Loop Execution Analysis Using Time Difference Segments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing application program behavior using intermittent sampling logs struggle to accurately represent loop execution order and time due to mismatched sampling intervals, leading to incomplete analysis results.
Innovation Solution
An information processing apparatus that converts sampling times into time differences from preceding start times, classifies logs into segments, and counts function executions within each segment to display a corrected order and execution rates, allowing for detailed loop analysis even with long sampling intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sampling logs are collected at shorter intervals, then analysis detail is improved, but computational resource consumption increases and program behavior may be altered by sampling load
Solution Approach 1:
The patent segments the continuous sampling data by identifying loop start times and dividing sampling logs into separate loop iteration groups. This segmentation allows the system to use longer sampling intervals while still capturing complete loop execution sequences, thereby reducing computational overhead while maintaining analysis precision.
Solution Approach 2:
The patent performs preliminary identification of loop start times and creates time difference segments before analyzing sampling logs. By pre-organizing the data structure with known loop boundaries, the system can use coarser sampling intervals without losing loop execution details, thus reducing the frequency of interrupts and computational resource usage.
2Stability of the object's composition
If sampling logs are arranged in time order, then chronological sequence is preserved, but loop execution order and time cannot be accurately determined
Solution Approach 1:
The patent introduces time difference segments as an intermediary between raw sampling timestamps and loop execution analysis. By calculating time differences from loop start times and grouping samples into segments, the system creates a new organizational layer that preserves chronological information while enabling accurate loop execution order determination through segment-based analysis.
3Productivity
If sampling interval is extended to conserve resources, then computational load is reduced, but loop behavior may be affected by sampling frequency
Solution Approach 1:
The patent performs preliminary marking of loop start times during program execution. This preliminary action creates reference points that allow the system to use extended sampling intervals while still accurately capturing loop behavior, because complete loop sequences can be reconstructed using the start time markers and time difference calculations.
Solution Approach 2:
The patent segments sampling logs into loop-specific groups based on time differences from loop start times. This segmentation allows the system to use longer sampling intervals since each segment represents a complete or partial loop iteration, preserving loop behavior accuracy while reducing the total number of samples needed and thus computational load.
Data Source
AI summary
An information processing apparatus obtains start time information indicating a plurality of start times that are calculated when a target program is executed. The target program includes a loop that repeats a process using a plurality of functions. The information processing apparatus obtains a plurality of sampling logs that are generated by intermittent sampling. The information processing apparatus converts each sampling time into a time difference from the immediately preceding start time. The information processing apparatus classifies the plurality of sampling logs into a plurality of time difference segments, and counts, for each of the time difference segments, the number of times execution of each of the plurality of functions is detected. The information processing apparatus displays time-series information indicating a corresponding relationship between the plurality of time difference segments and the functions that are executed, based on the counted number of times.