Abstract:
PROBLEM TO BE SOLVED: To provide a method, a system, a program, and a computer data structure for providing two server virtualization levels. SOLUTION: A first hypervisor provides a first virtualization level while allowing a plurality of logic sections to share a set of resources. A second hypervisor provide a second virtualization level while allowing a plurality of independent virtual machine to share resources assigned to a single logic section. All the events about all the virtual machines in the signal logic section are grouped into a single event queue possessed by a section for receiving an event notification from the shared resources about the single logic section. An interruption request is transmitted via a signal to the grouped events from the event queue possessed by the section, and demultiplexing of the grouped events from the event queue possessed by the section into individual virtualized event queues assigned on a virtual machine base is carried out by means of the machine. COPYRIGHT: (C)2008,JPO&INPIT
Abstract:
PROBLEM TO BE SOLVED: To provide the function of functionally enhancing input and output processing to reduce a host processing overhead associated with execution of input and output operation such as QDIO operation, whereas the function of functionally enhancing the input and output processing is needed to facilitate interpretation of an input and output operation request from a pageable guest. SOLUTION: An input and output operation request from a pageable storage mode guest is interpreted without intervention of a host. A request issued by the pageable storage mode guest is processed in a pageable mode virtual environment by one or more processors in the environment without intervention from one or more hosts in the environment. The processing of the request includes operation of buffer state information retained in a host storage for the guest by at least one processor. The operation is started by the guest, and executed through an instruction processed by one or more processors. COPYRIGHT: (C)2006,JPO&NCIPI
Abstract:
Disclosed is a method of interrupt handling in a logically partitioned data processing system. The system has at least two logical partition zones and at least one interrupt handling array, which includes bit values that change in accordance with the change of an interrupt pending status of the data processing system. A change of the interruption pending status is only performed for a logical partition zone having an updated bit value. The interrupt pending status is changed for only one partition at a time. The current interrupt pending status for all the zones may be stored as a vector or array of interrupt states, with one bit per zone. In the cases where a back-to-back writes to the same entry in the array a bypass may be used to bypass the array.