aboutsummaryrefslogtreecommitdiff
path: root/README.md
diff options
context:
space:
mode:
authorRob Austein <sra@hactrn.net>2015-06-19 13:56:03 -0400
committerRob Austein <sra@hactrn.net>2015-06-19 13:56:03 -0400
commitaaec57f3fc6775bdbf50965ae34e8a797542b7ee (patch)
tree18e7e6befd15052e5c570403f5ad4bfb23d30708 /README.md
parent79d3c33ee5f6fd0cfb24af68357e084e4f1fded0 (diff)
Add README.md.
Diffstat (limited to 'README.md')
-rw-r--r--README.md46
1 files changed, 46 insertions, 0 deletions
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..66669e3
--- /dev/null
+++ b/README.md
@@ -0,0 +1,46 @@
+libhal
+======
+
+This library combines a set of low-level API functions which talk to
+the Cryptech FPGA cores with a set of higher-level functions providing
+various cryptographic services.
+
+There's some overlap between the low-level code here and the low-level
+code in core/platform/novena, which will need sorting out some day,
+but at the time this library forked that code, the
+core/platform/novena code was all written to support a test harness
+rather than a higher-level API.
+
+Current contents of the library:
+
+* Low-level I/O code (EIM and I2C).
+
+* An implementation of AES Key Wrap using the Cryptech AES core.
+
+* An interface to the Cryptech CSPRNG.
+
+* An interface to the Cryptech hash cores, including HMAC.
+
+* An implementation of PBPDF2.
+
+* An implementation of RSA using the Cryptech ModExp core.
+
+* Test code for all of the above.
+
+Most of these are fairly well self-contained, although the PBKDF2
+implementation uses the hash-core-based HMAC implementation.
+
+The major exception is the RSA implementation, which uses an external
+bignum implementation (libtfm) to handle a lot of the arithmetic. In
+the long run, much or all of this may end up being implemented in
+Verilog, but for the moment all of the RSA math except for modular
+exponentiation is happening in software.
+
+The RSA implementation includes a compile-time option to bypass the
+ModExp core and do everything in software, because the ModExp core is
+a tad slow at the moment (others are hard at work fixing this).
+
+The RSA implementation includes optional blinding (enabled by default)
+and just enough ASN.1 code to read and write private keys; the
+expectation is that the latter will be used in combination with the
+AES Key Wrap code.