After encountering an issue with cts(cbc(aes)) in the Atmel AES module,
I have used tcrypt and libkcapi's test suite to validate my fix. This led
me to observe some other issues.
This series includes the IV issue correction for the Atmel AES crypto
engine, as well as a secondary issue observed when running
'insmod tcrypt.ko mode=10' and 'insmod tcrypt.ko mode=152' on a SAMA5D2
board.
The libkcapi test suite still reports some problems, for example when the
input data is too large to fit into an intermediate buffer in unaligned
cases. And it seems that with the v4.14 updates, new asynchronous tests
are enabled and report new issues.
Romain Izard (2):
crypto: atmel-aes - properly set IV after {en,de}crypt
crypto: atmel-aes - Reset the controller before each use
drivers/crypto/atmel-aes.c | 50 ++++++++++++++++++++++++++++++++++++----------
1 file changed, 40 insertions(+), 10 deletions(-)
--
2.14.1
