> >> + > >> + chained_irq_enter(chip, desc); > > > > Would you become interested to collaborate with another scoped guard > > for this programming interface? > > Collaborate in which way? What are you collaborating on? > > You are merely asking people to do work which you think is useful. You can do > that, but that does not make it useful. > > Making a guard variant of chained_irq_enter/exit needs some thought and a > general plan for cleaning the whole chained irq usage up. It's on the cleanup > list already with quite some other items. > > We are not adhoc adding a guard variant because guards are hip right now. > And no this does not need a scoped variant ever. > > guards are not the panacea for everything. > > > https://elixir.bootlin.com/linux/v6.12-rc2/source/include/linux/irqchi > > p/chained_irq.h#L13 > > Please refrain from these silly links. People know to find the functions on their > own. > > Kevin, please update the change log, add your SOB and move the local > variables (unsigned long bit, status;) into the scoped_guard() zone. OK. I will do this change. > > Leave chained_irq_enter/exit() alone and resubmit. OK. I will send the v4 patch with chained_irq_enter/exit() alone. Thanks a lot. > > Thanks, > > tglx