Ilia A. Petrov wrote:
> David Brownell wrote:
>> Ilia A. Petrov wrote:
>>
>>> When kernel is mounting root file system it is doing it too fast so
>>> usb-support have not ime to scan bus for mass-storage devices and
>>> connect them.
>>> ...
>>> or, imho better way, - when completing init of usb bus, first scans
>>> it and connect all devices and only after all devices were connected
>>> returns to main kernel code.
>>
>>
>>
>> That might not entirely solve the problem, since the relevant device
>> could drop off the bus temporarily, but it seems like it'd be a step
>> forward. How would you make root hub ("bus") initialization do that?
>
>
> i'm not familiar with linux usb implementation so may be it's wrong:
> after sending global reset over the bus you can manually check (not
> trough the hub driver) root port connection status and call enumeration
> if needed.
I had in mind an explanation that works with "patch -p1" ... :)
There's already a call to register a root hub. It'd be dangerous
to have some other thread try to share responsibilities with khubd,
so that call would likely need to be modified. (Although there's
been a bit of discussion about actually getting rid of that specific
single-point-of-failure, since not all hub operations need to be
serialized.)
- Dave
-
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/
This archive was generated by hypermail 2b29 : Mon Jul 07 2003 - 22:00:29 EST