--- lamssi_rpi.7 2012-11-08 18:58:00.147564313 +0000 +++ lamssi_rpi.7.new 2012-11-08 19:02:35.760106253 +0000 @@ -1,23 +1,23 @@ .TH lamssi_rpi 7 "March, 2006" "LAM 7.1.2" "LAM SSI RPI OVERVIEW" .SH NAME -LAM SSI RPI \- overview of LAM's RPI SSI modules +LAM_SSI_RPI \- overview of LAM's RPI SSI modules .SH DESCRIPTION The "kind" for RPI SSI modules is "rpi". Specifically, the string "rpi" (without the quotes) should be used to specify which RPI should -be used on the +be used on the .I mpirun command line with the -.I -ssi +.I \-ssi switch. For example: .TP 4 -mpirun -ssi rpi tcp C my_mpi_program +mpirun \-ssi rpi tcp C my_mpi_program Specifies to use the tcp RPI (and to launch a single copy of the executable "foo" on each node). .PP The "rpi" string is also used as a prefix send parameters to specific RPI modules. For example: .TP 4 -mpirun -ssi rpi tcp -ssi rpi_tcp_short 131072 C my_mpi_program +mpirun \-ssi rpi tcp \-ssi rpi_tcp_short 131072 C my_mpi_program Specifies to use the tcp RPI, and to pass in the value of 131072 (128K) as the short message length for TCP messages. See each RPI section below for a full description of parameters that are accepted @@ -33,15 +33,15 @@ .PP The kind for selecting an RPI is "rpi". For example: .TP 4 -mpriun -ssi rpi tcp C my_mpi_program +mpriun \-ssi rpi tcp C my_mpi_program Selects to use the tcp RPI and run a single copy of the foo exectuable -on each node. +on each node. .SH AVAILABLE MODULES As with all SSI modules, it is possible to pass parameters at run time. This section discusses the built-in LAM RPI modules, as well as the run-time parameters that they accept. .PP -In the discussion below, the parameters are discussed in terms of +In the discussion below, the parameters are discussed in terms of .I kind and .IR name . @@ -52,11 +52,11 @@ may be specified as command line arguments to the .I mpirun command with the -.I -ssi -switch, or they may be set in environment variables of the form +.I \-ssi +switch, or they may be set in environment variables of the form .RI LAM_MPI_SSI_ name = value . Note that using the -.I -ssi +.I \-ssi command line switch will take precendence over any environment variables. .PP @@ -80,11 +80,11 @@ gm RPI exists, but has not yet been optimized. It gives significantly better performance than TCP over Myrinet networks, but has not yet been properly tuned and instrumented in LAM. -.PP +.PP That being said, there are several tunable parameters in the gm RPI: .TP 4 rpi_gm_maxport N -If +If .I rpi_gm_port is not specified, LAM will attempt to find an open GM port to use for MPI communications starting with port 1 and ending with the @@ -113,9 +113,9 @@ continual progress within MPI. .TP rpi_gm_cr 1 -Enable checkpoint/restart behavior for gm. This can +Enable checkpoint/restart behavior for gm. This can .I only -be enabled if the gm rpi module was compiled with support for the +be enabled if the gm rpi module was compiled with support for the .I gm_get() function, which is disabled by default. See the LAM Installation and User's Guides for more information on this parameter before you use @@ -151,7 +151,7 @@ receiving side is not expecting them). Long messages use a rendevouz protocol (i.e., a three-way handshake) such that the message is not actually sent until the receiver is expecting it. This value defaults -to 64k. +to 64k. .SS sysv RPI The sysv RPI uses shared memory for communication between MPI processes on the same node, and TCP sockets for communication between @@ -159,7 +159,7 @@ lock the shared memory pools. This RPI is best used when running multiple MPI processes on uniprocessors (or oversubscribed SMPs) because of the blocking / yielding nature of semaphores. -.PP +.PP The sysv RPI has the following tunable parameters: .TP 4 rpi_tcp_short @@ -179,12 +179,12 @@ select() (with a 1ms timeout), depending what LAM's configure script finds at configuration time. This value defaults to 1. .TP -rpi_sysv_shmpoolsize +rpi_sysv_shmpoolsize The size of the shared memory pool that is used for long message transfers. It is allocated once on each node for each MPI parallel job. Specifically, if multiple MPI processes from the same parallel job are spawned on a single node, this pool will only be allocated -once. +once. The configure script will try to determine a default size for the pool if none is explicitly specified (you should always check this to see @@ -195,12 +195,12 @@ rpi_sysv_shmmaxalloc To prevent a single large message transfer from monopolizing the global pool, allocations from the pool are actually restricted to a -maximum of +maximum of .I rpi_sysv_shmmaxalloc bytes each. Even with this restriction, it is possible for the global -pool to temporarily become exhausted. In this case, the transport will +pool to temporarily become exhausted. In this case, the transport will fall back to using the postbox area to transfer the -message. Performance will be degraded, but the application will +message. Performance will be degraded, but the application will progress. The configure script will try to determine a default size for the @@ -217,7 +217,7 @@ processes on a single node is less than or equal to the number of processors because it allows LAM to fully occupy the processor while waiting for a message and never be swapped out. -.PP +.PP The usysv RPI has many of the same tunable parameters as the sysv RPI: .TP 4 @@ -234,7 +234,7 @@ .I rpi_sysv_pollyield in the sysv RPI. .TP -rpi_usysv_shmpoolsize +rpi_usysv_shmpoolsize Same meaning as .I rpi_sysv_shmpoolsize in the sysv RPI.