A deadlock vulnerability in the Linux kernel's SCSI libsas driver during high availability (HA) resume has been fixed. The issue occurs when smp_execute_task_sg() calls pm_runtime_get_sync() on the host during the HA resume path, causing a deadlock due to waiting on an ongoing resume while the resume waits for SMP IO completion. This deadlock specifically affects expander-attached SATA disks during the resume process.