Vmware, Inc.

Facebook Technologies, LLC

Meta Platforms Technologies, LLC

Christoph Klee

Adrian Drzewiecki

Bernhard Poess

Mukund Gunti

Sumit Kamath

Wenguang Wang

Christos Karamanolis

Richard Paul Spillane

Gabriel Tarasuk-Levin

Maxime Austruy

Chi-Hsiang Su

Aman Nijhawan

Robert Bosch

Vadim Victor Spivak

Jesse Pool

Mounesh Badiger

Nishant Yadav

Mark Lucovsky

In one embodiment, a microkernel of an operating system executing on a computing system receives, from an application, a first system call requesting to communicate with a service registry. The first system call being associated with an operation request. The microkernel sends a first instruction to the service registry. The microkernel receives, from the service registry, a second system call requesting to communicate with at least one of an application service or a protocol service. The second system call being associated with the operation request. The microkernel sends a second instruction to at least one of the application service or the protocol service. The microkernel receives, from at least one of the application service or the protocol service, a third system call requesting to communicate with a driver service. The third system call being associated with the operation request. The microkernel sends a third instruction to the driver service.

Microkernel architecture with enhanced reliability and security

In one embodiment, a method includes creating, by a computing device, an inter process communication (IPC) channel for communication from a producer process to a consumer process. The IPC channel includes a message buffer mapped to a first virtual address space of a kernel of an operating system and to a second virtual address space of the consumer process in a user space of the operating system. A sender handle for the message buffer is sent to the producer process. The kernel receives a request, which includes the sender handle, to send a message from the producer process to the consumer process through the IPC channel. Responsive to a determination that the request is permitted, the kernel writes the message to the message buffer using the sender handle, such that message written in the message buffer is accessible to the consumer process through the second address space.

Inter-process communication for microkernel operating system

In one embodiment, a kernel of an operating system receives a request to store a message in a message buffer, which includes a plurality of chunks. Each chunk of the message buffer is associated with a message-occupancy indicator. The kernel determines a start and end position of the message when stored in the message buffer. The kernel determines, based on the start and end positions, whether the message, when stored, will occupy more than one chunk. Responsive to determining that the message will occupy more than one chunk, the kernel determines each chunk that will contain the message when stored. If a message occupancy-indicator associated with each chunk indicates that at least each chunk after a first chunk associated with the start position is unoccupied, the kernel stores the message in the message buffer and updates the message-occupancy indicator for each chunk containing the stored message.

Memory buffer management for microkernel operating system

In one embodiment, a method includes receiving a request to create a data pipeline by an operating system executing on a computing device. The operating system allocates a shared virtual memory region for the data pipeline. The shared virtual memory region is mapped to a first virtual address space of a first process and a second virtual address space of a second process. The mapping enables the first process and the second process to share data through the shared virtual memory region. Membership information associated with the data pipeline is updated to include the first process and the second process. An access request for accessing the shared virtual memory region is received from the first process, and the access request is granted or denied based on one or more protection policies.

Data pipeline for microkernel operating system

In one embodiment, a method includes receiving, by a kernel of an operating system executing on a computing device, a request to store a message to communicate from a first process to a second process using a circular buffer. The circular buffer includes a first number of memory segments. An ownership of a first memory segment of the circular buffer is determined, based on a corresponding first ownership segment of an ownership array for the circular buffer. The ownership array includes a second number of ownership segments, where the second number is the same as the first number. Responsive to determining that the first segment is available to the first process, the message is stored in the first memory segment. After the message is stored in the first memory segment, the first ownership segment is changed to indicate the first memory segment is owned by the second process.

Circular queue for microkernel operating system

In one embodiment, a method includes allocating, by an operating system of a computing device, computer-implemented memory into a discardable portion and a durable portion; receiving, from a computer-executable program, a designation indicator of a particular data file that is associated with the computer-executable program indicating that the particular file is to be stored in the discardable portion of the memory and in response, storing the particular data file in one or more particular pages of the discardable portion of the memory; identifying an occurrence of a computing condition and in response, marking the one or more particular pages that include the particular data file as invalid for the computer-executable program; receiving, from the computer-executable program, a request for the particular data file; and in response to receiving the request, providing, to the computer-executable program, a notification that the particular data file is invalid for the computer-executable program.

Memory management of computing devices

The present disclosure provides techniques for managing a cache of a computer system using a cache management data structure. The cache management data structure includes a cold queue, a ghost queue, and a hot queue. The techniques herein improve the functioning of the computer because management of the cache management data structure can be performed in parallel with multiple cores or multiple processors, because a sequential scan will only pollute (i.e., add unimportant memory pages) cold queue, and to an extent, ghost queue, but not hot queue, and also because the cache management data structure has lower memory requirements and lower CPU overhead on cache hit than some prior art algorithms.

CPU-efficient cache replacment with two-phase eviction

In one embodiment, a method includes storing thread state information associated with the first user-mode thread into a memory space associated with the first user-mode thread when executing in kernel space in response to a first system call by a first-user mode thread, executing first operations corresponding to the first system call on a processor, where data associated with executing the first operations are stored on a kernel stack associated with the processor, determining to pause the execution of the first operations, enqueuing a workload on a schedule queue for resuming execution of the first operations in a future, where the workload comprises data associated with executing a remainder of the first operations, and executing second operations corresponding to a second system call from a second user-mode thread on the processor, where data associated with executing the second operations are stored on the kernel stack associated with the processor.

Operating system with a single kernel stack per processor

The disclosure provides an approach for dynamically reprogramming network and network infrastructure in response to VM mobility. The approach provides a hypervisor layer that can observe changes in VM-host relationships and reprogram the associated network and network infrastructure to maintain network communication. The hypervisor layer notifies a controller of a data center to migrate an ENI of the migrated VM to the same destination host as the migrated VM, in response to VM migration.

Reprogramming network infrastructure in response to VM mobility

The disclosure provides an approach for dynamically reprogramming network and network infrastructure in response to VM mobility. The approach provides a hypervisor layer that can observe changes in VM-host relationships and reprogram the associated network and network infrastructure to maintain network communication. The hypervisor layer notifies an elastic network interface of a new IP address to include within its whitelist in response to VM migration to that elastic network interface.

In one embodiment, a method includes creating, by an operating system executed by a computing device an inter process communication (IPC) channel and a port for a process executed in a user space of the operating system. The IPC channel is associated with a key, and the port comprises a port buffer mapped to a first virtual address space of a kernel of the operating system and to a second virtual address space of the process. A message for the process is written in a message buffer associated with the IPC channel. The kernel determines whether the process is actively consuming messages in the message buffer. Responsive to determining that the process is not actively consuming messages, a notification packet is written in the port buffer. The packet includes an action type and the key and causes the process to consume the message based on the action type and key.

Port configuration for microkernel operating system

The disclosure provides an approach for performing an operation by a first process on behalf of a second process, the method comprising: obtaining, by the first process, a memory handle from the second process, wherein the memory handle allows access, by the first process, to at least some of the address space of the second process; dividing the address space of the memory handle into a plurality of sections; receiving, by the first process, a request from the second process to perform an operation; determining, by the first process, a section of the plurality of sections that is to be mapped from the address space of the memory handle to the address space of the first process for the performance of the operation by the first process; mapping the section from the address space of the memory handle to the address space of the first process; and performing the operation by the first process on behalf of the second process.

System and methods of zero-copy data path among user level processes

The approaches described herein implement synchronous execution of a user space operation from a kernel context. A thread, executing on a computing device, initializes a second kernel stack based on a first kernel stack. The computing device executes an operating system having a user space and a kernel space. The thread, executing in kernel space, performs a non-blocking call (e.g., an upcall) to execute an upcall function in user space. The upcall function may further call other user space functions or system calls. The system calls are performed using the second kernel stack. Upon termination of the upcall function, the thread continues execution on the first kernel stack.

Synchronous user space function execution from a kernel context

Exemplary methods, apparatuses, and systems include a first input/output (I/O) filter receiving, from a first filter module within a virtualization stack of a host computer, an input/output (I/O) request originated by a virtual machine and directed to a first virtual disk. The first I/O filter determines to redirect the I/O request to a second virtual disk and, in response, forwards the I/O request to a second I/O filter associated with the second virtual disk. The first I/O filter is a part of a first instance of a filter framework within the host computer and the second I/O filter is part of a second, separate instance of the filter framework.

Input/output filter configuration of linked storage

Techniques for enabling filter-level access to a virtual disk (VMDK) are provided. In one set of embodiments, an application can invoke a first application programming interface (API) for opening the VMDK, the invoking of the first API causing an ordered group of filters associated with the VMDK to be instantiated. The application can further coordinate with a target filter in the ordered group of filters to establish a communication channel with the target filter and can receive, from the target filter, a handle to the target filter via the communication channel. The application can then issue an I/O request to the VMDK via the target filter using the handle, the issuing causing data associated with the I/O request to be filtered by other filters that are downstream from the target filter in the ordered group.

Growing community of inventors

Snoqualmie, WA, United States of America

Christoph Klee