V4L2SVDA/VAAPIVDA: use visible size from decoder and pass to client

media/gpu/h264_dpb.h

Index: media/gpu/h264_dpb.h
diff --git a/media/gpu/h264_dpb.h b/media/gpu/h264_dpb.h
index aef39a1bc5d517dcb86209b290671a856571ca0c..e8f119f6f6a1dd3983a3ff3e63b4ff0168b57073 100644
--- a/media/gpu/h264_dpb.h
+++ b/media/gpu/h264_dpb.h
@@ -15,6 +15,7 @@
 #include "base/macros.h"
 #include "base/memory/ref_counted.h"
 #include "media/filters/h264_parser.h"
+#include "ui/gfx/geometry/rect.h"
 
 namespace media {
 
@@ -84,6 +85,10 @@ class H264Picture : public base::RefCounted<H264Picture> {
   // Position in DPB (i.e. index in DPB).
   int dpb_position;
 
+  // The visible size of picture. This could be either parsed from SPS, or set
+  // to gfx::Rect(0, 0) for indicating invalid values or not available.
+  gfx::Rect visible_rect;
+
  protected:
   friend class base::RefCounted<H264Picture>;
   virtual ~H264Picture();