diff options
| author | Rich Felker <dalias@aerifal.cx> | 2013-11-20 17:40:33 -0500 |
|---|---|---|
| committer | Rich Felker <dalias@aerifal.cx> | 2013-11-20 17:40:33 -0500 |
| commit | 46db37289f917e23877a0e0df88cbb150805bc97 (patch) | |
| tree | e39e35854721b6912e3484207aeba5991dbc9ed6 /include/math.h | |
| parent | d8e8f1464cb02d6a62f01c7153ca4d7b0cd5c5e6 (diff) | |
| download | musl-46db37289f917e23877a0e0df88cbb150805bc97.tar.gz | |
write floating point limit constants to 21 significant decimal places
this is enough to produce the correct value even if the constant is
interpreted as 80-bit extended precision, which matters on archs with
excess precision (FLT_EVAL_METHOD==2) under at least some
interpretations of the C standard. the shorter representations, while
correct if converted to the nominal precision at translation time,
could produce an incorrect value at extended precision, yielding
results such as (double)DBL_MAX != DBL_MAX.
Diffstat (limited to 'include/math.h')
| -rw-r--r-- | include/math.h | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/include/math.h b/include/math.h index dc176012..3c2105ef 100644 --- a/include/math.h +++ b/include/math.h @@ -349,7 +349,7 @@ long double truncl(long double); #if defined(_XOPEN_SOURCE) || defined(_BSD_SOURCE) #undef MAXFLOAT -#define MAXFLOAT 3.40282347e+38F +#define MAXFLOAT 3.40282346638528859812e+38F #endif #if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) || defined(_BSD_SOURCE) @@ -379,7 +379,7 @@ double yn(int, double); #endif #if defined(_GNU_SOURCE) || defined(_BSD_SOURCE) -#define HUGE 3.40282347e+38F +#define HUGE 3.40282346638528859812e+38F double scalb(double, double); float scalbf(float, float); |