Re: [CFT][PATCH] long-living cache for block devices

Alexander Viro (viro@math.psu.edu)
Sun, 11 Nov 2001 14:09:09 -0500 (EST)


On Sun, 11 Nov 2001, Linus Torvalds wrote:

> If we end up removing the module, we'll call "unregister_blockdev()" or
> whatever, which in turn gets rid of _all_ pages, and that that time we
> will correctly get rid of buffers.
>
> And before you remove the module I see no advantage to have any guarantees
> of quiescence and removing the buffer heads. I only see extra code that
> doesn't seem to have any real purpose..

As it is, driver has a right to destroy any internal data structures as
soon as the last ->release() is called. None of them expect requests
coming in after that and frankly, that makes sense.

IOW, if there is any point in _having_ ->release() at all, we'd better
make sure that nothing will bother the driver between the final ->release()
and the next ->open().

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/