290 likes | 823 Views
Appendix A. Classifying Instruction Set Architecture Memory addressing mode Operations in the instruction set Control flow instructions Instruction format. CDA5155 Fall 2014, Peir / University of Florida. Classifying Architectures.
E N D
Appendix A Classifying Instruction Set Architecture Memory addressing mode Operations in the instruction set Control flow instructions Instruction format • CDA5155 Fall 2014, Peir / University of Florida
Classifying Architectures • One important classification scheme is by the type of addressing modes supported. • Stack architecture: Operands implicitly on top of a stack. (Early machines, Intel floating-point.) • Accumulator architecture: One operand is implicitly an accumulator (a special register). (Early machines.) • General-purpose register architecture: Operands may be any of a large (typically 10s-100s) # of registers. • Register-memory architectures: One op may be memory. • Load-store architectures: All ops are registers, except in special load and store instructions.
Four Architecture Classes • Each instruction has an opcode and one or more operands • Assembly for C:=A+B:
Number of Operands • A further classification is by the maximum number of operands, and # that can be memory: e.g., • 2-operand (e.g. a += b) • src/dest(reg), src(reg) • src/dest(reg), src(mem) IBM 360, x86, 68k • src/dest(mem), src(mem) VAX • 3-operand (e.g. a = b+c) • dest(reg), src1(reg), src2(reg) MIPS, PPC, SPARC, &c. • dest(reg), src1(reg), src2(mem) IBM 370 • dest(mem), src1(mem), src2(mem) IBM 370, VAX
Endians & Alignment • Increasing byteaddress • 7 • 6 • 5 • 4 • 3 • 2 • 1 • 0 • Byte addressable • memory • 4 • Word-aligned word at byte address 4. • 2 • Halfword-aligned word at byte address 2. • 1 • Byte-aligned (non-aligned) word, at byte address 1. • word • Little-endian byte order (least-significant byte “first”). • 3 (MSB) • 2 • 1 • 0 (LSB) • word • Big-endian byte order (most-significant byte “first”). • 0 (LSB) • 1 • 2 • 3 (MSB)
Addressing Modes • Scaled add r1, 100(r2)[r3] R[1] R[1] + M[100+R[2]+R[3]*d • In example assembly syntax in middle column, ( ) indicates memory access. (A typical syntax.) • In RTL syntax on right, [ ] denotes accessing a member of an array, Register or Memory.
Addressing Mode Usage • 3 SPEC89 on VAX
Displacement Distribution • Sign bit is not counted • SPEC CPU2000 on Alpha
Use of Immediate Operand • SPEC CPU2000 on Alpha
Distribution of Immediate • Sign bit is not counted • SPEC CPU2000 on Alpha
Instruction Type • (Same as A.12 in the 5th Edition)
Instruction Distribution • (Same as Fig. A.13) • (5 SPECint92)
Control Flow Instructions • Four basic types: • (Conditional) branches • (Unconditional) jumps • Procedure calls • Procedure returns • Control flow addressing modes: • Often PC-relative (PC + displacement). Relocatable. • Also useful: register indirect jumps (reg. has addr.). Uses for: • Procedure returns • Case / switch statements • Virtual functions / methods (abstract class method calls) • High-order functions / function pointers • Dynamically shared libraries
Conditional Branch Options • Need evaluate condition and branch to target • Condition Code (CC) Register • E.g.: X86, ARM, PPC, SPARC, … • ALU ops set condition code flags in the CCR • Branch just checks the flag • Condition register • E.g.: Alpha, MIPS • Comparison instruction puts result in a GPR • Branch instruction checks the register • Compare & Branch • E.g.: PA-RISC, VAX, BEQZ, BNE in MIPS • Compare & branch in 1 instruction.
Procedure Calling Conventions • Two major calling conventions (for saving & restoring registers): • Caller saves: • Before the call, procedure caller saves registers that will be needed later, even if callee did not use them • Callee saves: • Inside the call, called procedure saves registers that it will overwrite • Can be more efficient if many small procedures • Many architectures use a combination of schemes: • E.g., MIPS: Some registers caller-saves, some callee-saves
Three Classes of Control Instructions • SPEC CPU2000 on Alpha
Branch Distance Distribution • SPEC CPU2000 on Alpha
Branch Comparison Types • SPEC CPU2000 on Alpha
Example: MIPS32 • r0 • r1 • ° • ° • ° • r31 • 0 • Programmable storage • 2^32 x bytes • 31 x 32-bit GPRs (R0=0) • 32 x 32-bit FP regs (paired DP) • HI, LO, PC • Data types ? • Format ? • Addressing Modes? • PC • lo • hi Arithmetic logical Add, AddU, Sub, SubU, And, Or, Xor, Nor, SLT, SLTU, AddI, AddIU, SLTI, SLTIU, AndI, OrI, XorI, LUI SLL, SRL, SRA, SLLV, SRLV, SRAV Memory Access LB, LBU, LH, LHU, LW, LWL,LWR SB, SH, SW, SWL, SWR Control J, JAL, JR, JALR BEq, BNE, BLEZ,BGTZ,BLTZ,BGEZ,BLTZAL,BGEZAL • 32-bit instructions on word boundary
MIPS64 Architecture • RISC, load-store architecture, simple address • 32-bit instructions, fixed format • 32 64-bit GPRs, R0-R31. • Really, only 31 – R0 is just a constant 0. • 32 64-bit FPRs, F0-F31 • Can hold 32-bit floats also (with other ½ unused). • “SIMD” extensions operate on more floats in single FPR • A few special registers • Floating-point status register • Load/store 8-, 16-, 32-, 64-bit integers • All sign-extended to fill 64-bit GPR • Also 32- bit floats/doubles
MIPS64 Addressing Modes • Register (arith./logical ops only) – 64 bit integer, 32 or 64 floating-point; Inst. is fixed 32 bits • Immediate (arith./logical only) & Displacement (load/stores only) • 16-bit immediate / offset field • Register indirect: use 0 as displacement offset • Direct (absolute): use R0 as displacement base • Byte-addressed memory, 64-bit address • Software-settable big-endian/little-endian flag • Alignment required • Note, the course project works on MIPS32
Inst. Format: I-type Instructions • Immediate (16 bits) for imme. operand and for Displacement
Inst. Format: R-type Instructions • NOTE, R-R ALU only need 21 bits: opcode+rs+rt+rd • Shamt is for shift amount for bit-level operation • Funct is for function code; only one Opcode for RR ALU • More details see MIPS code map!!
MIPS Instruction Set • MIPS ISA in Figures A.23-26 • The MIPS32™ ISA manual provides more details for 32-bit arch.: http://www.cise.ufl.edu/class/cda5155fa12/projects/project1/mips.pdf • Branch and Jump Addresses • PC-relative addressing: • Branch target address = (PC+4) + (Displacement || “00”); Note instead of PC, (PC+4) for hardware convenience and the two zeros is added due to word displacement / alignment. • Jump (limited to 256MB range): • (Upper 4 bits of Current PC) || (26-bits address in Jump) || (“00”) • Long Jump: • Jump register: save 32-bit target address in register
MIPS Code Map • FIGURE A.10.2MIPS opcode map. The values of each field are shown to its left. The first column shows the values in base 10, and the second shows base 16 for the op field (bits 31 to 26) in the third column. This op field completely specifies the MIPS operation except for six op values: 0, 1, 16, 17, 18, and 19. These operations are determined by other fields, identified by pointers. The last field (funct) uses “f ” to mean “s” if rs = 16 and op = 17 or “d” if rs = 17 and op = 17. The second field (rs) uses “z” to mean “0”, “1”, “2”, or “3” if op = 16, 17, 18, or 19, respectively. If rs = 16, the operation is specified elsewhere: if z= 0, the operations are specified in the fourth field (bits 4 to 0); if z= 1, then the operations are in the last field with f= s. If rs = 17 and z= 1, then the operations are in the last field with f= d.