Hi,
On Mon, Jun 05, 2000 at 01:23:08PM +0100, Alan Cox wrote:
> > I am referring to the buffer cache on the disks. There will be a need to
> > flush/force write completion of data to the disk before power down if I
> > enable the intelligent write-caching in the hardware.
>
> If you do this then you won't be able to use IDE for journalling fs's. Before
> we can do cachable re-ordered writes on the device itself we have to
> implement store barrier points in the disk request queues so that the kernel
> can stop the drive re-ordering when it is not safe too
Ditto for any applications which rely on write ordering, such as
sendmail on a sync mail spool, or any database with log recovery.
Lying to the CPU about when the data is safe on disk is a bad thing
to do, in general.
--Stephen
-
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/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:26 EST