> > The BUG() would seem to be the one where the "free device state" > > method in the "uhci.c" driver (uhci_free_dev) doesn't actually do so. > > It's keeping some device state (refcounts) around after it claims to > > have gotten rid of all such state. > > Nope, it's because uhci_free_dev is a hold over from days gone by and > isn't necessary, assuming you use reference counting correctly.
Not hardly. I noticed you have yet to respond to the points I made the last time this came up ... that reads to me like you (still) have no good response to them. > > The fix discussed last time was to make "uhci.c" null out its copy > > of the device handle, and drop the refcounts, when unlinking URBs. > > That's not the right fix. I'm actually surprised you still think that's > acceptable. Actually, I'm surprised you have yet to respond to the points I made earlier ... and moreover, haven't even tried such a simple fix. - Dave _______________________________________________________________ Have big pipes? SourceForge.net is looking for download mirrors. We supply the hardware. You get the recognition. Email Us: [EMAIL PROTECTED] _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
