Add Block-Strided RMA Operations

content/backmatter.tex

@@ -654,12 +654,17 @@ \subsection{Table~\ref{p2psynctypes}: point-to-point synchronization types}
 \chapter{Changes to this Document}\label{sec:changelog}
 
 \section{Version 1.6}
-Major changes in \openshmem[1.6] include the addition of a new
-\FUNC{shmem\_team\_ptr} function.
+Major changes in \openshmem[1.6] include the addition of the new
+\FUNC{shmem\_team\_ptr}, \FUNC{shmem\_ibget}, and \FUNC{shmem\_ibput}
+functions.
 
 The following list describes the specific changes in \openshmem[1.6]:
 \begin{itemize}
 %
+\item Added interleaved block transfer APIs \FUNC{shmem\_ibget} and
+    \FUNC{shmem\_ibput}.
+\ChangelogRef{subsec:shmem_ibget, subsec:shmem_ibput}%
+%
 \item Added a team-based pointer query routine:
   \FUNC{shmem\_team\_ptr}.
 \ChangelogRef{subsec:shmem_team_ptr}%
@@ -686,6 +691,7 @@ \section{Version 1.6}
     2 to enable profiling with profile library defined effects and
     additional arguments.
 \ChangelogRef{subsec:shmem_pcontrol}
+>>>>>>> master
 %
 \end{itemize}
 

content/shmem_ibget.tex

@@ -30,18 +30,18 @@
         The type of \dest{} should match that implied in the SYNOPSIS section.}
     \apiargument{IN}{source}{Symmetric address of the source array data object.
         The type of \source{} should match that implied in the SYNOPSIS section.}
-    \apiargument{IN}{dst}{The stride between consecutive elements of the \dest{}
-        array. The stride must be greater than or equal to \CONST{1} and is
-        scaled by the element size of the \dest{} array. A value of \CONST{1}
+    \apiargument{IN}{dst}{The stride between consecutive blocks of the \dest{}
+        array. The stride must be greater than or equal to \VAR{bsize} and is
+        scaled by the element size of the \dest{} array. A value of \VAR{bsize}
         indicates contiguous data.}
-    \apiargument{IN}{sst}{The stride between consecutive elements of the \source{}
-        array. The stride must be greater than or equal to \CONST{1} and is
-        scaled by the element size of the \source{} array. A value of \CONST{1}
+    \apiargument{IN}{sst}{The stride between consecutive blocks of the \source{}
+        array. The stride must be greater than or equal to \VAR{bsize} and is
+        scaled by the element size of the \source{} array. A value of \VAR{bsize}
         indicates contiguous data.}
     \apiargument{IN}{bsize}{Number of elements per block in the \dest{} and \source{}
         arrays.}
-    \apiargument{IN}{nblocks}{Number of blocks in the \dest{} and \source{}
-        arrays.}
+    \apiargument{IN}{nblocks}{Number of blocks to be copied from the \source{} array
+        to the \dest{} array.}
     \apiargument{IN}{pe}{\ac{PE} number of the remote \ac{PE}.}
 \end{apiarguments}
 
@@ -56,4 +56,10 @@
     None.
 }
 
+\apinotes{
+    The \FUNC{shmem\_ibget} API provides a more general purpose interleaved
+    transfer API than \FUNC{shmem\_iget}. Calling \FUNC{shmem\_ibget} with a
+    block size of 1 is equivalent to the \FUNC{shmem\_iget} API.
+}
+
 \end{apidefinition}

content/shmem_ibput.tex

@@ -30,25 +30,25 @@
         The type of \dest{} should match that implied in the SYNOPSIS section.}
     \apiargument{IN}{source}{Local address of the array containing the data to be copied.
         The type of \source{} should match that implied in the SYNOPSIS section.}
-    \apiargument{IN}{dst}{The stride between consecutive elements of the \dest{}
-        array. The stride must be greater than or equal to \CONST{1} and is
-        scaled by the element size of the \dest{} array. A value of \CONST{1}
+    \apiargument{IN}{dst}{The stride between consecutive blocks of the \dest{}
+        array. The stride must be greater than or equal to \VAR{bsize} and is
+        scaled by the element size of the \dest{} array. A value of \VAR{bsize}
         indicates contiguous data.}
-    \apiargument{IN}{sst}{The stride between consecutive elements of the \source{}
-        array. The stride must be greater than or equal to \CONST{1} and is
-        scaled by the element size of the \source{} array. A value of \CONST{1}
+    \apiargument{IN}{sst}{The stride between consecutive blocks of the \source{}
+        array. The stride must be greater than or equal to \VAR{bsize} and is
+        scaled by the element size of the \source{} array. A value of \VAR{bsize}
         indicates contiguous data.}
     \apiargument{IN}{bsize}{Number of elements per block in the \dest{} and \source{}
         arrays.}
-    \apiargument{IN}{nblocks}{Number of blocks in the \dest{} and \source{}
-        arrays.}
+    \apiargument{IN}{nblocks}{Number of blocks to be copied from the \source{} array
+        to the \dest{} array.}
     \apiargument{IN}{pe}{\ac{PE} number of the remote \ac{PE}.}
 \end{apiarguments}
 
 
 \apidescription{
     The \FUNC{shmem\_ibput} routines provide a method  for  copying strided data
-    elements (specified by \VAR{sst}) of an array from a \source{} array on the
+    blocks (specified by \VAR{sst}) of an array from a \source{} array on the
     local \ac{PE} to locations specified by stride \VAR{dst} on a \dest{} array
     on specified remote \ac{PE}. The routines return when the data has
     been copied out of the \VAR{source} array on the local \ac{PE} but not
@@ -59,4 +59,10 @@
     None.
 }
 
+\apinotes{
+    The \FUNC{shmem\_ibput} API provides a more general purpose interleaved
+    transfer API than \FUNC{shmem\_iput}. Calling \FUNC{shmem\_ibput} with a
+    block size of 1 is equivalent to the \FUNC{shmem\_iput} API.
+}
+
 \end{apidefinition}