śr., 4 gru 2019 o 23:28 Andy Shevchenko <andy.shevchenko@xxxxxxxxx> napisał(a): > > On Wed, Dec 4, 2019 at 6:04 PM Bartosz Golaszewski <brgl@xxxxxxxx> wrote: > > > > From: Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx> > > > > Currently if the line-event kfifo is full, we just silently drop any new > > events. Add a ratelimited debug message so that we at least have some > > trace in the kernel log of event overflow. > > > > Hmm... I don't like prints in IRQ context (even threaded). > Can we rather switch to trace points at some point? > This is something that will be very rare and unlikely - I don't see how trace points will help here with all the boiler-plate. Bart