I thought modern drives "aged" the cached information and write it to disk
out of order w.r.t. the scheduling algorithm if it was too old. Otherwise
you would get some really interesting starvation problems under most os:s
with some bad corruption as a result on any heavily loaded machine.
The drives won't necessarily, but any caching controller should. Still,
without carefully crafting the failure case I'd say the probability is
vanishingly low for a command to be left uncompleted indefinitely.
Leonard
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu