2015-01-06 0:27 GMT+09:00 Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>: > On Mon, 5 Jan 2015, Akinobu Mita wrote: > >> While accessing a unusual usb storage (ums-alauda, ums-cypress, ...), >> the module reference count is not incremented. Because these drivers >> allocate scsi hosts with usb_stor_host_template defined in usb-storage >> module. So these drivers always can be unloaded. >> >> This fixes it by adjusting module reference after scsi host allocation. > >> diff --git a/drivers/usb/storage/alauda.c b/drivers/usb/storage/alauda.c >> index 62c2d9d..77660d6 100644 >> --- a/drivers/usb/storage/alauda.c >> +++ b/drivers/usb/storage/alauda.c >> @@ -1242,6 +1242,7 @@ static int alauda_probe(struct usb_interface *intf, >> us->transport = alauda_transport; >> us->transport_reset = usb_stor_Bulk_reset; >> us->max_lun = 1; >> + us_to_host(us)->module = THIS_MODULE; >> >> result = usb_stor_probe2(us); >> return result; > > ... etc. > > An easier and more foolproof approach would be to change > usb_store_probe2 to take an extra argument for the owner: > > int _usb_stor_probe2(struct us_data *us, struct module *owner) > ... > us_to_host(us)->module = owner; > ... > > Then in usb.h: > > extern int _usb_stor_probe2(struct us_data *us, struct module *owner); > #define usb_stor_probe2(us) _usb_stor_probe2(us, THIS_MODULE) > > This pattern is already used in several other places in the kernel. Sounds good. I'll take this idea and update this patch. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html