1. 1 Week 7 System Calls, Kernel Threads, Kernel Debugging Sarah Diesburg Florida State University

2. First… Any questions on  Part 1 – 5 system calls  Part 2 – xtime proc module 2

3. Story of Kernel Development Some context… 3

4. In the old days… There were no modules or virtual machines The kernel is a program  Has code, can compile, re-compile, make executable  When changes needed to be made, developers make changes in source and re-compile 4

5. How is the kernel different from a regular program? Mostly in how it is executed  Boot loader loads the kernel image/executable during boot time  Sets kernel mode  Jumps to the entry point in the image/executable Remember the generic booting sequence? 5

6. Quick Question How would you make changes to the kernel and run those changes? 1. Make changes to the source 2. Re-complie the kernel source 3. Re-install the kernel source 4. Make sure the bootloader sees the new kernel image (grub) 5. Reboot and profit! 6

7. Getting more modern.. Modules were created as bits of code that can be loaded and unloaded by the kernel in kernel mode Made development easier  Instead of re-compiling, re-installing, and rebooting into the new kernel, one could just re- compile and load a module 7

8. Quick Question How would you make changes to a module and run those changes? 1. Make changes to module source code 2. Re-compile the module 3. Load the new module 8

9. Present Day Reboots into new kernels and loading new modules often freezes machines Enter virtual machine software  Process that emulates the hardware necessary to run an OS in user-space  Guest OS is executed inside the virtual machine process! 9

10. New System Calls Fun but tricky! 10

11. Implementing System Calls Need to implement the functions above. But how? int start_elevator(void); int issue_request(int #1, int #2, int#3); int stop_elevator(void);

12. Adding a System Call to Kernel Files to add:  LINUX_DIR/PROJECT_NAME/Makefile  LINUX_DIR/PROJECT_NAME/PROJECT_NAME.c  LINUX_DIR/PROJECT_NAME/NEW_SYSCALLS.c Files to modify:  LINUX_DIR/arch/x86/kernel/syscall_table_32.S  LINUX_DIR/include/asm-generic/unistd.h  LINUX_DIR/include/linux/syscalls.h  LINUX_DIR/Makefile

13. Sample System Call Let’s add a sample module to the kernel that defines a sample system call test_newsyscall(int test_int);  Takes an int test_int and issues a printk on it  Returns test_int Problem – If our module isn’t loaded, what happens if we call our sample system call? 13

14. Project 2 System Call Model 14 Elevator moduleCore kernel User program

15. Project 2 System Call Model 15 Elevator moduleCore kernel User program User issues system call, core kernel looks up system call in system call table

16. Project 2 System Call Model 16 Elevator moduleCore kernel User program Elevator module performs system call action

17. Project 2 System Call Model 17 Elevator moduleCore kernel User program Elevator module returns result of system call

18. Project 2 System Call Model 18 Elevator moduleCore kernel User program Core kernel forwards result of system call to user program

19. What happens if elevator module is not loaded? 19 Core kernel User program

20. What happens if elevator module is not loaded? 20 Core kernel User program User issues system call, core kernel looks up system call in system call table

21. What happens if elevator module is not loaded? 21 Core kernel User program Elevator module is not loaded to perform the action…. OOPS!

22. Module System Calls We must create a wrapper system call!  Wrapper will call module function if module loaded, else returns an error  Must be created in a separate, built-in kernel file in the project folder 22

23. Function Pointers We will implement our system call wrapper with a function pointer  Pointer to a function Function pointer can point to any function that you implement that  Takes the same input variable types  Returns the same return type 23

24. Function Pointers long (*STUB_test_newsyscall)(int test_int) = NULL; Function pointer that  Returns a long  Name is STUB_test_newsyscall  Takes parameter int test_int  Function pointer set to NULL Can set function pointer to a local function you implement 24

25. Elevator Project Create a file in your elevator project that just contains the system call information  KERNEL_DIR/PROJECT_DIR/newsyscalls.c 25

26. KERNEL_DIR/PROJECT_DIR/ newsyscalls.c #include /* System call stub. We initialize the stub function to be NULL. */ long (*STUB_test_newsyscall)(int test_int) = NULL; EXPORT_SYMBOL(STUB_test_newsyscall); /* System call wrapper. If the stub is not NULL, it will be run, otherwise returns -ENOSYS */ asmlinkage long sys_test_newsyscall(int test_int) { if (STUB_test_newsyscall) return STUB_test_newsyscall(test_int) else return -ENOSYS; } 26

27. KERNEL_DIR/PROJECT_DIR/ newsyscalls.c #include /* System call stub. We initialize the stub function to be NULL. */ long (*STUB_test_newsyscall)(int test_int) = NULL; EXPORT_SYMBOL(STUB_test_newsyscall); /* System call wrapper. If the stub is not NULL, it will be run, otherwise returns -ENOSYS */ asmlinkage long sys_test_newsyscall(int test_int) { if (STUB_test_newsyscall) return STUB_test_newsyscall(test_int) else return -ENOSYS; } 27 Function pointer

28. KERNEL_DIR/PROJECT_DIR/ newsyscalls.c #include /* System call stub. We initialize the stub function to be NULL. */ long (*STUB_test_newsyscall)(int test_int) = NULL; EXPORT_SYMBOL(STUB_test_newsyscall); /* System call wrapper. If the stub is not NULL, it will be run, otherwise returns -ENOSYS */ asmlinkage long sys_test_newsyscall(int test_int) { if (STUB_test_newsyscall) return STUB_test_newsyscall(test_int) else return -ENOSYS; } 28 Export the pointer so we can access it later

29. KERNEL_DIR/PROJECT_DIR/ newsyscalls.c #include /* System call stub. We initialize the stub function to be NULL. */ long (*STUB_test_newsyscall)(int test_int) = NULL; EXPORT_SYMBOL(STUB_test_newsyscall); /* System call wrapper. If the stub is not NULL, it will be run, otherwise returns -ENOSYS */ asmlinkage long sys_test_newsyscall(int test_int) { if (STUB_test_newsyscall) return STUB_test_newsyscall(test_int) else return -ENOSYS; } 29 System call wrapper

30. Elevator Project Next create a separate file that  Holds your module code  Registers the system call pointer  Actually implements the system call behavior 30

31. Inside KERNEL_DIR/PROJECT_DIR/ PROJECT_NAME.C /* Extern system call stub declarations */ extern long (*STUB_test_newsyscall)(int test_int); long my_test_newsyscall(int test) { printk("%s: Your int is %i\n", __FUNCTION__, test); return test; } my_module_init() { STUB_test_newsyscall=&(my_test_newsyscall); return 0; } my_module_exit() { STUB_test_newsyscall=NULL; } 31

32. Inside KERNEL_DIR/PROJECT_DIR/ PROJECT_NAME.C /* Extern system call stub declarations */ extern long (*STUB_test_newsyscall)(int test_int); long my_test_newsyscall(int test) { printk("%s: Your int is %i\n", __FUNCTION__, test); return test; } my_module_init() { STUB_test_newsyscall=&(my_test_newsyscall); return 0; } my_module_exit() { STUB_test_newsyscall=NULL; } 32 Gain access to stub function pointer.

33. Inside KERNEL_DIR/PROJECT_DIR/ PROJECT_NAME.C /* Extern system call stub declarations */ extern long (*STUB_test_newsyscall)(int test_int); long my_test_newsyscall(int test) { printk("%s: Your int is %i\n", __FUNCTION__, test); return test; } my_module_init() { STUB_test_newsyscall=&(my_test_newsyscall); return 0; } my_module_exit() { STUB_test_newsyscall=NULL; } 33 Local function that implements syscall

34. Inside KERNEL_DIR/PROJECT_DIR/ PROJECT_NAME.C /* Extern system call stub declarations */ extern long (*STUB_test_newsyscall)(int test_int); long my_test_newsyscall(int test) { printk("%s: Your int is %i\n", __FUNCTION__, test); return test; } my_module_init() { STUB_test_newsyscall=&(my_test_newsyscall); return 0; } my_module_exit() { STUB_test_newsyscall=NULL; } 34 Set stub function pointer to local function in init

35. Inside KERNEL_DIR/PROJECT_DIR/ PROJECT_NAME.C /* Extern system call stub declarations */ extern long (*STUB_test_newsyscall)(int test_int); long my_test_newsyscall(int test) { printk("%s: Your int is %i\n", __FUNCTION__, test); return test; } my_module_init() { STUB_test_newsyscall=&(my_test_newsyscall); return 0; } my_module_exit() { STUB_test_newsyscall=NULL; } 35 Reset stub function pointer to NULL on module unload

36. KERNEL_DIR/PROJECT_DIR/ Makefile obj-m := my_module.o obj-y := newsyscalls.o KDIR := /lib/modules/2.6.32/build PWD := $(shell pwd) default: $(MAKE) -C $(KDIR) SUBDIRS=$(PWD) modules 36

37. KERNEL_DIR/PROJECT_DIR/ Makefile obj-m := my_module.o obj-y := newsyscalls.o KDIR := /lib/modules/2.6.32/build PWD := $(shell pwd) default: $(MAKE) -C $(KDIR) SUBDIRS=$(PWD) modules 37 Compile as a module

38. KERNEL_DIR/PROJECT_DIR/ Makefile obj-m := my_module.o obj-y := newsyscalls.o KDIR := /lib/modules/2.6.32/build PWD := $(shell pwd) default: $(MAKE) -C $(KDIR) SUBDIRS=$(PWD) modules 38 Compile as kernel built-in

39. Core Kernel Additions Add the new system call to the core kernel system call table  Modify three files Add the project directory to the main Makefile  Modify one file 39

40. Modifying syscall_table_32.S ….long sys_preadv.long sys_pwritev.long sys_rt_tgsigqueueinfo /* 335 */.long sys_perf_event_open

41. Modifying syscall_table_32.S ….long sys_preadv.long sys_pwritev.long sys_rt_tgsigqueueinfo /* 335 */.long sys_perf_event_open.long sys_test_newsyscall /* 337 */  Add new system call to the end of the file.  Remember the number – you’ll need it in userspace!

42. Modifying unistd.h /* midfile */ #define __NR_perf_event_open 241 __SYSCALL(__NR_perf_event_open, sys_perf_event_open) #undef __NR_syscalls #define __NR_syscalls 242 /* midfile */  Can be found around line 623…

43. Modifying unistd.h /* midfile */ #define __NR_perf_event_open 241 __SYSCALL(__NR_perf_event_open, sys_perf_event_open) #define __NR_test_newsyscall 242 __SYSCALL(__NR_test_newsyscall, sys_test_new_syscall) #undef __NR_syscalls #define __NR_syscalls 242 /* midfile */

44. Modifying unistd.h /* midfile */ #define __NR_perf_event_open 241 __SYSCALL(__NR_perf_event_open, sys_perf_event_open) #define __NR_test_newsyscall 242 __SYSCALL(__NR_test_newsyscall, sys_test_new_syscall) #undef __NR_syscalls #define __NR_syscalls 243 /* midfile */

45. Modifying syscalls.h asmlinkage long sys_perf_event_open( struct perf_event_attr __user *attr_uptr, pid_t pid, int cpu, int group_fd, unsigned long flags); #endif /* EOF */ 45

46. Modifying syscalls.h asmlinkage long sys_perf_event_open( struct perf_event_attr __user *attr_uptr, pid_t pid, int cpu, int group_fd, unsigned long flags); asmlinkage long sys_test_newsyscall(int test_int); #endif /* EOF */ 46

47. Modifying KERNEL_DIR/Makefile # Objects we will link into vmlinux / subdirs we need to visit init-y := init/ drivers-y := drivers/ sound/ firmware/ net-y := net/ libs-y := lib/ core-y := usr/ endif # KBUILD_EXTMOD 47

48. Modifying KERNEL_DIR/Makefile # Objects we will link into vmlinux / subdirs we need to visit init-y := init/ drivers-y := drivers/ sound/ firmware/ net-y := net/ libs-y := lib/ core-y := usr/ my_module/ endif # KBUILD_EXTMOD 48  Found around line 475…  Can replace “my_module” with the name of your PROJECT_DIR

49. Getting it all to work 1. Re-compile the kernel 2. Install modules, install kernel 3. Make new initramfs image 4. Reboot 5. Test with a user-space program…

50. Sample User-space Program #include #define __SYS_TEST_ELEVATOR 337 int main() { int test=5; long ret; ret=syscall(__SYS_TEST_ELEVATOR, test); if(ret<0) perror("system call error"); else printf("Function successful, returned %i\n", ret); return 0; } 50

51. syscall() int syscall(int number,...); Performs the system call based on the system call’s number  Number can be found in the syscall_table_32.S file (our example was 337) 51

52. User-space Program Output Output when my_module not loaded system call error: Function not implemented Output when my_module loaded Function successful, returned 5 52

53. Kthreads Run the main logic of your module in a kthread! 53

54. Refresher: hello.c #include MODULE_LICENSE(“Dual BSD/GPL”); static int hello_init(void) { printk(KERN_ALERT “Hello, world!\n”); return 0; } static void hello_exit(void) { printk(KERN_ALERT “Goodbye, sleepy world.\n”); } module_init(hello_init); module_exit(hello_exit); 54

55. Kernel Modules Remember, kernel modules are very event- based We need a way to start an independent thread of execution in response to an event  e.g. start_elevator() for project 2… 55

56. kthread_run kthread_run(threadfn, data, namefmt,...) Creates a new thread and tells it to run  threadfn – the name of the function the thread should run  data – data pointer for threadfn (can be NULL if the function does not take any args)  namefmt – name of the thread (displayed during “ps” command) Returns a task_struct 56

57. kthread_run example struct task_struct *t; t = kthread_run(run, NULL, “my_elevator"); if (IS_ERR(t)){ ret=PTR_ERR(t); } 57

58. kthread_stop int kthread_stop(struct task_struct * k); Sets kthread_should_stop for k to return true, wakes the thread, and waits for the thread to exit Returns the result of the thread function 58

59. kthread_stop_example ret=kthread_stop(t); if(ret != -EINTR) printk("Main logic tread stopped.\n“); 59

60. Thread Function Example static int run(void *arg) { /* Lock here */ while(!kthread_should_stop()) { /* Do stuff */ /* Unlock here */ schedule(); /* Lock here */ } /* Unlock here */ printk("%s: kernel thread exits.\n", __FUNCTION__); return 0; } 60

61. Thread Function Example static int run(void *arg) { /* Lock here */ while(!kthread_should_stop()) { /* Do stuff */ /* Unlock here */ schedule(); /* Lock here */ } /* Unlock here */ printk("%s: kernel thread exits.\n", __FUNCTION__); return 0; } 61 schedule() is very important here. Why?

62. Inefficient Solution Thread will continue to run even though it has nothing to do  Eats up resources Investigate the kthread interface to find ways to  Put thread to sleep  Wake up thread There is more than one way to do this… 62

63. Debugging 63

64. Kernel Debugging Configurations Timing info on printks __depreciated logic Detection of hung tasks SLUB debugging Kernel memory leak detector Mutex/lock debugging Kmemcheck Check for stack overflow Linked list debugging 64

65. Select Kernel Hacking 65

66. Enable Debugging Options 66

67. Debugging through procfs Necessary for elevator project! General process  Identify data to monitor in your module  Create a proc entry to monitor this data  Run your module  Query /proc/ for that information at any time 67

68. Kernel Oops and Other Errors Kernel errors often only appear on first tty (terminal interface)  Why? How can I see my first tty?  On regular system – CTRL+ALT+F1 CTRL+ALT+F7 to go back to X screen  On VMware – CTRL+ALT+SPACE+F1 CTRL+ALT+SPACE+F7 to go back to X screen 68

69. 69 Oops!

70. 70 Reason for failure

71. 71 Current drivers

72. 72 Call Trace

73. 73 Call Trace

74. 74 Failed command

75. Defensive Programming Infinite loops and deadlocks at the kernel level hang your machine – Ctrl-Alt-Del has NO effect – Ctrl-C does not matter – Ctrl-D does not matter – You may only reboot How do you protect yourself? – Use schedule() strategically – Use preemptable versions of functions

76. Debugging Tools not Covered  LTT – Linux Tracing Framework  gdb – Invoking gbd on the kernel image  kgdb – A remote debugger for the kernel  Magic SysRq  printk – Rate limiting, turning on/off

77. Next Time Locks Linked lists Elevator algorithms 77