[Intel-gfx] [PATCH] drm/i915: Always call fence-lost prior to removing the fence

Chris Wilson chris at chris-wilson.co.uk
Tue Mar 26 11:39:30 CET 2013


There is a minute window for a race between put-fence removing the fence
and for a new transaction by an external party on the GTT mmap. That is
we must zap the mmap prior to removing the fence and not afterwards.

Fixes regression from
commit 61050808bb019ebea966b7b5bfd357aaf219fb51
Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Tue Apr 17 15:31:31 2012 +0100

    drm/i915: Refactor put_fence() to use the common fence writing routine

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
Cc: stable at vger.kernel.org # regression introduced in v3.5
---
 drivers/gpu/drm/i915/i915_gem.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index a92d431..c9f9a92 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2124,11 +2124,11 @@ static void i915_gem_reset_fences(struct drm_device *dev)
 	for (i = 0; i < dev_priv->num_fence_regs; i++) {
 		struct drm_i915_fence_reg *reg = &dev_priv->fence_regs[i];
 
-		i915_gem_write_fence(dev, i, NULL);
-
 		if (reg->obj)
 			i915_gem_object_fence_lost(reg->obj);
 
+		i915_gem_write_fence(dev, i, NULL);
+
 		reg->pin_count = 0;
 		reg->obj = NULL;
 		INIT_LIST_HEAD(&reg->lru_list);
@@ -2760,10 +2760,10 @@ i915_gem_object_put_fence(struct drm_i915_gem_object *obj)
 	if (obj->fence_reg == I915_FENCE_REG_NONE)
 		return 0;
 
+	i915_gem_object_fence_lost(obj);
 	i915_gem_object_update_fence(obj,
 				     &dev_priv->fence_regs[obj->fence_reg],
 				     false);
-	i915_gem_object_fence_lost(obj);
 
 	return 0;
 }
-- 
1.7.10.4




More information about the Intel-gfx mailing list