Microsoft Technology Licensing, LLC

Facebook Technologies, LLC

Facebook, Inc.

Meta Platforms Technologies, LLC

Mircrosoft Corporation

Andrew Michael Rogers

Yadhu Nandh Gopalan

Sachin C Patel

Scott Randall Shell

Bor-Ming Hsieh

Arthur Zwiegincew

Upender Sandadi

Neil Laurence Coles

Matthew G Lyons

Mark Scott Tonkelowitz

David Michael Callaghan

Aman Kansal

John Mark Miller

Bernhard Poess

Mark Plagge

Michael E Markley

Keith C Bentley

Jon Vincent

David Fischer Kelley

Gleb Kurtsov

Angelo Renato Vals

Christopher Ross Jordan

Jeffrey D Glaum

Brian Hudson

Yangqing Jia

Frank James Eisenhart, Iii

Ariane Jansen

Hsiu-Tung Alex Yu

Avinashreddy Bathula

Joel Curtis McCall

Jimmy Alexander

Jeffery D Glaum

In certain embodiments, a method includes starting an application as a first process within a user space of an operating system. The application instantiates a key-value store library associated with the operating system. The method also includes managing, by a block device service running within the user space, one or more first data blocks of a persistent storage device allocated to the key-value store library. In addition, the method includes receiving, by a kernel of the operating system, a system call request comprising an access request generated by the key-value store library. The method further includes granting, by the block device service, and in response to the access request, the key-value store library access to the one or more first data blocks through an inter process communication channel between the application and the block device service via the kernel.

System and method for key value store

In certain embodiments, a method includes starting an application as a first process within a user space of an operating system. The application instantiates a data storage system associated with the operating system. The method also includes starting a block device service as a second process within the user space of the operating system, the block device service being configured to manage a persistent storage device of the computing device. In addition, the method includes receiving, by a kernel of the operating system, a system call request from the application to communicate with the block device service, the system call request is generated by the application using the data storage system and comprises an access request to access the persistent storage device. The method further includes providing the application, in response to the system call request, access to the block device service through the IPC channel.

System and method for split storage stack

In one embodiment, a method includes a system building a first machine-learning module and one or more secondary machine-learning modules for operating with an application. The first and second modules may be configured to utilize, in operation, particular types of processing hardware, respectively. The system may receive from a client device a request to download the application, and in response send to the device the application with the first module. The system may then receive another request to download a selected one of the secondary modules, which may be selected based on a determination by the application running on the device that the associated type of processing hardware is available. In response, the system may send the selected module to the device. The application may be configured to selectively use the first module or the selected module to perform an operation.

Dynamic integration of machine-learning modules

In certain embodiments, a method includes starting an application as a first process within a user space of an operating system. The application instantiates a key-value store and a file system associated with the operating system. The method also includes managing, by a block device service running within the user space, one or more first data blocks of a persistent storage allocated to the key value store and one or more second data blocks of the persistent storage allocated to the file system. In addition, the method includes receiving, by a kernel of the operating system, a system call request comprising an access request generated by the key-value store or the file system. The method further includes granting, by the block device service, and in response to the access request, the key value store or the file system access to the one or more first or second data blocks.

Authenticating a backup image stored on a storage device prior to restoring the backup image. A value such as a hash value is generated based on the image to be preserved. The value is stored on the storage device in a first area accessible for writing only when the computing device is in an update mode. The first area becomes read-only when the computing device is in an operating system mode. The image is stored on the storage device in a second area accessible for writing when the computing device is in the update mode or the operating system mode. When attempting to restore the image from the second area, the value from the first area is compared to another value generated at the time of the restoration for validation.

Authenticating a backup image with bifurcated storage

An inter-process messaging security management may be provided. A message comprising an operation to be performed may be sent from a process operating in a process chamber to a second process operating in another chamber. Before the message is allowed to be delivered, the validity of the operation contained in the message may be verified and a security policy may be examined to determine whether the message is permitted to be sent from the first process to the second process. If the security policy permits the second process to execute the operation requested by the first process, the message may be delivered to the second process. If the operation is not permitted, the message may not be delivered and an error message may be returned to the first process.

Inter-process message security

A transaction safe file system uses two sets of file allocation tables and bitmap images to perform file modifications on one of the sets while the other set remains a last known good set. After a modification is complete, a pointer is changed to the newly modified set, and the newly modified set becomes the last known good set. The sets are then synchronized. The file allocation table is used to define cluster chains while the bitmap image is used to determine if a cluster is free or not. In some operations, only the bitmap image may need to be manipulated. The file system may be used in a transaction safe mode as well as a non-transaction safe mode.

Storage system format for transaction safe file system

Directories in a file system are defined with a dummy cluster in a file allocation table as the initial entry. Subsequent clusters in a directory's definition may define any data for the directory that can be changed in a transaction-safe mode. A directory may be modified in a transaction-safe mode by modifying any of the subsequent clusters while tracking changes in a second file allocation table. When the changes have been made to the directory, a pointer to the second file allocation table may be switched to indicate that the second file allocation table is now last known good. The first file allocation table may then be synchronized with the second.

Fat directory structure for use in transaction safe file system

A method of managing memory of a computing device includes providing a first memory that can be allocated as cache memory or that can be used by a computing device component. A first memory segment can be allocated as cache memory in response to a cache miss. Cache size can be dynamically increased by allocating additional first memory segments as cache memory in response to subsequent cache misses. Cache memory size can be dynamically decreased by reallocating first memory cache segments for use by computing device components. The cache memory can be a cache for a second memory accessible to the computing device. The computing device can be a mobile device. The first memory can be an embedded memory and the second memory can comprise embedded, removable or external memory, or any combination thereof. The maximum size of the cache memory scales with the size of the first memory.

Efficient cache management

FAT directory structure for use in transaction safe file system

Embodiments provide a security infrastructure that may be configured to run on top of an existing operating system to control what resources can be accessed by an applications and what APIs an application can call. Security decisions are made by taking into account both the current thread's identity and the current thread's call chain context to enable minimal privilege by default. The current thread context is captured and a copy of it is created to be used to perform security checks asynchronously. Every thread in the system has an associated identity. To obtain access to a particular resource, all the callers on the current thread are analyzed to make sure that each caller and thread has access to that resource. Only when each caller and thread has access to that resource is the caller given access to that resource.

Granting least privilege access for computing processes

A computing device includes a processor, a storage device having an executable file, and a file system for executing the file in place on the storage device on behalf of the processor. The file is divided into multiple non-contiguous fragments on the storage device, and the computing device further includes a virtual address translator interposed between the processor and the storage device for translating between physical addresses of the fragments of the file on the storage device and corresponding virtual addresses employed by the processor.

Computing device with relatively limited storage space and operating / file system thereof

Fat directory structure for use in transaction safe file

A transaction log for flash recovery includes a chained sequence of blocks specifying the operations that have been performed, such as a write to a sector or an erase to a block. Checkpoints are performed writing the entire flash state to flash. Once a checkpoint is performed, all of the log entries prior to the checkpoint are deleted and the log processing on recovery begins with the latest checkpoint. If the system is able to safely shutdown, then a checkpoint may be performed before the driver unloads, and on initialization, the entire persisted flash state may be loaded into the flash memory with a minimal amount of flash scanning. If a power failure occurs during system operation, then on the next boot-up, only the sectors or blocks specified in the log entries after the latest checkpoint have to be scanned, rather than all the sectors on the part.

Flash recovery employing transaction log

Growing community of inventors

Bellevue, WA, United States of America

Andrew Michael Rogers