--- manual/s_software/text/sarch.tex 2006/04/05 01:12:02 1.24 +++ manual/s_software/text/sarch.tex 2010/08/30 23:09:22 1.26 @@ -1,12 +1,12 @@ -% $Header: /home/ubuntu/mnt/e9_copy/manual/s_software/text/sarch.tex,v 1.24 2006/04/05 01:12:02 jmc Exp $ +% $Header: /home/ubuntu/mnt/e9_copy/manual/s_software/text/sarch.tex,v 1.26 2010/08/30 23:09:22 jmc Exp $ This chapter focuses on describing the {\bf WRAPPER} environment within which both the core numerics and the pluggable packages operate. The description presented here is intended to be a detailed exposition and contains significant background material, as well as advanced details on working with the WRAPPER. The tutorial sections -of this manual (see sections \ref{sect:tutorials} and -\ref{sect:tutorialIII}) contain more succinct, step-by-step +of this manual (see sections \ref{sec:modelExamples} and +\ref{sec:tutorialIII}) contain more succinct, step-by-step instructions on running basic numerical experiments, of varous types, both sequentially and in parallel. For many projects simply starting from an example code and adapting it to suit a particular situation @@ -69,7 +69,7 @@ \begin{figure} \begin{center} -\resizebox{!}{2.5in}{\includegraphics{part4/mitgcm_goals.eps}} +\resizebox{!}{2.5in}{\includegraphics{s_software/figs/mitgcm_goals.eps}} \end{center} \caption{ The MITgcm architecture is designed to allow simulation of a wide range of physical problems on a wide range of hardware. The @@ -93,7 +93,7 @@ ``fit'' within the WRAPPER infrastructure. Writing code to ``fit'' within the WRAPPER means that coding has to follow certain, relatively straightforward, rules and conventions (these are discussed further in -section \ref{sect:specifying_a_decomposition}). +section \ref{sec:specifying_a_decomposition}). The approach taken by the WRAPPER is illustrated in figure \ref{fig:fit_in_wrapper} which shows how the WRAPPER serves to @@ -104,7 +104,7 @@ \begin{figure} \begin{center} -\resizebox{!}{4.5in}{\includegraphics{part4/fit_in_wrapper.eps}} +\resizebox{!}{4.5in}{\includegraphics{s_software/figs/fit_in_wrapper.eps}} \end{center} \caption{ Numerical code is written to fit within a software support @@ -118,7 +118,7 @@ \end{figure} \subsection{Target hardware} -\label{sect:target_hardware} +\label{sec:target_hardware} The WRAPPER is designed to target as broad as possible a range of computer systems. The original development of the WRAPPER took place @@ -136,11 +136,11 @@ IBM SP systems). In all cases numerical code, operating within the WRAPPER, performs and scales very competitively with equivalent numerical code that has been modified to contain native optimizations -for a particular system \ref{ref hoe and hill, ecmwf}. +for a particular system \cite{hoe-hill:99}. \subsection{Supporting hardware neutrality} -The different systems listed in section \ref{sect:target_hardware} can +The different systems listed in section \ref{sec:target_hardware} can be categorized in many different ways. For example, one common distinction is between shared-memory parallel systems (SMP and PVP) and distributed memory parallel systems (for example x86 clusters and @@ -165,7 +165,7 @@ scientific computing community. \subsection{Machine model parallelism} -\label{sect:domain_decomposition} +\label{sec:domain_decomposition} \begin{rawhtml} \end{rawhtml} @@ -211,7 +211,7 @@ \begin{figure} \begin{center} \resizebox{5in}{!}{ - \includegraphics{part4/domain_decomp.eps} + \includegraphics{s_software/figs/domain_decomp.eps} } \end{center} \caption{ The WRAPPER provides support for one and two dimensional @@ -240,13 +240,13 @@ domain it owns. Periodically processors will make calls to WRAPPER functions to communicate data between tiles, in order to keep the overlap regions up to date (see section -\ref{sect:communication_primitives}). The WRAPPER functions can use a +\ref{sec:communication_primitives}). The WRAPPER functions can use a variety of different mechanisms to communicate data between tiles. \begin{figure} \begin{center} \resizebox{5in}{!}{ - \includegraphics{part4/tiled-world.eps} + \includegraphics{s_software/figs/tiled-world.eps} } \end{center} \caption{ A global grid subdivided into tiles. @@ -280,7 +280,7 @@ call a function in the API of the communication library to communicate data from a tile that it owns to a tile that another CPU owns. By default the WRAPPER binds to the MPI communication library - \ref{MPI} for this style of communication. + \cite{MPI-std-20} for this style of communication. \end{itemize} The WRAPPER assumes that communication will use one of these two styles @@ -329,7 +329,7 @@ \end{figure} \subsection{Shared memory communication} -\label{sect:shared_memory_communication} +\label{sec:shared_memory_communication} Under shared communication independent CPUs are operating on the exact same global address space at the application level. This means @@ -356,7 +356,7 @@ appropriately. \subsubsection{Memory consistency} -\label{sect:memory_consistency} +\label{sec:memory_consistency} When using shared memory communication between multiple processors the WRAPPER level shields user applications from certain counter-intuitive @@ -382,7 +382,7 @@ particular platform. \subsubsection{Cache effects and false sharing} -\label{sect:cache_effects_and_false_sharing} +\label{sec:cache_effects_and_false_sharing} Shared-memory machines often have local to processor memory caches which contain mirrored copies of main memory. Automatic cache-coherence @@ -402,20 +402,24 @@ the standard mechanism for supporting shared memory that the WRAPPER utilizes. Configuring and launching code to run in multi-threaded mode on specific platforms is discussed in section -\ref{sect:multi-threaded-execution}. However, on many systems, +\ref{sec:multi_threaded_execution}. However, on many systems, potentially very efficient mechanisms for using shared memory communication between multiple processes (in contrast to multiple threads within a single process) also exist. In most cases this works by making a limited region of memory shared between processes. The -MMAP \ref{magicgarden} and IPC \ref{magicgarden} facilities in UNIX +MMAP %\ref{magicgarden} +and IPC %\ref{magicgarden} +facilities in UNIX systems provide this capability as do vendor specific tools like LAPI -\ref{IBMLAPI} and IMC \ref{Memorychannel}. Extensions exist for the +%\ref{IBMLAPI} +and IMC. %\ref{Memorychannel}. +Extensions exist for the WRAPPER that allow these mechanisms to be used for shared memory communication. However, these mechanisms are not distributed with the default WRAPPER sources, because of their proprietary nature. \subsection{Distributed memory communication} -\label{sect:distributed_memory_communication} +\label{sec:distributed_memory_communication} Many parallel systems are not constructed in a way where it is possible or practical for an application to use shared memory for communication. For example cluster systems consist of individual @@ -426,16 +430,16 @@ communication library used is MPI \cite{MPI-std-20}. However, it is relatively straightforward to implement bindings to optimized platform specific communication libraries. For example the work described in -\ref{hoe-hill:99} substituted standard MPI communication for a highly +\cite{hoe-hill:99} substituted standard MPI communication for a highly optimized library. \subsection{Communication primitives} -\label{sect:communication_primitives} +\label{sec:communication_primitives} \begin{figure} \begin{center} \resizebox{5in}{!}{ - \includegraphics{part4/comm-primm.eps} + \includegraphics{s_software/figs/comm-primm.eps} } \end{center} \caption{Three performance critical parallel primitives are provided @@ -518,7 +522,7 @@ \begin{figure} \begin{center} \resizebox{5in}{!}{ - \includegraphics{part4/tiling_detail.eps} + \includegraphics{s_software/figs/tiling_detail.eps} } \end{center} \caption{The tiling strategy that the WRAPPER supports allows tiles @@ -578,16 +582,16 @@ computing CPUs. \end{enumerate} This section describes the details of each of these operations. -Section \ref{sect:specifying_a_decomposition} explains how the way in +Section \ref{sec:specifying_a_decomposition} explains how the way in which a domain is decomposed (or composed) is expressed. Section -\ref{sect:starting_a_code} describes practical details of running +\ref{sec:starting_the_code} describes practical details of running codes in various different parallel modes on contemporary computer -systems. Section \ref{sect:controlling_communication} explains the +systems. Section \ref{sec:controlling_communication} explains the internal information that the WRAPPER uses to control how information is communicated between tiles. \subsection{Specifying a domain decomposition} -\label{sect:specifying_a_decomposition} +\label{sec:specifying_a_decomposition} At its heart much of the WRAPPER works only in terms of a collection of tiles which are interconnected to each other. This is also true of application @@ -624,7 +628,7 @@ \begin{figure} \begin{center} \resizebox{5in}{!}{ - \includegraphics{part4/size_h.eps} + \includegraphics{s_software/figs/size_h.eps} } \end{center} \caption{ The three level domain decomposition hierarchy employed by the @@ -639,7 +643,7 @@ dimensions of {\em sNx} and {\em sNy}. If, when the code is executed, these tiles are allocated to different threads of a process that are then bound to different physical processors ( see the multi-threaded -execution discussion in section \ref{sect:starting_the_code} ) then +execution discussion in section \ref{sec:starting_the_code} ) then computation will be performed concurrently on each tile. However, it is also possible to run the same decomposition within a process running a single thread on a single processor. In this case the tiles will be computed over sequentially. @@ -836,14 +840,14 @@ This set of values can be used for a cube sphere calculation. Each tile of size $32 \times 32$ represents a face of the cube. Initializing the tile connectivity correctly ( see section -\ref{sect:cube_sphere_communication}. allows the rotations associated with +\ref{sec:cube_sphere_communication}. allows the rotations associated with moving between the six cube faces to be embedded within the tile-tile communication code. \end{enumerate} \subsection{Starting the code} -\label{sect:starting_the_code} +\label{sec:starting_the_code} When code is started under the WRAPPER, execution begins in a main routine {\em eesupp/src/main.F} that is owned by the WRAPPER. Control is transferred to the application through a routine called {\em THE\_MODEL\_MAIN()} @@ -890,13 +894,13 @@ \end{figure} \subsubsection{Multi-threaded execution} -\label{sect:multi-threaded-execution} +\label{sec:multi_threaded_execution} Prior to transferring control to the procedure {\em THE\_MODEL\_MAIN()} the WRAPPER may cause several coarse grain threads to be initialized. The routine {\em THE\_MODEL\_MAIN()} is called once for each thread and is passed a single stack argument which is the thread number, stored in the variable {\em myThid}. In addition to specifying a decomposition with -multiple tiles per process ( see section \ref{sect:specifying_a_decomposition}) +multiple tiles per process ( see section \ref{sec:specifying_a_decomposition}) configuring and starting a code to run using multiple threads requires the following steps.\\ @@ -978,7 +982,7 @@ } \\ \subsubsection{Multi-process execution} -\label{sect:multi-process-execution} +\label{sec:multi_process_execution} Despite its appealing programming model, multi-threaded execution remains less common than multi-process execution. One major reason for @@ -990,7 +994,7 @@ Multi-process execution is more ubiquitous. In order to run code in a multi-process configuration a decomposition specification (see section -\ref{sect:specifying_a_decomposition}) is given (in which the at least +\ref{sec:specifying_a_decomposition}) is given (in which the at least one of the parameters {\em nPx} or {\em nPy} will be greater than one) and then, as for multi-threaded operation, appropriate compile time and run time steps must be taken. @@ -1010,7 +1014,7 @@ ALLOW\_USE\_MPI} and {\em ALWAYS\_USE\_MPI} flags in the {\em CPP\_EEOPTIONS.h} file.) More detailed information about the use of {\em genmake2} for specifying -local compiler flags is located in section \ref{sect:genmake}.\\ +local compiler flags is located in section \ref{sec:genmake}.\\ \fbox{ @@ -1114,6 +1118,7 @@ \subsection{Controlling communication} +\label{sec:controlling_communication} The WRAPPER maintains internal information that is used for communication operations and that can be customized for different platforms. This section describes the information that is held and used. @@ -1136,10 +1141,10 @@ a particular face. A value of {\em COMM\_MSG} is used to indicate that some form of distributed memory communication is required to communicate between these tile faces (see section - \ref{sect:distributed_memory_communication}). A value of {\em + \ref{sec:distributed_memory_communication}). A value of {\em COMM\_PUT} or {\em COMM\_GET} is used to indicate forms of shared memory communication (see section - \ref{sect:shared_memory_communication}). The {\em COMM\_PUT} value + \ref{sec:shared_memory_communication}). The {\em COMM\_PUT} value indicates that a CPU should communicate by writing to data structures owned by another CPU. A {\em COMM\_GET} value indicates that a CPU should communicate by reading from data structures owned @@ -1196,7 +1201,7 @@ the file {\em eedata}. If the value of {\em nThreads} is inconsistent with the number of threads requested from the operating system (for example by using an environment variable as described in - section \ref{sect:multi_threaded_execution}) then usually an error + section \ref{sec:multi_threaded_execution}) then usually an error will be reported by the routine {\em CHECK\_THREADS}. \fbox{ @@ -1213,7 +1218,7 @@ } \item {\bf memsync flags} - As discussed in section \ref{sect:memory_consistency}, a low-level + As discussed in section \ref{sec:memory_consistency}, a low-level system function may be need to force memory consistency on some shared memory systems. The routine {\em MEMSYNC()} is used for this purpose. This routine should not need modifying and the information @@ -1239,7 +1244,7 @@ \end{verbatim} \item {\bf Cache line size} - As discussed in section \ref{sect:cache_effects_and_false_sharing}, + As discussed in section \ref{sec:cache_effects_and_false_sharing}, milti-threaded codes explicitly avoid penalties associated with excessive coherence traffic on an SMP system. To do this the shared memory data structures used by the {\em GLOBAL\_SUM}, {\em @@ -1269,7 +1274,7 @@ CPP\_EEMACROS.h}. The \_GSUM macro is a performance critical operation, especially for large processor count, small tile size configurations. The custom communication example discussed in - section \ref{sect:jam_example} shows how the macro is used to invoke + section \ref{sec:jam_example} shows how the macro is used to invoke a custom global sum routine for a specific set of hardware. \item {\bf \_EXCH} @@ -1282,7 +1287,7 @@ the header file {\em CPP\_EEMACROS.h}. As with \_GSUM, the \_EXCH operation plays a crucial role in scaling to small tile, large logical and physical processor count configurations. The example in - section \ref{sect:jam_example} discusses defining an optimized and + section \ref{sec:jam_example} discusses defining an optimized and specialized form on the \_EXCH operation. The \_EXCH operation is also central to supporting grids such as the @@ -1323,7 +1328,7 @@ if this mechanism is unavailable then the work arrays can be extended with dimensions using the tile dimensioning scheme of {\em nSx} and {\em nSy} (as described in section - \ref{sect:specifying_a_decomposition}). However, if the + \ref{sec:specifying_a_decomposition}). However, if the configuration being specified involves many more tiles than OS threads then it can save memory resources to reduce the variable {\em MAX\_NO\_THREADS} to be equal to the actual number of threads @@ -1381,7 +1386,7 @@ how it can be used to adapt to new griding approaches. \subsubsection{JAM example} -\label{sect:jam_example} +\label{sec:jam_example} On some platforms a big performance boost can be obtained by binding the communication routines {\em \_EXCH} and {\em \_GSUM} to specialized native libraries (for example, the shmem library on CRAY @@ -1405,7 +1410,7 @@ pattern. \subsubsection{Cube sphere communication} -\label{sect:cube_sphere_communication} +\label{sec:cube_sphere_communication} Actual {\em \_EXCH} routine code is generated automatically from a series of template files, for example {\em exch\_rx.template}. This is done to allow a large number of variations on the exchange process @@ -1440,10 +1445,10 @@ Fitting together the WRAPPER elements, package elements and MITgcm core equation elements of the source code produces calling -sequence shown in section \ref{sect:calling_sequence} +sequence shown in section \ref{sec:calling_sequence} \subsection{Annotated call tree for MITgcm and WRAPPER} -\label{sect:calling_sequence} +\label{sec:calling_sequence} WRAPPER layer.