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

[Arnd Bergmann]

On Wednesday 22 October 2008, Eli Cohen wrote:
> In theory it would break Mellanox's guarantee for strict ordering on
> data, but in practice it will not since the only architecture that
> supports weak ordering is CELL. As Arnd suggested in his response
> email, here is the patch with a module parameter which by default will
> not configure weak ordering for data. Anyone wishing to benefit from
> weak ordering will have to set the module parameter accordingly.
> 

Thank you for following up with this patch, it looks good.

As a minor detail, I think we should make this module parameter
writable in order to allow switching the behaviour without
reloading the infiniband drivers.

so instead of 
 module_param(allow_weak_ordering, bool, 0444);
I think it would be better to use
 module_param(allow_weak_ordering, bool, 0644);

As mentioned before, I would personally also prefer to make
this attribute '1' by default instead of zero, but I trust
your judgement if you think the default should be '0'.

	Arnd <><