CPI (Cycle per Instruction) precise measurement method based on CPU (Central Processing Unit) hardware performance monitoring counter

Abstract

The invention discloses a CPI (Cycle Per Instruction) precise measurement method based on a CPU (Central Processing Unit) hardware performance monitoring counter, which comprises the following steps of: S1, sampling jump history records to obtain jump frequencies and obtain fine-grained instruction frequencies according to the jump frequencies; S2, obtaining clock cycles by a random sampling method; and S3, obtaining fine-grained CPI by dividing the fine-grained instruction frequencies by the fine-grained clock cycles. By sampling a register for storing the jump history records, the precise fine-grained instruction frequencies are obtained, and the resource overhead of the sampling method is low; and by the random sampling method, the precise fine-grained clock cycles are obtained. The fine-grained instruction frequencies are combined with the fine-grained clock cycles to obtain the precise fine-grained CPI.

Description

technical field [0001] The invention relates to the technical field of computer software performance analysis, in particular to a CPI accurate measurement method based on a CPU hardware performance monitoring counter. Background technique [0002] The number of clock cycles (Cycle Per Instruction, CPI) consumed by each instruction is an important observation index in the process of program performance analysis. This indicator can visually show the average number of clock cycles required for each instruction in the program. If this metric is large, it indicates a performance problem in the program. This also means that the program has a greater opportunity for performance optimization. [0003] In the traditional program optimization process, CPI is often the first indicator that programmers observe when performing performance tuning. When programmers get the specific CPI value, they will use the program performance counter to divide it into a CPI stack, and specifically c...

AI Technical Summary

Problems solved by technology

However, the accuracy of fine-grained CPI is poor

Even the CPI at the function level, its error is enough to mislead the programmer's tuning work

And for finer-grained CPI (such as basic block (the basic block refers to a continuous program statement sequence, the control flow enters from its beginning and leaves from its end, without interruption or branch in between) level CPI and instruction Level CPI), due to its poor accuracy, it has been completely unusable in the actual tuning process

Images