Intel Corporation

Robert P Colwell

Andrew F Glew

Glenn J Hinton

David B Papworth

Michael A Fetterman

David W Clift

Darrell D Boggs

Shantanu R Gupta

Robert W Martell

James M Arnold

Atiq A Bajwa

Frederick J Pollack

Narayan Hegde

Steven J Griffith

Haitham H Akkary

Bryant E Bigbee

Shahrokh Shahidzadeh

Frank Binns

Ashwani Kumar Gupta

James E Jacobson, Jr

James S Griffith

Grace C Chen

Stephen M Coward

Michael Alan Fetterman

A microprocessor for providing an extended linear address of more than 32 bits. The extended linear address may be provided by concatenating a linear address with a segment selector extension, or by concatenating the values from two registers. Hierarchical translation of a linear address to a physical address is performed in which the number of levels in the hierarchy depends upon whether the linear address is an extended linear address.

Linear address extension and mapping to physical memory using 4 and 8 byte page table entries in a 32-bit microprocessor

Maximum throughput or 'back-to-back' scheduling of dependent instructions in a pipelined processor is achieved by maximizing the efficiency in which the processor determines the availability of the source operands of a dependent instruction and provides those operands to an execution unit executing the dependent instruction. These two operations are implemented through number of mechanisms. One mechanism for determining the availability of source operands, and hence the readiness of a dependent instruction for dispatch to an available execution unit, relies on the prospective determination of the availability of a source operand before the operand itself is actually computed as a result of the execution of another instruction. Storage addresses of the source operands of an instruction are stored in a content addressable memory (CAM). Before an instruction is executed and its result data written back, the storage location address of the result is provided to the CAM and associatively compared with the source operand addresses stored therein. A CAM match and its accompanying match bit indicate that the result of the instruction to be executed will provide a source operand to the dependent instruction waiting in the reservation station. Using a bypass mechanism, if the operand is computed after dispatch of the dependent instruction, then the source operand is provided directly from the execution unit computing the source operand to a source operand input of the execution unit executing the dependent instruction.

Method and apparatus for maximum throughput scheduling of dependent

A processor is disclosed comprising a front end circuit that fetches a series of instructions according to a program sequence determined by at least one branch prediction, a register renaming circuit that allocates execution resources to each instruction, and an execution circuit that executes each instruction in the instruction stream. The execution circuit causes the front end circuit to refetch the series of instructions if a branch misprediction is detected. A stall signal disables the register renaming circuit until the execution circuit commits the branch result to an architectural state according to the program sequence.

Processor that redirects an instruction fetch pipeline immediately upon

Each member system of a distributed collection of self-monitoring hardware systems includes receiving logic operative to receive a wellness token from a first other hardware system of the distributed collection of hardware systems. Each member system also includes modification logic, communicatively coupled to the receiving logic, operative to modify the wellness token to create a modified wellness token in a manner that reflects the wellness of the member hardware system, and transmitting logic, communicatively coupled to the modification logic, operative to transmit the modified wellness token to a second other hardware system of the distributed collection of hardware systems.

Self-monitoring distributed hardware systems

A mechanism and method for renaming flags within a register alias table ('RAT') to increase processor parallelism and also providing and using flag masks associated with individual instructions. In order to reduce the amount of data dependencies between instructions that are concurrently processed, the flags used by these instructions are renamed. In general, a RAT unit provides register renaming to provide a larger physical register set than would ordinarily be available within a given macroarchitecture's logical register set (such as the Intel architecture or PowerPC or Alpha designs, for instance) to eliminate false data dependencies between instructions that reduce overall superscalar processing performance for the microprocessor. The renamed flag registers contain several flag bits and various flag bits may be updated or read by different instructions. Also, static and dynamic flag masks are associated with particular instructions and indicate which flags are capable of being updated by a particular instruction and also indicate which flags are actually updated by the instruction. Static flag masks are used in flag renaming and dynamic flag masks are used at retirement. The invention also discovers cases in which a flag register is required that is a superset of the previously renamed flag register portion.

Flag renaming and flag masks within register alias table

A method and circuitry for coordinating exceptions in a processor. The processor generates a result data value and an exception data value for each instruction wherein the exception data value specifies whether the corresponding instruction causes an exception. The processor commits the result data values to an architectural state of the processor in the sequential program order, and fetches an exception handler to processes the exception if the exception is indicated by one of the exception data values. The processor fetches an asynchronous event handler to processes an asynchronous event if the asynchronous event is detected while the result data values are committed to the architectural state of the processor.

Exception handling in a processor that performs speculative out-of-order

A mechanism for efficiently overlapping multiple operand types is used in a microprocessor which includes a plurality of execution units and a mechanism to provide operations, which include one or more operands, to the plurality of execution units. Each of the plurality of execution units interprets the one or more operands as different types of operands, and the mechanism to provide operations overlaps the different types of operands.

Mechanism for efficiently overlapping multiple operand types in a

This invention overcomes the address size backward compatibility problem by first subtracting the segment base address from the linear destination address of a branch instruction to generate a virtual destination address. It is assumed that the branch instruction destination address is n bits long with m most significant bits. It is desired to provide backward compatibility in the n-bit processor for branch instruction code written for processors utilizing instruction address fields of size (n-m) bits. After obtaining the virtual address, if any of the m most significant bits are non-zero, then those m bits are set to zero to thereby generate a corrected virtual address. If such a compatibility correction is necessary, then a clear signal is asserted to flush all state of the processor that resulted from instructions being fetched after the branch instruction was fetched. The corrected virtual address is added back to the segment base address to generate a corrected linear address. The next instruction is fetched at the corrected linear address.

Method and apparatus for providing address-size backward compatibility

A mechanism for indicating within a register alias table (RAT) that certain data has become architecturally visible so that the RAT contains the most recent location of the certain data. Upon receiving the indication that data associated with a particular register is architecturally visible, if a subsequent operation uses the particular register as a source, the data will be supplied from the architecturally visible buffer instead of from an internal buffer (not architecturally visible). The internal buffer is implemented by a reorder buffer (ROB) which contains information associated with instructions that have not yet retired. The architecturally visible buffer is a retirement register file (RRF) which contains information associated with retired instructions. When an instruction retires, the register alias table is searched for the retiring physical register and will indicate within the register alias table that the data associated with the retiring physical register is located within the RRF only if the register alias table has not already (or concurrently) reassigned a new physical register to the logical register associated with the retiring physical register. If the logical register associated with the retiring physical register as been reassigned by subsequent instructions, then no update of the register alias table is required. Also provided is an embodiment for providing the above features in a system wherein the register ordering of the buffers can be altered via register exchange operations.

Register alias table update to indicate architecturally visible state

A simplified method and apparatus for handling the change of instruction control flow in a microprocessor is provided. Rather than attempting to implement a change in the instruction flow immediately, the processor first recognizes that flow is to be redirected from a predicted instruction flow to a correct instruction flow according to a flow control indicator. The flow control indicator may be attached to instructions flowing down the pipeline or inserted as a separate instruction in the pipeline. The pipeline is cleared of state created by instructions that do not follow the correct instruction flow, i.e., instructions that were erroneously fetched after the instruction causing the change in flow. The change in flow as indicated by the flow control indicator is implemented later in the pipeline.

Method and apparatus for changing flow of control in a processor

A method and apparatus for dynamically allocating entries of microprocessor resources to particular instructions in an efficient manner to efficiently utilize buffer size and resources. The pipelined and superscalar microprocessor is capable of speculatively executing instructions and also out-of-order processing. Resources within the microprocessor include a store buffer, a load buffer, a reorder buffer and a reservation station. The reorder buffer contains a larger set of physical registers and also contains information related to speculative instructions and the reservation station comprises information related to instructions pending execution. The load buffer is only allocated to load instructions and is valid for an instruction from allocation pipestage to instruction retirement. The store buffer is only allocated to store instructions and is valid for an instruction from allocation to store performance. The reservation station is allocated to most instructions and is valid for an instruction from allocation to instruction dispatch. The reorder buffer is allocated to all instructions and is valid for a given instruction from allocation to retirement. The load buffer, store buffer, and reorder buffer are sequentially allocated while the reservation station is not. Resource allocation is performed dynamically (as needed by the operation) rather than as a full set of resources attached to each operation. Using the above allocation scheme, efficient usage of the microprocessor resources is accomplished.

Method and apparatus for dynamic allocation of multiple buffers in a

A circuit comprising a number of range registers and complimentary decoding/matching circuits is provided to a processor for determining the memory type of a physical address, thereby allowing memory operating characteristics to be determined as soon as the physical address is available in an execution stage preceding cache access. Additionally, a memory type field is provided to each address translation lookaside buffer entry of the data and instruction memory subsystem for storing the determined memory type, and the memory type determination circuit is disposed in the page miss handler, thereby allowing memory type to be determined at the same time when the physical address is determined.

Methods and apparatus for determining operating characteristics of a

A computer system including a processor having a inherently weakly-ordered memory model comprising a mechanism for emulating strong-ordering to produce self-consistent ordering on a system-wide basis. The processor snoops the system bus externally to determine if a STORE on the external bus hits a LOAD buffer inside the memory subsystem of the processor. If so, the situation is flagged as one which carries the risk of violating processor ordering rules. When the STORE hits the same LOAD address in the LOAD buffer of the processor's memory subsystem, the speculative state of the processor is erased. This cancels the LOAD operation in all subsequent operations. The processor then begins executing from the aborted LOAD; this time loading the newly updated value.

Computer system with self-consistent ordering mechanism

Method and apparatus for predicting, clearing and redirecting

A data processor includes a plurality of physical registers and a decoder that decodes a stream of instructions into micro-operations which include speculative operations specifying associated logical registers. The data processor further includes a register-alias table having a plurality of addressable entries corresponding to logical registers, specified by the speculative operations. Each entry of the register-alias table contains a register pointer to a corresponding physical register. The processor further includes a retirement register file that maintains register values of non-speculative operations, and a retirement array that maintains a retirement ordering for the retirement register file. Both the register-alias table and retirement array are updated by circuitry that is responsive to a register exchange operation; the circuitry swapping register pointers associated with first and second entries, respectively.

Growing community of inventors

Portland, OR, United States of America

Robert P Colwell