Method and apparatus for initiating the execution of instructions using a central pipeline execution unit

Claim 1

In a synchronous central processor of a digital data processing system having a cache unit, a plurality of execution units, a collector, and an instruction fetch unit; a central execution pipeline unit (CEPU) having a plurality of stages comprising: means for obtaining and loading an instruction word containing an address and instruction information into a basic instruction register (BIR); a first stage including first address means utilizing the address information of the instruction word in the basic instruction register for forming the carrys and sums of an effective and a virtual address, and control means for decoding the instruction information to produce memory command signals and data alignment control signals; a second stage including second address means for forming the effective address and to form the virtual address, which virtual address includes a word address portion; means for sending the word address portion of the virtual address to the cache unit, and second control means for issuing the memory command signals to the cache unit and for converting the instruction information to a set of execution code signals and signals designating the execution unit to execute the execution code; a third stage including third address means for converting the virtual address to a physical address including a physical page number, means for sending the physical page number to the cache unit, third control means for sending the execution code produced in the second stage to the designated execution unit, and means for converting the execution code into execution unit control signals if the central execution unit is the designated execution unit; a fourth stage including fourth control means for transmitting the data alignment control signals to a distributor means for aligning the addressed target word transmitted from the cache; a fifth stage including fifth control means to enable the addressed execution unit to receive the addressed target word from the distributor means, and for transmitting the instruction field to the collector, said central execution unit, if designated, executing the instruction on the addressed target word; and a sixth stage including control means for loading the results of the execution of an instruction by the central execution unit during the fifth stage on the addressed target word received from the cache unit during the fourth stage into a results stack, and for updating the indicators of the central execution pipeline unit.

Claim 8

The method of initiating the execution of instructions in a synchoronous central processor of a digital data processing system having a cache unit, a plurality of execution units, a collector, an instruction fetch unit; comprising the steps of: obtaining an instruction word containing address and instruction information from the instruction fetch unit; during a first cycle, utilizing the address information of the instruction word to form the carrys and sums of an effective and a virtual address, and decoding the instruction information to produce memory command signals and data alignment control signals; during a second cycle, completing the formation of the effective address and the virtual address, which virtual address includes a word address portion; sending the word address portion of the virtual address to the cache unit, issuing the memory command signals to the cache unit, and converting the instruction information to a set of execution code signals and signals designating the execution unit to execute the execution code; during a third cycle, converting the virtual address to a physical address including a physical page number, sending the physical page number to the cache unit, sending the execution code produced in the second stage to the designated execution unit, and converting the execution code into execution unit control signals if the central execution unit is the designated execution unit; during a fourth cycle, aligning the addressed target word transmitted from the cache; during a fifth cycle, enabling the addressed execution unit to receive the addressed target word, transmitting the instruction information to the collector, said central execution unit, if designated, executing the instruction on its addressed target word; and during a sixth cycle, loading the results of the execution of an instruction by the central execution unit during the fifth stage on the addressed target word received from the cache unit during the fourth stage into a results stack.

Claim 9

The method of initiating the execution of instructions in a synchoronous central processor of a digital data processing system having a cache unit, a plurality of execution units, a collector, an instruction fetch unit; comprising the steps of: obtaining an instruction word containing address and instruction information from the instruction fetch unit; during a first cycle, utilizing the address information of the instruction word to form the carrys and sums of an effective and a virtual address, and decoding the instruction information to produce memory command signals and data alignment control signals; during a second cycle, completing the formation of the effective address and the virtual address, which virtual address includes a word address portion; sending the word address portion of the virtual address to the cache unit, issuing the memory command signals to the cache unit, and converting the instruction information to a set of execution code signals and signals designating the execution unit to execute the execution code; during a third cycle, converting the virtual address to a physical address including a physical page number, sending the physical page number to the cache unit, sending the execution code produced in the second stage to the designated execution unit, and converting the execution code into execution unit control signals if the central execution unit is the designated execution unit; during a fourth cycle, aligning the addressed target word transmitted from the cache; during a fifth cycle, enabling the addressed execution unit to receive the addressed target word, transmitting the instruction information to the collector, said central execution unit, if designated, executing the instruction on its addressed target word; and during a sixth cycle, loading the results of the execution of an instruction by the central execution unit during the fifth stage on the addressed target word received from the cache unit during the fourth stage into a results stack.