OpenSSL: Buffer Overflow In EVP_EncodeBlock() (CVE-2026-121481, CVSS 9.8)
AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

An OpenSSL pull request proposes input checks to address a buffer overflow in EVP_EncodeBlock(), tracked as CVE-2026-121481 and listed in the prompt with a CVSS score of 9.8. The proposed changes also cover EVP_DecodeBlock(), a length check in http_client.c, documentation and tests; the supplied report says the PR was not yet merged when a November 20 backport reference described it as pending.

An OpenSSL pull request proposes changes to address a reported buffer overflow in EVP_EncodeBlock(), identified as CVE-2026-121481 and assigned a CVSS score of 9.8 in the supplied issue description. The patch also proposes input guards for the related decoder, a length-check correction in an HTTP client function, documentation changes and new tests; the provided material does not confirm that the changes have been merged or released.

Before you orderOffer from Amazon

Get privacy and security gear delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

The GitHub pull request says the encoder fix checks inputs that could cause undefined behavior and returns an error for invalid values. The report contains two descriptions of the proposed handling: one says the function returns -1 for invalid inputs, while another says it uses OPENSSL_assert() for inputs leading to undefined behavior. Those descriptions are not identical, and the supplied record does not establish which implementation is the final agreed approach.

The patch also applies a similar guard to EVP_DecodeBlock() and corrects the length check in base64encode() in http_client.c. The reported issue is that the existing check did not properly rule out lengths whose encoded output would not fit in an integer. A November 20 reference to a proposed backport describes rejecting negative lengths and lengths whose base64 output would exceed the representable range, and notes that the HTTP client path could be reached through a long s_client -proxy_pass value. That reference says the upstream change was not merged at the time.

The pull request record says the documentation was updated and tests were added, including tests that document behavior for zero-length inputs. Its activity log shows an approval on November 11, 2026, a request for changes on November 12, and a community-discussion hold added on November 13. These events indicate review activity, not a completed release or confirmation that users are protected.

At a glance
updateWhen: Developing; the supplied GitHub activit…
The developmentAn OpenSSL GitHub pull request proposes guards and related fixes for a reported buffer overflow in EVP_EncodeBlock().

Why Length Checks Matter

Base64 encoding and decoding are common operations in software that uses OpenSSL. A buffer overflow can cause memory corruption when a function writes beyond the space available for its output. The potential consequences depend on how an affected application calls the function and handles the input; the supplied source material does not report a confirmed exploit or describe a specific attack in the wild.

The related http_client.c correction matters because a length calculation can fail even when an input length appears to pass a simple upper-bound check. If the encoded result cannot be represented safely, a calculation or allocation based on that result may be wrong. The CVSS 9.8 score signals a high severity rating in the issue description, but the supplied material does not include the score’s vector or the details needed to assess its assumptions independently.

For maintainers and users, the practical point is to distinguish a proposed source-code fix from a released update. A pull request, a downstream backport reference and a fuzzer commit do not by themselves show which official OpenSSL versions are affected, which have been fixed, or whether a particular product has incorporated a remedy.

Products Worth Considering

Bose QuietComfort Ultra Earbuds (2nd Gen), Wireless Noise-Cancelling Earbuds, White Smoke

Bose QuietComfort Ultra Earbuds (2nd Gen), Wireless Noise-Cancelling Earbuds, White Smoke

As an affiliate, we earn on qualifying purchases.

Review and Backport Status

The source material centers on OpenSSL pull request #33652. It records review comments and a hold asking the community to reach consensus on how to proceed. The patch description links the encoder issue to additional input validation, decoder behavior, a client-side length check, documentation and tests.

On November 20, a separate pkgsrc commit referenced the pull request and described a backport to OpenSSL 3.6.5, while explicitly saying the upstream pull request was not merged and an advisory was pending. This is evidence of a downstream packaging change being discussed or prepared, not proof that the change was included in an official OpenSSL release. A separate fuzzer commit also referenced the work, but the supplied details do not establish that it constitutes a security fix or a release announcement.

The provided record gives no official security advisory, affected-version range, patched release number, or direct explanation from OpenSSL maintainers of the vulnerability’s discovery and impact. Those details are needed before users can determine exposure from version numbers alone.

“This fixes the buffer overflow in `EVP_EncodeBlock()` (CVE-2026-121481) by returning `-1` for invalid inputs.”

— OpenSSL pull request description

Affected Releases Still Unspecified

The supplied material does not identify the affected OpenSSL versions, a fixed release, or whether the pull request has since been merged. It also does not provide an official advisory or a complete technical description of conditions needed to trigger the overflow. The CVSS 9.8 figure appears in the issue title, but no scoring vector or attribution is included.

There is also a difference between the pull request’s descriptions of the encoder’s response: one refers to returning -1, while another refers to an assertion. Without the final patch and its review outcome, it is not possible to say which behavior was adopted. The backport note describes checks for negative or excessively large lengths, but does not establish the scope of exposure across other branches or products that bundle OpenSSL.

Watch for an Advisory and Release

The next evidence to look for is an updated status on OpenSSL pull request #33652, followed by any official security advisory or release notes naming affected and fixed versions. A merged patch would clarify the code changes; a release announcement would let users and distributors identify an available update.

Downstream maintainers may publish their own notices if they apply the proposed guards to packaged versions. Until such information appears, operators should not treat the pull request or the reported backport as confirmation that a fix is available in their installed build. They can track announcements from OpenSSL and their software vendors, then compare those notices with the versions they deploy.

Key Questions

What is CVE-2026-121481?

The supplied issue description identifies it as a buffer overflow in OpenSSL’s EVP_EncodeBlock() function and lists a CVSS score of 9.8. The source provided here does not include an official advisory or further technical details about triggering conditions.

Has OpenSSL released a fix?

The supplied information describes a proposed pull request, not a confirmed release. A November 20 downstream reference says the upstream change was not merged and an advisory was pending at that time; later status is not given.

Which OpenSSL versions are affected?

The source material does not provide an affected-version range. It mentions a proposed downstream backport to OpenSSL 3.6.5, but that does not establish which versions are vulnerable or fixed.

What other code does the proposed patch change?

Besides EVP_EncodeBlock(), the pull request describes a guard for EVP_DecodeBlock(), a corrected length check in http_client.c’s base64encode() function, updated documentation and new tests, including tests for zero-length inputs.

Source: github.com

You May Also Like

How Safe Is Paypal From Hackers? Learn the Facts!

Wondering if PayPal is truly secure? Discover the robust security measures and protection against hackers in this insightful guide!

Are Blink Cameras Safe From Hackers

Worried about the safety of your Blink cameras? Find out how recent vulnerabilities and security measures can impact your privacy.

Is Arlo Safe From Hackers? Protect Your Home Security!

Pondering how secure Arlo is from hackers? Unveil top-notch security measures and tips to fortify your home protection.

Malware 2.0: How Attacks Are Outsmarting Your Antivirus

As malware evolves with advanced stealth and evasion techniques, understanding these methods reveals how traditional antivirus tools are increasingly outmatched.