Mark McClelland wrote:
> Miles Lane wrote:
>
>
>> Mark McClelland wrote:
>>
>>
>>> Georg Acher wrote:
>>>
>>>>> On Sat, Mar 31, 2001 at 10:06:23PM -0500, Pete Zaitcev wrote:
>>>>>
>>>>>> many users complain (and file bugs) about a flood of
>>>>>> the following message:
>>>>>>
>>>>>> interrupt, status 3 frame# NNN
>>>>>
>
>>> Here are some things that reliably cause this condition:
>>> - Plugging in my D-Link DSB-C300 camera. Reloading HCD fixes it.
>>> - Switching between systems with my USB KVM. Only seems to happen when >2
>>> devices are plugged in to it. Reloading HCD fixes it.
>>> - Plugging >2 cameras into my (self-powered) D-Link DSB-H4 hub. Something goes
>>> wrong during enumeration and the hub shuts down, and the only way to fix it is
>>> to unplug a device.
>>
>> Okay, I'll ask the perennial question: What happens when you
>> use uhci.o? Both Georg and Johannes are working hard to
>> resolve all outstanding HCD bugs so we can standardize on
>> one UHCI HCD, so it helps to have a comprehensive list of
>> HCD problems.
>
>
> OK, I tried uhci.o on an unpatched 2.4.2 kernel:
>
> - Plugging in DSB-C300: Same result, no messages at all. Same fix.
> - Switching KVM: cannot reproduce reliably enough to test it.
> - Plugging in 3 cameras: works. Plugging in 4, however, gives me repeated "error
>-110"
> and "error -71" messages. I had to unplug all upstream and downstream cables from the
> hub to get it working again.
>
> So, it looks like this is a general Linux-USB problem. Either that, or it's a problem
> with my specific hardware (i82371SB PIIX3 USB [Natoma/Triton II] Rev. 1). I will try
> this on my trusty OHCI tomorrow.
In order to successfully debug these problems, Georg and Johannes
will likely need more contextual information. Also, it would
help if you ran each test after a fresh boot. If you are hanging
USB one way, then getting it back into a nominal state by
unloading and reloading the HCD, your hardware could still be in
a confused state. Subsequent testing results could be invalid.
When you reproduce each problem, send about thirty lines of your
logfile leading up to and including the messages shown in the
error state.
Also, please turn on debugging in uhci.c and usb-uhci.c.
uhci.c allows you to pass in the debug level as a parameter
at module load time:
MODULE_PARM(debug, "i");
MODULE_PARM_DESC(debug, "Debug level");
So try, "modprobe uhci debug=2"
With usb-uhci, it looks like you'll need to modify the
source code. Just change "#undef DEBUG" at line 61
to "#define DEBUG". OTOH, it looks like debugging may always
be enabled in the usb-uhci driver.
I hope this helps,
Miles
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
