Hi Apologies in advance if it's felt this is the wrong mailinglist. I'm developing g15daemon & libg15, which is a userspace driver for the logitech G15 keyboard and it's LCD, using libusb. Kernels up to and including the 2.6.17 series have been working fine, however users who've upgraded to 2.6.18-rc1 and above, and who use an intel (i945, possibly others too) chipset are seeing their syslog filled with:
usbfs: usb_submit_urb returned -22 usbfs: usb_submit_urb returned -22 usbfs: usb_submit_urb returned -22 usbfs: usb_submit_urb returned -22 etc.. to the tune of 10 or so a second. The userspace code (once the device is setup) which appears to provoke this warning simply polls the keyboard with usb_interrupt_read() every 50-100milliseconds. The keyboard still works fine, apparently. I've been unable to replicate this, it doesnt appear to affect VIA (and possibly other uhci) chipsets, and I dont currently have an intel board available for testing. Users who have gone to the trouble of adding 3rd-party pci usb ports report that the problem goes away when not using the onboard usb chipset. I'm attempting to narrow down an exact cause (and whether I can work around it in userspace) and would greatly appreciate a pointer or two on the most likely culprits in the kernel code (if any), as I realise there were numerous changes to the usb subsystem in 2.6.18. I know I'm asking for a shot in the dark here, due to the sparsity of information, but any ideas would be helpful. Thanks, Mike ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel