Hi!
> Interesting. Very interesting.
>
> Well, clearly one thing I can do is try to place locks around everything
> to eliminate this race condition. But this raises two concerns to me:
>
> (1) Lots of locks means poor performance
> (2) Lots of drivers probably suffer from this same problem. usb-storage
> is probably just one of the few that does so many repeated transfers over
> such a long period of time to make this visible.
>
> It occurs to me that the best answer is probably to make it so that
> usb_sndbulkpipe checks for a NULL device and make it so that
> usb_bulk_msg() can also handle a NULL device. Returning an error to
> indicate "invalid device" would probably be the best way to go.
What about not setting pusb_dev to NULL in disconnect? Usb layer should
handle that.
Pavel
--
I'm [EMAIL PROTECTED] "In my country we have almost anarchy and I don't care."
Panos Katsaloulis describing me w.r.t. patents me at [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
