Brice Videau

Hello, I have the game as well. I took a look a long time ago, but the format was very different from Platinum's other game. And thinking about it, it...

Thanks for this contribution. I don't know why the CI didn't trigger here. From what I understand it should have. To get around the issue, I pushed the branch on...

I think we should develop and advertise solutions like this one: https://github.com/KhronosGroup/OpenCL-Headers/pull/169 That could serve the same purpose, while being more future proof. @bashbaug has made similar proposals as well...

Thanks for implementing this @bashbaug, and for raising the very good questions. I think maybe a good way to look at it is through usage scenarios. I see two uses...

I see a problem with 2 & 3 in a scenario with layers. When freeing the layer memory, the dispatch tables that were passed to the layers are freed, so...

> Sure, I can do this. To be clear, to release in reverse order I would: > > 1. Free layers first, from the front of the list to the...

@bashbaug I though a bit more about the layer issues. Even if we do not release the layer library handle, there is no guaranty a layer could be reinitialized without...

I think it depends where you want to take it. If you want to enable the extension function, as in your original plan, then it may alleviate some failures, but...

Thanks for reporting the issue. The termination of the loader is an open problem. I'll try to summarize the issues here: - We leak memory, - Libraries are not unloaded...

I think you can also override the CMake variable `OPENCL_ICD_LOADER_HEADERS_DIR` to specify an alternate location.