In an SMP system, each processor has its own private cache. If Processor A changes the variable "X" in its cache, how does Processor B know that "X" has changed? Processor B might still be holding an old, "dirty" copy. This is the Cache Coherency Problem. Schimmel explains the hardware solutions (like the —Modified, Exclusive, Shared, Invalid) and the software workarounds, such as Memory Barriers .
The trade-offs between coarse-grained locking (easy to implement but causes CPU serialization) and fine-grained locking (highly scalable but prone to deadlocks and high overhead). 3. Kernel Structure and Reentrancy unix systems for modern architectures -1994- pdf
UNIX Systems for Modern Architectures provided the industry with a rigorous, systematic guide to solving these exact problems. Core Technical Themes of the Text In an SMP system, each processor has its own private cache
Modern architectures demanded sophisticated virtual memory handling. This is the Cache Coherency Problem
The book serves as a bridge between high-level UNIX system calls and the low-level reality of hardware execution. It is structured into three primary domains:
: You can find used copies from merchants like AmericanBookWarehouse or AbeBooks .
In 1994, the computing landscape was undergoing a massive seismic shift. The industry was rapidly transitioning from traditional, single-core uniprocessor systems to modern architectures, including Symmetric Multiprocessing (SMP) and Non-Uniform Memory Access (NUMA).