Gorokhov Dmitriy

@densamoilov In general the proposal looks good to me. Regarding opaque memory descriptor: acceptable options for OpenVINO are 1 and 2a. Option 2 is not acceptable at the same time...

@nshchego could you please start the review?

I see all comments are resolved. @nshchego Please update your review

@nshchego ping

@zhangYiIntel Please make CI green and I will merge

@usstq @maxnick Do you have any agreement on the solution?

@usstq I am just wondering is there a possibility to cover the issue by functional tests? It was caught by e2e testing so sounds like no specifics there.

@usstq As I can see the test was added. Can I proceed with the review?

@usstq just a kindly reminder

@iefode @maxnick @EgorDuplensky FYI