Download Now Free registration required
Code is code. In the entire application stack only a very small fraction of code is special in the sense that it requires the hardware to run in privileged mode. In theory all the rest can run in either the kernel or user domain, with the tradeoffs being well-known. Engineering an operating system core so that the same code can function both in user and kernel mode alike provides numerous benefits: testing and development, virtualization, stability and reusability of code in applications. In the current Unix style, code is divided to kernel code and userspace code. Although some limited code modules, such as byte order and address translation routines may be shared, functionality at large is not available in both domains.
- Format: PDF
- Size: 212.9 KB