David Brownell wrote: >On Thursday 28 April 2005 10:22 am, Brian Beardall wrote: > > >>Brian Beardall wrote: >> >> >> >>>I've been experiencing IRQ don't cares using webcams that use >>>isochronous transfers. ... >>> >>> >>This is how the isochronous transfer is setup: >> >>urb->number_of_packets = 10; >>urb->intervral = 1; >>urb->transfer_flags = USB_ISO_ASAP; >> >>Those two settings lockup the AMD 756 USB controller. >> >> > >But they don't lock up other OHCI implementations? Do the 756 errata >have any light to cast on this? > > > The 756 Errata doesn't expain this behavior. There are two errata's that affect the controller, and that is errata #4 which deals with low speed devices not being enumerated, and having the status registers 34h and above scewed on the data. There is errata #10 which deals with suspend related issues. Both of those issues were dealt with the patches in the 2.6.10 kernel.
I can't find another OHCI add-in-card or mainboard that uses OHCI. I know that it the above transfer flags will crash the controller. It seems as though it is a timing issue because the amount of time of living to dead is random. I ran it over eight hours once, and another time it the IRQ don't care occured in less than 30 seconds. ------------------------------------------------------- SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
