RFC: root mount enhancement (round 2)

[Andrew Thompson]

On 26 August 2010 08:44, M. Warner Losh <imp at bsdimp.com> wrote:
> In message: <54210.1282752073 at critter.freebsd.dk>
>            "Poul-Henning Kamp" <phk at phk.freebsd.dk> writes:
> : In message <34EF2360-1B68-4E0C-8CCE-409CE141D0B8 at mac.com>, Marcel Moolenaar wri
> : tes:
> :
> : >don't want to enhance: A USB disk cannot always be used as a root
> : >file system by virtue of the USB stack releasing the root mount
> : >lock after creating the umass device, but before CAM has created
> : >the corresponding da device.
> :
> : This is a bug which is entirely unrelated to how we find the
> : root filesystem:  It should simply be fixed by CAM grabing a
> : root mount lock when activated from USB and releasing it
> : only when all it's stuff is done.
>
> We already do this...  But it is insufficient since usb discovery is
> done asynchronously...

Its more that the usb disk appears and the root mount lock is dropped
without geom tasting taken into account. This was fixed with r190677
but then I was asked to back it out (r190878).

> Scott has a similar fix in the pipeline, but I don't know the state of
> it.

It would be great to get this finished, I believe the solution Scott
wanted was to properly use intr_config_hooks to kick off usb
enumeration.


Andrew