Due to the limited amount of memory resources in embedded systems, minimizing the memory requirements is an important goal of software synthesis. This paper presents a set of techniques to reduce the code and data siz...
详细信息
ISBN:
(纸本)0818686235
Due to the limited amount of memory resources in embedded systems, minimizing the memory requirements is an important goal of software synthesis. This paper presents a set of techniques to reduce the code and data size for software synthesis from graphical DSP programs based on the synchronous dataflow (SDF) model. By sharing the kernel code among multiple instances of a block, we can further reduce the code size below the single appearance schedule. And, a systematic approach is presented to give up single appearance schedules to reduce the data buffer requirements. Experimental results from two real examples prove the significance of the proposed techniques.
Data flow graph is a useful computational model to describe the functionality of a digital system. To execute a data flow graph on a target system, it should be synthesized to the code to be compiled on the target sys...
详细信息
ISBN:
(纸本)0818680644
Data flow graph is a useful computational model to describe the functionality of a digital system. To execute a data flow graph on a target system, it should be synthesized to the code to be compiled on the target system. Current research activities on software synthesis are mainly focused on Synchronous Data Flow (SDF) graph, which can not represent the control structure of the application. On the other hand, Dynamic Data Flow (DDF) graph can express the control structure, such as conditionals and data dependent iterations. This paper synthesizes a C code for a DDF graph which includes not only the functions associated with the nodes of the graph, but also the run-time scheduler and the run-time buffer manager which can execute a DDF graph without deadlock with bounded memory requirement. In addition, this paper suggests hierarchical implementation of DDF which enhances the efficiency of run-time scheduling by clustering SDF type nodes as a specific macro node.
暂无评论