My understanding is that 2D acceleration was made possible with support from AMD. My point was that adding OpenCL to open source drivers would require AMD assistance (assistance they have so far been unwilling to provide) not that it would be impossible.
Now that I can agree with.
The work required to add OpenCL support is pretty massive, and AFAIK there's pretty much 0 done on a open source OpenCL -> VLIWx shader asm compiler, several other large parts of the puzzle are also completely missing (runtime, runtime/opengl/driver integration/...).
But saying that the open source driver can't use shaders is wrong.
Now, for something like a dedicated miner for 5/6xxx you'd "only" have to handcraft a kernel for each arch in VLIWx ASM (fully documented in public AMD docs btw) and hack the radeon driver so it allocates input/output regions and loads/runs your shader program on command. Iirc there's still a open 200BTC bounty for a miner working like that. but it's a *lot* of work for very little benefit.
I think I understand. So while it wouldn't be possible to implement OpenCL (which is rather expansive API) hack together low level support just for mining and implement that in a fork of OpenSource drivers? The code wouldn't be openCL compatible but would function at a lower level.
Has anyone done a proof of concept? Something simpler than mining maybe something as simple as loading an value, increment it in a loop and then read the output?
I never thought about doing it that way. Hopefully someday AMD provides support for OpenCL in OpenSource drivers as that is a pretty ugly "hack".