Re: [syzbot] WARNING in __init_work

Quoting Thomas Gleixner (2021-09-19 05:41:18)

Again, it cannot do that from a softirq because
synchronize_rcu_expedited() might sleep.

quoted

Is it possible that object_is_on_stack() doesn't work in IRQ context?
I'm not really following along on x86 but I could see where
task_stack_page() gets the wrong "stack" pointer because the task has one
stack and the irq stack is some per-cpu dedicated allocation?

Even if debug objects would support objects on irq stacks, the above is
still bogus. But it does not and will not because the operations here
have to be fully synchronous:

    init() -> queue() or arm() -> wait() -> destroy()

because you obviously cannot queue work or arm a timer which are on stack
and then leave the function without waiting for the operation to complete.

Is there some way to make it more obvious that initializing a timer or
work on the stack in an irq context is a NONO because we can't wait for
it? Maybe some sort of debugobjects call to might_sleep() when it's
being told the object is on the stack, or throwing a might_sleep() into
the initialization of any stack based timer or workqueue, or both?

So these operations have to be synchronous which is a NONO when running
in hard or soft interrupt context because waiting for the operation to
complete is not possible there.