[Intel-gfx] [PATCH 04/11] drm/i915/gtt: Markup i915_ppgtt depth

[Mika Kuoppala]

Chris Wilson <chris at chris-wilson.co.uk> writes:

> Quoting Mika Kuoppala (2019-07-10 15:50:37)
>> Chris Wilson <chris at chris-wilson.co.uk> writes:
>> 
>> > Quoting Mika Kuoppala (2019-07-10 15:25:38)
>> >> Chris Wilson <chris at chris-wilson.co.uk> writes:
>> >> 
>> >> > Quoting Mika Kuoppala (2019-07-10 09:17:27)
>> >> >> Chris Wilson <chris at chris-wilson.co.uk> writes:
>> >> >> 
>> >> >> > This will be useful to consolidate recursive code.
>> >> >> >
>> >> >> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
>> >> >> > ---
>> >> >> >  drivers/gpu/drm/i915/i915_gem_gtt.c | 3 +++
>> >> >> >  drivers/gpu/drm/i915/i915_gem_gtt.h | 1 +
>> >> >> >  2 files changed, 4 insertions(+)
>> >> >> >
>> >> >> > diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
>> >> >> > index da4db76ce054..2fc60e8acd9a 100644
>> >> >> > --- a/drivers/gpu/drm/i915/i915_gem_gtt.c
>> >> >> > +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
>> >> >> > @@ -1537,6 +1537,8 @@ static void ppgtt_init(struct i915_ppgtt *ppgtt, struct intel_gt *gt)
>> >> >> >       ppgtt->vm.vma_ops.unbind_vma  = ppgtt_unbind_vma;
>> >> >> >       ppgtt->vm.vma_ops.set_pages   = ppgtt_set_pages;
>> >> >> >       ppgtt->vm.vma_ops.clear_pages = clear_pages;
>> >> >> > +
>> >> >> > +     ppgtt->vm.top = i915_vm_is_4lvl(&ppgtt->vm) ? 3 : 2;
>> >> >> 
>> >> >> Perhaps it becomes evident later in the series why top and
>> >> >> not level, so these would be 4 and 3 here.
>> >> >
>> >> > Because we use top and not level :)
>> >> 
>> >> You make me substract one with my biological processor.
>> >> It is hard.
>> >> 
>> >> Please do remake the i915_vm_is_4lvl() and include.
>> >
>> > I'm tempted to put the gtt_depth in the device info.
>> >
>> > How do you want i915_vm_is_4lvl() remade? The special case going
>> > forward is really is_3lvl?
>> 
>> No strong feelings here. How about i915_vm_get_lvl(vm)
>> { return top + 1; } ?
>
> Who's going to be calling get_lvl() though? The one time where it might
> be useful, we just use "<= top" instead.

Hmm right, prolly way too generic to query for lvl and compare. So
then it is the one which reads best on the few callsites it
will sit on.

-Mika