Nlnetlabs Unbound

NLnet Labs Unbound up to and including version 1.21.0 contains a vulnerability when handling replies with very large RRsets that it needs to perform name compression for. Malicious upstreams responses with very large RRsets can cause Unbound to spend a considerable time applying name compression to downstream replies. This can lead to degraded performance and eventually denial of service in well orchestrated attacks. The vulnerability can be exploited by a malicious actor querying Unbound for the specially crafted contents of a malicious zone with very large RRsets. Before Unbound replies to the query it will try to apply name compression which was an unbounded operation that could lock the CPU until the whole packet was complete. Unbound version 1.21.1 introduces a hard limit on the number of name compression calculations it is willing to do per packet. Packets that need more compression will result in semi-compressed packets or truncated packets, even on TCP for huge messages, to avoid locking the CPU for long. This change should not affect normal DNS traffic.

NLnet Labs Unbound version 1.18.0 up to and including version 1.19.1 contain a vulnerability that can cause denial of service by a certain code path that can lead to an infinite loop. Unbound 1.18.0 introduced a feature that removes EDE records from responses with size higher than the client's advertised buffer size. Before removing all the EDE records however, it would try to see if trimming the extra text fields on those records would result in an acceptable size while still retaining the EDE codes. Due to an unchecked condition, the code that trims the text of the EDE records could loop indefinitely. This happens when Unbound would reply with attached EDE information on a positive reply and the client's buffer size is smaller than the needed space to include EDE records. The vulnerability can only be triggered when the 'ede: yes' option is used; non default configuration. From version 1.19.2 on, the code is fixed to avoid looping indefinitely.

Certain DNSSEC aspects of the DNS protocol (in RFC 4033, 4034, 4035, 6840, and related RFCs) allow remote attackers to cause a denial of service (CPU consumption) via one or more DNSSEC responses, aka the "KeyTrap" issue. One of the concerns is that, when there is a zone with many DNSKEY and RRSIG records, the protocol specification implies that an algorithm must evaluate all combinations of DNSKEY and RRSIG records.

A vulnerability named 'Non-Responsive Delegation Attack' (NRDelegation Attack) has been discovered in various DNS resolving software. The NRDelegation Attack works by having a malicious delegation with a considerable number of non responsive nameservers. The attack starts by querying a resolver for a record that relies on those unresponsive nameservers. The attack can cause a resolver to spend a lot of time/resources resolving records under a malicious delegation point where a considerable number of unresponsive NS records reside. It can trigger high CPU usage in some resolver implementations that continually look in the cache for resolved NS records in that delegation. This can lead to degraded performance and eventually denial of service in orchestrated attacks. Unbound does not suffer from high CPU usage, but resources are still needed for resolving the malicious delegation. Unbound will keep trying to resolve the record until hard limits are reached. Based on the nature of the attack and the replies, different limits could be reached. From version 1.16.3 on, Unbound introduces fixes for better performance when under load, by cutting opportunistic queries for nameserver discovery and DNSKEY prefetching and limiting the number of times a delegation point can issue a cache lookup for missing records.

NLnet Labs Unbound, up to and including version 1.16.1, is vulnerable to a novel type of the "ghost domain names" attack. The vulnerability works by targeting an Unbound instance. Unbound is queried for a rogue domain name when the cached delegation information is about to expire. The rogue nameserver delays the response so that the cached delegation information is expired. Upon receiving the delayed answer containing the delegation information, Unbound overwrites the now expired entries. This action can be repeated when the delegation information is about to expire making the rogue delegation information ever-updating. From version 1.16.2 on, Unbound stores the start time for a query and uses that to decide if the cached delegation information can be overwritten.

NLnet Labs Unbound, up to and including version 1.16.1 is vulnerable to a novel type of the "ghost domain names" attack. The vulnerability works by targeting an Unbound instance. Unbound is queried for a subdomain of a rogue domain name. The rogue nameserver returns delegation information for the subdomain that updates Unbound's delegation cache. This action can be repeated before expiry of the delegation information by querying Unbound for a second level subdomain which the rogue nameserver provides new delegation information. Since Unbound is a child-centric resolver, the ever-updating child delegation information can keep a rogue domain name resolvable long after revocation. From version 1.16.2 on, Unbound checks the validity of parent delegation records before using cached delegation information.

Unbound before 1.9.5 allows an out-of-bounds write via a compressed name in rdata_copy. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an assertion failure via a compressed name in dname_pkt_copy. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an infinite loop via a compressed name in dname_pkt_copy. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an integer overflow in a size calculation in respip/respip.c. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an integer overflow in a size calculation in dnscrypt/dnscrypt.c. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an assertion failure and denial of service in dname_pkt_copy via an invalid packet. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an assertion failure and denial of service in synth_cname. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an out-of-bounds write in sldns_bget_token_par. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an integer overflow in sldns_str2wire_dname_buf_origin, leading to an out-of-bounds write. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an integer overflow in the regional allocator via the ALIGN_UP macro. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows an integer overflow in the regional allocator via regional_alloc. NOTE: The vendor disputes that this is a vulnerability. Although the code may be vulnerable, a running Unbound installation cannot be remotely or locally exploited

Unbound before 1.9.5 allows configuration injection in create_unbound_ad_servers.sh upon a successful man-in-the-middle attack against a cleartext HTTP session. NOTE: The vendor does not consider this a vulnerability of the Unbound software. create_unbound_ad_servers.sh is a contributed script from the community that facilitates automatic configuration creation. It is not part of the Unbound installation

NLnet Labs Unbound, up to and including version 1.12.0, and NLnet Labs NSD, up to and including version 4.3.3, contain a local vulnerability that would allow for a local symlink attack. When writing the PID file, Unbound and NSD create the file if it is not there, or open an existing file for writing. In case the file was already present, they would follow symlinks if the file happened to be a symlink instead of a regular file. An additional chown of the file would then take place after it was written, making the user Unbound/NSD is supposed to run as the new owner of the file. If an attacker has local access to the user Unbound/NSD runs as, she could create a symlink in place of the PID file pointing to a file that she would like to erase. If then Unbound/NSD is killed and the PID file is not cleared, upon restarting with root privileges, Unbound/NSD will rewrite any file pointed at by the symlink. This is a local vulnerability that could create a Denial of Service of the system Unbound/NSD is running on. It requires an attacker having access to the limited permission user Unbound/NSD runs as and point through the symlink to a critical file on the system.

Unbound before 1.10.1 has an infinite loop via malformed DNS answers received from upstream servers.

Unbound before 1.10.1 has Insufficient Control of Network Message Volume, aka an "NXNSAttack" issue. This is triggered by random subdomains in the NSDNAME in NS records.

Unbound 1.6.4 through 1.9.4 contain a vulnerability in the ipsec module that can cause shell code execution after receiving a specially crafted answer. This issue can only be triggered if unbound was compiled with `--enable-ipsecmod` support, and ipsecmod is enabled and used in the configuration.

Unbound before 1.9.4 accesses uninitialized memory, which allows remote attackers to trigger a crash via a crafted NOTIFY query. The source IP address of the query must match an access-control rule.

A flaw was found in the way unbound before 1.6.8 validated wildcard-synthesized NSEC records. An improperly validated wildcard NSEC record could be used to prove the non-existence (NXDOMAIN answer) of an existing wildcard record, or trick unbound into accepting a NODATA proof.

iterator.c in NLnet Labs Unbound before 1.5.1 does not limit delegation chaining, which allows remote attackers to cause a denial of service (memory and CPU consumption) via a large or infinite number of referrals.