Computer Architecture and Assembly Language Practical session outline

Computer Architecture and Assembly Language Practical session outline • Introduction • 80 x 86 assembly –Data storage –The registers –Flags –Instructions • Assignment 0

Introduction: Administration: - Background. - Guy’s office hours: Wednesday 16: 00 -18: 00, room -105/58. email: guyshat@cs…. - 4 practical assignments in the course, 1 theoretic.

Introduction: Why assembly? -Assembly is widely used in industry: - Embedded systems. - Real time systems. - Low level and direct access to hardware -Assembly is widely used not in industry: -Cracking software protections: , patching, patch-loaders and emulators (executable file compression, encryption, decryption) -Hacking into computer systems: buffer under/overflows (worms and trojans).

Byte structure: byte has 8 bits 7 6 5 msb (most significant bit) 4 3 2 1 0

Data storage in memory: NASM stores data using little endian order. Little endian means that the low-order byte of the number is stored in memory at the lowest address, and the high-order byte at the highest address. Example: You want to store 0 x 1 AB 3 (hex number) in the memory. This number has two bytes: 1 A and B 3. It would be stored this way: B 3 0 1 A memory block 1 2 bytes of memory Note: when read a stored data from the memory, it comes in the source order.

Registers: CPU contains a unit called “Register file”. This unit contains the registers of the following types: 1. 8 -bit general registers: AL, BL, CL, DL, AH, BH, CH, DH 2. 16 - bit general registers: AX, BX, CX, DX, SP, BP, SI, Dl 3. 32 -bit general registers: EAX, EBX, ECX, EDX, ESP, EBP, ESI, EDI (Accumulator, Base, Counter, Data, Stack pointer, Base pointer, Source index, Destination Index) 4. Segment registers: ES, CS , SS, DS, FS, GS 5. Floating-point registers: ST 0, ST 1, ST 2, ST 3, ST 4, ST 5, ST 6, ST 7 6. instruction pointer: IP Note: the registers above are the basic. There exist more registers.

Lets zoom in: IP - instruction pointer: contains offset (address) of the next instruction that is going to be executed. Exists only in run time. Can’t be reached. AX, BX, CX, DX - 16 -bit general registers: contains two 8 -bit registers: Example: AH, AL (for AX) high byte XH low byte XL EAX - 32 -bit general purpose register: lower 16 bits are AX. segment registers: we use flat memory model – 32 -bit 4 Gb address space, without segments. So for this course you can ignore segment registers. ST 0 - floating-point registers: we use it to do calculations on floating point numbers, you can ignore these registers. ESP - stack pointer: contains the next free address on a stack.

Lets zoom in: (2). Some instructions use only specific registers. Examples: 1. For DIV r/m 8 instruction, AX is divided by the given operand; the quotient is stored in AL and the remainder in AH. 2. LOOP imm, CX instruction uses CX register as a counter register. 3. LAHF instruction sets the AH register according to the contents of the low byte of the flags word. . We use ESP and EBP registers to work with stack.

Example for using registers: instruction: mov ax, 0 mov ah, 0 x 13 mov ax, 0 x 13 content of the register AX after the instruction execution: AH AL 00000000 AH AL 00010011 AH 00000000 AL 00010011

Status Flags: Flag is a bit (of the Flags Register). . The status flags provide some information about the result of the last (usually arithmetic) instruction that was executed. This information can be used by conditional instructions (such a JUPMcc and CMOVcc) as well as by some of the other instructions (such as ADC). There are 6 status flags: CF - Carry flag: set if an arithmetic operation generates a carry or a borrow out of the most-significant bit of the result; cleared otherwise. This flag indicates an overflow condition for unsigned-integer arithmetic. PF - Parity flag: set if the least-significant byte of the result contains an even number of ‘ 1’ bits; cleared otherwise.

Status Flags (2): AF - Adjust flag: set if an arithmetic operation generates a carry or a borrow out of bit 3 of the result; cleared otherwise. This flag is used in binary -coded decimal (BCD) arithmetic – not needed in our course. ZF - Zero flag: set if the result is zero; cleared otherwise. SF - Sign flag: set equal to the most-significant bit of the result, which is the sign bit of a signed integer. (0 indicates a positive value and 1 indicates a negative value). OF - Overflow flag: set if the integer result is too large a positive number or too small a negative number (excluding the sign-bit) to fit in the destination operand; cleared otherwise. This flag indicates an overflow condition for signed-integer (two's complement) arithmetic.

Instructions on Flags Register: We can’t reach Flags Register but there are few instructions that let us get and set its value: 1. LAHF: set the AH register according to the contents of the low byte of the flags word: AH: 7 6 5 4 3 2 1 0 SF ZF 0 AF 0 PF 1 CF 2. SAHF: set the low byte of the flags word according to the contents of the AH register. 3. SALC: set AL to zero if the carry flag is clear, or to 0 x. FF if it is set. 4. STC: sets the carry flag. 5. CLC: clears the carry flag. Note: this is not a complete set of the flag instructions. You can find more in the NASM tutorial.

Basic assembly instructions: Each NASM standard source line contains a combination of the 4 fields: label: (pseudo) instruction optional fields operands ; comment Either required or forbidden by an instruction Notes: 1. backslash () uses as the line continuation character: if a line ends with backslash, the next line is considered to be a part of the backslash-ended line. 2. no restrictions on white space within a line. 3. a colon after a label is optional. Examples: 1. mov ax, 2 ; moves constant 2 to the register ax 2. buffer: resb 64 ; reserves 64 bytes

Instruction arguments A typical instruction has 2 operands. The left operand is the target operand, while the right operand is the source operand 3 kinds of operands exists: 1. Immediate, i. e. a value 2. Register, such as AX, EBP, DL 3. Memory location; a variable or a pointer. 4. One should notice that the x 86 processor does not allow 5. both operands be memory locations. 6. mov [var 1], [var 2]

Move instructions: B. 4. 156: MOV – move data mov r/m 8, reg 8 (copies content of 8 -bit register (source) to 8 -bit register or 8 -bit memory unit (destination) ) mov reg 32, imm 32 (copies content of 32 -bit immediate (constant) to 32 -bit register) * for all the possible variants of operands look at NASM manual, B. 4. 156 - In all forms of the MOV instruction, the two operands are the same size Examples: mov EAX, 0 x 2334 AAFF mov word [buffer], ax * Note: NASM don’t remember the types of variables you declare. Whereas MASM will remember, on seeing var dw 0, that you declared var as a word-size variable, and will then be able to fill in the ambiguity in the size of the instruction mov var, 2, NASM will deliberately remember nothing about the symbol var except where it begins, and so you must explicitly code mov word [var], 2.

Move instructions (2): B. 4. 181 MOVSX, MOVZX: move data with sign or zero extend movsx reg 16, r/m 8 (sign-extends its source (second) operand to the length of its destination (first) operand, and copies the result into the destination operand) movzx reg 32, r/m 8 (does the same, but zero-extends rather than sign-extending) * for all the possible variants of operands look at NASM manual, B. 4. 181 Examples: movsx EAX, AX (if AX has 10… 0 b value, EAX would have 111… 1 100… 0 value) movzx EAX, BL (if AX has 10… 0 b value, EAX would have 000… 0 100… 0 value)

Basic arithmetical instructions: B. 4. 3 ADD: add integers add r/m 16, imm 16 (adds its two operands together, and leaves the result in its destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 3 Examples: add AX, BX B. 4. 2 ADC: add with carry adc r/m 16, imm 8 (adds its two operands together, plus the value of the carry flag, and leaves the result in its destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 2 • Examples: add AX, BX (AX gets a value of AX+BX+CF)

Basic arithmetical instructions (2): B. 4. 305 SUB: subtract integers sub reg 16, r/m 16 (subtracts its second operand from its first, and leaves the result in its destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 305 Examples: sub AX, BX B. 4. 285 SBB: subtract with borrow sbb r/m 16, imm 8 (subtracts its second operand, plus the value of the carry flag, from its first, and leaves the result in its destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 285 Examples: sbb AX, BX (AX gets a value of AX-BX-CF)

Basic arithmetical instructions (3): B. 4. 120 INC: increment integer inc r/m 16 (adds 1 to its operand) * does not affect the carry flag; affects all the other flags according to the result * for all the possible variants of operands look at NASM manual, B. 4. 120 Examples: inc AX B. 4. 58 DEC: decrement integer dec reg 16 (subtracts 1 from its operand) * does not affect the carry flag; affects all the other flags according to the result * for all the possible variants of operands look at NASM manual, B. 4. 58 Examples: dec byte [buffer]

Basic logical instructions: B. 4. 189 NEG, NOT: two's and one's complement neg r/m 16 (replaces the contents of its operand by the two's complement negation - invert all the bits, and then add one) not r/m 16 (performs one's complement negation- inverts all the bits) * for all the possible variants of operands look at NASM manual, B. 4. 189 Examples: neg AL (if AL = (11111110), it becomes (00000010)) not AL (if AL = (11111110), it becomes (00000001))

Basic logical instructions (2): B. 4. 191 OR: bitwise or or r/m 32, imm 32 (each bit of the result is 1 if and only if at least one of the corresponding bits of the two inputs was 1; stores the result in the destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 191 Example: or AL, BL (if AL = (11111100), BL= (00000010) => AL would be (11111110)) B. 4. 8 AND: bitwise and r/m 32, imm 32 (each bit of the result is 1 if and only if the corresponding bits of the two inputs were both 1; stores the result in the destination (first) operand) * for all the possible variants of operands look at NASM manual, B. 4. 8 Example: and AL, BL (if AL = (11111100), BL= (00000010) => AL would be (11111100))

Compare instruction: B. 4. 24 CMP: compare integers cmp r/m 32, imm 8 (performs a ‘mental’ subtraction of its second operand from its first operand, and affects the flags as if the subtraction had taken place, but does not store the result of the subtraction anywhere) * for all the possible variants of operands look at NASM manual, B. 4. 24 Example: cmp AL, BL (if AL = (11111100), BL= (00000010) => ZF would be 1) (if AL = (11111100), BL= (11111100) => ZF would be 0)

Labels definition (basic): . Each instruction of the code has its offset (address from the beginning of the address space). . If we want to refer to the specific instruction in the code, we should mark it with a label: my_loop 1: add ax, ax …. - label can be with or without colon - an instruction that follows it can be at the same or the next line - a code can’t contain two different non-local (as above) labels with the same name

Loop definition: B. 4. 142 LOOP, LOOPE, LOOPZ, LOOPNE, LOOPNZ: loop with counter * for all the possible variants of operands look at NASM manual, B. 4. 142 Example: mov ax, 1 mov cx, 3 my_ loop: add ax, ax loop my_ loop, cx 1. decrements its counter register (in this case it is CX register) 2. if the counter does not become zero as a result of this operation, it jumps to the given label Note: counter register can be either CX or ECX - if one is not specified explicitly, the BITS setting dictates which is used. LOOPE (or its synonym LOOPZ) adds the additional condition that it only jumps if the counter is nonzero and the zero flag is set. Similarly, LOOPNE (and LOOPNZ) jumps only if the counter is nonzero and the zero flag is clear.

Code ASCII The standard ASCII code defines 128 character codes (from 0 to 127), of which, the first 32 are control codes (non-printable), and the other 96 are representable characters: Example: the A character is located at the 4 throw and the 1 st column, for that it would be represented in hexadecimal as 0 x 41. Here you have an interactive Decimal-Hexadecimal-Octal-ASCII converter (at the bottom of the page).

Assignment 0 • You get a simple program which prints the Nth element of the Fibonacci series. • Add a function written in assembly to the program: – function_double: given 2 arguments: m and n, prints the number: m * 2 n (m, n > 0).

section. rodata LC 0: DB "the result is: %d", 10, 0 section. data an_2: DD 0 an_1: DD 1 helper: DD 0 section. text global function_fib extern printf function_fib: push ebp mov ebp, esp pushad mov ecx, dword [ebp+8] label_here: mov eax, [an_1] mov ebx, [an_2] add eax, ebx mov [helper], eax mov ebx, [an_1] mov [an_2], ebx mov ebx, [helper] mov [an_1], ebx loop label_here, ecx mov push call add eax, [an_2] eax dword LC 0 printf dword esp, 8 ; ; ; ; ; popad mov eax, [an_2] mov esp, ebp pop dword ebp ret ; add necessary modifications ; to the various sections function_double: push ebp mov ebp, esp pushad mov ecx, dword [ebp+8] mov edx, dword [ebp+12] ; ; insert your code here push eax push dword LC 0 call printf add dword esp, 8 popad mov esp, ebp pop dword ebp ret

Running NASM To assemble a file, you issue a command of the form > nasm -f <format> <filename> [-o <output>] [ -l listing] Example: > nasm -f elf mytry. s -o myelf. o It would create myelf. o file that has elf format (executable and linkable format). We use main. c file (that is written in C language) to start our program, and sometimes also for input / output from a user. So to compile main. c with our assembly file we should execute the following command: > cc main. c myelf. o -o myexe. out –l mylist. lst It would create executable file myexe. out and a listing file named mylist. lst. In order to run it you should write its name on the command line: > myexe. out