<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=windows-1255"
 http-equiv="Content-Type">
</head>
<body dir="ltr" bgcolor="#ffffff" text="#000000">
ronys wrote:
<blockquote cite="mid:AA1A2DBAAC28487EB77AF43428BEC572@LYDIA"
 type="cite">
  <meta content="text/html; charset=windows-1255"
 http-equiv="Content-Type">
  <meta name="GENERATOR" content="MSHTML 8.00.6001.18876">
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">Hi Shachar,</span></font></div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010"></span></font> </div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">Interesting problem. Here are some thoughts:</span></font></div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010"></span></font> </div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">Can you control the level of optimization
used by the customer? Does -O0 create identical object code?</span></font></div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010"></span></font> <br>
  </div>
</blockquote>
Havn't tested yet, but my gut feeling is "yes".<br>
<blockquote cite="mid:AA1A2DBAAC28487EB77AF43428BEC572@LYDIA"
 type="cite">
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">What are the differences in code? Perhaps
they're related to the different paths? (If the lengths of the paths
are different, and they're stored somewhere in the object code, that'll
change the results)</span></font></div>
</blockquote>
The question is why should identical source code produce different
paths when compiled with the same compiler?<br>
<blockquote cite="mid:AA1A2DBAAC28487EB77AF43428BEC572@LYDIA"
 type="cite">
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010"></span></font> </div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">You might also want to compare the
assembly-language output (-S) to narrow down the problem.</span></font></div>
</blockquote>
-d is better, as it does not overlay the source. If you read what I
wrote - that was what I was doing.<br>
<blockquote cite="mid:AA1A2DBAAC28487EB77AF43428BEC572@LYDIA"
 type="cite">
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010"></span></font> </div>
  <div><font color="#0000ff" face="Arial" size="2"><span
 class="750090008-09022010">If the customer needs cygwin for internal
use, I'd suggest using a virtualization solution such as VirtualBox
instead - the integration with Windows is quite impressive.</span></font></div>
</blockquote>
At this point in time, it will have to remain a dual environment.<br>
<br>
Shachar<br>
<pre class="moz-signature" cols="72">-- 
Shachar Shemesh
Lingnu Open Source Consulting Ltd.
<a class="moz-txt-link-freetext" href="http://www.lingnu.com">http://www.lingnu.com</a>
</pre>
</body>
</html>