Formula processor having cache means for storing and retrieving results of previously computed formulas based on identification and dependency information

摘要

A formula processor which computes results for a group of formulas in which at least one formula uses the result of another formula. The formula processor receives the formulas from a host and returns computed results for the formulas to the host for storage in a main memory. Formulas are received as a sequence of tokens, each token describing operands or operations. The formula processor is pipelined; as one token is interpreted and processed by the formula processor, subsequent tokens are received. The formula processor uses a computation element which is capable of performing only certain operations; before attempting each operation described by a token, the computation element determines whether the operation can be performed by the computation element--if it can be performed, it is; otherwise, the host may perform the operation itself, or rewrite the operation so the formula processor can perform it. After the formula processor computes a result for a formula, it stores this result in a cache memory. Thereafter, the formula processor can retrieve the result stored in the cache memory and use the retrieved result to compute a subsequent formula. The operands to a formula may be values or pointers indicating how to locate values; the formula processor converts pointers into values when required for a computation by requesting resolution from the host microprocessor, or, alternatively, by retrieval from the cache.