Dublin Port Tunnel

General

• Added AESNI emulation implementation
• Added AES-GCM multi-buffer implementation for AVX512 with VAES and VPCLMULQDQ
• Added flexible job chain order support #23
• GCM submit and flush functions moved into architecture MB manager modules
• GCM API's added to MB_MGR structure
• Added plain SHA support in JOB API
• Added architectural compiler optimizations for GCC/CC

LibTestApp

• Added option not to run GCM tests
• Added AESNI emulation tests
• Added plain SHA tests
• Updated to take advantage of new GCM macros

LibPerfApp

• Buffer alignment update
• Updated to take advantage of new GCM macros

Performance

• AES-GCM AAD performance improvement for AVX512, AVX2, AVX and SSE

Resolved issues

• Issue #24
• Issue #25
• Issue #26

Jack Lynch Tunnel

General

• Added zero-sized message support in GCM
• Stack execution flag disabled in new asm modules (pull request #19 )

LibTestApp

• Added AES vectors
• Added DOCSIS AES vectors
• Added CFB validation

LibPerfApp

• Smoke test option added

Performance

• AES-CMAC performance optimizations (intercept cycle cost)
• Implemented store to load optimizations in
  • AES-CMAC submit and flush jobs for SSE and AVX
  • HMAC-MD5, HMAC-SHA submit jobs for AVX
  • HMAC-MD5 submit job for AVX2

Resolved issues

• Issue #22
• Issue #21

Limerick Tunnel

API

• Replaced defines for API with symbols for binary compatibility
• Added support for compile time and runtime library version checking
• Added new API's to compute SHA1, SHA224, SHA256, SHA384 and SHA512 hashes
  to support key reduction cases where key is longer than a block size

Features

• Added support for full MD5 digest size
• Added support for zero cipher length in AES-CCM
• Extended support for HMAC full digest sizes for HMAC-SHA1, HMAC-SHA224,
  HMAC-SHA256, HMAC-SHA384 and HMAC-SHA512. Previously only truncated sizes
  were supported.
• Added AES-CMAC support for output digest size between 4 and 16 bytes
• Added GHASH support for output digest size up to 16 bytes

Performance

• Optimized submit job API's with store to load optimization in SSE, AVX,
  AVX2 (excluding MD5)

General

• Added HMAC-SHA & HMAC-MD5 vectors to LibTestApp
• Extended multi-thread features of LibPerfApp Windows version to match
  Linux version of the application
• Improved performance application accuracy by increase number of
  test iterations

Marmaray Tunnel

API

• Multiple API header files consolidated into one (intel-ipsec-mb.h)
• Multi-buffer manager instance allocate and free API's added

Features

• AES-CMAC-128 and AES-CMAC-96 support added
  • SSE implementation
  • AVX implementation
• 3DES support added
  • x86 implementation
  • AVX512 implementation
• LibPerfApp
  • Unhalted cycles support added to LibPerfApp (Linux at the moment)
  • Core affinity support added in LibPerfApp

Performance

• AVX512 DES performance smoothed out
• VZEROUPPER instruction issued after AVX2/AVX512 code to avoid expensive SSE<->AVX transitions

General

• Library compiles to SO/DLL by default
• Install/uninstall targets added to makefiles
• MAN page added
• README documentation extensions and updates

Bug and issue fixes

• Issue #16
• Issue #17
• Issue #18

West Rock Tunnel

Features

• AES CCM 128
  • SSE implementation
  • AVX implementation
  • RFC3610 tests added
• LibPerfApp
  • Multithread command line option added

General

• Linux SO compilation option added
• Windows DLL compilation option added
• Coding style fixes
• Coding style target added to Makefile

Siloam Tunnel

Features

• DES CBC
  • Generic implementation
  • Intel(R) AVX-512 Instructions implementation
  • Tests added
• DOCSIS DES
  • Generic implementation
  • Intel(R) AVX-512 Instructions implementation
  • Tests added

General

• package SPEC file created

Channel Tunnel

Features

• AES GCM optimizations for AVX2
• Change of AES GCM API
  - API's have been renamed
  - expanded keys have been separated from the operation context data
• New AES GCM API via job structure and job API's
  - use of the interface may simplify application design at the expense of some performance loss vs direct AES GCM API's
• AES GCM IV automatically padded with block counter (no need for application to do it)
• IV in AES CTR mode can be in 12 byte format (no block counter) or 16 byte format (with block counter)
• Macros added to ease access to job API for specific architecture
  - use of these macros can simplify application design but it may produce worse performance than calling architecture job API's directly
• 'submit_job_nocheck()' API's added to gain some CPU cycles by not validating job structure. Behavior of submit_job() API remains the same and it does validate job structure as part of its operation.
• Result stability improvements in LibPerfApp

Bug Fixes (vs v0.45)

• none

Seikan

Features

• Added Intel(R) AVX-512 optimized HMAC-SHA224 and HMAC-SHA256
• Added Intel(R) AVX-512 optimized HMAC-SHA384 and HMAC-SHA512
• GCM128 and GCM256 updates (with new API that is scatter gather list friendly)
• GCM192 added
• New DOCSIS SEC BPI V3.1 cipher
• Windows x64 compilation target
• Added library API benchmark tool 'ipsec_perf' and script to compare results 'ipsec_diff_tool.py'

Bug Fixes (vs v0.44)

• AES CTR mode fix to allow message size not to be multiple of AES block size
• RSI and RDI registers clobbered when running HMAC-SHA224 or HMAC-SHA256 on Windows using SHA extensions

GCM192 added

GCM192 added together with SGL support
API changes to use const keyword

GCM128/256 Update

GCM128/GCM256 updated to identical to ISA-L_crypto GCM implementation (https://github.com/01org/isa-l_crypto).
The differences are:

• gcm_data structure has to be 16 byte aligned
• no support for non-temporal loads and/or stores in the code