Commit 4a246cfc authored by Ben Widawsky's avatar Ben Widawsky Committed by Keith Packard

drm/i915: fix rc6 initialization on Ironlake

There is a race condition between setting PWRCTXA and executing
MI_SET_CONTEXT. PWRCTXA must not be set until a valid context has been
written (or else the GPU could possible go into rc6, and return to an
invalid context).
Reported-and-Tested-by: default avatarGu Rui <chaos.proton@gmail.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=28582Signed-off-by: default avatarBen Widawsky <ben@bwidawsk.net>
Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: default avatarKeith Packard <keithp@keithp.com>
parent 96f298aa
...@@ -7325,6 +7325,19 @@ void ironlake_enable_rc6(struct drm_device *dev) ...@@ -7325,6 +7325,19 @@ void ironlake_enable_rc6(struct drm_device *dev)
OUT_RING(MI_FLUSH); OUT_RING(MI_FLUSH);
ADVANCE_LP_RING(); ADVANCE_LP_RING();
/*
* Wait for the command parser to advance past MI_SET_CONTEXT. The HW
* does an implicit flush, combined with MI_FLUSH above, it should be
* safe to assume that renderctx is valid
*/
ret = intel_wait_ring_idle(LP_RING(dev_priv));
if (ret) {
DRM_ERROR("failed to enable ironlake power power savings\n");
ironlake_teardown_rc6(dev);
mutex_unlock(&dev->struct_mutex);
return;
}
I915_WRITE(PWRCTXA, dev_priv->pwrctx->gtt_offset | PWRCTX_EN); I915_WRITE(PWRCTXA, dev_priv->pwrctx->gtt_offset | PWRCTX_EN);
I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) & ~RCX_SW_EXIT); I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) & ~RCX_SW_EXIT);
mutex_unlock(&dev->struct_mutex); mutex_unlock(&dev->struct_mutex);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment