Mail Archives: djgpp/1999/01/08/12:06:50
From: | Charles Krug <charles AT mail DOT pentek DOT com>
|
Newsgroups: | comp.os.msdos.djgpp
|
Subject: | Still working on c++ std <limits> file . . .
|
Date: | Fri, 08 Jan 1999 11:57:29 -0500
|
Organization: | Pentek Corporation
|
Lines: | 26
|
Message-ID: | <369638F9.21FAB385@mail.pentek.com>
|
NNTP-Posting-Host: | mail.pentek.com
|
Mime-Version: | 1.0
|
X-Mailer: | Mozilla 4.5 [en] (WinNT; U)
|
X-Accept-Language: | en
|
To: | djgpp AT delorie DOT com
|
DJ-Gateway: | from newsgroup comp.os.msdos.djgpp
|
Reply-To: | djgpp AT delorie DOT com
|
Hello List:
I was able to find some resources for this that answered most of
my questions about the floating point sections, but a few remain
unresolved.
First, libm defines functions for nan() and nanf(). <limits>
asks for signaling_NaN and for quiet_NaN. Are the nan's defined
in libm signaling or quiet?
Second, do floating point numbers under djgpp have denormalized
forms.
I've a couple of pc <limits> files that I'm using as examples. I
would assume that these are the same, assuming identical floating
point formats, which is true for float and for double. But they
seem to be in disagreement as to the definitions for the NaN's
and for denormalization.
I'm learning a hell of a lot about floating point that I didn't
know here!
Charles
- Raw text -