Advanced Micro Devices Corporation

Globalfoundries Inc.

Kevin John McGrath

David S Christie

Michael Mantor

Geoffrey S Strongin

Rex Eldon McCrary

Philip J Rogers

Mark M Leather

Robert Scott Hartog

Nuwan S Jayasena

Ralph Clay Taylor

Sebastien Nussbaum

Thomas R Woller

William Alexander Hughes

Michael T Clark

Dale E Gulick

James B Keller

Gabriel H Loh

Andrew Gregory Kegel

Alexander Klaiber

Arkaprava Basu

Mark Henry Oskin

Ramsey W Haddad

Scott A White

Vydhyanathan Kalyanasundharam

James K Pickett

Sergey Blagodurov

Stephan G Meier

Ralph Clayton Taylor

Chetana Nagendra Keltcher

Creigton S Asato

Hongwen Gao

Bruce R Holloway

Amro Awad

I-Cheng K Chen

The described embodiments include a computing device with two or more translation lookaside buffers (TLB). During operation, the computing device updates an entry in the TLB based on a virtual address to physical address translation and metadata from a page table entry that were acquired during a page table walk. The computing device then computes, based on a lease length expression, a lease length for the entry in the TLB. Next, the computing device sets, for the entry in the TLB, a lease value to the lease length, wherein the lease value represents a time until a lease for the entry in the TLB expires, wherein the entry in the TLB is invalid when the associated lease has expired. The computing device then uses the lease value to control operations that are allowed to be performed using information from the entry in the TLB.

Adaptive extension of leases for entries in a translation lookaside buffer

Methods and apparatus are described. A method includes an accelerated processing device running a process. When a maximum time interval during which the process is permitted to run expires before the process completes, the accelerated processing device receives an operating-system-initiated instruction to stop running the process. The accelerated processing device stops the process from running in response to the received operating-system-initiated instruction.

Preemptive context switching of processes on an accelerated processing device (APD) based on time quanta

The described embodiments include a computing device with two or more translation lookaside buffers (TLB) that performs operations for handling entries in the TLBs. During operation, the computing device maintains lease values for entries in the TLBs, the lease values representing times until leases for the entries expire, wherein a given entry in the TLB is invalid when the associated lease has expired. The computing device uses the lease value to control operations that are allowed to be performed using information from the entries in the TLBs. In addition, the computing device maintains, in a page table, longest lease values for page table entries indicating when corresponding longest leases for entries in TLBs expire. The longest lease values are used to determine when and if a TLB shootdown is to be performed.

Using leases for entries in a translation lookaside buffer

Methods, systems, and computer readable media for preemptive context-switching of processes on an accelerated processing device are based upon a comparison of the running time of the process and a threshold time quanta. A method includes preempting a process running on an accelerated processing device based upon a running time of the process and a threshold time quanta.

Preemptive context switching of processes on ac accelerated processing device (APD) based on time quanta

A method resumes an accelerated processing device (APD) wavefront in which a subset of elements have faulted. A restore command for a job including a wavefront is received. A list of context states for the wavefront is read from a memory associated with a APD. An empty shell wavefront is created for restoring the list of context states. A portion of not acknowledged data is masked over a portion of acknowledged data within the restored wavefronts.

Method for resuming an APD wavefront in which a subset of elements have faulted

Methods, systems, and computer readable media embodiments are disclosed for preemptive context-switching of processes running on a accelerated processing device. Embodiments include, detecting by an accelerated processing device a memory exception, and preempting a process from running on the accelerated processing device based upon the detected exception.

Preemptive context switching

Methods, systems, and computer readable media embodiments are disclosed for preemptive context-switching of processes running on an accelerated processing device. A method includes, responsive to an exception upon access to a memory by a process running on a accelerated processing device, whether to preempt the process based on the exception, and preempting, based upon the determining, the process from running on the accelerated processing device.

Process device context switching

Embodiments describe herein provide a method of for managing task scheduling on a accelerated processing device. The method includes executing a first task within the accelerated processing device (APD), monitoring for an interruption of the execution of the first task, and switching to a second task when an interruption is detected.

Software mechanisms for managing task scheduling on an accelerated processing device (APD)

The desire to use an Accelerated Processing Device (APD) for general computation has increased due to the APD's exemplary performance characteristics. However, current systems incur high overhead when dispatching work to the APD because a process cannot be efficiently identified or preempted. The occupying of the APD by a rogue process for arbitrary amounts of time can prevent the effective utilization of the available system capacity and can reduce the processing progress of the system. Embodiments described herein can overcome this deficiency by enabling the system software to pre-empt a process executing on the APD for any reason. The APD provides an interface for initiating such a pre-emption. This interface exposes an urgency of the request which determines whether the process being preempted is allowed a grace period to complete its issued work before being forced off the hardware.

Method for urgency-based preemption of a process

Embodiments describe herein provide an apparatus, a computer readable medium and a method for simultaneously processing tasks within an APD. The method includes processing a first task within an APD. The method also includes reducing utilization of the APD by the first task to facilitate simultaneous processing of the second task, such that the utilization remains below a threshold.

Partitioning resources of a processor

In one embodiment, a processor is configured to operate in a first mode in which privilege level protection is disabled and paging is enabled. In another embodiment, a method is contemplated including intercepting a write to a control register by a guest executing in a processor; determining that the write attempts to establish a first mode in the processor in which privilege level protection is disabled and paging is disabled; and causing the guest to execute in a second mode in which privilege level protection is disabled and paging is enabled instead of the first mode. A computer accessible medium comprising instruction implementing at least a portion of the method is also described.

Virtualization of real mode execution

A computer system includes a processor which may initialize a secure execution mode by executing a security initialization instruction. Further, the processor may operate in the secure execution mode by executing a secure operating system code segment. The computer system also includes an input/output (I/O) interface coupled to the processor via an I/O link. The I/O interface may receive transactions performed as a result of the execution of the security initialization instruction. The transactions include at least a portion of the secure operating system code segment. The I/O interface may also determine whether the processor is a source of the transactions. The computer system further includes a security services processor coupled to the I/O interface via a peripheral bus. The I/O interface may convey the transactions to the security services processor dependent upon determining that the processor is the source of the transactions.

Computer system including a secure execution mode-capable CPU and a security services processor connected via a secure communication path

The initialization of a computer system including a secure execution mode-capable processor includes storing a secure operating system code segment loader to a plurality of locations corresponding to a particular range of addresses within a system memory. The method also includes executing a security initialization instruction. Executing the security initialization instruction may cause several operations to be performed including transmitting a start transaction including a base address of the particular range of addresses. In addition, executing the security instruction may also cause another operation to be performed including retrieving the secure operating system code segment loader from the system memory and transmitting the secure operating system code segment loader for validation as a plurality of data transactions.

Initialization of a computer system including a secure execution mode-capable processor

A method for controlling operation of a secure execution mode-capable processor includes receiving access requests to a plurality of addressable locations within a system memory. The method may further include preventing the access requests from completing in response to determining that the secure execution mode-capable processor is operating in a secure execution mode.

Method and apparatus for controlling operation of a secure execution mode-capable processor in system management mode

In one embodiment, a register in a processor is programmable with an intercept indication indicative of whether or not an event that would cause a transition by the processor to a first mode is to be intercepted during execution of a guest. Responsive to the intercept indication and further responsive to detecting the event, execution circuitry in the processor is configured to exit the guest. In another embodiment, a method comprises: detecting an event that would cause a processor to transition to a first mode, wherein first code is to be executed in the first mode; and causing the first code to be executed in a guest responsive to the detecting. In still another embodiment, a computer accessible medium comprising instructions which when executed in response to detecting the event, cause the first code to be executed in a guest.

Growing community of inventors

Los Gatos, CA, United States of America

Kevin John McGrath