Updated document now available on the website. I'm planning on sending this around to the forum at 5 pm eastern. -- Josh On Mon, Dec 19, 2011 at 1:44 PM, Josh Hursey <[email protected]> wrote:
A new version of the document is available on the following wiki page: �https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/rts_proposal_main I think this one is ready to go. If you have time, take a look over it and send notes to the list. I am planning on sending this out to the MPI Forum mailing list (and attaching it to the ticket) tonight or tomorrow morning.
I'm working on the slides for the meeting. I hope to have a version to circulate tomorrow afternoon.
Thanks, Josh Change Log:-----------�* Minor wording touch up clarifications�* 17.6: Added clarification to MPI_Comm_reenable_any_source() that is ok to pass MPI_GROUP_IGNORE.�* 17.6: Added clarification that MPI_Comm_drain will match across process failure notification. Similar to MPI_Comm_validate.�* 17.7.1: Added matching clarification to MPI_Comm_validate.�* 17.10: Clarify that MPI_Comm_disconnect can be passed a collectively inactive communicator.
On Fri, Dec 16, 2011 at 3:36 PM, Josh Hursey <[email protected]> wrote:
A new version of the document is available on the following wiki page: �https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/rts_proposal_main
We have to have this document finalized by COB Monday if at all possible. So take a look at it and send feedback to the list.
-- Josh
Change Log: ----------- �* Fix a few typos �* 17.5.1: Update the Process Failure handler section per discussions �* 17.6: Touch up the MPI_ANY_SOURCE wording �* 17.6: Clarify that if MPI_Comm_reenable_any_source is called with an intercommunicator then it returns the group of failed processes in the remote group. �* 17.6: Advice to implementors that MPI_Comm_drain maybe a locally collective operation. �* 17.6: Extend advice to users to note that MPI_FAILHANDLER_MODE_ALL does not guarantee that the failure handler will be called the same number of times unless they call a validation operation to synchronize the handlers. So they need to take care when using MPI_Comm_drain() in this operating mode. �* 17.6.1: Add a sentence allowing MPI_ERR_IN_STATUS to be returned from test and completion operations even if it is just the MPI_ERR_ANY_SOURCE_DISABLED warning. �* 17.8.2: Communicator creation must have a collectively active input communicator, and return uniformly at all processes. �* 17.8.2: Communicator construction operations will match across process failure. So they match similar to MPI_Comm_validate() and not like other collectives. �* 17.8.3: Inter-communicator creation operations have the same constraints as communicator creation (Previous 2 points). �* 17.8.4: Added example section with the communicator creation loop example. �* 17.9: Topology creation operations match semantics of communicator creation. �* 17.10: Dynamic creation operations (spawn and friends) match semantics of communicator creation. �* 17.11: Window creation operations match semantics of communicator creation. �* 17.12.2: File_open/close match semantics of communicator creation. �* A.1.1: Added the MPI_FAILHANDLER_MODE_*'s to the appendix
Open Discussion Items: ---------------------- �* 17.6: Rename MPI_Comm_reenable_any_source to *_validate_* (?) �* 3.10 & 17.6.2 : Do these sections conflict? Should the status only be associated with the 'source' since MPI_Recv would have returned the status value if the operations were called separately?
-- Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey
-- Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey
-- Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey