[openib-general] [PATCH] IPOIB: Use a GRH when appropriate for unicast packets

[Hal Rosenstock]

On Fri, 2007-01-26 at 16:45, Sean Hefty wrote:
> >Maybe the best fix is to have ib_init_ah_from_path() itself print a
> >warning if the GID index can't be found, just set the gid_index to 0
> >in that case, and change ib_init_ah_from_path() to return void?
> >
> >What do you think of doing something like this for 2.6.21:
> 
> Changing this to a void seems fine to me.  Hal, do you see any issues with this?
> Can the problem occur if we have an out of date path record?

There's some weird edge cases perhaps like subnet merge where the subnet
prefix might change and cause this sort of failure. It seems the
question is whether in these cases, is it better not to respond or to
respond with GID index 0 ? I'm not sure.

> And even if that can occur, can the problem be pushed off until the user calls ib_create_ah?

Not sure what you mean by this.

-- Hal

> 
> - Sean