def __init__(self, inf_bound, sup_bound):
self.inf_bound = inf_bound
self.sup_bound = sup_bound
self.zero_in_interval = 0 in sollya.Interval(
inf_bound, sup_bound)
self.min_exp = None if self.zero_in_interval else min(
sollya.ceil(sollya.log2(abs(inf_bound))),
sollya.ceil(sollya.log2(abs(sup_bound))))
self.max_exp = max(sollya.ceil(sollya.log2(abs(inf_bound))),
sollya.ceil(sollya.log2(abs(sup_bound))))
def get_smaller_format_min_error(self, ml_format):
""" return the maximal accuracy / minimal error
of the format just before @p ml_format in term
of size """
MIN_ERROR_MAP = {
ML_Binary64: -sollya.log2(
0
), # no format smaller than ML_Binary6(switch to fp32 not managed)
ML_DoubleDouble: S2**-53,
ML_TripleDouble: S2**-106,
ML_Binary32: -sollya.log2(0), # no format smaller than ML_Binary32
ML_SingleSingle: S2**-24,
ML_TripleSingle: S2**-48,
}
return MIN_ERROR_MAP[ml_format]
def split_domain(starting_domain, slivers):
in_domains = [starting_domain]
out_domains = list()
while len(in_domains) > 0:
I = in_domains.pop()
unround_e = sollya.log2(I)
e_low = sollya.floor(sollya.inf(unround_e))
e_high = sollya.floor(sollya.sup(unround_e))
#print("in: [{}, {}] ({}, {})".format(float(sollya.inf(I)), float(sollya.sup(I)), int(e_low), int(e_high)))
if e_low == e_high:
#print(" accepted")
out_domains.append(I)
continue
e_range = sollya.Interval(e_low, e_low+1)
I_range = 2**e_range
for _ in range(100):
mid = sollya.mid(I_range)
e = sollya.floor(sollya.log2(mid))
if e == e_low:
I_range = sollya.Interval(mid, sollya.sup(I_range))
else:
I_range = sollya.Interval(sollya.inf(I_range), mid)
divider_high = sollya.sup(I_range)
divider_low = sollya.inf(I_range)
lower_part = sollya.Interval(sollya.inf(I), divider_low)
upper_part = sollya.Interval(divider_high, sollya.sup(I))
#print(" -> [{}, {}]".format(float(sollya.inf(lower_part)), float(sollya.sup(lower_part))))
#print(" -> [{}, {}]".format(float(sollya.inf(upper_part)), float(sollya.sup(upper_part))))
in_domains.append(upper_part)
in_domains.append(lower_part)
in_domains = out_domains
# subdivide each section into 2**subd sections
for _ in range(slivers):
out_domains = list()
for I in in_domains:
mid = sollya.mid(I)
out_domains.append(sollya.Interval(sollya.inf(I), mid))
out_domains.append(sollya.Interval(mid, sollya.sup(I)))
in_domains = out_domains
in_domains = set(in_domains)
in_domains = sorted(in_domains, key=lambda x:float(sollya.inf(x)))
in_domains = [d for d in in_domains if sollya.inf(d) != sollya.sup(d)]
return in_domains
def get_lzc_output_width(width):
""" Compute the size of a standard leading zero count result for
a width-bit output
@param width [int] input width
@return output width (in bits) """
return int(floor(log2(width))) + 1
def get_integer_coding(self, value, language=C_Code):
if FP_SpecialValue.is_special_value(value):
return self.get_special_value_coding(value, language)
elif value == ml_infty:
return self.get_special_value_coding(FP_PlusInfty(self), language)
elif value == -ml_infty:
return self.get_special_value_coding(FP_MinusInfty(self), language)
else:
value = sollya.round(value, self.get_sollya_object(), sollya.RN)
# FIXME: managing negative zero
sign = int(1 if value < 0 else 0)
value = abs(value)
if value == 0.0:
Log.report(Log.Warning,
"+0.0 forced during get_integer_coding conversion")
exp_biased = 0
mant = 0
else:
exp = int(sollya.floor(sollya.log2(value)))
exp_biased = int(exp - self.get_bias())
if exp < self.get_emin_normal():
exp_biased = 0
mant = int((value / S2**self.get_emin_subnormal()))
else:
mant = int(
(value / S2**exp - 1.0) * (S2**self.get_field_size()))
return mant | (exp_biased << self.get_field_size()) | (
sign << (self.get_field_size() + self.get_exponent_size()))
def solve_format_Constant(optree):
""" Legalize Constant node """
assert isinstance(optree, Constant)
value = optree.get_value()
if FP_SpecialValue.is_special_value(value):
return optree.get_precision()
elif not optree.get_precision() is None:
# if precision is already set (manually forced), returns it
return optree.get_precision()
else:
# fixed-point format solving
frac_size = -1
FRAC_THRESHOLD = 100 # maximum number of frac bit to be tested
# TODO: fix
for i in range(FRAC_THRESHOLD):
if int(value*2**i) == value * 2**i:
frac_size = i
break
if frac_size < 0:
Log.report(Log.Error, "value {} is not an integer, from node:\n{}", value, optree)
abs_value = abs(value)
signed = value < 0
# int_size = max(int(sollya.ceil(sollya.log2(abs_value+2**frac_size))), 0) + (1 if signed else 0)
int_size = max(int(sollya.ceil(sollya.log2(abs_value + 1))), 0) + (1 if signed else 0)
if frac_size == 0 and int_size == 0:
int_size = 1
return fixed_point(int_size, frac_size, signed=signed)
def solve_format_CLZ(optree):
""" Legalize CountLeadingZeros precision
Args:
optree (CountLeadingZeros): input node
Returns:
ML_Format: legal format for CLZ
"""
assert isinstance(optree, CountLeadingZeros)
op_input = optree.get_input(0)
input_precision = op_input.get_precision()
if is_fixed_point(input_precision):
if input_precision.get_signed():
Log.report(Log.Warning , "signed format in solve_format_CLZ")
# +1 for carry overflow
int_size = int(sollya.floor(sollya.log2(input_precision.get_bit_size()))) + 1
frac_size = 0
return fixed_point(
int_size,
frac_size,
signed=False
)
else:
Log.report(Log.Warning , "unsupported format in solve_format_CLZ")
return optree.get_precision()
def numeric_emulate(self, vx, n):
""" Numeric emulation of n-th root """
if FP_SpecialValue.is_special_value(vx):
if is_nan(vx):
return FP_QNaN(self.precision)
elif is_plus_infty(vx):
return SOLLYA_INFTY
elif is_minus_infty(vx):
if int(n) % 2 == 1:
return vx
else:
return FP_QNaN(self.precision)
elif is_zero(vx):
if int(n) % 2 != 0 and n < 0:
if is_plus_zero(vx):
return FP_PlusInfty(self.precision)
else:
return FP_MinusInfty(self.precision)
elif int(n) % 2 == 0:
if n < 0:
return FP_PlusInfty(self.precision)
elif n > 0:
return FP_PlusZero(self.precision)
return FP_QNaN(self.precision)
else:
raise NotImplementedError
# OpenCL-C rootn, x < 0 and y odd: -exp2(log2(-x) / y)
S2 = sollya.SollyaObject(2)
if vx < 0:
if int(n) % 2 != 0:
if n > 0:
v = -bigfloat.root(
sollya.SollyaObject(-vx).bigfloat(), int(n))
else:
v = -S2**(sollya.log2(-vx) / n)
else:
return FP_QNaN(self.precision)
elif n < 0:
# OpenCL-C definition
v = S2**(sollya.log2(vx) / n)
else:
v = bigfloat.root(sollya.SollyaObject(vx).bigfloat(), int(n))
return sollya.SollyaObject(v)
def get_accuracy_from_epsilon(epsilon):
""" convert a numerical relative error into
a number of accuracy bits
:param epsilon: error to convert
:type epsilon: number
:return: accuracy corresponding to the error
:rtype: SollyaObject
"""
return sollya.floor(-sollya.log2(abs(epsilon)))
def get_fixed_type_from_interval(interval, precision):
""" generate a fixed-point format which can encode
@p interval without overflow, and which spans
@p precision bits """
lo = inf(interval)
hi = sup(interval)
signed = True if lo < 0 else False
msb_index = int(floor(sollya.log2(max(abs(lo), abs(hi))))) + 1
extra_digit = 1 if signed else 0
return fixed_point(msb_index + extra_digit,
-(msb_index - precision),
signed=signed)
def computeBoundPower(self, k, out_format, var_format):
# TODO: fix
epsilon = out_format.mp_node.epsilon
eps_target = out_format.eps_target
# to avoid derivating to larger and larger format when post-processing
# powerings, we over-estimate the error while matching eps_target
if eps_target > epsilon and epsilon > 0:
# limiting error to limit precision explosion
l_eps_target = sollya.log2(eps_target)
l_epsilon = sollya.log2(epsilon)
virtual_error_log = (l_eps_target + l_epsilon) / 2.0
virtual_error = sollya.evaluate(
sollya.SollyaObject(S2**(virtual_error_log)), 1)
print(
"lying on power_error target=2^{}, epsilon=2^{}, virtual_error=2^{} / {}"
.format(l_eps_target, l_epsilon, virtual_error_log,
virtual_error))
return virtual_error
else:
return sollya.SollyaObject(epsilon)
def __init__(self, low_exp_value, max_exp_value, field_bits, precision):
self.field_bits = field_bits
self.low_exp_value = low_exp_value
self.max_exp_value = max_exp_value
exp_bits = int(
sollya.ceil(sollya.log2(max_exp_value - low_exp_value + 1)))
assert exp_bits >= 0 and field_bits >= 0 and (exp_bits +
field_bits) > 0
self.exp_bits = exp_bits
self.split_num = (self.max_exp_value - self.low_exp_value +
1) * 2**(self.field_bits)
Log.report(Log.Debug, "split_num={}", self.split_num)
self.precision = precision
def generate_scheme(self):
lzc_width = int(floor(log2(self.width))) + 1
Log.report(Log.Info, "width of lzc out is {}".format(lzc_width))
input_precision = ML_StdLogicVectorFormat(self.width)
precision = ML_StdLogicVectorFormat(lzc_width)
# declaring main input variable
vx = self.implementation.add_input_signal("x", input_precision)
vr_out = Signal("lzc", precision=precision, var_type=Variable.Local)
tmp_lzc = Variable("tmp_lzc",
precision=precision,
var_type=Variable.Local)
iterator = Variable("i", precision=ML_Integer, var_type=Variable.Local)
lzc_loop = RangeLoop(
iterator,
Interval(0, self.width - 1),
ConditionBlock(
Comparison(VectorElementSelection(vx,
iterator,
precision=ML_StdLogic),
Constant(1, precision=ML_StdLogic),
specifier=Comparison.Equal,
precision=ML_Bool),
ReferenceAssign(
tmp_lzc,
Conversion(Subtraction(Constant(self.width - 1,
precision=ML_Integer),
iterator,
precision=ML_Integer),
precision=precision),
)),
specifier=RangeLoop.Increasing,
)
lzc_process = Process(Statement(
ReferenceAssign(tmp_lzc, Constant(self.width,
precision=precision)), lzc_loop,
ReferenceAssign(vr_out, tmp_lzc)),
sensibility_list=[vx])
self.implementation.add_process(lzc_process)
self.implementation.add_output_signal("vr_out", vr_out)
return [self.implementation]
def solve_format_Constant(optree):
""" Legalize Constant node """
assert isinstance(optree, Constant)
value = optree.get_value()
if FP_SpecialValue.is_special_value(value):
return optree.get_precision()
elif not optree.get_precision() is None:
# if precision is already set (manually forced), returns it
return optree.get_precision()
else:
# fixed-point format solving
assert int(value) == value
abs_value = abs(value)
signed = value < 0
int_size = max(int(sollya.ceil(sollya.log2(abs_value + 1))),
0) + (1 if signed else 0)
frac_size = 0
if frac_size == 0 and int_size == 0:
int_size = 1
return fixed_point(int_size, frac_size, signed=signed)
def computeNeededVariableFormat(self, I, epsTarget, variableFormat):
if epsTarget > 0:
# TODO: fix to support ML_Binary32
if epsTarget >= self.MIN_LIMB_ERROR or variableFormat.mp_node.precision is self.limb_format:
# FIXME: default to minimal precision (self.limb_format)
return variableFormat
else:
target_accuracy = sollya.ceil(-sollya.log2(epsTarget))
target_format = self.get_format_from_accuracy(
target_accuracy,
eps_target=epsTarget,
interval=variableFormat.mp_node.interval)
if target_format.mp_node.precision.get_bit_size(
) < variableFormat.mp_node.precision.get_bit_size():
return target_format
else:
# if variableFormat is smaller (less bits) and more accurate
# then we use it
return variableFormat
else:
return variableFormat
def determine_minimal_fixed_format_cst(value):
""" determine the minimal size format which can encode
exactly the constant value value """
# fixed-point format solving
frac_size = -1
FRAC_THRESHOLD = 100 # maximum number of frac bit to be tested
# TODO: fix
for i in range(FRAC_THRESHOLD):
if int(value * 2**i) == value * 2**i:
frac_size = i
break
if frac_size < 0:
Log.report(Log.Error, "value {} is not an integer, from node:\n{}",
value, optree)
abs_value = abs(value)
signed = value < 0
# int_size = max(int(sollya.ceil(sollya.log2(abs_value+2**frac_size))), 0) + (1 if signed else 0)
int_size = max(int(sollya.ceil(sollya.log2(abs_value + 1))),
0) + (1 if signed else 0)
if frac_size == 0 and int_size == 0:
int_size = 1
return fixed_point(int_size, frac_size, signed=signed)
def generate_scheme(self):
vx = self.implementation.add_input_variable("x",
self.get_input_precision())
sollya_precision = self.get_input_precision().get_sollya_object()
# local overloading of RaiseReturn operation
def ExpRaiseReturn(*args, **kwords):
kwords["arg_value"] = vx
kwords["function_name"] = self.function_name
return RaiseReturn(*args, **kwords)
# testing special value inputs
test_nan_or_inf = Test(vx,
specifier=Test.IsInfOrNaN,
likely=False,
debug=True,
tag="nan_or_inf")
test_nan = Test(vx,
specifier=Test.IsNaN,
debug=True,
tag="is_nan_test")
test_positive = Comparison(vx,
0,
specifier=Comparison.GreaterOrEqual,
debug=True,
tag="inf_sign")
test_signaling_nan = Test(vx,
specifier=Test.IsSignalingNaN,
debug=True,
tag="is_signaling_nan")
# if input is a signaling NaN, raise an invalid exception and returns
# a quiet NaN
return_snan = Statement(
ExpRaiseReturn(ML_FPE_Invalid,
return_value=FP_QNaN(self.precision)))
vx_exp = ExponentExtraction(vx, tag="vx_exp", debug=debugd)
int_precision = self.precision.get_integer_format()
# log2(vx)
# r = vx_mant
# e = vx_exp
# vx reduced to r in [1, 2[
# log2(vx) = log2(r * 2^e)
# = log2(r) + e
#
## log2(r) is approximated by
# log2(r) = log2(inv_seed(r) * r / inv_seed(r)
# = log2(inv_seed(r) * r) - log2(inv_seed(r))
# inv_seed(r) in ]1/2, 1] => log2(inv_seed(r)) in ]-1, 0]
#
# inv_seed(r) * r ~ 1
# we can easily tabulate -log2(inv_seed(r))
#
# retrieving processor inverse approximation table
dummy_var = Variable("dummy", precision=self.precision)
dummy_div_seed = DivisionSeed(dummy_var, precision=self.precision)
inv_approx_table = self.processor.get_recursive_implementation(
dummy_div_seed,
language=None,
table_getter=lambda self: self.approx_table_map)
# table creation
table_index_size = 7
log_table = ML_NewTable(dimensions=[2**table_index_size, 2],
storage_precision=self.precision,
tag=self.uniquify_name("inv_table"))
# value for index 0 is set to 0.0
log_table[0][0] = 0.0
log_table[0][1] = 0.0
for i in range(1, 2**table_index_size):
#inv_value = (1.0 + (self.processor.inv_approx_table[i] / S2**9) + S2**-52) * S2**-1
#inv_value = (1.0 + (inv_approx_table[i][0] / S2**9) ) * S2**-1
#print inv_approx_table[i][0], inv_value
inv_value = inv_approx_table[i][0]
value_high_bitsize = self.precision.get_field_size() - (
self.precision.get_exponent_size() + 1)
value_high = round(log2(inv_value), value_high_bitsize, sollya.RN)
value_low = round(
log2(inv_value) - value_high, sollya_precision, sollya.RN)
log_table[i][0] = value_high
log_table[i][1] = value_low
def compute_log(_vx, exp_corr_factor=None):
_vx_mant = MantissaExtraction(_vx,
tag="_vx_mant",
precision=self.precision,
debug=debug_lftolx)
_vx_exp = ExponentExtraction(_vx, tag="_vx_exp", debug=debugd)
# The main table is indexed by the 7 most significant bits
# of the mantissa
table_index = inv_approx_table.index_function(_vx_mant)
table_index.set_attributes(tag="table_index", debug=debuglld)
# argument reduction
# Using AND -2 to exclude LSB set to 1 for Newton-Raphson convergence
# TODO: detect if single operand inverse seed is supported by the targeted architecture
pre_arg_red_index = TypeCast(BitLogicAnd(
TypeCast(DivisionSeed(_vx_mant,
precision=self.precision,
tag="seed",
debug=debug_lftolx,
silent=True),
precision=ML_UInt64),
Constant(-2, precision=ML_UInt64),
precision=ML_UInt64),
precision=self.precision,
tag="pre_arg_red_index",
debug=debug_lftolx)
arg_red_index = Select(Equal(table_index, 0),
1.0,
pre_arg_red_index,
tag="arg_red_index",
debug=debug_lftolx)
_red_vx = FMA(arg_red_index, _vx_mant, -1.0)
_red_vx.set_attributes(tag="_red_vx", debug=debug_lftolx)
inv_err = S2**-inv_approx_table.index_size
red_interval = Interval(1 - inv_err, 1 + inv_err)
# return in case of standard (non-special) input
_log_inv_lo = TableLoad(log_table,
table_index,
1,
tag="log_inv_lo",
debug=debug_lftolx)
_log_inv_hi = TableLoad(log_table,
table_index,
0,
tag="log_inv_hi",
debug=debug_lftolx)
Log.report(Log.Verbose, "building mathematical polynomial")
approx_interval = Interval(-inv_err, inv_err)
poly_degree = sup(
guessdegree(
log2(1 + sollya.x) / sollya.x, approx_interval, S2**
-(self.precision.get_field_size() * 1.1))) + 1
sollya.settings.display = sollya.hexadecimal
global_poly_object, approx_error = Polynomial.build_from_approximation_with_error(
log2(1 + sollya.x) / sollya.x,
poly_degree, [self.precision] * (poly_degree + 1),
approx_interval,
sollya.absolute,
error_function=lambda p, f, ai, mod, t: sollya.dirtyinfnorm(
p - f, ai))
Log.report(
Log.Info, "poly_degree={}, approx_error={}".format(
poly_degree, approx_error))
poly_object = global_poly_object.sub_poly(start_index=1, offset=1)
#poly_object = global_poly_object.sub_poly(start_index=0,offset=0)
Attributes.set_default_silent(True)
Attributes.set_default_rounding_mode(ML_RoundToNearest)
Log.report(Log.Verbose, "generating polynomial evaluation scheme")
pre_poly = PolynomialSchemeEvaluator.generate_horner_scheme(
poly_object, _red_vx, unified_precision=self.precision)
_poly = FMA(pre_poly, _red_vx,
global_poly_object.get_cst_coeff(0, self.precision))
_poly.set_attributes(tag="poly", debug=debug_lftolx)
Log.report(
Log.Verbose, "sollya global_poly_object: {}".format(
global_poly_object.get_sollya_object()))
Log.report(
Log.Verbose, "sollya poly_object: {}".format(
poly_object.get_sollya_object()))
corr_exp = _vx_exp if exp_corr_factor == None else _vx_exp + exp_corr_factor
Attributes.unset_default_rounding_mode()
Attributes.unset_default_silent()
pre_result = -_log_inv_hi + (_red_vx * _poly + (-_log_inv_lo))
pre_result.set_attributes(tag="pre_result", debug=debug_lftolx)
exact_log2_hi_exp = Conversion(corr_exp, precision=self.precision)
exact_log2_hi_exp.set_attributes(tag="exact_log2_hi_hex",
debug=debug_lftolx)
_result = exact_log2_hi_exp + pre_result
return _result, _poly, _log_inv_lo, _log_inv_hi, _red_vx
result, poly, log_inv_lo, log_inv_hi, red_vx = compute_log(vx)
result.set_attributes(tag="result", debug=debug_lftolx)
# specific input value predicate
neg_input = Comparison(vx,
0,
likely=False,
specifier=Comparison.Less,
debug=debugd,
tag="neg_input")
vx_nan_or_inf = Test(vx,
specifier=Test.IsInfOrNaN,
likely=False,
debug=debugd,
tag="nan_or_inf")
vx_snan = Test(vx,
specifier=Test.IsSignalingNaN,
likely=False,
debug=debugd,
tag="vx_snan")
vx_inf = Test(vx,
specifier=Test.IsInfty,
likely=False,
debug=debugd,
tag="vx_inf")
vx_subnormal = Test(vx,
specifier=Test.IsSubnormal,
likely=False,
debug=debugd,
tag="vx_subnormal")
vx_zero = Test(vx,
specifier=Test.IsZero,
likely=False,
debug=debugd,
tag="vx_zero")
exp_mone = Equal(vx_exp,
-1,
tag="exp_minus_one",
debug=debugd,
likely=False)
vx_one = Equal(vx, 1.0, tag="vx_one", likely=False, debug=debugd)
# Specific specific for the case exp == -1
# log2(x) = log2(m) - 1
#
# as m in [1, 2[, log2(m) in [0, 1[
# if r is close to 2, a catastrophic cancellation can occur
#
# r = seed(m)
# log2(x) = log2(seed(m) * m / seed(m)) - 1
# = log2(seed(m) * m) - log2(seed(m)) - 1
#
# for m really close to 2 => seed(m) = 0.5
# => log2(x) = log2(0.5 * m)
# =
result_exp_m1 = (-log_inv_hi - 1.0) + FMA(poly, red_vx, -log_inv_lo)
result_exp_m1.set_attributes(tag="result_exp_m1", debug=debug_lftolx)
m100 = -100
S2100 = Constant(S2**100, precision=self.precision)
result_subnormal, _, _, _, _ = compute_log(vx * S2100,
exp_corr_factor=m100)
result_subnormal.set_attributes(tag="result_subnormal",
debug=debug_lftolx)
one_err = S2**-7
approx_interval_one = Interval(-one_err, one_err)
red_vx_one = vx - 1.0
poly_degree_one = sup(
guessdegree(
log(1 + x) / x, approx_interval_one, S2**
-(self.precision.get_field_size() + 1))) + 1
poly_object_one = Polynomial.build_from_approximation(
log(1 + sollya.x) / sollya.x, poly_degree_one,
[self.precision] * (poly_degree_one + 1), approx_interval_one,
absolute).sub_poly(start_index=1)
poly_one = PolynomialSchemeEvaluator.generate_horner_scheme(
poly_object_one, red_vx_one, unified_precision=self.precision)
poly_one.set_attributes(tag="poly_one", debug=debug_lftolx)
result_one = red_vx_one + red_vx_one * poly_one
cond_one = (vx < (1 + one_err)) & (vx > (1 - one_err))
cond_one.set_attributes(tag="cond_one", debug=debugd, likely=False)
# main scheme
pre_scheme = ConditionBlock(
neg_input,
Statement(ClearException(), Raise(ML_FPE_Invalid),
Return(FP_QNaN(self.precision))),
ConditionBlock(
vx_nan_or_inf,
ConditionBlock(
vx_inf,
Statement(
ClearException(),
Return(FP_PlusInfty(self.precision)),
),
Statement(ClearException(),
ConditionBlock(vx_snan, Raise(ML_FPE_Invalid)),
Return(FP_QNaN(self.precision)))),
ConditionBlock(
vx_subnormal,
ConditionBlock(
vx_zero,
Statement(
ClearException(),
Raise(ML_FPE_DivideByZero),
Return(FP_MinusInfty(self.precision)),
),
Statement(ClearException(), result_subnormal,
Return(result_subnormal))),
ConditionBlock(
vx_one,
Statement(
ClearException(),
Return(FP_PlusZero(self.precision)),
),
ConditionBlock(exp_mone, Return(result_exp_m1),
Return(result))))))
scheme = Statement(result, pre_scheme)
return scheme
def numeric_emulate(self, input_value):
""" Numeric emulation to generate expected value
corresponding to input_value input """
return log2(input_value)
def compute_log(_vx, exp_corr_factor=None):
_vx_mant = MantissaExtraction(_vx,
tag="_vx_mant",
precision=self.precision,
debug=debug_lftolx)
_vx_exp = ExponentExtraction(_vx, tag="_vx_exp", debug=debugd)
# The main table is indexed by the 7 most significant bits
# of the mantissa
table_index = inv_approx_table.index_function(_vx_mant)
table_index.set_attributes(tag="table_index", debug=debuglld)
# argument reduction
# Using AND -2 to exclude LSB set to 1 for Newton-Raphson convergence
# TODO: detect if single operand inverse seed is supported by the targeted architecture
pre_arg_red_index = TypeCast(BitLogicAnd(
TypeCast(DivisionSeed(_vx_mant,
precision=self.precision,
tag="seed",
debug=debug_lftolx,
silent=True),
precision=ML_UInt64),
Constant(-2, precision=ML_UInt64),
precision=ML_UInt64),
precision=self.precision,
tag="pre_arg_red_index",
debug=debug_lftolx)
arg_red_index = Select(Equal(table_index, 0),
1.0,
pre_arg_red_index,
tag="arg_red_index",
debug=debug_lftolx)
_red_vx = FMA(arg_red_index, _vx_mant, -1.0)
_red_vx.set_attributes(tag="_red_vx", debug=debug_lftolx)
inv_err = S2**-inv_approx_table.index_size
red_interval = Interval(1 - inv_err, 1 + inv_err)
# return in case of standard (non-special) input
_log_inv_lo = TableLoad(log_table,
table_index,
1,
tag="log_inv_lo",
debug=debug_lftolx)
_log_inv_hi = TableLoad(log_table,
table_index,
0,
tag="log_inv_hi",
debug=debug_lftolx)
Log.report(Log.Verbose, "building mathematical polynomial")
approx_interval = Interval(-inv_err, inv_err)
poly_degree = sup(
guessdegree(
log2(1 + sollya.x) / sollya.x, approx_interval, S2**
-(self.precision.get_field_size() * 1.1))) + 1
sollya.settings.display = sollya.hexadecimal
global_poly_object, approx_error = Polynomial.build_from_approximation_with_error(
log2(1 + sollya.x) / sollya.x,
poly_degree, [self.precision] * (poly_degree + 1),
approx_interval,
sollya.absolute,
error_function=lambda p, f, ai, mod, t: sollya.dirtyinfnorm(
p - f, ai))
Log.report(
Log.Info, "poly_degree={}, approx_error={}".format(
poly_degree, approx_error))
poly_object = global_poly_object.sub_poly(start_index=1, offset=1)
#poly_object = global_poly_object.sub_poly(start_index=0,offset=0)
Attributes.set_default_silent(True)
Attributes.set_default_rounding_mode(ML_RoundToNearest)
Log.report(Log.Verbose, "generating polynomial evaluation scheme")
pre_poly = PolynomialSchemeEvaluator.generate_horner_scheme(
poly_object, _red_vx, unified_precision=self.precision)
_poly = FMA(pre_poly, _red_vx,
global_poly_object.get_cst_coeff(0, self.precision))
_poly.set_attributes(tag="poly", debug=debug_lftolx)
Log.report(
Log.Verbose, "sollya global_poly_object: {}".format(
global_poly_object.get_sollya_object()))
Log.report(
Log.Verbose, "sollya poly_object: {}".format(
poly_object.get_sollya_object()))
corr_exp = _vx_exp if exp_corr_factor == None else _vx_exp + exp_corr_factor
Attributes.unset_default_rounding_mode()
Attributes.unset_default_silent()
pre_result = -_log_inv_hi + (_red_vx * _poly + (-_log_inv_lo))
pre_result.set_attributes(tag="pre_result", debug=debug_lftolx)
exact_log2_hi_exp = Conversion(corr_exp, precision=self.precision)
exact_log2_hi_exp.set_attributes(tag="exact_log2_hi_hex",
debug=debug_lftolx)
_result = exact_log2_hi_exp + pre_result
return _result, _poly, _log_inv_lo, _log_inv_hi, _red_vx
def generate_scheme(self):
# declaring target and instantiating optimization engine
vx = self.implementation.add_input_variable("x", self.precision)
Log.set_dump_stdout(True)
Log.report(Log.Info, "\033[33;1m generating implementation scheme \033[0m")
if self.debug_flag:
Log.report(Log.Info, "\033[31;1m debug has been enabled \033[0;m")
# local overloading of RaiseReturn operation
def ExpRaiseReturn(*args, **kwords):
kwords["arg_value"] = vx
kwords["function_name"] = self.function_name
return RaiseReturn(*args, **kwords)
C_m1 = Constant(-1, precision = self.precision)
test_NaN_or_inf = Test(vx, specifier = Test.IsInfOrNaN, likely = False, debug = debug_multi, tag = "NaN_or_inf", precision = ML_Bool)
test_NaN = Test(vx, specifier = Test.IsNaN, likely = False, debug = debug_multi, tag = "is_NaN", precision = ML_Bool)
test_inf = Comparison(vx, 0, specifier = Comparison.Greater, debug = debug_multi, tag = "sign", precision = ML_Bool, likely = False);
# Infnty input
infty_return = Statement(ConditionBlock(test_inf, Return(FP_PlusInfty(self.precision)), Return(C_m1)))
# non-std input (inf/nan)
specific_return = ConditionBlock(test_NaN, Return(FP_QNaN(self.precision)), infty_return)
# Over/Underflow Tests
precision_emax = self.precision.get_emax()
precision_max_value = S2**(precision_emax + 1)
expm1_overflow_bound = ceil(log(precision_max_value + 1))
overflow_test = Comparison(vx, expm1_overflow_bound, likely = False, specifier = Comparison.Greater, precision = ML_Bool)
overflow_return = Statement(Return(FP_PlusInfty(self.precision)))
precision_emin = self.precision.get_emin_subnormal()
precision_min_value = S2** precision_emin
expm1_underflow_bound = floor(log(precision_min_value) + 1)
underflow_test = Comparison(vx, expm1_underflow_bound, likely = False, specifier = Comparison.Less, precision = ML_Bool)
underflow_return = Statement(Return(C_m1))
sollya_precision = {ML_Binary32: sollya.binary32, ML_Binary64: sollya.binary64}[self.precision]
int_precision = {ML_Binary32: ML_Int32, ML_Binary64: ML_Int64}[self.precision]
# Constants
log_2 = round(log(2), sollya_precision, sollya.RN)
invlog2 = round(1/log(2), sollya_precision, sollya.RN)
log_2_cst = Constant(log_2, precision = self.precision)
interval_vx = Interval(expm1_underflow_bound, expm1_overflow_bound)
interval_fk = interval_vx * invlog2
interval_k = Interval(floor(inf(interval_fk)), ceil(sup(interval_fk)))
log2_hi_precision = self.precision.get_field_size() - 6
log2_hi = round(log(2), log2_hi_precision, sollya.RN)
log2_lo = round(log(2) - log2_hi, sollya_precision, sollya.RN)
# Reduction
unround_k = vx * invlog2
ik = NearestInteger(unround_k, precision = int_precision, debug = debug_multi, tag = "ik")
k = Conversion(ik, precision = self.precision, tag = "k")
red_coeff1 = Multiplication(k, log2_hi, precision = self.precision)
red_coeff2 = Multiplication(Negation(k, precision = self.precision), log2_lo, precision = self.precision)
pre_sub_mul = Subtraction(vx, red_coeff1, precision = self.precision)
s = Addition(pre_sub_mul, red_coeff2, precision = self.precision)
z = Subtraction(s, pre_sub_mul, precision = self.precision)
t = Subtraction(red_coeff2, z, precision = self.precision)
r = Addition(s, t, precision = self.precision)
r.set_attributes(tag = "r", debug = debug_multi)
r_interval = Interval(-log_2/S2, log_2/S2)
local_ulp = sup(ulp(exp(r_interval), self.precision))
print("ulp: ", local_ulp)
error_goal = S2**-1*local_ulp
print("error goal: ", error_goal)
# Polynomial Approx
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
Log.report(Log.Info, "\033[33;1m Building polynomial \033[0m\n")
poly_degree = sup(guessdegree(expm1(sollya.x), r_interval, error_goal) + 1)
polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_horner_scheme
poly_degree_list = range(0, poly_degree)
precision_list = [self.precision] *(len(poly_degree_list) + 1)
poly_object, poly_error = Polynomial.build_from_approximation_with_error(expm1(sollya.x), poly_degree, precision_list, r_interval, sollya.absolute, error_function = error_function)
sub_poly = poly_object.sub_poly(start_index = 2)
Log.report(Log.Info, "Poly : %s" % sub_poly)
Log.report(Log.Info, "poly error : {} / {:d}".format(poly_error, int(sollya.log2(poly_error))))
pre_sub_poly = polynomial_scheme_builder(sub_poly, r, unified_precision = self.precision)
poly = r + pre_sub_poly
poly.set_attributes(tag = "poly", debug = debug_multi)
exp_k = ExponentInsertion(ik, tag = "exp_k", debug = debug_multi, precision = self.precision)
exp_mk = ExponentInsertion(-ik, tag = "exp_mk", debug = debug_multi, precision = self.precision)
diff = 1 - exp_mk
diff.set_attributes(tag = "diff", debug = debug_multi)
# Late Tests
late_overflow_test = Comparison(ik, self.precision.get_emax(), specifier = Comparison.Greater, likely = False, debug = debug_multi, tag = "late_overflow_test")
overflow_exp_offset = (self.precision.get_emax() - self.precision.get_field_size() / 2)
diff_k = ik - overflow_exp_offset
exp_diff_k = ExponentInsertion(diff_k, precision = self.precision, tag = "exp_diff_k", debug = debug_multi)
exp_oflow_offset = ExponentInsertion(overflow_exp_offset, precision = self.precision, tag = "exp_offset", debug = debug_multi)
late_overflow_result = (exp_diff_k * (1 + poly)) * exp_oflow_offset - 1.0
late_overflow_return = ConditionBlock(
Test(late_overflow_result, specifier = Test.IsInfty, likely = False),
ExpRaiseReturn(ML_FPE_Overflow, return_value = FP_PlusInfty(self.precision)),
Return(late_overflow_result)
)
late_underflow_test = Comparison(k, self.precision.get_emin_normal(), specifier = Comparison.LessOrEqual, likely = False)
underflow_exp_offset = 2 * self.precision.get_field_size()
corrected_coeff = ik + underflow_exp_offset
exp_corrected = ExponentInsertion(corrected_coeff, precision = self.precision)
exp_uflow_offset = ExponentInsertion(-underflow_exp_offset, precision = self.precision)
late_underflow_result = ( exp_corrected * (1 + poly)) * exp_uflow_offset - 1.0
test_subnormal = Test(late_underflow_result, specifier = Test.IsSubnormal, likely = False)
late_underflow_return = Statement(
ConditionBlock(
test_subnormal,
ExpRaiseReturn(ML_FPE_Underflow, return_value = late_underflow_result)),
Return(late_underflow_result)
)
# Reconstruction
std_result = exp_k * ( poly + diff )
std_result.set_attributes(tag = "result", debug = debug_multi)
result_scheme = ConditionBlock(
late_overflow_test,
late_overflow_return,
ConditionBlock(
late_underflow_test,
late_underflow_return,
Return(std_result)
)
)
std_return = ConditionBlock(
overflow_test,
overflow_return,
ConditionBlock(
underflow_test,
underflow_return,
result_scheme)
)
scheme = ConditionBlock(
test_NaN_or_inf,
Statement(specific_return),
std_return
)
return scheme
def generate_scalar_scheme(self, vx):
# approximation the gamma function
abs_vx = Abs(vx, precision=self.precision)
FCT_LIMIT = 1.0
omega_value = self.precision.get_omega()
def sollya_wrap_bigfloat_fct(bfct):
""" wrap bigfloat's function <bfct> such that is can be used
on SollyaObject inputs and returns SollyaObject results """
def fct(x):
return sollya.SollyaObject(bfct(SollyaObject(x).bigfloat()))
return fct
sollya_gamma = sollya_wrap_bigfloat_fct(bigfloat.gamma)
sollya_digamma = sollya_wrap_bigfloat_fct(bigfloat.digamma)
# first derivative of gamma is digamma * gamma
bigfloat_gamma_d0 = lambda x: bigfloat.gamma(x) * bigfloat.digamma(x)
sollya_gamma_d0 = sollya_wrap_bigfloat_fct(bigfloat_gamma_d0)
# approximating trigamma with straightforward derivatives formulae of digamma
U = 2**-64
bigfloat_trigamma = lambda x: (
(bigfloat.digamma(x * (1 + U)) - bigfloat.digamma(x)) / (x * U))
sollya_trigamma = sollya_wrap_bigfloat_fct(bigfloat_trigamma)
bigfloat_gamma_d1 = lambda x: (bigfloat_trigamma(x) * bigfloat.gamma(
x) + bigfloat_gamma_d0(x) * bigfloat.digamma(x))
sollya_gamma_d1 = sollya_wrap_bigfloat_fct(bigfloat_gamma_d1)
def sollya_gamma_fct(x, diff_order, prec):
""" wrapper to use bigfloat implementation of exponential
rather than sollya's implementation directly.
This wrapper implements sollya's function API.
:param x: numerical input value (may be an Interval)
:param diff_order: differential order
:param prec: numerical precision expected (min)
"""
fct = None
if diff_order == 0:
fct = sollya_gamma
elif diff_order == 1:
fct = sollya_gamma_d0
elif diff_order == 2:
fct = sollya_gamma_d1
else:
raise NotImplementedError
with bigfloat.precision(prec):
if x.is_range():
lo = sollya.inf(x)
hi = sollya.sup(x)
return sollya.Interval(fct(lo), fct(hi))
else:
return fct(x)
# search the lower x such that gamma(x) >= omega
omega_upper_limit = search_bound_threshold(sollya_gamma, omega_value,
2, 1000.0, self.precision)
Log.report(Log.Debug, "gamma(x) = {} limit is {}", omega_value,
omega_upper_limit)
# evaluate gamma(<min-normal-value>)
lower_x_bound = self.precision.get_min_normal_value()
value_min = sollya_gamma(lower_x_bound)
Log.report(Log.Debug, "gamma({}) = {}(log2={})", lower_x_bound,
value_min, int(sollya.log2(value_min)))
# evaluate gamma(<min-subnormal-value>)
lower_x_bound = self.precision.get_min_subnormal_value()
value_min = sollya_gamma(lower_x_bound)
Log.report(Log.Debug, "gamma({}) = {}(log2={})", lower_x_bound,
value_min, int(sollya.log2(value_min)))
# Gamma is defined such that gamma(x+1) = x * gamma(x)
#
# we approximate gamma over [1, 2]
# y in [1, 2]
# gamma(y) = (y-1) * gamma(y-1)
# gamma(y-1) = gamma(y) / (y-1)
Log.report(Log.Info, "building mathematical polynomial")
approx_interval = Interval(1, 2)
approx_fct = sollya.function(sollya_gamma_fct)
poly_degree = int(
sup(
guessdegree(approx_fct, approx_interval, S2**
-(self.precision.get_field_size() + 5)))) + 1
Log.report(Log.Debug, "approximation's poly degree over [1, 2] is {}",
poly_degree)
sys.exit(1)
poly_degree_list = list(range(1, poly_degree, 2))
Log.report(Log.Debug, "poly_degree is {} and list {}", poly_degree,
poly_degree_list)
global_poly_object = Polynomial.build_from_approximation(
approx_fct, poly_degree_list,
[self.precision] * len(poly_degree_list), approx_interval,
sollya.relative)
Log.report(
Log.Debug, "inform is {}",
dirtyinfnorm(approx_fct - global_poly_object.get_sollya_object(),
approx_interval))
poly_object = global_poly_object.sub_poly(start_index=1, offset=1)
ext_precision = {
ML_Binary32: ML_SingleSingle,
ML_Binary64: ML_DoubleDouble,
}[self.precision]
pre_poly = PolynomialSchemeEvaluator.generate_horner_scheme(
poly_object, abs_vx, unified_precision=self.precision)
result = FMA(pre_poly, abs_vx, abs_vx)
result.set_attributes(tag="result", debug=debug_multi)
eps_target = S2**-(self.precision.get_field_size() + 5)
def offset_div_function(fct):
return lambda offset: fct(sollya.x + offset)
# empiral numbers
field_size = {ML_Binary32: 6, ML_Binary64: 8}[self.precision]
near_indexing = SubFPIndexing(eps_exp, 0, 6, self.precision)
near_approx = generic_poly_split(offset_div_function(sollya.erf),
near_indexing, eps_target,
self.precision, abs_vx)
near_approx.set_attributes(tag="near_approx", debug=debug_multi)
def offset_function(fct):
return lambda offset: fct(sollya.x + offset)
medium_indexing = SubFPIndexing(1, one_limit_exp, 7, self.precision)
medium_approx = generic_poly_split(offset_function(sollya.erf),
medium_indexing, eps_target,
self.precision, abs_vx)
medium_approx.set_attributes(tag="medium_approx", debug=debug_multi)
# approximation for positive values
scheme = ConditionBlock(
abs_vx < eps, Return(result),
ConditionBlock(
abs_vx < near_indexing.get_max_bound(), Return(near_approx),
ConditionBlock(abs_vx < medium_indexing.get_max_bound(),
Return(medium_approx),
Return(Constant(1.0,
precision=self.precision)))))
return scheme
def generate_payne_hanek(vx,
frac_pi,
precision,
n=100,
k=4,
chunk_num=None,
debug=False):
""" generate payne and hanek argument reduction for frac_pi * variable """
sollya.roundingwarnings = sollya.off
debug_precision = debug_multi
int_precision = {ML_Binary32: ML_Int32, ML_Binary64: ML_Int64}[precision]
p = precision.get_field_size()
# weight of the most significant digit of the constant
cst_msb = floor(log2(abs(frac_pi)))
# length of exponent range which must be covered by the approximation
# of the constant
cst_exp_range = cst_msb - precision.get_emin_subnormal() + 1
# chunk size has to be so than multiplication by a splitted <v>
# (vx_hi or vx_lo) is exact
chunk_size = precision.get_field_size() / 2 - 2
chunk_number = int(ceil((cst_exp_range + chunk_size - 1) / chunk_size))
scaling_factor = S2**-(chunk_size / 2)
chunk_size_cst = Constant(chunk_size, precision=ML_Int32)
cst_msb_node = Constant(cst_msb, precision=ML_Int32)
# Saving sollya's global precision
old_global_prec = sollya.settings.prec
sollya.settings.prec(cst_exp_range + n)
# table to store chunk of constant multiplicand
cst_table = ML_NewTable(dimensions=[chunk_number, 1],
storage_precision=precision,
tag="PH_cst_table")
# table to store sqrt(scaling_factor) corresponding to the
# cst multiplicand chunks
scale_table = ML_NewTable(dimensions=[chunk_number, 1],
storage_precision=precision,
tag="PH_scale_table")
tmp_cst = frac_pi
# cst_table stores normalized constant chunks (they have been
# scale back to close to 1.0 interval)
#
# scale_table stores the scaling factors corresponding to the
# denormalization of cst_table coefficients
# this loop divide the digits of frac_pi into chunks
# the chunk lsb weight is given by a shift from
# cst_msb, multiple of the chunk index
for i in range(chunk_number):
value_div_factor = S2**(chunk_size * (i + 1) - cst_msb)
local_cst = int(tmp_cst * value_div_factor) / value_div_factor
local_scale = (scaling_factor**i)
# storing scaled constant chunks
cst_table[i][0] = local_cst / (local_scale**2)
scale_table[i][0] = local_scale
# Updating constant value
tmp_cst = tmp_cst - local_cst
# Computing which part of the constant we do not need to multiply
# In the following comments, vi represents the bit of frac_pi of weight 2**-i
# Bits vi so that i <= (vx_exp - p + 1 -k) are not needed, because they result
# in a multiple of 2pi and do not contribute to trig functions.
vx_exp = ExponentExtraction(
vx, precision=vx.get_precision().get_integer_format())
vx_exp = Conversion(vx_exp, precision=ML_Int32)
msb_exp = -(vx_exp - p + 1 - k)
msb_exp.set_attributes(tag="msb_exp", debug=debug_multi)
msb_exp = Conversion(msb_exp, precision=ML_Int32)
# Select the highest index where the reduction should start
msb_index = Select(cst_msb_node < msb_exp, 0,
(cst_msb_node - msb_exp) / chunk_size_cst)
msb_index.set_attributes(tag="msb_index", debug=debug_multi)
# For a desired accuracy of 2**-n, bits vi so that i >= (vx_exp + n + 4) are not needed, because they contribute less than
# 2**-n to the result
lsb_exp = -(vx_exp + n + 4)
lsb_exp.set_attributes(tag="lsb_exp", debug=debug_multi)
lsb_exp = Conversion(lsb_exp, precision=ML_Int32)
# Index of the corresponding chunk
lsb_index = (cst_msb_node - lsb_exp) / chunk_size_cst
lsb_index.set_attributes(tag="lsb_index", debug=debug_multi)
# Splitting vx
half_size = precision.get_field_size() / 2 + 1
# hi part (most significant digit) of vx input
vx_hi = TypeCast(BitLogicAnd(
TypeCast(vx, precision=int_precision),
Constant(~int(2**half_size - 1), precision=int_precision)),
precision=precision)
vx_hi.set_attributes(tag="vx_hi_ph") #, debug = debug_multi)
vx_lo = vx - vx_hi
vx_lo.set_attributes(tag="vx_lo_ph") #, debug = debug_multi)
# loop iterator variable
vi = Variable("i", precision=ML_Int32, var_type=Variable.Local)
# step scaling factor
half_scaling = Constant(S2**(-chunk_size / 2), precision=precision)
i1 = Constant(1, precision=ML_Int32)
# accumulator to the output precision
acc = Variable("acc", precision=precision, var_type=Variable.Local)
# integer accumulator
acc_int = Variable("acc_int",
precision=int_precision,
var_type=Variable.Local)
init_loop = Statement(
vx_hi,
vx_lo,
ReferenceAssign(vi, msb_index),
ReferenceAssign(acc, Constant(0, precision=precision)),
ReferenceAssign(acc_int, Constant(0, precision=int_precision)),
)
cst_load = TableLoad(cst_table,
vi,
0,
tag="cst_load",
debug=debug_precision)
sca_load = TableLoad(scale_table,
vi,
0,
tag="sca_load",
debug=debug_precision)
# loop body
# hi_mult = vx_hi * <scale_factor> * <cst>
hi_mult = (vx_hi * sca_load) * (cst_load * sca_load)
hi_mult.set_attributes(tag="hi_mult", debug=debug_precision)
pre_hi_mult_int = NearestInteger(hi_mult,
precision=int_precision,
tag="hi_mult_int",
debug=(debuglld if debug else None))
hi_mult_int_f = Conversion(pre_hi_mult_int,
precision=precision,
tag="hi_mult_int_f",
debug=debug_precision)
pre_hi_mult_red = (hi_mult - hi_mult_int_f).modify_attributes(
tag="hi_mult_red", debug=debug_precision)
# for the first chunks (vx_hi * <constant chunk>) exceeds 2**k+1 and may be
# discard (whereas it may lead to overflow during integer conversion
pre_exclude_hi = ((cst_msb_node - (vi + i1) * chunk_size + i1) +
(vx_exp + Constant(-half_size + 1, precision=ML_Int32))
).modify_attributes(tag="pre_exclude_hi",
debug=(debugd if debug else None))
pre_exclude_hi.propagate_precision(ML_Int32,
[cst_msb_node, vi, vx_exp, i1])
Ck = Constant(k, precision=ML_Int32)
exclude_hi = pre_exclude_hi <= Ck
exclude_hi.set_attributes(tag="exclude_hi", debug=debug_multi)
hi_mult_red = Select(exclude_hi, pre_hi_mult_red,
Constant(0, precision=precision))
hi_mult_int = Select(exclude_hi, pre_hi_mult_int,
Constant(0, precision=int_precision))
# lo part of the chunk reduction
lo_mult = (vx_lo * sca_load) * (cst_load * sca_load)
lo_mult.set_attributes(tag="lo_mult") #, debug = debug_multi)
lo_mult_int = NearestInteger(lo_mult,
precision=int_precision,
tag="lo_mult_int") #, debug = debug_multi
lo_mult_int_f = Conversion(lo_mult_int,
precision=precision,
tag="lo_mult_int_f") #, debug = debug_multi)
lo_mult_red = (lo_mult - lo_mult_int_f).modify_attributes(
tag="lo_mult_red") #, debug = debug_multi)
# accumulating fractional part
acc_expr = (acc + hi_mult_red) + lo_mult_red
# accumulating integer part
int_expr = ((acc_int + hi_mult_int) + lo_mult_int) % 2**(k + 1)
CF1 = Constant(1, precision=precision)
CI1 = Constant(1, precision=int_precision)
# extracting exceeding integer part in fractionnal accumulator
acc_expr_int = NearestInteger(acc_expr, precision=int_precision)
# normalizing integer and fractionnal accumulator by subtracting then
# adding exceeding integer part
normalization = Statement(
ReferenceAssign(
acc, acc_expr - Conversion(acc_expr_int, precision=precision)),
ReferenceAssign(acc_int, int_expr + acc_expr_int),
)
acc_expr.set_attributes(tag="acc_expr") #, debug = debug_multi)
int_expr.set_attributes(tag="int_expr") #, debug = debug_multi)
red_loop = Loop(
init_loop, vi <= lsb_index,
Statement(acc_expr, int_expr, normalization,
ReferenceAssign(vi, vi + 1)))
result = Statement(lsb_index, msb_index, red_loop)
# restoring sollya's global precision
sollya.settings.prec = old_global_prec
return result, acc, acc_int
def generate_scalar_scheme(self, vx):
Log.set_dump_stdout(True)
Log.report(Log.Info, "\033[33;1m generating implementation scheme \033[0m")
if self.debug_flag:
Log.report(Log.Info, "\033[31;1m debug has been enabled \033[0;m")
index_size = 5
comp_lo = (vx < 0)
comp_lo.set_attributes(tag = "comp_lo", precision = ML_Bool)
sign = Select(comp_lo, -1, 1, precision = self.precision)
# as sinh is an odd function, we can simplify the input to its absolute
# value once the sign has been extracted
vx = Abs(vx)
int_precision = self.precision.get_integer_format()
# argument reduction
arg_reg_value = log(2)/2**index_size
inv_log2_value = round(1/arg_reg_value, self.precision.get_sollya_object(), sollya.RN)
inv_log2_cst = Constant(inv_log2_value, precision = self.precision, tag = "inv_log2")
# for r_hi to be accurate we ensure k * log2_hi_value_cst is exact
# by limiting the number of non-zero bits in log2_hi_value_cst
# cosh(x) ~ exp(abs(x))/2 for a big enough x
# cosh(x) > 2^1023 <=> exp(x) > 2^1024 <=> x > log(2^1024)
# k = inv_log2_value * x
# -1 for guard
max_k_approx = inv_log2_value * log(sollya.SollyaObject(2)**1024)
max_k_bitsize = int(ceil(log2(max_k_approx)))
Log.report(Log.Info, "max_k_bitsize: %d" % max_k_bitsize)
log2_hi_value_precision = self.precision.get_precision() - max_k_bitsize - 1
log2_hi_value = round(arg_reg_value, log2_hi_value_precision, sollya.RN)
log2_lo_value = round(arg_reg_value - log2_hi_value, self.precision.get_sollya_object(), sollya.RN)
log2_hi_value_cst = Constant(log2_hi_value, tag = "log2_hi_value", precision = self.precision)
log2_lo_value_cst = Constant(log2_lo_value, tag = "log2_lo_value", precision = self.precision)
k = Trunc(Multiplication(inv_log2_cst, vx), precision = self.precision)
k_log2 = Multiplication(k, log2_hi_value_cst, precision = self.precision, exact = True, tag = "k_log2", unbreakable = True)
r_hi = vx - k_log2
r_hi.set_attributes(tag = "r_hi", debug = debug_multi, unbreakable = True)
r_lo = -k * log2_lo_value_cst
# reduced argument
r = r_hi + r_lo
r.set_attributes(tag = "r", debug = debug_multi)
if is_gappa_installed():
r_eval_error = self.get_eval_error(r_hi, variable_copy_map =
{
vx: Variable("vx", interval = Interval(0, 715), precision = self.precision),
k: Variable("k", interval = Interval(0, 1024), precision = self.precision)
})
Log.report(Log.Verbose, "r_eval_error: ", r_eval_error)
approx_interval = Interval(-arg_reg_value, arg_reg_value)
error_goal_approx = 2**-(self.precision.get_precision())
poly_degree = sup(guessdegree(exp(sollya.x), approx_interval, error_goal_approx)) + 3
precision_list = [1] + [self.precision] * (poly_degree)
k_integer = Conversion(k, precision = int_precision, tag = "k_integer", debug = debug_multi)
k_hi = BitLogicRightShift(k_integer, Constant(index_size, precision=int_precision), tag = "k_int_hi", precision = int_precision, debug = debug_multi)
k_lo = Modulo(k_integer, 2**index_size, tag = "k_int_lo", precision = int_precision, debug = debug_multi)
pow_exp = ExponentInsertion(Conversion(k_hi, precision = int_precision), precision = self.precision, tag = "pow_exp", debug = debug_multi)
exp_table = ML_NewTable(dimensions = [2 * 2**index_size, 4], storage_precision = self.precision, tag = self.uniquify_name("exp2_table"))
for i in range(2 * 2**index_size):
input_value = i - 2**index_size if i >= 2**index_size else i
reduced_hi_prec = int(self.precision.get_mantissa_size() - 8)
# using SollyaObject wrapper to force evaluation by sollya
# with higher precision
exp_value = sollya.SollyaObject(2)**((input_value)* 2**-index_size)
mexp_value = sollya.SollyaObject(2)**((-input_value)* 2**-index_size)
pos_value_hi = round(exp_value, reduced_hi_prec, sollya.RN)
pos_value_lo = round(exp_value - pos_value_hi, self.precision.get_sollya_object(), sollya.RN)
neg_value_hi = round(mexp_value, reduced_hi_prec, sollya.RN)
neg_value_lo = round(mexp_value - neg_value_hi, self.precision.get_sollya_object(), sollya.RN)
exp_table[i][0] = neg_value_hi
exp_table[i][1] = neg_value_lo
exp_table[i][2] = pos_value_hi
exp_table[i][3] = pos_value_lo
# log2_value = log(2) / 2^index_size
# sinh(x) = 1/2 * (exp(x) - exp(-x))
# exp(x) = exp(x - k * log2_value + k * log2_value)
#
# r = x - k * log2_value
# exp(x) = exp(r) * 2 ^ (k / 2^index_size)
#
# k / 2^index_size = h + l * 2^-index_size, with k, h, l integers
# exp(x) = exp(r) * 2^h * 2^(l *2^-index_size)
#
# sinh(x) = exp(r) * 2^(h-1) * 2^(l *2^-index_size) - exp(-r) * 2^(-h-1) * 2^(-l *2^-index_size)
# S=2^(h-1), T = 2^(-h-1)
# exp(r) = 1 + poly_pos(r)
# exp(-r) = 1 + poly_neg(r)
# 2^(l / 2^index_size) = pos_value_hi + pos_value_lo
# 2^(-l / 2^index_size) = neg_value_hi + neg_value_lo
#
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
poly_object, poly_approx_error = Polynomial.build_from_approximation_with_error(exp(sollya.x), poly_degree, precision_list, approx_interval, sollya.absolute, error_function = error_function)
Log.report(Log.Verbose, "poly_approx_error: {}, {}".format(poly_approx_error, float(log2(poly_approx_error))))
polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_horner_scheme
poly_pos = polynomial_scheme_builder(poly_object.sub_poly(start_index = 1), r, unified_precision = self.precision)
poly_pos.set_attributes(tag = "poly_pos", debug = debug_multi)
poly_neg = polynomial_scheme_builder(poly_object.sub_poly(start_index = 1), -r, unified_precision = self.precision)
poly_neg.set_attributes(tag = "poly_neg", debug = debug_multi)
table_index = Addition(k_lo, Constant(2**index_size, precision = int_precision), precision = int_precision, tag = "table_index", debug = debug_multi)
neg_value_load_hi = TableLoad(exp_table, table_index, 0, tag = "neg_value_load_hi", debug = debug_multi)
neg_value_load_lo = TableLoad(exp_table, table_index, 1, tag = "neg_value_load_lo", debug = debug_multi)
pos_value_load_hi = TableLoad(exp_table, table_index, 2, tag = "pos_value_load_hi", debug = debug_multi)
pos_value_load_lo = TableLoad(exp_table, table_index, 3, tag = "pos_value_load_lo", debug = debug_multi)
k_plus = Max(
Subtraction(k_hi, Constant(1, precision = int_precision), precision=int_precision, tag="k_plus", debug=debug_multi),
Constant(self.precision.get_emin_normal(), precision = int_precision))
k_neg = Max(
Subtraction(-k_hi, Constant(1, precision=int_precision), precision=int_precision, tag="k_neg", debug=debug_multi),
Constant(self.precision.get_emin_normal(), precision = int_precision))
# 2^(h-1)
pow_exp_pos = ExponentInsertion(k_plus, precision = self.precision, tag="pow_exp_pos", debug=debug_multi)
# 2^(-h-1)
pow_exp_neg = ExponentInsertion(k_neg, precision = self.precision, tag="pow_exp_neg", debug=debug_multi)
hi_terms = (pos_value_load_hi * pow_exp_pos - neg_value_load_hi * pow_exp_neg)
hi_terms.set_attributes(tag = "hi_terms", debug=debug_multi)
pos_exp = (pos_value_load_hi * poly_pos + (pos_value_load_lo + pos_value_load_lo * poly_pos)) * pow_exp_pos
pos_exp.set_attributes(tag = "pos_exp", debug = debug_multi)
neg_exp = (neg_value_load_hi * poly_neg + (neg_value_load_lo + neg_value_load_lo * poly_neg)) * pow_exp_neg
neg_exp.set_attributes(tag = "neg_exp", debug = debug_multi)
result = Addition(
Subtraction(
pos_exp,
neg_exp,
precision=self.precision,
),
hi_terms,
precision=self.precision,
tag="result",
debug=debug_multi
)
# ov_value
ov_value = round(asinh(self.precision.get_max_value()), self.precision.get_sollya_object(), sollya.RD)
ov_flag = Comparison(Abs(vx), Constant(ov_value, precision = self.precision), specifier = Comparison.Greater)
# main scheme
scheme = Statement(
Return(
Select(
ov_flag,
sign*FP_PlusInfty(self.precision),
sign*result
)))
return scheme
def generate_scheme(self):
#func_implementation = CodeFunction(self.function_name, output_format = self.precision)
vx = self.implementation.add_input_variable("x", self.get_input_precision())
sollya_precision = self.get_sollya_precision()
# retrieving processor inverse approximation table
#dummy_var = Variable("dummy", precision = self.precision)
#dummy_div_seed = DivisionSeed(dummy_var, precision = self.precision)
#inv_approx_table = self.processor.get_recursive_implementation(dummy_div_seed, language = None, table_getter = lambda self: self.approx_table_map)
lo_bound_global = SollyaObject(0.0)
hi_bound_global = SollyaObject(0.75)
approx_interval = Interval(lo_bound_global, hi_bound_global)
approx_interval_size = hi_bound_global - lo_bound_global
# table creation
table_index_size = 7
field_index_size = 2
exp_index_size = table_index_size - field_index_size
table_size = 2**table_index_size
table_index_range = range(table_size)
local_degree = 9
coeff_table = ML_Table(dimensions = [table_size, local_degree], storage_precision = self.precision)
#local_interval_size = approx_interval_size / SollyaObject(table_size)
#for i in table_index_range:
# degree = 6
# lo_bound = lo_bound_global + i * local_interval_size
# hi_bound = lo_bound_global + (i+1) * local_interval_size
# approx_interval = Interval(lo_bound, hi_bound)
# local_poly_object, local_error = Polynomial.build_from_approximation_with_error(acos(x), degree, [self.precision] * (degree+1), approx_interval, absolute)
# local_error = int(log2(sup(abs(local_error / acos(approx_interval)))))
# print approx_interval, local_error
exp_lo = 2**exp_index_size
for i in table_index_range:
lo_bound = (1.0 + (i % 2**field_index_size) * S2**-field_index_size) * S2**(i / 2**field_index_size - exp_lo)
hi_bound = (1.0 + ((i % 2**field_index_size) + 1) * S2**-field_index_size) * S2**(i / 2**field_index_size - exp_lo)
local_approx_interval = Interval(lo_bound, hi_bound)
local_poly_object, local_error = Polynomial.build_from_approximation_with_error(acos(1 - x), local_degree, [self.precision] * (local_degree+1), local_approx_interval, sollya.absolute)
local_error = int(log2(sup(abs(local_error / acos(1 - local_approx_interval)))))
coeff_table
print local_approx_interval, local_error
for d in xrange(local_degree):
coeff_table[i][d] = sollya.coeff(local_poly_object.get_sollya_object(), d)
table_index = BitLogicRightShift(vx, vx.get_precision().get_field_size() - field_index_size) - (exp_lo << field_index_size)
print "building mathematical polynomial"
poly_degree = sup(sollya.guessdegree(acos(x), approx_interval, S2**-(self.precision.get_field_size())))
print "guessed polynomial degree: ", int(poly_degree)
#global_poly_object = Polynomial.build_from_approximation(log10(1+x)/x, poly_degree, [self.precision]*(poly_degree+1), approx_interval, absolute)
print "generating polynomial evaluation scheme"
#_poly = PolynomialSchemeEvaluator.generate_horner_scheme(poly_object, _red_vx, unified_precision = self.precision)
# building eval error map
#eval_error_map = {
# red_vx: Variable("red_vx", precision = self.precision, interval = red_vx.get_interval()),
# log_inv_hi: Variable("log_inv_hi", precision = self.precision, interval = table_high_interval),
# log_inv_lo: Variable("log_inv_lo", precision = self.precision, interval = table_low_interval),
#}
# computing gappa error
#poly_eval_error = self.get_eval_error(result, eval_error_map)
# main scheme
print "MDL scheme"
scheme = Statement(Return(vx))
return scheme
def ulp(v, format_):
""" return a 'unit in last place' value for <v> assuming precision is defined by format _ """
return sollya.S2**(sollya.ceil(sollya.log2(sollya.abs(v))) -
(format_.get_precision() + 1))
def generate_scheme(self):
# declaring CodeFunction and retrieving input variable
vx = self.implementation.add_input_variable("x", self.precision)
table_size_log = self.table_size_log
integer_size = 31
integer_precision = ML_Int32
max_bound = sup(abs(self.input_intervals[0]))
max_bound_log = int(ceil(log2(max_bound)))
Log.report(Log.Info, "max_bound_log=%s " % max_bound_log)
scaling_power = integer_size - max_bound_log
Log.report(Log.Info, "scaling power: %s " % scaling_power)
storage_precision = ML_Custom_FixedPoint_Format(1, 30, signed=True)
Log.report(Log.Info, "tabulating cosine and sine")
# cosine and sine fused table
fused_table = ML_NewTable(
dimensions=[2**table_size_log, 2],
storage_precision=storage_precision,
tag="fast_lib_shared_table") # self.uniquify_name("cossin_table"))
# filling table
for i in range(2**table_size_log):
local_x = i / S2**table_size_log * S2**max_bound_log
cos_local = cos(
local_x
) # nearestint(cos(local_x) * S2**storage_precision.get_frac_size())
sin_local = sin(
local_x
) # nearestint(sin(local_x) * S2**storage_precision.get_frac_size())
fused_table[i][0] = cos_local
fused_table[i][1] = sin_local
# argument reduction evaluation scheme
# scaling_factor = Constant(S2**scaling_power, precision = self.precision)
red_vx_precision = ML_Custom_FixedPoint_Format(31 - scaling_power,
scaling_power,
signed=True)
Log.report(
Log.Verbose, "red_vx_precision.get_c_bit_size()=%d" %
red_vx_precision.get_c_bit_size())
# red_vx = NearestInteger(vx * scaling_factor, precision = integer_precision)
red_vx = Conversion(vx,
precision=red_vx_precision,
tag="red_vx",
debug=debug_fixed32)
computation_precision = red_vx_precision # self.precision
output_precision = self.get_output_precision()
Log.report(Log.Info,
"computation_precision is %s" % computation_precision)
Log.report(Log.Info, "storage_precision is %s" % storage_precision)
Log.report(Log.Info, "output_precision is %s" % output_precision)
hi_mask_value = 2**32 - 2**(32 - table_size_log - 1)
hi_mask = Constant(hi_mask_value, precision=ML_Int32)
Log.report(Log.Info, "hi_mask=0x%x" % hi_mask_value)
red_vx_hi_int = BitLogicAnd(TypeCast(red_vx, precision=ML_Int32),
hi_mask,
precision=ML_Int32,
tag="red_vx_hi_int",
debug=debugd)
red_vx_hi = TypeCast(red_vx_hi_int,
precision=red_vx_precision,
tag="red_vx_hi",
debug=debug_fixed32)
red_vx_lo = red_vx - red_vx_hi
red_vx_lo.set_attributes(precision=red_vx_precision,
tag="red_vx_lo",
debug=debug_fixed32)
table_index = BitLogicRightShift(TypeCast(red_vx, precision=ML_Int32),
scaling_power -
(table_size_log - max_bound_log),
precision=ML_Int32,
tag="table_index",
debug=debugd)
tabulated_cos = TableLoad(fused_table,
table_index,
0,
tag="tab_cos",
precision=storage_precision,
debug=debug_fixed32)
tabulated_sin = TableLoad(fused_table,
table_index,
1,
tag="tab_sin",
precision=storage_precision,
debug=debug_fixed32)
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
Log.report(Log.Info, "building polynomial approximation for cosine")
# cosine polynomial approximation
poly_interval = Interval(0, S2**(max_bound_log - table_size_log))
Log.report(Log.Info, "poly_interval=%s " % poly_interval)
cos_poly_degree = 2 # int(sup(guessdegree(cos(x), poly_interval, accuracy_goal)))
Log.report(Log.Verbose, "cosine polynomial approximation")
cos_poly_object, cos_approx_error = Polynomial.build_from_approximation_with_error(
cos(sollya.x), [0, 2],
[0] + [computation_precision.get_bit_size()],
poly_interval,
sollya.absolute,
error_function=error_function)
#cos_eval_scheme = PolynomialSchemeEvaluator.generate_horner_scheme(cos_poly_object, red_vx_lo, unified_precision = computation_precision)
Log.report(Log.Info, "cos_approx_error=%e" % cos_approx_error)
cos_coeff_list = cos_poly_object.get_ordered_coeff_list()
coeff_C0 = cos_coeff_list[0][1]
coeff_C2 = Constant(cos_coeff_list[1][1],
precision=ML_Custom_FixedPoint_Format(-1,
32,
signed=True))
Log.report(Log.Info, "building polynomial approximation for sine")
# sine polynomial approximation
sin_poly_degree = 2 # int(sup(guessdegree(sin(x)/x, poly_interval, accuracy_goal)))
Log.report(Log.Info, "sine poly degree: %e" % sin_poly_degree)
Log.report(Log.Verbose, "sine polynomial approximation")
sin_poly_object, sin_approx_error = Polynomial.build_from_approximation_with_error(
sin(sollya.x) / sollya.x, [0, 2], [0] +
[computation_precision.get_bit_size()] * (sin_poly_degree + 1),
poly_interval,
sollya.absolute,
error_function=error_function)
sin_coeff_list = sin_poly_object.get_ordered_coeff_list()
coeff_S0 = sin_coeff_list[0][1]
coeff_S2 = Constant(sin_coeff_list[1][1],
precision=ML_Custom_FixedPoint_Format(-1,
32,
signed=True))
# scheme selection between sine and cosine
if self.cos_output:
scheme = self.generate_cos_scheme(computation_precision,
tabulated_cos, tabulated_sin,
coeff_S2, coeff_C2, red_vx_lo)
else:
scheme = self.generate_sin_scheme(computation_precision,
tabulated_cos, tabulated_sin,
coeff_S2, coeff_C2, red_vx_lo)
result = Conversion(scheme, precision=self.get_output_precision())
Log.report(
Log.Verbose, "result operation tree :\n %s " % result.get_str(
display_precision=True, depth=None, memoization_map={}))
scheme = Statement(Return(result))
return scheme
def get_value_exp(value):
""" return the binary exponent of value """
return sollya.ceil(sollya.log2(abs(value)))
def generate_scalar_scheme(self, vx):
abs_vx = Abs(vx, precision=self.precision)
FCT_LIMIT = 1.0
one_limit = search_bound_threshold(sollya.erf, FCT_LIMIT, 1.0, 10.0,
self.precision)
one_limit_exp = int(sollya.floor(sollya.log2(one_limit)))
Log.report(Log.Debug, "erf(x) = 1.0 limit is {}, with exp={}",
one_limit, one_limit_exp)
upper_approx_bound = 10
# empiral numbers
eps_exp = {ML_Binary32: -3, ML_Binary64: -5}[self.precision]
eps = S2**eps_exp
Log.report(Log.Info, "building mathematical polynomial")
approx_interval = Interval(0, eps)
# fonction to approximate is erf(x) / x
# it is an even function erf(x) / x = erf(-x) / (-x)
approx_fct = sollya.erf(sollya.x) - (sollya.x)
poly_degree = int(
sup(
guessdegree(approx_fct, approx_interval, S2**
-(self.precision.get_field_size() + 5)))) + 1
poly_degree_list = list(range(1, poly_degree, 2))
Log.report(Log.Debug, "poly_degree is {} and list {}", poly_degree,
poly_degree_list)
global_poly_object = Polynomial.build_from_approximation(
approx_fct, poly_degree_list,
[self.precision] * len(poly_degree_list), approx_interval,
sollya.relative)
Log.report(
Log.Debug, "inform is {}",
dirtyinfnorm(approx_fct - global_poly_object.get_sollya_object(),
approx_interval))
poly_object = global_poly_object.sub_poly(start_index=1, offset=1)
ext_precision = {
ML_Binary32: ML_SingleSingle,
ML_Binary64: ML_DoubleDouble,
}[self.precision]
pre_poly = PolynomialSchemeEvaluator.generate_horner_scheme(
poly_object, abs_vx, unified_precision=self.precision)
result = FMA(pre_poly, abs_vx, abs_vx)
result.set_attributes(tag="result", debug=debug_multi)
eps_target = S2**-(self.precision.get_field_size() + 5)
def offset_div_function(fct):
return lambda offset: fct(sollya.x + offset)
# empiral numbers
field_size = {ML_Binary32: 6, ML_Binary64: 8}[self.precision]
near_indexing = SubFPIndexing(eps_exp, 0, 6, self.precision)
near_approx = generic_poly_split(offset_div_function(sollya.erf),
near_indexing, eps_target,
self.precision, abs_vx)
near_approx.set_attributes(tag="near_approx", debug=debug_multi)
def offset_function(fct):
return lambda offset: fct(sollya.x + offset)
medium_indexing = SubFPIndexing(1, one_limit_exp, 7, self.precision)
medium_approx = generic_poly_split(offset_function(sollya.erf),
medium_indexing, eps_target,
self.precision, abs_vx)
medium_approx.set_attributes(tag="medium_approx", debug=debug_multi)
# approximation for positive values
scheme = ConditionBlock(
abs_vx < eps, Return(result),
ConditionBlock(
abs_vx < near_indexing.get_max_bound(), Return(near_approx),
ConditionBlock(abs_vx < medium_indexing.get_max_bound(),
Return(medium_approx),
Return(Constant(1.0,
precision=self.precision)))))
return scheme
def generate_scheme(self):
# declaring target and instantiating optimization engine
vx = self.implementation.add_input_variable("x", self.precision)
Log.set_dump_stdout(True)
Log.report(Log.Info,
"\033[33;1m generating implementation scheme \033[0m")
if self.debug_flag:
Log.report(Log.Info, "\033[31;1m debug has been enabled \033[0;m")
# local overloading of RaiseReturn operation
def ExpRaiseReturn(*args, **kwords):
kwords["arg_value"] = vx
kwords["function_name"] = self.function_name
if self.libm_compliant:
return RaiseReturn(*args, precision=self.precision, **kwords)
else:
return Return(kwords["return_value"], precision=self.precision)
test_nan_or_inf = Test(vx,
specifier=Test.IsInfOrNaN,
likely=False,
debug=debug_multi,
tag="nan_or_inf")
test_nan = Test(vx,
specifier=Test.IsNaN,
debug=debug_multi,
tag="is_nan_test")
test_positive = Comparison(vx,
0,
specifier=Comparison.GreaterOrEqual,
debug=debug_multi,
tag="inf_sign")
test_signaling_nan = Test(vx,
specifier=Test.IsSignalingNaN,
debug=debug_multi,
tag="is_signaling_nan")
return_snan = Statement(
ExpRaiseReturn(ML_FPE_Invalid,
return_value=FP_QNaN(self.precision)))
# return in case of infinity input
infty_return = Statement(
ConditionBlock(
test_positive,
Return(FP_PlusInfty(self.precision), precision=self.precision),
Return(FP_PlusZero(self.precision), precision=self.precision)))
# return in case of specific value input (NaN or inf)
specific_return = ConditionBlock(
test_nan,
ConditionBlock(
test_signaling_nan, return_snan,
Return(FP_QNaN(self.precision), precision=self.precision)),
infty_return)
# return in case of standard (non-special) input
# exclusion of early overflow and underflow cases
precision_emax = self.precision.get_emax()
precision_max_value = S2 * S2**precision_emax
exp_overflow_bound = sollya.ceil(log(precision_max_value))
early_overflow_test = Comparison(vx,
exp_overflow_bound,
likely=False,
specifier=Comparison.Greater)
early_overflow_return = Statement(
ClearException() if self.libm_compliant else Statement(),
ExpRaiseReturn(ML_FPE_Inexact,
ML_FPE_Overflow,
return_value=FP_PlusInfty(self.precision)))
precision_emin = self.precision.get_emin_subnormal()
precision_min_value = S2**precision_emin
exp_underflow_bound = floor(log(precision_min_value))
early_underflow_test = Comparison(vx,
exp_underflow_bound,
likely=False,
specifier=Comparison.Less)
early_underflow_return = Statement(
ClearException() if self.libm_compliant else Statement(),
ExpRaiseReturn(ML_FPE_Inexact,
ML_FPE_Underflow,
return_value=FP_PlusZero(self.precision)))
# constant computation
invlog2 = self.precision.round_sollya_object(1 / log(2), sollya.RN)
interval_vx = Interval(exp_underflow_bound, exp_overflow_bound)
interval_fk = interval_vx * invlog2
interval_k = Interval(floor(inf(interval_fk)),
sollya.ceil(sup(interval_fk)))
log2_hi_precision = self.precision.get_field_size() - (
sollya.ceil(log2(sup(abs(interval_k)))) + 2)
Log.report(Log.Info, "log2_hi_precision: %d" % log2_hi_precision)
invlog2_cst = Constant(invlog2, precision=self.precision)
log2_hi = round(log(2), log2_hi_precision, sollya.RN)
log2_lo = self.precision.round_sollya_object(
log(2) - log2_hi, sollya.RN)
# argument reduction
unround_k = vx * invlog2
unround_k.set_attributes(tag="unround_k", debug=debug_multi)
k = NearestInteger(unround_k,
precision=self.precision,
debug=debug_multi)
ik = NearestInteger(unround_k,
precision=self.precision.get_integer_format(),
debug=debug_multi,
tag="ik")
ik.set_tag("ik")
k.set_tag("k")
exact_pre_mul = (k * log2_hi)
exact_pre_mul.set_attributes(exact=True)
exact_hi_part = vx - exact_pre_mul
exact_hi_part.set_attributes(exact=True,
tag="exact_hi",
debug=debug_multi,
prevent_optimization=True)
exact_lo_part = -k * log2_lo
exact_lo_part.set_attributes(tag="exact_lo",
debug=debug_multi,
prevent_optimization=True)
r = exact_hi_part + exact_lo_part
r.set_tag("r")
r.set_attributes(debug=debug_multi)
approx_interval = Interval(-log(2) / 2, log(2) / 2)
approx_interval_half = approx_interval / 2
approx_interval_split = [
Interval(-log(2) / 2, inf(approx_interval_half)),
approx_interval_half,
Interval(sup(approx_interval_half),
log(2) / 2)
]
# TODO: should be computed automatically
exact_hi_interval = approx_interval
exact_lo_interval = -interval_k * log2_lo
opt_r = self.optimise_scheme(r, copy={})
tag_map = {}
self.opt_engine.register_nodes_by_tag(opt_r, tag_map)
cg_eval_error_copy_map = {
vx:
Variable("x", precision=self.precision, interval=interval_vx),
tag_map["k"]:
Variable("k", interval=interval_k, precision=self.precision)
}
#try:
if is_gappa_installed():
eval_error = self.gappa_engine.get_eval_error_v2(
self.opt_engine,
opt_r,
cg_eval_error_copy_map,
gappa_filename="red_arg.g")
else:
eval_error = 0.0
Log.report(Log.Warning,
"gappa is not installed in this environnement")
Log.report(Log.Info, "eval error: %s" % eval_error)
local_ulp = sup(ulp(sollya.exp(approx_interval), self.precision))
# FIXME refactor error_goal from accuracy
Log.report(Log.Info, "accuracy: %s" % self.accuracy)
if isinstance(self.accuracy, ML_Faithful):
error_goal = local_ulp
elif isinstance(self.accuracy, ML_CorrectlyRounded):
error_goal = S2**-1 * local_ulp
elif isinstance(self.accuracy, ML_DegradedAccuracyAbsolute):
error_goal = self.accuracy.goal
elif isinstance(self.accuracy, ML_DegradedAccuracyRelative):
error_goal = self.accuracy.goal
else:
Log.report(Log.Error, "unknown accuracy: %s" % self.accuracy)
# error_goal = local_ulp #S2**-(self.precision.get_field_size()+1)
error_goal_approx = S2**-1 * error_goal
Log.report(Log.Info,
"\033[33;1m building mathematical polynomial \033[0m\n")
poly_degree = max(
sup(
guessdegree(
expm1(sollya.x) / sollya.x, approx_interval,
error_goal_approx)) - 1, 2)
init_poly_degree = poly_degree
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_estrin_scheme
#polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_horner_scheme
while 1:
Log.report(Log.Info, "attempting poly degree: %d" % poly_degree)
precision_list = [1] + [self.precision] * (poly_degree)
poly_object, poly_approx_error = Polynomial.build_from_approximation_with_error(
expm1(sollya.x),
poly_degree,
precision_list,
approx_interval,
sollya.absolute,
error_function=error_function)
Log.report(Log.Info, "polynomial: %s " % poly_object)
sub_poly = poly_object.sub_poly(start_index=2)
Log.report(Log.Info, "polynomial: %s " % sub_poly)
Log.report(Log.Info, "poly approx error: %s" % poly_approx_error)
Log.report(
Log.Info,
"\033[33;1m generating polynomial evaluation scheme \033[0m")
pre_poly = polynomial_scheme_builder(
poly_object, r, unified_precision=self.precision)
pre_poly.set_attributes(tag="pre_poly", debug=debug_multi)
pre_sub_poly = polynomial_scheme_builder(
sub_poly, r, unified_precision=self.precision)
pre_sub_poly.set_attributes(tag="pre_sub_poly", debug=debug_multi)
poly = 1 + (exact_hi_part + (exact_lo_part + pre_sub_poly))
poly.set_tag("poly")
# optimizing poly before evaluation error computation
#opt_poly = self.opt_engine.optimization_process(poly, self.precision, fuse_fma = fuse_fma)
#opt_sub_poly = self.opt_engine.optimization_process(pre_sub_poly, self.precision, fuse_fma = fuse_fma)
opt_poly = self.optimise_scheme(poly)
opt_sub_poly = self.optimise_scheme(pre_sub_poly)
# evaluating error of the polynomial approximation
r_gappa_var = Variable("r",
precision=self.precision,
interval=approx_interval)
exact_hi_gappa_var = Variable("exact_hi",
precision=self.precision,
interval=exact_hi_interval)
exact_lo_gappa_var = Variable("exact_lo",
precision=self.precision,
interval=exact_lo_interval)
vx_gappa_var = Variable("x",
precision=self.precision,
interval=interval_vx)
k_gappa_var = Variable("k",
interval=interval_k,
precision=self.precision)
#print "exact_hi interval: ", exact_hi_interval
sub_poly_error_copy_map = {
#r.get_handle().get_node(): r_gappa_var,
#vx.get_handle().get_node(): vx_gappa_var,
exact_hi_part.get_handle().get_node():
exact_hi_gappa_var,
exact_lo_part.get_handle().get_node():
exact_lo_gappa_var,
#k.get_handle().get_node(): k_gappa_var,
}
poly_error_copy_map = {
exact_hi_part.get_handle().get_node(): exact_hi_gappa_var,
exact_lo_part.get_handle().get_node(): exact_lo_gappa_var,
}
if is_gappa_installed():
sub_poly_eval_error = -1.0
sub_poly_eval_error = self.gappa_engine.get_eval_error_v2(
self.opt_engine,
opt_sub_poly,
sub_poly_error_copy_map,
gappa_filename="%s_gappa_sub_poly.g" % self.function_name)
dichotomy_map = [
{
exact_hi_part.get_handle().get_node():
approx_interval_split[0],
},
{
exact_hi_part.get_handle().get_node():
approx_interval_split[1],
},
{
exact_hi_part.get_handle().get_node():
approx_interval_split[2],
},
]
poly_eval_error_dico = self.gappa_engine.get_eval_error_v3(
self.opt_engine,
opt_poly,
poly_error_copy_map,
gappa_filename="gappa_poly.g",
dichotomy=dichotomy_map)
poly_eval_error = max(
[sup(abs(err)) for err in poly_eval_error_dico])
else:
poly_eval_error = 0.0
sub_poly_eval_error = 0.0
Log.report(Log.Warning,
"gappa is not installed in this environnement")
Log.report(Log.Info, "stopping autonomous degree research")
# incrementing polynomial degree to counteract initial decrementation effect
poly_degree += 1
break
Log.report(Log.Info, "poly evaluation error: %s" % poly_eval_error)
Log.report(Log.Info,
"sub poly evaluation error: %s" % sub_poly_eval_error)
global_poly_error = None
global_rel_poly_error = None
for case_index in range(3):
poly_error = poly_approx_error + poly_eval_error_dico[
case_index]
rel_poly_error = sup(
abs(poly_error /
sollya.exp(approx_interval_split[case_index])))
if global_rel_poly_error == None or rel_poly_error > global_rel_poly_error:
global_rel_poly_error = rel_poly_error
global_poly_error = poly_error
flag = error_goal > global_rel_poly_error
if flag:
break
else:
poly_degree += 1
late_overflow_test = Comparison(ik,
self.precision.get_emax(),
specifier=Comparison.Greater,
likely=False,
debug=debug_multi,
tag="late_overflow_test")
overflow_exp_offset = (self.precision.get_emax() -
self.precision.get_field_size() / 2)
diff_k = Subtraction(
ik,
Constant(overflow_exp_offset,
precision=self.precision.get_integer_format()),
precision=self.precision.get_integer_format(),
debug=debug_multi,
tag="diff_k",
)
late_overflow_result = (ExponentInsertion(
diff_k, precision=self.precision) * poly) * ExponentInsertion(
overflow_exp_offset, precision=self.precision)
late_overflow_result.set_attributes(silent=False,
tag="late_overflow_result",
debug=debug_multi,
precision=self.precision)
late_overflow_return = ConditionBlock(
Test(late_overflow_result, specifier=Test.IsInfty, likely=False),
ExpRaiseReturn(ML_FPE_Overflow,
return_value=FP_PlusInfty(self.precision)),
Return(late_overflow_result, precision=self.precision))
late_underflow_test = Comparison(k,
self.precision.get_emin_normal(),
specifier=Comparison.LessOrEqual,
likely=False)
underflow_exp_offset = 2 * self.precision.get_field_size()
corrected_exp = Addition(
ik,
Constant(underflow_exp_offset,
precision=self.precision.get_integer_format()),
precision=self.precision.get_integer_format(),
tag="corrected_exp")
late_underflow_result = (
ExponentInsertion(corrected_exp, precision=self.precision) *
poly) * ExponentInsertion(-underflow_exp_offset,
precision=self.precision)
late_underflow_result.set_attributes(debug=debug_multi,
tag="late_underflow_result",
silent=False)
test_subnormal = Test(late_underflow_result,
specifier=Test.IsSubnormal)
late_underflow_return = Statement(
ConditionBlock(
test_subnormal,
ExpRaiseReturn(ML_FPE_Underflow,
return_value=late_underflow_result)),
Return(late_underflow_result, precision=self.precision))
twok = ExponentInsertion(ik,
tag="exp_ik",
debug=debug_multi,
precision=self.precision)
#std_result = twok * ((1 + exact_hi_part * pre_poly) + exact_lo_part * pre_poly)
std_result = twok * poly
std_result.set_attributes(tag="std_result", debug=debug_multi)
result_scheme = ConditionBlock(
late_overflow_test, late_overflow_return,
ConditionBlock(late_underflow_test, late_underflow_return,
Return(std_result, precision=self.precision)))
std_return = ConditionBlock(
early_overflow_test, early_overflow_return,
ConditionBlock(early_underflow_test, early_underflow_return,
result_scheme))
# main scheme
Log.report(Log.Info, "\033[33;1m MDL scheme \033[0m")
scheme = ConditionBlock(
test_nan_or_inf,
Statement(ClearException() if self.libm_compliant else Statement(),
specific_return), std_return)
return scheme
def generate_scheme(self):
# declaring CodeFunction and retrieving input variable
vx = self.implementation.add_input_variable("x", self.precision)
Log.report(Log.Info, "generating implementation scheme")
if self.debug_flag:
Log.report(Log.Info, "debug has been enabled")
# local overloading of RaiseReturn operation
def SincosRaiseReturn(*args, **kwords):
kwords["arg_value"] = vx
kwords["function_name"] = self.function_name
return RaiseReturn(*args, **kwords)
sollya_precision = self.precision.get_sollya_object()
hi_precision = self.precision.get_field_size() - 8
cw_hi_precision = self.precision.get_field_size() - 4
ext_precision = {
ML_Binary32: ML_Binary64,
ML_Binary64: ML_Binary64
}[self.precision]
int_precision = {
ML_Binary32: ML_Int32,
ML_Binary64: ML_Int64
}[self.precision]
if self.precision is ML_Binary32:
ph_bound = S2**10
else:
ph_bound = S2**33
test_ph_bound = Comparison(vx,
ph_bound,
specifier=Comparison.GreaterOrEqual,
precision=ML_Bool,
likely=False)
# argument reduction
# m
frac_pi_index = {ML_Binary32: 10, ML_Binary64: 14}[self.precision]
C0 = Constant(0, precision=int_precision)
C1 = Constant(1, precision=int_precision)
C_offset = Constant(3 * S2**(frac_pi_index - 1),
precision=int_precision)
# 2^m / pi
frac_pi = round(S2**frac_pi_index / pi, cw_hi_precision, sollya.RN)
frac_pi_lo = round(S2**frac_pi_index / pi - frac_pi, sollya_precision,
sollya.RN)
# pi / 2^m, high part
inv_frac_pi = round(pi / S2**frac_pi_index, cw_hi_precision, sollya.RN)
# pi / 2^m, low part
inv_frac_pi_lo = round(pi / S2**frac_pi_index - inv_frac_pi,
sollya_precision, sollya.RN)
# computing k
vx.set_attributes(tag="vx", debug=debug_multi)
vx_pi = Addition(Multiplication(vx,
Constant(frac_pi,
precision=self.precision),
precision=self.precision),
Multiplication(vx,
Constant(frac_pi_lo,
precision=self.precision),
precision=self.precision),
precision=self.precision,
tag="vx_pi",
debug=debug_multi)
k = NearestInteger(vx_pi,
precision=int_precision,
tag="k",
debug=debug_multi)
# k in floating-point precision
fk = Conversion(k,
precision=self.precision,
tag="fk",
debug=debug_multi)
inv_frac_pi_cst = Constant(inv_frac_pi,
tag="inv_frac_pi",
precision=self.precision,
debug=debug_multi)
inv_frac_pi_lo_cst = Constant(inv_frac_pi_lo,
tag="inv_frac_pi_lo",
precision=self.precision,
debug=debug_multi)
# Cody-Waite reduction
red_coeff1 = Multiplication(fk,
inv_frac_pi_cst,
precision=self.precision,
exact=True)
red_coeff2 = Multiplication(Negation(fk, precision=self.precision),
inv_frac_pi_lo_cst,
precision=self.precision,
exact=True)
# Should be exact / Sterbenz' Lemma
pre_sub_mul = Subtraction(vx,
red_coeff1,
precision=self.precision,
exact=True)
# Fast2Sum
s = Addition(pre_sub_mul,
red_coeff2,
precision=self.precision,
unbreakable=True,
tag="s",
debug=debug_multi)
z = Subtraction(s,
pre_sub_mul,
precision=self.precision,
unbreakable=True,
tag="z",
debug=debug_multi)
t = Subtraction(red_coeff2,
z,
precision=self.precision,
unbreakable=True,
tag="t",
debug=debug_multi)
red_vx_std = Addition(s, t, precision=self.precision)
red_vx_std.set_attributes(tag="red_vx_std", debug=debug_multi)
# To compute sine we offset x by 3pi/2
# which means add 3 * S2^(frac_pi_index-1) to k
if self.sin_output:
Log.report(Log.Info, "Computing Sin")
offset_k = Addition(k,
C_offset,
precision=int_precision,
tag="offset_k")
else:
Log.report(Log.Info, "Computing Cos")
offset_k = k
modk = Variable("modk",
precision=int_precision,
var_type=Variable.Local)
red_vx = Variable("red_vx",
precision=self.precision,
var_type=Variable.Local)
# Faster modulo using bitwise logic
modk_std = BitLogicAnd(offset_k,
2**(frac_pi_index + 1) - 1,
precision=int_precision,
tag="modk",
debug=debug_multi)
approx_interval = Interval(-pi / (S2**(frac_pi_index + 1)),
pi / S2**(frac_pi_index + 1))
red_vx.set_interval(approx_interval)
Log.report(Log.Info, "approx interval: %s\n" % approx_interval)
Log.report(Log.Info,
"building tabulated approximation for sin and cos")
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
# polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_estrin_scheme
polynomial_scheme_builder = PolynomialSchemeEvaluator.generate_horner_scheme
table_index_size = frac_pi_index + 1
cos_table = ML_NewTable(dimensions=[2**table_index_size, 1],
storage_precision=self.precision,
tag=self.uniquify_name("cos_table"))
for i in range(2**(frac_pi_index + 1)):
local_x = i * pi / S2**frac_pi_index
cos_local = round(cos(local_x), self.precision.get_sollya_object(),
sollya.RN)
cos_table[i][0] = cos_local
sin_index = Modulo(modk + 2**(frac_pi_index - 1),
2**(frac_pi_index + 1),
precision=int_precision,
tag="sin_index") #, debug = debug_multi)
tabulated_cos = TableLoad(cos_table,
modk,
C0,
precision=self.precision,
tag="tab_cos",
debug=debug_multi)
tabulated_sin = -TableLoad(cos_table,
sin_index,
C0,
precision=self.precision,
tag="tab_sin",
debug=debug_multi)
poly_degree_cos = sup(
guessdegree(cos(sollya.x), approx_interval, S2**
-self.precision.get_precision()) + 2)
poly_degree_sin = sup(
guessdegree(
sin(sollya.x) / sollya.x, approx_interval, S2**
-self.precision.get_precision()) + 2)
poly_degree_cos_list = range(0, int(poly_degree_cos) + 3)
poly_degree_sin_list = range(0, int(poly_degree_sin) + 3)
# cosine polynomial: limiting first and second coefficient precision to 1-bit
poly_cos_prec_list = [self.precision] * len(poly_degree_cos_list)
# sine polynomial: limiting first coefficient precision to 1-bit
poly_sin_prec_list = [self.precision] * len(poly_degree_sin_list)
error_function = lambda p, f, ai, mod, t: dirtyinfnorm(f - p, ai)
Log.report(Log.Info,
"building mathematical polynomials for sin and cos")
# Polynomial approximations
Log.report(Log.Info, "cos")
poly_object_cos, poly_error_cos = Polynomial.build_from_approximation_with_error(
cos(sollya.x),
poly_degree_cos_list,
poly_cos_prec_list,
approx_interval,
sollya.absolute,
error_function=error_function)
Log.report(Log.Info, "sin")
poly_object_sin, poly_error_sin = Polynomial.build_from_approximation_with_error(
sin(sollya.x),
poly_degree_sin_list,
poly_sin_prec_list,
approx_interval,
sollya.absolute,
error_function=error_function)
Log.report(
Log.Info, "poly error cos: {} / {:d}".format(
poly_error_cos, int(sollya.log2(poly_error_cos))))
Log.report(
Log.Info, "poly error sin: {0} / {1:d}".format(
poly_error_sin, int(sollya.log2(poly_error_sin))))
Log.report(Log.Info, "poly cos : %s" % poly_object_cos)
Log.report(Log.Info, "poly sin : %s" % poly_object_sin)
# Polynomial evaluation scheme
poly_cos = polynomial_scheme_builder(
poly_object_cos.sub_poly(start_index=1),
red_vx,
unified_precision=self.precision)
poly_sin = polynomial_scheme_builder(
poly_object_sin.sub_poly(start_index=2),
red_vx,
unified_precision=self.precision)
poly_cos.set_attributes(tag="poly_cos", debug=debug_multi)
poly_sin.set_attributes(tag="poly_sin",
debug=debug_multi,
unbreakable=True)
# TwoProductFMA
mul_cos_x = tabulated_cos * poly_cos
mul_cos_y = FusedMultiplyAdd(tabulated_cos,
poly_cos,
-mul_cos_x,
precision=self.precision)
mul_sin_x = tabulated_sin * poly_sin
mul_sin_y = FusedMultiplyAdd(tabulated_sin,
poly_sin,
-mul_sin_x,
precision=self.precision)
mul_coeff_sin_hi = tabulated_sin * red_vx
mul_coeff_sin_lo = FusedMultiplyAdd(tabulated_sin, red_vx,
-mul_coeff_sin_hi)
mul_cos = Addition(mul_cos_x,
mul_cos_y,
precision=self.precision,
tag="mul_cos") #, debug = debug_multi)
mul_sin = Negation(Addition(mul_sin_x,
mul_sin_y,
precision=self.precision),
precision=self.precision,
tag="mul_sin") #, debug = debug_multi)
mul_coeff_sin = Negation(Addition(mul_coeff_sin_hi,
mul_coeff_sin_lo,
precision=self.precision),
precision=self.precision,
tag="mul_coeff_sin") #, debug = debug_multi)
mul_cos_x.set_attributes(
tag="mul_cos_x", precision=self.precision) #, debug = debug_multi)
mul_cos_y.set_attributes(
tag="mul_cos_y", precision=self.precision) #, debug = debug_multi)
mul_sin_x.set_attributes(
tag="mul_sin_x", precision=self.precision) #, debug = debug_multi)
mul_sin_y.set_attributes(
tag="mul_sin_y", precision=self.precision) #, debug = debug_multi)
cos_eval_d_1 = (((mul_cos + mul_sin) + mul_coeff_sin) + tabulated_cos)
cos_eval_d_1.set_attributes(tag="cos_eval_d_1",
precision=self.precision,
debug=debug_multi)
result_1 = Statement(Return(cos_eval_d_1))
#######################################################################
# LARGE ARGUMENT MANAGEMENT #
# (lar: Large Argument Reduction) #
#######################################################################
# payne and hanek argument reduction for large arguments
ph_k = frac_pi_index
ph_frac_pi = round(S2**ph_k / pi, 1500, sollya.RN)
ph_inv_frac_pi = pi / S2**ph_k
ph_statement, ph_acc, ph_acc_int = generate_payne_hanek(vx,
ph_frac_pi,
self.precision,
n=100,
k=ph_k)
# assigning Large Argument Reduction reduced variable
lar_vx = Variable("lar_vx",
precision=self.precision,
var_type=Variable.Local)
lar_red_vx = Addition(Multiplication(lar_vx,
inv_frac_pi,
precision=self.precision),
Multiplication(lar_vx,
inv_frac_pi_lo,
precision=self.precision),
precision=self.precision,
tag="lar_red_vx",
debug=debug_multi)
C32 = Constant(2**(ph_k + 1), precision=int_precision, tag="C32")
ph_acc_int_red = Select(ph_acc_int < C0,
C32 + ph_acc_int,
ph_acc_int,
precision=int_precision,
tag="ph_acc_int_red")
if self.sin_output:
lar_offset_k = Addition(ph_acc_int_red,
C_offset,
precision=int_precision,
tag="lar_offset_k")
else:
lar_offset_k = ph_acc_int_red
ph_acc_int_red.set_attributes(tag="ph_acc_int_red", debug=debug_multi)
lar_modk = BitLogicAnd(lar_offset_k,
2**(frac_pi_index + 1) - 1,
precision=int_precision,
tag="lar_modk",
debug=debug_multi)
lar_statement = Statement(ph_statement,
ReferenceAssign(lar_vx,
ph_acc,
debug=debug_multi),
ReferenceAssign(red_vx,
lar_red_vx,
debug=debug_multi),
ReferenceAssign(modk, lar_modk),
prevent_optimization=True)
test_NaN_or_Inf = Test(vx,
specifier=Test.IsInfOrNaN,
likely=False,
tag="NaN_or_Inf",
debug=debug_multi)
return_NaN_or_Inf = Statement(Return(FP_QNaN(self.precision)))
scheme = ConditionBlock(
test_NaN_or_Inf, Statement(ClearException(), return_NaN_or_Inf),
Statement(
modk, red_vx,
ConditionBlock(
test_ph_bound, lar_statement,
Statement(
ReferenceAssign(modk, modk_std),
ReferenceAssign(red_vx, red_vx_std),
)), result_1))
return scheme