On Wed, Aug 25, 2021 at 02:13:41PM -0600, Alex Williamson wrote: > On Tue, 24 Aug 2021 21:42:26 -0300 > Jason Gunthorpe <jgg@xxxxxxxxxx> wrote: > > > On Tue, Aug 24, 2021 at 02:30:01PM -0600, Alex Williamson wrote: > > > On Mon, 23 Aug 2021 11:42:04 -0300 > > > Jason Gunthorpe <jgg@xxxxxxxxxx> wrote: > > > > > > > This is straightforward conversion, the ap_matrix_mdev is actually serving > > > > as the vfio_device and we can replace all the mdev_get_drvdata()'s with a > > > > simple container_of() or a dev_get_drvdata() for sysfs paths. > > > > > > > > Cc: Alex Williamson <alex.williamson@xxxxxxxxxx> > > > > Cc: Cornelia Huck <cohuck@xxxxxxxxxx> > > > > Cc: kvm@xxxxxxxxxxxxxxx > > > > Cc: Christoph Hellwig <hch@xxxxxx> > > > > Reviewed-by: Tony Krowiak <akrowiak@xxxxxxxxxxxxx> > > > > Reviewed-by: Christoph Hellwig <hch@xxxxxx> > > > > Signed-off-by: Jason Gunthorpe <jgg@xxxxxxxxxx> > > > > drivers/s390/crypto/vfio_ap_ops.c | 155 +++++++++++++++----------- > > > > drivers/s390/crypto/vfio_ap_private.h | 2 + > > > > 2 files changed, 91 insertions(+), 66 deletions(-) > > > > > > Jason & Tony, > > > > > > Would one of you please rebase on the other's series? The merge > > > conflict between this and 20210823212047.1476436-1-akrowiak@xxxxxxxxxxxxx > > > is more than I'd like to bury into a merge commit and I can't test beyond > > > a cross compile. Thanks, > > > > Tony, as you have the Hw to test it is probably best if you do it, all > > I can do is the same as Alex. > > Maybe we can do it this way, it seems easier to port Jason's single > patch on top of Tony's series than vice versa. The below is my attempt > at that. If Jason and Tony can give this a thumbs up, I'll run with it. It matches what I got when I did the same rebase Jason
