drm/i915: Check against the signaled bit for fences/requests

When dma_fence_signal() is called, it sets a flag to indicate the fence
is complete. Before the dma_fence is signaled, the seqno check will
first be passed. During an unlocked check (such as inside a waiter), it
is possible for the fence to be signaled even though the seqno has been
reset (by engine wraparound). In this case the waiter will be kicked,
but for an extra layer of protection we can check the persistent
signaled bit from the fence.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170223074422.4125-2-chris@chris-wilson.co.uk

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index e89f5647675978acc552f05d42016b5e05259669..9220a28fbaa87692816a0e95a01a16bf43dc3025 100644 (file)
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -4013,6 +4013,15 @@ __i915_request_irq_complete(struct drm_i915_gem_request *req)
 {
        struct intel_engine_cs *engine = req->engine;
 
+       /* Note that the engine may have wrapped around the seqno, and
+        * so our request->global_seqno will be ahead of the hardware,
+        * even though it completed the request before wrapping. We catch
+        * this by kicking all the waiters before resetting the seqno
+        * in hardware, and also signal the fence.
+        */
+       if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &req->fence.flags))
+               return true;
+
        /* Before we do the heavier coherent read of the seqno,
         * check the value (hopefully) in the CPU cacheline.
         */