Moe Wibble wrote:

http://mbox.bz/pub/lsusb-v-before.txt [17k, before timeout]
http://mbox.bz/pub/lsusb-v-after.txt  [16k, after timeout]

After the device went bad lsusb announces protocol errors
while querying it. That's not surprising but I put the after-
output there too, for completeness.

That one message seems to happen because the config descriptor's wTotalLength is one byte too big. Not enough quality assurance...


You are right. I overlooked it before but it really is a GeneSys.
Do you happen know an usb2->ide bridge (sold in a 2.5" external hdd-case?)
that definately works in linux 2.4?

I don't use them enough to recommend one, but only this newish batch of GeneSys-based devices seems to be generating significant problem reports.


You might be able to turn up some more information with this
EHCI diagnostic patch (for 2.4), which just dumps the hardware
data structures at the moment usb-storage times out the transfer.
What it's shown in similar cases is that the GeneSys adapter
got part way through a transfer and then stopped responding.

...


The additional debug messages reveal that the device behaves just
the way you described (if i parse them right that is). It seems to
block in the middle of transfer and times out.

That's my reading too. More specifically, the device locks up solid, part way through the write -- solid enough that it won't even reset properly.


One person had better luck after applying a patch that slowed
down EHCI on an NForce2 southbridge ... the fact that it even
mattered is in indication of problems in the GeneSys device.


I assume the wintendo drivers use a similar hack to make the

Not on that ALI controller they don't -- it doesn't support the relevant hardware feature, so it can't be turned off!

If I were to guess at hacks, my first one would be just to
throttle the write rate down by pausing periodically, on the
theory that it locks up when it can't buffer for some reason.
(This was a "big" series of writes, shorter ones were ok...)

It might be worth having usb-storage have a quirk for this
device.  Maybe just emit a warning when it's running at
high speed, if there's no workaround we can apply.


thing work then. What really pi**es me off right now is that the
case has the USB2.0-logo printed on it.
Yet another worthless cert...

Not entirely, but it's more focussed on basics (which have been problematic on occasion) than on bugs that somehow don't appear except with non-MSFT drivers.

Again, that's a vendor quality assurance issue:  they didn't
test with Linux.

- Dave




------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to