QNAN 传递到 C 标准库函数（例如 llrintf）：不清楚是否引发 FP 异常
QNAN passed into C standard library functions (ex. llrintf): not clear whether FP exceptions are raised or not
来自 math.h 的宏 NAN 安静 NAN:
ISO/IEC 9899:2011 (E)（重点添加）：
The macro
NAN
is defined if and only if the implementation supports quiet NaNs for the float type. It expands to a constant expression of type float representing a quiet NaN.
安静的 NaN 通常不会导致引发 FP 异常。示例：
- ISO/IEC 9899:2011 (E)（重点添加）：
5.2.4.2.2 Characteristics of floating types <float.h>
3 A quiet NaN propagates through almost every arithmetic operation without raising a floating-point exception; a signaling NaN generally raises a floating-point exception when occurring as an arithmetic operand.
- IEEE 754-2008（重点添加）：
5.11 Details of comparison predicates Programs that explicitly take account of the possibility of quiet NaN operands may use the unordered-quiet predicates in Table 5.3 which do not signal such an invalid operation exception.
但是：
llrintf() 既不是 算术运算 ，也不是 Table 5.3 中的 无序安静谓词。因此，5.2.4.2.2.3 和 5.11 不适用。7.12.9.5 The lrint and llrint functions（例如）在输入为 quiet NaN 的情况下未提及是否引发 FP 异常。
初步结论：由于一般做法“安静的 NaN 不会导致引发 FP 异常”，可以得出结论 lrint 和 llrint 函数应该 不 如果输入是安静的 NaN，则导致引发 FP 异常。
练习：
代码(t125.c):
#include <fenv.h>
#include <math.h>
#include <stdio.h>
#if _MSC_VER
#pragma fenv_access (on)
#else
#pragma STDC FENV_ACCESS ON
#endif
void show_fe_exceptions(void)
{
printf("exceptions raised: ");
if (fetestexcept(FE_DIVBYZERO)) printf(" FE_DIVBYZERO");
if (fetestexcept(FE_INEXACT)) printf(" FE_INEXACT");
if (fetestexcept(FE_INVALID)) printf(" FE_INVALID");
if (fetestexcept(FE_OVERFLOW)) printf(" FE_OVERFLOW");
if (fetestexcept(FE_UNDERFLOW)) printf(" FE_UNDERFLOW");
if (fetestexcept(FE_ALL_EXCEPT)==0) printf(" none");
printf("\n");
}
int main(void)
{
long long ll;
ll = llrintf(NAN);
show_fe_exceptions();
printf("ll %lld\n", ll);
return 0;
}
调用：
$ cl t125.c /std:c11 /fp:strict && t125
exceptions raised: FE_INEXACT FE_INVALID FE_OVERFLOW
ll 0
$ clang t125.c -std=c11 -ffp-model=strict -Wall -Wextra -pedantic && ./a.exe
t125.c:6:9: warning: unknown pragma ignored [-Wunknown-pragmas]
#pragma fenv_access (on)
^
1 warning generated.
exceptions raised: FE_INEXACT FE_INVALID FE_OVERFLOW
ll -9223372036854775808
$ gcc t125.c -std=c11 -Wall -Wextra -pedantic && ./a.exe
t125.c:8: warning: ignoring ‘#pragma STDC FENV_ACCESS’ [-Wunknown-pragmas]
8 | #pragma STDC FENV_ACCESS ON
|
exceptions raised: FE_INVALID
ll -9223372036854775808
问题：如果输入是 quiet NaN，有人可以澄清 C 标准库函数（例如 llrintf）引发 FP 异常的情况/行为吗？
OP 的印象是术语 conversion 仅暗示 隐式转换 和 显式转换 （投射操作）（参见 C11 6.3 Conversions）。这意味着 OP 认为术语 function 不是 conversion。
但是，C11 F.3 Operators and functions（强调）明确指出：
The lrint and llrint functions in <math.h> provide the IEC 60559 conversions.
因此，是的，C11 F.4 Floating to integer conversion（强调）回答了问题：
F.4 Floating to integer conversion
1 ... Otherwise, if the floating value is infinite or NaN or if the integral part of the floating value exceeds the range of the integer type, then the ‘‘invalid’’ floating-point exception is raised and the resulting value is unspecified. ...
来自 math.h 的宏 NAN 安静 NAN:
ISO/IEC 9899:2011 (E)（重点添加）：
The macro
NAN
is defined if and only if the implementation supports quiet NaNs for the float type. It expands to a constant expression of type float representing a quiet NaN.
安静的 NaN 通常不会导致引发 FP 异常。示例：
- ISO/IEC 9899:2011 (E)（重点添加）：
5.2.4.2.2 Characteristics of floating types <float.h>
3 A quiet NaN propagates through almost every arithmetic operation without raising a floating-point exception; a signaling NaN generally raises a floating-point exception when occurring as an arithmetic operand.
- IEEE 754-2008（重点添加）：
5.11 Details of comparison predicates Programs that explicitly take account of the possibility of quiet NaN operands may use the unordered-quiet predicates in Table 5.3 which do not signal such an invalid operation exception.
但是：
llrintf()既不是 算术运算 ，也不是 Table 5.3 中的 无序安静谓词。因此，5.2.4.2.2.3和5.11不适用。7.12.9.5 The lrint and llrint functions（例如）在输入为 quiet NaN 的情况下未提及是否引发 FP 异常。
初步结论：由于一般做法“安静的 NaN 不会导致引发 FP 异常”，可以得出结论 lrint 和 llrint 函数应该 不 如果输入是安静的 NaN，则导致引发 FP 异常。
练习：
代码(t125.c):
#include <fenv.h>
#include <math.h>
#include <stdio.h>
#if _MSC_VER
#pragma fenv_access (on)
#else
#pragma STDC FENV_ACCESS ON
#endif
void show_fe_exceptions(void)
{
printf("exceptions raised: ");
if (fetestexcept(FE_DIVBYZERO)) printf(" FE_DIVBYZERO");
if (fetestexcept(FE_INEXACT)) printf(" FE_INEXACT");
if (fetestexcept(FE_INVALID)) printf(" FE_INVALID");
if (fetestexcept(FE_OVERFLOW)) printf(" FE_OVERFLOW");
if (fetestexcept(FE_UNDERFLOW)) printf(" FE_UNDERFLOW");
if (fetestexcept(FE_ALL_EXCEPT)==0) printf(" none");
printf("\n");
}
int main(void)
{
long long ll;
ll = llrintf(NAN);
show_fe_exceptions();
printf("ll %lld\n", ll);
return 0;
}
调用：
$ cl t125.c /std:c11 /fp:strict && t125
exceptions raised: FE_INEXACT FE_INVALID FE_OVERFLOW
ll 0
$ clang t125.c -std=c11 -ffp-model=strict -Wall -Wextra -pedantic && ./a.exe
t125.c:6:9: warning: unknown pragma ignored [-Wunknown-pragmas]
#pragma fenv_access (on)
^
1 warning generated.
exceptions raised: FE_INEXACT FE_INVALID FE_OVERFLOW
ll -9223372036854775808
$ gcc t125.c -std=c11 -Wall -Wextra -pedantic && ./a.exe
t125.c:8: warning: ignoring ‘#pragma STDC FENV_ACCESS’ [-Wunknown-pragmas]
8 | #pragma STDC FENV_ACCESS ON
|
exceptions raised: FE_INVALID
ll -9223372036854775808
问题：如果输入是 quiet NaN，有人可以澄清 C 标准库函数（例如 llrintf）引发 FP 异常的情况/行为吗？
OP 的印象是术语 conversion 仅暗示 隐式转换 和 显式转换 （投射操作）（参见 C11 6.3 Conversions）。这意味着 OP 认为术语 function 不是 conversion。
但是，C11 F.3 Operators and functions（强调）明确指出：
The
lrintandllrintfunctions in<math.h>provide the IEC 60559 conversions.
因此，是的，C11 F.4 Floating to integer conversion（强调）回答了问题：
F.4 Floating to integer conversion
1 ... Otherwise, if the floating value is infinite or NaN or if the integral part of the floating value exceeds the range of the integer type, then the ‘‘invalid’’ floating-point exception is raised and the resulting value is unspecified. ...