X graphics system parallel acceleration method based on FT processor

Abstract

The invention discloses an X graphics system parallel acceleration method based on an FT processor. The implementation steps includes that 1) input and output devices are initialized through an X server main thread, an input event processing subsystem thread is created, an X client program request, a display card management and display output processing request and an input device event are respectively listened, the X client program request is responded through the X server main thread, a display card management drawing subsystem thread used for processing the display card management and processing display output is created by the X server main thread, and the display card management and processing display output request is executed by the display card management drawing subsystem thread; the input device event is read and responded by the input event processing subsystem thread. According to the X graphics system parallel acceleration method based on the FT processor, the X graphic system performance of the FT processor is improved through the multi-core and multi-thread advantage of the FT processor, and the method has the advantages that the utilization rate of hardware resource is high, the user experience is smooth, and the graphic processing performance is high.

Description

technical field [0001] The invention relates to a graphics system performance optimization technology of a Feiteng processor, in particular to a parallel acceleration method of an X graphics system based on a Feiteng processor. Background technique [0002] For an operating system with a graphics system, the graphics system is usually the most complex subsystem, and the performance of the graphics system greatly affects the user experience and has a crucial impact on the overall performance of the operating system. The Feiteng processor (FT processor) platform is a multi-core processor platform. A notable feature of this processor platform is that it contains multiple processor cores, and each processor core supports hardware threads. For such a processor platform, Said that using the method of application parallelization is an important way to obtain performance improvement. [0003] At present, the operating system used by the Phytium processor platform is the Kirin opera...

AI Technical Summary

Problems solved by technology

However, for an interactive graphics system, the processing of input device events and graphics update drawing are very frequent. When there are a large number of graphics applications (that is, X client programs) and more interactive processes in the system, The X server needs to process a large number of X client requests, input device requests, and notify the graphics card to redraw the graphics at the same time.

Traditionally, the X server is a single-threaded program. In this thread, requests from the X client, input device events, and graphics card drawing operations are managed. When there are many requests from the X client or user interaction operations, the X server will appear. It is too late to process all applications and some graphics applications cannot obtain enough CPU time. This situation will cause the X server to be blocked, and the requests of some X programs cannot be processed in time, which is manifested in user experience. "Card" phenomenon

Images