LiveActive security incident?Get immediate response
CWE Reference

CWE-1268: Policy Privileges are not Assigned Consistently… | Glexia

CWE-1268 (Policy Privileges are not Assigned Consistently Between Control and Data Agents) weakness overview with consequences, detection methods, mitigations,…

Release 4.20weaknessDraft

Glexia's Take · Automated analysis

CWE-1268: Policy Privileges are not Assigned Consistently Between Control and Data Agents

Policy Privileges are not Assigned Consistently Between Control and Data Agents represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.

Executive Impact

  • Confidentiality Integrity Availability Access Control: Modify Memory Read Memory DoS: Crash, Exit, or Restart Execute Unauthorized Code or Commands Gain Privileges or Assume Identity Bypass Protection Mechanism Read Files or Directories Reduce Reliability

Developer Pattern

CWE-1268 is the kind of defect developers can usually prevent with explicit validation, safer framework defaults, and tests that exercise hostile input or unsafe state transitions.

Automation confidence

high confidence from CWE-1268, 4.20.

Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.

Official CWE Definition

CWE-1268: Policy Privileges are not Assigned Consistently Between Control and Data Agents

The product's hardware-enforced access control for a particular resource improperly accounts for privilege discrepancies between control and write policies.

Integrated circuits and hardware engines may provide access to resources (device-configuration, encryption keys, etc.) belonging to trusted firmware or software modules (commonly set by a BIOS or a bootloader). These accesses are typically controlled and limited by the hardware. Hardware design access control is sometimes implemented using a policy. A policy defines which entity or agent may or may not be allowed to perform an action. When a system implements multiple levels of policies, a control policy may allow direct access to a resource as well as changes to the policies themselves. Resources that include agents in their control policy but not in their write policy could unintentionally allow an untrusted agent to insert itself in the write policy register. Inclusion in the write policy register could allow a malicious or misbehaving agent write access to resources. This action could result in security compromises including leaked information, leaked encryption keys, or modification of device configuration.

Type
weakness
Abstraction
Base
Status
Draft
Source
MITRE CWE definition

Developer And Remediation Guidance

How teams prevent and detect this weakness

Causes

  • Consider a system of seven registers for storing and configuring an AES key for encryption or decryption. Four 32-bit registers are used to store a 128-bit AES key. The names of those registers are AES_ENC_DEC_KEY_0, AES_ENC_DEC_KEY_1, AES_ENC_DEC_KEY_2, and AES_ENC_DEC_KEY_3. Collectively these are referred to as the AES Key registers. Register Field description AES_ENC_DEC_KEY_0 AES key [0:31] for encryption or decryption Default 0x00000000 AES_ENC_DEC_KEY_1 AES key [32:63] for encryption or decryption Default 0x00000000 AES_ENC_DEC_KEY_2 AES key [64:95] for encryption or decryption Default 0x00000000 AES_ENC_DEC_KEY_3 AES key [96:127] for encryption or decryption Default 0x00000000 Three 32-bit registers are used to define access control for the AES-key registers. The names of those registers are AES_KEY_CONTROL_POLICY, AES_KEY_READ_POLICY, and AES_KEY_WRITE_POLICY. Collectively these registers are referred to as the Policy registers, and their functions are explained next. The AES_KEY_CONTROL_POLICY register defines which agents can write to the AES_KEY_READ_POLICY or AES_KEY_WRITE_POLICY registers. The AES_KEY_READ_POLICY register defines which agents can read the AES-key registers. The AES_KEY_WRITE_POLICY register defines which agents can write the AES key registers. The preceding three policy registers encode access control at the bit level. Therefore a maximum of 32 agents can be defined (1 bit per agent). The value of the bit when set (i.e., "1") allows the respective action from an agent whose identity corresponds to the number of the bit. If clear (i.e., "0"), it disallows the respective action to that corresponding agent. For example, if bit 0 is set to "1" in the AES_KEY_READ_POLICY register, then agent 0 has permission to read the AES-key registers. Consider that there are 4 agents named Agent 1, Agent 2, Agent 3, and Agent 4. For access control purposes Agent 1 is assigned to bit 1, Agent 2 to bit 2, Agent 3 to bit 3, and Agent 4 to bit 4. All agents are trusted except for Agent 3 who is untrusted. Also consider the register values in the below table. The AES_KEY_CONTROL_POLICY register value is 0x00000018. In binary, the lower 8 bits will be 0001 1000, meaning that: Bits 3 and 4 are set, thus Agents 3 and 4 will have write access to AES_KEY_READ_POLICY or AES_KEY_WRITE_POLICY. All other bits are clear, hence agents other than 3 and 4 will not have access to write to AES_KEY_READ_POLICY or AES_KEY_WRITE_POLICY. The AES_KEY_READ_POLICY register value is 0x00000002. In binary, the lower 8 bits will be 0000 0010, meaning that: Bit 1 is set, thus Agent 1 will be able to read the AES key registers. The AES_KEY_WRITE_POLICY register value is 0x00000004. In binary, the lower 8 bits will be 0000 0100, meaning that: Bit 2 is set, thus Agent 2 will be able to write the AES Key registers. The configured access control policy for Agents 1,2,3,4 is summarized in the table below. Agent Read Write Control Agent 1 Allowed Not Allowed Not Allowed Agent 2 Not Allowed Allowed Not Allowed Agent 3 Not Allowed Not Allowed Allowed Agent 4 Not Allowed Not Allowed Allowed At this point, Agents 3 and 4 can only configure which agents can read AES keys and which agents can write AES keys. Agents 3 and 4 cannot read or write AES keys - just configure access control. Now, recall Agent 3 is untrusted. As explained above, the value of the AES_KEY_CONTROL_POLICY register gives agent 3 access to write to the AES_KEY_WRITE_POLICY register. Agent 3 can use this write access to add itself to the AES_KEY_WRITE_POLICY register. This is accomplished by Agent 3 writing the value 0x0000000C. In binary, the lower 8 bits are 0000 1100, meaning that bits 2 and 3 will be set. Thus, Agent 3 will be given the ability to write to the AES Key registers. If Agent 3 does not set bit 2, this would prevent Agent 2 from having write access, which could cause operational failures that could expose the attack (or be used for a denial of service). If the AES_KEY_CONTROL_POLICY register value is 0x00000010, the lower 8 bits will be 0001 0000. This will give Agent 4, a trusted agent, write access to AES_KEY_WRITE_POLICY, but Agent 3, who is untrusted, will not have write access. The Policy register values should therefore be as follows:

Remediation

  • Architecture and Design Implementation: Access-control-policy definition and programming flow must be sufficiently tested in pre-silicon and post-silicon testing.

Detection

  • Code review
  • SAST
  • DAST
  • Focused regression tests

Mappings

Related CVEs, CWEs, and ATT&CK context

Related CWEs

Related CVEs

Related CVE mappings appear after CVE records are cross-indexed.

Open CWE CVE mapping

ATT&CK Relevance

ATT&CK relevance is shown only when reviewed or responsibly inferred.