tmp/tmpagn54kcx/{from.md → to.md}
RENAMED
|
@@ -2,20 +2,20 @@
|
|
| 2 |
|
| 3 |
### Header `<cfenv>` synopsis <a id="cfenv.syn">[[cfenv.syn]]</a>
|
| 4 |
|
| 5 |
``` cpp
|
| 6 |
#define FE_ALL_EXCEPT see below
|
| 7 |
-
#define FE_DIVBYZERO see below
|
| 8 |
-
#define FE_INEXACT see below
|
| 9 |
-
#define FE_INVALID see below
|
| 10 |
-
#define FE_OVERFLOW see below
|
| 11 |
-
#define FE_UNDERFLOW see below
|
| 12 |
|
| 13 |
-
#define FE_DOWNWARD see below
|
| 14 |
-
#define FE_TONEAREST see below
|
| 15 |
-
#define FE_TOWARDZERO see below
|
| 16 |
-
#define FE_UPWARD see below
|
| 17 |
|
| 18 |
#define FE_DFL_ENV see below
|
| 19 |
|
| 20 |
namespace std {
|
| 21 |
// types
|
|
@@ -40,30 +40,31 @@ namespace std {
|
|
| 40 |
```
|
| 41 |
|
| 42 |
The contents and meaning of the header `<cfenv>` are the same as the C
|
| 43 |
standard library header `<fenv.h>`.
|
| 44 |
|
| 45 |
-
[*Note 1*: This
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
the
|
| 52 |
-
|
| 53 |
-
|
| 54 |
|
| 55 |
-
The floating-point environment has thread storage duration
|
| 56 |
-
[[basic.stc.thread]]
|
| 57 |
environment is the state of the floating-point environment of the thread
|
| 58 |
-
that constructs the corresponding `thread` object
|
| 59 |
-
[[thread.thread.class]]
|
|
|
|
| 60 |
|
| 61 |
[*Note 2*: That is, the child thread gets the floating-point state of
|
| 62 |
the parent thread at the time of the child’s creation. — *end note*]
|
| 63 |
|
| 64 |
-
A separate floating-point environment
|
| 65 |
-
|
| 66 |
-
|
| 67 |
|
| 68 |
-
ISO C 7.6
|
| 69 |
|
|
|
|
| 2 |
|
| 3 |
### Header `<cfenv>` synopsis <a id="cfenv.syn">[[cfenv.syn]]</a>
|
| 4 |
|
| 5 |
``` cpp
|
| 6 |
#define FE_ALL_EXCEPT see below
|
| 7 |
+
#define FE_DIVBYZERO see below // optional
|
| 8 |
+
#define FE_INEXACT see below // optional
|
| 9 |
+
#define FE_INVALID see below // optional
|
| 10 |
+
#define FE_OVERFLOW see below // optional
|
| 11 |
+
#define FE_UNDERFLOW see below // optional
|
| 12 |
|
| 13 |
+
#define FE_DOWNWARD see below // optional
|
| 14 |
+
#define FE_TONEAREST see below // optional
|
| 15 |
+
#define FE_TOWARDZERO see below // optional
|
| 16 |
+
#define FE_UPWARD see below // optional
|
| 17 |
|
| 18 |
#define FE_DFL_ENV see below
|
| 19 |
|
| 20 |
namespace std {
|
| 21 |
// types
|
|
|
|
| 40 |
```
|
| 41 |
|
| 42 |
The contents and meaning of the header `<cfenv>` are the same as the C
|
| 43 |
standard library header `<fenv.h>`.
|
| 44 |
|
| 45 |
+
[*Note 1*: This document does not require an implementation to support
|
| 46 |
+
the `FENV_ACCESS` pragma; it is *implementation-defined* [[cpp.pragma]]
|
| 47 |
+
whether the pragma is supported. As a consequence, it is
|
| 48 |
+
*implementation-defined* whether these functions can be used to test
|
| 49 |
+
floating-point status flags, set floating-point control modes, or run
|
| 50 |
+
under non-default mode settings. If the pragma is used to enable control
|
| 51 |
+
over the floating-point environment, this document does not specify the
|
| 52 |
+
effect on floating-point evaluation in constant
|
| 53 |
+
expressions. — *end note*]
|
| 54 |
|
| 55 |
+
The floating-point environment has thread storage duration
|
| 56 |
+
[[basic.stc.thread]]. The initial state for a thread’s floating-point
|
| 57 |
environment is the state of the floating-point environment of the thread
|
| 58 |
+
that constructs the corresponding `thread` object
|
| 59 |
+
[[thread.thread.class]] or `jthread` object [[thread.jthread.class]] at
|
| 60 |
+
the time it constructed the object.
|
| 61 |
|
| 62 |
[*Note 2*: That is, the child thread gets the floating-point state of
|
| 63 |
the parent thread at the time of the child’s creation. — *end note*]
|
| 64 |
|
| 65 |
+
A separate floating-point environment is maintained for each thread.
|
| 66 |
+
Each function accesses the environment corresponding to its calling
|
| 67 |
+
thread.
|
| 68 |
|
| 69 |
+
See also: ISO C 7.6
|
| 70 |
|