[common] Refactor Protected Files, part 7

[dimakuv]

Description of the changes

This commit refactors PF code without changing functionality (part 7 in a series of commits). In particular, this commit renames:

• "key_id" to a more correct "nonce"
• "encrypted_part_plain" to "metadata_decrypted" (i.e. decrypted conents of the encrypted part of the metadata node)
• "file" to a more specific "host_file_handle"
• "file_metadata" to a more correct "metadata_node"
• "root_mht" to a more specific "root_mht_node"

This PR supersedes #1894.

How to test this PR?

N/A.

---

This change is 

[mkow]

Reviewed 5 of 5 files at r1, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (2 more required), not enough approvals from different teams (1 more required, approved so far: Intel) (waiting on @dimakuv)

---

common/src/protected_files/protected_files_format.h line 57 at r1 (raw file):

    uint8_t    major_version;
    uint8_t    minor_version;
    pf_nonce_t metadata_key_nonce;

Key nonce? Nonce is unrelated to the key itself, it sounds confusing to me. Maybe "cipher_nonce"?

---

common/src/protected_files/protected_files_internal.h line 26 at r1 (raw file):

    metadata_node_t metadata_node; // plaintext and encrypted metadata from storage (bounce buffer)
    metadata_encrypted_t metadata_decrypted; // contains file path, size, etc.

Maybe decrypted_metadata? Sounds more natural. But I can also live with the current version if you don't like the proposal.

[dimakuv]

Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (2 more required), not enough approvals from different teams (1 more required, approved so far: Intel) (waiting on @mkow)

---

common/src/protected_files/protected_files_format.h line 57 at r1 (raw file):

Previously, mkow (Michał Kowalczyk) wrote…

Key nonce? Nonce is unrelated to the key itself, it sounds confusing to me. Maybe "cipher_nonce"?

What I'm trying to convey here is "the nonce with which the key to encrypt the metadata node is derived". I don't think metadata_cipher_nonce conveys this message better.

What about one of the following:

• metadata_nonce (a bit vague, but simple and uniform with metadata_gmac)
• metadata_nonce_for_key_derivation (super long but very specific)
• metadata_nonce_derive (same as before, but shorter but more cryptic)

---

common/src/protected_files/protected_files_internal.h line 26 at r1 (raw file):

Previously, mkow (Michał Kowalczyk) wrote…

Maybe decrypted_metadata? Sounds more natural. But I can also live with the current version if you don't like the proposal.

I like the current version better, for two reasons. First, metadata_decrypted conveys the meaning better (at least for me): it is the metadata node's contents, but decrypted. Whereas decrypted_metadata sounds to me more vague, as "some metadata which was decrypted".

Second, we have metadata_ prefix to denote everything that relates to the metadata node, not just some other metadata. So I'd like to keep all related-to-metadata-node variables/fields with this prefix.

[mkow]

Reviewable status: all files reviewed, 1 unresolved discussion, not enough approvals from maintainers (1 more required) (waiting on @dimakuv)

---

common/src/protected_files/protected_files_format.h line 57 at r1 (raw file):

Previously, dimakuv (Dmitrii Kuvaiskii) wrote…

What I'm trying to convey here is "the nonce with which the key to encrypt the metadata node is derived". I don't think metadata_cipher_nonce conveys this message better.

What about one of the following:

• metadata_nonce (a bit vague, but simple and uniform with metadata_gmac)
• metadata_nonce_for_key_derivation (super long but very specific)
• metadata_nonce_derive (same as before, but shorter but more cryptic)

Ah, I see. Ok, let's keep the current name then, I don't like the others.

[kailun-qin]

Reviewed 5 of 5 files at r1, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (1 more required) (waiting on @dimakuv)

---

-- commits line 7 at r1:
-> contents

Code quote:

conents

[dimakuv]

Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @kailun-qin)

---

-- commits line 7 at r1:

Previously, kailun-qin (Kailun Qin) wrote…

-> contents

Done.

[kailun-qin]

Reviewed all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved

[dimakuv]

Jenkins, retest Jenkins-SGX-22.04 please (very slow node, unrelated)

[mkow]

Reviewed all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved