How/When is 'mfxU16 mfxFrameData.Locked;' cleared?

How/When is 'mfxU16 mfxFrameData.Locked;' cleared?

 Question regarding field mfxU16 mfxFrameData.Locked;

I have a question about how/when mfxFrameData::Locked is cleared in the tutorials.

I can clearly see it being set, usually in  GetFreeSurfaceIndex(...),
but it is not clear when it is being cleared.
It seems as though the IMSDK is clearing mfxFrameData.Locked

My question, is how/when is mfxFrameData.Locked cleared?

It is cleared completely asynchronously by the IMSDK?
or 
Is it cleared by MFXVideoCORE_SyncOperation() ? [or another call?]

It seems to me if the IMSDK cleared mfxFrameData.Locked completely asynchronously, without regard to the
application, GetFreeSurfaceIndex() [in the samples] would be returning surfaces that have not yet had the frame data retrieved 
by the application, in effect trashing valid frame data. 

Thank you
Cameron

 

5 posts / 0 new
Last post
For more complete information about compiler optimizations, see our Optimization Notice.

mfxFrameData.Locked is cleared asynchronously by the IMSDK, when that's frame data became needless for encoder/decoder/etc.

Let me also add that "GetFreeSurfaceIndex()" does NOT modify the Locked parameter. The value is modified by SDK at the moment the surface is submitted to an SDK component (for instance via EncodeFrameAsync). And as dj_alek noted, Locked gets set to 0 again by the SDK asynchronously when the surface has been fully processed, thus allowing to to be reused.

What is the purpose of mfxU8/mfxU16/mfxU32? Do these serve same or similar purposes or are they completely different? Please provide a basic idea. I couldn't find their explanation in the documentation. TIA

The purpose of these types is to align the type to different pixel format, for example, align mfxU32 to RGBA format.

Mark

Leave a Comment

Please sign in to add a comment. Not a member? Join today