> On Mon, 1 Nov 1999, Gerard Roudier wrote:
> >I can understand that (1) perhaps allows to implement some features not
> >possible with the layering used in Linux, at the price of increasing FS
> >complexity probably.
> Agreed. This because at the blockdevice layer you lost all the semantics
> of the data and so at the ll_rw_block layer you can't implement any clever
> algorithm in order to relocate the data on optimal physical locations.
I am not sure that implementing mirroring (or RAID) with File System
interactions was the way to go given the number of File Systems supported
by Linux (in fact the current approach looks just fine to me). But I have
no objection about some FS controlling this feature by itself. Result is
indeed that this FS layer gets more complex.
May-be the block device layer should be moved to block peripheral drivers,
but I donnot see in what this will significantly change the semantic about
data from the FS layer.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to email@example.com
Please read the FAQ at http://www.tux.org/lkml/