mirror of
https://sourceware.org/git/glibc.git
synced 2024-11-21 20:40:05 +00:00
d9bef9c0a4
The tgmath.h macros return a real type not a complex type when an argument is of complex integer type (a GNU extension) and there are no arguments of complex floating type. It seems clear that just as real integers are mapped to double for tgmath.h, so complex integers should be mapped to _Complex double. This patch implements such a mapping. The main complication in fixing this bug is that the tgmath.h macros expand their arguments a large number of times, resulting in exponential blowup of the size of the expansion when calls to tgmath.h macros are used in the arguments of such macros; it would be unfortunate for fixing a bug with a fairly obscure extension to make the macros expand their arguments even more times. Thus, this patch optimizes the definitions of the relevant macros. __tgmath_real_type previously expanded its argument 7 times and now expands it 3 times. __tgmath_complex_type, used in place of __tgmath_real_type only for functions that might return either real or complex types, not for complex functions that always return real types or always return complex types, expands its argument 5 times. So the sizes of the macro expansions from nested macro calls are correspondingly reduced (remembering that each tgmath.h macro expands __tgmath_real_type, or sometimes now __tgmath_complex_type, several times). Sometimes the real return type resulted from calling a complex function and converting the result to a real type; sometimes it resulted from calling a real function, because the logic for determining whether arguments were real or complex, based on sizeof, was confused by integer promotions applying to e.g. short int but not _Complex short int. The relevant tests are converted to use a new macro __expr_is_real, which, by calling __builtin_classify_type rather than comparing the results of two calls to sizeof, also reduces the number of times macros expand their arguments. Although there are reductions in the number of times macros expand their arguments, I do not consider this to fix bug 21660, since a proper fix means each macro expanding its arguments only once (via using new compiler features designed for that purpose). Tested for x86_64. [BZ #21684] * math/tgmath.h (__floating_type): Simplify definitions. (__real_integer_type): New macro. (__complex_integer_type): Likewise. (__expr_is_real): Likewise. (__tgmath_real_type_sub): Update comment to describe handling of complex types. (__tgmath_complex_type_sub): New macro. (__tgmath_complex_type): Likewise. [__HAVE_FLOAT128 && __GLIBC_USE (IEC_60559_TYPES_EXT)] (__TGMATH_CF128): Use __expr_is_real. (__TGMATH_UNARY_REAL_IMAG): Use __tgmath_complex_type and __expr_is_real. (__TGMATH_BINARY_REAL_IMAG): Likewise. (__TGMATH_UNARY_REAL_IMAG_RET_REAL): Use __expr_is_real. * math/gen-tgmath-tests.py (Type.create_type): Create complex integer types.
626 lines
28 KiB
Python
Executable File
626 lines
28 KiB
Python
Executable File
#!/usr/bin/python
|
|
# Generate tests for <tgmath.h> macros.
|
|
# Copyright (C) 2017 Free Software Foundation, Inc.
|
|
# This file is part of the GNU C Library.
|
|
#
|
|
# The GNU C Library is free software; you can redistribute it and/or
|
|
# modify it under the terms of the GNU Lesser General Public
|
|
# License as published by the Free Software Foundation; either
|
|
# version 2.1 of the License, or (at your option) any later version.
|
|
#
|
|
# The GNU C Library is distributed in the hope that it will be useful,
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
# Lesser General Public License for more details.
|
|
#
|
|
# You should have received a copy of the GNU Lesser General Public
|
|
# License along with the GNU C Library; if not, see
|
|
# <http://www.gnu.org/licenses/>.
|
|
|
|
# As glibc does not support decimal floating point, the types to
|
|
# consider for generic parameters are standard and binary
|
|
# floating-point types, and integer types which are treated as double.
|
|
# The corresponding complex types may also be used (including complex
|
|
# integer types, which are a GNU extension, but are currently disabled
|
|
# here because they do not work properly with tgmath.h).
|
|
|
|
# The proposed resolution to TS 18661-1 DR#9
|
|
# <http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2149.htm#dr_9>
|
|
# makes the <tgmath.h> rules for selecting a function to call
|
|
# correspond to the usual arithmetic conversions (applied successively
|
|
# to the arguments for generic parameters in order), which choose the
|
|
# type whose set of values contains that of the other type (undefined
|
|
# behavior if neither type's set of values is a superset of the
|
|
# other), with interchange types being preferred to standard types
|
|
# (long double, double, float), being preferred to extended types
|
|
# (_Float128x, _Float64x, _Float32x).
|
|
|
|
# For the standard and binary floating-point types supported by GCC 7
|
|
# on any platform, this means the resulting type is the last of the
|
|
# given types in one of the following orders, or undefined behavior if
|
|
# types with both ibm128 and binary128 representation are specified.
|
|
|
|
# If double = long double: _Float16, float, _Float32, _Float32x,
|
|
# double, long double, _Float64, _Float64x, _Float128.
|
|
|
|
# Otherwise: _Float16, float, _Float32, _Float32x, double, _Float64,
|
|
# _Float64x, long double, _Float128.
|
|
|
|
# We generate tests to verify the return type is exactly as expected.
|
|
# We also verify that the function called is real or complex as
|
|
# expected, and that it is called for the right floating-point format
|
|
# (but it is OK to call a double function instead of a long double one
|
|
# if they have the same format, for example). For all the formats
|
|
# supported on any given configuration of glibc, the MANT_DIG value
|
|
# uniquely determines the format.
|
|
|
|
import string
|
|
|
|
class Type(object):
|
|
"""A type that may be used as an argument for generic parameters."""
|
|
|
|
# All possible argument or result types.
|
|
all_types_list = []
|
|
# All argument types.
|
|
argument_types_list = []
|
|
# All real argument types.
|
|
real_argument_types_list = []
|
|
# Real argument types that correspond to a standard floating type
|
|
# (float, double or long double; not _FloatN or _FloatNx).
|
|
standard_real_argument_types_list = []
|
|
# The real floating types by their order properties (which are
|
|
# tuples giving the positions in both the possible orders above).
|
|
real_types_order = {}
|
|
# The type double.
|
|
double_type = None
|
|
# The type _Complex double.
|
|
complex_double_type = None
|
|
# The type _Float64.
|
|
float64_type = None
|
|
# The type _Float64x.
|
|
float64x_type = None
|
|
|
|
def __init__(self, name, suffix=None, mant_dig=None, condition='1',
|
|
order=None, integer=False, complex=False, real_type=None):
|
|
"""Initialize a Type object, creating any corresponding complex type
|
|
in the process."""
|
|
self.name = name
|
|
self.suffix = suffix
|
|
self.mant_dig = mant_dig
|
|
self.condition = condition
|
|
self.order = order
|
|
self.integer = integer
|
|
self.complex = complex
|
|
if complex:
|
|
self.complex_type = self
|
|
self.real_type = real_type
|
|
else:
|
|
# complex_type filled in by the caller once created.
|
|
self.complex_type = None
|
|
self.real_type = self
|
|
|
|
def register_type(self, internal):
|
|
"""Record a type in the lists of all types."""
|
|
Type.all_types_list.append(self)
|
|
if not internal:
|
|
Type.argument_types_list.append(self)
|
|
if not self.complex:
|
|
Type.real_argument_types_list.append(self)
|
|
if not self.name.startswith('_Float'):
|
|
Type.standard_real_argument_types_list.append(self)
|
|
if self.order is not None:
|
|
Type.real_types_order[self.order] = self
|
|
if self.name == 'double':
|
|
Type.double_type = self
|
|
if self.name == '_Complex double':
|
|
Type.complex_double_type = self
|
|
if self.name == '_Float64':
|
|
Type.float64_type = self
|
|
if self.name == '_Float64x':
|
|
Type.float64x_type = self
|
|
|
|
@staticmethod
|
|
def create_type(name, suffix=None, mant_dig=None, condition='1', order=None,
|
|
integer=False, complex_name=None, complex_ok=True,
|
|
internal=False):
|
|
"""Create and register a Type object for a real type, creating any
|
|
corresponding complex type in the process."""
|
|
real_type = Type(name, suffix=suffix, mant_dig=mant_dig,
|
|
condition=condition, order=order, integer=integer,
|
|
complex=False)
|
|
if complex_ok:
|
|
if complex_name is None:
|
|
complex_name = '_Complex %s' % name
|
|
complex_type = Type(complex_name, condition=condition,
|
|
integer=integer, complex=True,
|
|
real_type=real_type)
|
|
else:
|
|
complex_type = None
|
|
real_type.complex_type = complex_type
|
|
real_type.register_type(internal)
|
|
if complex_type is not None:
|
|
complex_type.register_type(internal)
|
|
|
|
def floating_type(self):
|
|
"""Return the corresponding floating type."""
|
|
if self.integer:
|
|
return (Type.complex_double_type
|
|
if self.complex
|
|
else Type.double_type)
|
|
else:
|
|
return self
|
|
|
|
def real_floating_type(self):
|
|
"""Return the corresponding real floating type."""
|
|
return self.real_type.floating_type()
|
|
|
|
def __str__(self):
|
|
"""Return string representation of a type."""
|
|
return self.name
|
|
|
|
@staticmethod
|
|
def init_types():
|
|
"""Initialize all the known types."""
|
|
Type.create_type('_Float16', 'f16', 'FLT16_MANT_DIG',
|
|
complex_name='__CFLOAT16',
|
|
condition='defined HUGE_VAL_F16', order=(0, 0))
|
|
Type.create_type('float', 'f', 'FLT_MANT_DIG', order=(1, 1))
|
|
Type.create_type('_Float32', 'f32', 'FLT32_MANT_DIG',
|
|
complex_name='__CFLOAT32',
|
|
condition='defined HUGE_VAL_F32', order=(2, 2))
|
|
Type.create_type('_Float32x', 'f32x', 'FLT32X_MANT_DIG',
|
|
complex_name='__CFLOAT32X',
|
|
condition='defined HUGE_VAL_F32X', order=(3, 3))
|
|
Type.create_type('double', '', 'DBL_MANT_DIG', order=(4, 4))
|
|
Type.create_type('long double', 'l', 'LDBL_MANT_DIG', order=(5, 7))
|
|
Type.create_type('_Float64', 'f64', 'FLT64_MANT_DIG',
|
|
complex_name='__CFLOAT64',
|
|
condition='defined HUGE_VAL_F64', order=(6, 5))
|
|
Type.create_type('_Float64x', 'f64x', 'FLT64X_MANT_DIG',
|
|
complex_name='__CFLOAT64X',
|
|
condition='defined HUGE_VAL_F64X', order=(7, 6))
|
|
Type.create_type('_Float128', 'f128', 'FLT128_MANT_DIG',
|
|
complex_name='__CFLOAT128',
|
|
condition='defined HUGE_VAL_F128', order=(8, 8))
|
|
Type.create_type('char', integer=True)
|
|
Type.create_type('signed char', integer=True)
|
|
Type.create_type('unsigned char', integer=True)
|
|
Type.create_type('short int', integer=True)
|
|
Type.create_type('unsigned short int', integer=True)
|
|
Type.create_type('int', integer=True)
|
|
Type.create_type('unsigned int', integer=True)
|
|
Type.create_type('long int', integer=True)
|
|
Type.create_type('unsigned long int', integer=True)
|
|
Type.create_type('long long int', integer=True)
|
|
Type.create_type('unsigned long long int', integer=True)
|
|
Type.create_type('__int128', integer=True,
|
|
condition='defined __SIZEOF_INT128__')
|
|
Type.create_type('unsigned __int128', integer=True,
|
|
condition='defined __SIZEOF_INT128__')
|
|
Type.create_type('enum e', integer=True, complex_ok=False)
|
|
Type.create_type('_Bool', integer=True, complex_ok=False)
|
|
Type.create_type('bit_field', integer=True, complex_ok=False)
|
|
# Internal types represent the combination of long double with
|
|
# _Float64 or _Float64x, for which the ordering depends on
|
|
# whether long double has the same format as double.
|
|
Type.create_type('long_double_Float64', 'LDBL_MANT_DIG',
|
|
complex_name='complex_long_double_Float64',
|
|
condition='defined HUGE_VAL_F64', order=(6, 7),
|
|
internal=True)
|
|
Type.create_type('long_double_Float64x', 'FLT64X_MANT_DIG',
|
|
complex_name='complex_long_double_Float64x',
|
|
condition='defined HUGE_VAL_F64X', order=(7, 7),
|
|
internal=True)
|
|
|
|
@staticmethod
|
|
def can_combine_types(types):
|
|
"""Return a C preprocessor conditional for whether the given list of
|
|
types can be used together as type-generic macro arguments."""
|
|
have_long_double = False
|
|
have_float128 = False
|
|
for t in types:
|
|
t = t.real_floating_type()
|
|
if t.name == 'long double':
|
|
have_long_double = True
|
|
if t.name == '_Float128' or t.name == '_Float64x':
|
|
have_float128 = True
|
|
if have_long_double and have_float128:
|
|
# If ibm128 format is in use for long double, both
|
|
# _Float64x and _Float128 are binary128 and the types
|
|
# cannot be combined.
|
|
return '(LDBL_MANT_DIG != 106)'
|
|
return '1'
|
|
|
|
@staticmethod
|
|
def combine_types(types):
|
|
"""Return the result of combining a set of types."""
|
|
have_complex = False
|
|
combined = None
|
|
for t in types:
|
|
if t.complex:
|
|
have_complex = True
|
|
t = t.real_floating_type()
|
|
if combined is None:
|
|
combined = t
|
|
else:
|
|
order = (max(combined.order[0], t.order[0]),
|
|
max(combined.order[1], t.order[1]))
|
|
combined = Type.real_types_order[order]
|
|
return combined.complex_type if have_complex else combined
|
|
|
|
def list_product_initial(initial, lists):
|
|
"""Return a list of lists, with an initial sequence from the first
|
|
argument (a list of lists) followed by each sequence of one
|
|
element from each successive element of the second argument."""
|
|
if not lists:
|
|
return initial
|
|
return list_product_initial([a + [b] for a in initial for b in lists[0]],
|
|
lists[1:])
|
|
|
|
def list_product(lists):
|
|
"""Return a list of lists, with each sequence of one element from each
|
|
successive element of the argument."""
|
|
return list_product_initial([[]], lists)
|
|
|
|
try:
|
|
trans_id = str.maketrans(' *', '_p')
|
|
except AttributeError:
|
|
trans_id = string.maketrans(' *', '_p')
|
|
def var_for_type(name):
|
|
"""Return the name of a variable with a given type (name)."""
|
|
return 'var_%s' % name.translate(trans_id)
|
|
|
|
def vol_var_for_type(name):
|
|
"""Return the name of a variable with a given volatile type (name)."""
|
|
return 'vol_var_%s' % name.translate(trans_id)
|
|
|
|
def define_vars_for_type(name):
|
|
"""Return the definitions of variables with a given type (name)."""
|
|
if name == 'bit_field':
|
|
struct_vars = define_vars_for_type('struct s');
|
|
return '%s#define %s %s.bf\n' % (struct_vars,
|
|
vol_var_for_type(name),
|
|
vol_var_for_type('struct s'))
|
|
return ('%s %s __attribute__ ((unused));\n'
|
|
'%s volatile %s __attribute__ ((unused));\n'
|
|
% (name, var_for_type(name), name, vol_var_for_type(name)))
|
|
|
|
def if_cond_text(conds, text):
|
|
"""Return the result of making some text conditional under #if. The
|
|
text ends with a newline, as does the return value if not empty."""
|
|
if '0' in conds:
|
|
return ''
|
|
conds = [c for c in conds if c != '1']
|
|
conds = sorted(set(conds))
|
|
if not conds:
|
|
return text
|
|
return '#if %s\n%s#endif\n' % (' && '.join(conds), text)
|
|
|
|
class Tests(object):
|
|
"""The state associated with testcase generation."""
|
|
|
|
def __init__(self):
|
|
"""Initialize a Tests object."""
|
|
self.header_list = ['#define __STDC_WANT_IEC_60559_TYPES_EXT__\n'
|
|
'#include <float.h>\n'
|
|
'#include <stdbool.h>\n'
|
|
'#include <stdint.h>\n'
|
|
'#include <stdio.h>\n'
|
|
'#include <string.h>\n'
|
|
'#include <tgmath.h>\n'
|
|
'\n'
|
|
'struct test\n'
|
|
' {\n'
|
|
' void (*func) (void);\n'
|
|
' const char *func_name;\n'
|
|
' const char *test_name;\n'
|
|
' int mant_dig;\n'
|
|
' };\n'
|
|
'int num_pass, num_fail;\n'
|
|
'volatile int called_mant_dig;\n'
|
|
'const char *volatile called_func_name;\n'
|
|
'enum e { E, F };\n'
|
|
'struct s\n'
|
|
' {\n'
|
|
' int bf:2;\n'
|
|
' };\n']
|
|
float64_text = ('# if LDBL_MANT_DIG == DBL_MANT_DIG\n'
|
|
'typedef _Float64 long_double_Float64;\n'
|
|
'typedef __CFLOAT64 complex_long_double_Float64;\n'
|
|
'# else\n'
|
|
'typedef long double long_double_Float64;\n'
|
|
'typedef _Complex long double '
|
|
'complex_long_double_Float64;\n'
|
|
'# endif\n')
|
|
float64_text = if_cond_text([Type.float64_type.condition],
|
|
float64_text)
|
|
float64x_text = ('# if LDBL_MANT_DIG == DBL_MANT_DIG\n'
|
|
'typedef _Float64x long_double_Float64x;\n'
|
|
'typedef __CFLOAT64X complex_long_double_Float64x;\n'
|
|
'# else\n'
|
|
'typedef long double long_double_Float64x;\n'
|
|
'typedef _Complex long double '
|
|
'complex_long_double_Float64x;\n'
|
|
'# endif\n')
|
|
float64x_text = if_cond_text([Type.float64x_type.condition],
|
|
float64x_text)
|
|
self.header_list.append(float64_text)
|
|
self.header_list.append(float64x_text)
|
|
self.types_seen = set()
|
|
for t in Type.all_types_list:
|
|
self.add_type_var(t.name, t.condition)
|
|
self.test_text_list = []
|
|
self.test_array_list = []
|
|
|
|
def add_type_var(self, name, cond):
|
|
"""Add declarations of variables for a type."""
|
|
if name in self.types_seen:
|
|
return
|
|
t_vars = define_vars_for_type(name)
|
|
self.header_list.append(if_cond_text([cond], t_vars))
|
|
self.types_seen.add(name)
|
|
|
|
def add_tests(self, macro, ret, args, complex_func=None):
|
|
"""Add tests for a given tgmath.h macro."""
|
|
# 'c' means the function argument or return type is
|
|
# type-generic and complex only (a complex function argument
|
|
# may still have a real macro argument). 'g' means it is
|
|
# type-generic and may be real or complex; 'r' means it is
|
|
# type-generic and may only be real; 's' means the same as
|
|
# 'r', but restricted to float, double and long double.
|
|
have_complex = False
|
|
func = macro
|
|
if ret == 'c' or 'c' in args:
|
|
# Complex-only.
|
|
have_complex = True
|
|
complex_func = func
|
|
func = None
|
|
elif ret == 'g' or 'g' in args:
|
|
# Real and complex.
|
|
have_complex = True
|
|
if complex_func == None:
|
|
complex_func = 'c%s' % func
|
|
types = [ret] + args
|
|
for t in types:
|
|
if t != 'c' and t != 'g' and t != 'r' and t != 's':
|
|
self.add_type_var(t, '1')
|
|
for t in Type.argument_types_list:
|
|
if t.integer:
|
|
continue
|
|
if t.complex and not have_complex:
|
|
continue
|
|
if func == None and not t.complex:
|
|
continue
|
|
if ret == 's' and t.name.startswith('_Float'):
|
|
continue
|
|
if ret == 'c':
|
|
ret_name = t.complex_type.name
|
|
elif ret == 'g':
|
|
ret_name = t.name
|
|
elif ret == 'r' or ret == 's':
|
|
ret_name = t.real_type.name
|
|
else:
|
|
ret_name = ret
|
|
dummy_func_name = complex_func if t.complex else func
|
|
arg_list = []
|
|
arg_num = 0
|
|
for a in args:
|
|
if a == 'c':
|
|
arg_name = t.complex_type.name
|
|
elif a == 'g':
|
|
arg_name = t.name
|
|
elif a == 'r' or a == 's':
|
|
arg_name = t.real_type.name
|
|
else:
|
|
arg_name = a
|
|
arg_list.append('%s arg%d __attribute__ ((unused))'
|
|
% (arg_name, arg_num))
|
|
arg_num += 1
|
|
dummy_func = ('%s\n'
|
|
'(%s%s) (%s)\n'
|
|
'{\n'
|
|
' called_mant_dig = %s;\n'
|
|
' called_func_name = "%s";\n'
|
|
' return 0;\n'
|
|
'}\n' % (ret_name, dummy_func_name,
|
|
t.real_type.suffix, ', '.join(arg_list),
|
|
t.real_type.mant_dig, dummy_func_name))
|
|
dummy_func = if_cond_text([t.condition], dummy_func)
|
|
self.test_text_list.append(dummy_func)
|
|
arg_types = []
|
|
for t in args:
|
|
if t == 'g' or t == 'c':
|
|
arg_types.append(Type.argument_types_list)
|
|
elif t == 'r':
|
|
arg_types.append(Type.real_argument_types_list)
|
|
elif t == 's':
|
|
arg_types.append(Type.standard_real_argument_types_list)
|
|
arg_types_product = list_product(arg_types)
|
|
test_num = 0
|
|
for this_args in arg_types_product:
|
|
comb_type = Type.combine_types(this_args)
|
|
can_comb = Type.can_combine_types(this_args)
|
|
all_conds = [t.condition for t in this_args]
|
|
all_conds.append(can_comb)
|
|
any_complex = func == None
|
|
for t in this_args:
|
|
if t.complex:
|
|
any_complex = True
|
|
func_name = complex_func if any_complex else func
|
|
test_name = '%s (%s)' % (macro,
|
|
', '.join([t.name for t in this_args]))
|
|
test_func_name = 'test_%s_%d' % (macro, test_num)
|
|
test_num += 1
|
|
mant_dig = comb_type.real_type.mant_dig
|
|
test_text = '%s, "%s", "%s", %s' % (test_func_name, func_name,
|
|
test_name, mant_dig)
|
|
test_text = ' { %s },\n' % test_text
|
|
test_text = if_cond_text(all_conds, test_text)
|
|
self.test_array_list.append(test_text)
|
|
call_args = []
|
|
call_arg_pos = 0
|
|
for t in args:
|
|
if t == 'g' or t == 'c' or t == 'r' or t == 's':
|
|
type = this_args[call_arg_pos].name
|
|
call_arg_pos += 1
|
|
else:
|
|
type = t
|
|
call_args.append(vol_var_for_type(type))
|
|
call_args_text = ', '.join(call_args)
|
|
if ret == 'g':
|
|
ret_type = comb_type.name
|
|
elif ret == 'r' or ret == 's':
|
|
ret_type = comb_type.real_type.name
|
|
elif ret == 'c':
|
|
ret_type = comb_type.complex_type.name
|
|
else:
|
|
ret_type = ret
|
|
call_text = '%s (%s)' % (macro, call_args_text)
|
|
test_func_text = ('static void\n'
|
|
'%s (void)\n'
|
|
'{\n'
|
|
' extern typeof (%s) %s '
|
|
'__attribute__ ((unused));\n'
|
|
' %s = %s;\n'
|
|
'}\n' % (test_func_name, call_text,
|
|
var_for_type(ret_type),
|
|
vol_var_for_type(ret_type), call_text))
|
|
test_func_text = if_cond_text(all_conds, test_func_text)
|
|
self.test_text_list.append(test_func_text)
|
|
|
|
def add_all_tests(self):
|
|
"""Add tests for all tgmath.h macros."""
|
|
# C99/C11 real-only functions.
|
|
self.add_tests('atan2', 'r', ['r', 'r'])
|
|
self.add_tests('cbrt', 'r', ['r'])
|
|
self.add_tests('ceil', 'r', ['r'])
|
|
self.add_tests('copysign', 'r', ['r', 'r'])
|
|
self.add_tests('erf', 'r', ['r'])
|
|
self.add_tests('erfc', 'r', ['r'])
|
|
self.add_tests('exp2', 'r', ['r'])
|
|
self.add_tests('expm1', 'r', ['r'])
|
|
self.add_tests('fdim', 'r', ['r', 'r'])
|
|
self.add_tests('floor', 'r', ['r'])
|
|
self.add_tests('fma', 'r', ['r', 'r', 'r'])
|
|
self.add_tests('fmax', 'r', ['r', 'r'])
|
|
self.add_tests('fmin', 'r', ['r', 'r'])
|
|
self.add_tests('fmod', 'r', ['r', 'r'])
|
|
self.add_tests('frexp', 'r', ['r', 'int *'])
|
|
self.add_tests('hypot', 'r', ['r', 'r'])
|
|
self.add_tests('ilogb', 'int', ['r'])
|
|
self.add_tests('ldexp', 'r', ['r', 'int'])
|
|
self.add_tests('lgamma', 'r', ['r'])
|
|
self.add_tests('llrint', 'long long int', ['r'])
|
|
self.add_tests('llround', 'long long int', ['r'])
|
|
# log10 is real-only in ISO C, but supports complex arguments
|
|
# as a GNU extension.
|
|
self.add_tests('log10', 'g', ['g'])
|
|
self.add_tests('log1p', 'r', ['r'])
|
|
self.add_tests('log2', 'r', ['r'])
|
|
self.add_tests('logb', 'r', ['r'])
|
|
self.add_tests('lrint', 'long int', ['r'])
|
|
self.add_tests('lround', 'long int', ['r'])
|
|
self.add_tests('nearbyint', 'r', ['r'])
|
|
self.add_tests('nextafter', 'r', ['r', 'r'])
|
|
self.add_tests('nexttoward', 's', ['s', 'long double'])
|
|
self.add_tests('remainder', 'r', ['r', 'r'])
|
|
self.add_tests('remquo', 'r', ['r', 'r', 'int *'])
|
|
self.add_tests('rint', 'r', ['r'])
|
|
self.add_tests('round', 'r', ['r'])
|
|
self.add_tests('scalbn', 'r', ['r', 'int'])
|
|
self.add_tests('scalbln', 'r', ['r', 'long int'])
|
|
self.add_tests('tgamma', 'r', ['r'])
|
|
self.add_tests('trunc', 'r', ['r'])
|
|
# C99/C11 real-and-complex functions.
|
|
self.add_tests('acos', 'g', ['g'])
|
|
self.add_tests('asin', 'g', ['g'])
|
|
self.add_tests('atan', 'g', ['g'])
|
|
self.add_tests('acosh', 'g', ['g'])
|
|
self.add_tests('asinh', 'g', ['g'])
|
|
self.add_tests('atanh', 'g', ['g'])
|
|
self.add_tests('cos', 'g', ['g'])
|
|
self.add_tests('sin', 'g', ['g'])
|
|
self.add_tests('tan', 'g', ['g'])
|
|
self.add_tests('cosh', 'g', ['g'])
|
|
self.add_tests('sinh', 'g', ['g'])
|
|
self.add_tests('tanh', 'g', ['g'])
|
|
self.add_tests('exp', 'g', ['g'])
|
|
self.add_tests('log', 'g', ['g'])
|
|
self.add_tests('pow', 'g', ['g', 'g'])
|
|
self.add_tests('sqrt', 'g', ['g'])
|
|
self.add_tests('fabs', 'r', ['g'], 'cabs')
|
|
# C99/C11 complex-only functions.
|
|
self.add_tests('carg', 'r', ['c'])
|
|
self.add_tests('cimag', 'r', ['c'])
|
|
self.add_tests('conj', 'c', ['c'])
|
|
self.add_tests('cproj', 'c', ['c'])
|
|
self.add_tests('creal', 'r', ['c'])
|
|
# TS 18661-1 functions.
|
|
self.add_tests('roundeven', 'r', ['r'])
|
|
self.add_tests('nextup', 'r', ['r'])
|
|
self.add_tests('nextdown', 'r', ['r'])
|
|
self.add_tests('fminmag', 'r', ['r', 'r'])
|
|
self.add_tests('fmaxmag', 'r', ['r', 'r'])
|
|
self.add_tests('llogb', 'long int', ['r'])
|
|
self.add_tests('fromfp', 'intmax_t', ['r', 'int', 'unsigned int'])
|
|
self.add_tests('fromfpx', 'intmax_t', ['r', 'int', 'unsigned int'])
|
|
self.add_tests('ufromfp', 'uintmax_t', ['r', 'int', 'unsigned int'])
|
|
self.add_tests('ufromfpx', 'uintmax_t', ['r', 'int', 'unsigned int'])
|
|
self.add_tests('totalorder', 'int', ['r', 'r'])
|
|
self.add_tests('totalordermag', 'int', ['r', 'r'])
|
|
# The functions that round their result to a narrower type,
|
|
# and the associated type-generic macros, are not yet
|
|
# supported by this script or by glibc.
|
|
# Miscellaneous functions.
|
|
self.add_tests('scalb', 's', ['s', 's'])
|
|
|
|
def tests_text(self):
|
|
"""Return the text of the generated testcase."""
|
|
test_list = [''.join(self.test_text_list),
|
|
'static const struct test tests[] =\n'
|
|
' {\n',
|
|
''.join(self.test_array_list),
|
|
' };\n']
|
|
footer_list = ['static int\n'
|
|
'do_test (void)\n'
|
|
'{\n'
|
|
' for (size_t i = 0;\n'
|
|
' i < sizeof (tests) / sizeof (tests[0]);\n'
|
|
' i++)\n'
|
|
' {\n'
|
|
' called_mant_dig = 0;\n'
|
|
' called_func_name = "";\n'
|
|
' tests[i].func ();\n'
|
|
' if (called_mant_dig == tests[i].mant_dig\n'
|
|
' && strcmp (called_func_name,\n'
|
|
' tests[i].func_name) == 0)\n'
|
|
' num_pass++;\n'
|
|
' else\n'
|
|
' {\n'
|
|
' num_fail++;\n'
|
|
' printf ("Test %zu (%s):\\n"\n'
|
|
' " Expected: %s precision %d\\n"\n'
|
|
' " Actual: %s precision %d\\n\\n",\n'
|
|
' i, tests[i].test_name,\n'
|
|
' tests[i].func_name,\n'
|
|
' tests[i].mant_dig,\n'
|
|
' called_func_name, called_mant_dig);\n'
|
|
' }\n'
|
|
' }\n'
|
|
' printf ("%d pass, %d fail\\n", num_pass, num_fail);\n'
|
|
' return num_fail != 0;\n'
|
|
'}\n'
|
|
'\n'
|
|
'#include <support/test-driver.c>']
|
|
return ''.join(self.header_list + test_list + footer_list)
|
|
|
|
def main():
|
|
"""The main entry point."""
|
|
Type.init_types()
|
|
t = Tests()
|
|
t.add_all_tests()
|
|
print(t.tests_text())
|
|
|
|
if __name__ == '__main__':
|
|
main()
|