Skip to content

Commit

Permalink
Deploying to deploy-doc-site from @ 777cdc8 🚀
Browse files Browse the repository at this point in the history
  • Loading branch information
crtrott committed Dec 2, 2024
1 parent fb77ec9 commit 4c584cd
Show file tree
Hide file tree
Showing 7 changed files with 6 additions and 1 deletion.
Binary file modified .doctrees/environment.pickle
Binary file not shown.
Binary file modified .doctrees/known-issues.doctree
Binary file not shown.
2 changes: 2 additions & 0 deletions _sources/known-issues.rst.txt
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,8 @@ CUDA

Therefore, from version 4.5, the default behavior for Kokkos is to preventively disable `cudaMallocAsync.`

- CUDA 11.0 through 11.2 are not compatible with glibc 2.34 librt stubs. That issue is related to how the CMake package handles linking with librt. For more information please look at issue `#7512 <https://github.com/kokkos/kokkos/issues/7512>`_.


HIP
===
Expand Down
Binary file modified _static/_ext/__pycache__/cppkokkos.cpython-310.pyc
Binary file not shown.
2 changes: 2 additions & 0 deletions _static/known-issues.rst
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,8 @@ CUDA

Therefore, from version 4.5, the default behavior for Kokkos is to preventively disable `cudaMallocAsync.`

- CUDA 11.0 through 11.2 are not compatible with glibc 2.34 librt stubs. That issue is related to how the CMake package handles linking with librt. For more information please look at issue `#7512 <https://github.com/kokkos/kokkos/issues/7512>`_.


HIP
===
Expand Down
1 change: 1 addition & 0 deletions known-issues.html
Original file line number Diff line number Diff line change
Expand Up @@ -548,6 +548,7 @@ <h2>CUDA<a class="headerlink" href="#cuda" title="Permalink to this heading">#</
The user should set up the default memory pool to properly support IPC (<a class="reference external" href="https://developer.nvidia.com/blog/using-cuda-stream-ordered-memory-allocator-part-2/#library_composability">https://developer.nvidia.com/blog/using-cuda-stream-ordered-memory-allocator-part-2/#library_composability</a>).</p>
<p>Therefore, from version 4.5, the default behavior for Kokkos is to preventively disable <cite>cudaMallocAsync.</cite></p>
</li>
<li><p>CUDA 11.0 through 11.2 are not compatible with glibc 2.34 librt stubs. That issue is related to how the CMake package handles linking with librt. For more information please look at issue <a class="reference external" href="https://github.com/kokkos/kokkos/issues/7512">#7512</a>.</p></li>
</ul>
</section>
<section id="hip">
Expand Down
2 changes: 1 addition & 1 deletion searchindex.js

Large diffs are not rendered by default.

0 comments on commit 4c584cd

Please sign in to comment.