release builds done for cobalt and evbmips64-eb
Provide an __eprol symbol in TEXT_START_SYMBOLS. MIPS overrides
TEXT_START_SYMBOLS so it misses out on the default definition of
this that does provide an __eprol symbol. Fixes linking of profiled
binaries.
move the rearming of the cq after the refill of the rq.
this is the only real diff we have left outstanding on a box that
experienced rx lockups. since adding this change it's been happy
for the last 4 weeks and counting so far.
The previous code accepted '# 123 "file.c" 23' as specifying a system
header, just because that number ends with '3'. The original intention
was to compare the complete word, not its suffix. Fix that.
No practical change since the only flags that are used by GCC are all
single-digit.
This also changes the conditions to their positive form, which is easier
to read.
No functional change. The resulting binary would have been the same as
before, were it not for the changed line numbers in the lint_assert
calls further down in the code.
The intention of the getopt check was to analyze only those while loops
whose condition consists of the usual getopt call. For all other while
loops, ck.while_level was intended to stay at 0.
This was not the case in ckgetopt.c 1.2 and has been fixed in ckgetopt.c
1.3. The code did not document the intended invariants though, which it
should have done. This will be done in a follow-up commit.
*) Fixed the X509_issuer_and_serial_hash() function. It attempts
to create a unique hash value based on the issuer and serial
number data contained within an X509 certificate. However it
was failing to correctly handle any errors that may occur
while parsing the issuer field (which might occur if the issuer
field is maliciously constructed). This may subsequently result
in a NULL pointer deref and a crash leading to a potential
denial of service attack.
(CVE-2021-23841)
[Matt Caswell]
*) Fixed the RSA_padding_check_SSLv23() function and the
RSA_SSLV23_PADDING padding mode to correctly check for rollback
attacks. This is considered a bug in OpenSSL 1.1.1 because it
does not support SSLv2. In 1.0.2 this is CVE-2021-23839.
[Matt Caswell]
*) Fixed the EVP_CipherUpdate, EVP_EncryptUpdate and EVP_DecryptUpdate
functions. Previously they could overflow the output length
argument in some cases where the input length is close to the
maximum permissable length for an integer on the platform. In
such cases the return value from the function call would be
1 (indicating success), but the output length value would be
negative. This could cause applications to behave incorrectly
or crash.
(CVE-2021-23840)
[Matt Caswell]
*) Fixed SRP_Calc_client_key so that it runs in constant time.
The previous implementation called BN_mod_exp without setting
BN_FLG_CONSTTIME. This could be exploited in a side channel
attack to recover the password. Since the attack is local host
only this is outside of the current OpenSSL threat model and
therefore no CVE is assigned.
Thanks to Mohammed Sabt and Daniel De Almeida Braga for reporting
this issue.
[Matt Caswell]
Internally the code confuses the concept of "the user doesn't want
a backup file" and "the user hasn't defined a type of backup file".
Introduce a new "undefined" backup type to serve the purpose "none"
previously did, and make "none" not generate backup files, as expected.
http://mail-index.netbsd.org/tech-userlevel/2021/02/19/msg012901.html
XXX pullup?
5578. [protocol] Make "check-names" accept A records below "_spf",
"_spf_rate", and "_spf_verify" labels in order to cater
for the "exists" SPF mechanism specified in RFC 7208
section 5.7 and appendix D.1. [GL #2377]
5577. [bug] Fix the "three is a crowd" key rollover bug in KASP by
correctly implementing Equation (2) of the "Flexible and
Robust Key Rollover" paper. [GL #2375]
5575. [bug] When migrating to KASP, BIND 9 considered keys with the
"Inactive" and/or "Delete" timing metadata to be
possible active keys. This has been fixed. [GL #2406]
5572. [bug] Address potential double free in generatexml().
[GL #2420]
5571. [bug] named failed to start when its configuration included a
zone with a non-builtin "allow-update" ACL attached.
[GL #2413]
5570. [bug] Improve performance of the DNSSEC verification code by
reducing the number of repeated calls to
dns_dnssec_keyfromrdata(). [GL #2073]
5569. [bug] Emit useful error message when "rndc retransfer" is
applied to a zone of inappropriate type. [GL #2342]
5568. [bug] Fixed a crash in "dnssec-keyfromlabel" when using ECDSA
keys. [GL #2178]
5567. [bug] Dig now reports unknown dash options while pre-parsing
the options. This prevents "-multi" instead of "+multi"
from reporting memory usage before ending option parsing
with "Invalid option: -lti". [GL #2403]
5566. [func] Add "stale-answer-client-timeout" option, which is the
amount of time a recursive resolver waits before
attempting to answer the query using stale data from
cache. [GL #2247]
5565. [func] The SONAMEs for BIND 9 libraries now include the current
BIND 9 version number, in an effort to tightly couple
internal libraries with a specific release. [GL #2387]
5562. [security] Fix off-by-one bug in ISC SPNEGO implementation.
(CVE-2020-8625) [GL #2354]
5561. [bug] KASP incorrectly set signature validity to the value of
the DNSKEY signature validity. This is now fixed.
[GL #2383]
5560. [func] The default value of "max-stale-ttl" has been changed
from 12 hours to 1 day and the default value of
"stale-answer-ttl" has been changed from 1 second to 30
seconds, following RFC 8767 recommendations. [GL #2248]
5456. [func] Added "primaries" as a synonym for "masters" in
named.conf, and "primary-only" as a synonym for
"master-only" in the parameters to "notify", to bring
terminology up-to-date with RFC 8499. [GL #1948]
5362. [func] Limit the size of IXFR responses so that AXFR will
be used instead if it would be smaller. This is
controlled by the "max-ixfr-ratio" option, which
is a percentage representing the ratio of IXFR size
to the size of the entire zone. This value cannot
exceed 100%, which is the default. [GL #1515]
--- 9.16.11 released ---
5559. [bug] The --with-maxminddb=PATH form of the build-time option
enabling support for libmaxminddb was not working
correctly. This has been fixed. [GL #2366]
5557. [bug] Prevent RBTDB instances from being destroyed by multiple
threads at the same time. [GL #2317]
5556. [bug] Further tweak newline printing in dnssec-signzone and
dnssec-verify. [GL #2359]
5555. [bug] server->reload_status was not properly initialized.
[GL #2361]
5554. [bug] dnssec-signzone and dnssec-verify were missing newlines
between log messages. [GL #2359]
5553. [bug] When reconfiguring named, removing "auto-dnssec" did not
turn off DNSSEC maintenance. [GL #2341]
5552. [func] When switching to "dnssec-policy none;", named now
permits a safe transition to insecure mode and publishes
the CDS and CDNSKEY DELETE records, as described in RFC
8078. [GL #1750]
5551. [bug] named no longer attempts to assign threads to CPUs
outside the CPU affinity set. Thanks to Ole Bjørn
Hessen. [GL #2245]
5550. [func] dnssec-signzone and named now log a warning when falling
back to the "increment" SOA serial method. [GL #2058]
5545. [func] OS support for load-balanced sockets is no longer
required to receive incoming queries in multiple netmgr
threads. [GL #2137]
5543. [bug] Fix UDP performance issues caused by making netmgr
callbacks asynchronous-only. [GL #2320]
5542. [bug] Refactor netmgr. [GL #1920] [GL #2034] [GL #2061]
[GL #2194] [GL #2221] [GL #2266] [GL #2283] [GL #2318]
[GL #2321]
--- 9.16.10 released ---
5544. [func] Restore the default value of "nocookie-udp-size" to 4096
bytes. [GL #2250]
5541. [func] Adjust the "max-recursion-queries" default from 75 to
100. [GL #2305]
5540. [port] Fix building with native PKCS#11 support for AEP Keyper.
[GL #2315]
5539. [bug] Tighten handling of missing DNS COOKIE responses over
UDP by falling back to TCP. [GL #2275]
5538. [func] Add NSEC3 support to KASP. A new option for
"dnssec-policy", "nsec3param", can be used to set the
desired NSEC3 parameters. NSEC3 salt collisions are
automatically prevented during resalting. Salt
generation is now logged with zone context. [GL #1620]
5534. [bug] The CNAME synthesized from a DNAME was incorrectly
followed when the QTYPE was CNAME or ANY. [GL #2280]
--- 9.16.9 released ---
5533. [func] Add the "stale-refresh-time" option, a time window that
starts after a failed lookup, during which a stale RRset
is served directly from cache before a new attempt to
refresh it is made. [GL #2066]
5530. [bug] dnstap did not capture responses to forwarded UPDATE
requests. [GL #2252]
5527. [bug] A NULL pointer dereference occurred when creating an NTA
recheck query failed. [GL #2244]
5525. [bug] Change 5503 inadvertently broke cross-compilation by
replacing a call to AC_LINK_IFELSE() with a call to
AC_RUN_IFELSE() in configure.ac. This has been fixed,
making cross-compilation possible again. [GL #2237]
5523. [bug] The initial lookup in a zone transitioning to/from a
signed state could fail if the DNSKEY RRset was not
found. [GL #2236]
5522. [bug] Fixed a race/NULL dereference in TCPDNS send. [GL #2227]
5520. [bug] Fixed a number of shutdown races, reference counting
errors, and spurious log messages that could occur
in the network manager. [GL #2221]
5518. [bug] Stub zones now work correctly with primary servers using
"minimal-responses yes". [GL #1736]
5517. [bug] Do not treat UV_EOF as a TCP4RecvErr or a TCP6RecvErr.
[GL #2208]
--- 9.16.8 released ---
5516. [func] The default EDNS buffer size has been changed from 4096
to 1232 bytes. [GL #2183]
5515. [func] Add 'rndc dnssec -rollover' command to trigger a manual
rollover for a specific key. [GL #1749]
5514. [bug] Fix KASP expected key size for Ed25519 and Ed448.
[GL #2171]
5513. [doc] The ARM section describing the "rrset-order" statement
was rewritten to make it unambiguous and up-to-date with
the source code. [GL #2139]
5512. [bug] "rrset-order" rules using "order none" were causing
named to crash despite named-checkconf treating them as
valid. [GL #2139]
5511. [bug] 'dig -u +yaml' failed to display timestamps to the
microsecond. [GL #2190]
5510. [bug] Implement the attach/detach semantics for dns_message_t
to fix a data race in accessing an already-destroyed
fctx->rmessage. [GL #2124]
5509. [bug] filter-aaaa: named crashed upon shutdown if it was in
the process of recursing for A RRsets. [GL #1040]
5508. [func] Added new parameter "-expired" for "rndc dumpdb" that
also prints expired RRsets (awaiting cleanup) to the
dump file. [GL #1870]
5507. [bug] Named could compute incorrect SIG(0) responses.
[GL #2109]
5506. [bug] Properly handle failed sysconf() calls, so we don't
report invalid memory size. [GL #2166]
5505. [bug] Updating contents of a mixed-case RPZ could cause some
rules to be ignored. [GL #2169]
5503. [bug] Cleaned up reference counting of network manager
handles, now using isc_nmhandle_attach() and _detach()
instead of _ref() and _unref(). [GL #2122]
--- 9.16.7 released ---
5501. [func] Log CDS/CDNSKEY publication. [GL #1748]
5500. [bug] Fix (non-)publication of CDS and CDNSKEY records.
[GL #2103]
5499. [func] Add '-P ds' and '-D ds' arguments to dnssec-settime.
[GL #1748]
5497. [bug] 'dig +bufsize=0' failed to disable EDNS. [GL #2054]
5496. [bug] Address a TSAN report by ensuring each rate limiter
object holds a reference to its task. [GL #2081]
5495. [bug] With query minimization enabled, named failed to
resolve ip6.arpa. names that had extra labels to the
left of the IPv6 part. [GL #1847]
5494. [bug] Silence the EPROTO syslog message on older systems.
[GL #1928]
5493. [bug] Fix off-by-one error when calculating new hash table
size. [GL #2104]
5492. [bug] Tighten LOC parsing to reject a period (".") and/or "m"
as a value. Fix handling of negative altitudes which are
not whole meters. [GL #2074]
5491. [bug] rbtversion->glue_table_size could be read without the
appropriate lock being held. [GL #2080]
5489. [bug] Named erroneously accepted certain invalid resource
records that were incorrectly processed after
subsequently being written to disk and loaded back, as
the wire format differed. Such records include: CERT,
IPSECKEY, NSEC3, NSEC3PARAM, NXT, SIG, TLSA, WKS, and
X25. [GL !3953]
5488. [bug] NTA code needed to have a weak reference on its
associated view to prevent the latter from being deleted
while NTA tests were being performed. [GL #2067]
5486. [func] Add 'rndc dnssec -checkds' command, which signals to
named that the DS record for a given zone or key has
been updated in the parent zone. [GL #1613]
--- 9.16.6 released ---
5484. [func] Expire zero TTL records quickly rather than using them
for stale answers. [GL #1829]
5483. [func] A new configuration option "stale-cache-enable" has been
introduced to enable or disable keeping stale answers in
cache. [GL #1712]
5482. [bug] If the Duplicate Address Detection (DAD) mechanism had
not yet finished after adding a new IPv6 address to the
system, BIND 9 would fail to bind to IPv6 addresses in a
tentative state. [GL #2038]
5481. [security] "update-policy" rules of type "subdomain" were
incorrectly treated as "zonesub" rules, which allowed
keys used in "subdomain" rules to update names outside
of the specified subdomains. The problem was fixed by
making sure "subdomain" rules are again processed as
described in the ARM. (CVE-2020-8624) [GL #2055]
5480. [security] When BIND 9 was compiled with native PKCS#11 support, it
was possible to trigger an assertion failure in code
determining the number of bits in the PKCS#11 RSA public
key with a specially crafted packet. (CVE-2020-8623)
[GL #2037]
5479. [security] named could crash in certain query resolution scenarios
where QNAME minimization and forwarding were both
enabled. (CVE-2020-8621) [GL #1997]
5478. [security] It was possible to trigger an assertion failure by
sending a specially crafted large TCP DNS message.
(CVE-2020-8620) [GL #1996]
5477. [bug] The idle timeout for connected TCP sockets, which was
previously set to a high fixed value, is now derived
from the client query processing timeout configured for
a resolver. [GL #2024]
5476. [security] It was possible to trigger an assertion failure when
verifying the response to a TSIG-signed request.
(CVE-2020-8622) [GL #2028]
5475. [bug] Wildcard RPZ passthru rules could incorrectly be
overridden by other rules that were loaded from RPZ
zones which appeared later in the "response-policy"
statement. This has been fixed. [GL #1619]
5474. [bug] dns_rdata_hip_next() failed to return ISC_R_NOMORE
when it should have. [GL !3880]
5473. [func] The RBT hash table implementation has been changed
to use a faster hash function (HalfSipHash2-4) and
Fibonacci hashing for better distribution. Setting
"max-cache-size" now preallocates a fixed-size hash
table so that rehashing does not cause resolution
brownouts while the hash table is grown. [GL #1775]
5471. [bug] The introduction of KASP support inadvertently caused
the second field of "sig-validity-interval" to always be
calculated in hours, even in cases when it should have
been calculated in days. This has been fixed. (Thanks to
Tony Finch.) [GL !3735]
5469. [port] On illumos, a constant called SEC is already defined in
<sys/time.h>, which conflicts with an identically named
constant in libbind9. This conflict has been resolved.
[GL #1993]
5468. [bug] Addressed potential double unlock in process_fd().
[GL #2005]
5466. [bug] Addressed an error in recursive clients stats reporting.
[GL #1719]
5465. [func] Added fallback to built-in trust-anchors, managed-keys,
or trusted-keys if the bindkeys-file (bind.keys) cannot
be parsed. [GL #1235]
5464. [bug] Requesting more than 128 files to be saved when rolling
dnstap log files caused a buffer overflow. This has been
fixed. [GL #1989]
5462. [bug] Move LMDB locking from LMDB itself to named. [GL #1976]
5461. [bug] The STALE rdataset header attribute was updated while
the write lock was not being held, leading to incorrect
statistics. The header attributes are now converted to
use atomic operations. [GL #1475]
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
