Linux kernel (drivers/gpu/drm/xe): Same per-client accounting path, different failure - if the fdinfo read drops the

Impact

Same per-client accounting path, different failure - if the fdinfo read drops the last reference to a buffer object, destruction wants sleeping locks while the caller holds a spinlock. The result is a hard deadlock (and sleep-in-atomic) with driver locks held, so the xe device stops serving every tenant on it, not just the one that triggered it.

Who can reach it

Tenant holding /dev/dri/renderD* on Intel xe arranges for an fdinfo read to drop the final BO reference - achievable by freeing BOs from one thread while reading /proc/<pid>/fdinfo/<drmfd> from another. Reachable by the tenant itself and by any monitoring agent reading tenant fdinfo.

What to do

Update to a kernel with the fix (stable commits below; no fixed_in published). Interim: disable DRM fdinfo scraping in your GPU telemetry stack on xe nodes.

References