International Business Machines Corporation

Brian K Wade

Stephen E Record

Joel Alan Farrell

Peter Dana Driever

Mary Ellen Carollo

Seth E Lederer

Kevin Scott Adams

Donald P Wilton

Omkar Ulhas Javeri

John S Roman

A method, computer program product, and system include a processor(s) issues an instruction that includes processing core information that includes locations of processing cores of the computing system (logical cores and/or physical cores), and an operator selection. The processor(s) sets security parameters for information returned by the instruction which is topological information for mapping of the logical cores to the physical cores. The processor(s) obtains the topological information and utilizes an operating system to map the logical cores to the physical cores.

Accessing topological mapping of cores

A method of calculating a processing power available from a supervisor of a multi-programmed computing system by a first partition of a plurality of partitions, the method comprising collecting, by the first partition, state data from the supervisor, the state data including a processing capacity of the multi-programmed computing system. The method further comprises initializing a remaining capacity variable to the processing capacity of the multi-programmed computing system; initializing variables, including setting a binary variable to a first logic value for each of the plurality of partitions; iteratively computing an entitlement and amount of power to award for each of the plurality of partitions having their respective binary variables set to the first logic value; and requesting the processing power from the supervisor, based on the iterative computation.

Calculating expected maximum CPU power available for use

One or more measurements of processor utilization are taken. A utilization ceiling is calculated. One or more processing units (PUs) are added automatically if it is determined that the utilization ceiling is greater than an available PU capacity. One or more PUs are removed automatically responsive to determining that the utilization ceiling is at least one PU less than the available PU capacity.

Programmatic load-based management of processor population

Data processing system efficiency is improved by automatically determining whether to adjust for a next time interval a number N of processors running within the system for processing a workload. The automatically determining includes obtaining a measure of operating system overhead by evaluating one or more characteristics of processor time of the N processors consumed within the system for a time interval, and obtaining a measure of system utilization of the N processors running within the system for processing the workload for the time interval. The automatically determining further includes automatically ascertaining whether to adjust the number N of processors running within the system for the next time interval to improve system efficiency using the obtained measure of operating system overhead and the obtained measure of system utilization of the N processors.

System overhead-based automatic adjusting of number of running processors within a system

Application execution contexts within an untrusted computer system are classified as trusted or untrusted based on respective names assigned to the execution contexts. If an application runs in an untrusted execution context, an operating system within the untrusted computer system prevents the application from initiating a connection with a trusted computer system and accessing sensitive parts of the untrusted computer system. If the application runs in a trusted execution context, the operating system permits the application to initiate a connection with the trusted computer system.

Method and computer system for controlling access by applications to this and other computer systems

A multitasking operating system permits application programs (and their developers) to influence a schedule of execution of program threads derived from the application programs. Each of the program threads is assigned a priority level and a dispatch class in which the thread resides. Based on these parameters, the operating system schedules the threads for execution in the following manner. The operating system queues the highest priority thread which is available for execution from each dispatch class onto a run list for execution by a processor. The highest priority thread on the run list is executed first. An application program thread can change the dispatch class in which another program thread resides. An executing program thread can also voluntarily yield to a specified program thread in the same dispatch class or permit the highest priority available thread in the same dispatch class to be queued on the run list with itself being available and in contention for the run list.

Controlled scheduling of program threads in a multitasking operating

A multitasking operating system permits application programs (and their developers) to influence a schedule of execution of program threads which constitute the application programs by specifying parameters for the program threads. The parameters indicate each thread's priority level and dispatch class in which the thread resides. The application programs specify the thread's parameters based on the following principles of the operating system. The operating system queues the highest priority thread available for execution from each dispatch class onto a run list for execution by a processor. The highest priority thread on the run list is executed first. While this thread is dispatchable and being executed, no other thread from the same dispatch class can preempt it unless this executing thread voluntarily relinquishes control of the processor, even if the other thread has a higher priority. (This other thread would have been created or made available after the currently executing thread was selected for the run list.) However, the currently executing thread can be involuntarily preempted at any time by another higher priority, available thread from a different dispatch class. A thread can also voluntarily relinquish control of its processor at other appropriate points in the execution, for example, when data structures are valid, to share the processor with other lower priority threads from the same or different dispatch classes.

Growing community of inventors

Apalachin, NY, United States of America

Brian K Wade