Why assembler version of addmul_1 is so fast?

Andy borucki.andrzej at gmail.com
Sun Feb 2 01:41:49 UTC 2020

Problem is with high-level languages but I wonder, how low-level,
intermediate LLVM code will do in this case... (I we would write this in
LLVM code)
is special opcodes for this, is flexible enough and LLVM can optimize for
processor type?

More information about the gmp-discuss mailing list