at processor initialization, 4–83bit descriptions, 4–80instructions to read/write, 4–109operate instructions that use, 4–102saving and restoring, 4–83
trap disable bits in, 4–78Floating-point convert instructions, 3–14
Fa field requirements, 3–14Floating-point division, performance impact of,
A–10
Floating-point format, number representation (encodings), 4–65
Floating-point instructions
branch, 4–99faults, 4–62
function field format, 4–84introduced, 4–62memory format, 4–90
opcodes and format summarized, C–1operate, 4–102
rounding modes, 4–66terminology, 4–63trapping modes, 4–69traps, 4–62
Floating-point load instructions, 4–90
load F_floating, 4–91load G_floating, 4–92load S_floating, 4–93load T_floating, 4–94with non-finite values, 4–90
Floating-point operate instructions, 4–102
add (IEEE), 4–111add (VAX), 4–110compare (IEEE), 4–113compare (VAX), 4–112conditional move, 4–107
convert IEEE floating to integer, 4–117convert integer to IEEE floating, 4–118convert integer to integer, 4–106convert integer to VAX floating, 4–115convert S_floating to T_floating, 4–119convert T_floating to S_floating, 4–120convert VAX floating to integer, 4–114convert VAX floating to VAX floating, 4–116copy sign, 4–105
divide (IEEE), 4–122divide (VAX), 4–121format of, 3–13
from integer moves, 4–124move from/to FPCR, 4–109multiply (IEEE), 4–127multiply (VAX), 4–126subtract (IEEE), 4–131subtract (VAX), 4–130
to integer moves, 4–123
unused function codes with, 3–14
Floating-point registers, 3–2
Floating-point single-precision operations, 4–62
Floating-point store instructions, 4–90
store F_floating, 4–95store G_floating, 4–96store S_floating, 4–97store T_floating, 4–98with non-finite values, 4–90
Floating-point support
floating-point control (FP_C) quadword, B–4IEEE, 2–6
IEEE standard 754-1985,4–88instruction overview, 4–62longword integer, 2–11operate instructions, 4–102optional, 4–2
quadword integer, 2–12rounding modes, 4–66single-precision operations, 4–62trap modes, 4–69
VAX, 2–3
Floating-point to integer move, 4–123
Floating-point to integer move instructions, 3–14
Floating-point trapping modes, 4–69 See also Arithmetic traps
FNOP code form, A–11
FP_C quadword, B–4
FPCR. See Floating-point control register
FTOIS instruction, 4–123
FTOIT instruction, 4–123
Function codes
IEEE floating-point,C–6in numerical order, C–10independent floating-point,C–8VAX floating-point,C–7
See also Opcodes
G
G_floating data type, 2–4
alignment of, 2–5mapping, 2–5MAX/MIN, 4–65
GENTRAP (PALcode) instruction required recognition of, 6–4
gentrap (PALcode) instruction required recognition of, 6–4
H
HALT (PALcode) instruction required, 6–7
halt (PALcode) instruction required, 6–7
