On 8/11/2017 12:46 PM, Hal Rosenstock wrote:
On 8/11/2017 6:49 AM, Leon Romanovsky wrote:
From: Leon Romanovsky <leonro@xxxxxxxxxxxx>
There is no need to carry code, which is not implemented in any
underlying hardware. This patch removes unimplemented node_types and
their respective node transport layers.
We have supported IB switches and routers since day 1 of OpenIB and
linux-rdma. Although not the original intent, there are still no in-tree
consumers, but there are several out of tree ones. Every time this issue
has come up in the past, it has been decided to continue to support
them. Effectively, IB routers and switches are grandfathered.
I swear I remember having this exact conversation but can't find it in
my mail. Figures. Well regardless...
Rather than doing this, perhaps a better approach would be to supply
some skeleton drivers for IB switch and router.
If someone were to volunteer to do that I could see leaving the
functionality, no not functionality, the non-functioning code, leaving
that in on a _temporary_ basis. In reality though it's been years that
this has been sitting there.
I can see the grandfather clause applying to stuff in the kernel tree,
or even publicly available implementations at a stretch (even then still
probably not). If there any any out of tree switch/router drivers, what
is the likelihood that they support the latest and greatest kernel out
of the box anyway? Is this really going to be a huge problem for them? I
doubt it.
On the flip side, is this code that's been there for forever and a day
really that big of a problem? No not really. It would be good to clean
it up if no one has that big of a problem and I'm not going to argue
very strongly on one side or the other.
-Denny
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
