Computer Architecture Lab/SS2013/Group3/Assignment1

Description of the microcontroller architectures

edit

Although the AVR architecture comes in 3 main families of microcontrollers, plus other more specialized series, we will be discussing the AVR32 architecture. This architecture is a reduced instruction set computer (RISC) with a 32-bit datapath

Architecture

edit
  • data path on 32 bits
  • 3 to 7 stage pipeline
  • 15 general purpose registers
  • Big endianness
  • non-volatile flash memory for the program memory (has the advantage that the microcontroller keeps its program even after a power outage)
  • data memory consists of the register file, the I/O registers (for controlling the different peripherals on the microcontroller) and a SRAM memory
  • almost all AVR microcontroller have an internal EEPROM, which is for semi-permanent data storage
  • execution stage works in register-to-register mode

Instruction Set

edit
Atmel AVR instruction set overview, courtesy of Wikipedia
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 Instruction
0 0 0 0 0 0 0 1 d d d d r r r r MOVW Move register pair
0 0 0 0 0 0 1 d d d r r r Signed and fractional multiply (R16–R23 only)
0 0 0 0 0 1 r d d d d d r r r r 2-operand instructions
CPC, SBC, ADD, CPSE, CP,
SUB. ADC, AND, EOR, OR, MOV
0 0 0 0 1
0 0 0 1
0 0 1 0
0 0 1 1 K K K K h h h h K K K K Register-immediate instructions
CPI, SBCI, SUBI, ORI, ANDI
0 1
1 0 k 0 k k s d d d d d y k k k LDD/STD to Z+k or Y+k
1 0 0 1 0 0 s d d d d d LD/ST other
1 0 0 1 0 1 0 d d d d d 0 1-operand instructions (COM, NEG, SWAP, etc.)
1 0 0 1 0 1 0 0 b b b 1 0 0 0 SEx/CLx Status register clear/set bit
1 0 0 1 0 1 0 1 1 0 0 0 Misc instructions (RET, RETI, SLEEP, etc.)
1 0 0 1 0 1 0 c 0 0 0 1 0 0 1 Indirect jump/call to Z or EIND:Z
1 0 0 1 0 1 0 d d d d d 1 0 1 0 DEC Rd
1 0 0 1 0 1 0 0 k k k k 1 0 1 1 DES round k
1 0 0 1 0 1 0 k k k k k 1 1 c k JMP/CALL abs22
1 0 0 1 0 1 1 k k p p k k k k ADIW/SBIW Rp,uimm6
1 0 0 1 1 0 B a a a a a b b b I/O space bit operations
1 0 0 1 1 1 r d d d d d r r r r MUL, unsigned: R1:R0 = Rr×Rd
1 0 k 0 k k s d d d d d y k k k See 10k0 above
1 0 1 1 s a a d d d d d a a a a OUT/IN to I/O space
1 1 0 c 12 bit signed offset Relative jump/call to PC ± 2×simm12
1 1 1 0 K K K K h h h h K K K K LDI Rh,K
1 1 1 1 0 7-bit signed offset b b b Conditional branch on status register bit
1 1 1 1 1 0 s d d d d d 0 b b b BLD/BST register bit to STATUS.T
1 1 1 1 1 1 B d d d d d 0 b b b SBRC/SBRS skip if register bit equals B

Based on the high market share of the ARM-based chips, we can safely say that it is the most used reduced instruction set computer (RISC) today.

Architecture

edit
  • data path on 32 or 64 bits
  • 3 stage pipeline (up to ARM7 version), up to 13 stages for higher performance versions
  • 16 general purpose registers, some having special function (e.g. Stack Pointer, Link Register, Program Counter)
  • Bi-endianness, with little endian as default
  • support for 16 coprocessors, connected in a non-intrusive way
  • floating point based on VPF (Vector Floating Point) technology, as coprocessor
  • execution stage works in register-to-register mode

Instruction Set

edit
  • ARM relies on a load/store architecture
  • 32-bit wide instructions
  • mostly single clock cycle execution
  • link register for fast leaf function calls
  • 2 priority level interrupt subsystem
  • supports add, subtract and multiply instructions, with some cores implementing integer divide
  • fold shifts/rotates into the data processing, reducing memory accesses and therefore improving pipeline efficiency

ARM resources

MIPS

edit

MIPS (originally an acronym for Microprocessor without Interlocked Pipeline Stages) is a reduced instruction set computer (RISC) instruction set architecture (ISA) developed by MIPS Technologies (formerly MIPS Computer Systems, Inc.). Early versions had a 32-bit wide datapath, with latter ones at 64-bit.

Architecture (considering R2000 version)

edit
 
Pipeline diagram of the MIPS architecture - courtesy of Wikipedia
  • data path on 32 bits
  • 5 pipeline stages:
  1. Instruction Fetch
  2. Instruction Decode
  3. Execute
  4. Memory read/store
  5. Writeback
  • 32 32-bit general purpose registers (GPR)
  • could be booted either as big-endian or little-endian
  • support for up to 4 co-processors
  • could plug in the R2010 Floating Point Unit (FPU), having 32 32-bit registers (which could be used as 16 64-bit registers, for double precision)
  • external SRAM cache
  • internal DRAM (very slow writing time)
  • execution stage is of type register-to-register, the ALU obtaining values from the register file and writing back to the register file during writeback stage

Instruction Set

edit

We are assuming an 32-bit MIPS processor, therefore the instruction size is of 32 bits. The instruction set is based around 3 types of instructions:

  • R - for operations based around values in the registers (arithmetic, logical, data transfer, unconditional jump register)
  • I - for operations based around values in the registers as well as an immediate value
  • J - for jump instructions, having only an immediate value specified (due to the need of all 26 bits for the address space)

Instruction formats are as follows:

Type 31 format (bits) 0
R opcode (6) rs (5) rt (5) rd (5) shamt (5) funct(6)
I opcode (6) rs (5) rt (5) immediate (16)
J opcode (6) address (26)

For the complete list of instructions check this link.