finder are destroyed at the end of each test.
we may want to improve the reliability, but for now that should be sufficient.
JINMEI Tatuya [Fri, 21 Sep 2012 00:43:07 +0000 (17:43 -0700)]
[2219] move memory leak check from in-memory client destructor to client list.
now the memory segment is shared for multiple in-memory zones, we cannot check
it in each client.
JINMEI Tatuya [Fri, 21 Sep 2012 00:33:24 +0000 (17:33 -0700)]
[2219] editorial: move DataSourceInfo::getCacheClient closer to other methods
JINMEI Tatuya [Fri, 21 Sep 2012 00:24:05 +0000 (17:24 -0700)]
[2219] workaround for letting auth test refer to in-memory data source client
this is a hack, and we should eventually make it more cleanly.
JINMEI Tatuya [Fri, 21 Sep 2012 00:04:17 +0000 (17:04 -0700)]
[2219] updated ZoneFinderContextTest so it will use the new in-mem datasrc.
JINMEI Tatuya [Fri, 21 Sep 2012 00:02:02 +0000 (17:02 -0700)]
[2219] re-added findAll() version of finder context constructor.
since we don't yet support in-memory optimization for addtional section
processing, we need to have the other version at the moment.
JINMEI Tatuya [Thu, 20 Sep 2012 23:30:13 +0000 (16:30 -0700)]
[2219] regression fix: support glue_ok option at the zone cut in the main find
in the old implementation we did it for the specialized finder context, but
we are currently using the generic context, so we need to handle the case
in the main method. (and, technically, this case should be handled in the
main method even if we had a specialized context as it's an API contract.)
JINMEI Tatuya [Thu, 20 Sep 2012 23:13:08 +0000 (16:13 -0700)]
[2219] style fix: add () after return
JINMEI Tatuya [Thu, 20 Sep 2012 23:08:28 +0000 (16:08 -0700)]
[2219] bug fix: store real object of rrclass in memory zone finder, not a ref.
having it as a reference can easily cause a dangling reference (and it actually
happens in some tests).
we could probably have it as a reference without causing disruption if we
carefully, but that doesn't seem to be reliable.
copying RRclass should be cheap, so it should be okay in this case.
JINMEI Tatuya [Thu, 20 Sep 2012 22:23:33 +0000 (15:23 -0700)]
[2219] a bug fix to zone finder: make RRSIG visible only when DNSSEC is sought.
JINMEI Tatuya [Thu, 20 Sep 2012 22:01:37 +0000 (15:01 -0700)]
[2219] added unit tests TreeNodeRRset::getRRsig().
not directly related to this branch, but it should have been done when the
method was implemented.
Also made small optimization: just return NULL if it's not supposed to be
signed.
JINMEI Tatuya [Thu, 20 Sep 2012 21:02:02 +0000 (14:02 -0700)]
[2219] updated client list test so it works with new in-memory implementation.
the points of the test should be preserved, but some internal behaviors
have to be adjusted.
JINMEI Tatuya [Thu, 20 Sep 2012 17:44:24 +0000 (10:44 -0700)]
[master] cleanup: removed dead code.
I suspect it was a leftover of an intermediate attempt that ended up with
found = RdataSet::find(node->getData(), type);
at commi
32f6a873 .
Whatever the intent was, it's clear the code does nothing, so cleaning it
up shouldn't harm anyway.
okayed on jabber.
JINMEI Tatuya [Thu, 20 Sep 2012 16:20:59 +0000 (09:20 -0700)]
[master] Merge branch 'trac2267'
JINMEI Tatuya [Thu, 20 Sep 2012 16:19:36 +0000 (09:19 -0700)]
[2267] cleanup: removed a redundant 'return'
Tomek Mrugalski [Thu, 20 Sep 2012 11:55:42 +0000 (13:55 +0200)]
[2269] Initial subnet6 definitions for DHCPv6
Marcin Siodelski [Thu, 20 Sep 2012 11:25:35 +0000 (13:25 +0200)]
[2272] Replaced malloc with new in command options helper.
Michal 'vorner' Vaner [Thu, 20 Sep 2012 11:06:59 +0000 (13:06 +0200)]
[1357] Create only one Diff
It failed to commit the previous changes, as it created new Diff and the
old was just forgotten.
Marcin Siodelski [Thu, 20 Sep 2012 10:02:55 +0000 (12:02 +0200)]
[2272] Fixed typo in HAVE_OPTRESET in configure.ac.
Marcin Siodelski [Thu, 20 Sep 2012 09:34:34 +0000 (11:34 +0200)]
[2272] Check for HAVE_OPTRESET in the perfdhcp command options.
Michal 'vorner' Vaner [Thu, 20 Sep 2012 09:27:57 +0000 (11:27 +0200)]
[1357] Commit only after verifying the last TSIG
We can't commit during the stream of changes, they might be not signed
and that'd be a security risk (if someone stole the TCP connection and
pushed a change that'd put bogus data inside before getting to the last
message which would be rejected then).
Fixed the check: it now happens on the last message, not on each message
except the last. Returning true there does not mean to continue, but
that the current RR has been handled fully.
Marcin Siodelski [Thu, 20 Sep 2012 09:19:51 +0000 (11:19 +0200)]
[2272] Check for optreset declaration in unistd.h.
Naoki Kambe [Thu, 20 Sep 2012 09:02:34 +0000 (18:02 +0900)]
[2222] removed unnecessary tests from TestXfroutSession
`None` is not set to a counter handler as a default. These tests are
no longer effective. This change is due to a merge of #2158.
Naoki Kambe [Tue, 11 Sep 2012 11:28:00 +0000 (20:28 +0900)]
[2222] used find() instead of `in` operator
- used find() instead of `in` operator for exact pattern matching of string
- a style fix in the list
Naoki Kambe [Mon, 27 Aug 2012 09:48:42 +0000 (18:48 +0900)]
[2222] added 'ixfr_running' and 'axfr_running' descriptions into the manpage of b10-xfrout, and untabfy
Naoki Kambe [Mon, 27 Aug 2012 09:47:11 +0000 (18:47 +0900)]
[2222] added checking for the number of xfr_running and axfr_running
- added checking for the number of xfr_running and axfr_running
- added changes related to the previous changes of bind10_control.py
(the step method name and the 'between A and B' style)
Naoki Kambe [Mon, 27 Aug 2012 09:57:21 +0000 (18:57 +0900)]
[2222] updated lettuce step methods for statisitcs
- revised statistics tests to use 'between A and B' style
- revised the name of the step method for checking the statistics value
- updated copyright
Naoki Kambe [Mon, 27 Aug 2012 09:40:37 +0000 (18:40 +0900)]
[2222] added "axfr_running" and "ixfr_running" definitions for statistics into xfrout.spec
Naoki Kambe [Mon, 27 Aug 2012 09:56:26 +0000 (18:56 +0900)]
[2222] added tests for counting the number of AXFR/IXFR running
- added tests which check properly counting the number of AXFR/IXFR
running
- added changes related to the changes of XfroutCounter
- added misc refactoring for testing
- used validate_statistics() for checking statistics data format to be
returned to the stats module
- updated copyright
Naoki Kambe [Mon, 27 Aug 2012 09:55:31 +0000 (18:55 +0900)]
[2222] updated classes: XfroutCounter, XfroutServer and XfroutSession
- added incrementer/decrementer for the number of IXFR/AXFR running into
the
XfroutCounter class
- added getter methods of counter handlers for the XfroutSession/NotifyOut
class into the XfroutCounter class
- inserted incrementer/decrementer handler for the number of AXFR/IXFR
running into XfroutSession
- removed the unnecessary thread lock from the get_statistics()
- removed a commented-out code
- updated copyright
Naoki Kambe [Thu, 20 Sep 2012 06:14:11 +0000 (15:14 +0900)]
[stats_lettuce_fix] added configuration of stats
Changed the default time (60sec) of poll-interval to one second. This is
because it's so infrequent for lettuce test of stats.
Naoki Kambe [Thu, 20 Sep 2012 06:05:28 +0000 (15:05 +0900)]
[stats_lettuce_fix] fix for the lettuce error: Not found statistics counter notifyoutv4
http://git.bind10.isc.org/~tester/builder//BIND10-lettuce/
20120918121701 -DebianLinux-x86_64-GCC/logs/lettuce.out
AssertionError: Not found statistics counter notifyoutv4 for zone example.org.
Added waiting until the last requesting is finished instead of one-second
sleep. It should be done before checking the result of "Stats show" command via
bindctl.
JINMEI Tatuya [Thu, 20 Sep 2012 05:55:50 +0000 (22:55 -0700)]
[2218] editorial fixes: folded long lines, constify
Naoki Kambe [Thu, 20 Sep 2012 05:54:37 +0000 (14:54 +0900)]
Revert "[stats_lettuce_fix] fix for the lettuce error: Not found statistics counter notifyoutv4"
This reverts commit
14727f8945873937f353915eac0e918cdb58721f .
Naoki Kambe [Thu, 20 Sep 2012 04:56:08 +0000 (13:56 +0900)]
[stats_lettuce_fix] fix for the lettuce error: AssertionError: Got 3, expected less than 3 as counter xfrrej
http://git.bind10.isc.org/~tester/builder/BIND10-lettuce/
20120918210000 -MacOS/logs/lettuce.out
AssertionError: Got 3, expected less than 3 as counter xfrrej for zone _SERVER_
Removed the upper limits of the counter xfrrej. Because a xfer request might be
rejected greater than 3 times. This number of times might depend on something
like the environment or the platform or the rasing condition, where this
lettuce test is running.
Naoki Kambe [Thu, 20 Sep 2012 04:34:15 +0000 (13:34 +0900)]
[stats_lettuce_fix] fix for the lettuce error: Not found statistics counter notifyoutv4
http://git.bind10.isc.org/~tester/builder//BIND10-lettuce/
20120918121701 -DebianLinux-x86_64-GCC/logs/lettuce.out
AssertionError: Not found statistics counter notifyoutv4 for zone example.org.
It has to wait at least for one second since the query to
b10-stats. b10-stats doesn't request to other module, e.g. b10-xfrout,
for less than one second, even if show command is invoked via
bindctl. This is for avoiding too frequent requests to b10-stats.
JINMEI Tatuya [Thu, 20 Sep 2012 00:36:35 +0000 (17:36 -0700)]
[2219] use the new version in-memory data source client in client_list.
it compiles, but some tests currently fail.
JINMEI Tatuya [Wed, 19 Sep 2012 21:38:12 +0000 (14:38 -0700)]
[2219] Merge branch 'trac2218' into trac2219
JINMEI Tatuya [Wed, 19 Sep 2012 21:37:50 +0000 (14:37 -0700)]
[2219] Merge branch 'trac2267' into trac2219
JINMEI Tatuya [Wed, 19 Sep 2012 17:20:32 +0000 (10:20 -0700)]
[master] be sure to wait for getstats command in xfrin notify lettuce test.
as proposed by naokikambe. confirmed by myself. okayed on jabber.
JINMEI Tatuya [Wed, 19 Sep 2012 16:47:13 +0000 (09:47 -0700)]
[2267] unify the call to addRdataSet.
as suggested in review.
JINMEI Tatuya [Wed, 19 Sep 2012 16:38:05 +0000 (09:38 -0700)]
[2267] comment updates
Tomek Mrugalski [Wed, 19 Sep 2012 13:57:17 +0000 (15:57 +0200)]
[2238] Circular dependency (asiolink->log->util->asiolink) fixed.
Mukund Sivaraman [Wed, 19 Sep 2012 10:27:22 +0000 (15:57 +0530)]
[master] Update various .gitignore files
Mukund Sivaraman [Wed, 19 Sep 2012 10:23:12 +0000 (15:53 +0530)]
[master] Track origin of uninitalized memory on all Valgrind targets
Mukund Sivaraman [Wed, 19 Sep 2012 10:19:42 +0000 (15:49 +0530)]
[master] Move GTEST_LDADD to the bottom of the list
Mukund Sivaraman [Wed, 19 Sep 2012 10:07:16 +0000 (15:37 +0530)]
[master] Cleanup a Makefile.am (which didn't link to libb10-dhcp++)
Also move GTEST_LDADD to the bottom of the list of LDADDs.