6.9-stable review patch. If anyone has any objections, please let me know.
------------------
From: Vidya Srinivas vidya.srinivas@intel.com
commit 43e2b37e2ab660c3565d4cff27922bc70e79c3f1 upstream.
In some scenarios, the DPT object gets shrunk but the actual framebuffer did not and thus its still there on the DPT's vm->bound_list. Then it tries to rewrite the PTEs via a stale CPU mapping. This causes panic.
Cc: stable@vger.kernel.org Reported-by: Shawn Lee shawn.c.lee@intel.com Fixes: 0dc987b699ce ("drm/i915/display: Add smem fallback allocation for dpt") Signed-off-by: Vidya Srinivas vidya.srinivas@intel.com [vsyrjala: Add TODO comment] Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com Link: https://patchwork.freedesktop.org/patch/msgid/20240520165634.1162470-1-vidya... (cherry picked from commit 51064d471c53dcc8eddd2333c3f1c1d9131ba36c) Signed-off-by: Jani Nikula jani.nikula@intel.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/gpu/drm/i915/gem/i915_gem_object.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h @@ -284,7 +284,9 @@ bool i915_gem_object_has_iomem(const str static inline bool i915_gem_object_is_shrinkable(const struct drm_i915_gem_object *obj) { - return i915_gem_object_type_has(obj, I915_GEM_OBJECT_IS_SHRINKABLE); + /* TODO: make DPT shrinkable when it has no bound vmas */ + return i915_gem_object_type_has(obj, I915_GEM_OBJECT_IS_SHRINKABLE) && + !obj->is_dpt; }
static inline bool