1. Nachos Instructional OS CS 270, Tao Yang, Spring 2011

2. 11/9/20152 What is Nachos OS? Allow students to examine, modify and execute operating system software. A skeletal OS that supports kernel threads user-level processes Simulates MIPS instruction execution. Running on a virtual machine, executed as a single Unix process in a hosting OS. Over 9K lines of C++ code. Can understand its basic mechanisms by reading about 1- 2K lines of code.

3. System Layers Base Operating System (Linux for our class) Nachos kernel threads Thread 1Thread 2Thread N Nachos OS modules (Threads mgm, File System, Code execution/memory mapping, System calls/Interrupt) Simulated MIPS Machine (CPU, Memory, Disk, Console) User process

4. 11/9/20154 Steps to Install Nachos Obtain and install Nachos source code. Copy the source code from ~cs270t/nachosSept20.tar.gz Compile the source code using gmake Run threads demo under the threads subdirectory (just run kernel test threads). Run user program demo under the userprog subdirectory. http://www.cs.ucsb.edu/~cs270t/HW/warmu p.html http://www.cs.ucsb.edu/~cs270t/HW/warmu p.html

5. 11/9/20155 Nachos code directory machine --- Basic machine specification (MIPS simulator). threads --- threads management (HW1). userprog -- binary code execution and system calls (HW2). vm -- virtual memory (HW3). filesys -- file system (HW3) test -- binary test code network -- networking protocol bin -- utilities/tools (binary format conversion)

6. 11/9/20156 Source code reading and HW1 Objectives of next 2 weeks Scan through ~1,000-2,000 lines of code under threads directory Learn how context switch is accomplished among threads. Learn how thread scheduling is done. Learn how locks/synchronization are implemented and used. Complete HW1 Few hundred lines of code Sample solution for Task 1, 2, &3 are available

7. 11/9/20157 Single-threaded vs multithreaded Process

8. main () { Thread *t1 = new Thread("forked thread1"); Thread *t2 = new Thread("forked thread2"); t1->Fork(SimpleThread, 1); t2->Fork(SimpleThread, 2); SimpleThread(3); } SimpleThread(int i) { printf(“Hello %d\n”, i); currentThread->Yield(); } Sample Example of Nacho Threads Start to fork and execute a function in each child thread. Parent also executes the same function Function executed by threads Create 2 new threads.

9. 11/9/20159 Nachos Threads Nachos threads execute and share the same code, share same global variables. Nachos scheduler maintains a ready list, containing all threads that are ready to execute. Each thread is in one of four states: READY, RUNNING, BLOCKED, JUST_CREATED. Each thread object maintains a context block. Thread object supports the following operations: Thread(char *debugName). Create a thread. Fork(VoidFunctionPtr func, int arg). Let a thread execute a function. Yield(). Suspend the calling thread and select a new one for execution. Sleep(). Suspend the current thread, change its state to BLOCKED, and remove it from the ready list Finish()

10. Nachos Thread States and Transitions running (user) running (kernel) readyblocked Scheduler::Run Scheduler::ReadyToRun interrupt or exception Thread::Sleep Thread::Yield Machine::Run, ExceptionHandler When running in user mode, the thread executes within the machine simulator. HW 2 covers this. In HW 1 we are only concerned with the states in this box.

11. 11/9/201511 Thread Switching Switching involves suspending current thread, saving its state, and restoring the state of new thread. Following code involved in execution: the old code, the new code, and the code that performs switching. Switch(oldThread, newThread): Save all registers in oldThread's context block. Save the program address to be used when the old thread is resumed. Load new values into the registers from the context block of the new thread. Once the saved PC of the new thread is loaded, Switch() is no longer executing.

12. 11/9/201512 Scheduler object for thread scheduling A scheduler decides which thread to run next by scanning the ready list. The scheduler is invoked whenever the current thread gives up the CPU. The current Nachos scheduling policy is round-robin: new threads are appended to the end of the ready list, and the scheduler selects the front of the list. The Scheduler object has the following operations: ReadyToRun(Thread *thread). Make thread ready to run and place it on the ready list. Thread *FindNextToRun() Run(Thread *nextThread)

13. 11/9/201513 Semaphore object for thread synchronization Disable and re-enable interrupts to achieve mutual exclusion (e.g., by calling Interrupt::SetLevel()). Operations for a Semaphore object: Semaphore(char* debugName, int initialValue) P(): Decrement the semaphore's count, blocking the caller if the count is zero. V() :Increment the semaphore's count, releasing one thread if any are blocked waiting on the count.

14. 11/9/201514 Key steps when Nachos executes After you type ``nachos'' under threads subdirectory: It is executing as a single Unix process. The main() calls Initialize() to start up interrupt handling, create a scheduler for managing the ready queue. ThreadTest() (to be explained for HW 1). currentThread->Finish() to let other threads continue to run.

15. 11/9/201515 Key Calling graph when Nachos executes under thread directory All files are in threads directory. main() in main.cc Initialize() in system.cc ThreadTest () in threadtest.cc Thread:Fork () in thread.cc Thread:Yield () in thread.cc StackAllocate() in thread.cc FindNextToRun () in scheduler.cc ReadyToRun () in scheduler.cc Run () in scheduler.cc SWITCH () in switch.s ThreadRoot () in switch.s func() such as SimpleThread() in ThreadTest.cc currentThread->Finish () in threadtest.cc

16. 11/9/201516 ThreadRoot() Executed by SWITCH(oldThread, nextThread) jalStartupPC# call startup procedure. For a new thread, it is InterruptEnable(). movea0, InitialArg jalInitialPC# call main procedure jal WhenDonePC# when were done, call clean up procedure.

17. 11/9/201517 QA When will thread:Finish() be called? At the end of the forked thread. Check ThreadRoot assembly code. At the end of the main() thread. Thread:Finish() calls scheduler->run() to run a new thread. Will the old thread (that calls Thread:Finish()) be returned and continue to be executed? No. Because the following is called first. threadToBeDestroyed = currentThread; Scheduler->Run() will delete such TCB

18. 11/9/201518 QA In Sleep() or Yield(), the interrupt is turned off before calling scheduler->() to execute another thread. When will interrupt be turned on? In executing the switched thread, ThreadRoot() assembly code first executes StartupPC function which is machineState[StartupPCState] = (int) InterruptEnable;

19. 11/9/201519 HW 1: threads & synchronization Work under threads subdirectory. Modify ThreadTest() to do simple threads programming (spawning multiple threads). Implement locks and condition variables (missing from the file synch.cc). Workload: Read Nachos code and add few hundred lines of code. Undocumented sample solution is provided.

20. 11/9/201520 HW 1: Files involved Key files main.cc, threadtest.cc -- a simple test of our thread routines. thread.h thread.cc -- Nachos thread data structure and operations. scheduler.h scheduler.cc -- The thread ready list. synch.h synch.cc -- synchronization routines. Other related files. synchlist.h, synchlist.cc -- synchronized access to lists using locks/conditions (useful examples for your programming). list.h list.cc -- generic list management. system.h, system.cc -- Nachos startup/shutdown routines. utility.h utility.cc -- some useful definitions and debugging routines. interrupt.h interrupt.cc -- manage interrupts. time.h timer.cc -- clock emulation. switch.h, switch.s -- assembly code for thread switching. stats.h stats.cc -- collect interesting statistics.

21. 11/9/201521 HW Sample solution ~cs240t/sampleSolutionCode.tar.gz Has an old solution for HW1, HW2, HW3 HW1  threads subdirectory. ~400 lines of new code. ~50% are for Tasks 1/2/3. Code for task 4 is not useful. HW2 -> userprog subdirectory. ~1300 lines of new code. HW3 -> vm and filesys. ~1200 lines of new code. 800 may be enough. Caveat: Mixed benefits/problems in using other students’ code. e.g. Not well documented, not fully tested. Possibly awkward design. Still your responsibility to produce good solutions (correctness, performance, style).