Re: mmap on a device returns ENODEV

Raul Miller (moth@magenta.com)
Wed, 15 Dec 1999 14:25:00 -0500


Stephen C. Tweedie <sct@redhat.com> wrote:
> Absolutely, there's no problem there. From raid-5's point of view,
> the hashing of buffer_heads is at the minute mainly a performance
> question --- the correctness issue is easily fixed --- but it is a
> _serious_ performance question.

How serious? [Factor of two? Four? Less?]

If you look at this as a requirement for very-coarse-granularity updates,
what happens? [Do you run out of memory trying to cache the entire disk?
Do you wind up never writing anything to disk? Do you wind up writing
the same stuff too often? Some other set of possibilities?]

I'd think that raid5 just means that the size of a logical block
to be written must be larger, but I'm not the expert you folks are.

Thanks,

-- 
Raul

- 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/