CWE-1270 is about a product that uses security tokens to separate allowed actions from disallowed ones, but generates those tokens incorrectly. It is a hardware design issue for System-on-a-Chip teams.
How it happens
SoC designs give a security token to every agent that can generate or receive actions. Actions include read, write, program, reset, fetch and compute. Ideally each agent gets a unique token based on its trust level or privileges.
If the logic that produces these tokens is wrong, an agent can carry a token that does not match its real privileges. Every access control decision that trusts the token then inherits the error.
What an attacker gains
The record rates the impact high. Recorded consequences include modifying files, executing unauthorized code or commands, bypassing protection mechanisms, gaining privileges or assuming identity, reading and modifying memory, and crashing the device.
How to find it
No detection methods are listed. Verify token generation by enumerating every agent, its intended trust level, and the token it actually receives, then compare the two.
How to fix it
- Assign a unique token to each agent according to its privileges.
- Verify token generation logic in simulation and review, including unusual agents and reset paths.
- Confirm that destination agents act only on tokens that match the expected trust level.
Quick check
Why is a wrong token worse than a missing one?
A missing token can be denied by default. A wrong token looks valid, so the system grants access it should have refused.
