From 70ed11192bd9b46fc8e1438599729d896875802a Mon Sep 17 00:00:00 2001 From: Jeff Squyres Date: Mon, 11 Sep 2023 21:53:07 -0400 Subject: [PATCH] Restore the help-schizo-ompi.txt file Snip the [usage] section and simply refer to the mpirun(1) man page / HTML help. Signed-off-by: Jeff Squyres --- src/mca/schizo/ompi/Makefile.am | 4 + src/mca/schizo/ompi/help-schizo-ompi.txt | 471 +++++++++++++++++++++++ 2 files changed, 475 insertions(+) create mode 100644 src/mca/schizo/ompi/help-schizo-ompi.txt diff --git a/src/mca/schizo/ompi/Makefile.am b/src/mca/schizo/ompi/Makefile.am index 248f77f762..62945dbe40 100644 --- a/src/mca/schizo/ompi/Makefile.am +++ b/src/mca/schizo/ompi/Makefile.am @@ -24,6 +24,10 @@ rstdir = $(prtedatadir)/rst/schizo-ompi-rst-content dist_rst_DATA = \ schizo-ompi-cli.rst +# Install the help file +dist_prtedata_DATA = \ + help-schizo-ompi.txt + sources = \ schizo_ompi_component.c \ schizo_ompi.h \ diff --git a/src/mca/schizo/ompi/help-schizo-ompi.txt b/src/mca/schizo/ompi/help-schizo-ompi.txt new file mode 100644 index 0000000000..0d67cebfb3 --- /dev/null +++ b/src/mca/schizo/ompi/help-schizo-ompi.txt @@ -0,0 +1,471 @@ +# -*- text -*- +# +# Copyright (c) 2021-2022 Nanook Consulting. All rights reserved. +# Copyright (c) 2022 Cisco Systems, Inc. All rights reserved. +# Copyright (c) 2022 IBM Corporation. All rights reserved. +# Copyright (c) 2023 Jeffrey M. Squyres. All rights reserved. +# $COPYRIGHT$ +# +# Additional copyrights may follow +# +# $HEADER$ +# +# +[version] +%s (%s) %s + +%s +# +[usage] +%s (%s) %s + +Usage: %s [OPTION]... + +See the mpirun(1) man page or HTML help for a detailed list of command +line options that are available. + +%s +# +[dvm] +Utilize an existing persistent DVM instead of instantiating an +independent runtime infrastructure. This is provided as a convenience +option for those wanting to embed the "mpirun" command in a script that +can be optionally used to run either independently or under a persistent DVM. + +#include#help-schizo-cli#dvm +# +[prtemca] +Syntax: --prtemca +where arg0 is the parameter name and arg1 is the parameter value + +Pass a PRRTE MCA parameter +# +[pmixmca] +Syntax: --pmixmca +where arg0 is the parameter name and arg1 is the parameter value + +Pass a PMIx MCA parameter +# +[tune] +File(s) containing PRRTE and PMIx MCA params for tuning DVM and/or application operations. +Parameters in the file will be treated as _generic_ parameters and subject to the +translation rules/uncertainties. See "--help mca" for more information. + +Syntax in the file is: + +param=value + +with one parameter and its associated value per line. Empty lines and lines beginning +with the '#' character are ignored. + +There must not be any whitespace characters before or after the '=' in the param/value +specification. +# +[daemonize] +Daemonize the DVM daemons into the background +# +[set-sid] +Direct the DVM daemons to separate from the current session +# +[report-pid] +Display the PID a tool can use to connect to the mpirun process on stdout (-), stderr (+), or write it to +the specified file if anything else is specified. +# +[report-uri] +Display the URI that a tool can use to connect to the mpirun process on stdout (-), stderr (+), or write +it to a file if anything else is specified. +# +[test-suicide] +Suicide instead of clean abort after delay +# +[default-hostfile] +This option specifies a default set of nodes that are to be used. + +If neither the --hostfile or --host options are specified, this will be the set of nodes used. + +The --hostfile and --host options can be used to select a subset of nodes from the default set to be used. + +If both the --hostfile and --host options are specified, then the --hostfile option specifies the +initial subset of nodes used from the default-hostfile set, and the --host option specifies what subset of +nodes from that second subset will be used as the final subset of nodes used. + +If a resource manager is used, the resource manager determines the initial set of nodes. Then the +specifications for the --default-hostfile, --hostfile, and --host optons, in that order, are used to determine +the final subset of nodes that are used. +# +[singleton] +ID of the singleton process that started us +# +[keepalive] +Pipe to monitor - DVM will terminate upon closure of this pipe. +# +[launch-agent] +Name of daemon executable used to start processes on remote nodes (default: prted) +# +[max-vm-size] +Maximum number of daemons to start +# +[debug-daemons] +Debug daemon output enabled +# +[debug-daemons-file] +Enable debugging of any PRTE daemons used by this application, storing output in files +# +[leave-session-attached] +Do not discard stdout/stderr of remote PRTE daemons, mostly used when debugging the +application launcher. +# +[tmpdir] +Set the root for the session directory tree +# +[prefix] +This option specifies the path to be used as a prefix when resolving the path to Open MPI +executables. +# +[noprefix] +Ignores the path specified if Open MPI was configured with the --prefix option when built +from source code. +# +[forward-signals] +Comma-delimited list of additional signals (names or integers) to forward to application +processes ["none" => forward nothing]. Signals provided by default include SIGTSTP, +SIGUSR1, SIGUSR2, SIGABRT, SIGALRM, and SIGCONT +# +[allow-run-as-root] +Allow execution as root (STRONGLY DISCOURAGED) +# +[report-child-jobs-separately] +Return the exit status of the primary job only +# +[timeout] +Timeout the job if execution time exceeds the specified number of seconds +# +[report-state-on-timeout] +Report all job and process states upon timeout +# +[get-stack-traces] +Get stack traces of all application procs on timeout +# +[spawn-timeout] +Timeout the job if spawn takes more than the specified number of seconds +# +[np] +Specify number of application processes to be started +# +[n] +Specify number of application processes to be started +# +[N] +Specify number of application processes per node to be started +# +[app] +Provide an appfile; ignore all other command line options +# +[xterm] +Create a new xterm window and display output from the specified ranks there. +Ranks are specified as a comma-delimited list of ranges - e.g., "1,3-6,9", +or as "all"." +# +[stop-on-exec] +If supported, stop each process at start of execution +# +[stop-in-init] +Include the PMIX_DEBUG_STOP_IN_INIT attribute in the application's +job info directing that the ranks stop in PMIx_Init pending +release. +# +[stop-in-app] +Include the PMIX_DEBUG_STOP_IN_APP attribute in the application's +job info directing that the ranks stop at an application-determined +point pending release. +# +[x] +Export an environment variable, optionally specifying a value (e.g., "-x foo" exports the +environment variable foo and takes its value from the current environment; "-x foo=bar" +exports the environment variable name foo and sets its value to "bar" in the started +processes; "-x foo*" exports all current environmental variables starting with "foo") +# +[wdir] +Set the working directory of the started processes +# +[wd] +Synonym for --wdir +# +[set-cwd-to-session-dir] +Set the working directory of the started processes to their session directory +# +[path] +PATH to be used to look for executables to start processes +# +[show-progress] +Output a brief periodic report on launch progress +# +[pset] +User-specified name assigned to the processes in their given application +# +[hostfile] +Provide a hostfile +# +[machinefile] +Provide a hostfile (synonym for "hostfile") +# +[host] +Comma-separated list of hosts to invoke processes on +# +[personality] +Specify the personality to be used +# +[preload-files] +Syntax: --preload-files + +Preload the comma separated list of files to the remote machines current +working directory before starting the remote process. +# +[preload-binary] +Syntax: --preload-binary + +Preload the binary on the remote machine before starting the +remote process. +# +[output] +Comma-delimited list of case-insensitive options that control how output is generated. +Allowed values: + tag Mark each output line with the [job,rank] of the + process that generated it + timestamp Prefix each output line with a datetime stamp. Note + that the timestamp will be the time when the line + is output by the DVM - not the time when the source + output it + xml Format all output in XML + merge[-stderr-to-stdout] Merge stderr into stdout + dir=DIRNAME Redirect output from application processes into files of + the form DIRNAME/job/rank/std[out,err,diag]. + file=filename Redirect output from application processes into files of + the form filename.rank. + +In both the "dir" and "file" cases, the provided name will be converted to an absolute path. +Supported qualifiers include NOCOPY (i.e., output shall go only into the files - do not copy +the output to the stdout/err streams). + +Only enough characters are required to uniquely identify the option. +For example, TI is sufficient to identify TIMESTAMP option, while T cannot be used +since it does not uniquely identify an option. +# +[stream-buffering] +Adjust buffering for stdout/stderr [0 unbuffered] [1 line buffered] [2 fully buffered] +Can also be set via the MCA parameter: --mca ompi_stream_buffering VALUE +# +[stdin] +Specify procs to receive stdin [rank, "all", "none"] (default: 0, indicating rank 0) +# +[map-by] +Mapping Policy for job: + slot + hwthread + core (default: np <= 2) + l1cache + l2cache + l3cache + numa (default: np > 2) + package + node + seq + dist + ppr + rankfile +with supported colon-delimited qualifiers: + PE=y (for multiple cpus/proc) + SPAN + OVERSUBSCRIBE + NOOVERSUBSCRIBE + NOLOCAL + HWTCPUS + CORECPUS + DEVICE(for dist policy) + INHERIT + NOINHERIT + PE-LIST=a,b (comma-delimited ranges of cpus to use for this job) + FILE= for seq and rankfile options +# +[rank-by] +Ranking Policy for job: + slot (default: np <= 2) + hwthread + core + l1cache + l2cache + l3cache + numa (default: np > 2) + package + node +with supported colon-delimited qualifiers: + SPAN + FILL +# +[bind-to] +Binding Policy for job: + none (default: oversubscribed) + hwthread + core (default: np <= 2) + l1cache + l2cache + l3cache + numa (default: np > 2) + package +with supported colon-delimited qualifiers: + overload-allowed + if-supported +# +[rankfile] +Name of file to specify explicit task mapping +# +[display] +Comma-delimited list of case-insensitive options for displaying information about +the allocation and job. +Allowed values: + allocation + bind + map + map-devel + topo +Only enough characters are required to uniquely identify the option. +For example, A is sufficient to identify ALLOCATION option, while MAP cannot be used +to specify the MAP-DEVEL option since there is also a MAP option. +# +[do-not-launch] +Perform all necessary operations to prepare to launch the application, but do not actually +launch it (usually used to test mapping patterns) +# +[enable-recovery] +Enable recovery from process failure [Default = disabled] +# +[max-restarts] +Max number of times to restart a failed process +# +[disable-recovery] +Disable recovery (resets all recovery options to off) +# +[continuous] +Job is to run until explicitly terminated +[openmpi-install-path-not-found] +An internal Open MPI environment variable needed for Java support was not set. +Open MPI 5.0.0 or newer is required for this support. +# +[output-proctable] +Print the complete proctable to stdout after launch. This option accepts an optional argument, +which must be + to print to stdout, - to print to stderr, or a valid filename to print to +a file. For instance --output-proctable=+ + +If the optional argument is omitted, then the table is printed to stdout. + +The table includes rank, hostname, executable name and pid. +# +# DEPRECATED OPTIONS +# +[mca] +Syntax: --mca +where arg0 is the parameter name and arg1 is the parameter value + +Pass generic MCA parameters - i.e., parameters whose project affiliation +must be determined by PRRTE based on matching the name of the parameter with defined values +from various projects that PRRTE knows about. + +DEPRECATED: This translation can be incomplete (e.g., if known project adds or changes +parameters) - thus, it is strongly recommended that users use project-specific parameters +such as "prtemca" or "pmixmca". +# +[gmca] +Syntax: --gmca +where arg0 is the parameter name and arg1 is the parameter value. The "g" prefix +indicates that this parameter is to be applied to _all_ application contexts and +not just the one in which the directive appears. + +Pass generic MCA parameters - i.e., parameters whose project affiliation +must be determined by PRRTE based on matching the name of the parameter with defined values +from various projects that PRRTE knows about. This translation can be incomplete (e.g., if +a known project adds or changes parameters) - thus, it is strongly recommended that users +use project-specific parameters such as "prtemca" or "pmixmca". + +DEPRECATED: This translation can be incomplete (e.g., if known project adds or changes +parameters) - thus, it is strongly recommended that users use project-specific parameters +such as "gprtemca" or "gpmixmca". +# +[xml] +Provide all output in XML format + +DEPRECATED: please see "--help output" for details +# +[tag-output] +Tag all output with [job,rank] + +DEPRECATED: please see "--help output" for details +# +[timestamp-output] +Timestamp all application process output + +DEPRECATED: please see "--help output" for details +# +[output-directory] +Redirect output from application processes into filename/job/rank/std[out,err,diag]. A +relative path value will be converted to an absolute path. The directory name may include a +colon followed by a comma-delimited list of optional case-insensitive directives. Supported +directives currently include NOJOBID (do not include a job-id directory level) and NOCOPY +(do not copy the output to the stdout/err streams) + +DEPRECATED: please see "--help output" for details +# +[output-filename] +Redirect output from application processes into filename.rank. A relative path value will be +converted to an absolute path. The directory name may include a colon followed by a +comma-delimited list of optional case-insensitive directives. Supported directives currently +include NOCOPY (do not copy the output to the stdout/err streams) + +DEPRECATED: please see "--help output" for details +# +[merge-stderr-to-stdout] +Merge stderr to stdout for each process + +DEPRECATED: please see "--help output" for details +# +[display-devel-map] +Display a detailed process map (mostly intended for developers) +just before launch + +DEPRECATED: please see "--help display" for details +# +[display-topo] +Display the topology as part of the process map (mostly intended +for developers) just before launch + +DEPRECATED: please see "--help display" for details +# +[report-bindings] +Display process bindings to stderr + +DEPRECATED: please see "--help display" for details +# +[display-map] +Display the process map just before launch + +DEPRECATED: please see "--help display" for details +# +[display-allocation] +Display the allocation being used by this job + +DEPRECATED: please see "--help display" for details +# +[with-ft-bad-option] +The OMPI personality was passed the --with-ft directive, +but that directive was given an unsupported option: + + Option: %s + +Supported options include: + +* typical represenations for "true" ("yes", "true", or "1") + +* typical representations for "false" ("no", "false", or "0") + +* either "ulfm" or "mpi" + +Please provide one of the above values and try again.