
last December. :-( Add new MCA param: maffinity_libnuma_policy. Thanks to David Singleton for the suggestion. Here's the help text about it: {{{ MCA maffinity: parameter "maffinity_libnuma_policy" (current value: <loose>, data source: default value) Binding policy that determines what happens if memory is unavailable on the local NUMA node. A value of "strict" means that the memory allocation will fail; a value of "loose" means that the memory allocation will spill over to another NUMA node. }}} This commit was SVN r24290.
28 строки
648 B
Plaintext
28 строки
648 B
Plaintext
# -*- text -*-
|
|
#
|
|
# Copyright (c) 2011 Cisco Systems, Inc. All rights reserved.
|
|
# $COPYRIGHT$
|
|
#
|
|
# Additional copyrights may follow
|
|
#
|
|
# $HEADER$
|
|
#
|
|
# This is the US/English help file for Open MPI's libnuma support
|
|
#
|
|
[invalid policy]
|
|
WARNING: An invalid value was given for the maffinity_libnuma_policy
|
|
MCA parameter. The policy determines what happens when Open MPI tries
|
|
to allocate local memory, but no local memory is available. The value
|
|
provided was:
|
|
|
|
Value: %s
|
|
PID: %d
|
|
|
|
Valid values are:
|
|
|
|
strict: the memory allocation will fail
|
|
loose: the memory allocation will spill over to remote memory
|
|
|
|
Your job will now abort.
|
|
#
|