drm/i915: kill set_need_resched

danvet · danvet · commit 571c608d06df · 2013-09-12T22:40:36.000+02:00
This is just a remnant from the old days when our reset handling was
horribly racy, suffered from terribly locking issues and often happily
live-locked. Those days are now gone so we can drop the hacks and just
rip the reschedule-point out.

Reported-by: Peter Zijlstra &lt;peterz@infradead.org&gt;
Cc: Peter Zijlstra &lt;peterz@infradead.org&gt;
Cc: Linux Kernel Mailing List &lt;linux-kernel@vger.kernel.org&gt;
Reviewed-by: Chris Wilson &lt;chris@chris-wilson.co.uk&gt;
Acked-by: Peter Zijlstra &lt;peterz@infradead.org&gt;
Signed-off-by: Daniel Vetter &lt;daniel.vetter@ffwll.ch&gt;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
@@ -1390,14 +1390,11 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		if (i915_terminally_wedged(&dev_priv->gpu_error))
 			return VM_FAULT_SIGBUS;
 	case -EAGAIN:
-		/* Give the error handler a chance to run and move the
-		 * objects off the GPU active list. Next time we service the
-		 * fault, we should be able to transition the page into the
-		 * GTT without touching the GPU (and so avoid further
-		 * EIO/EGAIN). If the GPU is wedged, then there is no issue
-		 * with coherency, just lost writes.
+		/*
+		 * EAGAIN means the gpu is hung and we'll wait for the error
+		 * handler to reset everything when re-faulting in
+		 * i915_mutex_lock_interruptible.
 		 */
-		set_need_resched();
 	case 0:
 	case -ERESTARTSYS:
 	case -EINTR:
