On Tue, Apr 09, 2024 at 05:31:32PM +0200, Michal Wajdeczko wrote:

Although it's unlikely that drmm_mutex_init() will fail during
driver initialization, however we shouldn't ignore this case.

Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>

there was already a patch pending for this, but got stuck because it was
trying to do something similar for all the display mutexes, that
shouldn't be there in the first place. Not sure if we take that one with
conflict resolved or this.

If the patch mentioned is https://patchwork.freedesktop.org/series/131300/, it does not resolve the mutex mentioned above.

---
drivers/gpu/drm/xe/xe_device.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
index 9083f5e02dd9..ba3643906d9e 100644
--- a/drivers/gpu/drm/xe/xe_device.c
+++ b/drivers/gpu/drm/xe/xe_device.c
@@ -273,7 +273,9 @@ struct xe_device *xe_device_create(struct pci_dev *pdev,

    init_waitqueue_head(&xe->ufence_wq);

-    drmm_mutex_init(&xe->drm, &xe->usm.lock);
+    err = drmm_mutex_init(&xe->drm, &xe->usm.lock);
+    if (err)
+        goto err;

blank line missing

Lucas De Marchi

    xa_init_flags(&xe->usm.asid_to_vm, XA_FLAGS_ALLOC);

    if (IS_ENABLED(CONFIG_DRM_XE_DEBUG)) {
-- 
2.43.0