Abstract: The solution proposed in this invention is a nearest neighborhood access protocol, where not every processor is given access to every other memory block. It is shown by analyzing the pipeline that it is adequate to have no more than two masters (CPU's) in particular and 3 CPU's in general. In the case of the 2 CPU approach one of these CPU's is a producer, and the other CPU is a consumer. In the 3 CPU case the third owner may be a DMA channel.

Abstract: An accurate and simple benchmarking method for multiple processor systems. Instead of a central timer as used in the prior art, a counter is implemented in each processor that counts the processor's clock cycles. The counter may be read after the processor's completes a benchmark task. This eliminates the timing skew common in the prior art.

Abstract: A new mechanism submits multiple DMA requests that are becoming more common in the newer video codec standards. This feature improves system performance and allows bus accesses to be more efficient. An artificial burst is created by aggregating multiple requests which normally would be distributed to be more localized in time, thus creating a burst of traffic.

Abstract: This invention is applicable to filtering block artifacts of macroblock and block oriented video compression. This invention computes all possible filter results speculatively and simultaneously in parallel, computes conditions for application of corresponding filter results simultaneously in parallel, and writes filter results to memory conditionally dependent upon computed corresponding conditions. This invention permits effective block filtering on a very long instruction word data processor.

Abstract: A syndrome evaluation with partitioning of a received block of symbols into subsets and interleaved partial syndrome evaluations to overcome multiplier latency. Parallel syndrome evaluations with a parallel multiplier.

Abstract: This invention decodes a next significance symbol using a selected context. The invention operates in a loop for each symbol decode for a whole block until the number of decoded map elements reaches a maximum number of coefficients for the block type or a last significant coefficient marker is decoded updating loop variables accordingly. This invention counts the number of decoded significance symbols indicating a significant coefficient and stores the locations of such significant coefficients in an array. An embodiment of this invention estimates the number of significant coefficients in a block and selects the inventive method or a prior art decode method.

Abstract: This invention decodes coefficient magnitudes in compressed video data using a selected context and speculatively decodes a coefficient sign. The next context selection depends upon a number of iterations. This invention confirms the speculatively decoded coefficient sign upon completion of the magnitude decode. This invention operates in a loop until reaching the number of significant coefficients within the block. The method exits the loop and decodes an escape code if an iteration count is greater than a predetermined number. An embodiment of this invention collects both a count up and a count down in an escape code decode in one loop. An embodiment of this invention estimates the number of significant coefficients in a block and selects the inventive or a prior art decode.

Abstract: This invention is a method for inverse Wavelet transform using a breadth-first output data calculation which uses input data to calculate at least one output data for each iteration of a software loop even if the same input data is used in a later iteration for calculating other output data. This reduces data movement between memory and the data processor core thus reducing the possibility of cache misses and memory stalls due to access conflicts. The input data and computed output data are preferably stored as subwords packed within data words in memory. In inverse Wavelet transformation this method performs vertical spatial frequency expansion and horizontal spatial frequency expansion for each level of Wavelet encoding. This invention arranges data flow providing a more efficient use of memory bandwidth and cache space than other known methods.

Abstract: A data processing apparatus execution unit includes a multiplexer having inputs receiving data from sections of a source data register or registers. The multiplexer selects data from one section to store in a destination data register. The execution unit may zero extend or sign extend the remaining most significant bits of the destination data. In an alternative embodiment, the execution unit includes plural multiplexers, one for each section of the destination data. Each multiplexer received data from each section of the source data register or registers. Special codes in the sections of the second source data register may select 0 fill, 1 fill or sign extension from the next most significant section for each multiplexer.

Abstract: A computer implemented method accesses data in a look up table via an index divided into a coarse index and a fine index. The method calculates the coarse index and fetches plural look up table entries. While the fetch is in progress, the method calculates the fine index. The method determines and extracts the look up table entry corresponding to the combined coarse and fine indexes Fetching plural look up table entries includes executing a single word load instruction, a doubleword load instruction or a pair of sequential or simultaneous single word load instructions. Extracting the look up table entry may include a section select move instruction, an extract and zero-extend bit field instruction or a shift and mask operation.

Abstract: Context-based adapative binary arithmetic coding (CABAC), as used in video standards such as H.264/AVC, with a renormalization of the interval low value plus range that includes partitioning of the bits of the low value to provide output bits plus low value update without bit-level iterations or aggregation of output bits until a full byte can be output.

Abstract: An information carrier medium containing software that, when executed by a processor, causes the processor to receive information from circuit logic that is adapted to collect the information from caches on different cache levels, the information associated with a common address. The software also causes the processor to provide the information to a user of the software. The information comprises cache level and cache type information associated with a particular cache from one of the different cache levels.

Abstract: An information carrier medium containing software that, when executed by a processor, causes the processor to receive information from circuit logic that is adapted to collect the information from caches on different cache levels, at least some of the information from caches on different cache levels associated with a common address. The software also causes the processor to selectively bypass a portion of the information specified by a user of the software and to provide non-bypassed information to the user and not said bypassed portion.

Abstract: An information carrier medium containing software that, when executed by a processor, causes the processor to receive information from circuit logic that is adapted to collect the information from caches on different cache levels. At least some of the information is from caches on different cache levels associated with a common address. The processor also displays the information by way of a graphical user interface (GUI). The GUI displays a portion of the information using a mark-up technique different from that used to display remaining portions of the information.

Abstract: An information carrier medium containing software that, when executed by a processor, causes the processor to receive information from caches on different cache levels, at least some of the information from caches on different cache levels associated with a common address. The software also causes the processor to determine a difference between the information from caches on different cache levels associated with the common address and to provide the difference to a user of the software.

Abstract: This invention prevents illegal memory address faults on speculative data loads. Circular addressing of the address pointer limits memory access to a range of addresses including all addresses used by the address pointer and not including any invalid addresses. The invention uses circular addressing hardware, if available on the data processor. If not available, this invention simulates circular addressing. This invention permits loads to be issued earlier than if predication were used and allows already predicated loads to be speculated without the overhead of a compound predicate. This invention can be used on processors without hardware supporting speculation.

Abstract: This invention modifies an irregular software pipelined loop conditioned upon data in a condition register in a compiler scheduled very long instruction word data processor to prevent over-execution upon loop exit. The method replaces a register modifying instruction with an instruction conditional upon the inverse condition register if possible. The method inserts a conditional register move instruction to a previously unused register within the loop if possible without disturbing the schedule. Then a restoring instruction is added after the loop. Alternatively, both these two functions can be performed by a delayed register move instruction. Instruction insertion is into a previously unused instruction slot of an execute packet. These changes can be performed manually or automatically by the compiler.

Abstract: A data processing apparatus execution unit includes a multiplexer having inputs receiving data from sections of a source data register or registers. The multiplexer selects data from one section to store in a destination data register. The execution unit may zero extend or sign extend the remaining most significant bits of the destination data. In an alternative embodiment, the execution unit includes plural multiplexers, one for each section of the destination data. Each multiplexer received data from each section of the source data register or registers. Special codes in the sections of the second source data register may select 0 fill, 1 fill or sign extension from the next most significant section for each multiplexer.

Abstract: A computer implemented method accesses data in a look up table via an index divided into a coarse index and a fine index. The method calculates the coarse index and fetches plural look up table entries. While the fetch is in progress, the method calculates the fine index. The method determines and extracts the look up table entry corresponding to the combined coarse and fine indexes Fetching plural look up table entries includes executing a single word load instruction, a doubleword load instruction or a pair of sequential or simultaneous single word load instructions. Extracting the look up table entry may include a section select move instruction, an extract and zero-extend a bit field instruction or a shift and mask operation.

Abstract: This invention increases the available instruction level parallelism (IPC) of CABAC encoding by decoupling the re-normalization loop and the bit-insertion task required to create the encoded bit-stream. This makes all software implementations of CABAC based encoding significantly faster on digital signal processors that can exploit instruction level parallelism such as very long instruction word (VLIW) digital signal processors. In a joint hardware/software implementation, this invention employs existing Huffman variable length encoding hardware with minimum modifications. The de-coupling of these two tasks of this invention exposes previously hidden underlying instruction level parallelism and task level parallelism.