Roland Dreier wrote: > > The patch below is identical to Yossi's patch but without the copy of HA > in neigh_add_path. > > Why did Yossi include that copy? Does this patch still fix everything? > > - R. Yossi's intention was to save compares (gid size long) in ipoib_start_xmit(). The thought was that once the condition to start a new path query is met there is no need to meet it again (especially when the cost is high) The only thing is what happens when path query fails (I explained it above) and this is why I think its better to remove the copy.
The new patch still fixes the basic problem that it intends to (as explained by Yossi) thanks MoniS _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
