On Thu, Dec 10, 2015 at 09:58:51AM +0200, Moni Shoua wrote: > > creating a horrible API, or requiring all vendors to implement a > > network type flag. > > > Actually you haven't suggested anything yet besides a name to the function. > I already said that calculating gid_index from wc and grh requires > extra CPU cycles and is prone to mistakes. But, I might be wrong and > you have a better idea. Do you?
I told you, if a driver requires vendor specific information then include it as private data in the grh. I keep saying this: computing the gid index is not optional. If a drive can't do it efficiently then too bad, it has to go the long way around. >> I think mlx made a big mistake returning network_type instead of gid >> index, and I don't want to see that error enshrined in our API. >> > returning gid_index is wrong because it forces CQ pollers to be aware > of the entire table. Like I already mentioned, the GID table is a HW > resource that can be divided and handed to multiple VMs, Please. If HW virt stuff can't sort this out it is broken, there are many trivial solutions. Jason -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html