Re: [bgl-discuss] MPI failure, simplified

[Ray Loy <rloy@xxxxxxxxxxx>]

> andrew wrote:
> My understanding of the MPI standard is that the implementation is
> free to buffer or not, but that no matter what it does that this is
> guaranteed = to work


that's my interpretation as well.  EAGER is a heuristic, and if it
doesn't work out (i.e.  there isn't enough memory on the receive side)
then the MPI implementation should be falling back to the handshake
method (e.g. sending a NAK and re-doing it).  sure, it won't be
efficient, but it should work.


ray

- --------------------------------------------------------------------
To add or remove yourself from this mailing list, use the 'notifyme'
command on any BGL machine. To remove: notifyme -n, to add: notifyme -y.