> On Sun, 1 Aug 1999, Riley Williams wrote:
> > > The whole vfat design is "disaster waiting to happen" :-/
> >
> > The original design is, yes, but then, that came from Microsoft, so
> > what can one expect. The Linux implementation isn't too bad in a lot
> > of ways, but there are a few tweaks that are needed to polish it up to
Few tweaks??? Full rewrite of data-related part. I've found what
was wrong with my code. Finally. And it's not a pretty sight. Basically
with the serialization between write() and truncate() gone the whole
handling of cluster allocation (extending/truncating the chains) fell
apart. Wouldn't be too bad, if we wouldn't have to keep all data in page
cache when we extend the bloody thing. No holes in files, damnit... Oh,
well... At least it solves the problem with dmsdos - solution for normal
fatfs will work there too.
-
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/