Microsoft Technology Licensing, LLC

Jason T Cobb

Santosh Sharad Jodh

James George Cavalaris

Eugene Seichi Lin

Lonny Dean McMichael

Pierre-Yves Santerre

George Evangelos Roussos

Bjorn Benjamin Levidow

Mark Derbecker

Frederick Justus Smith

Raymond D Pedrizetti

Jamie Hunter

Vadim Bluvshteyn

Vadim Bluvshtevn

Techniques for device driver management/installation are provided. In at least some embodiments, a device driver management system can be employed by a user to selectively rollback a currently installed device driver to one or a plurality of previously installed device driver(s). Additionally, the system can be employed by the user to revert to a pristine state of not having the device driver installed at all, for example, the NULL driver (e.g., in the situation in which the first driver installed on the device causes machine instability). The system stores information associated with driver(s) running on a specific device and allows a user to selectively revert to any one of a plurality of previously installed device driver(s), for example, if they experience a problem with a newer driver. Rollback point(s) can be stored, for example, in the system registry.

Device driver rollback

A system and method for entity management is provided. In accordance with an aspect of the present invention, a system administrator is given control over device(s) that can be installed on a computer system(s), and/or how the installation can be performed. Thus, a system administrator can specify device installation policy that prevents the computer system from automatically installing device(s) (e.g., even when an appropriate driver is available to be installed). For example, this policy can be applied to substantially all new devices and/or only a subset of devices. Optionally, policy can also control how device(s) that are already installed will be made available to the computer system.

Privileged used control of device installation and/or availability

A system and method for device driver management/installation is provided. The device driver management system can be employed by a user to selectively rollback a currently installed device driver to one or a plurality of previously installed device driver(s). Additionally, the system can be employed by the user to revert to a pristine state of not having the device driver installed at all, for example, the NULL driver (e.g., in the situation in which the first driver installed on the device causes machine instability). The system stores information associated with driver(s) running on a specific device and allows a user to selectively revert to any one of a plurality of previously installed device driver(s), for example, if they experience a problem with a newer driver. Rollback point(s) can be stored, for example, in the system registry.

Systems and methods are disclosed that facilitate refining a list of candidate driver packages for a detected plug-and-play device. A device can be detected and a hardware ID there for can be determined and compared to hardware IDs in a data store to determine a preliminary list of driver packages that match the hardware device ID. The driver packages can be tagged with additional rich targeting criteria that must be satisfied in order for the driver package to be retained in the candidate list. If additional criteria in the driver packages are not satisfied, then the driver package can be excised from the candidate list to refine driver package compatibility with a computer to which the device is connected.

Rich targeting criteria for selection of driver packages

The subject invention relates to systems and methods that automatically monitor computer platform components and initiate automated resource recovery procedures based on detected periods of component inactivity. In one aspect, an automated maintenance system for computer resources is provided. The system includes a controller that monitors installation and removal of system components that cooperate to facilitate various operations of a computer. A threshold component supplies time out or decay values for the devices to determine inactive periods of the devices, whereby the controller removes persistent memory references for the devices based in part on the time out values. Supervisory threshold functions can be provided to cause the system to perform maintenance operations at other desired intervals in order to allow periods of device inactivity while maintaining desired component information on the respective system.

Un-installation of inactive or removed peripheral device drivers

Systems and methods are provided that prevent faulty drivers from being loaded and/or can prevent faulty drivers from being installed. Thus, instability of a computer system can be mitigated. Additionally, the occurrence of data corruption, system halting and the like can be reduced. A requested driver is compared to a list of faulty drivers from a faulty driver database. If the requested driver is in the list of faulty drivers, the requested driver is deemed faulty or defective, and is prevented from being loaded. Additionally, if the requested driver is in the list of faulty drivers, the requested driver can be prevented from being installed. Otherwise, the requested driver is operable and can be installed and/or loaded.

Faulty driver protection comparing list of driver faults

A method and system for avoiding the overwriting of drivers by subsequent versions or other commonly named drivers includes generating a unique identity for every eligible driver package. Driver files from the driver package, or the entire driver package itself, are then installed in a subdirectory location in a common storage based on the unique identity. The driver files may be loaded to a memory from the subdirectory location. Thus, multiple driver packages and driver files having the same name may be installed and loaded side-by-side.

Side-by-side drivers

Driver protection

The present invention is directed to tracking and managing failure-susceptible operations in a computer system. In a preferred embodiment, the performance of failure-susceptible operations is monitored such that, when execution of the operating system is interrupted and then restarted, it can be determined which failure-susceptible operation was in progress at the point at which execution of the operating system was interrupted. A list is made of the operations that failed to execute to completion and subsequent invocation of operations on the list is inhibited. A user may review the list and select operations, allowing the selected operations to be subsequently invoked.

Tracking and managing failure-susceptible operations in a computer system

The present invention is directed to tracking and managing failure-susceptible operations in a computer system. In a preferred embodiment, the performance of failure-susceptible operations is monitored, such that, when execution of the operating system is interrupted then restarted, it can be determined which failure-susceptible operation was in progress at the point at which execution of the operating system was interrupted. In one embodiment, this monitoring is performed using a stack data structure. When a failure-susceptible operation is invoked, an entry is pushed on the stack identifying the invoked failure-susceptible operation. On the other hand, when a failure-susceptible operation is completed, the entry identifying the completed failure-susceptible operation is popped off the stack. The failure-susceptible operation that was in progress at the point at which execution was last interrupted is then determined to be the failure-susceptible operation identified by the top entry on the stack when execution resumes. In a further preferred embodiment, subsequent attempted invocation of the operation determined to be in progress when execution was interrupted is inhibited.

Growing community of inventors

Sammamish, WA, United States of America

Jason T Cobb