At least on i965g and i965gm, performing a device reset clobbers the IER
resulting in loss of interrupts thereafter. So, run the irq_postinstall
hook to restore them.
Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrj...@linux.intel.com>
---
This feels a bit hairy, so maybe limit it to
INTEL_INFO()->reset_clobbers_display? Or even introduce
reset_clobbers_interrupt with the same platforms as clobbers_display?
-Chris
---
drivers/gpu/drm/i915/i915_reset.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/i915_reset.c
b/drivers/gpu/drm/i915/i915_reset.c
index 5a067a4b3d5d..451233709d1a 100644
--- a/drivers/gpu/drm/i915/i915_reset.c
+++ b/drivers/gpu/drm/i915/i915_reset.c
@@ -675,8 +675,14 @@ static int gt_reset(struct drm_i915_private *i915,
unsigned int stalled_mask)
{
struct intel_engine_cs *engine;
enum intel_engine_id id;
+ unsigned long flags;
int err;
+ /* Restore IER as it may get clobbered on some platforms! (i965g[m]) */
+ spin_lock_irqsave(&i915->irq_lock, flags);
+ i915->drm.driver->irq_postinstall(&i915->drm);
+ spin_unlock_irqrestore(&i915->irq_lock, flags);
+
/*
* Everything depends on having the GTT running, so we need to start
* there.
@@ -690,7 +696,7 @@ static int gt_reset(struct drm_i915_private *i915, unsigned
int stalled_mask)
i915_gem_restore_fences(i915);
- return err;
+ return 0;
}
static void reset_finish_engine(struct intel_engine_cs *engine)
--
2.20.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
