CVE-2026-26201

Summary

Multiple shared maps are accessed without consistent synchronization across goroutines. Under concurrent activity, Go runtime can trigger fatal error: concurrent map read and map write, causing C2 process crash (availability loss).

Vulnerable Component(with code examples)

Operator relay map had mixed access patterns (iteration and mutation without a single lock policy):

// vulnerable pattern (operator session map)
for sessionID, op := range OPERATORS { // iteration path
    ...
}
// concurrent mutation path elsewhere
OPERATORS[operatorSession] = &operator_t{...}
delete(OPERATORS, operatorSession)

Port-forwarding session map had read/write paths guarded inconsistently:

// vulnerable pattern (port forward map)
if sess, ok := PortFwds[id]; ok { // read path
    ...
}
PortFwds[id] = newSession // write path
delete(PortFwds, id)      // delete path

FTP stream map similarly mixed concurrent iteration with mutation:

// vulnerable pattern (FTP stream map)
for token, stream := range FTPStreams { // iteration path
    ...
}
FTPStreams[token] = stream // write path
delete(FTPStreams, token)  // delete path

Attack Vector

1. Attacker (or stress traffic in authenticated flows) triggers high concurrency in normal control paths.
2. Operator sessions connect/disconnect while message forwarding and file-transfer workflows are active.
3. Concurrent read/write hits shared maps.
4. Go runtime panics with concurrent map read/write error.
5. C2 component exits, producing denial of service.

Proof of Concept

1. Start C2 server with active operator session(s) in a lab environment.
2. Generate rapid operator session churn (connect/disconnect loops).
3. Simultaneously drive agent message tunnel traffic and/or file transfer activity.
4. Observe crash signature in logs: fatal error: concurrent map read and map write.
5. Optional: run with race detector in dev build to confirm race locations.

Impact

• C2 service interruption due to process panic/crash.
• Operational instability under load or deliberate churn.
• Repeated crash-restart cycles can degrade command reliability and incident response workflows.