diff options
Diffstat (limited to 'newlib/libm/math/w_gamma.c')
| -rw-r--r-- | newlib/libm/math/w_gamma.c | 12 |
1 files changed, 6 insertions, 6 deletions
diff --git a/newlib/libm/math/w_gamma.c b/newlib/libm/math/w_gamma.c index fad40496d..da0211555 100644 --- a/newlib/libm/math/w_gamma.c +++ b/newlib/libm/math/w_gamma.c @@ -76,9 +76,9 @@ $\mit ln\bigl(\Gamma(x)\bigr)$, the natural logarithm of the gamma function of <[x]>. The gamma function (<<exp(gamma(<[x]>))>>) is a generalization of factorial, and retains the property that -@ifnottex +@ifinfo <<exp(gamma(N))>> is equivalent to <<N*exp(gamma(N-1))>>. -@end ifnottex +@end ifinfo @tex $\mit \Gamma(N)\equiv N\times\Gamma(N-1)$. @end tex @@ -87,10 +87,10 @@ quickly. <<gamma>> is defined as @tex $\mit ln\bigl(\Gamma(x)\bigr)$ rather than simply $\mit \Gamma(x)$ @end tex -@ifnottex +@ifinfo the natural log of the gamma function, rather than the gamma function itself, -@end ifnottex +@end ifinfo to extend the useful range of results representable. The sign of the result is returned in the global variable <<signgam>>, @@ -145,11 +145,11 @@ Neither <<gamma>> nor <<gammaf>> is ANSI C. */ #endif { #ifdef _IEEE_LIBM - return __ieee754_gamma_r(x,&(_REENT_SIGNGAM(_REENT))); + return __ieee754_gamma_r(x,&(_REENT->_new._reent._gamma_signgam)); #else double y; struct exception exc; - y = __ieee754_gamma_r(x,&(_REENT_SIGNGAM(_REENT))); + y = __ieee754_gamma_r(x,&(_REENT->_new._reent._gamma_signgam)); if(_LIB_VERSION == _IEEE_) return y; if(!finite(y)&&finite(x)) { #ifndef HUGE_VAL |