bind-announce July 2011 archive
Main Archive Page > Month Archives  > bind-announce archives
bind-announce: BIND 9.8.1b2

BIND 9.8.1b2

From: Mark Andrews <marka_at_nospam>
Date: Fri Jul 08 2011 - 01:41:58 GMT
To: bind-announce@isc.org

Introduction

   BIND 9.8.1b2 is the second beta release of BIND 9.8.

   This document summarizes changes from BIND 9.8.0 to BIND 9.8.1b2.
   Please see the CHANGES file in the source code release for a complete
   list of all changes.

Download

   The latest versions of BIND 9 software can always be found on our web
   site at http://www.isc.org/downloads/all. There you will find
   additional information about each release, source code, and some
   pre-compiled versions for certain operating systems.

Support

   Product support information is available on
   http://www.isc.org/services/support for paid support options. Free
   support is provided by our user community via a mailing list.
   Information on all public email lists is available at
   https://lists.isc.org/mailman/listinfo.

New Features

9.8.1b2

     * Added a new include file with function typedefs for the DLZ
       "dlopen" driver. [RT #23629]
     * Added a tool able to generate malformed packets to allow testing of
       how named handles them. [RT #24096]

Security Fixes

9.8.1b2

     * If named is configured with a response policy zone (RPZ) and a
       query of type RRSIG is received for a name configured for RRset
       replacement in that RPZ, it will trigger an INSIST and crash the
       server. RRSIG. [RT #24280]
     * named, set up to be a caching resolver, is vulnerable to a user
       querying a domain with very large resource record sets (RRSets)
       when trying to negatively cache the response. Due to an off-by-one
       error, caching the response could cause named to crash. [RT #24650]
       [CVE-2011-1910]
     * Using Response Policy Zone (RPZ) to query a wildcard CNAME label
       with QUERY type SIG/RRSIG, it can cause named to crash. Fix is
       query type independant. [RT #24715]
     * Using Response Policy Zone (RPZ) with DNAME records and querying
       the subdomain of that label can cause named to crash. Now logs that
       DNAME is not supported. [RT #24766]
     * Change #2912 populated the message section in replies to UPDATE
       requests, which some Windows clients wanted. This exposed a latent
       bug that allowed the response message to crash named. With this
       fix, change 2912 has been reduced to copy only the zone section to
       the reply. A more complete fix for the latent bug will be released
       later. [RT #24777]

Feature Changes

9.8.1b2

     * Merged in the NetBSD ATF test framework (currently version 0.12)
       for development of future unit tests. Use configure --with-atf to
       build ATF internally or configure --with-atf=prefix to use an
       external copy. [RT #23209]
     * Added more verbose error reporting from DLZ LDAP. [RT #23402]
     * The DLZ "dlopen" driver is now built by default, no longer
       requiring a configure option. To disable it, use "configure
       --without-dlopen". (Note: driver not supported on win32.) [RT
       #23467]
     * Replaced compile time constant with STDTIME_ON_32BITS. [RT #23587]
     * Make --with-gssapi default for ./configure. [RT #23738]

Bug Fixes

9.8.1b2

     * During RFC5011 processing some journal write errors were not
       detected. This could lead to managed-keys changes being committed
       but not recorded in the journal files, causing potential
       inconsistencies during later processing. [RT #20256]
       A potential NULL pointer deference in the DNS64 code could cause
       named to terminate unexpectedly. [RT #20256]
       A state variable relating to DNSSEC could fail to be set during
       some infrequently-executed code paths, allowing it to be used
       whilst in an unitialized state during cache updates, with
       unpredictable results. [RT #20256]
       A potential NULL pointer deference in DNSSEC signing code could
       cause named to terminate unexpectedly [RT #20256]
       Several cosmetic code changes were made to silence warnings
       generated by a static code analysis tool. [RT #20256]
     * When using the -x (sign with only KSK) option on dnssec-signzone,
       it could incorrectly count the number of ZSKs in the zone. (And in
       9.9.0, some code cleanup and improved warning messages). [RT
       #20852]
     * When using _builtin in named.conf, named.conf changes were not
       found when reloading the config file. Now checks _builtin zone
       arguments to see if the zone is re-usable or not. [RT #21914]
     * Running dnssec-settime -f on an old-style key will now force the
       key to be rewritten to the new key format even if no other change
       has been specified, using "-P now -A now" as default values. [RT
       #22474]
     * After an external code review, a code cleanup was done. [RT #22521]
     * Cause named to terminate at startup or rndc reconfig reload to
       fail, if a log file specified in the conf file isn't a plain file.
       (RT #22771]
     * named now forces the ADB cache time for glue related data to zero
       instead of relying on TTL. This corrects problematic behavior in
       cases where a server was authoritative for the A record of a
       nameserver for a delegated zone and was queried to recursively
       resolve records within that zone. [RT #22842]
     * When a validating resolver got a NODATA response for DNSKEY, it was
       not caching the NODATA. Fixed and test added. [RT #22908]
     * Fixed a bug in which zone keys that were published and but not
       immediately activated, automatic signing could fail to trigger. [RT
       #22911]
     * Fixed precedence order bug with NS and DNAME records if both are
       present. (Also fixed timing of autosign test in 9.7+) [RT #23035]
     * When a DNSSEC signed dynamic zone's signatures need to be
       refreshed, named would first delete the old signatures in the zone.
       If a private key of the same algorithm isn't available to named,
       the signing would fail but the old signatures would already be
       deleted. named now checks if it can access the private key before
       deleting the old signatures and leaves the old signature if no
       private key is found. [RT #23136]
     * When using "auto-dnssec maintain" and rolling to a new key, a
       private-type record (only used internally by named) could be
       created and not marked as complete. [RT #23253]
     * Fixed last autosign test report. [RT #23256]
     * named didn't save gid at startup and later assumed gid 0. named now
       saves/restores the gid when creating creating named.pid at startup.
       [RT #23290]
     * If the server has an IPv6 address but does not have IPv6
       connectivity to the internet, dig +trace could fail attempting to
       use IPv6 addresses. [RT #23297]
     * If named is configured with managed zones, the managed key maint
       timer can exercise a race condition that can crash the server. [RT
       #23303]
     * Changing TTL did not cause dnssec-signzone to generate new
       signatures. [RT #23330]
     * Have the validating resolver use RRSIG original TTL to compute
       validated RRset and RRSIG TTL. [RT #23332]
     * In "make test" bin/tests/resolver, hold the socket manager lock
       while freeing the socket. [RT #23333]
     * If named encountered a CNAME instead of a DS record when walking
       the chain of trust down from the trust anchor, it incorrectly
       stopped validating. [RT #23338]
     * dns/view.h needed dns/rpz.h but it wasn't in the Makfile.in HEADERS
       variable. [RT #23342]
     * RRSIG records could have time stamps too far in the future. [RT
       #23356]
     * named stores cached data in an in-memory database and keeps track
       of how recently the data is used with a heap. The heap is stored
       within the cache's memory space. Under a sustained high query load
       and with a small cache size, this could lead to the heap exhausting
       the cache space. This would result in cache misses and SERVFAILs,
       with named never releasing the cache memory the heap used up and
       never recovering. This fix removes the heap into its own memory
       space, preventing the heap from exhausting the cache space and
       allowing named to recover gracefully when the high query load
       abates. [RT #23371]
     * Fully separated key management on a per view basis. [RT #23419]
     * If running on a powerpc CPU and with atomic operations enabled,
       named could lock up. Added sync instructions to the end of atomic
       operations. [RT #23469]
     * If OpenSSL was built without engine support, named would have
       compile errors and fail to build. [RT #23473]
     * If ./configure finds GOST but not elliptic curve, named fails to
       build. Added elliptic curve support check in GOST OpenSSL engine
       detection. [RT #23485]
     * "rndc secroots" would abort on the first error and so could miss
       remaining views. [RT #23488]
     * Handle isc_event_allocate failures in t_tasks test. [RT #23572]
     * ixfr-from-differences {master|slave}; failed to select the
       master/slave zones, resulting in on diff/journal file being
       created. [RT #23580]
     * If a DNAME substitution failed, named returned NOERROR. The correct
       response should be YXDOMAIN. [RT #23591]
     * dns_dnssec_findzonekeys{2} used a inconsistant timestamp when
       determining which keys are active. This could result in some RRsets
       not being signed/re-signed. [RT #23642]
     * Remove bin/tests/system/logfileconfig/ns1/named.conf and add
       setup.sh in order to resolve changing named.conf issue. [RT #23687]
     * NOTIFY messages were not being sent when generating a NSEC3 chain
       incrementally. [RT #23702]
     * DDNS updates using SIG(0) with update-policy match type "external"
       could cause a crash. Also fixed nsupdate core dump on shutdown when
       using a SIG(0) key, due to the key not being freed. [RT #23735]
     * Zones using automatic key maintenance could fail to check the key
       repository for updates. named now checks once per hour and the
       automatic check bug has been fixed. [RT #23744]
     * named now uses the correct strtok/strtok_r/strtok_s based on OS.
       [RT #23747]
     * Signatures for records at the zone apex could go stale due to an
       incorrect timer setting. [RT #23769]
     * The autosign tests attempted to open ports within reserved ranges.
       Test now avoids those ports. [RT #23957]
     * GSS TGIS test was failing, since log_cred() caused KRB5_KTNAME to
       be cached. Now sets KRB5_KTNAME before calling log_cred() in
       dst_gssapi_acceptctx(). [RT #24004]
     * named, acting as authoritative server for DLZ zones, was not
       correctly setting the authoritative (AA) bit. [RT #24146]
     * Clean up some cross-compiling issues and added two undocumented
       configure options, --with-gost and --with-rlimtype, to allow
       over-riding default settings (gost=no and rlimtype="long int") when
       cross-compiling. [RT #24367]
     * When trying sign with NSEC3, if dnssec-signzone couldn't find the
       KSK, it would give an incorrect error "NSEC3 iterations too big for
       weakest DNSKEY strength" rather than the correct "failed to find
       keys at the zone apex: not found" [RT #24369]
     * Configuring 'dnssec-validation auto' in a view instead of in the
       options statement could trigger an assertion failure in
       named-checkconf. [RT #24382]
     * Improved consistency checks for dnssec-enable and
       dnssec-validation, added test cases to the checkconf system test.
       [RT #24398]
     * If named is configured to be both authoritative and recursive and
       receives a recursive query for a CNAME in a zone that it is
       authoritative for, if that CNAME also points to a zone the server
       is authoritative for, the recursive part of name will not follow
       the CNAME change and the response will not be a complete CNAME
       chain. [RT #24455]
     * nsupdate could dump core on shutdown when using SIG(0) keys. [RT
       #24604]
     * Named could fail to validate zones list in a DLV that validated
       insecure without using DLV and had DS records in the parent zone.
       [RT #24631]

Known issues in this release

     * None.

Thank You

   Thank you to everyone who assisted us in making this release possible.
   If you would like to contribute to ISC to assist us in continuing to
   make quality open source software, please visit our donations page at
   http://www.isc.org/supportisc.
-- Mark Andrews, ISC 1 Seymour St., Dundas Valley, NSW 2117, Australia PHONE: +61 2 9871 4742 INTERNET: marka@isc.org _______________________________________________ bind-announce mailing list bind-announce@lists.isc.org https://lists.isc.org/mailman/listinfo/bind-announce