An IC includes an array of processor units, arranged in two or more subarrays. A subarray has a test generator, a multiplexer to apply a test vector to a datapath, and a test result output. It includes one or more processor units. A test result compressor is coupled with an output of the datapath, and compresses output data to obtain a test signature, which it stores in a signature register. The signature register is legible from outside the subarray. The datapath includes one or more memories and one or more ALUs. Test data travels through the full datapath, including the memories and the ALUs. ALU control registers are overridden during test to ensure a testable datapath.

A method and apparatus for generating a test bench for verifying a processor decoder are provided. The method including receiving an architecture description comprising processor decoder information, parsing the received architecture description into information for verifying the processor decoder, and generating the test bench to verify the processor decoder based on the parsed information.

A processor includes logic to implement a reconfigurable test access port with finite state machine control. A plurality of test access ports may each include a finite state machine for enabling implementation of different test interfaces to the processor, including JTAG IEEE 1149.1, JTAG IEEE 1149.7, and serial wire debug.

A debug-enabled processing device includes a processor, a communication transceiver circuit, and a debug support unit. The debug support unit has a plurality of dedicated debug registers to facilitate debugging a software program under execution by the processor. One of the plurality of debug registers is a control register having at least four bits, which are used to enable/disable a plurality of debugging operations. Others of the debug registers include a set of index registers that may be configured to pass data to and from the processor.

Temporary fault injection to existing hardware is performed using only software without changing an implementation of the hardware. A fault injection interrupt process starts on an operation of a CPU using an interrupt that is not used by software, and an internal state of hardware is updated to the same value as a result obtained when a fault has occurred during the interrupt process. A clock of the CPU during the interrupt process is accelerated so that a period of time of the interrupt process is smaller than a period of time until a fault becomes effective.

Embodiments relate to testing memory write operations. An aspect includes detecting a first write operation to a set of “n” divisions in a memory table, and defining a selected set of entries of an optimization checking table corresponding to the set of “n” divisions of the memory table. The aspect includes determining that at least one selected entry of the selected set of entries is not among an optimal set of entries of the checking table. The aspect further includes determining whether to generate an optimization error or to end an optimization analysis of the first write operation without generating the optimization error by comparing the first time stamps of one or both of the at least one selected entry and one or more optimal entries of the optimal set of entries to a temporal window defined by a predetermined duration.

An apparatus includes a main core processor configured to receive a first signal through a first main buffer, a second signal through a second main buffer, a third signal through a third main buffer and a fourth signal through a fourth main buffer, a shadow core processor configured to receive the first signal through a first shadow buffer, the second signal through a second shadow buffer, the third signal through a third shadow buffer and the fourth signal through a fourth shadow buffer, and a first glitch suppression buffer coupled to a common node of an input of the first main buffer and an input of the first shadow buffer.

Systems, apparatuses, and methods for implementing a safety monitor framework for a safety-critical computer vision (CV) application are disclosed. A system includes a safety-critical CV application, a safety monitor, and a CV accelerator engine. The safety monitor receives an input image, test data, and a CV graph from the safety-critical CV application. The safety monitor generates a modified image by adding additional objects outside of the input image. The safety monitor provides the modified image and CV graph to the CV accelerator which processes the modified image and provides outputs to the safety monitor. The safety monitor determines the likelihood of erroneous processing of the original input image by comparing the outputs for the additional objects with a known good result. The safety monitor complements the overall fault coverage of the CV accelerator engine and covers faults only observable at the level of the CV graph.

A fault recoverable computer system including an instruction table having a plurality of processor instructions. The system also includes at least one sensor arranged to monitor an environmental condition and output sensor data. A monitor module is arranged to receive sensor data and/or processor state information. A testing module is arranged to perform a plurality of self-tests including a first self-test of the computer system and, if the first self-test fails, output a failure notification. A recovery module is arranged to update the instruction table in response to receiving the failure notification. The update includes replacing a first processor instruction arranged to perform a first function with a replacement set of processor instructions configured to alternatively perform the first function.

A system to implement debugging for a multi-threaded processor is provided. The system includes a hardware thread scheduler configured to schedule processing of data, and a plurality of schedulers, each configured to schedule a given pipeline for processing instructions. The system further includes a debug control configured to control at least one of the plurality of schedulers to halt, step, or resume the given pipeline of the at least one of the plurality of schedulers for the data to enable debugging thereof. The system further includes a plurality of hardware accelerators configured to implement a series of tasks in accordance with a schedule provided by a respective scheduler in accordance with a command from the debug control. Each of the plurality of hardware accelerators is coupled to at least one of the plurality of schedulers to execute the instructions for the given pipeline and to a shared memory.