Re: USB-related panic in 8.2_STABLE

[Edgar Fuß <ef%math.uni-bonn.de@localhost>]

> list *(ugen_get_cdesc+0xb1)
0xffffffff802f8f2e is in ugen_get_cdesc (/usr/src-8/sys/dev/usb/ugen.c:1376).
1371            usb_config_descriptor_t *cdesc, *tdesc, cdescr;
1372            int len;
1373            usbd_status err;
1374    
1375            if (index == USB_CURRENT_CONFIG_INDEX) {
1376                    tdesc = usbd_get_config_descriptor(sc->sc_udev);
1377                    len = UGETW(tdesc->wTotalLength);
1378                    if (lenp)
1379                            *lenp = len;
1380                    cdesc = kmem_alloc(len, KM_SLEEP);

> list *(ugenioctl+0x9a4)
0xffffffff802f99d1 is in ugenioctl (/usr/src-8/sys/dev/usb/ugen.c:1668).
1663                            *usbd_get_device_descriptor(sc->sc_udev);
1664                    break;
1665            case USB_GET_CONFIG_DESC:
1666                    cd = (struct usb_config_desc *)addr;
1667                    cdesc = ugen_get_cdesc(sc, cd->ucd_config_index, &cdesclen);
1668                    if (cdesc == NULL)
1669                            return EINVAL;
1670                    cd->ucd_desc = *cdesc;
1671                    kmem_free(cdesc, cdesclen);
1672                    break;

Does that help?

What about the
	ohci0: 1 scheduling overruns
that preceded the detach that preceded the panic?