 Szabolcs Nagy
|
8f438115f2
math: fix rare underflow issue in fma
|
11 年之前 |
|
Szabolcs Nagy
|
63b9cc7773
math: remove *_WORD64 macros from libm.h
|
11 年之前 |
|
Szabolcs Nagy
|
34660d73bd
math: fix remaining old long double code (erfl, fmal, lgammal, scalbnl)
|
11 年之前 |
|
Szabolcs Nagy
|
1e5eb73545
math: add fma TODO comments about the underflow issue
|
12 年之前 |
|
Szabolcs Nagy
|
ffd8ac2dd5
math: fix two fma issues (only affects non-nearest rounding mode, x86)
|
12 年之前 |
|
Szabolcs Nagy
|
033a9d6ad2
math: use '#pragma STDC FENV_ACCESS ON' when fenv is accessed
|
12 年之前 |
|
nsz
|
e5fb6820a4
math: fix fma bug on x86 (found by Bruno Haible with gnulib)
|
13 年之前 |
|
nsz
|
2786c7d216
use scalbn or *2.0 instead of ldexp, fix fmal
|
13 年之前 |
|
nsz
|
682e471400
remove unnecessary TODO comments from fma.c
|
13 年之前 |
|
nsz
|
b1cbd70743
add fma implementation for x86
|
13 年之前 |
 Rich Felker
|
2e77dc13f8
make fma and lrint functions build without full fenv support
|
13 年之前 |
|
Rich Felker
|
b69f695ace
first commit of the new libm!
|
13 年之前 |
