Breaking News: Thousands Of GitHub Accounts Hijacked With wfx-loot CI Secret
Read the Report
Learn How MCP Is Bypassing a Decade of Cloud Security Best Practices
Read the Report
Are you exposed? Connect a repo in minutes.
Scan your repos free

CVE-2026-121619

The "private_ranges" Trusted Proxies Shortcut Trusts Clients Presenting an IPv4-Mapped, 6to4, Teredo or NAT64 Address as Reverse Proxies

Overview

Published
12 Nov 2026
Modified
17 Nov 2026
Medium CVE
3 Registry packages
0 OS packages
0 Linux distributions
No Fix available
CVSS score Medium, no score

Description

The Symfony FrameworkBundle framework.trusted_proxies option and the HttpFoundation Request::setTrustedProxies() method trust remote clients as reverse proxies when the private_ranges shortcut is used, allowing remote attackers to spoof the client IP address, host and scheme in Symfony FrameworkBundle versions 7.1 prior to 7.4.23 and 8.x prior to 8.1.11, and HttpFoundation versions 7.2 prior to 7.4.23 and 8.x prior to 8.1.11. The private_ranges and PRIVATE_SUBNETS keywords resolved to IpUtils::PRIVATE_SUBNETS, the deny list of NoPrivateNetworkHttpClient, which lists the private and special-purpose prefixes a server must not connect to. Since the fix for CVE-2026-48736 (7.4.13 and 8.0.13), this list also covers the IPv6 prefixes that embed an IPv4 address: IPv4-mapped (::ffff:0:0/96), IPv4-compatible (::/96), 6to4 (2002::/16), Teredo (2001::/32) and NAT64 (64:ff9b::/96). The IPv4-mapped entry has been in the list since the private_ranges shortcut was introduced, in 7.1 for framework.trusted_proxies and in 7.2 for Request::setTrustedProxies(). Over-blocking is safe for a deny list, but when the same entries are used as the allow list of trusted proxies, they trust remote clients. A client whose connection arrives with such a source address is treated as a reverse proxy, and the X-Forwarded-For, X-Forwarded-Host, X-Forwarded-Proto, X-Forwarded-Port and X-Forwarded-Prefix headers it sends are honoured. For example, with Request::setTrustedProxies(['private_ranges'], ...), a request with REMOTE_ADDR set to ::ffff:203.0.113.7 and the headers X-Forwarded-For: 10.0.0.9, X-Forwarded-Host: evil.example.com and X-Forwarded-Proto: https makes getClientIp() return 10.0.0.9, getHost() return evil.example.com and isSecure() return true. Two kinds of deployments are exposed: a server listening on the IPv6 wildcard address without IPV6_V6ONLY reports every IPv4 client in the IPv4-mapped form, so every client is trusted, and a server reachable over IPv6 trusts any client sourcing from 6to4 or Teredo space, which is globally routed, whatever its listener. A spoofed client address defeats IP allow lists, rate limiting and audit logs; a spoofed host poisons the URLs written into password reset and login link e-mails and into shared caches; a spoofed scheme satisfies requires_channel: https over plain HTTP. Symfony 6.4 does not point any trusted proxy setting at IpUtils::PRIVATE_SUBNETS and is not affected by default, but an application that passes that constant to Request::setTrustedProxies() itself could be exposed in the same way on any version. This issue is fixed in 7.4.23 and 8.1.11.

Source: OX vulnerability database

Severity

MEDIUM

No CVSS vector in the source. Severity comes from the advisory.

Weakness: CWE-290, CWE-348

Affected packages · 3

Packages affected by CVE-2026-121619
EcosystemPackageVulnerableFixed in
Composer (PHP) symfony/framework-bundle >=7.4.0, <7.4.23>=8.0.0, <8.1.0>=8.1.0, <8.1.11>=7.1.0, <7.2.0>=7.2.0, <7.3.0+1 more No fix yet
Composer (PHP) symfony/http-foundation >=8.1.0, <8.1.11>=7.2.0, <7.3.0>=7.3.0, <7.4.0>=7.4.0, <7.4.23>=8.0.0, <8.1.0 No fix yet
Composer (PHP) symfony/symfony >=7.2.0, <7.3.0>=7.3.0, <7.4.0>=7.4.0, <7.4.23>=8.0.0, <8.1.0>=8.1.0, <8.1.11+1 more No fix yet

Open a package to see all its vulnerabilities.

Operating systems

No Linux distribution reports this ID.

References

OX Radar and Response Center

Find threats in your stack.

OX Radar can check high-priority threats, like critical package CVEs, against your context.

Each incident is marked Detected or Not Detected with a remediation plan and proof all tracked in the Response Center.