[PATCH 3/7] drm/gpusvm: mark pages as dirty

[Matthew Auld]

If the memory is going to be accessed by the device, make sure we mark
the pages accordingly such that the kernel knows this. This aligns with
the xe-userptr code.

Signed-off-by: Matthew Auld <matthew.auld at intel.com>
Cc: Thomas Hellström <thomas.hellstrom at linux.intel.com>
Cc: Matthew Brost <matthew.brost at intel.com>
---
 drivers/gpu/drm/drm_gpusvm.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c
index 7f1cf5492bba..5b4ecd36dff1 100644
--- a/drivers/gpu/drm/drm_gpusvm.c
+++ b/drivers/gpu/drm/drm_gpusvm.c
@@ -1471,6 +1471,7 @@ int drm_gpusvm_range_get_pages(struct drm_gpusvm *gpusvm,
 			pages[i] = page;
 		} else {
 			dma_addr_t addr;
+			unsigned int k;
 
 			if (is_zone_device_page(page) || zdd) {
 				err = -EOPNOTSUPP;
@@ -1489,6 +1490,14 @@ int drm_gpusvm_range_get_pages(struct drm_gpusvm *gpusvm,
 			range->dma_addr[j] = drm_pagemap_device_addr_encode
 				(addr, DRM_INTERCONNECT_SYSTEM, order,
 				 dma_dir);
+
+			for (k = 0; k < 1u << order; k++) {
+				if (!ctx->read_only)
+					set_page_dirty_lock(page);
+
+				mark_page_accessed(page);
+				page++;
+			}
 		}
 		i += 1 << order;
 		num_dma_mapped = i;
-- 
2.48.1