26 Mar
2013
26 Mar
'13
11:09 a.m.
Hi Jeff, On 03/26/2013 07:26 AM US Central Time, Jeff Squyres (jsquyres) wrote:
2. Lines 28-31 is specifically talking about a non-thread-complaint implementation; it says that it must return THREAD_SINGLE and provide correct answers for functions like MPI_INITIALIZED even from multiple threads (assumedly even simultaneously) before MPI has been initialized.
Please note that the rationale there says that this would still be an incorrect application, but it allows the application to abort gracefully instead of creating garbage output. So, in that sense, the standard is still consistent, IMO. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji