home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: gnu.gcc.bug
- Path: sparky!uunet!usc!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!cygnus.com!mrs
- From: mrs@cygnus.com (Mike Stump)
- Subject: Re: floating point accuracy???
- Message-ID: <9209150813.AA01007@cygnus.com>
- Sender: gnulists@ai.mit.edu
- Organization: GNUs Not Usenet
- Distribution: gnu
- Date: Tue, 15 Sep 1992 08:13:23 GMT
- Approved: bug-gcc@prep.ai.mit.edu
- Lines: 37
-
- This is a problem with the backend I believe. If it is not, then the
- problem in the front-end has been fixed.
-
- On a sun4 I now get:
-
- pow (4,6)= 4096.000000
- pow (4,6)= 4096
-
- > Message-Id: <m0mUHwh-000AWzC@sophists.com>
- > Date: Mon, 14 Sep 92 11:02 EDT
- > From: lewis@sophists.com (Lewis G. Pringle)
- > To: bug-gcc@prep.ai.mit.edu
- > Reply-To: lewis@sophists.com
-
- > /*
- > * I'm not sure if this is a bug, a feature, or a configuration
- > * problem, but the results of these math calls were a bit
- > * surprising to me.
- > *
- > * This code we compiled with:
- > * gcc -o fred x.cc -lm
- > * on a 68k box running a sysVr2.2 variant using gcc 2.2.1.
- > *
- > * The output was:
- > * pow (4,6)= 4096.000000
- > * pow (4,6)= 4095
- > */
-
- > #include <math.h>
- > #include <stdio.h>
-
- > main ()
- > {
- > printf ("pow (4,6)= %f\n", pow (4,6));
- > printf ("pow (4,6)= %d\n", ((unsigned) pow (4,6)));
- > }
-
-