New graph transformation schemes in graph-based memory allocation method for an indirect addressing DSP

For indirect addressing DSPs, a novel memory address allocation method based on graph representation is presented. The method translates a given memory access sequence into a graph notation, and transforms it into line-shaped graphs. At the transformation, a cost evaluation measure is newly introduced, so that efficient memory allocation is given. The proposed cost functions are applied to the existing memory allocation method, and memory allocation results derived for several examples show its effectiveness.

[1]  B. Wess,et al.  DSP data memory layouts optimized for intermediate address pointer updates , 1998, IEEE. APCCAS 1998. 1998 IEEE Asia-Pacific Conference on Circuits and Systems. Microelectronics and Integrating Systems. Proceedings (Cat. No.98EX242).

[2]  Akinori Nishihara,et al.  Memory Allocation Method for Indirect Addressing DSPs with ±2 Update Operations , 1998 .

[3]  B. Wess,et al.  Optimal DSP memory layout generation as a quadratic assignment problem , 1997, Proceedings of 1997 IEEE International Symposium on Circuits and Systems. Circuits and Systems in the Information Age ISCAS '97.

[4]  Akinori Nishihara,et al.  DSP Code Optimization Methods Utilizing Addressing Operations at the Codes without Memory Accesses , 1997 .

[5]  N. Sugino,et al.  Computational ordering of adaptive digital networks under pipeline constraints and its application to DSP compilers , 1996, Proceedings of APCCAS'96 - Asia Pacific Conference on Circuits and Systems.

[6]  Rainer Leupers,et al.  Algorithms for address assignment in DSP code generation , 1996, ICCAD 1996.

[7]  B. Wess,et al.  Automatic code generation for integrated digital signal processors , 1991, 1991., IEEE International Sympoisum on Circuits and Systems.

[8]  Nobuo Fujii,et al.  DSP Code Optimization Utilizing Memory Addressing Operation (Special Section on Digital Signal Processing) , 1996 .