Re: [PATCH v27 0/4] MR346: tests/shader-runner: Test DXIL shaders.

[Henri Verbeet (＠hverbeet)]

> Oh I see now, Henri specifically requested the opposite.

I don't know if it's exactly the opposite, but if we're going to have "notimpl", I think that should have a consistent meaning across backends, yes. We could certainly decide that we don't care to distinguish between E_NOTIMPL and E_FAIL in general though, and just check for "FAILED(hr)" for "fail", but that seems like a change somewhat orthogonal to this MR.