[ofa-general] [PATCH] ib_core: Use weak ordering for data registered memory

[Arnd Bergmann]

On Tuesday 28 October 2008, Talpey, Thomas wrote:

> >
> >In the long run, I'd like to see something like that for all device drivers,
> >so that a user space library can tell the kernel about its requirements.
> 
> And vice-versa. If a driver or device cannot provide the requirement, it
> needs to communicate that back to the requester.

It all depends on how you define it. If the eager RDMA option asks specifically
for strict ordering, you need to handle errors. If you define it so that
not using eager RDMA passes a flag to allow relaxed ordering, you don't
need to communicate back.

	Arnd <><