Messages in this thread | | | From | "Tian, Kevin" <> | Subject | RE: [PATCH v4 5/5] vfio: Support IO page table replacement | Date | Mon, 13 Mar 2023 02:04:04 +0000 |
| |
> From: Nicolin Chen <nicolinc@nvidia.com> > Sent: Saturday, March 11, 2023 7:38 AM > > On Fri, Mar 10, 2023 at 11:53:56AM +0000, Tian, Kevin wrote: > > External email: Use caution opening links or attachments > > > > > > > From: Nicolin Chen <nicolinc@nvidia.com> > > > Sent: Wednesday, March 8, 2023 10:26 PM > > > > > > Now both the physical path and the emulated path should support an IO > > > page > > > table replacement. > > > > > > Call iommufd_device_replace() when vdev->iommufd_attached is true. > > > > > > > why is replace enabled only in physical path in this patch? > > The emulated pathway does not call iommufd_device_attach() but > iommufd_access_set_ioas() in the other patch, which internally > takes care of the replacement for the access pointer. >
I thought there is a similar check as in physical path which should be removed:
if (vdev->iommufd_attached) - return -EBUSY;
but looks it's not the case for the emulated path.
| |