By moving the firmware-loading to the transform initialisation,
the driver may be compiled in.
Notify the user if the firmware does not support AES, or even
no crypto at all.
Signed-off-by: Christian Hohnstaedt <[EMAIL PROTECTED]>
---
drivers/crypto/ixp4xx_crypto.c | 44 +++
Hi:
The newly added tcrypt infrastructure didn't help the simple cipher
users since we only test skcipher transforms. Most notably this
would mean that ansi_cprng would get an untested cipher as its
engine which is unacceptable. So here's a patch to fix that.
commit 413ffa5070e6c3c26e9c5cb59037
On Sun, Aug 17, 2008 at 10:20:47PM +1000, Herbert Xu wrote:
>
> There is also a new global rng object for everyone to use rather
> than having it allocated by every user separately.
And here is the new IV generator patch to use the global RNG object.
commit 1da64c8d0d573dcb1b67c42d41c0adf511f81d
On Thu, Aug 14, 2008 at 10:23:31PM +1000, Herbert Xu wrote:
>
> And here is the IV generator patch on top.
This also dead-locked if built as modules. I've split out the
IV generators from crypto_blkcipher so it doesn't interfere with
cryptomgr.
commit 106475b7c368794f22493fc67d98b874eb0f9383
Au
On Thu, Aug 14, 2008 at 08:21:37AM -0400, Neil Horman wrote:
>
> Looks great to me. Thanks Herbert!
Unfortunately it didn't quite work :)
First of all there was a module dependency loop between cryptomgr,
rng and crypto_blkcipher. This is fixed by putting krng in its
own module. Also I made a
On Fri, Jul 04, 2008 at 07:43:15PM +0800, Herbert Xu wrote:
> On Thu, Jul 03, 2008 at 02:03:57PM -0700, Harvey Harrison wrote:
> > Remove the private implementation of 32-bit rotation and unaligned
> > access with byteswapping.
> >
> > As a bonus, fixes sparse warnings:
> > crypto/camellia.c:602:2
