Hi,
I'm wondering if someone can help me out here, I have the following problem, but it seems to only exist on vanilla 2.4 kernels. My 2.6 works fine, and so do my Suse and Mandrake 2.4 kernels.
I wonder if those vanilla 2.4 kernels have some other bug for which this is just a symptom; I don't recall any EHCI changes that should help symptoms like that. Try using 2.6 until that problem in 2.4 gets resolved ... ;)
What is happening?
I'm running X in dualhead mode (2 Nvidia based cards), and shortly
after startup my usb mouse stops working. In /var/log/messages
I see a line:
"ehci-q.c: can't resubmit interrupt urb df60cc4c: status -32"
That's odd, yes. It means the interrupt endpoint reported an endpoint halt.
Virtually no drivers try to handle that kind of error sanely. (Handing work off to a helper thread, which will clear the halt and then restart the interrupt mode transfer.) Was that the only strange message appearing at that time?
I do have quite an extensive usb setup,...
Of which the relevant point was that the mouse was hooked up to the high speed hub, so presumably it's going through the USB 2.0 "Transaction Translator" logic. Maybe your config is abusing that code more than most?
Is there something else I can do to track down this problem?
See if it's caused by the mouse, and that it's caused when the mouse is going through the USB 2.0 hub, and that it doesn't happen when the mouse goes through a USB 1.1 hub (including a root hub).
- Dave
------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
