Dynamic DAT Table Splitting for Context-Aware Memory Access

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing memory architectures lack the ability to dynamically adjust and manage dynamic address translation (DAT) tables based on operational context, leading to inefficiencies in managing privileged and non-privileged access to storage spaces.

Innovation Solution

Implementing a computer-implemented method that uses radix tree-based DAT tables to dynamically join or split page tables based on context attributes, allowing for separate access to full and partial storage spaces, thereby managing two sets of DAT tables to ensure secure and efficient memory allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a single DAT table is used for all storage access, then device complexity is reduced, but security and access control are compromised

Engineering Contradiction:

ImproveDAT table structureVSAvoidaccess control security

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent divides the single DAT table into multiple context-specific DAT tables (e.g., trusted context table and non-trusted context table). Each table is segmented by context attributes, allowing different access control policies to be applied to different portions of the address space based on the operational context of the accessing entity.

Inventive Principle:

Principle #1Segmentation

2Reliability

If separate DAT tables are maintained for different access contexts, then access control security is improved, but device complexity increases

Engineering Contradiction:

Improveaccess control securityVSAvoidDAT table management

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements dynamic DAT table selection and merging based on runtime context attributes. The system dynamically determines which context attribute set applies to each access request and automatically selects or merges the appropriate DAT table, eliminating the need for static manual configuration and reducing management complexity.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent creates a universal DAT table management mechanism that handles multiple context types (trusted, non-trusted, privileged, non-privileged) through a single unified system. The same DAT table structure and access control logic serve multiple different access contexts, reducing the need for separate specialized tables for each context type.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Measurement precision

If context-specific DAT tables are implemented, then access control precision is improved, but storage requirements increase

Engineering Contradiction:

Improveaccess control granularityVSAvoidstorage memory

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent merges multiple context-specific DAT tables into a unified address space mapping structure when context attributes allow it. By merging tables based on matching context attributes, the system achieves fine-grained access control precision while minimizing duplicate storage of identical address mappings across different context tables.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS11321239B2Dynamically joining and splitting dynamic address translation (DAT) tables based on operational context

Publication Date: 2022.05.03 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

An aspect includes determining, via a processor, context attributes of a storage. Data address translation (DAT) tables are created, via the processor, to map virtual addresses to real addresses within the storage. When detecting, via the processor, that a context attribute of the storage has changed, and the DAT tables are updated based at least in part on the changed context attributes of the storage.