Pass decoded picture size from VDA to client

media/filters/gpu_video_decoder.cc

Index: media/filters/gpu_video_decoder.cc
diff --git a/media/filters/gpu_video_decoder.cc b/media/filters/gpu_video_decoder.cc
index 16a33eae54c6fc3cdff4f65b09e773af14392a01..bac46994b57bfb4aefd86c22c1d9f5cba980fead 100644
--- a/media/filters/gpu_video_decoder.cc
+++ b/media/filters/gpu_video_decoder.cc
@@ -53,9 +53,8 @@ GpuVideoDecoder::PendingDecoderBuffer::PendingDecoderBuffer(
 GpuVideoDecoder::PendingDecoderBuffer::~PendingDecoderBuffer() {}
 
 GpuVideoDecoder::BufferData::BufferData(
-    int32 bbid, base::TimeDelta ts, const gfx::Rect& vr, const gfx::Size& ns)
-    : bitstream_buffer_id(bbid), timestamp(ts), visible_rect(vr),
-      natural_size(ns) {
+    int32 bbid, base::TimeDelta ts)
+    : bitstream_buffer_id(bbid), timestamp(ts) {
 }
 
 GpuVideoDecoder::BufferData::~BufferData() {}
@@ -276,9 +275,7 @@ void GpuVideoDecoder::Decode(const scoped_refptr<DecoderBuffer>& buffer,
 void GpuVideoDecoder::RecordBufferData(const BitstreamBuffer& bitstream_buffer,
                                        const DecoderBuffer& buffer) {
   input_buffer_data_.push_front(BufferData(bitstream_buffer.id(),
-                                           buffer.timestamp(),
-                                           config_.visible_rect(),
-                                           config_.natural_size()));
+                                           buffer.timestamp()));
   // Why this value?  Because why not.  avformat.h:MAX_REORDER_DELAY is 16, but
   // that's too small for some pathological B-frame test videos.  The cost of
   // using too-high a value is low (192 bits per extra slot).
@@ -289,17 +286,13 @@ void GpuVideoDecoder::RecordBufferData(const BitstreamBuffer& bitstream_buffer,
     input_buffer_data_.pop_back();
 }
 
-void GpuVideoDecoder::GetBufferData(int32 id, base::TimeDelta* timestamp,
-                                    gfx::Rect* visible_rect,
-                                    gfx::Size* natural_size) {
+void GpuVideoDecoder::GetBufferData(int32 id, base::TimeDelta* timestamp) {
   for (std::list<BufferData>::const_iterator it =
            input_buffer_data_.begin(); it != input_buffer_data_.end();
        ++it) {
     if (it->bitstream_buffer_id != id)
       continue;
     *timestamp = it->timestamp;
-    *visible_rect = it->visible_rect;
-    *natural_size = it->natural_size;
     return;
   }
   NOTREACHED() << "Missing bitstreambuffer id: " << id;
@@ -424,10 +417,9 @@ void GpuVideoDecoder::PictureReady(const media::Picture& picture) {
 
   // Update frame's timestamp.
   base::TimeDelta timestamp;
-  gfx::Rect visible_rect;
-  gfx::Size natural_size;
-  GetBufferData(picture.bitstream_buffer_id(), &timestamp, &visible_rect,
-                &natural_size);
+  gfx::Rect visible_rect(picture.size());

[Pawel Osciak, 2014/08/10 00:02:22]

Please do not make any changes to this class. Unfortunately some of the VDAs,
notably DXVA and AVDA, don't distinguish between visible size and coded size, so
this would result in us using coded size for them here, which is not what we
want. We should keep using config information.

The only advantage of using the VDA-provided size here over the config size
would be if we were decoding a resolution changing stream without a container
(or with a broken container), but the former does not happen when we use this
class anyway, and the latter we can't help given the above.

Please only document that we are dropping picture.size() for the above reasons:
"Some of the VDAs don't support and thus don't provide us with visible size,
passing coded size instead, so drop this and use config information instead."

[kcwu, 2014/08/12 04:48:06]

Done.

+  gfx::Size natural_size(picture.size());
+  GetBufferData(picture.bitstream_buffer_id(), &timestamp);
   DCHECK(decoder_texture_target_);
 
   scoped_refptr<VideoFrame> frame(VideoFrame::WrapNativeTexture(