Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[fuse_ctrl] Generate partitions mmap #47

Draft
wants to merge 8 commits into
base: main
Choose a base branch
from

Conversation

andrea-caforio
Copy link

Regenerate partitions and assign the fuses (as documented here).

Copy link

linux-foundation-easycla bot commented Jan 13, 2025

CLA Not Signed

@andrea-caforio andrea-caforio force-pushed the fuse_ctrl_partitions branch 5 times, most recently from 40ddc68 to 24e06ae Compare January 15, 2025 10:30
{
name: "FIELD_ENTROPY_0",
inv_default: "<random>",
size: "8"
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we inline the documentation for each field? You can probably use desc: variables. The documentation should match the Caliptra spec.

variant: "Unbuffered",
size: "64", // in bytes
absorb: true,
secret: false,
sw_digest: true,
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@anjpar @ekarabu, when a partition is set to sw_digest: true, I believe there is no hardware digest checking taking place. We should document in the guidelines that the ROM is expected to check the digest before consuming the values.

Alternatively, you may want to add a hash digest field to this partition so that ROM can compare the hash after loading the data into SRAM. If you opt for the hash approach, you can use either SHA256 or SHA384.

This is a defense in depth recommendation. Without this, we are relying on the OTP ECC integrity checks.

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This comment also applies to other SW OTP partitions that contain critical values.

Another option I forgot to mention is that you can use multibit encodings for critical fields. If all fields are hardened, then the digest check becomes optional.

name: "UDS_SEED",
size: "48"
}
name: "MANUF_TOKEN",
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: It may be good to clarify this a bit more. Is this an unlock token? If so, can we call it MANUF_UNLOCK_TOKEN? The same applies for all the tokens below.

For cases where we have authorization tokens, it would be good to encode that in the name.

This will make the programming interface a bit more ergonomic.

read_lock: "CSR",
key_sel: "NoKey",
// This is a strike counter, hence we need to disable ECC integrity for this to work.
// Integrity is handled at a higher level by SW as described below.
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Integrity is handled at a higher level by SW as described below.

The description below is missing.

bkout_type: false,
items: [
{
name: "FMC_KEY_MANIFEST_SVN",
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Multi-bit encoding for redundancy is recommended. So 32-bits may result in 16 SVN versions. We should check if this is sufficient.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants