Support for CoVE's Deployment Model 3 #4
Conversation
This file contains 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
Utilize the correct NACL features by testing whether a feature required for NACL exploitation in nested virtualized environments is present. If nested virtualization is not present, use the NACL setup_shared_memory() ABIi. Otherwise, use the entire NACL ABI. Signed-off-by: Wojciech Ozga <[email protected]>
Signed-off-by: Wojciech Ozga <[email protected]>
Detect AIA presence by discovering that the TEE security monitor (TSM) supports AIA capability. If AIA is not present, inject external interrupts using the HVIP register when resuming execution of a virtual processor via the COVH tvm_vcpu_run() call. Signed-off-by: Wojciech Ozga <[email protected]>
Separate the TVM's init and finalized states to allow alternative ways of creating TVMs in future. These alternative TVM creation procedures require different behaviour during CoVE VM initialization, finalization and termination. Signed-off-by: Wojciech Ozga <[email protected]>
Specialize the current TVM creation procedure as a multi-step TVM creation to differentiate from other ways of creating TVMs. Signed-off-by: Wojciech Ozga <[email protected]>
Signed-off-by: Wojciech Ozga <[email protected]>
Support new type of CoVE VM that can be created in a single step. Preload VM pages into memory when creating the boot vcpu and request promotion when running the boot vcpu for the first time. Signed-off-by: Wojciech Ozga <[email protected]>
This patch is still under review and won’t be merged in this cycle. Signed-off-by: Wojciech Ozga <[email protected]>
Support the alternative way of creating TVM by letting the VM to decide when it should be promoted to TVM. The VM uses COVH promote_to_tvm() ABI at the time it wants to become TVM. Reflect this call and the VM architectural state to the TSM. This is an experimental feature, thus the patch is not targeted yet for merge. Signed-off-by: Wojciech Ozga <[email protected]>
Support sharing pages with a hypervisor in an environment where dynamic page conversion is not supported. When a single share_memory_region() call for a memory region that contains multiple 4KiB pages fails, execute multiple 4KiB share_memory_region() calls until the request is completed. Signed-off-by: Wojciech Ozga <[email protected]>
Allocate a placeholder in the kernel memory, so that the TVM image creator can embed there the TVM attestation payload (TAP). Signed-off-by: Wojciech Ozga <[email protected]>
Request isolation from the hypervisor early in the kernel boot process only if the kernel was built with the CONFIG_RISCV_COVE_GUEST_PROMOTE build parameter. Signed-off-by: Wojciech Ozga <[email protected]>
Signed-off-by: Wojciech Ozga <[email protected]>
Retrieve the TVM attestation payload (TAP) address passed by the VMM and reflect it to TEE security monitor (TSM) during the creation of the TEE virtual machine (TVM) via the promote call. Signed-off-by: Wojciech Ozga <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
This patchset introduces support for Deployment Model 3 of the RISC-V CoVE specification.
Subsequent patches add support for different deployment models and new CoVE features introduced in the CoVE spec version 0.6 and 0.7. Specifically, we support deployment models where (1) host runs in the HS mode, (2) TEE security monitor (TSM) does not support RISC-V AIA or dynamic allocation of confidential pages, (3) TVM is created in a single-step via promotion, (4) TVM requires local attestation via TVM attestation payload (TAP).
Please review the changes, and let me know if there are any questions or concerns.