Microsoft Technology Licensing, LLC

Stephane Plante

Jacob Kappeler Oshins

Andrew John Thornton

Tony Dwayne Pierce

Adam G Poulos

Andrew J Ritz

John D Vert

James A Schwartz, Jr

Adrian J Oney

Vijay Gajjala

Michael H Tsang

Nicholas Stephen Judge

Timothy Arthur Gill

Pierre N Martin

Robert B Nelson

James T Jacoby

Craig Alan Gick

Ajay S Kang

Vincent J Geglia Ii

Luke Jared Lenhart

Richard K Costleigh

Craig A Gick

Embodiments are disclosed for managing and providing access to a collection of digital resources. One embodiment provides a method comprising receiving a request to access a resource for a principal and determining one or more principal groups to which the principal belongs. The method further comprises obtaining resource set membership information indicating a resource set to which the resource belongs, and obtaining resource set access policy information for the resource set to which the resource belongs. The method yet further comprises determining whether the principal is allowed to access the resource based on the principal group membership information and the resource set access policy information, and, if the principal is allowed to access the resource, then permitting access to the resource by the principal.

Discretionary policy management in cloud-based environment

The exploitation of rolling back a system configuration to a previous system configuration is prevented by querying the update status of the system and comparing the received response with an expected response. If the comparison indicates that the update version of the system is older than the expected update version, the system is determined to have been rolled back. Accordingly, appropriate action is taken, such as sending a warning to the system, stopping the system from operating, disabling selected features, disconnecting the system from a network, banning the system from future connections to the network, and/or installing an update. The query can include a request for update version numbers of updates, times when updates were applied, predetermined questions, and an indication of the system (e.g., machine serial number, unique ID value).

Prevention of exploitation of update rollback

A base ACPI power management state is selected and a feature of a power management state that is to be created is defined. The definition of the power management state to be created is attached to an object of the base ACPI power management state in an ACPI namespace.

Creation of hybrid power management states

Transitioning to a suspend to RAM sleeping state while also protecting against power losses while sleeping is provided. System state context data is saved to non-volatile storage and components in the computer system prepare to transition to a suspend to disc sleeping. A transition to the suspend to RAM sleeping state is then effected. Alternatively, after the system context is saved and the components are prepared to transition, the system may wake to a working state. The components may be directed to prepare for transitioning to a suspend to RAM sleeping state, and then the BIOS may be directed to execute the transition. In either embodiment, if power to the system is lost while the system is in the suspend to RAM system state, then the system may resume to a working state by reading the context file stored to non-volatile storage.

Systems and methods for providing power-loss protection to sleeping computers systems

A system for adding multiple GPE blocks (in addition to the system/root GPE block device) to a computing system by creating a device entry in the ACPI namespace, and using a _CRS object to describe the system resources consumed by the device is described. The GPE block device may then access associated hardware devices through a well known mechanism (either I/O or Memory Mapped accesses). By creating additional GPE block devices within the ACPI namespace, general purpose events may be delivered using more traditional hardware interrupt mechanisms than with existing systems (e.g., wiring GPE blocks together). Moreover, by putting GPE block devices in the ACPI namespace, hardware components having hardware registers may be 'hot plugged' to the computing system.

System and method for adding hardware registers to a power management and configuration system

A method and system that enables customized computer machines to be more readily developed by removing the function of resource translation out of the hardware abstraction layer (HAL). A machine manufacturer describes a machine in firmware, such as accordance with the Advanced Configuration and Power Interface (ACPI) specification, using ACPI machine language (AML). Operating system components such as a Plug and Play (PnP) manager in the kernel, in conjunction with an ACPI driver, interpret the description information and locate resources (bus bridges) for which translation is needed. For any arbitrary bus architecture or CPU to PCI bridge implementation that can be expressed, e.g., in ACPI firmware, the invention provides a translator external to the HAL. In one implementation, a PnP driver communicates with the ACPI driver and various drivers in driver stacks via I/O request packets (IRPs) to look for resource translators. The ACPI driver analyzes the machine description and returns a translator if found for such a resource. The resource is then configured to output cycles based on the translator.

Dynamically configuring resources for cycle translation in a computer system

A system and method that utilizes a common hardware register pseudo-language are disclosed. The present invention employs a common platform to specify hardware functionality and to execute hardware action(s). Hardware actions can be effectuated by performing a series of instructions, which comprise hardware register primitives and resources utilized by the primitives. The hardware register primitives are operations defined according to the common hardware register pseudo-language. The series of instructions can be loaded prior to boot or during initialization.

System and method for specifying and utilizing hardware functionality by executing a common hardware register pseudo-language

A method and system for adaptively throttling a computer is provided. Prior CPU utilization is calculated when a CPU enters an idle state. If the prior CPU utilization warrants a change in the CPU performance level, an appropriate CPU performance level is calculated. Policies, including thermal policies and battery charge policies, may be applied to increase or decrease the appropriate CPU performance level. If prior CPU utilization and/or policies dictate a different CPU performance level, the CPU performance level is changed.

Method and system for using idle threads to adaptively throttle a computer

A mechanism for simulating the existence of hardware in a configuration and power management system is described. In one aspect, a simulator interfaces with the configuration and power management system to generate simulated events. In another aspect, accesses to hardware registers are simulated by registering the simulator with the configuration and power management system to handle accesses to a simulated hardware device. A component within the configuration and power management system may define the simulated hardware device such that accesses to the simulated hardware device occur with respect to a defined I/O space.

System and method for simulating hardware components in a configuration and power management system

A method and system for selectively enabling wake events in software of a computer system to overcome problems arising when hardware devices fail to clear a wake signal. The operating system manages wake events, and also distinguishes between events that are exclusively wake events, exclusively run-time events, and shared wake and run-time events. At boot time, the ACPI driver examines system tables provided by firmware to determine which GPEs are associated with wake-up events, either exclusively or shared with run-time events. These wake event associations are tracked and managed differently from events received on other hardware register pins. When the operating system receives events in a GPE Status hardware register that is enabled in a counterpart Enable register, the operating system runs an associated GPE method. When the GPE method has completed, the operating system selectively determines whether the event needs to be re-enabled.

Preventing general purpose event interrupt storms in a computer system

An advanced configuration and power management system is described which supports an unload command by identifying those objects that are to be unloaded and, rather than simply attempting to delete the objects, detaches them from the namespace while keeping them in memory in a location accessible by the system. In this way, the system avoids the problem of trying to synchronize accesses to the objects, yet the objects not longer exist in the tree, so namespace collisions are avoided.

System and method for unloading namespace devices

Described is a system and method by which a PCI device may be controlled by firmware in an Advanced Configuration and Power Management system. A device connected to the PCI bus is described in firmware with AML that declares a PCI BAR operation region associated with the PCI device. A generic driver is loaded and registers itself to handle any access to or from the PCI device my means of the PCI BAR operation region. Essentially, the generic driver is enumerated as a functional driver (FDO) for the PCI device. When a Plug-n-Play manager assigns base addresses to each PCI device on the PCI bus, the generic driver stores this information. Calls by the firmware to the PCI BAR operation region identify the PCI BAR number (i.e., the PCI device) and give an offset. The generic driver resolves that information into an absolute memory or I/O address based on the current BAR assigned by the PnP manager and performs the requested access.

PCI bar target operation region

A method and system for selectively enabling wake events (wake general purpose events, or GPEs) in software of a computer system to overcome problems arising when hardware devices fail to clear a wake signal, and as part of handling the event, to re-run control methods that fail to properly execute. The operating system intelligently manages wake events, and also distinguishes between events that are exclusively wake events, exclusively run-time events, and shared wake and run-time events, and handles control method failures. At boot time, the ACPI driver examines system tables provided by firmware to determine which GPEs are associated with wake-up events, either exclusively or shared with run-time events. These wake event associations are tracked, whereby they are managed differently from events received on other hardware register pins. In general, when the operating system receives events, the operating system runs an associated GPE method. If the control method fails, the operating system can re-run the GPE method, after a delay. When the GPE method has completed, the operating system selectively determines whether the event needs to be re-enabled.

Growing community of inventors

Kirkland, WA, United States of America

Stephane Plante