Abstract: A system comprises a physical processor to execute a virtual machine manager to run, on a logical core, a virtual machine including a guest user application and a virtual CPU. Circuitry coupled to an external device is to receive an interrupt request from the external device for the guest user application, locate a first interrupt data structure associated with the guest user application, generate a first interrupt with the first interrupt data structure based on a first interrupt vector for the interrupt request, locate a second interrupt data structure associated with the virtual CPU, and generate a first notification interrupt for the virtual CPU with the second interrupt data structure based on a first notification vector in the first interrupt data structure. The circuitry may generate a second notification interrupt for the logical core using a second notification vector and a logical core identifier from the second interrupt data structure.

Abstract: Systems and devices can include a controller and a command queue to buffer incoming write requests into the device. The controller can receive, from a client across a link, a non-posted write request (e.g., a deferred memory write (DMWr) request) in a transaction layer packet (TLP) to the command queue; determine that the command queue can accept the DMWr request; identify, from the TLP, a successful completion (SC) message that indicates that the DMWr request was accepted into the command queue; and transmit, to the client across the link, the SC message that indicates that the DMWr request was accepted into the command queue. The controller can receive a second DMWr request in a second TLP; determine that the command queue is full; and transmit a memory request retry status (MRS) message to be transmitted to the client in response to the command queue being full.

Abstract: Techniques for address translation cache (ATC) reservation in offload devices are disclosed. In the illustrative embodiment, a processor of a compute device sends a start ATC reservation descriptor to an offload device. The start ATC reservation descriptor includes an identifier associated with a virtual machine for which at least part of an address translation cache of the offload device should be reserved. The offload device establishes a zone in the ATC of the offload device that is reserved for address translations associated with the identifier. Such cache reservation may be used when, e.g., a priority of a task is high or there is a need for critical or important workload to have lower latency and higher throughput.

Abstract: In one embodiment, an apparatus includes a processor comprising an address translation cache (ATC); a shared work queue (SWQ) associated with the ATC, and a port to couple to a host processor over a Peripheral Component Interconnect Express (PCIe)-based link. The apparatus also includes circuitry to receive address translation information from a memory management unit of the host processor that includes virtual memory address to physical memory address translations, store the address translation information in the ATC, receive an invalidation command from the host processor indicating an invalidation of address translation information stored in the ATC, modify the address translation information in the ATC based on the invalidation command, and store completion data in a memory location indicated by the invalidation command.

Abstract: In an embodiment, a core includes at least one execution circuit. The core may be configured to: send a command for a first address translation cache (ATC) of a first device to perform an operation, the core to send the command to a first device queue of a shared memory, the first device queue associated with the first ATC; and send a register write directly to the first device to inform the first ATC regarding presence of the command in the first device queue. Other embodiments are described and claimed.

Abstract: Systems and devices can include a controller and a command queue to buffer incoming write requests into the device. The controller can receive, from a client across a link, a non-posted write request (e.g., a deferred memory write (DMWr) request) in a transaction layer packet (TLP) to the command queue; determine that the command queue can accept the DMWr request; identify, from the TLP, a successful completion (SC) message that indicates that the DMWr request was accepted into the command queue; and transmit, to the client across the link, the SC message that indicates that the DMWr request was accepted into the command queue. The controller can receive a second DMWr request in a second TLP; determine that the command queue is full; and transmit a memory request retry status (MRS) message to be transmitted to the client in response to the command queue being full.

Abstract: Aspects of the embodiments are directed to systems and methods for providing and using hints in data packets to perform memory transaction optimization processes prior to receiving one or more data packets that rely on memory transactions. The systems and methods can include receiving, from a device connected to the root complex across a PCIe-compliant link, a data packet; identifying from the received device a memory transaction hint bit; determining a memory transaction from the memory transaction hint bit; and performing an optimization process based, at least in part, on the determined memory transaction.

Abstract: As described herein, for a selected process identifier and virtual address, a page fault arising from multiple sources can be solved by a one-time operation. The selected process identifier can include a virtual function (VF) identifier or process address space identifier (PASID). In some examples, solving a page fault arising from multiple sources by a one-time operation comprises invoking a page fault handler to determine an address translation.

Abstract: Systems, methods, and devices for software-driven resource reservation of an input/output memory management unit (IOMMU) are provided. A system may include a peripheral device and a processing device. The peripheral device may be accessible to a virtual machine running on the processing device via direct memory access (DMA) that is translated by an IOMMU). The processing device may run the virtual machine and a virtual machine manager. The processing device also includes the IOMMU, which is configurable to reserve a subset of resources of the IOMMU to the virtual machine based on a descriptor provided by the virtual machine manager.

Abstract: Examples described herein relate to a packet processing device that includes circuitry to receive an address translation for a virtual to physical address prior to receipt of a GPUDirect remote direct memory access (RDMA) operation, wherein the address translation is provided at initiation of a process executed by a host system and circuitry to apply the address translation for a received GPUDirect RDMA operation.

Abstract: Aspects of the embodiments are directed to systems and methods for providing and using hints in data packets to perform memory transaction optimization processes prior to receiving one or more data packets that rely on memory transactions. The systems and methods can include receiving, from a device connected to the root complex across a PCIe-compliant link, a data packet; identifying from the received device a memory transaction hint bit; determining a memory transaction from the memory transaction hint bit; and performing an optimization process based, at least in part, on the determined memory transaction.

Abstract: In one embodiment, a read request is received from a peripheral device across an interconnect, with the read request including a process identifier and an encrypted virtual address. One or more keys are obtained based on the process identifier of the read request, and the encrypted virtual address of the read request is decrypted based on the one or more keys to obtain an unencrypted virtual address. Encrypted data is retrieved from memory based on the unencrypted virtual address, and the encrypted data is decrypted based on the one or more keys to obtain plaintext data. The plaintext data is transmitted to the peripheral device across the interconnect.

Abstract: Systems, methods, and devices can include circuitry or computer program products to receive a memory address translation request message from a downstream connected device; identify, from the memory address translation request message, a permission indication that the device intends to perform a write access to one or more memory address locations specified in the memory address translation request message; identify, from an address translation and protection table (ATPT), a dirty bit value associated with the one or more memory address locations; and transmit a translation of the one or more memory address locations and a read or read+write permission to the device based on the permission indication in the memory address translation request message and the dirty bit.

Abstract: Systems and devices can include a controller and a command queue to buffer incoming write requests into the device. The controller can receive, from a client across a link, a non-posted write request (e.g., a deferred memory write (DMWr) request) in a transaction layer packet (TLP) to the command queue; determine that the command queue can accept the DMWr request; identify, from the TLP, a successful completion (SC) message that indicates that the DMWr request was accepted into the command queue; and transmit, to the client across the link, the SC message that indicates that the DMWr request was accepted into the command queue. The controller can receive a second DMWr request in a second TLP; determine that the command queue is full; and transmit a memory request retry status (MRS) message to be transmitted to the client in response to the command queue being full.

Abstract: Techniques for offload device address translation fetching are disclosed. In the illustrative embodiment, a processor of a compute device sends a translation fetch descriptor to an offload device before sending a corresponding work descriptor to the offload device. The offload device can request translations for virtual memory address and cache the corresponding physical addresses for later use. While the offload device is fetching virtual address translations, the compute device can perform other tasks before sending the corresponding work descriptor, including operations that modify the contents of the memory addresses whose translation are being cached. Even if the offload device does not cache the translations, the fetching can warm up the cache in a translation lookaside buffer. Such an approach can reduce the latency overhead that the offload device may otherwise incur in sending memory address translation requests that would be required to execute the work descriptor.

Abstract: Systems and devices can include a controller and a command queue to buffer incoming write requests into the device. The controller can receive, from a client across a link, a non-posted write request (e.g., a deferred memory write (DMWr) request) in a transaction layer packet (TLP) to the command queue; determine that the command queue can accept the DMWr request; identify, from the TLP, a successful completion (SC) message that indicates that the DMWr request was accepted into the command queue; and transmit, to the client across the link, the SC message that indicates that the DMWr request was accepted into the command queue. The controller can receive a second DMWr request in a second TLP; determine that the command queue is full; and transmit a memory request retry status (MRS) message to be transmitted to the client in response to the command queue being full.

Abstract: Method and apparatus for per-agent control and quality of service of shared resources in a chip multiprocessor platform is described herein. One embodiment of a system includes: a plurality of core and non-core requestors of shared resources, the shared resources to be provided by one or more resource providers, each of the plurality of core and non-core requestors to be associated with a resource-monitoring tag and a resource-control tag; a mapping table to store the resource monitoring and control tags associated with each non-core requestor; and a tagging circuitry to receive a resource request sent from a non-core requestor to a resource provider, the tagging circuitry to responsively modify the resource request to include the resource-monitoring and resource-control tags associated with the non-core requestor in accordance to the mapping table and send the modified resource request to the resource provider.

Abstract: Embodiments of the present disclosure may be related to an electronic device that includes a root complex; and a processor coupled with the root complex. The root complex may identify a first direct memory access (DMA) transaction and a second DMA transaction respectively related to a first task and a second task of a device communicatively coupled with the root complex through an input/output (I/O) fabric. The root complex may further cache a first memory translation related to the first DMA transaction in a first micro translation lookaside buffer (uTLB) of the root complex. The root complex may further cache a second memory translation related to the second DMA transaction in a second uTLB of the root complex. Other embodiments may be described and/or claimed.

Abstract: Systems and devices can include a controller and a command queue to buffer incoming write requests into the device. The controller can receive, from a client across a link, a non-posted write request (e.g., a deferred memory write (DMWr) request) in a transaction layer packet (TLP) to the command queue; determine that the command queue can accept the DMWr request; identify, from the TLP, a successful completion (SC) message that indicates that the DMWr request was accepted into the command queue; and transmit, to the client across the link, the SC message that indicates that the DMWr request was accepted into the command queue. The controller can receive a second DMWr request in a second TLP; determine that the command queue is full; and transmit a memory request retry status (MRS) message to be transmitted to the client in response to the command queue being full.

Abstract: Systems, methods, and devices can include circuitry or computer program products to receive a memory address translation request message from a downstream connected device; identify, from the memory address translation request message, a permission indication that the device intends to perform a write access to one or more memory address locations specified in the memory address translation request message; identify, from an address translation and protection table (ATPT), a dirty bit value associated with the one or more memory address locations; and transmit a translation of the one or more memory address locations and a read or read+write permission to the device based on the permission indication in the memory address translation request message and the dirty bit.