Re: [PATCH] crypto: caam - fix ERA retrieval function

2014-02-07 Thread Kim Phillips
On Thu, 6 Feb 2014 11:41:28 +0200 Horia Geantă wrote: > On 2/6/2014 10:27 AM, Alex Porosanu wrote: > > Signed-off-by: Alex Porosanu > > --- > > drivers/crypto/caam/ctrl.c | 36 ++-- > > drivers/crypto/caam/ctrl.h | 2 +- > > 2 files changed, 11 insertions(+),

Re: [RFC PATCH 0/3] support for interleaving in generic chaining modes

2014-02-07 Thread Herbert Xu
On Fri, Feb 07, 2014 at 10:51:08AM +0100, Ard Biesheuvel wrote: > > So i get your point: I will have a go at passing the entire buffer to > ecb(%s), and then wrapping the chaining mode around that. > I guess you don't think the (avoidable) additional memory usage is a concern? You mean the tempor

Re: [RFC PATCH 0/3] support for interleaving in generic chaining modes

2014-02-07 Thread Ard Biesheuvel
On 7 February 2014 10:44, Herbert Xu wrote: > On Fri, Feb 07, 2014 at 10:42:14AM +0100, Ard Biesheuvel wrote: >> >> Another example is bit sliced AES like the implementation in >> arch/arm/crypto. It is 45% faster than the ordinary ARM asm >> implementation, but its natural chunk size is 8 blocks.

Re: [RFC PATCH 0/3] support for interleaving in generic chaining modes

2014-02-07 Thread Herbert Xu
On Fri, Feb 07, 2014 at 10:42:14AM +0100, Ard Biesheuvel wrote: > > Another example is bit sliced AES like the implementation in > arch/arm/crypto. It is 45% faster than the ordinary ARM asm > implementation, but its natural chunk size is 8 blocks. Passing fewer > blocks hurts performance, while pa

Re: [RFC PATCH 0/3] support for interleaving in generic chaining modes

2014-02-07 Thread Ard Biesheuvel
On 7 February 2014 10:23, Herbert Xu wrote: > On Fri, Feb 07, 2014 at 08:30:26AM +0100, Ard Biesheuvel wrote: >> >> I agree that it would be trivial for cbc(%s) to probe for ecb(%s) >> before settling on using plain '%s. >> But how to probe for an /accelerated/ ecb(%s), i.e., how to avoid >> using

Re: [RFC PATCH 0/3] support for interleaving in generic chaining modes

2014-02-07 Thread Herbert Xu
On Fri, Feb 07, 2014 at 08:30:26AM +0100, Ard Biesheuvel wrote: > > I agree that it would be trivial for cbc(%s) to probe for ecb(%s) > before settling on using plain '%s. > But how to probe for an /accelerated/ ecb(%s), i.e., how to avoid > using the generic ecb(%s) which adds nothing but overhea

Re: [PATCH] crypto: caam - Fix job ring discovery in controller driver

2014-02-07 Thread Horia Geantă
On 1/24/2014 11:03 AM, Nitesh Lal wrote: The SEC Controller driver creates platform devices for it's child job ring nodes. Currently the driver uses for_each_compatible routine which traverses the whole device tree to create the job rings for the platform device. The patch changes this to search