Use faster carries in decimal arithmetic

Recent x86-64 machines often include instructions for faster additions and multiplications with different carry flag behavior, which can increase performance of 128-bit and 256-bit arithmetic. See

https://www.intel.com/content/dam/www/public/us/en/documents/white-papers/ia-large-integer-arithmetic-paper.pdf

Impala could take advantage of this via Intel intrinsics (https://software.intel.com/sites/landingpage/IntrinsicsGuide/#othertechs=ADX,BMI2) or perhaps via codegen