<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html;
      charset=windows-1252">
  </head>
  <body>
    <p>Some comments ;)</p>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 4/8/20 2:56 PM, Narayanaswamy, Ravi
      via Openmp-dev wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:BCFA27423AEC4E41B6CBD7B6A8C696B6CE5812C6@ORSMSX105.amr.corp.intel.com">
      <pre class="moz-quote-pre" wrap="">Next Meeting :  April 22nd

Opens:

  *   Johannes requested developers to participate in reviews to make code robust.
  *   stream pool

  *   Shilei has posted support for fixing the correctness of device synchronization with streams here <a class="moz-txt-link-freetext" href="https://reviews.llvm.org/D77412">https://reviews.llvm.org/D77412</a>

  *   Async support in target library
     *   To support async,  the current _tgt_ API for plugin was changed to add an additional argument. <a class="moz-txt-link-freetext" href="https://reviews.llvm.org/D77005">https://reviews.llvm.org/D77005</a>

       We had agreed previous any exposed API should not change but new APIs will be added and the old API can internally call the new  API

  *   AMD plugin
     *   Nice if AMD plugin can be committed to llvm.org.  This will enable testing the rewrite of device rtl into common/device specific partition</pre>
    </blockquote>
    <p>Especially the plugin can than be split into common and device
      specific.</p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:BCFA27423AEC4E41B6CBD7B6A8C696B6CE5812C6@ORSMSX105.amr.corp.intel.com">
      <pre class="moz-quote-pre" wrap="">
     *   Greg said they are working on upstreaming the code.  This code is currently available for all to access in their github
  *   NEC SX-Aurora plugin
     *   New plugin for NEC SX-Aurora is being developed.  Can be found here <a class="moz-txt-link-freetext" href="https://reviews.llvm.org/D76843">https://reviews.llvm.org/D76843</a>
  *   The activity of partitioning device rtl was on hold as Ron was spending quality time as proud dad.  Congratulations.</pre>
    </blockquote>
    <p>Jon, not Ron ;)<br>
    </p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:BCFA27423AEC4E41B6CBD7B6A8C696B6CE5812C6@ORSMSX105.amr.corp.intel.com">
      <pre class="moz-quote-pre" wrap="">
     *   Cuda requirement has been removed</pre>
    </blockquote>
    <p>Not yet. We have a plan to do so. Work is needed though.<br>
    </p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:BCFA27423AEC4E41B6CBD7B6A8C696B6CE5812C6@ORSMSX105.amr.corp.intel.com">
      <pre class="moz-quote-pre" wrap="">
     *   Shared data is supported using declare target and device code using target pragmas.</pre>
    </blockquote>
    <p>That's the plan I mention above and in the RFC, yes.</p>
    <p><br>
    </p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:BCFA27423AEC4E41B6CBD7B6A8C696B6CE5812C6@ORSMSX105.amr.corp.intel.com">
      <pre class="moz-quote-pre" wrap="">



Development Activity:-

  *   Unshackled threads
     *   Shilei presented the design of a prototype to add unshackled threads in openmp runtime to support async target tasks. The presentation is attached

        There are still issue with how to synchronize with other task and with task wait.

         Alex mentioned we already have support for detach task for use with target task.

         We need to create examples and check if they are handled in both the schemes. (unshackled or existing implementation)

  *   Fat Static Library
     *   George has uploaded the support in offload bundler here <a class="moz-txt-link-freetext" href="https://reviews.llvm.org/D74262">https://reviews.llvm.org/D74262</a>
     *   He will be uploading the driver changes.
  *   math support in target regions
     *   Support has been added.
  *   omp loop, omp metadirective, omp interop
     *   Johannes has prototypes implemented omp loop and metadirectives
     *   Was asking if anyone want to participate in omp interop.
  *   Declare variants
     *   Johannes has implemented support for begin/end declare variants.

Roll Call :

Company

Attendees

8-Jan

29-Jan

12-Feb

26-Feb

11-Mar

25-Mar

8-Apr

AMD



















Greg Rodgers



x









x



Ashwin Aji

















Jan Sjodin



x













Ron Lieberman



x









x



sameer Sahasrabuddhe

















Jonathan Chesterfield

x

x

x

x

x

x





Andrey Kasaurov















ANL

Hal Finkel











x

x



Michael Kruse





x

x

x

x

x



Johannes Doerfert

x

x

x







x



Ye Luo













x

ARM

Andrzej Warzynski



x













Francesco Petrogalli



x

x

x



x



IBM

Alexandre Eichenberger

















Carlo Bertolli

















Kelvin Li

x





x

x







Doru

















Alexey Bataev

x

x

x

x

x

x

x

Intel

Andrey Churbanov











x

x



Ravi Narayanaswamy

x

x

x

x

x

x

x



Serguei Dmitriev

















Rajiv Deodhar

















Lorri Menard

















Terry Wilmarth

















Rao, Prem

















Hansang Bae

















Xinmin Tian

















Sirram Aananthakrishnan





x

x









Jeff Hammond











x





George Rokos

x

x

x

x

x

x

x



Alex Duran













x



Micheal Klemm







x







HPE

Deepak Eachempati



x

x

x



x

x

Micron

John Leidel















Nvidia

James Beyer















ORNL

Graham Lopez

















Joel Denny



x



x

x

x

x



Geoffroy Vallee

















Fady Ghanim

x

x

x

x

x







Oscar Hernandez















SBU/BNL

Lingda Li





x











Jose Monlsave

















Vivek Kale



x













Shilei Tian













x

TU Darmstadt

Lukas Sommer

x



x

x





x

BrookHavens

Abid Malik

x













Universidata EAFIT Colombia

Hamilton Tobon Mosquera













x



.........................................................................................................................................
[Blue Jeans]<a class="moz-txt-link-rfc2396E" href="https://bluejeans.com"><https://bluejeans.com></a>


Johannes Doerfert<a class="moz-txt-link-rfc2396E" href="mailto:jdoerfert@anl.gov"><mailto:jdoerfert@anl.gov></a> has invited you to a meeting.




Join Meeting <a class="moz-txt-link-rfc2396E" href="https://bluejeans.com/544112769?src=htmlEmail"><https://bluejeans.com/544112769?src=htmlEmail></a>




(Join from computer or phone)




  ________________________________



Phone Dial-in
+1.408.317.9254<a class="moz-txt-link-rfc2396E" href="tel:+1.408.317.9254"><tel:+1.408.317.9254></a> (US (San Jose))
+1.866.226.4650<a class="moz-txt-link-rfc2396E" href="tel:+1.866.226.4650"><tel:+1.866.226.4650></a> (US Toll Free)
(Global Numbers<a class="moz-txt-link-rfc2396E" href="https://www.bluejeans.com/premium-numbers"><https://www.bluejeans.com/premium-numbers></a>)

Meeting ID: 544 112 769


Room System
199.48.152.152 or bjn.vc

Meeting ID: 544 112 769


  ________________________________



Description:
Bi-weekly meeting to discuss OpenMP support in LLVM.


  ________________________________



Want to test your video connection?
<a class="moz-txt-link-freetext" href="https://bluejeans.com/111">https://bluejeans.com/111</a><a class="moz-txt-link-rfc2396E" href="https://bluejeans.com/111?ll=en"><https://bluejeans.com/111?ll=en></a>







</pre>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <pre class="moz-quote-pre" wrap="">_______________________________________________
Openmp-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Openmp-dev@lists.llvm.org">Openmp-dev@lists.llvm.org</a>
<a class="moz-txt-link-freetext" href="https://lists.llvm.org/cgi-bin/mailman/listinfo/openmp-dev">https://lists.llvm.org/cgi-bin/mailman/listinfo/openmp-dev</a>
</pre>
    </blockquote>
  </body>
</html>