Re: POLL in char drivers

[Shankar Ganesh <shankarganeshk@xxxxxxxxx>]

On Thu, Aug 13, 2009 at 5:23 PM, Daniel Baluta <daniel.baluta@xxxxxxxxx> wrote:

On Thu, Aug 13, 2009 at 2:17 PM, Shankar Ganesh<shankarganeshk@xxxxxxxxx> wrote:
> Hi ,
>
> I have a doubt regarding poll function in char driver . Let us consider
> following code
>
> --------------------------------------------------------------------------------------
> static unsigned int random_poll(struct file *file, poll_table * wait)
> {
>      unsigned int mask;
>    poll_wait(file, &random_read_wait, wait);
>    poll_wait(file, &random_write_wait, wait);
>    mask = 0;
> ........
>      return mask;
> }
>
>
> read()  / write()
> {
>        wait_event_interuptable()
> }
> --------------------------------------------------------------------------------------
>
> Does the above code (poll function) is in sleep state until both
> random_read_wait & random_write_wait are awakened ? . If i have use case If
> i want my user space application to poll only for read , then can i remove
> one of the above  ?

t beautiful thing we can experience is the mysterious. It is the
> source of all true art and science."
>

AFAIK your actual random_poll function nevers sleeps. The calls for
poll_wait just inform the VFS layer
about the existence of random_read_wait and random_write_wait wait
queues. An upper layer will use the return mask
to actual put your current process to sleep if there are no data
available for reading/writing.

Thanks. How do VFS know which of the queue is woke-up (read -WQ or write - WQ ) ? . As any ISR or relevant function wil wake-up the wait-queue. 

I think you should not care about what type of operation a user space
application is watching for.

Daniel.

-- 
With Regards,
ShankarGanesh K .







"The most beautiful thing we can experience is the mysterious. It is the source of all true art and science."