METHOD FOR IMPROVING EXECUTION PERFORMANCE OF MULTIPLY-ADD INSTRUCTION DURING COMPILING

摘要

The present invention relates to a method for improving execution performance of multiply-add instructions during compiling, comprising the following steps of: compiling a source code by a compiler to acquire internal representation; optimizing; generating a machine code on the basis of a target processor, and allocating a physical register to a pseudo-register in the machine code; and improving results of register allocation to multiply-accumulate instructions. The method for improving execution performance of multiply-add instructions during compiling provided by the present invention has the following advantages: the compiler is allowed to realize procedure optimization by acquiring the optimal MAC (multiply-accumulate) instruction use gain.

主权项

1. A method for improving execution performance of multiply-add instructions during compiling, comprising the following steps of:
A) compiling a source code by a compiler, converting the source code into an internal representation, and optimizing the internal representation; B) generating a machine code on the basis of a target processor, and allocating a physical register to a pseudo-register in the machine code; and C) improving allocation of registers to multiply-accumulate instructions, respectively: allocating at least one idle accumulator to the pseudo-register in the current multiply-accumulate instruction according to set conditions if there is at least one idle accumulator, so that the pseudo-register is allowed to execute multiply-accumulation using the accumulator; and allocating a general register to the current multiply-accumulate instruction if there is no idle accumulator currently, so that the multiply-accumulate instruction is allowed to be realized through two independent operations.