Mateusz Jablonski

@jszuppe could you look at this PR?

Hi @kylelutz, the problem is scenario with sub devices. The current implementation is that boost queries number of context devices, then creates a vector of `compute::device` (default constructor of `compute::device`...

@kylelutz @jszuppe any comments?

@kylelutz ping

@kylelutz @jszuppe ping

kindly reminder @kylelutz

@kylelutz @jszuppe could you merge this PR?

@eero-t For Linux, renderDXX and cardX may be associated with same GPU but AFAIK each of them has different privileges. We use renderDXXX as UMD shouldn't use cardX. I don't...

I see that AMD plugin BDF implementation is strongly vendor specific as it bases on `/sys/module/amdgpu/drivers`. However, isn't BDF the most generic mechanism, which is independent on vendor? e.g. for...

Once we already expose BDF, would that mechanism work for you?