This is the mail archive of the newlib@sources.redhat.com mailing list for the newlib project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Error of tan/tanf in libm.


Are you using #include <math.h> correctly?  The only libm functions in newlib which I've found to be substandard are acosf() / asinf().   If you were writing code for compatibility with MSVC, you wouldn't dare use any of the float nor long double functions.  If you were looking for best efficiency or accuracy on either i386 (short of P4, which I don't think is supported specifically in open source) or m68k, you would be using something resembling <bits/mathinline.h> from glibc.
----- Original Message -----
From: Cliff Tsai
Sent: Friday, April 06, 2001 4:22 AM
Subject: Error of tan/tanf in libm.

Hi:
    I'm testing libm on m68k platform.
    But I found that there are erros in some functions in libm.
    Like tan/tanf.For example,tanf(0.25) shoud be  close to 0.2553419.
    But when I test tan/tanf in native platform(cygwin on x86)
    And tan/tanf on m68k platform.
    The return is 2.0.Why?
    Do anyone know other errors in libm?
 
    Thanks!
 
   
 

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]