lkml.org 
[lkml]   [2002]   [Jan]   [10]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
DateThu, 10 Jan 2002 01:29:57 -0500
From kevin@koconnor ...
SubjectRe: lock order in O(1) scheduler
On Thu, Jan 10, 2002 at 12:26:08AM -0500, Robert Love wrote:
> On Thu, 2002-01-10 at 00:10, kevin@koconnor.net wrote:
> 
> > I was unable to figure out what the logic of the '(smp_processor_id() <
> > p->cpu)' test is..  (Why should the CPU number of the process being awoken
> > matter?)  My best guess is that this is to enforce a locking invariant -
> > but if so, isn't this test backwards?  If p->cpu > current->cpu then
> > p->cpu's runqueue is locked first followed by this_rq - locking greatest to
> > least, where the rest of the code does least to greatest..
> 
> Not so sure of the validity, but it is to respect lock order.
[...]

Right.  It is a confusing though - depending on the value of p->cpu
relative to current->cpu, the synchronous flag may be ignored.  Since the
relationship between p->cpu and current->cpu is essentially random, this
makes the behavior of the synchronous flag random - why bother?

I can see that re-grabbing the locks in the proper order would be a pain.
Also, after a quick grep, it appears only fs/pipe.c is interested in the
wake_up_sync() variant.  Maybe this feature should just be culled?

-Kevin


The try_to_wake_up function in pre11 (just for reference):


static int try_to_wake_up(task_t * p, int synchronous)
{
        unsigned long flags;
        int success = 0;
        runqueue_t *rq;
        lock_task_rq(rq, p, flags);
        p->state = TASK_RUNNING;
        if (!p->array) {
                if (!rt_task(p) && synchronous && (smp_processor_id() < p->cpu)) {
                        spin_lock(&this_rq()->lock);
                        p->cpu = smp_processor_id();
                        activate_task(p, this_rq());
                        spin_unlock(&this_rq()->lock);
                } else {
                        activate_task(p, rq);
                        if ((rq->curr == rq->idle) ||
                                        (p->prio < rq->curr->prio))
                                resched_task(rq->curr);
                }
                success = 1;
        }
        unlock_task_rq(rq, p, flags);
        return success;
}
-- 
 ------------------------------------------------------------------------
 | Kevin O'Connor                     "BTW, IMHO we need a FAQ for      |
 | kevin@koconnor.net                  'IMHO', 'FAQ', 'BTW', etc. !"    |
 ------------------------------------------------------------------------
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2005-03-22 12:18    [from the cache]
©2003-2008