Re: lack of raw disk devices

jhohertz@golden.net
Mon, 5 Jan 1998 17:11:11 -0500


Hey.... I may have said this to the list already, but, is it possible to just
make an ioctl() for the block devices that turns off the buffering? Seems a
more elegant solution than having a two device files for every single block
device.

IMHO at least. Probably can't be much worse to implement that the seperate
device file way of doing it.

All this from someone who's never tried acutally writting kernel code himself
of course. :P (learning though....)

Joe

On Jan 5, 10:04pm, Alan Cox wrote:
> Subject: Re: lack of raw disk devices
> > So you can disable the write caching if you want. There seems some
> > dissense whether applications want to disable the read cache as well.
>
> Unusual but specific examples exist and are quite valid. So I don't think
> its an item of dissent just a "has anyone got around to submitting Linus
> a diff yet".
>
> The main one cited is a two system database sharing the same disk across
> both machines. Here read caching is a catastrophe