Floating point instructions

• Separate set of registers for fp operations ($f0, $f1, ..., $f31)

• For double-precision, uses a pair of registers. So $f0 used in double-precision operation is really $f0 and $f1.

• Single-precision indicated by .s and double by .d
• Instructions:

  l.s  $f1, 0($t0)   # load word at address stored in $t0 into $f1
  s.s  $f1, 0($t0)   # store contents of $f1 into word at address stored in $t0
  add.s $f0, $f1, $f2 # add contents of $f1 and $f2 as single-precision fp 
                      #  numbers and store in $f0
  add.d $f0, $f2, $f4 # add contents of $f2$f3 and $f4$f5 as double-precision fp 
                      #  numbers and store in $f0$f1
  

  Others on p. 288