[Intel-gfx] [PATCH 2/2] drm/i915/gt: Yield the timeslice if waiting on a semaphore

[Chris Wilson]

Quoting Chris Wilson (2020-01-06 11:21:26)
> If we find ourselves waiting on a MI_SEMAPHORE_WAIT, either within the
> user batch or in our own preamble, the engine raises a
> GT_WAIT_ON_SEMAPHORE interrupt. We can unmask that interrupt and so
> respond to a semaphore wait by yielding the timeslice (if we have
> another process to yield to!)
> 
> The only real complication is that the interrupt is only generated for
> the start of the semaphore wait, and is asynchronous to our
> process_csb() -- that is we may not have registered the timeslice before
> we see the interrupt. To ensure we don't miss a potential semaphore
> (e.g. selftests/live_timeslice_preempt) we mark the interrupt and apply
> it to the next timeslice regardless of whether it was active at the
> time.

Sigh. Preempt-to-busy uses a semaphore so if we do have multiple
contexts to slice amongst, we keep on switching as we charge the
preempt-to-busy semaphore against the next.

Left in a bit of dilemma, we can excuse the preempt-to-busy, but then we
are liable to miss a genuine stall (e.g. live_timeslice_preempt).
-Chris