Skip to content

Latest commit

 

History

History
112 lines (80 loc) · 4.9 KB

Project1.md

File metadata and controls

112 lines (80 loc) · 4.9 KB
Raw

Project 1

  • Assigned: 2021-09-23
  • Due: 2021-10-08 Fri 23:59:59 KST
  • Help document: doc

Introduction

This assignment includes implementing a new system call in Linux. It returns the process tree information in depth-first-search order.

This is a team project. Each team will have access to its own GitHub repository (e.g. [githubID of one member]/osfall2021-team1) for collaboration and submission. Your final code and README.md document have to be committed into the proj1 branch for submission. Start from kernel source and make incremental changes to get things done.

1. Writing ptree System Call

The system call you write should take two arguments and return the process tree information in a depth-first-search order.

The prototype for your system call will be:

int ptree(struct prinfo *buf, int *nr);

You should define struct prinfo as:

struct prinfo {
  int64_t state;            /* current state of process */
  pid_t   pid;              /* process id */
  pid_t   parent_pid;       /* process id of parent */
  pid_t   first_child_pid;  /* pid of oldest child */
  pid_t   next_sibling_pid; /* pid of younger sibling */
  int64_t uid;              /* user id of process owner */
  char    comm[64];         /* name of program executed */
};

in include/linux/prinfo.h as part of your solution.

The argument buf points to a buffer for the process data, and nr points to the size of this buffer (number of struct prinfo entries). The system call copies at most that many entries to the buffer in pre-order and stores the number of entries actually copied to variable pointed by nr.

pid_t value for a non-existing entity is 0. For example, first_child_pid should be 0 if the process does not have a child.

Each system call must be assigned a number. Your system call should be assigned number 398.

Return Value

Your system call should return the total number of entries on success (this may be bigger than the actual number of entries copied).

Your code should handle errors that can occur but not handle any errors that cannot occur. At a minimum, your system call should detect and report the following error conditions:

  • -EINVAL: if buf or nr are null, or if the number of entries is less than 1
  • -EFAULT: if buf or nr are outside the accessible address space. The referenced error codes are defined in include/uapi/asm-generic/errno-base.h

tasklist_lock

Linux maintains a list of all processes in a doubly linked list. Each entry in this list is a task_struct structure, which is defined in include/linux/sched.h. When traversing the process tree data structures, it is necessary to prevent the data structures from changing in order to ensure consistency. For this purpose the kernel relies on a special lock, the tasklist_lock. You should grab this lock before you begin the traversal, and only release the lock when the traversal is completed. While holding the lock, your code may not perform any operations that may result in a sleep, such as memory allocation, copying of data into and out from the kernel etc. Use the following code to grab and then release the lock (make sure you include the required header file):

#include <linux/sched/task.h>

read_lock(&tasklist_lock);
/* do the job... */
read_unlock(&tasklist_lock);

2. Test your new system call

Write a simple C program which calls ptree system call. Your program should print the entire process tree in pre-order using tabs to indent children with respect to their parents. For each process, it should use the following format for program output:

printf("%s,%d,%lld,%d,%d,%d,%lld\n", p.comm, p.pid, p.state,
  p.parent_pid, p.first_child_pid, p.next_sibling_pid, p.uid);

You can invoke ptree system call using syscall function. (See here for details.)

Example program output

systemd,1,1,0,156,2,0
  systemd-journal,156,1,1,0,185,0
  systemd-udevd,185,1,1,0,484,0
  syslogd,484,1,1,0,495,0
    ...
  deviced,802,1,1,1612,857,0
    systemctl,1612,64,802,0,1613,0
    systemctl,1613,64,802,0,1614,0
    systemctl,1614,64,802,0,31175,0
    ...
kthreadd,2,1,0,3,0,0
  ksoftirqd/0,3,1,2,0,5,0
  kworker/0:0H,5,1,2,0,6,0
  kworker/u8:0,6,1,2,0,7,0

Compiling test program

arm-linux-gnueabi-gcc -I<your kernel path>/include test.c -o test

Running test program

/root/test

(Switch to root if Permission denied error occurs.)

Save your C program as : test/test_ptree.c in your team repository.

We're Here to Help You

Any troubles? Questions on issue board are more than welcome. Discussions between students are also encouraged.

Start early, be collaborative, and still most importantly, have fun!