Make it easier for vendors to declare the Package Manager flag
that matches the version of KeyMint that their device implements.
Test: TreeHugger
Bug: 381125092
Change-Id: I13f781836b4b06c43811b315ca78da076ecd928f
Changes:
1) The aidl_format pre-upload hook complains about some line wrapping.
2) ASN.1 comments begin with "--". This only makes a difference in this case
if a vendor copy-pastes the schema in these comments somewhere that
expects valid a ASN.1 schema. It's also nice to be consistent.
As a drive-by improvement, I also added comments for all the fields in
KeyDescription. The statements about how attestationSecurityLevel and
keymintSecurityLevel must match are tested in VTS (and CTS).
Bug: 381383980
Test: n/a, comment update
Change-Id: I6a331fd3ed18e4a54ff8d34e96ae7dc62e3e3a0c
adds a test to VtsRemotelyProvisionedComponentTests that generates 256
P256 keypairs and adds functionality that makes sure the points (public
keys) lie on the curve, are not infinity, and are not equal to the base
point (generator).
Test: atest VtsHalRemotelyProvisionedComponentTargetTest
Bug: 327026543
Change-Id: I4002b9e5ed5dbc366a14ff5c16e2c4cfd7d14485
Updating the comment to match the test expectation. The attested Verified
Boot key should contain the user-set root of trust if the Verified Boot
state is "SelfSigned" (i.e. "yellow").
Bug: 376832222
Test: n/a (comment update)
Change-Id: Ie3c43157f05e40fa42b107768af561644ff5d30f
There are a lot of upcoming requirements (see below) that place
restrictions on the CSR and the DICE chain. This is a first step to
make some of the infrastructure necessary to utilize
the already-existing functionality inside hwtrust.
Bug: 372843164
Bug: 376297620
A lot of parsing happens in `remote_prov_utils`. It would be nice to
have all parsing and validation occur in hwtrust.
Test: atest VtsHalRemotelyProvisionedComponentTargetTest
atest libkeymint_remote_prov_support_test
Change-Id: I52383c0c62a6bb73554fd0771d2e7a8b366246c0
`android.hardware.security.keymint-service.nonsecure` is part of
`com.android.hardware.keymint.rust_nonsecure` apex. However, listing
android.hardware.hardware_keystore.xml in `required` of the binary
installs two copies of this permission file on device - one inside the
apex, and another it the /vendor/etc/permission (outside the apex).
With this CL, the latter will no longer be installed. This CL is
motivated by having the vendor.img building with soong as part of mk->bp
migration.
Test: m com.android.hardware.keymint.rust_nonsecure
Test: deapexer list
out/target/product/vsoc_x86_64/vendor/apex/com.android.hardware.keymint.rust_nonsecure.apex
# verified that etc/permissions/android.hardware.hardware_keystore.xml
is present
Bug: 374371755
Bug: 376110962
Change-Id: I15ba92c81296f869277940767ff3a76c28214c18
key_transform field not being reset during DeleteKey() causing
keygen test to skip the key generation.
Bug: 375151780
Test: VtsAidlKeyMintBenchmarkTest
Change-Id: I10428dc9f84831a44917fb180d16bb97e2db2257
Revert submission 3312283-2024-10-18-schuffelen-sepolicy
Reason for revert: Droidmonitor created revert due to b/375059051. Will be verifying through ABTD before submission.
Reverted changes: /q/submissionid:3312283-2024-10-18-schuffelen-sepolicy
Change-Id: Icebd34eb9e2941a6480d9943986a104cc5249729
add a flag to verifyFactoryCsr that enforces the presence
of UDS certs in the presented CSR
Bug: 366147625
Test: m rkp_factory_extraction_tool
Test: m librkp_factory_extraction_test
Test: m libkeymint_remote_prov_support_test
Test: atest system/security/provisioner
Test: atest hardware/interfaces/security/keymint/support
Change-Id: I1b6c85b858d76b1ea418400342c89553cbd4bc1d
