2ce0274303
Use a 32-bit limb size ("like a digit, but bigger") for floating-point
conversion. This cuts the number of multiplications per constant by a
factor of four.
This means supporting fractional-limb-sized numbers, so while we're at
it, add support for 8-bit floating point numbers (apparently used in
graphics and in audio compression applications.)
57 lines
476 B
Text
57 lines
476 B
Text
#
|
|
# Tokens other than instructions and registers
|
|
#
|
|
|
|
% TOKEN_PREFIX, 0, P_*
|
|
a16
|
|
a32
|
|
a64
|
|
asp
|
|
lock
|
|
o16
|
|
o32
|
|
o64
|
|
osp
|
|
rep
|
|
repe
|
|
repne
|
|
repnz
|
|
repz
|
|
times
|
|
|
|
% TOKEN_SPECIAL, 0, S_*
|
|
abs
|
|
byte
|
|
dword
|
|
far
|
|
long
|
|
near
|
|
nosplit
|
|
oword
|
|
qword
|
|
rel
|
|
short
|
|
strict
|
|
to
|
|
tword
|
|
word
|
|
|
|
% TOKEN_FLOAT, 0, 0
|
|
__infinity__
|
|
__nan__
|
|
__qnan__
|
|
__snan__
|
|
|
|
% TOKEN_FLOATIZE, 0, FLOAT_{__float*__}
|
|
__float8__
|
|
__float16__
|
|
__float32__
|
|
__float64__
|
|
__float80m__
|
|
__float80e__
|
|
__float128l__
|
|
__float128h__
|
|
|
|
% TOKEN_*, 0, 0
|
|
seg
|
|
wrt
|