Multiple data patch CPU architecture

Claim 2

A CPU architecture for use in data processing systems wherein an instruction causes operands to be processed during execution of a microcode word, including: arithmetic logic means in a first data path for performing arithmetic logic operations on operands as specified by a first microcode filed of the microcode word, at least one selectable function means in a second data path for performing operations on operands simultaneously with the arithmetic logic operations as specified by a second microcode filed of the microcode word, a plurality of first register means for holding operands for processing by said arithmetic logic means and said function means, a plurality of first select logic means, responsive to a third microcode field of the microcode word, for selectively connecting one of said first register means to said arithmetic logic means and one of said first register means to said function means, first return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, and second return bus means for returning results from an operations of said function means to at least one of said first register means,

Claim 3

A CPU architecture for use in data processing systems wherein an instruction causes operands to be processed during execution of a microcode word, including: arithmetic logic operations on operands as specified by a first microcode field of the microcode word, at least one selectable function means in a second data path for performing operations on operands simultaneously with the arithmetic logic operations as specified by a second microcoded field of the microcode word, a plurality of first register means for holding operands for processing by said arithmetic logic means and said function means, a plurality of first select logic means, responsive to a third microcode field of the microcode word, for selectively connecting one of said first register means to said arithmetic logic means and one of said first register means to said function means, first return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operations of said function means to at least one of said first register means, and at least one second register means connected to at least one of said first register means for holding operands for processing by said arithmetic logic means and said function means and selectively supplying said operands to said first register means in response to a fourth microcode field of the microcode word.

Claim 4

A CPU architecture for use in data processing systems wherein an instruction causes operands to be processed during execution of a microcode word, including: arithmetic logic operations on operands as specified by a first microcode field of the microcode word, at least one selectable function means in a second data path for performing operations on operands simultaneously with the arithmetic logic operations as specified by a second microcoded field of the microcode word, a plurality of first register means for holding operands for processing by said arithmetic logic means and said function means, a plurality of first select logic means, responsive to a third microcode field of the microcode word, for selectively connecting one of said first register means to said arithmetic logic means and one of said first register means to said function means, first return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operations of said function means to at least one of said first register means, and at least one second register means connected to at least one of said first register means for holding operands for processing by said arithmetic logic means and said function means and selectively supplying said operands to said first register means in response to a fourth microcode field of the microcode word.

Claim 5

The CPU architecture of claim 2 further including: bus means, connected between said first and second register means, for communicating operands form at least one of said first register means to said at least one of second register means.

Claim 6

A CPU architecture for use in a data processing system wherein an instruction causes operands to be processed during execution of a microcode word having at least first, second, third, fourth, and fifth microcode fields, including: arithmetic logic means in a first data path for performing arithmetic logic operations on operands as specified by the first microcode field, at least one selectable function means in a second data path for performing operations on operands simultaneously with the arithmetic logic operations as specified by the second microcode field, a plurality of first register means for holding operands for processing by said arithmetic logic means and said function means, a plurality of first select logic means, responsive to the third microcode field, for selectively connecting one of said first register means to said arithmetic logic means and one of said first register means to said function means, first return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operation of said function means to at least one of said first register means, at least one second register means connected to at least one of said first register means for holding operands for processing by said arithmetic logic means and said function means and selectively supplying said operands to said first register means in response to the fourth microcode field, and at least one second select logic means connected to at least one of said first register means for selectively connecting one of a plurality of data sources to said first register means in response to the fifth microcode field.

Claim 7

A CPU architecture for use in a data processing system wherein an instruction causes operands to be processed during execution of a microcode word having at least first, second, third, fourth, and fifth microcode fields, including: arithmetic logic means in a first data path for performing arithmetic logic operations on operands as specified by the first microcode field, at least one selectable function means in a second data path for performing operations on operands simultaneously with the arithmetic logic operations as specified by the second microcode field, a plurality of first register means for holding operands for processing by said arithmetic logic means and said function means, a plurality of first select logic means, responsive to the third microcode field, for selectively connecting one of said first register means to said arithmetic logic means and one of said first register means to said function means, first return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operation of said arithmetic logic means to at least one of said first register means, second return bus means for returning results from an operation of said function means to at least one of said first register means, at least one second register means connected to at least one of said first register means for holding operands for processing by said arithmetic logic means and said function means and selectively supplying said operands to said first register means in response to the fourth microcode field, and at least one second select logic means connected to at least one of said first register means for selectively connecting one of a plurality of data sources to said first register means in response to the fifth microcode field.

Claim 8

A pipelined CPU architecture for use in data processing systems including: arithmetic logic means in a first data path for performing arithmetic logic processes on operands, a plurality of selectable function means in a second data path for processing operands simultaneously with the arithmetic logic means, said function means selected in response to a first microcode field in a microinstruction, the function means including sense means for indicating whether one of a plurality of test conditions selected by a test selection microcode field of the microcode word has been met, a plurality of first register means for storing operands for processing, a plurality of first multiplexer means, connected to the first register means and the data paths and responsive to a second microcode field in the microinstruction, for simultaneously placing selected ones of the operands on the first and second data paths, first return bus means for returning results from the arithmetic logic means to at least one of the first register means, second return bus means for returning results from an operation of the function means to at least one of the first register means, and microbranching means, connected to a microsequencer in said CPU, and responsive to the sense means, for performing microbranching when the sense means indicates the selected test condition has been met.

Claim 9

A pipelined CPU architecture for use in data processing systems including: arithmetic logic means in a first data path for performing arithmetic logic processes on operands, a plurality of selectable function means in a second data path for processing operands simultaneously with the arithmetic logic means, said function means selected in response to a first microcode field in a microinstruction, the function means including sense means for indicating whether one of a plurality of test conditions selected by a test selection microcode field of the microcode word has been met, a plurality of first register means for storing operands for processing, a plurality of first multiplexer means, connected to the first register means and the data paths and responsive to a second microcode field in the microinstruction, for simultaneously placing selected ones of the operands on the first and second data paths, first return bus means for returning results from the arithmetic logic means to at least one of the first register means, second return bus means for returning results from an operation of the function means to at least one of the first register means, and microbranching means, connected to a microsequencer in said CPU, and responsive to the sense means, for performing microbranching when the sense means indicates the selected test condition has been met.

Claim 10

The CPU architecture of claim 5, wherein said plurality of function means includes an arithmetic logic unit.

Claim 11

The CPU architecture of claim 5, wherein said plurality of function means includes a multiple address scratchpad.

Claim 12

The CPU architecture of claim 5, wherein said plurality of first register means includes a literal constant register.

Claim 13

The CPU architecture of claim 5, further including: parity generating means, including parity bus means, in said plurality of function means for generating parity data for operands processed by said plurality of function means, the parity generating means also including delay means for placing said parity data on said parity bus means on a clock cycle occurring later than a clock cycle in which said results from an operation of said function means are placed on said second return bus means.

Claim 14

The CPU architecture of claims 2, 4, or 3, wherein said plurality of second register means includes a memory cache.

Claim 15

The CPU architecture of claims 5, 1, 2, 4, 3, 6, 7 or 8, further including: parity checking means for checking parity accompanying said operands, including means for synchronizing said operands and accompanying parity causing said operands and accompanying parity to arrive at said parity checking means during a signal clock period.

Claim 16

The CPU architecture of claim 1, 2, 4, or 3, including: sense means in said at least one selectable function means for indicating whether one of a plurality of test conditions selected by a test selection microcode field of the microcode word has been met, and microbranching means, connected to a microsequencer in said CPU, and responsive to said sense means, for performing microbranching when said sense means indicates said selected test condition has been met.

Claim 17

A central processing unit of a data processing system having an arithmetic logic unit in a data path, wherein an instruction causes operands to be processed during an instruction cycle, including: a plurality of selectable function means, in a data path separate from that of said arithmetic logic unit, for performing operations on operands during said instruction cycle in accordance with said instruction, said instruction causing the selection of one of said function means and defining its function and its operands; sense means in said function means for indicating that a test condition has been met; and microbranching means, connected to a microsequencer in said CPU, responsive to said sense means, for performing microbranching when aids sense means indicates said test condition has been met.

Claim 18

A central processing unit of a data processing system having an arithmetic logic unit in a data path, wherein an instruction causes operands to be processed during an instruction cycle, including: a plurality of selectable function means, in a data path separate from that of said arithmetic logic unit, for performing operations on operands during said instruction cycle in accordance with said instruction, said instruction causing the selection of one of said function means and defining its function and its operands; sense means in said function means for indicating that a test condition has been met; and microbranching means, connected to a microsequencer in said CPU, responsive to said sense means, for performing microbranching when aids sense means indicates said test condition has been met.

Claim 19

The central processing unit as in claim 13, further including parity generating means, including parity bus means, in each of said plurality of selectable function means for generating parity data for results of operands processed by said plurality of selectable function means and delay means for placing said parity data on said parity bus means on a clock cycle occurring later than a clock cycle in which said results from an operation of said function means are placed at the output of said function means.

Claim 20

The central processing limit of claim 14, further including parity checking means for checking parity, including delay means for synchronizing said results and said parity data so that they arrive at said parity checking means during a single clock period.