[openib-general] mthca and DDR not hidden

[Sean Hefty]

On Mon, 04 Oct 2004 13:36:26 -0400
Hal Rosenstock <halr at voltaire.com> wrote:

> On Sat, 2004-10-02 at 15:26, Michael S. Tsirkin wrote:
> > I'd like to suggest that the mad layer could expose an allocator
> > function that the user will call to grab the memory for the mad.
> > This function would return the memory pointer and the appropriate
> > lkey value.
> 
> Wouldn't the triple of lkey, start address, and size need to be returned

I would think that the size would need to be an input into the allocator routine.

> There was a discussion about a MAD pool or even multiple pools before
> the ib_mad API was agreed upon. It seemed that consensus at the time was
> against this, although I think the consensus might have been 51 to 49 on
> this.

Having an allocator routine might force users to perform data copies when sending data.

Do all of the existing MAD implementations have routines to allocate MADs when sending data, and require those routines to be used?