On Wed, 1 Apr 2020 00:16:13 +0530
Kirti Wankhede <kwankhede@xxxxxxxxxx> wrote:
> On 3/31/2020 3:04 AM, Alex Williamson wrote:
> > On Mon, 30 Mar 2020 22:20:42 +0530
> > Kirti Wankhede <kwankhede@xxxxxxxxxx> wrote:
> >
> >> DMA mapped pages, including those pinned by mdev vendor drivers, might
> >> get unpinned and unmapped while migration is active and device is still
> >> running. For example, in pre-copy phase while guest driver could access
> >> those pages, host device or vendor driver can dirty these mapped pages.
> >> Such pages should be marked dirty so as to maintain memory consistency
> >> for a user making use of dirty page tracking.
> >>
> >> To get bitmap during unmap, user should allocate memory for bitmap, set
> >> size of allocated memory, set page size to be considered for bitmap and
> >> set flag VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP.
> >>
> >> Signed-off-by: Kirti Wankhede <kwankhede@xxxxxxxxxx>
> >> Reviewed-by: Neo Jia <cjia@xxxxxxxxxx>
> >> ---
> >> drivers/vfio/vfio_iommu_type1.c | 55 ++++++++++++++++++++++++++++++++++++++---
> >> include/uapi/linux/vfio.h | 10 ++++++++
> >> 2 files changed, 61 insertions(+), 4 deletions(-)
> >>
> >> diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
> >> index 5efebc2b60e1..266550bd7307 100644
> >> --- a/drivers/vfio/vfio_iommu_type1.c
> >> +++ b/drivers/vfio/vfio_iommu_type1.c
> >> @@ -983,7 +983,8 @@ static int verify_bitmap_size(uint64_t npages, uint64_t bitmap_size)
> >> }
> >>
> >> static int vfio_dma_do_unmap(struct vfio_iommu *iommu,
> >> - struct vfio_iommu_type1_dma_unmap *unmap)
> >> + struct vfio_iommu_type1_dma_unmap *unmap,
> >> + struct vfio_bitmap *bitmap)
> >> {
> >> uint64_t mask;
> >> struct vfio_dma *dma, *dma_last = NULL;
> >> @@ -1034,6 +1035,10 @@ static int vfio_dma_do_unmap(struct vfio_iommu *iommu,
> >> * will be returned if these conditions are not met. The v2 interface
> >> * will only return success and a size of zero if there were no
> >> * mappings within the range.
> >> + *
> >> + * When VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP flag is set, unmap request
> >> + * must be for single mapping. Multiple mappings with this flag set is
> >> + * not supported.
> >> */
> >> if (iommu->v2) {
> >> dma = vfio_find_dma(iommu, unmap->iova, 1);
> >> @@ -1041,6 +1046,14 @@ static int vfio_dma_do_unmap(struct vfio_iommu *iommu,
> >> ret = -EINVAL;
> >> goto unlock;
> >> }
> >> +
> >> + if ((unmap->flags & VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP) &&
> >> + dma &&
> >> + (dma->iova != unmap->iova || dma->size != unmap->size)) {
> >
> >
> > I think your intention was to return error if the user asked for the
> > dirty bitmap and the requested unmap range doesn't exactly match the
> > vfio_dma. Not finding a vfio_dma should therefore also be an error.
> > For example, if we had a single mapping at {0x1000-0x1fff} and the user
> > unmapped with dirty bitmap {0x0-0x2fff}, that should return an error,
> > but it's not caught by the above because there is no vfio_dma @0x0.
> > Therefore I think you want:
> >
> > ((unmap->flags & VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP) &&
> > (!dma || dma->iova != unmap->iova || dma->size != unmap->size))
> >
> > Right? Thanks,
> >
>
>
> Yes, updating check.
>
> Is !dma here also error case when VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP
> flag is not set?
No it's not.
> DMA_UNMAP ioctl returns how much was unmapped, from user space
> perspective this would be from start of range (unmap->iova), right?
It's the actual amount of memory that was unmapped within the provided
range, without getting the dirty bitmap the range doesn't need to start
or end at a vfio_dma, it just cannot bisect a vfio_dma. Thanks,
Alex
