CPU: add traits for new AVX512 support
Add several standard traits for the support of new CPU AVX512 instructions, which includes AVX512VBMI [1], AVX512IFMA [2], AVX512VBMI2 [3], AVX512BITALG [4], AVX512VAES [5], AVX512GFNI [6], AVX512VPCLMULQDQ [7], AVX512VPOPCNTDQ [4]. [1] https://en.wikipedia.org/wiki/AVX-512#BW,_DQ_and_VBMI [2] https://en.wikipedia.org/wiki/AVX-512#IFMA [3] https://en.wikipedia.org/wiki/AVX-512#VBMI2 [4] https://en.wikipedia.org/wiki/AVX-512#VPOPCNTDQ_and_BITALG [5] https://en.wikipedia.org/wiki/AVX-512#VAES [6] https://en.wikipedia.org/wiki/AVX-512#GFNI [7] https://en.wikipedia.org/wiki/AVX-512#VPCLMULQDQ Change-Id: I58cba02f204d2e19a1a838852c628694edd47d73
This commit is contained in:
parent
77bf79e2aa
commit
066a2a972d
@ -41,6 +41,14 @@ TRAITS = [
|
||||
'AVX512BW', # byte + word
|
||||
'AVX512DQ', # double word + quad word
|
||||
'AVX512VNNI', # vector neural network instructions
|
||||
'AVX512VBMI', # vector byte manipulation instructions
|
||||
'AVX512IFMA', # integer fused multiply add
|
||||
'AVX512VBMI2', # vector byte manipulation instructions 2
|
||||
'AVX512BITALG', # bit algorithms
|
||||
'AVX512VAES', # vector aes instructions
|
||||
'AVX512GFNI', # galois field new instructions
|
||||
'AVX512VPCLMULQDQ', # carry-less multiplication of quadwords
|
||||
'AVX512VPOPCNTDQ', # vector population count instruction
|
||||
# ref: https://en.wikipedia.org/wiki/Bit_Manipulation_Instruction_Sets
|
||||
'ABM',
|
||||
'BMI',
|
||||
|
Loading…
Reference in New Issue
Block a user