<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic MKL question in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828167#M5307</link>
    <description>The subroutine by itself gets compiled with no error. The problem lies, then, in the part of the source code that you have &lt;B&gt;not&lt;/B&gt; shown. &lt;BR /&gt;&lt;BR /&gt;Given that the compiler is telling you that there are conflicting subroutines, in your code, switching to F77 would be pointless: it would shift the detection of the bug from compile time to run time.</description>
    <pubDate>Tue, 20 Sep 2011 01:07:41 GMT</pubDate>
    <dc:creator>mecej4</dc:creator>
    <dc:date>2011-09-20T01:07:41Z</dc:date>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828161#M5301</link>
      <description>I haven't used any MKL routines before, and I'm trying to use the spevd eigenvalue routine.&lt;BR /&gt;&lt;BR /&gt;With INCLUDE 'LAPACK.F90' I get 975 warning #6738messages, and the linker says it can't find _spevd.&lt;BR /&gt;&lt;BR /&gt;I must be missing something simple. Can you help?&lt;BR /&gt;&lt;BR /&gt;Thanks!&lt;BR /&gt;Neal</description>
      <pubDate>Mon, 19 Sep 2011 13:44:55 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828161#M5301</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-19T13:44:55Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828162#M5302</link>
      <description>You left out details of where the INCLUDE statement is placed, how you call MKL and how you compiled and linked the program. Without those, it is difficult to tell you what went wrong.&lt;BR /&gt;&lt;BR /&gt;Here is a short example of calling SPEVD using the Lapack95 interface. The upper triangle of the symmetric matrix is packed by columns into the 1-D array Ap.&lt;BR /&gt;&lt;BR /&gt;&lt;PRE&gt;[fortran]program spevdx
use lapack95
integer, parameter :: N=4
real, dimension(N*(N+1)/2) :: Ap= &amp;amp;
   (/-4., &amp;amp;
     2.,-6., &amp;amp;
     0.5,2.,-8., &amp;amp;
     0.,0.3,2.,-10./), &amp;amp;
   w(N)

call spevd(Ap,w)
write(*,10)(i,w(i),i=1,N)

10 format(1x,I3,2x,F10.5)

end program spevdx
[/fortran]&lt;/PRE&gt; Using the current version of IFort (Win32), which comes with the Lapack95 libraries precompiled, build using the command&lt;BR /&gt;&lt;BR /&gt;&lt;PRE&gt;[bash]ifort /Qmkl spevdx.f90 mkl_lapack95.lib
[/bash]&lt;/PRE&gt; and run to obtain the results:&lt;BR /&gt;&lt;BR /&gt;&lt;PRE&gt;[bash]S:LANG&amp;gt;spevdx
   1   -11.37360
   2    -8.43331
   3    -5.91824
   4    -2.27485
[/bash]&lt;/PRE&gt;</description>
      <pubDate>Mon, 19 Sep 2011 16:50:31 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828162#M5302</guid>
      <dc:creator>mecej4</dc:creator>
      <dc:date>2011-09-19T16:50:31Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828163#M5303</link>
      <description>&lt;P&gt;Thanks!&lt;BR /&gt;&lt;BR /&gt;That helped a lot, but when I try it with visual studio I get&lt;BR /&gt;&lt;BR /&gt;"Generic procedure reference has two or more specific procedure with the same type/rank/keyword signature. [SPEVD]"&lt;BR /&gt;&lt;BR /&gt;&lt;EM&gt;&lt;BR /&gt;&lt;/EM&gt;I haven't tried with the command line yet as you suggested because I'm not quite set up for that, but shouldn't it work with visual studio?&lt;/P&gt;</description>
      <pubDate>Mon, 19 Sep 2011 17:59:55 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828163#M5303</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-19T17:59:55Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828164#M5304</link>
      <description>There may be a problem with your installation of Visual Studio or Intel Fortran. All that I had to do was to create a new Fortran console solution with the default options, to which I added these options:&lt;BR /&gt;&lt;BR /&gt;Project=&amp;gt;Properties=&amp;gt;Fortran=&amp;gt;Libraries=&amp;gt;Use Intel MKL   {make your selection}&lt;BR /&gt;Project=&amp;gt;Properties=&amp;gt;Linker=&amp;gt; Additional Dependencies mkl_lapack95.lib&lt;BR /&gt;&lt;BR /&gt;and my example ran fine with IFort 12.1 and VS2010.&lt;BR /&gt;&lt;BR /&gt;&lt;I&gt;&amp;gt; I haven't tried with the command line yet as you suggested because I'm not quite set up for that,..&lt;BR /&gt;&lt;BR /&gt;&lt;/I&gt;There is no set up required. Every installation of Intel Fortran gives you links in the Start Menu to one or more properly configured command windows.</description>
      <pubDate>Mon, 19 Sep 2011 18:20:14 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828164#M5304</guid>
      <dc:creator>mecej4</dc:creator>
      <dc:date>2011-09-19T18:20:14Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828165#M5305</link>
      <description>Thanks a million!&lt;BR /&gt;&lt;BR /&gt;I got it to work. Now I can work it into my programs...&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;NEal&lt;BR /&gt;</description>
      <pubDate>Mon, 19 Sep 2011 19:53:17 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828165#M5305</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-19T19:53:17Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828166#M5306</link>
      <description>Unfortunately, while I can get it to work in it's own project, when I try it in my real project I still get&lt;BR /&gt;&lt;BR /&gt;"Generic procedure reference has two or more specific procedure with the same type/rank/keyword signature. [SPEVD]"&lt;BR /&gt;&lt;BR /&gt;there a way to track this down? A way to resolve it? Why would it think there are 2 routines witht he same name. Would it work better to use the Fortran 77 version? The test routine in the larger project is:&lt;BR /&gt;&lt;BR /&gt;&lt;P&gt;subroutine test&lt;BR /&gt;use lapack95 &lt;/P&gt;&lt;P&gt;integer, parameter :: N=4 &lt;/P&gt;&lt;P&gt;real, dimension(N*(N+1)/2) :: Ap= (/-4., 2.,-6.,0.5,2.,-8.,0.,0.3,2.,-10./),w(N) &lt;/P&gt;&lt;P&gt;call spevd(Ap,w)&lt;/P&gt;&lt;P&gt;write(*,10)(i,w(i),i=1,N) &lt;/P&gt;&lt;P&gt;10 format(1x,I3,2x,F10.5) &lt;BR /&gt;end&lt;/P&gt;</description>
      <pubDate>Mon, 19 Sep 2011 23:30:42 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828166#M5306</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-19T23:30:42Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828167#M5307</link>
      <description>The subroutine by itself gets compiled with no error. The problem lies, then, in the part of the source code that you have &lt;B&gt;not&lt;/B&gt; shown. &lt;BR /&gt;&lt;BR /&gt;Given that the compiler is telling you that there are conflicting subroutines, in your code, switching to F77 would be pointless: it would shift the detection of the bug from compile time to run time.</description>
      <pubDate>Tue, 20 Sep 2011 01:07:41 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828167#M5307</guid>
      <dc:creator>mecej4</dc:creator>
      <dc:date>2011-09-20T01:07:41Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828168#M5308</link>
      <description>&lt;P&gt;Well, sure, but I don't have any other reference to spevd, and I've added all other unique use statements used eleswhere to the standalone project with no problems, so is it a library conflict?&lt;/P&gt;</description>
      <pubDate>Tue, 20 Sep 2011 05:21:23 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828168#M5308</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-20T05:21:23Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828169#M5309</link>
      <description>&lt;I&gt;&amp;gt; is it a library conflict?&lt;BR /&gt;&lt;BR /&gt;&lt;/I&gt;Cannot say, without seeing the whole source code. The message from the compiler indicates that it is a usage conflict, not a library conflict.</description>
      <pubDate>Tue, 20 Sep 2011 07:36:56 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828169#M5309</guid>
      <dc:creator>mecej4</dc:creator>
      <dc:date>2011-09-20T07:36:56Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828170#M5310</link>
      <description>Do you still have an include 'LAPACK.F90' in some place in your project? Try removing it.</description>
      <pubDate>Tue, 20 Sep 2011 09:58:49 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828170#M5310</guid>
      <dc:creator>Andrew_Smith</dc:creator>
      <dc:date>2011-09-20T09:58:49Z</dc:date>
    </item>
    <item>
      <title>MKL question</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828171#M5311</link>
      <description>I removed everything from the main project except the test routine--got same problem. So I compared the properties of the cut-down main project with the one that works, side-by-side and made all properties match the one that works.&lt;BR /&gt;&lt;BR /&gt;I still got the same error message. &lt;BR /&gt;&lt;BR /&gt;I cleaned the original main project, rebuilt it and it cleared the problem!&lt;BR /&gt;&lt;BR /&gt;I think something got left around from my first abortive attempts...&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 20 Sep 2011 11:54:08 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/MKL-question/m-p/828171#M5311</guid>
      <dc:creator>nvaneck</dc:creator>
      <dc:date>2011-09-20T11:54:08Z</dc:date>
    </item>
  </channel>
</rss>

