tech-kern archive

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

Re: automatic vnd'ing on kernel boot

On Tue, Mar 02, 2010 at 08:29:04PM +0900, Toru Nishimura wrote:
> Someone once said internal NAND controller chip fabricated in modern
> sillicon storage, like CompactFlash, SDMMC and others, are optimized
> for FAT FS runtime dynamism.  It might be a myth.  As It's not
> disclosed how the controller HW manages badblock and wear-levelling,
> it'd be unlikely validatable whether "made (just) for FAT or not."
> So, my idea is "let us go FFS in a plain large image of FAT, and
> don't care to do disklabel+newfs similicon storage themselves
> anymore."

I doubt that would do anything useful.  The "FAT support" in any NAND
 controller is likely to be dependent on files actually being stored in
a FAT filesystem, but the I/O behaviour of a FFS filesystem in a vnd is
still going to look like FFS, regardless of what the underlying file
is.  At that point you're not using the FAT FS as a filesystem, but just
as a complicated disklabel.


Home | Main Index | Thread Index | Old Index