[Intel-gfx] [PATCH 02/11] drm/i915/fence: Avoid del_timer_sync() from inside a timer

[Chris Wilson]

Quoting Tvrtko Ursulin (2017-09-19 11:19:25)
> 
> On 11/09/2017 09:41, Chris Wilson wrote:
> > A fence may be signaled from any context, including from inside a timer.
> > One example is timer_i915_sw_fence_wake() which is used to provide a
> > safety-net when waiting on an external fence. If the external fence is
> > not signaled within a timely fashion, we signal our fence on its behalf,
> > and so we then may process subsequent fences in the chain from within
> > that timer context.
> > 
> > Given that dma_i915_sw_fence_cb() may be from inside a timer, we cannot
> 
> s/dma_i915_sw_fence_cb/dma_i915_sw_fence_wake/ ?
> 
> > then use del_timer_sync() as that requires the timer lock for itself. To
> > circumvent this, while trying to keep the signal propagation as low
> > latency as possible, move the completion into a worker and use a bit of
> > atomic switheroo to serialise the timer-callback and the dma-callback.
> > 
> > Testcase: igt/gem_eio/in-flight-external
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>

> Okay it took me a few minutes to figure it out. Looks correct.
> 
> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>

Ta, fixed up the bad reference and pushed this along with the igt that
was happily deadlocking the kernel.
-Chris