It might be useful to be able to update your primary disk
driver on the fly, without rebooting. It might be useful to keep your
system running while you switch from a failing drive to a good one.
Impossible? Not if you have enough memory! Might take a
little hackery at the VFS and block cache levels, though:
Setp 1: mount primary filesystem
Step 2: read all blocks into cache
Step 3: disconnect filesystem from driver
Step 4: unload driver
... run for a while ...
Step 5: reload driver (load a new driver?)
Step 6: reconnect filesystem to driver
Step 7: update filesystem on disk
a) if reconnected to the same partition, update only changed blocks
b) if connected to a different partition, write all blocks
You may have noticed that I've combined two different proposed
features. The ability to read a filesystem into memory, switch its
drive/partition, and write it out again, is independent of the feature
of unloading the primary driver.
Craig Milo Rogers
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/