From help-request at octave dot org Fri Sep 23 12:52:23 2005
Subject: RE: lo_ieee_init: unrecognized floating point format! on ARM platform
From: "John W. Eaton" <jwe at bevo dot che dot wisc dot edu>
To: Simon Pickering <S dot G dot Pickering at bath dot ac dot uk>
Cc: help-octave at bevo dot che dot wisc dot edu
Date: Fri, 23 Sep 2005 13:51:43 -0400

On 23-Sep-2005, Simon Pickering wrote:

| One final interesting fact - although the ARM processors can be either
| little- or bigendian (well there are 4 modes actually), mine is running in
| little-endian mode. That said, to make R compile correctly I had to specify
| that it was big-endian and as you've said above - the output looks to be
| big-endian too. I think I need to do some reading up on ARM in general and
| look at how this affects the machar code, etc.

You might also try applying the following patch to machar.c, then
compiling with

  gcc -DDP -DTEST machar.c

then run the resulting binary.  Here is what I see (x86):

  $ ./a.out
  Double  precision MACHAR constants
  ibeta  = 2
  it     = 53
  irnd   = 5
  ngrd   = 0
  machep = -52
  negep  = -53
  iexp   = 11
  minexp = -1022
  maxexp = 1024
  eps      2.2204460492503131e-16          0   3CB00000
  epsneg   1.1102230246251565e-16          0   3CA00000
  xmin    2.2250738585072014e-308          0     100000
  xmax    1.7976931348623157e+308   FFFFFFFF   7FEFFFFF

If you have IEEE floating point, I'd expect the same numbers, but with
the hex values swapped if the byte ordering is different from x86.

jwe


Index: machar.c
===================================================================
RCS file: /cvs/octave/libcruft/misc/machar.c,v
retrieving revision 1.7
diff -u -r1.7 machar.c
--- machar.c	1 Mar 2005 18:01:54 -0000	1.7
+++ machar.c	23 Sep 2005 17:44:09 -0000
 at @ -4,7 +4,9 @@
 
 #include <float.h>
 
+#ifndef TEST
 #include "f77-fcn.h"
+#endif
 
 /*
 
 at @ -364,6 +366,8 @@
 
 }
 
+#ifndef TEST
+
 F77_RET_T
 F77_FUNC (machar, MACHAR) (REAL *xmin, REAL *xmax, REAL *epsneg,
 			   REAL *eps, REAL *log10_ibeta)
 at @ -390,3 +394,79 @@
 
   F77_RETURN (0)
 }
+
+#else
+
+
+/*
+
+This program prints hardware-determined double-precision machine
+constants obtained from rmachar.  Dmachar is a C translation of the
+Fortran routine MACHAR from W. J. Cody, "MACHAR: A subroutine to
+dynamically determine machine parameters".  TOMS (14), 1988.
+
+Descriptions of the machine constants are given in the prologue
+comments in rmachar.
+
+Subprograms called
+
+  rmachar
+
+Original driver: Richard Bartels, October 16, 1985
+
+Modified by: W. J. Cody
+	     July 26, 1988
+ 
+*/
+int
+main (void)
+{
+
+  int ibeta, iexp, irnd, it, machep, maxexp, minexp, negep, ngrd;
+
+  int i;
+
+  REAL eps, epsneg, xmax, xmin;
+
+  union wjc
+  {
+    long int jj[REALSIZE];
+    REAL xbig;
+  } uval;
+
+  rmachar  (&ibeta, &it, &irnd, &ngrd, &machep, &negep, &iexp,
+	    &minexp, &maxexp, &eps, &epsneg, &xmin, &xmax);
+
+  printf (PREC);
+  printf (" precision MACHAR constants\n");
+  printf ("ibeta  = %d\n", ibeta);
+  printf ("it     = %d\n", it);
+  printf ("irnd   = %d\n", irnd);
+  printf ("ngrd   = %d\n", ngrd);
+  printf ("machep = %d\n", machep);
+  printf ("negep  = %d\n", negep);
+  printf ("iexp   = %d\n", iexp);
+  printf ("minexp = %d\n", minexp);
+  printf ("maxexp = %d\n", maxexp);
+
+#define DISPLAY(s, x) \
+  do \
+    { \
+      uval.xbig = x ; \
+      printf (s); \
+      printf (" %24.16e ", (double) x) ; \
+      for (i = 0; i < REALSIZE; i++) \
+	printf (" %9X ", uval.jj[i]) ; \
+      printf ("\n"); \
+    } \
+  while (0)
+			
+  DISPLAY ("eps   ", eps);
+  DISPLAY ("epsneg", epsneg);
+  DISPLAY ("xmin  ", xmin);
+  DISPLAY ("xmax  ", xmax);
+
+  return 0;
+}
+
+#endif



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------