If we fail to recover the HW state upon resume (i.e. our attempt to clear the wedged bit and reset during i915_gem_sanitize() fails), then skip the HW restart inside i915_gem_init_hw(). We will ultimate do the the HW restart when sucessfully unwedgeding and reseting the HW later, but attempting to restore a wedged device upon resume is risky as the HW is in an unknown state.
Signed-off-by: Chris Wilson <[email protected]> Cc: Mika Kuoppala <[email protected]> --- drivers/gpu/drm/i915/i915_gem.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index d9d39b309ce8..5993222c81ae 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -4835,6 +4835,10 @@ int i915_gem_init_hw(struct drm_i915_private *dev_priv) init_unused_rings(dev_priv); BUG_ON(!dev_priv->kernel_context); + if (i915_terminally_wedged(&dev_priv->gpu_error)) { + ret = -EIO; + goto out; + } ret = i915_ppgtt_init_hw(dev_priv); if (ret) { -- 2.15.0.rc0 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
