Am Sonntag 02 Dezember 2001 03:40 schrieb Pete Zaitcev:
> > From: [EMAIL PROTECTED] (Oliver Neukum)
> > To: <[EMAIL PROTECTED]>
> > cc: David Brownell <[EMAIL PROTECTED]>
> > Date: Sun, 2 Dec 2001 02:10:03 +0100 (CET)
> >
> > this should cure the abuse of sleep_on as a delay.
> >
> > @@ -198,7 +195,8 @@
> > }
> > if (retval != USB_ST_TIMEOUT)
> > break;
> > - interruptible_sleep_on_timeout (&camera->wait, RETRY_TIMEOUT);
> > + set_current_state(TASK_INTERRUPTIBLE);
> > + schedule_timeout (RETRY_TIMEOUT);
> >
> > dbg ("read (%Zd) - retry", len);
> > }
>
> I think you must restore state to TASK_RUNNING after the
> return from schedule(). BTW, you did it correctly in printer.c -
> thinko here?
No, here I am guaranteed to sleep. schedule_timeout() will restore the task
state to TASK_RUNNING. In printer.c there's a chance of not calling
schedule() at all.
Regards
Oliver
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
