Hi,
On Mon, Apr 24, 2000 at 01:04:57PM -0600, Andreas Dilger wrote:
>
> Is any work being done (or do you have a version yourself) to allow ext3
> as a module? This would simplify testing with ext3 a bit. I imagine the
> jfs changes would have to stay in the kernel, as it changes the task struct
> and more...
It should be possible, but only with a bit of extra help. There are
one or two pieces of core kernel functionality which are necessary to
support journaling properly --- in particular, callbacks from the VM
to the filesystem to let the filesystem flush dirty buffers to disk if
we run short of available memory (because we cannot rely on the normal
bdflush wakeup to flush buffers if the buffers are pinned by a
transaction).
However, the core fs code will be modularisable very easily. The only
thing preventing it right now is some debugging code in current ext3
trees which is there to detect buffers being written to disk out-of-
order.
--Stephen
-
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/
This archive was generated by hypermail 2b29 : Sun Apr 30 2000 - 21:00:08 EST