PCIe Port Link Restart for Automatic Abnormal Device Repair
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for repairing devices have a limited adaptive range and are prone to device identification errors, leading to inefficient resource allocation and increased labor and time consumption.
Innovation Solution
A method involving obtaining link state and performance information of PCIe devices, determining abnormal ports based on transmission rate and bandwidth inconsistencies, and performing a link restart on the link disable bit in the link control register to automatically repair the devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hot update on contents or functions using loaded resource packages is performed to repair devices, then operating stability of devices is ensured, but adaptive range of repairable devices is relatively small and devices at end of device link cannot be repaired
Solution Approach 1:
The patent changes the repair approach from hot updating resource packages to manipulating link control register parameters (specifically the link disable bit). This parameter change enables the repair method to work across different device types and link configurations, expanding the adaptive range while maintaining operating stability through controlled link restarts.
Solution Approach 2:
Instead of trying to update device contents or functions to repair devices, the patent inverts the approach by disabling and re-enabling the link through register manipulation. This inverted method bypasses the limitations of content-based repair and successfully repairs devices at the end of device links that were previously unreachable.
2Reliability
If hot update method is used for device repair, then device stability is maintained, but resource reservation accuracy deteriorates due to device identification errors
Solution Approach 1:
The patent replaces the mechanical/content-based hot update process with a register-based control mechanism. By substituting the complex resource package loading and device identification process with direct manipulation of the link disable bit in the link control register, the system achieves both device stability and accurate resource reservation without identification errors.
3Ease of manufacture
If manual resource allocation and device repair methods are used, then resource allocation can be customized, but labor costs and time consumption increase significantly
Solution Approach 1:
The patent implements self-service by enabling the system to automatically detect abnormal ports and execute repair operations through programmatic control of the link disable bit. This eliminates the need for manual resource allocation and device repair operations, significantly reducing labor costs and time consumption while maintaining resource allocation flexibility through software control.
Solution Approach 2:
The patent incorporates feedback mechanisms that automatically detect link abnormalities and trigger repair operations. This closed-loop system continuously monitors device status and performs repairs when needed, replacing manual intervention with automated feedback-driven operations that reduce labor costs while maintaining flexible resource allocation.
Data Source
AI summary
The present disclosure discloses a method for automatically repairing a device, an electronic device and a storage medium. The method includes: obtaining first link state information and first link performance information of a port corresponding to a peripheral component interconnect express (PCIe) device; if a first transmission rate in the first link state information is inconsistent with a second transmission rate in the first link performance information, determining that the port corresponding to the PCIe device is a first abnormal port; if first bandwidth information in the first link state information is less than a preset bandwidth threshold corresponding to second bandwidth information in the first link performance information, determining that the port corresponding to the PCIe device is the first abnormal port; performing a link restart on a link disable bit in a link control register of the first abnormal port to complete a repair of the PCIe device.


