SIC only knows how to do basic operations, such as working with registers,
integer arithmetic (all in register A), comparing register A to a word in
memory, conditional jumping, and calling and returning from a subroutine using
register L.
LDA m: Load a word from memory at address m into register A (A <- (m..m+2))LDCH m: Load a byte (character) from memory at address m into the lowest
byte of register A (A.low <- (m))LDL m: Load a word from memory at address m into register L (L <- (m..m+2))LDX m: Load a word from memory at address m into register X (X <- (m..m+2))Warning: LDCH only sets the lowest byte of register A, it leaves the other
bits unchanged, so you have to be careful when reading the value of register A
after using LDCH.
STA m: Store a word from register A into memory at address m (m..m+2 <- (A))STCH m: Store the lowest byte (character) from register A into memory at
address m (m <- (A.low))STL m: Store a word from register L into memory at address m (m..m+2 <- (L))STSW m: Store a word from register SW into memory at address m (m..m+2 <- (SW))STX m: Store a word from register X into memory at address m (m..m+2 <- (X))(m..m+2) means the value in memory from address (byte) m to m+2,
which is 3 bytes, i.e. a word.
All arithmetic instructions use the register A and store the result back into
it.
ADD m: Sum the value in register A with a word from memory at address m
and store the result back into register A (A <- (A) + (m..m+2))DIV m: Divide the value in register A with a word from memory at address
m and store the result back into register A (A <- (A) / (m..m+2))MUL m: Multiply the value in register A with a word from memory at address
m and store the result back into register A (A <- (A) * (m..m+2))SUB m: Subtract a word from memory at address m from the value in register
A and store the result back into register A (A <- (A) - (m..m+2))AND m: Do a bitwise AND operation on the value of register A and a word
from memory at address m, and store the result into register A
(A <- (A) & (m..m+2))OR m: Do a bitwise OR operation on the value of register A and a word from
memory at address m, and store the result into register A
(A <- (A) | (m..m+2))COMP m: Compare the value of register A with a word in memory at address
m and set the value of the CC bits according to the result
(CC <- (A) : (m..m+2))TIX m: Increment the value of register X by 1, store it back into the
register, compare the value of the register to a word from memory at address m,
and set the value of the CC bits according to the result
(X <- (X) + 1; CC <- (X) : (m..m+2))There are three possible outcomes of the comparison:
(A/X) = (m..m+2) then CC is (=)(A/X) > (m..m+2) then CC is (>)(A/X) < (m..m+2) then CC is (<)The actual underlying value for each of the results isn’t mentioned in the SIC
specification, so it’s up to each simulator to implement it and then consistently
use it (e.g 00 for =, 01 for > and 10 for <).
The TIX instruction is commonly used when counting or doing the assembly
version of a for loop.
All jump instructions set the value of the PC register to the address to jump to.
J m: Unconditionally jump to address m (PC <- m)JEQ m: Jump to address m if CC is = (PC <- m if CC is =)JGT m: Jump to address m if CC is > (PC <- m if CC is >)JLT m: Jump to address m if CC is < (PC <- m if CC is <)JSUB m: Store the value of the PC register to the L register and then
jump to address m (set the value of the PC register to the address m -
L <- (PC); PC <- m).RSUB: Set the value of the PC register to the value in register L, i.e.
restore the previous value and thus jump back to the parent routine at the
place you left off (PC <- (L)).TODO: Check if PC is incremented before it is stored in JSUB, since we want to go to the next command, not the same one
SIC supports writing to and reading from external devices (usually text files and/or standard input/output in simulator implementations). Each device has an 8-bit ID (so there can be at most 256 devices), which has to be stored somewhere in memory.
RD m: Read a byte (character) from the device with ID written at address
m and store it into the lowest byte of register A (A.low <- dev[(m)])TD m: Test device with ID written at address m (CC <- test(dev[(m)]))
CC is set to <CC is set to =WD m: Write a byte (character) from register A to the device with ID
written at address m (dev[(m)] <- A.low)dev[(m)] means device, whose ID is stored in memory at the address m.
SIC/XE can use all of the instructions that SIC supports, but it can use advanced addressing modes to adapt them to its larger memory size. It also introduces a few new instructions, which are not supported on the bare SIC.
Things to keep in mind:
r2 <- (r2) op (r1))F register deal with 48 bit values (6 bytes, 2 words)SIC/XE adds a few new register instructions, but those that are used in arithmetic operations will be listed in the next section.
CLEAR r1: Sets the value of the r1 register to 0 - resets it (r1 <- 0)RMO r1, r2: Sets r2 to the value of r1 (r2 <- (r1))SHIFTL r1, n: Shifts the value of r1 by n bits to the left (r1 <- (r1) << n)SHIFTR r1, n: Shifts the value of r1 by n bits to the right (r1 <- (r1) >> n)LDB m: Load a word from memory at address m into register B (B <- (m..m+2))LDF m: Load two words from memory at address m into register F (F <- (m..m+5))LDS m: Load a word from memory at address m into register S (S <- (m..m+2))LDT m: Load a word from memory at address m into register T (T <- (m..m+2))STB: Store a word from register B into memory at address m (m..m+2 <- (B))STF: Store two words from register F into memory at address m (m..m+5 <- (F))STS: Store a word from register S into memory at address m (m..m+2 <- (S))STT: Store a word from register T into memory at address m (m..m+2 <- (T))FIX: Stores the lowest word of register F in register A - converts the
float to an int, but just cuts the top part (A <- int(F) or A <- (F.low))FLOAT: Stores the value of register A in register F - converts the int
to a float, but it doesn’t just copy the value, it converts it into the correct
format, with a sign, exponent and mantissa (F <- float(A))NORM: Normalizes the value of the F register (F <- norm(F))New (SIC/XE-only) arithmetic instructions use registers for calculating and storing the result (of course SIC/XE can also use the existing SIC arithmetic instructions as well).
ADDF m: Sum the value in register F with 2 words (6 bytes) from memory at
address m and store the result back into register F (F <- (F) + (m..m+5))ADDR r1, r2: Sum the values of the two registers and store the result into
the 2nd register (r2 <- (r2) + (r1))DIVF m: Divide the value in register F with two words from memory at address
m and store the result back into register F (F <- (F) / (m..m+5))DIVR r1, r2: Divide the value in register 2 with the value in register 1 and
store the result in register 2 (r2 <- (r2) / (r1))MULF m: Multiply the value in register F with two words from memory at address
m and store the result back into register F (F <- (F) * (m..m+5))MULR r1, r2: Multiply the values of the two registers and store the result into
the 2nd register (r2 <- (r2) * (r1))SUBF m: Subtract two words from memory at address m from the value in register
F and store the result back into register F (F <- (F) - (m..m+5))SUBR r1, r2: Subtract the value in register 1 from the value in register 2 and
store the result back in register 2 (r2 <- (r2) - (r1))COMPF m: Compare the value in register F with two words from memory at address
m and set the value of the CC bits according to the result (CC <- (F) : (m..m+5))COMPR r1, r2: Compare the values of registers 1 and 2 and set the value of
the CC bits according to the result (CC <- (r1) : (r2))TIXR r1: Increment the value of register X by 1, store it back into the
register, and compare the value in register X with the value in register r1,
and set the value of the CC bits according to the result
(X <- (X) + 1; CC <- (X) : (r1))The value of CC is set in the same way as with SIC.
HIO: TODOLPS m: Restore the values of registers after returning from an interrupt
routine - load processor status from memory beginning at address m (PS <- (m..m+2))SIO: TODOSSK m: TODOSTI m: TODOSVC n: TODOTIO: TODO