Abstract: A method, system and computer program product are provided for implementing enhanced error handling for a hardware I/O adapter, such as a Single Root Input/Output Virtualization (SRIOV) adapter, in a virtualized system. The hardware I/O adapter is partitioned into multiple endpoints, with each Partitionable Endpoint (PE) corresponding to a function, and there is an adapter PE associated with the entire adapter. The endpoints are managed both independently for actions limited in scope to a single function, and as a group for actions with the scope of the adapter. An error or failure of the adapter PE freezes the adapter PE and propagates to the VF PEs associated with the adapter, causing the VF PEs to be frozen. An adapter driver and VF device drivers are informed of the error, and start recovery. The hypervisor locks out the VF device drivers at key points enabling adapter recovery to successfully complete.

Abstract: A method, system and computer program product are provided for implementing concurrent adapter firmware update of a Single Root Input/Output Virtualization (SRIOV) adapter in a virtualized system. An adapter driver is used to update adapter firmware concurrent with normal I/O operations. When configuration is stored in a scratchpad buffer, the adapter driver detects virtual functions (VFs) configured and operating. An enhanced error handling (EEH) process is initiated, freezing the VFs, and an updated adapter firmware image is loaded to the adapter. The adapter driver completes the EEH recovery, the adapter is restarted using the new updated adapter firmware. The VFs device drivers unfreeze the VFs, and complete the EEH recovery.

Abstract: A method, apparatus and program product implement a failover of a communication channel in a cluster fabric that transfers a state of the communication channel between windows resident in a hardware fabric interface device. The failover is desirably implemented by updating a plurality of mappings between memory resources in a host memory and hardware resources in the fabric interface device, and typically without modifying the memory resources such that involvement of a client that utilizes the communication channel in the failover is minimized or eliminated.

Abstract: Systems and methods to manage access to shared resources are provided. A particular method may include receiving a request to access a shared resource from a first client of a plurality of clients and determining whether the shared resource is being used. A first window credential associated with the first client may be retrieved. The first window credential may be one of a plurality of window credentials associated with the plurality of clients. The first window credential may be used to access the shared resource.

Abstract: Processing multicast messages in a data processing system that includes compute nodes, each of which includes a communications controller, the communications controllers operatively coupled for data communications amongst the compute nodes, each compute node in the data processing system executing at least one logical partition, the data processing system including a hypervisor, where multicast messages are processed by receiving, by more than one of the communications controllers, a multicast message; responsive to receipt of the multicast message, signaling, by one of the communications controllers, an interrupt in the hypervisor without signaling an interrupt by any other communications controller; determining, by the hypervisor, that the message received by the interrupt signaling communications controller is a multicast message; and providing, by the hypervisor, the multicast message to at least one logical partition configured to receive multicast messages.

Abstract: A method, system and computer program product are provided for implementing memory migration of large system memory pages in a computer system. A large page to be migrated from a current location to a target location is converted into a plurality of smaller subpages for a processor or system page table. The migrated page is divided into first, second and third segments, each segment composed of the smaller subpages and each respective segment changes as each individual subpage is migrated. CPU and I/O accesses to respective subpages of the first segment are directed to corresponding subpages of the target page or new page. I/O accesses to respective subpages of the second segment use a dual write mode targeting corresponding subpages of both the current page and the target page. CPU and I/O accesses to the subpages of the third segment access the corresponding subpages of the current page.

Abstract: Processing multicast messages in a data processing system that includes compute nodes, each of which includes a communications controller, the communications controllers operatively coupled for data communications amongst the compute nodes, each compute node in the data processing system executing at least one logical partition, the data processing system including a hypervisor, where multicast messages are processed by receiving, by more than one of the communications controllers, a multicast message; responsive to receipt of the multicast message, signaling, by one of the communications controllers, an interrupt in the hypervisor without signaling an interrupt by any other communications controller; determining, by the hypervisor, that the message received by the interrupt signaling communications controller is a multicast message; and providing, by the hypervisor, the multicast message to at least one logical partition configured to receive multicast messages.

Abstract: DMA in a computing environment that includes several computers and DMA engines, the computers adapted to one another for data communications by an data communications fabric, each computer executing an application, where DMA includes: pinning, by a first application, a memory region, including providing, to all applications, information describing the memory region; effecting, by a second application in dependence upon the information describing the memory region, DMA transfers related to the memory region, including issuing DMA requests to a particular DMA engine for processing; and unpinning, by the first application, the memory region, including insuring, prior to unpinning, that no additional DMA requests related to the memory region are issued, that all outstanding DMA requests related to the memory region are provided to a DMA engine, and that processing of all outstanding DMA requests related to the memory region and provided to a DMA engine has been completed.

Abstract: DMA in a computing environment that includes several computers and DMA engines, the computers adapted to one another for data communications by an data communications fabric, each computer executing an application, where DMA includes pinning, by a first application, a memory region, including providing, to all applications, information describing the memory region; effecting, by a second application in dependence upon the information describing the memory region, DMA transfers related to the memory region, including issuing DMA requests to a particular DMA engine for processing; and unpinning, by the first application, the memory region, including insuring, prior to unpinning, that no additional DMA requests related to the memory region are issued, that all outstanding DMA requests related to the memory region are provided to a DMA engine, and that processing of all outstanding DMA requests related to the memory region and provided to a DMA engine has been completed.

Abstract: An optimal configuration of Internet Small Computer Systems Interface (iSCSI) sessions is automatically determined. A number of iSCSI target ports and a number of iSCSI initiator ports in a configuration are determined, and an ideal session count is determined for each iSCSI initiator port in order to maximize efficiency. All iSCSI initiator ports are traversed, and a determination is made which iSCSI initiator ports have less than the ideal session count. For those iSCSI initiator ports that have less than the ideal session count, the iSCSI target ports that are currently not logged into an iSCSI initiator ports that have fewer sessions than the ideal sessions. The steps of finding and attempting are repeated until a current total number of active sessions equals the number of iSCSI target ports.