[Mpi3-rma] Next RMA Telecon
Torsten Hoefler
htor at illinois.edu
Sun Nov 27 13:14:07 CST 2011
Hello All,
The doodle poll indicates that everybody can make it on Friday 12/9
2pm (I hope Brian can make it). We may need a bit more than 1 hour (it
would be good to discuss every item on the agenda, so please plan for
1.5 hours if possible). The usual call-in information is (thanks to
Rajeev):
Dial-in Number: 866-654-6744
International Dial-in Number: 1-517-308-8226
Participant Access code: 6086006
The current agenda draft includes:
1) Adam's suggestion that datatype offsets have to be relative to MPI_BOTTOM
2) disp_unit should not be fixed (one can create one window for each)
3) allocate_shared proposal with MPI_Win_lock_all(exclusive)
4) ticket 300: https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/300
5) ticket 283: https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/283
6) Adam suggested: p23, 6-7 Regarding status, we should either say that
all fields are undefined (except for the error) or we should specify
all values, e.g., what about values returned for MPI_GET_COUNT,
MPI_GET_ELEMENTS, or MPI_TEST_CANCELLED? Replace "The values of the
MPI_SOURCE and MPI_TAG fields are undefined." with "All other fields
in status are undefined." -- Question: Would it be useful to set
fields for count and get elements?
Additions are welcome anytime!
Some of the questions come from Adam's review (see comments ticket 270).
I fixed several ticket 0 things in the proposal. Rajeev reviewed but
additional reviews are welcome!
All the Best,
Torsten
--
bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ -----
Torsten Hoefler | Performance Modeling and Simulation Lead
Blue Waters Directorate | University of Illinois (UIUC)
1205 W Clark Street | Urbana, IL, 61801
NCSA Building | +01 (217) 244-7736
More information about the mpiwg-rma
mailing list