Tagging Logic for Selective Activity Tracing in Data Processing Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing systems face complexity in tracing and configuring multiple trace sources across multiple components, leading to excessive trace data generation and difficulty in managing bandwidth, especially when tracing activity across a chip with multiple CPUs and buses.
Innovation Solution
The implementation of tagging logic that selects and tags specific items for monitoring, allowing trace logic to detect and output relevant trace information, reducing the need for complex configuration of individual trace sources and automatically managing trace data generation across devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple trace sources are configured to monitor different devices on the chip, then the coverage of traced activities is improved, but the complexity of configuration and data management increases
Solution Approach 1:
The patent introduces a trace controller as an intermediary component that manages multiple trace sources. Instead of directly configuring each trace source individually, the trace controller acts as a mediator that receives trace control signals and distributes them to appropriate trace sources, thereby reducing configuration complexity while maintaining comprehensive coverage of multiple devices.
Solution Approach 2:
The trace controller is designed as a universal component that can manage multiple different types of trace sources (ETM, bus trace macrocell, etc.) through a single interface. This multi-functional design allows the system to monitor various devices without requiring separate configuration mechanisms for each trace source type, thus reducing overall configuration complexity.
2Measurement precision
If detailed trace configuration is applied to control trace region tightly, then the relevance of trace data is improved, but the amount of trace data generated increases
Solution Approach 1:
The patent applies preliminary action by configuring trace sources with trace control signals before actual tracing begins. The trace controller pre-establishes which trace sources should be active and what their monitoring parameters should be, based on the current operational context. This allows the system to generate only relevant trace data from the outset, avoiding unnecessary data generation while maintaining precision in trace monitoring.
Solution Approach 2:
The trace configuration is made dynamic rather than static. The trace controller can adjust trace control signals in real-time based on system state, enabling the trace region boundaries and monitoring parameters to adapt dynamically. This dynamic approach ensures that trace data remains highly relevant to current operations while preventing the generation of obsolete or irrelevant trace data, thus optimizing the balance between data relevance and volume.
3Adaptability or versatility
If multiple trace sources are enabled simultaneously, then the monitoring capability is improved, but the bandwidth requirements for trace data increase
Solution Approach 1:
The patent segments the trace data output by introducing separate trace data outputs for different trace sources. Each trace source has its own dedicated output path, allowing trace data to be segmented and transmitted independently. This segmentation enables the system to monitor multiple devices simultaneously while managing bandwidth by allowing selective transmission of trace data from different sources based on current needs, rather than requiring all trace data to share a single bandwidth resource.
Data Source
AI summary
A data processing apparatus is disclosed, said data processing apparatus comprising a plurality of devices, trace logic associated with at least one of said plurality of devices, and tagging logic associated with at least one of said plurality of devices, said tagging logic being operable to: select at least one item, said at least one item comprising an activity to be monitored; provide said at least one selected item with tag data identifying said at least one item as an item to be monitored; and said trace logic being operable to: detect tagged items processed by said at least one device; and output trace information relating to at least some of said detected tagged items.


