Re: Clarification required about select vs wake_up race condition

["Dmitry Adamushko" <dmitry.adamushko@gmail.com>]

On 12/03/07, Ravinandan Arakali (rarakali) <rarakali@cisco.com> wrote:
> Hi,
> I am facing following problem and was wondering if somebody could help
> me out.
> Our char driver(pretty much like all other char drivers) does a
> poll_wait()
> and returns status depending on whether data is available to be read.
> Even though some data is available to be read(verified using one of our
> internal
> commands), the select() never wakes up, inspite of any no. of messages
> sent.
>
> To understand this, I was looking at the code of select vs
> wake_up_interruptible().
> I feel I am misunderstanding some part of the kernel code but will be
> glad if
> somebody can point it out.
>
> My understanding:
> The do_select() sets the state of task to TASK_INTERRUPTIBLE and calls
> the driver's
> poll entry point. In our poll(), let's say immediately after we
> determine that there's
> nothing to be read, some data arrives causing a wake_up_interruptible()
> on another CPU.
> The wake up happens in the context of process sending the data. Since
> the receiving
> process was already added to the list of listeners, from looking at the
> code of
> try_to_wake_up(), it looks like it can set the state of the receiving
> process to
> TASK_RUNNING(I don't see any lock preventing this). After this happens,
> the receiving
> process goes to sleep (because of schedule_timeout called by do_select)
> but
> state is still set to TASK_RUNNING.

No, it's not going to sleep then.

The effect of schedule() being called with current->state ==
TASK_RUNNING is a re-scheduling to another task with a higher prio (if
any) or just getting back (iow, the task doesn't lose a cpu). For both
cases, the task is on the runqueue.

/ Look how/when deactivate_task() is called in schedule() /

Maybe there is a race in your code between (1) how you check "data is
available" in poll and (2) a part that sets this fact (data is
available)...


-- 
Best regards,
Dmitry Adamushko