TRAPB (trap barrier) instruction

described, 4–144with FPCR, 4–84

True result, 4–64

True zero, 4–65

U

UMULH instruction, 4–36with MULQ, 4–35

UNALIGNED data objects, 1–8Unconditional long jump, 4–23

UNDEFINED operations, 1–7

Underflow enable (UNFE) FP_C quadword bit, B–6

Underflow status (UNFS) FP_C quadword bit, B–5

UNDZ bit. See Trap disable bits, underflow to zero

UNF bit

See also Arithmetic traps, underflow UNFD bit. See Trap disable bits, underflow

UNOP code form, A–11

UNORDERED memory references, 5–10

Unpack to bytes instructions, 4–156

UNPKBL (Unpack bytes to longwords) instruction, 4–156

UNPKBW (Unpack bytes to words) instruction, 4–156

UNPREDICTABLE results, 1–7

Updated datum, 5–6

V

VAX compatibility instructions, restrictions for, 4–149

VAX compatibility register, 3–3

VAX floating-point

D_floating, 2–5

F_floating, 2–3

G_floating, 2–4

See also Floating-point instructions VAX floating-point instructions

add instructions, 4–110

compare instructionsCMPGEQ instruction, 4–112

convert from integer instructions, 4–115convert to integer instructions, 4–114convert VAX floating format instructions,

4–116

divide instructions, 4–121from integer move, 4–124function codes for, C–7

function field format, 4–87multiply instructions, 4–126operate instructions, 4–102square root instructions, 4–128subtract instructions, 4–130

VAX rounding modes, 4–66

Vector instructions

byte and word maximum, 4–152byte and word minimum, 4–152

Virtual D-cache,5–4

Virtual I-cache,5–4maintaining coherency of, 5–5

Visibility, defined, 5–14

W

Waivers, E–1

WH64 (Write hint) instruction, 4–145

WH64 instruction lock_flag with, 4–10

Windows NT Alpha PALcode, instruction summary, C–17

WMB (Write memory barrier) instruction, 4–147

atomic operations with, 5–8compared with MB, 4–148with shared data structures, 5–9

Word data type, 2–1atomic access of, 5–3

Write buffers, requirements for, 5–5

Write-back caches, requirements for, 5–5

wrunique (PALcode) instruction required recognition of, 6–4

X

x MOD y operator, 3–8

X_floating data type, 2–9

alignment of, 2–10big-endian format, 2–10MAX/MIN, 4–65

Index–12

Page 370
Image 370
Compaq ECQD2KCTE manual Index-12