[openib-general] ib_req_ncomp_notif in core_ layer

[Michael S. Tsirkin]

Hello!
Quoting r. Roland Dreier (roland at topspin.com) "Re: [openib-general] ib_req_ncomp_notif in core_ layer":
>     Sean> Since it's an optional
>     Sean> call though, it shouldn't assume that the call exists, and
>     Sean> should check first.

I'd like to point out that any device can always implement
req_ncomp_notif by means of req_comp_notif and cq peek
or possibly even just alias req_ncomp_notif to req_comp_notif
since the user will be always prepared to get a spurious
event, right?

So I would expect the core code to privide a software implementation
using one of these two approaches if the harware driver
set the pointer to 0 since device hardware does not know how to handle
such event.

This way the user would not need to worry if hardware can
handle this feature, and thats the point of abstraction layer in
my opinion - to abstract hardware.

MST