HAPI FHIR: Incomplete fix for CVE-2026-45367: DSTU2 FHIRPathEngine.matches() missing RegexTimeout protection allows ReDoSGHSA-fxj4-p9xp-37v5
HIGHGHSA-fxj4-p9xp-37v5 is a high-severity (CVSS 7.5) CWE-1333 vulnerability in ca.uhn.hapi.fhir:org.hl7.fhir.dstu2. A fix is available for ca.uhn.hapi.fhir:org.hl7.fhir.dstu2 — see the affected versions and patch details below.
Exploitation Status
Proof-of-concept exploit code exists
- CISA’s SSVC triage found public proof-of-concept exploit code for this CVE, though no confirmed active exploitation.
- CISA assesses this as automatable — exploitation doesn’t require manual, per-target effort, which raises the odds of mass scanning and opportunistic attacks.
Exploitation and automatability from CISA’s SSVC triage for GHSA-fxj4-p9xp-37v5.
EPSS Exploitation Probability
Probability of exploitation in the next 30 days, from FIRST.org EPSS.
How urgent is this, really
GHSA-fxj4-p9xp-37v5 by exploitation likelihood (EPSS) against impact (CVSS). Outside the shaded patch-first corner.
Where this sits among everything scored
Of 384,189 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Counts from FIRST.org, log-scaled.
Real-World Exposure
ca.uhn.hapi.fhir:org.hl7.fhir.dstu2☕ca.uhn.hapi.fhir:org.hl7.fhir.convertors☕ca.uhn.hapi.fhir:org.hl7.fhir.validation☕ca.uhn.hapi.fhir:org.hl7.fhir.validation.cliReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Maven packages — download data is not available via public APIs for these ecosystems.
Description
Summary
The fix for CVE-2026-45367 added RegexTimeout protection to the matches() function in DSTU2016MAY, DSTU3, R4, R4B, and R5, but the DSTU2 module was incompletely patched. In org.hl7.fhir.dstu2, replaceMatches() was updated while matches() at line 2462 still calls the raw String.matches(sw) without any timeout, allowing an unauthenticated attacker to trigger catastrophic regex backtracking and exhaust server CPU.
Details
Incomplete patch
Within the same file
(org.hl7.fhir.dstu2/utils/FHIRPathEngine.java), the two functions were
patched inconsistently:
Line 2226 — replaceMatches() — PATCHED:
result.add(new StringType(
RegexTimeout.replaceAll(
convertToString(focus.get(0)), regex, repl, regexTimeoutMillis)));
Line 2462 — matches() — NOT PATCHED:
result.add(new BooleanType(
convertToString(focus.get(0)).matches(sw)));
// ↑ raw String.matches() — no RegexTimeout, no complexity check
DSTU3 line 2447 — matches() — PATCHED (for comparison):
result.add(new BooleanType(
RegexTimeout.matches(st, sw, regexTimeoutMillis)));
Module-by-module status
| Module | matches() | replaceMatches() |
|---|---|---|
| DSTU2 | ❌ raw str.matches(sw) | ✅ RegexTimeout.replaceAll() |
| DSTU2016MAY | ✅ RegexTimeout.matches() | ✅ |
| DSTU3 | ✅ RegexTimeout.matches() | ✅ |
| R4 | ✅ RegexTimeout.matches() | ✅ |
| R4B | ✅ RegexTimeout.matches() | ✅ |
| R5 | ✅ RegexTimeout.matches() | ✅ |
PoC
Requirements: Java 17+, Maven 3.8+
pom.xml dependencies:
<dependency>
<groupId>ca.uhn.hapi.fhir</groupId>
<artifactId>org.hl7.fhir.utilities</artifactId>
<version>6.9.7</version>
</dependency>
Test code (reproduces the exact behaviour of DSTU2 line 2462):
import org.hl7.fhir.utilities.regex.RegexTimeout;
import java.util.concurrent.*;
String regex = "((a|b){0,5}){20}";
String input = "a".repeat(25) + "c"; // no match → full backtracking
// ① Patched approach — RegexTimeout terminates at 500 ms
long t1 = System.currentTimeMillis();
try {
RegexTimeout.matches(input, regex, 500);
} catch (TimeoutException e) {
System.out.println("RegexTimeout blocked in " +
(System.currentTimeMillis() - t1) + " ms");
}
// ② DSTU2 line 2462 — raw String.matches(), no timeout
long t2 = System.currentTimeMillis();
input.matches(regex); // equivalent to what FHIRPathEngine does
System.out.println("str.matches() ran for " +
(System.currentTimeMillis() - t2) + " ms with no timeout");
Verified output (JDK 25.0.3, Linux):
RegexTimeout blocked in 508 ms ← patched modules: attack stopped
str.matches() ran for 1410 ms ← DSTU2: no timeout, CPU exhausted
The patched approach cuts off the evaluation at 508 ms. The unpatched DSTU2 code runs for 1410 ms on this input with no mechanism to stop it. Longer inputs or more complex patterns produce proportionally worse results.
Impact
Vulnerability type: Regular Expression Denial of Service (ReDoS) causing CPU exhaustion and service disruption.
Who is impacted: Any application using the ca.uhn.hapi.fhir:org.hl7.fhir.dstu2 module that evaluates user-supplied FHIRPath expressions — including the FHIR Validator HTTP endpoint, FHIR servers applying FHIRPath invariants from user-provided resources or profiles, and any system embedding FHIRPathEngine from the DSTU2 module. No authentication is required; an attacker needs only to submit a FHIR resource or FHIRPath expression whose matches() argument contains a catastrophically backtracking regular expression.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| ☕Maven | ca.uhn.hapi.fhir:org.hl7.fhir.dstu2 | all versions | 6.9.10ca.uhn.hapi.fhir:org.hl7.fhir.dstu2:6.9.10 |
| ☕Maven | ca.uhn.hapi.fhir:org.hl7.fhir.convertors | all versions | 6.9.10ca.uhn.hapi.fhir:org.hl7.fhir.convertors:6.9.10 |
| ☕Maven | ca.uhn.hapi.fhir:org.hl7.fhir.validation | all versions | 6.9.10ca.uhn.hapi.fhir:org.hl7.fhir.validation:6.9.10 |
| ☕Maven | ca.uhn.hapi.fhir:org.hl7.fhir.validation.cli | all versions | 6.9.10ca.uhn.hapi.fhir:org.hl7.fhir.validation.cli:6.9.10 |
Affected Products
hl7 fhir corehapifhirDetection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for ca.uhn.hapi.fhir:org.hl7.fhir.dstu2, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update ca.uhn.hapi.fhir:org.hl7.fhir.dstu2 to 6.9.10 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-fxj4-p9xp-37v5 is resolved across your whole dependency graph.
Workarounds
Cap what an attacker can consume: apply request size, rate and timeout limits in front of the affected component, and run it with memory and CPU limits so exhaustion degrades one worker rather than the whole service.
Fixing This On Your OS
If you run this on a Linux distribution, patch through your package manager against the distro's own security advisory below — it tracks the exact backported fix for your release, which can ship on a different timeline (and sometimes a different severity) than the upstream project.
This flaw is rated as Important because an unauthenticated, remote attacker can exploit an incomplete patch in the HAPI FHIR DSTU2 module. By submitting a specially crafted FHIRPath expression, an attacker can trigger a Regular Expression Denial of Service (ReDoS) attack, leading to significant resource exhaustion and…
Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.Source: Red Hat security advisory for GHSA-fxj4-p9xp-37v5 (CC BY 4.0)
| Product | Fixed in | Advisory |
|---|---|---|
| Red Hat Build of Apache Camel 4.18 for Quarkus 3.33 | org.hl7.fhir.dstu2 | RHSA-2026:54776 |
Frequently Asked Questions
Is GHSA-fxj4-p9xp-37v5 in your dependencies?
Find it across Maven, including transitive dependencies.