[Intel-gfx] [PATCH] drm/i915: Perform GGTT restore much earlier during resume

Chris Wilson chris at chris-wilson.co.uk
Sat Sep 7 08:35:13 UTC 2019


As soon as we re-enable the various functions within the HW, they may go
off and read data via a GGTT offset. Hence, if we have not yet restored
the GGTT PTE before then, they may read and even *write* random locations
in memory.

Detected by DMAR faults during resume.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala at linux.intel.com>
Cc: Martin Peres <martin.peres at linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
Cc: stable at vger.kernel.org
---
 drivers/gpu/drm/i915/gem/i915_gem_pm.c | 3 ---
 drivers/gpu/drm/i915/i915_drv.c        | 5 +++++
 2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pm.c b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
index b3993d24b83d..9b1129aaacfe 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pm.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
@@ -242,9 +242,6 @@ void i915_gem_resume(struct drm_i915_private *i915)
 	mutex_lock(&i915->drm.struct_mutex);
 	intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
 
-	i915_gem_restore_gtt_mappings(i915);
-	i915_gem_restore_fences(i915);
-
 	if (i915_gem_init_hw(i915))
 		goto err_wedged;
 
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 7b2c81a8bbaa..1af4eba968c0 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -1877,6 +1877,11 @@ static int i915_drm_resume(struct drm_device *dev)
 	if (ret)
 		DRM_ERROR("failed to re-enable GGTT\n");
 
+	mutex_lock(&dev_priv->drm.struct_mutex);
+	i915_gem_restore_gtt_mappings(dev_priv);
+	i915_gem_restore_fences(dev_priv);
+	mutex_unlock(&dev_priv->drm.struct_mutex);
+
 	intel_csr_ucode_resume(dev_priv);
 
 	i915_restore_state(dev_priv);
-- 
2.23.0



More information about the Intel-gfx mailing list