[ofa-general] Re: [RFC][PATCH] last wqe event handler patch

[Shirley Ma]

Roland Dreier <rdreier at cisco.com> wrote on 06/25/2008 06:33:00 PM:

>  > My understand it could happen from the comments under
>  >  static void ipoib_cm_start_rx_drain()
>  > {
>  >         /*
>  >          * QPs on flush list are error state.  This way, a "flush
>  >          * error" WC will be immediately generated for each WR we
post.
>  >          */
>
> I don't see what this comment has to do with anything... the QP is in
> the error state so posting a send request will generate a completion
> with flush error status.  But that doesn't say anything about whether
> earlier completions may become visible in a CQ after a last WQE reached
> event is reported to the consumer.

I thought handling last wqe event drain WRs one by one is more safe. But if
there is no such cqes left in this case, then what I hit could be a
different bug. More debug is needed. Thanks for helping this.

Thanks
Shirley
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20080625/6c0012d5/attachment.html>