The Linux kernel vulnerability CVE-2026-64326 involves the block device function bdev_mark_dead() calling sync_blockdev() unconditionally during surprise removal of a device (surprise == true). Since the device is already gone, sync_blockdev() blocks indefinitely waiting for writeback completion that cannot occur, causing hangs in kernel reset workers and dependent tasks. The fix modifies bdev_mark_dead() to skip sync_blockdev() when surprise removal is detected, aligning with the filesystem callback fs_bdev_mark_dead() behavior. This prevents the kernel from hanging during unexpected device removals. The issue was found by fuzzing NVMe device reset paths.

The vulnerability can cause indefinite kernel hangs during surprise removal of block devices, specifically affecting NVMe device reset workflows. This results in wedging the reset worker thread and any tasks waiting on it, potentially leading to system instability or denial of service conditions. There is no indication of code execution or data corruption from the provided data.

Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. The fix involves skipping sync_blockdev() on surprise removal in bdev_mark_dead(), preventing hangs. Until an official patch is available, avoid scenarios causing surprise removal or device resets that trigger this code path if possible.