I tried to change the current RML preemptive patch for 2.4.19-rc2 to work
with the O(1) scheduler patch applied. The only changes I made were in
sched.c - Not sure if this is a correct change:
Anton
=============================================================================
diff -ru linux-2.4.19-rc2/kernel/sched.c linux/kernel/sched.c
--- linux-2.4.19-rc2/kernel/sched.c Mon Jul 22 19:02:43 2002
+++ linux/kernel/sched.c Mon Jul 22 18:52:54 2002
@@ -345,11 +345,13 @@
#if CONFIG_SMP
int need_resched;
+ preempt_disable();
need_resched = p->need_resched;
wmb();
set_tsk_need_resched(p);
if (!need_resched && (p->cpu != smp_processor_id()))
smp_send_reschedule(p->cpu);
+ preempt_enable();
#else
set_tsk_need_resched(p);
#endif
@@ -367,6 +369,7 @@
runqueue_t *rq;
repeat:
+ preempt_disable();
rq = task_rq(p);
if (unlikely(task_running(rq, p))) {
cpu_relax();
@@ -375,14 +378,17 @@
* a preemption point - we are busy-waiting
* anyway.
*/
+ preempt_enable();
goto repeat;
}
rq = task_rq_lock(p, &flags);
if (unlikely(task_running(rq, p))) {
task_rq_unlock(rq, &flags);
+ preempt_enable();
goto repeat;
}
task_rq_unlock(rq, &flags);
+ preempt_enable();
}
#endif
@@ -519,7 +525,7 @@
p->sleep_avg) / (EXIT_WEIGHT + 1);
}
-#if CONFIG_SMP
+#if CONFIG_SMP || CONFIG_PREEMPT
asmlinkage void schedule_tail(task_t *prev)
{
finish_arch_switch(this_rq(), prev);
@@ -1078,6 +1084,7 @@
BUG();
need_resched:
+ preempt_disable();
prev = current;
rq = this_rq();
@@ -1085,6 +1092,14 @@
prev->sleep_timestamp = jiffies;
spin_lock_irq(&rq->lock);
+#ifdef CONFIG_PREEMPT
+ /*
+ * entering from preempt_schedule, off a kernel preemption,
+ * go straight to picking the next task.
+ */
+ if (unlikely(preempt_get_count() & PREEMPT_ACTIVE))
+ goto pick_next_task;
+#endif
switch (prev->state) {
case TASK_INTERRUPTIBLE:
if (unlikely(signal_pending(prev))) {
@@ -1096,9 +1111,9 @@
case TASK_RUNNING:
;
}
-#if CONFIG_SMP
+
pick_next_task:
-#endif
+
if (unlikely(!rq->nr_running)) {
#if CONFIG_SMP
load_balance(rq, 1);
@@ -1151,10 +1166,30 @@
spin_unlock_irq(&rq->lock);
reacquire_kernel_lock(current);
+
if (need_resched())
goto need_resched;
+ preempt_enable_no_resched();
+
+}
+
+#ifdef CONFIG_PREEMPT
+
+ /*
+ * this is is the entry point to schedule() from in-kernel preemption.
+ */
+ asmlinkage void preempt_schedule(void)
+ {
+ do {
+ current->preempt_count += PREEMPT_ACTIVE;
+ schedule();
+ current->preempt_count -= PREEMPT_ACTIVE;
+ barrier();
+ } while (current->need_resched);
}
+#endif /* CONFIG_PREEMPT */
+
/*
* The core wakeup function. Non-exclusive wakeups (nr_exclusive == 0) just
* wake everything up. If it's an exclusive wakeup (nr_exclusive == small
+ve
@@ -1923,6 +1958,9 @@
double_rq_unlock(idle_rq, rq);
set_tsk_need_resched(idle);
__restore_flags(flags);
+
+ /* Set the preempt count _outside_ the spinlocks! */
+ idle->preempt_count = (idle->lock_depth >= 0);
}
#if CONFIG_SMP
@@ -1968,6 +2006,7 @@
if (!new_mask)
BUG();
+ preempt_disable();
rq = task_rq_lock(p, &flags);
p->cpus_allowed = new_mask;
/*
@@ -1995,7 +2034,7 @@
down(&req.sem);
out:
- return;
+ preempt_enable();
}
static __initdata int master_migration_thread;
On Mon, 22 Jul 2002, anton wilson wrote:
> I tried to change the current RML preemptive patch for 2.4.19-rc2 to
> work with the O(1) scheduler patch applied. The only changes I made were
> in sched.c - Not sure if this is a correct change:
looks good at first sight.
this one:
> +
> + /* Set the preempt count _outside_ the spinlocks! */
> + idle->preempt_count = (idle->lock_depth >= 0);
needs to be #if CONFIG_PREEMPT.
Ingo
Strangely enough I'm getting processes exiting with preemption count of 1
when I use my patch.
What causes such a problem?
Anton
=================================================================
diff -ru linux-2.4.19-rc2/kernel/sched.c linux/kernel/sched.c
--- linux-2.4.19-rc2/kernel/sched.c Mon Jul 22 19:02:43 2002
+++ linux/kernel/sched.c Tue Jul 23 09:44:54 2002
@@ -345,11 +345,13 @@
#if CONFIG_SMP
int need_resched;
+ preempt_disable();
need_resched = p->need_resched;
wmb();
set_tsk_need_resched(p);
if (!need_resched && (p->cpu != smp_processor_id()))
smp_send_reschedule(p->cpu);
+ preempt_enable();
#else
set_tsk_need_resched(p);
#endif
@@ -367,6 +369,7 @@
runqueue_t *rq;
repeat:
+ preempt_disable();
rq = task_rq(p);
if (unlikely(task_running(rq, p))) {
cpu_relax();
@@ -375,14 +378,17 @@
* a preemption point - we are busy-waiting
* anyway.
*/
+ preempt_enable();
goto repeat;
}
rq = task_rq_lock(p, &flags);
if (unlikely(task_running(rq, p))) {
task_rq_unlock(rq, &flags);
+ preempt_enable();
goto repeat;
}
task_rq_unlock(rq, &flags);
+ preempt_enable();
}
#endif
@@ -519,7 +525,7 @@
p->sleep_avg) / (EXIT_WEIGHT + 1);
}
-#if CONFIG_SMP
+#if CONFIG_SMP || CONFIG_PREEMPT
asmlinkage void schedule_tail(task_t *prev)
{
finish_arch_switch(this_rq(), prev);
@@ -1078,6 +1084,7 @@
BUG();
need_resched:
+ preempt_disable();
prev = current;
rq = this_rq();
@@ -1085,6 +1092,14 @@
prev->sleep_timestamp = jiffies;
spin_lock_irq(&rq->lock);
+#ifdef CONFIG_PREEMPT
+ /*
+ * entering from preempt_schedule, off a kernel preemption,
+ * go straight to picking the next task.
+ */
+ if (unlikely(preempt_get_count() & PREEMPT_ACTIVE))
+ goto pick_next_task;
+#endif
switch (prev->state) {
case TASK_INTERRUPTIBLE:
if (unlikely(signal_pending(prev))) {
@@ -1096,9 +1111,8 @@
case TASK_RUNNING:
;
}
-#if CONFIG_SMP
+
pick_next_task:
-#endif
if (unlikely(!rq->nr_running)) {
#if CONFIG_SMP
load_balance(rq, 1);
@@ -1151,13 +1165,33 @@
spin_unlock_irq(&rq->lock);
reacquire_kernel_lock(current);
+
if (need_resched())
goto need_resched;
+ preempt_enable_no_resched();
+
+}
+
+#ifdef CONFIG_PREEMPT
+
+ /*
+ * this is is the entry point to schedule() from in-kernel preemption.
+ */
+ asmlinkage void preempt_schedule(void)
+ {
+ do {
+ current->preempt_count += PREEMPT_ACTIVE;
+ schedule();
+ current->preempt_count -= PREEMPT_ACTIVE;
+ barrier();
+ } while (current->need_resched);
}
+#endif /* CONFIG_PREEMPT */
+
/*
* The core wakeup function. Non-exclusive wakeups (nr_exclusive == 0) just
- * wake everything up. If it's an exclusive wakeup (nr_exclusive == small
+ve
+ * wake everything up. If it's an exclusive wakeup (nr_exclusive == small
* number) then we wake all the non-exclusive tasks and one exclusive task.
*
* There are circumstances in which we can try to wake a task which has
already
@@ -1923,6 +1957,11 @@
double_rq_unlock(idle_rq, rq);
set_tsk_need_resched(idle);
__restore_flags(flags);
+
+#if CONFIG_PREEMPT
+ /* Set the preempt count _outside_ the spinlocks! */
+ idle->preempt_count = (idle->lock_depth >= 0);
+#endif
}
#if CONFIG_SMP
@@ -1968,6 +2007,7 @@
if (!new_mask)
BUG();
+ preempt_disable();
rq = task_rq_lock(p, &flags);
p->cpus_allowed = new_mask;
/*
@@ -1995,7 +2035,7 @@
down(&req.sem);
out:
- return;
+ preempt_enable();
}
static __initdata int master_migration_thread;
=====================================================
On Tuesday 23 July 2002 04:49 am, Ingo Molnar wrote:
> On Mon, 22 Jul 2002, anton wilson wrote:
> > I tried to change the current RML preemptive patch for 2.4.19-rc2 to
> > work with the O(1) scheduler patch applied. The only changes I made were
> > in sched.c - Not sure if this is a correct change:
>
> looks good at first sight.
>
> this one:
> > +
> > + /* Set the preempt count _outside_ the spinlocks! */
> > + idle->preempt_count = (idle->lock_depth >= 0);
>
> needs to be #if CONFIG_PREEMPT.
>
> Ingo
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
have you added the "#if CONFIG_PREEMPT" to the place i suggested?
Ingo
On Tue, 23 Jul 2002, anton wilson wrote:
>
> Strangely enough I'm getting processes exiting with preemption count of 1
> when I use my patch.
> What causes such a problem?
>
> Anton
>
>
> =================================================================
> diff -ru linux-2.4.19-rc2/kernel/sched.c linux/kernel/sched.c
> --- linux-2.4.19-rc2/kernel/sched.c Mon Jul 22 19:02:43 2002
> +++ linux/kernel/sched.c Tue Jul 23 09:44:54 2002
> @@ -345,11 +345,13 @@
> #if CONFIG_SMP
> int need_resched;
>
> + preempt_disable();
> need_resched = p->need_resched;
> wmb();
> set_tsk_need_resched(p);
> if (!need_resched && (p->cpu != smp_processor_id()))
> smp_send_reschedule(p->cpu);
> + preempt_enable();
> #else
> set_tsk_need_resched(p);
> #endif
> @@ -367,6 +369,7 @@
> runqueue_t *rq;
>
> repeat:
> + preempt_disable();
> rq = task_rq(p);
> if (unlikely(task_running(rq, p))) {
> cpu_relax();
> @@ -375,14 +378,17 @@
> * a preemption point - we are busy-waiting
> * anyway.
> */
> + preempt_enable();
> goto repeat;
> }
> rq = task_rq_lock(p, &flags);
> if (unlikely(task_running(rq, p))) {
> task_rq_unlock(rq, &flags);
> + preempt_enable();
> goto repeat;
> }
> task_rq_unlock(rq, &flags);
> + preempt_enable();
> }
> #endif
>
> @@ -519,7 +525,7 @@
> p->sleep_avg) / (EXIT_WEIGHT + 1);
> }
>
> -#if CONFIG_SMP
> +#if CONFIG_SMP || CONFIG_PREEMPT
> asmlinkage void schedule_tail(task_t *prev)
> {
> finish_arch_switch(this_rq(), prev);
> @@ -1078,6 +1084,7 @@
> BUG();
>
> need_resched:
> + preempt_disable();
> prev = current;
> rq = this_rq();
>
> @@ -1085,6 +1092,14 @@
> prev->sleep_timestamp = jiffies;
> spin_lock_irq(&rq->lock);
>
> +#ifdef CONFIG_PREEMPT
> + /*
> + * entering from preempt_schedule, off a kernel preemption,
> + * go straight to picking the next task.
> + */
> + if (unlikely(preempt_get_count() & PREEMPT_ACTIVE))
> + goto pick_next_task;
> +#endif
> switch (prev->state) {
> case TASK_INTERRUPTIBLE:
> if (unlikely(signal_pending(prev))) {
> @@ -1096,9 +1111,8 @@
> case TASK_RUNNING:
> ;
> }
> -#if CONFIG_SMP
> +
> pick_next_task:
> -#endif
> if (unlikely(!rq->nr_running)) {
> #if CONFIG_SMP
> load_balance(rq, 1);
> @@ -1151,13 +1165,33 @@
> spin_unlock_irq(&rq->lock);
>
> reacquire_kernel_lock(current);
> +
> if (need_resched())
> goto need_resched;
> + preempt_enable_no_resched();
> +
> +}
> +
> +#ifdef CONFIG_PREEMPT
> +
> + /*
> + * this is is the entry point to schedule() from in-kernel preemption.
> + */
> + asmlinkage void preempt_schedule(void)
> + {
> + do {
> + current->preempt_count += PREEMPT_ACTIVE;
> + schedule();
> + current->preempt_count -= PREEMPT_ACTIVE;
> + barrier();
> + } while (current->need_resched);
> }
>
> +#endif /* CONFIG_PREEMPT */
> +
> /*
> * The core wakeup function. Non-exclusive wakeups (nr_exclusive == 0) just
> - * wake everything up. If it's an exclusive wakeup (nr_exclusive == small
> +ve
> + * wake everything up. If it's an exclusive wakeup (nr_exclusive == small
> * number) then we wake all the non-exclusive tasks and one exclusive task.
> *
> * There are circumstances in which we can try to wake a task which has
> already
> @@ -1923,6 +1957,11 @@
> double_rq_unlock(idle_rq, rq);
> set_tsk_need_resched(idle);
> __restore_flags(flags);
> +
> +#if CONFIG_PREEMPT
> + /* Set the preempt count _outside_ the spinlocks! */
> + idle->preempt_count = (idle->lock_depth >= 0);
> +#endif
> }
>
> #if CONFIG_SMP
> @@ -1968,6 +2007,7 @@
> if (!new_mask)
> BUG();
>
> + preempt_disable();
> rq = task_rq_lock(p, &flags);
> p->cpus_allowed = new_mask;
> /*
> @@ -1995,7 +2035,7 @@
>
> down(&req.sem);
> out:
> - return;
> + preempt_enable();
> }
>
> static __initdata int master_migration_thread;
>
>
> =====================================================
> On Tuesday 23 July 2002 04:49 am, Ingo Molnar wrote:
> > On Mon, 22 Jul 2002, anton wilson wrote:
> > > I tried to change the current RML preemptive patch for 2.4.19-rc2 to
> > > work with the O(1) scheduler patch applied. The only changes I made were
> > > in sched.c - Not sure if this is a correct change:
> >
> > looks good at first sight.
> >
> > this one:
> > > +
> > > + /* Set the preempt count _outside_ the spinlocks! */
> > > + idle->preempt_count = (idle->lock_depth >= 0);
> >
> > needs to be #if CONFIG_PREEMPT.
> >
> > Ingo
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
>
> >?+#if CONFIG_PREEMPT
> >?+ ? ? ? /* Set the preempt count _outside_ the spinlocks! */
> >?+ ? ? ? idle->preempt_count = (idle->lock_depth >= 0);
> >?+#endif
This is the update I added.
Anton
On Tue, 2002-07-23 at 08:55, anton wilson wrote:
> Strangely enough I'm getting processes exiting with preemption count of 1
> when I use my patch.
> What causes such a problem?
It probably means you are off-by-1 somewhere in your scheduler locking
that is keeping the preempt_count at 1 instead of at 0 in normal
situations.
This also means you are not preempting.
A good check to see if your patch is right would be to diff, for
example, kernel/sched.c in 2.4-ac against yours... apply my patch to a
recent 2.4-ac and see the differences.
I have also attached a preempt-kernel patch for O(1) scheduler on 2.4.18
- if this applies cleanly then everything should be fine. The patch is
not entirely up-to-date but may be of use.
Robert Love