Doc: TriSolver (dist): move sorting permutation from CPU to GPU (#1118)

eth-cscs · Jun 21, 2024 · 2969eb1 · 2969eb1
1 parent 3002c02
commit 2969eb1
Show file tree

Hide file tree

Showing 9 changed files with 2,640 additions and 2,738 deletions.
diff --git a/master/eigensolver_2tridiag__solver_2impl_8h_source.html b/master/eigensolver_2tridiag__solver_2impl_8h_source.html
diff --git a/master/merge_8h_source.html b/master/merge_8h_source.html
diff --git a/master/permutations_2general_2impl_8h_source.html b/master/permutations_2general_2impl_8h_source.html
diff --git a/master/permutations_2general_8h.html b/master/permutations_2general_8h.html
@@ -255,7 +255,7 @@ <h2 class="memtitle"><span class="permalink"><a href="#af301d1e323dbafdc8503756c
 <p>Permutes the columns or rows of an input sub-matrix mat_in[i_begin:i_end][i_begin:i_end] into an output sub-matrix mat_out[i_begin:i_end][i_begin:i_end] using the index map of permutations perms[i_begin:i_end].</p>
 <dl class="params"><dt>Parameters</dt><dd>
   <table class="params">
-    <tr><td class="paramname">perms</td><td>is the index map of permutations represented as a tiled column vector. Indices are in the range [0, n) where <code>n</code> is the size of the submatrix (i.e. the indices are local to the submatrix, they are not global). Only tiles whose row tile coords are in the range [i_begin,i_end) are accessed in read-only mode. </td></tr>
+    <tr><td class="paramname">perms</td><td>is the index map of permutations represented as a tiled column vector. Indices are in the range [0, n) where <code>n</code> is the local size of the submatrix (i.e. the indices are local to the submatrix, they are not global). Only tiles whose row tile coords are in the range [i_begin,i_end) are accessed in read-only mode. </td></tr>
   </table>
   </dd>
 </dl>
@@ -270,9 +270,7 @@ <h2 class="memtitle"><span class="permalink"><a href="#af301d1e323dbafdc8503756c
   </table>
   </dd>
 </dl>
-<dl class="section pre"><dt>Precondition</dt><dd><code>mat_in</code> is not distributed </dd>
-<dd>
-<code>mat_in</code> has size (N x N) </dd>
+<dl class="section pre"><dt>Precondition</dt><dd><code>mat_in</code> has size (N x N) </dd>
 <dd>
 <code>mat_in</code> has blocksize (NB x NB) </dd>
 <dd>

diff --git a/master/permutations_2general_8h_source.html b/master/permutations_2general_8h_source.html
diff --git a/master/structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror-members.html b/master/structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror-members.html
@@ -84,9 +84,10 @@
   <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>e2</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
   <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>i2</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
   <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>i5</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>i6</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>z0</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>z1</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>i5b</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>i6</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>z0</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>z1</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html">dlaf::eigensolver::internal::DistWorkSpaceHostMirror&lt; T, D &gt;</a></td><td class="entry"></td></tr>
 </table></div><!-- contents -->
 <!-- start footer part -->
 <hr class="footer"/><address class="footer"><small>

diff --git a/master/structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html b/master/structdlaf_1_1eigensolver_1_1internal_1_1_dist_work_space_host_mirror.html
@@ -103,6 +103,9 @@
 <tr class="memitem:ac433e44a91977fd97b31dd5933a787a3" id="r_ac433e44a91977fd97b31dd5933a787a3"><td class="memItemLeft" align="right" valign="top"><a id="ac433e44a91977fd97b31dd5933a787a3" name="ac433e44a91977fd97b31dd5933a787a3"></a>
 HostMirrorMatrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i5</b></td></tr>
 <tr class="separator:ac433e44a91977fd97b31dd5933a787a3"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr class="memitem:afaf663d88a7c98128c7740d1508ad548" id="r_afaf663d88a7c98128c7740d1508ad548"><td class="memItemLeft" align="right" valign="top"><a id="afaf663d88a7c98128c7740d1508ad548" name="afaf663d88a7c98128c7740d1508ad548"></a>
+HostMirrorMatrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i5b</b></td></tr>
+<tr class="separator:afaf663d88a7c98128c7740d1508ad548"><td class="memSeparator" colspan="2">&#160;</td></tr>
 <tr class="memitem:a599d71f4d24993e8208fa6d72e8ac4ce" id="r_a599d71f4d24993e8208fa6d72e8ac4ce"><td class="memItemLeft" align="right" valign="top"><a id="a599d71f4d24993e8208fa6d72e8ac4ce" name="a599d71f4d24993e8208fa6d72e8ac4ce"></a>
 HostMirrorMatrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i6</b></td></tr>
 <tr class="separator:a599d71f4d24993e8208fa6d72e8ac4ce"><td class="memSeparator" colspan="2">&#160;</td></tr>

diff --git a/master/structdlaf_1_1eigensolver_1_1internal_1_1_work_space-members.html b/master/structdlaf_1_1eigensolver_1_1internal_1_1_work_space-members.html
@@ -85,9 +85,10 @@
   <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>e2</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
   <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>i2</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
   <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>i5</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>i6</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>z0</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
-  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>z1</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>i5b</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>i6</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="even"><td class="entry"><b>z0</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
+  <tr bgcolor="#f0f0f0" class="odd"><td class="entry"><b>z1</b> (defined in <a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a>)</td><td class="entry"><a class="el" href="structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html">dlaf::eigensolver::internal::WorkSpace&lt; T, D &gt;</a></td><td class="entry"></td></tr>
 </table></div><!-- contents -->
 <!-- start footer part -->
 <hr class="footer"/><address class="footer"><small>

diff --git a/master/structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html b/master/structdlaf_1_1eigensolver_1_1internal_1_1_work_space.html
@@ -106,6 +106,9 @@
 <tr class="memitem:a14324e0620063401f04d70331ff142d0" id="r_a14324e0620063401f04d70331ff142d0"><td class="memItemLeft" align="right" valign="top"><a id="a14324e0620063401f04d70331ff142d0" name="a14324e0620063401f04d70331ff142d0"></a>
 Matrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i5</b></td></tr>
 <tr class="separator:a14324e0620063401f04d70331ff142d0"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr class="memitem:a393015149f02ed30f500e6861e3413a8" id="r_a393015149f02ed30f500e6861e3413a8"><td class="memItemLeft" align="right" valign="top"><a id="a393015149f02ed30f500e6861e3413a8" name="a393015149f02ed30f500e6861e3413a8"></a>
+Matrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i5b</b></td></tr>
+<tr class="separator:a393015149f02ed30f500e6861e3413a8"><td class="memSeparator" colspan="2">&#160;</td></tr>
 <tr class="memitem:a09f12fdd541d4dd8127cdb3f55e74049" id="r_a09f12fdd541d4dd8127cdb3f55e74049"><td class="memItemLeft" align="right" valign="top"><a id="a09f12fdd541d4dd8127cdb3f55e74049" name="a09f12fdd541d4dd8127cdb3f55e74049"></a>
 Matrix&lt; SizeType, D &gt;&#160;</td><td class="memItemRight" valign="bottom"><b>i6</b></td></tr>
 <tr class="separator:a09f12fdd541d4dd8127cdb3f55e74049"><td class="memSeparator" colspan="2">&#160;</td></tr>