Floating Point Instructions

Question 1

Which bits are used by S and D registers?

Answer 1

S registers use the low order 32 bits, while D registers use the low order 64 bits

Question 2

How many FP registers does ARMv8 have?

Answer 2

32 128-bit registers

Question 3

How is memory allocated and initialized for FP numbers?

Answer 3

Using the .single or .double pseudo-op, with the prefix 0r
ie. .single 0r5.0

Question 4

Can FP arithmetic use immediate vlaues?

Answer 4

No, FP instructions cannot use immediates, only registers

Question 5

Which immediate values can be moved into FP registers?

Answer 5

The immediate must be expressible as +/- n / (16 x 2^r)
n in range: 16 to 31
r in range: -3 to 4

Question 6

What is the instruction to convert from a FP register to a W/X register

Answer 6

fcvtns or fcvtnu