Hi Hans, On Fri, Mar 03, 2023 at 12:58:52PM +0200, Sakari Ailus wrote: > > devm_request_irq() is used a lot in the kernel, so if this is a > > common issue, then just fixing it in two drivers isn't going to make > > much of a difference. > > I came to think of this after sending the patch as well. It's memory that > is the problem, any hardware access needs to end before remove is called. > > I'll drop the devm removal. Actually, the reason why devm_request_irq() isn't a great idea is that if an interrupt arrives between device's remove function and actually releasing that IRQ, bad things will (again) happen. So it's only safe to use devm_request_irq() if you can guarantee you're actually not going to get an interrupt then. You generally shouldn't be getting one, but that's not the same thing than being sure. -- Sakari Ailus
