International Business Machines Corporation

Richard John Moore

Damian Leo Osisek

Brian Robert Prasky

James J Bonanno

Louis P Gomes

Suparna Bhattacharya

Mark John Lorenc

Thomas R Zanussi

Richard Victor Kisley

Michael D Hocker

Klaus Werner

Eric David Rossman

Volker Urban

Harald Freudenberger

Vamsi Krishna Sangavarapu

Tracy Ann Krein

A method, computer program product, and computer system are provided. A machine hypervisor builds and enqueue a zeroize adjunct processor (AP) domain (ZAPD) common command request message to a target AP domain queue. Machine firmware validates the enqueued common command request message on the target AP domain queue. The machine firmware converts the ZAPD command request to the mode-specific form of the domain zeroize request. A command request message is executed at an AP crypto adapter of the target AP domain. Executing includes zeroizing a storage area of the AP crypto adapter. The execution results are returned to the machine firmware. The machine firmware converts the crypto card mode's ZAPD command reply message into the common ZAPD command reply message. The final execution results including the converted common ZAPD command reply message are returned to the machine hypervisor.

Adjunct processor (AP) domain zeroize

A method, computer program product, and computer system are provided for supporting lossless transitions between interruption and polling mode of a resource. In response to a state variable setting or a timed trigger, launching a task to process replies for a resource. The task processes each ready reply for each resource. A timed trigger is established to relaunch the task based on the task executing in polling mode. The task is exited.

Switching between polling-mode and interruption-driven modes without loss of I/O replies

A message limit value to be used in enqueuing one or more messages on a queue of a device of the computing environment is obtained. The message limit value indicates whether an extended maximum message length is supported by the device. The extended maximum message length is different from a default maximum message length supported by the device. Based on determining that the extended maximum message length is supported and that the obtained message limit value has a defined relationship with a select value, at least one message of an extended length is enqueued on the queue of the device.

Adjunct processor extended message limit facility

Dynamic relocation of virtual machines among hypervisors in a relocation domain is provided. A hypervisor is initialized in a subdomain of the relocation domain. A record of architecture characteristics is retrieved for each hypervisor in the relocation domain. A new canonical architectural description (ARD) is created for each subdomain in the relocation domain. An effective adapter characteristic representation is created for each virtual machine defined to the hypervisor. The record of architecture characteristics for each hypervisor in the relocation domain is updated.

Live virtual machine relocation to accommodate reversible relocations in a heterogeneous cluster of hypervisor versions

A method, computer program product, and computer system are provided. An adjunct processor receives an indication to reset a message queue. The reset removes all messages from the message queue on the adjunct processor. The reset is in response to a hardware failure or an external manual operation. An operation to enqueue a message to the message queue is received. Based on completing a successful enqueue operation, the adjunct processor sets a status that includes an indication that the enqueue operation is a first operation following the reset. A requestor receives the indication of the first operation. Based on the internal count of pending messages being greater than one, the requestor requeues to the adjunct processor, the message requests, except for the first message, having outstanding replies, and resets the indication of the first enqueue operation upon the requeuing of the message requests being complete.

Enhanced recovery from externally initiated adjunct processor queue reset

Branch prediction using a correlating event, such as an unconditional branch that calls a routine including the branch, instead of the branch itself, to predict the behavior of the branch. The circumstances in which the branch is employed, and not the actual branch itself, is used to predict how strongly taken or not taken the branch is to behave. An anchor point associated with the branch (e.g., an address of the instruction calling a routine that includes the branch), an address of the branch, and a value that represents the number of selected branch instructions between the anchor point and the branch are used to select information to be used to predict the direction of the branch.

Branch prediction based on correlating events

Dynamic relocation of virtual machines among hypervisors in a domain is provided. A domain management module is executed on a computer processor and is configured to receive a request to add a new hypervisor to a domain. Each hypervisor and each domain comprises an architecture description (ARD). A new canonical ARD is generated for all hypervisors in the domain. A virtual machine executing on a first system is relocated to any other hypervisor in the domain. Virtual machines may also be moved to out-of-domain hypervisors through the use of override masks. In addition, under certain conditions, new hypervisors additions, as well as hypervisor and system updates may be performed without disrupting the virtual machines that are controlled by the hypervisors.

Dynamic virtual machine domain configuration and virtual machine relocation management

A high-level language, architecture-independent probe program compiler is disclosed. A base program is executable by one or processors, and has one or more breakpoints. A probe program is associated with each breakpoint and is also indirectly executable by the one or more processors. The probe program is independent of the architecture of the processors, and is generated from source code written in a high-level language. The probe program associated with each breakpoint is executed when the breakpoint is reached during execution of the base program. The compiler may employ an abstract syntax tree to switch between an address space of the probe program and an address space of the base program, by traversing the tree. Some of the nodes of the tree may more simply represent address space-specific objects of the base program. The probe program may be able to pass messages by manipulating the state of the base program.

High-level language, architecture-independent probe program compiler

A method for testing a software product comprises initiating a breakpoint in the software product, and executing a probe routine. The executing of the probe routine comprises creating a single step state stack, and repetitively; identifying data to be protected, disabling interrupts, writing the data to the single step state stack, enabling interrupts, restoring a state according to the single step state stack, and single-stepping a probe instruction, until the instructions in the probe are completed.

Testing a software product by initiating a breakpoint and executing a probe routine

A method is provided that allows a general set of watchpoints to be defined for a computer system (a watchpoint is a memory address that triggers an interrupt for debugging or tracing purposes). This is accomplished by modifying the system page table for the memory page containing a watchpoint, such that a page fault interrupt is triggered whenever said memory page is accessed (for example by marking the page as not present). The paging mechanism of the computer system is then adapted, so that responsive to a page fault interrupt, a determination is made as to whether such interrupt has resulted from an access to the watchpoint, and if so, control is passed to a watchpoint handler. Alternatively, if there is no watchpoint, normal paging operations can be resumed if necessary.

Computer system with watchpoint support

A method, an apparatus, and a computer program product for detecting one or more zombie global breakpoints for debugging computer software are disclosed. The method includes the steps of: checking a breakpoint data structure to determine if a breakpoint known to a debugging process is at an address where a breakpoint fired; if a known breakpoint cannot be determined at the address, verifying if a breakpoint condition continues to exist at the address where the breakpoint fired; and if the breakpoint condition does not exist, identifying the breakpoint as a zombie breakpoint.

Method of detecting zombie breakpoints

A system for intercepting API calls in a virtual memory system comprises an activation module comprising an executable program and an interception module comprising a dynamic link library (DLL). The activation module is first executed at system initialization time, its prime purposes being to: parse user configuration information supplied in a configuration file; act as a daemon to launch and thus retain the interception module in shared memory; and hook system APIs to redirect calls via the interception module.

Growing community of inventors

Waterlooville, United Kingdom

Richard John Moore