Code Generation Techniques for Irregular Architectures

نویسنده

  • Steven Bashford
چکیده

The fast development of many different ASIPs make demands of rapid availability of dedicated compilers. Fast retargeting is a major aspect, while fast compilation times are of minor importance. There are also new demands in the quality of the generated code. Irregular properties together with fine–grain parallelism given by a target architecture have to be effectively supported by the compiler. This report is focused on the traditional tasks of code generation — code selection, register allocation, and instruction scheduling. The major subject is to expose the tendencies of research of code generation techniques in recent years, and survey their features with regards to support for irregular architectures, fine–grain parallelism, retargetability, and phase coupling. The report outlines the preferable techniques involved in code generators. Features of irregular architectures being sufficiently supported by these techniques are examined. The insufficiencies with regards to irregular architectures are described and approaches to overcome them are described. The essential problems arising are due to mutual dependencies among the tasks of code generation. Thus, phase ordering problems and phase coupling approaches are a very important issue of the report. Retargeting is discussed with regards to retargetability of the described techniques, but also with regards to the quality of the generated code. Relations of structural and behavioral models are exposed, addressing the issue of supporting both, the design process of the target architecture and effective retargeting of all tasks of code generation.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

ILP-based Approximations for Retargetable Code Optimization

Embedded systems are characterized by high performance requirements and are subject to severe cost and power consumption restrictions. This has led to the development of specialized irregular hardware architectures for which traditional code generation and optimization techniques usually fail to generate machine code of satisfactory quality. The PROPAN system has been developed as a framework f...

متن کامل

Scheduling, Compaction and Binding in a Retargetable Code Generator using Constraint Logic Programming

Code generation for embedded programmable processors is becoming increasingly important. Many of these processors have irregular architectures and offer instruction-level parallelism (e.g. DSPs). In order to generate code for a wide range of architectures, a code generator should be retargetable. Most of the previous code generation approaches concentrate on the datapath, not taking the peculia...

متن کامل

Integrated Scheduling and Register Assignment for Vliw-dsp Architectures

This paper describes code generation techniques for VLIW-DSP architectures. We focus on architectures with heterogeneous functional units and heterogeneous register sets. When generating code, scheduling and register allocation/assignment are typically done in separate steps. This is due to the fact that these tasks are complex combinatorial optimization problems particularly in case of irregul...

متن کامل

An effective and efficient code generation algorithm for uniform loops on non-orthogonal DSP architecture

To meet ever-increasing demands for higher performance and lower power consumption, many high-end digital signal processors (DSPs) commonly employ non-orthogonal architecture. This architecture typically is characterized by irregular data paths, heterogeneous registers, and multiple memory banks. Moreover, sufficient compiler support is obviously important to harvest its benefits. However, usua...

متن کامل

Register and Memory Assignment for Non-orthogonal Architectures via Graph Coloring and MST Algorithms

Finding an optimal assignment of program variables into registers and memory is prohibitively difficult in code generation for application specific instruction-set processors (ASIPs). This is mainly because, in order to meet stringent speed and power requirements for embedded applications, ASIPs commonly employ non-orthogonal architectures which are typically characterized by irregular data pat...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1995