When cache memory usage is over the configured cache size (overmem) and
we are cleaning unused entries, it might not be enough to clean just two
entries if the entries to be expired are smaller than the newly added
rdata. This could be abused by an attacker to cause a remote Denial of
Service by possibly running out of the operating system memory.
Currently, the addrdataset() tries to do a single TTL-based cleaning
considering the serve-stale TTL and then optionally moves to overmem
cleaning if we are in that condition. Then the overmem_purge() tries to
do another single TTL based cleaning from the TTL heap and then continue
with LRU-based cleaning up to 2 entries cleaned.
Squash the TTL-cleaning mechanism into single call from addrdataset(),
but ignore the serve-stale TTL if we are currently overmem.
Then instead of having a fixed number of entries to clean, pass the size
of newly added rdatasetheader to the overmem_purge() function and
cleanup at least the size of the newly added data. This prevents the
cache going over the configured memory limit (`max-cache-size`).
Additionally, refactor the overmem_purge() function to reduce for-loop
nesting for readability.
Add a test ensuring that the amount of work fctx_getaddresses() performs
for any encountered delegation is limited: delegate example.net to a set
of 1,000 name servers in the redirect.com zone, the names of which all
resolve to IP addresses that nothing listens on, and query for a name in
the example.net domain, checking the number of times the findname()
function gets executed in the process; fail if that count is excessively
large.
Since the size of the referral response sent by ans3 is about 20 kB, it
cannot be sent back over UDP (EMSGSIZE) on some operating systems in
their default configuration (e.g. FreeBSD - see the
net.inet.udp.maxdgram sysctl). To enable reliable reproduction of
CVE-2022-2795 (retry patterns vary across BIND 9 versions) and avoid
false positives at the same time (thread scheduling - and therefore the
number of fetch context restarts - vary across operating systems and
across test runs), extend bin/tests/system/resolver/ans3/ans.pl so that
it also listens on TCP and make "ns1" in the "resolver" system test
always use TCP when communicating with "ans3".
Also add a test (foo.bar.sub.tld1/TXT) that ensures the new limitations
imposed on the resolution process by the mitigation for CVE-2022-2795 do
not prevent valid, glueless delegation chains from working properly.
Limit the amount of database lookups that can be triggered in
fctx_getaddresses() (i.e. when determining the name server addresses to
query next) by setting a hard limit on the number of NS RRs processed
for any delegation encountered. Without any limit in place, named can
be forced to perform large amounts of database lookups per each query
received, which severely impacts resolver performance.
The limit used (20) is an arbitrary value that is considered to be big
enough for any sane DNS delegation.
Compound literals are not used in BIND 9.11, in order to ensure backward
compatibility with ancient compilers. Rework the relevant parts of the
BIND 9.11 backport of the CVE-2021-25220 fix so that compound literals
are not used.
When caching glue, we need to ensure that there is no closer
source of truth for the name. If the owner name for the glue
record would be answered by a locally configured zone, do not
cache.
When caching additional and glue data *not* from a forwarder, we must
check that there is no "forward only" clause covering the owner name
that would take precedence. Such names would normally be allowed by
baliwick rules, but a "forward only" zone introduces a new baliwick
scope.
If we are using a fowarder, in addition to checking that names to
be cached are subdomains of the forwarded namespace, we must also
check that there are no subsidiary forwarded namespaces which would
take precedence. To be safe, we don't cache any responses if the
forwarding configuration has changed since the query was sent.
Previously, when lame cache would be disabled by setting lame-ttl to 0,
it would also disable lame answer detection. In this commit, we enable
the lame response detection even when the lame cache is disabled. This
enables stopping answer processing early rather than going through the
whole answer processing flow.
GCC 11 produced the following warning:
iterated_hash.c:21:33: warning: argument 1 of type ‘unsigned char[20]’ with mismatched bound [-Warray-parameter=]
21 | isc_iterated_hash(unsigned char out[ISC_SHA1_DIGESTLENGTH],
| ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from iterated_hash.c:18:
./include/isc/iterated_hash.h:33:37: note: previously declared as ‘unsigned char[155]’
33 | int isc_iterated_hash(unsigned char out[NSEC3_MAX_HASH_LENGTH],
| ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
Add REQUIRE(type == dns_rdatatype_nsec3 || firstp != NULL); so
that dereferences of *firstp is not flagged as a NULL pointer
dereference.
(cherry picked from commit 683ad6e4bd)
Add missing DNS results from 9.16. We only care about
DNS_R_NSEC3ITERRANGE but better to sync and use the same numbers than
use a different number for this result code than what is used in 9.16.
* The location of the digest type field has changed to where the
reserved field was.
* The reserved field is now called scheme and is where the digest
type field was.
* Digest type 2 has been defined (SHA256).
(cherry picked from commit 8510ccaa54)
If a zone transfer results in a zone not having any NS records, named
stops serving it because such a zone is broken. Do the same if an
incoming zone transfer results in a zone lacking an SOA record at the
apex or containing more than one SOA record.
An IXFR containing SOA records with owner names different than the
transferred zone's origin can result in named serving a version of that
zone without an SOA record at the apex. This causes a RUNTIME_CHECK
assertion failure the next time such a zone is refreshed. Fix by
immediately rejecting a zone transfer (either an incremental or
non-incremental one) upon detecting an SOA record not placed at the apex
of the transferred zone.
BIND 9 attempts to look up GSSAPI OIDs for the Kerberos 5 and SPNEGO
mechanisms in the relevant header files provided by the Kerberos/GSSAPI
library used. Due to the differences between various Kerberos/GSSAPI
implementations, if any of the expected preprocessor macros
(GSS_KRB5_MECHANISM, GSS_SPNEGO_MECHANISM) is not defined in the header
files provided by the library used, the code in lib/dns/gssapictx.c
defines its own version of each missing macro, so that BIND 9 can
attempt to use the relevant security mechanisms anyway.
Commit 7f08c756aa, which contains a
partial backport of the changes introduced in commit
978c7b2e89, left a block of code in the
lib/dns/include/dst/gssapi.h header which defines the
GSS_SPNEGO_MECHANISM preprocessor macro to NULL if it is not defined by
any header file provided by the Kerberos/GSSAPI library used. This
causes the gss_add_oid_set_member() call in the mech_oid_set_create()
helper function to always return an error. This in turn causes the
dst_gssapi_acquirecred() function to also always return an error, which
ultimately prevents any named instance whose configuration includes the
"tkey-gssapi-credential" option from starting.
Remove the offending conditional definition of the GSS_SPNEGO_MECHANISM
preprocessor macro from lib/dns/include/dst/gssapi.h, so that a proper
GSSAPI OID is assigned to that macro in lib/dns/gssapictx.c when the
Kerberos/GSSAPI library used does not define it.
(cherry picked from commit 648ef3a2b4)
add matching macros to pass arguments from called methods
to generic methods. This will reduce the amount of work
required when extending methods.
Also cleanup unnecessary UNUSED declarations.
(cherry picked from commit a88d3963e2)
Even if a call to gss_accept_sec_context() fails, it might still cause a
GSS-API response token to be allocated and left for the caller to
release. Make sure the token is released before an early return from
dst_gssapi_acceptctx().
(cherry picked from commit d954e152d9)
Fix race between zone_maintenance and dns_zone_notifyreceive functions,
zone_maintenance was attempting to read a zone flag calling
DNS_ZONE_FLAG(zone, flag) while dns_zone_notifyreceive was updating
a flag in the same zone calling DNS_ZONE_SETFLAG(zone, ...).
The code reading the flag in zone_maintenance was not protected by the
zone's lock, to avoid a race the zone's lock is now being acquired
before an attempt to read the zone flag is made.
Previously, every function had it's own #ifdef GSSAPI #else #endif block
that defined shim function in case GSSAPI was not being used. Now the
dummy shim functions have be split out into a single #else #endif block
at the end of the file.
This makes the gssapictx.c similar to 9.17.x code, making the backports
and reviews easier.
The Heimdal Kerberos library handles the OID sets in a different manner.
Unify the handling of the OID sets between MIT and Heimdal
implementations by dynamically creating the OID sets instead of using
static predefined set. This is how upstream recommends to handle the
OID sets.
The custom ISC SPNEGO mechanism implementation is no longer needed on
the basis that all major Kerberos 5/GSSAPI (mit-krb5, heimdal and
Windows) implementations support SPNEGO mechanism since 2006.
This commit removes the custom ISC SPNEGO implementation, and removes
the option from both autoconf and win32 Configure script. Unknown
options are being ignored, so this doesn't require any special handling.
The 'checknames' field wasn't initialized in dns_view_create(), but it
should otherwise AddressSanitizer identifies the following runtime error
in query_test.c.
runtime error: load of value 190, which is not a valid value for type '_Bool'
(cherry picked from commit 0c6fa16477)
The ISC SPNEGO implementation is based on mod_auth_kerb code. When
CVE-2006-5989 was disclosed, the relevant fix was not applied to the
BIND 9 codebase, making the latter vulnerable to the aforementioned flaw
when "tkey-gssapi-keytab" or "tkey-gssapi-credential" is set in
named.conf.
The original description of CVE-2006-5989 was:
Off-by-one error in the der_get_oid function in mod_auth_kerb 5.0
allows remote attackers to cause a denial of service (crash) via a
crafted Kerberos message that triggers a heap-based buffer overflow
in the component array.
Later research revealed that this flaw also theoretically enables remote
code execution, though achieving the latter in real-world conditions is
currently deemed very difficult.
This vulnerability was responsibly reported as ZDI-CAN-12302 ("ISC BIND
TKEY Query Heap-based Buffer Overflow Remote Code Execution
Vulnerability") by Trend Micro Zero Day Initiative.