On Wed, Mar 24, 2021 at 11:00:46AM -0300, Jason Gunthorpe wrote:
> On Wed, Mar 24, 2021 at 03:47:34PM +0200, Leon Romanovsky wrote:
> > On Tue, Mar 23, 2021 at 06:59:52PM -0500, Shiraz Saleem wrote:
> > > From: Mustafa Ismail <mustafa.ismail@xxxxxxxxx>
> > >
> > > Register auxiliary drivers which can attach to auxiliary RDMA
> > > devices from Intel PCI netdev drivers i40e and ice. Implement the private
> > > channel ops, and register net notifiers.
> > >
> > > Signed-off-by: Mustafa Ismail <mustafa.ismail@xxxxxxxxx>
> > > Signed-off-by: Shiraz Saleem <shiraz.saleem@xxxxxxxxx>
> > > drivers/infiniband/hw/irdma/i40iw_if.c | 229 +++++++++++++
> > > drivers/infiniband/hw/irdma/main.c | 382 ++++++++++++++++++++++
> > > drivers/infiniband/hw/irdma/main.h | 565 +++++++++++++++++++++++++++++++++
> > > 3 files changed, 1176 insertions(+)
> > > create mode 100644 drivers/infiniband/hw/irdma/i40iw_if.c
> > > create mode 100644 drivers/infiniband/hw/irdma/main.c
> > > create mode 100644 drivers/infiniband/hw/irdma/main.h
> >
> > <...>
> >
> > > +/* client interface functions */
> > > +static const struct i40e_client_ops i40e_ops = {
> > > + .open = i40iw_open,
> > > + .close = i40iw_close,
> > > + .l2_param_change = i40iw_l2param_change
> > > +};
> > > +
> > > +static struct i40e_client i40iw_client = {
> > > + .ops = &i40e_ops,
> > > + .type = I40E_CLIENT_IWARP,
> > > +};
> > > +
> > > +static int i40iw_probe(struct auxiliary_device *aux_dev, const struct auxiliary_device_id *id)
> > > +{
> > > + struct i40e_auxiliary_device *i40e_adev = container_of(aux_dev,
> > > + struct i40e_auxiliary_device,
> > > + aux_dev);
> > > + struct i40e_info *cdev_info = i40e_adev->ldev;
> > > +
> > > + strncpy(i40iw_client.name, "irdma", I40E_CLIENT_STR_LENGTH);
> > > + cdev_info->client = &i40iw_client;
> > > + cdev_info->aux_dev = aux_dev;
> > > +
> > > + return cdev_info->ops->client_device_register(cdev_info);
> >
> > Why do we need all this indirection? I see it as leftover from previous
> > version where you mixed auxdev with your peer registration logic.
>
> I think I said the new stuff has to be done sanely, but the i40iw
> stuff is old and already like this.
They declared this specific "ops" a couple of lines above and all the
functions are static. At least for the new code, in the irdma, this "ops"
thing is not needed.
>
> Though I would be happy to see this fixed too.
>
> Jason
