CWE-1326: Missing Immutable Root of Trust in Hardware
View customized information:
For users who are interested in more notional aspects of a weakness. Example: educators, technical writers, and project/program managers.
For users who are concerned with the practical application and details about the nature of a weakness and how to prevent it from happening. Example: tool developers, security researchers, pen-testers, incident response analysts.
For users who are mapping an issue to CWE/CAPEC IDs, i.e., finding the most appropriate CWE for a specific issue (e.g., a CVE record). Example: tool developers, security researchers.
For users who wish to see all available information for the CWE/CAPEC entry.
For users who want to customize what details are displayed.
×
Edit Custom FilterA missing immutable root of trust in the hardware results in the ability to bypass secure boot or execute untrusted or adversarial boot code.
A System-on-Chip (SoC) implements secure boot by verifying or authenticating signed boot code. The signing of the code is achieved by an entity that the SoC trusts. Before executing the boot code, the SoC verifies that the code or the public key with which the code has been signed has not been tampered with. The other data upon which the SoC depends are system-hardware settings in fuses such as whether "Secure Boot is enabled". These data play a crucial role in establishing a Root of Trust (RoT) to execute secure-boot flows. One of the many ways RoT is achieved is by storing the code and data in memory or fuses. This memory should be immutable, i.e., once the RoT is programmed/provisioned in memory, that memory should be locked and prevented from further programming or writes. If the memory contents (i.e., RoT) are mutable, then an adversary can modify the RoT to execute their choice of code, resulting in a compromised secure boot. Note that, for components like ROM, secure patching/update features should be supported to allow authenticated and authorized updates in the field. ![]()
![]() ![]()
![]() ![]()
![]()
![]() Languages Class: Not Language-Specific (Undetermined Prevalence) Operating Systems Class: Not OS-Specific (Undetermined Prevalence) Architectures Class: Not Architecture-Specific (Undetermined Prevalence) Technologies Security Hardware (Undetermined Prevalence) Class: Not Technology-Specific (Undetermined Prevalence) Example 1 The RoT is stored in memory. This memory can be modified by an adversary. For example, if an SoC implements "Secure Boot" by storing the boot code in an off-chip/on-chip flash, the contents of the flash can be modified by using a flash programmer. Similarly, if the boot code is stored in ROM (Read-Only Memory) but the public key or the hash of the public key (used to enable "Secure Boot") is stored in Flash or a memory that is susceptible to modifications or writes, the implementation is vulnerable. In general, if the boot code, key materials and data that enable "Secure Boot" are all mutable, the implementation is vulnerable. Good architecture defines RoT as immutable in hardware. One of the best ways to achieve immutability is to store boot code, public key or hash of the public key and other relevant data in Read-Only Memory (ROM) or One-Time Programmable (OTP) memory that prevents further programming or writes. Example 2 The example code below is a snippet from the bootrom of the HACK@DAC'19 buggy OpenPiton SoC [REF-1348]. The contents of the bootrom are critical in implementing the hardware root of trust. It performs security-critical functions such as defining the system's device tree, validating the hardware cryptographic accelerators in the system, etc. Hence, write access to bootrom should be strictly limited to authorized users or removed completely so that bootrom is immutable. In this example (see the vulnerable code source), the boot instructions are stored in bootrom memory, mem. This memory can be read using the read address, addr_i, but write access should be restricted or removed. (bad code)
Example Language: Verilog
... always_ff @(posedge clk_i) begin
...if (req_i) begin
endif (!we_i) begin
endraddr_q <= addr_i[$clog2(RomSize)-1+3:3];
end else begin
mem[addr_i[$clog2(RomSize)-1+3:3]] <= wdata_i;
end
// this prevents spurious Xes from propagating into the speculative fetch stage of the core
assign rdata_o = (raddr_q < RomSize) ? mem[raddr_q] : '0;
... The vulnerable code shows an insecure implementation of the bootrom where bootrom can be written directly by enabling write enable, we_i, and using write address, addr_i, and write data, wdata_i. To mitigate this issue, remove the write access to bootrom memory. [REF-1349] (good code)
Example Language: Verilog
... always_ff @(posedge clk_i) begin
...if (req_i) begin
endraddr_q <= addr_i[$clog2(RomSize)-1+3:3];
end
// this prevents spurious Xes from propagating into the speculative fetch stage of the core
assign rdata_o = (raddr_q < RomSize) ? mem[raddr_q] : '0;
...
![]()
More information is available — Please edit the custom filter or select a different filter. |
Use of the Common Weakness Enumeration (CWE™) and the associated references from this website are subject to the Terms of Use. CWE is sponsored by the U.S. Department of Homeland Security (DHS) Cybersecurity and Infrastructure Security Agency (CISA) and managed by the Homeland Security Systems Engineering and Development Institute (HSSEDI) which is operated by The MITRE Corporation (MITRE). Copyright © 2006–2025, The MITRE Corporation. CWE, CWSS, CWRAF, and the CWE logo are trademarks of The MITRE Corporation. |