tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: write alignment matters?

On Fri, Jun 24, 2011 at 07:50:35AM +0000, Michael van Elst wrote:
> Similar to raw devices in performance, I/O caching, and without the
> hardware constraints regarding block sizes or alignment. I.e. something
> that lets you treat disk contents like a regular file.

At least for NetBSD, that's never been true.  The most glaring problem
is that there's no protection against causing the same underlying disk
blocks to be multiply cached by accessing the buffer cache with a different
stride.  And no way to keep those multiple cachings coherent...

I don't know (I guess I could look) whether the original Unix code had
that problem but the replacement code in NetBSD had them from the very
beginnings.  The filesystem doesn't *do* that, so it doesn't suffer...


Home | Main Index | Thread Index | Old Index