2013-07-21 01:48:23 +04:00
|
|
|
#!/usr/bin/perl
|
|
|
|
#
|
|
|
|
# Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
|
|
|
|
# University Research and Technology
|
|
|
|
# Corporation. All rights reserved.
|
|
|
|
# Copyright (c) 2004-2005 The University of Tennessee and The University
|
|
|
|
# of Tennessee Research Foundation. All rights
|
|
|
|
# reserved.
|
|
|
|
# Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
|
|
|
|
# University of Stuttgart. All rights reserved.
|
|
|
|
# Copyright (c) 2004-2005 The Regents of the University of California.
|
|
|
|
# All rights reserved.
|
2013-11-03 23:42:16 +04:00
|
|
|
# Copyright (c) 2013 Intel, Inc. All rights reserved.
|
2013-07-21 01:48:23 +04:00
|
|
|
# $COPYRIGHT$
|
|
|
|
#
|
|
|
|
# Additional copyrights may follow
|
|
|
|
#
|
|
|
|
# $HEADER$
|
|
|
|
#
|
|
|
|
|
|
|
|
use File::Find;
|
|
|
|
use File::Basename;
|
|
|
|
use File::Compare;
|
2013-11-03 23:42:16 +04:00
|
|
|
use File::Copy;
|
|
|
|
use File::Path;
|
As per the RFC, bring in the ORTE async progress code and the rewrite of OOB:
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE ***
Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro.
***************************************************************************************
I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week.
The code is in https://bitbucket.org/rhc/ompi-oob2
WHAT: Rewrite of ORTE OOB
WHY: Support asynchronous progress and a host of other features
WHEN: Wed, August 21
SYNOPSIS:
The current OOB has served us well, but a number of limitations have been identified over the years. Specifically:
* it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code)
* we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface.
* the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients
* there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort
* only one transport (i.e., component) can be "active"
The revised OOB resolves these problems:
* async progress is used for all application processes, with the progress thread blocking in the event library
* each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on")
* multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC.
* a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions.
* opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object
* NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions
* obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel
* the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport
* routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active
* all blocking send/recv APIs have been removed. Everything operates asynchronously.
KNOWN LIMITATIONS:
* although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline
* the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker
* routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways
* obviously, not every error path has been tested nor necessarily covered
* determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost.
* reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways
* the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC
This commit was SVN r29058.
2013-08-22 20:37:40 +04:00
|
|
|
use Getopt::Long;
|
2013-09-06 23:08:12 +04:00
|
|
|
use Text::Diff;
|
2013-07-21 01:48:23 +04:00
|
|
|
|
2013-09-06 23:08:12 +04:00
|
|
|
my $src_arg;
|
|
|
|
my $tgt_arg;
|
|
|
|
my $src_dir;
|
|
|
|
my $target_dir;
|
2013-07-21 01:48:23 +04:00
|
|
|
my @src_tree = ();
|
|
|
|
my @tgt_tree = ();
|
|
|
|
my $flag;
|
2013-09-06 23:08:12 +04:00
|
|
|
my $help_arg = 0;
|
|
|
|
my $diff_file = "";
|
|
|
|
my $diff_arg;
|
2013-11-03 23:42:16 +04:00
|
|
|
my $update_arg;
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
my $modified_arg;
|
2014-08-25 21:04:13 +04:00
|
|
|
my $repo_type;
|
|
|
|
my $cmd;
|
As per the RFC, bring in the ORTE async progress code and the rewrite of OOB:
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE ***
Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro.
***************************************************************************************
I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week.
The code is in https://bitbucket.org/rhc/ompi-oob2
WHAT: Rewrite of ORTE OOB
WHY: Support asynchronous progress and a host of other features
WHEN: Wed, August 21
SYNOPSIS:
The current OOB has served us well, but a number of limitations have been identified over the years. Specifically:
* it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code)
* we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface.
* the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients
* there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort
* only one transport (i.e., component) can be "active"
The revised OOB resolves these problems:
* async progress is used for all application processes, with the progress thread blocking in the event library
* each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on")
* multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC.
* a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions.
* opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object
* NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions
* obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel
* the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport
* routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active
* all blocking send/recv APIs have been removed. Everything operates asynchronously.
KNOWN LIMITATIONS:
* although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline
* the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker
* routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways
* obviously, not every error path has been tested nor necessarily covered
* determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost.
* reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways
* the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC
This commit was SVN r29058.
2013-08-22 20:37:40 +04:00
|
|
|
|
2013-07-21 01:48:23 +04:00
|
|
|
sub construct {
|
|
|
|
# don't process directories or links, and dont' recurse down
|
|
|
|
# "special" directories
|
|
|
|
if ( -l $_ ) { return; }
|
|
|
|
if ( -d $_ ) {
|
2015-03-05 19:24:56 +03:00
|
|
|
if ((/\.deps/) || (/\.libs/) || (/\.git/) || (/\.dSYM/) || ($_ eq "autom4te.cache") || ($_ eq "libltdl")) {
|
2013-07-21 01:48:23 +04:00
|
|
|
$File::Find::prune = true;
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
# $File::Find::name is the path relative to the starting point.
|
|
|
|
# $_ contains the file's basename. The code automatically changes
|
|
|
|
# to the processed directory, so we want to add the full pathname.
|
|
|
|
|
|
|
|
# ignore some obvious files we don't care about
|
2014-02-12 02:53:14 +04:00
|
|
|
if (($_ =~ /\.dirstamp$/i) || ($_ =~ /\.DS_Store$/i) || ($_ =~ /\.lo$/i) || ($_ =~ /\.la$/i) || ($_ =~ /\.o$/i) || ($_ =~ /\.\d$/i)) {
|
2013-07-21 01:48:23 +04:00
|
|
|
$File::Find::prune = true;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (($_ eq "Makefile") || ($_ eq "Makefile.in") || ($_ eq "config.log") || ($_ eq "config.status")) {
|
|
|
|
$File::Find::prune = true;
|
|
|
|
return;
|
|
|
|
}
|
2014-02-12 02:53:14 +04:00
|
|
|
# ignore executables
|
|
|
|
if (-x $File::Find::name) {
|
|
|
|
$File::Find::prune = true;
|
|
|
|
return;
|
|
|
|
}
|
2013-07-21 01:48:23 +04:00
|
|
|
|
|
|
|
if ($flag == 0) {
|
|
|
|
push(@src_tree, $File::Find::name);
|
|
|
|
} else {
|
|
|
|
push(@tgt_tree, $File::Find::name);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-09-06 23:08:12 +04:00
|
|
|
# Command line parameters
|
|
|
|
|
|
|
|
my $ok = Getopt::Long::GetOptions("help|h" => \$help_arg,
|
|
|
|
"src=s" => \$src_arg,
|
|
|
|
"tgt=s" => \$tgt_arg,
|
|
|
|
"diff=s" => \$diff_arg,
|
2013-11-03 23:42:16 +04:00
|
|
|
"update" => \$update_arg,
|
2015-03-05 19:24:56 +03:00
|
|
|
"update-modified" => \$modified_arg
|
2013-09-06 23:08:12 +04:00
|
|
|
);
|
|
|
|
|
|
|
|
if (!$ok || $help_arg) {
|
|
|
|
print "Invalid command line argument.\n\n"
|
|
|
|
if (!$ok);
|
|
|
|
print "Options:
|
2013-11-03 23:42:16 +04:00
|
|
|
--diff | -diff Output diff of changed files to specified file
|
|
|
|
--src | -src Head of source directory
|
|
|
|
--tgt | -tgt Head of target directory
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
--update | -update Apply changes to update target
|
2015-03-05 19:24:56 +03:00
|
|
|
--update-modified Only update modified files (do not add/delete files)\n";
|
2013-09-06 23:08:12 +04:00
|
|
|
exit($ok ? 0 : 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!$src_arg || !$tgt_arg) {
|
|
|
|
print "Missing src or tgt directory\n";
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2013-11-03 23:42:16 +04:00
|
|
|
$src_dir = File::Spec->rel2abs($src_arg);
|
|
|
|
$target_dir = File::Spec->rel2abs($tgt_arg);
|
|
|
|
my @srcpth = ();
|
|
|
|
my @newpth = ();
|
|
|
|
my $spath;
|
|
|
|
my $npath;
|
|
|
|
# if we are updating, then identify the
|
|
|
|
# leading elements of the src_dir path that
|
|
|
|
# must be replaced when pointing to the
|
|
|
|
# target location for a file copy
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
if ($update_arg || $modified_arg) {
|
2013-11-03 23:42:16 +04:00
|
|
|
my $s;
|
|
|
|
my $t;
|
|
|
|
my @srcpath = File::Spec->splitdir($src_dir);
|
|
|
|
my @tgtpath = File::Spec->splitdir($target_dir);
|
|
|
|
# find the place where they first differ - since
|
|
|
|
# they cannot be identical, they must differ
|
|
|
|
# somewhere
|
|
|
|
my $found = 0;
|
|
|
|
while ($found == 0) {
|
|
|
|
$s = shift(@srcpath);
|
|
|
|
$t = shift(@tgtpath);
|
|
|
|
push(@srcpth, $s);
|
|
|
|
push(@newpth, $t);
|
|
|
|
if ($s ne $t) {
|
|
|
|
$found = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
# if either path has been exhausted, then we are done
|
|
|
|
if (0 != scalar(@srcpath) && 0 != scalar(@tgtpath)) {
|
|
|
|
# find the place where they re-converge - this
|
|
|
|
# might be nowhere, e.g., if they provided the
|
|
|
|
# top of two different source trees
|
|
|
|
}
|
|
|
|
$spath = join("/", @srcpth);
|
|
|
|
$npath = join("/", @newpth);
|
|
|
|
print "Source: " . $spath . " New: " . $npath . "\n";
|
|
|
|
}
|
2013-09-06 23:08:12 +04:00
|
|
|
|
|
|
|
if ($diff_arg) {
|
|
|
|
$diff_file = File::Spec->rel2abs($diff_arg);
|
|
|
|
unlink ($diff_file);
|
|
|
|
open(MYFILE, ">$diff_file");
|
|
|
|
}
|
|
|
|
|
|
|
|
my $len_src_dir = length($src_dir);
|
|
|
|
my $len_tgt_dir = length($target_dir);
|
|
|
|
|
2013-07-21 01:48:23 +04:00
|
|
|
# construct a tree of all files in the source directory tree
|
|
|
|
$flag = 0;
|
|
|
|
find(\&construct, $src_dir);
|
|
|
|
|
|
|
|
# construct a tree of all files in the target directory tree
|
|
|
|
$flag = 1;
|
|
|
|
find(\&construct, $target_dir);
|
|
|
|
|
|
|
|
print "size of src_tree: " . @src_tree . ".\n";
|
|
|
|
print "size of tgt_tree: " . @tgt_tree . ".\n";
|
|
|
|
|
|
|
|
# print a list of files in the source tree that need to be added to the target
|
|
|
|
my $found;
|
|
|
|
my $src_file;
|
|
|
|
my $tgt_file;
|
|
|
|
my @modified = ();
|
|
|
|
my @src_pared = ();
|
|
|
|
my $i;
|
|
|
|
foreach $src (@src_tree) {
|
|
|
|
# strip the leading elements of the path that was given to us
|
|
|
|
$src_file = substr($src, $len_src_dir);
|
|
|
|
$found = 0;
|
|
|
|
$i = -1;
|
|
|
|
foreach $tgt (@tgt_tree) {
|
|
|
|
$i = $i + 1;
|
|
|
|
$tgt_file = substr($tgt, $len_tgt_dir);
|
|
|
|
if ($src_file eq $tgt_file) {
|
|
|
|
# printf "Matched: " . $src_file . " " . $tgt_file . "\n";
|
|
|
|
# file has been found - ignore it
|
|
|
|
$found = 1;
|
|
|
|
if (compare($src, $tgt) != 0) {
|
2013-09-06 23:08:12 +04:00
|
|
|
if ($diff_arg) {
|
2013-11-03 23:42:16 +04:00
|
|
|
my $diff = diff $tgt, $src, { STYLE => "Unified" };
|
2013-09-06 23:08:12 +04:00
|
|
|
print MYFILE $diff . "\n";
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
push(@modified, $src);
|
|
|
|
} elsif ($update_arg || $modified_arg) {
|
2014-01-28 23:37:27 +04:00
|
|
|
print "Updating $src to $tgt\n";
|
2013-11-03 23:42:16 +04:00
|
|
|
copy("$src", "$tgt") or die "Copy failed: src=$src tgt=$tgt\n";
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
} else {
|
|
|
|
push(@modified, $src);
|
2013-09-06 23:08:12 +04:00
|
|
|
}
|
2013-07-21 01:48:23 +04:00
|
|
|
}
|
|
|
|
# remove this file from the target tree as it has been found
|
|
|
|
# splice @tgt_tree, $i, 1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
if (!$modified_arg && $found == 0) {
|
2013-11-03 23:42:16 +04:00
|
|
|
if ($update_arg) {
|
|
|
|
my $targetpath = $src;
|
|
|
|
$targetpath =~ s/$spath/$npath/;
|
|
|
|
my $tpath = dirname($targetpath);
|
|
|
|
if (! -d $tpath) {
|
|
|
|
my $dirs = eval { mkpath($tpath) };
|
|
|
|
if (!$dirs) {
|
|
|
|
print "Failed to create path $tpath\n";
|
|
|
|
exit;
|
|
|
|
}
|
2014-08-25 21:04:13 +04:00
|
|
|
print "Adding $tpath to repo\n";
|
2015-03-05 19:24:56 +03:00
|
|
|
$cmd = "pushd $target_dir >& /dev/null; git add $tpath >& /dev/null; popd >& /dev/null";
|
2013-11-03 23:42:16 +04:00
|
|
|
system($cmd);
|
|
|
|
}
|
2014-08-25 21:04:13 +04:00
|
|
|
print "Adding $src to repo\n";
|
2013-11-03 23:42:16 +04:00
|
|
|
copy("$src", "$targetpath") or die "Update failed: src=$src tgt=$targetpath\n";
|
2015-03-05 19:24:56 +03:00
|
|
|
$cmd = "pushd $target_dir >& /dev/null; git add $targetpath >& /dev/null; popd >& /dev/null";
|
2013-11-03 23:42:16 +04:00
|
|
|
system($cmd);
|
|
|
|
} else {
|
|
|
|
print "Add: " . $src . "\n";
|
|
|
|
}
|
2013-09-06 23:08:12 +04:00
|
|
|
} else {
|
|
|
|
push(@src_pared, $src);
|
|
|
|
}
|
2013-07-21 01:48:23 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
print "\n";
|
|
|
|
|
|
|
|
# print a list of files in the target tree that need to be deleted
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
if (!$modified_arg) {
|
|
|
|
foreach $tgt (@tgt_tree) {
|
|
|
|
$found = 0;
|
|
|
|
$tgt_file = substr($tgt, $len_tgt_dir);
|
|
|
|
foreach $src (@src_pared) {
|
|
|
|
$src_file = substr($src, $len_src_dir);
|
|
|
|
if ($src_file eq $tgt_file) {
|
|
|
|
# file has been found - ignore it
|
|
|
|
$found = 1;
|
|
|
|
break;
|
|
|
|
}
|
2013-09-06 23:08:12 +04:00
|
|
|
}
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
if ($found == 0) {
|
|
|
|
if ($update_arg) {
|
2014-08-25 21:04:13 +04:00
|
|
|
print "Removing $tgt_file from repo\n";
|
2015-03-05 19:24:56 +03:00
|
|
|
$cmd = "pushd $target_dir >& /dev/null; git rm .$tgt_file >& /dev/null; popd >& /dev/null";
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
system($cmd);
|
|
|
|
} else {
|
|
|
|
print "Delete: " . $tgt . "\n";
|
|
|
|
}
|
2013-11-03 23:42:16 +04:00
|
|
|
}
|
2013-09-06 23:08:12 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
print "\n";
|
2013-07-21 01:48:23 +04:00
|
|
|
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
# print a list of files that have been modified
|
|
|
|
foreach $tgt (@modified) {
|
|
|
|
print "Modified: " . $tgt . "\n";
|
2013-09-06 23:08:12 +04:00
|
|
|
}
|
|
|
|
|
George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL
All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.
This commit was SVN r32317.
2014-07-26 04:47:28 +04:00
|
|
|
|
2013-09-06 23:08:12 +04:00
|
|
|
if ($diff_arg) {
|
|
|
|
close(MYFILE);
|
|
|
|
}
|