> -----Original Message----- > From: Jason Gunthorpe <jgg@xxxxxxxx> > Sent: Friday, January 11, 2019 5:45 PM > To: Leon Romanovsky <leon@xxxxxxxxxx> > Cc: Doug Ledford <dledford@xxxxxxxxxx>; Leon Romanovsky > <leonro@xxxxxxxxxxxx>; RDMA mailing list <linux-rdma@xxxxxxxxxxxxxxx>; > Parav Pandit <parav@xxxxxxxxxxxx> > Subject: Re: [PATCH rdma-next 1/7] RDMA: Rename port_callback to > port_init > > On Tue, Dec 18, 2018 at 02:28:30PM +0200, Leon Romanovsky wrote: > > From: Parav Pandit <parav@xxxxxxxxxxxx> > > > > Most provider routines are callback routines which ib core invokes. > > _callback suffix doesn't convey information about when such callback > > is invoked. Therefore, rename port_callback to port_init. > > > > Additionally, store the port_init function pointer to ib_device, so > > that it can be checked in subsequent patch when binding rdma device to > > net namespace. Currently devices which exposes their own port specific > > files cannot be bind to net namespace. Once infrastructure is built to > > init and uninit the ports, this limitation will be removed. > > > > Signed-off-by: Parav Pandit <parav@xxxxxxxxxxxx> > > Signed-off-by: Leon Romanovsky <leonro@xxxxxxxxxxxx> > > --- > > I've been meaning to do this for a while, applied to for-next thanks > > > diff --git a/drivers/infiniband/hw/qib/qib_verbs.c > > b/drivers/infiniband/hw/qib/qib_verbs.c > > index 611a6b5ef83f..669b25496a9f 100644 > > --- a/drivers/infiniband/hw/qib/qib_verbs.c > > +++ b/drivers/infiniband/hw/qib/qib_verbs.c > > @@ -1496,6 +1496,7 @@ static void qib_fill_device_attr(struct > > qib_devdata *dd) static const struct ib_device_ops qib_dev_ops = { > > .modify_device = qib_modify_device, > > .process_mad = qib_process_mad, > > + .port_init = qib_create_port_files, > > These lists should be kept sorted, I fixed it > I noticed that I did this method as object_method(); while rest of the function pointers are method_object(). So it should be dev ops name should be init_port() to be consistent with rest of them. Please let me know if I should send the patch or you are changing it since you already applied this patch. > Jason