Method for dynamically loading embedded type Java processor microcode instruction set

Abstract

The invention discloses a java processor microcode instruction set dynamic loading technique, which belongs to the embedded processor design field. The invention increases an instruction set loading module on the basis of an original processor. The technique comprises: firstly reading microcode instructions from an external memory by a loading module, registering the microcode instructions in a temporary register of the loading module, then, solving the problem that the digit numbers of the external memory and an internal memory disaccord through the microcode mosaic technique, finally, writing the microcode instructions which are spliced well into the internal memory, and starting executing starting codes to enter the normal working state by the system when all internal memories finish initialization, namely after an instruction set is loaded to the inside of a chip. The invention not only enables Java programs of other platforms to operate accurately on the platform to guarantee the universality of the upper application, but also can update and optimize a local instruction set aiming at different embedded application environments to modify the starting codes on real time and to improve the efficiency of the processor.

Description

technical field [0001] The invention belongs to the field of embedded processor design, and in particular relates to a dynamic loading technology of a Java processor microcode instruction set. Background technique [0002] Existing Java processors or Java coprocessors solidify the local instruction set in the chip, and cannot optimize the instruction set. For a general-purpose processor, such as a processor in a PC, this method can meet application requirements, because an upper-layer application based on a general-purpose processor is supported by a fully functional operating system. Java processors are often used in embedded systems. Considering the operating speed and system cost, the corresponding operating system or control functions are relatively simple, and corresponding adjustments are required for different applications. [0003] The process of the operating system executing a program is directly mapped to the process of executing the native instruction set at the...

AI Technical Summary

Problems solved by technology

However, these solutions can only be achieved by reprogramming the FPGA through changes in the HDL code, not dynamic loading in the true sense.

Moreover, the cost and performance of FPGA, compared with special-purpose processors, are not suitable for use in embedded systems as mass-produced products, and can only be used as a verification platform in the research and development process.

Images