Add Intel-linux to list of toolsets that have no working long double support in Boost.Math.

[SVN r47966]

status/explicit-failures-markup.xml

@@ -3181,6 +3181,9 @@ for more information.
        <mark-expected-failures>
           <test name="test_tr1_long_double"/>
           <toolset name="darwin*"/>
+          <toolset name="intel-linux-10.0"/>
+          <toolset name="intel-linux-9*"/>
+          <toolset name="intel-linux-8*"/>
           <note author="John Maddock">
              Some versions of the Darwin platform have insufficient long double support
              for us to be able to run this test.