On Wed, 4 Dec 2002 [EMAIL PROTECTED] wrote:

> On Tue, 3 Dec 2002, David Brownell wrote:
>
> > Yes, any new driver's probe() normally handles #3.  For
> > usbfs it's necessarily factored a bit differently, since
> > there's no probe(), but it's still the responsibility of
> > that driver.
>
> Provided that it can do so from every state. This is a very huge
> assumption. Imagine a disconnect() during firmware download.
> You'll have no problems finding devices you can hang this way.
> We have eg. storage devices known buggy if control messages
> arrive at an imopportune moment and HID devices which need
> specific sequences of control messages.

Indeed.  And should the user unplug the USB cable at just such a moment,
there's very little that can be done when the device gets plugged back in.
No device driver can handle a scenario like this, but it is reasonable to
require that a logical disconnect (caused e.g. by rmmod, as opposed to a
physical disconnect) be handled with a little more grace.

Alan Stern



-------------------------------------------------------
This SF.net email is sponsored by: Microsoft Visual Studio.NET 
comprehensive development tool, built to increase your 
productivity. Try a free online hosted session at:
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr0003en
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to