Re: Should raw I/O be added to the kernel?

Peter Svensson (petersv@psv.nu)
Fri, 18 Dec 1998 22:26:22 +0100 (CET)


On Fri, 18 Dec 1998, Linus Torvalds wrote:

> The only argument for raw disk IO is the caching policy issue, and I very
> strongly agre with Ingo that we should consider it as such, and not get
> dragged into the rathole of thinking that raw devices are somehow worth it
> for some other reason.

Raw-io is basically a solution to a problem. As with most solutions there
are advantages and disadvantages. One of the main advantages I see is that
it is a by now well known solution to one set of problems. Perhaps someone
with a deep understanding of the kernel and the problems should take a
step back and look at the problems instead of the various solutions
proposed.

The database problem is one I find the most interesting. Using raw io
basically means we cannot find an abstracted way of describing what we
require of the system, so we do everything by ourself. Of course, that way
anything can be done.

Peter

--
Peter Svensson      ! Pgp key available by finger, fingerprint:
<petersv@psv.nu>    ! 8A E9 20 98 C1 FF 43 E3  07 FD B9 0A 80 72 70 AF
<petersv@df.lth.se> !
------------------------------------------------------------------------
Remember, Luke, your source will be with you... always...

- 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/