On Mon, Jun 07, 2004 at 01:43:56PM +0200, Wichert Akkerman wrote: > I still haven't managed to get wcusb working but at least I'm making > some progress in improving things with the help of #kernelnewbiews. > Below is my current patch. Changes: > Forgive me, but for some reason I missed your first message that you wrote about this. I just looked back in my mbox and found it.
I (as of yet) have not tried to get it running in 2.6.x. Looks like I'm going to have to compile a new kernel and see what's up. > * drop support for < 2.4.20 USB stack. If anyone is running a kernel > that old they deserve problems and keeping all the legacy cruft around > makes the driver harder to read No comment :-) > * use usb_fill_control_urb in prepare_transfer_urbs. This makes the > code simpler and also makes sure all relevant fields are initalized > properly (the spinklock wasn't initalized for example) Good idea. > * Add a #define for the read and write interfaces instead of keeping the > magic numbers 6 and 7 in the source Good idea too. > > Some comments: > > * the code still submits bogus URBs; at some point an URB without a > completion handler is submitted > > * the completion handlers blindly resubmit URBs without checking for > failures, which is a good way to enter a tight loop One question. It looks like the patch that you submitted is for the wcfxsusb.c driver as opposed to the wcusb.c driver. Have you tried using both yet? I'm a little out of the loop on what's happened with these drivers of late, but the wcusb driver was written as a replacement for the wcfxsusb driver. Matthew Fredrickson _______________________________________________ Asterisk-Dev mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-dev To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-dev
