> From: Liu, Yi L <yi.l.liu@xxxxxxxxx> > Sent: Thursday, August 18, 2022 4:13 PM > > Hi Kevin, > > On 2022/8/18 15:46, Tian, Kevin wrote: > >> From: Jason Gunthorpe <jgg@xxxxxxxxxx> > >> Sent: Tuesday, August 16, 2022 12:50 AM > >> > >> This counts the number of devices attached to a vfio_group, ie the > number > >> of items in the group->device_list. > >> > >> It is only read in vfio_pin_pages(), however that function already does > >> vfio_assert_device_open(). Given an opened device has to already be > >> properly setup with a group, this test and variable are redundant. Remove > >> it. > > > > I didn't get the rationale behind. The original check was for whether > > the group is singleton. Why is it equivalent to the condition of an > > opened device? > > > > Though I do think this check is unnecessary. All the devices in the group > > share the container and iommu domain which is what the pinning > > operation applies to. I'm not sure why the singleton restriction was > > added in the first place. > > see if your confusion is addressed in below link? > > https://lore.kernel.org/kvm/BN9PR11MB5276F9EB0295CBD485A58D308C6D > 9@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/T/#m42eaf154823581082 > 0d7e2ad1491092c4b0bbcba > Ah, yes. I didn't note that this is already addressed. 😊