Processor that decodes a multi-cycle instruction into single-cycle micro-instructions and schedules execution of the micro-instructions

Claim 2

A processor comprising: a plurality of execution units, each execution unit being adapted to complete a single-cycle operation; a decode buffer which is partitioned into a plurality of entries, wherein each entry includes: a plurality of micro-instruction fields for storage of micro-instructions; and a field for identifying which of the micro-instruction is to be executed first; and a decoder adapted to decode instructions, wherein for each instruction decoded, the decoder writes one or more micro-instructions in the micro-instruction fields of an entry associated with the instruction, each micro-instruction identifying an operation to be executed by one of the execution units during execution of the instruction by the processor.

Claim 3

The processor of claim 1, wherein a first of the micro-instruction fields for a first micro-instruction includes a sub-field identifying a second of the micro-instruction fields for a second micro-instruction that is to be executed after the first micro-instruction.

Claim 4

The processor of claim 2, further comprising a scheduler coupled to the decode buffer, wherein the scheduler comprises: issuing logic that attempts to issue micro-instructions from the decode buffer to the execution units; and instruction dependency checking logic to prevent the issuing logic from issuing a micro-instruction that would interfere with in-order execution of the first and second micro-instructions.

Claim 5

The processor of claim 3, wherein instruction dependency checking logic further comprises logic coupled to the micro-instruction fields that generates from the micro-instructions stored the micro-instruction fields that signals that control the relative order in which the issuing logic can issue micro-instructions from an entry.

Claim 6

The processor of claim 1, wherein each entry further comprises a first operand field that identifies a source operand for an instruction associated with the entry and a second operand field that identifies a destination operand for the instruction.

Claim 7

The processor of claim 1, wherein each micro-instruction field is dedicated for use by an associated one of the execution units.

Claim 8

The processor of claim 6, wherein each micro-instruction field comprises a sub-field for a value indicating whether the micro-instruction field is storing a valid micro-instruction for execution by the execution unit associated with the micro-instruction field.

Claim 9

The processor of claim 1, further comprising a scheduler that is coupled to the buffer and schedules execution of micro-instructions in the buffer.

Claim 10

A processor comprising: a plurality of execution units, each execution unit being adapted to complete a single-cycle operation; a decode buffer which is partitioned into a plurality of entries, wherein each entry includes: a plurality of micro-instruction fields for storage of micro-instructions; a first operand field that identifies a source operand for an instruction associated with the entry; and a second operand field that identifies a destination operand for the instruction; a decoder adapted to decode instructions, wherein for each instruction decoded, the decoder writes one or more micro-instructions in the micro-instruction fields of an entry associated with the instruction, each micro-instruction identifying an operation to be executed by one of the execution units during execution of the instruction by the processor; and a scheduler coupled to the decode buffer, wherein the scheduler comprises: issuing logic that attempts to issue micro-instructions from the decode buffer to the execution units; and operand dependency logic coupled to compare the first operand field in a first of the entries to the second operand field in a second of the entries, the first entry being associated with an instruction that is newer than an instruction that is associated with the second entry, wherein the operand dependency logic prevents the issuing logic from issuing a micro-instruction from the first entry if the first field in the first entry identifies the same operand as the second field in the second entry.

Claim 11

A processor comprising: a plurality of execution units, each execution unit being adapted to complete a single-cycle operation; a decode buffer which is partitioned into a plurality of entries, wherein each entry includes: a plurality of micro-instruction fields for storage of micro-instructions; a first operand field that identifies a source operand for an instruction associated with the entry; and a second operand field that identifies a destination operand for the instruction; a decoder adapted to decode instructions, wherein for each instruction decoded, the decoder writes one or more micro-instructions in the micro-instruction fields of an entry associated with the instruction, each micro-instruction identifying an operation to be executed by one of the execution units during execution of the instruction by the processor; and a scheduler coupled to the decode buffer, wherein the scheduler comprises: issuing logic that attempts to issue micro-instructions from the decode buffer to the execution units; and operand dependency logic coupled to compare the first operand field in a first of the entries to the second operand field in a second of the entries, the first entry being associated with an instruction that is newer than an instruction that is associated with the second entry, wherein the operand dependency logic prevents the issuing logic from issuing a micro-instruction from the first entry if the first field in the first entry identifies the same operand as the second field in the second entry.

Claim 12

A processor comprising: plurality of execution units, each execution unit being adapted to complete a single-cycle operation; a decode buffer which is partitioned into a plurality of entries, wherein each entry includes a plurality of micro-instruction fields for storage of micro-instructions; a decoder adapted to decode instructions, wherein for each instruction decoded, the decoder writes one or more micro-instructions in the micro-instruction fields of an entry associated with the instruction, each micro-instruction identifying an operation to be executed by one of the execution units during execution of the instruction by the processor; and a scheduler coupled to the decode buffer, wherein the scheduler comprises: issuing logic that attempts to issue micro-instructions from the decode buffer to the execution units; and instruction dependency checking logic that controls the relative order in which the issuing logic can issue micro-instructions from an entry.

Claim 13

A processor comprising: plurality of execution units, each execution unit being adapted to complete a single-cycle operation; a decode buffer which is partitioned into a plurality of entries, wherein each entry includes a plurality of micro-instruction fields for storage of micro-instructions; a decoder adapted to decode instructions, wherein for each instruction decoded, the decoder writes one or more micro-instructions in the micro-instruction fields of an entry associated with the instruction, each micro-instruction identifying an operation to be executed by one of the execution units during execution of the instruction by the processor; and a scheduler coupled to the decode buffer, wherein the scheduler comprises: issuing logic that attempts to issue micro-instructions from the decode buffer to the execution units; and instruction dependency checking logic that controls the relative order in which the issuing logic can issue micro-instructions from an entry.

Claim 14

The processor of claim 10, wherein each entry further comprises one or more fields that define the relative order in which the scheduler issues micro-instructions from the entry.

Claim 17

The method of claim 12, wherein scheduling the micro-instructions permits execution micro-instructions out of the program order for the instructions.

Claim 18

The method of claim 13, wherein scheduling the micro-instructions limits execution of micro-instructions that are decoded from the same instruction to the relative order for that instruction.

Claim 19

The method of claim 12, further comprising simultaneously executing micro-instructions from a plurality of entries.