On Mon, Mar 30, 2020 at 02:02:07PM +0200, Greg KH wrote: > On Mon, Mar 30, 2020 at 05:22:43PM +0530, madhuparnabhowmik10@xxxxxxxxx wrote: > > Hi, > > > > This bug is found by Linux Driver Verification project (linuxtesting.org). > > > > The bug is related to the parallel execution of u132_probe() function and u132_hcd_exit() function in u132_hcd.c. In case the module is unloaded when the probe function is executing there can be data race as the mutex lock u132_module_lock is not used properly. > > Please note that module unloading, while a nice thing to have, is never > something that happens automatically :) > > > i) Usage of mutex lock only when writing into the u132_exiting variable in u132_hcd_exit(). The lock is not used when this variable is read in u132_probe(). > > > > Moreover, this variable does not serve its purpose, as even if locking is used while the u132_exiting variable is read in probe(), the function may still miss that exit function is executing if it acquires the mutex before exit() function does. > > > > How to fix this? > > > > ii) Usage of mutex while adding entries in u132_static_list in probe function but not in exit function while unregistering. > > This should be easy to fix by holding the mutex in the exit function as well. > > > > There can be other synchronization problems related to the usage of u132_module_lock in this module, I have only spotted these so far. > > > > Please let me know if this bug report is helpful and I can send a patch fixing it. > > Please just send a patch, no need to ever ask if you should, that's the > best way to report and fix anything. > Sure, I will do it. Thank you, Madhuparna > thanks, > > greg k-h
