<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7652.24">
<TITLE>[gmx-developers] double precision, gmx-3.3.1 and 3.3.2 do not work for me</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<P><FONT SIZE=2>Hi,<BR>
<BR>
Sorry for the delay but Im having problems with sending email to this list - some spam filters must be at work.<BR>
<BR>
<BR>
I think Ive had the same problem - at least the symptoms were the same - a running single precision dataset<BR>
when supplied to a double precision program gave instant infinite forces as shown<BR>
- more specifically for me the LJ-14 and Coul-14 were the ones which gave rise to the problem having stupidly<BR>
high energies.<BR>
<BR>
However, I had another double precision version on another machine and when I tried there it worked.<BR>
<BR>
So it was obviously the build - Id copied the double precision source from a single precision compile.<BR>
<BR>
I redid the build with two important steps<BR>
<BR>
make clean<BR>
<BR>
and<BR>
<BR>
make distclean<BR>
<BR>
re-did the configure using the configure from the other machine ( but as far as I could it was the same as the<BR>
current machine)<BR>
<BR>
and the program now works.<BR>
<BR>
I dont really know what happened and why the new build works but Im suspicous of the make distclean - I could<BR>
well have not done that previously.<BR>
<BR>
There was the issue of a system upgrade around this time which updated the gcc version.<BR>
<BR>
<BR>
David<BR>
<BR>
PS this is probably html email - because of the spam issues Im having to run this using crappy webmail.<BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>