"value":"In the Linux kernel, the following vulnerability has been resolved:\n\nscsi: core: Move scsi_host_busy() out of host lock for waking up EH handler\n\nInside scsi_eh_wakeup(), scsi_host_busy() is called & checked with host\nlock every time for deciding if error handler kthread needs to be waken up.\n\nThis can be too heavy in case of recovery, such as:\n\n - N hardware queues\n\n - queue depth is M for each hardware queue\n\n - each scsi_host_busy() iterates over (N * M) tag/requests\n\nIf recovery is triggered in case that all requests are in-flight, each\nscsi_eh_wakeup() is strictly serialized, when scsi_eh_wakeup() is called\nfor the last in-flight request, scsi_host_busy() has been run for (N * M -\n1) times, and request has been iterated for (N*M - 1) * (N * M) times.\n\nIf both N and M are big enough, hard lockup can be triggered on acquiring\nhost lock, and it is observed on mpi3mr(128 hw queues, queue depth 8169).\n\nFix the issue by calling scsi_host_busy() outside the host lock. We don't\nneed the host lock for getting busy count because host the lock never\ncovers that.\n\n[mkp: Drop unnecessary 'busy' variables pointed out by Bart]"
