This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] problems in hypotf math function



On Aug 30, 2005, at 3:47 PM, Ulrich Drepper wrote:


I've changed the constant value which was clearly wrong.  But the bias
change has to be justified here in detail.

Fair enough, but you still need to remove the exponent bias from the increment value for 2^60. (The value is being directly added to a floating-point number that already has the exponent bias, so the increment value cannot include the bias.) Here is a patch for that.


2005-08-31 Bob Wilson <bob.wilson@acm.org>

* sysdeps/ieee754/flt-32/e_hypotf.c (__ieee754_hypotf): Remove the
exponent bias from the increment value for scaling by 2^60.


Attachment: hypot.diff
Description: Binary data


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