diff --git a/en_US.ISO8859-1/books/arch-handbook/book.sgml b/en_US.ISO8859-1/books/arch-handbook/book.sgml index a3d6113f2b..f407d4c9a7 100644 --- a/en_US.ISO8859-1/books/arch-handbook/book.sgml +++ b/en_US.ISO8859-1/books/arch-handbook/book.sgml @@ -1,509 +1,504 @@ %bookinfo; %man; %chapters; %authors; ]> FreeBSD Developers' Handbook The FreeBSD Documentation Project August 2000 2000 2001 The FreeBSD Documentation Project &bookinfo.legalnotice; Welcome to the Developers' Handbook. Introduction Developing on FreeBSD This will need to discuss FreeBSD as a development platform, the vision of BSD, architectural overview, layout of /usr/src, history, etc. Thank you for considering FreeBSD as your development platform! We hope it will not let you down. The BSD Vision Architectural Overview The Layout of /usr/src The complete source code to FreeBSD is available from our public CVS repository. The source code is normally installed in /usr/src which contains the following subdirectories. Directory Description bin/ Source for files in /bin contrib/ Source for files from contributed software. crypto/ DES source etc/ Source for files in /etc games/ Source for files in /usr/games gnu/ Utilities covered by the GNU Public License include/ Source for files in /usr/include kerberosIV/ Source for Kerbereros version IV kerberos5/ Source for Kerbereros version 5 lib/ Source for files in /usr/lib libexec/ Source for files in /usr/libexec release/ Files required to produce a FreeBSD release sbin/ Source for files in /sbin secure/ FreeSec sources share/ Source for files in /sbin sys/ Kernel source files tools/ Tools used for maintenance and testing of FreeBSD usr.bin/ Source for files in /usr/bin usr.sbin/ Source for files in /usr/sbin Basics &chap.tools; &chap.secure; Kernel History of the Unix Kernel Some history of the Unix/BSD kernel, system calls, how do processes work, blocking, scheduling, threads (kernel), context switching, signals, interrupts, modules, etc. &chap.locking; Memory and Virtual Memory Virtual Memory VM, paging, swapping, allocating memory, testing for memory leaks, mmap, vnodes, etc. I/O System UFS UFS, FFS, Ext2FS, JFS, inodes, buffer cache, labeling, locking, metadata, soft-updates, LFS, portalfs, procfs, vnodes, memory sharing, memory objects, TLBs, caching Interprocess Communication Signals Signals, pipes, semaphores, message queues, shared memory, ports, sockets, doors Networking Sockets Sockets, bpf, IP, TCP, UDP, ICMP, OSI, bridging, firewalling, NAT, switching, etc Network Filesystems AFS AFS, NFS, SANs etc] Terminal Handling Syscons Syscons, tty, PCVT, serial console, screen savers, etc Sound OSS OSS, waveforms, etc Device Drivers &chap.driverbasics; &chap.isa; &chap.pci; &chap.scsi; &chap.usb; NewBus This chapter will talk about the FreeBSD NewBus architecture. Architectures - - IA-32 - - Talk about the architectural specifics of FreeBSD/x86. - - + &chap.x86; Alpha Talk about the architectural specifics of FreeBSD/alpha. Explanation of allignment errors, how to fix, how to ignore. Example assembly language code for FreeBSD/alpha. IA-64 Talk about the architectural specifics of FreeBSD/ia64. Debugging Truss various descriptions on how to debug certain aspects of the system using truss, ktrace, gdb, kgdb, etc Compatibility Layers Linux Linux, SVR4, etc Appendices Dave A Patterson John L Hennessy 1998Morgan Kaufmann Publishers, Inc. 1-55860-428-6 Morgan Kaufmann Publishers, Inc. Computer Organization and Design The Hardware / Software Interface 1-2 W. Richard Stevens 1993Addison Wesley Longman, Inc. 0-201-56317-7 Addison Wesley Longman, Inc. Advanced Programming in the Unix Environment 1-2 Marshall Kirk McKusick Keith Bostic Michael J Karels John S Quarterman 1996Addison-Wesley Publishing Company, Inc. 0-201-54979-4 Addison-Wesley Publishing Company, Inc. The Design and Implementation of the 4.4 BSD Operating System 1-2 Aleph One Phrack 49; "Smashing the Stack for Fun and Profit" Chrispin Cowan Calton Pu Dave Maier StackGuard; Automatic Adaptive Detection and Prevention of Buffer-Overflow Attacks Todd Miller Theo de Raadt strlcpy and strlcat -- consistent, safe string copy and concatenation. diff --git a/en_US.ISO8859-1/books/arch-handbook/chapters.ent b/en_US.ISO8859-1/books/arch-handbook/chapters.ent index aca3e3bb3c..6e2d0c3f6f 100644 --- a/en_US.ISO8859-1/books/arch-handbook/chapters.ent +++ b/en_US.ISO8859-1/books/arch-handbook/chapters.ent @@ -1,61 +1,61 @@ - + diff --git a/en_US.ISO8859-1/books/developers-handbook/book.sgml b/en_US.ISO8859-1/books/developers-handbook/book.sgml index a3d6113f2b..f407d4c9a7 100644 --- a/en_US.ISO8859-1/books/developers-handbook/book.sgml +++ b/en_US.ISO8859-1/books/developers-handbook/book.sgml @@ -1,509 +1,504 @@ %bookinfo; %man; %chapters; %authors; ]> FreeBSD Developers' Handbook The FreeBSD Documentation Project August 2000 2000 2001 The FreeBSD Documentation Project &bookinfo.legalnotice; Welcome to the Developers' Handbook. Introduction Developing on FreeBSD This will need to discuss FreeBSD as a development platform, the vision of BSD, architectural overview, layout of /usr/src, history, etc. Thank you for considering FreeBSD as your development platform! We hope it will not let you down. The BSD Vision Architectural Overview The Layout of /usr/src The complete source code to FreeBSD is available from our public CVS repository. The source code is normally installed in /usr/src which contains the following subdirectories. Directory Description bin/ Source for files in /bin contrib/ Source for files from contributed software. crypto/ DES source etc/ Source for files in /etc games/ Source for files in /usr/games gnu/ Utilities covered by the GNU Public License include/ Source for files in /usr/include kerberosIV/ Source for Kerbereros version IV kerberos5/ Source for Kerbereros version 5 lib/ Source for files in /usr/lib libexec/ Source for files in /usr/libexec release/ Files required to produce a FreeBSD release sbin/ Source for files in /sbin secure/ FreeSec sources share/ Source for files in /sbin sys/ Kernel source files tools/ Tools used for maintenance and testing of FreeBSD usr.bin/ Source for files in /usr/bin usr.sbin/ Source for files in /usr/sbin Basics &chap.tools; &chap.secure; Kernel History of the Unix Kernel Some history of the Unix/BSD kernel, system calls, how do processes work, blocking, scheduling, threads (kernel), context switching, signals, interrupts, modules, etc. &chap.locking; Memory and Virtual Memory Virtual Memory VM, paging, swapping, allocating memory, testing for memory leaks, mmap, vnodes, etc. I/O System UFS UFS, FFS, Ext2FS, JFS, inodes, buffer cache, labeling, locking, metadata, soft-updates, LFS, portalfs, procfs, vnodes, memory sharing, memory objects, TLBs, caching Interprocess Communication Signals Signals, pipes, semaphores, message queues, shared memory, ports, sockets, doors Networking Sockets Sockets, bpf, IP, TCP, UDP, ICMP, OSI, bridging, firewalling, NAT, switching, etc Network Filesystems AFS AFS, NFS, SANs etc] Terminal Handling Syscons Syscons, tty, PCVT, serial console, screen savers, etc Sound OSS OSS, waveforms, etc Device Drivers &chap.driverbasics; &chap.isa; &chap.pci; &chap.scsi; &chap.usb; NewBus This chapter will talk about the FreeBSD NewBus architecture. Architectures - - IA-32 - - Talk about the architectural specifics of FreeBSD/x86. - - + &chap.x86; Alpha Talk about the architectural specifics of FreeBSD/alpha. Explanation of allignment errors, how to fix, how to ignore. Example assembly language code for FreeBSD/alpha. IA-64 Talk about the architectural specifics of FreeBSD/ia64. Debugging Truss various descriptions on how to debug certain aspects of the system using truss, ktrace, gdb, kgdb, etc Compatibility Layers Linux Linux, SVR4, etc Appendices Dave A Patterson John L Hennessy 1998Morgan Kaufmann Publishers, Inc. 1-55860-428-6 Morgan Kaufmann Publishers, Inc. Computer Organization and Design The Hardware / Software Interface 1-2 W. Richard Stevens 1993Addison Wesley Longman, Inc. 0-201-56317-7 Addison Wesley Longman, Inc. Advanced Programming in the Unix Environment 1-2 Marshall Kirk McKusick Keith Bostic Michael J Karels John S Quarterman 1996Addison-Wesley Publishing Company, Inc. 0-201-54979-4 Addison-Wesley Publishing Company, Inc. The Design and Implementation of the 4.4 BSD Operating System 1-2 Aleph One Phrack 49; "Smashing the Stack for Fun and Profit" Chrispin Cowan Calton Pu Dave Maier StackGuard; Automatic Adaptive Detection and Prevention of Buffer-Overflow Attacks Todd Miller Theo de Raadt strlcpy and strlcat -- consistent, safe string copy and concatenation. diff --git a/en_US.ISO8859-1/books/developers-handbook/chapters.ent b/en_US.ISO8859-1/books/developers-handbook/chapters.ent index aca3e3bb3c..6e2d0c3f6f 100644 --- a/en_US.ISO8859-1/books/developers-handbook/chapters.ent +++ b/en_US.ISO8859-1/books/developers-handbook/chapters.ent @@ -1,61 +1,61 @@ - + diff --git a/en_US.ISO8859-1/books/developers-handbook/x86/chapter.sgml b/en_US.ISO8859-1/books/developers-handbook/x86/chapter.sgml new file mode 100644 index 0000000000..8f60c7db9a --- /dev/null +++ b/en_US.ISO8859-1/books/developers-handbook/x86/chapter.sgml @@ -0,0 +1,2904 @@ + + + + + x86 Assembly Language Programming + + + This chapter written by G. Adam Stanislav. + Whiz Kid Technomagic + Modifications for the Handbook made by &a.wylie;, &.logo;, + and &a.murray; + + + Synopsis + + Assembly language programing under Unix is highly + undocumented. It is generally assumed that no one would ever + want to use it because various Unix systems run on different + microprocessors, so everything should be written in C for + portability. + + + In reality, C portability is quite a myth. Even C programs need + to be modified when ported from one Unix to another, regardless + of what processor each runs on. Typically, such a program is + full of conditional statements depending on the system it is + compiled for. + + + Even if we believe that all of Unix software should be written + in C, or some other high-level language, we still need assembly + language programmers: Who else would write the section of C + library that accesses the kernel? + + + In this tutorial, which is quite brief at this time, I will + attempt to show you how you can use assembly language writing + Unix programs, specifically under FreeBSD. I hope to turn it + into a complete course of FreeBSD assembly language + eventually. + + + This tutorial does not explain the basics of assembly + language. There are enough resources about that (for a complete + online course in assembly language, see Randall Hyde's Art of Assembly + Language; or if you prefer a printed book, take a look + at Jeff Duntemann's Assembly Language Step-by-Step. However, + once the tutorial is finished, any assembly language programmer + will be able to write programs for FreeBSD quickly and + efficiently. + + + Copyright © 2000 G. Adam Stanislav.All rights reserved. + + + + The Tools + + + The Assembler + + + The most important tool for assembly language programming is + the assembler, the software that converts assembly language + code into machine language. + + + Two very different assemblers are available for FreeBSD. One + is &man.as.1;, which uses the traditional Unix assembly + language syntax. It comes with the system. + + + The other is /usr/ports/devel/nasm. It + uses the Intel syntax. Its main advantage is that it can + assemble code for many operating systems. It needs to be + installed separately, but is completely free. + + + This tutorial uses nasm syntax because + most assembly language programmers coming to FreeBSD from + other operating systems will find it easier to + understand. And, because, quite frankly, that is what I am + used to. + + + + + The Linker + + + The output of the assembler, like that of any compiler, needs + to be linked to form an executable file. + + + The standard &man.ld.1; linker comes with FreeBSD. It works + with the code assembled with either assembler. + + + + + + System Calls + + + Default Calling Convention + + + By default, the FreeBSD kernel uses the C calling + convention. Further, although the kernel is accessed using int + 80h, it is assumed the program will call a function that + issues int 80h, rather than issuing int 80h directly. + + + This convention is very convenient, and quite superior to the + Microsoft convention used by MS DOS. Why? Because the Unix + convention allows any program written in any language to + access the kernel. + + + An assembly language program can do that as well. For example, + we could open a file: + + + kernel: + int 80h ; Call kernel + ret + + open: + push dword mode + push dword flags + push dword path + mov eax, 5 + call kernel + add esp, byte 12 + ret + + + + This is a very clean and portable way of coding. If you need + to port the code to a Unix system which uses a different + interrupt, or a different way of passing parameters, all you + need to change is the kernel procedure. + + + But assembly language programmers like to shave off + cycles. The above example requires a + call/ret combination. We can eliminate it + by pushing an extra dword: + + + open: + push dword mode + push dword flags + push dword path + mov eax, 5 + push eax ; Or any other dword + int 80h + add esp, byte 16 + + + + The 5 that we have placed in + EAX identifies the kernel function, in + this case open. + + + + + Alternate Calling Convention + + FreeBSD is an extremely flexible system. It offers other ways + of calling the kernel. For it to work, however, the system + must have Linux emulation installed. + + + Linux is a Unix-like system. However, its kernel uses the + Microsoft system-call convention of passing parameters in + registers. As with the Unix convention, the function number is + placed in EAX. The parameters, however, + are not passed on the stack but EBX, ECX, EDX, ESI, + EDI, EBP: + + + open: + mov eax, 5 + mov ebx, path + mov ecx, flags + mov edx, mode + int 80h + + + + This convention has a great disadvantage over the Unix way, at + least as far as assembly language programming is concerned: + Every time you make a kernel call you must + push the registers, then + pop them later. This makes your code + bulkier and slower. Nevertheless, FreeBSD gives you a + choice. + + + If you do choose the Microsoft/Linux convention, you must let + the system know about it. After your program is assembled and + linked, you need to brand the executable: + + &prompt.user; brandelf -f Linux filename + + + + + + Which convention should you use? + + If you are coding specifically for FreeBSD, you should always + use the Unix convention: It is faster, you can store global + variables in registers, you do not have to brand the + executable, and you do not impose the installation of the + Linux emulation package on the target system. + + + If you want to create portable code that can also run on + Linux, you will probably still want to give the FreeBSD users + as efficient a code as possible. I will show you how you can + accomplish that after I have explained the basics. + + + + + Call Numbers + + To tell the kernel which system service you are calling, + place its number in EAX. Of course, you + need to know what the number is. + + + The Syscalls File + + The numbers are listed in + syscalls. locate + syscalls finds this file in several different + formats, all produced automatically from + syscalls.master. + + You can find the master file for the default Unix + calling convention in + /usr/src/sys/kern/syscalls.master. If + you need to use the Microsoft convention implemented in the + Linux emulation mode, read + /usr/src/sys/i386/linux/syscalls.master. + + N.B.: Not only do FreeBSD and Linux + use different calling conventions, they sometimes use + different numbers for the same functions. + + syscalls.master describes how the + call is to be made: + + 0 STD NOHIDE { int nosys(void); } syscall nosys_args int + 1 STD NOHIDE { void exit(int rval); } exit rexit_args void + 2 STD POSIX { int fork(void); } + 3 STD POSIX { ssize_t read(int fd, void *buf, size_t nbyte); } + 4 STD POSIX { ssize_t write(int fd, const void *buf, size_t nbyte); } + 5 STD POSIX { int open(char *path, int flags, int mode); } + 6 STD POSIX { int close(int fd); } + etc... + + It is the leftmost column that tells us the number to + place in EAX. + + The rightmost column tells us what parameters to + push. They are + pushed from right to left. + + EXAMPLE 3.1: For example, to + open a file, we need to + push the mode + first, then flags, then the address at + which the path is stored. + + + + + + + + Return Values + + A system call would not be useful most of the time if it did not + return some kind of a value: The file descriptor of an open + file, the number of bytes read to a buffer, the system time, + etc. + + + Additionally, the system needs to inform us if an error occurs: + A file does not exist, system resources are exhausted, we passed + an invalid parameter, etc. + + + man-pages + + The traditional place to look for information about various + system calls under Unix systems are the man pages. FreeBSD + describes its system calls in section 2, sometimes in section + 3. + + + For example, open(2) says: + + + If successful, open() returns a + non-negative integer, termed a file descriptor. It returns + -1 on failure, and sets + errno to indicate the error. + + + The assembly language programmer new to Unix and FreeBSD will + immediately ask the puzzling question: Where is + errno and how do I get to it? + + + N.B.: The information presented in the + man pages applies to C programs. The assembly language + programmer needs additional information. + + + + + Where are the return values? + + Unfortunately, it depends... For most system calls it is in + EAX, but not for all. A good rule of + thumb, when working with a system call for the first time, + look for the return value in EAX. If it + is not there, you need further research. + + + N.B.: I am aware of one system call that + returns the value in EDX: + SYS_fork. All others I have worked with + use EAX. But I have not worked with them + all yet. + + + TIP: If you cannot find the answer here + or anywhere else, study libc source code and see how it + interfaces with the kernel. + + + + + Where is <function>errno</function>>? + Actually, nowhere... + + + errno is part of the C language, not the + Unix kernel. When accessing kernel services directly, the + error code is returned in EAX, the same + register the proper return value generally ends up in. + + + This makes perfect sense. If there is no error, there is no + error code. If there is an error, there is no return + value. One register can contain either. + + + + + Determining an Error Occurred + + When using the standard FreeBSD calling convention, the + carry flag is cleared upon success, set + upon failure. + + + When using the Linux emulation mode, the signed value in + EAX is non-negative upon success, and + contains the return value. In case of an error, the value is + negative, i.e., -errno. + + + + + + + Creating Portable Code + + Portability is generally not one of the strengths of assembly + language. Yet, writing assembly language programs for different + platforms is possible, especially with + nasm. I have written assembly language + libraries that can be assembled for such different operating + systems as Windows and FreeBSD. + + + It is all the more possible when you want your code to run on + two platforms which, while different, are based on similar + architectures. + + + For example, FreeBSD is Unix, Linux is Unix-like. I only + mentioned three differences between them (from an assembly + language programmerâs perspective): The calling convention, the + function numbers, and the way of returning values. + + + Dealing With Function Numbers + + In many cases the function numbers are the same. However, even + when they are not, the problem is easy to deal with: Instead + of using numbers in your code, use constants which you have + declared differently depending on the target + architecture: + + + %ifdef LINUX + %define SYS_execve 11 + %else + %define SYS_execve 59 + %endif + + + + + Dealing With Conventions + + Both, the calling convention, and the return value (the + errno problem) can be resolved with + macros: + + + + %ifdef LINUX + + %macro system 0 + call kernel + %endmacro + + align 4 + kernel: + push ebx + push ecx + push edx + push esi + push edi + push ebp + + mov ebx, [esp+32] + mov ecx, [esp+36] + mov edx, [esp+40] + mov esi, [esp+44] + mov ebp, [esp+48] + int 80h + + pop ebp + pop edi + pop esi + pop edx + pop ecx + pop ebx + + or eax, eax + js .errno + clc + ret + + .errno: + neg eax + stc + ret + + %else + + %macro system 0 + int 80h + %endmacro + + %endif + + + + + + Dealing With Other Portability Issues + + The above solutions can handle most cases of writing code + portable between FreeBSD and Linux. Nevertheless, with some + kernel services the differences are deeper. + + + In that case, you need to write two different handlers for + those particular system calls, and use conditional + assembly. Luckily, most of your code does something other than + calling the kernel, so usually you will only need a few such + conditional sections in your code. + + + + + Using a Portable Library + + You can avoid portability issues in your main code altogether + by writing a library of system calls. Create a separate + library for FreeBSD, a different one for Linux, and yet other + libraries for more operating systems. + + + In your library, write a separate function (or procedure, if + you prefer the traditional assembly language terminology) for + each system call. Use the C calling convention of passing + parameters. But still use EAX to pass + the call number in. In that case, your FreeBSD library can be + very simple, as many seemingly different functions can be just + labels to the same code: + + + sys.open: + sys.close: + [etc...] + int 80h + ret + + + + Your Linux library will require more different functions. But + even here you can group system calls using the same number of + parameters: + + + sys.exit: + sys.close: + [etc... one-parameter functions] + push ebx + mov ebx, [esp+12] + int 80h + pop ebx + jmp sys.return + + ... + + sys.return: + or eax, eax + js sys.err + clc + ret + + sys.err: + neg eax + stc + ret + + + + The library approach may seem inconvenient at first because it + requires you to produce a separate file your code depends + on. But it has many advantages: For one, you only need to + write it once and can use it for all your programs. You can + even let other assembly language programmers use it, or + perhaps use one written by someone else. But perhaps the + greatest advantage of the library is that your code can be + ported to other systems, even by other programmers, by simply + writing a new library without any changes to your code. + + + If you do not like the idea of having a library, you can at + least place all your system calls in a separate assembly + language file and link it with your main program. Here, again, + all porters have to do is create a new object file to link + with your main program. + + + + + Using a Portable Include File + + If you are releasing your software as (or with) source code, + you can use macros and place them in a separate file, which + you include in your code. + + + Porters of your software will simply write a new include + file. No library or external object file is necessary, yet + your code is portable without any need to edit the + code. + + + N.B.: This is the approach we will use + throughout this tutorial. We will name our include file + system.inc, and add to it whenever we + deal with a new system call. + + + We can start our system.inc by declaring + the standard file descriptors: + + + %define stdin 0 + %define stdout 1 + %define stderr 2 + + + + Next, we create a symbolic name for each system call: + + + %define SYS_nosys 0 + %define SYS_exit 1 + %define SYS_fork 2 + %define SYS_read 3 + %define SYS_write 4 + ; [etc...] + + + + We add a short, non-global procedure with a long name, so we + do not accidentally reuse the name in our code: + + + section .code + align 4 + access.the.bsd.kernel: + int 80h + ret + + + + We create a macro which takes one argument, the syscall + number: + + + %macro system 1 + mov eax, %1 + call access.the.bsd.kernel + %endmacro + + + + Finally, we create macros for each syscall. These macros take + no arguments. + + + %macro sys.exit 0 + system SYS_exit + %endmacro + + %macro sys.fork 0 + system SYS_fork + %endmacro + + %macro sys.read 0 + system SYS_read + %endmacro + + %macro sys.write 0 + system SYS_write + %endmacro + + ; [etc...] + + + + Go ahead, enter it into your editor and save it as + system.inc. We will add more to it as we + discuss more syscalls. + + + + + + + Our First Program + + We are now ready for our first program, the mandatory + Hello, World! + + + 1: %include 'system.inc' + 2: + 3: section .data + 4: hello db 'Hello, World!', 0Ah + 5: hbytes equ $-hello + 6: + 7: section .code + 8: global _start + 9: _start: + 10: push dword hbytes + 11: push dword hello + 12: push dword stdout + 13: sys.write + 14: + 15: push dword 0 + 16: sys.exit + + + + Here is what it does: Line 1 includes the defines, the macros, + and the code from system.inc. + + + Lines 3-5 are the data: Line 3 starts the data + section/segment. Line 4 contains the string "Hello, World!" + followed by a new line (0Ah). Line 5 creates a constant that + contains the length of the string from line 4 in bytes. + + + Lines 7-16 contain the code. Note that FreeBSD uses the elf file + format for its executables, which requires every program to + start at the point labeled _start (or, more precisely, the + linker expects that). This label has to be global. + + + Lines 10-13 ask the system to write hbytes + bytes of the hello string to + stdout. + + + Lines 15-16 ask the system to end the program with the return + value of 0. The + SYS_exit syscall never returns, so the code + ends there. + + + N.B.: If you have come to Unix from MS DOS + assembly language background, you may be used to writing + directly to the video hardware. You will never have to worry + about this in FreeBSD, or any other flavor of Unix. As far as + you are concerned, you are writing to a file known as + stdout. This can be the video screen, or a + Telnet terminal, or an actual file, or even the input of another + program. Which it is, is for the system to figure out. + + + Assembling the Code + + Type the code (except the line numbers) in an editor, and + save it in a file named hello.asm. You + need nasm to assemble it. + + + Installing NASM + + If you do not have nasm, type: + + &prompt.user; su +Password: your root password +&prompt.root; cd /usr/ports/devel/nasm +&prompt.root; make install +&prompt.root; exit +&prompt.user; + + &prompt.user; su +Password: your root password +&prompt.root; cd /usr/ports/devel/nasm +&prompt.root; make install +&prompt.root; exit +&prompt.user; + + You may type make install clean + instead of just make install if you do + not want to keep nasm source + code. + + Either way, FreeBSD will automatically download + nasm from the Internet, compile it, and + install it on your system. + + N.B.: If your system is not + FreeBSD, you need to get nasm from its + home + page. You can still use it to assemble FreeBSD + code. + + + + + + Assemble, link and run + + Now you can assemble, link, and run the code: + + &prompt.user; nasm -f elf hello.asm +&prompt.user; ld -s -o hello hello.o +&prompt.user; ./hello +Hello, World! +&prompt.user; + + + + + + Writing Unix Filters + + A common type of Unix application is a filterâa program that + reads data from the stdin, processes it + somehow, then writes the result to + stdout. + + In this chapter, we shall develop a simple filter, and learn + how to read from stdin and write to + stdout. This filter will convert each byte + of its input into a hexadecimal number followed by a blank + space. + + %include 'system.inc' + + section .data + hex db '0123456789ABCDEF' + buffer db 0, 0, ' ' + + section .code + global _start + _start: + ; read a byte from stdin + push dword 1 + push dword buffer + push dword stdin + sys.read + add esp, byte 12 + or eax, eax + je .done + + ; convert it to hex + movzx eax, byte [buffer] + mov edx, eax + shr dl, 4 + mov dl, [hex+edx] + mov [buffer], dl + and al, 0Fh + mov al, [hex+eax] + mov [buffer+1], al + + ; print it + push dword 3 + push dword buffer + push dword stdout + sys.write + add esp, byte 12 + jmp short _start + + .done : + push dword 0 + sys.exit + + In the data section we create an array called + hex. It contains the 16 hexadecimal digits + in ascending order. The array is followed by a buffer which we + will use for both input and output. The first two bytes of the + buffer are initially set to 0. This is + where we will write the two hexadecimal digits (the first byte + also is where we will read the input). The third byte is a + space. + + The code section consists of four parts: Reading the byte, + converting it to a hexadecimal number, writing the result, and + eventually exiting the program. + + To read the byte, we ask the system to read one byte from + stdin, and store it in the first byte of + the buffer. The system returns the number + of bytes read in EAX. This will be + 1 while data is coming, or + 0, when no more input data is + available. Therefore, we check the value of + EAX. If it is 0, we + jump to .done, otherwise we + continue. + + N.B.: For simplicity sake, we are + ignoring the possibility of an error condition at this + time. + + The hexadecimal conversion reads the byte from the + buffer into EAX, or + actually just AL, while clearing the + remaining bits of EAX to zeros. We also + copy the byte to EDX because we need to + convert the upper four bits (nibble) separately from the lower + four bits. We store the result in the first two bytes of the + buffer. + + Next, we ask the system to write the three bytes of the + buffer, i.e., the two hexadecimal digits and the blank space, to + stdout. We then jump back to the beginning + of the program and process the next byte. + + Once there is no more input left, we ask the system to exit + our program, returning a zero, which is the traditional value + meaning the program was successful. + + Go ahead, and save the code in a file named + hex.asm, then type the following (the + ^D means press the control key and type + D while holding the control key + down): + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + N.B.: If you are migrating to Unix from + MS DOS, you may be wondering why each line ends with + 0A instead of 0D + 0A. This is because Unix does not use the cr/lf + convention, but a ânew lineâ convention, which is + 0A in hexadecimal. + + Can we improve this? Well, for one, it is a bit confusing + because once we have converted a line of text, our input no + longer starts at the begining of the line. We can modify it to + print a new line instead of a space after each + 0A: + + %include 'system.inc' + +section .data +hex db '0123456789ABCDEF' +buffer db 0, 0, ' ' + +section .code +global _start +_start: + mov cl, ' ' + +.loop: + ; read a byte from stdin + push dword 1 + push dword buffer + push dword stdin + sys.read + add esp, byte 12 + or eax, eax + je .done + + ; convert it to hex + movzx eax, byte [buffer] + mov [buffer+2], cl + cmp al, 0Ah + jne .hex + mov [buffer+2], al + +.hex: + mov edx, eax + shr dl, 4 + mov dl, [hex+edx] + mov [buffer], dl + and al, 0Fh + mov al, [hex+eax] + mov [buffer+1], al + + ; print it + push dword 3 + push dword buffer + push dword stdout + sys.write + add esp, byte 12 + jmp short .loop + +.done: + push dword 0 + sys.exit + + We have stored the space in the CL + register. We can do this safely because, unlike Microsoft + Windows, Unix system calls do not modify the value of any + register they do not use to return a value in. + + That means we only need to set CL + once. We have, therefore, added a new label + .loop and jump to it for the next byte + instead of jumping at _start. We have also + added the .hex label so we can either have + a blank space or a new line as the third byte of the + buffer. + + Once you have changed hex.asm to + reflect these changes, type: + + &prompt.user nasm -f elf hex.asm +&prompt.user ld -s -o hex hex.o +&prompt.user ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + That looks better. But this code is quite inefficient! We + are making a system call for every single byte twice (once to + read it, another time to write the output). + + + + + Buffered Input and Output + + We can improve the efficiency of our code by buffering our + input and output. We create an input buffer and read a whole + sequence of bytes at one time. Then we fetch them one by one + from the buffer. + + We also create an output buffer. We store our output in it + until it is full. At that time we ask the kernel to write the + contents of the buffer to stdout. + + The program ends when there is no more input. But we still + need to ask the kernel to write the contents of our output + buffer to stdout one last time, otherwise + some of our output would make it to the output buffer, but never + be sent out. Do not forget that, or you will be wondering why + some of your output is missing. + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +global _start +_start: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword stdin + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword stdout + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + We now have a third section in the source code, named + .bss. This section is not included in our + executable file, and, therefore, cannot be initialized. We use + resb instead of db. It + simply reserves the requested size of uninitialized memory for + our use. + + We take advantage of the fact that the system does not + modify the registers: We use registers for what, otherwise, + would have to be global variables stored in the + .data section. This is also why the Unix + convention of passing parameters to system calls on the stack is + superior to the Microsoft convention of passing them in the + registers: We can keep the registers for our own use. + + We use EDI and ESI + as pointers to the next byte to be read from or written to. We + use EBX and ECX to + keep count of the number of bytes in the two buffers, so we know + when to dump the output to, or read more input from, the + system. + + Let us see how it works now: + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +Here I come! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + Not what you expected? The program did not print the output + until we pressed ^D. That is easy to fix by + inserting three lines of code to write the output every time we + have converted a new line to 0A. I have + marked the three lines with > (do not copy the > in your + hex.asm + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +global _start +_start: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar +> cmp al, 0Ah +> jne .loop +> call write + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword stdin + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword stdout + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + Now, let us see how it works: + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + Not bad for a 644-byte executable, is it! + + + How to Unread a Character + + WARNING: This may be a somewhat advanced + topic, mostly of interest to programmers familiar with the + theory of compilers. If you wish, you may skip to the next chapter, and + perhaps read this later. + + While our sample program does not require it, more + sophisticated filters often need to look ahead. In other + words, they may need to see what the next character is (or + even several characters). If the next character is of a + certain value, it is part of the token currently being + processed. Otherwise, it is not. + + For example, you may be parsing the input stream for a + textual string (e.g., when implementing a language compiler): + If a character is followed by another character, or perhaps a + digit, it is part of the token you are processing. If it is + followed by white space, or some other value, then it is not + part of the current token. + + This presents an interesting problem: How to return the + next character back to the input stream, so it can be read + again later? + + One possible solution is to store it in a character + variable, then set a flag. We can modify + getchar to check the flag, and if it is + set, fetch the byte from that variable instead of the input + buffer, and reset the flag. But, of course, that slows us + down. + + The C language has an ungetc() + function, just for that purpose. Is there a quick way to + implement it in our code? I would like you to scroll back up + and take a look at the getchar procedure + and see if you can find a nice and fast solution before + reading the next paragraph. Then come back here and see my own + solution. + + The key to returning a character back to the stream is in + how we are getting the characters to start with: + + First we check if the buffer is empty by testing the value + of EBX. If it is zero, we call the + read procedure. + + If we do have a character available, we use + lodsb, then decrease the value of + EBX. The lodsb + instruction is effectively identical to: + + mov al, [esi] + inc esi + + The byte we have fetched remains in the buffer until the + next time read is called. We do not know + when that happens, but we do know it will not happen until the + next call to getchar. Hence, to âreturnâ + the last-read byte back to the stream, all we have to do is + decrease the value of ESI and increase + the value of EBX: + + ungetc: + dec esi + inc ebx + ret + + But, be careful! We are perfectly safe doing this if our + look-ahead is at most one character at a time. If we are + examining more than one upcoming character and call + ungetc several times in a row, it will + work most of the time, but not all the time (and will be tough + to debug). Why? + + Because as long as getchar does not + have to call read all of the pre-read + bytes are still in the buffer, and our + ungetc works without a glitch. But the + moment getchar calls + read, the contents of the buffer + change. + + We can always rely on ungetc working + properly on the last character we have read with + getchar, but not on anything we have read + before that. + + If your program reads more than one byte ahead, you have + at least two choices: + + If possible, modify the program so it only reads one byte + ahead. This is the simplest solution. + + If that option is not available, first of all determine + the maximum number of characters your program needs to return + to the input stream at one time. Increase that number + slightly, just to be sure, preferably to a multiple of 16âso + it aligns nicely. Then modify the .bss + section of your code, and create a small âspareâ buffer right + before your input buffer, something like this: + + section .bss + resb 16 ; or whatever the value you came up with +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + + You also need to modify your ungetc + to pass the value of the byte to unget in + AL: + + ungetc: + dec esi + inc ebx + mov [esi], al + ret + + With this modification, you can call + ungetc up to 17 times in a row safely + (the first call will still be within the buffer, the remaining + 16 may be either within the buffer or within the + âspareâ). + + + + + + Command-line Arguments + + Our hex program will be more useful if + it can read the names of an input and output file from its + command line, i.e., if it can process the command line + arguments. But... Where are they? + + Before a Unix system starts a program, it pushes some data + on the stack, then jumps at the _start + label of the program. Yes, I said jumps, not calls. That means + the data can be accessed by reading + [esp+offset], or by simply + popping it. + + The value at the top of the stack contains the number of + command line arguments. It is traditionally called + argc, for âargument count.â + + Command line arguments follow next, all + argc of them. These are typically referred + to as argv, for âargument value(s).â That + is, we get argv[0], + argv[1], ..., + argv[argc-1]. These are not the actual + arguments, but pointers to arguments, i.e., memory addresses of + the actual arguments. The arguments themselves are + NUL-terminated character strings. + + The argv list is followed by a + NULL pointer, which is simply a + 0. There is more, but this is enough for + our purposes right now. + + N.B.: If you have come from the MS DOS + programming environment, the main difference is that each + argument is in a separate string. The second difference is that + there is no practical limit on how many arguments there can + be. + + Armed with this knowledge, we are almost ready for the next + version of hex.asm. First, however, we need + to add a few lines to system.inc: + + First, we need to add two new entries to our list of system + call numbers: + + %define SYS_open 5 +%define SYS_close 6 + + Then we add two new macros at the end of the file: + + %macro sys.open 0 + system SYS_open +%endmacro + +%macro sys.close 0 + system SYS_close +%endmacro + + Here, then, is our modified source code: + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +fd.in dd stdin +fd.out dd stdout +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +align 4 +err: + push dword 1 ; return failure + sys.exit + +align 4 +global _start +_start: + add esp, byte 8 ; discard argc and argv[0] + + pop ecx + jecxz .init ; no more arguments + + ; ECX contains the path to input file + push dword 0 ; O_RDONLY + push ecx + sys.open + jc err ; open failed + + add esp, byte 8 + mov [fd.in], eax + + pop ecx + jecxz .init ; no more arguments + + ; ECX contains the path to output file + push dword 420 ; file mode (644 octal) + push dword 0200h | 0400h | 01h + ; O_CREAT | O_TRUNC | O_WRONLY + push ecx + sys.open + jc err + + add esp, byte 12 + mov [fd.out], eax + +.init: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from input file or stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar + cmp al, dl + jne .loop + call write + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword [fd.in] + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + + ; close files + push dword [fd.in] + sys.close + + push dword [fd.out] + sys.close + + ; return success + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword [fd.out] + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + In our .data section we now have two + new variables, fd.in and + fd.out. We store the input and output file + descriptors here. + + In the .code section we have replaced + the references to stdin and + stdout with [fd.in] + and [fd.out]. + + The .code section now starts with a + simple error handler, which does nothing but exit the program + with a return value of 1. The error handler + is before _start so we are within a short + distance from where the errors occur. + + Naturally, the program execution still begins at + _start. First, we remove + argc and argv[0] from + the stack: They are of no interest to us (in this program, that + is). + + We pop argv[1] to + ECX. This register is particularly suited + for pointers, as we can handle NULL + pointers with jecxz. If + argv[1] is not NULL, + we try to open the file named in the first argument. Otherwise, + we continue the program as before: Reading from + stdin, writing to + stdout. If we fail to open the input file + (e.g., it does not exist), we jump to the error handler and + quit. + + If all went well, we now check for the second argument. If + it is there, we open the output file. Otherwise, we send the + output to stdout. If we fail to open the + output file (e.g., it exists and we do not have the write + permission), we, again, jump to the error handler. + + The rest of the code is the same as before, except we close + the input and output files before exiting, and, as mentioned, we + use [fd.in] and + [fd.out]. + + Our executable is now a whopping 768 bytes long. + + Can we still improve it? Of course! Every program can be + improved. Here are a few ideas of what we could do: + + + Have our error handler print a message to + stderr. + + Add error handlers to the + read and write + functions. + + Close stdin when we open an + input file, stdout when we open an output + file. + + Add command line switches, such as + -i and -o, so we can + list the input and output files in any order, or perhaps read + from stdin and write to a + file. + + Print a usage message if command line arguments + are incorrect. + + + + I shall leave these enhancements as an exercise to the + reader: You already know everything you need to know to implement + them. + + + + Unix Environment + + + An important Unix concept is the environment, which is defined + by environment variables. Some are set by + the system, others by you, yet others by the + shell, or any program that loads another + program. + + + How to Find Environment Variables + + I said earlier that when a program starts executing, the stack + contains argc followed by the + NULL-terminated argv + array, followed by something else. The âsomething elseâ is the + environment, or, to be more precise, a + NULL-terminated array of pointers to + environment variables. This is often + referred to as env. + + + The structure of env is the same as that + of argv, a list of memory addresses + followed by a NULL + (0). In this case, there is no + âenvcââwe figure out where the array ends + by searching for the final NULL. + + + The variables usually come in the + name=value format, but sometimes the + =value part may be missing. We need to account for that + possibility. + + + + Webvar + + + I could just show you some code that prints the environment + the same way the Unix env command + does. But I thought it would be more interesting to write a + simple assembly language CGI utility. + + + CGI: A Quick Overview + + 10.2.1. CGI: A Quick Overview + + + I have a detailed + CGI tutorial on my web site, but here is a very + quick overview of CGI: + + + + The web server communicates with the CGI + program by setting environment + variables. + + The CGI program sends its output to + stdout. The web server reads it from + there. + + It must start with an HTTP header followed + by two blank lines. + + It then prints the HTML code, or whatever + other type of data it is producing. + + + + + N.B.: While certain + environment variables use standard + names, others vary, depending on the web server. That makes + webvars quite a useful diagnostic + tool. + + + + + Webvar continued... + + + Our webvar program, then, must send out + the HTTP header followed by some HTML mark-up. It then must + read the environment variable one by one + and send them out as part of the HTML page. + + + The code follows. I placed comments and explanations right + inside the code: + + +;;;;;;; webvars.asm ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; +; Copyright (c) 2000 G. Adam Stanislav +; All rights reserved. +; +; Redistribution and use in source and binary forms, with or without +; modification, are permitted provided that the following conditions +; are met: +; 1. Redistributions of source code must retain the above copyright +; notice, this list of conditions and the following disclaimer. +; 2. Redistributions in binary form must reproduce the above copyright +; notice, this list of conditions and the following disclaimer in the +; documentation and/or other materials provided with the distribution. +; +; THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND +; ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +; IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +; ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE +; FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +; DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +; OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +; HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +; LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +; OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +; SUCH DAMAGE. +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; +; Version 1.0 +; +; Started: 8-Dec-2000 +; Updated: 8-Dec-2000 +; +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +%include 'system.inc' + +section .data +http db 'Content-type: text/html', 0Ah, 0Ah + db '<?xml version="1.0" encoding="UTF-8"?>', 0Ah + db '<!DOCTYPE html PUBLIC "-//W3C/DTD XHTML Strict//EN" ' + db '"DTD/xhtml1-strict.dtd">', 0Ah + db '<html xmlns="http://www.w3.org/1999/xhtml" ' + db 'xml.lang="en" lang="en">', 0Ah + db '<head>', 0Ah + db '<title>Web Environment</title>', 0Ah + db '<meta name="author" content="G. Adam Stanislav" />', 0Ah + db '</head>', 0Ah, 0Ah + db '<body bgcolor="#ffffff" text="#000000" link="#0000ff" ' + db 'vlink="#840084" alink="#0000ff">', 0Ah + db '<div class="webvars">', 0Ah + db '<h1>Web Environment</h1>', 0Ah + db '<p>The following <b>environment variables</b> are defined ' + db 'on this web server:</p>', 0Ah, 0Ah + db '<table align="center" width="80" border="0" cellpadding="10" ' + db 'cellspacing="0" class="webvars">', 0Ah +httplen equ $-http +left db '<tr>', 0Ah + db '<td class="name"><tt>' +leftlen equ $-left +middle db '</tt></td>', 0Ah + db '<td class="value"><tt><b>' +midlen equ $-middle +undef db '<i>(undefined)</i>' +undeflen equ $-undef +right db '</b></tt></td>', 0Ah + db '</tr>', 0Ah +rightlen equ $-right +wrap db '</table>', 0Ah + db '</div>', 0Ah + db '</body>', 0Ah + db '</html>', 0Ah, 0Ah +wraplen equ $-wrap + +section .code +global _start +_start: + ; First, send out all the http and xhtml stuff that is + ; needed before we start showing the environment + push dword httplen + push dword http + push dword stdout + sys.write + + ; Now find how far on the stack the environment pointers + ; are. We have 12 bytes we have pushed before "argc" + mov eax, [esp+12] + + ; We need to remove the following from the stack: + ; + ; The 12 bytes we pushed for sys.write + ; The 4 bytes of argc + ; The EAX*4 bytes of argv + ; The 4 bytes of the NULL after argv + ; + ; Total: + ; 20 + eax * 4 + ; + ; Because stack grows down, we need to ADD that many bytes + ; to ESP. + lea esp, [esp+20+eax*4] + cld ; This should already be the case, but let's be sure. + + ; Loop through the environment, printing it out +.loop: + pop edi + or edi, edi ; Done yet? + je near .wrap + + ; Print the left part of HTML + push dword leftlen + push dword left + push dword stdout + sys.write + + ; It may be tempting to search for the '=' in the env string next. + ; But it is possible there is no '=', so we search for the + ; terminating NUL first. + mov esi, edi ; Save start of string + sub ecx, ecx + not ecx ; ECX = FFFFFFFF + sub eax, eax +repne scasb + not ecx ; ECX = string length + 1 + mov ebx, ecx ; Save it in EBX + + ; Now is the time to find '=' + mov edi, esi ; Start of string + mov al, '=' +repne scasb + not ecx + add ecx, ebx ; Length of name + + push ecx + push esi + push dword stdout + sys.write + + ; Print the middle part of HTML table code + push dword midlen + push dword middle + push dword stdout + sys.write + + ; Find the length of the value + not ecx + lea ebx, [ebx+ecx-1] + + ; Print "undefined" if 0 + or ebx, ebx + jne .value + + mov ebx, undeflen + mov edi, undef + +.value: + push ebx + push edi + push dword stdout + sys.write + + ; Print the right part of the table row + push dword rightlen + push dword right + push dword stdout + sys.write + + ; Get rid of the 60 bytes we have pushed + add esp, byte 60 + + ; Get the next variable + jmp .loop + +.wrap: + ; Print the rest of HTML + push dword wraplen + push dword wrap + push dword stdout + sys.write + + ; Return success + push dword 0 + sys.exit + + + + This code produces a 1,396-byte executable. Most of it is + data, i.e., the HTML mark-up we need to send out. + + + Assemble and link it as usual: + + &prompt.user; nasm -f elf webvars.asm +&prompt.user; ld -s -o webvars webvars.o + + + To use it, you need to upload webvars to + your web server. Depending on how your web server is set up, + you may have to store in a special + cgi-bin directory, or perhaps rename it + with a .cgi extension. + + + Then you need to use your browser to view its output. To see + its output on my web server, please instruct your browser to + go to http://www.int80h.org/webvars/. I + am deliberately not placing a regular link here because I do + not want its output to appear on all the search + engines... + + + + + + Working with Files + + We have already done some basic file work: We know how to open + and close them, how to read and write them using buffers. But + Unix offers much more functionality when it comes to files. We + will examine some of it in this section, and end up with a nice + file conversion utility. + + + Indeed, let us start at the end, that is, with the file + conversion utility. It always makes programming easier when we + know from the start what the end product is supposed to + do. + + + One of the first programs I wrote for Unix was tuc, + a text-to-Unix file converter. It converts a text file from + other operating systems to a Unix text file. In other words, it + changes from different kind of line endings to the newline + convention of Unix. It saves the output in a different + file. Optionally, it converts a Unix text file to a DOS text + file. + + I have used tuc extensively, but always + only to convert from some other OS to Unix, never the other + way. I have always wished it would just overwrite the file + instead of me having to send the output to a different + file. Most of the time, I end up using it like this: + + &prompt.user; tuc myfile tempfile +&prompt.user; mv tempfile myfile + + It would be nice to have a ftuc, i.e., + fast tuc, and use it like this: + + &prompt.user; ftuc myfile + + In this chapter, then, we will write + ftuc in assembly language (the original + tuc is in C), and + study various file-oriented kernel services in the + process. + + At first sight, such a file conversion is very simple: All + you have to do is strip the carriage returns, right? + + If you answered yes, think again: That approach will work + most of the time (at least with MS DOS text files), but will + fail occasionally. + + The problem is that not all non-Unix text files end their + line with the carriage return / line feed sequence. Some use + carriage returns without line feeds. Others combine several + blank lines into a single carriage return followed by several + line feeds. And so on. + + A text file converter, then, must be able to handle any + possible line endings: + + + carriage return / line feed + carriage return + line feed / carriage return + line feed + + + + It should also handle files that use some kind of a combination + of the above (e.g., carriage return followed by several line + feeds). + + + Finite State Machine + + The problem is easily solved by the use of a technique called + finite state machine, originally + developed by the designers of digital electronic circuits. A + finite state machine is a digital circuit + whose output is dependent not only on its input but on its + previous input, i.e., on its state. The microprocessor is an + example of a finite state machine: Our + assembly language code is assembled to machine language in + which some assembly language code produces a single byte of + machine language, while others produce several bytes. As the + microprocessor fetches the bytes from the memory one by one, + some of them simply change its state rather than produce some + output. When all the bytes of the op code are fetched, the + microrpocessor produces some output, or changes the value of a + register, etc. + + + Because of that, all software is essentially a sequence of + state instructions for the microprocessor. Nevertheless, the + concept of finite state machine is useful + in software design as well. + + + Our text file converter can be designed as a finite + state machine with three possible states. We could + call them states 0-2, but it will make our life easier if we + give them symbolic names: + + + ordinary + cr + lf + + + + Our program will start in the ordinary + state. During this state, the program action depends on its + input as follows: + + + If the input is anything other than a carriage + return or line feed, the input is simply passed on to the + output. The state remains unchanged. + + If the input is a carriage return, the state + is changed to cr. The input is then + discarded, i.e., no output is made. + + If the input is a line feed, the state is + changed to lf. The input is then + discarded. + + + + + Whenever we are in the cr state, it is + because the last input was a carriage return, which was + unprocessed. What our software does in this state again + depends on the current input: + + + + If the input is anything other than a carriage return + or line feed, output a line feed, then output the input, + then change the state to + ordinary. + + + If the input is a carriage return, we have received + two (or more) carriage returns in a row. We discard the + input, we output a line feed, and leave the state + unchanged. + + + If the input is a line feed, we output the line feed + and change the state to ordinary. Note + that this is not the same as the first case above â if we + tried to combine them, we would be outputting two line feeds + instead of one. + + + + + Finally, we are in the lf state after we + have received a line feed that was not preceded by a carriage + return. This will happen when our file already is in Unix + format, or whenever several lines in a row are expressed by a + single carriage return followed by several line feeds, or when + line ends with a line feed / carriage return sequence. Here is + how we need to handle our input in this state: + + + + If the input is anything other than a carriage return + or line feed, we output a line feed, then output the input, + then change the state to ordinary. This + is exactly the same action as in the cr + state upon receiving the same kind of + input. + + + If the input is a carriage return, we discard the + input, we output a line feed, then change the state to + ordinary. + + + If the input is a line feed, we output the line feed, + and leave the state unchanged. + + + + + + The Final State + + The above finite state machine works + for the entire file, but leaves the possibility that the + final line end will be ignored. That will happen whenever + the file ends with a single carriage return or a single line + feed. I did not think of it when I wrote + tuc, just to discover that occasionally + it strips the last line ending. + + + This problem is easily fixed by checking the state after the + entire file was processed. If the state is not + ordinary, we simply need to output one + last line feed. + + + N.B.: Now that we have expressed our + algorithm as a finite state machine, we + could easily design a dedicated digital electronic circuit + (a âchipâ) to do the conversion for us. Of course, doing so + would be considerably more expensive than writing an + assembly language program. + + + + + The Output Counter + + Because our file conversion program may be combining two + characters into one, we need to use an output counter. We + initialize it to 0, and increase it every time we send a + character to the output. At the end of the program, the + counter will tell us what size we need to set the file + to. + + + + + + Implementing FSM in Software + + The hardest part of working with a finite state + machine is analyzing the problem and expressing it + as a finite state machine. That + accomplished, the software almost writes itself. + + + In a high-level language, such as C, there are several main + approaches. One is to use a switch + statement which chooses what function should be run. For + example, + + + switch (state) { + default: + case REGULAR: + regular(inputchar); + break; + case CR: + cr(inputchar); + break; + case LF: + lf(inputchar); + break; + } + + + + Another approach is by using an array of function pointers, + something like this: + + + (output[state])(inputchar); + + + + Yet another is to have state be a + function pointer, set to point at the appropriate + function: + + + (*state)(inputchar); + + + + This is the approach we will use in our program because it is + very easy to do in assembly language, and very fast, too. We + will simply keep the address of the right procedure in + EBX, and then just issue: + + + call ebx + + + + This is possibly faster than hardcoding the address in the + code because the microprocessor does not have to fetch the + address from the memoryâit is already stored in one of its + registers. I said possibly because with + the caching modern microprocessors do, either way may be + equally fast. + + + + + Memory Mapped Files + + Because our program works on a single file, we cannot use the + approach that worked for us before, i.e., to read from an + input file and to write to an output file. + + + Unix allows us to map a file, or a section of a file, into + memory. To do that, we first need to open the file with the + appropriate read/write flags. Then we use the + mmap system call to map it into the + memory. One nice thing about mmap is that + it automatically works with virtual memory: We can map more of + the file into the memory than we have physical memory + available, yet still access it through regular memory op + codes, such as mov, + lods, and + stos. Whatever changes we make to the + memory image of the file will be written to the file by the + system. We do not even have to keep the file open: As long as + it stays mapped, we can read from it and write to it. + + + The 32-bit Intel microprocessors can access up to four + gigabytes of memory â physical or virtual. The FreeBSD system + allows us to use up to a half of it for file mapping. + + + For simplicity sake, in this tutorial we will only convert + files that can be mapped into the memory in their + entirety. There are probably not too many text files that + exceed two gigabytes in size. If our program encounters one, + it will simply display a message suggesting we use the + original tuc instead. + + + If you examine your copy of + syscalls.master, you will find two + separate syscalls named mmap. This is + because of evolution of Unix: There was the traditional BSD + mmap, syscall 71. That one was superceded + by the POSIX mmap, syscall 197. The + FreeBSD system supports both because older programs were + written by using the original BSD version. But new software + uses the POSIX version, which is what we will use. + + + The syscalls.master file lists the POSIX + version like this: + + +197 STD BSD { caddr_t mmap(caddr_t addr, size_t len, int prot, \ + int flags, int fd, long pad, off_t pos); } + + + + This differs slightly from what mmap(2) + says. That is because mmap(2) describes + the C version. + + + The difference is in the long pad + argument, which is not present in the C version. However, the + FreeBSD syscalls add a 32-bit pad after pushing a 64-bit + argument. In this case, off_t is a 64-bit + value. + + + When we are finished working with a memory-mapped file, we + unmap it with the munmap syscall: + + + TIP: For an in-depth treatment of + mmap, see W. Richard Stevensâ Unix + Network Programming, Volume 2, Chapter 12. + + + + + Determining File size + + Because we need to tell mmap how many + bytes of the file to map into the memory, and because we want + to map the entire file, we need toq determine the size of the + file. + + + We can use the fstat syscall to get all + the information about an open file that the system can give + us. That includes the file size. + + + Again, syscalls.master lists two versions + of fstat, a traditional one (syscall 69), + and a POSIX one (syscall 189)Naturally, we will use the POSIX + version: + + +189 STD POSIX { int fstat(int fd, struct stat *sb); } + + + + This is a very straightforward call: We pass to it the address + of a stat structure and the descriptor of + an open file. It will fill out the contents of the + stat structure. + + + I do, however, have to say that I tried to declare the + stat structure in the + .bss section, and + fstat did not like it: It set the carry + flag indicating an error. After I changed the code to allocate + the structure on the stack, everything was working + fine. + + + + + Changing the File Size + + 11.5. Changing the File Size + + + Because our program may combine carriage return / line feed + sequences into straight line feeds, our output may be smaller + than our input. However, since we are placing our output into + the same file we read the input from, we may have to change + the size of the file. + + + The ftruncate system call allows us to do + just that. Despite its somewhat misleading name, the + ftruncate system call can be used to both + truncate the file (make it smaller) and to grow it. + + + And yes, we will find two versions of + ftruncate in + syscalls.master, an older one (130), and + a newer one (201). We will use the newer one: + + +201 STD BSD { int ftruncate(int fd, int pad, off_t length); } + + + + Please note that this one contains a int + pad again. + + + + + ftuc + + We now know everything we need to write + ftuc. We start by adding some new lines + in system.inc. First, we define some + constants and structures, somewhere at or near the beginning + of the file: + + +;;;;;;; open flags +%define O_RDONLY 0 +%define O_WRONLY 1 +%define O_RDWR 2 + +;;;;;;; mmap flags +%define PROT_NONE 0 +%define PROT_READ 1 +%define PROT_WRITE 2 +%define PROT_EXEC 4 +;; +%define MAP_SHARED 0001h +%define MAP_PRIVATE 0002h + +;;;;;;; stat structure +struc stat +st_dev resd 1 ; = 0 +st_ino resd 1 ; = 4 +st_mode resw 1 ; = 8, size is 16 bits +st_nlink resw 1 ; = 10, ditto +st_uid resd 1 ; = 12 +st_gid resd 1 ; = 16 +st_rdev resd 1 ; = 20 +st_atime resd 1 ; = 24 +st_atimensec resd 1 ; = 28 +st_mtime resd 1 ; = 32 +st_mtimensec resd 1 ; = 36 +st_ctime resd 1 ; = 40 +st_ctimensec resd 1 ; = 44 +st_size resd 2 ; = 48, size is 64 bits +st_blocks resd 2 ; = 56, ditto +st_blksize resd 1 ; = 64 +st_flags resd 1 ; = 68 +st_gen resd 1 ; = 72 +st_lspare resd 1 ; = 76 +st_qspare resd 4 ; = 80 +endstruc + + + We define the new syscalls: + + +%define SYS_mmap 197 +%define SYS_munmap 73 +%define SYS_fstat 189 +%define SYS_ftruncate 201 + +We add the macros for their use: + +%macro sys.mmap 0 + system SYS_mmap +%endmacro + +%macro sys.munmap 0 + system SYS_munmap +%endmacro + +%macro sys.ftruncate 0 + system SYS_ftruncate +%endmacro + +%macro sys.fstat 0 + system SYS_fstat +%endmacro + + +And here is our code: + + +;;;;;;; Fast Text-to-Unix Conversion (ftuc.asm) ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +;; +;; Started: 21-Dec-2000 +;; Updated: 22-Dec-2000 +;; +;; Copyright 2000 G. Adam Stanislav. +;; All rights reserved. +;; +;;;;;;; v.1 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +%include 'system.inc' + +section .data + db 'Copyright 2000 G. Adam Stanislav.', 0Ah + db 'All rights reserved.', 0Ah +usg db 'Usage: ftuc filename', 0Ah +usglen equ $-usg +co db "ftuc: Can't open file.", 0Ah +colen equ $-co +fae db 'ftuc: File access error.', 0Ah +faelen equ $-fae +ftl db 'ftuc: File too long, use regular tuc instead.', 0Ah +ftllen equ $-ftl +mae db 'ftuc: Memory allocation error.', 0Ah +maelen equ $-mae + +section .code + +align 4 +memerr: + push dword maelen + push dword mae + jmp short error + +align 4 +toolong: + push dword ftllen + push dword ftl + jmp short error + +align 4 +facerr: + push dword faelen + push dword fae + jmp short error + +align 4 +cantopen: + push dword colen + push dword co + jmp short error + +align 4 +usage: + push dword usglen + push dword usg + +error: + push dword stderr + sys.write + + push dword 1 + sys.exit + +align 4 +global _start +_start: + pop eax ; argc + pop eax ; program name + pop ecx ; file to convert + jecxz usage + + pop eax + or eax, eax ; Too many arguments? + jne usage + + ; Open the file + push dword O_RDWR + push ecx + sys.open + jc cantopen + + mov ebp, eax ; Save fd + + sub esp, byte stat_size + mov ebx, esp + + ; Find file size + push ebx + push ebp ; fd + sys.fstat + jc facerr + + mov edx, [ebx + st_size + 4] + + ; File is too long if EDX != 0 ... + or edx, edx + jne near toolong + mov ecx, [ebx + st_size] + ; ... or if it is above 2 GB + or ecx, ecx + js near toolong + + ; Do nothing if the file is 0 bytes in size + jecxz .quit + + ; Map the entire file in memory + push edx + push edx ; starting at offset 0 + push edx ; pad + push ebp ; fd + push dword MAP_SHARED + push dword PROT_READ | PROT_WRITE + push ecx ; entire file size + push edx ; let system decide on the address + sys.mmap + jc near memerr + + mov edi, eax + mov esi, eax + push ecx ; for SYS_munmap + push edi + + ; Use EBX for state machine + mov ebx, ordinary + mov ah, 0Ah + cld + +.loop: + lodsb + call ebx + loop .loop + + cmp ebx, ordinary + je .filesize + + ; Output final lf + mov al, ah + stosb + inc edx + +.filesize: + ; truncate file to new size + push dword 0 ; high dword + push edx ; low dword + push eax ; pad + push ebp + sys.ftruncate + + ; close it (ebp still pushed) + sys.close + + add esp, byte 16 + sys.munmap + +.quit: + push dword 0 + sys.exit + +align 4 +ordinary: + cmp al, 0Dh + je .cr + + cmp al, ah + je .lf + + stosb + inc edx + ret + +align 4 +.cr: + mov ebx, cr + ret + +align 4 +.lf: + mov ebx, lf + ret + +align 4 +cr: + cmp al, 0Dh + je .cr + + cmp al, ah + je .lf + + xchg al, ah + stosb + inc edx + + xchg al, ah + ; fall through + +.lf: + stosb + inc edx + mov ebx, ordinary + ret + +align 4 +.cr: + mov al, ah + stosb + inc edx + ret + +align 4 +lf: + cmp al, ah + je .lf + + cmp al, 0Dh + je .cr + + xchg al, ah + stosb + inc edx + + xchg al, ah + stosb + inc edx + mov ebx, ordinary + ret + +align 4 +.cr: + mov ebx, ordinary + mov al, ah + ; fall through + +.lf: + stosb + inc edx + ret + + + + WARNING: Do not use this program on files + stored on a disk formated by MS DOS or Windows. There seems to + be a subtle bug in the FreeBSD code when using + mmap on these drives mounted under + FreeBSD: If the file is over a certain size, + mmap will just fill the memory with + zeros, and then copy them to the file overwriting its + contents. + + + + + + Caveats + + Assembly language programmers who "grew up" under MS DOS and + Windows often tend to take shortcuts. Reading the keyboard scan + codes and writing directly to video memory are two classical + examples of practices which, under MS DOS are not frowned upon + but considered the right thing to do. + + The reason? Both the PC BIOS and MS DOS are notoriously slow + when performing these operations. + + You may be tempted to continue similar practices in the Unix + environment. For example, I have seen a web site which explains + how to access the keyboard scan codes on a popular Unix + clone. + + That is generally a very bad idea in + Unix environment! Let me explain why. + + + Unix Is Protected + + For one thing, it may simply not be possible. Unix runs in + protected mode. Only the kernel and device drivers are allowed + to access hardware directly. Perhaps a particular Unix clone + will let you read the keyboard scan codes, but chances are a + real Unix operating system will not. And even if one version + may let you do it, the next one may not, so your carefully + crafted software may become a dinosaur overnight. + + + + + Unix is an Abstraction + + But there is a much more important reason not to try + accessing the hardware directly (unless, of course, you are + writing a device driver), even on the Unix-like systems that + let you do it: + + Unix is an abstraction! + + There is a major difference in the philosophy of design + between MS DOS and Unix. MS DOS was designed as a + single-user system. It is run on a computer with a keyboard + and a video screen attached directly to that computer. User + input is almost guaranteed to come from that keyboard. Your + programâs output virtually always ends up on that + screen. + + This is NEVER guaranteed under Unix. It is quite common + for a Unix user to pipe and redirect program input and + output: + + &prompt.user program1 | program2 | program3 > file1 + + + If you have written program2, your + input does not come from the keyboard but from the + outputprogram1. Similarly, your output + does not go to the screen but becomes the input for + program3 whose output, in turn, goes to + file1. + + But there is more! Even if you made sure that your input + comes from, and your output goes to, the terminal, there is + no guarantee the terminal is a PC: It may not have its video + memory where you expect it, nor may its keyboard be + producing PC-style scan codes. It may be a Macintosh, or any + other computer. + + Now you may be shaking your head: My software is in + assembly language, how can it run on a Macintosh? But I did + not say your software would be running on a Macintosh, only + that its terminal may be a Macintosh. + + Under Unix, the terminal does not have to be directly + attached to the computer that runs your software, it can + even be on another continent, or, for that matter, on + another planet. It is perfectly possible that a Macintosh + user in Australia connects to a Unix system in North America + (or anywhere else) via telnet. The + software then runs on one computer, while the terminal is on + a different computer: If you try to read the scan codes, you + will get the wrong input! + + Same holds true about any other hardware: A file you are + reading may be on a disk you have no direct access to. A + camera you are reading images from may be on a space + shuttle, connected to you via satellites. + + That is why under Unix you must never make any + assumptions about where your data is coming from and going + to. Always let the system handle the physical access to the + hardware. + + N.B.: These are caveats, not + absolute rules. Exceptions are possible. For example, if a + text editor has determined it is running on a local machine, + it may want to read the scan codes directly for improved + control. I am not mentioning these caveats to tell you what + to do or what not to do, just to make you aware of certain + pitfalls that await you if you have just arrived to Unix + form MS DOS. Of course, creative people often break rules, + and it is OK as long as they know they are breaking them and + why. + + + + + + + Acknowledgements + + + This tutorial would never have been possible without the help of + many experienced FreeBSD programmers from the FreeBSD hackers + mailing list, many of whom have patiently answered my questions, + and pointed me in the right direction in my attempts to explore + the inner workings of Unix system programming in general and + FreeBSD in particular. + + + Thomas M. Sommers opened the door for me. His How do I write + "Hello, world" in FreeBSD assembler? web page was my + first encounter with an example of assembly language programming + under FreeBSD. + + + Jake Burkholder has kept the door open by willingly answering + all of my questions and supplying me with example assembly + language source code. Copyright © 2000 G. Adam + Stanislav.All rights reserved. + + diff --git a/en_US.ISO_8859-1/books/developers-handbook/book.sgml b/en_US.ISO_8859-1/books/developers-handbook/book.sgml index a3d6113f2b..f407d4c9a7 100644 --- a/en_US.ISO_8859-1/books/developers-handbook/book.sgml +++ b/en_US.ISO_8859-1/books/developers-handbook/book.sgml @@ -1,509 +1,504 @@ %bookinfo; %man; %chapters; %authors; ]> FreeBSD Developers' Handbook The FreeBSD Documentation Project August 2000 2000 2001 The FreeBSD Documentation Project &bookinfo.legalnotice; Welcome to the Developers' Handbook. Introduction Developing on FreeBSD This will need to discuss FreeBSD as a development platform, the vision of BSD, architectural overview, layout of /usr/src, history, etc. Thank you for considering FreeBSD as your development platform! We hope it will not let you down. The BSD Vision Architectural Overview The Layout of /usr/src The complete source code to FreeBSD is available from our public CVS repository. The source code is normally installed in /usr/src which contains the following subdirectories. Directory Description bin/ Source for files in /bin contrib/ Source for files from contributed software. crypto/ DES source etc/ Source for files in /etc games/ Source for files in /usr/games gnu/ Utilities covered by the GNU Public License include/ Source for files in /usr/include kerberosIV/ Source for Kerbereros version IV kerberos5/ Source for Kerbereros version 5 lib/ Source for files in /usr/lib libexec/ Source for files in /usr/libexec release/ Files required to produce a FreeBSD release sbin/ Source for files in /sbin secure/ FreeSec sources share/ Source for files in /sbin sys/ Kernel source files tools/ Tools used for maintenance and testing of FreeBSD usr.bin/ Source for files in /usr/bin usr.sbin/ Source for files in /usr/sbin Basics &chap.tools; &chap.secure; Kernel History of the Unix Kernel Some history of the Unix/BSD kernel, system calls, how do processes work, blocking, scheduling, threads (kernel), context switching, signals, interrupts, modules, etc. &chap.locking; Memory and Virtual Memory Virtual Memory VM, paging, swapping, allocating memory, testing for memory leaks, mmap, vnodes, etc. I/O System UFS UFS, FFS, Ext2FS, JFS, inodes, buffer cache, labeling, locking, metadata, soft-updates, LFS, portalfs, procfs, vnodes, memory sharing, memory objects, TLBs, caching Interprocess Communication Signals Signals, pipes, semaphores, message queues, shared memory, ports, sockets, doors Networking Sockets Sockets, bpf, IP, TCP, UDP, ICMP, OSI, bridging, firewalling, NAT, switching, etc Network Filesystems AFS AFS, NFS, SANs etc] Terminal Handling Syscons Syscons, tty, PCVT, serial console, screen savers, etc Sound OSS OSS, waveforms, etc Device Drivers &chap.driverbasics; &chap.isa; &chap.pci; &chap.scsi; &chap.usb; NewBus This chapter will talk about the FreeBSD NewBus architecture. Architectures - - IA-32 - - Talk about the architectural specifics of FreeBSD/x86. - - + &chap.x86; Alpha Talk about the architectural specifics of FreeBSD/alpha. Explanation of allignment errors, how to fix, how to ignore. Example assembly language code for FreeBSD/alpha. IA-64 Talk about the architectural specifics of FreeBSD/ia64. Debugging Truss various descriptions on how to debug certain aspects of the system using truss, ktrace, gdb, kgdb, etc Compatibility Layers Linux Linux, SVR4, etc Appendices Dave A Patterson John L Hennessy 1998Morgan Kaufmann Publishers, Inc. 1-55860-428-6 Morgan Kaufmann Publishers, Inc. Computer Organization and Design The Hardware / Software Interface 1-2 W. Richard Stevens 1993Addison Wesley Longman, Inc. 0-201-56317-7 Addison Wesley Longman, Inc. Advanced Programming in the Unix Environment 1-2 Marshall Kirk McKusick Keith Bostic Michael J Karels John S Quarterman 1996Addison-Wesley Publishing Company, Inc. 0-201-54979-4 Addison-Wesley Publishing Company, Inc. The Design and Implementation of the 4.4 BSD Operating System 1-2 Aleph One Phrack 49; "Smashing the Stack for Fun and Profit" Chrispin Cowan Calton Pu Dave Maier StackGuard; Automatic Adaptive Detection and Prevention of Buffer-Overflow Attacks Todd Miller Theo de Raadt strlcpy and strlcat -- consistent, safe string copy and concatenation. diff --git a/en_US.ISO_8859-1/books/developers-handbook/chapters.ent b/en_US.ISO_8859-1/books/developers-handbook/chapters.ent index aca3e3bb3c..6e2d0c3f6f 100644 --- a/en_US.ISO_8859-1/books/developers-handbook/chapters.ent +++ b/en_US.ISO_8859-1/books/developers-handbook/chapters.ent @@ -1,61 +1,61 @@ - + diff --git a/en_US.ISO_8859-1/books/developers-handbook/x86/chapter.sgml b/en_US.ISO_8859-1/books/developers-handbook/x86/chapter.sgml new file mode 100644 index 0000000000..8f60c7db9a --- /dev/null +++ b/en_US.ISO_8859-1/books/developers-handbook/x86/chapter.sgml @@ -0,0 +1,2904 @@ + + + + + x86 Assembly Language Programming + + + This chapter written by G. Adam Stanislav. + Whiz Kid Technomagic + Modifications for the Handbook made by &a.wylie;, &.logo;, + and &a.murray; + + + Synopsis + + Assembly language programing under Unix is highly + undocumented. It is generally assumed that no one would ever + want to use it because various Unix systems run on different + microprocessors, so everything should be written in C for + portability. + + + In reality, C portability is quite a myth. Even C programs need + to be modified when ported from one Unix to another, regardless + of what processor each runs on. Typically, such a program is + full of conditional statements depending on the system it is + compiled for. + + + Even if we believe that all of Unix software should be written + in C, or some other high-level language, we still need assembly + language programmers: Who else would write the section of C + library that accesses the kernel? + + + In this tutorial, which is quite brief at this time, I will + attempt to show you how you can use assembly language writing + Unix programs, specifically under FreeBSD. I hope to turn it + into a complete course of FreeBSD assembly language + eventually. + + + This tutorial does not explain the basics of assembly + language. There are enough resources about that (for a complete + online course in assembly language, see Randall Hyde's Art of Assembly + Language; or if you prefer a printed book, take a look + at Jeff Duntemann's Assembly Language Step-by-Step. However, + once the tutorial is finished, any assembly language programmer + will be able to write programs for FreeBSD quickly and + efficiently. + + + Copyright © 2000 G. Adam Stanislav.All rights reserved. + + + + The Tools + + + The Assembler + + + The most important tool for assembly language programming is + the assembler, the software that converts assembly language + code into machine language. + + + Two very different assemblers are available for FreeBSD. One + is &man.as.1;, which uses the traditional Unix assembly + language syntax. It comes with the system. + + + The other is /usr/ports/devel/nasm. It + uses the Intel syntax. Its main advantage is that it can + assemble code for many operating systems. It needs to be + installed separately, but is completely free. + + + This tutorial uses nasm syntax because + most assembly language programmers coming to FreeBSD from + other operating systems will find it easier to + understand. And, because, quite frankly, that is what I am + used to. + + + + + The Linker + + + The output of the assembler, like that of any compiler, needs + to be linked to form an executable file. + + + The standard &man.ld.1; linker comes with FreeBSD. It works + with the code assembled with either assembler. + + + + + + System Calls + + + Default Calling Convention + + + By default, the FreeBSD kernel uses the C calling + convention. Further, although the kernel is accessed using int + 80h, it is assumed the program will call a function that + issues int 80h, rather than issuing int 80h directly. + + + This convention is very convenient, and quite superior to the + Microsoft convention used by MS DOS. Why? Because the Unix + convention allows any program written in any language to + access the kernel. + + + An assembly language program can do that as well. For example, + we could open a file: + + + kernel: + int 80h ; Call kernel + ret + + open: + push dword mode + push dword flags + push dword path + mov eax, 5 + call kernel + add esp, byte 12 + ret + + + + This is a very clean and portable way of coding. If you need + to port the code to a Unix system which uses a different + interrupt, or a different way of passing parameters, all you + need to change is the kernel procedure. + + + But assembly language programmers like to shave off + cycles. The above example requires a + call/ret combination. We can eliminate it + by pushing an extra dword: + + + open: + push dword mode + push dword flags + push dword path + mov eax, 5 + push eax ; Or any other dword + int 80h + add esp, byte 16 + + + + The 5 that we have placed in + EAX identifies the kernel function, in + this case open. + + + + + Alternate Calling Convention + + FreeBSD is an extremely flexible system. It offers other ways + of calling the kernel. For it to work, however, the system + must have Linux emulation installed. + + + Linux is a Unix-like system. However, its kernel uses the + Microsoft system-call convention of passing parameters in + registers. As with the Unix convention, the function number is + placed in EAX. The parameters, however, + are not passed on the stack but EBX, ECX, EDX, ESI, + EDI, EBP: + + + open: + mov eax, 5 + mov ebx, path + mov ecx, flags + mov edx, mode + int 80h + + + + This convention has a great disadvantage over the Unix way, at + least as far as assembly language programming is concerned: + Every time you make a kernel call you must + push the registers, then + pop them later. This makes your code + bulkier and slower. Nevertheless, FreeBSD gives you a + choice. + + + If you do choose the Microsoft/Linux convention, you must let + the system know about it. After your program is assembled and + linked, you need to brand the executable: + + &prompt.user; brandelf -f Linux filename + + + + + + Which convention should you use? + + If you are coding specifically for FreeBSD, you should always + use the Unix convention: It is faster, you can store global + variables in registers, you do not have to brand the + executable, and you do not impose the installation of the + Linux emulation package on the target system. + + + If you want to create portable code that can also run on + Linux, you will probably still want to give the FreeBSD users + as efficient a code as possible. I will show you how you can + accomplish that after I have explained the basics. + + + + + Call Numbers + + To tell the kernel which system service you are calling, + place its number in EAX. Of course, you + need to know what the number is. + + + The Syscalls File + + The numbers are listed in + syscalls. locate + syscalls finds this file in several different + formats, all produced automatically from + syscalls.master. + + You can find the master file for the default Unix + calling convention in + /usr/src/sys/kern/syscalls.master. If + you need to use the Microsoft convention implemented in the + Linux emulation mode, read + /usr/src/sys/i386/linux/syscalls.master. + + N.B.: Not only do FreeBSD and Linux + use different calling conventions, they sometimes use + different numbers for the same functions. + + syscalls.master describes how the + call is to be made: + + 0 STD NOHIDE { int nosys(void); } syscall nosys_args int + 1 STD NOHIDE { void exit(int rval); } exit rexit_args void + 2 STD POSIX { int fork(void); } + 3 STD POSIX { ssize_t read(int fd, void *buf, size_t nbyte); } + 4 STD POSIX { ssize_t write(int fd, const void *buf, size_t nbyte); } + 5 STD POSIX { int open(char *path, int flags, int mode); } + 6 STD POSIX { int close(int fd); } + etc... + + It is the leftmost column that tells us the number to + place in EAX. + + The rightmost column tells us what parameters to + push. They are + pushed from right to left. + + EXAMPLE 3.1: For example, to + open a file, we need to + push the mode + first, then flags, then the address at + which the path is stored. + + + + + + + + Return Values + + A system call would not be useful most of the time if it did not + return some kind of a value: The file descriptor of an open + file, the number of bytes read to a buffer, the system time, + etc. + + + Additionally, the system needs to inform us if an error occurs: + A file does not exist, system resources are exhausted, we passed + an invalid parameter, etc. + + + man-pages + + The traditional place to look for information about various + system calls under Unix systems are the man pages. FreeBSD + describes its system calls in section 2, sometimes in section + 3. + + + For example, open(2) says: + + + If successful, open() returns a + non-negative integer, termed a file descriptor. It returns + -1 on failure, and sets + errno to indicate the error. + + + The assembly language programmer new to Unix and FreeBSD will + immediately ask the puzzling question: Where is + errno and how do I get to it? + + + N.B.: The information presented in the + man pages applies to C programs. The assembly language + programmer needs additional information. + + + + + Where are the return values? + + Unfortunately, it depends... For most system calls it is in + EAX, but not for all. A good rule of + thumb, when working with a system call for the first time, + look for the return value in EAX. If it + is not there, you need further research. + + + N.B.: I am aware of one system call that + returns the value in EDX: + SYS_fork. All others I have worked with + use EAX. But I have not worked with them + all yet. + + + TIP: If you cannot find the answer here + or anywhere else, study libc source code and see how it + interfaces with the kernel. + + + + + Where is <function>errno</function>>? + Actually, nowhere... + + + errno is part of the C language, not the + Unix kernel. When accessing kernel services directly, the + error code is returned in EAX, the same + register the proper return value generally ends up in. + + + This makes perfect sense. If there is no error, there is no + error code. If there is an error, there is no return + value. One register can contain either. + + + + + Determining an Error Occurred + + When using the standard FreeBSD calling convention, the + carry flag is cleared upon success, set + upon failure. + + + When using the Linux emulation mode, the signed value in + EAX is non-negative upon success, and + contains the return value. In case of an error, the value is + negative, i.e., -errno. + + + + + + + Creating Portable Code + + Portability is generally not one of the strengths of assembly + language. Yet, writing assembly language programs for different + platforms is possible, especially with + nasm. I have written assembly language + libraries that can be assembled for such different operating + systems as Windows and FreeBSD. + + + It is all the more possible when you want your code to run on + two platforms which, while different, are based on similar + architectures. + + + For example, FreeBSD is Unix, Linux is Unix-like. I only + mentioned three differences between them (from an assembly + language programmerâs perspective): The calling convention, the + function numbers, and the way of returning values. + + + Dealing With Function Numbers + + In many cases the function numbers are the same. However, even + when they are not, the problem is easy to deal with: Instead + of using numbers in your code, use constants which you have + declared differently depending on the target + architecture: + + + %ifdef LINUX + %define SYS_execve 11 + %else + %define SYS_execve 59 + %endif + + + + + Dealing With Conventions + + Both, the calling convention, and the return value (the + errno problem) can be resolved with + macros: + + + + %ifdef LINUX + + %macro system 0 + call kernel + %endmacro + + align 4 + kernel: + push ebx + push ecx + push edx + push esi + push edi + push ebp + + mov ebx, [esp+32] + mov ecx, [esp+36] + mov edx, [esp+40] + mov esi, [esp+44] + mov ebp, [esp+48] + int 80h + + pop ebp + pop edi + pop esi + pop edx + pop ecx + pop ebx + + or eax, eax + js .errno + clc + ret + + .errno: + neg eax + stc + ret + + %else + + %macro system 0 + int 80h + %endmacro + + %endif + + + + + + Dealing With Other Portability Issues + + The above solutions can handle most cases of writing code + portable between FreeBSD and Linux. Nevertheless, with some + kernel services the differences are deeper. + + + In that case, you need to write two different handlers for + those particular system calls, and use conditional + assembly. Luckily, most of your code does something other than + calling the kernel, so usually you will only need a few such + conditional sections in your code. + + + + + Using a Portable Library + + You can avoid portability issues in your main code altogether + by writing a library of system calls. Create a separate + library for FreeBSD, a different one for Linux, and yet other + libraries for more operating systems. + + + In your library, write a separate function (or procedure, if + you prefer the traditional assembly language terminology) for + each system call. Use the C calling convention of passing + parameters. But still use EAX to pass + the call number in. In that case, your FreeBSD library can be + very simple, as many seemingly different functions can be just + labels to the same code: + + + sys.open: + sys.close: + [etc...] + int 80h + ret + + + + Your Linux library will require more different functions. But + even here you can group system calls using the same number of + parameters: + + + sys.exit: + sys.close: + [etc... one-parameter functions] + push ebx + mov ebx, [esp+12] + int 80h + pop ebx + jmp sys.return + + ... + + sys.return: + or eax, eax + js sys.err + clc + ret + + sys.err: + neg eax + stc + ret + + + + The library approach may seem inconvenient at first because it + requires you to produce a separate file your code depends + on. But it has many advantages: For one, you only need to + write it once and can use it for all your programs. You can + even let other assembly language programmers use it, or + perhaps use one written by someone else. But perhaps the + greatest advantage of the library is that your code can be + ported to other systems, even by other programmers, by simply + writing a new library without any changes to your code. + + + If you do not like the idea of having a library, you can at + least place all your system calls in a separate assembly + language file and link it with your main program. Here, again, + all porters have to do is create a new object file to link + with your main program. + + + + + Using a Portable Include File + + If you are releasing your software as (or with) source code, + you can use macros and place them in a separate file, which + you include in your code. + + + Porters of your software will simply write a new include + file. No library or external object file is necessary, yet + your code is portable without any need to edit the + code. + + + N.B.: This is the approach we will use + throughout this tutorial. We will name our include file + system.inc, and add to it whenever we + deal with a new system call. + + + We can start our system.inc by declaring + the standard file descriptors: + + + %define stdin 0 + %define stdout 1 + %define stderr 2 + + + + Next, we create a symbolic name for each system call: + + + %define SYS_nosys 0 + %define SYS_exit 1 + %define SYS_fork 2 + %define SYS_read 3 + %define SYS_write 4 + ; [etc...] + + + + We add a short, non-global procedure with a long name, so we + do not accidentally reuse the name in our code: + + + section .code + align 4 + access.the.bsd.kernel: + int 80h + ret + + + + We create a macro which takes one argument, the syscall + number: + + + %macro system 1 + mov eax, %1 + call access.the.bsd.kernel + %endmacro + + + + Finally, we create macros for each syscall. These macros take + no arguments. + + + %macro sys.exit 0 + system SYS_exit + %endmacro + + %macro sys.fork 0 + system SYS_fork + %endmacro + + %macro sys.read 0 + system SYS_read + %endmacro + + %macro sys.write 0 + system SYS_write + %endmacro + + ; [etc...] + + + + Go ahead, enter it into your editor and save it as + system.inc. We will add more to it as we + discuss more syscalls. + + + + + + + Our First Program + + We are now ready for our first program, the mandatory + Hello, World! + + + 1: %include 'system.inc' + 2: + 3: section .data + 4: hello db 'Hello, World!', 0Ah + 5: hbytes equ $-hello + 6: + 7: section .code + 8: global _start + 9: _start: + 10: push dword hbytes + 11: push dword hello + 12: push dword stdout + 13: sys.write + 14: + 15: push dword 0 + 16: sys.exit + + + + Here is what it does: Line 1 includes the defines, the macros, + and the code from system.inc. + + + Lines 3-5 are the data: Line 3 starts the data + section/segment. Line 4 contains the string "Hello, World!" + followed by a new line (0Ah). Line 5 creates a constant that + contains the length of the string from line 4 in bytes. + + + Lines 7-16 contain the code. Note that FreeBSD uses the elf file + format for its executables, which requires every program to + start at the point labeled _start (or, more precisely, the + linker expects that). This label has to be global. + + + Lines 10-13 ask the system to write hbytes + bytes of the hello string to + stdout. + + + Lines 15-16 ask the system to end the program with the return + value of 0. The + SYS_exit syscall never returns, so the code + ends there. + + + N.B.: If you have come to Unix from MS DOS + assembly language background, you may be used to writing + directly to the video hardware. You will never have to worry + about this in FreeBSD, or any other flavor of Unix. As far as + you are concerned, you are writing to a file known as + stdout. This can be the video screen, or a + Telnet terminal, or an actual file, or even the input of another + program. Which it is, is for the system to figure out. + + + Assembling the Code + + Type the code (except the line numbers) in an editor, and + save it in a file named hello.asm. You + need nasm to assemble it. + + + Installing NASM + + If you do not have nasm, type: + + &prompt.user; su +Password: your root password +&prompt.root; cd /usr/ports/devel/nasm +&prompt.root; make install +&prompt.root; exit +&prompt.user; + + &prompt.user; su +Password: your root password +&prompt.root; cd /usr/ports/devel/nasm +&prompt.root; make install +&prompt.root; exit +&prompt.user; + + You may type make install clean + instead of just make install if you do + not want to keep nasm source + code. + + Either way, FreeBSD will automatically download + nasm from the Internet, compile it, and + install it on your system. + + N.B.: If your system is not + FreeBSD, you need to get nasm from its + home + page. You can still use it to assemble FreeBSD + code. + + + + + + Assemble, link and run + + Now you can assemble, link, and run the code: + + &prompt.user; nasm -f elf hello.asm +&prompt.user; ld -s -o hello hello.o +&prompt.user; ./hello +Hello, World! +&prompt.user; + + + + + + Writing Unix Filters + + A common type of Unix application is a filterâa program that + reads data from the stdin, processes it + somehow, then writes the result to + stdout. + + In this chapter, we shall develop a simple filter, and learn + how to read from stdin and write to + stdout. This filter will convert each byte + of its input into a hexadecimal number followed by a blank + space. + + %include 'system.inc' + + section .data + hex db '0123456789ABCDEF' + buffer db 0, 0, ' ' + + section .code + global _start + _start: + ; read a byte from stdin + push dword 1 + push dword buffer + push dword stdin + sys.read + add esp, byte 12 + or eax, eax + je .done + + ; convert it to hex + movzx eax, byte [buffer] + mov edx, eax + shr dl, 4 + mov dl, [hex+edx] + mov [buffer], dl + and al, 0Fh + mov al, [hex+eax] + mov [buffer+1], al + + ; print it + push dword 3 + push dword buffer + push dword stdout + sys.write + add esp, byte 12 + jmp short _start + + .done : + push dword 0 + sys.exit + + In the data section we create an array called + hex. It contains the 16 hexadecimal digits + in ascending order. The array is followed by a buffer which we + will use for both input and output. The first two bytes of the + buffer are initially set to 0. This is + where we will write the two hexadecimal digits (the first byte + also is where we will read the input). The third byte is a + space. + + The code section consists of four parts: Reading the byte, + converting it to a hexadecimal number, writing the result, and + eventually exiting the program. + + To read the byte, we ask the system to read one byte from + stdin, and store it in the first byte of + the buffer. The system returns the number + of bytes read in EAX. This will be + 1 while data is coming, or + 0, when no more input data is + available. Therefore, we check the value of + EAX. If it is 0, we + jump to .done, otherwise we + continue. + + N.B.: For simplicity sake, we are + ignoring the possibility of an error condition at this + time. + + The hexadecimal conversion reads the byte from the + buffer into EAX, or + actually just AL, while clearing the + remaining bits of EAX to zeros. We also + copy the byte to EDX because we need to + convert the upper four bits (nibble) separately from the lower + four bits. We store the result in the first two bytes of the + buffer. + + Next, we ask the system to write the three bytes of the + buffer, i.e., the two hexadecimal digits and the blank space, to + stdout. We then jump back to the beginning + of the program and process the next byte. + + Once there is no more input left, we ask the system to exit + our program, returning a zero, which is the traditional value + meaning the program was successful. + + Go ahead, and save the code in a file named + hex.asm, then type the following (the + ^D means press the control key and type + D while holding the control key + down): + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + N.B.: If you are migrating to Unix from + MS DOS, you may be wondering why each line ends with + 0A instead of 0D + 0A. This is because Unix does not use the cr/lf + convention, but a ânew lineâ convention, which is + 0A in hexadecimal. + + Can we improve this? Well, for one, it is a bit confusing + because once we have converted a line of text, our input no + longer starts at the begining of the line. We can modify it to + print a new line instead of a space after each + 0A: + + %include 'system.inc' + +section .data +hex db '0123456789ABCDEF' +buffer db 0, 0, ' ' + +section .code +global _start +_start: + mov cl, ' ' + +.loop: + ; read a byte from stdin + push dword 1 + push dword buffer + push dword stdin + sys.read + add esp, byte 12 + or eax, eax + je .done + + ; convert it to hex + movzx eax, byte [buffer] + mov [buffer+2], cl + cmp al, 0Ah + jne .hex + mov [buffer+2], al + +.hex: + mov edx, eax + shr dl, 4 + mov dl, [hex+edx] + mov [buffer], dl + and al, 0Fh + mov al, [hex+eax] + mov [buffer+1], al + + ; print it + push dword 3 + push dword buffer + push dword stdout + sys.write + add esp, byte 12 + jmp short .loop + +.done: + push dword 0 + sys.exit + + We have stored the space in the CL + register. We can do this safely because, unlike Microsoft + Windows, Unix system calls do not modify the value of any + register they do not use to return a value in. + + That means we only need to set CL + once. We have, therefore, added a new label + .loop and jump to it for the next byte + instead of jumping at _start. We have also + added the .hex label so we can either have + a blank space or a new line as the third byte of the + buffer. + + Once you have changed hex.asm to + reflect these changes, type: + + &prompt.user nasm -f elf hex.asm +&prompt.user ld -s -o hex hex.o +&prompt.user ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + That looks better. But this code is quite inefficient! We + are making a system call for every single byte twice (once to + read it, another time to write the output). + + + + + Buffered Input and Output + + We can improve the efficiency of our code by buffering our + input and output. We create an input buffer and read a whole + sequence of bytes at one time. Then we fetch them one by one + from the buffer. + + We also create an output buffer. We store our output in it + until it is full. At that time we ask the kernel to write the + contents of the buffer to stdout. + + The program ends when there is no more input. But we still + need to ask the kernel to write the contents of our output + buffer to stdout one last time, otherwise + some of our output would make it to the output buffer, but never + be sent out. Do not forget that, or you will be wondering why + some of your output is missing. + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +global _start +_start: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword stdin + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword stdout + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + We now have a third section in the source code, named + .bss. This section is not included in our + executable file, and, therefore, cannot be initialized. We use + resb instead of db. It + simply reserves the requested size of uninitialized memory for + our use. + + We take advantage of the fact that the system does not + modify the registers: We use registers for what, otherwise, + would have to be global variables stored in the + .data section. This is also why the Unix + convention of passing parameters to system calls on the stack is + superior to the Microsoft convention of passing them in the + registers: We can keep the registers for our own use. + + We use EDI and ESI + as pointers to the next byte to be read from or written to. We + use EBX and ECX to + keep count of the number of bytes in the two buffers, so we know + when to dump the output to, or read more input from, the + system. + + Let us see how it works now: + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +Here I come! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + Not what you expected? The program did not print the output + until we pressed ^D. That is easy to fix by + inserting three lines of code to write the output every time we + have converted a new line to 0A. I have + marked the three lines with > (do not copy the > in your + hex.asm + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +global _start +_start: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar +> cmp al, 0Ah +> jne .loop +> call write + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword stdin + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword stdout + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + Now, let us see how it works: + + &prompt.user; nasm -f elf hex.asm +&prompt.user; ld -s -o hex hex.o +&prompt.user; ./hex +Hello, World! +48 65 6C 6C 6F 2C 20 57 6F 72 6C 64 21 0A +Here I come! +48 65 72 65 20 49 20 63 6F 6D 65 21 0A +^D +&prompt.user; + + Not bad for a 644-byte executable, is it! + + + How to Unread a Character + + WARNING: This may be a somewhat advanced + topic, mostly of interest to programmers familiar with the + theory of compilers. If you wish, you may skip to the next chapter, and + perhaps read this later. + + While our sample program does not require it, more + sophisticated filters often need to look ahead. In other + words, they may need to see what the next character is (or + even several characters). If the next character is of a + certain value, it is part of the token currently being + processed. Otherwise, it is not. + + For example, you may be parsing the input stream for a + textual string (e.g., when implementing a language compiler): + If a character is followed by another character, or perhaps a + digit, it is part of the token you are processing. If it is + followed by white space, or some other value, then it is not + part of the current token. + + This presents an interesting problem: How to return the + next character back to the input stream, so it can be read + again later? + + One possible solution is to store it in a character + variable, then set a flag. We can modify + getchar to check the flag, and if it is + set, fetch the byte from that variable instead of the input + buffer, and reset the flag. But, of course, that slows us + down. + + The C language has an ungetc() + function, just for that purpose. Is there a quick way to + implement it in our code? I would like you to scroll back up + and take a look at the getchar procedure + and see if you can find a nice and fast solution before + reading the next paragraph. Then come back here and see my own + solution. + + The key to returning a character back to the stream is in + how we are getting the characters to start with: + + First we check if the buffer is empty by testing the value + of EBX. If it is zero, we call the + read procedure. + + If we do have a character available, we use + lodsb, then decrease the value of + EBX. The lodsb + instruction is effectively identical to: + + mov al, [esi] + inc esi + + The byte we have fetched remains in the buffer until the + next time read is called. We do not know + when that happens, but we do know it will not happen until the + next call to getchar. Hence, to âreturnâ + the last-read byte back to the stream, all we have to do is + decrease the value of ESI and increase + the value of EBX: + + ungetc: + dec esi + inc ebx + ret + + But, be careful! We are perfectly safe doing this if our + look-ahead is at most one character at a time. If we are + examining more than one upcoming character and call + ungetc several times in a row, it will + work most of the time, but not all the time (and will be tough + to debug). Why? + + Because as long as getchar does not + have to call read all of the pre-read + bytes are still in the buffer, and our + ungetc works without a glitch. But the + moment getchar calls + read, the contents of the buffer + change. + + We can always rely on ungetc working + properly on the last character we have read with + getchar, but not on anything we have read + before that. + + If your program reads more than one byte ahead, you have + at least two choices: + + If possible, modify the program so it only reads one byte + ahead. This is the simplest solution. + + If that option is not available, first of all determine + the maximum number of characters your program needs to return + to the input stream at one time. Increase that number + slightly, just to be sure, preferably to a multiple of 16âso + it aligns nicely. Then modify the .bss + section of your code, and create a small âspareâ buffer right + before your input buffer, something like this: + + section .bss + resb 16 ; or whatever the value you came up with +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + + You also need to modify your ungetc + to pass the value of the byte to unget in + AL: + + ungetc: + dec esi + inc ebx + mov [esi], al + ret + + With this modification, you can call + ungetc up to 17 times in a row safely + (the first call will still be within the buffer, the remaining + 16 may be either within the buffer or within the + âspareâ). + + + + + + Command-line Arguments + + Our hex program will be more useful if + it can read the names of an input and output file from its + command line, i.e., if it can process the command line + arguments. But... Where are they? + + Before a Unix system starts a program, it pushes some data + on the stack, then jumps at the _start + label of the program. Yes, I said jumps, not calls. That means + the data can be accessed by reading + [esp+offset], or by simply + popping it. + + The value at the top of the stack contains the number of + command line arguments. It is traditionally called + argc, for âargument count.â + + Command line arguments follow next, all + argc of them. These are typically referred + to as argv, for âargument value(s).â That + is, we get argv[0], + argv[1], ..., + argv[argc-1]. These are not the actual + arguments, but pointers to arguments, i.e., memory addresses of + the actual arguments. The arguments themselves are + NUL-terminated character strings. + + The argv list is followed by a + NULL pointer, which is simply a + 0. There is more, but this is enough for + our purposes right now. + + N.B.: If you have come from the MS DOS + programming environment, the main difference is that each + argument is in a separate string. The second difference is that + there is no practical limit on how many arguments there can + be. + + Armed with this knowledge, we are almost ready for the next + version of hex.asm. First, however, we need + to add a few lines to system.inc: + + First, we need to add two new entries to our list of system + call numbers: + + %define SYS_open 5 +%define SYS_close 6 + + Then we add two new macros at the end of the file: + + %macro sys.open 0 + system SYS_open +%endmacro + +%macro sys.close 0 + system SYS_close +%endmacro + + Here, then, is our modified source code: + + %include 'system.inc' + +%define BUFSIZE 2048 + +section .data +fd.in dd stdin +fd.out dd stdout +hex db '0123456789ABCDEF' + +section .bss +ibuffer resb BUFSIZE +obuffer resb BUFSIZE + +section .code +align 4 +err: + push dword 1 ; return failure + sys.exit + +align 4 +global _start +_start: + add esp, byte 8 ; discard argc and argv[0] + + pop ecx + jecxz .init ; no more arguments + + ; ECX contains the path to input file + push dword 0 ; O_RDONLY + push ecx + sys.open + jc err ; open failed + + add esp, byte 8 + mov [fd.in], eax + + pop ecx + jecxz .init ; no more arguments + + ; ECX contains the path to output file + push dword 420 ; file mode (644 octal) + push dword 0200h | 0400h | 01h + ; O_CREAT | O_TRUNC | O_WRONLY + push ecx + sys.open + jc err + + add esp, byte 12 + mov [fd.out], eax + +.init: + sub eax, eax + sub ebx, ebx + sub ecx, ecx + mov edi, obuffer + +.loop: + ; read a byte from input file or stdin + call getchar + + ; convert it to hex + mov dl, al + shr al, 4 + mov al, [hex+eax] + call putchar + + mov al, dl + and al, 0Fh + mov al, [hex+eax] + call putchar + + mov al, ' ' + cmp dl, 0Ah + jne .put + mov al, dl + +.put: + call putchar + cmp al, dl + jne .loop + call write + jmp short .loop + +align 4 +getchar: + or ebx, ebx + jne .fetch + + call read + +.fetch: + lodsb + dec ebx + ret + +read: + push dword BUFSIZE + mov esi, ibuffer + push esi + push dword [fd.in] + sys.read + add esp, byte 12 + mov ebx, eax + or eax, eax + je .done + sub eax, eax + ret + +align 4 +.done: + call write ; flush output buffer + + ; close files + push dword [fd.in] + sys.close + + push dword [fd.out] + sys.close + + ; return success + push dword 0 + sys.exit + +align 4 +putchar: + stosb + inc ecx + cmp ecx, BUFSIZE + je write + ret + +align 4 +write: + sub edi, ecx ; start of buffer + push ecx + push edi + push dword [fd.out] + sys.write + add esp, byte 12 + sub eax, eax + sub ecx, ecx ; buffer is empty now + ret + + In our .data section we now have two + new variables, fd.in and + fd.out. We store the input and output file + descriptors here. + + In the .code section we have replaced + the references to stdin and + stdout with [fd.in] + and [fd.out]. + + The .code section now starts with a + simple error handler, which does nothing but exit the program + with a return value of 1. The error handler + is before _start so we are within a short + distance from where the errors occur. + + Naturally, the program execution still begins at + _start. First, we remove + argc and argv[0] from + the stack: They are of no interest to us (in this program, that + is). + + We pop argv[1] to + ECX. This register is particularly suited + for pointers, as we can handle NULL + pointers with jecxz. If + argv[1] is not NULL, + we try to open the file named in the first argument. Otherwise, + we continue the program as before: Reading from + stdin, writing to + stdout. If we fail to open the input file + (e.g., it does not exist), we jump to the error handler and + quit. + + If all went well, we now check for the second argument. If + it is there, we open the output file. Otherwise, we send the + output to stdout. If we fail to open the + output file (e.g., it exists and we do not have the write + permission), we, again, jump to the error handler. + + The rest of the code is the same as before, except we close + the input and output files before exiting, and, as mentioned, we + use [fd.in] and + [fd.out]. + + Our executable is now a whopping 768 bytes long. + + Can we still improve it? Of course! Every program can be + improved. Here are a few ideas of what we could do: + + + Have our error handler print a message to + stderr. + + Add error handlers to the + read and write + functions. + + Close stdin when we open an + input file, stdout when we open an output + file. + + Add command line switches, such as + -i and -o, so we can + list the input and output files in any order, or perhaps read + from stdin and write to a + file. + + Print a usage message if command line arguments + are incorrect. + + + + I shall leave these enhancements as an exercise to the + reader: You already know everything you need to know to implement + them. + + + + Unix Environment + + + An important Unix concept is the environment, which is defined + by environment variables. Some are set by + the system, others by you, yet others by the + shell, or any program that loads another + program. + + + How to Find Environment Variables + + I said earlier that when a program starts executing, the stack + contains argc followed by the + NULL-terminated argv + array, followed by something else. The âsomething elseâ is the + environment, or, to be more precise, a + NULL-terminated array of pointers to + environment variables. This is often + referred to as env. + + + The structure of env is the same as that + of argv, a list of memory addresses + followed by a NULL + (0). In this case, there is no + âenvcââwe figure out where the array ends + by searching for the final NULL. + + + The variables usually come in the + name=value format, but sometimes the + =value part may be missing. We need to account for that + possibility. + + + + Webvar + + + I could just show you some code that prints the environment + the same way the Unix env command + does. But I thought it would be more interesting to write a + simple assembly language CGI utility. + + + CGI: A Quick Overview + + 10.2.1. CGI: A Quick Overview + + + I have a detailed + CGI tutorial on my web site, but here is a very + quick overview of CGI: + + + + The web server communicates with the CGI + program by setting environment + variables. + + The CGI program sends its output to + stdout. The web server reads it from + there. + + It must start with an HTTP header followed + by two blank lines. + + It then prints the HTML code, or whatever + other type of data it is producing. + + + + + N.B.: While certain + environment variables use standard + names, others vary, depending on the web server. That makes + webvars quite a useful diagnostic + tool. + + + + + Webvar continued... + + + Our webvar program, then, must send out + the HTTP header followed by some HTML mark-up. It then must + read the environment variable one by one + and send them out as part of the HTML page. + + + The code follows. I placed comments and explanations right + inside the code: + + +;;;;;;; webvars.asm ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; +; Copyright (c) 2000 G. Adam Stanislav +; All rights reserved. +; +; Redistribution and use in source and binary forms, with or without +; modification, are permitted provided that the following conditions +; are met: +; 1. Redistributions of source code must retain the above copyright +; notice, this list of conditions and the following disclaimer. +; 2. Redistributions in binary form must reproduce the above copyright +; notice, this list of conditions and the following disclaimer in the +; documentation and/or other materials provided with the distribution. +; +; THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND +; ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +; IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +; ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE +; FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +; DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +; OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +; HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +; LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +; OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +; SUCH DAMAGE. +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +; +; Version 1.0 +; +; Started: 8-Dec-2000 +; Updated: 8-Dec-2000 +; +;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +%include 'system.inc' + +section .data +http db 'Content-type: text/html', 0Ah, 0Ah + db '<?xml version="1.0" encoding="UTF-8"?>', 0Ah + db '<!DOCTYPE html PUBLIC "-//W3C/DTD XHTML Strict//EN" ' + db '"DTD/xhtml1-strict.dtd">', 0Ah + db '<html xmlns="http://www.w3.org/1999/xhtml" ' + db 'xml.lang="en" lang="en">', 0Ah + db '<head>', 0Ah + db '<title>Web Environment</title>', 0Ah + db '<meta name="author" content="G. Adam Stanislav" />', 0Ah + db '</head>', 0Ah, 0Ah + db '<body bgcolor="#ffffff" text="#000000" link="#0000ff" ' + db 'vlink="#840084" alink="#0000ff">', 0Ah + db '<div class="webvars">', 0Ah + db '<h1>Web Environment</h1>', 0Ah + db '<p>The following <b>environment variables</b> are defined ' + db 'on this web server:</p>', 0Ah, 0Ah + db '<table align="center" width="80" border="0" cellpadding="10" ' + db 'cellspacing="0" class="webvars">', 0Ah +httplen equ $-http +left db '<tr>', 0Ah + db '<td class="name"><tt>' +leftlen equ $-left +middle db '</tt></td>', 0Ah + db '<td class="value"><tt><b>' +midlen equ $-middle +undef db '<i>(undefined)</i>' +undeflen equ $-undef +right db '</b></tt></td>', 0Ah + db '</tr>', 0Ah +rightlen equ $-right +wrap db '</table>', 0Ah + db '</div>', 0Ah + db '</body>', 0Ah + db '</html>', 0Ah, 0Ah +wraplen equ $-wrap + +section .code +global _start +_start: + ; First, send out all the http and xhtml stuff that is + ; needed before we start showing the environment + push dword httplen + push dword http + push dword stdout + sys.write + + ; Now find how far on the stack the environment pointers + ; are. We have 12 bytes we have pushed before "argc" + mov eax, [esp+12] + + ; We need to remove the following from the stack: + ; + ; The 12 bytes we pushed for sys.write + ; The 4 bytes of argc + ; The EAX*4 bytes of argv + ; The 4 bytes of the NULL after argv + ; + ; Total: + ; 20 + eax * 4 + ; + ; Because stack grows down, we need to ADD that many bytes + ; to ESP. + lea esp, [esp+20+eax*4] + cld ; This should already be the case, but let's be sure. + + ; Loop through the environment, printing it out +.loop: + pop edi + or edi, edi ; Done yet? + je near .wrap + + ; Print the left part of HTML + push dword leftlen + push dword left + push dword stdout + sys.write + + ; It may be tempting to search for the '=' in the env string next. + ; But it is possible there is no '=', so we search for the + ; terminating NUL first. + mov esi, edi ; Save start of string + sub ecx, ecx + not ecx ; ECX = FFFFFFFF + sub eax, eax +repne scasb + not ecx ; ECX = string length + 1 + mov ebx, ecx ; Save it in EBX + + ; Now is the time to find '=' + mov edi, esi ; Start of string + mov al, '=' +repne scasb + not ecx + add ecx, ebx ; Length of name + + push ecx + push esi + push dword stdout + sys.write + + ; Print the middle part of HTML table code + push dword midlen + push dword middle + push dword stdout + sys.write + + ; Find the length of the value + not ecx + lea ebx, [ebx+ecx-1] + + ; Print "undefined" if 0 + or ebx, ebx + jne .value + + mov ebx, undeflen + mov edi, undef + +.value: + push ebx + push edi + push dword stdout + sys.write + + ; Print the right part of the table row + push dword rightlen + push dword right + push dword stdout + sys.write + + ; Get rid of the 60 bytes we have pushed + add esp, byte 60 + + ; Get the next variable + jmp .loop + +.wrap: + ; Print the rest of HTML + push dword wraplen + push dword wrap + push dword stdout + sys.write + + ; Return success + push dword 0 + sys.exit + + + + This code produces a 1,396-byte executable. Most of it is + data, i.e., the HTML mark-up we need to send out. + + + Assemble and link it as usual: + + &prompt.user; nasm -f elf webvars.asm +&prompt.user; ld -s -o webvars webvars.o + + + To use it, you need to upload webvars to + your web server. Depending on how your web server is set up, + you may have to store in a special + cgi-bin directory, or perhaps rename it + with a .cgi extension. + + + Then you need to use your browser to view its output. To see + its output on my web server, please instruct your browser to + go to http://www.int80h.org/webvars/. I + am deliberately not placing a regular link here because I do + not want its output to appear on all the search + engines... + + + + + + Working with Files + + We have already done some basic file work: We know how to open + and close them, how to read and write them using buffers. But + Unix offers much more functionality when it comes to files. We + will examine some of it in this section, and end up with a nice + file conversion utility. + + + Indeed, let us start at the end, that is, with the file + conversion utility. It always makes programming easier when we + know from the start what the end product is supposed to + do. + + + One of the first programs I wrote for Unix was tuc, + a text-to-Unix file converter. It converts a text file from + other operating systems to a Unix text file. In other words, it + changes from different kind of line endings to the newline + convention of Unix. It saves the output in a different + file. Optionally, it converts a Unix text file to a DOS text + file. + + I have used tuc extensively, but always + only to convert from some other OS to Unix, never the other + way. I have always wished it would just overwrite the file + instead of me having to send the output to a different + file. Most of the time, I end up using it like this: + + &prompt.user; tuc myfile tempfile +&prompt.user; mv tempfile myfile + + It would be nice to have a ftuc, i.e., + fast tuc, and use it like this: + + &prompt.user; ftuc myfile + + In this chapter, then, we will write + ftuc in assembly language (the original + tuc is in C), and + study various file-oriented kernel services in the + process. + + At first sight, such a file conversion is very simple: All + you have to do is strip the carriage returns, right? + + If you answered yes, think again: That approach will work + most of the time (at least with MS DOS text files), but will + fail occasionally. + + The problem is that not all non-Unix text files end their + line with the carriage return / line feed sequence. Some use + carriage returns without line feeds. Others combine several + blank lines into a single carriage return followed by several + line feeds. And so on. + + A text file converter, then, must be able to handle any + possible line endings: + + + carriage return / line feed + carriage return + line feed / carriage return + line feed + + + + It should also handle files that use some kind of a combination + of the above (e.g., carriage return followed by several line + feeds). + + + Finite State Machine + + The problem is easily solved by the use of a technique called + finite state machine, originally + developed by the designers of digital electronic circuits. A + finite state machine is a digital circuit + whose output is dependent not only on its input but on its + previous input, i.e., on its state. The microprocessor is an + example of a finite state machine: Our + assembly language code is assembled to machine language in + which some assembly language code produces a single byte of + machine language, while others produce several bytes. As the + microprocessor fetches the bytes from the memory one by one, + some of them simply change its state rather than produce some + output. When all the bytes of the op code are fetched, the + microrpocessor produces some output, or changes the value of a + register, etc. + + + Because of that, all software is essentially a sequence of + state instructions for the microprocessor. Nevertheless, the + concept of finite state machine is useful + in software design as well. + + + Our text file converter can be designed as a finite + state machine with three possible states. We could + call them states 0-2, but it will make our life easier if we + give them symbolic names: + + + ordinary + cr + lf + + + + Our program will start in the ordinary + state. During this state, the program action depends on its + input as follows: + + + If the input is anything other than a carriage + return or line feed, the input is simply passed on to the + output. The state remains unchanged. + + If the input is a carriage return, the state + is changed to cr. The input is then + discarded, i.e., no output is made. + + If the input is a line feed, the state is + changed to lf. The input is then + discarded. + + + + + Whenever we are in the cr state, it is + because the last input was a carriage return, which was + unprocessed. What our software does in this state again + depends on the current input: + + + + If the input is anything other than a carriage return + or line feed, output a line feed, then output the input, + then change the state to + ordinary. + + + If the input is a carriage return, we have received + two (or more) carriage returns in a row. We discard the + input, we output a line feed, and leave the state + unchanged. + + + If the input is a line feed, we output the line feed + and change the state to ordinary. Note + that this is not the same as the first case above â if we + tried to combine them, we would be outputting two line feeds + instead of one. + + + + + Finally, we are in the lf state after we + have received a line feed that was not preceded by a carriage + return. This will happen when our file already is in Unix + format, or whenever several lines in a row are expressed by a + single carriage return followed by several line feeds, or when + line ends with a line feed / carriage return sequence. Here is + how we need to handle our input in this state: + + + + If the input is anything other than a carriage return + or line feed, we output a line feed, then output the input, + then change the state to ordinary. This + is exactly the same action as in the cr + state upon receiving the same kind of + input. + + + If the input is a carriage return, we discard the + input, we output a line feed, then change the state to + ordinary. + + + If the input is a line feed, we output the line feed, + and leave the state unchanged. + + + + + + The Final State + + The above finite state machine works + for the entire file, but leaves the possibility that the + final line end will be ignored. That will happen whenever + the file ends with a single carriage return or a single line + feed. I did not think of it when I wrote + tuc, just to discover that occasionally + it strips the last line ending. + + + This problem is easily fixed by checking the state after the + entire file was processed. If the state is not + ordinary, we simply need to output one + last line feed. + + + N.B.: Now that we have expressed our + algorithm as a finite state machine, we + could easily design a dedicated digital electronic circuit + (a âchipâ) to do the conversion for us. Of course, doing so + would be considerably more expensive than writing an + assembly language program. + + + + + The Output Counter + + Because our file conversion program may be combining two + characters into one, we need to use an output counter. We + initialize it to 0, and increase it every time we send a + character to the output. At the end of the program, the + counter will tell us what size we need to set the file + to. + + + + + + Implementing FSM in Software + + The hardest part of working with a finite state + machine is analyzing the problem and expressing it + as a finite state machine. That + accomplished, the software almost writes itself. + + + In a high-level language, such as C, there are several main + approaches. One is to use a switch + statement which chooses what function should be run. For + example, + + + switch (state) { + default: + case REGULAR: + regular(inputchar); + break; + case CR: + cr(inputchar); + break; + case LF: + lf(inputchar); + break; + } + + + + Another approach is by using an array of function pointers, + something like this: + + + (output[state])(inputchar); + + + + Yet another is to have state be a + function pointer, set to point at the appropriate + function: + + + (*state)(inputchar); + + + + This is the approach we will use in our program because it is + very easy to do in assembly language, and very fast, too. We + will simply keep the address of the right procedure in + EBX, and then just issue: + + + call ebx + + + + This is possibly faster than hardcoding the address in the + code because the microprocessor does not have to fetch the + address from the memoryâit is already stored in one of its + registers. I said possibly because with + the caching modern microprocessors do, either way may be + equally fast. + + + + + Memory Mapped Files + + Because our program works on a single file, we cannot use the + approach that worked for us before, i.e., to read from an + input file and to write to an output file. + + + Unix allows us to map a file, or a section of a file, into + memory. To do that, we first need to open the file with the + appropriate read/write flags. Then we use the + mmap system call to map it into the + memory. One nice thing about mmap is that + it automatically works with virtual memory: We can map more of + the file into the memory than we have physical memory + available, yet still access it through regular memory op + codes, such as mov, + lods, and + stos. Whatever changes we make to the + memory image of the file will be written to the file by the + system. We do not even have to keep the file open: As long as + it stays mapped, we can read from it and write to it. + + + The 32-bit Intel microprocessors can access up to four + gigabytes of memory â physical or virtual. The FreeBSD system + allows us to use up to a half of it for file mapping. + + + For simplicity sake, in this tutorial we will only convert + files that can be mapped into the memory in their + entirety. There are probably not too many text files that + exceed two gigabytes in size. If our program encounters one, + it will simply display a message suggesting we use the + original tuc instead. + + + If you examine your copy of + syscalls.master, you will find two + separate syscalls named mmap. This is + because of evolution of Unix: There was the traditional BSD + mmap, syscall 71. That one was superceded + by the POSIX mmap, syscall 197. The + FreeBSD system supports both because older programs were + written by using the original BSD version. But new software + uses the POSIX version, which is what we will use. + + + The syscalls.master file lists the POSIX + version like this: + + +197 STD BSD { caddr_t mmap(caddr_t addr, size_t len, int prot, \ + int flags, int fd, long pad, off_t pos); } + + + + This differs slightly from what mmap(2) + says. That is because mmap(2) describes + the C version. + + + The difference is in the long pad + argument, which is not present in the C version. However, the + FreeBSD syscalls add a 32-bit pad after pushing a 64-bit + argument. In this case, off_t is a 64-bit + value. + + + When we are finished working with a memory-mapped file, we + unmap it with the munmap syscall: + + + TIP: For an in-depth treatment of + mmap, see W. Richard Stevensâ Unix + Network Programming, Volume 2, Chapter 12. + + + + + Determining File size + + Because we need to tell mmap how many + bytes of the file to map into the memory, and because we want + to map the entire file, we need toq determine the size of the + file. + + + We can use the fstat syscall to get all + the information about an open file that the system can give + us. That includes the file size. + + + Again, syscalls.master lists two versions + of fstat, a traditional one (syscall 69), + and a POSIX one (syscall 189)Naturally, we will use the POSIX + version: + + +189 STD POSIX { int fstat(int fd, struct stat *sb); } + + + + This is a very straightforward call: We pass to it the address + of a stat structure and the descriptor of + an open file. It will fill out the contents of the + stat structure. + + + I do, however, have to say that I tried to declare the + stat structure in the + .bss section, and + fstat did not like it: It set the carry + flag indicating an error. After I changed the code to allocate + the structure on the stack, everything was working + fine. + + + + + Changing the File Size + + 11.5. Changing the File Size + + + Because our program may combine carriage return / line feed + sequences into straight line feeds, our output may be smaller + than our input. However, since we are placing our output into + the same file we read the input from, we may have to change + the size of the file. + + + The ftruncate system call allows us to do + just that. Despite its somewhat misleading name, the + ftruncate system call can be used to both + truncate the file (make it smaller) and to grow it. + + + And yes, we will find two versions of + ftruncate in + syscalls.master, an older one (130), and + a newer one (201). We will use the newer one: + + +201 STD BSD { int ftruncate(int fd, int pad, off_t length); } + + + + Please note that this one contains a int + pad again. + + + + + ftuc + + We now know everything we need to write + ftuc. We start by adding some new lines + in system.inc. First, we define some + constants and structures, somewhere at or near the beginning + of the file: + + +;;;;;;; open flags +%define O_RDONLY 0 +%define O_WRONLY 1 +%define O_RDWR 2 + +;;;;;;; mmap flags +%define PROT_NONE 0 +%define PROT_READ 1 +%define PROT_WRITE 2 +%define PROT_EXEC 4 +;; +%define MAP_SHARED 0001h +%define MAP_PRIVATE 0002h + +;;;;;;; stat structure +struc stat +st_dev resd 1 ; = 0 +st_ino resd 1 ; = 4 +st_mode resw 1 ; = 8, size is 16 bits +st_nlink resw 1 ; = 10, ditto +st_uid resd 1 ; = 12 +st_gid resd 1 ; = 16 +st_rdev resd 1 ; = 20 +st_atime resd 1 ; = 24 +st_atimensec resd 1 ; = 28 +st_mtime resd 1 ; = 32 +st_mtimensec resd 1 ; = 36 +st_ctime resd 1 ; = 40 +st_ctimensec resd 1 ; = 44 +st_size resd 2 ; = 48, size is 64 bits +st_blocks resd 2 ; = 56, ditto +st_blksize resd 1 ; = 64 +st_flags resd 1 ; = 68 +st_gen resd 1 ; = 72 +st_lspare resd 1 ; = 76 +st_qspare resd 4 ; = 80 +endstruc + + + We define the new syscalls: + + +%define SYS_mmap 197 +%define SYS_munmap 73 +%define SYS_fstat 189 +%define SYS_ftruncate 201 + +We add the macros for their use: + +%macro sys.mmap 0 + system SYS_mmap +%endmacro + +%macro sys.munmap 0 + system SYS_munmap +%endmacro + +%macro sys.ftruncate 0 + system SYS_ftruncate +%endmacro + +%macro sys.fstat 0 + system SYS_fstat +%endmacro + + +And here is our code: + + +;;;;;;; Fast Text-to-Unix Conversion (ftuc.asm) ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +;; +;; Started: 21-Dec-2000 +;; Updated: 22-Dec-2000 +;; +;; Copyright 2000 G. Adam Stanislav. +;; All rights reserved. +;; +;;;;;;; v.1 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; +%include 'system.inc' + +section .data + db 'Copyright 2000 G. Adam Stanislav.', 0Ah + db 'All rights reserved.', 0Ah +usg db 'Usage: ftuc filename', 0Ah +usglen equ $-usg +co db "ftuc: Can't open file.", 0Ah +colen equ $-co +fae db 'ftuc: File access error.', 0Ah +faelen equ $-fae +ftl db 'ftuc: File too long, use regular tuc instead.', 0Ah +ftllen equ $-ftl +mae db 'ftuc: Memory allocation error.', 0Ah +maelen equ $-mae + +section .code + +align 4 +memerr: + push dword maelen + push dword mae + jmp short error + +align 4 +toolong: + push dword ftllen + push dword ftl + jmp short error + +align 4 +facerr: + push dword faelen + push dword fae + jmp short error + +align 4 +cantopen: + push dword colen + push dword co + jmp short error + +align 4 +usage: + push dword usglen + push dword usg + +error: + push dword stderr + sys.write + + push dword 1 + sys.exit + +align 4 +global _start +_start: + pop eax ; argc + pop eax ; program name + pop ecx ; file to convert + jecxz usage + + pop eax + or eax, eax ; Too many arguments? + jne usage + + ; Open the file + push dword O_RDWR + push ecx + sys.open + jc cantopen + + mov ebp, eax ; Save fd + + sub esp, byte stat_size + mov ebx, esp + + ; Find file size + push ebx + push ebp ; fd + sys.fstat + jc facerr + + mov edx, [ebx + st_size + 4] + + ; File is too long if EDX != 0 ... + or edx, edx + jne near toolong + mov ecx, [ebx + st_size] + ; ... or if it is above 2 GB + or ecx, ecx + js near toolong + + ; Do nothing if the file is 0 bytes in size + jecxz .quit + + ; Map the entire file in memory + push edx + push edx ; starting at offset 0 + push edx ; pad + push ebp ; fd + push dword MAP_SHARED + push dword PROT_READ | PROT_WRITE + push ecx ; entire file size + push edx ; let system decide on the address + sys.mmap + jc near memerr + + mov edi, eax + mov esi, eax + push ecx ; for SYS_munmap + push edi + + ; Use EBX for state machine + mov ebx, ordinary + mov ah, 0Ah + cld + +.loop: + lodsb + call ebx + loop .loop + + cmp ebx, ordinary + je .filesize + + ; Output final lf + mov al, ah + stosb + inc edx + +.filesize: + ; truncate file to new size + push dword 0 ; high dword + push edx ; low dword + push eax ; pad + push ebp + sys.ftruncate + + ; close it (ebp still pushed) + sys.close + + add esp, byte 16 + sys.munmap + +.quit: + push dword 0 + sys.exit + +align 4 +ordinary: + cmp al, 0Dh + je .cr + + cmp al, ah + je .lf + + stosb + inc edx + ret + +align 4 +.cr: + mov ebx, cr + ret + +align 4 +.lf: + mov ebx, lf + ret + +align 4 +cr: + cmp al, 0Dh + je .cr + + cmp al, ah + je .lf + + xchg al, ah + stosb + inc edx + + xchg al, ah + ; fall through + +.lf: + stosb + inc edx + mov ebx, ordinary + ret + +align 4 +.cr: + mov al, ah + stosb + inc edx + ret + +align 4 +lf: + cmp al, ah + je .lf + + cmp al, 0Dh + je .cr + + xchg al, ah + stosb + inc edx + + xchg al, ah + stosb + inc edx + mov ebx, ordinary + ret + +align 4 +.cr: + mov ebx, ordinary + mov al, ah + ; fall through + +.lf: + stosb + inc edx + ret + + + + WARNING: Do not use this program on files + stored on a disk formated by MS DOS or Windows. There seems to + be a subtle bug in the FreeBSD code when using + mmap on these drives mounted under + FreeBSD: If the file is over a certain size, + mmap will just fill the memory with + zeros, and then copy them to the file overwriting its + contents. + + + + + + Caveats + + Assembly language programmers who "grew up" under MS DOS and + Windows often tend to take shortcuts. Reading the keyboard scan + codes and writing directly to video memory are two classical + examples of practices which, under MS DOS are not frowned upon + but considered the right thing to do. + + The reason? Both the PC BIOS and MS DOS are notoriously slow + when performing these operations. + + You may be tempted to continue similar practices in the Unix + environment. For example, I have seen a web site which explains + how to access the keyboard scan codes on a popular Unix + clone. + + That is generally a very bad idea in + Unix environment! Let me explain why. + + + Unix Is Protected + + For one thing, it may simply not be possible. Unix runs in + protected mode. Only the kernel and device drivers are allowed + to access hardware directly. Perhaps a particular Unix clone + will let you read the keyboard scan codes, but chances are a + real Unix operating system will not. And even if one version + may let you do it, the next one may not, so your carefully + crafted software may become a dinosaur overnight. + + + + + Unix is an Abstraction + + But there is a much more important reason not to try + accessing the hardware directly (unless, of course, you are + writing a device driver), even on the Unix-like systems that + let you do it: + + Unix is an abstraction! + + There is a major difference in the philosophy of design + between MS DOS and Unix. MS DOS was designed as a + single-user system. It is run on a computer with a keyboard + and a video screen attached directly to that computer. User + input is almost guaranteed to come from that keyboard. Your + programâs output virtually always ends up on that + screen. + + This is NEVER guaranteed under Unix. It is quite common + for a Unix user to pipe and redirect program input and + output: + + &prompt.user program1 | program2 | program3 > file1 + + + If you have written program2, your + input does not come from the keyboard but from the + outputprogram1. Similarly, your output + does not go to the screen but becomes the input for + program3 whose output, in turn, goes to + file1. + + But there is more! Even if you made sure that your input + comes from, and your output goes to, the terminal, there is + no guarantee the terminal is a PC: It may not have its video + memory where you expect it, nor may its keyboard be + producing PC-style scan codes. It may be a Macintosh, or any + other computer. + + Now you may be shaking your head: My software is in + assembly language, how can it run on a Macintosh? But I did + not say your software would be running on a Macintosh, only + that its terminal may be a Macintosh. + + Under Unix, the terminal does not have to be directly + attached to the computer that runs your software, it can + even be on another continent, or, for that matter, on + another planet. It is perfectly possible that a Macintosh + user in Australia connects to a Unix system in North America + (or anywhere else) via telnet. The + software then runs on one computer, while the terminal is on + a different computer: If you try to read the scan codes, you + will get the wrong input! + + Same holds true about any other hardware: A file you are + reading may be on a disk you have no direct access to. A + camera you are reading images from may be on a space + shuttle, connected to you via satellites. + + That is why under Unix you must never make any + assumptions about where your data is coming from and going + to. Always let the system handle the physical access to the + hardware. + + N.B.: These are caveats, not + absolute rules. Exceptions are possible. For example, if a + text editor has determined it is running on a local machine, + it may want to read the scan codes directly for improved + control. I am not mentioning these caveats to tell you what + to do or what not to do, just to make you aware of certain + pitfalls that await you if you have just arrived to Unix + form MS DOS. Of course, creative people often break rules, + and it is OK as long as they know they are breaking them and + why. + + + + + + + Acknowledgements + + + This tutorial would never have been possible without the help of + many experienced FreeBSD programmers from the FreeBSD hackers + mailing list, many of whom have patiently answered my questions, + and pointed me in the right direction in my attempts to explore + the inner workings of Unix system programming in general and + FreeBSD in particular. + + + Thomas M. Sommers opened the door for me. His How do I write + "Hello, world" in FreeBSD assembler? web page was my + first encounter with an example of assembly language programming + under FreeBSD. + + + Jake Burkholder has kept the door open by willingly answering + all of my questions and supplying me with example assembly + language source code. Copyright © 2000 G. Adam + Stanislav.All rights reserved. + +