Hi,
It seems that we should test the new USB code. So, I did a bit
of testing with the various HCD drivers and the irda-usb dongle
driver. Bad news.
The agravating circonstances :
o kernel 2.5.20
o one IrDA-USB dongle, irda-usb
o SMP (dual PPro 200)
o Shared irq with the network card (both ohci/uhci)
ohci-hcd :
o First Tx URB never complete or timeout
o Removing module OOPS
usb-uhic-hcd :
o Deadlock at the end of first IrDA file transfer
o Tx URB never complete or timeout
o Whole networking layer frozen
o Removing module OOPS
uhci-hcd :
o Deadlock at the beggining of first IrDA file transfer
o Tx URB never complete or timeout
o Whole networking layer frozen
o Removing module OOPS
Of course, clean shutdown/power-off between each of the
test. The difference between uhci-hcd and usb-uhci-hcd is probably
only due to the different amount of time before I start the transfer
or other random factors, but to me they suffer from the same same
problem.
And ohci-hcd is definitely bad. But I saw the same problem in
usb-ohci in 2.5.15 (that's why I upgraded to 2.5.20).
On the other hand, a quick test with uhci-hcd indicate that it
support Short Packet insertion (USB_ZERO_PACKET) in the Tx path. I
guess the other HCD probably do it as well.
And of course it use to work properly in 2.4.X :-(
Regards,
Jean
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
