Re: Problem with Excel import on ARM

[Wookey <wookey aleph1 co uk>]

+++ Morten Welinder [04-05-05 12:40 -0400]:
> 
> Gnumeric runs fine of Sparc so big-endian machines are handled.  That
> is not to say that there might not be a screw-up somewhere.
> 
> I cannot actually load your sheet, so I cannot see what is in it.
> (HTTP request sent, awaiting response... 403 Forbidden)

Sorry - Permissions fixed now.

> It would be useful to see what kind of byte layout is being used
> for doubles.  Please do something like this:
> 
> #include <math.h>
> #include <stdio.h>
> int main ()
> {
>   double x = M_PI;
>   const unsigned char *p = (const unsigned char *)&x;
>   int i;
>   for (i = 0; i < sizeof (x); i++)
>     printf ("%02x ", *p++);
>   printf ("\n");
>   return 0;
> }
> 
> I get...
> 
> sparc> ./a.out 
> 40 09 21 fb 54 44 2d 18 
> 
> x86> ./a.out
> 18 2d 44 54 fb 21 09 40 

thanx for that - makes my life very easy :-)

> Please let me know what you get...  If you don't get anything like the
> above, perhaps you know of a reference that tells the layout of doubles
> on ARM?

  arm> ./a.out
  fb 21 09 40 18 2d 44 54
  
(using gcc 3.3.3 (native))

Looks like little-endian within the word but the words are stored
big-endian. That may be what IEEE_DOUBLE_MIXED_ENDIAN refers to in the patch
referred-to below.

I'm still trying to find a definitive reference for the format.
The best I've found so far is a clue here suggests that it is either ieee754
or ieee854:
http://lists.arm.linux.org.uk/pipermail/linux-arm/2004-February/007055.html

HTH

Wookey



-- 
Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK  Tel +44 (0) 1223 811679
work: http://www.aleph1.co.uk/     play: http://www.chaos.org.uk/~wookey/