On Thu, 2006-03-16 at 18:34, Roland Dreier wrote:
> This patch has a chunk:
>
> > if (mgmt_class == IB_MGMT_CLASS_SUBN_ADM)
> > return IB_MGMT_SA_HDR;
> > + else if (mgmt_class == IB_MGMT_CLASS_DEVICE_MGMT)
> > + return IB_MGMT_DM_HDR;
> > else if ((mgmt_class >= IB_MGMT_CLASS_VENDOR_RANGE2_START) &&
> > (mgmt_class <= IB_MGMT_CLASS_VENDOR_RANGE2_END))
> > return IB_MGMT_VENDOR_HDR;
>
> and...
>
> > if (mgmt_class == IB_MGMT_CLASS_SUBN_ADM)
> > return IB_MGMT_SA_HDR;
> > + else if (mgmt_class == IB_MGMT_CLASS_DEVICE_MGMT)
> > + return IB_MGMT_DM_HDR;
> > else if ((mgmt_class >= IB_MGMT_CLASS_VENDOR_RANGE2_START) &&
> > (mgmt_class <= IB_MGMT_CLASS_VENDOR_RANGE2_END))
> > return IB_MGMT_VENDOR_HDR;
>
> We should consolidate this identical code into one place.
Yes, I think Sean pointed this out before when this duplication was
originally created.
> Among other
> benefits, this would make this sort of maintenance easier and less
> error prone.
Sure; I can reissue an updated patch for this.
> In fact, there's also the following chunk which maybe could tie in too:
>
> > + } else if (rmpp_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_DEVICE_MGMT) {
> > + hdr_len = IB_MGMT_DM_HDR;
Yes, that might be possible too.
-- Hal
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
