aboutsummaryrefslogtreecommitdiff
path: root/ks.c
AgeCommit message (Collapse)Author
2016-09-01Move in-memory keystore from client to server. Whack with club until compiles.Rob Austein
Fixes for various minor issues found while integrating with sw/stm32. Moving the in-memory keystore (PKCS #11 session objects, etc) from the client library to the HSM was on the near term to-do list in any case, doing it now turned out to be the easiest way to solve one of the build problems.
2016-06-23ks_flash returns all-ones instead of all-zeros for "blank" memory, cope.Rob Austein
This will need refactoring once we have a proper test for whether the HSM is initializing after receiving a fresh software load.
2016-06-10Allow host-side libhal build without access to secure hardware toRob Austein
store unencrypted public keys (we don't allow this for private keys). Yet another screwball feature to support PKCS #11, sigh. Anyway, with this change, mixed-mode builds should work again.
2016-05-25Start cleaning up PIN code.Rob Austein
2016-05-17Increment return count, not pointer to return count. Feh, C.Rob Austein
2016-05-15Tweak keystore API to allow update-in-place, so hal_ks_rename() will work.Rob Austein
2016-05-15Add hal_rpc_pkey_rename(); allow null string as (temporary) key name.Rob Austein
Temporary nature of null string as key name is not enforced by the keystore code, it's just a convention to allow callers to generate a keypair, obtain the public key, hash that to a Subject Key Identifier (SKI), and rename the key using the SKI as the new name. This is a compromise to let us use SKI-based key names in PKCS #11 while keeping the keystore code simple.
2016-05-14Trailing whitespace cleanup.Rob Austein
2016-03-11First round of debugging based on RPC pkey tests: mostly ASN.1Rob Austein
silliness, with a bit of PKCS #1.5 padding silliness for desert.
2016-03-03Initial implementations of ks_get_kek(). Untested, and none of theseRob Austein
are secure (the one in ks_flash.c is a stub, and the others are for cases where we have no secure hardware in which to store the KEK). These are primarily for testing, since in the long run the entire software implementation of AES-keywrap will be replaced by Verilog which never lets software see the unwrapped key. Or so says current theory. For the moment, we just need something that will let us test the rest of the RPC and keystore mechanisms.
2016-02-25RPC over loopback socket, just to work out the mechanics for serialization ↵Paul Selkirk
and dispatch.
2015-12-24More work on PIN/login/logout code. Access control still missing,Rob Austein
committing now so Paul has a chance to look at the current RPC API.
2015-12-23RPC interface to TRNG and (incomplete) PIN code.Rob Austein
2015-12-20RPC server stuff mostly written. Compiles, not yet tested. RPCRob Austein
public key extraction functions on hold pending ASN.1 cleanup.