[openib-general] Re: [PATCH] process locked in D state.

[Roland Dreier]

    Michael> Right, but it seems this introduces the race for apps
    Michael> doing deregister_mr properly, too, which used to work
    Michael> fine.

True.

    Michael> Right, but at least lets not introduce the hole for the
    Michael> good case.  When I call deregister_mr I have the right to
    Michael> expect that I can lock the memory immediately after that.
    Michael> So deregister mr should always wait till it has the
    Michael> semaphore.

    Michael> In the case of close, its the buggy app that suffers, and
    Michael> I'm ok with handling it in a work queue since you think
    Michael> its worth it - but lets not try to optimise that case, it
    Michael> should be ok to always use the workqueue.

OK, this looks fine.  Applied (with slight renames of function names).

Thanks,
  Roland