Skip to content

crypto: aria: add ARIA AES-NI/AVX/x86_64/GFNI implementation

Taehee Yoo linux-crypto 4 versions – In review

Get series with b4
b4 am [email protected]

Versions newest first · linked by author and title

  1. v4 Latest3 patches 3 repliesIn review
  2. v33 patches 4 repliesSuperseded
  3. v23 patches 10 repliesSuperseded
Earlier versions · 1
VersionPostedPatchesRepliesStatus
v13 patches2 repliesSuperseded

Changes in v4 from the cover letter

- Fix sparse warning.
- Remove .align statement for .text
- https://lkml.kernel.org/r/[email protected]

Patches in v4 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/3crypto: aria: prepare generic module for optimized implementationscrypto/Makefile crypto/aria_generic.c include/crypto/aria.h+40 -180 replies
2/3crypto: aria-avx: add AES-NI/AVX/x86_64/GFNI assembler implementation of aria cipherarch/x86/crypto/Kconfig arch/x86/crypto/Makefile arch/x86/crypto/aria-aesni-avx-asm_64.S +2 more+1553 -02 replies
3/3crypto: tcrypt: add async speed test for aria ciphercrypto/tcrypt.c+30 -00 replies

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v4 yet.

Read the v4 thread

Cover letter [PATCH v4 0/3] crypto: aria: add ARIA AES-NI/AVX/x86_64/GFNI implementation

The purpose of this patchset is to support the implementation of ARIA-AVX. Many of the ideas in this implementation are from Camellia-avx, especially byte slicing.
Like Camellia, ARIA also uses a 16way strategy.

ARIA cipher algorithm is similar to AES.
There are four s-boxes in the ARIA spec and the first and second s-boxes are the same as AES's s-boxes.
Almost functions are based on aria-generic code except for s-box related function.
The aria-avx doesn't implement the key expanding function. it supports only encrypt() and decrypt().

Encryption and Decryption are actually the same but it should use separated keys(encryption key and decryption key). En/Decryption steps are like below:
1. Add-Round-Key
2. S-box.
3. Diffusion Layer.

There is no special thing in the Add-Round-Key step.

There are some notable things in s-box step.
Like Camellia, it doesn't use a lookup table, instead, it uses AES-NI. There are 2 implementations for that.
One is to use AES-NI and affine transformation, which is the same as Camellia, sm4, and others.
Another is to use GFNI.
GFNI implementation is faster than AES-NI implementation. So, it uses GFNI implementation if the running CPU supports GFNI.

To calculate the first s-box(S1), it just uses the aesenclast and then inverts shift_row. No more process is needed for this job because the first s-box is the same as the AES encryption s-box.

To calculate the second s-box(X1, invert of S1), it just uses the aesdeclast and then inverts shift_row. No more process is needed for this job because the second s-box is the same as the AES decryption s-box.

To calculate the third s-box(S2), it uses the aesenclast, then affine transformation, which is combined AES inverse affine and ARIA S2.

To calculate the last s-box(X2, invert of S2), it uses the aesdeclast, then affine transformation, which is combined X2 and AES forward affine.

The optimized third and last s-box logic and GFNI s-box logic are implemented by Jussi Kivilinna.

The aria-generic implementation is based on a 32-bit implementation, not an 8-bit implementation.
The aria-avx Diffusion Layer implementation is based on aria-generic implementation because 8-bit implementation is not fit for parallel implementation but 32-bit is fit for this.

The first patch in this series is to export functions for aria-avx. The aria-avx uses existing functions in the aria-generic code. The second patch is to implement aria-avx.
The last patch is to add async test for aria.

Benchmarks:
The tcrypt is used.
cpu: i3-12100

How to test:
modprobe aria-generic
tcrypt mode=610 num_mb=8192

Result:
testing speed of multibuffer ecb(aria) (ecb(aria-generic)) encryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 534 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 2006 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 3674 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 52374 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 608 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2586 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 4707 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 69794 cycles

testing speed of multibuffer ecb(aria) (ecb(aria-generic)) decryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 545 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 1995 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 3673 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 52359 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 615 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2588 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 4712 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 69916 cycles

How to test:
modprobe aria
tcrypt mode=610 num_mb=8192

AVX with AES-NI:
testing speed of multibuffer ecb(aria) (ecb-aria-avx) encryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 629 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 2060 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 1223 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 11931 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 686 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2616 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 1439 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 15488 cycles

testing speed of multibuffer ecb(aria) (ecb-aria-avx) decryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 609 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 2027 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 1211 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 12040 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 684 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2614 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 1445 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 15478 cycles

AVX with GFNI:
testing speed of multibuffer ecb(aria) (ecb-aria-avx) encryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 730 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 2056 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 1028 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 9223 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 685 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2603 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 1179 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 11728 cycles

testing speed of multibuffer ecb(aria) (ecb-aria-avx) decryption
test 0 (128 bit key, 16 byte blocks): 1 operation in 617 cycles test 2 (128 bit key, 128 byte blocks): 1 operation in 2057 cycles test 3 (128 bit key, 256 byte blocks): 1 operation in 1020 cycles test 6 (128 bit key, 4096 byte blocks): 1 operation in 9280 cycles test 7 (256 bit key, 16 byte blocks): 1 operation in 687 cycles test 9 (256 bit key, 128 byte blocks): 1 operation in 2599 cycles test 10 (256 bit key, 256 byte blocks): 1 operation in 1176 cycles test 13 (256 bit key, 4096 byte blocks): 1 operation in 11909 cycles

v4:
- Fix sparse warning.
- Remove .align statement for .text
- https://lkml.kernel.org/r/[email protected]

v3:
- Use ECB macro instead of opencode.
- Implement ctr(aria-avx).
- Improve performance(20% ~ 30%) with combined affine transformation for S2 and X2.
- Implemented by Jussi Kivilinna.
- Improve performance( ~ 55%) with GFNI.
- Implemented by Jussi Kivilinna.
- Add aria-ctr async speed test.
- Add aria-gcm multi buffer speed test
- Rebase and fix Kconfig

v2:
- Do not call non-FPU functions(aria_{encrypt | decrypt}()) in the FPU context.
- Do not acquire FPU context for too long.

Taehee Yoo (3):
crypto: aria: prepare generic module for optimized implementations
crypto: aria-avx: add AES-NI/AVX/x86_64/GFNI assembler implementation
of aria cipher
crypto: tcrypt: add async speed test for aria cipher

 arch/x86/crypto/Kconfig                 |   18 +
 arch/x86/crypto/Makefile                |    3 +
 arch/x86/crypto/aria-aesni-avx-asm_64.S | 1303 +++++++++++++++++++++++
 arch/x86/crypto/aria-avx.h              |   16 +
 arch/x86/crypto/aria_aesni_avx_glue.c   |  213 ++++
 crypto/Makefile                         |    2 +-
 crypto/{aria.c => aria_generic.c}       |   39 +-
 crypto/tcrypt.c                         |   30 +
 include/crypto/aria.h                   |   17 +-
 9 files changed, 1623 insertions(+), 18 deletions(-)
 create mode 100644 arch/x86/crypto/aria-aesni-avx-asm_64.S
 create mode 100644 arch/x86/crypto/aria-avx.h
 create mode 100644 arch/x86/crypto/aria_aesni_avx_glue.c
 rename crypto/{aria.c => aria_generic.c} (86%)
Signature · 1 line
2.17.1