Security Advisory

CVE-2026-3549: wolfSSL TLS 1.3 ECH Heap Overflow

Last updated:
  • CVE-2026-3549

CVE-2026-3549 is a heap buffer overflow in the TLS 1.3 Encrypted Client Hello (ECH) parsing of the wolfSSL embedded TLS library: an integer underflow in a buffer-length calculation writes beyond an allocated buffer. NVD scores it 9.8, wolfSSL rates it High, and 5.9.0 fixes it; only builds with ECH enabled are affected, and ECH is off by default.

Vulnerability summary

CVSS base score

9.8

Critical severity

EPSS score (FIRST)

0.5%

Probability of exploitation in the wild in the next 30 days

CISA KEV catalog

No

Not listed

Published

Vulnerability ID: CVE-2026-3549 · Source: NVD

Affected components

ComponentCPEVersion range
wolfSSL (TLS 1.3 Encrypted Client Hello parsing)cpe:2.3:a:wolfssl:wolfssl:*:*:*:*:*:*:*:*NVD: before 5.9.0; only builds with ECH (--enable-ech / HAVE_ECH); fixed in 5.9.0 (PR 9817)
OpenWrt libwolfssl package—5.7.6 in OpenWrt 24.10.0-24.10.6 and 5.8.4 in 25.12.0-25.12.2 (stock build has no --enable-ech); 5.9.1 from 24.10.7 and 25.12.3

What does this mean for affected devices?

CVE-2026-3549 is a heap-based buffer overflow (CWE-122, per wolfSSL as CNA). The 5.9.0 section of the wolfSSL ChangeLog (18 March 2026) says "an integer underflow existed in ECH extension parsing logic when calculating a buffer length, which resulted in writing beyond the bounds of an allocated buffer" and adds that "in wolfSSL, ECH is off by default, and the ECH standard is still evolving". It was found by internal wolfSSL testing, with thanks to Oleh Konko, and fixed in pull request 9817. The wolfSSL vulnerabilities page rates it High.

NVD gives it a Primary CVSS v3.1 score of 9.8 with vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H and covers all versions before 5.9.0 (NVD). wolfSSL's own CVSS v4.0 score is 8.3 with AT:P, an attack requirement that reflects the ECH build condition. NVD's score does not account for that condition, so the build decides whether a device is exposed.

In wolfSSL's configure.ac the --enable-ech option is "default: disabled" and defines HAVE_ECH when switched on. OpenWrt lists CVE-2026-3549 among the CVEs fixed by its wolfSSL 5.9.0 update, but the stock OpenWrt wolfSSL Makefile does not pass --enable-ech, so its libwolfssl builds (5.7.6 in 24.10.0 to 24.10.6, 5.8.4 in 25.12.0 to 25.12.2) carry the version without the vulnerable feature. For an RTOS image, look for HAVE_ECH in user_settings.h.

Debian fixed it in sid with 5.9.0-0.1 and marks bookworm and trixie as minor issues (Debian security tracker).

CVE-2026-3549 is not in the CISA KEV catalog as of 2026-11-16 (catalog version 2026.11.13), and FIRST EPSS gives it 0.0054, the 45.8th percentile, in the 2026-11-15 dataset (CISA KEV, FIRST EPSS).

If the build defines HAVE_ECH, update to 5.9.0 or later (the current release is 5.9.4); if it does not, record a not-affected statement with that justification. Our Mbed TLS and wolfSSL firmware guide covers how to read the build options, and our memory corruption fix guide covers the bug class. Other wolfSSL flaws that depend on an optional feature are CVE-2026-55962 and CVE-2026-94418.

How to fix this class of issueThis advisory's weakness class is Memory corruption. Our practical guide covers how to detect, fix, verify and disclose it: How to fix memory corruption in embedded firmware.

Evidence

Related advisories

Sources

  1. NVD
  2. wolfSSL: Security Vulnerabilities
  3. wolfSSL ChangeLog.md (release 5.9.0)
  4. wolfSSL pull request 9817
  5. wolfSSL configure.ac (v5.9.2-stable)
  6. OpenWrt commit: wolfssl update to 5.9.0
  7. OpenWrt wolfSSL package Makefile (v24.10.8)
  8. Debian security tracker
  9. CISA KEV
  10. FIRST EPSS

Frequently asked questions

Is every wolfSSL build affected by CVE-2026-3549?

No. The overflow is in Encrypted Client Hello parsing, and wolfSSL says ECH is off by default; configure.ac lists --enable-ech as disabled by default, and it defines HAVE_ECH when enabled (per the wolfSSL ChangeLog and configure.ac).

Why is the NVD score 9.8 when wolfSSL rates it High?

NVD's CVSS v3.1 vector is network, low complexity, no privileges and full impact. wolfSSL's CVSS v4.0 score of 8.3 adds an attack requirement (AT:P) for the ECH build condition (per NVD).

Does OpenWrt build wolfSSL with ECH?

No. OpenWrt listed CVE-2026-3549 in its 5.9.0 update, but its stock wolfSSL Makefile does not pass --enable-ech (per the OpenWrt Makefile and commit).

Is CVE-2026-3549 in the CISA KEV catalog?

No, not as of 2026-11-16 (catalog version 2026.11.13). FIRST EPSS gives it 0.0054, the 45.8th percentile, in the 2026-11-15 dataset (per CISA KEV and FIRST EPSS).