[ofa-general] Re: [PATCH] opensm: provide methods for getting the OpenSM and Log

[Sasha Khapyorsky]

Hi Tim,

On 08:48 Mon 10 Mar     , Timothy A. Meier wrote:
> I understand, and agree in principal with the philosophy.

Ok.

> In practice, however, this leads to many instances where a pointer to the
> opensm object is just being casually added to the argument list of a 
> function.

It is not really needed in most cases. In OpenSM most functions work with
some objects, just keep the reference to osm_opensm_t there.

> * I just feel that if a function has a legitimate need for the opensm 
> object,
> it shouldn't impose that requirement on its entire calling tree.

See above.

> No matter how it is created or initialized, the basic design assumes one 
> and only
> one OpenSM object.

I don't think so, it is rather opposite IMO. Of course there could be
some code which violates this, but looking over history I think the
basic design was multi instance ready.

> In addition, the OpenSM object contains other objects 
> (such as
> the log) that are also unique.

It also can be replaced "on the fly" (I even played with it in a past).

> If there comes a time when more than one opensm per thread of execution is 
> required,
> I suspect many things will need to be re-examined.

That is true. :(

Sasha