Optimize ExtendSign()

Open chfast opened this issue 4 years ago • 0 comments

The current implementation is done by bits manipulation on whole 256-bit values (as you would do it for architecture native integers). There exists more efficient implementation which handles each of 4 64-bit words separately. This has been implemented in evmone in https://github.com/ethereum/evmone/pull/390.

Nov 15 '21 12:11 chfast