[ofa-general] New proposal for memory management

[Aaron Fabbri]

Jeff Squyres <jsquyres <at> cisco.com> writes:
...
> 
> Introduction:
> =============
> 
> MPI has long had a problem maintaining its own verbs memory
> registration cache in userspace.  The main issue is that user
> applications are responsible for allocating/freeing their own data
> buffers -- the MPI layer does not (usually) have visibility when
> application buffers are allocated or freed.

I'm late to the debate, so sorry if you've already covered this...

Have you considered changing the MPI API to require applications to use MPI to 
allocate any/all buffers that may be used for network I/O?  That is, instead of 
calling malloc() et al., call a new mpi_malloc() which allocates from pre-
registered memory.

I'm sure it is hard to just up and change MPI, but it seems like the right 
thing to do.

(While you're at it, change the sockets interface too.)

-AF