drm/xe: Open-code GGTT MMIO access protection

GGTT MMIO access is currently protected by hotplug (drm_dev_enter),
which works correctly when the driver loads successfully and is later
unbound or unloaded. However, if driver load fails, this protection is
insufficient because drm_dev_unplug() is never called.

Additionally, devm release functions cannot guarantee that all BOs with
GGTT mappings are destroyed before the GGTT MMIO region is removed, as
some BOs may be freed asynchronously by worker threads.

To address this, introduce an open-coded flag, protected by the GGTT
lock, that guards GGTT MMIO access. The flag is cleared during the
dev_fini_ggtt devm release function to ensure MMIO access is disabled
once teardown begins.

Cc: stable@vger.kernel.org
Fixes: 919bb54e989c ("drm/xe: Fix missing runtime outer protection for ggtt_remove_node")
Reviewed-by: Zhanjun Dong <zhanjun.dong@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260310225039.1320161-8-zhanjun.dong@intel.com

1 files changed, 9 insertions, 6 deletions

diff --git a/drivers/gpu/drm/xe/xe_ggtt.c b/drivers/gpu/drm/xe/xe_ggtt.c
index 0f2e3af49912..21071b64b09d 100644
--- a/drivers/gpu/drm/xe/xe_ggtt.c
+++ b/drivers/gpu/drm/xe/xe_ggtt.c
@@ -66,6 +66,9 @@
  * give us the correct placement for free.
  */
 
+#define XE_GGTT_FLAGS_64K	BIT(0)
+#define XE_GGTT_FLAGS_ONLINE	BIT(1)
+
 /**
  * struct xe_ggtt_node - A node in GGTT.
  *
@@ -117,6 +120,8 @@ struct xe_ggtt {
 	 * @flags: Flags for this GGTT
 	 * Acceptable flags:
 	 * - %XE_GGTT_FLAGS_64K - if PTE size is 64K. Otherwise, regular is 4K.
+	 * - %XE_GGTT_FLAGS_ONLINE - is GGTT online, protected by ggtt->lock
+	 *   after init
 	 */
 	unsigned int flags;
 	/** @scratch: Internal object allocation used as a scratch page */
@@ -367,6 +372,8 @@ static void dev_fini_ggtt(void *arg)
 {
 	struct xe_ggtt *ggtt = arg;
 
+	scoped_guard(mutex, &ggtt->lock)
+		ggtt->flags &= ~XE_GGTT_FLAGS_ONLINE;
 	drain_workqueue(ggtt->wq);
 }
 
@@ -437,6 +444,7 @@ int xe_ggtt_init_early(struct xe_ggtt *ggtt)
 	if (err)
 		return err;
 
+	ggtt->flags |= XE_GGTT_FLAGS_ONLINE;
 	return devm_add_action_or_reset(xe->drm.dev, dev_fini_ggtt, ggtt);
 }
 ALLOW_ERROR_INJECTION(xe_ggtt_init_early, ERRNO); /* See xe_pci_probe() */
@@ -465,13 +473,10 @@ static void ggtt_node_fini(struct xe_ggtt_node *node)
 static void ggtt_node_remove(struct xe_ggtt_node *node)
 {
 	struct xe_ggtt *ggtt = node->ggtt;
-	struct xe_device *xe = tile_to_xe(ggtt->tile);
 	bool bound;
-	int idx;
-
-	bound = drm_dev_enter(&xe->drm, &idx);
 
 	mutex_lock(&ggtt->lock);
+	bound = ggtt->flags & XE_GGTT_FLAGS_ONLINE;
 	if (bound)
 		xe_ggtt_clear(ggtt, xe_ggtt_node_addr(node), xe_ggtt_node_size(node));
 	drm_mm_remove_node(&node->base);
@@ -484,8 +489,6 @@ static void ggtt_node_remove(struct xe_ggtt_node *node)
 	if (node->invalidate_on_remove)
 		xe_ggtt_invalidate(ggtt);
 
-	drm_dev_exit(idx);
-
 free_node:
 	ggtt_node_fini(node);
 }