[Intel-gfx] [PATCH 1/2] drm/i915/selftests: Calculate maximum ring size for preemption chain

[Caz Yokoyama]

Reviewed-by: Yokoyama, Caz <caz.yokoyama at intel.com>
-caz

On Thu, 2019-03-21 at 18:42 +0000, Chris Wilson wrote:
> Quoting Chris Wilson (2019-03-21 18:38:53)
> > Quoting Caz Yokoyama (2019-03-21 18:41:10)
> > > inline
> > > -caz
> > > On Thu, 2019-03-21 at 07:37 +0000, Chris Wilson wrote:
> > > > 32 is too many for the likes of kbl, and in order to insert
> > > > that many
> > > 
> > > Not only kbl. ring_size is 25 on my cfl.
> > > 
> > > > requests into the ring requires us to declare the first few
> > > > hung --
> > > 
> > > The hung is not caused by 32. It is caused by accumulation of
> > > requests
> > > for all prime numbers.
> > 
> > Sure, but the design of the test is that we don't care for more
> > than
> > ring_size.
> 
> And you can't have more than ring_size requests without hanging on
> the
> spinner. If we pick the maximum that allows for a prime number larger
> than ring_size, it hangs irrespective of whether or not you remember
> to
> flush the lo.ctx in between.
> -Chris