Backmerge tag 'v4.16-rc7' into drm-next
Linux 4.16-rc7 This was requested by Daniel, and things were getting a bit hard to reconcile, most of the conflicts were trivial though.
This commit is contained in:
@@ -1338,6 +1338,19 @@ static void __vmw_svga_disable(struct vmw_private *dev_priv)
|
||||
*/
|
||||
void vmw_svga_disable(struct vmw_private *dev_priv)
|
||||
{
|
||||
/*
|
||||
* Disabling SVGA will turn off device modesetting capabilities, so
|
||||
* notify KMS about that so that it doesn't cache atomic state that
|
||||
* isn't valid anymore, for example crtcs turned on.
|
||||
* Strictly we'd want to do this under the SVGA lock (or an SVGA mutex),
|
||||
* but vmw_kms_lost_device() takes the reservation sem and thus we'll
|
||||
* end up with lock order reversal. Thus, a master may actually perform
|
||||
* a new modeset just after we call vmw_kms_lost_device() and race with
|
||||
* vmw_svga_disable(), but that should at worst cause atomic KMS state
|
||||
* to be inconsistent with the device, causing modesetting problems.
|
||||
*
|
||||
*/
|
||||
vmw_kms_lost_device(dev_priv->dev);
|
||||
ttm_write_lock(&dev_priv->reservation_sem, false);
|
||||
spin_lock(&dev_priv->svga_lock);
|
||||
if (dev_priv->bdev.man[TTM_PL_VRAM].use_type) {
|
||||
|
Reference in New Issue
Block a user