Re: [PATCH v3 2/4] vhost-vdpa: introduce STOP backend feature bit

[Eugenio Perez Martin <eperezma@redhat.com>]

On Thu, May 26, 2022 at 11:07 AM Stefano Garzarella <sgarzare@redhat.com> wrote:
>
> On Thu, May 26, 2022 at 10:57:03AM +0200, Eugenio Perez Martin wrote:
> >On Wed, May 25, 2022 at 1:23 PM Dawar, Gautam <gautam.dawar@amd.com> wrote:
> >>
> >> [AMD Official Use Only - General]
> >>
> >> -----Original Message-----
> >> From: Eugenio Pérez <eperezma@redhat.com>
> >> Sent: Wednesday, May 25, 2022 4:29 PM
> >> To: Michael S. Tsirkin <mst@redhat.com>; netdev@vger.kernel.org; linux-kernel@vger.kernel.org; kvm@vger.kernel.org; virtualization@lists.linux-foundation.org; Jason Wang <jasowang@redhat.com>
> >> Cc: Zhu Lingshan <lingshan.zhu@intel.com>; martinh@xilinx.com; Stefano Garzarella <sgarzare@redhat.com>; ecree.xilinx@gmail.com; Eli Cohen <elic@nvidia.com>; Dan Carpenter <dan.carpenter@oracle.com>; Parav Pandit <parav@nvidia.com>; Wu Zongyong <wuzongyong@linux.alibaba.com>; dinang@xilinx.com; Christophe JAILLET <christophe.jaillet@wanadoo.fr>; Xie Yongji <xieyongji@bytedance.com>; Dawar, Gautam <gautam.dawar@amd.com>; lulu@redhat.com; martinpo@xilinx.com; pabloc@xilinx.com; Longpeng <longpeng2@huawei.com>; Piotr.Uminski@intel.com; Kamde, Tanuj <tanuj.kamde@amd.com>; Si-Wei Liu <si-wei.liu@oracle.com>; habetsm.xilinx@gmail.com; lvivier@redhat.com; Zhang Min <zhang.min9@zte.com.cn>; hanand@xilinx.com
> >> Subject: [PATCH v3 2/4] vhost-vdpa: introduce STOP backend feature bit
> >>
> >> [CAUTION: External Email]
> >>
> >> Userland knows if it can stop the device or not by checking this feature bit.
> >>
> >> It's only offered if the vdpa driver backend implements the stop() operation callback, and try to set it if the backend does not offer that callback is an error.
> >>
> >> Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
> >> ---
> >>  drivers/vhost/vdpa.c             | 16 +++++++++++++++-
> >>  include/uapi/linux/vhost_types.h |  2 ++
> >>  2 files changed, 17 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 1f1d1c425573..32713db5831d 100644
> >> --- a/drivers/vhost/vdpa.c
> >> +++ b/drivers/vhost/vdpa.c
> >> @@ -347,6 +347,14 @@ static long vhost_vdpa_set_config(struct vhost_vdpa *v,
> >>         return 0;
> >>  }
> >>
> >> +static bool vhost_vdpa_can_stop(const struct vhost_vdpa *v) {
> >> +       struct vdpa_device *vdpa = v->vdpa;
> >> +       const struct vdpa_config_ops *ops = vdpa->config;
> >> +
> >> +       return ops->stop;
> >> [GD>>] Would it be better to explicitly return a bool to match the return type?
> >
> >I'm not sure about the kernel code style regarding that casting. Maybe
> >it's better to return !!ops->stop here. The macros likely and unlikely
> >do that.
>
> IIUC `ops->stop` is a function pointer, so what about
>
>      return ops->stop != NULL;
>

I'm ok with any method proposed. Both three ways can be found in the
kernel so I think they are all valid (although the double negation is
from bool to integer in (0,1) set actually).

Maybe Jason or Michael (as maintainers) can state the preferred method here.

Generally I prefer explicit conversions, both signed and from/to
different types length. But I find conversion to bool to be simple
enough to be an exception to the rule. Same with void *. Let's see!

Sending v4 without this changed, waiting for answers.

Thanks!