>> Under which circumstances would you become interested to apply a statement >> like “guard(spinlock_irqsave)(&crtc->dev->event_lock);”? > > I just follow the kerneldoc of drm_crtc_send_vblank_event() to > hold event lock like many other drivers do. Please take another look at software design possibilities to benefit more from further applications of scope-based resource management. Regards, Markus
