Re: [PATCH fwd] poll: allow f_op->poll to sleep

[Andrew Morton]

On Sat, 22 Nov 2008 21:43:51 +0900 Tejun Heo <htejun@xxxxxxxxx> wrote:

> Matthew Wilcox wrote:
> > On Sat, Nov 22, 2008 at 09:58:33AM +0100, Miklos Szeredi wrote:
> >> +int poll_schedule_timeout(struct poll_wqueues *pwq, int state,
> >> +			  ktime_t *expires, unsigned long slack)
> > 
> > All callers of poll_schedule() and poll_schedule_timeout() pass
> > TASK_INTERRUPTIBLE.  We can elide the 'state' argument.
> 
> Well, I wanted to keep it as to keep it more consistent with other
> schedule() functions but both Miklos and you don't seem to like it, so I
> might as well just drop it.  Andrew, what do you think?

I guess that if any poll/select syscall were to sleep in
uninterruptible state, people would get upset about the effect upon their
load average and we'd have to go in and fix it.

So, yup, I expect that hard-coding TASK_INTERRUPTIBLE would be OK.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/