Dzmitry Malyshau

> If building with the 'metal' command line tool, add the options '-gline-tables-only' and '-MO' to your compilation step.

I looked more at this, and concluded that there is no official mechanism to achieve this when going through `newLibraryWithSource`. Filed FB8247902 to Apple (and added to our list of...

I don't think it makes sense to patch the released GL backend for this, since the backend is heavily in flux (especially for the Web target), and not yet ready...

Found something in https://docs.nvidia.com/drive/active/5.1.6.0L/nvvib_docs/DRIVE_OS_Linux_SDK_Development_Guide/baggage/GL_EXT_buffer_storage.html > What commands are affected by the relaxed errors for persistently mapped buffers? > RESOLVED: In GL 4.5 the following commands have the relaxed language BufferSubData,...

Ok, so this is valid when we have `private_caps.buffer_storage` enabled. However, when we don't, `gfx-memory` still treats us as persistently mappable... This may be solved with https://github.com/gfx-rs/wgpu/pull/1015 In the meantime,...

Wow, that's quite something... I don't know right off the bat how we should be handling this. gfx doesn't control threading, and we can't launch a helper thread just to...

This appears to be more problematic than I thought. The `wait_events` are supposed to receive a list of the sets of barriers now, as opposed to the just a set...

I'm not sure about `api: ll` feature. Everything non-deprecated is basically `ll` by default.

Yes, at this point we should just go ahead and close them.

Quad example doesn't properly synchronize things. Could you instead try running [wgpu-rs examples](https://github.com/gfx-rs/wgpu-rs) and report if it's also a problem? (to clarify, wgpu uses gfx-rs with the new swapchain model,...