Re: ensure device private pages have an owner v2
On Thu, Mar 19, 2020 at 08:16:33AM +0100, Christoph Hellwig wrote: > On Wed, Mar 18, 2020 at 09:28:49PM -0300, Jason Gunthorpe wrote: > > > Changes since v1: > > > - split out the pgmap->owner addition into a separate patch > > > - check pgmap->owner is set for device private mappings > > > - rename the dev_private_owner field in struct migrate_vma to src_owner > > > - refuse to migrate private pages if src_owner is not set > > > - keep the non-fault device private handling in hmm_range_fault > > > > I'm happy enough to take this, did you have plans for a v3? > > I think the only open question is if merging 3 and 4 might make sense. > It's up to you if you want it resent that way or not. Okay, I kept it as is and elaborated the commit messages a bit based on the discussion It doesn't seem like the changes outside hmm are significant enough to need more acks Thanks, Jason ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: ensure device private pages have an owner v2
On Thu, Mar 19, 2020 at 08:16:33AM +0100, Christoph Hellwig wrote: > On Wed, Mar 18, 2020 at 09:28:49PM -0300, Jason Gunthorpe wrote: > > > Changes since v1: > > > - split out the pgmap->owner addition into a separate patch > > > - check pgmap->owner is set for device private mappings > > > - rename the dev_private_owner field in struct migrate_vma to src_owner > > > - refuse to migrate private pages if src_owner is not set > > > - keep the non-fault device private handling in hmm_range_fault > > > > I'm happy enough to take this, did you have plans for a v3? > > I think the only open question is if merging 3 and 4 might make sense. > It's up to you if you want it resent that way or not. Now that I understand that amdgpu doesn't set the 'do not return device_private pages' flag, I think the split is fine, I'll grab it as is then today Thanks, Jason ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: ensure device private pages have an owner v2
On Mon, Mar 16, 2020 at 08:32:12PM +0100, Christoph Hellwig wrote: > When acting on device private mappings a driver needs to know if the > device (or other entity in case of kvmppc) actually owns this private > mapping. This series adds an owner field and converts the migrate_vma > code over to check it. I looked into doing the same for > hmm_range_fault, but as far as I can tell that code has never been > wired up to actually work for device private memory, so instead of > trying to fix some unused code the second patch just remove the code. > We can add it back once we have a working and fully tested code, and > then should pass the expected owner in the hmm_range structure. > > Changes since v1: > - split out the pgmap->owner addition into a separate patch > - check pgmap->owner is set for device private mappings > - rename the dev_private_owner field in struct migrate_vma to src_owner > - refuse to migrate private pages if src_owner is not set > - keep the non-fault device private handling in hmm_range_fault I'm happy enough to take this, did you have plans for a v3? Thanks, Jason ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel