Abstract: A data processing system is described in which trace signals are provided upon a trace bus 12 to track the address of an instruction code currently being executed and the latest address to which a data access was made. The system incorporates a central processing unit core 14 and an instruction pipeline 16 via which instruction codes are fed to the central processing unit core 14. When a non-sequential instruction code fetch is made, a number of cycles must pass before that non-sequential instruction has made its way along the instruction pipeline 16 to the central processing unit core 14. This period is utilised to output the address of the non-sequential instruction code fetch upon the trace bus. The multiple cycles available for this allow a time division multiplexing technique to be employed for different portions of the address thereby enabling the trace bus to be narrower.

Abstract: An integrated circuit includes a plurality of data handling devices and a data buffer for enabling transfer of data between the internal data handling devices and one or more external data handling devices external to the integrated circuit. A controller responds to an original clock signal for supplying a clock signal to control data transfer between the data handling devices. The controller includes a delay circuit operable to delay the original clock signal to generate a delayed clock signal, and includes a selector for inhibiting operation of the delay circuit and for selecting the original clock signal for controlling data transfer from an internal data handling device to another data handling device. The selector also enables operation of the delay circuit and selects the delayed clock signal for controlling data transfer from an external data handling device to an internal data handling device.

Abstract: An integrated circuit test mechanism based upon the JTAG standard utilises serial scan chains for applying signals to and capturing signals from predetermined nodes within an integrated circuit (2). Multiple independent scan chains (12, 14, 16) are provided for different circuit units (4, 6, 8, 10) within the integrated circuit, i.e. individual scan chains (12, 14) for circuit elements such as a central processing core (4) or a cache memory (8). The scan chain controller (18) is responsive to a scan chain selecting instruction (Scan-N) received at its serial input (20) to capture a scan chain specifying value at the serial input. The scan chain specifying value is then used to control the position of a scan chain multiplexer (28) that selects one of the multiple scan chains to which subsequent instructions received at the serial input are applied.

Abstract: A data memory having an addressable array of memory cells which can be accessed as predetermined groups of memory cells, comprises output buffer means for storing the contents of at least the most recently read group of memory cells and another previously read group of memory cells; and reading means, responsive to an indication that the contents of the group of memory cells containing the required memory cell is not stored in the output buffer means, for reading the contents of the group of memory cells containing the required memory cell into the output buffer means; the contents of at least the required memory cell being supplied as an output from the output buffer means.

Abstract: A data processing system is provided in which the central processing unit clock signal (mclk, fclk) to a central processing unit core (14) may be suspended to reduce power consumption. This suspension is controlled by a suspend controller (20) that responds to a write request to a predetermined address (0x0320001C) to hold asserted a bus request signal (REQ) that cooperates with a bus controller (18) to block the central processing unit clock signal. The central processing unit core sees the suspend mode as a write request of an indefinite length. The suspend controller is responsive to an asynchronous input signal (FIQ, IRQ, EVENT1) to exit the suspend mode by issuing a bus acknowledge signal (ACK) and removing the block on the central processing unit clock signal.

Abstract: An integrated circuit is described having a scan chain of the JTAG type in which there are provided a plurality of serially connected test cells 2. The test cells serve the additional function of operating during the normal operation of the integrated circuit to store signal values that are logically combined (compared) with signal values generated by the integrated circuit to yield control signals into for controlling the operation of the integrated circuit. This allows the storage capacity of the test cells to be utilised during normal operation when they would otherwise be idle.

Abstract: A data memory includes a plurality of memory cells that are selectively operable during a memory read operation to generate a data signal indicative of data stored in that memory cell, and one or more sense amplifiers are connected to receive data signals from a plurality of the memory cells. An address decoder is responsive to a memory address supplied to the data memory for controlling one of said plurality of memory cells connected to each of the sense amplifiers to generate a respective data signal under the control of a first control signal. The data memory also includes circuitry for delaying the first control signal to generate a second control signal, and for energizing the one or more sense amplifiers for operation in response to the second control signal.

Abstract: A data processing system is described utilizes a multiplier-accumulator 108 that performs both a first class of multiply-accumulate instructions and a second class of multiply-accumulate instructions. The first class of multiply-accumulate instructions are of the form N*N+N.fwdarw.N and the second class of multiply-accumulate instructions are of the form N*N+2N.fwdarw.2N. The second class of multiply-accumulate instructions provide a greater precision of arithmetic in a single instruction and avoid the use of excessive instruction set space by being constrained that the result is written back into the two registers from which the 2N-bit accumulate value was taken. The multiplier-accumulator also provides N*N.fwdarw.N and N*N.fwdarw.2N multiplication operations.

Abstract: Data processing apparatus for executing successive data processing instructions comprises a processing core having a current operational state selected from a predetermined set of possible operational states, the current operational state being defined by a control state signal supplied to the core; a synchronous state machine circuit for generating an output state signal, indicating a provisionally valid next operational state of the core, in response to a predetermined phase of a current clock cycle of a clocking signal, the output state signal being dependent upon a current operational state of the core and control signals generated by the core before the predetermined phase of the current clock cycle indicative of a next data processing instruction to be executed by the core; and an asynchronous logic circuit for generating the control state signal in response to the output state signal and late control signals received after the predetermined phase of the current clock cycle.

Abstract: A data processing system is described utilising multiple instruction sets. The program instruction words are supplied to a processor core 2 via an instruction pipeline 6. As program instruction words of a second instruction set pass along the instruction pipeline, they are mapped to program instruction words of the first instruction set. The second instruction set has program instruction words of a smaller bit size than those of the first instruction set and is a subset of the first instruction set. Smaller bit size improves code density, whilst the nature of the second instruction set as a subset of the first instruction set enables a one-to-one mapping to be efficiently performed and so avoid the need for a dedicated instruction decoder for the second instruction set.

Abstract: A data memory includes a memory cell that operates to generate two complementary data outputs indicative of a data bit stored in the memory cell. First and second sense amplifiers set an output latch to a first output logic state in response to a first state of the complementary data output signals, and set the output latch to a second output logic state in response to a second state of the data output signals. One or the other, but not both of the sense amplifiers change state in response to the complimentary data outputs received thereby during a memory read operation.

Abstract: An iterative multiplier having a multiplier core generating partial results upon each iteration. When an early terminate of a multiply instruction occurs, at least one of the partial results is passed to a general purpose barrel shifter for bit realignment dependent upon the number of iterations performed before the early terminate occurred. The bit realigned partial results are then passed to an arithmetic logic unit where they are added to yield the final result.

Abstract: A multiply-accumulate circuit is described in which upon each multiply iteration some of the accumulate value bits are incorporated into the result with lower order subsequently non-changing bits being latched. In this way, a wide accumulate value can be dealt with without incurring a correspondingly wide data path through the multiply accumulate circuit. Initialisation of the multiply accumulate circuit with one of the carry value and save value as the first partial summand and the other as at least part of the accumulate value is performed to reduce the total number of iterative cycles required to produce the result.

Abstract: An integrated circuit comprises a plurality of data handling devices interconnected to exchange data by an interconnection bus during a first mode of operation; and diagnostic control means, responsive to a diagnostic device external to the integrated circuit, for controlling one or more of the data handling devices to exchange diagnostic data with the diagnostic device via the interconnection bus during a second, diagnostic, mode of operation.

Abstract: A CPU core 4 can operate at either an internal clock frequency fclk or an external clock frequency mclk. When operating at the internal clock frequency fclk, write request signals are buffered in a write buffer 10. When operating at the external clock frequency mclk, write request signals are unbuffered. In order to avoid write request signals reaching a signal bus 6 out of order, an interlock is provided between the two paths so that any pending write request signals in the write buffer 10 will serve to hold off any write request signals that may issue through the other path. When a write request signal generated at the external clock frequency is blocked, this serves to stall the CPU core 4 since the blocked external clock write request signal may give rise to an externally generated abort which would alter subsequent processing.

Abstract: A data processing system having a first circuit and a second circuit that together control a third circuit by a respective first control signal and a second control signal. The first circuit issues a request signal to the second circuit to trigger initiation of the operation of the third circuit and the second circuit returns a grant signal to the first circuit to indicate that operation of the third circuit has completed. An advance controller within the second circuit serves to start to synchronize the grant signal back to the clock signal of the first circuit at one of a plurality of possible times that is selected to match the relative frequencies of the clock signals driving the first circuit and the second circuit.

Abstract: A pipeline processor 2 having an associated branch cache 4 is provided. Each cache line 12 of the branch cache stores a cache TAG, a next branch data value R, a target address value TA and a target instruction value TI. The next branch data value indicates when the next branch instruction will be encountered in the stream of instructions fed to the pipeline processor. This data is used such that following a branch cache hit, no further reading of the branch cache is made until the next branch data indicates that the next branch instruction should have been reached. At this stage, the branch cache 4 is read to see if it contains corresponding data for that next branch instruction that will avoid the need to decode that next branch instruction before instructions from the target address of that branch instruction can be fed into the pipeline. The avoiding of the need to read the branch cache for every instruction fed into the pipeline saves power.

Abstract: A memory circuit, in which test data are compared with stored data, comprises a plurality of memory cells each having two complementary data outputs indicative of a respective stored bit of the stored data. The two complementary outputs are selectively interchanged, in response to a respective test bit of the test data. An output signal is then generated (e.g. by a sense amplifier) in response to the relative polarities of the two complementary data outputs. The output signal is indicative of whether the stored bit is equal to the test bit. Where a multi-bit word is stored in a plurality of the memory cells, the output signals generated by a comparison of each stored bit of the multi-bit word and respective bits of the test data are combined by, for example, an AND gate. The output of the AND gate indicates whether the test data matches the stored multi-bit word.

Abstract: An output signal driver is described in which a functional block 10, such as a random access memory, provides a data signal to one input of a multiplexer 26 and then via a buffer stage 24 to an output signal line 6. The other input of the multiplexer 26 receives a recirculated value from the output signal line 6. In operation, the buffer stage 24 is enabled as soon as the functional block 10 is enabled. A data valid signal dv indicating when the data signal from the functional block is ready for output is used to switch the multiplexer 26 from selecting the recirculated value to selecting the new data signal value. The output signal on the output signal line 6 is made to always have a defined state derived from the existing output signal level or, when it becomes available, the new data signal. Power wastage through the driving of spurious signals on the output signal line 6 is thus avoided.

Abstract: An apparatus for mapping a logical pixel value 2 defining pixel appearance to a plurality of physical appearance component values for driving a display device is described. Read addresses for a plurality of palettes 4, 6, 8 and 10 are derived from bits of the logical pixel value. The portions of the logical pixel value which are used to provide these read addresses overlap. Multiple storage of given physical appearance values GPV0, GPV1, . . . within the palettes is provided so as to ensure the appropriate output irrespective of what particular value a bit has that is non-significant for that palette. Three component color palettes 4, 6 and 8 addressed with 8-bit addresses are provided together with an effects palette 10 addressed by a 4-bit address. In the case of a 16-bit logical pixel value, the system has the flexibility to support any of 6-5-5, 5-6-5 or 5-5-6 depending upon what is stored within the color palettes.