On Sun, 10 Jan 2010, Michael S. Tsirkin wrote: > Not sure what you mean by irqfd locking. IMO we must take waitqueue > lock, otherwise we can not prevent wait queue callback from being > invoked, right? Note that if we take our own lock under wait queue > callback, we won't be able to use this lock around remove_wait_queue. > Also note how fs/evetfd.c uses __remove_wait_queue after taking wait > queue lock - we'd like to do the same with our wait queue. > > Could you clarify pls? Wait a second. Looking at the current code in Linus tree, when you deassign an irqfd, you are actually destroying it, so the idea of saving the counter on deassign is not going to work. - Davide -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
