On (12/12/18 14:54), Dmitry Safonov wrote: > > In this particular case we probably can just move free_page() > > out of uart_port lock scope. Note that free_page()->MM can printk() > > on its own. > > > > > > Something like this (not tested): > > Looks good to me. > Probably, it's worth to update comment about freeing just to make sure > no one will "refactor"/"simplify" it some day. > > Does it make sense to add this to your patch? Makes perfect sense, thanks! I'll send a patch a bit later. -ss
