On Tue, Sep 22, 2020 at 01:21:19PM +0200, Oliver Neukum wrote: > Stress testing has shown that CDC-WDM has some issues with hangs > and error reporting > > 1. wakeups are not correctly handled in multhreaded environments > 2. unresponsive hardware is not handled > 3. errors are not correctly reported. This needs flush() to be > implemented. > > This version makes wdm_flush() use interruptible sleep. > > For easier review all squashed together: > I have like 3 or 4 different "RFC" series here from you for this driver, which one is the "newest"? And can you send a series that isn't RFC so that I can know you feel it is good enough to be merged? thanks, greg k-h
