iommufd/iova_bitmap: Consider page offset for the pages to be pinned

For small bitmaps that aren't PAGE_SIZE aligned *and* that are less than
512 pages in bitmap length, use an extra page to be able to cover the
entire range e.g. [1M..3G] which would be iterated more efficiently in a
single iteration, rather than two.

Fixes: b058ea3ab5af ("vfio/iova_bitmap: refactor iova_bitmap_set() to better handle page boundaries")
Link: https://lore.kernel.org/r/20240202133415.23819-10-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

diff --git a/drivers/iommu/iommufd/iova_bitmap.c b/drivers/iommu/iommufd/iova_bitmap.c
index b370e8ee88665409a6b0a7eb7f42b6233a6b7a31..db8c46bee1559ac46fb148d2474668b5a994ae15 100644 (file)
--- a/drivers/iommu/iommufd/iova_bitmap.c
+++ b/drivers/iommu/iommufd/iova_bitmap.c
@@ -178,18 +178,19 @@ static int iova_bitmap_get(struct iova_bitmap *bitmap)
                               bitmap->mapped_base_index) *
                               sizeof(*bitmap->bitmap), PAGE_SIZE);
 
-       /*
-        * We always cap at max number of 'struct page' a base page can fit.
-        * This is, for example, on x86 means 2M of bitmap data max.
-        */
-       npages = min(npages,  PAGE_SIZE / sizeof(struct page *));
-
        /*
         * Bitmap address to be pinned is calculated via pointer arithmetic
         * with bitmap u64 word index.
         */
        addr = bitmap->bitmap + bitmap->mapped_base_index;
 
+       /*
+        * We always cap at max number of 'struct page' a base page can fit.
+        * This is, for example, on x86 means 2M of bitmap data max.
+        */
+       npages = min(npages + !!offset_in_page(addr),
+                    PAGE_SIZE / sizeof(struct page *));
+
        ret = pin_user_pages_fast((unsigned long)addr, npages,
                                  FOLL_WRITE, mapped->pages);
        if (ret <= 0)