drm/vmwgfx: Fix cpu updates of coherent multisample surfaces

In cases where the dirty linear memory range spans multiple sample sheets
in a surface, the dirty surface region is incorrectly computed.
To do this correctly and in an optimized fashion  we would have to compute
the dirty region of each sample sheet and compute the union of those
regions.

But assuming that cpu writing to a multisample surface is rather a corner
case than a common case, just set the dirty region to the full surface.

This fixes OpenGL piglit errors with SVGA_FORCE_COHERENT=1
and the piglit test:

fbo-depthstencil blit default_fb -samples=2 -auto

Fixes: 9ca7d19f ("drm/vmwgfx: Add surface dirty-tracking callbacks")
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210505035740.286923-4-zackr@vmware.com

drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h

@@ -1467,6 +1467,7 @@ struct svga3dsurface_cache {
 
 /**
  * struct svga3dsurface_loc - Surface location
+ * @sheet: The multisample sheet.
  * @sub_resource: Surface subresource. Defined as layer * num_mip_levels +
  * mip_level.
  * @x: X coordinate.
@@ -1474,6 +1475,7 @@ struct svga3dsurface_cache {
  * @z: Z coordinate.
  */
 struct svga3dsurface_loc {
+	u32 sheet;
 	u32 sub_resource;
 	u32 x, y, z;
 };
@@ -1566,8 +1568,8 @@ svga3dsurface_get_loc(const struct svga3dsurface_cache *cache,
 	u32 layer;
 	int i;
 
-	if (offset >= cache->sheet_bytes)
-		offset %= cache->sheet_bytes;
+	loc->sheet = offset / cache->sheet_bytes;
+	offset -= loc->sheet * cache->sheet_bytes;
 
 	layer = offset / cache->mip_chain_bytes;
 	offset -= layer * cache->mip_chain_bytes;
@@ -1631,6 +1633,7 @@ svga3dsurface_min_loc(const struct svga3dsurface_cache *cache,
 		      u32 sub_resource,
 		      struct svga3dsurface_loc *loc)
 {
+	loc->sheet = 0;
 	loc->sub_resource = sub_resource;
 	loc->x = loc->y = loc->z = 0;
 }
@@ -1652,6 +1655,7 @@ svga3dsurface_max_loc(const struct svga3dsurface_cache *cache,
 	const struct drm_vmw_size *size;
 	u32 mip;
 
+	loc->sheet = 0;
 	loc->sub_resource = sub_resource + 1;
 	mip = sub_resource % cache->num_mip_levels;
 	size = &cache->mip[mip].size;

drivers/gpu/drm/vmwgfx/vmwgfx_surface.c

@@ -1804,6 +1804,19 @@ static void vmw_surface_tex_dirty_range_add(struct vmw_resource *res,
 	svga3dsurface_get_loc(cache, &loc2, end - 1);
 	svga3dsurface_inc_loc(cache, &loc2);
 
+	if (loc1.sheet != loc2.sheet) {
+		u32 sub_res;
+
+		/*
+		 * Multiple multisample sheets. To do this in an optimized
+		 * fashion, compute the dirty region for each sheet and the
+		 * resulting union. Since this is not a common case, just dirty
+		 * the whole surface.
+		 */
+		for (sub_res = 0; sub_res < dirty->num_subres; ++sub_res)
+			vmw_subres_dirty_full(dirty, sub_res);
+		return;
+	}
 	if (loc1.sub_resource + 1 == loc2.sub_resource) {
 		/* Dirty range covers a single sub-resource */
 		vmw_subres_dirty_add(dirty, &loc1, &loc2);