Secure boot and measured boot explained
Secure boot enforces: each stage verifies the next against a trusted key and refuses to run anything that fails. Measured boot records: each stage hashes the next into a tamper-evident register so a remote party can see what ran. They answer different questions and are frequently confused.
Secure boot enforces
The root of trust verifies the first mutable stage against a key fused at manufacture. That stage verifies the next, and so on. A signature that fails to verify stops the boot.
The property gained is that only code signed by the holder of the signing key runs. The property not gained is any statement about whether that code is correct.
Measured boot records
Each stage hashes the next and extends that hash into a register that can be appended to but never rewritten — a PCR in TPM terms. The boot proceeds regardless of what the measurements are.
At the end, the register values are a fingerprint of exactly what ran. A remote party can ask for them, signed, and decide whether to trust the device. Measured boot does not prevent anything; it makes what happened undeniable.
Why both exist
- Secure boot alone stops unsigned code, and cannot tell you which signed version booted — including an older signed image with a known vulnerability.
- Measured boot alone tells you precisely what ran, after it has already run.
- Together they stop unsigned code and produce evidence of which signed code ran, which is what remote attestation actually needs.
Where the practical difficulty is
Not the mechanism, but key management and revocation. If a signing key is compromised, every device trusting it needs updating, and the update has to be signed by something. If an older signed image is vulnerable, secure boot happily runs it unless anti-rollback counters are in place.
Anti-rollback is where many implementations are thin, and it is a question worth asking of any device claiming secure boot.
The entropy connection
Attestation requires the device to sign its measurements with a device-unique key. That key was generated at some point, from some source, usually at first provisioning. If it was predictable, the attestation is forgeable and the entire chain reports on a device an attacker can impersonate — which is the provisioning problem.
Keep reading
- Foundations
Device identity and secure provisioning
A device identity is a key pair generated inside the device and bound to it, which the device uses to prove what it is. Its strength is fixed at the moment of generation: entropy available at first provisioning determines whether the identity is unique and unpredictable for the device's entire life.
- Foundations
What is a hardware root of trust, and what it actually protects
A hardware root of trust is an immutable anchor — typically a key and verification logic in silicon — that the rest of a system's trust chain is built on. It establishes what code was allowed to run and protects the key material underneath, and it cannot verify anything about its own correctness.
- Foundations
Hardware or software root of trust
A software root of trust relies on the platform's existing protections to keep its key secret; a hardware one puts the key somewhere software cannot reach. Software is a defensible choice when the threat model excludes physical access and the platform already isolates well, and it stops being one when either assumption fails.