Table of Contents
- Due Dates and Mark Distribution
- Setting Up Your Assignment
- Concurrent Programming with OS/161
- Tutorial Exercises
- Code reading
- Coding the Assignment
Due Date & Time: 4pm (16:00), March 11 (Week 4)
Marks: Worth 30 marks (of the class mark component of the course)
The 2% per day bonus for each day early applies, capped at 10%, as per course outline.
In this assignment you will solve a number of synchronisation problems within the software environment of the OS/161 kernel. By the end of this assignment you will gain the skills required to write concurrent code within the OS/161 kernel. While the synchronisation problems themselves are only indirectly related to the services that OS/161 provides, they solve similar concurrency problems that you would encounter when writing OS code.
The Week 3 tutorial contains various synchronisation familiarisation exercises. Please prepare for it. Additionally, feel free to ask any assignment related questions in the tutorial.
We assume after ASST0 that you now have some familiarity with setting up for OS/161 development. The following is a brief setup guide. If you need more detail, refer back to ASST0.
Clone the ASST1 source repository from gitlab.cse.unsw.edu.au.
% cd ~/cs3231 % git clone https://zXXXXXXX@gitlab.cse.unsw.edu.au/COMP3231/22T1/zXXXXXXX-asst1.git asst1-src
Configure your new sources as follows.
% cd ~/cs3231/asst1-src % ./configure && bmake && bmake install
We have provided you with a framework to run your solutions for ASST1. This framework consists of tester code (found in kern/asst1) and menu items you can use to execute the code and your solutions from the OS/161 kernel boot menu.
You have to configure your kernel itself before you can use this framework. The procedure for configuring a kernel is the same as in ASST0, except you will use the ASST1 configuration file:
% cd ~/cs3231/asst1-src/kern/conf % ./config ASST1
You should now see an ASST1 directory in the kern/compile directory.
When you built OS/161 for ASST0, you ran bmake in compile/ASST0. In ASST1, you run bmake from (you guessed it) compile/ASST1.
% cd ../compile/ASST1 % bmake depend % bmake % bmake install
If you are told that the compile/ASST1 directory does not exist, make sure you ran config for ASST1.
Tip: Once you start modifying the OS/161 kernel, you can quickly rebuild and re-install with the following command sequence. It will install the kernel if the build succeeds.
% bmake && bmake install
The sys161.conf should be already be installed in the ~/cs3231/root directory from assignment 0. If not, follow the instructions below to obtain another copy. A pre-configured sys161 configuration is available here: sys161.conf.
% cd ~/cs3231/root % wget http://cgi.cse.unsw.edu.au/~cs3231/22T1/assignments/asst1/sys161.conf
Run the previously built kernel:
% cd ~/cs3231/root % sys161 kernel sys161: System/161 release 2.0.8, compiled Feb 25 2019 09:34:40 OS/161 base system version 2.0.3 (with locks&CVs solution) Copyright (c) 2000, 2001-2005, 2008-2011, 2013, 2014 President and Fellows of Harvard College. All rights reserved. Put-your-group-name-here's system version 0 (ASST1 #1) 16220k physical memory available Device probe... lamebus0 (system main bus) emu0 at lamebus0 ltrace0 at lamebus0 ltimer0 at lamebus0 beep0 at ltimer0 rtclock0 at ltimer0 lrandom0 at lamebus0 random0 at lrandom0 lser0 at lamebus0 con0 at lser0 cpu0: MIPS/161 (System/161 2.x) features 0x0 OS/161 kernel [? for menu]:
If your code is properly synchronised, the timing of context switches, the location of kprintf() calls, and the order in which threads run should not influence the correctness of your solution. Of course, your threads may print messages in different orders, but you should be able to verify that they implement the functionality required and that they do not deadlock.
thread_yield() is automatically called for you at intervals that vary randomly. thread_yield() context switches between threads via the scheduler to provide multi-threading in the OS/161 kernel. While the randomness is fairly close to reality, it complicates the process of debugging your concurrent programs.
The random number generator used to vary the time between these thread_yield() calls uses the same seed as the random device in System/161. This means that you can reproduce a specific execution sequence by using a fixed seed for the random number generator. You can pass an explicit seed into the random device by editing the "random" line in your sys161.conf file. For example, to set the seed to 1, you would edit the line to look like:
28 random seed=1
We recommend that while you are writing and debugging your solutions you start the kernel via command line arguments and pick a seed and use it consistently. Once you are confident that your threads do what they are supposed to do, set the random device to autoseed. This should allow you to test your solutions under varying timing that may expose scenarios that you had not anticipated.
To reproduce your test cases, you need to run your tests via the command line arguments to sys161 as described above, otherwise system behaviour will depend on your precise typing speed (and not be reproducible for debugging).
The aim of the week 3 tutorial is to have you implement synchronised data structures using the supplied OS synchronisation primitives. See the Week 03 Tutorial for details.
It is useful to be prepared to discuss both the questions and the following assignment in your tutorial.
The following questions aim to guide you through OS/161's implementation of threads and synchronisation primitives in the kernel itself for those interested in a deeper understanding of OS/161. A deeper understanding can be useful when debugging, but is not strictly required, though recommended especially for Extended OS students.
For those interested in gaining a deeper understanding of how synchronisation primitives are implemented, it is helpful to understand the operation of the threading system in OS/161. After which, walking through the implementation of the synchronisation primitives themselves should be relatively straightforward.
- What happens to a thread when it exits (i.e., calls thread_exit())? What about when it sleeps?
- What function(s) handle(s) a context switch?
- How many thread states are there? What are they?
- What does it mean to turn interrupts off? How is this accomplished? Why is it important to turn off interrupts in the thread subsystem code?
- What happens when a thread wakes up another thread? How does a sleeping thread get to run again?
- What function is responsible for choosing the next thread to run?
- How does that function pick the next thread?
- What role does the hardware timer play in scheduling? What hardware independent function is called on a timer interrupt?
- What is a wait channel? Describe how wchan_sleep() and wchan_wakeone() are used to implement semaphores.
- Why does the lock API in OS/161 provide lock_do_i_hold(), but not lock_get_holder()?
We know: you've been itching to get to the coding. Well, you've finally arrived!
This is the assessable component of this assignment.
The following problems will give you the opportunity to write some fairly straightforward concurrent systems and get a practical understanding of how to use concurrency mechanisms to solve problems. We have provided you with basic tester code that starts a predefined number of threads that execute a predefined activity (in the form of calling functions that you must implement or modify).
Note: In this assignment, you are restricted to the lock, semaphore, and condition variable primitives provided in OS/161. The use of other primitives such as thread_yield(), spinlocks, interrupt disabling (spl), atomic instructions, and the like are prohibited. Moreover, they usually result in a poor solution involving busy waiting.
Note: In some instances, the comments within the code also form part of the specification and give guidance as to what is required. Make sure you read the provided code carefully.
Check that you have specified a seed to use in the random number generator by examining your sys161.conf file, and run your tests using System/161 command line arguments. It is much easier to debug initial problems when the sequence of execution and context switches are reproducible.
When you configure your kernel for ASST1, the tester code and extra menu options for executing the problems (and your solutions) are automatically compiled in.
For the first problem, we ask you to solve a mutual exclusion problem. The code in kern/asst1/counter.c is an incomplete implementation of an interface specified in kern/asst1/counter.h. The interface specifies functions to initialise (counter_initialise()), increment (counter_increment()), decrement (counter_decrement()), and read and cleanup a synchronised counter (counter_read_and_destroy()). All functions can be called concurrently by multiple threads and are unsynchronised.
The testing code provided in kern/asst1/counter_tester.c exercises a subset of the counter code and produces an incorrect result similar to the following. Note that the final count of the incomplete implementation is dependent on scheduling and hence will vary.
OS/161 kernel [? for menu]: 1a Starting 10 incrementer threads The final count value was 5083 (expected 10000)
Your task is to modify kern/asst1/counter.c and kern/asst1/counter.h by synchronising the code appropriately such that synchronised counters can be created, destroyed, incremented and decremented correctly.
You can assume counter_read_and_destroy() is called after a call to counter_initialise() with a valid pointer originally provided by counter_initialise(). For each valid pointer provided by counter_initialise(), counter_read_and_destroy() will only be called once. counter_increment() and counter_decrement() are only ever called (multiple times) with a valid pointer originally provided by counter_initialise() and before the final call to counter_read_and_destroy().
To test your solution, use the 1a menu choice. Sample output from a correct solution in included below.
OS/161 kernel [? for menu]: 1a Starting 10 incrementer threads The final count value was 10000 (expected 10000)
When we mark your assignment, we will replace the testing code provided in kern/asst1/counter_tester.c to test your implementation more extensively than the provided code.
This task involves modifying an example such that the example no longer deadlocks and is able to finish. The example is in twolocks.c.
In the example, bill(), bruce(), bob() and ben() are threads that need to hold one or two locks at various times to make progress: lock_a and lock_b. While holding one or two locks, the threads call holds_lockX that just consumes some CPU. The way the current code is written, the code deadlocks and triggers OS/161's deadlock detection code, as shown below.
OS/161 kernel: 1b Locking frenzy starting up Hi, I'm Bill Hi, I'm Ben Hi, I'm Bruce Hi, I'm Bob hangman: Detected lock cycle! hangman: in ben thread (0x80031ed8); hangman: waiting for lock_a (0x80032d04), but: lockable lock_a (0x80032d04) held by actor bill thread (0x80031f58) waiting for lockable lock_b (0x80032cc4) held by actor ben thread (0x80031ed8) panic: Deadlock. sys161: trace: software-requested debugger stop sys161: Waiting for debugger connection...
You task is to modify the existing code such that:
- you apply resource-ordering deadlock prevention such that the code no longer deadlocks, and runs to completion as shown below (the ordering may vary);
- the modified solution still calls the holds_lockX functions in the same places, and only the locks indicated are held by the thread at that point in the code;
- your deadlock free solution only uses the existing locks and calls them the same number of times; and
- you document the overall resource order chosen in the comment indicated in the code.
OS/161 kernel: 1b Locking frenzy starting up Hi, I'm Bill Hi, I'm Bruce Hi, I'm Ben Hi, I'm Bob Bruce says 'bye' Bob says 'bye' Ben says 'bye' Bill says 'bye' Locking frenzy finished
Your next task in this part is to synchronise a solution to a producer/consumer problem. In this producer/consumer problem, one or more producer threads allocate data structures, and call producer_send(), which copies pointers to the data structures into a fixed-sized buffer, while one or more consumer threads retrieve those pointers using consumer_receive(), and inspect and de-allocate the data structures.
The code in kern/asst1/producerconsumer_tester.c starts up a number of producer and consumer threads. The producer threads attempt to send pointers to the consumer threads by calling the producer_send() function with a pointer to the data structure as an argument. In turn, the consumer threads attempt to receive pointers to the data structure from the producer threads by calling consumer_receive(). These functions are currently partially implemented. Your job is to synchronise them.
Here's what you might see before you have implemented any code:
OS/161 kernel [? for menu]: 1c run_producerconsumer: starting up Waiting for producer threads to exit... Consumer started Producer started Consumer started Producer finished Consumer started Producer started *** Error! Unexpected data -2147287680 and -2147287680 Consumer started *** Error! Unexpected data -2147287712 and -2147287712 Consumer started *** Error! Unexpected data -2147287648 and -2147287648 *** Error! Unexpected data -2147287712 and -2147287712 *** Error! Unexpected data -2147287648 and -2147287648 *** Error! Unexpected data -2147287648 and -2147287648 *** Error! Unexpected data -2147287648 and -2147287648 *** Error! Unexpected data -2147287712 and -2147287712 *** Error! Unexpected data -2147287664 and -2147287664 *** Error! Unexpected data -2147287664 and -2147287664 *** Error! Unexpected data -2147287600 and -2147287600 *** Error! Unexpected data -2147287600 and -2147287600 *** Error! Unexpected data -2147287664 and -2147287664 *** Error! Unexpected data -2147287600 and -2147287600 panic: Assertion failed: fl != fl->next, at ../../vm/kmalloc.c:1134 (subpage_kfree)
Note that code will panic (crash) in different ways depending on the timing.
And here's what you will see with a (possibly) correct solution:
OS/161 kernel: 1c run_producerconsumer: starting up Consumer started Waiting for producer threads to exit... Producer started Consumer started Consumer started Producer started Consumer started Consumer started Producer finished Producer finished All producer threads have exited. Consumer finished normally Consumer finished normally Consumer finished normally Consumer finished normally Consumer finished normally
- producerconsumer_tester.c: Starts the producer/consumer simulation by creating producer and consumer threads that will call producer_send() and consumer_receive(). You are welcome to modify this simulation when testing your implementation — in fact, you are encouraged to — but remember that it will be overwritten when we test your solution is tested, so you can't rely on any changes you make in this file.
- producerconsumer.h: Contains prototypes for the functions in producerconsumer.c, as well as the description of the data structure that is passed from producer to consumer (the uninterestingly-named data_item_t). This file will also be overwritten when your solution is tested by us.
- producerconsumer.c: Contains the implementation of producer_send() and consumer_receive(). It also contains the functions producerconsumer_startup() and producerconsumer_shutdown(), which you can implement to initialise any variables and any synchronisation primitives you may need.
The provided data structure represents a bounded buffer capable that is capable of holding BUFFER_ITEMS data_item_t pointers. This means that calling producer_send() BUFFER_ITEMS times should not block (or overwrite existing items, of course), but calling producer_send() one more time should block, until an item has been removed from the buffer using consumer_receive(). We have provided an unsynchronised skeleton of circular buffer code, though you will have to use appropriate synchronisation primitives to ensure that concurrent access is handled safely.
The data structure should function as a circular buffer with first-in, first-out semantics.
This part simulates a simple cafe with customer threads and barista threads. The customers grab a ticket when they want to order. The tickets are basically a number assigned to customers when they want to order as typical in delicatessens, cafes, and take-away cafes. Ticket numbers are used once only with the next ticket being one more than the previous ticket.
After receiving a ticket, the customer waits for their ticket's turn (by blocking on a synchronisation primitive), until a barista announces their ticket number thus waking them up so they can then order.
Baristas continuously serve ticket numbers in ascending order. The ticket number a particular barista selects is the one that they are responsible for announcing and serving. They block while there are still customers in the cafe and future orders to serve, but no customer is waiting on the current number yet.
Note: The simulation is a little unrealistic in several ways.
- Customers repeatedly order coffees by obtaining more tickets until they are seriously over caffeinated.
- Baristas serve a fixed number orders in a session based on the number of baristas, customers, and repeat orders configured for a particular simulation.
- The ordering, waiting, making coffee, and drinking coffee part of the real world is not simulated.
The code that drives the system is in cafe_tester.c. You should review the code to develop an understanding of the system. You'll see it starts a number of customer and barista threads and then waits for the customers to issue all their requests and leave. The code also waits for the baristas exit after the orders have been made and the customers all leave.
The functions of particular interest are customer_thread and barista_thread which document and show the behaviour of customers and baristas.
The customer and barista threads interact with each other via the functions provided in cafe.c. At a high level, these functions manage ticket allocation, customer waiting, the number a barista is server, and barista waiting. A more detailed specification of the each function is provided in the comments in the code itself and form part of the assignment specification.
Your task is to implement the functions such that the cafe system will execute correctly with all the tickets processed.
- Your solution should not busy-wait when a thread can't make progress.
- You should not rely on any changes to code in the cafe_tester.c or cafe.h files. They will be changed for testing purposes after your final submission. You can vary the code for your own testing purposes, but we'll replace them for our own testing.
A sample of how the code can fail follows. Notice the baristas serving customers without the correct ticket, and serving customers that have left. The system is not synchronised.
cpu0: MIPS/161 (System/161 2.x) features 0x0 OS/161 kernel: 1d run_cafe: starting up Waiting for barista threads to exit... Barista thread 0 started Customer thread 2 started Customer 2 got ticket 1 Customer 2 ordered from barista 0 Barista thread 1 started Customer thread 1 started Customer thread 0 started Barista 0 serving ticket 1 Barista 0 accepted customer 0 with ticket 1 Barista 1 serving ticket 2 Customer 1 got ticket 3 Customer 0 got ticket 4 Customer 0 ordered from barista 0 Customer 0 got ticket 5 Customer 0 ordered from barista 0 Customer 1 ordered from barista 0 Customer 1 got ticket 6 Barista 0 serving ticket 3 Barista 0 accepted customer 0 with ticket 3 Customer 0 leaving cafe Customer 2 got ticket 2 Customer 2 ordered from barista 0 Customer 2 leaving cafe Barista 0 serving ticket 4 Barista 0 accepted customer 0 with ticket 4 Barista 1 accepted customer 0 with ticket 2 Customer 1 ordered from barista 0 Customer 1 leaving cafe Barista 1 serving ticket 5 Barista 1 accepted customer 0 with ticket 5 Barista 1 serving ticket 6 Barista 1 accepted customer 0 with ticket 6 All barista threads have exited. Operation took 3.006829880 seconds
A potentially correct solution generates output similar to that below.
Note: The interleaving of progress of individual threads can vary between each run of the simulation. However, baristas always accept orders from the customer holding the matching ticket, and tickets are unique and only used once.
cpu0: MIPS/161 (System/161 2.x) features 0x0 OS/161 kernel: 1d run_cafe: starting up Waiting for barista threads to exit... Barrista thread 0 started Customer thread 2 started Customer 2 got ticket 1 Barista 0 serving ticket 1 Barrista thread 1 started Customer thread 1 started Customer 1 got ticket 2 Barista 0 accepted customer 2 with ticket 1 Customer 2 ordered from barista 0 Barista 1 serving ticket 2 Customer thread 0 started Barista 0 serving ticket 3 Customer 2 got ticket 3 Barista 1 accepted customer 1 with ticket 2 Customer 1 ordered from barista 1 Customer 1 got ticket 5 Barista 0 accepted customer 2 with ticket 3 Customer 2 ordered from barista 0 Barista 1 serving ticket 4 Customer 0 got ticket 4 Barista 0 serving ticket 5 Customer 2 leaving cafe Barista 1 accepted customer 0 with ticket 4 Customer 0 ordered from barista 1 Barista 0 accepted customer 1 with ticket 5 Customer 1 ordered from barista 0 Customer 1 leaving cafe Customer 0 got ticket 6 Barista 1 serving ticket 6 Barista 1 accepted customer 0 with ticket 6 Customer 0 ordered from barista 1 Customer 0 leaving cafe All barista threads have exited. Operation took 3.062782800 seconds
- This problem is simpler to solve in OS/161 using locks and condition variables.
- Identify the shared state between threads in the system and ensure it is synchronised to avoid races.
- A barista needs to be able to find a customer with the sought after ticket. A new shared data structure needs to facilitate that. Observe customers have that information to share when they wait.
- A barista needs to wake a specific customer matching a ticket. Again, a data structure combined with one or per-customer condition variables might provide the desired functionality.
- For the customer, it helps to be able to identify what is the condition that requires cv_wait to be called. What is the condition that triggers a cv_signal in a barista.
- For the barista, it helps to be able to identify what is the condition that requires cv_wait to be called. What is the condition that triggers a cv_signal in a customer.
The submission instructions are available on the Wiki. Like ASST0, you will be submitting the git repository bundle via CSE's give system. For ASST1, the submission system will do a test build and run a simple test to confirm your bundle at least compiles. It does not exhaustively test you submission
Don't ignore the submission system! If your submission fails the simple test in the submission process, you may not receive any marks.
To submit your bundle:
% cd ~ % give cs3231 asst1 asst1.bundle
You're now done.
Even though the generated bundle should represent all the changes you have made to the supplied code, occasionally students do something "ingenious". So always keep your git repository so that you may recover your assignment should something go wrong. We recommend to git push it back to gitlab.cse.unsw.edu.au for safe keeping.