CVE-2020-25681 is a heap overflow in the DNSSEC code of dnsmasq before 2.83: forged DNS replies can corrupt heap memory and possibly run code. NVD scores it 8.1. It is part of DNSpooq and needs DNSSEC compiled in and enabled. CERT/CC lists Cradlepoint, Digi, NETGEAR and OpenWrt dnsmasq-full builds as affected.
Vulnerability summary
CVSS base score
8.1
High severity
EPSS score (FIRST)
88.3%
Probability of exploitation in the wild in the next 30 days
CISA KEV catalog
No
Not listed
Published
Vulnerability ID: CVE-2020-25681 · Source: NVD
Affected components
| Component | CPE | Version range |
|---|---|---|
| dnsmasq (DNSSEC validation, RRSet sorting) | cpe:2.3:a:thekelleys:dnsmasq:*:*:*:*:*:*:*:* | before 2.83, only with DNSSEC compiled in and enabled (dnsmasq CHANGELOG 2.83) |
| OpenWrt dnsmasq-full package | — | OpenWrt 19.07.0 to 19.07.5; fixed in 19.07.6 (dnsmasq 2.80-16.2); dnsmasq and dnsmasq-dhcpv6 variants not affected (OpenWrt advisory 2021-01-19-1) |
| Cradlepoint NetCloud OS, Digi DAL-based routers, NETGEAR RAX35 and RAX40 | — | NCOS up to 7.21.20; Digi fixed in firmware 21.2.x; NETGEAR before v1.0.3.88 (vendor statements in CERT/CC VU#434904) |
What does this mean for affected devices?
CVE-2020-25681 is a heap-based buffer overflow (CWE-122) in dnsmasq. NVD's description says it occurs "in the way RRSets are sorted before validating with DNSSEC data". An attacker on the network "who can forge DNS replies such as that they are accepted as valid" could overflow "a heap memory segment, possibly executing code on the machine". NVD scores it 8.1 (high), vector CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H, for dnsmasq before 2.83 (NVD).
The dnsmasq CHANGELOG entry for 2.83 states the condition directly: "Any dnsmasq with DNSSEC compiled in and enabled is vulnerable to this", and groups this CVE with CVE-2020-25682, -25683 and -25687. JSOF found the bugs, and CERT/CC published them on 2021-01-19 as DNSpooq.
The device impact comes from the vendor statements collected in CERT/CC VU#434904, which also tracks the set as ICS-VU-668462:
- Cradlepoint: NetCloud OS up to 7.21.20 affected. With the default (DNSSEC disabled) only the cache-poisoning CVEs apply. With DNSSEC enabled, Cradlepoint rates the risk Medium/High, and Critical if DNS is exposed on WAN interfaces.
- Digi International: fixed in firmware 21.2.x on all DAL-based products, including the Digi IX10, IX14, IX15, IX20, EX12, EX15, TX54, TX64 and LR54 routers and the Connect EZ and AnywhereUSB Plus products.
- NETGEAR: RAX40 and RAX35 before firmware v1.0.3.88.
- OpenWrt: "Only package dnsmasq-full, which is not installed by default, is affected." OpenWrt's advisory 2021-01-19-1 lists 19.07.0 to 19.07.5 as affected and the fix in 19.07.6 (dnsmasq 2.80-16.2). Its interim mitigation is
uci set dhcp.@dnsmasq[0].dnssec='0'. - Not affected: Cisco, Arista, Check Point and Juniper reported this CVE as not affecting their products. Red Hat said RHEL 5, 6 and 7 were not affected "because they are not compiled with DNSSEC support".
That split is the whole triage. A version scan flags every dnsmasq below 2.83, but only builds whose compile options include DNSSEC, with --dnssec switched on, run the vulnerable code. Our dnsmasq vulnerabilities in embedded firmware guide shows how to read the compile-options string and runtime configuration from an image. The cache-poisoning CVEs in the same set, such as CVE-2020-25684, apply to every variant regardless of DNSSEC. The 2017 overflow CVE-2017-14491 is the earlier example of a dnsmasq memory bug reachable from DNS replies.
CVE-2020-25681 is not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives it 0.883, the 99.8th percentile, in the 2026-11-15 dataset (CISA KEV, FIRST EPSS). EPSS does not know whether DNSSEC is enabled on a given device.
To fix it, upgrade to dnsmasq 2.83 or later, or to a vendor firmware that includes it. The Debian Security Tracker shows distribution backports such as 2.80-1+deb10u1 (DSA-4844-1). Until then, disable DNSSEC validation in dnsmasq, keep port 53 closed on WAN, and see our memory corruption fix guide for the weakness class.
Evidence
- Heap overflow sorting RRSets before DNSSEC validation; NVD CVSS 3.1 8.1 (AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H); CWE-122 (Red Hat); dnsmasq before 2.83. — NVD
- 2.83: "Any dnsmasq with DNSSEC compiled in and enabled is vulnerable to this", CVE-2020-25681, -25682, -25683, -25687. — dnsmasq CHANGELOG
- Vendor statements: Cradlepoint NCOS up to 7.21.20; Digi fixed in 21.2.x on DAL products; NETGEAR RAX40/RAX35 before v1.0.3.88; OpenWrt only dnsmasq-full; Cisco, Arista, Check Point, Juniper not affected by 25681; RHEL 5-7 not compiled with DNSSEC. — CERT/CC VU#434904
- OpenWrt 19.07.0-19.07.5 affected, fixed in 19.07.6 (dnsmasq 2.80-16.2); DNSSEC CVEs only affect dnsmasq-full; mitigation dnssec=0. — OpenWrt Security Advisory 2021-01-19-1
- Buster fixed in 2.80-1+deb10u1 (DSA-4844-1), stretch in 2.76-5+deb9u3 (DLA-2604-1). — Debian Security Tracker
- Not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13). — CISA KEV
- EPSS 0.88256 (99.8th percentile), dataset 2026-11-15, read 2026-11-16. — FIRST EPSS

Related advisories
Sources
Frequently asked questions
Does CVE-2020-25681 affect dnsmasq builds without DNSSEC?
No. The dnsmasq 2.83 changelog says any dnsmasq with DNSSEC compiled in and enabled is vulnerable. OpenWrt reports that only its dnsmasq-full package is affected, and Red Hat reported RHEL 5 to 7 unaffected because they are built without DNSSEC (per the dnsmasq CHANGELOG, OpenWrt and CERT/CC).
Which router and gateway firmware is affected by CVE-2020-25681?
Vendor statements in CERT/CC VU#434904 name Cradlepoint NetCloud OS up to 7.21.20, Digi DAL-based routers (fixed in 21.2.x), NETGEAR RAX40 and RAX35 before v1.0.3.88, and OpenWrt 19.07.0 to 19.07.5 with dnsmasq-full (per CERT/CC and OpenWrt).
Which dnsmasq version fixes CVE-2020-25681?
dnsmasq 2.83. OpenWrt shipped the fix in 19.07.6 as dnsmasq 2.80-16.2, and Debian backported it to buster as 2.80-1+deb10u1 (per the dnsmasq CHANGELOG, OpenWrt and the Debian Security Tracker).
Is CVE-2020-25681 in the CISA KEV catalog?
No, not as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives it 0.883, the 99.8th percentile, in the 2026-11-15 dataset (per CISA KEV and FIRST EPSS).
Questions about this page?