METHOD FOR IMPLEMENTING DYNAMIC LIFETIME RELIABILITY EXTENSION FOR MICROPROCESSOR ARCHITECTURES

• US 20090178051A1
• Filed: 05/09/2008
• Published: 07/09/2009
• Est. Priority Date: 01/04/2008
• Status: Abandoned Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Zyuban, Victor, Shin, Jeonghee
• Application Number:
  US12/118,050
• Publication Number:
  US 20090178051A1
• Time in Patent Office: Days
• US Class Current: 718/104
• CPC Class Codes: G06F 11/004 Error avoidance G06F11/07 and subgroups take precedence
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: G06F 11/2025 using centralised failover control functionality
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 11/2051 in regular structures
• CPC Class Codes: G06F 12/0846 Cache with multiple tag or data arrays being simultaneously accessible
• CPC Class Codes: G06F 2209/5011 Pool
• CPC Class Codes: G06F 2209/503 Resource availability
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]

Abstract

A method for implementing dynamic lifetime reliability extension for microprocessor architectures having a plurality of primary resources and a secondary resource pool of one or more secondary resources includes configuring a resource operational mode controller to selectively switch of the primary and secondary resources between an operational mode and a non-operational mode, wherein the non-operational mode corresponds to a lifetime extension process; configuring a resource mapper associated with the secondary resource pool and in communication with the resource operational mode controller to map a secondary resource placed into the operational mode to a corresponding primary resource placed into the non-operational mode; and configuring a transaction decoder to receive incoming transaction requests and direct the requests to one of a primary resource in the operational mode and a secondary resource in the operational mode, the secondary resource mapped to an associated primary resource placed in the non-operational mode.

Claims

1. A method for implementing dynamic lifetime reliability extension for microprocessor architectures, the method comprising:

configuring a plurality of primary resources;

configuring a secondary resource pool having one or more secondary resources;

configuring a resource operational mode controller to selectively switch each of the primary and secondary resources between an operational mode and a non-operational mode, wherein the operational mode corresponds to performance of one or more tasks for which a given resource is designed to execute with respect to a microprocessor system and wherein the non-operational mode corresponds to a temporary lifetime extension process for at least one of suspending the aging of resources and extending the aging of resources;

configuring a resource mapper associated with the secondary resource pool and in communication with the resource operational mode controller to map a secondary resource placed into the operational mode to a corresponding primary resource placed into the non-operational mode, and to un-map a secondary resource from a corresponding primary resource in the event the corresponding primary resource is placed back into the operational mode; and

configuring a transaction decoder to receive incoming transaction requests and, responsive to the resource operational mode controller, direct the requests to one of a primary resource in the operational mode and a secondary resource in the operational mode, the secondary resource mapped to an associated primary resource placed in the non-operational mode.

2. The method of claim 1, further comprising:

configuring a lifetime predictor to monitor usage of the primary resources, wherein the lifetime predictor predicts a remaining lifetime thereof;

wherein, in the event that any of the primary resources are predicted as having a lifetime shorter than a defined threshold lifetime, the lifetime predictor alerts the resource operational mode controller and resource mapper so that an identified primary resource designated to be placed in the non-operational mode.

3. The method of claim 1, wherein the resource mapper is configured to:

find and allocate available secondary resources in the pool upon a request, and to map an allocated secondary resource to a corresponding one of the identified primary resources that is designated for lifetime extension treatment, such that decoded transactions by the decoder originally intended for the identified primary resource are thereafter executed by the allocated secondary resource;

wherein, in the event that there is more than one available secondary resource found in the pool, the resource mapper select a secondary resource based on one or more of;

a round-robin order, a remaining lifetime order of the secondary resources themselves, and a maximum recovery based order; and

in the event there is no available secondary resource found, the resource mapper is further configured to implement one of;

rejecting the request, and commandeering a secondary resource already mapped to another primary resource.

4. The method of claim 3, wherein the resource mapper is further configured to:

evaluate a priority of a new request with respect to lifetime extension treatment;

de-allocate a secondary resource mapped to a primary resource of lower priority in favor of another primary resource of higher priority;

notify the resource operational mode controller to interrupt the non-operational mode for the lower priority primary resource;

map the de-allocated secondary resource to the other primary resource of higher priority and notify the resource operational mode controller of the new request so as to result in the placement of the higher priority primary resource into the non-operational mode for lifetime extension treatment; and

in the event a lower priority primary resource with respect to the request is not found, the resource mapper is further configured to implement one or more of;

sending a rejection of lifetime extension treatment for the new request, and adding the request to a reservation table.

5. The method of claim 1, wherein the primary and secondary resources comprise one or more of:

static random access memory (SRAM) arrays, embedded dynamic random access memory (EDRAM) arrays, register files, execution units, and processor cores.