Systems and methods for dynamic low latency optimization

Abstract

The present invention provides low-latency optimized systems and methods configured to execute from the hardware layer of the entire operating system to application programs. Low-latency operation implemented according to embodiments is optimized for a specific application, which interfaces with specific hardware, executes on a host processor-based system, and is configured for low-latency optimization. For example, a low-latency optimized implementation may include various modules implemented in user space and kernel space, where these modules cooperate to obtain information about the services and hardware used by the application and provide this information for low-latency operation of the application . When running according to an embodiment, low-latency execution is dynamically enabled or disabled by a low-latency optimization implementation, on a per-application basis, when appropriate or as needed on an application.

Description

technical field [0001] The present invention relates to processor-based system operation, and more particularly to dynamic low-latency optimization for processor-based system operation. Background technique [0002] The use of various forms of processor-based platforms has exploded in recent years. For example, the use of processor-based personal communication and data processing platforms, commonly referred to as smartphones, has grown dramatically over the past few years and is expected to continue growing for the foreseeable future. For example, in addition to being used to provide cellular telephone communication services, smartphones are commonly used to provide data communication, data processing, viewing and / or creating streaming media, gaming and entertainment, and the like. Applications that can provide the aforementioned functionality range from those that use less processing power and other resources (e.g., thin clients and simple web browser applications) to tho...

AI Technical Summary

Problems solved by technology

[0007] Similarly, user-space low-latency techniques are not optimized for any particular application, are implemented statically, and may or may not have any significant impact on the latency experienced by the application, depending on the resources used by the application. influences

For example, improving runtime performance by using optimized compilation options (for example, using the "-O3" compilation option of the GNU compiler toolchain) is applicable during compilation, but cannot be reverted to the original option during runtime.

In addition, existing user-space low-latency techniques are not tightly coupled with existing kernel-space low-latency techniques, so instead of providing low-latency optimizations for a combination of operating systems and applications, they try to address the operating system separately and independently and the latency of one of the applications

For example, a command line to be able to start / stop system services may be available to developers, but these command line tools are usually not accessible to application developers

Also, existing user-space low-latency techniques focus on application-related tasks rather than the entire data path

In addition, existing user-space low-latency techniques do not directly access processor-based system hardware or its information, but instead interface through a hardware abstraction layer (HAL) that blocks access to the understanding of specific hardware

Images