Changes to let KFD use render node VMs

[Christian König]

Hi Felix,

patch #3 looks good to me in general, but why do you need to cache the 
pd_phys_addr?

The BO already does that anyway, so you can just use 
amdgpu_bo_gpu_addr() for this which also makes some additional checks on 
debug builds.

patch #5 well mixing power management into the VM functions is a clear NAK.

That certainly doesn't belong there, but we can have a counter how many 
compute VMs we have in the manager. amdgpu_vm_make_compute() can then 
return if this was the first VM to became a compute VM or not.

The rest of the patch looks good to me.

Regards,
Christian.

Am 01.03.2018 um 23:58 schrieb Felix Kuehling:
> Hi Christian,
>
> I have a working patch series against amd-kfg-staging that lets KFD use
> VMs from render node FDs, as we discussed. There are two patches in that
> series that touch amdgpu_vm.[ch] that I'd like your feedback on before I
> commit the changes to amd-kfd-staging and include them in my upstream
> patch series for KFD GPUVM support. See attached.
>
> Thanks,
>    Felix
>