Methods, devices, and executable instructions are provided for allocating external memory. A request to allocate a portion of memory may be received at a first device. A memory appliance and the first client device are connected by a network. A region of memory of the memory appliance may be allocated as external memory prior to receipt of the request to allocate the portion of memory at the first client device. A subset of the region of memory for the portion of memory allocated may be selected at the first client device and in coordination with a second client device on the network, wherein the portion of memory allocated is external to the first client device and yet is primary memory to the first client device. At least the subset of the region of memory may be mapped at the first client device to a virtual address space.

Various embodiments are generally directed to virtualized systems. A first guest memory page may be identified based at least in part on a number of accesses to a page table entry for the first guest memory page in a page table by an application executing in a virtual machine (VM) on the processor, the first guest memory page corresponding to a first byte-addressable memory. The execution of the VM and the application on the processor may be paused. The first guest memory page may be migrated to a target memory page in a second byte-addressable memory, the target memory page comprising one of a target host memory page and a target guest memory page, the second byte-addressable memory having an access speed faster than an access speed of the first byte-addressable memory.

Implementations described provide hardware support for the co-existence of restricted and non-restricted encryption keys on a computing system. Such hardware support may comprise a processor having a core, a hardware register to store a bit range to identify a number of bits, of physical memory addresses, that define key identifiers (IDs) and a partition key ID identifying a boundary between non-restricted and restricted key IDs. The core may allocate at least one of the non-restricted key IDs to a software program, such as a hypervisor. The core may further allocate a restricted key ID to a trust domain whose trust computing base does not comprise the software program. A memory controller coupled to the core may allocate a physical page of a memory to the trust domain, wherein data of the physical page of the memory is to be encrypted with an encryption key associated with the restricted key ID.

A processor includes a global register to store a value of an interrupted block count. A processor core, communicably coupled to the global register, may, upon execution of an instruction to flush blocks of a cache that are associated with a security domain: flush the blocks of the cache sequentially according to a flush loop of the cache; and in response to detection of a system interrupt: store a value of a current cache block count to the global register as the interrupted block count; and stop execution of the instruction to pause the flush of the blocks of the cache. After handling of the interrupt, the instruction may be called again to restart the flush of the cache.

A memory appliance may be provided comprising a processor, a communication interface, a memory, and a region access unit. The memory may be configured in an address space addressable by the processor. The communication interface may be configured to provide the client access to the region of the memory via client-side memory access before initialization of all of the region. A method to create a virtual copy of memory accessible by client-side memory access is also provided. A system may be provided that memory maps at least a portion of a file to a memory region, wherein a virtual address addressable is generated, and the at least a portion of file is accessible through the memory region at the virtual address. The virtual address may be registered with the communication interface, where registration of the virtual address provides client-side memory access to the memory region.

An apparatus comprises transaction handling circuitry to issue memory access transactions, each memory access transaction specifying an epoch identifier indicative of a current epoch in which the memory access transaction is issued; transaction tracking circuitry to track, for each of at least two epochs, a number of outstanding memory access transactions issued in that epoch; barrier termination circuitry to signal completion of a barrier termination command when the transaction tracking circuitry indicates that there are no outstanding memory access transactions remaining which were issued in one or more epochs preceding a barrier point; and epoch changing circuitry to change the current epoch to a next epoch, in response to a barrier point signal representing said barrier point. This helps to reduce the circuit area overhead for tracking completion of memory access transactions preceding a barrier point.

Systems, apparatuses, and methods for implementing translation lookaside buffer (TLB) striping to enable efficient invalidation operations are described. TLB sizes are growing in width (more features in a given page table entry) and depth (to cover larger memory footprints). A striping scheme is proposed to enable an efficient and high performance method for performing TLB maintenance operations in the face of this growth. Accordingly, a TLB stores first attribute data in a striped manner across a plurality of arrays. The striped manner allows different entries to be searched simultaneously in response to receiving an invalidation request which identifies a particular attribute of a group to be invalidated. Upon receiving an invalidation request, the TLB generates a plurality of indices with an offset between each index and walks through the plurality of arrays by incrementing each index and simultaneously checking the first attribute data in corresponding entries.

Examples herein relate to a system that includes a first memory device; a second memory device; and an input-output memory management unit (IOMMU). The IOMMU can search for a virtual-to-physical address translation entry in a first table for a received virtual address and based on a virtual-to-physical address translation entry for the received virtual address not being present in the first table, search a second table for a virtual-to-physical address translation entry for the received virtual address, wherein the first table is stored in the first memory device and the second table is stored in the second memory device. In some examples, based on a virtual-to-physical address translation entry for the received virtual address not being present in the second table, a page table walk is performed to determine a virtual-to-physical address translation for the received virtual address. In some examples, the first table includes an IO translation lookaside buffer (IOTLB).

Various embodiments are generally directed to virtualized systems. A first guest memory page may be identified based at least in part on a number of accesses to a page table entry for the first guest memory page in a page table by an application executing in a virtual machine (VM) on the processor, the first guest memory page corresponding to a first byte-addressable memory. The execution of the VM and the application on the processor may be paused. The first guest memory page may be migrated to a target memory page in a second byte-addressable memory, the target memory page comprising one of a target host memory page and a target guest memory page, the second byte-addressable memory having an access speed faster than an access speed of the first byte-addressable memory.

One disclosed embodiment includes a method for memory management. The method includes receiving a first request to clear one or more entries of a translation lookaside buffer (TLB), receiving a second request to clear one or more entries of the TLB, bundling the first request with the second request, determining that a processor associated with the TLB transitioned to an inactive mode, and dropping the bundled first and second requests based on the determination.