$24
In this project, you will work on two programs that help you (1) to be familiar with the concepts of process management and system calls; (2) to understand the functionalities and internals of the shell; and (3) to gain hand-on experience of writing system programs.
Task A: Concurrent Prime Sieve (30 Points)
<code>Write a concurrent version of prime sieve using pipes. The program shall run on a Linux machine
in the School of Computing computer labs. </code>
Overview
Your job for this task is to write a concurrent version of prime sieve program using pipes. In mathematics, the Sieve of Eratosthenes is an ancient algorithm for finding all prime numbers up to any given limit. A classical Sieve program finds all prime numbers up to a given integer *N*
by repeatedly dropping the multiples of newly founded prime numbers. If you are unfamiliar with the algorithm, you can read its wikipage at https://en.wikipedia.org/wiki/Sieve_of_Eratosthenes.
The idea of conncurent prime sieve is due to Doug McIlroy, inventor of Unix pipes. The following figure and text borrowed from the page (https://swtch.com/~rsc/thread/) created by Russ Cox, one of the XV6 authors.
A generating process can feed the numbers 2, 3, 4, ..., N into the left end of the pipeline: the first process in the line eliminates the multiples of 2, the second eliminates the multiples of 3, the third eliminates the multiples of 5, and so on:
![Illustration of Concurrent Sieve Using Pipes](./sieve.gif)
In your program, you are required to use pipe and fork to set up the pipeline. The length of the pipeline depends on N, which is passed into the program as its first argument.
The first process feeds the numbers 2 through N into the pipeline.
For each prime number, you will arrange to create one process that reads from its left neighbor over a pipe and writes to its right neighbor over another pipe.
Specification
1. You provide a program `primes.c` and put it into the folder `TaskA`.
1. Invoking the command `make` shall create a executable binary `primes`, which takes a single positive integer *N* which is larger than 2.
1. You can assume the test will use a reasonable *N* such that your program will complete successfully on a system with a limited number of file descriptors and processes.
1. The program `primes` shall output all primes from 2 through *N* with the following format:
```
pid=<pid of of the process that drops of multiples of 2> prime 2
pid=<pid of of the process that drops of multiples of 3> prime 3
...
pid=<pid of of the process that drops of multiples of p> prime p
```
Here *p* is the largest prime number smaller than or equal to *N*.
1. The program must stop after it output the last prime number.
Example Output
Your program shall generate output like the following examples except that
your program may output different PIDs for the processes.
```
$ make
gcc -o primes primes.c
$ ./primes 6
pid=70819 prime 2
pid=70820 prime 3
pid=70821 prime 5
$ ./primes 30
pid=70829 prime 2
pid=70830 prime 3
pid=70831 prime 5
pid=70832 prime 7
pid=70833 prime 11
pid=70834 prime 13
pid=70835 prime 17
pid=70836 prime 19
pid=70837 prime 23
pid=70838 prime 29
```
Hints
+ Because the program could not know the length of
the pipelines, the program may need some kind of **recursion**
and must decide when a process should fork a new process
to be its right neighbor. The body of the recursion has the following logic. You shall also build pipe and properly exit and wait.
```
prime_sieve()
...
p = get a number from left neighbor
print p
loop:
n = get a number from left neighbor
if (p does not divide n)
if ( not has_right_neighbor)
create right neighbor process
if (right neighbor process is newly created)
prime_sieve()
send n to right neighbor
```
+ Each process must determine when there wil be no more data from
its left neighbor.
+ You can use the first process to generate all the integers
from 2 through N and there is no need to
store them in an array.
+ You can read and write an integer in binary format as follows:
<pre>
read(fd, &num, sizeof(num));
write(fd, &num, sizeof(num));
// OR when num is of type int
read(fd, &num, sizeof(int));
write(fd, &num, sizeof(int));
</pre>
Task B: Simple xv6 Shell (70 Points)
<code>Write a simple shell for XV6. The program shall run on the xv6-riscv OS in the QEMU environment
in the School of Computing computer labs. </code>
Overview
Your job for this task is to write a simple shell for xv6. A shell is a a command line interpreter (CLI) that accepts user commands
and then executes each command in one or more separate processes. Your shell shall be able to:
1. run commands with arguments
1. handle input and output redirection
1. set up two-command pipelines.
1. handle internal commands including **cd** and **exit**
1. read commands from a text file and execute the commands in the same way as they are read from the console.
Specification
Specific to xv6
1. You build the shell program for xv6, which has fewer systems calls than mainstream Linux distributions. Also, the
system calls on XV6 may have slightly different syntax as their Linux peers.
1. You are **NOT allowed to use malloc()** in your `tsh` implementation. Allocating memory without freeing them
properly can lead to memory leak or corrupted data, causing the shell to crash.
1. Therefore, all the data structures in your shell implementations are either statically allocated or on the stack.
1. The `tsh` uses the limits defined in the header file `tsh.h` as a reasonable constraint to the command lines entered
by a user. Use these Macros instead of numbers in your program to make the program use to read and maintain.
<pre>
define TSH_MAX_CMD_LINE_LENGTH 255 The maximum number of characters in a command line
define TSH_MAX_NUM_ARGUMENTS 5 The maximum number of arguments in a simple command
<!-- define TSH_MAX_CMD_LIST_LENGTH 5 The maximum number of simple commands in a command list -->
define TSH_MAX_PIPELINE_LENGTH 5 The maximum number of simple commands in a pipelined command
define TSH_MAX_FILENAME_LENGTH 63 The maximum number of characters in a filename
</pre>
Basic Shell
1. Your basic shell, called `tsh` (Tiger Shell), is basically an interactive loop: it repeatedly prints a
prompt `tsh> ` (note the space after the greater-than sign), parses the input, executes
the command specified on a line of input, and waits for the command to finish. This process is
repeated until the user types the `exit` command.
1. You should put your shell source in user/tsh.c, and modify the Makefile to compile it.
1. An xv6 session with your shell might look like this:
```
xv6 kernel is booting
$ tsh
tsh> echo hello
tsh> exit
$
```
1. You should structure your shell such that it creates a new process for each
external command (an external command that is implemented as a separate program.
In contrast, a built-in command must be executed in the same process as your
current shell session).
1. Your can assume that given an external command, either there is a program in current
directory (root directory) or there is no such program. For the first case,
execute the command as the shell normally does. For the second case, prints an error message
"An error has occurred" and then proceeds to next iteration.
Note that a normal shell like *Bash* will search the executable in a list of directories
specified in the path variable and execute the first executable found. We make this
simplification because we are using a primitive version of the xv6 and hasn't studied
file system yet. For example,
```
tsh> grep Copyright README
Copyright 2006-2019 Frans Kaashoek, Robert Morris, and Russ Cox.
tsh> unimplemented_command
An error has occurred
tsh>
```
Built-in Commands
1. Whenever your shell accepts a command, it should check whether the command is a built-in
command or an external command. If the command is a built-in command, your shell shall
invoke the function that implements the command. For example, when shell sees the command **cd destpath**, it invokes the function chdir("destpath"). Note that shell doesn't fork a new process for built-in commands such as **cd**.
1. While most Unix shells have numerous built-in commands, in this project, you shall implement
**exit** and **cd**. The syntax and semantics of the
three commands are described as follows.
1. **exit [n]**
* Cause the shell to exit with a status of n.
* If n is omitted, the exit status is that of the last command executed.
1. **cd [dir]**
* Change the current directory to dir.
* If dir is not supplied, the value of the HOME shell variable is the default.
* If dir does not exit, prints an error message "An error has occurred" and then
proceeds to next iteration.
<!--
1. **pwd**
* Print the absolute pathname of the current working directory.
1. **set [var=value]**
* Assign `value` to the shell variable `var`.
* If the set command has no argument, print all shell variables at the time the command is executed.
* To simplify the implementation, you can assume:
1. no more than five variables will be defined;
1. both the name and value of a variable are string type;
1. a variable name has less than 16 characters; and
1. the value of a variable has less than 64 characters.
-->
Commands
1. You can assume all commands will be provided in a single command line.
1. Each command line has less than **TSH_MAX_CMD_LINE_LENGTH** characters.
1. Each command line has less than **TSH_MAX_PIPELINE_LENGTH** simple commands, connected
by '|'.
1. A simple command is a command that consists of a word followed by a list of arguments and
optional input and output redirections.
1. Each simple command has no more than **TSH_MAX_NUM_ARGUMENTS** arguments.
1. If a command line contains multiple simple commands separated by '|', then only the first simple
command can redirect its standard input and
only the last simple command can redirect its standard output.
I/O Redirection
Many times, a shell user prefers to send the output of a program to a file rather than
to the screen or to read the input to a program from a file rather than from the console.
This feature is called I/O redirection. You shell will implement a simplified version of I/O redirection with
the following syntax.
1. syntax: **command [argument ...] < input**
* Redirect standard input to file input
* If the file input does not exist, prints an error message "An error has occurred" and then proceeds
to next iteration of the shell's main loop.
1. syntax: **command [argument ...] > output**
* Redirect standard output to file output, new file mode.
* If the file output already exists, that file will be overwritten; otherwise it will be created.
<!--1. syntax: **command [argument ...] >> output**
* Redirect standard output to file output, append mode.
* If the file output already exists, the output from the command will be appended to that file;
otherwise it will be created.
1. syntax: **command [argument ...] 2> error**
* Redirect standard error to file error, new file mode.
* If the file error already exists, that file will be overwritten;
otherwise it will be created.
1. syntax: **command [argument ...] 2>> error**
* Redirect standard error to file output, appending mode.
* If the file error already exists, the error from command will be appended to that file;
otherwise it will be created.
-->
1. You don't need to implement I/O redirection for internal commands.
<!-- List commands
1. A list command is a list of simple commands separated by ';'. For example,
```
tsh> echo a > b; wc b; wc README
1 1 2 b
43 286 1982 README
tsh>
```
1. Given a list command, you shell should execute each simple command included in the list command in the order
from left to right.
-->
Pipeline commands
1. A pipeline command is a list of commands connected by '|'. For example,
```
tsh> cat README | grep Copy | wc
1 9 65
tsh>
```
1. For a pipeline command, a command reads input from its left neighbor and writes output to its right neighbor
when such neighbor exists. The leftmost process does not have a left neighbor and
the rightmost neighbor does not have a right neighbor.
1. In a pipeline command, only the leftmost command can redirect its standard input from a file.
1. In a pipeline command, only the rightmost command can redirect its standard output to a file.
1. Below are some examples.
```
tsh> cat < README | wc
43 286 1982
tsh> cat | wc < README
An error has occurred
tsh> cat > output | wc
An error has occurred
```
<!--
Batch mode execution
1. Your shell should be able to read the commands from a batch script file and execute the commands in sequence.
For example, assume the file `batch.sh` contains a command line `cat README | grep Copy | wc`, you can run the script and
the output will be as follows.
```
tsh> tsh batch.sh
1 9 65
tsh> exit
$ tsh batch.sh
1 9 65
$
```
1. A straightforward implementation of batch mode execution is to redirect the
standard input of the shell process to the batch script file, which i
s passed into the `tsh` program as an argument.
-->
Error Handling
Reporting clear and accurate error messages and handling program errors properly helps you quickly
locate the problem and fix it. In all programming projects, you should always try to develop
solid skills and habits to handle programming, user input, and runtime errors elegantly.
In this project, we have provided functions that you can use.
+ For user-related errors such as incorrectly formatted commands or long commands, use `void ErrorU(char *cause);`
+ For program-related errors such as the program doesn't handle special cases yet, use: `void ErrorP(char *cause);`
+ For system-related error such as system call failures, use `void ErrorS(char *cause);`
+ For program debug information, use `void Debug(char *fmt, ...);` The Debug() function accepts the same type of
arguments as the printf() function.
In your shell program, you should at least check and report the following errors:
+ An incorrect number of command line arguments to your shell.
+ A command does not exist or cannot be executed.
+ A very long command line (over **TSH_MAX_CMD_LINE_LENGTH** bytes).
+ A system call failed.
When your shell detects an error in a command entered by the user, it should print the error message and then
move to the next iteration of reading, parsing, and executing a new line of command.
The following scenarios should not be treated as errors:
+ A command line is empty.
+ There are multiple white spaces before, within, or after a command line.
Test
Test cases
You can use the following test cases during your development.
<pre>
test a simple command without argument
ls
test a simple command with arguments
echo hello there
test a simple command with I/O redirect
echo something > file.txt
<!-- test a list command
echo a > b; wc b; wc README
-->
test a 2-element pipeline command
ls | grep READ
test a 2-element pipeline command whose one or two sub commands have I/O redirection
grep lion < data.txt | wc > count
<!-- test a 3 element pipeline command
cat README | grep Copy | wc
-->
test tsh as an external command
echo echo hello | tsh
<!-- test tsh run in batch mode
echo echo hello > batch.sh
tsh batch.sh
-->
</pre>
Test program
The xv6 authors have provided an xv6 test program testsh, source in user/testsh.c. We have modified the program
for this project. You can use the program to verify if your shell can passes all the tests. The output may
be shown as follows:
<pre>
$ testsh tsh
simple echo: PASS
simple grep: PASS
two commands: PASS
output redirection: PASS
input redirection: PASS
both redirections: PASS
simple pipe: PASS
pipe and redirects: PASS
lots of commands: PASS
passed all tests
</pre>
Hints
+ You don't need to implement features that are not required by the tests; for example, you don't have to
implement parentheses or quoting.
+ `tsh` is much simpler than xv6's `sh` in terms of features, so you are likely best off writing tsh from scratch.
But, you can look at the xv6 shell source (in user/sh.c) for ideas, and borrow code, as long as you include
comments saying where you got the code.
+ To help you get started, we provided a sample code in user/tsh0.c, which implements the functions of parsing user input into a command structure. You can write your shell based on this code in your project make as many changes as you want to suit your program's need. To see how this program works, execute:
```
make qemu
xv6 kernel is booting
init: starting sh
$ tsh0
tsh>
```
+ Xv6 supplies you with a small library of C functions in user/ulib.c; feel free to use them. As mentioned above,
however, you are not allowed to use malloc().
+ Remember to close un-needed file descriptors, both to avoid running out, and because a process reading a
pipe won't see end-of-file until all the write descriptors for that pipe are closed.
+ Every system call in your code should check whether the call returned an error.
+ Remember that a C string of N characters requires N + 1 bytes for memory storage (the last bit will be \'\0').
+ The testsh redirects your shell's standard output, which means you won't see it. Your shell should print
error and debug messages using the ErrorU(), ErrorP(), ErrorS() and Debug() function.
+ You can modify the testsh.c program if necessary. We will use a separate test program in the grading, which is
identical to testsh.c but may use different but equivalent test cases.
Submission
Please following the procedure below in your submission.
1. Switch to your project folder:
```cd ~/path/to/your-projects-folder```
2. Clean up the folders
```
cd TaskA
make clean
cd ../TaskB
make clean
```
3. List files that you have added or modified: ```git status```
4. Stage the new and modified files in your local repo: ```git add files-you-created-or-modified```
5. Commit the changes to your local repo: ```git commit -m "Finished Project 2"```
6. Push the changes to the remote repo (i.e. github): ```git push``` Or ```git push origin master```.
7. Verify all changes have been push into the repo on github.
+ You can run "git status", "git log", and "git show" to the see the changes and commits you have made
+ You can also log into github to see if your changes have been actually pushed into your project repo on github.
8. (Optional but recommended) For a further validation, you can check out the repo in a separate folder
on your computer and then verify that it has all the files for the programs to work correctly.