[keymgr_dpe] Replace keymgr in EG - PR8 crypto driver#30677
Draft
rroth-lowrisc wants to merge 12 commits into
Draft
[keymgr_dpe] Replace keymgr in EG - PR8 crypto driver#30677rroth-lowrisc wants to merge 12 commits into
rroth-lowrisc wants to merge 12 commits into
Conversation
Split the otp_keymgr_key interface into its respective components. Future version of the keymgr(_dpe) may receive the information from different IP's rather than combined from the otp_ctrl. To avoid a circular dependencies the DevIdWidth instantiated in keymgr_dpe_pkg replaces the otp_ctrl_pkg::DeviceIdWidth. As the keymgr still expects the combined message the earlgrey top manually combines the key material. This fix will be removed when the keymgr is replaced by the keymgr_dpe. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit updates the `otp_ctrl` dv to reflect the new interface between the `keymgr` and the `otp_ctrl`. The secret / seed verification inside the otp_ctrl scoreboard template are no longer dynamically generated. The reason is that the previously shared port signal has been split into its respective components, which does not interacts well with the dynamic generation anymore. However, adding a secret to the otp now results in a port-list change anyways the resulting overhead is deemed acceptable. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit removes all otp_ctrl dependency in the keymgr_dpe dv. This change ensures that the dv architecture accurately reflects the dependency in the RTL implementation. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit improves the block level dv by: - Introducing more verbose messages - Latching the correct DEFAULT_UDS_POLICY - Prevent ambiguous state after the disable operation by restricting any advance call Signed-off-by: Raphael Roth <rroth@lowrisc.org>
The scoreboard loads the UDS directly from the DUT via interface. The problem is the keymgr_dpe initializes the slot for the UDS with randomness. The initally loaded UDS is xored with the randomness already present in the destination slot as countermeasure for SCA. The scoreboard however can currently not recreate this randomness thus the backdoor load. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
Introduction of the third bootstage defined in the opentitan identities-and-root-keys (https://opentitan.org/book/doc/security/specs/identities_and_root_keys/index.html) strategy. The `creator_seed` is now consumed by BootStageOwnerInt rather than by BootStageCreator. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
Author
|
DO NOT MERGE THIS PR This PR cannot be merged until the following PR passes the CI:
The introduced keymgr_dpe drivers are not yet referenced in any build system until the TOP integration PR. Therefore none of the CI checks really verifies anything in this PR. The only reason for this PR is to break down the review effort into several smaller chunks of work along "logical" lines of code spaces. |
nasahlpa
approved these changes
Jul 7, 2026
nasahlpa
left a comment
nasahlpa
left a comment
Contributor
There was a problem hiding this comment.
Thanks Raphael, I had a look into the last three commits.
The first commit I already have reviewed in another repo.
The other two commits LGTM!
| @@ -0,0 +1,212 @@ | |||
| // Copyright lowRISC contributors (OpenTitan project). | |||
Contributor
There was a problem hiding this comment.
I think you should be able to already add this test to the BUILD file but just mark it as broken such that it is not build and executed.
nasahlpa
approved these changes
Jul 7, 2026
nasahlpa
left a comment
nasahlpa
left a comment
Contributor
There was a problem hiding this comment.
Thanks Raphael - I had a look into the last three commits:
- The first commit I already have reviewed in another repo
- The last two commits look good to me, they streamline the cryptolib drivers with upstream as well as adding a new test
This commit allows for a limited parametrization by the top without having a fully templated IP. The top can define the following: - Number of bootstages (either two or three) - Number of ROM digest values - Number of HW slots instanciated (up to 8 slots) The block level dv is extended to test two different configurations, called earlgrey and darjeeling. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
Add three missing features to the dif of the keymgr_dpe. - The disable function to transition the keymgr_dpe into the disabled state - The configuration function to set the re-seeding interval of the keymgr_dpe - Any sideloaded key on the sideload interface can now be cleared by sw Add an empty `dif_keymgr_dpe_unittest.cc` file to avoid triggering error during the top integration, tracked in Issue lowRISC#30609. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit introduces a first version of the keymgr_dpe sw in the silicon creator lib. It covers the basic keymgr_dpe features. Add source / header file **without** linking them in the appropriate *BUILD* file as some dependencies do not exist yet! Add an empty keymgr_dpe_unittest.cc file to avoid triggering error during the top integration, tracked in Issue lowRISC#30609. Signed-off-by: Raphael Roth <rroth@lowrisc.org>
To ensure backwards compatibility, the file has been split into two sections due to the different boot processes. The first part covers the boot process for darjeeling and the second part covers the boot process for earlgrey. This `if define ...` should either be replaced with two testutils files for the keymgr_dpe (linked by Bazel according to the top) or with a finer granularity of defines (e.g. function-wise) Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit introduces a first version of the keymgr_dpe sw in the crypto lib. It covers the basic keymgr_dpe features. Add source / header file **without** linking them in the appropriate BUILD file as some dependencies do not exist yet! This commit contains the backported changes from the following git commit hashes (changes done in `keymgr` but relevant for `keymgr_dpe` too): - 9dcbb05 - 4afe7cd - ab8c166 - b10d2fe - e587303 - 068c37e Signed-off-by: Raphael Roth <rroth@lowrisc.org>
This commit introduces the test for the `keymgr_dpe` driver inside the crypto lib. It covers the basic `keymgr_dpe` features. Add source / header file **without** linking them in the appropriate BUILD file as some dependencies do not exist yet! Signed-off-by: Raphael Roth <rroth@lowrisc.org>
rroth-lowrisc
force-pushed
the
keymgr_dpe_PR8_cryptolib
branch
from
e167c75 to
910b63b
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
PR8 - crypto driver
This PR is the eight in a series which will replace the keymgr with the kemgr_dpe in earlgrey. The replacement was approved by this RFC.
Merge-Dependencies
Relevant Commits
Last
threetwo commitDescription
This PR will add a first version of the crypto lib driver for the
keymgr_dpe. It adds the basic functionality to interact with thekeymgr_dpe. A first draft version of thekeymgr_dpe_testis already present, however some commits needs to be backported from themasterbranch first.It is not yet included in the bazel build system as several dependencies are missing until the top integration is finished.