Multi-target-oriented fine-grained compilation self-tuning method

A fine-grained, multi-objective technology, applied in the field of compilation and optimization, which can solve problems such as difficulty in meeting high-performance user requirements, large compilation and optimization combination solution space, and difficulty in fine-grained tuning.

Active Publication Date: 2021-03-19
JIANGNAN INST OF COMPUTING TECH
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

On the one hand, the solution space of compilation and optimization combinations is huge, making it difficult to perform fine-grained tuning for spe

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Multi-target-oriented fine-grained compilation self-tuning method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0031] Embodiment: a fine-grained compiling self-tuning method oriented to multi-objectives, comprising the following steps:

[0032] S1. According to the tuning requirement q, determine the goal of compilation optimization, as follows:

[0033] S11. The user specifies the tuning code p and tuning requirement q;

[0034] S12. According to the tuning requirement q, in the configuration file of the performance analysis tool, search the corresponding relationship table between the tuning requirement q and the performance counter through a script, determine the goal of compilation optimization, and set the feedback parameters of self-tuning;

[0035] S2. Identify the core segment of the code, that is, identify the core segment function with a long execution time, as follows:

[0036] S21. Call the performance analysis tool configured in the software stack of the target system to analyze the dynamic characteristics of the tuning code p, and obtain the calling relationship of the f...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a multi-target-oriented fine-grained compilation self-tuning method, which comprises the following steps: S1, determining a compilation optimization target according to a userdemand; s2, performing code core segment identification, namely splitting the execution time of the program, sorting core segment functions in the program according to the execution time of the program, and selecting a plurality of core segment functions with the longest execution time of the program according to a set threshold n; s3, performing instrumentation on the code, and selecting a corresponding instrumentation interface for quantitatively evaluating a compiling optimization effect according to a compiling optimization target; s4, modifying the configuration file, and starting a self-tuning operation; and S5, performing clustering analysis on the historical information of the tuning result, and modifying a heuristic strategy of an iterative process. According to the invention, theperformance potential of compiling optimization options can be mined, the user friendliness of a compiler is improved, and a learning curve of a user for application tuning is shortened.

Description

technical field [0001] The invention relates to a multi-objective fine-grained compilation self-tuning method, which belongs to the technical field of compilation optimization. Background technique [0002] In order to give full play to the performance advantages of the computer system, users often need to tune the application program for the target architecture. Mainstream compilers provide a variety of optimization techniques to compile and optimize the code to achieve the purpose of improving code execution performance, reducing the size of target code, and reducing code execution power consumption. However, there are a large number of mainstream compilation and optimization passes, and different optimization passes have different effects on different code structures, and there are interactions between optimization passes. How to select compilation optimization options for the target program has also become an optimization problem. [0003] There are hundreds of compilat...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F8/41
CPCG06F8/443G06F8/445Y02D10/00
Inventor 朱琪吴伟王飞钱宏周文浩杨涛
Owner JIANGNAN INST OF COMPUTING TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap