Ati International Srl

Hal Computer Systems, Inc.

Other

Fujitsu Corporation

Advanced Silicon Technologies LLC

Niteen Aravind Patkar

Shalesh Thusoo

Korbin S Van Dyke

Michael C Shebanow

Gene W Shen

John R Szeto

Stephen C Purcell

T R Ramesh

John S Yates, Jr

Paul Henry Hohensee

David Reese

Gurjeet Singh Saund

Jaspal Kohli

Akira Katsuno

Don A Van Dyke

Ronak Patel

Govind V Malalur

Nital P Patwa

Ali Alasti

Sanjay Mansingh

Sandeep Nijhawan

Farnad Sajjadian

Michael A Simone

Stephen C Hale

Ravi Swami

Dale R Jurich

David L Reese

Sunil W Savkar

Jim Lin

Hideki Osone

Takumi Maruyama

Tr Ramesh

Greg L Ries

Dee Tovey

Paul H Hohensee

Paul Campbell

A method and a computer for performance of the method. While executing a program on a computer, profileable events occurring in the instruction pipeline are detected. The instruction pipeline is directed to record profile information describing the profileable events essentially concurrently with the occurrence of the profileable events. The detecting and recording occur under control of hardware of the computer without software intervention.

Profiling execution of computer programs

A computer. An instruction pipeline and memory access unit execute instructions in a logical address space of a memory of the computer. An address translation circuit translates address references generated by the program from the program's logical address space to the computer's physical address space. Profile circuitry is cooperatively interconnected with the instruction pipeline and configured to detect, without compiler assistance for execution profiling, occurrence of profilable events occurring in the instruction pipeline, and is cooperatively interconnected with the memory access unit to record profile information describing physical memory addresses referenced during an execution interval of the program.

Profiling of computer programs executing in virtual memory systems

A method and computer for performance of the method. While executing a program on a computer, the computer uses registers of a general register file for storage of instruction results. Profile information describing the profileable events is recorded into the general register file as the profileable events occur, without first capturing the information into a main memory of the computer.

Profiling program execution into registers of a computer

A method and apparatus for out-of-order memory processing within an in-order processing device includes processing that allows a plurality of memory transactions to be processed in a pipeline manner until a dependency arises between two or more memory transactions. Such processing includes, for each of the plurality of memory transactions, determining whether data associated with the transaction is stored in local cache. If the data is stored in local cache, it is written into a data register in a next pipeline interval. The processing continues by storing the memory transaction in a miss buffer when the data associated with the memory transaction is not stored in the local cache. The processing continues by writing the associated data for the memory transaction identified in the missed buffer into the data register when the data is received without regard to the pipeline manner.

Method and apparatus for out of order memory processing within an in order processor

An apparatus that provides configurable processing includes a fetch module, a decoder, and a dynamic arithmetic unit. The fetch module is operable to fetch at least one instruction and provide it to the decoder. The decoder receives the instruction and decodes it. The dynamic arithmetic logic unit receives the decoded instruction and configures at least one configurable arithmetic logic unit to perform an operation contained within the decoded instruction.

Method and apparatus of configurable processing

A multiple instruction set processor and method dynamically activates one of a plurality of branch prediction processes depending upon which one of a multiple instruction set is operational. Shared branch history table structures are used and are indexed differently depending upon which instruction set is operational. The apparatus and method also allows switching between instruction set index generators for each of the plurality of instruction sets. Accordingly, different indexes to branch prediction data are used depending upon which of the plurality of instruction sets is operational. Shared memory may be used to contain branch prediction table data for instructions from each of the plurality of instruction sets in response to selection of an instruction set. Shared memory is also used to contain branch target buffer data for instructions from each of the plurality of instruction sets in response to selection of one of the instruction sets.

Switching between a plurality of branch prediction processes based on which instruction set is operational wherein branch history data structures are the same for the plurality of instruction sets

An integrated computing system includes at least one processor formed on a substrate, wherein the processor operates at a processor rate. The integrated computing system further includes a global bus that is coupled to the at least one processor and is formed on the substrate. The global bus supports transactions (e.g., data, operational instructions, and/or control signaling conveyances) at a rate that is equal to or greater than the processing rate. The integrated computing system further includes a device gateway and memory gateway that are operably coupled to the global bus and formed on the substrate. The device gateway provides an interface for at least one device (e.g., internal or external) to the global bus. The memory gateway provides an interface between the global bus and memory.

Method of manufacture and apparatus of an integrated computing system

A branch resolution logic for an in-order processor is provided which scans the stages of processor pipeline to determine the oldest branch instruction having sufficient condition codes for resolution. The stages are scanned in order from the latter stages to the earlier stages, which allows quick and simple branch resolution. Therefore, because branches are resolved as soon as the necessary condition codes are generated in a specific stage, branch mispredict penalties are minimized.

Multi-branch resolution

A method and apparatus for busing data elements within a computing system includes processing that begins by providing, on a shared bus, a first control signal relating to a first transaction during a first bus cycle. The processing continues by providing a second control signal relating to a second transaction and a first address signal relating to the first transaction during a second bus cycle. The processing continues by providing a third control signal relating to a third transaction and a second address signal relating to a second transaction during a third bus cycle. The processing then continues by providing a first status relating to the first transaction and a third addressing signal relating to the third transaction during a fourth bus cycle. The processing then continues by providing a second status relating to the second transaction during a fifth bus cycle. The processing then continues by providing first data relating to the first transaction when the first status is a hit and providing third status relating to the third transaction during a sixth bus cycle.

Method and apparatus for busing data elements

A method and apparatus for interfacing a processor with a bus includes processing that begins by storing transactions initiated by the processor into a buffer. The processing then continues by selecting one of the transactions stored in the buffer and placing the selected transaction on the bus. The processing continues by monitoring progress of fulfillment of each transaction in the buffer and flagging a transaction when it has been successfully completed. The processing also includes processing at least two related transactions prior to selecting one of the transactions from the buffer where, if transactions can be processed locally, they do not need to be transported on the bus. In addition, the processing includes monitoring the bus for related transactions initiated by another processor such that these transactions can be more efficiently processed. The related transaction on the bus would correspond to a transaction queued in the buffer.

Method and apparatus for interfacing a processor with a bus

A method and apparatus for both facilitating access to shared memory addresses over a common bus by a plurality of data processors includes detecting, by at least a first processor, that two access addresses are boundary addresses on either side of an address boundary. The method and apparatus locks the common bus in response to detecting the two access addresses. In addition, the method and apparatus locks the two detected addresses based on address probe inquiry data communicated by the first processor. Accordingly, at least one processor employs probe based bus lock and address lock control to facilitate efficient access to shared memory addresses. Preferably, each processor includes probe-based bus lock and address locking control. The method and apparatus provides a type of address locking with deterministic bus locking when needed.

Method and apparatus for providing probe based bus locking and address locking

A high-performance processor is disclosed with structure and methods for: (1) aggressively scheduling long latency instructions including load/store instructions while maintaining precise state; (2) maintaining and restoring state at any instruction boundary; (3) tracking instruction status; (4) checkpointing instructions; (5) creating, maintaining, and using a time-out checkpoint; (6) tracking floating-point exceptions; (7) creating, maintaining, and using a watchpoint for plural, simultaneous, unresolved-branch evaluation; and (9) increasing processor throughput while maintaining precise state. In one embodiment of the invention, a method of restoring machine state in a processor at any instruction boundary is disclosed. For any instruction which may modify control registers, the processor is either synchronized prior to execution or an instruction checkpoint is stored to preserve state; and for any instruction that creates a program counter discontinuity an instruction checkpoint is stored. Instruction execution status is monitored to detect a fault condition and if a fault occurs, the instruction identifier is saved according to predetermined rules and used as an endpoint to backup the program counter. If the faulting instruction is a checkpointed instruction, then the state information is restored from the checkpoint prior to reexecuting the faulting instruction, but if one of the instructions for which state was stored is sequentially between the faulting instruction and the last issued instruction, then (i) backing-up said processor to the closest checkpoint after the faulting instruction, (ii) back stepping the processor to restore state, and (iii) decrementing the program counter to the faulting instruction.

Structure and method for instruction boundary machine state restoration

Apparatus and method provide for tracking and maintaining precise state by assigning a unique identification tag to each instruction at the time of issue, associating the tag with a storage location in a first active instruction data structure, updating the data stored in the storage location in response to instruction activity status changes for each instruction, and maintaining a plurality of pointers to the storage locations that move in response to the instruction activity status. Status information includes an activity data item, such as an activity bit, that is set at the time the instruction is issued and cleared when execution completes without error. Pointers are established that point to the last issued instruction, the last committed instruction pointer, and reclaimed instruction pointer. These three pointers are moved forward toward the later issued (newer) instructions along the data structure based on comparisons of the active-bit for each location associated with one instruction in the data structure and predetermined rules. Exceptions or error conditions for any instruction prevent changing the active-bit so that movement of the pointers is controlled and prevented under these error conditions.

Processor structure and method for tracking instruction status to

A clocked instruction flow is managed subject to issue and fetch constraints through a plurality of instruction latches which receive instructions from selected memory locations. By checking the number of instructions fetched and issued, the fetch program counter is adjusted responsive to the status of selected state variables indicating instructions issued and fetched. The instruction latches are fully scheduled from cycle to cycle with instructions, by fetching instructions in accordance with a fetch program counter.

Method for adjusting fetch program counter in response to the number of

In a microprocessor, an apparatus is included for coordinating the use of physical registers in the microprocessor. Upon receiving an instruction, the coordination apparatus extracts source and destination logical registers from the instruction. For the destination logical register, the apparatus assigns a physical address to correspond to the logical register. In so doing, the apparatus stores the former relationship between the logical register and another physical register. Storing this former relationship allows the apparatus to backstep to a particular instruction when an execution exception is encountered. Also, the apparatus checks the instruction to determine whether it is a speculative branch instruction. If so, then the apparatus creates a checkpoint by storing selected state information. This checkpoint provides a reference point to which the processor may later backup if it is determined that a speculated branch was incorrectly predicted. Overall, the apparatus coordinates the use of physical registers in the processor in such a way that: (1) logical/physical register relationships are easily changeable; and (2) backup and backstep procedures are accommodated.

Growing community of inventors

Sunnyvale, CA, United States of America

Niteen Aravind Patkar