Abstract: Disclosed is a circuit for simultaneously searching two ends of a vector. The circuit comprises at least one input for receiving a vector of head pointers. A first input of a memory latch receives the vector of head pointers. An input of a first priority decoder receives the vector of head pointers from the memory latch. The first priority decoder traverses the vector of head pointers from a first end of the vector for identifying one active bit in the vector. An input of a first reverse module also receives the vector of head pointers from the memory latch. An input of second priority decoder receives the vector of head pointers (in reverse order) from the first reverse module. The second priority decoder traverses the vector received from the first reverse module from a first end of the reversed vector for identifying one active bit in the vector.

Abstract: A method and a device for processing instructions based on register group size information includes a pipelined processor, an instruction memory unit and a register file, whereas the pipelined processor includes a write-back unit and an execution unit. The device is characterized by including a controller that is adapted to receive a first register group size information and a first register identification information that define a first group of source registers associated with a first instruction; and to determine an execution related operation of the first instruction in response to the first register group size information, the first register identification information, a second register group size information and a second register identification information. The second register group size information and the second register identification information define a second group of target registers associated with a second instruction.

Abstract: An extended DRAIN instruction is used to stall processing within a computing environment. The instruction includes an indication of the one or more processing stages at which processing is to be stalled. It also includes a control that allows processing to be stalled for additional cycles, as desired. The enhanced DRAIN instruction enables its use to be more granular and to have minimal impact with respect to each individual use.

Abstract: In-lane vector shuffle operations are described. In one embodiment a shuffle instruction specifies a field of per-lane control bits, a source operand and a destination operand, these operands having corresponding lanes, each lane divided into corresponding portions of multiple data elements. Sets of data elements are selected from corresponding portions of every lane of the source operand according to per-lane control bits. Elements of these sets are copied to specified fields in corresponding portions of every lane of the destination operand. Another embodiment of the shuffle instruction also specifies a second source operand, all operands having corresponding lanes divided into multiple data elements. A set selected according to per-lane control bits contains data elements from every lane portion of a first source operand and data elements from every corresponding lane portion of the second source operand. Set elements are copied to specified fields in every lane of the destination operand.

Abstract: A multiport semiconductor memory device having a processor wake-up function and multiprocessor system, the multiprocessor system including a first processor configured to perform a first predetermined task; a second processor configured to perform a second predetermined task; and a multiport semiconductor memory device coupled to the first and second processors. The multiport semiconductor memory device includes a memory cell array having at least one shared memory area; a first port coupled to the at least one shared memory area; a second port coupled to the at least one shared memory area; and a wake-up signal generator. The first processor is coupled to the at least one shared memory area via the first port, the second processor is coupled to the at least one shared memory area via the second port, and the wake-up signal generator is coupled to the first processor and the second processor.

Abstract: A method for recovering global history shift register (GHSR) and return address stack (RAS) is provided, which is applicable to an instruction pipeline of a processor and includes the following steps. First, provide a branch recovery table (BRT) and a backup stack. Whenever a branch instruction enters a predetermined stage of the instruction pipeline, add a record in the BRT according to the branch instruction. Whenever a return address is popped from the RAS of the instruction pipeline, push the return address into the backup stack. When flushing the instruction pipeline, determine a removal range of the BRT according to the condition which triggers the pipeline flush. Recover the RAS according to the records in the removal range and the backup stack. Remove all records in the removal range. Recover the GHSR of the instruction pipeline according to the removed records.

Abstract: A data processing device includes a program execution section to supply an operation direction signal to a peripheral device based on an executed program and execute a branch operation in response to a branch direction signal, and a branch wait operation section to receive the branch direction signal and a peripheral device status notification signal indicating whether an operation performed in the peripheral device is being executed. The branch wait operation section outputs an instruction issue stop signal directing waiting of the branch operation to the program execution section if the branch direction signal is input during a period when the peripheral device status notification signal is active indicating that the operation in the peripheral device is being executed.

Abstract: Systems and methods for accelerators which may execute a program in conjunction with a PC are disclosed. In one embodiment, an accelerator includes a plurality of calculation units, each calculation unit operable to execute the program in parallel, an operation control unit configured to control an operation capability or a processing capability of at least one of the plurality of calculation units, and a control unit configured to determine a corresponding operation capability or processing capability for the plurality of calculation units based on load information on the program to be executed and control the operation control unit accordingly.

Abstract: A processor comprising fetch logic adapted to fetch instructions from memory and decode logic coupled to the fetch logic and adapted to decode the fetched instructions. If a bit in the decode logic is in a first state, a particular fetched instruction is skipped and a group of one or more instructions is executed in lieu of the particular fetched instruction. If the bit is in a second state, both the group and the particular fetched instruction are executed.

Abstract: Provided are a central processing unit (CPU) and method for executing a branch instruction of a CPU, which can protect user's data by preventing an error due to a computer virus and a hacker is provided. The CPU includes: a branch instruction verification unit which verifies whether a branch instruction is valid; and a branch instruction execution unit which executes the branch instruction when the branch instruction is valid. The method includes: verifying whether the branch instruction is valid; and not executing the branch instruction when the branch instruction is invalid.

Abstract: A system to prevent livelock. An outcome of an event is predicted to form an event outcome prediction. The event outcome prediction is compared with a correct value for a datum to be accessed. An instruction is appended with a real event outcome when the outcome of the event is mispredicted to form an appended instruction. A prediction override bit is set on the appended instruction. Then, the appended instruction is executed with the real event outcome.

Abstract: An asynchronous ripple pipeline has a plurality of stages, each with a controller (18) and a register (16). The controller has a register control output (21), and a combined acknowledgement and request output (20), together with a request input (22) and an acknowledgement input (24). The protocol used has a single signal, output on the combined acknowledgement and request output (20) of a stage (30), that functions both as a request to the next stage (32) and an acknowledgement to the previous stage (34).

Abstract: Trap flags and a pointer trap are associated with registers in a processor. Each trap flag indicates whether a corresponding register has been written with valid data. If not, the trap flag is set to indicate that the register corresponding to the trap flag contains invalid data. During instruction processing, the pointer trap receives control signals from instruction fetch/decode logic on the processor indicating an instruction being processed calls for a register to be used as a pointer. If the specified pointer register has its corresponding trap flag set, then the pointer trap indicates that a processing exception has occurred. The interrupt logic/exception processing logic then causes a trap interrupt service routine (ISR) to be executed in response to the exception. The ISR prevents errors from being introduced in the instruction processing due to invalid pointer values.

Abstract: A device and method may fetch an instruction or micro-operation for execution. An indication may be made as to whether the instruction is dependent upon any source values corresponding to a set of previously fetched instructions. A value may be stored corresponding to each source value from which the first instruction depends. An indication may be made for each of the set of sources of the instruction, whether the source depends on a previously loaded value or source, where indicating may include storing a value corresponding to the indication. The instruction may be executed after the stored values associated with the instruction indicate the dependencies are satisfied.

Abstract: A pipeline processor has circuits to detect the presence of a register access instruction in an issue stage of the pipeline. A load-miss occurring at a later stage may cause the register access instruction to be marked with an associated bit. The register access instruction progresses down the pipeline and when the flush stage is reached, the processor checks the associated bit and flushes the register access instruction.

Abstract: An electronic apparatus includes a plurality of stages serially interconnected as a pipeline to perform sequential processings on input operands. A shortening circuit associated with at least one stage of the pipeline recognizes when one or more of input operands for the stage has been predetermined as appropriate for shortening and execute the shortening when appropriate.

Abstract: In one embodiment, the present invention includes a retirement unit to receive and retire executed instructions. The retirement unit may include a first array to receive information at allocation and a second array to receive information after execution. The retirement unit may further include logic to calculate an event associated with an executed instruction if information associated with the executed instruction is stored in an on-demand portion of at least one of arrays. Other embodiments are described and claimed.

Abstract: A processor performs a prefetch operation on non-sequential instruction addresses. If a first instruction address misses in an instruction cache and accesses a higher-order memory as part of a fetch operation, and a branch instruction associated with the first instruction address or an address following the first instruction address is detected and predicted taken, a prefetch operation is performed using a predicted branch target address, during the higher-order memory access. If the predicted branch target address hits in the instruction cache during the prefetch operation, associated instructions are not retrieved, to conserve power. If the predicted branch target address misses in the instruction cache during the prefetch operation, a higher-order memory access may be launched, using the predicted branch instruction address. In either case, the first instruction address is re-loaded into the fetch stage pipeline to await the return of instructions from its higher-order memory access.

Abstract: An input/output system transfers data packets to and from a SIMD array of processing elements (PEs) such that different sizes of data packets are transferred to respective ones of the PEs. The packets are transferred in batches to respective different addresses in the array under the control of the PEs. Transfer to or from the array may be carried out when either a batch or part of a batch is ready for transfer. The decision to transfer either full or part batches is made in dependence upon the speed of the PEs and the speed and intermittency of the data packets.

Abstract: In general, in one aspect, a processing unit includes an input buffer to store data received by the processing unit, a memory, an arithmetic logic unit coupled to the input buffer and to the memory, an output buffer; and control logic having access to a control store of program instructions, the control logic to process instructions including an instruction to transfer data from the input buffer to the memory and an instruction to cause the arithmetic logic unit to perform an operation on operands provided by at least one of the memory and the input buffer, the instruction to output results of the operation to at least one of the memory and the output buffer.