Hi On Tue, Apr 15, 2008 at 4:42 PM, Christian Fischlin <cfischlin@xxxxxx> wrote: > Is there also a possibility that a process gets woken up by an interrupt from > a state mentioned above without being registered in any waitqueue? > And if so, how could we prevent the kernel from doing so? > > The problem is that we suspect the kernel to wake some processes up even if it > is not registered in any waitqueue at all. If this really happens, it > corrupts an extension we are trying to integrate into the kernel. And we > would look for a way to prevent the kernel from such unwanted wake ups. To the best I know, TASK_UNINTERRUPTIBLE should do the job ... or...I think you should add checking code after you're woken up to make sure it's not triggered due to unexpected condition...and then put your process back to sleep as soon as possible. regards, Mulyadi. -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
