Messages in this thread | | | Date | Mon, 9 Jan 2012 12:01:49 -0500 (EST) | From | Alan Stern <> | Subject | Re: Problems with get_driver() and driver_attach() (and new_id too) |
| |
On Mon, 9 Jan 2012, Dmitry Torokhov wrote:
> > > drivers/ssb/main.c seems like needs some protection but does it > > > incorrectly as we do not wait for drivers to drop all references before > > > unloading modules. > > > > Possibly it needs to be replaced with try_module_get. I'll send out an > > email to the maintainers of these drivers to see what they think. > > No, I am not that try_module_get() [alone] is quite what is needed, as > strictly speaking driver lifetime does not need to be the same as module > lifetime. But maybe I am just splitting hair as all drivers are > statically initialized and are tied to their modules...
In theory you're right; what matters is when the driver is registered and unregistered. (And in fact, although the driver structs are indeed statically initialized, there also is a struct driver_private part which doesn't get allocated or initialized until the driver is registered.)
However, I don't know of any drivers that are unregistered before their module is unloaded. We could add an API to the driver core for this (i.e., to block a driver from being unregistered), but there doesn't seem to be any real point.
Alan Stern
| |