Threats Tagged 'cve-2026-56416'

This update for unbound fixes the following issues: Update to version 1.25.2. Security issues fixed: - CVE-2026-14586: DoS via assertion in `libngtcp2` for DNS-over-QUIC environments with high concurrency (bsc#1271879). - CVE-2026-32665: Remote DNS-over-QUIC denial of service due to `quic-size` budget bypass (bsc#1271873). - CVE-2026-40691: DoS due to heap overflow via single bad DNSCrypt query over TCP (bsc#1271875). - CVE-2026-41637: Degradation of resolution service due to improperly accounted client-terminated DNS-over-QUIC queries (bsc#1271891). - CVE-2026-42955: Ghost domain window can be extended by up to one cached TTL configured value for A/AAAA glue records (bsc#1271892). - CVE-2026-44621: Libunbound applications configured with `unwanted-reply-threshold` could eventually be abruptly terminated (bsc#1271876). - CVE-2026-44687: Off-by-one error in `harden-below-nxdomain` logic can shadow a stub/forward zone by a legitimate parent's `NXDOMAIN` (bsc#1271893). - CVE-2026-44690: Cross-zone wildcard cache poisoning via `RRSIG.labels` manipulation (bsc#1271877). - CVE-2026-46582: Replay of a wildcard `rrset` as another piece of data triggers poisoning in the server expired reply path (bsc#1271894). - CVE-2026-50045: `max-global-quota` bypass via single client queries for a deeply nested name under a DNSSEC-signed parent (bsc#1271878). - CVE-2026-50046: Possible heap use-after-free in an error path when a DoT forwarded query is jostled out (bsc#1271882). - CVE-2026-50243: `response-ip`/`rpz` can rewrite BOGUS answers instead of returning SERVFAIL (bsc#1271880). - CVE-2026-50248: BOGUS configured primary hostname accepted for XFR in auth/rpz zones (bsc#1271881). - CVE-2026-50251: Attacker supplied `0.0.0.0`/`::` glue triggers defensive full-cache flush (bsc#1271883). - CVE-2026-50252: Possible cache poisoning attack by mapping source port population per thread (bsc#1271884). - CVE-2026-54478: DNS Cookie bypass when proxy-protocol with with `answer-cookie:yes` is used (bsc#1271895). - CVE-2026-55708: Privacy/configuration issue when adding local data in views through `unbound-control` (bsc#1271896). - CVE-2026-55717: `serve-expired-client-timeout` and `response-ip` CNAME redirect could lead to a crash (bsc#1271886). - CVE-2026-55973: DoS due to stack overflow triggered crash when `dns-error-reporting: yes` is set (bsc#1271874). - CVE-2026-55990: Crash via crafted client UDP query due to DNSCrypt faulty configuration (bsc#1271887). - CVE-2026-55991: Remote DNS-over-QUIC (DoQ) flow-control reacheable assertion failure in `libngtcp2` (bsc#1271888). - CVE-2026-56416: Possible heap buffer overflow when validator canonicalizes RDATA that contains domain name (bsc#1271889). - CVE-2026-56444: Degradation of resolution service when `discard-timeout` and `serve-expired-client-timeout` are combined in unusual configuration (bsc#1271890). Other updates and bugfixes: - Version 1.25.2: + Features: * TTL behavior changes: cached records reaching TTL 0 are expired; TTL 0 upstream answers are no longer cached by cachedb; serve-expired-reply-ttl is now capped by the original TTL value; TTL decoding updated to adhere to RFC 8767 section 4 * Add new statistics: num.queries.replyaddr_limit and requestlist.current.replies * Add 'log-thread-id' configuration option to log the system-wide Linux thread ID for easier debugging * Add ECC-GOST12 support per RFC 9558 (available as contrib/gost12.patch) * Allow synthesized DNAME TTL=0 to be served from cache within a 1-second grace period, reducing recursion for TTL=0 DNAMEs (RFC 2308) * Fix DoT/DoH/DoQ to reload certificates on config reload without requiring a full restart; fast_reload now supports changes to tls-service-key, tls-service-pem and tls-cert-bundle * Allow ip@port notation in control-interface configuration * Add iter-scrub-rrsig option (default: 8) to limit the number of RRSIGs processed by the scrubber * Add 'tls-protocols' configuration option to select which TLS protocol versions are used; TLSv1.2 is re-enabled by default * Add pthread_setname_np support for named threads + Bug Fixes: * Fix handle leak in pythonmod on pythonmod_init * Fix crash when mesh_detect_cycle_found() is called with no mesh state * Fix modstack_call_init to use the original string when it has changed * Fix fr_atomic_copy_cfg * Fix auth-zone empty label for $ORIGIN when downloading via HTTP * Fix respip and dns64 to be usable simultaneously; RPZ now works with DNS64 * Fix HTTPS and QUIC not being enabled when port is listed in interface-automatic-ports * Allow wait-limit-cookie: 0 to disable cookie-validated wait limits * Fix FIPS mode in OpenSSL causing unit test failure * Fix discard-timeout to only drop UDP, not stream connections * Reply with SERVFAIL when the wait-limit is exceeded * A

NLnet Labs Unbound up to and including version 1.25.1 contains a heap buffer overflow vulnerability in the DNSSEC validator when processing certain RRSIG-covered PX, RP, MINFO, or SOA resource records. The issue arises because the validator does not verify the presence of a second embedded domain name before passing its address to a function that converts domain names to lowercase, potentially causing out-of-bounds memory reads and heap corruption. This vulnerability has a medium severity rating and a CVSS score of 4.8.

In NLnet Labs Unbound up to and including version 1.25.1, when the validator builds the canonical RDATA form for an RRSIG-covered PX/RP/MINFO/SOA RRset, it computes the address of the second embedded domain name as 'datstart + dname_valid(datstart, ...)' and passes it straight to 'query_dname_tolower()' without checking that a second name is actually present in the RDATA. The wire-format parser accepts multi-dname RRs whose RDATA ends after the first name, so an attacker who runs a DNSSEC-signed authoritative server can deliver a record with an absent second domain name (e.g. SOA record) and cause 'query_dname_tolower()' to walk label-by-label through stale bytes in the per-worker 'env->scratch_buffer', past the end of that heap allocation if 'msg-buffer-size' has been lowered from the default. This leads to heap buffer overflow and on a release build the outcome relies heavily on the contents of the buffer tail and the adjacent heap chunk.