-
Notifications
You must be signed in to change notification settings - Fork 943
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Interop with underlying graphics API #965
Comments
There isn't currently any mechanism to do so. I assume that you are talking about native-only support, outside of WebGPU proper. In this case, first step would be implementing the "external memory objects" in gfx-hal. Then we'd be able to hook it up here. |
Is there any progress on this issue? My use case is the following: I want to do video processing on the GPU, I have the textures from OpenGL, Metal, DirectX and Vulkan and I would like to pass them to wgpu, process the pixels and draw it back, all without going through CPU memory. I use Qt and it supports different graphic APIs via the QRhi, so it'd be nice to be able to support the same graphic APIs to do the processing directly, and for that I need to be able to feed the |
There is quite a bit of progress overall, although more may be needed. |
I have a use case where i would like to export a wgpu texture, so that i can use it in opengl. meanwhile, opengl works great already for transparency. but ofcourse, all the unsafe and thread-safety of openGL :( is horrible to deal with in rust. my plan is to
I can then release the above opengl-wgpu hybrid version, and a normal full wgpu version. those who have issues with vk drivers can use hybrid version. this way, i can stay with wgpu 90% of the time and only use opengl in specific / backwards compatible places, waiting to deprecate it as issues are fixed or wgpu getting GL backend. |
Note that wgpu does have an OpenGL backend, so you shouldn't need to share between native apis at all, you can just run us from that same context |
that's nice. but i need vulkan backend.
|
#1460 might be a duplicate of this |
Closing in favor of #4067 which is a concrete plan |
Is there a user-facing mechanism for accessing the native API handles underlying given wgpu objects so that interop can be performed? If not, are there any plans for such an escape hatch? Specifically I'm interested in using external memory objects from Vulkan.
The text was updated successfully, but these errors were encountered: