I&#39;ve been running various builds of VC11 for a bit now and though there are a few more failures than VC10, its not too bad. However, I just got a machine up with VC11 using the address-model=64 flag and every single test fails.  <div>

<br></div><div>If you look at the trunk results (<a href="http://www.boost.org/development/tests/trunk/developer/summary.html">http://www.boost.org/development/tests/trunk/developer/summary.html</a>), you&#39;ll see the problem child runner: teeks99-3-win7-64on64. There are two other testers (teeks99-2-win7-32on64 and teeks99-7-win7-32on64) that are running from the same VM image but without the address-model variable and they seem to be nominal. </div>

<div><br></div><div>I&#39;m not sure what I should check from here. I was thinking it might have been a problem with the bjam executable not being updated for VC11 correctly, but I was able to build the binaries for VC11 with the address-model variable just fine, so that makes me think it might be something more on the regression tests side. Any ideas?</div>

<div><br></div><div>Tom Kent</div>