Does this failure path exist in 1.5? -Nathan
On Fri, 21 Oct 2011, r...@osl.iu.edu wrote:
Author: rhc Date: 2011-10-21 10:44:48 EDT (Fri, 21 Oct 2011) New Revision: 25350 URL: https://svn.open-mpi.org/trac/ompi/changeset/25350 Log: Fix a minor issue seen by Jeff in specific failure pathway Text files modified: trunk/orte/mca/debugger/base/debugger_base_fns.c | 8 ++++++-- 1 files changed, 6 insertions(+), 2 deletions(-) Modified: trunk/orte/mca/debugger/base/debugger_base_fns.c ============================================================================== --- trunk/orte/mca/debugger/base/debugger_base_fns.c (original) +++ trunk/orte/mca/debugger/base/debugger_base_fns.c 2011-10-21 10:44:48 EDT (Fri, 21 Oct 2011) @@ -90,10 +90,14 @@ orte_process_name_t rank0; int rc; - if (MPIR_proctable) { + /* if we couldn't get thru the mapper stage, we might + * enter here with no procs. Avoid the "zero byte malloc" + * message by checking here + */ + if (MPIR_proctable || 0 == jdata->num_procs) { /* already initialized */ opal_output_verbose(5, orte_debugger_base.output, - "%s: debugger already initialized", + "%s: debugger already initialized or zero procs", ORTE_NAME_PRINT(ORTE_PROC_MY_NAME)); return; } _______________________________________________ svn mailing list s...@open-mpi.org http://www.open-mpi.org/mailman/listinfo.cgi/svn
