Stall Detection Circuit for Mesh Network Deadlock Resolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Detecting deadlocks in mesh networks is challenging due to the complexity of monitoring cyclic dependencies and the risk of false-positive detections, which can waste resources.

Innovation Solution

Incorporating a stall detection circuit in each node of the mesh network to detect cyclical deadlocks by generating a stall output when linked input and output pipeline circuits are deadlocked, and when an upstream node is stalled.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If all possible loops in the mesh network are monitored to detect deadlocks, then deadlock detection accuracy is improved, but device complexity increases significantly

Engineering Contradiction:

Improvedeadlock detection accuracyVSAvoidmonitoring complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The mesh network is divided into individual nodes, each equipped with its own stall detection circuit. Each circuit monitors only its local pipelines and generates stall outputs independently, rather than having a centralized system monitor all loops. This segmentation reduces the complexity of monitoring while maintaining detection accuracy through distributed detection.

Inventive Principle:

Principle #1Segmentation

2Loss of energy

If false-positive deadlock detection is avoided, then resource waste is reduced, but detection sensitivity decreases

Engineering Contradiction:

Improveresource wasteVSAvoiddeadlock detection precision

Core Design Contradiction:

Loss of energyVSMeasurement precision

Solution Approach 1:

The stall detection circuit uses locally-generated stall outputs from each node's own pipelines and upstream nodes, rather than relying on global monitoring signals. Each node's stall output is asserted only when its specific pipelines are deadlocked, creating a localized detection mechanism that reduces false positives while maintaining precision through node-specific conditions.

Inventive Principle:

Principle #3Local quality

3Reliability

If stall detection is extended to upstream nodes, then cyclical deadlock detection capability is improved, but device complexity increases

Engineering Contradiction:

Improvecyclical deadlock detection capabilityVSAvoiddetection circuit complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The stall detection circuit incorporates feedback by asserting the stall output not only when local pipelines are deadlocked but also when upstream nodes generate stall signals. This feedback mechanism allows the circuit to detect cyclical deadlocks by propagating stall information through the network, improving detection capability without requiring complex centralized monitoring.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20250112845A1Deadlock detection and removal for mesh network for a processor-based system

Publication Date: 2025.04.03 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Systems and methods are disclosed for detecting a deadlock in a cyclical dependency between a set of the plurality of nodes in a mesh network. In some aspects, each of the nodes having a stall detection circuit. The stall detection circuit of each of the nodes operates by providing a stall output that is asserted not only when linked input and output pipeline circuits are stalled but when a stall input from an upstream node indicates that the upstream node is stalled. The stall output is provided as a stall input to the downstream node. In this manner, the stall outputs of the stall detection circuits are stable and asserted when there is a deadlock in a cyclical dependency between a closed loop of nodes in the mesh network.