[Openib-windows] post send request in invalid QP state

[Fabian Tillier]

Hi Yossi,

On 3/13/06, Yossi Leybovich <sleybo at mellanox.co.il> wrote:
>
> Yes we can track that state . I add the check.
> Pls fix the documantation

Can you include the change to the docs with the patch that changes the behavior?

> > If we remove the value, what is the behavior of a post
> > operation when the QP is in the reset state?  If it fails
> > silently but the post call succeeded, we have a problem.  A
> > simple flag would solve the problem.
>
> Yes it silently drop.

We don't want a silent failure.

> > Heck, even an assertion in debug builds during post would be
> > fine, so that we don't have the overhead on a release build.
>
> I add a check.

Generally, we should decide whether to transition to assertions for
many of the runtime checks.  The IBAL code initially was that way, but
was changed to runtime checks for a lot of things by Mellanox a ways
back.

Replacing return codes with assertions simplifies unit testing, since
it removes the need to test a bunch of error handling.

- Fab