CVE-2025-71314

*** Pendiente de traducción *** In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/panthor: Recover from panthor_gpu_flush_caches() failures<br /> <br /> We have seen a few cases where the whole memory subsystem is blocked<br /> and flush operations never complete. When that happens, we want to:<br /> <br /> - schedule a reset, so we can recover from this situation<br /> - in the reset path, we need to reset the pending_reqs so we can send<br /> new commands after the reset<br /> - if more panthor_gpu_flush_caches() operations are queued after<br /> the timeout, we skip them and return -EIO directly to avoid needless<br /> waits (the memory block won&amp;#39;t miraculously work again)<br /> <br /> Note that we drop the WARN_ON()s because these hangs can be triggered<br /> with buggy GPU jobs created by the UMD, and there&amp;#39;s no way we can<br /> prevent it. We do keep the error messages though.<br /> <br /> v2:<br /> - New patch<br /> <br /> v3:<br /> - Collect R-b<br /> - Explicitly mention the fact we dropped the WARN_ON()s in the commit<br /> message<br /> <br /> v4:<br /> - No changes