Weenix Operating System Kernel

A comprehensive, Unix-like operating system kernel built from the ground up as part of the CSCI 402: Operating Systems course at the University of Southern California.

Note: Per university policy, the source code for this project is hosted in a private repository. This repository serves as a portfolio to demonstrate the architectural design and features implemented during the MSCS program.

---

🚀 Project Overview

Weenix is a teaching operating system designed to provide hands-on experience with the internal workings of a Unix kernel. Over a single semester, I implemented several critical subsystems, culminating in a kernel capable of booting into a user-space shell and executing complex concurrent processes.

🛠 Subsystems Implemented

1. Threads and Processes (PROCS)

Implemented the core execution model for a uniprocessor system.

• Thread Lifecycle: Managed thread creation, execution, and termination.
• Scheduler: Built a non-preemptive FIFO scheduler.
• Synchronization: Developed Mutexes, Condition Variables, and Wait Queues to manage concurrency.
• Context Switching: Handled low-level state saving and restoring of CPU registers.

2. Virtual File System (VFS)

Developed an abstraction layer that allows the kernel to interact with different file system implementations transparently.

• System Calls: Implemented open(), read(), write(), close(), lseek(), stat(), and mkdir().
• Abstractions: Managed vnode and vfs structures, and per-process file descriptor tables.

3. S5 File System (S5FS)

Built a simplified version of the System V File System to handle persistent storage.

• Disk Management: Implemented Inode management, data block allocation, and directory entry traversal.
• Consistency: Managed the free-list and ensured proper synchronization during disk I/O.

4. Virtual Memory (VM)

The final and most complex phase, enabling user-space memory management.

• Address Spaces: Managed memory mappings using anonymous and file-backed objects.
• Demand Paging: Implemented a page fault handler to load data only when accessed.
• Copy-on-Write (COW): Optimized memory usage for the fork() system call.
• Process Pipeline: Implemented fork(), exec(), and brk(), allowing the kernel to boot into a functional Unix-like shell.

---

🔍 Verification & Testing

The stability of the kernel was verified through rigorous stress tests:

• faber_thread_test: Validated thread synchronization and scheduling.
• vfstest: Confirmed VFS abstraction and file system integrity.
• memtest: Verified virtual memory allocation and page fault handling.
• forkbomb: Stress-tested resource limits and process cleanup under heavy load.

---

💻 Technical Stack

• Language: C
• Tools: GDB (Kernel-level debugging), QEMU, GCC, Make
• Concepts: Kernel-mode vs User-mode, Interrupt Handling, Paging, Deadlock Avoidance.

---

🔗 References & Credits

• Project Source: The Weenix Project (Originally developed at Brown University).
• Instructor: Prof. William (Bill) Cheng
• Course: CSCI 402: Operating Systems, University of Southern California