(adding the list to CC) Hi Andreas,
What do you mean exactly by "context should be created to encompass the subregions"? At the moment of context creation the subregions don't exist yet. On Thu, Jan 16, 2020 at 9:55 PM Andreas Kloeckner <li...@informa.tiker.net> wrote: > Bogdan Opanchuk <bog...@opanchuk.net> writes: > > > Hello all, > > > > I have an old macbook that has a discrete GeForce on it, and have run > into > > the following problem. The simplified example is here: > > https://gist.github.com/fjarri/9aff0474868e2faf438f7e8229d194ec > > > > Basically, what I'm trying to do: > > - create a two-device context > > - create a buffer > > - split it into two subregions to use on each device > > - run a kernel on each device in parallel working with the corresponding > > subregion > > - get the result back on the host > > (the expected result is [0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7]) > > > > First, it turned out that if the context includes an nVidia card, the > > Buffer must be necessarily created with the cl.mem_flags.ALLOC_HOST_PTR > > flag, otherwise if one uses its subregion in a kernel, the program > crashes. > > If the context is created on a CPU + Iris Pro (the other two devices > > available), everything works fine without this flag, giving the expected > > result. > > > > After fixing that, the program finishes without crashing when run on a > CPU > > + GeForce or Iris Pro + GeForce context, but the result is [0 1 2 3 4 5 > 6 7 > > 0 0 0 0 0 0 0 0] - that is, the second kernel (on the GeForce device) > > either did not run, or its changes to the subregion were not incorporated > > into the whole buffer. Uncommenting the explicit migration in the end > does > > not help either. Does anyone know what I'm missing here? Or is it an > > nVidia/Apple bug? > > I feel like maybe the context should be created to encompass the > subregions? I've never used these features much, so that's a bit of a > shot in the dark. > > HTH, > Andreas > >
_______________________________________________ PyOpenCL mailing list -- pyopencl@tiker.net To unsubscribe send an email to pyopencl-le...@tiker.net
