I think the filesystem layer needs to communicate more with the driver layer
to handle this, but it's not really up to the FS.
A reasonable API should allow any fs to be layered on top of an MD / LVM
layer (think of LVM as being a variant of striping); if the fs doesn't know
about md a default strategy could be used, but if it's md-aware then the fs
can request specific handling for things. If the md layer is sufficiently
flexible as to allow variable-sized logical extents (that is, fixed size
within a particular logical volume but variable between volumes, presumably
always a multiple of the physical extent size), one could (for example)
create an md volume and then let "md-aware ext3" designate the use of
logical extents the size of cylinder groups. And a log-structured
filesystem could request that log extents be allocated on a separate device
from file extents if possible.
-- brandon s. allbery [os/2][linux][solaris][japh] allbery@kf8nh.apk.net system administrator [WAY too many hats] allbery@ece.cmu.edu carnegie mellon / electrical and computer engineering KF8NH Kiss my bits, Billy-boy.
- 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/