[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [usb] Clarification on USB



The preferred method is for the host to tell the hub to shut off the port.
The port will not take action on it's own.

There is no specific method defined for the host to determine which USB
device is bad.  The host will most likely retry the communication to the
device it was talking to when the error occurs.  If the error continues it
will probably shut it off assuming it had problems.  Then if the bad device
continues to create problems for other devices they will be shut off until
the bad one is found and silenced.

This may not be fair to good devices, but you are talking about an anomalous
situation where you have a bad device (one that is not following the bus
protocol) attached.  So the same problems that would occur on Ethernet, PCI,
ISA, SCSI, etc. apply to USB.

Marc Reinig
System Solutions


> -----Original Message-----
> From: owner-usb@opencores.org [mailto:owner-usb@opencores.org]On Behalf
> Of Vishal Namshikar
> Sent: Thursday, March 06, 2003 8:55 PM
> To: usb@opencores.org
> Subject: Re: [usb] Clarification on USB
>
>
> Hi Marc,
> If this is the case then Hub will be responsible for disabling the port,
> right? Does the hub block upstream traffic after timeout period expires?
> If answer to first question is "No" then how does the host controller know
> which device on hub is responsible for babble?
> If the hub doesnot block upstream traffic then how does the host
> controller
> know whether its a  device on hub that is responsible for babble
> and not the
> hub?
> --vishaln
>
>
> ----- Original Message -----
> From: "Marc Reinig" <mreinig@pacbell.net>
> To: <usb@opencores.org>
> Sent: Thursday, March 06, 2003 8:49 PM
> Subject: RE: [usb] Clarification on USB
>
>
> > The port that is shut off is the port to which the device is directly
> > connected, i.e., the hub, not the root hub, unless it is connected
> directly
> > to the root hub.
> >
> > Marc Reinig
> > System Solutions
> >
> >
> > > -----Original Message-----
> > > From: owner-usb@opencores.org
> [mailto:owner-usb@opencores.org]On Behalf
> > > Of Vishal Namshikar
> > > Sent: Wednesday, March 05, 2003 7:41 PM
> > > To: usb@opencores.org
> > > Subject: Re: [usb] Clarification on USB
> > >
> > >
> > > Suppose the babbling device is connected to hub which in turn is
> connected
> > > to root hub, then the hub port is disabled or the root hub port
> > > is disabled?
> > >
> > > ----- Original Message -----
> > > From: "Marc Reinig" <mreinig@pacbell.net>
> > > To: <usb@opencores.org>
> > > Sent: Wednesday, March 05, 2003 8:23 PM
> > > Subject: RE: [usb] Clarification on USB
> > >
> > >
> > > > The contention is avoided by only have devices that work
> > > plugged in to the
> > > > USB.  Certainly a device that does not adhere to the turn
> > > around time spec
> > > > is not a working device.
> > > >
> > > > The contention is controlled by the host shutting off the port that
> the
> > > > device is connected to and not talking to the device any
> more, if this
> > > > occurs more than a few times.
> > > >
> > > > Marc Reinig
> > > > System Solutions
> > > >
> > > >
> > > > > -----Original Message-----
> > > > > From: owner-usb@opencores.org
> > > [mailto:owner-usb@opencores.org]On Behalf
> > > > > Of Ravi Kumar
> > > > > Sent: Tuesday, March 04, 2003 10:41 PM
> > > > > To: usb@opencores.org
> > > > > Subject: [usb] Clarification on USB
> > > > >
> > > > >
> > > > > Hi Folks
> > > > >
> > > > > I have  a Query on Usb.
> > > > > suppose host puts IN token and send it  to device .
> > > > > 2.Device takes more than the Bus turn around Time ( or
> > > > > more than the bit times required to respond )
> > > > > 3.Host wil assume itz the Bus time out and intimate to
> > > > > software.
> > > > > 4.Assuming this the host wil retry to the same Device
> > > > > .
> > > > > 5.By that time whn Device responds for the previous IN
> > > > > token it wil lead to Contention
> > > > >
> > > > >
> > > > >
> > > > > How the above said Contention is avoided or
> > > > > controlled.
> > > >
> > > >
> > > >
> > > > --
> > > > To unsubscribe from usb mailing list please visit
> > > http://www.opencores.org/mailinglists.shtml
> > >
> > > --
> > > To unsubscribe from usb mailing list please visit
> > http://www.opencores.org/mailinglists.shtml
> >
> >
> > --
> > To unsubscribe from usb mailing list please visit
> http://www.opencores.org/mailinglists.shtml
>
> --
> To unsubscribe from usb mailing list please visit
> http://www.opencores.org/mailinglists.shtml


--
To unsubscribe from usb mailing list please visit http://www.opencores.org/mailinglists.shtml