VDA implementation for Exynos, using V4L2

content/public/common/content_switches.cc

Index: content/public/common/content_switches.cc
diff --git a/content/public/common/content_switches.cc b/content/public/common/content_switches.cc
index aaa236d920df661c4d199205c6cca91cca62a952..19e43d858d4188a449906bb0f4d832daab2d6671 100644
--- a/content/public/common/content_switches.cc
+++ b/content/public/common/content_switches.cc
@@ -760,6 +760,9 @@ const char kDefaultTileHeight[]             = "default-tile-height";
 const char kMaxUntiledLayerWidth[]          = "max-untiled-layer-width";
 const char kMaxUntiledLayerHeight[]         = "max-untiled-layer-height";
 
+// Use ExynosVideoDecodeAccelerator for video decode (instead of SECOMX)
+const char kUseExynosVda[]                  = "use-exynos-vda";

[piman, 2013/01/12 03:24:58]

Note: it would be nice at some point to have a selection switch
--use-vda=exynos|omx|libva|dxva|... like we have --use-gl, rather than an
explosion of switches with undocumented precedence as we add more paths. Not
necessarily for this CL, but maybe as a follow-up?

[sheu, 2013/01/14 23:49:49]

This is just a temporary flag, which I intend to remove as soon as we finish the
switchover from sec_omx.

+
 const char kEnableFixedPositionCreatesStackingContext[]
     = "enable-fixed-position-creates-stacking-context";
 const char kDisableFixedPositionCreatesStackingContext[]