On Tue, 2019-06-18 at 10:10 -0300, Jason Gunthorpe wrote:
> On Tue, Jun 18, 2019 at 03:17:09PM +0300, Leon Romanovsky wrote:
> > > /**
> > > * ib_set_client_data - Set IB client context
> > > * @device:Device to set context for
> > > diff --git a/drivers/infiniband/core/nldev.c
> > > b/drivers/infiniband/core/nldev.c
> > > index 69188cbbd99bd5..55eccea628e99f 100644
> > > +++ b/drivers/infiniband/core/nldev.c
> > > @@ -120,6 +120,9 @@ static const struct nla_policy
> > > nldev_policy[RDMA_NLDEV_ATTR_MAX] = {
> > > [RDMA_NLDEV_ATTR_DEV_PROTOCOL] = { .type =
> > > NLA_NUL_STRING,
> > > .len = RDMA_NLDEV_ATTR_ENTRY_STRLEN
> > > },
> > > [RDMA_NLDEV_NET_NS_FD] = { .type =
> > > NLA_U32 },
> > > + [RDMA_NLDEV_ATTR_CHARDEV_TYPE] = { .type =
> > > NLA_NUL_STRING,
> > > + .len = 128 },
> > > + [RDMA_NLDEV_ATTR_PORT_INDEX] = { .type = NLA_U32 },
> >
> > It is wrong, we already have RDMA_NLDEV_ATTR_PORT_INDEX declared in
> > nla_policy.
> > But we don't have other RDMA_NLDEV_ATTR_CHARDEV_* declarations
> > here.
>
> Doug can you fix it?
I haven't pushed my wip to for-next yet, so yeah, I can fix it. We
just need to decide on what the full fix is ;-)
Drop the duplicate ATTR_PORT_INDEX, but what about a final decision on
including the outputs for possible future type checking? You and Leon
seem to be going back and forth, and I don't have strong feelings
either way on this one. It's just a definition statement, not like
it's a dead subroutine.
--
Doug Ledford <dledford@xxxxxxxxxx>
GPG KeyID: B826A3330E572FDD
Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57
2FDD
Attachment:
signature.asc
Description: This is a digitally signed message part
