II. Systems Programming using C (Process Control Subsystem)

Transcription

1 II. Systems Programming using C (Process Control Subsystem) 129

2 Intended Schedule Date Lecture Hand out Submission Introduction to Operating Systems Course registration Systems Programming using C (File Subsystem) 1. Assignment Systems Programming using C (Process Control) 2. Assignment 1. Assignment Processes Scheduling 3. Assignment 2. Assignment Process Synchronization 4. Assignment 3. Assignment Inter Process Communication 5. Assignment 4. Assignment Pfingstmontag 6. Assignment 5. Assignment Input / Output 7. Assignment 6. Assignment Memory Management 8. Assignment 7. Assignment Assignment 8. Assignment Filesystems Assignment 9. Assignment Special subject: Transactional Memory 10. Assignment Special subject: XQuery your Filesystem Wrap up session First examination date Second examination date 130

3 Process Control Subsystem 131

4 !"#$%"& *+%,-.('#")-/ UNIX system architecture (System V)!"+5#<"$=$;6$1))" -."#&'"("%!"#$"%&'"("% :$;6$1))&%&A%;#?"D.('#")15E$5E4.>?+%##'#"AA"-8.3.4"5&8*%%&0$4"#/*8"9 71#"%4.5&'('#")-8/0%"& "59 B+#"$4:$;<"''4 C;))5+%D1#%;+ 8;<=9 Files :5EE"$ //"#&8*89"9 :$;<"'''#"5"$5+6'4.5&'('#")-86#78".. 87$4#7%& "59 :$;<"'';$4 <5#"%A "+1%0$:9 Processes 85"57#3 5*$*:"5"$49 01$231$"4.#"5"$5+6!"#$"%&'"("% )*#+,*#"&'"("% 01$231$"!"#$%&'()*&+,*(-.)/ 132

5 Program and Processes A program is an executable file residing on disk in a directory read into memory executed by the kernel as a result of one of the six exec(3) functions. An executing instance of a program is called a process. Every process has a unique numeric identifier called process ID. The process ID is always a non-negative integer. Although unique, process ID are reused (kernel delays re-usage). 133

6 Running process printing pid 134

7 Status information of processes man 1 ps describes the user command program /bin/ps 135

8 Obtaining process state with ps(1) 136

9 Getting process identifiers 137

10 Virtual Memory Each process has its own address space in virtual memory. Processes: do not know about each other (from their perspective) run alone on the machine their address spaces are independent from each other For inter-process communication special kernel mechanisms have to be invoked ( lecture 5) 138

11 Virtual Memory on Linux Memory locations are addressed through pointers. As such the natural word size of a processor sets an upper bound for the largest address to reference. On 32-bit systems: 2 32 Bytes = Bytes = 4 GiB Often this exceeds the physical memory available and gives name to the term virtual memory. Linux divides virtual memory in kernel and user space. Each user process has its own address space from 0 to TASK_SIZE. TASK_SIZE depends on architecture. kernel space user space 2 32 / 2 64 TASK_SIZE 0 139

12 Building programs 140

13 From source to executable code From Source Code To Executable Code Translation Steps (multi-phase compilation) Compilation HLL source code to assembler source code Assembly Assembler source code to object code Linking Object code to executable code Compilers and assemblers create object files containing the generated binary code and data for a source file. Linkers combine multiple object files into one, loaders take object files and load them into memory. Goal: An executable binary file (a.out) From high-level language (HLL) source code to executable code, i.e., concrete processor instructions in combination with data

14 Translation step using gcc(1) Translation Steps Using gcc(1) Quellcode C/C++ Assembler-Quellcode Objektdatei, Bibliotheksdatei Eingabedateien *.c/*.cc/*.cpp *.s *.o/*.a Präprozessor Compiler Assembler Binder Ausgabedateien *.i/*.ii *.s *.o a.out Vorverarbeiteter C/C++-Quellcode Assembler-Quellcode Objektdatei Ausführbare Datei (ungebunden) (= Objektdatei, ladbar)

15 Creation of an executable file Creation Of An Executable File (Filename).c Kompilieren gcc (Filename).s = Operation = Kommando = Eingang oder Ausgang Assemblieren gas Object/Library Files (Filename).o ld Binden a.out

16 Linking an executable binary Linking An Executable Binary OBJ1 OBJ2 OBJ3.text1.data1.bss1.text2.bss2.text3.data3.bss3 Eingabedaten: ungebundene Objektdateien Bindung (linking).text: Code.data: initialisierte Variablen.bss: nicht initialisierte Variablen OBJtotal.text1.text2.text3.data1.data3.bss1.bss2.bss3 Verarbeitungsresultat: ausführbare Datei (gebunden, reloziert) Each object code (compiled seperately) starts at address 0 Linking them together involves centralization of sections relocation of adresses

17 Program file organization Typical Program Organisation A typical program divides naturally in sections Code machine instructions, should be unmodifiable, size is known after compilation, does not change (.text) Data static data initialized (.data) /uninitialized (.bss) constant address in memory permanent life time dynamic data stack or heap storage space not known volatile life time

18 Memory Layout 146

19 (Virtual) Memory Layout Virtual Memory And Segments Virtual Memory Whenever a process is created, the kernel provides a chunk of physical memory which can be located anywhere Through the magic of virtual memory (VM), the process believes it has all the memory on the computer ( lecture 8) Typically the VM space is laid out in a similar manner: Text Segment (.text) Initialized Data Segment (.data) Uninitialized Data Segment (.bss) The Stack The Heap

20 { Memory Segments (TASK_SIZE) 148

21 Memory segments Memory Segments Text Segment. The text segment contains the actual code (including constants) to be executed. It s usually sharable, so multiple instances of a program can share the text segment to lower memory requirements. This segment is usually marked read-only so a program can t modify its own instructions. Initialized Data Segment. This segment contains global variables which are initialized by the programmer. Uninitialized Data Segment. Also named.bss (block started by symbol) which was an operator used by an old assembler. This segment contains uninitialized global variables. All variables in this segment are initialized to 0 or NULL pointers before the program begins to execute

22 Memory segments Memory Segments (cont.) The Stack The stack is a collection of stack frames which we will discuss later. When a new frame needs to be added (as a result of a newly called function), the stack grows downward. The Heap Dynamic memory, where storage can be (de-)allocated via C s free(3)/malloc(3). The C library also gets dynamic memory for its own personal workspace from the heap as well. As more memory is requested on the fly, the heap grows upward

23 Loading into memory segments Variable Placement Variables (outside a function) Globally declared variables go to the Uninitialized Data Segment if they are not initialized, to Initialized Data Segment otherwise. Necessary for the OS to decide if storage has to be loaded with initialization data from the executable binary. Variables (inside a function) Implicit assumption of auto, go to The Stack. Declared as static, see above. Constants (const) Text Segment Function Parameters Are pushed on The Stack or stored in registers. If pointers are passed, data is elsewhere

24 Variable placement, life time, visibility Variable Placement And Life Time (Code) int a; /* Permanent life time */ static int b; /* dito, but reduced scope */ void func ( void) { char c; /* only for the life time of func() */ /* but 2x; visible only in func() */ static int d; /* i m unique, exist once at a stable */ /* address, visible only in func() */ } int main ( void) { int e; /* life time of main() */ } int *pi = ( int*) malloc ( sizeof( int )); /* newborn */ func (); func (); free (pi ); /* RIP, pi points to an invalid address */ return (0);

25 Variable placement in address space Variable Placement And Life Time (Diagram) Adresse 0 PC(t=0) PC(t=x) pi 1. Instruktion 2. Instruktion 3. Instruktion 4. Instruktion... a b d int Code Daten Halde (Heap) SP(t=x) SP(t=0) max. c pi e Stapel (Stack) t=0: Programmausführung wird gestartet, d.h., Ausführungsumgebung ist bereits initialisiert t=x: beliebiger Zeitpunkt während der Programmausführung

26 Multitasking 154

27 Multitasking Systems In terms of hardware parallelism as much processes may be executed concurrently as processors are available. Multitasking systems allow for (apparently) simultaneous execution of multiple processes on a single processor. The kernel switches between processes in short intervals to simulate the parallel execution (task switch). The kernel decides in what way the computing time is assigned to different processes (scheduling lecture 3). 155

28 Processes and Threads Threads: Parallel running activities inside a process. Each thread belongs to dedicated process !9 7:9 Single Process 7<)32173()*+255 7<)32173()*+255.2C)2)23()*+2552 Multiple Processes 7;9.2C)2)23()*+2552 Single Control Flow 7<)3%3MC)2">.2C)2)23MC)2">5 Three Control Flows (Threads) Single Control Flow 8237<)32173MC)2"> Multiple Control Flows (Threads) 823.2C)2)23MC)2">5!"#$%$&'(")()*+,-. /*01234

29 Pseudo parallel execution!"#$%"&''('#")"!"%'*%"+,-.$,/"0$*$12$3))&"#$%"& Conceptually: Each process has its own machine -."+),//'012'.*' ,4$16"'',7, ,4$16"'',:,845:9 45,4$16"'',;,845;9 45,4$16"'',<,845<9 4$16"''?)> task switch 45D,4$12$3))> 6E0+"$,8!"#$"%&'(#)*+,"9 PC: program counter PCB: Process Control Block ",%/'012'.*' ,4$16"'',7 45,4$16"'',: 45,4$16"'',; 45,4$16"'',< 45,4$16"'',7 45,4$16"'',:?'AB CCD3E2F: ()*+2;;23G!"#$"%&'()*+,-."!!"!H E=>0:;;-*0A2):6A26Q !"#$%$&'(")()*+,-. /*01234 ="%# time time ="%# 4$16"'',7 4$16"'',: 4$16"'',; 4$16"'',< Time-Multiplexing

30 !"#$%"&''('#")" Task Switch and PCB *$+,"''-)'./01#-23"24-254*6! *$+,"''47 *$+,"''48 *64%24:*67;4'%./"$24 :*68;4%24*64<+=%"$"24 *$+,"''41>-?#42%./# :*67;4%24*64<+=%"$"24 *64%24:*68;4'%./"$24 :*67;B4:*68;A4C%$#-"11"4*6 5"$4*$+,"''"474D48 9"%# *674D4*684'%254%24*6!74D *6!84"2#/01#"2 *6!A4*$+."''46+2#$+14!1+.< :*$+,"''C"$E01#-23'50#"2;!"#$%$&'(")()*+,-. Glatz 2005 /*

31 Process Life Cycle!"#$%"&''('#")" *&+,-./0,'"12"%1"'23$45"''"' 56"713892:2;7#<"72;=321;273()* "A2;3A273()*C)"..73DE39"A2#<"72F G?")?2;3A273()*C)"..73DE3G?")?#<"72F ()*C)"..":0"6-3DE3H2?)12:7#<"72F ()*C)"..:22;A1C6;C3DE3I2).1;12)6;C7#<"72F 6&7"8#9,#"%.4$),# "67-K<):")2;3J"?213":C202C?371;A MMN3U1;3:27?1..?273H2?)12:77Q7?2.3!21;321;3*A2)3.2<)2)23/*)."?236;?2)7?K?+2; kann!"#$%$&'(")()*+,-. /*0123%4 159

32 Process Termination 160

33 Process Termination!"#$%"&''('#")" *$+,"''#"$)%-%"$.-/ /*).25362)3()* :59; <$=3A*)+21B1923?225619: !"B"7B)*#D"02.3/2D02)32)!"55B36:)CD3 >?;3IJ:-197B2)3/"00;3<%=3>*)."023?225619:

34 Eight way for a process to terminate Normal termination Returning from main Calling exit(3) Calling _exit(2) (POSIX.1) or _Exit(3) (ISO C) Return of the last thread from its start routine Calling pthread_exit(3) from the last thread Abnormal termination Calling abort(3) Receipt of a signal ( lecture 5) Response of the last thread to a cancellation request 162

35 Normal termination details Normal termination Returning from main Equivalent to calling exit(3) Calling exit(3) closes/flushes all standard I/O streams calls all exit handlers registered with atexit(3) ISO C does not deal with file descriptors, multiple processes (parents and children), and job control, as such its definition is incomplete for UNIX systems. 163

36 Normal termination details Calling _exit(2) (POSIX.1) or _Exit(3) (ISO C) ISO C defines _Exit(3) to provide a way for a process to terminate without exit or signal handlers. Whether standard I/O streams are flushed is system dependent. On UNIX systems _exit(2) and _Exit(3) are synonymous and do not flush standard I/O streams. The _exit(2) function is called by exit(2) and handles the UNIX system-specific details Return of the last thread from its start routine The return value of the thread is not used as the return value of the process. When the last thread returns, the process exits with 0. Calling pthread_exit(3) from the last thread Same as above 164

37 Starting and terminating a C program 165

38 Process Creation 166

39 Process Creation *+,+,-.$/0"''"$0" "$)%3%"$132!"#$%"&''('#")" A0023()* B19:12)263>*63A6-"6<3"63C6D30"C-2632E1<F3D,5,39*0"6<23489:2.30GC-: ()*+29923E2)D263H213I2D")-32)+2C<:3C6D3<202<26:01753E12D2)3:2).1612): ;%?3489:2.9:"):3;L61:1"01912)C6<? ;&?3489:2."C-)C-3+C)3()*+2992)+2C<C6<3DC)7531)<26D "C-26D263()*+299 ;$?3I26C:+2)"6-*)D2)C6<3+C.34:"): C263()* ;A##01!":1*699:"):? ;M?3AC90J9C6< :"#20"C-:)"<93;#,(23%41#?!"#$%$&'(")()*+, /*0123%&

40 What is a process?!"#$%"&''('#")" *+',%'#,"%-,.$/0"'' ()*:)"..2;7<15!02)93=1;3>?@ >A()*:)"..B38,6,3main()A/C;!71*; D27)12E99F972.93=G8)299)"C.E202:C;:?@ H"72;E2)21563=1;!0,347"5!3C;83I2"#? ()*:)"..5*82E2)2156 H"72;97)C!7C)3+C)3()*+2992J2)<"07C;:38C)5638"93 D27)12E99F972.3=!"#$%!&'()**%"'+,&'-%#-'(./$%01#1 2%!34%5!&'0)**.)++6+7/ 2%8.,6)--)&%!&'0)**06*,9+: 2%!&'0)**.'+,);, kernel space 232 / 2 64 PCB user space argv, env ()*+299A 5*82 ()*+299A 8"72; (>D G8)299)"C. TASK_SIZE linux/sched.h: struct task_struct 0 168

41 Process Hierarchy 169

42 !"#$%"&''('#")" *+,+-./$01"''2%"$3$42%".5!"%'6%"78.9:%;< UNIX process hierarchy /=>?A %:%# B"##( B"##( B"##( '2 B"##( C%:D '2!"#$%$&'(")()*+,-. /*0123$$ 170

43 UNIX system start (simplified) Bootstrap loads system code to memory. Initialization of internal kernel data structures. Mount root filesystem Environment for the first process is created. System code becomes process with PID=0 Process 0 forks and creates process 1 (kernel space) Process 1 creates environment in user space and turns over PID=0 (kernel space), PID=1 (user space) 171

44 UNIX system start (simplified) Process 1 uses exec on, e.g., /sbin/init & becomes init init is responsible for bringing up a UNIX system after the kernel has been bootstrapped. It usually reads system-dependent initialization files (such as /etc/rc* or /etc/inittab, /etc/init.d) and brings the system to a certain state, such as multiuser. For instance, the terminals for users to login (getty) are started. It is a normal user process running with superuser privileges. init never dies. Meanwhile, process 0 starts kernel services in kernel space and finally becomes the swapper, i.e., the scheduler process. 172

45 Process Hierarchy on Linux Linux has a hierarchical process schema, i.e., each process is connected to its parent process. The kernel starts init as first process. The process hierarchy is a result of the process creation used on UNIX systems: fork(2) exec(3) family 173

46 Process creation and joining *$+,"'''#-$#.+$)"/ 3456*$+,"''0"$7"##2/1 3!56*$+,"''0"$1-&"82/1 3956*$+,"''"$,"212/1!"#$%"&''('#")" 45"16 -*)! 4)2"72 *$+,"''0"$"%/%12/1'.+$)"/ 3:56*$+,"''0"$"%/%12/1 3;56*$+,"''#$".."/ 8* :" Operating!"#$%$&'(")()*+,-. Systems Prof. Dr. Marc H. Scholl DBIS U KN Summer Term 2009 /*0123%$ 174

47 Related system calls!"#$%"&''('#")" *('#")+,-$,-".-/$.0$12"''3456$"+73*#+$# B6+%9 exec() $C fork() B$"+#" 5 pthread_create() CreateProcess() CreateThread() *('#")+,-$,-".-/$.0$12"''3456$"+738"$"%9%:,9: D1%9 wait()63waitpid() pthread_join() 5 5 A+%# 5 5 WaitForSingleObject() WaitForSingleObject()!"#$%$&'(")()*+,-. /*0123%4 175

48 Initialization of processes/threads!"#$%"&''('#")" *"%#"$+,&"-./0-1,#"0-,0-0"2"-3$/4"''"567$",8' !21:2;< H21D#12023!*;!)2:2)3C.#02.2;:12)?;72;31;3H2:)12>DDID:2.2;< 90%:;3$/4"''" 3<=>?;67$",8' 3$/4"''" A"$"$&20+ fork() J CreateProcess() J >0#"$B$/4C/)D popen() J J J *%08/@'; 67$",8' >0%#%,EB,$,)D J pthread_create() J CreateThread() 176

49 Inheritance with fork / exec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perating!"#$%$&'(")()*+,-. Systems Prof. Dr. Marc H. Scholl DBIS U KN Summer Term 2009 /*0123$% 177

50 Forking processes 178

51 *+,+*-.$/0"''"-12#"$-32%45-.$/0"''6"$78&"9127-:;/$<%27=!"#$%"&''('#")" ()1:+1#8215#120D int main() { int k, status; pid_t pid; k=fork(); if (k == 0) {... // Anweisungen, die nur Kindprozess } Forking processes on UNIX... // ausführen soll exit(0); // Kindprozess terminiert } else {... // Anweisungen, die nur Elternprozess... // ausführen soll pid = wait(&status); // Auf Ende des Kindprozesses warten exit(0); }!"#$%$&'(")()*+,-. /*0123%4 179

52 Forking processes on UNIX!"#$%"&''('#")" *$+,"''-"$./&" *$+,"''-"$5+66"012.37/&38+$9:;18$18< =0#"$26$+,"'' >%256$+,"'' *$%2,%6/&0/18 main() { int k, status; pid_t pid; main() { int k, status; pid_t pid; k=fork(); if (k == 0) { exit(0); } else { pid = wait(&status); } } k=fork(); if (k == 0) { exit(0); } else { pid = wait(&status); } } -*)! #")286 9:10; 5" !"#$%$&'(")()*+,-. /*0123%4 180

53 Process system calls in a nutshell *+%,-.('#")/01$01"213$24$56"''"$6" "$)%+%"$0+7!"#$%"&''('#")" system call pid = fork () pid = wait (&statloc) pid = waitpid (pid, &statloc, options) s = execve (name, argv, environp) exit (status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wait() /3waitpid()3A2.3>0?2):#)* ;M:;01673;2."67?=!"#$%$&'(")()*+,-. /*0123&& 181

54 The fork(2) system call PID is returned to parent, because a process can have more than one child, and there is no function that allows a process to obtain the PID of its children. Returning 0 to the child is apt, since PID=0 is reserved by the kernel, so it s not possible for 0 to be the PID of a child. The child can always obtain the PID of its parent, calling getppid(2). 182

55 !""*+%,-*,'./-*+/0-'123$-*,''#4#-'!"#$%"&''('#")" 92>263:100D 41232)?"0;263:1) I6-*).";1*6263A*.3JK7;2.3+B927;200;L FFM );23!*.>1612);3A1"3N7;";0*COF(")".2;2)35273wait()Pwaitpid()FEB-)B-7 FFM3N7;";B7O3!"663.1;;2073A*) );2)3R"!)*73"B792:2);2;3:2)526< 33333WIFEXITED(status)<3STUV3-H)36*)."0W3/EXJV3-H)3A*)+21;193">92>)*C? #120< Obtaining termination status pid_t pid; int status; pid = wait(&status); if (WIFEXITED(status)) printf( Rückgabewert ist %d\n, WEXITSTATUS(status);!"#$%$&'(")()*+,-. /*0123&$ 183

56 Zombie Processes!"#$%"&''('#")" *$+,"'''#-$#.)%#./"$,0"%1"2.34+)&%"5*$+&6")7 = :35062)7#)*+28839") !7:3:17;#)* ).1712)6<3 "E-357;23;283:17;#)* >2F*)35062)7#)*+28839")626 ( C ( C ( C ( C fork() ( C ( D ( C fork() ( ( C ( D D ( C ( D exit() exit() B"627A wait() =*.>12? #)* !"#$%$&'(")()*+,-. /*0123%4

57 Zombie prevention *+)&%","$-%./"$0.12)%##"3'24.5"36$+7"'' G%H G&H G$H GIH GKH fork() 91:;% 91:;% 91:;% wait()j exit() fork() 91:;& 91:;& 91:;& /*0123&4 185

58 Process chaining 186

59 Process chaining!"#$%"&''('#")" *$+,"''#-$#.)%#./"$0"## %2% ": <"%# /"$0"##123 * = * = ;-#"2: execve() *! *! :3272;<=>?@-)@-253)2"01A12)8 6582)A;B )3:3272;<=>C")1"5825D3?)83@593?5+"B0392)3?@-)@-#")".282) <15392)3/@5!81*53A1593"0023E021;BF39,B,3()*+2AAG2)!288@5E=!"#$%$&'(")()*+, /*0123&%

60 exec(3) family ok = execl(pathname, arg0, arg1,..., NULL); Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * pathname: Pfadname der Datei const char * arg0, * arg1,..: 1./2./.. Kommandozeilenargumente (Liste mit NULL abschließen) Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execle(pathname, arg0, arg1,..., NULL, envp[]); Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * pathname: Pfadname der Datei const char * arg0, * arg1,..: 1./2./.. Kommandozeilenargumente (Liste mit NULL abschließen) const char * envp[]: Adresse einer neuen Umgebungsvariablenliste Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execlp(filename, arg0, arg1,..., NULL) Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * filename: Dateiname (Suchpfad entsprechend Umgebungsvariable PATH) const char * arg0, * arg1,..: 1./2./.. Kommandozeilenargumente (Liste mit NULL abschließen) Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execv(pathname, argv[]); Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess 188

61 ok = execlp(filename, arg0, arg1,..., NULL) Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * filename: Dateiname (Suchpfad entsprechend Umgebungsvariable PATH) const char * arg0, * arg1,..: 1./2./.. Kommandozeilenargumente (Liste mit NULL abschließen) exec(3) family Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execv(pathname, argv[]); Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * pathname: Pfadname der Datei const char * argv[]: Adresse einer Liste mit allen Kommandozeilenargumenten Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execve(pathname, argv[], envp[]); Tab. 3 5 Unix-Systemaufrufe execl(), execle(), execlp() und execv() Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * pathname: Pfadname der Datei const char * argv[]: Adresse einer Liste mit allen Kommandozeilenargumenten const char * envp[]: Adresse einer neuen Umgebungsvariablenliste Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 ok = execvp(filename, argv[]); Zweck: Startet ausführbare Datei als neuen Prozess, ersetzt (terminiert) damit laufenden Prozess Parameter: const char * filename: Dateiname (Suchpfad entsprechend Umgebungsvariable PATH) const char * argv[]: Adresse einer Liste mit allen Kommandozeilenargumenten Rückgabewert int: keine Rückkehr falls okay; im Fehlerfall Rückgabe von -1 Tab. 3 6 Unix-Systemaufrufe execve() und execvp() 189

62 Differences among the exec functions 190

63 main function 191

64 Program Startup The main function Prototype for the main function The function called at program startup is named main It shall be defined with a return type of int And either with no parameters: int main ( void); Or with two parameters: int main ( int argc, char * argv []); 192

65 Constraints For main Function Constraints of main function Terminology (though any names may be used) argc stands for argument count argv stands for argument vector If argc and argv are declared The value of argc shall be nonnegative argv[0] represents the program name or argv[0][0] shall be the null character if the program name is not available argv[1] to argv[argc-1] represent program parameters argv[argc] shall be a NULL pointer 193

66 Command-line Arguments main: Argument vector When a program is executed, the process that does the exec can pass command-line arguments to the new program Normal operation for UNIX system shells argv: NULL echo\0 hello,\0 world\0 194

67 Environment Environment Each program is also passed an environment list Like the argument list, it is an array of character pointers Each pointing to a null-terminated C string The address of the array is contained in a global variable: extern char ** environ ; environ: NULL HOME=/home/holu\0 SHELL=/bin/ksh\0 PS1=\w \$\

68 Environment (cont.) Environment (cont.) Terminology environ is called environment pointer The array of pointers is the environment list The strings they point to are the envoronment strings By convention, name=value string are used Historical third argument to main (not ISO C) int main ( int argc, char * argv [], char * envp ); Most UNIX systems have provided a third argument to main ISO C specifies main with two arguments Posix.1 specifies environ to be used instead of 3rd arg

69 Echo Command-line Arguments Echo command-line arguments # include <stdio.h> int main ( int argc, char * argv []) { int i; /* for (i = 0; argv[i]!= NULL; i ++) */ for ( i = 0; i < argc ; i ++) printf (" argv [%d]: %s\n", i, argv [i ]); } return (0); $./a. out -a 1 -arg2 -- arg3 argv [0]:./a. out argv [1]: -a argv [2]: 1 argv [3]: -arg2 argv [4]: -- arg

70 Building a shell 198

71 (Very) basic sketch of a shell *+,+-./012#%31'4"%'".5"$.61%789:";; int main () { char kdo[100]; pid_t pid; int status;!"#$%"&''('#")" } while(1) { printf("$>"); gets(kdo); pid = fork(); if (pid==0) execl(kdo,null); else wait(&status); } 56738"93:*;2-)"<.2=>3+21<>3=?)3; @091=>2)#)2>">1*=3A;,@,3!21=23B2)")C21>?=< =<2C"?>2)3D@200E62-2@023?=;3!21=23D!)1#>1=>2)#)2>">1*=F!"#$%$&'(")()*+,-. /*0123&4 199

72 How a shell (basically) works... *+,-#%.,'/"%'"01"$023" *.."9:*19;2)#)2;2)373()*+2<<31.3=29>;+2).*:><!"#$%"&''('#" 8*.."9:*?2)")@21;>9A3BC9:0*<<D6021-2EF B%E3C1902<2938*.."9:*+21023"@35;"9:"):219A"@23BC19A"@232:1;12)@")E B&E38*.."9:*19;2)#)2;";1*93B9"D63G21029"@<D60><<3.1;3H)2;>)9IE B$E3C19A2@">;2)3=2-2603B!"#$%&#'()*++,'-EJ 33333KKI3L"F3M><-N6)>9A3:>)D <20@<; BOE3/"00<391D6;3B$EF35!)1#;:";21J 33333KKI3L"F3M@")@21;293:2)35!)1#;@ B4E3/"00<391D6;3BOEF35;");293"0<32P;2)92<38*.."9:*3B"><-N6)@")23Q";21E 200

73 *+",,-./0'12+$&3$".43#"%.'#3$#"5 56")637*)82)9):;8#)*+2<<= Start up of an executable!"#$%"&''('#")" 8:9 8=9 8?9 '+!"#$%& '+!"#$%& '+!"#$%& 16$789 '+ '()*+ '()*+ ;)< ">";.;)< /*0123&4 201

74 Next lecture: Next lecture/tutorial slots Process Scheduling Mo, , 12:00-14:00, C 252 (new room!) Next tutorial: Introduction to the C Programming Language Part II Tomorrow, Tue, , 16:00-18:00, C 252 Slides: arbeitsgruppen/dbis/lehre/operating-systems/ Have fun and see you next week! 202