The BinGuestSecret type provides no benefits. The public GuestSecret
struct can handle everything. Therefore, move the two functions from bin
to the non-bin variant. While at it, use a struct to define the binary
structure instead of copy numbers to some positions in a Vec. This
simplifies the addition of further secret types.
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Reviewed-by: Julian Ruess <julianr@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
This streamlines and unifies the use and (de)serialization of structs
using a secret id. As a bonus, the hidden `for_pv` module is not longer
needed.
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Reviewed-by: Julian Ruess <julianr@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
The content of `pv/src/uvsecret/uvc.rs` was moved with
9b51b8b882 ("rust/pv: Refactor pv crate") to pv_core.
The content was unused, but the file was not deleted.
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Reviewed-by: Julian Ruess <julianr@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Fix clippy waring `warning: very complex type used.` by introducing a
new struct containing the tuple, that was returned before.
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Reviewed-by: Julian Ruess <julianr@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
New IBM signing keys will have Armonk as locality in the subject.
Ensure that CRLs with Poughkeepsie as issuer locality are still
discovered if they are signed with the signing keys private key.
Also, drop the check for issuer/subject comparison and only rely on
validity period and cryptographic signatures.
Reviewed-by: Christoph Schlameuss <schlameuss@linux.ibm.com>
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
* Get rid of Mockito
* create certs with AKID
* simplify things in the `create_certs.py` script
Signed-off-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Add a function to read a private key in PEM or DER format.
While at it, fix some documentation issues in read_{certs, crls}.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
User-data can be added to add-secret requests. User-data can be unsigned
or containing a signature of the request using a custom private key.
The user-data is placed at offset 0x218 and 0x200 bytes long.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Add four new user-data types an add-secret request could have:
Unsigned, Signed(EcSECP521R1, Rsa(2048, 3072))
As the user-data enum was not marked as non-exaustive this might be a
breaking change for users. (Not for any crate in this repo though).
The addition of such user-data is provided by following patches.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Recent changes makes it reasonable to update the version string of pv
and pv_base to 1.0.0. Recent changes introduced some non-backwards
compatible changes, like move some Error definition from pv to the new
pv_core crate. Also, the pv crates seem to be in a good shape to be
considered released.
Therefore, move up the Semver to 1.0.0.
See also:
https://doc.rust-lang.org/cargo/reference/specifying-dependencies.html
NOTE: As these crates are not published to `crates.io` as of now, the
version number is meaningless, as cargo will not pull them from
anywhere and just use the files provided by this repository.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Update the zerocopy dependency from v0.6.* to the non forward compatible
v0.7.*. Incompatible changes are the requirment of the FromZeroes trait
for FromBytes trait and the non-default derive feature which this
patches handles as well.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Use a static Mockito server instead of creating multiple on-the-fly
during testing. Add the `mockito_server_available` test to quickly
verify that the mockito server could be constructed. Relocate the
verifier API test to the test folder as it tests the public API. Also,
make the Debug impl for CertVerifier public. The test relocation
requires this. Before, there could be a test local implementation of the
Debug impl as it was in the same crate.
While at it, get rid of the lazy_static crate in favor of the once_cell
crate. As the new std lib implementation is very near to the one from
once cell. The project will switch to the std impl when v1.70 is
available in all supported distros. See:
https://doc.rust-lang.org/std/sync/struct.OnceLock.html
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Replace all libc references to rust-std references if available.
This eliminates the need to include libc in the pv crate.
However, pv_base still refers to libc::ioctl and libc::ENOTTY.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
The rust std lib already provides functionality to compare two slies.
Replace all `memeq` invocations with == and remove the `memeq` function.
As a side effect this eliminates some unsafe code in this crate.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Big refactoring patch of the pv crate. The main reason behind this
refactoring is to simplify testing and maintaining the pv crate while
keeping OpenSSL/libcurl dependencies optional. Using crate features
increases the number of targets that have to be tested. This refactoring
eliminates the use of features by splitting the functionality of pv into
a use OpenSSL and no-use-OpenSSL crate.
Split off some code from the pv crate into a pv_core crate. pv requires
pv_core and reexports all symbols. pv_base contains all code from former
pv that does not use OpenSSL or libcurl functionalities. The refactored
pv crate contains functionalities to generate requests and validate host
key documents. All features from pv are dropped as they are not needed
anymore and to streamline the codebase for easier use and testing. While
at it fix some documentation issues.
Users (pvsecret & pvapconfig) have next to no code change, besides the
different import of the crate.
Acked-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
A function does the job as well. This improves code readability and
useability. While at it remove the implicit Buffer wrapper. Users are
currently not use the benefits of a buffered write. Also, streamline the
write_out helper function.
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
Adds getter for SecretList and SecretEntry.
Adds enum to represent secret types.
Add Iterator functionality for SecretList.
While at it, make the datatype of the capacity of the list transparent
for users.
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
The crate dependencies were a bit to slack. Due to the rust dependency
resolver's strategy of always selecting the latest version this never
lead to any issues.
This has no impact on the workspaces Cargo.lock
Reviewed-by: Marc Hartmayer <marc@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
A workspaces simplifies the build and packaging process significantly.
All build artifacts and binaries are now built in a single location
(e.g., rust/target/release/*), and a unified dependency resolution is
used. Hence one Cargo.lock for all crates at rust/Cargo.lock.
Closes: https://github.com/ibm-s390-linux/s390-tools/issues/156
Reviewed-by: Marc Hartmayer <marc@linux.ibm.com>
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Fix a valgrind finding. Fix an invalid read/write of one byte after the
actual struct to clear. Not fixing this may result in a illegal write or
memory corruption of the program. Fortunately, for the actual only user,
pvsecret this is not the case.
Fixes: c6f621d0 ("rust: Add library for pv tools")
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Add a `pv` crate that bundles useful functions and structs for creating
requests like `Attestation`, `Add Secret`, or even `Boot` a.k.a.
Secure Execution Image.
Note pv includes a subcrate `openssl_extensions` that (temporarily)
bundles some needed `openssl-rust` functionalities that are not
upstream yet. The plan is to remove these, when they become
upstream.
The pv crate has multiple features:
* request - code to generate requests
* uvsecret - code to access the UV-secret api
with request enabled also generating requests is
possible
Signed-off-by: Steffen Eiden <seiden@linux.ibm.com>
Acked-by: Jan Höppner <hoeppner@linux.ibm.com>
Acked-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
