mbox series

[v2,0/2] crypto: aegis128 SIMD improvements

Message ID 20191014161645.1961-1-ard.biesheuvel@linaro.org (mailing list archive)
Headers show
Series crypto: aegis128 SIMD improvements | expand

Message

Ard Biesheuvel Oct. 14, 2019, 4:16 p.m. UTC 
Refactor the aegis128 code to get rid of indirect calls, and implement
SIMD versions of the init() and final() hooks. This results in a ~2x
speedup on ARM Cortex-A57 for ~1500 byte inputs.

Changes since v1:
- fix missing Sbox loads for plain SIMD on GCC
- fix endianness issue in final_simd() routine

Cc: Ondrej Mosnacek <omosnace@redhat.com>

Ard Biesheuvel (2):
  crypto: aegis128 - avoid function pointers for parameterization
  crypto: aegis128 - duplicate init() and final() hooks in SIMD code

 crypto/aegis128-core.c       | 125 ++++++++++----------
 crypto/aegis128-neon-inner.c |  50 ++++++++
 crypto/aegis128-neon.c       |  21 ++++
 3 files changed, 134 insertions(+), 62 deletions(-)

Comments

Herbert Xu Oct. 25, 2019, 3:18 p.m. UTC | #1 
On Mon, Oct 14, 2019 at 06:16:43PM +0200, Ard Biesheuvel wrote:
> Refactor the aegis128 code to get rid of indirect calls, and implement
> SIMD versions of the init() and final() hooks. This results in a ~2x
> speedup on ARM Cortex-A57 for ~1500 byte inputs.
> 
> Changes since v1:
> - fix missing Sbox loads for plain SIMD on GCC
> - fix endianness issue in final_simd() routine
> 
> Cc: Ondrej Mosnacek <omosnace@redhat.com>
> 
> Ard Biesheuvel (2):
>   crypto: aegis128 - avoid function pointers for parameterization
>   crypto: aegis128 - duplicate init() and final() hooks in SIMD code
> 
>  crypto/aegis128-core.c       | 125 ++++++++++----------
>  crypto/aegis128-neon-inner.c |  50 ++++++++
>  crypto/aegis128-neon.c       |  21 ++++
>  3 files changed, 134 insertions(+), 62 deletions(-)

All applied.  Thanks.