You’re a legend
2 Likes
Never mind. I made a mistake. Forgot a long time ago I had overridden my render function to go in smaller chunks than the full buffer size. 
You shoud have a look at shared memory https://www.intel.com/content/www/us/en/developer/articles/technical/opencl-20-shared-virtual-memory-overview.html
Also this old but gold for a general overview on opencl memory spaces
https://www.codeproject.com/Articles/122405/Part-2-OpenCL-Memory-Spaces
1 Like
Never mind my mistake.
Should you check the error code of enqueueReadBuffer in case it’s not operating as expected and returning quickly ?
Sorry it was my mistake. Years ago I had overridden my renderNextBlock to go in smaller chunks while testing something so it was running way too often and I just remembered now. I think it will work once I get rid of that. Completely forgot I had done that.
Sorry for spamming nonsense. Back to it.