Fast CRC Computation Using PCLMULQDQ Instruction
4
Contents
Background ............................................................................................................. 5
Overview of CRC computation ................................................................................... 5
PCLMULQDQ Instruction ......................................................................... 7
Fast CRC Computation .............................................................................................. 7
Overview of the Folding Method .............................................................. 7
Single fold of a 128-bit Data chunk .............................................. 10
Parallel Folding of 128-bit Data chunks ........................................ 11
Detailed Implementation Steps for 32-bit CRC ........................................ 11
Step 1 – Iteratively Fold by 4: .................................................... 11
Step 2 – Iteratively Fold by 1: .................................................... 12
Step 3 - Final Reduction of 128-bits ............................................. 13
Overall flow ............................................................................... 14
Example of 32-bit CRC (IEEE 802.3) ............................................ 16
16-bit CRC ............................................................................................................ 16
64-bit CRC ............................................................................................................ 17
Final Reduction of 128-bits ......................................................... 17
Bit-Reflection ........................................................................................................ 18
Example of 32-bit CRC with bit-reflected data (gzip CRC) ............... 21
Conclusion ............................................................................................................ 22
