Ben Ashbaugh

Can we call this a bugfix and just add the `const` qualifier to the `atomic_load` functions? I think that was almost certainly the original intent, although the const qualifier has...

We need to support OpFRem and OpFMod because these are standard SPIR-V instructions, so removing them is not an option. We do have options how they should behave, though. ####...

Yeah, our implementation of the precise fmod is similar (we don't have a dedicated HW instruction). In some ways the only reason we're having this discussion is because we currently...

Discussed in the November 29th teleconference. We're trending towards matching Vulkan for the SPIR-V `OpFMod` and `OpFRem`. I will make a PR to clarify this in the OpenCL SPIR-V Environment...

I've converted this to a draft while we determine whether there is a precision issue or not.

Feedback from discussion on April 11th: it would be best to have a test for the desired behavior first.

This is a really great question! One of the nice properties (at least, from an implementor's perspective) of the current command buffer extension is that a command buffer is entirely...

I lost track of this one, my apologies. Is it ready to go?

One more: * `subdevice` or `sub-device`

Another: * `out of order` or `out-of-order`