Abstract:
A method for slowing down a high speed microprocessor with an internal cache to maintain compatibility with application software written for slower speed microprocessors. The internal cache of the processor is invalidated during the slowdown and the cache address comparison circuitry is directed to evaluate external addresses for a preset interval, preventing the processor from the accessing the cache, thereby slowing down the processor. The external address evaluation direction is released when a bus requesting device indicates a bus request during the slowdown, allowing the processor to respond to the bus request promptly to prevent possible latency problems from occurring, but still maintaining the processor in a halted state.
Abstract:
A memory mapping and module enabling circuit for allowing logical 128 kbyte memory blocks to be defined for any location in any module (58) connected to a memory system. A RAM (126) is addressed by the system address lines defining 128 kbyte blocks, with the output data (TA, RASEN*) providing the row address strobe enable signals for a particular memory module and the address values necessary to place the 128 kbyte block within the module (58). Various other parameters such as write protect status (HWP) and memory location (HLOCMEM*) are also provided by the RAM. Circuits and techniques for programming and reading the RAM are provided.
Abstract:
A memory mapping and module enabling circuit for allowing logical 128 kbyte memory blocks to be defined for any location in any module (58) connected to a memory system. A RAM (126) is addressed by the system address lines defining 128 kbyte blocks, with the output data (TA, RASEN*) providing the row address strobe enable signals for a particular memory module and the address values necessary to place the 128 kbyte block within the module (58). Various other parameters such as write protect status (HWP) and memory location (HLOCMEM*) are also provided by the RAM. Circuits and techniques for programming and reading the RAM are provided.
Abstract:
A bus arbitration protocol and accompanying bus arbitration logic for multiple-processor computer systems is described in which each processing module (12,12a) has a local cache (16,16a). Several bus arbitration policies are enforced on contending devices which effectively introduce delay states into the arbitration behavior exhibitied by each device. The bus arbitration protocol employs a distributed method of arbitration control involving an essentially fixed prioritization of arbitrating devices.
Abstract:
A data processing system is disclosed which permits the use of a memory mapped floating point coprocessor or other such peripheral, by a process which is unable to address the peripheral's resident memory location but which is able to address a specified location. Memory mapping means intercept attempted accesses by the process to the specified location and diverts the process's access instruction to cause it to access the corresponding mapped address instead. An illustrative implementation is also described, namely the use of a Weitek 1167 memory-mapped, floating point coprocessor by processes running in the "virtual mode" of an Intel 80386 CPU.
Abstract:
A method for slowing down a high speed microprocessor with an internal cache to maintain compatibility with application software written for slower speed microprocessors. The internal cache of the processor is invalidated during the slowdown and the cache address comparison circuitry is directed to evaluate external addresses for a preset interval, preventing the processor from the accessing the cache, thereby slowing down the processor. The external address evaluation direction is released when a bus requesting device indicates a bus request during the slowdown, allowing the processor to respond to the bus request promptly to prevent possible latency problems from occurring, but still maintaining the processor in a halted state.
Abstract:
A memory controller (48) of a personal computer has an asynchronous portion and a synchronous portion. The synchronous portion is used wher the system processor (20) is accessing the memory, while the asynchronous portion is used when control of the memories is held by a DMA controller (48) or a bus master (34) located on a standardised bus (44).
Abstract:
A bus arbitration protocol and accompanying bus arbitration logic for multiple-processor computer systems is described in which each processing module (12,12a) has a local cache (16,16a). Several bus arbitration policies are enforced on contending devices which effectively introduce delay states into the arbitration behavior exhibitied by each device. The bus arbitration protocol employs a distributed method of arbitration control involving an essentially fixed prioritization of arbitrating devices.