[ofa-general] RE: [PATCH] IB/mad: Allow tuning of QP0 and QP1 sizes

[Hal Rosenstock]

On Wed, Aug 12, 2009 at 1:09 PM, Sean Hefty <sean.hefty at intel.com> wrote:

> >+      mad_recvq_size = roundup_pow_of_two(mad_recvq_size);
> >+      mad_recvq_size = min(mad_recvq_size, IB_MAD_QP_MAX_SIZE);
> >+      mad_recvq_size = max(mad_recvq_size, IB_MAD_QP_MIN_SIZE);
> >+
> >+      mad_sendq_size = roundup_pow_of_two(mad_sendq_size);
> >+      mad_sendq_size = min(mad_sendq_size, IB_MAD_QP_MAX_SIZE);
> >+      mad_sendq_size = max(mad_sendq_size, IB_MAD_QP_MIN_SIZE);
>
> Why round up to a power of two or have min/max restrictions?


power of two is arbitrary and could be removed. min is also somewhat
arbitrary but didn't want to allow it too much smaller than it already is
(default for this patch). max truly is a maximum as create QP fails with
larger size (didn't try this across all HCAs though).


>
>
> _______________________________________________
> general mailing list
> general at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
>
> To unsubscribe, please visit
> http://openib.org/mailman/listinfo/openib-general
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20090812/b47d2d57/attachment.html>