Am Samstag, den 30.05.2020, 13:58 +0900 schrieb Tetsuo Handa:

Hi,

sorry for taking this long. There has been a family emergency.

> > The kernel most definitely does need to protect itself against 
> > misbehaving hardware.  Let's just leave it at that.  If you don't 
> > believe me, ask Greg KH.
> 
> I've made many locations killable (in order to reduce damage caused by OOM
> condition). But I can't make locations killable where handling SIGKILL case is
> too difficult to implement.

We can make flush interruptible. But that will not do the job. We would
get a file that cannot be closed.

> "struct file_operations"->flush() is called from filp_close() when there is
> something which has to be done before "struct file_operations"->release() is
> called.

Yes, in particular error reporting. Without flush() there is no way to
know whether the last write() has actually worked.

> As far as I read this thread, what you are trying to do sounds like allow
> "not waiting for completion of wdm_out_callback()" with only
> 's/wait_event/wait_event_intrruptible/' in wdm_flush(). Then, please do remove
> wdm_flush() call itself.

That would break error reporting. That flush() waits for IO to complete
is basically a side effect. You can know whether IO has worked after it
is finished.

> Therefore, again, please show me as a patch first.

Sure, attached. The difficulty here is that I see three possible
interacting errors, two of which are races.

	Regards
		Oliver