Most uses of synchronization code in multi-threaded applications fall into a small number of high-level “usage patterns”, or what can be called generic synchronization policies (GSPs). This paper illustrates how the use of such GSPs simplify the writing of thread-safe classes. In addition, this paper presents a C++ class library that implements commonly-used GSPs.
Very lightweight stackless threads; give linear code execution for event-driven systems, designed to use little memory; library is pure C, no platform-specific Assembly; usable with or without OS. Open source, BSD-type license.
Interprocess communication is an essential component of modern software engineering. Often, lock-free IPC is accomplished via special processor commands. This article propose a communication type that requires only atomic writing of processor word from processor cache into main memory and atomic processor word reading from main memory into the processor register or processor cache.
Explains why in the concurrent world, locality is a first-order issue that trumps most other performance considerations. Now locality is no longer just about fitting well into cache and RAM, but to avoid scalability busters by keeping tightly coupled data physically close together and separately used data far, far apart.
Presents a solution to races and deadlocks based on a well-known deadlock-avoidance protocol and shows how it can be enforced by the compiler. It can be applied to programs in which the number of locks is fixed and known up front.
So far multithreaded file I/O is a under-researched field. Although its simple to measure, there is not much common knowledge about it. The measurements presented here show that multithreading can improve performance of file access directly, as well as indirectly by utilizing available cores to process the data read.
Sharing requires waiting and overhead, and is a natural enemy of scalability. This article focuses on one important case, namely mutable (writable) shared objects in memory, which are an inherent bottleneck to scalability on multicore systems.
A thread pool hides a lot of details, but to use it effectively some awareness of some things a pool does under the covers is needed to avoid inadvertently hitting performance and correctness pitfalls.